*Complete removal of patient data listings may mean that ... · CONFIDENTIAL 2017N324761_00 206215 6 1. INTRODUCTION 1.1. Study Rationale This study was designed to assess the benefits

In February 2013, GlaxoSmithKline (GSK) announced a commitment to further clinical transparency through the public disclosure of GSK Clinical Study Reports (CSRs) on the GSK Clinical Study Register.

The following guiding principles have been applied to the disclosure: Information will be excluded in order to protect the privacy of patients and all named

persons associated with the study Patient data listings will be completely removed* to protect patient privacy. Anonymized

data from each patient may be made available subject to an approved research proposal. For further information please see the Patient Level Data section of the GSK Clinical Study Register.

Aggregate data will be included; with any direct reference to individual patients excluded

*Complete removal of patient data listings may mean that page numbers are no longer consecutively

numbered

CONFIDENTIAL 2017N324761_00The GlaxoSmithKline group of companies 206215

1

Division: Worldwide DevelopmentInformation Type: Clinical Study ReportControl: Placebo

Title:A randomized, open-label, cross-over, placebo-device study investigating critical and over all errors, training/teaching time, and preference attributes of the ELLIPTA dry powder Inhaler (DPI) as compared to HandiHaler DPI used in combination with either DISKUS DPI or Turbuhaler DPI, in adult patients with Chronic Obstructive Pulmonary Disease (COPD).

Phase: IV

Compound Number: GSK573719+GW642444+GW685698 (GSK2834425)

Effective Date: 07-DEC-2017

Subject: COPD, ELLIPTA, DPI, HandiHaler, Turbuhaler, DISKUS (ACCUHALER)

Author(s): (Complete Regulatory);

Indication Studied: COPD

Initiation Date: 30-DEC-2016

Completion Date: 19-JUN-2017

Sponsor Signatory:(and Medical Officer)

Steve PascoeVice President, Respiratory Head Unit Physician+MDLMDC Global Clinical Devices USGlaxoSmithKline

This study was performed in compliance with Good Clinical Practices and GlaxoSmithKline Standard Operating Procedures for all processes involved, including the archiving of essential documents. This study complies with US 21 CFR 312.120.

Copyright 2017 the GlaxoSmithKline group of companies. All rights reserved. Unauthorised copying or use of this information is prohibited.

1

PPD

PPD

Table of Contents Page TITLE PAGE .......................................................................................................... ABBREVIATIONS .................................................................................................. ETHICS AND GOOD CLINICAL PRACTICE ......................................................... 1. INTRODUCTION ................................................................................................

1.1. Study Rationale ............................................................................................... 1.2. Background .....................................................................................................

2. STUDY OBJECTIVES ........................................................................................ 3. INVESTIGATORS AND STUDY ADMINISTRATIVE STRUCTURE .................. 4. INVESTIGATIONAL PLAN ................................................................................

4.1. Study Design ................................................................................................... 4.2. Discussion of Study Design ............................................................................. 4.3. Protocol Amendment(s) ................................................................................... 4.4. Selection of Study Population ..........................................................................

4.4.1. Inclusion/Exclusion Criteria ......................................................................... 4.4.2. Withdrawal Criteria ......................................................................................

4.5. Treatments ...................................................................................................... 4.5.1. Investigational Product(s) and Reference Therapy ..................................... 4.5.2. Treatment Assignment ................................................................................ 4.5.3. Blinding ........................................................................................................ 4.5.4. Prior and Concomitant Medications and Non-Drug Therapies .................... 4.5.5. Compliance .................................................................................................

4.6. Study Assessments and Procedures ............................................................... 4.6.1. Efficacy Assessment ................................................................................... 4.6.2. Safety Assessments ....................................................................................

4.7. Data Quality Assurance ................................................................................... 4.8. Statistical Analyses ..........................................................................................

4.8.1. Sample Size Considerations ....................................................................... 4.8.2. Analysis Populations ................................................................................... 4.8.3. Study Population Analysis ........................................................................... 4.8.4. Efficacy Analysis ......................................................................................... 4.8.5. Safety Analysis ............................................................................................ 4.8.6. Interim Analyses .......................................................................................... 4.8.7. Final Analyses ............................................................................................. 4.8.8. Changes in Conduct of the Study or Planned Analyses ..............................

5. STUDY POPULATION RESULTS ..................................................................... 5.1. Subject Disposition ..........................................................................................

5.1.1. Screen Failures ........................................................................................... 5.1.2. Randomised Subjects ..................................................................................

5.2. Protocol Deviations .......................................................................................... 5.3. Populations Analysed ...................................................................................... 5.4. Demographics and Baseline Characteristics ...................................................

5.4.1. Demographics ............................................................................................. 5.4.2. Baseline Characteristics ..............................................................................

1678889101010111111121213131314141415151717181818191920202020222222222223232324

2

CONFIDENTIAL 2017N324761_00206215

5.5. Prior and Concomitant Medications ................................................................. 5.5.1. COPD Medications ...................................................................................... 5.5.2. Non-COPD Medications ..............................................................................

5.6. Prior Inhaler Use .............................................................................................. 5.7. Exposure and Treatment Compliance .............................................................

6. EFFICACY RESULTS ........................................................................................ 6.1. Critical Inhaler Use Errors ................................................................................

6.1.1. Critical Inhaler Use Errors after Reading the PIL ......................................... 6.1.2. Critical Inhaler Use Errors after HCP Instruction ......................................... 6.1.3. Summary of Critical Inhaler Use Errors .......................................................

6.2. Overall Inhaler Use Errors ............................................................................... 6.2.1. Overall Inhaler Use Errors after Reading the PIL ........................................ 6.2.2. Overall Inhaler Use Errors after HCP Instruction ......................................... 6.2.3. Summary of Overall Inhaler Use Errors .......................................................

6.3. Sensitivity Analysis .......................................................................................... 6.4. Number of Instructions Required from HCP to Demonstrate Correct Inhaler

Use .......................................................................................................... 6.5. Time Taken to Use Inhaler .............................................................................. 6.6. Inhaler Preference ...........................................................................................

7. SAFETY RESULTS ............................................................................................ 7.1. Adverse Events ............................................................................................... 7.2. Serious and Other Significant Adverse Events ................................................ 7.3. Pregnancies ..................................................................................................... 7.4. Combination Device/Drug Product Near-Incidents, Malfunctions and

Remedial Action ....................................................................................... 8. DISCUSSION AND CONCLUSIONS .................................................................

8.1. Discussion ....................................................................................................... 8.2. Conclusions .....................................................................................................

9. REFERENCES ................................................................................................... STUDY POPULATION DATA SOURCE TABLES ................................................

Table 1.01 Summary of Subject Population (All Subjects Enrolled Population) ...... Table 1.02 Summary of Screening Status and Reasons for Screen Failures (All

Subjects Enrolled Population) .................................................................. Table 1.03 Summary of Age Ranges (All Subjects Enrolled Population) ................ Table 1.04 Summary of Attendance at Each Clinic Visit (Intent-to-Treat

Population) ............................................................................................... Table 1.05 Summary of End of Study Record (All Subjects Enrolled Population) .. Table 1.06 Summary of Demographic Characteristics (Intent-to-Treat Population) Table 1.07 Summary of Demographic Characteristics by Country (Intent-to-Treat

Population) ............................................................................................... Table 1.08 Summary of Number of Subjects by Country and Centre (Intent-to-

Treat Population) ..................................................................................... Table 1.09 Summary of Race and Racial Combinations (Intent-to-Treat

Population) ............................................................................................... Table 1.10 Summary of Race and Racial Combination Details (Intent-to-Treat

Population) ...............................................................................................

262627272829292931343636373940

41424647474747

47484850515353

5455

565758

60

64

65

66

3

CONFIDENTIAL 2017N324761_00206215

Table 1.11 Summary of Important Protocol Deviations (Intent-to-Treat Population) .................................................................................................................

Table 1.12 Summary of Inclusion/Exclusion Criteria Deviations (Intent-to-Treat Population) ...............................................................................................

Table 1.13 Summary of Current Medical Conditions (Intent-to-Treat Population) .. Table 1.14 Summary of Past Medical Conditions (Intent-to-Treat Population) ....... Table 1.15 Summary of Family History of Cardiovascular Risk Factors (Intent-to-

Treat Population) ..................................................................................... Table 1.16 Summary of COPD History (Intent-to-Treat Population) ....................... Table 1.17 Summary of Smoking History and Smoking Status at Screening

(Intent-to-Treat Population) ...................................................................... Table 1.18 Summary of COPD Medications (Intent-to-Treat Population) ............... Table 1.19 Summary of Non-COPD Concomitant Medications (Intent-to-Treat

Population) ............................................................................................... Table 1.20 Summary of Naive to Inhaler Devices (Intent-to-Treat Population) .......

EFFICACY DATA SOURCE FIGURES.................................................................. Figure 2.01 Percentage of Subjects with at Least One Critical Error by

Assessment (Intent-to-Treat Population) ................................................. Figure 2.02 Percentage of Subjects with at Least One Error by Assessment

(Intent-to-Treat Population) ...................................................................... Figure 2.03 Number of Instructions required from a HCP to Demonstrate Correct

Use (Intent-to-Treat Population) .............................................................. Figure 2.04 Kaplan-Meier Plot of Time Taken to Demonstrate Correct Inhaler Use

(Intent-to-Treat Population) ...................................................................... EFFICACY DATA SOURCE TABLES ...................................................................

Table 2.01 Summary of Errors on ELLIPTA (Intent-to-Treat Population)................ Table 2.02 Summary of Errors on ELLIPTA by Country (Intent-to-Treat

Population) ............................................................................................... Table 2.03 Summary of Errors on DISKUS + HandiHaler (Intent-to-Treat

Population) ............................................................................................... Table 2.04 Summary of Errors on DISKUS + HandiHaler by Country (Intent-to-

Treat Population) ..................................................................................... Table 2.05 Summary of Errors on Turbuhaler + HandiHaler (Intent-to-Treat

Population) ............................................................................................... Table 2.06 Summary of Errors on Turbuhaler + HandiHaler by Country (Intent-to-

Treat Population) ..................................................................................... Table 2.07 Summary of Critical Errors (Intent-to-Treat Population) ........................ Table 2.08 Summary of Critical Errors by Country (Intent-to-Treat Population) ...... Table 2.09 Analysis of Percentage of Subjects making at least one Critical Error

(Intent-to-Treat Population) ...................................................................... Table 2.10 Sensitivity Analysis of Percentage of Subjects making at least one

Error after Reading the PIL(s) (Intent-to-Treat Population) ...................... Table 2.11 Analysis of Percentage of Subjects making at least one Overall Error

(Intent-to-Treat Population) ...................................................................... Table 2.12 Summary and Analysis of Time (Minutes) Taken to Demonstrate

Correct Inhaler Use (Intent-to-Treat Population) ......................................

67

686971

7374

7576

809091

91

93

95

97103103

105

109

112

118

121127133

145

151

153

159

4

CONFIDENTIAL 2017N324761_00206215

Table 2.13 Summary of Treatment Preference (Intent-to-Treat Population) ........... Table 2.14 Summary and Analysis of Number of Instructions from the HCP

(Intent-to-Treat Population) ...................................................................... Table 2.15 Probability of Total Time (Minutes) Taken to Demonstrate Correct

Inhaler Use by 5 Minute Intervals (Intent-to-Treat Population) .................

165

167

169

5

CONFIDENTIAL 2017N324761_00206215

CONFIDENTIAL 2017N324761_00206215

4

ABBREVIATIONS

AE(s) Adverse Event(s)ASE All Subjects Enrolled CI Confidence IntervalCOPD Chronic Obstructive Pulmonary DiseaseCSR Clinical Study ReportDPI Dry Powder InhalereCRF Electronic Case Record FormGCP Good Clinical PracticeGSK GlaxoSmithKlineHCP Healthcare ProfessionalICH International Council for Harmonisation of Technical Requirements

for Registration of Pharmaceuticals for Human UseICS Inhaled CorticosteroidIEC Independent Ethics CommitteesITT Intent-to-treatIWRS Interactive Web Response SystemLABA Long-Acting Beta2-AgonistLAMA Long-Acting Muscarinic AntagonistMax MaximumMin MinimumN Sample Sizen Subset of Sample SizePI Principal InvestigatorPIL Patient Instruction LeafletPREF Preference questionnaireRMC Respiratory medication classSABA Short Acting Beta-Adrenergic AgonistSAE(s) Serious Adverse Event(s)SD Standard DeviationUK United KingdomV Visit

Trademark Information

Trademarks of the GlaxoSmithKline group of companies

Trademarks not owned by the GlaxoSmithKline group of

companies

ANORO ELLIPTA SASSERETIDE DISKUS Seebri BreezhalerINCRUSE ELLIPTA Spiriva HandiHalerBREO ELLIPTA Symbicort Turbuhaler

Ultibro Breezhaler

6

CONFIDENTIAL 2017N324761_00206215

5

ETHICS AND GOOD CLINICAL PRACTICE

The study protocol, any amendments, the informed consent, and other information that required pre-approval were reviewed and approved by a national, regional, or investigational centre ethics committee, in accordance with the International Council forHarmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH) Good Clinical Practice (GCP) and applicable country-specific requirements, including US 21 Code of Federal Regulations (CFR) 312.3(b) for constitution of independent ethics committees. Ethics committee approvals are maintained in the Sponsor’s study file.

Investigators were trained to conduct the study in accordance with GCPs and the study protocol, as defined in ICH E3, Section 9.6. Written commitments were obtained from investigators to conduct the study in accordance with ICH GCP and all applicable subject privacy requirements, and the ethical principles that are outlined in the Declaration of Helsinki and to conduct the study in accordance with the protocol.

The study was monitored in accordance with ICH E6, Section 5.18. If significant findings (e.g., potential serious misconduct or noncompliance with GCP, including potential Serious Breaches) were identified during monitoring or auditing of a site, these are presented in Section 5.2, Protocol Deviations of this report.

Written informed consent was obtained from each subject prior to the performance of any study-specific procedures. The investigator agreed to provide the subject as much time as necessary to review the document, to inquire about details of the trial, and to decide whether or not to participate in the study. The informed consent was signed and dated by the study subject and by the person who conducted the informed consent discussion. Case report forms were provided for each subject’s data to be recorded.

7

CONFIDENTIAL 2017N324761_00206215

6

1. INTRODUCTION

1.1. Study Rationale

This study was designed to assess the benefits of delivering triple therapy using a single ELLIPTA Dry Powder Inhaler (DPI) (referred to as closed triple therapy), versus delivering triple therapy using 2 different types of inhalers (referred to as open triple therapy) to patients with chronic obstructive pulmonary disease (COPD). It assessed the proportion of COPD subjects who made critical errors when using a single ELLIPTA DPI versus those using combinations of commercially available and commonly used DPIs: DISKUS used in combination with HandiHaler, or Turbuhaler used in combination with HandiHaler. This study also assessed training/teaching time and preference attributes for receiving closed triple therapy compared with receiving open triple therapy.

1.2. Background

COPD is a progressive disease of the airways. Noxious particles or gases lead to a modification in subject’s airways which develop a chronic inflammatory response to these environmental factors. This leads to increasing airflow limitation, breathlessness and other symptoms. The current treatment goals for COPD are to reduce and relieve symptoms and thereby improve exercise tolerance and health status, whilst reducing risk by preventing disease progression, preventing and treating exacerbations, and reducing mortality.

The current Global Initiative for Chronic Obstructive Lung Disease (GOLD) treatment guidelines, place patients in 4 groups from A to D, dependent on severity of symptoms and assessment of risk; with Group A patients having the fewest symptoms and a low risk, as assessed by degree of airflow limitation and/or exacerbation history. COPD patients in Group D, who have many symptoms and a high risk of exacerbations and/or high airflow limitation, may need multiple inhaled therapies (inhaled corticosteroids [ICS], long-acting beta2-agonist [LABA] and long-acting muscarinic antagonist[LAMA]); this triple therapy can be prescribed in separate inhalers. Use of different inhaler types with different inhalation techniques and dosing regimens can add to treatment complexity, and also increase the potential for errors in inhaler use that reduce or preclude drug delivery to the site of action in the lungs [Cochrane, 2000, Van der Palen, 1999]. Fixed-dose combination inhalers that minimise the number of inhalers required would simplify treatment, improve adherence, reduce errors in inhaler use, and potentially lead to better outcomes [GOLD 2017].

The skill and ability of the COPD patient to use the prescribed inhaler/s correctly coupled with adequate training in inhaler technique are also critical to ensure effective drug delivery [Cochrane, 2000, Melani, 2011]. Given the time demands on healthcare professionals, a device which is simple to use and requires minimal time to train would be desirable [Bonini, 2015]. Further, patients may prefer easy-to-use inhalers having fewer steps to deliver drug; this has the potential to improve compliance and thereby impact outcomes.

8

CONFIDENTIAL 2017N324761_00206215

7

ELLIPTA DPI has been designed to be simple for patients to use. In COPD patients’ naïve to ELLIPTA and comparator inhalers, data has shown that patients make fewer critical and overall errors when using ELLIPTA as compared to other common DPIs tested [Van Der Palen, 2016(a)]. ELLIPTA DPI was also shown to be preferred by patients for a number of attributes, including number of steps and training time required to receive therapy [Van Der Palen, 2016(b), Van Der Palen, 2016(c), Komase, 2014]. ELLIPTA is already available to deliver a LABA/ICS combination (BREO ELLIPTA), LAMA (INCRUSE ELLIPTA), and LAMA/LABA combination (ANORO ELLIPTA). Closed triple therapy (ICS/LAMA/LABA) is currently being assessed in clinical studies to deliver all 3 active treatments from a single ELLIPTA DPI.

Please note that Turbohaler and ACCUHALER are the names commonly used within the United Kingdom (UK) for these inhalers. However, in other regions and for the purposes of this study, these inhalers are referred to as Turbuhaler and DISKUS, respectively, in this clinical study report (CSR).

2. STUDY OBJECTIVES

Primary objective: To compare the number of critical errors made by COPD patients, after they had read the respective patient information leaflet(s) (PIL[s]), for each treatment option tested.

Secondary objectives:

To compare the number of critical errors made by COPD patients after instruction from the healthcare professional (HCP) for each treatment option tested;

To compare the number of overall (critical and non-critical) errors made by COPD patients, after a subject had read the PIL(s) or after instruction from the HCP foreach treatment option tested;

To compare the number of instructions (maximum of 2) from a HCP which wereneeded to demonstrate correct inhaler use;

To compare the training/teaching time required to demonstrate correct inhaler use;

Preference attributes for each treatment option tested.

9

CONFIDENTIAL 2017N324761_00206215

8

3. INVESTIGATORS AND STUDY ADMINISTRATIVE STRUCTURE

This clinical study was sponsored by GlaxoSmithKline (GSK). A total of 5 centres in 2 countries (the UK and the Netherlands) randomised subjects. This study was initiated on 30 December 2016 (first subject, first visit) and completed on 19 June 2017 (last subject, last visit).

GSK was responsible for the administration, supply of investigational products, monitoring, analyses, and reporting of this study. No independent data monitoring committee was involved.

This CSR was prepared by at Complete Regulatory, McCann Complete Medical, 19-21 King Edward Street, Macclesfield, UK, SK10 1AQ, under the guidance of GSK.

4. INVESTIGATIONAL PLAN

4.1. Study Design

This study was a multi-centre, randomised, open-label, placebo-device crossover study with a 2x2 complete block design, comprised of 2 sub-studies:

Sub-study 1 compared ELLIPTA (Treatment Option 1) to DISKUS plus HandiHaler combination (Treatment Option 2);

Sub-study 2 compared ELLIPTA (Treatment Option 1) to Turbuhaler plusHandiHaler combination (Treatment Option 3).

Each sub-study ran independently and in parallel with the other sub-study. All placebo DPIs were available for the start of each sub-study. The data from the 2 sub-studies wasreported as each sub-study completed, after the data for that sub-study had been cleaned. The database was locked when all of the sub-studies had completed and all data werecleaned.

The study had 2 visits (Visit 0 [V0] and Visit 1 [V1]); both visits could be completed on the same day.

The study subjects were randomised to order of treatment and preference questionnaire in each of the 2 sub-studies as noted in Figure 1.

10

PPD

CONFIDENTIAL 2017N324761_00206215

9

Figure 1 Study Schematic

Abbreviations: COPD = chronic obstructive pulmonary disease; HCP = health care professional; PIL = patient information leaflet; PREF = preference questionnaire; V = visit.

4.2. Discussion of Study Design

This study design has been used in previous studies assessing critical errors and/or preference of inhalers in patients with COPD who were naïve (defined as no experience within the preceding 24 months) to the inhalers tested. The cross-over design was chosen to allow the subjects to serve as their own control in their ability to use both the ELLIPTA DPI and the other combination DPIs. A single visit was suitable for assessing these endpoints and reduced the inconvenience of multiple visits for subjects.

Placebo inhalers were used to remove any bias due to treatment effect, avoid the need for wash-in/out periods, and avoid the need to withhold or discontinue current COPD medications, which could have affected the subject’s clinical status (improvement or decline) and perceptions.

4.3. Protocol Amendment(s)

The study protocol was not amended.

4.4. Selection of Study Population

Key inclusion and exclusion criteria are noted below. Full details for these criteria are provided in the Protocol (see Protocol Section 6.1 and Protocol Section 6.2, respectively).

11

CONFIDENTIAL 2017N324761_00206215

10

4.4.1. Inclusion/Exclusion Criteria

4.4.1.1. Key Inclusion Criteria

Males or non-pregnant females who were current or former smokers, ≥40 years of age at Visit 1, with a diagnosis of COPD and a documented history of COPD in accordance with the definition by the European Respiratory Society [Celli, 2004].

Current COPD therapy: Currently receiving maintenance therapy with a fixed dose combination of a LABA and ICS. Subject could also be receiving a LAMA. Subjects were required to continue using their prescribed COPD maintenance inhaler therapy throughout the study, and could use short acting beta-adrenergic agonist (SABA) and/or short acting muscarinic antagonist rescue medications as needed.

Subjects were required to have been on current maintenance ICS/LABA COPD treatment for at least 4 weeks prior to V0 and were evaluated as unlikely to change treatment within 4 weeks of V1.

4.4.1.2. Key Exclusion Criteria

Subjects who had a current diagnosis of asthma were excluded. Subjects with a prior history of asthma were eligible if they had a current diagnosis of COPD.

Subjects who had recent experience with the ELLIPTA inhaler or any capsule system inhaler (e.g., Spiriva HandiHaler, Seebri/Ultibro Breezhaler), or who had participated in any clinical studies using these inhalers (including placebo inhalers) within 24 months prior to V0. Dependent on which sub-study a subject was included in, they were not tohave had any recent experience (within 24 months of V0) of the following inhalers: Sub-study 1, DISKUS inhaler (e.g., SERETIDE DISKUS or placebo DISKUS); Sub-study 2, Turbuhaler (e.g., Symbicort Turbuhaler or placebo Turbuhaler).

Subjects who had a history of hypersensitivity to any component of the study inhaler (e.g., lactose, magnesium stearate) and subjects with a history of severe milk protein allergy that, in the opinion of the study physician, contraindicated participation were also excluded. In the opinion of the investigator, any subject who was unable to read and/or not able to complete a questionnaire and understand verbal instructions was excluded from the study.

4.4.2. Withdrawal Criteria

Subjects could withdraw from the study at any time at his/her own request, or could be withdrawn at any time at the discretion of the investigator for safety, behavioural, compliance or administrative reasons.

If the participant withdrew consent for disclosure of future information, the sponsor could retain and continue to use any data collected before such a withdrawal of consent. As this was a single visit, placebo-only study, if a subject withdrew during this single visit, no assessments were required following withdrawal.

12

CONFIDENTIAL 2017N324761_00206215

11

4.5. Treatments

4.5.1. Investigational Product(s) and Reference Therapy

The study involved the use of placebo inhalers (placebo ELLIPTA, placebo DISKUS, placebo Turbuhaler and placebo HandiHaler) that did not contain active treatments; subjects continued to take their own prescribed COPD medication and other concomitant medications for the duration of the study.

Placebo treatments were administered to subjects as directed in the respective PILs; adescription of the DPIs and their batch numbers are provided in Table 1.

Table 1 Description of DPIs

TreatmentELLIPTA Placebo DPI

DISKUS Placebo DPI

HandiHaler DPI and Placebo Capsules

Turbuhaler Placebo DPI

Dosage Formulation

Placebo DPI containing 2 strips with 30 blisters per strip.1st strip: lactose monohydrate2nd strip: lactose monohydrate blended with magnesium stearate

Placebo DPI with 1 blister strip containing lactose monohydrate.

DPI with placebo capsules containing lactose monohydrate

Placebo DPI containing lactose monohydrate

Manufacturer GSK GSKBoehringer -Ingelheim

AstraZeneca

Device NA NA HandiHaler NARoute of Administration

Oral Inhalation Oral Inhalation Oral Inhalation Oral Inhalation

Dosing Instructions

As directed As directed As directed As directed

Storage Instructions

Up to 25°C (77°F) 2-25°C (36-77°F)

Up to 30°C (86°F) At or below 25°C (77°F). Not to be frozen.

Batch/Lot Numbers

162397455 6ZP5067 YZ0005 HAAM

4.5.2. Treatment Assignment

All participants were centrally randomised using an Interactive Web Response System (IWRS). Participants were assigned in equal numbers to 1 of 4 sequences for the sub-study they were included in, according to the randomisation schedule generated using the Randomisation and Medication Ordering System (RAMOS NG) prior to the start of the study. Subjects were randomised to treatment and preference questionnaire to reduceany potential bias.

13

CONFIDENTIAL 2017N324761_00206215

12

Treatment Option 1 required subjects to demonstrate use of a single inhaler (ELLIPTA), while Treatment Option 2 and 3 required the subject to demonstrate use of 2 inhalers (DISKUS plus HandiHaler in Sub-study 1 or Turbuhaler plus HandiHaler in Sub-study 2, respectively).

The possible sequences for each sub-study are shown in Table 2 and Table 3.

Table 2 Sub-study 1: Treatment Sequences

Sequence Period 1 Period 2Preference

QuestionnaireA ELLIPTA DISKUS + HandiHaler 1B DISKUS + HandiHaler ELLIPTA 2C ELLIPTA DISKUS + HandiHaler 2D DISKUS + HandiHaler ELLIPTA 1

Table 3 Sub-study 2: Treatment Sequence


QuestionnaireE ELLIPTA Turbuhaler + HandiHaler 3F Turbuhaler + HandiHaler ELLIPTA 4G ELLIPTA Turbuhaler + HandiHaler 4H Turbuhaler + HandiHaler ELLIPTA 3

4.5.3. Blinding

This was an open-label study.

4.5.4. Prior and Concomitant Medications and Non-Drug Therapies

A detailed history of previous and ongoing COPD medications, in particular any types of inhalers used to deliver these medications for the 24 months prior to V0, was recorded in order to inform on the subject’s eligibility.

As V0 and V1 could occur on a single visit, subjects were to continue taking their prescribed COPD and other medications.

Any medication or vaccine (including over-the-counter or prescription medicines, vitamins, and/or herbal supplements) that the participant was receiving at the time of enrolment or received during the study were recorded.

4.5.5. Compliance

Participants received placebo study treatment only at the clinic directly under the supervision of the investigator or designee.

14

CONFIDENTIAL 2017N324761_00206215

13

4.6. Study Assessments and Procedures

Full details of the study efficacy assessments, including the schedule of activities areprovided in the Protocol, Section 2.

4.6.1. Efficacy Assessment

For full details of the efficacy assessments see Protocol, Section 9.1.

The primary endpoint was to assess the percentage of subjects making at least 1 critical error after reading the PIL(s).

The secondary endpoints were to assess:

The percentage of subjects making at least 1 critical error after receiving the first, and second, instruction from a HCP;

The percentage of subjects making at least 1 overall error after reading the PIL(s), following the first, and second, instruction from a HCP;

The number of instructions (0, 1 or 2) required from the HCP to demonstrate correct inhaler use;

The amount of teacher/training time taken to: 1) read the PIL and demonstrate correct inhaler use (referred to as T1); 2) be given instruction on use of the inhaler by the HCP (up to 2 times) and to demonstrate correct inhaler use (referred to as T2); and 3) the total amount of time taken to demonstrate correct inhaler use (T1+T2);

Treatment preference based on responses to the preference questionnaire, which considered the number of steps required to take the COPD medication and overall treatment preference.

Further details for each of the endpoints are given in the following sections.

4.6.1.1. Assessment of Errors in Use of Device

Subjects were provided with the relevant section of the PIL which explained the correct use of each inhaler they were to be tested on. A critical error was defined as an error that was most likely to result in no, or a significantly reduced amount, of medication being inhaled by the subject. For further information on the critical and overall errors for each of the inhalers and how these were defined see the Protocol, Appendix 2. The process flow for assessment of errors in inhaler use is illustrated in Figure 2.

15

CONFIDENTIAL 2017N324761_00206215

14

Figure 2 Assessment of Errors in Use for Each DPI Tested

Abbreviations: DPI = dry powder inhaler; HCP = health care professional; PIL = patient information leaflet.

4.6.1.2. Number of Instructions Needed to Demonstrate Inhaler Use

As previously noted in Figure 2, subjects were asked to read the PILs for each inhaler prior to demonstrating use. If a subject made any error in the use of the inhaler, the HCP provided instruction to the subject, and the subject then demonstrated the inhaler use again. The number of instructions from a HCP (maximum of 2) required to demonstrate correct inhaler use was captured.

4.6.1.3. Time Taken to Use Device

Time taken to use the device was recorded as follows:

T1: the time from when the subject started to read the PIL(s) until they hadcompleted demonstration of device use (i.e., with no investigator support);

T2: the time from when the HCP started to instruct the subject until correct use was demonstrated (up to a maximum of 2 attempts only);

T1+T2: the time from when the subject started to read the PIL(s) until correct use was demonstrated (up to a maximum of 3 attempts, once after reading the PIL(s) and twice following instruction by the HCP).

4.6.1.4. Assessments of Preference

After completing the errors-in-use assessment for both treatments, the HCP asked the subject questions from an assigned preference questionnaire. The wording of the questionnaires is provided in the Protocol, Appendix 3.

16

CONFIDENTIAL 2017N324761_00206215

15

4.6.2. Safety Assessments

There were no mandated safety assessments in the study, other than monitoring for adverse events (AEs) and serious adverse events (SAEs) as described in the Protocol,Section 9.2.

The investigator or site staff were responsible for detecting, documenting and reporting events that met the definition of an AE or SAE. AE information volunteered by the subject, discovered by investigator questioning or detected by other means was collected from the start of study treatment until the follow-up contact. The following information on AEs was obtained:

Duration (start and stop dates);

Severity (mild, moderate, severe);

Causality (reasonable possibility yes/no);

Actions taken and outcome.

The AE and SAE definitions are provided in the Protocol, Appendix 5.

4.7. Data Quality Assurance

Subject data was entered into GSK-defined electronic case record forms (eCRFs) andtransmitted electronically to GSK.

Verification of data accuracy and adherence to protocol requirements was achieved through regular monitoring visits at each investigational site. Subsequent data handling and reporting processes were performed according to processes detailed in GSK’s Standard Operating Procedures. AEs and concomitant medications were coded using company standard dictionaries, Medical Dictionary for Regulatory Activities and GSKDrug.

SAEs (if observed) were entered into the database and quality assured, including reconciliation with the Global Clinical Safety and Pharmacovigilance database.

All investigators and responsible study site staff attended an investigator training meeting and/or separate study site initiation visit to review study protocol procedures, study requirements, and Good Clinical Practice (GCP) responsibilities. Investigators and staff were given opportunity to discuss any aspect of the study protocol and GCP requirements. Training records were reviewed to ensure investigators and staff were qualified to conduct the study and to document training in GCP. Any staff lacking in GCP training were either sent to a GCP training course or provided an electronic GCP training module. Documentation of GCP training was confirmed prior to staff participation in the study.

Principal investigators (PIs) signed the investigator page of the protocol to confirm their commitment to conduct the study in accord with the protocol and GCP. The signed documents have been archived within individual investigator study files.

17

CONFIDENTIAL 2017N324761_00206215

16

In accordance with applicable regulations, GCP and GSK procedures, GSK monitors contacted the site prior to the start of the study to review with the site staff the protocol, study requirements, and their responsibilities to satisfy regulatory, ethical, and GSK requirements. When reviewing data collection procedures, the discussion included identification, agreement and documentation of data items for which the eCRF served as the source document. GSK monitored the study and site activity to verify that: (1) the data are authentic, accurate, and complete; (2) the safety and rights of subjects were protected; (3) the study was conducted in accordance with the currently approved protocol and any other study agreements, GCP and all applicable regulatory requirements.

All protocol deviations collected during the study were reviewed by the GSK study team in order to identify important protocol deviations. Consistent with ICH E3 guidance, only protocol deviations identified as “important” are provided in this clinical study report. Important deviations are defined as deviations that were likely to affect the interpretation of the results and/or led to exclusion of any subject data from an analysis. Important deviations include, but are not limited to, those related to study inclusion or exclusion criteria, adherence to the protocol, conduct of the study, subject management or subject assessment.

4.8. Statistical Analyses

Full details of all data handling conventions and statistical analyses conducted for this study are provided in the Reporting and Analysis Plan (RAP). All programming was performed in a HARP environment using SAS Version 9 or a later release. All analyses were conducted after the database was locked. All endpoints were analysed by individual sub-study.

4.8.1. Sample Size Considerations

The sample size calculation for each sub-study was based on the primary endpoint. Based on the results from GSK studies 201301 and 201330 [GSK Document Number2015N230775_00 and GSK Document Number 2015N230821_01], a range of critical error rates (at least 1 critical error using each device after reading the PIL(s) for each of the treatment options (ELLIPTA, DISKUS plus HandiHaler and Turbuhaler plusHandiHaler) were explored. Using 10,000 simulations, a total of 72 subjects in each sub-study would provide at least 90% power to show a statistically significant difference between the critical error rate of each of the paired treatment options assuming the critical error rates specified in the RAP, Section 1. Conditional logistic regression and a 2-sided 5% significance level were used as the analysis method in the simulations. No withdrawals were expected, however up to 80 subjects were planned to be randomised to each sub-study to ensure 72 subjects were evaluable in each sub study.

Further details regarding the simulations for determination of sample size are provided in the Protocol, Section 10.2 and the RAP.

4.8.2. Analysis Populations

Details of the study analysis populations are provided in Table 4.

18

CONFIDENTIAL 2017N324761_00206215

17

Table 4 Analysis Populations

Population Definition/CriteriaAll Subjects Enrolled (ASE)

All participants who signed the informed consent form and for whom a record existed in the study database, including screen failures and any subject who was not screened but experienced an SAE between the date of informed consent and the planned date of the Screening visit.

Randomised All participants who were randomised.Intent-to-Treat (ITT) All randomised subjects (excluding those who were randomised in error) who

made at least 1 critical error assessment from 1 treatment option device. A subject who was recorded as a screen or run-in failure and was also randomised was considered to be randomised in error provided they had not performed any error assessments.

Safety This population was the same as the ITT population.

4.8.3. Study Population Analysis

Summaries of subject disposition, protocol deviations, populations analysed, demographic, baseline characteristics and concomitant medications were produced as described in Section 6 of the RAP. Unless stated otherwise, all study population summaries, analyses, figures and listings were performed on the ITT Population.

4.8.4. Efficacy Analysis

All efficacy analyses were based on the ITT Population.

The primary endpoint of the percentage of subjects making at least 1 critical error after reading the PIL(s) was planned to be analysed using conditional logistic regression with subject as fixed strata, and treatment option and period as fixed effects (exact logistic regression was actually used, see Section 4.8.8 for details). The odds ratio, 95% confidence interval (CI) and p-value were presented for the comparison between treatment options and were based on a 2-sided hypothesis testing approach of superiority. This analysis approach was also used for the secondary endpoints of number of critical errors made after the first HCP instruction and number of overall errors made (see RAP, Section 7 and Section 8). The endpoint of number of instructions from the HCP was analysed using the Wilcoxon signed rank test.

Error data were summarised for each sub-study overall. In addition, data were summarised for each sub-study by country. It was considered that data by country (the UK and the Netherlands) did not provide any additional information to that of the ITT Population and, therefore is not presented in this CSR.

A sensitivity analysis was performed on the primary endpoint and the secondary endpoint of overall errors after reading the PIL using the Cochran-Mantel-Haenszel test. This test served as a stratified approximation to the Mainland-Gart test, a variation of a one-sample chi-square test that accounted for study inhaler sequence (period). A subject who had an error with both treatment options or who had no errors with both treatment options did not provide any information about the superiority of either treatment option. Only those

19

CONFIDENTIAL 2017N324761_00206215

18

subjects who had error(s) in 1 treatment option and had no error in the other treatment option were accounted for in the Mainland-Gart test.

Refer to RAP, Section 8 and Section 9 for further efficacy analysis details.

4.8.5. Safety Analysis

All safety summaries and listings were produced for the ITT Population. Any AEs (regardless of causality), SAEs and AEs leading to early withdrawal from the study were summarised and listed. Any AEs/SAEs were reported for each sub-study by system organ class and preferred term. A summary for cardiovascular events was also produced. Full details of the safety analyses are provided in Section 9.2 of the RAP.

4.8.6. Interim Analyses

No interim analyses were planned or conducted.

4.8.7. Final Analyses

All analysis was performed after the completion of the following sequential steps:

1. All subjects had completed the study as defined in the Protocol for both Sub-study 1 and Sub-study 2;

2. All required database cleaning activities had been completed and final database release and database freeze had been declared by Data Management;

3. All criteria for unblinding the randomisation codes had been met;

4. Randomisation code schedules for both sub-studies had been distributed according to RandAll NG procedures.

4.8.8. Changes in Conduct of the Study or Planned Analyses

One change was applied to the statistical analysis specified in the Protocol, dated 15 Sep 2016. For the secondary endpoint of “the number of instructions (0, 1 or 2 times) from the HCP which were needed to demonstrate correct inhaler use”, changes made to the draft Protocol were inadvertently omitted from the final Protocol. As such, the analysis stated in the Protocol (logistic regression) was not correct. The correct analysis, using the Wilcoxon signed rank test, was planned and documented in the RAP prior to database freeze.

In addition, there were 2 changes made to the analysis detailed in the RAP:

The planned analysis for the primary endpoint was conditional logistic regression. Due to zero cell counts this model failed to converge and so the method was amended to exact logistic regression. This modification was made following database freeze.

20

CONFIDENTIAL 2017N324761_00206215

19

A post-hoc table presenting the probabilities of total time to demonstrate correct inhaler use by 5-minute intervals was produced. This table supported the pre-planned figure of Kaplan-Meier probabilities.

21

CONFIDENTIAL 2017N324761_00206215

20

5. STUDY POPULATION RESULTS

In this section, where appropriate, data has been presented for the total population.

5.1. Subject Disposition

5.1.1. Screen Failures

A total of 160 subjects were screened for this study, none of whom were screen failures(Table 1.02).

5.1.2. Randomised Subjects

Five investigational centres (3 in the Netherlands and 2 in the UK) randomised a total of 160 subjects; 80 to Sub-study 1 and 80 to Sub-study 2. Eighty subjects in Sub-study 1 and 79 subjects in Sub-study 2 were included in the ITT Population. The number of subjects randomised at each centre ranged from 13 to 76 (Table 1.08).

Most subjects completed this single-visit study (159 subjects [>99%]). One subject (<1%) in Sub-study 2 withdrew from the study (Table 5); the subject withdrew consent prior to device assessment (Table 1.04).

Table 5 Subject Disposition (ASE Population)

Number (%) of Subjects

Sub-study 1(N=80)

Sub-study 2(N=80)

Total(N=160)

Completion StatusCompleted 80 (100) 79 (99) 159 (>99)Withdrawn 0 1 (1) 1 (<1)

Primary Reason1/Sub Reason2 for WithdrawalWithdrew consent 0 1 (1) 1 (<1)

Source: Table 1.051. Subjects only record 1 primary reason for withdrawal.2. Subjects were not required to indicate sub-reasons.

5.2. Protocol Deviations

In total, 2 subjects (1%), who had not met all of the inclusion/exclusion criteria were randomised and performed at least 1 critical error assessment from 1 treatment option device (Table 1.12). Subject (Sub-study 1) had experience of using the ELLIPTA inhaler within the 24 months prior to the study and Subject (Sub-study 2) had a current diagnosis of asthma (Listing 4). All subjects completed the study and were included in the ITT Population.

In addition, 2 subjects (Subject and Subject ) both in Sub-study 1, were identified after database freeze as not taking an ICS/LABA fixed dose combination as per the inclusion criteria. These inclusion criteria deviations were not recorded in the database

22

PPD PPD

PPD

PPD

CONFIDENTIAL 2017N324761_00206215

21

and thus are not included in Table 1.12 (see Section 5.5.1). Both subjects completed the study and were included in the ITT Population.

In total, important protocol deviations were reported for 6 subjects (4%, Table 1.11); these included informed consent deviations (2 subjects [1%]; the PI was not present at the informed consent procedure and signed after the subject visit date); eligibility criteria not met (2 subjects [1%], as described earlier in this section); assessment or time point completion deviations (1 subject [<1%] did not have vital signs assessed at V1); and wrong study treatment/administration/dose (1 subject [<1%] filled in Questionnaire 1, but was randomised to Questionnaire 2) (Listing 3).

At the time of this report, no GCP noncompliance issues were identified by monitoring or audit.

5.3. Populations Analysed

Definitions for the populations analysed are provided in Section 4.8.2. A total of 160 subjects were included in both the ASE and randomised populations and all but 1 of these subjects (159 subjects [>99%]) were included in the ITT Population. For this study, the safety population is identical to the ITT population (Table 6).

Table 6 Summary of Subject Populations

PopulationNumber (%) of Subjects

Sub-study 1 Sub-study 2 TotalASE 160Randomised 80 80 160ITT1 80 (100) 79 (99) 159 (>99)Safety1,2 80 (100) 79 (99) 159 (>99)Source: Table 1.011. Percentages are of the randomised population.2. The safety and ITT populations are identical.Abbreviations: ASE = all subjects enrolled; ITT = intent-to-treat.

5.4. Demographics and Baseline Characteristics

5.4.1. Demographics

Demographic characteristics were similar across the 2 sub-studies (Table 7). The majority of subjects were White (74%) and there was a similar proportion of males (52%) and females (48%) in the study overall. The mean age was 65.0 years.

23

CONFIDENTIAL 2017N324761_00206215

22

Table 7 Demographic Characteristics (ITT Population)

Sub-study 1(N=80)

Sub-study 2(N=79)

Total(N=159)

Age (years)Mean (SD) 64.3 (8.69) 65.7 (8.49) 65.0 (8.60)Median (Min – Max) 65.5 (48 – 82) 66.0 (48 – 92) 66.0 (48 – 92)

Sex, n (%)Female 39 (49) 37 (47) 76 (48)Male 41 (51) 42 (53) 83 (52)

Race, n (%)African American/African Heritage 1 (1) 0 1 (<1)American Indian or Alaskan Native 0 0 0Asian 21 (26) 20 (25) 41 (26)

Central/South Asian Heritage 20 (25) 20 (25) 40 (25)Japanese/East AsianHeritage/South East Asian Heritage

1 (1) 0 1 (<1)

Native Hawaiian or Other Pacific Islander

0 0 0

White 58 (73) 59 (75) 117 (74)Arabic/North African Heritage 5 (6) 4 (5) 9 (6)White/Caucasian/European Heritage

53 (66) 55 (70) 108 (68)

Ethnicity, n (%)Hispanic/Latino 0 0 0Not Hispanic/Latino 80 (100) 79 (100) 159 (100)

Height (cm)n 80 78 158Mean (SD) 168.4 (9.86) 169.6 (9.64) 169.0 (9.74)Median (Min – Max) 169.5 (147 – 191) 170.0 (150 – 199) 170.0 (147 – 199)

Weight (kg)n 80 78 158Mean (SD) 75.95 (18.098) 78.10 (14.883) 77.01 (16.571)Median (Min – Max) 73.90 (43.6 – 122.4) 75.85 (50.6 – 125.4) 74.35 (43.6 – 125.4)

Body Mass Index (kg/m2)n 80 78 158Mean (SD) 26.69 (5.607) 27.12 (4.568) 26.90 (5.109)Median (Min – Max) 26.08 (16.5 – 43.4) 26.21 (19.1 – 40.5) 26.21 (16.5 – 43.4)

Source: Table 1.06, Table 1.09 and Table 1.10Note: Subject performed the informed consent procedure and was randomised, however the subject's vital signs were not collected at V1.Abbreviations: ITT = intent-to-treat; Max = maximum; Min = minimum; N = sample size; n = subset of sample size; SD = standard deviation; V = visit.

5.4.2. Baseline Characteristics

5.4.2.1. Current and Past Medical Conditions

Sixty-two percent of subjects in the ITT Population had a current medical condition in addition to COPD: 51% of subjects in Sub-study 1 and 73% of subjects in Sub-study 2

24

PPD

CONFIDENTIAL 2017N324761_00206215

23

(Table 1.13). The most common current medical conditions ongoing at baseline (10% or higher incidence) were in the categories of metabolism and nutrition disorders (50% of subjects, with hypercholesterolaemia and diabetes mellitus being the most commonly reported conditions in this category [reported in 34% and 25% of subjects, respectively]), vascular disorders (34% of subjects, all of whom had hypertension) and cardiac disorders (14% of subjects, with coronary artery disease being the most commonly reported condition in this category [reported in 10% of subjects]).

Past medical conditions were reported by 38% of subjects: 36% of subjects in Sub-study 1 and 39% of subjects in Sub-study 2 (Table 1.14). The highest incidence was in the category of infections and infestations (22% of subjects, all of whom had pneumonia).

5.4.2.2. Family History of Cardiovascular Risk Factors

In the ITT Population, 39 subjects (25%) had a family history of cardiovascular risk factors (Table 1.15); 87 subjects (55%) had no family history and, for 33 subjects (21%) their family history was unknown.

5.4.2.3. COPD and Exacerbation History

The majority of subjects (104 subjects [65%]) had COPD between 1 and 10 years (Table 8). A higher proportion of subjects in Sub-study 2 (21 subjects [27%]) had COPD between 10 and 15 years compared with Sub-study 1 (12 subjects [15%]); conversely a higher proportion of subjects in Sub-study 1 (30 subjects [38%]) had COPD between 1 and 5 years compared with Sub-study 2 (21 subjects [27%]). Two subjects (1%, both in Sub-study 1) had COPD for less than a year and 8 subjects (5%) had COPD for 20 years.

Table 8 Summary of COPD History (ITT Population)

Duration of COPD


Sub-study 1(N=80)

Sub-study 2(N=79)

Total(N=159)

6 months to <1 year 2 (3) 0 2 (1)

1 year to <5 years 30 (38) 21 (27) 51 (32)

5 years to <10 years 26 (33) 27 (34) 53 (33)

10 years to <15 years 12 (15) 21 (27) 33 (21)

15 years to <20 years 6 (8) 6 (8) 12 (8)

20 years to <25 years 1 (1) 2 (3) 3 (2)

25 years 3 (4) 2 (3) 5 (3)Source: Table 1.16

5.4.2.4. Smoking History

In the ITT Population, 85 subjects (53%) were current smokers and 74 subjects (47%) were former smokers. Subjects had smoked a mean of 17.9 cigarettes a day and had smoked for a mean of 41.1 years (36.4 pack years) (Table 1.17).

25

CONFIDENTIAL 2017N324761_00206215

24

5.5. Prior and Concomitant Medications

5.5.1. COPD Medications

All subjects were taking concomitant COPD medications during the study (Table 9). However the following data discrepancies were noted post-database freeze. Four subjects had errors in COPD concomitant medication records and were not coded/positioned correctly for taking an inhaled ICS, in combination with a LABA.

Subject (Sub-study 1): Recorded fostair in the eCRF (ICS/LABA); route incorrectly recorded in the eCRF as oral; Respiratory medication class (RMC) coding methodology for all steroids considers the route and thus this did not code as an as ICS.

Subject (Sub-study 2): Recorded fluticasone propionate in a combination with salbutamol, via the Evohaler in the eCRF. This coded as an ICS and SABA. However, this combination was not available and the site confirmed this should have been fluticasone propionate in combination with salmeterol and thus coded as ICS and LABA.

Subject (Sub-study 1): Recorded symbicort (ICS/LABA) in the eCRF; route incorrectly recorded in the eCRF as gastrostomy tube; RMC coding methodology for all steroids considers the route and thus this did not code as an ICS.

Subject (Sub-study 2): Recorded duoresp (ICS/LABA) in the eCRF as a non-COPD medication. Non-COPD medications do not undergo RMC coding and so was not coded as an ICS or LABA for respiratory medication reporting.

Further to this, 2 subjects (Subject and Subject ), both in Sub-study 1, were identified as not taking an ICS/LABA fixed dose combination as per the inclusion criteria and therefore should have been recorded as protocol deviations in the eCRF. Both subjects were taking a LABA (striverdi) but were not taking an ICS. Both these subjects were also taking tiotropium (LAMA).

These errors are reflected in Table 9. However, for Sub-study 1 there were 80 subjects taking a LABA; of those subjects, 78 were taking it in a fixed dose combination with an ICS. For Sub-study 2, all 79 subjects were taking a LABA/ICS fixed dose combination. These errors are recorded here but will not be corrected in the relevant source tables (Table 1.12 and Table 1.18).

26

PPD

PPD

PPD

PPD

PPD

PPD

CONFIDENTIAL 2017N324761_00206215

25

Table 9 Summary of COPD Medications (ITT Population)

Respiratory Class


Sub-study 1(N=80)

Sub-study 2(N=79)

Total(N=159)

Any medication 80 (100) 79 (100) 159 (100)LABA 80 (100) 77 (97)1 157 (99)ICS 76 (95)1 78 (99)1 154 (97)Short-acting beta-2 agonist 51 (64) 62 (78) 113 (71)Long-acting anticholinergic 34 (43) 29 (37) 63 (40)Short-acting anticholinergic 21 (26) 14 (18) 35 (22)Other COPD medication 5 (6) 4 (5) 9 (6)Corticosteroid (systemic, oral, parenteral, intra-articular)

7 (9) 0 7 (4)

Anti-infectives (antibiotics, antifungals, antivirals, antiseptics)

2 (3) 2 (3) 4 (3)

Corticosteroid (other) 2 (3) 1 (1) 3 (2)Leukotriene receptor antagonist 2 (3) 0 2 (1)Mucolytics 0 1 (1) 1 (<1)

Source: Table 1.181. Due to an error in coding respiratory medication routes, not all subjects taking ICS or LABA were included in the

source table. For Sub-study 1, 78 subjects (98%) were taking a fixed dose combination with an ICS. For Sub-study 2, 79 subjects (100%) were taking a LABA and an ICS fixed dose combination.

Abbreviations: COPD = chronic obstructive pulmonary disease; ICS = inhaled corticosteroid; LABA = long-acting beta2-agonist.

5.5.2. Non-COPD Medications

Most subjects (135 subjects [85%]) were taking a concomitant non-COPD medication during the study (Table 1.19). The most frequently used non-COPD medications were in the Anatomical Therapeutic Chemical classifications of alimentary tract and metabolism (62% of subjects; omeprazole was the most commonly used medication [18% of subjects]) and cardiovascular system (60% of subjects; simvastatin was the most commonly used medication [19% of subjects]).

5.6. Prior Inhaler Use

Subjects were allocated to a sub-study based on which inhalers they were naïve to (defined as no experience within the preceding 24 months). All subjects in bothsub-studies were reported as naïve to the ELLIPTA inhaler, which was required for participation in the study (Table 10); however, post analysis, Subject in Sub-study 1 was identified as having experience with ELLIPTA in the previous 24 months (see Section 5.2). All subjects in Sub-study 1 and Sub-study 2 were also naïve to the other inhalers being assessed in that particular sub-study.

27

PPD

CONFIDENTIAL 2017N324761_00206215

26

Table 10 Subjects Naïve to Inhaler Devices (ITT Population)


Sub-study 1(N=80)

Sub-study 2(N=79)

Total(N=159)

ELLIPTA 80 (100)1 79 (100) 159 (100)DISKUS + HandiHaler 80 (100) 36 (46) 116 (73)Turbuhaler + HandiHaler 53 (66 ) 79 (100) 132 (83)Source: Table 1.201. Subject had recent experience with the ELLIPTA inhaler within 24 months prior to screening. However, the

subject had been recorded as being naïve to ELLIPTA in the clinical trial database and thus has been reported as such.

5.7. Exposure and Treatment Compliance

As this was a single visit, placebo-only study, exposure and compliance data were not collected.

28

PPD

CONFIDENTIAL 2017N324761_00206215

27

6. EFFICACY RESULTS

6.1. Critical Inhaler Use Errors

6.1.1. Critical Inhaler Use Errors after Reading the PIL

In both sub-studies, after reading the PIL, fewer subjects made 1 or more critical errorswhen using ELLIPTA (7 subjects [9%] in both sub-studies) compared with DISKUS plusHandiHaler (60 subjects [75%], Sub-study 1) or Turbuhaler plus HandiHaler (58 subjects [73%], Sub-study 2) (Figure 3, Figure 4 and Table 2.07).

These differences were statistically significant for both of the sub-studies (p<0.001 for both ELLIPTA versus DISKUS plus HandiHaler and ELLIPTA versus Turbuhaler plus HandiHaler) (Table 11).

All subjects who made critical errors when using ELLIPTA (7 subjects in both sub-studies) also made errors when using the comparator inhaler combination for the respective sub-study (Table 2.07).

Table 11 Analysis of Percentage of Subjects making at least 1 Critical Error after Reading the PIL (ITT Population)

Sub-study 1ELLIPTA(N=80)

DISKUS + HandiHaler (N=80)

At least 1 critical error, n (%) 7 (9) 60 (75)Zero critical errors, n (%) 73 (91) 20 (25)

DISKUS + HandiHaler vs. ELLIPTAOdds ratio (95% CI) 29.114 (11.047, infinity)p-value <0.001

Sub-study 2ELLIPTA (N=79)

Turbuhaler + HandiHaler (N=79)


Turbuhaler + HandiHaler vs. ELLIPTAOdds ratio (95% CI) 27.744 (10.512, infinity)p-value <0.001

Source: Table 2.09Note: Odds ratio, 95% CI and p-value obtained from a stratified exact logistic model. Subject was included in the model as fixed strata, treatment option was included in the exact statement and period was included as fixed effects.

On the error checklists, 4 types of critical error were possible with ELLIPTA, 5 with DISKUS and 7 with HandiHaler. The types of critical error reported in Sub-study 1 for each inhaler after reading the PIL are shown in Table 12. After reading the PIL, a total of 8 critical errors were made when using ELLIPTA compared with 139 critical errors when using DISKUS plus HandiHaler (41 critical errors using DISKUS and 98 critical errors using HandiHaler, Table 2.01 and Table 2.03).

29

CONFIDENTIAL 2017N324761_00206215

28

Three types of critical error were reported for subjects using ELLIPTA: exhaled directly into the mouthpiece (5/19 subjects [26%]), failed to open cover (2/19 subjects [11%]),and no seal by the lips round the mouthpiece during the inhalation (1/19 subjects [5%]).Five types of critical error were reported for DISKUS use, the most common of which was the lever was not pushed back (27/44 subjects [61%]); and 7 types of critical errorwere reported for HandiHaler use, the most common of which was the capsule did not rattle (43/57 subjects [75%]) (Table 12).

Table 12 Summary of Critical Errors after Reading PIL: Sub-study 1(ITT Population)

On the error checklists, 4 types of critical error were possible with ELLIPTA, 6 with Turbuhaler and 7 with HandiHaler. The types of critical error reported in Sub-study 2 for each inhaler after reading the PIL are shown in Table 13. After reading the PIL, a total of 13 critical errors were made when using ELLIPTA compared with 134 critical errors when using Turbuhaler plus HandiHaler (50 critical errors using Turbuhaler and 84 critical errors using HandiHaler, Table 2.01 and Table 2.05).

Four types of critical error were reported for subjects using ELLIPTA: exhaled directly into the mouthpiece (7/17 subjects [41%]), failed to open cover (3/17 subjects [18%]), no seal by the lips round the mouthpiece during the inhalation (2/17 subjects [12%]), and shook the device after dose preparation (1/17 subjects [6%]). Five types of critical errorwere reported for Turbuhaler use, the most common of which was the base was not

ELLIPTA vs. DISKUS + HandiHalerSub-study 1


Errors on ELLIPTA (19 subjects with any error)

Failed to open cover 2 (11)Exhaled directly into mouthpiece 5 (26)No seal by the lips round the mouthpiece during the inhalation 1 (5)

Errors on DISKUS (44 subjects with any error)

Failed to open cover 5 (11)Lever not pushed back 27 (61)Shook the device after dose preparation 1 (2)Exhaled directly into mouthpiece 7 (16)No seal by the lips round the mouthpiece during the inhalation 1 (2)

Errors on HandiHaler (57 subjects with any error)

Failed to remove capsule 7 (12)Failed to insert capsule into chamber 6 (11)Did not completely close device capsule chamber 17 (30)Capsule did not rattle 43 (75)Did not pierce the capsule 18 (32)Exhaled directly into mouthpiece 4 (7)No seal by the lips round the mouthpiece during the inhalation 3 (5)

Source: Table 2.01 and Table 2.03Note: Percentages for type of errors are calculated based on the total number of subjects with any errors.

30

CONFIDENTIAL 2017N324761_00206215

29

twisted fully backwards and forwards (27/45 subjects [60%]); and 7 types of critical error were reported for HandiHaler use, the most common of which was the capsule did not rattle (33/50 subjects [66%]) (Table 13).

Table 13 Summary of Critical Errors after Reading PIL: Sub-study 2 (ITT Population)

6.1.2. Critical Inhaler Use Errors after HCP Instruction

6.1.2.1. Critical Inhaler User Errors after First HCP Instruction

Subjects who made no errors after reading the PIL were not given instruction by the HCP and so were counted as having zero critical errors after the first instruction from the HCP. In both sub-studies, few subjects made a critical error when using any inhaler after the first instruction from the HCP. Despite this, a difference was still seen between the treatment options in terms of the critical errors reported. Fewer subjects made at least 1 critical error when using ELLIPTA (1 subject [1%] in Sub-study 1 and 2 subjects [3%] in Sub-study 2) compared with DISKUS plus HandiHaler (9 subjects [11%], Sub-study 1) or Turbuhaler plus HandiHaler (11 subjects [14%], Sub-study 2) (Figure 3, Figure 4 and Table 2.07). These differences were statistically significant for both of the sub-studies (p<0.05 for both ELLIPTA versus DISKUS plus HandiHaler and ELLIPTA versus Turbuhaler plus HandiHaler) (Table 14).

ELLIPTA vs. Turbuhaler + HandiHalerSub-study 2


Errors on ELLIPTA (17 subjects with any error)

Failed to open cover 3 (18)Shook the device after dose preparation 1 (6)Exhaled directly into mouthpiece 7 (41)No seal by the lips round the mouthpiece during the inhalation 2 (12)

Errors on Turbuhaler (45 subjects with any error)

Failed to remove cap 3 (7)Did not hold device upright during dose preparation 14 (31)Base not twisted fully backwards and forwards 27 (60)Exhaled directly into mouthpiece 5 (11)No seal by the lips round the mouthpiece during the inhalation 1 (2)

Errors on HandiHaler (50 subjects with any error)

Failed to remove capsule 7 (14)Failed to insert capsule into chamber 6 (12)Did not completely close device capsule chamber 16 (32)Capsule did not rattle 33 (66)Did not pierce the capsule 14 (28)Exhaled directly into mouthpiece 6 (12)No seal by the lips round the mouthpiece during the inhalation 2 (4)

Source: Table 2.01 and Table 2.05Note: Percentages for type of errors are calculated based on the total number of subjects with any errors.

31

CONFIDENTIAL 2017N324761_00206215

30

Table 14 Analysis of Percentage of Subjects making at least 1 Critical Error after First HCP Instruction (ITT Population)




DISKUS + HandiHaler vs. ELLIPTAOdds ratio (95% CI) 4.248 (1.416, infinity)p-value 0.029




Turbuhaler + HandiHaler vs. ELLIPTAOdds ratio (95% CI) 3.855 (1.394, infinity)p-value 0.026

Source: Table 2.09Note: Odds ratio, 95% CI and p-value obtained from a stratified exact logistic model. Subject was included in the model as fixed strata, treatment option was included in the exact statement and period was included as fixed effects.Subjects who made no errors after reading the PIL were not given instruction by the HCP and so were counted as having zero critical errors after the first instruction from the HCP.

In Sub-study 1, a single critical error was made when using ELLIPTA compared with 15 critical errors when using DISKUS plus HandiHaler (6 critical errors using DISKUS and 9 critical errors using HandiHaler, Table 2.01 and Table 2.03).

The types of critical error observed after the first HCP instruction in Sub-study 1 were(data presented out of number of subjects with any error):

ELLIPTA: Subject exhaled directly into the mouthpiece (1/4 subjects [25%]);

DISKUS: Lever was not pushed back (4/8 subjects [50%]), and subject exhaled directly into the mouthpiece (2/8 subjects [25%]);

HandiHaler: Device did not pierce the capsule (5/9 subjects [56%]), the capsule did not rattle (2/9 subjects [22%]), and subject failed to remove capsule and subject exhaled directly into mouth piece (1/9 subjects [11%] each).

In Sub-study 2, 2 critical errors were made when using ELLIPTA compared with 16 critical errors when using Turbuhaler plus HandiHaler (11 critical errors using Turbuhaler and 5 critical errors using HandiHaler, Table 2.01 and Table 2.05).

The types of critical error observed after the first HCP instruction in Sub-study 2 were(data presented out of number of subjects with any error):

ELLIPTA: Subject exhaled directly into the mouthpiece (2/4 subjects [50%]);

32

CONFIDENTIAL 2017N324761_00206215

31

Turbuhaler: Base of device not twisted fully backwards and forwards, so no click was heard (5/12 subjects [42%]), subject did not hold the device upright during dose preparation (3/12 subjects [25%]), and subject failed to remove cap, subject shook the device after dose preparation and subject exhaled directly into the mouthpiece (1/12 subjects [8%] each);

HandiHaler: The capsule did not rattle (3/7 subjects [43%]), and device did not pierce the capsule and subject exhaled directly into the mouthpiece (1/7 subjects [14%] each).

6.1.2.2. Critical Inhaler User Errors after Second HCP Instruction

Subjects who made no errors after reading the PIL or the first HCP instruction were counted as having zero critical errors after the second instruction from the HCP. In both sub-studies, very few subjects made a critical error when using any inhaler after the second instruction from the HCP. No subjects using ELLIPTA made a critical error in either sub-study. In Sub-study 1, 5 subjects (6%) made at least 1 critical error when using DISKUS plus HandiHaler and, in Sub-study 2, 4 subjects (5%) made at least 1 critical error when using Turbuhaler plus HandiHaler (Figure 3, Figure 4 and Table 2.07). Analyses of these data (where possible) did not show any statistically significant differences between ELLIPTA and either of the other inhaler combinations(Table 15), however, as the number of subjects was small, this analysis should be interpreted with caution.

Table 15 Analysis of Percentage of Subjects making at least 1 Critical Error after Second HCP Instruction (ITT Population)



At least 1 critical error, n (%) 0 5 (6)Zero critical errors, n (%) 80 (100) 75 (94)

DISKUS + HandiHaler vs. ELLIPTAOdds Ratio (95% CI) 2.000 (0.459, infinity)p-value 0.400



At least 1 critical error, n (%) 0 4 (5)Zero critical errors, n (%) 79 (100) 75 (95)

Turbuhaler + HandiHaler vs. ELLIPTAOdds Ratio (95% CI) 1.732 (0.397, infinity)p-value 0.500

Source: Table 2.09Note: Odds ratio, 95% CI and p-value obtained from a stratified exact logistic model. Subject was included in the model as fixed strata, treatment option was included in the exact statement and period was included as fixed effects.Subjects who made no errors after reading the PIL or the first HCP instruction were counted as having zero critical errors after the second instruction from the HCP.

33

CONFIDENTIAL 2017N324761_00206215

32

In Sub-study 1, 9 critical errors were made when using DISKUS plus HandiHaler (3 critical errors using DISKUS and 6 critical errors using HandiHaler, Table 2.03).

The types of critical error observed after the second HCP instruction in Sub-study 1 were(data presented out of number of subjects with any error):

DISKUS: Lever was not pushed back (3/3 subjects [100%]);

HandiHaler: Device did not pierce the capsule (3/4 subjects [75%]), subject did not completely close the device capsule chamber (2/4 subjects [50%]), and the capsule did not rattle (1/4 subjects [25%]).

In Sub-study 2, 7 critical errors were made when using Turbuhaler plus HandiHaler (4 critical errors using Turbuhaler and 3 critical errors using HandiHaler, Table 2.05).

The types of critical error observed after the second HCP instruction in Sub-study 2 were(data presented out of number of subjects with any error):

Turbuhaler: Subject did not hold the device upright during dose preparation andsubject shook the device after dose preparation (2/4 subjects [50%] each);

HandiHaler: The capsule did not rattle (2/2 subjects [100%]), and no seal by the lips round the mouthpiece during the inhalation (1/2 subjects [50%]).

6.1.3. Summary of Critical Inhaler Use Errors

Figure 3 and Figure 4 illustrate the number of subjects having at least 1 critical error after reading the PIL and following the first and second instruction from the HCP for Sub-study 1 and Sub-study 2, respectively. Across all inhaler types, the number of subjects with critical errors declined following each instruction.

Critical errors are summarised by country in Table 2.02, Table 2.04, Table 2.06 and Table 2.08.

34

CONFIDENTIAL 2017N324761_00206215

33

Figure 3 Proportion of Subjects with at least 1 Critical Error by Assessment: Sub-study 1, ELLIPTA vs. DISKUS plus HandiHaler (ITT Population)

Source: Figure 2.01

Figure 4 Proportion of Subjects with at least 1 Critical Error by Assessment: Sub-study 2, ELLIPTA vs. Turbuhaler plus HandiHaler (ITT Population)

Source: Figure 2.01

35

CONFIDENTIAL 2017N324761_00206215

34

6.2. Overall Inhaler Use Errors

6.2.1. Overall Inhaler Use Errors after Reading the PIL

The results for overall errors followed a similar trend to critical errors. In both sub-studies, after reading the PIL, fewer subjects had at least 1 overall error when using ELLIPTA (19 subjects [24%] in Sub-study 1 and 17 subjects [22%] in Sub-study 2) compared with DISKUS plus HandiHaler (64 subjects [80%], Sub-study 1) or Turbuhaler plus HandiHaler (63 subjects [80%], Sub-study 2) (Figure 5, Figure 6 and Table 2.01, Table 2.03 and Table 2.05). These differences were statistically significant for both of the sub-studies (p<0.001 for both ELLIPTA versus DISKUS plus HandiHaler and ELLIPTA versus Turbuhaler plus HandiHaler) (Table 16).

Table 16 Analysis of Percentage of Subjects making at least 1 Overall Error after Reading the PIL (ITT Population)



At least 1 overall error, n (%) 19 (24) 64 (80)Zero overall errors, n (%) 61 (76) 16 (20)

DISKUS + HandiHaler vs. ELLIPTAOdds ratio (95% CI) 24.539 (9.268, infinity)p-value <0.001




Turbuhaler + HandiHaler vs. ELLIPTAOdds ratio (95% CI) 17.974 (7.239, infinity)p-value <0.001

Source: Table 2.11Note: Odds ratio, 95% CI and p-value obtained from a stratified exact logistic model. Subject was included in the model as fixed strata, treatment option was included in the exact statement and period was included as fixed effects.

On the error checklists, 9 types of overall error were possible with ELLIPTA and DISKUS, 12 with Turbuhaler and 13 with HandiHaler.

In Sub-study 1, a total of 38 overall errors were made after reading the PIL when using ELLIPTA compared with 242 errors when using DISKUS plus HandiHaler (90 errors using DISKUS and 152 errors using HandiHaler, Table 2.01 and Table 2.03).

In Sub-study 2, a total of 43 overall errors were made after reading the PIL when using ELLIPTA compared with 239 errors when using Turbuhaler plus HandiHaler (106 errors using Turbuhaler and 133 errors using HandiHaler, Table 2.01 and Table 2.05).

36

CONFIDENTIAL 2017N324761_00206215

35

6.2.2. Overall Inhaler Use Errors after HCP Instruction

6.2.2.1. Overall Inhaler Use Errors after First HCP Instruction

Subjects who made no errors after reading the PIL were not given instruction by the HCP and so were counted as having zero overall errors after the first instruction from the HCP. In both sub-studies, few subjects made overall errors when using any inhaler after the first instruction from the HCP. Despite this, a difference was still seen between the treatment options in terms of the overall errors reported. Fewer subjects made an error when using ELLIPTA (4 subjects [5%] in both sub-studies) compared with DISKUS plus HandiHaler (12 subjects [15%], Sub-study 1) or Turbuhaler plus HandiHaler (16 subjects [20%], Sub-study 2) (Figure 5, Figure 6 and Table 2.01, Table 2.03 and Table 2.05). The difference was statistically significant for Sub-study 2 only (p<0.05 ELLIPTA versus Turbuhaler plus HandiHaler) (Table 17).

Table 17 Analysis of Percentage of Subjects making at least 1 Overall Error after First HCP Instruction (ITT Population)




DISKUS + HandiHaler vs. ELLIPTAOdds ratio (95% CI) 3.237 (1.124, infinity)p-value 0.067




Turbuhaler + HandiHaler vs. ELLIPTAOdds ratio (95% CI) 6.357 (2.219, infinity)p-value 0.003

Source: Table 2.11Note: Odds ratio, 95% CI and p-value obtained from a stratified exact logistic model. Subject was included in the model as fixed strata, treatment option was included in the exact statement and period was included as fixed effects. Subjects who made no errors after reading the PIL were not given instruction by the HCP and so were counted as having zero overall errors after the first instruction from the HCP.

In Sub-study 1, 6 overall errors were made after the first HCP instruction when using ELLIPTA compared with 29 errors when using DISKUS plus HandiHaler (13 errors using DISKUS and 16 errors using HandiHaler, Table 2.01 and Table 2.03).

In Sub-study 2, 6 overall errors were made after the first HCP instruction when using ELLIPTA compared with 36 errors when using Turbuhaler plus HandiHaler (21 errors using Turbuhaler and 15 errors using HandiHaler, Table 2.01 and Table 2.05).

37

CONFIDENTIAL 2017N324761_00206215

36

6.2.2.2. Overall Inhaler Use Errors after Second HCP Instruction

Subjects who made no errors after reading the PIL or the first HCP instruction were counted as having zero critical errors after the second instruction from the HCP. In bothsub-studies, very few subjects made overall errors when using any inhaler after the second instruction from the HCP. Despite this, a difference was still seen between the inhalers used in terms of the overall errors reported. Fewer subjects made an error when using ELLIPTA (3 subjects [4%] in Sub-study 1 and 1 subject [1%] in Sub-study 2)compared with DISKUS plus HandiHaler (6 subjects [8%], Sub-study 1) or Turbuhaler plus HandiHaler (5 subjects [6%], Sub-study 2) (Figure 5, Figure 6 and Table 2.01, Table 2.03 and Table 2.05). Analyses of these data (where possible) did not show any statistically significant differences between ELLIPTA and either of the other inhaler combinations (Table 18), however, as the number of subjects was small, this analysis should be interpreted with caution.

Table 18 Analysis of Percentage of Subjects making at least 1 Overall Error after Second HCP Instruction (ITT Population)




DISKUS + HandiHaler vs. ELLIPTAOdds Ratio (95% CI) NAp-value NA




Turbuhaler + HandiHaler vs. ELLIPTAOdds Ratio (95% CI) 1.732 (0.397, infinity)p-value 0.500

Source: Table 2.11Note: Odds ratio, 95% CI and p-value obtained from a stratified exact logistic model. Statistics are only presented when the model successfully converged. Subject was included in the model as fixed strata, treatment option was included in the exact statement and period included as fixed effects. Subjects who made no errors after reading the PIL or the first HCP instruction were counted as having zero critical errors after the second instruction from the HCP.

In Sub-study 1, 5 overall errors were made after the second HCP instruction when using ELLIPTA compared with 15 errors when using DISKUS plus HandiHaler (5 errors using DISKUS and 10 errors using HandiHaler, Table 2.01 and Table 2.03).

In Sub-study 2, 1 overall error was made after the second HCP instruction when using ELLIPTA compared with 13 errors when using Turbuhaler plus HandiHaler (6 errors using Turbuhaler and 7 errors using HandiHaler, Table 2.01 and Table 2.05).

38

CONFIDENTIAL 2017N324761_00206215

37

6.2.3. Summary of Overall Inhaler Use Errors

Figure 5 and Figure 6 illustrate the number of subjects having at least 1 overall error after reading the PIL and following the first and second instruction from the HCP for Sub-study 1 and Sub-study 2, respectively. Across all inhaler types, the number of subjects with overall errors declined following each instruction.

For overall errors, there was a single subject in Sub-study 2 who made an error using ELLIPTA who did not make an error when using Turbuhaler plus HandiHaler(Table 2.10).

Figure 5 Proportion of Subjects with at least 1 Error by Assessment: Sub-study 1, ELLIPTA vs. DISKUS plus HandiHaler (ITT Population)

Source: Figure 2.02

39

CONFIDENTIAL 2017N324761_00206215

38

Figure 6 Proportion of Subjects with at least 1 Error by Assessment: Sub-study 2, ELLIPTA vs. Turbuhaler plus HandiHaler (ITT Population)

Source: Figure 2.02

6.3. Sensitivity Analysis

Sensitivity analyses were performed on the number of critical errors and overall errors in each sub-study (further details are provided in the RAP, Section 7.1.2). The sensitivity analyses supported the primary and secondary endpoint analyses for critical and overall errors after reading the PIL. The number of subjects making at least 1 critical or overall error after reading the PIL remaining statistically significantly lower with ELLIPTA in both sub-studies (Table 19).

40

CONFIDENTIAL 2017N324761_00206215

39

Table 19 Sensitivity Analysis of Percentage of Subjects making at least 1 Error after Reading the PIL (ITT Population)

Sub-study 1Result

Critical ErrorsNumber of subjects with a discordant result1, n 53

At least 1 critical error on ELLIPTA and 0 critical errors on DISKUS + HandiHaler, n 0At least 1 critical error on DISKUS + HandiHaler and 0 critical errors on ELLIPTA, n (%) 53 (100)p-value2 <0.001

Overall ErrorsNumber of subjects with a discordant result1, n 45

At least 1 overall error on ELLIPTA and 0 critical errors on DISKUS + HandiHaler, n 0At least 1 overall error on DISKUS + HandiHaler and 0 critical errors on ELLIPTA, n (%)

45 (100)

p-value2 <0.001Sub-study 2

ResultCritical Errors

Number of subjects with a discordant result1 51At least 1 critical error on ELLIPTA and 0 critical errors on Turbuhaler + HandiHaler, n 0At least 1 critical error on Turbuhaler + HandiHaler and 0 critical errors on ELLIPTA, n (%)

51 (100)

p-value2 <0.001Overall Errors

Number of subjects with a discordant result1 48At least 1 overall error on ELLIPTA and 0 critical errors on Turbuhaler + HandiHaler, n 1 (2)At least 1 overall error on Turbuhaler + HandiHaler and 0 critical errors on ELLIPTA, n (%)

47 (98)

p-value2 <0.001Source: Table 2.10 Note: Subjects were only included in the analysis if they had discordant data.1. Defined as an error with 1 treatment and not the other.2. Analysis performed using the Cochran-Mantel-Haenszel test.

6.4. Number of Instructions Required from HCP to Demonstrate Correct Inhaler Use

After reading the PIL, the majority of subjects made no errors using ELLIPTA (76% and 78% in Sub-studies 1 and 2, respectively) and thus only 24% and 22% of subjects (Sub-study 1 and 2, respectively) required further instruction (in addition to reading the PIL) from the HCP (Table 20). For the comparator inhaler combinations used in each sub-study, the majority of subjects made an error and required further HCP instruction (Sub-study 1 [DISKUS plus HandiHaler] and Sub-study 2 [Turbuhaler plus HandiHaler]80% of subjects each).

Of the subjects who required instruction, the majority only required 1 instruction from the HCP to demonstrate correct inhaler use (Table 20). In each sub-study, fewer subjects required more than 1 instruction for ELLIPTA (4 subjects in each sub-study) compared with DISKUS plus HandiHaler (12 subjects) and with Turbuhaler plus HandiHaler

41

CONFIDENTIAL 2017N324761_00206215

40

(16 subjects). There was a statistically significant difference in the number of HCP instructions required to demonstrate correct use of the inhaler(s) between ELLIPTA and both of the inhaler combinations (p<0.001 for both sub-studies). Three subjects in Sub-study 1 and a single subject in Sub-study 2 failed to demonstrate correct inhaler use with ELLIPTA; 6 subjects failed to demonstrate correct use of DISKUS plus HandiHaler (Sub-study 1) and 5 subjects failed to demonstrate correct use of Turbuhaler plus HandiHaler (Sub-study 2).

Table 20 Summary and Analysis of Number of Instructions from the HCP (ITT Population)

Sub-study 1 Sub-study 2ELLIPTA

(N=80)

DISKUS + HandiHaler

(N=80)

ELLIPTA

(N=79)

Turbuhaler + HandiHaler

(N=79)Median, n (95% CI) 0 1.0 (1.0, 1.0) 0 1.0 (1.0, 1.0)Min, n 0 0 0 0Max, n 3 3 3 3Number of Instructions, n (%)

0 61 (76) 16 (20) 62 (78) 16 (20)1 15 (19) 52 (65) 13 (16) 47 (59)2 1 (1) 6 (8) 3 (4) 11 (14)Failed to demonstrate correct use

3 (4) 6 (8) 1 (1) 5 (6)

p-value1 <0.001 <0.001Source: Table 2.14 Note: This analysis did not take into account sequence of treatment options. For the DISKUS plus HandiHaler and Turbuhaler plus HandiHaler groups, if a subject failed to demonstrate correct use on 1 device and not the other, the subject was counted as having failed to demonstrate correct use.1. p-value has come from the Wilcoxon signed rank test.

6.5. Time Taken to Use Inhaler

Three time intervals were defined in assessing time taken to use the inhaler device (T1, T2, T1+T2; for a definition of these time intervals see Section 4.6.1.3). For DISKUS plus HandiHaler or Turbuhaler plus HandiHaler, the time for each device was added together to form the total time needed to demonstrate correct use at T1 and T2, respectively. T1, T2 and T1+T2 were censored for subjects who failed to demonstrate correct use at the end of each time period. For subjects who demonstrated correct inhaler use after reading the PIL, T2 was recorded as 0 minutes for the appropriate inhaler.

In both sub-studies, the median T1 could not be calculated for DISKUS plus HandiHaler and Turbuhaler plus HandiHaler, as more than 50% of data was censored for these inhalers (Table 21). The median T2, in both sub-studies, was shorter for subjects using ELLIPTA (0 minutes in both sub-studies) compared with DISKUS plus HandiHaler(2.89 minutes, Sub-study 1) and with Turbuhaler plus HandiHaler (3.12 minutes,Sub-study 2). Median T1+T2 was also shorter for subjects using ELLIPTA (2.68 and 2.58 minutes in Sub-studies 1 and 2, respectively) compared with DISKUS plus

42

CONFIDENTIAL 2017N324761_00206215

41

HandiHaler (10.57 minutes, Sub-study 1) and with Turbuhaler plus HandiHaler (11.30 minutes, Sub-study 2).

Table 21 Summary and Analysis of Time Taken to Demonstrate CorrectInhaler Use (ITT Population)

Sub-study 1 Sub-study 2ELLIPTA

(N=80)

DISKUS + HandiHaler

(N=80)

ELLIPTA

(N=79)


(N=79)Time taken to read the PIL and demonstrate correct inhaler use (T1)

Subjects with correct use, n (%) 61 (76) 16 (20) 62 (78) 16 (20)Subjects censored, n (%) 19 (24) 64 (80) 17 (22) 63 (80)Median time1, minutes 2.88 NA 2.78 NA

Time taken for HCP instruction and demonstrate correct inhaler use (T2)Subjects with correct use, n (%) 77 (96) 74 (93) 77 (99) 74 (94)Subjects censored, n (%) 3 (4) 6 (8) 1 (1) 5 (6)Median time1, minutes 0 2.89 0 3.12

Total time to demonstrate correct use (T1+T2)Subjects with correct use, n (%) 77 (96) 74 (93) 77 (99) 74 (94)Subjects censored, n (%) 3 (4) 6 (8) 1 (1) 5 (6)Median time1, minutes 2.68 10.57 2.58 11.30

Source: Table 2.12 Note: For the DISKUS plus HandiHaler and Turbuhaler plus HandiHaler groups, if a subject demonstrated correct use on only 1 of the 2 devices, the time to demonstrate correct use was censored at the total of the reading and/or instruction time for both devices. Subject did not demonstrate correct use for the ELLIPTA device after reading the PIL, but was able to after receiving the first HCP instruction. The time taken to demonstrate correct use at the HCP assessment was not recorded, therefore this subject's ELLIPTA data could only be included in the time taken to demonstrate correct use after reading the PIL assessment.1. Taken from Kaplan-Meier analysis. If more than 50% of the data was censored then the median was not

applicable.

A further post hoc analysis of the time taken to use the inhaler was produced (Table 22). For both Sub-study 1 and 2, 86% of subjects demonstrated correct use of the ELLIPTA device within the first 5 minutes compared with 10% of subjects using DISKUS plus HandiHaler (Sub-study 1) and 9% of subjects using Turbuhaler plus HandiHaler (Sub-study 2). In Sub-study 1, all subjects had demonstrated correct use of the ELLIPTA device within 15 minutes compared with 67% of subjects demonstrating the correct use of the DISKUS plus HandiHaler combination. In Sub-study 2, 97% of subjects had demonstrated correct use of the ELLIPTA device within 20 minutes compared with 82% of subjects demonstrating the correct use of the Turbuhaler plus HandiHaler combination.

The Kaplan-Meier plots showing the total time taken to demonstrate correct use of the respective inhalers for the 2 sub-studies are shown in Figure 7 and Figure 8.

43

PPD

CONFIDENTIAL 2017N324761_00206215

42

Table 22 Probability of Total Time Taken to Demonstrate Correct Inhaler Use by 5 Minute Intervals (ITT Population)


(%)


(%)

Probability of demonstrating correct use in:5 minutes or less 86 1010 minutes or less 97 4615 minutes or less 100 6720 minutes or less 100 82


(%)


(%)

Probability of demonstrating correct use in:5 minutes or less 86 910 minutes or less 97 4315 minutes or less 97 6920 minutes or less 97 82

Source: Table 2.15Note: Probabilities are from the Kaplan-Meier analysis and account for censoring in the data for subjects who were unable to demonstrate correct use after the second HCP instruction. For the DISKUS plus HandiHaler and Turbuhaler plus HandiHaler groups, if a subject demonstrated correct use on only 1 of the 2 devices, the time to demonstrate correct use was censored at the total of the reading and/or instruction time for both devices. Subject (Sub-study 2) demonstrated correct use for the ELLIPTA device following the first HCP instruction,however, the time taken to demonstrate correct use was not recorded and therefore this subject was not included for the ELLIPTA arm of Sub-study 2.

44

PPD

CONFIDENTIAL 2017N324761_00206215

43

Figure 7 Kaplan-Meier Plot of Total Time Taken to Demonstrate Correct Inhaler Use: Sub-study 1, ELLIPTA vs. DISKUS plus HandiHaler (ITT Population)

Source: Figure 2.04Note: For the DISKUS plus HandiHaler and Turbuhaler plus HandiHaler groups, if a subject demonstrated correct use on only 1 of the 2 devices, the time to demonstrate correct use was censored at the total of the reading and/or instruction time for both devices.

Figure 8 Kaplan-Meier Plot of Total Time Taken to Demonstrate Correct Inhaler Use: Sub-study 2, ELLIPTA vs. Turbuhaler plus HandiHaler (ITT Population)

Source: Figure 2.04Note: For the DISKUS plus HandiHaler and Turbuhaler plus HandiHaler groups, if a subject demonstrated correct use on only 1 of the 2 devices, the time to demonstrate correct use was censored at the total of the reading and/or instruction time for both devices.

45

CONFIDENTIAL 2017N324761_00206215

44

6.6. Inhaler Preference

The majority of subjects in Sub-study 1 and Sub-study 2 preferred to take their COPD medication using ELLIPTA (81% and 84% in Sub-studies 1 and 2, respectively) compared with DISKUS plus HandiHaler (9%, Sub-study 1; 10% of subjects had no preference) and with Turbuhaler plus HandiHaler (4% Sub-study 2; 13% of subjects had no preference) (Table 23 and Table 24). Preference for ELLIPTA was also observed in both sub-studies based on the number of steps required to take the COPD medication; 89% and 91% of subjects (Sub-studies 1 and 2, respectively) preferred ELLIPTA compared with DISKUS plus HandiHaler (8%, Sub-study 1; 4% of subjects had no preference) and with Turbuhaler plus HandiHaler (5%, Sub-study 2; 4% of subjects had no preference). The differences between ELLIPTA and the comparator inhalercombinations for both preference questions in both sub-studies, were statistically significant (p<0.001, for both sub-studies).

Table 23 Summary of Treatment Preference: Sub-study 1 (ITT Population)

ELLIPTA vs. DISKUS + HandiHaler(N=80)

p-value1

Preference based on number of steps needed to take COPD medication, n (%)

ELLIPTA 71 (89)<0.001DISKUS + HandiHaler 6 (8)

No preference 3 (4)Preference for taking COPD medication, n (%)

ELLIPTA 65 (81)<0.001DISKUS + HandiHaler 7 (9)

No preference 8 (10)Source: Table 2.13 1. Taken from Cochran-Mantel-Haenszel test.

Table 24 Summary of Treatment Preference: Sub-study 2 (ITT Population)

ELLIPTA vs. Turbuhaler + HandiHaler

(N=79)p-value1

Preference based on number of steps needed to take COPD medication, n (%)

ELLIPTA 72 (91)<0.001Turbuhaler + HandiHaler 4 (5)

No preference 3 (4)Preference for taking COPD medication, n (%)

ELLIPTA 66 (84)<0.001Turbuhaler + HandiHaler 3 (4)

No preference 10 (13)Source: Table 2.13 1. Taken from Cochran-Mantel-Haenszel test.

46

CONFIDENTIAL 2017N324761_00206215

45

7. SAFETY RESULTS

7.1. Adverse Events

One on-study treatment-related AE was reported: Subject had an AE of laceration (verbatim term, cuts on both thumbs) in Sub-study 1 (Table 3.02 and Listing 9); the event resolved after 6 days. No AEs were reported in Sub-study 2.

7.2. Serious and Other Significant Adverse Events

No SAEs or deaths were reported during this study (Table 3.04 and Listing 32).

7.3. Pregnancies

No pregnancies were reported during this study (Listing 33).

7.4. Combination Device/Drug Product Near-Incidents, Malfunctions and Remedial Action

No incidents, near incidents or malfunctions were reported with the use of the medical device(s) manufactured or marketed, by GSK or by a third party for GSK, for this study.

47

PPD

CONFIDENTIAL 2017N324761_00206215

46

8. DISCUSSION AND CONCLUSIONS

8.1. Discussion

This study was designed to assess the benefits of delivering triple therapy to patients with COPD using a single ELLIPTA DPI versus delivering triple therapy using 2 different types of inhalers. To achieve this, the number of critical and overall errors made whenusing a single ELLIPTA DPI versus those made when using inhaler combinations of DISKUS plus HandiHaler or Turbuhaler plus HandiHaler were assessed. These errors were analysed after reading the PIL and then, if required, following the first and second instructions from the HCP. This study also assessed training/teaching time (based on the number of instructions required and the time taken to demonstrate correct inhaler use) as well as preference attributes for each inhaler device. Placebo drugs were used to prevent any confounding influence of medication and subjects had to be naïve (no experience in previous 24 months) to the inhaler(s) within the sub-study they were allocated to, in orderto eliminate previous experience influencing error rates. It should be noted that 1 subject was determined as not naïve to ELLIPTA during analysis.

The demographic characteristics of the study population were generally similar across each sub-study and were consistent with a COPD population in clinical practice.

The primary objective of this study was to compare the number of critical errors made bysubjects with COPD, after reading the PIL, for each treatment option tested. Both Sub-study 1 and Sub-study 2 met the primary objective by demonstrating that, after reading the PIL, fewer subjects demonstrated 1 or more critical errors when using ELLIPTA (9% in both sub-studies) compared with DISKUS plus HandiHaler (75%, Sub-study 1) or with Turbuhaler plus HandiHaler (73%, Sub-study 2), and that the differences between ELLIPTA (in Sub-studies 1 and 2) and each of the inhaler combinations used in the respective sub-studies, were statistically significant (p<0.001). In addition, following the first HCP instruction, for those subjects who made an error after reading the PIL, fewer critical errors were made by subjects using ELLIPTA (1%and 3%, Sub-studies 1 and 2, respectively), compared with subjects using DISKUS plus HandiHaler (11%) or Turbuhaler plus HandiHaler (14%); these differences were also statistically significant. Few subjects from either sub-study required a second HCP instruction (required by subjects who made an error following the first HCP instruction). Of these subjects, none made critical errors in either sub-study using ELLIPTA; however, critical errors were made by subjects using DISKUS plus HandiHaler (6%) and using Turbuhaler plus HandiHaler (5%). These differences were not statistically significant; however, it should be noted that the proportion of subjects requiring a second HCP instruction was small, thus limiting the conclusions that could be drawn from analysis of these data. Of note, critical error rates observed for ELLIPTA were similar across both sub-studies and all subjects who had critical errors when using ELLIPTA also had critical errors when using the comparator inhaler combinations.

The observed critical error rates for ELLIPTA (9% in both sub-studies) were comparable to those seen in previous GSK studies [GSK Document Number 2015N230775_00 and GSK Document Number 2015N230821_01].

48

CONFIDENTIAL 2017N324761_00206215

47

On the error checklists, 4 types of critical error were possible with ELLIPTA, 5 with DISKUS, 6 with Turbuhaler and 7 with HandiHaler. Similar to the proportion of subjects making errors, the actual number of critical errors recorded after reading the PIL was lower with ELLIPTA (8 and 13 errors in each sub-study) compared with each of the other inhaler combinations (139 errors with DISKUS plus HandiHaler; 134 errors with Turbuhaler plus HandiHaler).

The results for overall errors after reading the PIL and after the first and second HCP instruction followed the same trend as critical errors. Statistically significant differences were observed between ELLIPTA and both inhaler combinations after reading the PIL(both sub-studies) and after the first HCP instruction (Sub-study 2). Differences were not statistically significant in Sub-study 1 following the first HCP instruction and in either sub-study following the second HCP instruction; of note, the proportions of subjects requiring instructions were small, thus limiting the conclusions that could be drawn from analysis of these data.

After reading the PIL, the majority of subjects made no errors using ELLIPTA, thus did not require instruction from the HCP. The converse was observed for the comparator inhaler combinations, where the majority of subjects made errors and required HCP instruction. Few subjects required more than 1 instruction from the HCP to demonstrate correct inhaler use. The differences in the overall number of HCP instructions required between ELLIPTA and the comparator inhaler combinations used were statistically significant (p<0.001). That said, across all inhaler types, the number of subjects with overall errors declined following each instruction. These findings indicate that training, in conjunction with clear written instructions, is an important element in reducing error rates.

The median total time from when the subject started to read the PIL(s) until correct use was demonstrated was shorter for subjects using ELLIPTA (2.68 and 2.58 minutes in Sub-studies 1 and 2, respectively) compared with DISKUS plus HandiHaler (10.57 minutes, Sub-study 1) and with Turbuhaler plus HandiHaler (11.30 minutes, Sub-study 2). The time from when the HCP started to instruct the subject was recorded as zero for subjects who demonstrated correct use after reading the PIL. For both Sub-study 1 and 2, the median time from when the HCP started to instruct the subjectuntil correct use was demonstrated was also shorter for subjects using ELLIPTA (0 minutes in both sub-studies) compared with DISKUS plus HandiHaler (2.89 minutes, Sub-study 1) and with Turbuhaler plus HandiHaler (3.12 minutes, Sub-study 2). In both Sub-study 1 and 2, 86% of subjects demonstrated correct use of the ELLIPTA device within the first 5 minutes compared with 10% of subjects using DISKUS plus HandiHaler (Sub-study 1) and 9% of subjects using Turbuhaler plus HandiHaler (Sub-study 2). In Sub-study 1, all subjects had demonstrated correct use of the ELLIPTA device within 15 minutes compared with 67% of subjects demonstrating the correct use of the DISKUS plus HandiHaler combination. In Sub-study 2, 97% of subjects had demonstrated correct use of the ELLIPTA device within 20 minutes compared with 82% of subjects demonstrating the correct use of the Turbuhaler plus HandiHaler combination. Overall, subjects took less time to demonstrate correct inhaler use with ELLIPTA following reading of the PIL and receiving HCP instruction compared with the comparator inhaler combinations.

49

CONFIDENTIAL 2017N324761_00206215

48

Patient preference for a particular inhalation device depends on multiple factors. These factors can include, for example, ease of use, size, shape, ability to track doses remaining, comfort and past inhaler experience. Thus, it is important to assess specific attributes that are related to preference, not only overall preference for a particular device. This can be helpful in patient education and in prescriber-patient dialogue about the choice of device, with the goal of engaging patients in their treatment selection which could potentially influence their medication taking behaviour. The majority of subjects in this study preferred to take their COPD medication using ELLIPTA and preferred ELLIPTA to the other inhaler combinations based on the number of steps required to take their COPD (p<0.001).

The low error rates observed for subjects using ELLIPTA compared with the other inhaler device combinations could be due to the fact that fewer steps are required when using a single inhaler compared with using multiple inhalers, and possibly a consequence of the more intuitive design of ELLIPTA resulting in fewer errors. There is also a possibility that, when using 2 inhalers, a set of instructions for each inhaler can cause confusion and lead to errors in use. This study also highlights the importance of thorough instruction and training, provided by the HCP, to ensure subjects fully understand how to take their medication. In addition, this study illustrates that the allotted time given to teach/train and correct errors should be considered for inhalers, in particular when looking at the average length of an appointment scheduled per patient in general practice.

A single on-study treatment-related AE of laceration was observed in this study; no SAEs were reported.

8.2. Conclusions

In both sub-studies, after reading the PIL, a statistically significantly lower percentage of subjects with COPD made critical and overall errors with ELLIPTA compared with DISKUS plus HandiHaler or Turbuhaler plus HandiHaler. Overall, subjects using ELLIPTA required fewer HCP instructions, made fewer errors following each HCP instruction and took less time to demonstrate the correct use of the inhaler device, compared with either DISKUS plus HandiHaler or Turbuhaler plus HandiHaler. In addition, a statistically significantly higher percentage of subjects preferred the ELLIPTA device compared with the other inhaler combinations based on both preference questions; for taking their COPD medication, and based on the number of steps required to take their COPD medication.

50

CONFIDENTIAL 2017N324761_00206215

49

9. REFERENCES

Bonini M, Usmani O. The importance of inhaler devices in the treatment of COPD. COPD Res Pract. 2015;1:9.

Celli BR, MacNee W. Standards of the diagnosis and treatment of patients with COPD: a summary of the ATS/ERS position paper. Eur Respir J. 2004;23:932-46.

Cochrane M, Bala M, Downs K, Mauskopf J, Ben-Joseph R. Inhaled corticosteroids for asthma therapy:patient compliance, devices, and inhalation technique. Chest. 2000;117:542-50.

Global Initiative for Chronic Obstructive Lung Disease (GOLD 2017). Global strategy for the diagnosis, management, and prevention of COPD (Updated 2017). Available from http://www.goldcopd.org.

GSK Document Number 2015N230775_00. An open-label study of inhaler device attributes investigating critical and overall errors, ease of use, and preference between a number of inhaler devices (ELLIPTA, TURBUHALER, HANDIHALER, BREEZHALER, MDI and DISKUS/ACCUHALER) in adult subjects with chronic obstructive pulmonary disease (COPD) (GSK Study 200301) (December 2015).

GSK Document Number 2015N230821_01. An open-label study of inhaler device attributes investigating critical and overall errors, and ease of use and preference between a number of inhaler devices (ELLIPTA, DISKUS/ACCUHALER, TURBUHALER, and MDI) in adult subjects with asthma (GSK Study 200330) (January 2016).

Komase Y, Asako A, Kobyyashi A, Sharma R. Ease-of-use preference for the ELLIPTAdry powder inhaler over a commonly used single-dose capsule dry powder inhaler by inhalation device-naïve Japanese volunteers aged 40 years or older. Int J Chron Obstruct Pulmon Dis. 2014:9;1365-75.

Melani A, Bonavia M, Cilenti V, Cinti C, Lodi M, Martucci P, et al. Inhaler mishandling remains common in real life and is associated with reduced disease control. Resp Med. 2011;105:930-8.

Van der Palen J, Klein J, Van Herwaarden, Zielhuis, Seydel E. Multiple inhalers confuse asthma patients. Eur Respir J. 1999;14:1034-7.

Van Der Palen J, Thomas M, Chrystyn H, Sharma R, Imber V, Zhu C-Q, et al. Inhaler Errors After Reading The Patient Information Leaflet In Patients With COPD: A Comparison Of ElliptaWith Five Inhaler Devices. Am J Respir Crit Care Med.2016;193:A6811 (a).

Van Der Palen J, Thomas M, Chrystyn H, Sharma R, Svedsater H, Zhu C-Q, et al. Inhaler Preference In Patients With COPD: A Comparison Of Ellipta With Five Inhaler Devices. Am J Respir Crit Care Med. 2016:193;A6813 (b).

51

CONFIDENTIAL 2017N324761_00206215

50

Van Der Palen J, Thomas M, Chrystyn H, Sharma R, Imber V, Zhu C-Q, et al. Trainingand Time To Achieve Correct Inhaler Use: A Comparison Between Inhalers In Patients With COPD. Am J Respir Crit Care Med. 2016:193;A6811 (c).

52

Protocol: 206215 Page 1 of 1Population: All Subjects Enrolled Table 1.01 Summary of Subject Population

Sub Sub Population Study 1 Study 2 Total ----------------------------------------------------------------------- All Subjects Enrolled (ASE) 160 Screen Failures [1] 0 Randomised 80 80 160 Intent-to-Treat (ITT) [2] 80 (100%) 79 (99%) 159 (>99%) Safety [2] 80 (100%) 79 (99%) 159 (>99%)

[1] Percentages are of the ASE population.[2] Percentages are of the randomised population.ASE: All subjects who signed the informed consent form and for whom a record exists on the study database.Randomised: All subjects who were randomised.ITT: All randomised subjects who made at least one critical error assessment from one treatmentoption device.Safety: This population is the same as the Intent-to-Treat population.

53

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 1Population: All Subjects Enrolled Table 1.02 Summary of Screening Status and Reasons for Screen Failures

Total (N=160) ------------------------------------------------------------------ Screening Status Enrolled 160 (100%) Failed 0

Reason for Screen Failure [1] Study closed/terminated (unspecified) 0 Did not meet inclusion/exclusion criteria 0 Adverse event (unspecified) 0 Investigator discretion 0 Withdrew consent 0

Note: Subjects may have more than one reason for failure, so percentages may sum to more than 100%.[1] Percentage is calculated based on screening failures.

54

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 1Population: All Subjects Enrolled Table 1.03 Summary of Age Ranges

Sub Sub Study 1 Study 2 Total Age Range (N=80) (N=80) (N=160) -------------------------------------------------------------------------------------------- 18 - 64 Years 34 (43%) 36 (45%) 70 (44%) 65 - 84 Years 46 (58%) 43 (54%) 89 (56%) >=85 Years 0 1 (1%) 1 (<1%)

55

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 1Population: Intent-to-Treat Table 1.04 Summary of Attendance at Each Clinic Visit

Sub Sub Study 1 Study 2 Total Visit (N=80) (N=79) (N=159) ------------------------------------------------------------------------ Screening 80 (100%) 79 (100%) 159 (100%) Randomisation/End of study 80 (100%) 79 (100%) 159 (100%)

Note: Subject performed the ICF procedure and was randomised, however withdrew consentprior to any inhaler assessments.

56

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

Protocol: 206215 Page 1 of 1Population: All Subjects Enrolled Table 1.05 Summary of End of Study Record

Sub Sub Study 1 Study 2 Total (N=80) (N=80) (N=160) ----------------------------------------------------------------------- Completion status Completed 80 (100%) 79 (99%) 159 (>99%) Withdrawn 0 1 (1%) 1 (<1%)

Primary reason [1]/sub-reason [2] for withdrawal Adverse event 0 0 0 Protocol deviation 0 0 0 Lack of adherence 0 0 0 Pregnancy 0 0 0 Prohibited medication use 0 0 0 Study closed/terminated 0 0 0 Investigator site closed 0 0 0 Investigator discretion 0 0 0 Withdrew consent 0 1 (1%) 1 (<1%)

[1] Subjects only record one primary reason for withdrawal.[2] Subjects are not required to indicate sub-reasons.

57

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 2Population: Intent-to-Treat Table 1.06 Summary of Demographic Characteristics

Sub Sub Study 1 Study 2 Total (N=80) (N=79) (N=159) ------------------------------------------------------------------------------

Age (yrs) n 80 79 159 Mean 64.3 65.7 65.0 SD 8.69 8.49 8.60 Median 65.5 66.0 66.0 Min. 48 48 48 Max. 82 92 92

Sex n 80 79 159 Female 39 (49%) 37 (47%) 76 (48%) Male 41 (51%) 42 (53%) 83 (52%)

Ethnicity n 80 79 159 Hispanic or Latino 0 0 0 Not Hispanic or Latino 80 (100%) 79 (100%) 159 (100%)

Height (cm) n 80 78 158 Mean 168.4 169.6 169.0 SD 9.86 9.64 9.74 Median 169.5 170.0 170.0 Min. 147 150 147 Max. 191 199 199

Note: Subject performed the ICF procedure and was randomised, however the subject's vital signswere not collected at visit 1.

58

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

Protocol: 206215 Page 2 of 2Population: Intent-to-Treat Table 1.06 Summary of Demographic Characteristics

Sub Sub Study 1 Study 2 Total (N=80) (N=79) (N=159) ------------------------------------------------------------------------------

Weight (kg) n 80 78 158 Mean 75.95 78.10 77.01 SD 18.098 14.883 16.571 Median 73.90 75.85 74.35 Min. 43.6 50.6 43.6 Max. 122.4 125.4 125.4

BMI (kg/m^2) n 80 78 158 Mean 26.69 27.12 26.90 SD 5.607 4.568 5.109 Median 26.08 26.21 26.21 Min. 16.5 19.1 16.5 Max. 43.4 40.5 43.4


59

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

Protocol: 206215 Page 1 of 4Population: Intent-to-Treat Table 1.07 Summary of Demographic Characteristics by Country

Country: Netherlands Sub Sub Study 1 Study 2 Total (N=62) (N=60) (N=122) ------------------------------------------------------------------------ Age (yrs) n 62 60 122 Mean 64.9 65.0 65.0 SD 8.65 8.30 8.44 Median 66.0 65.0 66.0 Min. 48 48 48 Max. 82 81 82

Sex n 62 60 122 Female 28 (45%) 27 (45%) 55 (45%) Male 34 (55%) 33 (55%) 67 (55%)




60

CONFIDENTIAL 2017N324761_00206215

PPD

PPD


Country: Netherlands Sub Sub Study 1 Study 2 Total (N=62) (N=60) (N=122) ------------------------------------------------------------------------ Weight (kg) n 62 60 122 Mean 75.30 77.62 76.44 SD 17.749 15.172 16.504 Median 72.90 74.25 73.75 Min. 43.6 50.7 43.6 Max. 121.2 125.4 125.4



61

CONFIDENTIAL 2017N324761_00206215

PPD

PPD


Country: United Kingdom Sub Sub Study 1 Study 2 Total (N=18) (N=19) (N=37) ------------------------------------------------------------------------ Age (yrs) n 18 19 37 Mean 62.2 68.1 65.2 SD 8.75 8.88 9.20 Median 63.5 66.0 65.0 Min. 48 54 48 Max. 80 92 92

Sex n 18 19 37 Female 11 (61%) 10 (53%) 21 (57%) Male 7 (39%) 9 (47%) 16 (43%)




62

CONFIDENTIAL 2017N324761_00206215

PPD

PPD


Country: United Kingdom Sub Sub Study 1 Study 2 Total (N=18) (N=19) (N=37) ------------------------------------------------------------------------ Weight (kg) n 18 18 36 Mean 78.21 79.71 78.96 SD 19.616 14.172 16.882 Median 75.85 84.10 79.10 Min. 47.2 50.6 47.2 Max. 122.4 96.3 122.4



63

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

Protocol: 206215 Page 1 of 1Population: Intent-to-Treat Table 1.08 Summary of Number of Subjects by Country and Centre

Sub Sub Investigator at Investigator Study 1 Study 2 Total Country Centre Name (N=80) (N=79) (N=159) ----------------- ------------------------------------------------------------------- Netherlands Soest 0 14 (18%) 14 (9%) Hoek 46 (58%) 30 (38%) 76 (48%) Moeskops-van 16 (20%) 16 (20%) 32 (20%) Beurden

United Kingdom Collier 14 (18%) 10 (13%) 24 (15%) Van Den Berg 4 (5%) 9 (11%) 13 (8%)

64

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

Protocol: 206215 Page 1 of 1Population: Intent-to-Treat Table 1.09 Summary of Race and Racial Combinations

Sub Sub Study 1 Study 2 Total (N=80) (N=79) (N=159) ------------------------------------------------------------------------------------ n 80 79 159 African American/African Heritage 1 (1%) 0 1 (<1%) American Indian or Alaskan Native 0 0 0 Asian 21 (26%) 20 (25%) 41 (26%) Central/South Asian Heritage 20 (25%) 20 (25%) 40 (25%) Japanese/East Asian Heritage/ 1 (1%) 0 1 (<1%) South East Asian Heritage Native Hawaiian or Other Pacific Islander 0 0 0 White 58 (73%) 59 (75%) 117 (74%)

65

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 1Population: Intent-to-Treat Table 1.10 Summary of Race and Racial Combination Details

Sub Sub Study 1 Study 2 Total (N=80) (N=79) (N=159) ----------------------------------------------------------------------- n 80 79 159 African American/African Heritage 1 (1%) 0 1 (<1%) American Indian or Alaskan Native 0 0 0 Asian - Central/South Asian 20 (25%) 20 (25%) 40 (25%) Heritage Asian - East Asian Heritage 0 0 0 Asian - Japanese Heritage 0 0 0 Asian - South East Asian Heritage 1 (1%) 0 1 (<1%) Native Hawaiian or Other Pacific 0 0 0 Islander White - Arabic/North African 5 (6%) 4 (5%) 9 (6%) Heritage White - White/Caucasian/European 53 (66%) 55 (70%) 108 (68%) Heritage

66

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 1Population: Intent-to-Treat Table 1.11 Summary of Important Protocol Deviations

Sub Sub Study 1 Study 2 Total Protocol Deviations (N=80) (N=79) (N=159) -------------------------------------------------------------------------------------------------------- Any Protocol Deviations 4 (5%) 2 (3%) 6 (4%)

Informed consent 2 (3%) 0 2 (1%) Informed consent/assent not signed and/or dated by appropriate 2 (3%) 0 2 (1%) site staff

Eligibility criteria not met 1 (1%) 1 (1%) 2 (1%)

Not withdrawn after developing withdrawal criteria 0 0 0

Visit completion 0 0 0

Assessment or time point completion 0 1 (1%) 1 (<1%) Missed assessment 0 1 (1%) 1 (<1%)

Wrong study treatment/administration/dose 1 (1%) 0 1 (<1%) Wrong study treatment or assignment administered 1 (1%) 0 1 (<1%)

Study Procedures 0 0 0

Failure to report safety events per protocol 0 0 0

Note: A subject may have more than one protocol deviation.

67

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 1Population: Intent-to-Treat Table 1.12 Summary of Inclusion/Exclusion Criteria Deviations

Sub Sub Study 1 Study 2 Total Criterion (N=80) (N=79) (N=159) ----------------------------------------------------------------------------------------- Any Criteria Deviations 1 (1%) 1 (1%) 2 (1%)

Exclusion criteria Asthma 0 1 (1%) 1 (<1%) Recent experience with the ELLIPTA inhaler 1 (1%) 0 1 (<1%) Recent experience with any capsule inhaler 0 0 0 No recent experience with the DISKUS inhaler or 0 0 0 Turbuhaler within 24 months of screening for the sub study a subject is included Drug/alcohol abuse 0 0 0 Drug/Food Allergy 0 0 0 Investigational Product 0 0 0 Inability to Read 0 0 0

Inclusion criteria Age 0 0 0 Diagnosis of COPD with a documented history of 0 0 0 COPD Current COPD Therapy 0 0 0 Current maintenance ICS/LABA COPD treatment for 0 0 0 at least 4 weeks prior to screening Smoking History 0 0 0 Gender 0 0 0 Informed consent 0 0 0

Note: A subject may have more than one protocol deviation.

68

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 2Population: Intent-to-Treat Table 1.13 Summary of Current Medical Conditions

Sub Sub Study 1 Study 2 Total (N=80) (N=79) (N=159) ------------------------------------------------------------------------------- Any condition 41 (51%) 58 (73%) 99 (62%)

Cardiac disorders Any condition 9 (11%) 14 (18%) 23 (14%) Arrhythmia 3 (4%) 7 (9%) 10 (6%) Congestive heart failure 0 2 (3%) 2 (1%) Coronary artery disease 8 (10%) 8 (10%) 16 (10%) Myocardial infarction 1 (1%) 0 1 (<1%)

Endocrine disorders Any condition 0 0 0 Adrenal suppression 0 0 0 Cushing's syndrome 0 0 0

Eye disorders Any condition 3 (4%) 7 (9%) 10 (6%) Cataract 3 (4%) 5 (6%) 8 (5%) Glaucoma 1 (1%) 2 (3%) 3 (2%)

Vascular disorders Any condition 18 (23%) 36 (46%) 54 (34%) Cerebrovascular accident 0 0 0 Hypertension 18 (23%) 36 (46%) 54 (34%)

Metabolism and nutrition disorders Any condition 33 (41%) 46 (58%) 79 (50%) Diabetes mellitus 19 (24%) 20 (25%) 39 (25%) Hypercholesterolemia 20 (25%) 34 (43%) 54 (34%) Osteoporosis 6 (8%) 5 (6%) 11 (7%)

69

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 2 of 2Population: Intent-to-Treat Table 1.13 Summary of Current Medical Conditions

Sub Sub Study 1 Study 2 Total (N=80) (N=79) (N=159) ------------------------------------------------------------------------------- Infections and infestations Any condition 0 1 (1%) 1 (<1%) Pneumonia 0 1 (1%) 1 (<1%)

70

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 2Population: Intent-to-Treat Table 1.14 Summary of Past Medical Conditions

Sub Sub Study 1 Study 2 Total (N=80) (N=79) (N=159) ------------------------------------------------------------------------------- Any condition 29 (36%) 31 (39%) 60 (38%)

Cardiac disorders Any condition 15 (19%) 9 (11%) 24 (15%) Arrhythmia 2 (3%) 2 (3%) 4 (3%) Congestive heart failure 0 0 0 Coronary artery disease 8 (10%) 3 (4%) 11 (7%) Myocardial infarction 9 (11%) 6 (8%) 15 (9%)

Endocrine disorders Any condition 0 0 0 Adrenal suppression 0 0 0 Cushing's syndrome 0 0 0

Eye disorders Any condition 1 (1%) 10 (13%) 11 (7%) Cataract 1 (1%) 10 (13%) 11 (7%) Glaucoma 0 0 0

Vascular disorders Any condition 1 (1%) 11 (14%) 12 (8%) Cerebrovascular accident 1 (1%) 10 (13%) 11 (7%) Hypertension 0 2 (3%) 2 (1%)

Metabolism and nutrition disorders Any condition 3 (4%) 0 3 (2%) Diabetes mellitus 1 (1%) 0 1 (<1%) Hypercholesterolemia 2 (3%) 0 2 (1%) Osteoporosis 0 0 0

71

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 2 of 2Population: Intent-to-Treat Table 1.14 Summary of Past Medical Conditions

Sub Sub Study 1 Study 2 Total (N=80) (N=79) (N=159) ------------------------------------------------------------------------------- Infections and infestations Any condition 17 (21%) 18 (23%) 35 (22%) Pneumonia 17 (21%) 18 (23%) 35 (22%)

72

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 1Population: Intent-to-Treat Table 1.15 Summary of Family History of Cardiovascular Risk Factors

Sub Sub Study 1 Study 2 Total (N=80) (N=79) (N=159) ---------------------------------------------------------------- Family History [1] Yes 13 (16%) 26 (33%) 39 (25%) No 44 (55%) 43 (54%) 87 (55%) Unknown 23 (29%) 10 (13%) 33 (21%)

[1] Family history of premature coronary artery disease in women < 65 years or men < 55 years in firstdegree relatives only (e.g., biological mother or father, biological brother or sister, biological son ordaughter).

73

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 1Population: Intent-to-Treat Table 1.16 Summary of COPD History

Sub Sub Study 1 Study 2 Total (N=80) (N=79) (N=159) ----------------------------------------------------------------------------- Duration of COPD n 80 79 159 >=6 months to <1 year 2 (3%) 0 2 (1%) >=1 year to <5 years 30 (38%) 21 (27%) 51 (32%) >=5 years to <10 years 26 (33%) 27 (34%) 53 (33%) >=10 years to <15 years 12 (15%) 21 (27%) 33 (21%) >=15 years to <20 years 6 (8%) 6 (8%) 12 (8%) >=20 years to <25 years 1 (1%) 2 (3%) 3 (2%) >=25 years 3 (4%) 2 (3%) 5 (3%)

74

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 1Population: Intent-to-Treat Table 1.17 Summary of Smoking History and Smoking Status at Screening

Sub Sub Study 1 Study 2 Total (N=80) (N=79) (N=159) -------------------------------------------------------------------------------------------- Years Smoked n 80 79 159 Mean 41.3 41.0 41.1 SD 11.50 13.11 12.29 Median 40.5 40.0 40.0 Min. 13 15 13 Max. 63 67 67

Cigarettes/Day n 80 79 159 Mean 18.2 17.7 17.9 SD 7.79 6.53 7.17 Median 20.0 20.0 20.0 Min. 7 6 6 Max. 50 40 50

Smoking Pack Years[1] n 80 79 159 Mean 36.4 36.4 36.4 SD 16.76 18.72 17.70 Median 34.0 34.0 34.0 Min. 12 10 10 Max. 112 100 112

Smoking Status at Screening n 80 79 159 Current smoker 45 (56%) 40 (51%) 85 (53%) Former smoker 35 (44%) 39 (49%) 74 (47%)

[1] Smoking Pack Years = (Number of cigarettes smoked per day/20) x number of years smoked.

75

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 4Population: Intent-to-Treat Table 1.18 Summary of COPD Medications

Sub Sub Respiratory Class Study 1 Study 2 Total Ingredient (N=80) (N=79) (N=159) ---------------------------------------------------------------------------------------------- Any medication 80 (100%) 79 (100%) 159 (100%)

Long-acting beta-2 agonist Any medication 80 (100%) 77 (97%) 157 (99%) SALMETEROL XINAFOATE + FLUTICASONE PROPIONATE 22 (28%) 40 (51%) 62 (39%) BECLOMETASONE DIPROPIONATE + FORMOTEROL FUMARATE 29 (36%) 26 (33%) 55 (35%) BUDESONIDE + FORMOTEROL FUMARATE 17 (21%) 2 (3%) 19 (12%) SALMETEROL + FLUTICASONE 5 (6%) 8 (10%) 13 (8%) TIOTROPIUM BROMIDE + OLODATEROL HYDROCHLORIDE 6 (8%) 2 (3%) 8 (5%) FORMOTEROL FUMARATE 5 (6%) 1 (1%) 6 (4%) FORMOTEROL FUMARATE + FLUTICASONE PROPIONATE 2 (3%) 3 (4%) 5 (3%) OLODATEROL HYDROCHLORIDE 4 (5%) 1 (1%) 5 (3%) SALMETEROL + FLUTICASONE PROPIONATE 1 (1%) 2 (3%) 3 (2%) SALMETEROL XINAFOATE 0 3 (4%) 3 (2%) BECLOMETASONE + FORMOTEROL 1 (1%) 0 1 (<1%) GLYCOPYRRONIUM BROMIDE + INDACATEROL MALEATE 1 (1%) 0 1 (<1%)

Corticosteroid - Inhaled Any medication 76 (95%) 78 (99%) 154 (97%) SALMETEROL XINAFOATE + FLUTICASONE PROPIONATE 22 (28%) 40 (51%) 62 (39%) BECLOMETASONE DIPROPIONATE + FORMOTEROL FUMARATE 28 (35%) 26 (33%) 54 (34%) BUDESONIDE + FORMOTEROL FUMARATE 16 (20%) 2 (3%) 18 (11%) SALMETEROL + FLUTICASONE 5 (6%) 8 (10%) 13 (8%) FORMOTEROL FUMARATE + FLUTICASONE PROPIONATE 2 (3%) 3 (4%) 5 (3%) CICLESONIDE 3 (4%) 1 (1%) 4 (3%) FLUTICASONE PROPIONATE 0 4 (5%) 4 (3%) SALMETEROL + FLUTICASONE PROPIONATE 1 (1%) 2 (3%) 3 (2%) BECLOMETASONE + FORMOTEROL 1 (1%) 0 1 (<1%) SALBUTAMOL + FLUTICASONE 0 1 (1%) 1 (<1%)

Note: Combination products are listed as combinations and included in all applicable respiratory classes.

76

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Sub Respiratory Class Study 1 Study 2 Total Ingredient (N=80) (N=79) (N=159) ---------------------------------------------------------------------------------------------- Short-acting beta-2 agonist Any medication 51 (64%) 62 (78%) 113 (71%) SALBUTAMOL 44 (55%) 60 (76%) 104 (65%) FENOTEROL HYDROBROMIDE + IPRATROPIUM BROMIDE 7 (9%) 4 (5%) 11 (7%) SALBUTAMOL + IPRATROPIUM 2 (3%) 0 2 (1%) SALBUTAMOL + FLUTICASONE 0 1 (1%) 1 (<1%) SALBUTAMOL SULFATE + IPRATROPIUM BROMIDE 0 1 (1%) 1 (<1%) TERBUTALINE 1 (1%) 0 1 (<1%)

Long-acting anticholinergic Any medication 34 (43%) 29 (37%) 63 (40%) TIOTROPIUM BROMIDE 24 (30%) 26 (33%) 50 (31%) TIOTROPIUM BROMIDE + OLODATEROL HYDROCHLORIDE 6 (8%) 2 (3%) 8 (5%) ACLIDINIUM BROMIDE 3 (4%) 0 3 (2%) TIOTROPIUM 0 2 (3%) 2 (1%) GLYCOPYRRONIUM BROMIDE + INDACATEROL MALEATE 1 (1%) 0 1 (<1%)

Short-acting anticholinergic Any medication 21 (26%) 14 (18%) 35 (22%) IPRATROPIUM BROMIDE 8 (10%) 8 (10%) 16 (10%) FENOTEROL HYDROBROMIDE + IPRATROPIUM BROMIDE 7 (9%) 4 (5%) 11 (7%) IPRATROPIUM 5 (6%) 1 (1%) 6 (4%) SALBUTAMOL + IPRATROPIUM 2 (3%) 0 2 (1%) SALBUTAMOL SULFATE + IPRATROPIUM BROMIDE 0 1 (1%) 1 (<1%)

Other COPD medication Any medication 5 (6%) 4 (5%) 9 (6%) AERIUS (NOS) 1 (1%) 0 1 (<1%) ALENDRONATE SODIUM 1 (1%) 0 1 (<1%) ATORVASTATIN 0 1 (1%) 1 (<1%) AZELASTINE HYDROCHLORIDE + FLUTICASONE PROPIONATE 1 (1%) 0 1 (<1%)


77

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Sub Respiratory Class Study 1 Study 2 Total Ingredient (N=80) (N=79) (N=159) ---------------------------------------------------------------------------------------------- CALCIUM CARBONATE + COLECALCIFEROL 0 1 (1%) 1 (<1%) CHONDROITIN SULFATE + GLUCOSAMINE SULFATE 1 (1%) 0 1 (<1%) CODEINE PHOSPHATE 0 1 (1%) 1 (<1%) LEVOCETIRIZINE 0 1 (1%) 1 (<1%) MORPHINE SULFATE 1 (1%) 0 1 (<1%)

Corticosteroid - Systemic, oral, parenteral and intra-articular Any medication 7 (9%) 0 7 (4%) PREDNISOLONE 4 (5%) 0 4 (3%) BECLOMETASONE DIPROPIONATE + FORMOTEROL FUMARATE 1 (1%) 0 1 (<1%) BUDESONIDE + FORMOTEROL FUMARATE 1 (1%) 0 1 (<1%) PREDNISONE 1 (1%) 0 1 (<1%)

Antiinfectives (antibiotics, antifungals, antivirals, antiseptics) Any medication 2 (3%) 2 (3%) 4 (3%) AZITHROMYCIN 2 (3%) 0 2 (1%) COLISTIN MESILATE SODIUM 0 1 (1%) 1 (<1%) SULFAMETHOXAZOLE + TRIMETHOPRIM 0 1 (1%) 1 (<1%)

Corticosteroid - Other Any medication 2 (3%) 1 (1%) 3 (2%) AZELASTINE HYDROCHLORIDE + FLUTICASONE PROPIONATE 1 (1%) 0 1 (<1%) FLUTICASONE 0 1 (1%) 1 (<1%) FLUTICASONE FUROATE 1 (1%) 0 1 (<1%)

Leukotriene Receptor Antagonist Any medication 2 (3%) 0 2 (1%) MONTELUKAST 2 (3%) 0 2 (1%)


78

CONFIDENTIAL 2017N324761_00206215

PPDPPD


Sub Sub Respiratory Class Study 1 Study 2 Total Ingredient (N=80) (N=79) (N=159) ---------------------------------------------------------------------------------------------- Mucolytics Any medication 0 1 (1%) 1 (<1%) CARBOCISTEINE 0 1 (1%) 1 (<1%)


79

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 10Population: Intent-to-Treat Table 1.19 Summary of Non-COPD Concomitant Medications

Sub Sub ATC Level 1 Study 1 Study 2 Total Ingredient (N=80) (N=79) (N=159) ------------------------------------------------------------------------------------------------- Any medication 66 (83%) 69 (87%) 135 (85%)

ALIMENTARY TRACT AND METABOLISM Any medication 50 (63%) 48 (61%) 98 (62%) OMEPRAZOLE 16 (20%) 13 (16%) 29 (18%) METFORMIN 10 (13%) 12 (15%) 22 (14%) PANTOPRAZOLE 10 (13%) 12 (15%) 22 (14%) ACETYLSALICYLIC ACID 6 (8%) 10 (13%) 16 (10%) ESOMEPRAZOLE 7 (9%) 9 (11%) 16 (10%) GLICLAZIDE 4 (5%) 6 (8%) 10 (6%) COLECALCIFEROL 6 (8%) 3 (4%) 9 (6%) CALCIUM CARBONATE + COLECALCIFEROL 5 (6%) 2 (3%) 7 (4%) MACROGOL + ELECTROLYTES NOS 4 (5%) 2 (3%) 6 (4%) PREDNISONE 3 (4%) 1 (1%) 4 (3%) ESOMEPRAZOLE MAGNESIUM 2 (3%) 1 (1%) 3 (2%) LANSOPRAZOLE 2 (3%) 1 (1%) 3 (2%) CALCIUM 1 (1%) 1 (1%) 2 (1%) GLIMEPIRIDE 1 (1%) 1 (1%) 2 (1%) MACROGOL 0 2 (3%) 2 (1%) METFORMIN HYDROCHLORIDE + SITAGLIPTIN 1 (1%) 1 (1%) 2 (1%) RANITIDINE 0 2 (3%) 2 (1%) TOLBUTAMIDE 1 (1%) 1 (1%) 2 (1%) ASCORBIC ACID + PYRIDOXINE HYDROCHLORIDE + CALCIUM 1 (1%) 0 1 (<1%) CARBONATE + COLECALCIFEROL CALCIUM CARBONATE 1 (1%) 0 1 (<1%) COLECALCIFEROL + CALCIUM 0 1 (1%) 1 (<1%) COLECALCIFEROL + CALCIUM CITRATE 1 (1%) 0 1 (<1%) DAPAGLIFLOZIN PROPANEDIOL 1 (1%) 0 1 (<1%) DEXAMETHASONE 0 1 (1%) 1 (<1%) EMPAGLIFLOZIN 0 1 (1%) 1 (<1%) ENZYME, NOS 1 (1%) 0 1 (<1%)

Note: Combination products are listed as combinations and included in all applicable ATC level 1 categories.

80

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Sub ATC Level 1 Study 1 Study 2 Total Ingredient (N=80) (N=79) (N=159) ------------------------------------------------------------------------------------------------- GALVUS (NOS) 0 1 (1%) 1 (<1%) GAVISCON NOS 1 (1%) 0 1 (<1%) HYDROCORTISONE BUTYRATE 0 1 (1%) 1 (<1%) INSULIN ASPART 0 1 (1%) 1 (<1%) INSULIN DETEMIR 0 1 (1%) 1 (<1%) INSULIN LISPRO 0 1 (1%) 1 (<1%) LINAGLIPTIN 0 1 (1%) 1 (<1%) METFORMIN HYDROCHLORIDE + DAPAGLIFLOZIN PROPANEDIOL 0 1 (1%) 1 (<1%) METFORMIN HYDROCHLORIDE + VILDAGLIPTIN 1 (1%) 0 1 (<1%) PREDNISOLONE 0 1 (1%) 1 (<1%) PYRIDOXINE 1 (1%) 0 1 (<1%) RABEPRAZOLE SODIUM 1 (1%) 0 1 (<1%) SITAGLIPTIN 0 1 (1%) 1 (<1%) THIAMINE 1 (1%) 0 1 (<1%) VITAMIN D NOS 1 (1%) 0 1 (<1%)

CARDIOVASCULAR SYSTEM Any medication 40 (50%) 56 (71%) 96 (60%) SIMVASTATIN 13 (16%) 17 (22%) 30 (19%) ATORVASTATIN 9 (11%) 18 (23%) 27 (17%) GLYCERYL TRINITRATE 5 (6%) 6 (8%) 11 (7%) HYDROCHLOROTHIAZIDE 3 (4%) 8 (10%) 11 (7%) METOPROLOL SUCCINATE 4 (5%) 7 (9%) 11 (7%) AMLODIPINE 5 (6%) 5 (6%) 10 (6%) PERINDOPRIL 6 (8%) 4 (5%) 10 (6%) FUROSEMIDE 3 (4%) 5 (6%) 8 (5%) METOPROLOL TARTRATE 1 (1%) 7 (9%) 8 (5%) LOSARTAN POTASSIUM 2 (3%) 4 (5%) 6 (4%) NIFEDIPINE 0 6 (8%) 6 (4%) BUMETANIDE 2 (3%) 3 (4%) 5 (3%) FELODIPINE 2 (3%) 3 (4%) 5 (3%)


81

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Sub ATC Level 1 Study 1 Study 2 Total Ingredient (N=80) (N=79) (N=159) ------------------------------------------------------------------------------------------------- ISOSORBIDE MONONITRATE 1 (1%) 4 (5%) 5 (3%) METOPROLOL 2 (3%) 3 (4%) 5 (3%) ROSUVASTATIN CALCIUM 1 (1%) 4 (5%) 5 (3%) SIMVASTATIN + EZETIMIBE 2 (3%) 3 (4%) 5 (3%) VALSARTAN 1 (1%) 4 (5%) 5 (3%) BISOPROLOL 1 (1%) 3 (4%) 4 (3%) SPIRONOLACTONE 2 (3%) 2 (3%) 4 (3%) VERAPAMIL 2 (3%) 2 (3%) 4 (3%) BENDROFLUMETHIAZIDE 0 3 (4%) 3 (2%) CARVEDILOL 2 (3%) 1 (1%) 3 (2%) ENALAPRIL MALEATE 2 (3%) 1 (1%) 3 (2%) IRBESARTAN 2 (3%) 1 (1%) 3 (2%) LOSARTAN 0 3 (4%) 3 (2%) TADALAFIL 1 (1%) 2 (3%) 3 (2%) AMLODIPINE BESILATE + VALSARTAN 0 2 (3%) 2 (1%) BARNIDIPINE 0 2 (3%) 2 (1%) CANDESARTAN 1 (1%) 1 (1%) 2 (1%) DILTIAZEM 1 (1%) 1 (1%) 2 (1%) DOXAZOSIN 1 (1%) 1 (1%) 2 (1%) ENALAPRIL 0 2 (3%) 2 (1%) LISINOPRIL 1 (1%) 1 (1%) 2 (1%) PERINDOPRIL ARGININE + AMLODIPINE BESILATE 2 (3%) 0 2 (1%) PERINDOPRIL ERBUMINE 1 (1%) 1 (1%) 2 (1%) RAMIPRIL 0 2 (3%) 2 (1%) AMIODARONE 0 1 (1%) 1 (<1%) BARNIDIPINE HYDROCHLORIDE 1 (1%) 0 1 (<1%) BETAHISTINE 0 1 (1%) 1 (<1%) BISOPROLOL FUMARATE 0 1 (1%) 1 (<1%) DEXAMETHASONE 0 1 (1%) 1 (<1%) DIGOXIN 0 1 (1%) 1 (<1%) DILTIAZEM HYDROCHLORIDE 0 1 (1%) 1 (<1%)


82

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Sub ATC Level 1 Study 1 Study 2 Total Ingredient (N=80) (N=79) (N=159) ------------------------------------------------------------------------------------------------- ENZYME, NOS 1 (1%) 0 1 (<1%) EPROSARTAN MESILATE 0 1 (1%) 1 (<1%) FLECAINIDE 0 1 (1%) 1 (<1%) HYDRALAZINE 0 1 (1%) 1 (<1%) HYDROCHLOROTHIAZIDE + CAPTOPRIL 0 1 (1%) 1 (<1%) HYDROCHLOROTHIAZIDE + ENALAPRIL 1 (1%) 0 1 (<1%) HYDROCHLOROTHIAZIDE + LOSARTAN 0 1 (1%) 1 (<1%) HYDROCHLOROTHIAZIDE + LOSARTAN POTASSIUM 0 1 (1%) 1 (<1%) HYDROCHLOROTHIAZIDE + METOPROLOL 0 1 (1%) 1 (<1%) HYDROCHLOROTHIAZIDE + TELMISARTAN 0 1 (1%) 1 (<1%) ISOSORBIDE 0 1 (1%) 1 (<1%) LERCANIDIPINE 0 1 (1%) 1 (<1%) LIDOCAINE 1 (1%) 0 1 (<1%) NEBIVOLOL 1 (1%) 0 1 (<1%) OLMESARTAN 1 (1%) 0 1 (<1%) PERINDOPRIL + AMLODIPINE 1 (1%) 0 1 (<1%) PRAVASTATIN 1 (1%) 0 1 (<1%) PREDNISOLONE 0 1 (1%) 1 (<1%) PROPRANOLOL 0 1 (1%) 1 (<1%) ROSUVASTATIN 1 (1%) 0 1 (<1%) SOTALOL 0 1 (1%) 1 (<1%) TELMISARTAN 1 (1%) 0 1 (<1%) TRIAMTERENE 0 1 (1%) 1 (<1%)

NERVOUS SYSTEM Any medication 37 (46%) 34 (43%) 71 (45%) CARBASALATE CALCIUM 14 (18%) 11 (14%) 25 (16%) ACETYLSALICYLIC ACID 6 (8%) 10 (13%) 16 (10%) CITALOPRAM 5 (6%) 3 (4%) 8 (5%) OXAZEPAM 3 (4%) 2 (3%) 5 (3%) AMITRIPTYLINE 2 (3%) 2 (3%) 4 (3%)


83

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Sub ATC Level 1 Study 1 Study 2 Total Ingredient (N=80) (N=79) (N=159) ------------------------------------------------------------------------------------------------- PARACETAMOL + TRAMADOL HYDROCHLORIDE 3 (4%) 1 (1%) 4 (3%) ZOPICLONE 3 (4%) 1 (1%) 4 (3%) CODEINE PHOSPHATE + PARACETAMOL 1 (1%) 2 (3%) 3 (2%) DIAZEPAM 2 (3%) 1 (1%) 3 (2%) PREGABALIN 2 (3%) 1 (1%) 3 (2%) GABAPENTIN 1 (1%) 1 (1%) 2 (1%) PARACETAMOL 1 (1%) 1 (1%) 2 (1%) SERTRALINE 2 (3%) 0 2 (1%) ZOLPIDEM TARTRATE 1 (1%) 1 (1%) 2 (1%) ALPRAZOLAM 0 1 (1%) 1 (<1%) AMITRIPTYLINE HYDROCHLORIDE 1 (1%) 0 1 (<1%) ARIPIPRAZOLE 1 (1%) 0 1 (<1%) BETAHISTINE 0 1 (1%) 1 (<1%) DIHYDROCODEINE 1 (1%) 0 1 (<1%) FENTANYL 1 (1%) 0 1 (<1%) LIDOCAINE 1 (1%) 0 1 (<1%) METHADONE HYDROCHLORIDE 0 1 (1%) 1 (<1%) MIRTAZAPINE 1 (1%) 0 1 (<1%) NORTRIPTYLINE 1 (1%) 0 1 (<1%) NORTRIPTYLINE HYDROCHLORIDE 0 1 (1%) 1 (<1%) OXYCODONE 1 (1%) 0 1 (<1%) PARACETAMOL + TRAMADOL 1 (1%) 0 1 (<1%) PAROXETINE 1 (1%) 0 1 (<1%) PILOCARPINE 0 1 (1%) 1 (<1%) ROPINIROLE 1 (1%) 0 1 (<1%) TEMAZEPAM 1 (1%) 0 1 (<1%) TRAMADOL HYDROCHLORIDE 1 (1%) 0 1 (<1%)

BLOOD AND BLOOD FORMING ORGANS Any medication 30 (38%) 36 (46%) 66 (42%) CARBASALATE CALCIUM 14 (18%) 11 (14%) 25 (16%)


84

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Sub ATC Level 1 Study 1 Study 2 Total Ingredient (N=80) (N=79) (N=159) ------------------------------------------------------------------------------------------------- ACETYLSALICYLIC ACID 6 (8%) 10 (13%) 16 (10%) CLOPIDOGREL 3 (4%) 2 (3%) 5 (3%) DIPYRIDAMOLE 2 (3%) 2 (3%) 4 (3%) PHENPROCOUMON 1 (1%) 3 (4%) 4 (3%) FOLIC ACID 2 (3%) 1 (1%) 3 (2%) APIXABAN 0 2 (3%) 2 (1%) DABIGATRAN ETEXILATE 0 2 (3%) 2 (1%) FERROUS FUMARATE 2 (3%) 0 2 (1%) FERROUS SULPHATE 1 (1%) 1 (1%) 2 (1%) TICAGRELOR 1 (1%) 1 (1%) 2 (1%) WARFARIN 1 (1%) 1 (1%) 2 (1%) ACENOCOUMAROL 0 1 (1%) 1 (<1%) CLOPIDOGREL BESYLATE 0 1 (1%) 1 (<1%) CLOPIDOGREL BISULFATE 0 1 (1%) 1 (<1%) CYANOCOBALAMIN 1 (1%) 0 1 (<1%) EDOXABAN 0 1 (1%) 1 (<1%) ENZYME, NOS 1 (1%) 0 1 (<1%) RIVAROXABAN 1 (1%) 0 1 (<1%)

MUSCULO-SKELETAL SYSTEM Any medication 24 (30%) 19 (24%) 43 (27%) ACETYLSALICYLIC ACID 6 (8%) 10 (13%) 16 (10%) ALENDRONATE SODIUM 4 (5%) 2 (3%) 6 (4%) ALLOPURINOL 1 (1%) 3 (4%) 4 (3%) NAPROXEN 2 (3%) 2 (3%) 4 (3%) COLECALCIFEROL + ALENDRONIC ACID 3 (4%) 0 3 (2%) ALENDRONIC ACID 1 (1%) 1 (1%) 2 (1%) CELECOXIB 2 (3%) 0 2 (1%) METHOTREXATE 1 (1%) 1 (1%) 2 (1%) COLCHICINE 0 1 (1%) 1 (<1%) DICLOFENAC 1 (1%) 0 1 (<1%)


85

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Sub ATC Level 1 Study 1 Study 2 Total Ingredient (N=80) (N=79) (N=159) ------------------------------------------------------------------------------------------------- DICLOFENAC SODIUM + MISOPROSTOL 1 (1%) 0 1 (<1%) ENZYME, NOS 1 (1%) 0 1 (<1%) FEBUXOSTAT 1 (1%) 0 1 (<1%) HYDROXYCHLOROQUINE SULFATE 0 1 (1%) 1 (<1%) QUININE 1 (1%) 0 1 (<1%)

GENITO URINARY SYSTEM AND SEX HORMONES Any medication 13 (16%) 13 (16%) 26 (16%) NAPROXEN 2 (3%) 2 (3%) 4 (3%) TAMSULOSIN 3 (4%) 1 (1%) 4 (3%) SOLIFENACIN SUCCINATE 2 (3%) 1 (1%) 3 (2%) TADALAFIL 1 (1%) 2 (3%) 3 (2%) DOXAZOSIN 1 (1%) 1 (1%) 2 (1%) TAMSULOSIN HYDROCHLORIDE + SOLIFENACIN SUCCINATE 0 2 (3%) 2 (1%) ALFUZOSIN 0 1 (1%) 1 (<1%) ESTRADIOL + DYDROGESTERONE 0 1 (1%) 1 (<1%) ESTRADIOL VALERATE 1 (1%) 0 1 (<1%) ESTROGENS CONJUGATED 0 1 (1%) 1 (<1%) SILDENAFIL 1 (1%) 0 1 (<1%) SOLIFENACIN 0 1 (1%) 1 (<1%) TAMSULOSIN + DUTASTERIDE 1 (1%) 0 1 (<1%) TIBOLONE 1 (1%) 0 1 (<1%)

RESPIRATORY SYSTEM Any medication 8 (10%) 14 (18%) 22 (14%) DESLORATADINE 1 (1%) 5 (6%) 6 (4%) FLUTICASONE FUROATE 0 2 (3%) 2 (1%) LEVOCETIRIZINE 1 (1%) 1 (1%) 2 (1%) BUDESONIDE + FORMOTEROL FUMARATE 0 1 (1%) 1 (<1%) CARBOCISTEINE 0 1 (1%) 1 (<1%) CETIRIZINE 1 (1%) 0 1 (<1%)


86

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Sub ATC Level 1 Study 1 Study 2 Total Ingredient (N=80) (N=79) (N=159) ------------------------------------------------------------------------------------------------- CETIRIZINE HYDROCHLORIDE 0 1 (1%) 1 (<1%) CLEMASTINE 0 1 (1%) 1 (<1%) DEXAMETHASONE 0 1 (1%) 1 (<1%) DIHYDROCODEINE 1 (1%) 0 1 (<1%) FLUTICASONE PROPIONATE 1 (1%) 0 1 (<1%) FORMOTEROL FUMARATE 1 (1%) 0 1 (<1%) LIDOCAINE 1 (1%) 0 1 (<1%) LORATADINE 1 (1%) 0 1 (<1%) MOMETASONE FUROATE 1 (1%) 0 1 (<1%) MONTELUKAST 0 1 (1%) 1 (<1%) PREDNISOLONE 0 1 (1%) 1 (<1%) RUPATADINE FUMARATE 1 (1%) 0 1 (<1%) SALBUTAMOL 1 (1%) 0 1 (<1%) TIOTROPIUM BROMIDE 0 1 (1%) 1 (<1%)

DERMATOLOGICALS Any medication 8 (10%) 9 (11%) 17 (11%) GLYCERYL TRINITRATE 5 (6%) 6 (8%) 11 (7%) BETAMETHASONE + CALCIPOTRIOL 0 1 (1%) 1 (<1%) CLEMASTINE 0 1 (1%) 1 (<1%) DEXAMETHASONE 0 1 (1%) 1 (<1%) EMOLLIENTS AND PROTECTIVES 0 1 (1%) 1 (<1%) ENZYME, NOS 1 (1%) 0 1 (<1%) HYDROCORTISONE BUTYRATE 0 1 (1%) 1 (<1%) LIDOCAINE 1 (1%) 0 1 (<1%) MOMETASONE FUROATE 1 (1%) 0 1 (<1%) PREDNISOLONE 0 1 (1%) 1 (<1%) ZINC OXIDE 1 (1%) 0 1 (<1%)


87

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Sub ATC Level 1 Study 1 Study 2 Total Ingredient (N=80) (N=79) (N=159) ------------------------------------------------------------------------------------------------- SYSTEMIC HORMONAL PREPARATIONS, EXCL. SEX HORMONES AND INSULINS Any medication 8 (10%) 7 (9%) 15 (9%) LEVOTHYROXINE 5 (6%) 0 5 (3%) PREDNISONE 3 (4%) 1 (1%) 4 (3%) LEVOTHYROXINE SODIUM 0 3 (4%) 3 (2%) DEXAMETHASONE 0 1 (1%) 1 (<1%) HYDROCORTISONE BUTYRATE 0 1 (1%) 1 (<1%) PREDNISOLONE 0 1 (1%) 1 (<1%)

SENSORY ORGANS Any medication 3 (4%) 5 (6%) 8 (5%) MACROGOL 0 2 (3%) 2 (1%) AZITHROMYCIN 1 (1%) 0 1 (<1%) DEXAMETHASONE 0 1 (1%) 1 (<1%) DICLOFENAC 1 (1%) 0 1 (<1%) ISOSORBIDE 0 1 (1%) 1 (<1%) LIDOCAINE 1 (1%) 0 1 (<1%) PILOCARPINE 0 1 (1%) 1 (<1%) PREDNISOLONE 0 1 (1%) 1 (<1%)

ANTINEOPLASTIC AND IMMUNOMODULATING AGENTS Any medication 3 (4%) 4 (5%) 7 (4%) METHOTREXATE 1 (1%) 1 (1%) 2 (1%) ADALIMUMAB 0 1 (1%) 1 (<1%) ESTRADIOL VALERATE 1 (1%) 0 1 (<1%) ESTROGENS CONJUGATED 0 1 (1%) 1 (<1%) GOSERELIN 0 1 (1%) 1 (<1%) HYDROXYCARBAMIDE 1 (1%) 0 1 (<1%)


88

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Sub ATC Level 1 Study 1 Study 2 Total Ingredient (N=80) (N=79) (N=159) ------------------------------------------------------------------------------------------------- ANTIINFECTIVES FOR SYSTEMIC USE Any medication 1 (1%) 1 (1%) 2 (1%) AMOXICILLIN 0 1 (1%) 1 (<1%) AMOXICILLIN + CLAVULANIC ACID 0 1 (1%) 1 (<1%) AZITHROMYCIN 1 (1%) 0 1 (<1%) CLARITHROMYCIN 0 1 (1%) 1 (<1%)

ANTIPARASITIC PRODUCTS, INSECTICIDES AND REPELLENTS Any medication 1 (1%) 1 (1%) 2 (1%) HYDROXYCHLOROQUINE SULFATE 0 1 (1%) 1 (<1%) QUININE 1 (1%) 0 1 (<1%)

VARIOUS Any medication 1 (1%) 1 (1%) 2 (1%) TOLBUTAMIDE 1 (1%) 1 (1%) 2 (1%)


89

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 1Population: Intent-to-Treat Table 1.20 Summary of Naive to Inhaler Devices

Sub Sub Study 1 Study 2 Total (N=80) (N=79) (N=159) ------------------------------------------------------------------- n 80 79 159 ELLIPTA 80(100%) 79(100%) 159(100%) DISKUS + HandiHaler 80(100%) 36 (46%) 116 (73%) Turbuhaler + HandiHaler 53 (66%) 79(100%) 132 (83%)

Note: Subject had recent experience with the ELLIPTA inhaler within 24 months prior to screening.However, the subject has been recorded as being naive to ELLIPTA in the clinical trial databaseand thus reported as such.

90

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

Protocol: 206215 Page 1 of 2Population: Intent-to-Treat

Figure 2.01Percentage of Subjects with at Least One Critical Error by Assessment

Sub Study 1: ELLIPTA vs DISKUS + HandiHaler

Inhaler ELLIPTA DISKUS + HandiHaler

Per

cent

age

of S

ubje

cts

0

10

20

30

40

50

60

70

80

90

100

After Reading Leaflet After 1st Instruction from HCP After 2nd Instruction from HCP

91

CONFIDENTIAL 2017N324761_00206215

PPD


Figure 2.01Percentage of Subjects with at Least One Critical Error by Assessment

Sub Study 2: ELLIPTA vs Turbuhaler + HandiHaler

Inhaler ELLIPTA Turbuhaler + HandiHaler

Per

cent

age

of S

ubje

cts

0

10

20

30

40

50

60

70

80

90

100


92

CONFIDENTIAL 2017N324761_00206215

PPD


Figure 2.02Percentage of Subjects with at Least One Error by Assessment



Per

cent

age

of S

ubje

cts

0

10

20

30

40

50

60

70

80

90

100


93

CONFIDENTIAL 2017N324761_00206215

PPD


Figure 2.02Percentage of Subjects with at Least One Error by Assessment



Per

cent

age

of S

ubje

cts

0

10

20

30

40

50

60

70

80

90

100


94

CONFIDENTIAL 2017N324761_00206215

PPD


Figure 2.03Number of Instructions required from a HCP to Demonstrate Correct Use


use on one device and not the other, the subject would be counted as having failed to demonstrate correct use.Note: For the DISKUS+Handihaler and Turbuhaler+HandiHaler groups, if a subject failed to demonstrate correct


Per

cent

age

of S

ubje

cts

0

10

20

30

40

50

60

70

80

90

100

No HCP Instructions(PIL Only)

1 HCP Instruction 2 HCP Instructions Failed to demonstratecorrect use

95

CONFIDENTIAL 2017N324761_00206215

PPD


Figure 2.03Number of Instructions required from a HCP to Demonstrate Correct Use


use on one device and not the other, the subject would be counted as having failed to demonstrate correct use.Note: For the DISKUS+Handihaler and Turbuhaler+HandiHaler groups, if a subject failed to demonstrate correct


Per

cent

age

of S

ubje

cts

0

10

20

30

40

50

60

70

80

90

100

No HCP Instructions(PIL Only)

1 HCP Instruction 2 HCP Instructions Failed to demonstratecorrect use

96

CONFIDENTIAL 2017N324761_00206215

PPD

Pro

babi

lity

of D

emon

stra

ting

Cor

rect

Use

(%

)

0

10

20

30

40

50

60

70

80

90

100

Time (Minutes)

0 5 10 15 20 25 30 35 40 45 50 55 60


Figure 2.04Kaplan-Meier Plot of Time Taken to Demonstrate Correct Inhaler Use

Sub Study 1: ELLIPTA vs DISKUS + HandiHalerAssessment:Time taken to read the PIL and Demonstrate Correct Use

to demonstrate correct use after reading the PIL assessment.assessment was not recorded, therefore this subject’s ELLIPTA data could only be included in the time takenafter receiving the first HCP instruction. The time taken to demonstrate correct use at the HCPSubject did not demonstrate correct use for the ELLIPTA device after reading the PIL, but was able tothe reading and/or instruction time for both devices.use on only one of the two devices, the time to demonstrate correct use is censored at the total ofNote: For the DISKUS+HandiHaler and Turbuhaler+HandiHaler groups, if a subject demonstrated correct


97

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

Pro

babi

lity

of D

emon

stra

ting

Cor

rect

Use

(%

)

0

10

20

30

40

50

60

70

80

90

100

Time (Minutes)

0 5 10 15 20 25 30 35 40 45 50 55 60



Sub Study 1: ELLIPTA vs DISKUS + HandiHalerAssessment:Time for HCP Instruction and to Demonstrate Correct Use



98

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

Pro

babi

lity

of D

emon

stra

ting

Cor

rect

Use

(%

)

0

10

20

30

40

50

60

70

80

90

100

Time (Minutes)

0 5 10 15 20 25 30 35 40 45 50 55 60



Sub Study 1: ELLIPTA vs DISKUS + HandiHalerAssessment:Total Time to demonstrate correct use



99

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

Pro

babi

lity

of D

emon

stra

ting

Cor

rect

Use

(%

)

0

10

20

30

40

50

60

70

80

90

100

Time (Minutes)

0 5 10 15 20 25 30 35 40 45 50 55 60



Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerAssessment:Time taken to read the PIL and Demonstrate Correct Use



100

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

Pro

babi

lity

of D

emon

stra

ting

Cor

rect

Use

(%

)

0

10

20

30

40

50

60

70

80

90

100

Time (Minutes)

0 5 10 15 20 25 30 35 40 45 50 55 60



Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerAssessment:Time for HCP Instruction and to Demonstrate Correct Use



101

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

Pro

babi

lity

of D

emon

stra

ting

Cor

rect

Use

(%

)

0

10

20

30

40

50

60

70

80

90

100

Time (Minutes)

0 5 10 15 20 25 30 35 40 45 50 55 60



Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerAssessment:Total Time to demonstrate correct use



102

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

Protocol: 206215 Page 1 of 2Population: Intent-to-Treat Table 2.01 Summary of Errors on ELLIPTA

Sub Study 1: ELLIPTA vs DISKUS + HandiHaler After After 1st After 2nd Reading Instruction Instruction Leaflet from HCP from HCP Inhaler errors test (N=80) (N=80) (N=80) ----------------------------------------------------------------------------------------- Number of Subjects with Errors 19 (24%) 4 (5%) 3 (4%) Number of Subjects with Critical Errors 7 (9%) 1 (1%) 0

Total Number of Errors 38 6 5 Total Number of Critical Errors 8 1 0

Failed to open cover [1] 2 (11%) 0 0 Inhalation manoeuvre: long, steady, deep 8 (42%) 1 (25%) 1 (33%) Blocked air inlet during inhalation manoeuvre 0 0 0 Shook the device after dose preparation [1] 0 0 0 No exhalation before an inhalation 11 (58%) 2 (50%) 3 (100%) Exhaled directly into mouthpiece [1] 5 (26%) 1 (25%) 0 No seal by the lips round the mouthpiece during 1 (5%) 0 0 the inhalation [1] Did not hold breath 7 (37%) 2 (50%) 1 (33%) Did not close the device (Note: this is an error 4 (21%) 0 0 but one which does not affect the medication that is inhaled)

[1] Indicates a Critical Error.Note: Percentages for number of subjects are calculated from the total number of subjects who used thedevice(s), percentages for type of errors are calculated from the number of subjects with errors.

103

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 2 of 2Population: Intent-to-Treat Table 2.01 Summary of Errors on ELLIPTA

Sub Study 2: ELLIPTA vs Turbuhaler + HandiHaler After After 1st After 2nd Reading Instruction Instruction Leaflet from HCP from HCP Inhaler errors test (N=79) (N=79) (N=79) ----------------------------------------------------------------------------------------- Number of Subjects with Errors 17 (22%) 4 (5%) 1 (1%) Number of Subjects with Critical Errors 7 (9%) 2 (3%) 0


Failed to open cover [1] 3 (18%) 0 0 Inhalation manoeuvre: long, steady, deep 5 (29%) 1 (25%) 0 Blocked air inlet during inhalation manoeuvre 1 (6%) 0 0 Shook the device after dose preparation [1] 1 (6%) 0 0 No exhalation before an inhalation 14 (82%) 3 (75%) 1 (100%) Exhaled directly into mouthpiece [1] 7 (41%) 2 (50%) 0 No seal by the lips round the mouthpiece during 2 (12%) 0 0 the inhalation [1] Did not hold breath 8 (47%) 0 0 Did not close the device (Note: this is an error 2 (12%) 0 0 but one which does not affect the medication that is inhaled)


104

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 4Population: Intent-to-Treat Table 2.02 Summary of Errors on ELLIPTA by Country

Sub Study 1: ELLIPTA vs DISKUS + HandiHalerCountry: Netherlands After After 1st After 2nd Reading Instruction Instruction Leaflet from HCP from HCP Inhaler errors test (N=62) (N=62) (N=62) ----------------------------------------------------------------------------------------- Number of Subjects with Errors 12 (19%) 1 (2%) 0 Number of Subjects with Critical Errors 3 (5%) 0 0


Failed to open cover [1] 0 0 0 Inhalation manoeuvre: long, steady, deep 6 (50%) 1 (100%) 0 Blocked air inlet during inhalation manoeuvre 0 0 0 Shook the device after dose preparation [1] 0 0 0 No exhalation before an inhalation 8 (67%) 0 0 Exhaled directly into mouthpiece [1] 2 (17%) 0 0 No seal by the lips round the mouthpiece during 1 (8%) 0 0 the inhalation [1] Did not hold breath 5 (42%) 1 (100%) 0 Did not close the device (Note: this is an error 0 0 0 but one which does not affect the medication that is inhaled)


105

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 1: ELLIPTA vs DISKUS + HandiHalerCountry: United Kingdom After After 1st After 2nd Reading Instruction Instruction Leaflet from HCP from HCP Inhaler errors test (N=18) (N=18) (N=18) ----------------------------------------------------------------------------------------- Number of Subjects with Errors 7 (39%) 3 (17%) 3 (17%) Number of Subjects with Critical Errors 4 (22%) 1 (6%) 0


Failed to open cover [1] 2 (29%) 0 0 Inhalation manoeuvre: long, steady, deep 2 (29%) 0 1 (33%) Blocked air inlet during inhalation manoeuvre 0 0 0 Shook the device after dose preparation [1] 0 0 0 No exhalation before an inhalation 3 (43%) 2 (67%) 3 (100%) Exhaled directly into mouthpiece [1] 3 (43%) 1 (33%) 0 No seal by the lips round the mouthpiece during 0 0 0 the inhalation [1] Did not hold breath 2 (29%) 1 (33%) 1 (33%) Did not close the device (Note: this is an error 4 (57%) 0 0 but one which does not affect the medication that is inhaled)


106

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerCountry: Netherlands After After 1st After 2nd Reading Instruction Instruction Leaflet from HCP from HCP Inhaler errors test (N=60) (N=60) (N=60) ----------------------------------------------------------------------------------------- Number of Subjects with Errors 12 (20%) 2 (3%) 0 Number of Subjects with Critical Errors 5 (8%) 1 (2%) 0


Failed to open cover [1] 2 (17%) 0 0 Inhalation manoeuvre: long, steady, deep 4 (33%) 1 (50%) 0 Blocked air inlet during inhalation manoeuvre 1 (8%) 0 0 Shook the device after dose preparation [1] 1 (8%) 0 0 No exhalation before an inhalation 11 (92%) 1 (50%) 0 Exhaled directly into mouthpiece [1] 5 (42%) 1 (50%) 0 No seal by the lips round the mouthpiece during 1 (8%) 0 0 the inhalation [1] Did not hold breath 7 (58%) 0 0 Did not close the device (Note: this is an error 1 (8%) 0 0 but one which does not affect the medication that is inhaled)


107

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerCountry: United Kingdom After After 1st After 2nd Reading Instruction Instruction Leaflet from HCP from HCP Inhaler errors test (N=19) (N=19) (N=19) ----------------------------------------------------------------------------------------- Number of Subjects with Errors 5 (26%) 2 (11%) 1 (5%) Number of Subjects with Critical Errors 2 (11%) 1 (5%) 0


Failed to open cover [1] 1 (20%) 0 0 Inhalation manoeuvre: long, steady, deep 1 (20%) 0 0 Blocked air inlet during inhalation manoeuvre 0 0 0 Shook the device after dose preparation [1] 0 0 0 No exhalation before an inhalation 3 (60%) 2 (100%) 1 (100%) Exhaled directly into mouthpiece [1] 2 (40%) 1 (50%) 0 No seal by the lips round the mouthpiece during 1 (20%) 0 0 the inhalation [1] Did not hold breath 1 (20%) 0 0 Did not close the device (Note: this is an error 1 (20%) 0 0 but one which does not affect the medication that is inhaled)


108

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 3Population: Intent-to-Treat Table 2.03 Summary of Errors on DISKUS + HandiHaler

Sub Study 1: ELLIPTA vs DISKUS + HandiHalerInhaler: DISKUS + HandiHaler After After 1st After 2nd Reading Instruction Instruction Leaflet from HCP from HCP Inhaler errors test (N=80) (N=80) (N=80) --------------------------------------------------------------------------------------------------- Number of Subjects with Errors 64 (80%) 12 (15%) 6 (8%) Number of Subjects with Critical Errors 60 (75%) 9 (11%) 5 (6%)



109

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 1: ELLIPTA vs DISKUS + HandiHalerInhaler: DISKUS After After 1st After 2nd Reading Instruction Instruction Leaflet from HCP from HCP Inhaler errors test (N=80) (N=80) (N=80) --------------------------------------------------------------------------------------------------- Number of Subjects with Errors 44 (55%) 8 (10%) 3 (4%) Number of Subjects with Critical Errors 32 (40%) 4 (5%) 3 (4%)


Failed to open cover [1] 5 (11%) 0 0 Lever is not pushed back [1] 27 (61%) 4 (50%) 3 (100%) Shook the device after dose preparation [1] 1 (2%) 0 0 No exhalation before an inhalation 20 (45%) 4 (50%) 1 (33%) Exhaled directly into mouthpiece [1] 7 (16%) 2 (25%) 0 No seal by the lips round the mouthpiece during the 1 (2%) 0 0 inhalation [1] Inhalation manoeuvre: steady, deep 9 (20%) 0 0 Did not hold breath 10 (23%) 1 (13%) 0 Did not close the device (Note: this is an error but one 10 (23%) 2 (25%) 1 (33%) which does not affect the medication that is inhaled)


110

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 1: ELLIPTA vs DISKUS + HandiHalerInhaler: HandiHaler After After 1st After 2nd Reading Instruction Instruction Leaflet from HCP from HCP Inhaler errors test (N=80) (N=80) (N=80) --------------------------------------------------------------------------------------------------- Number of Subjects with Errors 57 (71%) 9 (11%) 4 (5%) Number of Subjects with Critical Errors 54 (68%) 7 (9%) 3 (4%)


Blocked air inlet during inhalation manoeuvre 3 (5%) 0 0 Failed to remove capsule [1] 7 (12%) 1 (11%) 0 Failed to insert capsule into chamber [1] 6 (11%) 0 0 Did not completely close device capsule chamber (heard click 17 (30%) 0 2 (50%) when satisfactory) [1] Shook the device after dose preparation 4 (7%) 0 0 Capsule did not rattle [1] 43 (75%) 2 (22%) 1 (25%) Did not check inside the capsule chamber if powder was left 23 (40%) 4 (44%) 1 (25%) / did not make a second inhalation Did not pierce the capsule (HCP should check capsule was 18 (32%) 5 (56%) 3 (75%) pierced) [1] Inhalation manoeuvre: was not slow, deep 10 (18%) 0 0 No exhalation before an inhalation 8 (14%) 1 (11%) 2 (50%) Exhaled directly into mouthpiece [1] 4 (7%) 1 (11%) 0 No seal by the lips round the mouthpiece during the 3 (5%) 0 0 inhalation [1] Did not hold breath 6 (11%) 2 (22%) 1 (25%)


111

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 6Population: Intent-to-Treat Table 2.04 Summary of Errors on DISKUS + HandiHaler by Country

Sub Study 1: ELLIPTA vs DISKUS + HandiHalerCountry: NetherlandsInhaler: DISKUS + HandiHaler After After 1st After 2nd Reading Instruction Instruction Leaflet from HCP from HCP Inhaler errors test (N=62) (N=62) (N=62) --------------------------------------------------------------------------------------------------- Number of Subjects with Errors 50 (81%) 5 (8%) 2 (3%) Number of Subjects with Critical Errors 48 (77%) 3 (5%) 2 (3%)



112

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 1: ELLIPTA vs DISKUS + HandiHalerCountry: NetherlandsInhaler: DISKUS After After 1st After 2nd Reading Instruction Instruction Leaflet from HCP from HCP Inhaler errors test (N=62) (N=62) (N=62) --------------------------------------------------------------------------------------------------- Number of Subjects with Errors 34 (55%) 4 (6%) 2 (3%) Number of Subjects with Critical Errors 25 (40%) 2 (3%) 2 (3%)


Failed to open cover [1] 3 (9%) 0 0 Lever is not pushed back [1] 23 (68%) 2 (50%) 2 (100%) Shook the device after dose preparation [1] 1 (3%) 0 0 No exhalation before an inhalation 18 (53%) 1 (25%) 0 Exhaled directly into mouthpiece [1] 5 (15%) 1 (25%) 0 No seal by the lips round the mouthpiece during the 1 (3%) 0 0 inhalation [1] Inhalation manoeuvre: steady, deep 8 (24%) 0 0 Did not hold breath 9 (26%) 1 (25%) 0 Did not close the device (Note: this is an error but one 5 (15%) 0 0 which does not affect the medication that is inhaled)


113

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 1: ELLIPTA vs DISKUS + HandiHalerCountry: NetherlandsInhaler: HandiHaler After After 1st After 2nd Reading Instruction Instruction Leaflet from HCP from HCP Inhaler errors test (N=62) (N=62) (N=62) --------------------------------------------------------------------------------------------------- Number of Subjects with Errors 44 (71%) 2 (3%) 0 Number of Subjects with Critical Errors 44 (71%) 1 (2%) 0


Blocked air inlet during inhalation manoeuvre 2 (5%) 0 0 Failed to remove capsule [1] 1 (2%) 0 0 Failed to insert capsule into chamber [1] 4 (9%) 0 0 Did not completely close device capsule chamber (heard click 13 (30%) 0 0 when satisfactory) [1] Shook the device after dose preparation 3 (7%) 0 0 Capsule did not rattle [1] 40 (91%) 1 (50%) 0 Did not check inside the capsule chamber if powder was left 13 (30%) 1 (50%) 0 / did not make a second inhalation Did not pierce the capsule (HCP should check capsule was 13 (30%) 0 0 pierced) [1] Inhalation manoeuvre: was not slow, deep 8 (18%) 0 0 No exhalation before an inhalation 5 (11%) 0 0 Exhaled directly into mouthpiece [1] 3 (7%) 0 0 No seal by the lips round the mouthpiece during the 2 (5%) 0 0 inhalation [1] Did not hold breath 4 (9%) 0 0


114

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 1: ELLIPTA vs DISKUS + HandiHalerCountry: United KingdomInhaler: DISKUS + HandiHaler After After 1st After 2nd Reading Instruction Instruction Leaflet from HCP from HCP Inhaler errors test (N=18) (N=18) (N=18) --------------------------------------------------------------------------------------------------- Number of Subjects with Errors 14 (78%) 7 (39%) 4 (22%) Number of Subjects with Critical Errors 12 (67%) 6 (33%) 3 (17%)



115

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 1: ELLIPTA vs DISKUS + HandiHalerCountry: United KingdomInhaler: DISKUS After After 1st After 2nd Reading Instruction Instruction Leaflet from HCP from HCP Inhaler errors test (N=18) (N=18) (N=18) --------------------------------------------------------------------------------------------------- Number of Subjects with Errors 10 (56%) 4 (22%) 1 (6%) Number of Subjects with Critical Errors 7 (39%) 2 (11%) 1 (6%)


Failed to open cover [1] 2 (20%) 0 0 Lever is not pushed back [1] 4 (40%) 2 (50%) 1 (100%) Shook the device after dose preparation [1] 0 0 0 No exhalation before an inhalation 2 (20%) 3 (75%) 1 (100%) Exhaled directly into mouthpiece [1] 2 (20%) 1 (25%) 0 No seal by the lips round the mouthpiece during the 0 0 0 inhalation [1] Inhalation manoeuvre: steady, deep 1 (10%) 0 0 Did not hold breath 1 (10%) 0 0 Did not close the device (Note: this is an error but one 5 (50%) 2 (50%) 1 (100%) which does not affect the medication that is inhaled)


116

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 1: ELLIPTA vs DISKUS + HandiHalerCountry: United KingdomInhaler: HandiHaler After After 1st After 2nd Reading Instruction Instruction Leaflet from HCP from HCP Inhaler errors test (N=18) (N=18) (N=18) --------------------------------------------------------------------------------------------------- Number of Subjects with Errors 13 (72%) 7 (39%) 4 (22%) Number of Subjects with Critical Errors 10 (56%) 6 (33%) 3 (17%)


Blocked air inlet during inhalation manoeuvre 1 (8%) 0 0 Failed to remove capsule [1] 6 (46%) 1 (14%) 0 Failed to insert capsule into chamber [1] 2 (15%) 0 0 Did not completely close device capsule chamber (heard click 4 (31%) 0 2 (50%) when satisfactory) [1] Shook the device after dose preparation 1 (8%) 0 0 Capsule did not rattle [1] 3 (23%) 1 (14%) 1 (25%) Did not check inside the capsule chamber if powder was left 10 (77%) 3 (43%) 1 (25%) / did not make a second inhalation Did not pierce the capsule (HCP should check capsule was 5 (38%) 5 (71%) 3 (75%) pierced) [1] Inhalation manoeuvre: was not slow, deep 2 (15%) 0 0 No exhalation before an inhalation 3 (23%) 1 (14%) 2 (50%) Exhaled directly into mouthpiece [1] 1 (8%) 1 (14%) 0 No seal by the lips round the mouthpiece during the 1 (8%) 0 0 inhalation [1] Did not hold breath 2 (15%) 2 (29%) 1 (25%)


117

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 3Population: Intent-to-Treat Table 2.05 Summary of Errors on Turbuhaler + HandiHaler

Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerInhaler: Turbuhaler + HandiHaler After After 1st After 2nd Reading Instruction Instruction Leaflet from HCP from HCP Inhaler errors test (N=79) (N=79) (N=79) --------------------------------------------------------------------------------------------------- Number of Subjects with Errors 63 (80%) 16 (20%) 5 (6%) Number of Subjects with Critical Errors 58 (73%) 11 (14%) 4 (5%)



118

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerInhaler: Turbuhaler After After 1st After 2nd Reading Instruction Instruction Leaflet from HCP from HCP Inhaler errors test (N=79) (N=79) (N=79) --------------------------------------------------------------------------------------------------- Number of Subjects with Errors 45 (57%) 12 (15%) 4 (5%) Number of Subjects with Critical Errors 39 (49%) 8 (10%) 3 (4%)


Blocked air inlet during inhalation manoeuvre 0 0 0 Failed to remove cap [1] 3 (7%) 1 (8%) 0 Did not hold device upright (+/=45% OK) during dose 14 (31%) 3 (25%) 2 (50%) preparation [1] Base not twisted fully backwards and forwards, no click 27 (60%) 5 (42%) 0 heard [1] Device tipped downwards after dose preparation 13 (29%) 1 (8%) 1 (25%) Inhalation manoeuvre: forceful, deep (HCP: it is important 11 (24%) 3 (25%) 0 that the inhalation is forceful and deep from the start for this inhaler) Shook the device after dose preparation [1] 0 1 (8%) 2 (50%) No exhalation before an inhalation 15 (33%) 3 (25%) 0 Exhaled directly into mouthpiece [1] 5 (11%) 1 (8%) 0 No seal by the lips round the mouthpiece during the 1 (2%) 0 0 inhalation [1] Did not hold breath 11 (24%) 2 (17%) 1 (25%) Did not close the device (Note: this is an error but one 6 (13%) 1 (8%) 0 which does not affect the medication that is inhaled)


119

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerInhaler: HandiHaler After After 1st After 2nd Reading Instruction Instruction Leaflet from HCP from HCP Inhaler errors test (N=79) (N=79) (N=79) --------------------------------------------------------------------------------------------------- Number of Subjects with Errors 50 (63%) 7 (9%) 2 (3%) Number of Subjects with Critical Errors 45 (57%) 4 (5%) 2 (3%)


Blocked air inlet during inhalation manoeuvre 4 (8%) 0 0 Failed to remove capsule [1] 7 (14%) 0 0 Failed to insert capsule into chamber [1] 6 (12%) 0 0 Did not completely close device capsule chamber (heard click 16 (32%) 0 0 when satisfactory) [1] Shook the device after dose preparation 1 (2%) 0 1 (50%) Capsule did not rattle [1] 33 (66%) 3 (43%) 2 (100%) Did not check inside the capsule chamber if powder was left 20 (40%) 5 (71%) 0 / did not make a second inhalation Did not pierce the capsule (HCP should check capsule was 14 (28%) 1 (14%) 0 pierced) [1] Inhalation manoeuvre: was not slow, deep 8 (16%) 2 (29%) 1 (50%) No exhalation before an inhalation 10 (20%) 1 (14%) 1 (50%) Exhaled directly into mouthpiece [1] 6 (12%) 1 (14%) 0 No seal by the lips round the mouthpiece during the 2 (4%) 0 1 (50%) inhalation [1] Did not hold breath 6 (12%) 2 (29%) 1 (50%)


120

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 6Population: Intent-to-Treat Table 2.06 Summary of Errors on Turbuhaler + HandiHaler by Country

Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerCountry: NetherlandsInhaler: Turbuhaler + HandiHaler After After 1st After 2nd Reading Instruction Instruction Leaflet from HCP from HCP Inhaler errors test (N=60) (N=60) (N=60) --------------------------------------------------------------------------------------------------- Number of Subjects with Errors 46 (77%) 9 (15%) 1 (2%) Number of Subjects with Critical Errors 42 (70%) 7 (12%) 0



121

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerCountry: NetherlandsInhaler: Turbuhaler After After 1st After 2nd Reading Instruction Instruction Leaflet from HCP from HCP Inhaler errors test (N=60) (N=60) (N=60) --------------------------------------------------------------------------------------------------- Number of Subjects with Errors 30 (50%) 8 (13%) 1 (2%) Number of Subjects with Critical Errors 26 (43%) 5 (8%) 0


Blocked air inlet during inhalation manoeuvre 0 0 0 Failed to remove cap [1] 0 1 (13%) 0 Did not hold device upright (+/=45% OK) during dose 6 (20%) 2 (25%) 0 preparation [1] Base not twisted fully backwards and forwards, no click 20 (67%) 3 (38%) 0 heard [1] Device tipped downwards after dose preparation 6 (20%) 0 0 Inhalation manoeuvre: forceful, deep (HCP: it is important 10 (33%) 3 (38%) 0 that the inhalation is forceful and deep from the start for this inhaler) Shook the device after dose preparation [1] 0 0 0 No exhalation before an inhalation 11 (37%) 3 (38%) 0 Exhaled directly into mouthpiece [1] 5 (17%) 1 (13%) 0 No seal by the lips round the mouthpiece during the 1 (3%) 0 0 inhalation [1] Did not hold breath 7 (23%) 2 (25%) 1 (100%) Did not close the device (Note: this is an error but one 3 (10%) 0 0 which does not affect the medication that is inhaled)


122

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerCountry: NetherlandsInhaler: HandiHaler After After 1st After 2nd Reading Instruction Instruction Leaflet from HCP from HCP Inhaler errors test (N=60) (N=60) (N=60) --------------------------------------------------------------------------------------------------- Number of Subjects with Errors 37 (62%) 3 (5%) 0 Number of Subjects with Critical Errors 34 (57%) 3 (5%) 0


Blocked air inlet during inhalation manoeuvre 4 (11%) 0 0 Failed to remove capsule [1] 1 (3%) 0 0 Failed to insert capsule into chamber [1] 3 (8%) 0 0 Did not completely close device capsule chamber (heard click 15 (41%) 0 0 when satisfactory) [1] Shook the device after dose preparation 0 0 0 Capsule did not rattle [1] 28 (76%) 2 (67%) 0 Did not check inside the capsule chamber if powder was left 13 (35%) 2 (67%) 0 / did not make a second inhalation Did not pierce the capsule (HCP should check capsule was 10 (27%) 1 (33%) 0 pierced) [1] Inhalation manoeuvre: was not slow, deep 6 (16%) 1 (33%) 0 No exhalation before an inhalation 6 (16%) 0 0 Exhaled directly into mouthpiece [1] 5 (14%) 1 (33%) 0 No seal by the lips round the mouthpiece during the 2 (5%) 0 0 inhalation [1] Did not hold breath 4 (11%) 0 0


123

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerCountry: United KingdomInhaler: Turbuhaler + HandiHaler After After 1st After 2nd Reading Instruction Instruction Leaflet from HCP from HCP Inhaler errors test (N=19) (N=19) (N=19) --------------------------------------------------------------------------------------------------- Number of Subjects with Errors 17 (89%) 7 (37%) 4 (21%) Number of Subjects with Critical Errors 16 (84%) 4 (21%) 4 (21%)



124

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerCountry: United KingdomInhaler: Turbuhaler After After 1st After 2nd Reading Instruction Instruction Leaflet from HCP from HCP Inhaler errors test (N=19) (N=19) (N=19) --------------------------------------------------------------------------------------------------- Number of Subjects with Errors 15 (79%) 4 (21%) 3 (16%) Number of Subjects with Critical Errors 13 (68%) 3 (16%) 3 (16%)


Blocked air inlet during inhalation manoeuvre 0 0 0 Failed to remove cap [1] 3 (20%) 0 0 Did not hold device upright (+/=45% OK) during dose 8 (53%) 1 (25%) 2 (67%) preparation [1] Base not twisted fully backwards and forwards, no click 7 (47%) 2 (50%) 0 heard [1] Device tipped downwards after dose preparation 7 (47%) 1 (25%) 1 (33%) Inhalation manoeuvre: forceful, deep (HCP: it is important 1 (7%) 0 0 that the inhalation is forceful and deep from the start for this inhaler) Shook the device after dose preparation [1] 0 1 (25%) 2 (67%) No exhalation before an inhalation 4 (27%) 0 0 Exhaled directly into mouthpiece [1] 0 0 0 No seal by the lips round the mouthpiece during the 0 0 0 inhalation [1] Did not hold breath 4 (27%) 0 0 Did not close the device (Note: this is an error but one 3 (20%) 1 (25%) 0 which does not affect the medication that is inhaled)


125

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerCountry: United KingdomInhaler: HandiHaler After After 1st After 2nd Reading Instruction Instruction Leaflet from HCP from HCP Inhaler errors test (N=19) (N=19) (N=19) --------------------------------------------------------------------------------------------------- Number of Subjects with Errors 13 (68%) 4 (21%) 2 (11%) Number of Subjects with Critical Errors 11 (58%) 1 (5%) 2 (11%)


Blocked air inlet during inhalation manoeuvre 0 0 0 Failed to remove capsule [1] 6 (46%) 0 0 Failed to insert capsule into chamber [1] 3 (23%) 0 0 Did not completely close device capsule chamber (heard click 1 (8%) 0 0 when satisfactory) [1] Shook the device after dose preparation 1 (8%) 0 1 (50%) Capsule did not rattle [1] 5 (38%) 1 (25%) 2 (100%) Did not check inside the capsule chamber if powder was left 7 (54%) 3 (75%) 0 / did not make a second inhalation Did not pierce the capsule (HCP should check capsule was 4 (31%) 0 0 pierced) [1] Inhalation manoeuvre: was not slow, deep 2 (15%) 1 (25%) 1 (50%) No exhalation before an inhalation 4 (31%) 1 (25%) 1 (50%) Exhaled directly into mouthpiece [1] 1 (8%) 0 0 No seal by the lips round the mouthpiece during the 0 0 1 (50%) inhalation [1] Did not hold breath 2 (15%) 2 (50%) 1 (50%)


126

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 6Population: Intent-to-Treat Table 2.07 Summary of Critical Errors

Sub Study 1: ELLIPTA vs DISKUS + HandiHalerTimepoint: After Reading Leaflet Total (N=80)------------------------------------------------------------------------------------------------------------At least one Critical Error on ELLIPTA 7 (9%)Zero Critical Errors on ELLIPTA 73 (91%)

At least one Critical Error on DISKUS + HandiHaler 60 (75%)Zero Critical Errors on DISKUS + HandiHaler 20 (25%)

At least one Critical Error on ELLIPTA and DISKUS + HandiHaler 7 (9%)Zero Critical Errors on ELLIPTA or DISKUS + HandiHaler 73 (91%)

Number of subjects with discordant results [1] 53 (66%)At least one Critical Error on ELLIPTA and Zero Critical Errors on DISKUS + HandiHaler [2] 0At least one Critical Error on DISKUS + HandiHaler and Zero Critical Errors on ELLIPTA [2] 53 (100%)

[1] Defined as an error in one device and not the other.[2] Percentage is out of the number of subjects with discordant results.Subjects who made no errors after reading PIL were not given instruction by HCP and so are counted as havingzero critical errors after 1st and 2nd instruction from HCP

127

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 1: ELLIPTA vs DISKUS + HandiHalerTimepoint: After 1st Instruction from HCP Total (N=80)------------------------------------------------------------------------------------------------------------At least one Critical Error on ELLIPTA 1 (1%)Zero Critical Errors on ELLIPTA 79 (99%)





128

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 1: ELLIPTA vs DISKUS + HandiHalerTimepoint: After 2nd Instruction from HCP Total (N=80)------------------------------------------------------------------------------------------------------------At least one Critical Error on ELLIPTA 0Zero Critical Errors on ELLIPTA 80 (100%)


At least one Critical Error on ELLIPTA and DISKUS + HandiHaler 0Zero Critical Errors on ELLIPTA or DISKUS + HandiHaler 80 (100%)



129

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerTimepoint: After Reading Leaflet Total (N=79)------------------------------------------------------------------------------------------------------------At least one Critical Error on ELLIPTA 7 (9%)Zero Critical Errors on ELLIPTA 72 (91%)

At least one Critical Error on Turbuhaler + HandiHaler 58 (73%)Zero Critical Errors on Turbuhaler + HandiHaler 21 (27%)

At least one Critical Error on ELLIPTA and Turbuhaler + HandiHaler 7 (9%)Zero Critical Errors on ELLIPTA or Turbuhaler + HandiHaler 72 (91%)

Number of subjects with discordant results [1] 51 (65%)At least one Critical Error on ELLIPTA and Zero Critical Errors on Turbuhaler + HandiHaler [2] 0At least one Critical Error on Turbuhaler + HandiHaler and Zero Critical Errors on ELLIPTA [2] 51 (100%)


130

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerTimepoint: After 1st Instruction from HCP Total (N=79)------------------------------------------------------------------------------------------------------------At least one Critical Error on ELLIPTA 2 (3%)Zero Critical Errors on ELLIPTA 77 (97%)



Number of subjects with discordant results [1] 11 (14%)At least one Critical Error on ELLIPTA and Zero Critical Errors on Turbuhaler + HandiHaler [2] 1 (9%)At least one Critical Error on Turbuhaler + HandiHaler and Zero Critical Errors on ELLIPTA [2] 10 (91%)


131

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerTimepoint: After 2nd Instruction from HCP Total (N=79)------------------------------------------------------------------------------------------------------------At least one Critical Error on ELLIPTA 0Zero Critical Errors on ELLIPTA 79 (100%)


At least one Critical Error on ELLIPTA and Turbuhaler + HandiHaler 0Zero Critical Errors on ELLIPTA or Turbuhaler + HandiHaler 79 (100%)



132

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 12Population: Intent-to-Treat Table 2.08 Summary of Critical Errors by Country

Sub Study 1: ELLIPTA vs DISKUS + HandiHalerCountry: NetherlandsTimepoint: After Reading Leaflet Total (N=62)------------------------------------------------------------------------------------------------------------At least one Critical Error on ELLIPTA 3 (5%)Zero Critical Errors on ELLIPTA 59 (95%)





133

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 1: ELLIPTA vs DISKUS + HandiHalerCountry: United KingdomTimepoint: After Reading Leaflet Total (N=18)------------------------------------------------------------------------------------------------------------At least one Critical Error on ELLIPTA 4 (22%)Zero Critical Errors on ELLIPTA 14 (78%)





134

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 1: ELLIPTA vs DISKUS + HandiHalerCountry: NetherlandsTimepoint: After 1st Instruction from HCP Total (N=62)------------------------------------------------------------------------------------------------------------At least one Critical Error on ELLIPTA 0Zero Critical Errors on ELLIPTA 62 (100%)





135

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 1: ELLIPTA vs DISKUS + HandiHalerCountry: United KingdomTimepoint: After 1st Instruction from HCP Total (N=18)------------------------------------------------------------------------------------------------------------At least one Critical Error on ELLIPTA 1 (6%)Zero Critical Errors on ELLIPTA 17 (94%)





136

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 1: ELLIPTA vs DISKUS + HandiHalerCountry: NetherlandsTimepoint: After 2nd Instruction from HCP Total (N=62)------------------------------------------------------------------------------------------------------------At least one Critical Error on ELLIPTA 0Zero Critical Errors on ELLIPTA 62 (100%)





137

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 1: ELLIPTA vs DISKUS + HandiHalerCountry: United KingdomTimepoint: After 2nd Instruction from HCP Total (N=18)------------------------------------------------------------------------------------------------------------At least one Critical Error on ELLIPTA 0Zero Critical Errors on ELLIPTA 18 (100%)





138

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerCountry: NetherlandsTimepoint: After Reading Leaflet Total (N=60)------------------------------------------------------------------------------------------------------------At least one Critical Error on ELLIPTA 5 (8%)Zero Critical Errors on ELLIPTA 55 (92%)





139

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerCountry: United KingdomTimepoint: After Reading Leaflet Total (N=19)------------------------------------------------------------------------------------------------------------At least one Critical Error on ELLIPTA 2 (11%)Zero Critical Errors on ELLIPTA 17 (89%)





140

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerCountry: NetherlandsTimepoint: After 1st Instruction from HCP Total (N=60)------------------------------------------------------------------------------------------------------------At least one Critical Error on ELLIPTA 1 (2%)Zero Critical Errors on ELLIPTA 59 (98%)



Number of subjects with discordant results [1] 8 (13%)At least one Critical Error on ELLIPTA and Zero Critical Errors on Turbuhaler + HandiHaler [2] 1 (13%)At least one Critical Error on Turbuhaler + HandiHaler and Zero Critical Errors on ELLIPTA [2] 7 (88%)


141

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerCountry: United KingdomTimepoint: After 1st Instruction from HCP Total (N=19)------------------------------------------------------------------------------------------------------------At least one Critical Error on ELLIPTA 1 (5%)Zero Critical Errors on ELLIPTA 18 (95%)





142

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerCountry: NetherlandsTimepoint: After 2nd Instruction from HCP Total (N=60)------------------------------------------------------------------------------------------------------------At least one Critical Error on ELLIPTA 0Zero Critical Errors on ELLIPTA 60 (100%)

At least one Critical Error on Turbuhaler + HandiHaler 0Zero Critical Errors on Turbuhaler + HandiHaler 60 (100%)


Number of subjects with discordant results [1] 0At least one Critical Error on ELLIPTA and Zero Critical Errors on Turbuhaler + HandiHaler [2] 0At least one Critical Error on Turbuhaler + HandiHaler and Zero Critical Errors on ELLIPTA [2] 0


143

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerCountry: United KingdomTimepoint: After 2nd Instruction from HCP Total (N=19)------------------------------------------------------------------------------------------------------------At least one Critical Error on ELLIPTA 0Zero Critical Errors on ELLIPTA 19 (100%)





144

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 6Population: Intent-to-Treat Table 2.09 Analysis of Percentage of Subjects making at least one Critical Error

Sub Study 1: ELLIPTA vs DISKUS + HandiHalerAssessment: After Reading Leaflet ELLIPTA DISKUS + HandiHaler (N=80) (N=80) ------------------------------------------------------------------------------ At least one Critical Error 7 (9%) 60 (75%) Zero Critical Errors 73 (91%) 20 (25%)

DISKUS + HandiHaler vs ELLIPTA Odds Ratio 29.114 95% CI (11.047,Inf) p-value <0.001

Note: Odds ratio, 95% CI and P-value obtained from a stratified exact logistic model.Subject is included in the model as fixed strata, treatment option was included in the exact statement andperiod included as fixed effects.Subjects who made no errors after reading PIL were not given instruction by HCP and so are counted as havingzero critical errors after 1st and 2nd instruction from HCP

145

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 1: ELLIPTA vs DISKUS + HandiHalerAssessment: After 1st Instruction from HCP ELLIPTA DISKUS + HandiHaler (N=80) (N=80) ------------------------------------------------------------------------------ At least one Critical Error 1 (1%) 9 (11%) Zero Critical Errors 79 (99%) 71 (89%)

DISKUS + HandiHaler vs ELLIPTA Odds Ratio 4.248 95% CI (1.416,Inf) p-value 0.029


146

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 1: ELLIPTA vs DISKUS + HandiHalerAssessment: After 2nd Instruction from HCP ELLIPTA DISKUS + HandiHaler (N=80) (N=80) ------------------------------------------------------------------------------ At least one Critical Error 0 5 (6%) Zero Critical Errors 80 (100%) 75 (94%)



147

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerAssessment: After Reading Leaflet ELLIPTA Turbuhaler + HandiHaler (N=79) (N=79) ------------------------------------------------------------------------------ At least one Critical Error 7 (9%) 58 (73%) Zero Critical Errors 72 (91%) 21 (27%)

Turbuhaler + HandiHaler vs ELLIPTA Odds Ratio 27.744 95% CI (10.512,Inf) p-value <0.001


148

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerAssessment: After 1st Instruction from HCP ELLIPTA Turbuhaler + HandiHaler (N=79) (N=79) ------------------------------------------------------------------------------ At least one Critical Error 2 (3%) 11 (14%) Zero Critical Errors 77 (97%) 68 (86%)

Turbuhaler + HandiHaler vs ELLIPTA Odds Ratio 3.855 95% CI (1.394,Inf) p-value 0.026


149

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerAssessment: After 2nd Instruction from HCP ELLIPTA Turbuhaler + HandiHaler (N=79) (N=79) ------------------------------------------------------------------------------ At least one Critical Error 0 4 (5%) Zero Critical Errors 79 (100%) 75 (95%)



150

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 2Population: Intent-to-Treat Table 2.10 Sensitivity Analysis of Percentage of Subjects making at least one Error after Reading the PIL(s)

Sub Study 1: ELLIPTA vs DISKUS + HandiHalerTimepoint: After Reading Leaflet Result------------------------------------------------------------------------------------------------------------Critical Error Number of subjects with discordant results [1] 53 At least one Critical Error on ELLIPTA and Zero Critical Errors on DISKUS + HandiHaler 0 At least one Critical Error on DISKUS + HandiHaler and Zero Critical Errors on ELLIPTA 53 (100%) p-value [2] <0.001

Overall Error Number of subjects with discordant results [1] 45 At least one Overall Error on ELLIPTA and Zero Overall Errors on DISKUS + HandiHaler 0 At least one Overall Error on DISKUS + HandiHaler and Zero Overall Errors on ELLIPTA 45 (100%) p-value [2] <0.001

[1] Defined as an error in one device and not the other.[2] P-value is from Cochran-Mantel-Haenszel test.Note: Subjects are only included in the analysis if they have discordant data.Subjects who made no errors after reading PIL were not given instruction by HCP and so are counted as havingzero critical/overall errors after 1st and 2nd instruction from HCP

151

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 2 of 2Population: Intent-to-Treat Table 2.10 Sensitivity Analysis of Percentage of Subjects making at least one Error after Reading the PIL(s)

Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerTimepoint: After Reading Leaflet Result------------------------------------------------------------------------------------------------------------Critical Error Number of subjects with discordant results [1] 51 At least one Critical Error on ELLIPTA and Zero Critical Errors on Turbuhaler + HandiHaler 0 At least one Critical Error on Turbuhaler + HandiHaler and Zero Critical Errors on ELLIPTA 51 (100%) p-value [2] <0.001

Overall Error Number of subjects with discordant results [1] 48 At least one Overall Error on ELLIPTA and Zero Overall Errors on Turbuhaler + HandiHaler 1 (2%) At least one Overall Error on Turbuhaler + HandiHaler and Zero Overall Errors on ELLIPTA 47 (98%) p-value [2] <0.001

[1] Defined as an error in one device and not the other.[2] P-value is from Cochran-Mantel-Haenszel test.Note: Subjects are only included in the analysis if they have discordant data.Subjects who made no errors after reading PIL were not given instruction by HCP and so are counted as havingzero critical/overall errors after 1st and 2nd instruction from HCP

152

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 6Population: Intent-to-Treat Table 2.11 Analysis of Percentage of Subjects making at least one Overall Error

Sub Study 1: ELLIPTA vs DISKUS + HandiHalerAssessment: After Reading Leaflet ELLIPTA DISKUS + HandiHaler (N=80) (N=80) ------------------------------------------------------------------------------ At least one Overall Error 19 (24%) 64 (80%) Zero Overall Errors 61 (76%) 16 (20%)

DISKUS + HandiHaler vs ELLIPTA Odds Ratio 24.539 95% CI (9.268,Inf) p-value <0.001

Note: Odds ratio, 95% CI and P-value obtained from a stratified exact logistic model. These statistics areonly presented when the model has successfully converged.Subject is included in the model as fixed strata, treatment option was included in the exact statement andperiod included as fixed effects.Subjects who made no errors after reading PIL were not given instruction by HCP and so are counted as havingzero overall errors after 1st and 2nd instruction from HCP

153

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 1: ELLIPTA vs DISKUS + HandiHalerAssessment: After 1st Instruction from HCP ELLIPTA DISKUS + HandiHaler (N=80) (N=80) ------------------------------------------------------------------------------ At least one Overall Error 4 (5%) 12 (15%) Zero Overall Errors 76 (95%) 68 (85%)



154

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 1: ELLIPTA vs DISKUS + HandiHalerAssessment: After 2nd Instruction from HCP ELLIPTA DISKUS + HandiHaler (N=80) (N=80) ------------------------------------------------------------------------------ At least one Overall Error 3 (4%) 6 (8%) Zero Overall Errors 77 (96%) 74 (93%)

DISKUS + HandiHaler vs ELLIPTA Odds Ratio NA 95% CI NA p-value NA


155

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerAssessment: After Reading Leaflet ELLIPTA Turbuhaler + HandiHaler (N=79) (N=79) ------------------------------------------------------------------------------ At least one Overall Error 17 (22%) 63 (80%) Zero Overall Errors 62 (78%) 16 (20%)

Turbuhaler + HandiHaler vs ELLIPTA Odds Ratio 17.974 95% CI (7.239,Inf) p-value <0.001


156

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerAssessment: After 1st Instruction from HCP ELLIPTA Turbuhaler + HandiHaler (N=79) (N=79) ------------------------------------------------------------------------------ At least one Overall Error 4 (5%) 16 (20%) Zero Overall Errors 75 (95%) 63 (80%)



157

CONFIDENTIAL 2017N324761_00206215

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerAssessment: After 2nd Instruction from HCP ELLIPTA Turbuhaler + HandiHaler (N=79) (N=79) ------------------------------------------------------------------------------ At least one Overall Error 1 (1%) 5 (6%) Zero Overall Errors 78 (99%) 74 (94%)



158

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 6Population: Intent-to-Treat Table 2.12 Summary and Analysis of Time (Minutes) Taken to Demonstrate Correct Inhaler Use

Sub Study 1: ELLIPTA vs DISKUS + HandiHalerAssessment: Time taken to read the PIL and Demonstrate Correct Use ELLIPTA DISKUS + HandiHaler (N=80) (N=80) ---------------------------------------------------------------------------------------------------------- n 61 16 Mean 2.76 7.67 SD 1.582 5.365 Median 2.38 5.50 Min. 0.8 1.5 Max. 7.2 19.2

Number of Subjects who demonstrated correct use 61 (76%) 16 (20%) Number of Subjects who failed to demonstrate correct use (censored) 19 (24%) 64 (80%) Median Time to demonstrate correct inhaler use (minutes) [1] 2.88 NA

[1] The Median Time to demonstrate correct inhaler use (minutes) is taken from the Kaplan-Meieranalysis. If more than 50% of the data is censored then the median is not applicable.Note: For the DISKUS+HandiHaler and Turbuhaler+HandiHaler groups, if a subject demonstrated correctuse on only one of the two devices, the time to demonstrate correct use is censored at the total ofthe reading and/or instruction time for both devices.Subject did not demonstrate correct use for the ELLIPTA device after reading the PIL, but was able toafter receiving the first HCP instruction. The time taken to demonstrate correct use at the HCPassessment was not recorded, therefore this subject's ELLIPTA data could only be included in the time takento demonstrate correct use after reading the PIL assessment.

159

CONFIDENTIAL 2017N324761_00206215

PPD

PPD


Sub Study 1: ELLIPTA vs DISKUS + HandiHalerAssessment: Time for HCP Instruction and to Demonstrate Correct Use ELLIPTA DISKUS + HandiHaler (N=80) (N=80) ---------------------------------------------------------------------------------------------------------- n 77 74 Mean 0.26 3.53 SD 0.630 3.629 Median 0.00 2.69 Min. 0.0 0.0 Max. 3.3 16.5

Number of Subjects who demonstrated correct use 77 (96%) 74 (93%) Number of Subjects who failed to demonstrate correct use (censored) 3 (4%) 6 (8%) Median Time to demonstrate correct inhaler use (minutes) [1] 0.00 2.89


160

CONFIDENTIAL 2017N324761_00206215

PPD

PPD


Sub Study 1: ELLIPTA vs DISKUS + HandiHalerAssessment: Total Time to demonstrate correct use ELLIPTA DISKUS + HandiHaler (N=80) (N=80) ---------------------------------------------------------------------------------------------------------- n 77 74 Mean 3.08 12.06 SD 2.088 7.562 Median 2.63 9.95 Min. 0.8 1.5 Max. 12.9 42.4



161

CONFIDENTIAL 2017N324761_00206215

PPD

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerAssessment: Time taken to read the PIL and Demonstrate Correct Use ELLIPTA Turbuhaler + HandiHaler (N=79) (N=79) ---------------------------------------------------------------------------------------------------------- n 62 16 Mean 2.58 7.46 SD 1.544 3.362 Median 2.06 7.66 Min. 0.7 3.3 Max. 8.6 15.1

Number of Subjects who demonstrated correct use 62 (78%) 16 (20%) Number of Subjects who failed to demonstrate correct use (censored) 17 (22%) 63 (80%) Median Time to demonstrate correct inhaler use (minutes) [1] 2.78 NA


162

CONFIDENTIAL 2017N324761_00206215

PPD

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerAssessment: Time for HCP Instruction and to Demonstrate Correct Use ELLIPTA Turbuhaler + HandiHaler (N=79) (N=79) ---------------------------------------------------------------------------------------------------------- n 77 74 Mean 0.36 3.66 SD 0.964 3.524 Median 0.00 2.83 Min. 0.0 0.0 Max. 6.3 14.3



163

CONFIDENTIAL 2017N324761_00206215

PPD

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerAssessment: Total Time to demonstrate correct use ELLIPTA Turbuhaler + HandiHaler (N=79) (N=79) ---------------------------------------------------------------------------------------------------------- n 77 74 Mean 3.21 13.31 SD 3.088 9.482 Median 2.37 11.11 Min. 0.7 3.3 Max. 25.2 55.8



164

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

Protocol: 206215 Page 1 of 2Population: Intent-to-Treat Table 2.13 Summary of Treatment Preference

Sub Study 1: ELLIPTA vs DISKUS + HandiHaler Total (N=80) P-value ------------------------------------------------------------------------------- Which treatment do you prefer based on the number of steps needed to take your COPD medication? n 80 <0.001 ELLIPTA 71 (89%) DISKUS + HandiHaler 6 (8%) No Preference 3 (4%)

Which treatment do you prefer for taking your COPD medication? n 80 <0.001 ELLIPTA 65 (81%) DISKUS + HandiHaler 7 (9%) No Preference 8 (10%)

Note: The p-value is from the Cochran-Mantel-Haenszel Test.

165

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 2 of 2Population: Intent-to-Treat Table 2.13 Summary of Treatment Preference

Sub Study 2: ELLIPTA vs Turbuhaler + HandiHaler Total (N=79) P-value ------------------------------------------------------------------------------- Which treatment do you prefer based on the number of steps needed to take your COPD medication? n 79 <0.001 ELLIPTA 72 (91%) Turbuhaler + HandiHaler 4 (5%) No Preference 3 (4%)

Which treatment do you prefer for taking your COPD medication? n 79 <0.001 ELLIPTA 66 (84%) Turbuhaler + HandiHaler 3 (4%) No Preference 10 (13%)

Note: The p-value is from the Cochran-Mantel-Haenszel Test.

166

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 2Population: Intent-to-Treat Table 2.14 Summary and Analysis of Number of Instructions from the HCP

Sub Study 1: ELLIPTA vs DISKUS + HandiHaler ELLIPTA DISKUS + HandiHaler Difference (ELLIPTA minus (N=80) (N=80) DISKUS + HandiHaler) ----------------------------------------------------------------------------------------------------- n 80 80 80

Median (95% CI) 0.0 (0.0,0.0) 1.0 (1.0,1.0) -1.0 (-1.0,0.0) Min. 0 0 -3 Max. 3 3 1

Number of Instructions 0 61 (76%) 16 (20%) 1 15 (19%) 52 (65%) 2 1 (1%) 6 (8%) Failed to demonstrate correct use 3 (4%) 6 (8%)

DISKUS + HandiHaler vs ELLIPTA p-value <0.001

Note: P-value has come from the Wilcoxon signed rank test.This analysis did not take into account sequence of treatment options.For the DISKUS+HandiHaler and Turbuhaler+HandiHaler groups, if a subject failed to demonstrate correctuse on one device and not the other, the subject would be counted as having failed to demonstrate correctuse.

167

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 2 of 2Population: Intent-to-Treat Table 2.14 Summary and Analysis of Number of Instructions from the HCP

Sub Study 2: ELLIPTA vs Turbuhaler + HandiHaler Turbuhaler + ELLIPTA HandiHaler Difference (ELLIPTA minus (N=79) (N=79) Turbuhaler + HandiHaler) ---------------------------------------------------------------------------------------------------- n 79 79 79

Median (95% CI) 0.0 (0.0,0.0) 1.0 (1.0,1.0) -1.0 (-1.0,-1.0) Min. 0 0 -3 Max. 3 3 1

Number of Instructions 0 62 (78%) 16 (20%) 1 13 (16%) 47 (59%) 2 3 (4%) 11 (14%) Failed to demonstrate correct use 1 (1%) 5 (6%)

Turbuhaler + HandiHaler vs ELLIPTA p-value <0.001

Note: P-value has come from the Wilcoxon signed rank test.This analysis did not take into account sequence of treatment options.For the DISKUS+HandiHaler and Turbuhaler+HandiHaler groups, if a subject failed to demonstrate correctuse on one device and not the other, the subject would be counted as having failed to demonstrate correctuse.

168

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 2Population: Intent-to-Treat Table 2.15 Probability of Total Time (Minutes) Taken to Demonstrate Correct Inhaler Use by 5 Minute Intervals

Sub Study 1: ELLIPTA vs DISKUS + HandiHaler ELLIPTA DISKUS + HandiHaler (N=80) (N=80) ---------------------------------------------------------------------------------------------------------- Probability of demonstrating correct use in 5 minutes or less (%) 86 10 Probability of demonstrating correct use in 10 minutes or less (%) 97 46 Probability of demonstrating correct use in 15 minutes or less (%) 100 67 Probability of demonstrating correct use in 20 minutes or less (%) 100 82

Note: Probabilities are from the Kaplan-Meier analysis and account for censoring in the data for subjectswho were unable to demonstrate correct use after the second HCP instruction.For the DISKUS+HandiHaler and Turbuhaler+HandiHaler groups, if a subject demonstrated correctuse on only one of the two devices, the time to demonstrate correct use is censored at the total ofthe reading and/or instruction time for both devices.Subject (Sub Study 2) demonstrated correct use for the ELLIPTA device following the first HCPinstruction however the time taken to demonstrate correct use was not recorded and therefore this subjectis not included for the ELLIPTA arm of Sub Study 2.

169

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

Protocol: 206215 Page 2 of 2Population: Intent-to-Treat Table 2.15 Probability of Total Time (Minutes) Taken to Demonstrate Correct Inhaler Use by 5 Minute Intervals

Sub Study 2: ELLIPTA vs Turbuhaler + HandiHaler ELLIPTA Turbuhaler + HandiHaler (N=79) (N=79) ---------------------------------------------------------------------------------------------------------- Probability of demonstrating correct use in 5 minutes or less (%) 86 9 Probability of demonstrating correct use in 10 minutes or less (%) 97 43 Probability of demonstrating correct use in 15 minutes or less (%) 97 69 Probability of demonstrating correct use in 20 minutes or less (%) 97 82

Note: Probabilities are from the Kaplan-Meier analysis and account for censoring in the data for subjectswho were unable to demonstrate correct use after the second HCP instruction.For the DISKUS+HandiHaler and Turbuhaler+HandiHaler groups, if a subject demonstrated correctuse on only one of the two devices, the time to demonstrate correct use is censored at the total ofthe reading and/or instruction time for both devices.Subject (Sub Study 2) demonstrated correct use for the ELLIPTA device following the first HCPinstruction however the time taken to demonstrate correct use was not recorded and therefore this subjectis not included for the ELLIPTA arm of Sub Study 2.

170

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

Protocol: 206215 Page 1 of 1Population: Intent-to-Treat Table 3.01 Summary of Adverse Events

Sub Sub System Organ Class Study 1 Study 2 Preferred Term (N=80) (N=79) -------------------------------------------------- ANY EVENT 1 (1%) 0

Injury, poisoning and procedural complications Any event 1 (1%) 0 Laceration 1 (1%) 0

171

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 1Population: Intent-to-Treat Table 3.02 Summary of On-Study Adverse Events

Sub Sub System Organ Class Study 1 Study 2 Preferred Term (N=80) (N=79) -------------------------------------------------- ANY EVENT 1 (1%) 0

Injury, poisoning and procedural complications Any event 1 (1%) 0 Laceration 1 (1%) 0

172

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 1Population: Intent-to-Treat Table 3.03 Summary of Post-Study Adverse Events

No data to report

173

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 1Population: Intent-to-Treat Table 3.04 Summary of Serious Adverse Events

No data to report

174

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 1Population: Intent-to-Treat Table 3.05 Summary of On-Treatment Adverse Events Leading to Discontinuation of Study Treatment or Withdrawal from the Study

No data to report

175

CONFIDENTIAL 2017N324761_00206215

PPD

Protocol: 206215 Page 1 of 1Population: Intent-to-Treat Table 3.06 Summary of Serious Adverse Events by System Organ Class and Preferred Term (Number of Subjects and Occurrences)

No data to report

176

CONFIDENTIAL 2017N324761_00206215

PPD

CONFIDENTIAL 2017N324761_00206215

Synopsis

Name of company: GlaxoSmithKline Research & Development Limited

Name of finished product:ELLIPTA, HandiHaler, Turbuhaler and DISKUS

Name of active substance:Not applicable, placebo versions were used for this study.

Study Number: 206215

Title: A randomized, open-label, cross-over, placebo-device study investigating critical and over all errors, training/teaching time, and preference attributes of the ELLIPTA dry powder Inhaler (DPI) as compared to HandiHaler DPI used in combination with either DISKUS DPI or Turbuhaler DPI, in adult patients with Chronic Obstructive Pulmonary Disease (COPD).

Investigator(s): Multi-centre study.

Study centre(s): Five centres in 2 countries (the United Kingdom and the Netherlands).

Publication(s): None at the time of this report

Study period: 30-Dec-2016 to 19-Jun-2017

Phase of development: IV

Objectives:

Primary:

To compare the number of critical errors made by COPD patients, after a subject had read the respective patient information leaflet(s) (PIL), for each treatment option tested.

Secondary:

To compare the number of critical errors made by COPD patients after instruction from the healthcare professional (HCP) for each treatment option tested;

To compare the number of overall (critical and non-critical) errors made by COPD patients, after a subject had read the PIL(s) or after instruction from the HCP for each treatment option tested;

To compare the number of instructions (maximum of 2) from a HCP which was needed to demonstrate correct inhaler use;

To compare the training/teaching time required to demonstrate correct inhaler use;

Preference attributes for each treatment option tested.

1

CONFIDENTIAL 2017N324761_00206215

Methodology: This study was a multi-centre, randomised, open-label, placebo-device crossover study with a 2x2 complete block design, comprised of 2 sub-studies:

Sub-study 1: Compared ELLIPTA (Treatment Option 1) to DISKUS plusHandiHaler combination (Treatment Option 2);

Sub-study 2: Compared ELLIPTA (Treatment Option 1) to Turbuhaler plusHandiHaler combination (Treatment Option 3).

Each sub-study was run independently and in parallel to the other sub-study. There was no active treatment and subjects continued to take their own prescribed COPD medication for the duration of the study.

The study had 2 visits (Visit 0 [V0] and Visit 1 [V1]); both visits could be completed on the same day.

Following completion of all randomised device assessments, subjects completed 1 of 4 possible inhaler preference questionnaires assigned at randomisation (Version 1 or 2 [Sub-study 1] or Version 3 or 4 [Sub-study 2]). The 2 preference questionnaires used for each sub-study asked the same 2 questions, but the ordering of the inhalers in the responses was different in order to reduce potential bias.

Number of subjects: Sufficient patients with COPD were screened to ensure approximately 160 participants were randomised, such that approximately 144 evaluable participants completed the study, with approximately 72 participants completing each sub-study. A total of 160 subjects were screened and were randomised; 80 to Sub-study 1 and 80 to Sub-study 2. All subjects, except 1, completed the study (1 subject in Sub-study 2 withdrew from the study prior to device assessment).

Diagnosis and main criteria for inclusion: Males or non-pregnant females who were current or former smokers, ≥40 years of age at V1, with a diagnosis of COPD and a documented history of COPD in accordance with the definition by the European Respiratory Society. Subjects were to be receiving maintenance therapy with a fixed dose combination of a long-acting beta2-agonist (LABA) and inhaled corticosteroid (ICS). Subjects could also be receiving a long-acting muscarinic antagonist. Subjects were required to continue using their prescribed COPD maintenance inhaler therapy throughout the study, and use short acting beta-adrenergic agonist and/or short acting muscarinic antagonist rescue medications as needed. Subjects were required to have been on current maintenance ICS/LABA COPD treatment for at least 4 weeks prior to V0 and evaluated as unlikely to change treatment within 4 weeks of V1.

2

CONFIDENTIAL 2017N324761_00206215

Treatment administration: Subjects were randomised to receive placebo treatments as noted below, dependent on which sub-study they were included in:

Sub-study 1 Treatment Sequence

Sequence Period 1 Period 2 Preference QuestionnaireA ELLIPTA DISKUS + HandiHaler 1B DISKUS + HandiHaler ELLIPTA 2C ELLIPTA DISKUS + HandiHaler 2D DISKUS + HandiHaler ELLIPTA 1Sub-study 2 Treatment SequenceSequence Period 1 Period 2 Preference QuestionnaireE ELLIPTA Turbuhaler + HandiHaler 3F Turbuhaler + HandiHaler ELLIPTA 4G ELLIPTA Turbuhaler + HandiHaler 4H Turbuhaler + HandiHaler ELLIPTA 3

Batch Numbers of Investigational ProductsInvestigational Product Batch/Lot NumbersHandiHaler - Placebo YZ0005ELLIPTA - Placebo 162397455DISKUS - Placebo 6ZP5067Turbuhaler - Placebo HAAM

Criteria for evaluation: As an open-label, placebo-device study, efficacy endpoints were defined in the context of the assessment of errors in the use of the device(s) under evaluation.

The primary endpoint was to assess the percentage of subjects making at least 1 critical error after reading the PIL(s).

The secondary endpoints were to assess:

The percentage of subjects making at least 1 critical error after receiving the first, and second, instruction from a HCP;

The percentage of subjects making at least 1 overall error after reading the PIL(s), following the first, and second, instruction from a HCP;

The number of instructions (0, 1 or 2) required from the HCP to demonstrate correct inhaler use;

The amount of teacher/training time taken to: 1) read the PIL and demonstrate correct inhaler use (referred to as T1); 2) be given instruction on use of the inhaler by the HCP (up to 2 times) and to demonstrate correct inhaler use (referred to as T2);and 3) the total amount of time taken to demonstrate correct inhaler use (T1+T2);

Treatment preference based on responses to the preference questionnaire, which considered the number of steps required to take the COPD medication and overall treatment preference.

3

CONFIDENTIAL 2017N324761_00206215

There were no mandated safety assessments, other than monitoring for adverse events (AEs) and serious adverse events.

Statistical methods: The primary purpose of this superiority study was to assess the number of critical errors made by subjects with COPD after they had read the PIL(s) for each treatment option tested. There were 2 sub-studies, both analysed separately. There was no overlap of subjects; the sub-studies were considered independent, hence there was no adjustment for multiplicity.

The sample size calculation for each sub-study was based on the primary endpoint. Based on the results from GlaxoSmithKline studies 201301 and 201330, a range of critical error rates (at least 1 critical error using each device after reading the PIL) for each of the treatment options (ELLIPTA, DISKUS plus HandiHaler and Turbuhaler plusHandiHaler) were explored. Using 10,000 simulations, a total of 72 subjects in each sub-study would provide at least 90% power to show a statistically significant difference between the critical error rate of each of the paired treatment options. Conditional logistic regression and a 2-sided 5% significance level were used as the analysis method in the simulations. No withdrawals were expected, however, up to 80 subjects were planned to be randomised to each sub-study to ensure 72 subjects were evaluable in each sub-study.

The primary endpoint of percentage of subjects making at least 1 critical error after reading the PIL(s) was analysed using exact logistic regression with subject as fixed strata, and treatment option and period as fixed effects. The odds ratio, 95% confidence interval and p-values were presented for the comparison between treatment options and were based on a 2-sided hypothesis testing approach of superiority.

The primary population used for analyses of efficacy and safety measures was the Intent-to-treat (ITT) Population, which comprised all randomised subjects who made at least 1 critical error assessment from 1 treatment option device. A subject who was recorded as a screen or run-in failure and also randomised, was considered to be randomised in error, provided they had not performed any error assessments.

Summary:

Demographics: As demographic characteristics were similar across the 2 sub-studies, data has been presented for the total population in text. The majority of subjects were White (74%) and there was a similar proportion of males (52%) and females (48%) in the study overall. The mean age was 65.0 years. Key demographic information for the ITT Population are summarised in the following table.

4

CONFIDENTIAL 2017N324761_00206215

Summary of Demographic Characteristics (ITT Population)Sub-study 1

(N=80)Sub-study 2

(N=79) Total

(N=159)Age (years)

Mean (SD) 64.3 (8.69) 65.7 (8.49) 65.0 (8.60)Median (Min – Max) 65.5 (48 – 82) 66.0 (48 – 92) 66.0 (48 – 92)

Sex, n (%)Female 39 (49) 37 (47) 76 (48)Male 41 (51) 42 (53) 83 (52)

Race, n (%)African American/African Heritage 1 (1) 0 1 (<1)American Indian or Alaskan Native 0 0 0Asian 21 (26) 20 (25) 41 (26)

Central/South Asian Heritage 20 (25) 20 (25) 40 (25)Japanese/East AsianHeritage/South East AsianHeritage

1 (1) 0 1 (<1)


0 0 0

White 58 (73) 59 (75) 117 (74)Arabic/North African Heritage 5 (6) 4 (5) 9 (6)White/Caucasian/EuropeanHeritage

53 (66) 55 (70) 108 (68)

Ethnicity, n (%)Hispanic/Latino 0 0 0Not Hispanic/Latino 80 (100) 79 (100) 159 (100)

Height (cm)n 80 78 158Mean (SD) 168.4 (9.86) 169.6 (9.64) 169.0 (9.74)Median (Min – Max) 169.5 (147 – 191) 170.0 (150 – 199) 170.0 (147 – 199)

Weight (kg)n 80 78 158Mean (SD) 75.95 (18.098) 78.10 (14.883) 77.01 (16.571)Median (Min – Max) 73.90 (43.6 – 122.4) 75.85 (50.6 – 125.4) 74.35 (43.6 – 125.4)

Body Mass Index (kg/m2)n 80 78 158Mean (SD) 26.69 (5.607) 27.12 (4.568) 26.90 (5.109)Median (Min – Max) 26.08 (16.5 – 43.4) 26.21 (19.1 – 40.5) 26.21 (16.5 – 43.4)

Note: One subject's vital signs were not collected at V1.Abbreviations: ITT = intent-to-treat; Max = maximum; Min = minimum; N = sample size; n = subset of sample size; SD = standard deviation; V = visit.

Efficacy:

All percentages were calculated based on the total number of subjects in the respectivesub-study (Sub-study 1, N=80; Sub-study 2, N=79).

Critical inhaler use errors: Both Sub-study 1 and Sub-study 2 met the primary objective by demonstrating that, after reading the PIL, fewer subjects demonstrated 1 or more critical errors when using ELLIPTA (9% in both sub-studies) compared with

5

CONFIDENTIAL 2017N324761_00206215

DISKUS plus HandiHaler (75%, Sub-study 1) or Turbuhaler plus HandiHaler (73%, Sub-study 2), and that the differences between ELLIPTA (in Sub-studies 1 and 2) and each of the inhaler combinations used in the respective sub-studies, were statistically significant (p<0.001). In addition, following the first HCP instruction, for those subjects who made an error after reading the PIL, fewer critical errors were made by subjects using ELLIPTA (1% and 3%, Sub-studies 1 and 2, respectively), compared with subjects using DISKUS plus HandiHaler (11%) or Turbuhaler plus HandiHaler (14%); these differences were also statistically significant (p<0.05). Few subjects from either sub-study required a second HCP instruction (required by subjects who made an error following the first HCP instruction). Of these subjects, none made critical errors in either sub-study using ELLIPTA; however, critical errors were made by subjects using DISKUS plus HandiHaler (6%) and using Turbuhaler plus HandiHaler (5%); these differences were not statistically significant.

On the error checklists, 4 types of critical error were possible with ELLIPTA, 5 with DISKUS, 6 with Turbuhaler and 7 with HandiHaler. The total number of critical errors recorded after reading the PIL was lower with ELLIPTA (8 and 13 errors, Sub-studies 1 and 2, respectively) compared with each of the other inhaler combinations (139 errors with DISKUS plus HandiHaler; 134 errors with Turbuhaler plus HandiHaler).

Overall inhaler use errors: The results for overall errors followed a similar trend to critical errors. After reading the PIL, fewer subjects demonstrated 1 or more overall errors when using ELLIPTA (24% and 22% in Sub-studies 1 and 2, respectively) compared with DISKUS plus HandiHaler (80%, Sub-study 1) or Turbuhaler plus HandiHaler (80%, Sub-study 2); these differences, between ELLIPTA (in Sub-studies 1 and 2) and each of the inhaler combinations used in the respective sub-studies, were statistically significant (p<0.001). In addition, following the first HCP instruction, for those subjects who made an error after reading the PIL, fewer overall errors were made by subjects using ELLIPTA (5% in both sub-studies), compared with subjects using DISKUS plus HandiHaler (15%) or Turbuhaler plus HandiHaler (20%);only the difference between ELLIPTA and Turbuhaler plus HandiHaler was statistically significant (p<0.05). As for critical errors, few subjects required a second HCP instruction (required by subjects who made an error following the first HCP instruction). Of these subjects, fewer overall errors were made by subjects using ELLIPTA (4% and 1% in Sub-studies 1 and 2, respectively), compared with subjects using DISKUS plus HandiHaler (8%) or Turbuhaler plus HandiHaler (6%); these differences were not statistically significant.

Number of instructions required from HCP to demonstrate correct inhaler use: The data demonstrate that, after reading the PIL, the majority of subjects made no errors using ELLIPTA (76% and 78% in Sub-studies 1 and 2, respectively) and thus only 24% and 22% of subjects (Sub-study 1 and 2, respectively) required further instruction (in additionto reading the PIL) from the HCP. The converse was observed for the comparator inhaler combinations used in each sub-study, where the majority of subjects made at least 1 overall error and therefore required further HCP instruction (Sub-study 1 [DISKUS plus HandiHaler] and Sub-study 2 [Turbuhaler plus HandiHaler], 80% of subjects each).

6

CONFIDENTIAL 2017N324761_00206215

Of the subjects who required instruction, the majority only required 1 instruction from the HCP to demonstrate correct inhaler use. In each sub-study, fewer subjects required 2 HCP instructions for ELLIPTA (4 subjects in each sub-study) compared with DISKUS plus HandiHaler (12 subjects) and with Turbuhaler plus HandiHaler (16 subjects). Therewas a statistically significant difference in the number of HCP instructions required to demonstrate correct use of the inhaler(s) between ELLIPTA and both of the inhaler combinations (p<0.001 for both sub-studies). Three subjects in Sub-study 1 and a singlesubject in Sub-study 2 failed to demonstrate correct inhaler use with ELLIPTA; 6 subjects failed to demonstrate correct use of DISKUS plus HandiHaler (Sub-study 1) and 5 subjects failed to demonstrate correct use of Turbuhaler plus HandiHaler (Sub-study 2).

Time taken to use inhaler device: In both sub-studies, the median time from when the subject started to read the PIL until correct device use was demonstrated without any HCP support (T1) could not be calculated for DISKUS plus HandiHaler and Turbuhaler plus HandiHaler, as more than 50% of data was censored for these inhalers. T1, T2 (the time from when the HCP started to instruct the subject until correct use was demonstrated) and T1+T2 (the time from when the subject started to read the PIL(s) until correct use was demonstrated [up to a maximum of 3 attempts, once after reading the PIL(s) and twice following instruction by the HCP]) were censored for subjects who failed to demonstrate correct use at the end of each time period. For subjects who demonstrated correct inhaler use after reading the PIL, T2 was recorded as 0 minutes for the appropriate inhaler. Median T2 was shorter for subjects using ELLIPTA (0 minutes in both sub-studies) compared with DISKUS plus HandiHaler (2.89 minutes, Sub-study 1) and with Turbuhaler plus HandiHaler (3.12 minutes, Sub-study 2). Mediantotal time (T1+T2) was also shorter for subjects using ELLIPTA (2.68 and 2.58 minutes in Sub-studies 1 and 2, respectively) compared with DISKUS plus HandiHaler (10.57 minutes, Sub-study 1) and with Turbuhaler plus HandiHaler (11.30 minutes, Sub-study 2). In addition, for both Sub-study 1 and 2, 86% of subjects demonstrated correct use of the ELLIPTA device within the first 5 minutes compared with 10% of subjects using DISKUS plus HandiHaler (Sub-study 1) and 9% of subjects using Turbuhaler plus HandiHaler (Sub-study 2). In Sub-study 1, all subjects had demonstrated correct use of the ELLIPTA device within 15 minutes compared with 67% of subjects demonstrating the correct use of the DISKUS plus HandiHaler combination. In Sub-study 2, 97% of subjects had demonstrated correct use of the ELLIPTA device within 20 minutes compared with 82% of subjects demonstrating the correct use of the Turbuhaler plus HandiHaler combination.

Inhaler preference: The majority of subjects in both sub-studies preferred to take their COPD medication using ELLIPTA (81% and 84% in Sub-studies 1 and 2, respectively) compared with DISKUS plus HandiHaler (9%, Sub-study 1; 10% of subjects had no preference) and with Turbuhaler plus HandiHaler (4% Sub-study 2; 13% of subjects had no preference). Preference for ELLIPTA was also observed in both sub-studies based on the number of steps required to take the COPD medication; 89% and 91% of subjects (Sub-studies 1 and 2, respectively) preferred ELLIPTA compared with DISKUS plus HandiHaler (8%, Sub-study 1; 4% of subjects had no preference) and with Turbuhaler plus HandiHaler (5%, Sub-study 2; 4% of subjects had no preference). The differences between ELLIPTA and the comparator inhaler combinations for both preference

7

CONFIDENTIAL 2017N324761_00206215

questions in both sub-studies, were statistically significant (p<0.001, for both sub-studies).

Safety: One on-study treatment-related AE was reported: Subject had an AE of laceration (verbatim term, cuts on both thumbs) in Sub-study 1; the event resolved after 6 days. No AEs were reported in Sub-study 2.

Conclusions: In both sub-studies, after reading the PIL, a statistically significantly lower percentage of subjects with COPD made critical and overall errors with ELLIPTA compared with DISKUS plus HandiHaler or Turbuhaler plus HandiHaler. Overall, subjects using ELLIPTA required fewer HCP instructions, made fewer errors following each HCP instruction and took less time to demonstrate the correct use of the inhaler device, compared with either DISKUS plus HandiHaler or Turbuhaler plus HandiHaler. In addition, a statistically significantly higher percentage of subjects preferred the ELLIPTA device compared with the other inhaler combinations based on both preference questions; for taking their COPD medication, and based on the number of steps required to take their COPD medication.

Effective Date: 07-DEC-2017

8

PPD

2016N292801_01 CONFIDENTIALGlaxoSmithKline group of companies 206215

1

TITLE PAGE

Protocol Title: A randomized, open-label, cross-over, placebo-device study investigating critical and over all errors, training/teaching time, and preference attributes of the ELLIPTA dry powder Inhaler (DPI) as compared to HandiHaler DPI used in combination with either DISKUS DPI or Turbuhaler DPI, in adult patients with Chronic Obstructive Pulmonary Disease (COPD)

Protocol Number: 206215

Short Title: A clinical study assessing critical errors, training/teaching time, and preference attributes of the ELLIPTA dry powder inhaler, in comparison to combinations of dry powder inhalers used to provide triple therapy, in patients with COPD.

Compound Number: GSK573719+GW642444+GW685698 (GSK2834425)

Sponsor Name and Legal Registered Address:

GlaxoSmithKline Research & Development Limited980 Great West RoadBrentfordMiddlesex, TW8 9GSUK

Medical Monitor Name and Contact Information can be found in the Study Reference Manual

Regulatory Agency Identifying Number(s): Not Applicable

Approval Date: 15-SEP-2016


2017N324761_00

2017N324761_00

PPD

PPD

PPD

2016N292801_01 CONFIDENTIAL206215

3

TABLE OF CONTENTS

PAGE

1. SYNOPSIS...............................................................................................................5

2. SCHEDULE OF ACTIVITIES (SOA).........................................................................8

3. INTRODUCTION......................................................................................................93.1. Study Rationale ............................................................................................93.2. Background ..................................................................................................93.3. Benefit/Risk Assessment ............................................................................11

3.3.1. Risk Assessment .........................................................................113.3.2. Benefit Assessment .....................................................................123.3.3. Overall Benefit: Risk Conclusion..................................................12

4. OBJECTIVES AND ENDPOINTS...........................................................................12

5. STUDY DESIGN ....................................................................................................135.1. Overall Design ............................................................................................135.2. Number of Participants ...............................................................................145.3. Participant and Study Completion...............................................................145.4. Scientific Rationale for Study Design ..........................................................155.5. Dose Justification........................................................................................15

6. STUDY POPULATION ...........................................................................................156.1. Inclusion Criteria .........................................................................................156.2. Exclusion Criteria........................................................................................166.3. Lifestyle Restrictions...................................................................................17

6.3.1. Meals and Dietary Restrictions ....................................................176.3.2. Caffeine, Alcohol, and Tobacco ...................................................176.3.3. Activity .........................................................................................17

6.4. Screen Failures...........................................................................................17

7. TREATMENTS.......................................................................................................177.1. Treatments Administered............................................................................18

7.1.1. Medical Devices...........................................................................187.2. Method of Treatment Assignment ...............................................................197.3. Blinding.......................................................................................................197.4. Preparation/Handling/Storage/Accountability ..............................................207.5. Treatment Compliance................................................................................207.6. Concomitant Therapy..................................................................................207.7. Treatment after the End of the Study ..........................................................21

8. DISCONTINUATION CRITERIA.............................................................................218.1. Discontinuation of Study Treatment ............................................................21

8.1.1. Liver Chemistry Stopping Criteria ................................................218.1.2. QTc Stopping Criteria ..................................................................21

8.2. Withdrawal from the Study..........................................................................218.3. Lost to Follow Up ........................................................................................21

9. STUDY ASSESSMENTS AND PROCEDURES .....................................................219.1. Efficacy Assessments.................................................................................22

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

4

9.1.1. Assessment of Errors in Use of Device........................................229.1.1.1. Assessment of ELLIPTA error in use..........................239.1.1.2. Assessment of DISKUS-HandiHaler in Sub

Study 1, or Turbuhaler-HandiHaler in Sub Study 2, for error in use ..............................................23

9.1.1.3. Time Taken to Use Device .........................................249.1.2. Assessments of Preference .........................................................24

9.2. Adverse Events...........................................................................................259.2.1. Time Period and Frequency for Collecting AE and SAE

Information...................................................................................259.2.2. Method of Detecting AEs and SAEs.............................................259.2.3. Follow-up of AEs and SAEs.........................................................259.2.4. Regulatory Reporting Requirements for SAEs .............................269.2.5. Cardiovascular and Death Events................................................26

9.3. Safety Assessments ...................................................................................269.4. Pharmacokinetics .......................................................................................269.5. Pharmacodynamics ....................................................................................279.6. Genetics .....................................................................................................279.7. Biomarkers .................................................................................................279.8. Health Economics OR Medical Resource Utilization and Health

Economics ..................................................................................................27

10. STATISTICAL CONSIDERATIONS........................................................................2710.1. Hypotheses.................................................................................................2710.2. Sample Size Determination ........................................................................2710.3. Populations for Analyses ............................................................................2810.4. Statistical Analyses.....................................................................................29

10.4.1. Efficacy Analyses.........................................................................2910.4.2. Safety Analyses ...........................................................................3010.4.3. Other Analyses ............................................................................3010.4.4. Interim Analyses ..........................................................................30

11. REFERENCES.......................................................................................................31

12. APPENDICES ........................................................................................................3212.1. Appendix 1: Abbreviations and Trademarks................................................3212.2. Appendix 2: Inhaler Specific Errors for the inhalers used in the study.........3312.3. Appendix 3: Inhaler Preference Questionnaires..........................................3612.4. Appendix 4: Study Governance Considerations..........................................3812.5. Appendix 5: Adverse Events: Definitions and Procedures for

Recording, Evaluating, Follow-up, and Reporting .......................................42

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

5

1. SYNOPSIS

Protocol Title: A randomized, open-label, cross-over, placebo-device study investigating critical and over all errors, training/teaching time, and preference attributes of the ELLIPTA dry powder Inhaler (DPI) as compared to HandiHaler DPI used in combination with either DISKUS DPI or Turbuhaler DPI, in adult patients with Chronic ObstructivePulmonary Disease (COPD)

Short Title: A clinical study assessing critical errors, training/teaching time, and preference attributes of the ELLIPTA dry powder inhaler, in comparison to combinations of dry powder inhalers used to provide triple therapy, in patients with COPD.

Rationale: This study is designed to assess the benefits of delivering triple therapy using a single ELLIPTATM DPI (Closed Triple therapy) versus delivering triple therapy using two different types of inhalers (open triple therapy) to patients with Chronic Obsttructive Pulmonay Disease (COPD). It will assess the proportion of COPD subjects who make critical errors when using a single ELLIPTA™ DPI versus those using combinations of commercially available and commonly used DPIs: DISKUS used in combination with HandiHaler, or Turbuhaler used in combination with HandiHaler. This study would also assess training/teaching time and preference attributes for closed triple therapy as compared to the open triple therapy.

Objectives and Endpoints:

Objectives Endpoints

Primary

To compare the number of critical errors made by COPD patients, after a subject has read the respective patient information leaflet(s) (PIL), for each treatment option tested

The percentage of subjects making at least one critical error after reading the PIL(s)

Secondary

To compare the number of critical errors made by COPD patients after instruction from the Healthcare Professional (HCP) for each treatment option tested

The percentage of subjects making at least one critical error after the: first instruction from the HCP

second instruction from the HCP

To compare the number of overall(critical and non-critical) errors made by COPD patients, after a subject has read the PIL(S) or after Instruction from the HCP for each treatment optiontested

The percentage of subjects making at least one overall error after reading the PIL(s)

The percentage of subjects making at least one overall error after the first instruction from the HCP

The percentage of subjects making at least one overall error after the second instruction from the HCP

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

6


To compare the number of instructions (maximum of 2) from a HCP which is needed to demonstrate correct inhaler use

The number of instructions (0, 1 or 2 times) from the HCP which are needed to demonstrate correct inhaler use

To compare the Training/Teaching Time required to demonstrate correct inhaler use

The total amount of time taken to demonstrate correct inhaler use (T1+T2).

The amount of time taken to read the patient information leaflet and demonstrate correct inhaler use (T1)

The amount of time taken to be given instruction by the HCP (up to 2 times) on use of the inhaler and to demonstrate correct inhaler use (T2)

Preference attributes for each treatment option tested

Treatment preference, from questionnaire for: Number of steps required to take COPD

medication

Over all treatment preference

Overall Design: The study will be conducted as a multi-centre, randomized, open-label, placebo-device, cross-over study, with a 2x2 complete block design.

It will comprise of two sub-studies:

Sub-study 1: Will compare ELLIPTA DPI to DISKUS-HandiHaler DPI combination

Sub-study 2: Will compare ELLIPTA DPI to Turbuhaler-HandiHaler DPI combination.

The study has 2 visits (V0 and V1) and both can be completed on the same day.

Each sub-study may run independently or in parallel of the other sub study. Each will start dependent on the availability of the placebo DPIs required for that sub-study. The data from the 2 sub studies may be reported independently and as each sub study completes.

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

7

Study Schematic

Number of Participants: Sufficient patients with COPD will be screened to ensure approximately 160 participants will be randomized, such that approximately 144evaluable participants complete the study. Approximately 72 participants completing in each sub study.

Treatment Groups and Duration: Duration is a single visit and subjects are randomised to receive treatments below; dependent on which sub-study they are included in.

There is no active treatment and subjects will continue to take their own prescribed COPD medication for the duration of the study.

Sub Study 1 Treatment Sequences



2017N324761_00

2016N292801_01 CONFIDENTIAL206215

8

Sub Study 2 Treatment Sequence



2. SCHEDULE OF ACTIVITIES (SOA)

Visit Number V0 V1 Notes

Study Day 1 1 V0 can take place on the same day as V1. V1 should be completed no later than 30 days after consent.

Procedure:

Screening Assessments Completed prior to randomisation

Written informed consent X Informed consent may take place prior to V0for logistical reasons. Subjects should be included and randomised within 30 days of providing consent.

Subject demography X Age, height, weight, year of birth, sex, ethnicity and geographic ancestry will be recorded

Medical/disease history including Chronic Obstructive Pulmonary Disease (COPD)

X Subject will have a medical history of COPD, previously confirmed by spirometry.

Concomitant medication history including COPD Therapy History

X Current concomitant medication will be recorded. A minimum COPD therapy history for the preceding 2 years from inclusion will be recorded

Inclusion/exclusion criteria X All criteria must be meet prior to randomisation at V1

Study Assessments Completed once a subject is included on study

Randomisation X Randomised to treatment order and preference questionnaire

Assess the number of inhaler errors (critical and overall) on each treatment after reading the Patient information leaflet (PIL) for Inhaler tested

X No instruction is provided by the Health Care Professional (HCP) for this assessment.

Assess the number of inhaler errors (overall and critical) on each treatment after each of 2 attempts following instruction by HCP

If a subject cannot show correct use after reading the PIL, then the HCP has up to 2attempts to instruct the subject to attain this.

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

9


Teaching-Training Time for each inhaler includes the following: The amount of time taken to

read the patient information leaflet and demonstrate inhaler use

The amount of time taken to be given instruction by the HCP on use of the inhaler and demonstrate inhaler use

The total amount of time taken to demonstrate inhaler use

X First assessment is attempt one, including time to read PIL and demonstrate use.

Second assessment will be time for HCP to correct errors and subject to show use for up to 2 more attempts

The Cumulative time for all assessments needed by subject and any HCPinstruction to demonstrate no errors will also be captured.

Preference Questionnaire X Subject will complete version of Preference Questionnaire they have been randomised to.

SAE/AE assessment X Collected until completion of final study assessment at V1.

3. INTRODUCTION

3.1. Study Rationale

Please note that Turbohaler and ACCUHALER are the names commonly used within the United Kingdom for these inhalers. However, in other regions and for the purposes of this study these inhalers are referred to as Turbuhaler and DISKUS respectively in this protocol.

This study is designed to assess the benefits of delivering triple therapy using a single ELLIPTA TM Dry Powder Inhaler (DPI) (Closed Triple therapy) versus delivering triple therapy using two different types of inhalers (open triple therapy) to patients with COPD. It will assess the proportion of COPD subjects who make critical errors when using a single ELLIPTA DPI versus those using combinations of commercially available and commonly used DPIs: DISKUS used in combination with HandiHaler, or Turbuhalerused in combination with HandiHaler. This study would also assess training/teaching time and preference attributes for closed triple therapy as compared to the open triple therapy.

3.2. Background

Chronic Obstructive Pulmonary Disease (COPD) is a progressive disease of the airways. Noxious particles or gases lead to a modification in subject’s airways which develop a chronic inflammatory response to these environmental factors. This leads to increasing airflow limitation, breathlessness and other symptoms. Despite being both treatable and preventable, COPD is a leading cause of morbidity and mortality worldwide. The economic and social burden of this disease has not been reduced and is increasing despite advances in diagnosis and treatment. [GOLD 2016]

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

10

The current treatment goals for COPD are to reduce and relieve symptoms and thereby improve exercise tolerance and health status, whilst reducing risk by preventing disease progression, preventing and treating exacerbations, and reducing mortality

Smoking cessation, inhaled pharmacological therapy, primarily long and short term bronchodilators (beta2-agonists and anti-cholinergics) and inhaled corticosteroids, and other non pharmacological interventions are currently used to achieve the treatment goals.

The current GOLD treatment guidelines, place patients in 4 groups from A to D, dependent on severity of symptoms and assessment of risk; with Group A patients having the fewest symptoms and a low risk as assessed by degree of airflow limitation and/or exacerbation history. COPD patients in Group A may just need a short acting bronchodilator as needed to relieve symptoms. However, as disease progresses, other inhaled therapies are added on to maintain control, such that COPD patients in Group D, who have many symptoms and a high risk of exacerbations and/or high airflow limitation, may need multiple inhaled therapies (inhaled corticosteroids (ICS), long-acting beta2-agonist (LABA) and long-acting anticholinergic (LAMA)).

Currently, subjects requiring triple therapy can be prescribed ICS/LABA and LAMA in separate inhalers. The specific ICS/LABA and LAMA prescribed determine whether the inhaler types (and thereby the inhalation techniques) and the dosing regimens are similar or different. Use of different inhaler types with different inhalation techniques and dosing regimens can add to treatment complexity, and also increase the potential for errors in inhaler use that reduce or preclude drug delivery to the site of action in the lungs[Cochrane, 2000, Van der Palen, 1999]. Fixed-dose combination inhalers that minimisethe number of inhalers required would simplify treatment, improve adherence, reduce errors in inhaler use, and potentially lead to better outcomes [GOLD 2016].

The skill and ability of the COPD patient to use the prescribed inhaler/s correctly coupled with adequate training in inhaler technique are also critical to ensure effective drug delivery [Cochrane, 2000, Melani, 2011]. For any prescribed inhaler, the patient needs to follow all the steps in the patient leaflet correctly in order to ensure optimal drug delivery. In this regard, the time needed for a primary care nurse or physician, or a community pharmacist to train a patient in correct use of an inhaler at the time of initial prescription, and for any subsequent retraining becomes important. Given the time demands on healthcare professionals, a device which is simple to use and requires minimal time to train would be desirable [Bonini, 2015]. Further, patients may prefer easy-to-use inhalers having fewer steps to deliver drug; this has the potential to improve compliance and thereby impact outcomes.

ELLIPTA DPI has been designed to be simple for patients to use. In COPD patients’naïve to Ellipta and comparator inhalers, data has shown that patients make fewer critical and overall errors when using ELLIPTA as compared to other common DPIs tested [Van Der Palen, 2016(a)]. ELLIPTA DPI was also shown to be preferred by patients for a number of attributes, including number of steps and training time required to receive therapy [Van Der Palen, 2016(b), Van Der Palen, 2016(c), Komase, 2014]. ELLIPTA is already available to deliver a LABA/ICS combination (BREO ELLIPTA), LAMA

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

11

(INCRUSE ELLIPTA), and LAMA/LABA combination (ANORO ELLIPTA). Closed triple therapy (ICS/LAMA/LABA) is currently being assessed in clinical studies to deliver all 3 active treatments from a single ELLIPTA DPI.

3.3. Benefit/Risk Assessment

The study involves use of placebo inhalers (placebo ELLIPTA, placebo DISKUS, placebo Turbuhaler and placebo HandiHaler) that do not contain active treatments and subjects will continue to take their own prescribed COPD medication and other concomitant medication for the duration of the study.

The placebo inhalers and placebo capsules for HandiHaler contain the excipients lactose or lactose blended with magnesium stearate. Excipients of the study inhaler are noted in Section 7.1. Subjects with a known hypersensitivity to any of these or severe milk protein allergy that could contraindicate study participation are excluded from the study (Section 6). Subjects who meet the inclusion/exclusion criteria will continue their COPD treatment as prescribed by their healthcare provider during their participation in the study. Subjects should continue to follow up with their regular physician for their COPD healthcare during the study.

The study procedures include reading of the inhaler patient information leaflets and demonstration of inhaler use by the study subjects and instruction in correct-inhaler use by the site staff.

3.3.1. Risk Assessment

Potential Risk of Clinical

Significance

Summary of Data/Rationale for Risk

Mitigation Strategy

Paradoxical bronchospasm, which may occur with an immediate increase in wheezing after inhaling.

As with other inhalation therapy, paradoxical bronchospasm may occur with an immediate increase in wheezing after dosing. From post-marketing data, paradoxical bronchospasm has been reported at afrequency of <1/10,000 including isolated reports.

This should be treated immediately with a fast and short acting inhaled bronchodilator. Investigators will be instructed to assess the subject’s condition to determine their eligibility to continue in the study and the need for alternative therapy.

Allergic reaction due to hypersensitivity to placebo excipients.

The placebo inhalers andplacebo capsules (for HandiHaler) contain theexcipients lactose andlactose blended with magnesium stearate. Thereare known allergies to these ingredients.

Subjects with a known hypersensitivity to any of these, or severe milk protein allergy that could contraindicate study participation are excluded from the study (Section 6.2). If an allergic reaction occurs, it should be treated immediately with a fast and short acting inhaled bronchodilator. Investigators will be instructed to assess the subject’s condition to determine their eligibility to continue in the study and the need for alternative therapy

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

12

3.3.2. Benefit Assessment

As this is a placebo study, no benefit to the subject is expected. No active treatment is being administered. The subjects will continue to receive their own COPD therapy as prescribed.

3.3.3. Overall Benefit: Risk Conclusion

The overall potential risk identified is minimal, due to the nature of the study.

4. OBJECTIVES AND ENDPOINTS

The study will compare critical errors, training time, and patient preference for closed triple therapy received from a single ELLIPTA inhaler versus open triple therapy received from HandiHaler in combination with either DISKUS or Turbuhaler. These comparisons will be performed as separate sub studies and the endpoints and objectives listed here will be the same for the treatment comparisons in each sub study:

Sub study 1: ELLIPTA (Treatment option 1) versus the combination of DISKUS-HandiHaler (Treatment Option 2)

Sub study 2: ELLIPTA ((Treatment option 1) versus the combination of Turbuhaler-HandiHaler (Treatment Option 3)


Primary

To compare the number of critical errors a made by COPD patients, after a subject has read the respective patient information leaflet(s) (PIL), for each treatment option tested


Secondary

To compare the number of critical errors made by COPD patients after instruction from the Healthcare Professional (HCP) for each treatment option tested

The percentage of subjects making at least one critical error after the: first instruction from the HCP

second instruction from the HCP d

To compare the number of overall a

(critical and non-critical) errors made by COPD patients, after a subject has read the PIL(S) or after Instruction from the HCP for each treatment optiontested



The percentage of subjects making at least one overall error after the second instruction from the HCP d

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

13


To compare the number of instructions (maximum of 2) from a HCP which is needed to demonstrate correct inhaler usec




The amount of time taken to read the patient information leaflet and demonstrate correct inhaler use (T1) d

The amount of time taken to be given instruction by the HCP (up to 2 times) on use of the inhaler and to demonstrate correct inhaler use (T2) d

Preference attributes for each treatment option tested b

Treatment preference, from questionnaire for: Number of steps required to take COPD

medication


a. Over all and critical errors for each inhaler tested, as well as an explanation of how these endpoints were defined, are listed in Appendix 2

b. Preference Questionnaires for each sub-study are included in Appendix 3c. To show correct inhaler use a subject demonstrates the use of the inhaler without making any critical or non-

critical errord. Endpoint will be treated as ‘other endpoint’ and not secondary for the statistical analysis

5. STUDY DESIGN

5.1. Overall Design

The study will be conducted as a multi-centre, randomized, open-label, placebo-device, cross-over study, with a 2x2 complete block design.

It will comprise of two sub-studies:

Sub-study 1: Will compare ELLIPTA DPI (Treatment option 1) to DISKUS-HandiHaler DPI combination (Treatment option 2)

Sub-study 2: Will compare ELLIPTA DPI (Treatment option 1) to Turbuhaler-HandiHaler DPI combination (Treatment option 3).

The study has 2 visits (V0 and V1) and both can be completed on the same day.

Each sub-study may run independently or in parallel of the other sub study. Each will start dependent on the availability of the placebo DPIs required for that sub-study. The data from the 2 sub studies will be reported independently and as each sub study completes and the data have been cleaned. The database will be locked when all of the sub-studies have completed and data are cleaned.

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

14

The study subjects will be randomised to order of treatment and preference questionnaire in each of the 2 sub studies. This is shown in further detail in Figure 1.

Figure 1 Study Schematic

5.2. Number of Participants

Sufficient patients with COPD will be screened to ensure approximately 160 participants will be randomized, such that approximately 144 evaluable participants complete the study. An evaluable subject is defined in the intent to treat population in Section 10.3. Approximately 72 participants completing in each sub study.

If participants prematurely discontinue the study, additional replacement participants maybe recruited at the discretion of the Sponsor.

5.3. Participant and Study Completion

A participant is considered to have completed the study if he/she has completed all phases of the study including the last scheduled procedure shown in the Schedule of Activities.

The end of the study is defined as the date of the last visit of the last participant in the study.

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

15

5.4. Scientific Rationale for Study Design

This design has been used in previous studies looking at critical errors and/or preference of inhalers in patients with COPD and naive to the inhalers tested. The cross-over design was chosen to allow the subjects to serve as their own control in their ability to use both the ELLIPTA DPI and the other combination DPIs. A single visit is suitable for assessing these endpoints and reduces inconvenience of multiple visits for subject included.

Placebo inhalers are used to remove any bias due to treatment effect, avoid the need for wash-in/-out periods, and to avoid the need to withhold or discontinue current COPD medications, which may affect the subject’s clinical status (improvement or decline) and perceptions, as noted above.

5.5. Dose Justification

Not applicable as this is a placebo-only study.

6. STUDY POPULATION

Prospective approval of protocol deviations to recruitment and enrolment criteria, also known as protocol waivers or exemptions, is not permitted.

6.1. Inclusion Criteria

Participants are eligible to be included in the study only if all of the following criteria apply:

AGE

1. ≥40 years of age at Visit 1

TYPE OF SUBJECT AND DIAGNOSIS

2. Diagnosis of COPD with a documented history of COPD, in accordance with thedefinition by the European Respiratory Society [Celli, 2004].

3. Current COPD Therapy: Currently receiving maintenance therapy with a fixed dose combination of a long-acting beta 2-agonist (LABA) and inhaled corticosteroid (ICS). Subject may also be receiving long-acting muscarinic antagonist (LAMA; also known as a long-acting anti-cholinergic). Subjects must be able to continue using their currently prescribed COPD maintenance inhaler therapy throughout the study and as needed short acting beta-adrenergic agonist (SABA) and/or short acting muscarinic antagonist (SAMA) for rescue use.

4. Has been on current maintenance ICS/LABA COPD treatment for at least 4 weeksprior to V0 and evaluated as unlikely to change treatment within 4 weeks of Visit 1.

SMOKING HISTORY

5. Smoking History: Current or former (defined as subjects who have quit smoking for at least 3 months prior to V0/V1) cigarette smokers with a >10 pack-year smoking history [Number of pack years = (number of cigarettes per day ÷ 20) x number of

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

16

years smoked (e.g., 10 pack-years is equal to 20 cigarettes per day for 10 years, or 10 cigarettes per day for 20 years].

SEX

6. Males or

Females who are not pregnant or not planning a pregnancy during the study or not lactating

INFORMED CONSENT

7. Capable of giving signed informed consent as described in Appendix 4 which includes compliance with the requirements and restrictions listed in the informed consent form (ICF) and in this protocol.

6.2. Exclusion Criteria

Participants are excluded from the study if any of the following criteria apply:

CONCURRENT CONDITIONS/MEDICAL HISTORY

1. Asthma: Subjects with a current diagnosis of asthma. Subjects with a prior history of asthma are eligible if they have a current diagnosis of COPD.

CONCOMITANT MEDICATIONS

2. Recent experience with the ELLIPTA inhaler: Subjects who used any ELLIPTA inhaler (e.g., RELVAR ELLIPTA, ANORO ELLIPTA, ARNUITY ELLIPTA, INCRUSE ELLIPTA, participated in a clinical study of GW685698, GW642444, GSK573719 [fluticasone furoate, vilanterol, umeclidinium bromide], or any combination thereof, or placebo in an ELLIPTA inhaler study) within 24 monthsprior to Visit 0.

3. Recent experience with any capsule inhaler: Subjects who used any capsule systeminhaler (e.g. Spiriva HandiHaler, Seebri/Ultibro Breezhaler, or participated in a clinical studies of these, including placebo inhalers) within 24 months prior to Visit 0.

4. Dependent on which sub-study a subject is included on they should not have any recent experience, within 24 months of V0 of the following inhaler for the sub study included on:

Sub Study 1: DISKUS inhaler (e.g. Seretide DISKUS or placebo DISKUS) Sub Study 2:Turbuhaler (e.g. Symbicort Turbuhaler or placebo Turbuhaler)

RELEVANT HABITS

5. Drug/alcohol abuse: Subjects with a known or suspected alcohol or drug abuse at Visit 1 which in the opinion of the investigator could interfere with the subject’s proper completion of the protocol requirement

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

17

CONTRAINDICATIONS

6. Drug/Food Allergy: A history of hypersensitivity to any components of the study inhaler (e.g., lactose, magnesium stearate). In addition, subjects with a history of severe milk protein allergy that, in the opinion of the study physician, contraindicates participation will also be excluded.

DIAGNOSTIC ASSESSMENTS AND OTHER CRITERIA

7. Investigational Product: Subjects who have received an investigational drug and/or medical device within 30 days of entry into this study (Screening/Visit 1), or within five drug half-lives of the investigational drug, whichever is longer

8. Inability to Read: In the opinion of the investigator, any subject who is unable to read and/or would not be able to complete a questionnaire and understand verbal instructions.

6.3. Lifestyle Restrictions

There are no lifestyle restrictions.

6.3.1. Meals and Dietary Restrictions

There are no meals or dietary restrictions.

6.3.2. Caffeine, Alcohol, and Tobacco

There are no caffeine, alcohol, and tobacco restrictions, other than those for a subject’sinclusion and any local restrictions whilst the subject is in the clinic.

6.3.3. Activity

There are no restrictions on activity.

6.4. Screen Failures

Screen failures are defined as participants who consent to participate in the clinical study but are not subsequently randomized. A minimal set of screen failure information is required to ensure transparent reporting of screen failure participants to meet the Consolidated Standards of Reporting Trials (CONSORT) publishing requirements and to respond to queries from regulatory authorities. Minimal information includes demography, screen failure details, eligibility criteria, and any serious adverse events (SAEs).

Individuals who do not meet the criteria for participation in this study (screen failure) may be rescreened. Rescreened participants should be assigned a new subject number.

7. TREATMENTS

Study treatment is defined as any investigational treatment(s), marketed product(s), placebo, or medical device(s) intended to be administered to a study participant according to the study protocol.

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

18

7.1. Treatments Administered

Study Treatment Name:

ELLIPTA placebo DPI

DISKUSplacebo DPI

HandiHaler DPI and placebo Capsules

Turbuhaler placebo DPI

Dosage formulation:

Placebo DPI with two strips with 30 blisters per strip.

First strip: lactose monohydrateSecond strip: lactose monohydrate blended with magnesium stearate

Placebo DPIwith one blister strip containing lactose monohydrate.

DPI with placebocapsules containing lactose monohydrate

Placebo DPI containing lactose monohydrate

Route of Administration

Oral Inhalation Oral Inhalation Oral Inhalation Oral Inhalation

Dosing instructions:

As directed As directed As directed As directed

Packaging and Labelling

Study Treatment will be provided in the appropriate container and will be labelled as required per country requirement.




Manufacturer GSK GSK Boehringer Ingelheim

AstraZeneca

Device NA NA HandiHaler NA

7.1.1. Medical Devices

Other medical device (not manufactured by or for GSK) provided for use in thisstudy is HandiHaler Dry Powder Inhaler

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

19

Any extra instructions for medical device will be provided in the Study Reference Manual.

7.2. Method of Treatment Assignment

All participants will be centrally randomized using an Interactive Web Response System (IWRS). Before the study is initiated, the log in information & directions for the IWRS will be provided to each site.

Participants will be assigned in equal numbers to one of four sequences for the sub-study they are included on, in accordance with the randomization schedule generated by Clinical Statistics, prior to the start of the study and using validated internal software.

The possible sequences for each sub study are shown in Table 1 and Table 2

Table 1 Sub Study 1 Treatment Sequences



Table 2 Sub Study 2 Treatment Sequence



Study treatments will be dispensed at the study visit summarized in SOA

7.3. Blinding

This is an open-label study; however, the order of treatment and preference questionnaire answered by a participant will be assigned using an IWRS. The site will contact the IWRS prior to study treatment administration for each participant. The site will record the treatment assignment on the applicable case report form, if required. Potential bias will be reduced by randomising the order of treatment and the preference questionnaire version answered by subjects.

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

20

7.4. Preparation/Handling/Storage/Accountability

The investigator or designee must confirm appropriate temperature conditions have been maintained during transit for all study treatment received and any discrepancies are reported and resolved before use of the study treatment.Details of transit and storage conditions for study treatments will be included in the SRM.

Only participants enrolled in the study may receive study treatment and only authorized site staff may supply study treatment. All study treatments must be stored in a secure, environmentally controlled, and monitored (manual or automated) area in accordance with the labeled storage conditions with access limited to the investigator and authorized site staff.

The investigator, institution, or the head of the medical institution (where applicable) is responsible for study treatment accountability, reconciliation, and record maintenance (i.e., receipt, reconciliation, and final disposition records).

Further guidance and information for the final disposition of unused study treatment are provided in the Study Reference Manual.

Under normal conditions of handling and administration, study treatment is not expected to pose significant safety risks to site staff.

7.5. Treatment Compliance

Participants receive study treatment only at the clinical site and they will receive study treatment directly under supervision of the investigator or designee.

The date and time of each dose administered in the clinic will be recorded in the source documents.

7.6. Concomitant Therapy

A detailed history of previous and ongoing COPD medications, in particular any types of inhalers used to deliver these medications for the previous 24 months from V0, should be recorded in order to inform on the subject’s inclusion.

This is a single visit study, and the subject should continue to take their usual COPD and other medications through the conduct of the study.

However, any medication or vaccine (including over-the-counter or prescription medicines, vitamins, and/or herbal supplements) that the participant is receiving at the time of enrolment or receives during the study must be recorded along with:

reason for use

dates of administration including start and end dates

dosage information including dose and frequency

The Medical Monitor should be contacted if there are any questions regarding concomitant or prior therapy.

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

21

7.7. Treatment after the End of the Study

There is no active treatment; this is a single visit, placebo only study and participants will not receive any specific post study treatments.

8. DISCONTINUATION CRITERIA

8.1. Discontinuation of Study Treatment

Not applicable as this is a single-visit study with no active treatment; all subjects will receive placebo treatment only.

8.1.1. Liver Chemistry Stopping Criteria

Not applicable as this is a single-visit study with no active treatment; all subjects will receive placebo treatment only. .

8.1.2. QTc Stopping Criteria

Not applicable as this is a single-visit study with no active treatment; all subjects will receive placebo treatment only.

8.2. Withdrawal from the Study

A participant may withdraw from the study at any time at his/her own request, or may be withdrawn at any time at the discretion of the investigator for safety, behavioural, compliance or administrative reasons.

If the participant withdraws consent for disclosure of future information, the sponsor may retain and continue to use any data collected before such a withdrawal of consent.

This is a single visit, placebo only study, if a subjects withdrawals during this single visit there are no assessments that would need to be completed following withdrawal.

8.3. Lost to Follow Up

Not applicable as this is a single visit study.

9. STUDY ASSESSMENTS AND PROCEDURES

Study procedures and their timing are summarized in the SoA.

Protocol waivers or exemptions are not allowed

Immediate safety concerns should be discussed with the sponsor immediately upon occurrence or awareness to determine if the participant should continue or discontinue study treatment.

Adherence to the study design requirements, including those specified in the SoA, is essential and required for study conduct.

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

22

All screening evaluations must be completed and reviewed to confirm that potential participants meet all eligibility criteria. The investigator will maintain a screening log to record details of all participants screened and to confirm eligibility or record reasons for screening failure, as applicable.

Procedures conducted as part of the participant’s routine clinical management (e.g.,to confirm diagnosis of COPD) and obtained before signing of ICF may be utilized for screening provided the procedure met any protocol-specified criteria and was performed within the time frame defined in the SoA.

Prior to randomisation, at V1, all assessments to confirm inclusion at V0 will be completed. The details of these assessment are provided in the SoA

9.1. Efficacy Assessments

9.1.1. Assessment of Errors in Use of Device

Within each sub-study, subjects will be randomised for the order of treatment. One treatment will require subjects to demonstrate use of a single inhaler (ELLIPTA), while the other treatment will require the subject to demonstrate use of 2 inhalers (DISKUS-HandiHaler in Sub Study 1 or Turbuhaler-HandiHaler in Sub Study 2).

Subjects will be provided with the relevant section of the PIL, explaining correct use, for each inhaler they are to be tested on. The critical and over all errors per inhaler and how these have been defined are included in Appendix 2.

The errors listed will be in aligned with the correct use information from the respective PILs, in a checklist for each inhaler and these will be provided to HCPs for scoring errors during the study conduct. The checklists will be provided in the Study Reference Manual.

Figure 2 depicts the process flow for assessing inhaler use errors and further details areprovided, for each treatment, in Section 9.1.1.1 and Section 9.1.1.2.

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

23

Figure 2 Assessment of errors in use for each DPI tested

9.1.1.1. Assessment of ELLIPTA error in use

Subject will be asked to read the patient instruction leaflet for ELLIPTA and then demonstrate ELLIPTA use. Any errors (critical or non-critical) made by the subjectwhile using the ELLIPTA will be recorded by the HCP on the checklist provided. If the subject makes no errors, this will also be recorded by the HCP, and there will be no further assessments. If the subject makes any error in the use of the ELLIPTA, the HCP will provide instruction in the correct use of the ELLIPTA to the subject. The subjectwill then demonstrate ELLIPTA use again. If the subject makes no errors, this will be recorded by the HCP and there will be no further assessments. If the subject again makeserrors in the use of the ELLIPTA any errors will be recorded by the HCP. The HCP will again provide instruction in the correct use for the final time, following which the subjectwill then demonstrate ELLIPTA use one final time. Any errors made during this demonstration will be recorded. There will be no further assessments. In total, the HCP can provide instruction in the use of the inhaler up to two times.

9.1.1.2. Assessment of DISKUS-HandiHaler in Sub Study 1, or Turbuhaler-HandiHaler in Sub Study 2, for error in use

Subjects will be provided with 2 inhalers and 2 PILs dependent on which sub study they are included in.

Subjects should be tested on each DPI provided consecutively, they will be tested in the following order for each sub study.

Sub Study 1: DISKUS followed by HandiHaler

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

24

Sub Study 2: Turbuhaler followed by HandiHaler

Subjects will be asked to read the patient instruction leaflet for the first DPI (DISKUS or Turbuhaler) and then demonstrate first DPI use. Any errors (critical or non-critical) made by the subject while using the first DPI will be recorded by the HCP on the checklists provided. If the subject makes no errors, this will also be recorded by the HCP and there will be no further assessment for this first DPI. If the subject makes any error in the use of the first DPI, the HCP will provide instruction in the correct use of the inhaler to the subject. The subject will then demonstrate inhaler use again. If the subject makes no errors, this will also be recorded by the HCP and there will be no further assessments for this DPI. If the subject again makes errors in the use of the DPI, the HCP will record the errors and provide instruction in correct use for the final time. The subject will then demonstrate use of first DPI for one final time. Any errors or correct use after this demonstration will be recorded, and there will be no further assessments for this DPI. In total, the HCP can provide instruction in the use of the first DPI up to two times.

This assessment will then be repeated with HandiHaler.

9.1.1.3. Time Taken to Use Device

Time taken to use the device will be recorded as follows:

T1: the time from when the subject starts to read the patient instruction leaflet until they have completed demonstration of device use (i.e., with no investigator support).

T2: the time from when the HCP starts to instruct subject until correct use is demonstrated (up to a maximum of two attempts only).

T1+T2: the time from when the subject starts to read the patient information leaflet until correct use is demonstrated (up to a maximum of three attempts, once after reading the PIL and following instruction, up to 2 times, by HCP).

The HCP will start recording (clock start) the time from when T1 begins and will stop when device use has been demonstrated. The clock re-starts when T2 begins and stops once correct use has been demonstrated, or when three failed attempts have been made. Note: T2 includes the time used by the investigator for re-instructing the Subjectsthroughout.

For subjects who demonstrate correct inhaler use after reading the PIL, T2 will be 0.

9.1.2. Assessments of Preference

After completing the errors in use assessment for both treatments, the HCP will ask the subject questions from the assigned preference questionnaire. The wording of the questionnaires is included in Appendix 3.

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

25

9.2. Adverse Events

The definitions of an AE or SAE can be found in Appendix 5.

The investigator and any designees are responsible for detecting, documenting, and reporting events that meet the definition of an AE or SAE and remain responsible for following up AEs that are serious, considered related to the study treatment or the study,or that caused the participant to discontinue the study (see Section 8).

9.2.1. Time Period and Frequency for Collecting AE and SAE Information

All SAEs will be collected from consent until the time point specified in the SoA (Section 2).

All AEs will be collected from randomisation until the time point specified in the SoA (Section 2).

Medical occurrences that begin before the start of study treatment but after obtaining informed consent will be recorded on the Medical History/Current Medical Conditions section of the case report form (CRF) not the AE section.

All SAEs will be recorded and reported to the sponsor or designee within 24 hours, as indicated in Appendix 5. The investigator will submit any updated SAE data to the sponsor within 24 hours of it being available.

Investigators are not obligated to actively seek AEs or SAEs in former study participants. However, if the investigator learns of any SAE, including a death, at any time after a participant has been discharged from the study, and he/she considers the event to be reasonably related to the study treatment or study participation, the investigator must promptly notify the sponsor.

The method of recording, evaluating, and assessing causality of AEs and SAEsand the procedures for completing and transmitting SAE reports are provided in Appendix 5.

9.2.2. Method of Detecting AEs and SAEs

Care will be taken not to introduce bias when detecting AE and/or SAE. Open-ended and non-leading verbal questioning of the participant is the preferred method to inquire about AE occurrence.

9.2.3. Follow-up of AEs and SAEs

After the initial AE/SAE report, the investigator is required to proactively follow each participant at subsequent visits/contacts. All SAEs, will be followed until the event is resolved, stabilized, otherwise explained, or the participant is lost to follow-up). Further information on follow-up procedures is given in Appendix 5.

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

26

9.2.4. Regulatory Reporting Requirements for SAEs

Prompt notification by the investigator to the sponsor of a SAE is essential so that legal obligations and ethical responsibilities towards the safety of participants and the safety of a study treatment under clinical investigation are met.

The sponsor has a legal responsibility to notify both the local regulatory authority and other regulatory agencies about the safety of a study treatment under clinical investigation. The sponsor will comply with country-specific regulatory requirements relating to safety reporting to the regulatory authority, Institutional Review Boards (IRB)/Independent Ethics Committees (IEC), and investigators.

Investigator safety reports must be prepared for suspected unexpected serious adverse reactions (SUSAR) according to local regulatory requirements and sponsor policy and forwarded to investigators as necessary.

An investigator who receives an investigator safety report describing a SAE or other specific safety information eg, summary or listing of SAE) from the sponsor will review and then file it along with the Investigator’s Brochure and will notify the IRB/IEC, if appropriate according to local requirements.

9.2.5. Cardiovascular and Death Events

For any cardiovascular events and all deaths, whether or not they are considered SAEs, specific Cardiovascular (CV) and Death sections of the CRF will be required to be completed. These sections include questions regarding cardiovascular (including sudden cardiac death) and non-cardiovascular death.

The CV CRFs are presented as queries in response to reporting of certain CV Medical Dictionary of Regulatory activities (MedDRA) terms. The CV information should be recorded in the specific cardiovascular section of the CRF within one week of receipt of a CV Event data query prompting its completion.

The Death CRF is provided immediately after the occurrence or outcome of death is reported. Initial and follow-up reports regarding death must be completed within one week of when the death is reported.

9.3. Safety Assessments

There are no mandated safety assessments, other than monitoring for AEs and SAEs and

this is described in Section 9.2.

9.4. Pharmacokinetics

PK parameters are not evaluated in this study.

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

27

9.5. Pharmacodynamics

Pharmacodynamic parameters are not evaluated in this study.

9.6. Genetics

Genetics are not evaluated in this study.

9.7. Biomarkers

Biomarkers are not evaluated in this study.

9.8. Health Economics OR Medical Resource Utilization and Health Economics

Health Economics/Medical Resource Utilization and Health Economics parameters are not evaluated in this study.

10. STATISTICAL CONSIDERATIONS

10.1. Hypotheses

The primary purpose of this study is to assess the number of critical errors made by COPD patients, after a subject has read the patient information leaflet(s) (PIL) for each treatment option tested. This is a superiority study.

There are two sub-studies and these will be analysed separately. There is no overlap of subjects and so these are considered independent. Hence there will be no adjustment for multiplicity.

The primary endpoint is the percentage of subjects making at least one critical error after reading the PIL(s) on Treatment Option 1 compared with each of Treatment Option 2 (Sub-study 1) and Treatment Option 3 (Sub-study 2).

For each sub-study, the null hypotheses are no difference between treatment options:

H0: p1=pi ; i=2, 3

The alternative hypothesis is that there is a difference between treatment options.

HA: p1≠pi; i=2,3

10.2. Sample Size Determination

The sample size calculation for each sub-study is based on the primary endpoint, the percentage of subjects making at least one critical error in each device combination (Ellipta or DISKUS - HandiHaler or Turbuhaler - HandiHaler) after reading the patient information leaflet(s).

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

28

Based on the results from studies 201301 and 201330, a range of critical error rates (at least one critical error using each device after reading the patient information leaflet) for each of the treatment options (ELLIPTA, DISKUS - HandiHaler and Turbuhaler -HandiHaler) were explored.

Using 10000 simulations, a total of 72 subjects in each sub-study will provide at least 90% power to show a statistically significant difference between the critical error rate of each of the paired treatment options (Sub-study 1: Treatment option 1 vs. Treatment Option 2; Sub-study 2: Treatment option 1 vs. Treatment Option 3) assuming the following true critical error rates.

Treatment Option 1: Treatment Option 2: Treatment Option 3:ELLIPTA Critical Error Rate

DISKUS - HandiHalerCritical Error rate

Turbuhaler - HandiHalerCritical Error rate

11% >=37% >=37%10% >=35% >=35%9% >=34% >=34%8% >=32% >=32%7% >=31% >=31%6% >=29% >=29%

The error rates for the combined treatment options were based on the studies 201301 and 201330 and a conservative assumption that the critical error rate for the combined treatment was the same as the critical error rate for a single device i.e. subjects who made errors made them on both devices.

Conditional logistic regression and a two-sided 5% significance level were used as the analysis method in the simulations. No withdrawal is expected in this single visit studyhowever up to 80 subjects may be randomised to each sub-study to ensure we have 72 evaluable subjects in each sub-study.

10.3. Populations for Analyses

All analysis will be performed by individual sub-study. Each sub-study data may be analysed when the sub-study has completed and the data have been cleaned. Thedatabase will be locked when all of the sub-studies have completed and data are cleaned.

For purposes of analysis, the following populations are defined for each sub-study:

Population Description

All Subjects Enrolled (ASE) All participants who sign the ICF and for whom a record exists in the study database, including screen failures and any subject who was not screened but experienced an SAE between the date of informed consent and the planned date of the Screening visit

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

29

Population Description

Randomised All participants who were randomised.

Intent-to-treat (ITT) All randomised subjects, excluding those who were randomised in error and made at least one critical error assessment from one treatment option device. A subject who is recorded as a screen or run-in failure and also randomized will be considered to be randomized in error. Any other subject who receives a randomization number will be considered to have been randomized.

Displays will be based on the treatment to which the subject was randomized.

Safety This population will be the same as the Intent-to-treat population.

10.4. Statistical Analyses

10.4.1. Efficacy Analyses

Endpoint Statistical Analysis Methods

Primary The primary endpoint is the percentage of subjects making at least one critical error after reading the PIL(s).

This endpoint will be analysed using conditional logistic regression with subject as fixed strata, and treatment option and period as fixed effects. The odds ratio, 95% CI and p-value will be presented for the comparison between treatment options. It will be based on a two-sided hypothesis testing approach of superiority.

Secondary The following endpoints will be analysed in the same way as for the primary endpoint if there are sufficient subjects making errors otherwise the data will be summarised only:

The percentage of subjects making at least one critical error after thefirst instruction from the HCP


The percentage of subjects making at least one overall error after thefirst instruction from the HCP

The number of instructions (0, 1 or 2 times) from the HCP which are needed to demonstrate correct inhaler use will be analysed using the Wilcoxon signed rank test.

The total amount of time taken to demonstrate correct inhaler use (T1+T2)

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

30

Endpoint Statistical Analysis Methods

will be analysed using Kaplan-Meier methods.

Treatment preference from questionnaire will be analysed using a Cochran-Mantel-Haelszel test, adjusted for study inhaler use sequence.

Other Analysis of the following other endpoints will be defined in the Reporting and Analysis Plan.

The percentage of subjects making at least one critical error after the second instruction from the HCP


The amount of time taken to read the patient information leaflet and demonstrate correct inhaler device use (T1)

The amount of time taken to be given instruction by the HCP on use of the inhaler and demonstrate correct inhaler use (T2)

10.4.2. Safety Analyses

All safety analyses will be performed on the Safety Population.

Adverse events (AEs) will be coded using the standard GSK dictionary, MedicalDictionary for Regulatory Activities (MedDRA), and grouped by body system. Thenumber and percentage of subjects experiencing at least one AE of any type, AEs withineach body system and AEs within each preferred term will be presented for eachtreatment group. Separate summaries will be provided for all AEs, drug related AEs, fatalAEs, non-fatal SAEs, AESIs and AEs leading to withdrawal.

Deaths and SAEs, if applicable, will be documented in case narrative format.

10.4.3. Other Analyses

Any other analyses will be described in the reporting and analysis plan

10.4.4. Interim Analyses

No interim analyses are planned.

The protocol consists of two sub-studies to assess the following treatment comparisons: ELLIPTA DPI (treatment option 1) versus the combination of DISKUS-HandiHaler (treatment option 2) in Sub-study 1 and ELLIPTA DPI (treatment option 1) versus the combination of Turbuhaler-HandiHaler (treatment option 3) in Sub-study 2.

Sub-study 1 will report out independently of Sub-study 2.

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

31

11. REFERENCES

Bonini M, Usmani O. The importance of inhaler devices in the treatment of COPD. COPD Research and Practice (2015) 1:9

Celli BR, MacNee W. Standards of the diagnosis and treatment of patients with COPD: a summary of the ATS/ERS position paper. Eur Respir J. 2004;23:932-46.

Cochrane M, Bala M, Downs K, Mauskopf J, Ben-Joseph R. Inhaled Corticosteroids for Asthma Therapy:Patient Compliance, Devices, and Inhalation Technique. CHEST 2000; 117:542–550

Global Initiative for Chronic Obstructive Lung Disease (GOLD 2016). Global strategy for the diagnosis, management, and prevention of COPD (Updated 2016). Available from http://www.goldcopd.org

Komase Y, Asako A, Kobyyashi A, Sharma R. E Ease-of-use preference for the ELLIPTA® dry powder inhaler over a commonly used single-dose capsule dry powder inhaler by inhalation device-naïve Japanese volunteers aged 40 years or older. International Journal of COPD 2014:9

Melani A, Bonavia M, Cilenti V, Cinti C, Lodi M, Martucci P. Inhaler mishandling remains common in real life and is associated with reduced disease control. Respiratory Medicine (2011) 105, 930-938

Van der PalenJ, Klein J, Van Herwaarden ,Zielhuis, Seydel E. Multiple inhalers confuse asthma patients. Eur Respir J 1999; 14: 1034-1037

Van Der Palen J, Thomas M, Chrystyn H, Sharma R, Imber V, Zhu C.-Q, BarnesN. . Inhaler Errors After Reading The Patient Information Leaflet In Patients With COPD: A Comparison Of Ellipta® With Five Inhaler Devices. Am J Respir Crit Care Med 193; 2016:A6811 (a)

Van Der Palen J, Thomas M, Chrystyn H, Sharma R, Svedsater H, Zhu C.-Q, Barnes N. Inhaler Preference In Patients With COPD: A Comparison Of Ellipta® With Five Inhaler Devices. Am J Respir Crit Care Med 193; 2016:A6813 (b)

Van Der Palen J, Thomas M, Chrystyn H, Sharma R, Imber V , Zhu C.-Q, Svedsater H . Training And Time To Achieve Correct Inhaler Use: A Comparison Between Inhalers In Patients With COPD. Am J Respir Crit Care Med 193; 2016:A6811 (c)

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

32

12. APPENDICES

12.1. Appendix 1: Abbreviations and Trademarks

Abbreviations

ASE All Subjects Enrolled COPD Chronic Obstructive Pulmonary DiseaseCONSORT Consolidated Standards of Reporting TrialsCRF Case Report FormCV CardiovascularDPI Dry Powder InhalerHCP Healthcare ProfessionalICS Inhaled CorticosteroidsIEC Independent Ethics CommitteesIRB Institutional Review BoardsITT Intent-to-treatIWRS Interactive Web Response SystemLABA Long-Acting Beta2-AgonistLAMA Long-Acting AnticholinergicMedRA Medical Dictionary of Regulatory activitiesPIL Patient Instruction LeafletSABA Short Acting Beta-Adrenergic AgonistSAE Serious Adverse EventsSAMA Short Acting Muscarinic AntagonistSRM Study Reference ManualSUSAR Suspected Unexpected Serious Adverse Reactions

Trademark Information

Trademarks of the GlaxoSmithKline group of companies

Trademarks not owned by the GlaxoSmithKline group of companies

ANORO ELLIPTA SASARNUITY ELLIPTA Seebri BreezhalerINCRUSE ELLIPTA Spiriva HandiHalerRELVAR ELLIPTA Symbicort TurbuhalerSERETIDE DISKUS Symbicort TurbohalerBREO ELLIPTA Ultibro BreezhalerSERETIDE ACCUHALER

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

33

12.2. Appendix 2: Inhaler Specific Errors for the inhalers used in the study

There are no universally agreed checklists that define critical errors and over all errors for specific inhalers. The checklist and critical errors, for use in this study have been developed by GSK for each inhaler based upon:

A review of the patient information leaflet for each inhaler and the steps defined therein for correct use

The available literature which is exhaustive for a number of the commonly used inhalers

Review of these errors with a group of external inhaler experts

The critical errors checklist are, therefore, as robust as possible. Furthermore, GSK has used selected sites with trained assessors to ensure as much consistency as possible in the valuation of errors in study subjects.

Checklist of instructions for correct use will be based on the steps listed in the patient information leaflets/package insert for each device. Some of the steps outlined in the patient information leaflets/package inserts require several actions to be identified and checked by the HCP.

Critical Errors are identified in bold text in the list below. A critical error is defined as an error that is most likely to result in no or significantly reduced medication being inhaled.

These errors will be captured in a checklist provided for HCP assessment of DPI use.

DISKUS/ACCUHALERFailed to open coverLever is not pushed back Shook the device after dose preparationNo exhalation before an inhalationExhaled directly into mouthpieceNo seal by the lips round the mouthpiece during the inhalationInhalation manoeuvre:

- steady- deep

Did not hold breathDid not close the device (Note: this is an error but one which does not affect the medication that is inhaled)Any other comments: [free text box]

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

34

ELLIPTAFailed to open coverShook the device after dose preparationNo exhalation before an inhalationExhaled directly into mouthpieceNo seal by the lips round the mouthpiece during the inhalationInhalation manoeuvre:

- long- steady- deep

Blocked air inlet during inhalation manoeuvreDid not hold breathDid not close the device (Note: this is an error but one which does not affect the medication that is inhaled)Any other comments: [free text box]

TURBUHALERFailed to remove capDid not hold device upright (±45% OK) during dose preparationBase not twisted fully backwards and forwards, no click heardDevice tipped downwards after dose preparationShook the device after dose preparationNo exhalation before an inhalationExhaled directly into mouthpieceNo seal by the lips round the mouthpiece during the inhalationInhalation manoeuvre:

- forceful- deep

Note to HCP: it is important that the inhalation is forceful and deep from the start for this inhalerBlocked air inlet during inhalation manoeuvreDid not hold breathDid not close the device (Note: this is an error but one which does not affect the medication that is inhaled)Any other comments: [free text box]

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

35

HANDIHALERFailed to remove capsuleFailed to insert capsule into the chamberDid not completely close device capsule chamber (heard click when satisfactory)Did not pierce the capsule (HCP should check capsule was pierced)Shook the device after dose preparation No exhalation before an inhalationExhaled directly into mouthpieceNo seal by the lips round the mouthpiece during the inhalationInhalation manoeuvre:

- slow- deep

Capsule did not rattle Blocked air inlet during inhalation manoeuvreDid not hold breathDid not check inside the capsule chamber if powder was left / did not make a second inhalationAny other comments: [free text box]

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

36

12.3. Appendix 3: Inhaler Preference Questionnaires

Inhaler Preference Questionnaire Version 1

Instructions: Please complete the following questions related to the treatmentsyou used during this study. Check only one response for each question.

1. Which treatment do you prefer based on the number of steps needed to take your COPD medication?

� DISKUS and HandiHaler � ELLIPTA � No preference

2. Which treatment do you prefer for taking your COPD medication?

� DISKUS and HandiHaler � ELLIPTA � No preference




� ELLIPTA� DISKUS and HandiHaler� No preference


� ELLIPTA� DISKUS and HandiHaler� No preference

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

37




� Turbuhaler and HandiHaler � ELLIPTA � No preference


� Turbuhaler and HandiHaler� ELLIPTA � No preference




� ELLIPTA � DISKUS and HandiHaler� No preference


� ELLIPTA � DISKUS and HandiHaler� No preference

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

38

12.4. Appendix 4: Study Governance Considerations

Regulatory and Ethical Considerations

This study will be conducted in accordance with the protocol and with:

Consensus ethical principles derived from international guidelines including the Declaration of Helsinki and Council for International Organizations of Medical Sciences (CIOMS) International Ethical Guidelines

Applicable ICH Good Clinical Practice (GCP) Guidelines

Applicable laws and regulations

The protocol, protocol amendments, ICF, Investigator Brochure, and other relevant documents (eg, advertisements) must be submitted to an IRB/IEC by the investigator and reviewed and approved by the IRB/IEC before the study is initiated.

Any amendments to the protocol will require IEC/IRB approval before implementation of changes made to the study design, except for changes necessary to eliminate an immediate hazard to study participants.

The investigator will be responsible for the following:

Providing written summaries of the status of the study to the IRB/IEC annually or more frequently in accordance with the requirements, policies, and procedures established by the IRB/EC

Notifying the IRB/IEC of SAE or other significant safety findings as required by IRB/IEC procedures

Providing oversight of the conduct of the study at the site and adherence to requirements of 21 CFR, ICH guidelines, the IRB/IEC, European regulation 536/2014 for clinical studies (if applicable), and all other applicable local regulations

Financial Disclosure

Investigators and sub-investigators will provide the sponsor with sufficient, accurate financial information as requested to allow the sponsor to submit complete and accurate financial certification or disclosure statements to the appropriate regulatory authorities.Investigators are responsible for providing information on financial interests during the course of the study and for 1 year after completion of the study.

Informed Consent Process

The investigator or his/her representative will explain the nature of the study to the participant or his/her legally authorized representative and answer all questions regarding the study.

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

39

Participants must be informed that their participation is voluntary. Participants or their legally authorized representative will be required to sign a statement of informed consent that meets the requirements of 21 CFR 50, local regulations, ICH guidelines, Health Insurance Portability and Accountability Act (HIPAA) requirements, where applicable, and the IRB/IEC or study centre.

The medical record must include a statement that written informed consent was obtained before the participant was enrolled in the study and the date the written consent was obtained. The authorized person obtaining the informed consent must also sign the ICF.

Participants must be re-consented to the most current version of the ICF(s) during their participation in the study.

A copy of the ICF(s) must be provided to the participant or the participant’s legally authorized representative.

Data Protection

Participants will be assigned a unique identifier by the sponsor. Any participant records or datasets that are transferred to the sponsor will contain the identifier only; participant names or any information which would make the participant identifiable will not be transferred.

The participant must be informed that his/her personal study-related data will be used by the sponsor in accordance with local data protection law. The level of disclosure must also be explained to the participant.

The participant must be informed that his/her medical records may be examined by Clinical Quality Assurance auditors or other authorized personnel appointed by the sponsor, by appropriate IRB/IEC members, and by inspectors from regulatory authorities.

Publication Policy

The results of this study may be published or presented at scientific meetings. If this is foreseen, the investigator agrees to submit all manuscripts or abstracts to the sponsor before submission. This allows the sponsor to protect proprietary information and to provide comments.

The sponsor will comply with the requirements for publication of study results. In accordance with standard editorial and ethical practice, the sponsor will generally support publication of multicenter studies only in their entirety and not as individual site data. In this case, a coordinating investigator will be designated by mutual agreement.

Authorship will be determined by mutual agreement and in line with International Committee of Medical Journal Editors authorship requirements.

Dissemination of Clinical Study Data

Where required by applicable regulatory requirements, an investigator signatory will be identified for the approval of the clinical study report. The investigator will be provided

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

40

reasonable access to statistical tables, figures, and relevant reports and will have the opportunity to review the complete study results at a GSK site or other mutually-agreeable location.

GSK will also provide the investigator with the full summary of the study results. The investigator is encouraged to share the summary results with the study subjects, as appropriate.

The procedures and timing for public disclosure of the results summary and for development of a manuscript for publication will be in accordance with GSK Policy.

Data Quality Assurance

All participant data relating to the study will be recorded on printed or electronic CRF unless transmitted to the sponsor or designee electronically (eg, laboratory data). The investigator is responsible for verifying that data entries are accurate and correct by physically or electronically signing the CRF.

The investigator must maintain accurate documentation (source data) that supports the information entered in the CRF.

The investigator must permit study-related monitoring, audits, IRB/IEC review, and regulatory agency inspections and provide direct access to source data documents.

The sponsor or designee is responsible for the data management of this study including quality checking of the data.

Study monitors will perform ongoing source data verification to confirm that data entered into the CRF by authorized site personnel are accurate, complete, and verifiable from source documents; that the safety and rights of participants are being protected; and that the study is being conducted in accordance with the currently approved protocol and any other study agreements, ICH GCP, and all applicable regulatory requirements.

Records and documents, including signed ICF, pertaining to the conduct of this study must be retained by the investigator for 25 years after study completion unless local regulations or institutional policies require a longer retention period. No records may be destroyed during the retention period without the written approval of the sponsor. No records may be transferred to another location or party without written notification to the sponsor.

Source Documents

Source documents provide evidence for the existence of the participant and substantiate the integrity of the data collected. Source documents are filed at the investigator’s site.

Data reported on the CRF or entered in the eCRF that are transcribed from source documents must be consistent with the source documents or the discrepancies must be explained. The investigator may need to request previous

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

41

medical records or transfer records, depending on the study. Also, current medical records must be available.

Definition of what constitutes source data can be found in source data agreement or source data verification form

Study and Site Closure

GSK or its designee reserves the right to close the study site or terminate the study at any time for any reason at the sole discretion of GSK. Study sites will be closed upon study completion. A study site is considered closed when all required documents and study supplies have been collected and a study-site closure visit has been performed.

The investigator may initiate study-site closure at any time, provided there is reasonable cause and sufficient notice is given in advance of the intended termination.

Reasons for the early closure of a study site by the sponsor or investigator may include but are not limited to:

Failure of the investigator to comply with the protocol, the requirements of the IRB/IEC or local health authorities, the sponsor's procedures, or GCP guidelines

Inadequate recruitment of participants by the investigator

Discontinuation of further study treatment development

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

42

12.5. Appendix 5: Adverse Events: Definitions and Procedures for Recording, Evaluating, Follow-up, and Reporting

Definition of AE

AE Definition

An AE is any untoward medical occurrence in a clinical study participant, temporally associated with the use of a study treatment, whether or not considered related to the study treatment.

NOTE: An AE can therefore be any unfavorable and unintended sign (including an abnormal laboratory finding), symptom, or disease (new or exacerbated) temporally associated with the use of a study treatment.

Events Meeting the AE Definition

Any abnormal laboratory test results (hematology, clinical chemistry, or urinalysis) or other safety assessments (eg, ECG, radiological scans, vital signs measurements), including those that worsen from baseline, considered clinically significant in the medical and scientific judgment of the investigator (ie, not related to progression of underlying disease).

Exacerbation of a chronic or intermittent pre-existing condition including either an increase in frequency and/or intensity of the condition.

New conditions detected or diagnosed after study treatment administration even though it may have been present before the start of the study.

Signs, symptoms, or the clinical sequelae of a suspected drug-drug interaction.

Signs, symptoms, or the clinical sequelae of a suspected overdose of either study treatment or a concomitant medication. Overdose per se will not be reported as an AE/SAE unless it is an intentional overdose taken with possible suicidal/self-harming intent. Such overdoses should be reported regardless of sequelae.

"Lack of efficacy" or "failure of expected pharmacological action" per se will not be reported as an AE or SAE. Such instances will be captured in the efficacy assessments. However, the signs, symptoms, and/or clinical sequelae resulting from lack of efficacy will be reported as AE or SAE if they fulfill the definition of an AE or SAE.

The signs, symptoms, and/or clinical sequelae resulting from lack of efficacy will be reported as AE or SAE if they fulfill the definition of an AE or SAE. Also, "lack of efficacy" or "failure of expected pharmacological action" constitutes an AE or SAE.

Events NOT Meeting the AE Definition

Any clinically significant abnormal laboratory findings or other abnormal safety

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

43

assessments which are associated with the underlying disease, unless judged by the investigator to be more severe than expected for the participant’s condition.

The disease/disorder being studied or expected progression, signs, or symptoms of the disease/disorder being studied, unless more severe than expected for the participant’s condition.

Medical or surgical procedure (eg, endoscopy, appendectomy): the condition that leads to the procedure is the AE.

Situations in which an untoward medical occurrence did not occur (social and/or convenience admission to a hospital).

Anticipated day-to-day fluctuations of pre-existing disease(s) or condition(s) present or detected at the start of the study that do not worsen.

Definition of SAE

If an event is not an AE per definition above, then it cannot be an SAE even if serious conditions are met (eg, hospitalization for signs/symptoms of the disease under study, death due to progression of disease).

A SAE is defined as any untoward medical occurrence that, at any dose:

a. Results in death

b. Is life-threatening

The term 'life-threatening' in the definition of 'serious' refers to an event in which the participant was at risk of death at the time of the event. It does not refer to an event, which hypothetically might have caused death, if it were more severe.

c. Requires inpatient hospitalization or prolongation of existing hospitalization

In general, hospitalization signifies that the participant has been detained (usually involving at least an overnight stay) at the hospital or emergency ward for observation and/or treatment that would not have been appropriate in the physician’s office or outpatient setting. Complications that occur during hospitalization are AE. If a complication prolongs hospitalization or fulfills any other serious criteria, the event is serious. When in doubt as to whether “hospitalization” occurred or was necessary, the AE should be considered serious.

Hospitalization for elective treatment of a pre-existing condition that did not worsen from baseline is not considered an AE.

d. Results in persistent disability/incapacity

The term disability means a substantial disruption of a person’s ability to conduct normal life functions.

This definition is not intended to include experiences of relatively minor medical significance such as uncomplicated headache, nausea, vomiting, diarrhea, influenza,

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

44

and accidental trauma (eg, sprained ankle) which may interfere with or prevent everyday life functions but do not constitute a substantial disruption.

e. Is a congenital anomaly/birth defect

f. Other situations:

Medical or scientific judgment should be exercised in deciding whether SAE reporting is appropriate in other situations such as important medical events that may not be immediately life-threatening or result in death or hospitalization but may jeopardize the participant or may require medical or surgical intervention to prevent one of the other outcomes listed in the above definition. These events should usually be considered serious.

Examples of such events include invasive or malignant cancers, intensive treatment in an emergency room or at home for allergic bronchospasm, blood dyscrasias or convulsions that do not result in hospitalization, or development of drug dependency or drug abuse.

Recording AE and SAE

AE and SAE Recording

When an AE/SAE occurs, it is the responsibility of the investigator to review all documentation (eg, hospital progress notes, laboratory, and diagnostics reports) related to the event.

The investigator will then record all relevant AE/SAE information in the CRF.

It is not acceptable for the investigator to send photocopies of the participant’s medical records to GSK in lieu of completion of the GSK /AE/SAE CRF page.

There may be instances when copies of medical records for certain cases are requested by GSK. In this case, all participant identifiers, with the exception of the participant number, will be redacted on the copies of the medical records before submission to GSK.

The investigator will attempt to establish a diagnosis of the event based on signs, symptoms, and/or other clinical information. Whenever possible, the diagnosis (not the individual signs/symptoms) will be documented as the AE/SAE.

Assessment of Intensity

The investigator will make an assessment of intensity for each AE and SAE reported during the study and assign it to 1 of the following categories:

Mild: An event that is easily tolerated by the participant, causing minimal discomfort and not interfering with everyday activities.

Moderate: An event that causes sufficiently discomfort and interferes with normal everyday activities.

Severe: An event that prevents normal everyday activities. An AE that is assessed as

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

45

severe should not be confused with an SAE. Severe is a category utilized for rating the intensity of an event; and both AE and SAE can be assessed as severe.

An event is defined as ‘serious’ when it meets at least 1 of the predefined outcomes as described in the definition of an SAE, NOT when it is rated as severe.

Assessment of Causality

The investigator is obligated to assess the relationship between study treatment and each occurrence of each AE/SAE.

A "reasonable possibility" of a relationship conveys that there are facts, evidence, and/or arguments to suggest a causal relationship, rather than a relationship cannot be ruled out.

The investigator will use clinical judgment to determine the relationship.

Alternative causes, such as underlying disease(s), concomitant therapy, and other risk factors, as well as the temporal relationship of the event to study treatment administration will be considered and investigated.

The investigator will also consult the Investigator’s Brochure (IB) and/or Product Information, for marketed products, in his/her assessment.

For each AE/SAE, the investigator must document in the medical notes that he/she has reviewed the AE/SAE and has provided an assessment of causality.

There may be situations in which an SAE has occurred and the investigator has minimal information to include in the initial report to GSK. However, it is very important that the investigator always make an assessment of causality for every event before the initial transmission of the SAE data to GSK.

The investigator may change his/her opinion of causality in light of follow-up information and send an SAE follow-up report with the updated causality assessment.

The causality assessment is one of the criteria used when determining regulatory reporting requirements.

Follow-up of AE and SAE

The investigator is obligated to perform or arrange for the conduct of supplemental measurements and/or evaluations as medically indicated or as requested by GSK to elucidate the nature and/or causality of the AE or SAE as fully as possible. This may include additional laboratory tests or investigations, histopathological examinations, or consultation with other health care professionals.

If a participant dies during participation in the study or during a recognized follow-up period, the investigator will provide GSK with a copy of any post-mortem findings including histopathology.

2017N324761_00

2016N292801_01 CONFIDENTIAL206215

46

New or updated information will be recorded in the originally completed CRF.

The investigator will submit any updated SAE data to GSK within 24 hours of receipt of the information.

Reporting of SAE to GSK

SAE Reporting to GSK via Electronic Data Collection Tool

The primary mechanism for reporting SAE to GSK will be the electronic data collection tool.

If the electronic system is unavailable for more than 24 hours, then the site will use the paper SAE data collection tool (see next section).

The site will enter the SAE data into the electronic system as soon as it becomes available.

After the study is completed at a given site, the electronic data collection tool will be taken off-line to prevent the entry of new data or changes to existing data.

If a site receives a report of a new SAE from a study participant or receives updated data on a previously reported SAE after the electronic data collection tool has been taken off-line, then the site can report this information on a paper SAE form (see next section) or to the medical monitor by telephone.

Contacts for SAE reporting can be found in the SRM.

SAE Reporting to GSK via Paper CRF

Facsimile transmission, or a scanned version sent by email, of the SAE paper CRF is the preferred method to transmit this information to the medical monitor.

In rare circumstances and in the absence of facsimile equipment, notification by telephone is acceptable with a copy of the SAE data collection tool sent by overnight mail or courier service.

Initial notification via telephone does not replace the need for the investigator to complete and sign the SAE CRF pages within the designated reporting time frames.

Contacts for SAE reporting can be found in the SRM.

2017N324761_00

CONFIDENTIALThe GlaxoSmithKline group of companies 206215

1

Division : Worldwide Development

Information Type : Reporting and Analysis Plan

Title : Reporting and Analysis Plan for Study 206215: A randomized, open-label, cross-over, placebo-device study investigating critical and over all errors, training/teaching time, and preference attributes of the ELLIPTA dry powder Inhaler (DPI) as compared to HandiHaler DPI used in combination with either DISKUS DPI or Turbuhaler DPI, in adult patients with Chronic Obstructive Pulmonary Disease (COPD)

Compound Number : GSK573719+GW642444+GW685698 (GSK2834425)

Effective Date : 18-JUL-2017

Description :

The purpose of this Reporting and Analysis Plan (RAP) is to describe the planned analyses and output to be included in the Clinical Study Report for Protocol 2016N292801_01.

This RAP is intended to describe the planned efficacy and safety analyses required for the study.

This document will be provided to the study team members to convey the content of the final Statistical Analysis Complete (SAC) deliverable.

Author’s Name and Functional Area:

18-JUL-2017Statistician (Clinical Statistics)

Approved by:

18-JUL-2017Manager, Statistics (Clinical Statistics)


2017N324761_00

PPD

PPD

CONFIDENTIAL206215

2

TABLE OF CONTENTS

PAGE

1. REPORTING & ANALYSIS PLAN SYNPOSIS.........................................................4

2. SUMMARY OF KEY PROTOCOL INFORMATION ..................................................62.1. Changes to the Protocol Defined Statistical Analysis Plan ............................62.2. Study Objective(s) and Endpoint(s)...............................................................72.3. Study Design ................................................................................................82.4. Statistical Hypotheses...................................................................................9

3. PLANNED ANALYSES ..........................................................................................103.1. Final Analyses ............................................................................................10

4. ANALYSIS POPULATIONS ...................................................................................104.1. Protocol Deviations.....................................................................................11

5. CONSIDERATIONS FOR DATA ANALYSES AND DATA HANDLING CONVENTIONS.....................................................................................................11

6. STUDY POPULATION ANALYSES .......................................................................126.1. Disposition ..................................................................................................136.2. Medical Conditions .....................................................................................136.3. Concomitant Medications............................................................................13

7. PRIMARY STATISTICAL ANALYSES....................................................................137.1. Efficacy Analyses........................................................................................13

7.1.1. Overview of Planned Efficacy Analyses .......................................137.1.2. Planned Efficacy Statistical Analyses...........................................14

8. SECONDARY STATISTICAL ANALYSES .............................................................158.1. Efficacy Analyses........................................................................................15

8.1.1. Overview of Planned Secondary Efficacy Analyses .....................158.1.2. Planned Secondary Efficacy Statistical Analyses.........................16

8.1.2.1. Number of critical errors made after 1st

Healthcare Professional (HCP) instructions................168.1.2.2. Number of overall errors made ...................................168.1.2.3. Number of instructions (maximum of 2) from

the HCP......................................................................168.1.2.4. Training/Teaching time required .................................178.1.2.5. Preference attributes ..................................................18

9. OTHER STATISTICAL ANALYSES .......................................................................189.1. Efficacy Analyses........................................................................................18

9.1.1. Overview of Planned Other Efficacy Analyses .............................189.1.1.1. Number of critical errors/overall errors made

after 2nd Healthcare Professional (HCP) instructions .................................................................19

9.1.1.2. Training/Teaching time required .................................209.2. Safety Analyses ..........................................................................................20

9.2.1. Overview of Planned Safety Analyses .........................................209.2.1.1. Adverse Events and Serious Adverse Events.............21

2017N324761_00

CONFIDENTIAL206215

3

9.2.1.2. Cardiovascular Events................................................229.2.2. Pregnancies.................................................................................22

10. REFERENCES.......................................................................................................23

11. APPENDICES ........................................................................................................2411.1. Appendix 1: Protocol Deviation Management .............................................2511.2. Appendix 2: Time & Events.........................................................................26

11.2.1. Protocol Defined Time & Events ..................................................2611.3. Appendix 3: Treatment Phases...................................................................28

11.3.1. Treatment Phases .......................................................................2811.3.2. Treatment Phase Definitions........................................................28

11.4. Appendix 4: Data Display Standards & Handling Conventions....................2911.4.1. Study Treatment & Sub-group Display Descriptors ......................2911.4.2. Reporting Process & Standards...................................................29

11.5. Appendix 5: Derived and Transformed Data ...............................................3111.5.1. Study Population..........................................................................3111.5.2. Date of Completion and Withdrawal.............................................3111.5.3. Critical and Overall Errors............................................................3111.5.4. Responses to Preference Questions............................................3211.5.5. Number of instructions.................................................................32

11.6. Appendix 6: Premature Withdrawals & Handling of Missing Data ...............3311.6.1. Premature Withdrawals................................................................3311.6.2. Handling of Missing Data .............................................................33

11.6.2.1. Handling of Missing/Partial Dates ...............................3411.6.2.2. Handling of Missing Data for Statistical

Analysis......................................................................3511.7. Appendix 7: Multicenter Studies..................................................................36

11.7.1. Methods for Handling Centres .....................................................3611.8. Appendix 8: Examination of Covariates, Subgroups & Other Strata............37

11.8.1. Handling of Covariates.................................................................3711.9. Appendix 9: Model Checking and Diagnostics for Statistical

Analyses .....................................................................................................3811.10. Appendix 10: Abbreviations & Trade Marks ................................................39

11.10.1. Abbreviations...............................................................................3911.10.2. Trademarks .................................................................................39

11.11. Appendix 11: List of Data Displays..............................................................4011.11.1. Study Population Tables ..............................................................4111.11.2. Efficacy Tables ............................................................................4311.11.3. Safety Tables...............................................................................4511.11.4. Efficacy Figures ...........................................................................4611.11.5. ICH Listings .................................................................................4711.11.6. Non ICH Listings..........................................................................49

11.12. Appendix 12: Example Mock Shells for Data Displays ................................51

2017N324761_00

CONFIDENTIAL206215

4

1. REPORTING & ANALYSIS PLAN SYNPOSIS

Overview Key Elements of the RAPPurpose This RAP details all planned analyses and outputs required for the final Clinical

Study Report (CSR) of study 206215.

Protocol This RAP is based on the original protocol [(Dated: 15-Sep-2016) of study 206215 (GSK Document No. : 2016N292801_01] and eCRF Version 1

Primary Objective

To compare the number of critical errors made by COPD patients, after a subject has read the respective patient information leaflet(s) (PIL), for each treatment option tested.

Primary Endpoint

The percentage of subjects making at least one critical error after reading the PIL(s).

Secondary Objectives

To compare the number of critical errors made by COPD patients after instructions from the Healthcare Professional (HCP) for each treatment option tested.

To compare the number of overall (critical and non-critical) errors made by COPD patients, after a subject has read the PIL(S) or after Instruction from the HCP for each treatment option tested



Preference attributes for each treatment option tested

Secondary Endpoints

The percentage of subjects making at least one critical error after the first instruction from the HCP





Treatment preference, from questionnaire for:

Number of steps required to take COPD medication


Study Design

The study will be conducted as a multi-centre, randomised, open-label, placebo-device, cross-over study, with a 2x2 complete block design. The study will comprise of 2 sub studies. Sub study 1 will compare ELLIPTA to DISKUS + HandiHaler and Sub Study 2 will compare ELLIPTA to Turbuhaler + Handihaler.

2017N324761_00

CONFIDENTIAL206215

5

Overview Key Elements of the RAP

Subjects will be randomised 1:1:1:1 to each of the four treatment sequences in both sub study 1 and sub study 2. The randomisation schedules for each sub study will be created independently. 160 participants will be randomised (80 in each sub study), such that approximately 144 are evaluable (72 in each sub study).

The study has three treatment periods. The first two treatment periods are a 2x2 complete block design to test the ELLIPTA against DISKUS + HandiHaler or Turbuhaler + HandiHaler.The third treatment period contains a questionnaire to assess subject preference.

Using 10000 simulations, a total of 72 subjects in each sub-study will provide at least 90% power to show a statistically significant difference between the critical error rate of each of the paired treatment options (Sub-study 1: Treatment option 1 vs. Treatment Option 2; Sub-study 2: Treatment option 1 vs. Treatment Option 3) assuming the following true critical error rates.

Treatment Option 1: Treatment Option 2: Treatment Option 3:ELLIPTA Critical Error Rate

DISKUS + HandiHaler Critical Error Rate

Turbuhaler + HandiHaler Critical Error Rate

11% >=37% >=37%10% >=35% >=35%9% >=34% >=34%8% >=32% >=32%7% >=31% >=31%6% >=29% >=29%

Planned Analyses

No interim analyses are planned for this study.

Sub study 1 and sub study 2 will be reported at the same time.

All decisions regarding final analysis as defined in this RAP document will be made prior to Database Freeze.

Analysis Populations

Subjects Enrolled population (ASE) will consist of all participants who sign the ICF and for whom a record exists in the study database, including screen failures and any subject who was not screened but experienced an SAE between the date of informed consent and the planned date of the Screening visit. All Subjects Enrolled population (ASE) will be used for subject disposition, reason for withdrawal prior to randomisation, inclusion, exclusion and randomisation criteria deviations and SAEs for non randomised subjects.

The Intent-to-Treat population will consist of all randomised subjects, excluding those who were randomised in error and made at least one critical error assessment from one treatment option device. A subject who is recorded as a screen or run-in failure and also randomized will be considered to be randomized in error. Any other subject who receives a randomization number will be considered to have been randomized. The Intent-to-treat population (ITT) will be used for study population, efficacy and safety endpoints.

The Safety population will be the same as the ITT population.

2017N324761_00

CONFIDENTIAL206215

6

Overview Key Elements of the RAPHypothesis For each sub-study, the null hypotheses are no difference between treatment

options:H0: p1=pi; i=2, 3

The alternative hypothesis is that there is a difference between treatment options:HA: p1≠pi; i=2,3

Where treatment p1 = ELLIPTA, p2 = DISKUS + HandiHaler and p3 = Turbuhaler + HandiHaler

Primary Analyses

The primary endpoint of the percentage of subjects making at least one critical error after reading the PIL(s) for each treatment option tested will be analysed using conditional logistic regression with subject as fixed strata, and treatment option and period as fixed effects.

Secondary Analyses

The number of critical errors made by COPD patients after instruction from the Healthcare Professional (HCP) for each treatment option tested will be analysed using conditional logistic regression with subject as fixed strata, and treatment option and period as fixed effects.

The number of overall (critical and non-critical) errors made by COPD patients, after a subject has read the PIL(S) or after Instruction from the HCP for each treatment option tested, will be analysed using conditional logistic regression with subject as fixed strata, and treatment option and period as fixed effects.

The number of instructions (maximum of 2) from a HCP which is needed to demonstrate correct inhaler use will be analysed using Wilcoxon sum rank test.

The Training/Teaching Time required to demonstrate correct inhaler use will be analysed using Kaplan-Meier plots and summary statistics.

Preference attributes for each treatment option tested will be analysed using the Cochran-Mantel-Haenszel test.

2. SUMMARY OF KEY PROTOCOL INFORMATION

2.1. Changes to the Protocol Defined Statistical Analysis Plan

There is one change to the originally planned statistical analysis specified in the protocol (Dated: 15-Sep-2016). For the secondary endpoint of the number of instructions (0, 1 or 2 times) from the HCP which are needed to demonstrate correct inhaler use, tracked changes were not copied through to the final document. As such the analysis stated in the protocol (logistic regression) is not appropriate and instead the analysis will be performed using Wilcoxon signed rank test as this was the originally planned analysis.

2017N324761_00

CONFIDENTIAL206215

7

2.2. Study Objective(s) and Endpoint(s)

Objectives EndpointsPrimary Objectives Primary Endpoints To compare the number of critical

errors made by COPD patients, after a subject has read the respective patient information leaflet(s) (PIL), for each treatment option tested


Secondary Objectives Secondary Endpoints

To compare the number of criticalerrors made by COPD patients after instruction from the Healthcare Professional (HCP) for each treatment option tested

The percentage of subjects making at least one critical error after the:

first instruction from the HCP second instruction from the HCP

To compare the number of overall (critical and non-critical) errors made by COPD patients, after a subject has read the PIL(S) or after Instruction from the HCP for each treatment option tested








The amount of time taken to read the patient information leaflet and demonstrate correct inhaler use (T1)The amount of time taken to be given instruction by the HCP (up to 2 times) on use of the inhaler and to demonstrate correct inhaler use (T2)

Preference attributes for eachtreatment option tested

Treatment preference, from questionnaire for: Number of steps required to take COPD medication Over all treatment preference

2017N324761_00

CONFIDENTIAL206215

8

2.3. Study Design

Overview of Study Design and Key Features

DesignFeatures

The study will be conducted as a multi-centre, randomised, open-label, placebo-device, cross-over study, with a 2x2 complete block design.

It will comprise of two sub studies: Sub-study 1 will compare ELLIPTA (Treatment option 1) to DISKUS +

HandiHaler combination (Treatment option 2)

Sub-study 2 will compare ELLIPTA (Treatment option 1) to Turbuhaler +

HandiHaler combination (Treatment option 3).

The study has 2 visits (V0 and V1) and both can be completed on the same day.Each sub-study may run independently or in parallel of the other sub study. Each will start dependent on the availability of the placebo DPIs required for that sub-study. The data from the 2 sub studies may be reported independently and as each sub study completes and the data for that sub study have been cleaned. The database will be locked when all of the sub-studies have completed and all data are cleaned.

Dosing Not applicable as this is a placebo study.

2017N324761_00

CONFIDENTIAL206215

9

Overview of Study Design and Key FeaturesTreatment Assignment

Sub-study 1 Treatment Sequences



Sub-study 2 Treatment Sequence


Questionnaire

E ELLIPTA Turbuhaler + HandiHaler 3

F Turbuhaler + HandiHaler ELLIPTA 4G ELLIPTA Turbuhaler + HandiHaler 4H Turbuhaler + HandiHaler ELLIPTA 3

2.4. Statistical Hypotheses

The primary purpose of this study is to assess the number of critical errors made by COPD patients, after a subject has read the patient information leaflet(s) (PIL) for each treatment option tested. This is a superiority study.

There are two sub-studies and these will be analysed separately. There is no overlap of subjects and so these are considered independent. Hence there will be no adjustment for multiplicity.

The primary endpoint is the percentage of subjects making at least one critical error after reading the PIL(s) on Treatment Option 1 compared with each of Treatment Option 2 (Sub-study 1) and Treatment Option 3 (Sub-study 2).

For each sub-study, the null hypotheses are no difference between treatment options:H0: p1=pi; i=2, 3

The alternative hypothesis is that there is a difference between treatment options.HA: p1≠pi; i=2,3

Where treatment p1 = ELLIPTA, p2 = DISKUS + HandiHaler and p3 = Turbuhaler + HandiHaler

2017N324761_00

CONFIDENTIAL206215

10

3. PLANNED ANALYSES

3.1. Final Analyses

No interim analysis is planned.

The final planned primary analyses will be performed after the completion of the following sequential steps:

1. All subjects have completed the study as defined in the protocol for both sub-study 1and sub-study 2.

2. All required database cleaning activities have been completed and final database release and database freeze has been declared by Data Management.

3. All criteria for unblinding the randomisation codes have been met.

4. Randomisation code schedules for both sub-studies have been distributed according to RandAll NG procedures.

4. ANALYSIS POPULATIONS

Population Definition / Criteria Analyses EvaluatedAll Subjects Enrolled (ASE)

All participants who sign the ICF and for whom a record exists in the study database, including screen failures and any subject who was not screened but experienced an SAE between the date of informed consent and the planned date of the Screening visit.

Study population Reason for

withdrawal prior to randomisation

Randomised All participants who were randomised. No formal analysis will be performed on this population

Intent-to-treat (ITT)

All randomised subjects who made at least one critical error assessment from one treatment option device. A subject who is recorded as a screen or run-in failure and also randomised will be considered to be randomised in error provided they have not performed any error assessments.

Study population Efficacy Subject disposition Safety Inclusion, exclusion

and randomisation criteria deviations

Safety This population will be the same as the Intent-to-treat population.

NOTES : Please refer to Appendix 11: List of Data Displays which details the population to be used for each displays

being generated.

2017N324761_00

CONFIDENTIAL206215

11

4.1. Protocol Deviations

Important protocol deviations (including deviations related to study inclusion/exclusion criteria, conduct of the trial, patient management or patient assessment) will be summarised and listed.

Protocol deviations will be tracked by the study team throughout the conduct of the study in accordance with the Protocol Deviation Management Plan.

o Data will be reviewed prior to unblinding and freezing the database to ensure all important deviations are captured and categorised on the protocol deviations dataset.

o This dataset will be the basis for the summaries and listings of protocol deviations.

A separate summary and listing of all inclusion/exclusion criteria deviations will also be provided. This summary will be based on data as recorded on the inclusion/exclusion page of the eCRF.

5. CONSIDERATIONS FOR DATA ANALYSES AND DATA HANDLING CONVENTIONS

Table 1 provides an overview of appendices within the RAP for outlining general considerations for data analyses and data handling conventions.

Table 1 Overview of Appendices

Section Component

11.1 Appendix 1: Protocol Deviation Management

11.2 Appendix 2: Time & Events

11.3 Appendix 3: Treatment Phases

11.4 Appendix 4: Data Display Standards & Handling Conventions

11.5 Appendix 5: Derived and Transformed Data

11.6 Appendix 6: Premature Withdrawals & Handling of Missing Data

11.7 Appendix 7: Multicenter Studies

11.8 Appendix 8: Examination of Covariates, Subgroups & Other Strata

11.9 Appendix 9: Model Checking and Diagnostics for Statistical Analyses

11.10 Appendix 10: Abbreviations & Trade Marks

11.11 Appendix 11: List of Data Displays

11.12 Appendix 12: Example Mock Shells for Data Displays

2017N324761_00

CONFIDENTIAL206215

12

6. STUDY POPULATION ANALYSES

The study population analyses will be based on the ITT population unless otherwise stated.

Table 2 provides an overview of the planned study population analyses, with full details of data displays being presented in Appendix 11: List of Data Displays.

Table 2 Overview of Planned Study Population Analyses

Display Type Data Displays GeneratedTable Figure Listing

Subject Disposition and DemographyStudy Populations (ASE) YAttendance at Each Clinic Visit YScreening Failures Y YTreatment Status and Reasons for Discontinuation of Study Treatment

Y Y

End of Study Record by Treatment period YSubject Disposition at Each Treatment Period YReasons for withdrawal Y Y(ASE)Number of Subjects by Country and Centre YSummary of Age Ranges Y (ASE)Demographic Characteristics Y YDemographic Characteristics by Country YRace and Racial Combinations YRace and Racial Combinations Details Y YProtocol DeviationsImportant deviations Y YInclusion/Exclusion/Randomisation Criteria Deviations

Y Y

Treatment Misallocations YMedical Condition & Concomitant MedicationsMedical Conditions (Current/Past) Y YFamily History of Cardiovascular Risk Factors Y YCOPD History and COPD Exacerbation History (Only one listing required)

Y Y

Smoking History and Status Y YCOPD Medications Y YNon-COPD Medications Y YRelationship between ATC Level 1, ingredient and verbatim text non-COPD medications only

Y

DevicesNaivety to Inhaler Devices Y Y

2017N324761_00

CONFIDENTIAL206215

13

6.1. Disposition

The study population summary will show the number of subjects overall who were enrolled, the number of screen failures and the number with each reason for screen failure. It will also show the number of subjects in each sub-study and overall who were randomised and who were in the ITT population.

The reasons for withdrawal summary will show the number and percentage of subjects who completed the study, who withdrew prematurely from the study and who reported each primary and sub-reason for withdrawal. This will be presented for each sub-study and overall.

6.2. Medical Conditions

The number and percentage of subjects reporting each current medical condition will be presented by sub-study and overall. This table will include a category of ‘Cardiovascular Risk Factors’. All medical conditions must be summarised on this table regardless of frequency.

This will be repeated for past medical conditions.

6.3. Concomitant Medications

Non-COPD medications will be summarised by Anatomical-Therapeutic-Chemical (ATC) level 1 and ingredient. COPD medications will be summarised by Respiratory Medication Class (RMC) and ATC, and will be derived for each COPD concomitant medication. Multi-ingredient medications will be presented according to their combination ATC classification rather than the classifications of the ingredients. These tables will be presented by sub-study and overall.

COPD and non-COPD medications will be listed together.

A listing of the relationship between ATC Level 1, ingredient and verbatim text will be produced for non-COPD medications only.

7. PRIMARY STATISTICAL ANALYSES

7.1. Efficacy Analyses

7.1.1. Overview of Planned Efficacy Analyses

The primary efficacy analyses will be based on the ITT population, unless otherwise specified.

Table 3 provides an overview of the planned efficacy analyses, with full details of data displays being presented in Appendix 11: List of Data Displays.

2017N324761_00

CONFIDENTIAL206215

14

Table 3 Overview of Planned Primary Efficacy Analyses

AbsoluteStats Analysis Summary

T F T FAt least one critical errorPercentage of subjects making at least one critical error after reading the PIL(s)

Y Y Y

NOTES : T = Table, F = Figure, Y = Yes display generated. Stats Analysis = Represents TF related to any formal statistical analyses (i.e. modelling) conducted. Summary = Represents TF related to any summaries (i.e. descriptive statistics) of the observed raw data.

7.1.2. Planned Efficacy Statistical Analyses

Primary Statistical AnalysesEndpoint(s)

Percentage of subjects making at least one critical error after reading the PIL(s)Model Specification

The primary endpoint of the percentage of subjects making at least one critical error after reading the PIL(s) will be analysed using the Intent-to-treat population.

This endpoint will be analysed using conditional logistic regression with subject as fixed strata, and treatment option and period as fixed effects.

Country will be included in the model providing there are sufficient subjects distributed to enable the model to run. If the model will not run with country in the model then country will be removed from the model.

Model Checking & Diagnostics

Refer to Appendix 9: Model Checking and Diagnostics for Statistical Analyses.Model Results Presentation

The odds ratio, 95% CI and p-value will be presented for the comparison between treatment options. It will be based on a two-sided hypothesis testing approach of superiority.

SAS Code

Logistic regression will be used to calculate the odds ratio with 95% CI and p-value. The following SAS code will be used:

proc logistic data=errors; class device period country; strata subject; model error = country device period / expb; oddsratio device; run;

error is derived as shown in Section 11.5.3

2017N324761_00

CONFIDENTIAL206215

15

Sensitivity and Supportive Statistical Analyses

A sensitivity analysis will be performed using the Cochran-Mantel-Haenszel (CMH) test. The CMH test serves as a stratified approximation to the Mainland-Gart test, a variation of a one-sample chi-square test that accounts for study inhaler sequence (period). A subject who had an error with both devices or who had no errors with both devices does not provide any information about the superiority of either device. Only those subjects who had error(s) in one device and had no error in the other device are counted for in the Mainland-Gart test.

SAS code for the CMH test is provided below:

proc freq data=errors (where=(err_ord^=0)); tables seq*err_ord/cmh;

run;

As with the main analysis, country will be included as a covariate if sufficient data is available as described above. err_ord is derived as shown in Section 11.5.3

8. SECONDARY STATISTICAL ANALYSES


8.1.1. Overview of Planned Secondary Efficacy Analyses

The secondary efficacy analyses will be based on the Intent-to-treat population, unless otherwise specified.

Table 4 provides an overview of the planned secondary efficacy analyses, with further details of data displays being presented in Appendix 11: List of Data Displays.

Table 4 Overview of Planned Secondary Efficacy Analyses

AbsoluteStats Analysis Summary Individual

T T F L

Number of critical errors made after 1st Healthcare Professional (HCP) instructionsThe percentage of subjects making at least one critical error after the first instruction from the HCP

Y Y Y

Number of overall errors madePercentage of subjects making at least one overall error after reading the PIL(s)

Y Y Y

Percentage of subjects making at least one overall error after the first instruction from the HCP

Y Y Y

Number of instructions (maximum of 2) from the HCPNumber of instructions from the HCP Y Y Y Y

Total Training/Teaching time requiredTotal time taken to demonstrate correct inhaler use (T1+T2)

Y Y Y Y

2017N324761_00

CONFIDENTIAL206215

16


T T F LPreference attributesTreatment preference for number of steps required to take medication

Y Y Y

Treatment preference for taking COPD medication Y Y Y

NOTES : T = Table, F = Figure, L = Listing, Y = Yes display generated. Stats Analysis = Represents TFL related to any formal statistical analyses (i.e. modelling) conducted. Summary = Represents TFL related to any summaries (i.e. descriptive statistics) of the observed raw data.

Individual = Represents TFL related to any displays of individual subject observed raw data.

8.1.2. Planned Secondary Efficacy Statistical Analyses

8.1.2.1. Number of critical errors made after 1st Healthcare Professional (HCP) instructions

Secondary Statistical AnalysesEndpoint(s)The percentage of subjects making at least one critical error after the: first instruction from the HCPModel Specification, Model Checking & Diagnostics, Model Results Presentation

Same method as percentage of subjects making at least one critical error after reading the PIL(s), see Section 7.1.2

Note this model will only be fitted if there is sufficient data. If there is insufficient data then only summary statistics will be produced.

8.1.2.2. Number of overall errors made

Secondary Statistical AnalysesEndpoint(s)

Percentage of subjects making at least one overall error after reading the PIL(s) Percentage of subjects making at least one overall error after the first instruction from the HCPModel Specification, Model Checking & Diagnostics, Model Results Presentation


Sensitivity analysis will also be performed as detailed in Section 7.1.2 Note this model will only be fitted if there is sufficient data. If there is insufficient data then only

summary statistics will be produced.

8.1.2.3. Number of instructions (maximum of 2) from the HCP


Number of instructions from the HCP

2017N324761_00

CONFIDENTIAL206215

17

Secondary Statistical AnalysesModel Specification

The endpoint of number of instructions from the HCP will be analysed using the Intent-to-treat population.

Those who were unable to demonstrate correct use will not be included in the analysis, but will be summarised in tables

This endpoint will be analysed using the Wilcoxon signed rank testModel Results Presentation

Summary statistics on the number of instructions from the HCP will be provided for each device along with the p-value of the Wilcoxon signed rank test.

SAS CodeThe following SAS code will be used: proc univariate data=no_error;

var diffinst; run;

diffinst is derived as shown in Section 11.5.5

8.1.2.4. Training/Teaching time required


Total demonstration time (T1+T2)Model Specification

The endpoint of training/teaching time will be analysed using the Intent-to-treat population. This endpoint will be analysed using Kaplan-Meier analysis.Model Results Presentation

For DISKUS + HandiHaler or Turbuhaler + HandiHaler the time for each device will be added together to form the total time needed to demonstrate correct use at T1 and T2 respectively.

Median time to correctly use the inhaler after reading the patient information leaflet and HCP’s instruction (T1+T2) will be presented for each inhaler group. For subjects who correctly usedthe inhaler after reading the patient information leaflet set T2=0. T1+T2 is censored for subjects who did not use the inhaler correctly after 3 attempts. Kaplan-Meier survivor functions of T1+T2 will be obtained for each inhaler group and plotted on the same figure.

In the above analysis the median time will be taken from the Kaplan Meier model.SAS Code

Kaplan-Meier survivor functions of the proportion of subjects with a time to correctly use theinhaler will be obtained for each inhaler group separately and will be plotted on the same figure. The following SAS code will be used:

ods output ProductLimitEstimates = plest(keep = trtcd timeto failed left) quartiles=median CensoredSummary=cen; proc lifetest data=t_device outsurv=survest; time timeto*event(0); strata treatment;

run;

2017N324761_00

CONFIDENTIAL206215

18

8.1.2.5. Preference attributes


Treatment preference from questionnaire for number of steps required to take medication Treatment preference from questionnaire preferred treatment for overall treatment preference.Model Specification

The endpoint of treatment preference will be analysed using the Intent-to-treat population. This endpoint will be analysed using the Cochran-Mantel-Haelszel test.Model Checking & Diagnostics

Refer to Appendix 9: Model Checking and Diagnostics for Statistical Analyses.Model Results Presentation

The percentage of preference for each specific attribute (i.e., preferring ELLIPTA inhaler, preferring the other inhaler, and no preference) will be summarised by inhaler use sequence.

Subject preference will also be analysed using a Cochran-Mantel-Haenszel test. The Cochran-Mantel-Haenszel test serves as a stratified approximation to Prescott’s test, a variation of a one-sample chi-square test that accounts for study inhaler sequence and subjects who indicate no preference. P-values from this analysis will be presented in summary and analysis tables.

SAS Code

SAS code for stratified approximation to Prescott’s test for each preference question:

ods output CMH = cmhi (where = (upcase(AltHypothesis) =: 'ROW MEAN SCORES') keep = AltHypothesis Prob) ;proc freq data = pref_data ; tables seq * pref_ord country / cmh ;run ;

Country will be included in the model providing there are sufficient subjects distributed to enable the model to run. If the model will not run with country in the model then country will be removed from the model.

pref_ord is derived as shown in Section 11.5.4

9. OTHER STATISTICAL ANALYSES


9.1.1. Overview of Planned Other Efficacy Analyses

The other efficacy analyses will be based on the Intent-to-treat population, unless otherwise specified.

Table 5 provides an overview of the planned other efficacy analyses, with further details of data displays being presented in Appendix 11: List of Data Displays.

2017N324761_00

CONFIDENTIAL206215

19

Table 5 Overview of Planned Other Efficacy Analyses


T T F L

Number of critical errors made after Healthcare Professional (HCP) instructionsThe percentage of subjects making at least one critical error after the second instruction from the HCP

Y Y Y

Number of overall errors madePercentage of subjects making at least one overall error after the second instruction from the HCP

Y Y Y

Training/Teaching time requiredTime to read PIL(s) (T1) Y Y Y Y

Instruction time by the HCP (T2) Y Y Y Y

NOTES : T = Table, F = Figure, L = Listing, Y = Yes display generated. Stats Analysis = Represents TFL related to any formal statistical analyses (i.e. modelling) conducted. Summary = Represents TFL related to any summaries (i.e. descriptive statistics) of the observed raw data.

Individual = Represents TFL related to any displays of individual subject observed raw data.

Summary statistics will be generated for the amount of time needed to read the PIL(s) (T1) and the total instruction time by the HCP (T2) separately. These will be presented in the same format as the total amount of time taken to demonstrate correct inhaler use.

9.1.1.1. Number of critical errors/overall errors made after 2nd Healthcare Professional (HCP) instructions

Other Statistical AnalysesEndpoint(s)

The percentage of subjects making at least one critical error after the second instruction from the HCP

Percentage of subjects making at least one overall error after the second instruction from the HCP

Model Specification, Model Checking & Diagnostics, Model Results Presentation


Note this model will only be fitted if there is sufficient data. If there is insufficient data then only summary statistics will be produced.

2017N324761_00

CONFIDENTIAL206215

20

9.1.1.2. Training/Teaching time required

Other Statistical AnalysesEndpoint(s)

Time to read PIL(s) and demonstrate correct inhaler use (T1) Instruction time by the HCP (including demonstration of correct inhaler use) (T2)Model Specification

The endpoint of training/teaching time will be analysed using the Intent-to-treat population. This endpoint will be analysed using Kaplan-Meier analysis.Model Results Presentation

Median time to correctly use the inhaler after reading the patient information leaflet (T1) will be presented for each inhaler group. Time will be censored for subjects who did not use the inhaler correctly. Kaplan-Meier survivor functions of T1 will be obtained for each inhaler group and plotted together on the same graph. This analysis will then be repeated with no censoring, only allowing subjects who correctly used the inhaler to be included in the model.

Median time to correctly use the inhaler following HCP’s instruction (T2) will be presented for each inhaler group. Time will be set to 0 for subjects who correctly used the inhaler after reading the patient information leaflet, and will be censored for subjects who did not use the inhaler correctly after 3 attempts. Kaplan-Meier survivor functions of T2 will be obtained for each inhaler group and plotted on the same figure.

In all of the above analysis the median time will be taken from the Kaplan Meier model.SAS Code

See Section 8.1.2.4

9.2. Safety Analyses

9.2.1. Overview of Planned Safety Analyses

These other safety analyses will be based on the ITT populations, unless otherwise specified.

Table 6 provides an overview of the planned safety analyses, with further details of data displays being presented in Appendix 11: List of Data Displays.

2017N324761_00

CONFIDENTIAL206215

21

Table 6 Overview of Other Planned Safety Analyses

Endpoint AbsoluteSummary Individual

T LAdverse EventsOverview of AEs YOn-treatment AEs, drug related AE’s, fatal/non-fatal SAE’s, fatal/non-fatal drug related SAE’s

Y Y

Summary of Serious Adverse Events by System Organ Class and Preferred Term (Number of Subjects and Occurrences)

Y

On-treatment serious AEs experienced by 3% (before rounding) or more of subjects in any treatment group

Y

On-treatment non-serious AEs experienced by 3% (before rounding) or more of subjects in any treatment group

Y

Subject numbers for individual AEs (ASE) YAll AEs including identification of whether each AE occurred pre-treatment or on-treatment (ASE)

Y

Non-fatal SAEs (ASE) YFatal AEs (ASE) YOn-treatment drug related SAEs YAEs leading to discontinuation of study treatment or withdrawal from the study Y YCardiovascular Events YRelationship between adverse event system organ class, preferred term and verbatim text (ASE)

Y

NOTES : T = Table, L = Listing, Y = Yes display generated. Summary = Represents TL related to any summaries (i.e. descriptive statistics) of the observed raw data.

Individual = Represents TL related to any displays of individual subject observed raw data.

9.2.1.1. Adverse Events and Serious Adverse Events

A summary of the following AEs and SAEs will be provided:

Any AE (pre treatment -, on- treatment or post-treatment) Any on-treatment drug related AE Any AE (pre-treatment or on-treatment) leading to permanent discontinuation of

study treatment or withdrawal from study Any non-fatal SAE Any fatal SAE Any on-treatment non-fatal drug related SAE Any on-treatment fatal drug related SAE

AE incidence will be summarised overall using the primary System Organ Class (SOC) and preferred term. All listings of AEs/SAEs will identify whether each adverse event occurred during screening, on treatment or after treatment.

2017N324761_00

CONFIDENTIAL206215

22

9.2.1.2. Cardiovascular Events

For subjects who report a cardiovascular event, individual patient profiles will be produced for one or more of the following categories:

Myocardial infarction/unstable angina Congestive heart failure Arrhythmias Valvulopathy Pulmonary hypertension Cerebrovascular events/stroke and transient ischemic attack Peripheral arterial thrombosis Deep Venous Thrombosis Revascularization

9.2.2. Pregnancies

Any pregnancies reported during the study will be summarised in case narratives. Any pregnancy complication or elective termination of a pregnancy for medical reasons will be recorded as an AE or SAE and included in summaries and listings of AEs/SAEs.

2017N324761_00

CONFIDENTIAL206215

23

10. REFERENCES

GlaxoSmithKline Document Number 2016N292801_01, Protocol: A randomized, open-label, cross-over, placebo-device study investigating critical and over all errors, training/teaching time, and preference attributes of the ELLIPTA dry powder Inhaler (DPI) as compared to HandiHaler DPI used in combination with either DISKUS DPI or Turbuhaler DPI, in adult patients with Chronic Obstructive Pulmonary Disease (COPD), 15 September 2016

Prescott, RS. The comparison of success rates in cross-over trials in the presence of anorder effect. Appl Stat. 1981; 30:9-15.

Senn, Stephen. Cross-over Trials in Clinical Research, 2nd ed. Chichester: John Wiley & Sons, Ltd, 2002. pp 128-131, 202-203.

2017N324761_00

CONFIDENTIAL206215

24

11. APPENDICES

Section Appendix

RAP Section 4 : Analysis Populations

Section 11.1 Appendix 1: Protocol Deviation Management

RAP Section 5 : General Considerations for Data Analyses & Data Handling Conventions

Section 11.2 Appendix 2: Time & Events

Section 11.3 Appendix 3: Treatment Phases

Section 11.4 Appendix 4: Data Display Standards & Handling Conventions

Section 11.5 Appendix 5: Derived and Transformed Data

Section 11.6 Appendix 6: Premature Withdrawals & Handling of Missing Data

Premature Withdrawals

Handling of Missing Data

Section 11.7 Appendix 7: Multicenter Studies

Section 11.8 Appendix 8: Examination of Covariates, Subgroups & Other Strata

Section 11.9 Appendix 9: Model Checking and Diagnostics for Statistical Analyses

Other RAP Appendices

Section 11.10 Appendix 10: Abbreviations & Trade Marks

Section 11.11 Appendix 11: List of Data Displays

Section 11.12 Appendix 12: Example Mock Shells for Data Displays

2017N324761_00

CONFIDENTIAL206215

25

11.1. Appendix 1: Protocol Deviation Management

The full list of protocol deviations collected on the eCRF is in the Protocol Deviation Management Plan (PDMP). Please refer to this document for current guidance.

2017N324761_00

CONFIDENTIAL206215

26

11.2. Appendix 2: Time & Events

11.2.1. Protocol Defined Time & Events


Study Day 1 1 V0 can take place on the same day as V1. V1 should be completed no later than 30 days after consent.

Procedure:

Screening Assessments Completed prior to randomisation

Written informed consent X Informed consent may take place prior to V0 for logistical reasons. Subjects should be included and randomised within 30 days of providing consent.

Subject demography X Age, height, weight, year of birth, sex, ethnicity and geographic ancestry will be recorded


X Subject will have a medical history of COPD, previously confirmed by spirometry.

Concomitant medication history including COPD Therapy History

X Current concomitant medication will be recorded. A minimum COPD therapy history for the preceding 2 years from inclusion will be recorded

Inclusion/exclusion criteria X All criteria must be meet prior to randomisation at V1

Study Assessments Completed once a subject is included on study

Randomisation X Randomised to treatment order and preference questionnaire

Assess the number of inhaler errors (critical and overall) on each treatment after reading the Patient information leaflet (PIL) for Inhaler tested

X No instruction is provided by the Health CareProfessional (HCP) for this assessment.

Assess the number of inhaler errors (overall and critical) on each treatment after each of 2 attempts following instruction by HCP

X If a subject cannot show correct use after reading the PIL, then the HCP has up to 2 attempts to instruct the subject to attain this.

2017N324761_00

CONFIDENTIAL206215

27


Teaching-Training Time for each inhaler includes the following: The amount of time taken to

read the patient information leaflet and demonstrate inhaler use

The amount of time taken to be given instruction by the HCP on use of the inhaler and demonstrate inhaler use

The total amount of time taken to demonstrate inhaler use

X First assessment is attempt one, including time to read PIL and demonstrate use.

Second assessment will be time for HCP to correct errors and subject to show use for up to 2 more attempts

The Cumulative time for all assessments needed by subject and any HCP instruction to demonstrate no errors will also be captured.

Preference Questionnaire X Subject will complete version of Preference Questionnaire they have been randomised to.

SAE/AE assessment X Collected until completion of final study assessment at V1.

2017N324761_00

CONFIDENTIAL206215

28

11.3. Appendix 3: Treatment Phases

11.3.1. Treatment Phases

All data if necessary will be categorised according to the following treatment phases; pre-study, during- study and post- study. These definitions will be defined based on the date of Visit 1, as this is the visit where subjects complete all study assessments. In addition, as date and not time is captured in the eCRF the treatment phase can only be applied if an event spans multiple days.

11.3.2. Treatment Phase Definitions

DefinitionTreatment Phase

Pre-study During-study Post-study

Before Visit 1 YAt Visit 1 Y

After Visit 1 Y

2017N324761_00

CONFIDENTIAL206215

29

11.4. Appendix 4: Data Display Standards & Handling Conventions

11.4.1. Study Treatment & Sub-group Display Descriptors

Treatment Group Descriptions

RandAll NG Data Displays for Reporting

Code Description Description Order [1]

A ELLIPTA Inhaler ELLIPTA 1

B DISKUS Inhaler + HandiHaler Inhaler DISKUS + HandiHaler 2

C Turbuhaler Inhaler + HandiHaler Inhaler Turbuhaler + HandiHaler 3

Q1 Preference Questionnaire 1 Q1 4




NOTES:

1. Order represents treatments being presented in TFL, as appropriate.

11.4.2. Reporting Process & Standards

Reporting Process

Software

The currently supported versions of SAS software will be used.

The currently supported versions of TSCG software will be used.

Reporting Area

HARP Server : uk1salx00175

HARP Area /arenv/arprod/gsk2834425/mid206215/final_01

QC Spreadsheet /arenv/arwork/gsk2834425/mid206215/final_01/documents

Analysis Datasets

Analysis datasets will be created according to Legacy GSK A&R dataset standards.

Generation of RTF Files

RTF files will be generated for use in writing the CSR.

2017N324761_00

CONFIDENTIAL206215

30

Reporting Standards

General

The current GSK Integrated Data Standards Library (IDSL) will be applied for reporting, unless otherwise stated:o 4.03 to 4.23: General Principleso 5.01 to 5.08: Principles Related to Data Listingso 6.01 to 6.11: Principles Related to Summary Tableso 7.01 to 7.13: Principles Related to Graphics

Formats

All data will be reported according to the actual treatment option the subject received unless otherwise stated.

GSK IDSL Statistical Principles (5.03 & 6.06.3) for decimal places (DP’s) will be adopted for reporting of data based on the raw data collected.

Numeric data on listings will be reported at the precision collected on the eCRF.

Planned and Actual Time

Reporting for Data Listings:

Unscheduled or unplanned readings will be presented within the subject’s listings but not presented in summary tables.

Descriptive Summary Statistics

Continuous Data Refer to IDSL Statistical Principle 6.06.1

Categorical Data N, n, frequency, %

Graphical Displays

Refer to IDSL Statistical Principals 7.01 to 7.13.

2017N324761_00

CONFIDENTIAL206215

31

11.5. Appendix 5: Derived and Transformed Data

11.5.1. Study Population

Demographics

Age

GSK standard IDSL algorithms will be used for calculating age where birth date will be imputed as follows:o Any subject with a missing day will have this imputed as day ‘15’. o Any subject with a missing date and month will have this imputed as ‘30th June’.

Birth date will be presented in listings as ‘YYYY’.

Body Mass Index (BMI)

Calculated as Weight (kg) / [Height (m)]2

11.5.2. Date of Completion and Withdrawal

This study only has one on-treatment visit. Subjects either complete or withdraw on the day. There is no follow up visit.

11.5.3. Critical and Overall Errors

Critical and Overall errors

An ordinal variable (error) will be derived to indicate where a participant had any error, or any critical error whilst demonstrating correct device use. If the subject has an overall/critical error the error=Ywhilst if they did not have an overall/critical error then error=N. The variables errty/errtycd will be used to determine whether the variable error is referring to a critical or overall error.

An ordinal variable (err_ord) will be derived to indicate the responses of critical error based on comparing responses for the device sequences used. This variable will be used in the sensitivity analysis for critical error data.

If the sequence of inhaler use is ELLIPTA in the 1st period and DISKUS + HandiHaler or Turbuhaler + HandiHaler in the 2nd period, then for critical error in relation to sequence:

err_ord = -1, if subject had critical error in the 1st period but NOT in the 2nd period err_ord = 1, if subject had critical error in the 2nd period but NOT in the 1st period err_ord = 0, if subject had critical errors in BOTH periods or had NO critical errors in

BOTH periods

If the sequence of inhaler use is and DISKUS + HandiHaler or Turbuhaler + HandiHaler in the 1st

period and ELLIPTA in the 2nd period, then for each preference question: err_ord = 1, if subject had critical error in the 1st period but NOT in the 2nd period err_ord = -1, if subject had critical error in the 2nd period but NOT in the 1st period err_ord = 0, if subject had critical errors in BOTH periods or had NO critical errors in

BOTH periods

The above method also applies to overall errors.

2017N324761_00

CONFIDENTIAL206215

32

11.5.4. Responses to Preference Questions

Preference Questions

An ordinal variable (pref_ord) will be derived to indicate the responses to preference based on the sequence of inhaler use. This variable will be used in the statistical analysis for preference data.

If the sequence of inhaler use is ELLIPTA in the 1st period and DISKUS + HandiHaler or Turbuhaler + HandiHaler in the 2nd period, then for each preference question:

pref_ord = -1, if subject prefer ELLIPTA pref_ord = 1, if subject prefer DISKUS + HandiHaler or Turbuhaler + HandiHaler pref_ord = 0, if subject has no preference

If the sequence of inhaler use is DISKUS + HandiHaler or Turbuhaler + HandiHaler in 1st period and ELLIPTA in 2nd period, then for each preference question:

pref_ord = 1, if subject prefer ELLIPTA pref_ord = -1, if subject prefer DISKUS + HandiHaler or Turbuhaler + HandiHaler pref_ord = 0, if subject has no preference

11.5.5. Number of instructions

Number of Instructions

The number of instructions from the HCP which are needed todemonstrate adequate inhalation technique for each inhaler will be summarised. The number of instruction from the HCP will be assigned as follows:

NUMINSTR = 0, if no instruction needed correct after reading leaflet NUMINSTR = 1, if the participant received 1st instruction and had no error NUMINSTR = 2, if the participant received the 2nd instruction and had no error NUMINSTR = 3, if the participant received 2nd instruction and had at least one error

From this an ordinal variable (diffinst) will be derived to indicate the difference between the number of instructions required to demonstrate correct use on the ELLIPTA device compared to the DISKUS + HandiHaler or Turbuhaler + HandiHaler devices.

2017N324761_00

CONFIDENTIAL206215

33

11.6. Appendix 6: Premature Withdrawals & Handling of Missing Data

11.6.1. Premature Withdrawals

Element Reporting Detail

General For reporting purposes, subject study completion will be defined as completion of all three periods (both devices and answered questionnaire) on the day of study visit.

Withdrawn subjects will be defined as those who used at least one device but did not complete all assessments.

Withdrawn subjects will not be replaced in the study.

All available data from subjects who were withdrawn from the study will be listed and all available scheduled data will be included in summary tables and figures, unless otherwise specified.

11.6.2. Handling of Missing Data


General Missing data occurs when any requested data is not provided, leading to blank fields on the collection instrument :o These data will be indicated by the use of a “blank” in subject listing

displays. Unless all data for a specific visit are missing in which case the data is excluded from the table.

o Answers such as “Not applicable” and “Not evaluable” are not considered to be missing data and should be displayed as such.

As subjects need to be assessed for critical errors on both devices in order for the analysis to be performed, if they only complete one period they will not be included in the analysis. Subjects who complete both devices, but do not answer the questionnaire will be included in all analysis excluding those related to preference.

Outliers Whilst outliers are not anticipated within this study any subjects excluded from the summaries and/or statistical analyses will be documented along with the reason for exclusion in the clinical study report. Any potential exclusions would be for exploratory purposes only and not for the primary analysis.

2017N324761_00

CONFIDENTIAL206215

34

11.6.2.1. Handling of Missing/Partial Dates


General Partial dates will be displayed as captured in subject listing displays.

Adverse Events

The eCRF allows for the possibility of partial dates (i.e., only month and year) to be recorded for AE start and end dates; that is, the day of the month may be missing. In such a case, the following conventions will be applied for calculating the time to onset and the duration of the event:o Missing Start Day: First of the month will be used unless this is before the

start date of study treatment; in this case the study treatment start date will be used and hence the event is considered On-treatment as per Appendix 3: Treatment Phases.

o Missing Stop Day: Last day of the month will be used, unless this is after the stop date of study treatment; in this case the study treatment stop date will be used.

Completely missing start or end dates will remain missing, with no imputation applied. Consequently, time to onset and duration of such events will be missing.

Start or end dates which are completely missing (i.e. no year specified) will remain missing, with no imputation applied.

Concomitant Medications

Partial dates for any concomitant medications recorded in the eCRF will be imputed using the following convention:o If the partial date is a start date, a '01' will be used for the day and 'Jan' will

be used for the montho If the partial date is a stop date, a '28/29/30/31' will be used for the day

(dependent on the month and year) and 'Dec' will be used for the month.

The recorded partial date will be displayed in listings.

Adverse Events

Any partial dates for adverse events will be raised to data management. If the full date cannot be ascertained, the following assumptions will be made:o If the partial date is a start date, a '01' will be used for the day and 'Jan' will

be used for the month. o However, if these results in a date prior to Week 1 Day 1 and the event

could possibly have occurred during treatment from the partial information, then the Week 1 Day 1 date will be assumed to be the start date.

o The AE will then be considered to start on-treatment (worst case).o If the partial date is a stop date, a '28/29/30/31' will be used for the day

(dependent on the month and year) and 'Dec' will be used for the month.

The recorded partial date will be displayed in listings.

2017N324761_00

CONFIDENTIAL206215

35

11.6.2.2. Handling of Missing Data for Statistical Analysis


Critical Errors If any of the critical error assessment elements (sub-questions) have a missing response then, for the purposes of the analysis of the percentage of subjects making at least one critical error, the subject will be defined as having one or more critical errors. All summarised data and listed data at the individual question level will show that response to be missing.Note that if a subject already has a critical error in another question then the missing response will not impact their error status.

Overall Errors If any of the error assessment elements have a missing response then, for the purposes of the analysis of the percentage of subjects making at least oneerror, the subject will be defined as having one or more errors. All summarised data and listed data at the individual question level will show that response to be missing.Note that if a subject already has an error in another question then the missing response will not impact their error status.

2017N324761_00

CONFIDENTIAL206215

36

11.7. Appendix 7: Multicenter Studies

11.7.1. Methods for Handling Centres


Centres Central randomisation was used for this study. Data from all participating centres will be pooled prior to analysis.

2017N324761_00

CONFIDENTIAL206215

37

11.8. Appendix 8: Examination of Covariates, Subgroups & Other Strata

11.8.1. Handling of Covariates

The following is a list of covariates that may be used in descriptive summaries and statistical analyses. Whilst inhaler sequence will be used throughout, the inclusion of country in the model relies on sufficient data being collected in each country to allow the model to remain statistically valid. If the model is valid the data the analysis will be performed.

Country will be included as a covariate as part of a sensitivity analysis if enough subjects from each country are recruited, within each of the possible outcomes (critical error/no critical error) by device and randomised sequence, to allow the statistical model to converge without errors.

Category Covariates and / or Subgroups

Inhaler Sequence The study inhaler use sequence (ELLIPTA/Other or Other/ELLIPTA) which include information on the treatment and period it was received in will be adjusted for in the statistical analysis. Other refers to either DISKUS + HandiHaler or Turbuhaler + HandiHaler.

Country This study will be carried out in two countries (the United Kingdom and the Netherlands), summary tables will be provided by country and it may be included in statistical sensitivity analysis if enough subjects from each country are recruited, within each of the possible outcomes (critical error/no critical error) by device and randomised sequence, to allow the statistical model to converge without errors.

2017N324761_00

CONFIDENTIAL206215

38

11.9. Appendix 9: Model Checking and Diagnostics for Statistical Analyses

Model Diagnostic Checks

Logistic Regression Residual vs Fitted plots will be generated to help assess the constant variance assumption. Deviance and Pearson goodness of fit statistics from the model will be examined to help assess the null hypothesis that the model is an adequate fit to the data.

Cochran-Mantel-Haenszel Homogeneity across groups will be assessed using the Breslow-Day test.

SAS Code:proc freq data = pref_data (where = (pref_ord ne 0)) ; tables country * seq * pref_ord / cmh ;run;

All model checking statistics will be assessed but not formally reported in a table, listing or figure.

2017N324761_00

CONFIDENTIAL206215

39

11.10. Appendix 10: Abbreviations & Trade Marks

11.10.1. Abbreviations

Abbreviation DescriptionA&R Analysis and Reporting AE Adverse EventCI Confidence IntervalCSR Clinical Study ReportDOB Date of BirthDP Decimal PlaceseCRF Electronic Case Record FormGSK GlaxoSmithKlineICH International Conference on HarmonisationIDSL Integrated Data Standards LibraryITT Intent-To-TreatPDMP Protocol Deviation Management PlanQC Quality ControlRAMOS Randomization & Medication Ordering System RAP Reporting & Analysis PlanSAC Statistical Analysis CompleteSOP Standard Operation ProcedureTFL Tables, Figures & Listings

11.10.2. Trademarks

Trademarks of the GlaxoSmithKline Group of Companies

Trademarks not owned by the GlaxoSmithKline Group of Companies

DISKUS HandiHalerELLIPTA SAS

Turbuhaler

2017N324761_00

CONFIDENTIAL206215

40

11.11. Appendix 11: List of Data Displays

If enough subjects are recruited in both the Netherlands and the United Kingdom then the additional analysis tables including country will be included as detailed in the main RAP, and table numbers will be adjusted accordingly.

2017N324761_00

CONFIDENTIAL206215

41

11.11.1. Study Population Tables

Study Population Tables

No. PopulationIDSL / TST ID / Example Shell

Title Programming NotesDeliver-

able

Subject Disposition and Demography

1.01 ASE IDSL_SP01 Summary of Subject PopulationsPage by sub-study, total column only

SAC

1.02 ASE IDSL_ES6Summary of Screening Status and Reasons for Screen Failures

Page by sub-study SAC

1.03 ASE IDSL_DM1 Summary of Age Ranges Page by sub-study SAC

1.04 ITT IDSL_SP02 Summary of Attendance at Each Clinic VisitPage by sub-study, total column only

SAC

1.05 ITT IDSL_SD1Summary of Treatment Status and Reasons for Discontinuation of Study Treatment


1.06 ITT IDSL_ES1 Summary of End of Study Record Page by sub-study SAC

1.07 ITT IDSL_DM2 Summary of Demographic Characteristics Page by sub-study SAC

1.08 ITT IDSL_DM2Summary of Demographic Characteristics by Country


1.09 ITT IDSL_NS1Summary of Number of Subjects by Country and Centre


1.10 ITT IDSL_DM5 Summary of Race and Racial Combinations Page by sub-study SAC

1.11 ITT IDSL_DM6Summary of Race and Racial Combination Details


2017N324761_00

CONFIDENTIAL206215

42

Study Population Tables



able

Protocol Deviations

1.12 ITT IDSL_SP04 Summary of Important Protocol Deviations Page by sub-study SAC

1.13 ITT IDSL_IE1Summary of Inclusion/ Exclusion/ Randomisation Criteria Deviations for Intent-to-treat


Medical Condition & Concomitant Medications

1.14 ITT IDSL_MH4 Summary of Current Medical Conditions Page by sub-study SAC

1.15 ITT IDSL_MH4 Summary of Past Medical Conditions Page by sub-study SAC

1.16 ITT IDSL – SP07Summary of Family History of Cardiovascular Risk Factors


1.17 ITTIDSL_SP08 Summary of COPD History and COPD

Exacerbation History at ScreeningPage by sub-study SAC

1.18 ITT IDSL_SP10 Summary of Smoking Status Page by sub-study SAC

1.19 ITT IDSL_SP11 Summary of COPD Medications Page by sub-study SAC

1.20 ITT IDSL_CM1 Summary of Non-COPD Medications Page by sub-study SAC

2017N324761_00

CONFIDENTIAL206215

43

11.11.2. Efficacy Tables

Overall Summary of Errors

No. PopulationIDSL / TST

ID / Example Shell


able

2.01 ITT EFF_T01 Summary of Errors on ELLIPTA Page by sub-study SAC

2.02 ITT EFF_T01 Summary of Errors on ELLIPTA by Country Page by sub-study and country SAC

2.03 ITT EFF_T01 Summary of Errors on DISKUS + HandiHalerPage by DISKUS + HandiHaler Errors, DISKUS Errors and HandiHaler Errors

SAC

2.04 ITT EFF_T01Summary of Errors on DISKUS + HandiHaler by Country

Page by DISKUS + HandiHaler Errors, DISKUS Errors, HandiHaler Errors and Country

SAC

2.05 ITT EFF_T01 Summary of Errors on Turbuhaler + HandiHalerPage by Turbuhaler + HandiHaler Errors, Turbuhaler Errors and HandiHaler Errors

SAC

2.06 ITT EFF_T01Summary of Errors on Turbuhaler + HandiHaler by Country

Page by Turbuhaler + HandiHaler Errors, Turbuhaler Errors, HandiHaler Errors and country

SAC

2.07 ITT EFF_T06 Summary of Critical ErrorsPage by sub-study and assessment

SAC

2.08 ITT EFF_T06 Summary of Critical Errors by Country.Page by sub-study, assessmentand country

SAC

2017N324761_00

CONFIDENTIAL206215

44

Overall Summary of Errors


ID / Example Shell


able

Primary Analysis Tables

2.09 ITT EFF_T04Analysis of Percentage of Subjects making atleast one Critical Error

Include analysis only if sufficient data is available for models to run. Page by sub-study and assessment

SAC

2.10 ITT EFF_T07Sensitivity Analysis of Percentage of Subjects making at least one Error after Reading the PIL(s)


Secondary Analysis Tables

Number of Overall Errors

2.11 ITT EFF_T04Analysis of Percentage of Subjects making at least one Overall Error

Include analysis only if sufficient data is available for models to run. Page by sub-study and assessment.

SAC

Training/Teaching Time Required

2.12 ITT EFF_T02Summary and Analysis of Time (Minutes) Taken to Demonstrate Correct Inhaler Use

Page by sub-study and assessment.

SAC

Preference Attributes

2.13 ITT EFF_T03 Summary of Treatment Preference Page by sub-study. SAC

2.14 ITT EFF_T05Summary and Analysis of Number of Instructions from the HCP

Only if sufficient data is available. Page by sub-study.

SAC

2017N324761_00

CONFIDENTIAL206215

45

11.11.3. Safety Tables


ID / Example Shell


able

Adverse Events

3.01 ITT IDSL_AE1 Summary of Adverse Events Page by sub-study. SAC

3.02 ITT IDSL_AE1 Summary of On-study Adverse Events Page by sub-study. SAC

3.03 ITT IDSL_AE1 Summary of Post-study Adverse Events Page by sub-study. SAC

3.04 ITT IDSL_AE1 Summary of Serious Adverse Events Page by sub-study. SAC

3.05 ITT IDSL_AE1On-treatment AEs leading to discontinuation of study treatment or withdrawal from the study

Page by sub-study. SAC

3.06 ITT IDSL_AE1Summary of Serious Adverse Events by System Organ Class and Preferred Term (Number of Subjects and Occurrences)


2017N324761_00

CONFIDENTIAL206215

46

11.11.4. Efficacy Figures

Primary Efficacy Analysis


ID / Example Shell

Title Programming Notes Deliverable

2.01 ITT EFF_F01Percentage of Subjects with at Least One Critical Error by Assessment


Secondary Efficacy Analysis

2.02 ITT EFF_F01Percentage of Subjects with at Least One Error by Assessment


2.03 ITT EFF_F02 Number of Instructions required from a HCP Page by sub-study. SAC

2.04 ITT EFF_F03Kaplan-Meier Plot of Time Taken to Demonstrate Correct Inhaler Use

Page by sub-studyand assessment.

SAC

2017N324761_00

CONFIDENTIAL206215

47

11.11.5. ICH Listings

ICH : Listings



Study Population: Subject Disposition and Demography

01 ASE IDSL_ES7 Listing of Screen Failures Page by sub-study. SAC

02 ITT IDSL_ES2 Listing of Reasons for Study Withdrawal Page by sub-study. SAC

Study Population: Protocol Deviations

03 ITT IDSL_SP01 Listing of Important Protocol Deviations Page by sub-study. SAC

04 ASE IDSL_IE3Listing of Inclusion/ Exclusion/ Randomisation Criteria Deviations


Study Population: Treatment

05 ITT IDSL_TA1Listing of Randomised and Actual TreatmentSequence


Study Population: Demography

06 ITT IDSL_DM2 Listing of Demographic Characteristics

Include BMI as the optional measurement. In addition, include a country column as the first sort variable. Page by sub-study.

SAC

07 ITT IDSL_DM9 Listing of Race

Include a countrycolumn as the first sort variable. Page by sub-study.

SAC

2017N324761_00

CONFIDENTIAL206215

48

ICH : Listings



Safety: Adverse Events

08 ASE IDSL_AE7Listing of Subject Numbers for Individual Adverse Events


09 ASE IDSL_AE8 Listing of All Adverse Events Page by sub-study. SAC

Efficacy: Primary Endpoint

10 ITT EFF_L01 Listing of Subject Errors by Assessment Page by sub-study. SAC

2017N324761_00

CONFIDENTIAL206215

49

11.11.6. Non ICH Listings

Non-ICH : Listings

No. PopulationIDSL / TST ID

/ Example Shell

Title Programming NotesDelivera

ble

Study Population: Subject Disposition

11 ITT Listing of Subjects by Country and Centres Page by sub-study. SAC

Study Population: Protocol Deviations

12 ITT IDSL_TA1 Listing of Treatment Sequence Misallocations Page by sub-study. SAC

Study Population: Medical Conditions & Concomitant Medications

13 ITT IDSL_MH2 Listing of Medical Conditions Page by sub-study. SAC

14 ITT IDSL_SP05 Listing of Family History of Cardiovascular Risk Factors Page by sub-study. SAC

15 ITTIDSL_SP06 Listing of COPD History and COPD Exacerbation

HistoryPage by sub-study. SAC

16 ITT IDSL_SP07 Listing of Smoking History and Smoking Status Page by sub-study. SAC

17 ITT IDSL_CM2 Listing of COPD and non-COPD Medications Page by sub-study. SAC

18 ITT IDSL_CM6Relationship between ATC Level 1, Ingredient and Verbatim Text


Study Population: Devices

19 ITT IDSL_SP08 Listing of Naivety to Inhaler Devices Page by sub-study. SAC

Efficacy: Secondary Efficacy Analysis

20 ITT EFF_L02Listing of Time Taken to Demonstrate Correct Inhaler Use


21 ITT EFF_L03 Listing of Preference Questionnaire Page by sub-study. SAC

2017N324761_00

CONFIDENTIAL206215

50

Non-ICH : Listings

No. PopulationIDSL / TST ID

/ Example Shell

Title Programming NotesDelivera

ble

Safety

22 ITT IDSL_VS4 Listing of Myocardial infarction/unstable angina Page by sub-study. SAC

23 ITT IDSL_S06 Listing of Congestive heart failure Page by sub-study. SAC

24 ITT IDSL_VS4 Listing of Arrhythmias Page by sub-study. SAC

25 ITT IDSL_S06 Listing of Valvulopathy Page by sub-study. SAC

26 ITT IDSL_VS4 Listing of Pulmonary hypertension Page by sub-study. SAC

27 ITT IDSL_S06Listing of Cerebrovascular events/stroke and transient ischemic attack


28 ITT IDSL_S06 Listing of Peripheral arterial thromboembolism Page by sub-study. SAC

29 ITT IDSL_VS4 Listing of Deep venous thrombosis/pulmonary embolism Page by sub-study. SAC

30 ITT IDSL_S06 Listing of Revascularisation Page by sub-study. SAC

31 ITT IDSL_VS4 Listing of All cause deaths Page by sub-study. SAC

32 ITT IDSL_PREG1aListing of All subjects who became pregnant during the study


2017N324761_00

CONFIDENTIAL206215

51

11.12. Appendix 12: Example Mock Shells for Data Displays

Example : EFF_T01 Page 1 of 1Protocol : 206215Population : Intent to Treat

Table X.XSummary of errors on ELLIPTA

Sub-study 1: ELLIPTA vs DISKUS + HandiHaler

After Reading After 1st Instruction After 2nd Instruction Leaflet from HCP from HCP Inhaler errors test (N=XXX) (N=XXX) (N=XXX) - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Number of Subjects with Errors XXX (XX%) XXX (XX%) XXX (XX%)Number of Subjects with Critical Errors XXX (XX%) XXX (XX%) XXX (XX%)

Total Number of Errors XXX XXX XXXTotal Number of Critical Errors XXX XXX XXX

Failed to open cover [1] XXX (XX%) XXX (XX%) XXX (XX%)Shook the device upside down after dose preparation [1] XXX (XX%) XXX (XX%) XXX (XX%)Inhalation manoeuvre: long, steady, deep XXX (XX%) XXX (XX%) XXX (XX%)Blocked air inlet during inhalation manoeuvre XXX (XX%) XXX (XX%) XXX (XX%)No exhalation before an inhalation XXX (XX%) XXX (XX%) XXX (XX%)Exhaled directly into mouthpiece [1] XXX (XX%) XXX (XX%) XXX (XX%)No seal by the lips round the mouthpiece during the inhalation [1] XXX (XX%) XXX (XX%) XXX (XX%)

Did not hold breath XXX (XX%) XXX (XX%) XXX (XX%)Did not close the device XXX (XX%) XXX (XX%) XXX (XX%)

2017N324761_00

CONFIDENTIAL206215

52

[1] Indicates a Critical Error. Note: Percentages for number of subjects are calculated from the total number of subjects who used the device(s), percentage of type of errors are calculated based on the number of subjects with errors.Programming notes: Percentages for rows 1 and 2 are calculate from the total number of subjects in the ITT population shown in the heading, the remaining percentages are calculated based on the total number of subjects who had errors. Page by sub-study. For DISKUS + HandiHaler/Turbuhaler + HandiHaler the first page should only contain the first four rows which will the overall summary of number of subjects with errors/critical errors and total number of errors/critical errors. The remaining two pages will be a replicate of the shell above but for two devices separately.

2017N324761_00

CONFIDENTIAL206215

53


Table X.XSummary and Analysis of Time (Minutes) Taken to Demonstrate Correct Inhaler Use

Sub-study 1: ELLIPTA vs DISKUS + HandiHalerAssessment: Time to read the PIL and Demonstrate Correct Use

ELLIPTA DISKUS + HandiHaler(N=XXX) (N=XXX)

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

n XXX XXX Mean XX.XX XX.XX SD XX.XXX XX.XXX Median XX.XX XX.XX

Min XX.X XX.X Max XX.X XX.X

Number of Subjects who demonstrated correct use XXX XXXNumber of Subjects who failed to demonstrate correct use (censored) XXX XXXMedian Time to demonstrate correct inhaler use (minutes) [1] XXX XXX

[1] The Median Time to demonstrate correct inhaler use (minutes) is taken from the Kaplan-Meier analysis. If more than 50% of the data is censored then the median is not applicable.Note: For the Diskus+Handihaler and Turbuhaler +HandiHaler groups, if a subject made an error on one device and not the other the time to demonstrate correct use is censored at the total of time taken for both devices.

Programming notes: Page by sub-study and assessment (Time taken to read the PIL and Demonstrate Correct Use/ Time for HCP Instruction and to Demonstrate Correct Use /Total Time to demonstrate correct use).

2017N324761_00

CONFIDENTIAL206215

54


Table X.XSummary and Analysis of Treatment Preference


Total P-Value(N=XXX)

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Which treatment do you prefer based on the number of the number of steps needed to take your medication?

n XXX X.XXXELLIPTA XXX (XX%)DISKUS + HandiHaler XXX (XX%)No Preference XXX (XX%)

Which treatment do you prefer for taking your medication?

n XXX X.XXXELLIPTA XXX (XX%)DISKUS + HandiHaler XXX (XX%)No Preference XXX (XX%)

Note: The p-value is from the Cochran-Mantel-Haenszel Test.Programming notes: Page by sub-study. %’s calculated out of the small n for each question.

2017N324761_00

CONFIDENTIAL206215

55


Table X.XAnalysis of Percentage of Subjects making at least one Critical Error

Sub-study 1: ELLIPTA vs DISKUS + HandiHalerAssessment: After reading the PIL(s)

ELLIPTA DISKUS + HandiHaler(N=XXX) (N=XXX)

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

At least one Critical Error XXX (XX%) XXX (XX%)Zero Critical Errors XXX (XX%) XXX (XX%)

ELLIPTA vs DISKUS + HandiHaler

Odds Ratio X.XX95% CI (X.XX, X.XX)p-value X.XXX

Note: Odds ratio, 95% CI and p-value obtained from a conditional logistic regression model. Subject is included in the model as fixed strata, treatment option, country and period included as fixed effects.

Programming notes: Page by sub-study and assessment

2017N324761_00

CONFIDENTIAL206215

56


Table X.XSummary and Analysis of Number of Instructions from the HCP


ELLIPTA DISKUS + HandiHaler Difference (ELLIPTA – DISKUS+HandiHaler)(N=XXX) (N=XXX)

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

n X X X

Median (95% CI) XXX (XXX – XXX) XXX (XXX – XXX) XXX (XXX – XXX)Min XXX XXX XXXMax XXX XXX XXX

Number of Insturctions0 XXX (XX%) XXX (XX%)1 XXX (XX%) XXX (XX%)2 XXX (XX%) XXX (XX%)Failed to demonstrate XXX (XX%) XXX (XX%)correct use

ELLIPTA vs DISKUS + HandiHaler

p-value X.XXX

For the Diskus+Handihaler and Turbuhaler+HandiHaler groups, if a subject failed to demonstrate correct use on one device and not the other, the subject would be counted as failed to demonstrate correct use.Note: P-value has come from the Wilcoxon signed rank test. This analysis did not take into account sequence of treatment options. Programming notes: Page by sub-study.

2017N324761_00

CONFIDENTIAL206215

57

Example : EFF_T06 Page 1 of XProtocol : 206215Population : Intent to Treat

Table X.XSummary of Critical Errors

Timepoint: After reading PILTotal(N=XXX)

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

At least one Critical Error on ELLIPTA XXX (XX%)Zero Critical Errors on ELLIPTA XXX (XX%)

At least one Critical Error on DISKUS + HandiHaler XXX (XX%)Zero Critical Errors on DISKUS + HandiHaler XXX (XX%)

At least one Critical Error on ELLIPTA and DISKUS + HandiHaler XXX (XX%)Zero Critical Errors on ELLIPTA or DISKUS + HandiHaler XXX (XX%)

Number of subjects with discordant results [1] XXX (XX%)At least one Critical Error on ELLIPTA and Zero Critical Errors on DISKUS + HandiHaler [2] XXX (XX%)At least one Critical Error on DISKUS + HandiHaler and Zero Critical Errors on ELLIPTA [2] XXX (XX%)

[1] Defined as an error in one device and not the other.[2] Percentage is out of the number of subjects with discordant results.

Programming note: Repeat by sub-study and remaining timepoints, e.g. after first HCP instruction, second etc and by

2017N324761_00

CONFIDENTIAL206215

58


Table X.XSensitivity Analysis of Percentage of Subjects making at least one Error after Reading the PIL(s)

Timepoint: After reading PILResult

_____________________________________________________________________________________________________________________________________

Critical ErrorsNumber of subjects with a discordant results [1] XXX

Number of subjects with a critical error on ELLIPTA XXX (XX%)Number of subjects with a critical error on DISKUS + HandiHaler XXX (XX%)p-Value [2] X.XXX

Overall ErrorsNumber of subjects with a discordant results [1] XXX

Number of subjects with a error on ELLIPTA XXX (XX%)Number of subjects with a error on DISKUS + HandiHaler XXX (XX%)p-Value [2] X.XXX

[1] Defined as an error in one device and not the other.[2] Analysis performed using the Cochran-Mantel-Haenszel test adjusted for country.Note: Subjects are only included in the analysis if they have discordant data.

Programming note: Page by sub-study

2017N324761_00

CONFIDENTIAL206215

59

Example : EFF_F01 Page 1 of 1Protocol : 206215Population : Intent to Treat

Figure X.XProportion of Subjects with at Least One Critical Error by Assessment

Sub-study 1: ELLIPTA vs DISKUS + HandiHalerAssessment: After Reading the PIL(s)

Programming notes: Page by sub-study.

2017N324761_00

CONFIDENTIAL206215

60


Figure X.XNumber of Instructions required from a HCP



2017N324761_00

CONFIDENTIAL206215

61


Figure X.XKaplan-Meier Plot of Total Time Taken to Demonstrate Correct Inhaler Use

Sub-study 1: ELLIPTA vs DISKUS + HandiHalerAssessment: Time to read the PIL and Demonstrate Correct Use

Note: For the Diskus+Handihaler and Turbuhaler +HandiHaler groups, if a subject made an error on one device and not the other the time to demonstrate correct use is censored at the total of time taken for both devicesProgramming notes: Page by sub-study and assessment (Time taken to read the PIL and Demonstrate Correct Use/ Time for HCP Instruction and to Demonstrate Correct Use /Total Time to demonstrate correct use).

2017N324761_00

CONFIDENTIAL206215

62

Example : EFF_L01 Page 1 of nProtocol : 206215Population : Intent to Treat

Listing X.XListing of Subject Errors by Assessment


Subject ID Period Treatment Assessment Total Errors Device Type of Error- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

XXXX 1 ELLIPTA Reading the PIL 2 ELLIPTA Exhaled directly into mouthpieceFailed to open cover

After 1st instruction from HCP 1 ELLIPTA Did not hold breathAfter 2nd instruction from HCP 0

2 DISKUS + Reading the PIL 3 DISKUS Failed to open coverHandiHaler Lever is not pushed back

HandiHaler Did not hold breathAfter 1st instruction from HCP 2 DISKUS Failed to open cover

HandiHaler Did not hold breath [1]After 2nd instruction from HCP 1 HandiHaler Did not hold breath [1]

...

[1] Indicates a Critical Error.


2017N324761_00

CONFIDENTIAL206215

63


Listing X.XListing of Preference Questionnaire


Subject ID Treatment Sequence Question 1 Question 2- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

XXXX ELLITA/DISKUS + HandiHaler/Q1 ELLIPTA ELLIPTAXXXX DISKUS + HandiHaler/ELLIPTA/Q2 DISKUS + HandiHaler DISKUS + HandiHalerXXXX ELLITA/DISKUS + HandiHaler/Q2 No Preference No Preference

...

Note: Q1 = Questionnaire 1, Q2 = Questionnaire 2 Question 1 = Which treatment do you prefer based on the number of the number of steps needed to take your medication? Question 2 = Which treatment do you prefer for taking your medication?


2017N324761_00

CONFIDENTIAL206215

64


Listing X.XListing of Time Taken to Demonstrate Correct Inhaler Use (Minutes)

Sub-Study 1: ELLIPTA vs DISKUS + HandiHaler

Time to Read Instruction Time Total TimeSubject ID Treatment Sequence PIL (T1) by the HCP (T2) (T1 + T2)

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

XXXX ELLITA/DISKUS + HandiHaler XXX XXX XXXXXXX DISKUS + HandiHaler/ELLIPTA XXX XXX XXXXXXX ELLITA/DISKUS + HandiHaler XXX XXX XXX


2017N324761_00

Annotated Study Book for Study Design: 206215

Study Design Version: 1.1

Protocol: 206215

Study Design

Generated by Central Designer TM

May 9, 2017 8:45AM

Page 1 of 66Annotated Study Book - 206215

1

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: NON-SERIOUS ADVERSE EVENT (AE) - Repeating Form

# Sequence Number Event Start Date Outcome Frequency Maximum Intensity Action taken Subject withdrawn? Relationship

1

NON-SERIOUS ADVERSE EVENT

1. Sequence Number

[Sequence Number]

2. Event Diagnosis Only (if known) Otherwise Sign/Symptom

[Event]

3. Start Date

[Start Date]

/ /

4. Outcome / End Date

[Outcome]

Recovered/Resolved, provide End Date

/ /

Recovering/Resolving

Not recovered/Not resolved

Recovered/Resolved with sequelae, provide End Date

/ /

5. Frequency

[Frequency]

Single Episode

Intermittent

6. Maximum Intensity

Record maximum intensity throughout duration of event

[Maximum Intensity]

Mild

Moderate

Severe

Not applicable

7. Most Clinically Significant Action Taken with Study Treatment(s) as a Result of the AE[Action taken]

Study Treatment(s) withdrawn

Not applicable

8. Did the subject withdraw from study as a result of this AE?

[Subject withdrawn?]

Yes

No

9. Is there a reasonable possibility that the AE may have been caused by the Study Treatment?

Use best judgment at initial entry. May be amended when additional information becomes available.

[Relationship]

No

Yes

Note: Hidden items are not displayed.


2

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: ARRHYTHMIAS (ARRHYTHMIAS) - Repeating Form

# Seq Date/Time Cardiac

Arrest

Heart

rate

Palpitations Lightheadedness Syncope Was subject hospitalised due to

arrhythmias?

Was an ECG

Performed?

Is a rhythm / in-hospital telemetry strip

available?

Is a pacemaker / ICD strip printout

available?

Is a loop recorder printout

available?

Is a holter monitor report / printout

available?

Is an Electrophysiology Study (EP) report

available?

Was any of the following therapy

administered?

1

Clinical presentation

1. Sequence Number

[Seq]

2. Date and time of clinical presentation

[Date/Time]

/ /

: 24-hour clock

3. Cardiac arrest

[Cardiac Arrest]

Yes

No

4. Heart rate

[Heart rate]

beats/min

Symptoms

5. Palpitations

[Palpitations]

Yes

No

6. Lightheadedness

[Lightheadedness]

Yes

No

7. Syncope

[Syncope]

Yes

No

Hospitalisation

8. Was subject hospitalised due to arrhythmias?

[Was subject hospitalised due to arrhythmias?]

No

Yes

Admission date and time

/ /

: 24-hour clock

ECG Standard 12-Lead

9. Was an ECG Performed?[Was an ECG Performed?]

YesRecord date and time of ECG

/ /

: 24-hour clock

Rhythm

Check all that apply:

Sinus Bradycardia

Atrial flutter

Atrial fibrillation

Sustained ventricular tachycardia

Non-sustained ventricular tachycardia

Other abnormal rhythm

Ventricular fibrillation sustained

Ventricular fibrillation non-sustained

No

Please provide the clinical diagnosis of the arrhythmias?

Evaluation

10. Is a rhythm / in-hospital telemetry strip available?

[Is a rhythm / in-hospital telemetry strip available?]

No

Yes, has event been reported as an SAE?

Yes

No, please record details below if event has not been reported as an SAE:

11. Is a pacemaker / ICD strip printout available?

[Is a pacemaker / ICD strip printout available?]

No


Yes


12. Is a loop recorder printout available?

[Is a loop recorder printout available?]

No


Yes


13. Is a holter monitor report / printout available?

[Is a holter monitor report / printout available?]

No


Yes


14. Is an Electrophysiology Study (EP) report available?

[Is an Electrophysiology Study (EP) report available?]

No


Yes


Therapy

15. Was any of the following therapy administered?

[Was any of the following therapy administered?]

Medication

Yes

No

Cardioversion

Yes

No

Defibrillation

Yes

No

Defibrillator/pacemaker insertion

Yes

No

Radiofrequency ablation

Yes

No


3

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: CONGESTIVE HEART FAILURE (CHF) - Repeating Form

# Seq Has the subject

had a previous

episode of heart

failure?

Did heart failure lead

to an unplanned visit

to a health care

clinic?


to an unplanned visit

to the Emergency

Room?


to an unplanned

admission to the

hospital?

Functional

status

Lab

Name

Lab

Address

Brain

natriuretic

peptide

N-terminal pro-

Brain

natriuretic

peptide

Atrial

natriuretic

peptide

Heart

failure

signs

Heart failure

symptoms

Was an

Echocardiogram

performed?

Was a right heart

catheterisation

performed?

Was a Chest X-

ray

performed?

Was a Multiple

Gated Acquisition

Scan (MUGA)

performed?

Were any changes to the subject's

treatment for heart failure required

during their clinic or emergency room

visit or hospitalisation?

Escalation of a

chronic

cardiovascular

medication

Increase in

oral

diuretics

New class of

medication

added

Intravenous

diuretics

Inotropes Intravenous

vasodilators

Mechanical

assisted

device

Final

primary

diagnosis

1

Past medical history

1. Sequence Number

[Seq]

2. Has the subject had a previous episode of heart failure?[Has the subject had a previous episode of heart failure?]

Yes

No

Hospitalisation

3. Did heart failure lead to an unplanned visit to a health care clinic?

[Did heart failure lead to an unplanned visit to a health care clinic?]

No

Yes


/ /

: 24-hour clock

Discharge date and time

/ /

: 24-hour clock

4. Did heart failure lead to an unplanned visit to the Emergency Room?

[Did heart failure lead to an unplanned visit to the Emergency Room?]

No

Yes


/ /

: 24-hour clock


/ /

: 24-hour clock

5. Did heart failure lead to an unplanned admission to the hospital?

[Did heart failure lead to an unplanned admission to the hospital?]

No

Yes


/ /

: 24-hour clock


/ /

: 24-hour clock

NYHA Score

6. Functional status

[Functional status]

Select one:

Class I - No symptoms and no limitation in ordinary physical activity

Class II - Mild symptoms and slight limitation during ordinary activity; comfortable at rest

Class III - Marked limitation in activity due to symptoms, even during less-than-ordinary activity. Comfortable only at rest

Class IV - Severe limitations. Experiences symptoms even while at rest

Laboratory Data

7. Lab Name[Lab Name]

8. Lab Address

[Lab Address]

9. Brain natriuretic peptide Date and time sample taken

/ /

: 24-hour clock

Results:

Enter ’NR’ if the laboratory results is not available to report or if a lab error occurred.

Units:

High

Low

10. N-terminal pro-Brain natriuretic peptide Date and time sample taken

/ /

: 24-hour clock

Results:


Units:

High

Low

11. Atrial natriuretic peptide Date and time sample taken

/ /

: 24-hour clock

Results:Enter ’NR’ if the laboratory results is not available to report or if a lab error occurred.

Units:

High

Low

Current clinical signs and symptoms of heart failure

12. Heart failure signs[Heart failure signs]


>2+ lower extremity edema

Pulmonary crackles

Jugular venous distention

Tachypnea with respiratory rate >20

S3 cardiac gallop

Increasing abdominal distention and/or ascites

Weight gain

Hepatojugular reflux

13. Heart failure symptoms

[Heart failure symptoms]


Dyspnea

Orthopnea

Paroxysmal Nocturnal Dyspnea

Increasing fatigue/worsening exercise tolerance

Echocardiography

14. Was an Echocardiogram performed?[Was an Echocardiogram performed?]

No

Yes, complete the following:

Date and Time of Echocardiogram


4

CONFID

ENTIA

L2017N

324761_00206215

PPD

/ /

: 24-hour clock

Ejection Fraction Assessment

No

Yes, record percentage

Evidence of diastolic dysfunction

Yes

No

Evidence of significant valvular disease

Yes

No

Right heart catheterisation

15. Was a right heart catheterisation performed?

[Was a right heart catheterisation performed?]

No

Yes, did RHC demonstrate Pulmonary Capillary Wedge Pressure >18mmHg or Cardiac output <2.2 l/min/m2?

Yes

No

Chest X-ray

16. Was a Chest X-ray performed?

[Was a Chest X-ray performed?]

No


Date and Time of the Chest X-ray

/ /

: 24-hour clock

Evidence of pulmonary edema

Yes

No

Pleural Effusion

Yes

No

Evidence of cardiomegaly

Yes

No

MUGA

17. Was a Multiple Gated Acquisition Scan (MUGA) performed?

[Was a Multiple Gated Acquisition Scan (MUGA) performed?]

No


Date and Time of the MUGA

/ /

: 24-hour clock

Ejection Fraction Assessment

No

Yes, record percentage

Therapy

18. Were any changes to the subject's treatment for heart failure required during their clinic or emergency room visit or hospitalisation?

[Were any changes to the subject's treatment for heart failure required during their clinic or emergency room visit or hospitalisation?]

Yes

No

19. Escalation of a chronic cardiovascular medication

[Escalation of a chronic cardiovascular medication]

Yes

No

20. Increase in oral diuretics

[Increase in oral diuretics]

Yes

No

21. New class of medication added

[New class of medication added]

Yes

No

22. Intravenous diuretics

[Intravenous diuretics]

Yes

No

23. Inotropes

[Inotropes]

Yes

No

24. Intravenous vasodilators

[Intravenous vasodilators]

Yes

No

25. Mechanical assisted device

[Mechanical assisted device]

No


Intra-aortic balloon placement

Yes

No

Ventricular assist device insertion

Yes

No

Final primary diagnosis

26. Final primary diagnosis

[Final primary diagnosis]

Select one:

Dilated cardiomyopathy

Hypertrophic cardiomyopathy

Restrictive cardiomyopathy

Arrhythmogenic right ventricular cardiomyopathy

Unclassified cardiomyopathies

Myocarditis

Other, Specify:


5

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: STUDY CONCLUSION (CONC)

STUDY CONCLUSION

1. Date of subject completion or withdrawal

[Date of subject completion or withdrawal]

/ /

2. Was the subject withdrawn from the study?

[Was the subject withdrawn from the study?]

No

Yes

Primary reason for withdrawal

Adverse event

Record details on the Non-Serious Adverse Events or Serious Adverse Events forms as appropriate

Protocol Deviation

Check all that apply. If none, select "No Subreasons":

No subreasons

Pregnancy

Lack of adherence

Prohibited medication use

Study closed/terminated

Investigator site closed, specify

Investigator discretion, specifySelect this reason if none of the other primary reasons are appropriate.

Withdrew consent

Select this reason if none of the other primary reasons are appropriate.

Specify:



6

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: CONCOMITANT MEDICATIONS (CONMEDS) - Repeating Form

# Drug name

(Trade name preferred)

Unit Dose Units Frequency Route code Device used to administer medication Reason for medication Start date Taken prior to study? Ongoing medication? Medication type

1

CONCOMITANT MEDICATIONS

1. Drug Name

(Trade Name preferred)

[Drug name

(Trade name preferred)]

2. Unit Dose

[Unit Dose]

3. Units

[Units]

4. Frequency

[Frequency]

5. Route code

[Route code]

6. Device used to administer medication

[Device used to administer medication]

7. Reason for Medication

[Reason for medication]

8. Start date

[Start date]

/ /

9. Taken prior to study?

[Taken prior to study?]

Yes

No

10. Ongoing medication?

[Ongoing medication?]

Yes

No, please enter End date

/ /

11. Medication type

[Medication type]

COPD

Non-COPD



7

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: CEREBROVASCULAR EVENTS STROKE AND TRANSIENT ISCHEMIC ATTACK (CVA / TIA) - Repeating Form

# Seq History and/or clinical

examination which clearly

defines the new onset of

focal or global neurological

deficit

Newly defined

brain lesion

consistent with

signs and

symptoms

Was subject

hospitalised due

to event?

Was event related

to occurrence of

arrhythmia(s)?

Was event

due to

trauma?

Motor and/or

sensory loss in

face, arm, leg

right side

Motor and/or

sensory loss in

face, arm, leg

left side

Dysphasia/Aphasia

(difficulty with

language)

Dysarthria/Dysphagia

(difficulty with speech and

swallowing)

Hemianopsia/Dizziness/Vertigo Ataxia Nystagmus Diplopia Acute

confusion/cognitive

change

Decreased

consciousness

Did subject have

abnormal neurologic

exam or history

prior to event?

Was a CT

Scan

performed?

Was a MRI

Scan

performed?

Was a MRA

Scan

performed?

If no CT, MRI or MRA examination was

performed, what was the clinical

diagnosis of the cause of the event

(e.g., cerebral thrombosis,

hemorrhage, embolus).

Were there

any long-term

sequelae?

Final

Diagnosis

1

CEREBROVASCULAR EVENTS STROKE (CVA) AND TRANSIENT ISCHEMIC ATTACK (TIA)

1. Sequence Number

[Seq]

2. History and/or clinical examination which clearly defines the new onset of focal or global neurological deficit

[History and/or clinical examination which clearly defines the new onset of focal or global neurological deficit]

No

Yes, Date and time of onset of symptoms

/ /

: 24-hour clock

Did symptoms resolve?

No

Yes, Date and Time of Resolution of Symptoms

/ /

: 24-hour clock

3. Newly defined brain lesion consistent with signs and symptoms

[Newly defined brain lesion consistent with signs and symptoms]

Yes

No

Hospitalisation

4. Was subject hospitalised due to event?

[Was subject hospitalised due to event?]

No

Yes, Admission date and time

/ /

: 24-hour clock


/ /

: 24-hour clock

5. Was event related to occurrence of arrhythmia(s)?

[Was event related to occurrence of arrhythmia(s)?]

Yes

No

If Yes, complete the Arrhythmias form

6. Was event due to trauma?[Was event due to trauma?]

Yes

No

Symptoms

7. Motor and/or sensory loss in face, arm, leg right side

[Motor and/or sensory loss in face, arm, leg right side]

No


Yes


8. Motor and/or sensory loss in face, arm, leg left side

[Motor and/or sensory loss in face, arm, leg left side]

No


Yes


9. Dysphasia/Aphasia (difficulty with language)

[Dysphasia/Aphasia (difficulty with language)]

No


Yes


10. Dysarthria/Dysphagia (difficulty with speech and swallowing)

[Dysarthria/Dysphagia (difficulty with speech and swallowing)]

No


Yes


11. Hemianopsia/Dizziness/Vertigo[Hemianopsia/Dizziness/Vertigo]

No


Yes


12. Ataxia

[Ataxia]

No


Yes


13. Nystagmus

[Nystagmus]

No


Yes


14. Diplopia

[Diplopia]

No


Yes


15. Acute confusion/cognitive change

[Acute confusion/cognitive change]

No


Yes


16. Decreased consciousness

[Decreased consciousness]

No


Yes



8

CONFID

ENTIA

L2017N

324761_00206215

PPD

17. Did subject have abnormal neurologic exam or history prior to event?

[Did subject have abnormal neurologic exam or history prior to event?]

No


Yes


Brain Imaging

18. Was a CT Scan performed?

[Was a CT Scan performed?]

No

Yes, date and time of the CT scan

/ /

: 24-hour clock

Evidence of hemorrhage

Yes

No

Evidence of hemorrhagic conversion

Yes

No

Evidence of infarction

Yes

No

Evidence of tumor

Yes

No

Evidence of aneurysm

Yes

No

19. Was a MRI Scan performed?

[Was a MRI Scan performed?]

No

Yes, date and time of the MRI scan

/ /

: 24-hour clock

Result of MRI scan

Normal

Abnormal, evidence of acute/sub acute event

Yes

No

Abnormal, evidence of chronic event

Yes

No

20. Was a MRA Scan performed?

[Was a MRA Scan performed?]

No

Yes, date and time of the MRA scan

/ /

: 24-hour clock

Result of MRA scan

Normal

Abnormal, evidence of acute/sub acute event

Yes

No

Abnormal, evidence of chronic event

Yes

No

21. If no CT, MRI or MRA examination was performed, what was the clinical diagnosis of the cause of the event (e.g., cerebral thrombosis, hemorrhage, embolus).

[If no CT, MRI or MRA examination was performed, what was the clinical diagnosis of the cause of the event (e.g., cerebral thrombosis, hemorrhage, embolus).]

Outcome

22. Were there any long-term sequelae?[Were there any long-term sequelae?]

No


Able to perform ADL’s (activities of daily living) without assistance?

Yes

No

Was the subject confined to bed?

Yes

No

Final Diagnosis

23. Final Diagnosis

[Final Diagnosis]

Select one:

Ischemic stroke

Hemorrhagic stroke, intracerebral

Hemorrhagic stroke, subarachnoid

Stroke - type uncertain

TIA


9

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: DEATH (DEATH)

DEATH CERTIFICATE

1. Date of death

[Date of death]

/ /

2. Cause of death listed on Death Certificate:

List the Immediate cause; mode of dying, such as cardiac arrest, respiratory arrest is insufficient information and should only be used when followed by conditions such as arrhythmia due to ischemic cardiac disease.[Cause of death listed on Death Certificate: ]

3. List the condition that gave rise to the immediate cause, e.g. renal failure (immediate cause) due to diabetes (condition) or myocardial infarction (immediate cause) due to coronary artery disease (condition).

[List the condition that gave rise to the immediate cause, ]

4. List the underlying cause.

[List the underlying cause.]

5. Certifier from death certificate:[Certifier from death certificate:]


Investigator

Primary Care Physician

Treating Physician

Medical Examiner

Other, specify:

6. Was autopsy performed?

[Was autopsy performed?]

No

Yes

If Yes, summarise findings in General Narrative of SAE form.

COURSE OF DEATH

7. Was the death witnessed?

[Was the death witnessed?]

No, the death was unwitnessed

When was subject last seen or heard alive?

/ /

: 24-hour clock

Yes, describe signs and symptoms that preceded death:

8. Was resuscitation attempted?

[Was resuscitation attempted?]

Yes

No, was this because there was a DNR order?

Yes

No

9. Where did death occur (e.g., hospital, home, street)?

[Where did death occur (e.g., hospital, home, street)?]

10. Did the subject have a condition that made him/her terminal or pre-terminal?

[Did the subject have a condition that made him/her terminal or pre-terminal?]

Yes

No

11. Was death sudden (without warning or within 1 hr of onset of symptoms)?

[Was death sudden (without warning or within 1 hr of onset of symptoms)?]

No

Yes, what were the circumstances?

During Sleep

Normal daily activity

During exertion

12. Did the subject have any new clinical signs or symptoms that may have indicated a potential cause of death?

[Did the subject have any new clinical signs or symptoms that may have indicated a potential cause of death?]

No

Yes, specify:

13. Was death attributed to a cardiovascular procedure or cardiovascular surgery?

[Was death attributed to a cardiovascular procedure or cardiovascular surgery?]

Yes

No

If Yes, complete Revascularization CRF.

14. Was death attributed to a non-cardiovascular procedure or surgery?

[Was death attributed to a non-cardiovascular procedure or surgery?]

Yes

No

PRIMARY CAUSE OF DEATH

15. Primary cause of death

[Primary cause of death]

Select only one:

Myocardial infarction

Heart failure

Cardiac arrhythmia

Stroke

Pulmonary Embolism (PE)

Haemorrhage - specify organ:

Other cardiovascular diagnosis - specify:

Cancer

Disease under study

Other cancer - specify:

Sepsis - note source of sepsis (e.g., pneumonia, etc.)

Trauma - specify:

Suicide - specify mode of death:

Did the subject have risk factors for suicide?

No

Yes, specify:

Other non-cardiovascular diagnosis - specify:

SECONDARY CAUSE OF DEATH

16. Were there any secondary cause(s) of death?

[Were there any secondary cause(s) of death?]

No

Yes, (Check all that apply)

Myocardial infarction (Complete MI/Unstable Angina eCRF.)

Heart failure (Complete Congestive Heart Failure eCRF.)

Cardiac arrhythmia (Complete Arrhythmias eCRF.)

Stroke (Complete Cerebrovascular Events Stroke/TIA eCRF.)

Pulmonary Embolism (PE) (Complete Deep Venous Thrombosis/PE eCRF.)

Haemorrhage - specify organ:


10

CONFID

ENTIA

L2017N

324761_00206215

PPD

Other cardiac diagnosis - specify:

Cancer

Disease under study

Other cancer - specify:

Sepsis - note source of sepsis (e.g., pneumonia, etc.)

Trauma - specify:

Suicide - specify mode of death:

Did the subject have risk factors for suicide?

No

Yes, specify:

Other non-cardiovascular diagnosis - specify:

MEDICAL / CLINICAL HISTORY

17. Was there a particular risk of death from clinical history?

[Was there a particular risk of death from clinical history?]

No

Yes

If Yes, is death associated with SAE?

Yes

No, death is not associated with SAE, please record details:

Ensure relevant clinical history is recorded on the Medical Conditions form.

18. Is there any other information (test results, medical history, etc.) you think would help understand why this subject died?

[Is there any other information (test results, medical history, etc.) you think would help understand why this subject died?]

No

Yes


Yes

No, death is not associated with an SAE, please record details:

19. Were there any new medications or changes in doses of chronic medications within the four weeks prior to death that may have contributed to death?

[Were there any new medications or changes in doses of chronic medications within the four weeks prior to death that may have contributed to death?]

No

Yes


Yes

No, death is not associated with an SAE, please provide drug name and provide reason:

Ensure all medications up till the time of death are recorded on the Concomitant Medication form, including the new medications or changes in chronic medications. For studies where applicable, ensure the

column,’Was medication stopped due to toxicity?’ has been completed.


11

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: DEMOGRAPHY (DEMO)

DEMOGRAPHY

1. Year of birth

[Year of birth]

2. Sex

[Sex]

Male

Female

3. Ethnicity

[Ethnicity]

Hispanic or Latino

Not Hispanic or Latino

4. Geographic Ancestry

[Geographic Ancestry]

Check all that apply

African American/African Heritage

American Indian or Alaskan Native

Asian - Central/South Asian Heritage

Asian - East Asian Heritage

Asian - Japanese Heritage

Asian - South East Asian Heritage


White - Arabic/North African Heritage

White - White/Caucasian/European Heritage

SUBJECT STATUS



12

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: DISEASE DURATION (DISDUR)

DISEASE DURATION

1. Duration of disease

[Duration of disease]



13

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: DISKUS/ACCUHALER ERRORS AFTER FIRST HCP INSTRUCTION (DISKUS HCP1)

DISKUS/ACCUHALER ERRORS

1. Did the subject make an error while using the device?

[Did the subject make an error while using the device?]

No

Yes

Failed to open cover

Completed correctly

Not completed correctly

Lever is not pushed back

Completed correctly


Shook the device after dose preparation

Completed correctly


No exhalation before an inhalation

Completed correctly


Exhaled directly into mouthpiece

Completed correctly


No seal by the lips round the mouthpiece during the inhalation

Completed correctly


Inhalation manoeuvre was not:

- steady

- deep

Completed correctly


Did not hold breath

Completed correctly


Did not close the device (Note: this is an error but one which does not affect the medication that is inhaled)

Completed correctly


Any other comments:


14

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: DISKUS/ACCUHALER ERRORS AFTER SECOND HCP INSTRUCTION (DISKUS HCP2)




No

Yes


Completed correctly



Completed correctly



Completed correctly



Completed correctly



Completed correctly



Completed correctly



- steady

- deep

Completed correctly


Did not hold breath

Completed correctly



Completed correctly


Any other comments:


15

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: DISKUS/ACCUHALER ERRORS AFTER READING LEAFLET (DISKUS LEAFLET)




No

Yes


Completed correctly



Completed correctly



Completed correctly



Completed correctly



Completed correctly



Completed correctly



- steady

- deep

Completed correctly


Did not hold breath

Completed correctly



Completed correctly


Any other comments:


16

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: DATE OF VISIT/ASSESSMENT (DOV)

Record in the LOGS Visit details of any new Serious Adverse Events related to study procedures from the time of signing the Informed Consent Form. From Randomization onwards, record in the LOGS Visit details of any new Adverse Events observed or reported by the subject or any changes to ongoing Adverse Events in the appropriate Non-Serious Adverse Event/Serious Adverse Event eCRF.

Record any new medication taken or changes to the subject's concomitant medication since the last visit in the appropriate Concomitant Medication eCRF form in the LOGS visit.

DATE OF VISIT/ASSESSMENT

1. Date of visit/assessment

[DOV]

/ /


17

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: DEEP VEIN THROMBOSIS (DVT) /PULMONARY EMBOLISM (PE) (DVT) - Repeating Form

# Seq Was the subject

hospitalised due

to event?

Date and

time of

onset

Calf

tenderness

Calf

swelling

Femoral

vein signs

Surgical

procedures within

the past 12 weeks

Other typical

symptoms and

signs of DVT

Date and

time of

onset

Surgical

procedures within

the past 12 weeks

Hypotension Requiring

vasopressor

support

Shortness

of breath

Pleuritic

chest pain

Tachycardia, Heart

rate >100/minute

Other typical sign

and symptoms

consistent with PE

Known

hypercoagulable

state

Prolonged

immobilisation

Postoperative Recent

severe

trauma

History of

prior DVT

or PE

Impedance

plethysmography

Lower extremity

compression

ultrasonograph

Venography MRI

Scan

CT

Scan

Angiography Ventilation -

Perfusion

Scan

D-

dimer

Did the subject

require the

following

treatment?

1

DEEP VEIN THROMBOSIS (DVT) /PULMONARY EMBOLISM (PE)

1. Sequence Number

[Seq]

2. Was the subject hospitalised due to event?[Was the subject hospitalised due to event?]

No


/ /

: 24-hour clock

Deep Vein Thrombosis (DVT)

3. Date and time of onset

[Date and time of onset]

/ /

: 24-hour clock

4. Calf tenderness

[Calf tenderness]

Yes

No

5. Calf swelling[Calf swelling]

Yes

No

6. Femoral vein signs

[Femoral vein signs]

Yes

No

7. Surgical procedures within the past 12 weeks

[Surgical procedures within the past 12 weeks]

Yes

No

8. Other typical symptoms and signs of DVT

[Other typical symptoms and signs of DVT]

Yes

No

Pulmonary Embolism (PE)

9. Date and time of onset

[Date and time of onset]

/ /

: 24-hour clock

10. Surgical procedures within the past 12 weeks

[Surgical procedures within the past 12 weeks]

Yes

No

11. Hypotension

[Hypotension]

Yes

No

12. Requiring vasopressor support

[Requiring vasopressor support]

Yes

No

13. Shortness of breath

[Shortness of breath]

No

Yes, complete the following

Mild

Moderate

Severe

14. Pleuritic chest pain[Pleuritic chest pain]

Yes

No

15. Tachycardia, Heart rate >100/minute

[Tachycardia, Heart rate >100/minute]

Yes

No

16. Other typical sign and symptoms consistent with PE

[Other typical sign and symptoms consistent with PE]

Yes

No

Risk Factors

17. Known hypercoagulable state

[Known hypercoagulable state]

Yes

No

18. Prolonged immobilisation

[Prolonged immobilisation]

Yes

No

19. Postoperative

[Postoperative]

Yes

No

20. Recent severe trauma

[Recent severe trauma]

Yes

No

21. History of prior DVT or PE

[History of prior DVT or PE]

Yes

No

Diagnostic tests

22. Impedance plethysmography

[Impedance plethysmography]

Was test performed?

No

Yes, Date and Time of Test

/ /

: 24-hour clock

Consistent with DVT?

Yes

No

23. Lower extremity compression ultrasonograph

[Lower extremity compression ultrasonograph]

Was test performed?

No

Date and Time of Test

/ /

: 24-hour clock


Yes

No

24. Venography

[Venography]

Was test performed?

No


/ /

: 24-hour clock


Yes

No

25. MRI Scan

[MRI Scan]

Was test performed?

No


/ /

: 24-hour clock


Yes

No

26. CT Scan

[CT Scan]

Was test performed?

No


/ /

: 24-hour clock

Evidence of PE?

Yes

No

27. Angiography

[Angiography]

Was test performed?

No


18

CONFID

ENTIA

L2017N

324761_00206215

PPD


/ /

: 24-hour clock

Evidence of PE?

Yes

No

28. Ventilation - Perfusion Scan

[Ventilation - Perfusion Scan]

Was test performed?

No


/ /

: 24-hour clock

Evidence of PE?

Yes

No

29. D-dimer

[D-dimer]

Was test performed?

No


/ /

: 24-hour clock

Elevated?

No

Yes

Medications and Procedures

30. Did the subject require the following treatment?

[Did the subject require the following treatment?]

Thrombolytics

Yes

No

Thrombectomy

Yes

No

Anti-coagulation

Yes

No


19

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: ELIGIBILITY (ELIG)

ELIGIBILITY REVIEW

1. Protocol version

Note: Refer to the protocol cover page for the protocol version (YYYYN000000_00).

[Protocol version]

2. Did the subject meet all the entry criteria?

[Did the subject meet all the entry criteria?]

Yes

No. Please complete an entry below for each inclusion/exclusion criteria not met.

Criteria type Criteria number failed

3.

�

Entry

3.1 Criteria type

[Criteria type]

Inclusion criteria

Exclusion criteria

3.2 Criteria number failed

[Criteria number failed]


20

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: ELLIPTA ERRORS AFTER FIRST HCP INSTRUCTION (ELLIPTA HCP1)

ELLIPTA ERRORS

Critical Errors are identified in bold text in the list below.



No

Yes


Completed correctly



Completed correctly



Completed correctly



Completed correctly



Completed correctly



- long

- steady

- deep

Completed correctly


Blocked air inlet during inhalation manouevre

Completed correctly


Did not hold breath

Completed correctly



Completed correctly


Any other comments:


21

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: ELLIPTA ERRORS AFTER SECOND HCP INSTRUCTION (ELLIPTA HCP2)

ELLIPTA ERRORS




No

Yes


Completed correctly



Completed correctly



Completed correctly



Completed correctly



Completed correctly



- long

- steady

- deep

Completed correctly



Completed correctly


Did not hold breath

Completed correctly



Completed correctly


Any other comments:


22

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: ELLIPTA ERRORS AFTER READING LEAFLET (ELLIPTA LEAFLET)

ELLIPTA ERRORS




No

Yes


Completed correctly



Completed correctly



Completed correctly



Completed correctly



Completed correctly



- long

- steady

- deep

Completed correctly



Completed correctly


Did not hold breath

Completed correctly



Completed correctly


Any other comments:


23

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: INFORM ENROLLMENT (ENROL)

SUBJECT NUMBER

1. Subject number

[Subject number]


24

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: FAMILY HISTORY (FAMHIST)

FAMILY HISTORY-CARDIOVASCULAR RISK FACTORS

1. Is there a family history of premature coronary artery disease in women <65 years or men < 55 years in first degree relatives only (e.g., biological mother or father, biological brother or sister, biological son or daughter)?

Half siblings are considered first degree relatives

[Is there a family history of premature coronary artery disease?]

Yes

No

Unknown


25

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: PROTOCOL DEVIATION (GSK) - Repeating Form

# Protocol deviation category Protocol Deviation Description Date/time of deviation

1

1. Protocol deviation category

[Protocol deviation category]

Informed consent

Signed informed consent/assent not available on site

Wrong informed consent/assent version signed

Informed consent/assent not signed and/or dated by subject (parent/Legally Acceptable representative, if applicable)

Informed consent/assent not signed and/or dated by appropriate site staff

Informed consent/assent not signed prior to any study procedure

Other informed consent/assent deviations

Eligibility criteria not met

Not withdrawn after developing withdrawal criteria

Not withdrawn from study

Not discontinued from study treatment

Visit Completion

Out of window visit/phone contact

Assessment or time point completion

Missed assessment

Assessment not properly performed

Wrong study treatment/administration/dose

Study treatment not administered per protocol

Wrong study treatment or assignment administered

Expired study treatment administered

Use of study treatment impacted by a temperature excursion which was not reported or approved or which was disapproved for further use

Study treatment not available at site for administration

Other deviations related to wrong study treatment/administration/dose

Study Procedures

Randomisation procedures (e.g. subject assigned to wrong stratum, subject randomised out of order)

Equipment procedures

Other deviation from study procedures

Failure to report safety events per protocol:

SAE not reported within the expected time frame

Failure to confirm causality assessment within the expected time frame

Pregnancy

Other

2. Protocol Deviation Description

[Protocol Deviation Description]

3. Date/time of deviation

[Date/time of deviation]

/ /

: 24-hour clock



26

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: HANDIHALER ERRORS AFTER FIRST HCP INSTRUCTION (HANDIHALER HCP1)

HANDIHALER ERRORS



No

Yes

Failed to remove capsule

Completed correctly


Failed to insert capsule into the chamber

Completed correctly


Did not completely close device capsule chamber (heard click when satisfactory)

Completed correctly


Did not pierce the capsule (HCP should check capsule was pierced)

Completed correctly



Completed correctly



Completed correctly



Completed correctly



Completed correctly



- slow

- deep

Completed correctly


Capsule did not rattle

Completed correctly



Completed correctly


Did not hold breath

Completed correctly


Did not check inside the capsule chamber if powder was left / did not make a second inhalation

Completed correctly


Any other comments:


27

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: HANDIHALER ERRORS AFTER SECOND HCP INSTRUCTION (HANDIHALER HCP2)

HANDIHALER ERRORS



No

Yes


Completed correctly



Completed correctly



Completed correctly



Completed correctly



Completed correctly



Completed correctly



Completed correctly



Completed correctly



- slow

- deep

Completed correctly



Completed correctly



Completed correctly


Did not hold breath

Completed correctly



Completed correctly


Any other comments:


28

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: HANDIHALER ERRORS AFTER READING LEAFLET (HANDIHALER LEAFLET)

HANDIHALER ERRORS



No

Yes


Completed correctly



Completed correctly



Completed correctly



Completed correctly



Completed correctly



Completed correctly



Completed correctly



Completed correctly



- slow

- deep

Completed correctly



Completed correctly



Completed correctly


Did not hold breath

Completed correctly



Completed correctly


Any other comments:


29

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: INHALER USE STATUS (INH USE STATUS)

NAIVE TO DEVICE

1. Subject is naive to which inhaler devices?

[Subject is naive to which inhaler devices?]

ELLIPTA

DISKUS + HandiHaler


INHALERS USED

2. Which sequence of inhalers was used by the subject?

[Which sequence of inhalers was used by the subject?]

Sub-Study 1 Seq A: ELLIPTA first then DISKUS + HandiHaler (Preference Qx 1)

Sub-Study 1 Seq B: DISKUS + HandiHaler first then ELLIPTA (Preference Qx 2)

Sub-Study 1 Seq C: ELLIPTA first then DISKUS + HandiHaler (Preference Qx 2)

Sub-Study 1 Seq D: DISKUS + HandiHaler first then ELLIPTA (Preference Qx 1)

Sub-Study 2 Seq E: ELLIPTA first then Turbuhaler + HandiHaler (Preference Qx 3)

Sub-Study 2 Seq F: Turbuhaler + HandiHaler first then ELLIPTA (Preference Qx 4)

Sub-Study 2 Seq G: ELLIPTA first then Turbuhaler + HandiHaler (Preference Qx 4)

Sub-Study 2 Seq H: Turbuhaler + HandiHaler first then ELLIPTA (Preference Qx 3)


30

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: INHALER PREFERENCE QUESTIONNAIRE VERSION 1 (IPQV1)

Instructions: Please complete the following questions related to the treatments you used during this study. Check only one response for each question.


[Which treatment do you prefer based on the number of steps needed to take your COPD medication?]

DISKUS and HandiHaler

ELLIPTA

No Preference


[Which treatment do you prefer for taking your COPD medication?]


ELLIPTA

No Preference

3. ISO language code

[ISO language code]

Dutch; Flemish

English


31

CONFID

ENTIA

L2017N

324761_00206215

PPD





ELLIPTA


No Preference



ELLIPTA


No Preference


[ISO language code]

Dutch; Flemish

English


32

CONFID

ENTIA

L2017N

324761_00206215

PPD





Turbuhaler and HandiHaler

ELLIPTA

No Preference




ELLIPTA

No Preference


[ISO language code]

Dutch; Flemish

English


33

CONFID

ENTIA

L2017N

324761_00206215

PPD





ELLIPTA


No Preference



ELLIPTA


No Preference


[ISO language code]

Dutch; Flemish

English


34

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: LOG STATUS (LOG STATUS)

LOG STATUS

1. Did the subject experience any non-serious adverse events during the study?

[Any AEs?]

No

Yes

2. Did the subject experience a cardiovascular event during the study?

[Any CV event during the study?]

No

Yes



35

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: MEDICAL CONDITIONS (MEDHX)

CARDIAC DISORDERS

1. Coronary Artery Disease

[Coronary Artery Disease]

Current

Past

Not Assessed

No Medical Condition

2. Myocardial Infarction

[Myocardial Infarction]

Current

Past

Not Assessed


3. Arrhythmia

[Arrhythmia]

Current

Past

Not Assessed


4. Congestive Heart Failure

[Congestive Heart Failure]

Current

Past

Not Assessed


ENDOCRINE DISORDERS

5. Cushing's Syndrome

[Cushing's Syndrome]

Current

Past

Not Assessed


6. Adrenal Suppression

[Adrenal Suppression]

Current

Past

Not Assessed


EYE DISORDERS

7. Cataract

[Cataract]

Current

Past

Not Assessed


8. Glaucoma[Glaucoma]

Current

Past

Not Assessed


VASCULAR DISORDERS

9. Hypertension

[Hypertension]

Current

Past

Not Assessed


10. Cerebrovascular accident

[Cerebrovascular accident]

Current

Past

Not Assessed


METABOLISM AND NUTRITION DISORDERS

11. Hypercholesterolemia

[Hypercholesterolemia]

Current

Past

Not Assessed


12. Diabetes Mellitus

[Diabetes Mellitus]

Current

Past

Not Assessed


13. Osteoporosis

[Osteoporosis]

Current

Past

Not Assessed


INFECTIONS AND INFESTATIONS

14. Pneumonia

[Pneumonia]

Current

Past

Not Assessed




36

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: MYOCARDIAL INFARCTION (MI) / UNSTABLE ANGINA (UA) (MI) - Repeating Form

# Seq Date and time of onset of

Myocardial

Infarction/Unstable

Angina symptoms

Duration of

symptoms at

time of

presentation

New onset of

severe angina

or accelerated

angina

Angina

at rest

Exertional

angina

Atypical

symptoms

Did the

anginal/infarction

occur after medical or

surgical procedure?

Did the subject

visit the

emergency

room/chest pain

centre?

Was subject

admitted to

the

hospital?

Was the subject on any of the

following medications (anti-

angina, antithrombotic agents,

anti-arrhythmias or other

relevant drugs) at the time the

event occured?

Are symptoms

consistent with

MI (e.g.,

prolonged chest

pain, etc.)

Was an ECG

performed?

Is there an ECG

prior to current

event available for

comparison?

Lab

Name

Lab

Address

Peak

Creatine

Kinase

Peak Creatine

Kinase MB - mass

(concentration)

Peak Creatine

Kinase MB -

mass

(percentage)

Peak Creatine

Kinase MB -

activity

(concentration

Peak

Troponin

I

Peak

Troponin

T

Stress

Test

Echo Nuclear MRI Coronary

Angiogram

Angioplasty Coronary

Artery

Bypass

Graft

Final Diagnosis

If this event is

reported as a Serious

Adverse Event (SAE),

ensure that this

diagnosis is reported on the SAE form

1

Myocardial Infarction/Unstable Angina

1. Sequence Number

[Seq]

2. Date and time of onset of Myocardial Infarction/Unstable Angina symptoms

[Date and time of onset of Myocardial Infarction/Unstable Angina symptoms]

/ /

: 24-hour clock

3. Duration of symptoms at time of presentation

[Duration of symptoms at time of presentation]

hrs min

Angina Symptoms

4. New onset of severe angina or accelerated angina

[New onset of severe angina or accelerated angina]

Yes

No

5. Angina at rest

[Angina at rest]

Yes

No

6. Exertional angina

[Exertional angina ]

Yes

No

7. Atypical symptoms

[Atypical symptoms]

Yes

No

8. Did the anginal/infarction occur after medical or surgical procedure?

[Did the anginal/infarction occur after medical or surgical procedure?]

Yes

No

Urgent care and/or hospitalisation

9. Did the subject visit the emergency room/chest pain centre?[Did the subject visit the emergency room/chest pain centre?]

No

Yes, Date and Time of the emergency room/chest pain center visit

/ /

: 24-hour clock

10. Was subject admitted to the hospital?

[Was subject admitted to the hospital?]

No


/ /

: 24-hour clock

11. Was the subject on any of the following medications (anti-angina, antithrombotic agents, anti-arrhythmias or other relevant drugs) at the time the event

occured?

[Was the subject on any of the following medications (anti-angina, antithrombotic agents, anti-arrhythmias or other relevant drugs) at the time the event

occured?]

No

Yes, Specify:

Drug name 1


Drug name 2


Drug name 3


Drug name 4


Drug name 5


Drug name 6


Drug name 7


12. Are symptoms consistent with MI (e.g., prolonged chest pain, etc.)

[Are symptoms consistent with MI (e.g., prolonged chest pain, etc.)]

Yes

No

ECG Standard 12-LEAD

13. Was an ECG performed?

[Was an ECG performed?]

No

Non Evaluable


Date of ECG

/ /


Conduction

Left bundle branch block

ECG Findings

Myocardial infarction, old

Non-specific ST-T changes

ST elevation

ST depression

T-wave flattening/inversion

Pathological Q waves

Persistent ST elevation

Transient ST elevation

Dynamic horizontal/down-sloping depression

14. Is there an ECG prior to current event available for comparison?[Is there an ECG prior to current event available for comparison?]

No


Date of ECG

/ /

Previous ECG Findings


Myocardial infarction, old

Non-specific ST-T changes

ST elevation

ST depression

T-wave flattening/inversion

Pathological Q waves

Laboratory Data

15. Lab Name

[Lab Name]


37

CONFID

ENTIA

L2017N

324761_00206215

PPD

16. Lab Address

[Lab Address]

17. Peak Creatine Kinase

[Peak Creatine Kinase]

Date Sample Taken

/ /

Results:


Units:

High

Low

18. Peak Creatine Kinase MB - mass (concentration)

[Peak Creatine Kinase MB - mass (concentration)]

Date Sample Taken

/ /

Results:


Units:

High

Low

19. Peak Creatine Kinase MB - mass (percentage)

[Peak Creatine Kinase MB - mass (percentage)]

Date Sample Taken

/ /

Results:


Units:

High

Low

20. Peak Creatine Kinase MB - activity (concentration)[Peak Creatine Kinase MB - activity (concentration]

Date Sample Taken

/ /

Results:


Units:

High

Low

21. Peak Troponin I

[Peak Troponin I]

Date Sample Taken

/ /

Results:


Units:

High

Low

22. Peak Troponin T

[Peak Troponin T]

Date Sample Taken

/ /

Results:


Units:

High

Low

Imaging Reports

23. Stress Test

[Stress Test]

Was test done?

No


/ /

: 24-hour clock

What is the interpretation of result?

Normal

Abnormal

Is there evidence of ischemia?

Yes

No

Is there evidence of infarction?

Yes

No

24. Echo

[Echo]

Was test done?

No


/ /

: 24-hour clock


Normal

Abnormal


Yes

No


Yes

No

25. Nuclear

[Nuclear]

Was test done?

No


/ /

: 24-hour clock



38

CONFID

ENTIA

L2017N

324761_00206215

PPD

Normal

Abnormal


Yes

No


Yes

No

26. MRI

[MRI]

Was test done?

No


/ /

: 24-hour clock


Normal

Abnormal


Yes

No


Yes

No

Surgical/Medical Procedures

27. Coronary Angiogram

[Coronary Angiogram]

Was the procedure done?

No

Yes

Date and Time of Surgery/Procedure

/ /

: 24-hour clock

What is the interpretation of the result?

Normal

Abnormal

Is there evidence of significant lesion in any major epicardial

(50% Left Main Coronary Artery or 70% in any vessel)?

Yes

No

Number of vessels affected

Is there evidence of one or more stents?

Yes

No

Is there evidence of any stent thrombosis?

Yes

No

Unknown

Has ejection fraction been evaluated?

No

Yes, percentage of ejection fraction?

Surgical/Medical Procedures (Continued)

28. Angioplasty

[Angioplasty]


No

Yes, Date and Time of Surgery/Procedure

/ /

: 24-hour clock

29. Coronary Artery Bypass Graft

[Coronary Artery Bypass Graft]


No

Yes, Date and Time of Surgery/Procedure

/ /

: 24-hour clock

Final Diagnosis

30. Final Diagnosis

If this event is reported as a Serious Adverse Event (SAE), ensure that this diagnosis is reported on the SAE form

[Final Diagnosis

If this event is reported as a Serious Adverse Event (SAE), ensure that this diagnosis is reported on the SAE form]

Select one:

Unstable angina

Myocardial Infarction - ST segment elevation

Myocardial Infarction - Non-ST segment elevation

Non-cardiac chest pain


39

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: PERIPHERAL ARTERIAL THROMBOEMBOLISM (PAT) - Repeating Form

# Seq Date and time of onset of thromboembolism Typical signs and symptoms of acute vascular occlusion Risk factors present Medication Percutaneous Surgical Ultrasound CT MRI Angiography

1

Peripheral arterial thromboembolism

1. Sequence Number

[Seq]

2. Date and time of onset of thromboembolism

[Date and time of onset of thromboembolism]

/ /

: 24-hour clock

Symptoms

3. Typical signs and symptoms of acute vascular occlusion

[Typical signs and symptoms of acute vascular occlusion]

No


Loss of palpable pulse

Yes

No

Acute signs and symptoms

Yes

No

Chronic + subchronic signs and symptoms

Yes

No

Risk factors

4. Risk factors present

[Risk factors present]

No


Afib/Flutter

Yes

No

Hypercoagulable state

Yes

No

Malignancy

Yes

No

Known atherosclerotic process

Yes

No

Other risk factors

Yes

No

Intervention required or given for this event

5. Medication[Medication]

Yes

No

6. Percutaneous

[Percutaneous]

Yes

No

7. Surgical

[Surgical]

Yes

No

Diagnostic Tests

8. Ultrasound

[Ultrasound]

Was test performed?

No


Consistent with peripheral arterial thromboembolism

Yes

No

Thromboembolism within a stent?

Yes

No

Location

Upper extremity

Lower extremity

Renal

Mesentric

Splenic

Hepatic

Occular/Retinal

Stent thrombosis

Other location, specify


/ /

: 24-hour clock

9. CT

[CT]

Was test performed?

No



Yes

No


Yes

No

Location

Upper extremity

Lower extremity

Renal

Mesentric

Splenic

Hepatic

Occular/Retinal

Stent thrombosis



/ /

: 24-hour clock

10. MRI

[MRI]

Was test performed?

No



Yes

No


Yes

No

Location

Upper extremity

Lower extremity

Renal

Mesentric


40

CONFID

ENTIA

L2017N

324761_00206215

PPD

Splenic

Hepatic

Occular/Retinal

Stent thrombosis



/ /

: 24-hour clock

11. Angiography

[Angiography]

Was test performed?

No



Yes

No


Yes

No

Location

Upper extremity

Lower extremity

Renal

Mesentric

Splenic

Hepatic

Occular/Retinal

Stent thrombosis



/ /

: 24-hour clock


41

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: PREGNANCY INFORMATION (PREG)

PREGNANCY INFORMATION

1. Did the subject become pregnant during the study?

If Yes, complete the paper Pregnancy Notification form

[Did the subject become pregnant during the study?

If Yes, complete the paper Pregnancy Notification form]

No

Yes


42

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: PULMONARY HYPERTENSION (PUL HYP) - Repeating Form

# Seq Shortness of

breath

Chronic obstructive

pulmonary disease

Mitral

Stenosis

History of

pulmonary emboli

Current acute

pulmonary emboli

History of medication known to cause pulmonary hypertension

(e.g., Fenfluramine, Amphetamine, Ergotamine)

Other Clinical signs of right

ventricular hypertrophy

BP Was a Chest X-ray

performed?

Was an Echocardiogram

performed?

Was a ventilation/perfusion scan

performed?

Were pulmonary function

tests performed?

Was a right heart catheterisation

test performed?

Was a study other than echocardiogram performed that

reported left ventricular function?

1

History

1. Sequence Number

[Seq]

2. Shortness of breath

[Shortness of breath]

No

Yes, record details below:

Mild

Moderate

Severe

Underlying disorder predisposing to pulmonary hypertension

3. Chronic obstructive pulmonary disease

[Chronic obstructive pulmonary disease]

Yes

No

4. Mitral Stenosis

[Mitral Stenosis]

Yes

No

5. History of pulmonary emboli

[History of pulmonary emboli]

Yes

No

6. Current acute pulmonary emboli

[Current acute pulmonary emboli]

Yes

No

If Yes, complete the Deep Venous Thrombosis (DVT) Pulmonary Embolism (PE) form.

7. History of medication known to cause pulmonary hypertension (e.g., Fenfluramine, Amphetamine, Ergotamine)

[History of medication known to cause pulmonary hypertension (e.g., Fenfluramine, Amphetamine, Ergotamine)]

Yes

No

8. Other

[Other]

No

Yes, Specify:

Physical

9. Clinical signs of right ventricular hypertrophy

[Clinical signs of right ventricular hypertrophy]

Yes

No

10. Blood pressure

[BP]

/ mmHg mmHg

(systolic/diastolic)

Chest X-ray

11. Was a Chest X-ray performed?

[Was a Chest X-ray performed?]

No


Date and Time of the Chest X-ray

/ /

: 24-hour clock

Was Chest X-ray consistent with pulmonary hypertension?

Yes

No

Echocardiogram

12. Was an Echocardiogram performed?

[Was an Echocardiogram performed?]

No



/ /

: 24-hour clock

Was echocardiography consistent with pulmonary hypertension?

Yes

No

Right ventricular systolic blood pressure (mmHg)

Left ventricular ejection fraction (%)

Ventilation/perfusion scan

13. Was a ventilation/perfusion scan performed?

[Was a ventilation/perfusion scan performed?]

No


Date and Time of most recent ventilation/perfusion scan

/ /

: 24-hour clock

Was ventilation/perfusion scan consistent with pulmonary hypertension?

No


Acute pulmonary embolism

Yes

No

Acute massive pulmonary embolism

Yes

No

Chronic pulmonary embolism syndrome

Yes

No

Pulmonary function test

14. Were pulmonary function tests performed?[Were pulmonary function tests performed?]

No


Date and time of most recent pulmonary function test

/ /

: 24-hour clock

Evidence of obstructive disease?

No

Yes

Evidence of restrictive disease?

No

Yes

Right heart catheterisation

15. Was a right heart catheterisation test performed?

[Was a right heart catheterisation test performed?]

No


Date and time of catheterisation

/ /

: 24-hour clock

Was right heart catheterisation consistent with pulmonary hypertension

Yes

No

Was pulmonary vascular resistance increased

No


Pulmonary artery pressure:

Systolic (increased relative to normal pressure)

[b]


43

CONFID

ENTIA

L2017N

324761_00206215

PPD

Yes

No

Diastolic (increased relative to normal pressure)

Yes

No

Mean (increased relative to normal pressure)

Yes

No

Right ventricular systolic blood pressure (mmHg)

Was left ventricular end diastolic pressure increased?

No

Yes

Yes, please provide left ventricular end diastolic blood pressure (mmHg)

Other means of left ventricular assessment

16. Was a study other than echocardiogram performed that reported left ventricular function?

[Was a study other than echocardiogram performed that reported left ventricular function?]

No


Left heart catherization

No

Yes, date and time of study

/ /

: 24-hour clock

MUGA

No


/ /

: 24-hour clock

MRI

No


/ /

: 24-hour clock

Left ventricular ejection fraction (%)


44

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: RANDOMISATION (RAND)

RANDOMISATION NUMBER

1. Was the subject able to be randomised?

[Was subject randomised?]

No

Yes, provide:

Randomisation Number

Date randomisation number obtained

/ /


45

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: RESCREEN (RESCR)

RESCREEN

1. Was the subject rescreened?

[Was the subject rescreened?]

No

Yes, list all previous subject numbers

1.

2.


46

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: REVASCULARISATION (REVASCULARISATION) - Repeating Form

# Seq Was coronary revascularisation performed? Was peripheral revascularisation performed? Peripheral Revascularisation

Create a new row if procedure was performed on more than one location.

1

Coronary revascularisation

1. Sequence Number

[Seq]

2. Was coronary revascularisation performed?

[Was coronary revascularisation performed?]

No


Was a Coronary artery bypass graft procedure performed?

No


Elective/Urgent

Elective

Urgent

Number of Grafts

Date and Time of Procedure

/ /

: 24-hour clock

Was a Percutaneous coronary intervention procedure performed?

No


Elective/Urgent

Elective

Urgent

Vessel

Native

Vein graft

Mammary graft


/ /

: 24-hour clock

Was a Balloon angioplasty procedure performed?

No


Elective/Urgent

Elective

Urgent

Vessel

Native

Vein graft

Mammary graft


/ /

: 24-hour clock

Was a Drug-eluting coronary stent placement procedure performed?

No


Elective/Urgent

Elective

Urgent

Vessel

Native

Vein graft

Mammary graft


/ /

: 24-hour clock

Was a Bare metal coronary stent placement procedure performed?

No


Elective/Urgent

Elective

Urgent

Vessel

Native

Vein graft

Mammary graft


/ /

: 24-hour clock

Was an Atherectomy procedure performed?

No


Elective/Urgent

Elective

Urgent

Vessel

Native

Vein graft

Mammary graft


/ /

: 24-hour clock

Peripheral Revascularisation

3. Was peripheral revascularisation performed?

[Was peripheral revascularisation performed?]

No


Was Bypass surgery performed?

No


Elective

Urgent

Elective/Urgent

Location

Number of Grafts

/ / Date and Time of Procedure

: 24-hour clock

Was a Percutaneous revascularisation intervention procedure performed?

No


Elective/Urgent

Elective

Urgent

Location


47

CONFID

ENTIA

L2017N

324761_00206215

PPD

Vessel

Native

Graft


/ /

: 24-hour clock

Was an Angioplasty procedure performed?

No


Elective/Urgent

Elective

Urgent

Location

Vessel

Native

Graft


/ /

: 24-hour clock

Was a Vascular stent insertion procedure performed?

No


Elective/Urgent

Elective

Urgent

Location

Vessel

Native

Graft


/ /

: 24-hour clock

Specific procedure Elective/Urgent Location Indicate Vessel or Number of Grafts

(Vessel only applies to Percutaneous revascularisation intervention, Angioplasty and Vascular stent insertion. Number of grafts only applies to bypass surgery.)


4.

�

Peripheral Revascularisation

Create a new row if procedure was performed on more than one location. Entry

4.1 Specific procedure

[Specific procedure]

Bypass surgery

Percutaneous revascularisation intervention

Angioplasty

Vascular stent insertion

4.2 Elective/Urgent

[Elective/Urgent]

Elective

Urgent

4.3 Location[Location]

4.4 Indicate Vessel or Number of Grafts

(Vessel only applies to Percutaneous revascularisation intervention, Angioplasty and Vascular stent insertion. Number of grafts only applies to bypass surgery.)

[Indicate Vessel or Number of Grafts

(Vessel only applies to Percutaneous revascularisation intervention, Angioplasty and Vascular stent insertion. Number of grafts only applies to bypass surgery.)]

Vessel

Native

Graft

Number of Grafts

4.5 Date and Time of Procedure

[Date and Time of Procedure]

/ /

: 24-hour clock


48

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: SERIOUS ADVERSE EVENTS (SAE) - Repeating Form

# Did SAE occur after initiation of study

medication?

SERIOUS ADVERSE

EVENT

INTENSITY

CHANGES

Seriousness? RELEVANT CONCOMITANT/TREATMENT

MEDICATIONS

RELEVANT MEDICAL CONDITIONS/RISK

FACTORS

RELEVANT DIAGNOSTIC

RESULTS

Relevant diagnostic results not noted

above

If study treatment(s) were stopped temporarily, did the reported event(s) recur after study treatments products

were restarted?

General narrative

comments

1

If you wish to record a new SAE please determine if the new SAE is clinically or temporally related to an SAE previously entered on this form. If yes, record the details below using the 'Add Entry' button in this form. If not clinically or temporally related, create a new SAE form for this subject by clicking on the 'New' button at the top of the page. Do not record pre and post randomisation events on the same form.

TYPE OF REPORT

RANDOMISATION

1. Did SAE occur after initiation of study medication?

[Did SAE occur after initiation of study medication?]

No

Yes

Sequence Number Event Start Date Outcome / End Date Maximum Intensity Action taken Subject withdrawn? Relationship Was the SAE caused by activities related to study participation other than Study Treatment (e.g. procedures, blood draws, washout, etc)? Investigator reviewed

2.

�

SERIOUS ADVERSE EVENT Entry

Use the 'Add Entry' button to enter details of the SAE. For additional SAEs that are clinically or temporally related (e.g., SAEs that occur during the same hospitalisation) use the 'Add Entry' button to create a new row for entry of the additional SAE. Enter ONE event per row.

2.1 Sequence Number

[Sequence Number]

2.2 Event Diagnosis Only (if known) Otherwise Sign/Symptom

[Event]

2.3 Start Date

[Start Date]

/ /

2.4 Outcome / End Date[Outcome / End Date]

Recovered/Resolved, provide End Date

/ /

Recovering/Resolving

Not recovered/Not resolved

Recovered/Resolved with sequelae, provide End Date

/ /

Fatal, record Date of Death

/ /

2.5 Maximum Intensity

Record maximum intensity throughout duration of event

[Maximum Intensity]

Mild

Moderate

Severe

Not applicable

2.6 Most Clinically Significant Action Taken with Study Treatment(s) as a Result of the AE

[Action taken]

Study Treatment(s) withdrawn

Not applicable

2.7 Did the subject withdraw from study as a result of this AE?

[Subject withdrawn?]

Yes

No

2.8 Is there a reasonable possibility that the AE may have been caused by the Study Treatment?

Use best judgment at initial entry. May be amended when additional information becomes available.

[Relationship]

No

Yes

2.9 Was the SAE caused by activities related to study participation other than Study Treatment (e.g. procedures, blood draws, washout, etc)?

[Was the SAE caused by activities related to study participation other than Study Treatment (e.g. procedures, blood draws, washout, etc)?]

Yes

No

2.10 The ticking of this box confirms that the Relationship to Study Treatment and/or Study Participation entered for this SAE has been reviewed.

If any changes are made to the Relationship to Study Treatment or Study Participation after review, the investigator will be required to review and re-tick the box.

[Investigator reviewed]

Reviewed

3.

�

INTENSITY CHANGES Entry

For each change in intensity/grade, enter the corresponding SAE sequence number from the SAE SEQ number assigned to the event term

SERIOUSNESS

Specify the reason for considering this an SAE. Check all that apply.

4. Seriousness?

[Seriousness?]

Results in death

Is life-threatening

Requires hospitalisation or prolongation of existing hospitalisation

Results in disability/incapacity

Congenital anomaly/birth defect

Other, specify within general narrative comment

Possible drug-induced liver injury (see definition in SAE section of protocol).

Drug name Dose Unit Frequency Route Start Date Ongoing? Primary Indication Drug Type

5.

�

RELEVANT CONCOMITANT/TREATMENT MEDICATIONS Entry

Include details of any medication that may help explain the SAE, may have caused the SAE or was used to treat the SAE. Ensure each concomitant medication recorded in this section is also recorded in the appropriate concomitant medication form.

5.1 Drug name

[Drug name]

5.2 Dose

[Dose]

5.3 Unit[Unit]

5.4 Frequency

[Frequency]

5.5 Route[Route]

5.6 Start Date

[Start Date]

/ /

5.7 Ongoing?

[Ongoing?]

Yes

No, specify End Date

/ /

5.8 Primary Indication

Enter a medical diagnosis not description[Primary Indication]

5.9 Drug Type

[Drug Type]

Specific Condition Name Date of onset Continuing?

6.

�

RELEVANT MEDICAL CONDITIONS/RISK FACTORS Entry

Use the 'Add Entry' button for entry of each past or current medical disorder, allergy or surgery that may be RELEVANT to the SAE. Enter a diagnosis, not description. Relevant family or social history should be described in the 'General Narrative Comments' at the bottom of this form. Ensure each medical condition/risk factor recorded in this section is also recorded in the appropriate medical conditions form.

6.1 Specific Condition Name

Enter a medical diagnosis not description.

[Specific Condition Name]

6.2 Date of onset

[Date of onset]

/ /

6.3 Continuing?

[Continuing?]

Yes

No, specify date of last occurrence

/ /

Unknown

Test Name Test Date Test Result Test Units Normal Low Range Normal High Range

7.

�

RELEVANT DIAGNOSTIC RESULTS Entry

Provide details of any tests or procedures carried out to diagnose the SAE, if any of the diagnostic tests (e.g. a lab test) for this SAE has not been previously entered, and the eCRF includes an item for this test, ensure the data is also entered either into the associated visit or an unscheduled visit.

7.1 Test Name

[Test Name]

7.2 Test Date / /


49

CONFID

ENTIA

L2017N

324761_00206215

PPD

[Test Date]

7.3 Test Result

[Test Result]

7.4 Test Units

[Test Units]

7.5 Normal Low Range

[Normal Low Range]

7.6 Normal High Range

[Normal High Range]

8. Relevant diagnostic results not noted above

[Relevant diagnostic results not noted above]

RECHALLENGE

9. If Study Treatment(s) were stopped temporarily, did the reported event(s) recur after Study Treatments were restarted?

[If study treatment(s) were stopped temporarily, did the reported event(s) recur after study treatments products were restarted?]

No

Yes

Unknown at this time

Not applicable

GENERAL NARRATIVE COMMENTS

Provide a brief narrative description of SAE, possible other causes of the event (e.g. lack of efficacy, withdrawal of study treatment(s), the disease under study or other medical conditions) and details of the treatment.

10. General narrative comments

NON CLINICAL



50

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: SAE / CONMEDS STATUS (SAE/CONMEDS STATUS)

SAE / CONMEDS STATUS

1. Did the subject experience a serious adverse event during the study?

[Any SAEs?]

No

Yes

2. Were any concomitant medications taken by the subject prior to screening and/or during the study?

[Any Conmeds?]

No

Yes

3. Did the subject die during the study?

[Subject died?]

No

Yes



51

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: INFORM SCREENING (SCREEN)

INFORM SCREENING

1. Year of birth

[Year of birth]

2. Are you willing to tell us how you learned about this study?

[Learned about study?]

No

If yes, how did the subject initially hear about the study or how was the subject recruited into the study?

Prior Clinic Patient

Physician or Colleague Referral

Word of Mouth

Posters, Flyers or Brochures (RECRUITMENT KIT)

Community Outreach

Internet / Website / Email

Local Advertising (PLACED BY SITE)

Central / National Media Campaign (PLACED BY GSK)

Call Center (Organized by GSK)

Other, specify



52

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: SCREEN FAILURE (SCRFAIL)

SCREEN FAILURE

1. Was this subject a screen failure?

[Was this subject a screen failure?]

No

Yes

Screen failure date

/ /

Study closed/terminated (unspecified)

Did not meet inclusion/exclusion criteria

Adverse Event (unspecified)

Investigator discretion, specify

Withdrew consent, specify


53

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: INVESTIGATOR SIGNATURE (SIGN)

INVESTIGATOR SIGNATURE

1. Is this casebook ready to sign?

If not, click on the RETURN button below

[Casebook ready to sign?]

Yes



54

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: CV EVENT STATUS (STATUS)

CV EVENTS

Select Yes to display form that is required to be completed for a particular CV event.Select No if form was not required during the study.

1. Arrhythmias

[Arrhythmias]

Yes

No

2. Congestive heart failure

[Congestive heart failure]

Yes

No

3. Cerebrovascular events stroke (CVA) and Transient ischemic attack (TIA)[Cerebrovascular events stroke (CVA) and Transient ischemic attack (TIA)]

Yes

No

4. Deep vein thrombosis (DVT)/Pulmonary embolism (PE)

[Deep vein thrombosis (DVT)/Pulmonary embolism (PE)]

Yes

No

5. Myocardial Infarction/Unstable Angina

[Myocardial Infarction/Unstable Angina]

Yes

No

6. Peripheral arterial thromboembolism

[Peripheral arterial thromboembolism]

Yes

No

7. Pulmonary Hypertension

[Pulmonary Hypertension]

Yes

No

8. Revascularisation

[Revascularisation]

Yes

No

9. Valvulopathy

[Valvulopathy]

Yes

No


55

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: SUBJECT IDENTIFICATION (SUB ID)

SUBJECT IDENTIFICATION

If Subject number was entered incorrectly, please correct and submit; otherwise disregard

1. Subject number

[Subject number]


56

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: TIME PERFORMANCE MEASUREMENTS (TIME)

ELLIPTA INHALER

1. The amount of time taken to read the patient information leaflet and demonstrate correct inhaler use (T1).

[The amount of time taken to read the patient information leaflet and demonstrate correct inhaler use (T1).]

mins secs

2. The amount of time taken to be given instruction by the HCP on use of the inhaler and demonstrate correct inhaler use (T2).

[The amount of time taken to be given instruction by the HCP on use of the inhaler and demonstrate correct inhaler use (T2).]

mins secs

TURBUHALER



mins secs



mins secs

HANDIHALER

5. The amount of time taken to read the patient information leaflet and demonstrate correct inhaler use (T1).[The amount of time taken to read the patient information leaflet and demonstrate correct inhaler use (T1).]

mins secs



mins secs


57

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: TIME PERFORMANCE MEASUREMENTS (TIME)

ELLIPTA INHALER



mins secs



mins secs

DISKUS



mins secs



mins secs

HANDIHALER

5. The amount of time taken to read the patient information leaflet and demonstrate correct inhaler use (T1).[The amount of time taken to read the patient information leaflet and demonstrate correct inhaler use (T1).]

mins secs



mins secs


58

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: HISTORY OF TOBACCO USE (TOBACCO HISTORY)

HISTORY OF TOBACCO USE

1. What is the subject's history of smoking use?

[What is the subject's history of smoking use?]

Never smoked

Current smoker

Former smoker, when did the subject last smoke?

/ /


59

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: HISTORY OF TOBACCO USE (TOBACCO HISTORY DETAIL)

HISTORY OF TOBACCO USE

1. Average number of cigarettes smoked per day?

[Average number of cigarettes smoked per day?]

cigarettes per day

2. Number of years during which the subject has smoked/used tobacco?

[Number of years during which the subject has smoked/used tobacco?]

years

3. Number of pack years?

[Number of pack years?]


60

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: TURBUHALER ERRORS AFTER FIRST HCP INSTRUCTION (TURBUHALER HCP1)

TURBUHALER ERRORS



No

Yes

Failed to remove cap

Completed correctly


Did not hold device upright (+/- 45° OK) during dose preparation

Completed correctly Not completed correctly

Base not twisted fully backwards and forwards, no click heard

Completed correctly


Device tipped downwards after dose preparation

Completed correctly



Completed correctly



Completed correctly



Completed correctly



Completed correctly



- forceful

- deep

-(Note to HCP: it is important that the inhalation is forceful and deep from the start for this inhaler)

Completed correctly



Completed correctly


Did not hold breath

Completed correctly



Completed correctly


Any other comments:


61

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: TURBUHALER ERRORS AFTER SECOND HCP INSTRUCTION (TURBUHALER HCP2)

TURBUHALER ERRORS



No

Yes


Completed correctly



Completed correctly Not completed correctly


Completed correctly



Completed correctly



Completed correctly



Completed correctly



Completed correctly



Completed correctly



- forceful

- deep


Completed correctly



Completed correctly


Did not hold breath

Completed correctly



Completed correctly


Any other comments:


62

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: TURBUHALER ERRORS AFTER READING LEAFLET (TURBUHALER LEAFLET)

TURBUHALER ERRORS



No

Yes


Completed correctly



Completed correctly



Completed correctly



Completed correctly



Completed correctly



Completed correctly



Completed correctly



Completed correctly



- forceful - deep


Completed correctly



Completed correctly


Did not hold breath

Completed correctly



Completed correctly


Any other comments:


63

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: VALVULOPATHY (VALVULOPATHY) - Repeating Form

# Seq Did subject require hospitalization as a result of the

valvulopathy?

Did subject have any prior medical condition known to cause or contribute to

valvulopathy?

Fenfluramine Selective serotonin reuptake

inhibitor

Amphetamine Ergot

derivatives

Did subject have an abnormal physical

exam?

Was an Echocardiogram

performed?

Cardiac

catheterisation

Was medication required for treatment of

valvulopathy?

Was surgical therapy required for treatment of

valvulopathy?

1

Presentation

1. Sequence Number

[Seq]

2. Did subject require hospitalization as a result of the valvulopathy?

[Did subject require hospitalization as a result of the valvulopathy?]

No


/ /

: 24-hour clock

History

3. Did subject have any prior medical condition known to cause or contribute to valvulopathy?

[Did subject have any prior medical condition known to cause or contribute to valvulopathy?]

No


Yes


Medications used in the past known to cause/contribute to valvulopathy

4. Fenfluramine

[Fenfluramine]

Yes

No

Unknown

5. Selective serotonin reuptake inhibitor

[Selective serotonin reuptake inhibitor]

Yes

No

Unknown

6. Amphetamine

[Amphetamine]

Yes

No

Unknown

7. Ergot derivatives

[Ergot derivatives]

Yes

No

Unknown

Physical exam

8. Did subject have an abnormal physical exam?

[Did subject have an abnormal physical exam?]

No


Abnormal mitral valve exam

No

Yes, describe:

Abnormal tricuspid valve exam

No

Yes, describe:

Abnormal aortic valve exam

No

Yes, describe:

Abnormal pulmonic valve exam

No

Yes, describe:

Diagnostic tests

9. Was an Echocardiogram performed?

[Was an Echocardiogram performed?]

No


/ /

: 24-hour clock

Mitral

Appearance of valves

Normal

Abnormal

Technically inadequate study

Stenosis

No

Yes, degree of stenosis

Mild

Moderate

Severe

Regurgitation

No

Yes, degree of regurgitation

Mild

Moderate

Severe

Tricuspid


Normal

Abnormal


Stenosis

No


Mild

Moderate

Severe

Regurgitation

No


Mild

Moderate

Severe

Aortic


Normal

Abnormal


Stenosis


64

CONFID

ENTIA

L2017N

324761_00206215

PPD

No


Mild

Moderate

Severe

Regurgitation

No


Mild

Moderate

Severe

Pulmonic


Normal

Abnormal


Stenosis

No


Mild

Moderate

Severe

Regurgitation

No


Mild

Moderate

Severe

10. Cardiac catheterisation

[Cardiac catheterisation]

No

Yes, date and time of cardiac catheterisation

/ /

: 24-hour clock

Therapy required

11. Was medication required for treatment of valvulopathy?

[Was medication required for treatment of valvulopathy?]

Yes

No

Unknown

12. Was surgical therapy required for treatment of valvulopathy?[Was surgical therapy required for treatment of valvulopathy?]

No

Unknown

Yes, note type of surgical therapy below:

Mitral valve repair

Yes

No

Tricuspid valve repair

Yes

No

Aortic valve repair

Yes

No

Pulmonary valve repair

Yes

No

Mitral valve replacement

Yes

No

Tricuspid valve replacement

Yes

No

Aortic valve replacement

Yes

No

Pulmonary valve replacement

Yes

No

Other surgical therapy

No

If Yes, specify


65

CONFID

ENTIA

L2017N

324761_00206215

PPD

206215: VITAL SIGNS (VS)

VITAL SIGNS

1. Height

[Height]

cm

2. Weight

[Weight]

kg


[b]

[b]


66

CONFID

ENTIA

L2017N

324761_00206215

PPD

CONFIDENTIAL

- 1 -

LIST OF INVESTIGATORS AND IECS/IRBS FOR 206215

Investigator Sub-Investigator Investigator no./Centre no. Description of Research

Facility, Hospital/ Institution,

and Address

Name of IEC/IRB Committee,

Address, Committee Chair

Netherlands

Hoek, Boudewijn A. MD Mangal, Radjesh V.

MD/PhD

Q Clinical, Kleiweg 78,

ROTTERDAM, 3051 GV,

Netherlands

St. Antonius Ziekenhuis,

Koekoekslaan 1, NIEUWEGEIN,

3435 CM, Netherlands

Chairperson:

Moeskops-van

Beurden, Wendy J.C.

MD/PhD

Veen, Helena P.A.A.

van (Ilonka). MD/PhD

Medisch Spectrum Twente,

Haaksbergerstraat 55,

ENSCHEDE, 7513 ER,

Netherlands




Chairperson:

Soest, Jacob A. (Jaap)

van. MD

None Huisartsenmaatschap MCN

Nijkerkendijk 38-01

NIJVERDAL




2017N324761_00206215

PPD

PPD

PPD

PPD

PPD

CONFIDENTIAL

- 2 -



and Address



7442 LS

Netherlands Chairperson:

United Kingdom

Collier, David. MBBS/

PhD

Pfeffer, Paul. PhD

Shiel, Julian. MBChB

Barts and the London school of

Dentistry and Medicine, Clinical

Research Centre William

Harvey Heart Cen., Charter

House Square, London, EC1M

6BQ, United Kingdom

NRES Committee East

Midlands - Leicester The Old

Chapel, Royal Standard Place,

Nottingham, NG1 6FS, United

Kingdom

Chairperson:

Van Den Berg, Frans.

MBChB

Adeloye, Temitope.

MBChB

Dennison, Jeremy.

MBChB

Loewenstein, Inge.

Hammersmith Medicines

Research, Cumberland Avenue,

London, North west 10 7East

West, United Kingdom

NRES Committee East

Midlands - Leicester The Old

Chapel, Royal Standard Place,

Nottingham, NG1 6FS, United

Kingdom

2017N324761_00206215

PPD

PPD

PPD

PPD

CONFIDENTIAL

- 3 -



and Address



MD

Mbow, Fatou. DFFP

Puri, Adeep. MPH

Wilkes, Denisa. MD

Chairperson:

2017N324761_00206215

PPD

Page(s) removed- Out of Scope of phase 1 of Policy 0070 – Investigator CVs

206215 CONFIDENTIAL Model ICF Version No: 1.0Date: 28-09-16

- 1 -

Model ICF Review and Approval Documentation: (Delete from final Model ICF template)Clinical Study Identifier: 206215Version and Date of Model ICF: final draft dated 28SEP16Model ICF Author (Name/Role): (CIL)Aligned with Protocol Version/Date: Version 01 dated 15SEP16Reviewed/Approved By (Name/Role/Date):Statement of Review/Approval: Final approval

A placebo Inhaler study to investigate critical and overall errors, preference and teaching/training time, in adult subjects with COPD

Study doctor: <insert>Institution name: <insert>Site address: <insert>Phone number: <insert>Subject identification/number: <insert>

Introduction: You have been provided the information booklet explaining clinical trials.This document is the Informed Consent Form. It contains specific information about this clinical trial. To keep the information in this form simple we shall refer to a Clinical Trial as a “study”.

This consent form has been reviewed and approved by an Ethics Committee (EC). This committee reviews research studies to protect the rights and well-being of the people taking part. Some of the information in this consent form is required by law.

Why is this research study being done?We are asking if you want to take part in this study because you have Chronic Obstructive Pulmonary Disease (COPD) and because you have used an inhaler to treat your COPD.

There are many types of medicines, delivered by inhaler, for COPD that doctors can prescribe. People may use one, two or sometimes threedifferent types of inhaler to take all their different COPD medicines. Onecommonly used type of inhaler is the metered dose inhaler (MDI) and you may have one for your rescue medication such as Ventolin (salbutamol)or Atrovent (ipratropium). Another type is a dry powder inhaler (DPI), usually used to deliver a maintenance medicine, such as Seretide DISKUS, Symbicort Turbuhaler or Spiriva HandiHaler. In this study, we will compare the ELLIPTA inhaler with combinations of other inhalers which are commonly used to take COPD medicines.

The purpose of this study is to see how many errors are made by a COPD patient when using inhalers and to see how COPD patients would prefer to take their medicine.

The ELLIPTA inhaler can be used to deliver 3 different medicines, all from just one inhaler. Currently patients would need to be prescribed 2

2017N324761_00

PPD

PPD


- 2 -

different inhalers to take all 3 different medicines.

In this study and associated documents we will use the term ‘Treatment’. However, all the inhalers in this study will contain placebo (dummy) powder, or use placebo capsules; that means they do not contain any active medicine. The placebo ELLIPTA inhaler, or combination of other placebo inhalers you test would, match the inhalers used by COPD patients’ to deliver the 3 active medicines but with no active medicine in them.

About 144 people from 2 countries will take part in this study.

How does the study work?To enter in this study, you must have COPD and you must be on treatment for your COPD. You should not have used the ELLIPTA inhalerand HandiHaler inhaler in the last 2 years. Also, you must not have used one of either DISKUS/ACCUHALER inhaler or Turbuhaler inhaler in the past 2 years. It is also okay if you have used neither of them.

This study will be divided into 2 parts, called sub-study 1 and sub-study 2. Each part will compare the number of errors that are made when using the ELLIPTA inhaler to one of the following 2 combination treatments; DISKUS/ACCUHALER with HandiHaler (Sub-Study 1), or Turbuhaler with HandiHaler (Sub-Study 2).

You will be placed in either Sub-Study 1 or 2 dependent on which inhaler, DISKUS/ACCUHALER or Turbuhaler, you have not usedpreviously. If you have used neither of these 2 inhalers and if both sub-studies are recruiting, then you will be placed in the sub-study that hasstarted first.

A computer will put people into groups, deciding by chance on the order of the inhalers you receive. Neither you nor the study doctor can choose a group, but everybody can see the order of the inhalers they then receive.

In the study, you will use either the ELLIPTA inhaler first followed by one of the inhaler combinations, or you will receive the inhaler combination first followed by the ELLIPTA inhaler.

You will demonstrate the use of the inhalers one by one, and any errors made will be recorded by the study doctor or nurse, who will then help instruct you to use the inhalers correctly, if needed. You will also be asked some questions, by the study doctor or nurse, about which of the treatments (single inhaler or combination of inhalers) you preferred. The study doctor or nurse will also time how long it takes for you to use each inhaler correctly.

2017N324761_00


- 3 -

What am I expected to do in this study? How will being part of this study affect my lifestyle? How long will I be in the study?This study requires only one clinic visit (one day), which is expected to last approximately 2 hours. To help you with scheduling, this single visit may be divided over 2 separate visits (V0 and V1). These 2 visits should take place within 30 days.

The screening and study assessments, as well as when they can be scheduled are shown in the table at the end of this section.

Please keep in mind how the study tests and visits described here will affect your work and family schedules. Consider if you need transportation to and from the clinic. You may find that these tests and visits need some planning. Ask the study doctor or nurse if you have any questions about the tests and procedures for the study.

You will continue to take all your usual medicines during the study.

If you need to have two separate visits (V0 and V1), and you start any new medications between V0 and V1, you will need to tell the study doctor.

If, based on the assessments performed at screening you qualify to take part in the study, then a computer will “put” you into a treatment groupin the sub-study you are included in. This decides which order you will receive the inhalers in your sub-study and also which preference questionnaire you will answer.

Once you have been allocated, by computer, to the appropriate group,you will have up to 3 attempts to demonstrate the use of each inhaler.

Attempt 1

You will be asked to read the Patient Information Leaflet (PIL) (the “how to use” section) of the first inhaler. You may take as long as needed to read this and to then feel comfortable to demonstrate how to use thatinhaler.

Any errors made will be recorded by the study doctor or nurse. If you do not make any errors, this will also be recorded by the study doctor or nurse and you will not need to demonstrate use again with this inhaler.

Attempt 2

If you made any errors, the study doctor or nurse will instruct you incorrect use of the first inhaler. Then you will be asked to demonstrate the use of the inhaler again. Any errors will be recorded by the study doctor or nurse. If you do not make any errors, this will also be recorded by the

2017N324761_00


- 4 -

study doctor or nurse and you will not need to demonstrate any more with this inhaler.

Attempt 3

If you made any errors during the second attempt, then the study doctor or nurse will instruct you once more in correct use of the first inhaler.Then you will be asked to demonstrate correct use one final time again.Any errors will be recorded by the study doctor or nurse. If you do notmake any errors, this will also be recorded by the study doctor or nurse.

After completing the procedure with the first inhaler, you will be asked to do the same with the second inhaler and third inhaler.

The study doctor or nurse will also record the time it takes for you to read the PIL and demonstrate the use of each inhaler for the first attempt; If you make errors in the use of the inhaler at the first attempt, then the time taken for study doctor or nurse to instruct you on the use of the inhaler up to 2 more times, if needed, will also be recorded.

After completing all the inhaler demonstration assessments for all inhalers, the study doctor or nurse will ask you two questions from a questionnaire on which of the treatments you preferred. Once you have answered these questions you will have completed the study.

The table here shows all scheduled tests you would have to do to be included on the study and would have to complete if you are then suitable to take part.

2017N324761_00


- 5 -

Visit V0 V1 Notes

Study Day 1 1 or 2 V0 can take place on the same day as V1. Or V1 can be completed within 30 days of consent and V0.

Procedure:

Screening Assessments All completed prior to randomisation

Written informed consent X Study doctor/nurse will talk to you about the study. You will need to read, ask any questions and sign this consent form.

Subject demography X The study doctor/nurse will ask about your health and review your medical and COPD history. Information about you such as age, ethnicity, and gender will be collected.


X

Medication history including COPD Therapy History

X The study doctor/nurse will ask about your health or any illnesses including history of COPD and current treatment, and what type of inhalers you used previously.

Inclusion/exclusion criteria X The study doctor/nurse checks you are suitable to be included on the study.

Study Assessments These tests are completed if you are included on the study

Randomisation X A computer will decide the order of your inhalers

Assess the number of inhaler errors on each treatment after reading the PIL for Inhaler tested

X You will read the instruction on Inhaler use in the PIL and then show the study doctor or nurse how to use the inhaler. No instruction is provided by the study doctor/nurse for this assessment.

Assess the number of inhaler errors on each treatment after each of up to 2 attempts following instruction by Site Staff

X If the study doctor/nurse observe an incorrect use of the inhaler they will be able to give you instructions twice on the correct use of the inhaler

Teaching-Training Time for each inhaler includes thefollowing:

X Study doctor/nurse will measure: The amount of time taken for you to read

the patient information leaflet and demonstrate inhaler use

The amount of time you need to be given instruction by the site staff on use of the inhaler and then demonstrate inhaler use

Preference Questionnaire X Study doctor/nurse will ask you 2 questions from a Preference Questionnaire on which of the treatments you preferred

2017N324761_00


- 6 -

Insurance Cover

Before participating you should consider if this will affect any insurance you currently have, or may purchase in the future, and seek advice if necessary from your insurance company.

What side effects can I expect from this study?Since the placebo inhalers provided do not contain any active medication, and you will not receive any other medication as a result of participation in this study, it is very unlikely that you would have side effects due to study participation.

The inhaler or capsules for inhalation in this study contain placebo, a powder made with lactose (which contains milk protein) and magnesium stearate. The milk protein in lactose may cause side effects (for example an allergic reaction) in some patients, which could include rash, wheezing, difficulty in breathing, tissue swelling, feeling faint or changes in your heart rhythm. Tell your study doctor before the study (at screening) if you are allergic to lactose, milk protein or magnesium stearate.

If you experience any unusual symptoms in the clinic inform the study doctor/research nurse immediately. If any symptoms begin after you have left the clinic you should seek medical advice straight away and let the study doctor know what has happened.

Most side effects may go away as soon as you stop using the inhalers. In unusual cases, the side effects can be serious, lasting or may never go away. Ask your study doctor or nurse if you have any questions about the side effects.

In the unlikely event that during the demonstration of inhalation technique you may have shortness of breath, coughing, light-headedness or fainting, and/or chest tightness you will receive appropriate medical care.

What benefits can I expect from this study?Since you will not receive any medication as a result of participation in this study, taking part in this study will not make your health / condition better, and will not have any direct benefit to you.

Knowledge from this study may help doctors to better understand which inhaler or inhaler combinations are easier to use without making any mistakes, for patients like you. Eventually, this could lead to creating more clear instructions for inhaler use for patients and doctors.

When the study is completed at all the study sites, the data will be

2017N324761_00


- 7 -

analyzed. You will have an opportunity to learn of the results. You may ask your study doctor for the results and to have them explained to you.

Are there alternatives to taking part in this study?Taking part in this study will not make your health better. You may choose to:

Continue to get normal care from your own doctor

Take part in another study

Not take part in this study

Talk with the study doctor, or your family doctor, about your choices.

Will I receive payment to be part of this study?We will reimburse you for the cost of travelling to your study visit(s). You may receive up to <amount> per visit for travel.

You will receive <amount> for your time and effort in taking part in the study.

Will I have to pay anything to be part of this study?There are no costs for you if you participate in the study.

You and your insurance company/the National Health Service will continue to pay for your regular health care.

What happens to my personal and medical information?It is very important that your personal and medical information stay confidential and secure. GSK will protect your information in accordance with current law.

When you sign this consent form you agree that GSK can use your personal and medical information as described here:

Your personal and medical information may be checked by GSK and others (like agencies that approve and monitor studies). This is to make sure that the study is being run properly.

Besides that, only the researchers at this study site can use information that identifies you (such as name and address) and only for the purpose of the study.

Your study information will be labelled with a code number (for example, It will not include your name or address. The study doctor will have the link between your name and the code number.

The link between your name and the code number will not be shared. Only the code number and coded information will be sent to GSK.

GSK will use your coded information for research only. This may

2017N324761_00

PPD


- 8 -

include research looking at improving the quality and efficiency in conducting clinical research trials in general.

GSK may:

Keep your coded information electronically, and analyse it by computer to find out what the study is telling us. This may be done by GSK or a third party, in which case GSK will ensure that the third party is required to keep your data secure,

share the information with regulatory agencies that approve new medicines,

share the information with people who check that the study is done properly (like the ethics committee or review boards),

combine the information with results from other studies to learn more about the medicine and other medicines, and this diseaseand other diseases and conditions. This may help us to assess the risks and benefits of GSK (or other) medicines, or to improve disease understanding,

publish study results for medical journals, meetings and on the internet for other researchers to use; your name will not appear in any publication,

share coded information with other companies, organisations or universities to carry out research. This may include research looking at improving the quality and efficiency in conducting clinical research trials in general.

Personal and medical data collected during the study may be moved, stored and used in the country where you live or another country where GSK or those working with GSK work.

Use of this information may take place in countries with lower data protection rules than the country where you live. GSK will make sure that if your data are moved to another country, it will still be treated as stated in this Informed Consent Form.

A description of this clinical study will be available on the GSK Clinical Study Register (http://www.gsk-clinicalstudyregister.com/) and may also appear in clinical trial/study registries in countries in which the clinical study is conducted.

GSK will be the owner of the study results. GSK plans to use the results, and may get patents, or sell the drug in the future, or make profits other ways. You will not be paid any part of this.

At any time, you may ask the study doctor to see your personal information and correct it, if necessary. In some circumstances, you may not be able to access your study information while the study is ongoing.However, the study doctor will share any important medical information

2017N324761_00


- 9 -

if it is relevant to your health during the course of the study.

What happens if I withdraw my consent?If you decide to withdraw consent, this means you decided that no more information about your health could be collected for this study. You and the study doctor will need to discuss the best way to do this. All the information you gave us before you left the study will still be used for the study.

You may need to leave the study if: The results of certain tests show that you are not right for this

study.

You do not follow study instructions for treatment.

You get new health problems during the study that might not work well with the study set-up.

The study doctor thinks it is best for you to stop.

GSK (the study sponsor), the regulatory authority, or the study doctor may choose to stop the study at any time. We will give you the reason at that time.

What happens if I decide to leave the study? If you decide to leave the study for any other reason before the end of the study, you and the study doctor will discuss the best way to do this.

All the data collected before you left the study will still be used for the study.

What happens if I get hurt while taking part in this study? If you become ill or are hurt while you are in the study, you will get the medical care that you need right away.

<Insert legally approved country – specific (compensation) clause.>

GSK will help pay for your care if you are hurt by a procedure that is done to you only because you are part of this study. Your study doctor can give you a copy of the guidelines for this kind of injury.

Signing this consent form does not change any legal rights you may have.

Whom should I call if I have questions?You can talk with the study doctor, <name> about any questions or concerns you have about this study. Call him/her at <number>.

If you have any questions about the rights you have while taking part in the study, call the <ethics board > at <number>.

If you think you have been hurt from taking part in this study, or have

2017N324761_00


- 10 -

any questions about side effects, call <study staff> at <number>.

2017N324761_00


- 11 -

Study Number:Centre Name:Study Subject

A placebo Inhaler study to investigate critical and overall errors, preference and teaching/training time, in adult subjects with COPD

Patient Consent Form

Study 206215: An open-label study to evaluate the preference attributes of the ELLIPTA™ dry powder inhaler (DPI) compared to the HandiHaler™ DPI in subjects with Chronic Obstructive Pulmonary Disease (COPD)

Name of Principal Researcher:……………………………………

If you agree with each section below, please INITIAL the box:

NB. All boxes below need to be initialled in order for the patient to be enrolled in to the study.

INITIALS

1) I have received and reviewed the general information booklet.

2) I have read and understood the information/consentsheet (<insert version & date>) for the above study and I confirm that the study procedures and information have been explained to me. I have had the opportunity to ask questions and I am satisfied with the answers and explanations provided.

3) I have been given time and opportunity to read the information carefully, to discuss it with others and to decide whether or not to take part in the study. I understand that my participation in this study is voluntary and that I am free to withdraw from the study at any time, without giving any reason, without my medical care or legal rights being affected, the study doctor may ask me to leave the study at any time.

2017N324761_00


- 12 -

4) I understand that sections of my medical notes may be looked at by responsible individuals from GlaxoSmithKline, companies acting on their behalf, NHS Trust/Health Board or regulatory authorities where it is relevant to my taking part in research. I give permission for these individuals to have access to my records.

5) I have been given the names of study staff who I can call.

6) I agree to my GP being informed about my participation in this study.

7) I agree that my data can be transferred outside the European Union.

8) I freely agree to take part in this study.

Please print and sign your name below and add today’s date:

__________________ ___________________ _________________

Name of patient Signature Date

__________________

Name of person taking consent

__________________

Signature

__________________

Date

N.B. The patient must date his/her own signature

1 copy for patient; 1 copy for study file; 1 copy to be kept with hospital/clinic notes.

2017N324761_00

1

CONFIDENTIAL 2017N324761_00206215

PPD

PPDPPD PPD

GSK CONFIDENTIAL GlaxoSmithKline Research& Development LimitedPark RoadWareHertfordshireSG12 0DP

Tel. Fax.www.gsk.com

COA No.: 162397455-01 (RoW)

TMP2011N116425v5 Page 1 of 2 Registered in England & WalesNo. 835139

Registered office980 Great West RoadBrentford, Middlesex. TW8 9GS

Product DevelopmentDrug Product Certificate of Analysis

Product: Placebo

Strength / Dose Form: Placebo to match Umeclidinium Inhalation Powder or Umeclidinium/Vilanterol Inhalation Powder

Batch Number: 162397455

Date of Manufacture: 05-November-2015

Site of Manufacture: GSK GMS WARE, UNITED KINGDOM

Test Acceptance Criteria Results

Description A two tone grey inhaler in moulded plastic with a dose counter. The inhaler contains two foil blister strips. Each strip contains 30

regularly distributed blisters each containing a white free flowing

powder. The foil strips consist of a formed silver coloured base foil with

a peelable lid.

Conforms

Absence of Umeclidinium by HPLC (for the 0.6% Magnesium Stearate in lactose strip)

Not detected (less than 0.06 mcg/blister)

Conforms

Absence of Umeclidinium by HPLC (for the 1.0% Magnesium Stearate in lactose strip)

Not detected (less than 0.06 mcg/blister)

Conforms

Absence of Vilanterol by HPLC (for the 0.6% Magnesium Stearate in lactose strip)

Not detected (less than 0.002mcg/blister)

Conforms

Absence of Vilanterol by HPLC (for the 1.0% Magnesium Stearate in lactose strip)

Not detected (less than 0.002mcg/blister)

Conforms

2

CONFIDENTIAL 2017N324761_00206215

PPD

GSK CONFIDENTIAL GlaxoSmithKline Research& Development LimitedPark RoadWareHertfordshireSG12 0DP

Tel. Fax.www.gsk.com

COA No.: 162397455-01 (RoW)

TMP2011N116425v5 Page 2 of 2 Registered in England & WalesNo. 835139

Registered office980 Great West RoadBrentford, Middlesex. TW8 9GS

Test Acceptance Criteria Results

Microbial limits test (for the 0.6% Magnesium Stearate in lactose strip)

Total Aerobic Microbial Count (TAMC) (CFU/g)Total Yeast and Mould Count (TYMC) (CFU/g)

Absence of Specified microorganisms:Bile-tolerant Gram negative bacteriaPseudomonas aeruginosaStaphylococcus aureus

Not greater than 102


Absent in 1gAbsent in 1gAbsent in 1g

<2<2

ConformsConformsConforms

Microbial limits test (for the 1.0% Magnesium Stearate in lactose strip)

Total Aerobic Microbial Count (TAMC) (CFU/g)Total Yeast and Mould Count (TYMC) (CFU/g)

Absence of Specified microorganisms:Bile-tolerant Gram negative bacteriaPseudomonas aeruginosaStaphylococcus aureus



Absent in 1gAbsent in 1gAbsent in 1g

<2<2

ConformsConformsConforms

This material conforms to Specification S0017136v05

This material was manufactured and tested in accordance with current Good Manufacturing Practices.

Name: Role: Signature and Date:

Quality Control Approval

Qualified Person Approval

This document has been digitally signed using the SAFE-BioPharma Digital Credential system.

3

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

PPD

4

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

PPD

5

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

PPD

6

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

PPD

Electronic signatures are located on the last page of the PDF.The EU Qualified Person (QP) or QP Delegate signature appears as ‘Quality Assurance Approval’.

1(1)

CERTIFICATE OF ANALYSIS FOR PLACEBO TURBUHALER, WHITE, 60 DOSES, BATCH NUMBER HAAM

Article/Material ID: 215409900/190005954

Batch/Lot ID: HAAM/L005937

Specification: Specification for Placebo for Turbuhaler, 60 dosesFNC.000-209-492.2.0

Analysis approved: October 2016

Reference: 467157 HAAM

Test procedure Acceptance criteria Results

Description The colour of the turning grip is white.There is no embossment on the turning grip.The colour of the cover is white. Inside the cover 5 fins are present.The figure 60 is visible in the dose indicator window.The mouthpiece has 4 bars, and can be rotated.

Contents:The contents are white to off-white, predominantly in the form of rounded granules.

Complies

Microbiological Quality Meets the requirements in Ph Eur/USP/JP Complies

Document ID: Status: Version Number: Document Status Date:

Doc ID-003375926 Approved1.0 20/Oct/2016

7

CONFIDENTIAL 2017N324761_00206215

Document Name:

Document Title:

Document ID:

Version Label:

Server Date (dd-MMM-yyyy HH:mm ‘GMT’Z) Signed By Meaning of Signature

Notes: (1) Document details as stored in ANGEL, an AstraZeneca document management system.

batch-analyses-haam

Certificate of Analysis, batch HAAM, placebo for Turbuhaler, 60 doses, white,GSK

Doc ID-003375926

1.0ApprovedCURRENTLATEST

20-Oct-2016 07:17 GMT+0100Quality AssuranceApproval

8

CONFIDENTIAL 2017N324761_00206215

PPD

ICH Data Listings Page

Listing 1 Listing of Screen Failures (All Subjects Enrolled Population) ................... Listing 2 Listing of Reasons for Study Withdrawal (All Subjects Enrolled

Population) ................................................................................................... Listing 3 Listing of Important Protocol Deviations (Intent-to-Treat Population) ....... Listing 4 Listing of Subjects with Inclusion and Exclusion Criteria Deviations

(Intent-to-Treat Population) .......................................................................... Listing 5 Listing of Randomised and Actual Treatment Sequence (Intent-to-Treat

Population) ................................................................................................... Listing 6 Listing of Demographic Characteristics (Intent-to-Treat Population) ........ Listing 7 Listing of Race (Intent-to-Treat Population) .............................................. Listing 8 Listing of Subject Numbers for Individual Adverse Events (All Subjects

Enrolled Population) ..................................................................................... Listing 9 Listing of All Adverse Events (All Subjects Enrolled Population) .............. Listing 10 Listing of Subject Errors by Assessment (Intent-to-Treat Population) ....

2

35

8

103244

616365

1

CONFIDENTIAL 2017N324761_00206215

Page(s) removed- Out of Scope of phase 1 of Policy 0070 - ICH Listings

Other Data Listings Page

Listing 11 Listing of Subjects by Country and Centres (Intent-to-Treat Population) Listing 12 Listing of Treatment Sequence Misallocations (Intent-to-Treat

Population) ................................................................................................... Listing 13 Listing of Medical Conditions (Intent-to-Treat Population) ...................... Listing 14 Listing of Family History of Cardiovascular Risk Factors (Intent-to-Treat

Population) ................................................................................................... Listing 15 Listing of COPD History (Intent-to-Treat Population) .............................. Listing 16 Listing of Smoking History and Smoking Status (Intent-to-Treat

Population) ................................................................................................... Listing 17 Listing of COPD Concomitant Medications (Intent-to-Treat Population) . Listing 18 Listing of Non-COPD Concomitant Medications (Intent-to-Treat

Population) ................................................................................................... Listing 19 Relationship between ATC Level 1, Ingredient and Verbatim Text for

Non-COPD Medications (Intent-to-Treat Population) ................................... Listing 20 Listing of Naivety to Inhaler Devices (Intent-to-Treat Population) ........... Listing 21 Listing of Time Taken to Demonstrate Correct Inhaler Use (Intent-to-

Treat Population) ......................................................................................... Listing 22 Listing of Preference Questionnaire (Intent-to-Treat Population) ........... Listing 23 Listing of Myocardial Infarction/Unstable Angina (Intent-to-Treat

Population) ................................................................................................... Listing 24 Listing of Congestive Heart Failure (Intent-to-Treat Population) ............. Listing 25 Listing of Arrhythmias (Intent-to-Treat Population) ................................. Listing 26 Listing of Valvulopathy (Intent-to-Treat Population) ................................ Listing 27 Listing of Pulmonary Hypertension (Intent-to-Treat Population) ............. Listing 28 Listing of Cerebrovascular Events/Stroke and Transient Ischemic

Attack (Intent-to-Treat Population) ............................................................... Listing 29 Listing of Peripheral Arterial Thromboembolism (Intent-to-Treat

Population) ................................................................................................... Listing 30 Listing of Deep Venous Thrombosis/Pulmonary Embolism (Intent-to-

Treat Population) ......................................................................................... Listing 31 Listing of Revascularisation (Intent-to-Treat Population) ........................ Listing 32 Listing of All Cause Deaths (Intent-to-Treat Population) ........................ Listing 33 Listing of All Subjects who Became Pregnant During the Study (Intent-

to-Treat Population) .....................................................................................

2

810

3143

5672

194

364395

407487

505506507508509

510

511

512513514

515

1

CONFIDENTIAL 2017N324761_00206215

Page(s) removed- Out of Scope of phase 1 of Policy 0070 - Other Listings

CONFIDENTIAL

1

Table 1: Protocol Level Administration

Study Number: 206215 Study Title: A randomized, open-label, cross-over, placebo-device study investigating critical and over all errors,

training/teaching time, and preference attributes of the ELLIPTA dry powder Inhaler (DPI) as compared to HandiHaler DPI used

in combination with either DISKUS DPI or Turbuhaler DPI, in adult patients with Chronic Obstructive Pulmonary Disease (COPD)

Study Sponsor Medical Writing:

Protocol

Authorship

Statistics

Group

Data

Management

Group

Randomisation Site of

Manufacture &

Assembly

Medical Writing:

Clinical Report

Authorship

Site of EU

release

Location of

Study Master

File

GCP

Regulatory

inspection

Address Names of authors

and their Address

Address Address Company and

Address

Address Names of authors

and their Address

Address Address Name of

Agency

GlaxoSmithKline

Research &

Development

Limited, 980

Great West

Road, Brentford,

Middlesex, TW8

9GS, United

Kingdom

GlaxoSmithKline,

Research &

Development

Projects Clinical

Platforms and

Sciences,

Research &

Development

Projects Clinical

Platforms and

Sciences Therapy

Area Delivery

Respiratory 3,

GlaxoSmithKline,

Research &

Development

Projects Clinical

Platforms and

Sciences,

Research &

Development

Projects Clinical

Platforms and

Sciences Qsci

Clinical Statistics

10G, Building 11,

DOCS, Suite

300, 79 TW

Alexander Drive,

4401 Research

Commons,

Durham, North

Carolina, 27709,

United States

GlaxoSmithKline,

7333

Mississauga

Road North,

GlaxoSmithKline

Registration and

Medication

Ordering

System

(RAMOS), New

Frontiers

Science Park,

Third Avenue,

Harlow, Essex,

CM19 5AW,

United Kingdom

AstraZeneca AB,

Forskargatan 18,

Södertälje,

Sweden

GlaxoSmithKline

Research &

Development

Limited, New

Frontiers

Science Park

(South), Third

Avenue, Harlow,

Essex, CM19

GlaxoSmithKline,

Global Medical,

Chief Medical

Office Global

Clinical Safety &

Pharmacovigilance,

Safety Evaluation

and Risk

Management

United Kingdom 3,

Stockley Park,

United Kingdom

GlaxoSmithKline

Research &

Development

Limited, New

Frontiers

Science Park

(South), Third

Avenue, Harlow,

Essex, CM19

5AW, United

Kingdom

GlaxoSmithKline

Research &

Development,

Medicines

Research

Centre, Gunnels

Wood Road,

Stevenage,

Hertfordshire,

SG1 2NY,

United Kingdom

None

2017N324761_00206215

PPD PPD PPD

PPD

CONFIDENTIAL

2





Protocol

Authorship

Statistics

Group

Data

Management

Group


Manufacture &

Assembly

Medical Writing:

Clinical Report

Authorship

Site of EU

release

Location of

Study Master

File

GCP

Regulatory

inspection

Stockley Park,

United Kingdom

GlaxoSmithKline,

Global Medical,

Chief Medical

Office Global

Clinical Safety &

Pharmacovigilance,

Safety Evaluation

and Risk

Management

United Kingdom 3,

Stockley Park,

United Kingdom

Stockley Park,

United Kingdom

GlaxoSmithKline,

Research &

Development

Projects Clinical

Platforms and

Sciences,

Research &

Development

Projects Clinical

Platforms and

Sciences Qsci

Clinical Statistics

10G, Building 10.

Mississauga,

Ontario, L5N

6L4, Canada

5AW, United

Kingdom

GlaxoSmithKline

Research &

Development

Limited, Park

Road, Ware,

Hertfordshire,

SG12 0DP,

United Kingdom

GlaxoSmithKline,

1011 North

Arendell Avenue,

PO Box 1217,

Zebulon, North

Carolina, 27597,

GlaxoSmithKline,

Respiratory TAU &

Flexible Discovery

Unit, Research &

Development

Respiratory HUP

Clin Dev 2,

Stockley Park,

United Kingdom

GlaxoSmithKline,

Respiratory TAU &

Flexible Discovery

Unit, Research &

Development

Respiratory HUP

2017N324761_00206215

PPD

PPD PPD

PPD

PPD

PPD

CONFIDENTIAL

3





Protocol

Authorship

Statistics

Group

Data

Management

Group


Manufacture &

Assembly

Medical Writing:

Clinical Report

Authorship

Site of EU

release

Location of

Study Master

File

GCP

Regulatory

inspection

GlaxoSmithKline,

Respiratory TAU &

Flexible Discovery

Unit, Research &

Development

Respiratory HUP

Clin Dev 2,

Stockley Park,

United Kingdom

;

GlaxoSmithKline,

Research &

Development

Projects Clinical

Platforms and

Sciences,

Research &

Ground floor,

Stockley Park,

United Kingdom

;

GlaxoSmithKline,

Research &

Development

Projects Clinical

Platforms and

Sciences,

Research &

Development

Projects Clinical

Platforms and

Sciences,

Stockley Park,

United Kingdom

United States

GlaxoSmithKline,

1061 Mountain

Highway,

Boronia, Victoria,

3155, Australia

Clin Dev Medical

Writing 1, 5.3317,

Research Triangle

Park, United States

GlaxoSmithKline,

Research &

Development

Projects Clinical

Platforms and

Sciences,

Research &

Development

Projects Clinical

Platforms and

Sciences Qsci

Clinical Statistics

2017N324761_00206215

PPD

PPD

PPD

CONFIDENTIAL

4





Protocol

Authorship

Statistics

Group

Data

Management

Group


Manufacture &

Assembly

Medical Writing:

Clinical Report

Authorship

Site of EU

release

Location of

Study Master

File

GCP

Regulatory

inspection

Development

Projects Clinical

Platforms and

Sciences Qsci

Clinical Statistics

10G, Building 10

Ground floor,

Stockley Park,

United Kingdom

;

GlaxoSmithKline,

Respiratory TAU &

Flexible Discovery

Unit, Research &

Development

Respiratory HUP

Clin Dev 2, Building

;

GlaxoSmithKline,

Research &

Development

Projects Clinical

Platforms and

Sciences,

Research &

Development

Projects Clinical

Platforms and

Sciences Qsci

Clinical Statistics

10G, Stockley

Park, United

Kingdom

10G, Building 10.

Ground floor,

Stockley Park,

United Kingdom

;

GlaxoSmithKline,

Research &

Development

Projects Clinical

Platforms and

Sciences,

Research &

Development

Projects Clinical

Platforms and

Sciences, Stockley

Park, United

2017N324761_00206215

PPD

PPD

PPD

CONFIDENTIAL

5





Protocol

Authorship

Statistics

Group

Data

Management

Group


Manufacture &

Assembly

Medical Writing:

Clinical Report

Authorship

Site of EU

release

Location of

Study Master

File

GCP

Regulatory

inspection

11, Stockley Park,

United Kingdom

;

GlaxoSmithKline,

Research &

Development

Projects Clinical

Platforms and

Sciences,

Research &

Development

Projects Clinical

Platforms and

Sciences Therapy

Area Delivery,

2239, Mississauga,

Canada

Kingdom

;

Complete

Regulatory,

Complete House,

19-21 King Edward

street,

Macclesfield,

Cheshire,

SK101AQ, United

Kingdom

;

GlaxoSmithKline,

Respiratory TAU &

Flexible Discovery

Unit, Research &

2017N324761_00206215

PPD

PPD

PPD

CONFIDENTIAL

6





Protocol

Authorship

Statistics

Group

Data

Management

Group


Manufacture &

Assembly

Medical Writing:

Clinical Report

Authorship

Site of EU

release

Location of

Study Master

File

GCP

Regulatory

inspection

;

GlaxoSmithKline,

Global Medical,

Respiratory

Medical Franchise,

Stockley Park,

GlaxoSmithKline

House, United

Kingdom

;

GlaxoSmithKline,

Research &

Development

Platform

Technology &

Science, PDS -

Development

Respiratory HUP

Clin Dev 2, Building

11, Stockley Park,

United Kingdom

;

GlaxoSmithKline,

Research &

Development

Projects Clinical

Platforms and

Sciences,

Research &

Development

Projects Clinical

Platforms and

Sciences Therapy

2017N324761_00206215

PPD

PPD

PPD

CONFIDENTIAL

7





Protocol

Authorship

Statistics

Group

Data

Management

Group


Manufacture &

Assembly

Medical Writing:

Clinical Report

Authorship

Site of EU

release

Location of

Study Master

File

GCP

Regulatory

inspection

T2PS - Supply

Chain Manager

Group, 15.209C,

Zebulon, United

States

;

GlaxoSmithKline,

Respiratory TAU &

Flexible Discovery

Unit, Respiratory,

Upper Merion,

United States

Area Delivery,

2239, Mississauga,

Canada

;

GlaxoSmithKline,

Global Medical,

Respiratory

Medical Franchise,

Stockley Park,

GlaxoSmithKline

House, United

Kingdom

;

GlaxoSmithKline,

Research &

Development

2017N324761_00206215

PPD

PPD

PPD

CONFIDENTIAL

8





Protocol

Authorship

Statistics

Group

Data

Management

Group


Manufacture &

Assembly

Medical Writing:

Clinical Report

Authorship

Site of EU

release

Location of

Study Master

File

GCP

Regulatory

inspection

Platform

Technology &

Science, PDS -

T2PS - Supply

Chain Manager

Group, 15.209C,

Zebulon, United

States

GlaxoSmithKline,

Respiratory TAU &

Flexible Discovery

Unit, Respiratory,

Upper Merion,

United States

2017N324761_00206215

PPD

PPD

CONFIDENTIAL

9





Protocol

Authorship

Statistics

Group

Data

Management

Group


Manufacture &

Assembly

Medical Writing:

Clinical Report

Authorship

Site of EU

release

Location of

Study Master

File

GCP

Regulatory

inspection

GlaxoSmithKline,

Research &

Development

Projects Clinical

Platforms and

Sciences,

Respiratory,

Projects Clinical

Platforms and

Sciences, Stockley

Park, United

Kingdom

2017N324761_00206215

CONFIDENTIAL

10

Table 2: Country Level Administration


Study Title: A randomized, open-label, cross-over, placebo-device study investigating critical and over all errors,

training/teaching time, and preference attributes of the ELLIPTA dry powder Inhaler (DPI) as compared to

HandiHaler DPI used in combination with either DISKUS DPI or Turbuhaler DPI, in adult patients with Chronic

Obstructive Pulmonary Disease (COPD)

Participating

Countries Monitoring

Responsible for

which sites

Laboratory

Assessments

Sites of

distribution in

Europe Audits

Were Audits

done?

Centre Number

of Site(s) Conducted by:

Country Name

Address of

Country Medical

Department

responsible for

monitoring Centre Numbers

Laboratory name

and address Address yes/no Centre numbers

Name and

address of audit

group

Netherlands

GlaxoSmithKline

BV Netherlands,

Huis ter

Heideweg 62,

None GlaxoSmithKline

Research &

Development

Limited, New

Frontiers Science

No Not Applicable Not Applicable

2017N324761_00206215

PPDPPD

CONFIDENTIAL

11






Participating


Responsible for

which sites

Laboratory

Assessments

Sites of

distribution in

Europe Audits

Were Audits

done?

Centre Number


3705 LZ Zeist,

Netherlands

Park (South),

Third Avenue,

Harlow, Essex,

CM19 5AW,

United Kingdom

United Kingdom

GlaxoSmithKline

United Kingdom,

Building 10,

Stockley Park

West, Uxbridge,

None GlaxoSmithKline

Research &

Development

Limited, New

Frontiers Science

Park (South),

Third Avenue,

No Not Applicable Not Applicable

2017N324761_00206215

PPDPPD

PPD

CONFIDENTIAL

12






Participating


Responsible for

which sites

Laboratory

Assessments

Sites of

distribution in

Europe Audits

Were Audits

done?

Centre Number


Middlesex, UB11

1BT, United

Kingdom

Harlow, Essex,

CM19 5AW,

United Kingdom

2017N324761_00206215

1

PPD

PPD

PPD

1

PPD

PPD

Protocol: 206215 Page 1 of 22Population: Intent-to-Treat Listing 5 Listing of Randomised and Actual Treatment Sequence

Sub Study 1: ELLIPTA vs DISKUS + HandiHalerCountry: NetherlandsInvestigator ID:Investigator Name:Investigator at centre: Randomisation Randomised Treatment Subject Number Sequence Actual Treatment Sequence Deviation ---------------------------------------------------------------------- ELLIPTA/DISKUS + ELLIPTA/DISKUS + No HandiHaler/Q1 HandiHaler/Q1

DISKUS + DISKUS + No HandiHaler/ELLIPTA/Q2 HandiHaler/ELLIPTA/Q2



ELLIPTA/DISKUS + ELLIPTA/DISKUS + No HandiHaler/Q1 HandiHaler/Q1





Note: Q1 = Questionnaire 1, Q2 = Questionnaire 2, Q3 = Questionnaire 3, Q4 = Questionnaire 4.

10

CONFIDENTIAL 2017N324761_00206215

PPD

PPDPPD

PPD

PPD


Sub Study 1: ELLIPTA vs DISKUS + HandiHalerCountry: NetherlandsInvestigator ID:Investigator Name:Investigator at centre: Randomisation Randomised Treatment Subject Number Sequence Actual Treatment Sequence Deviation ----------------------------------------------------------------------- ELLIPTA/DISKUS + ELLIPTA/DISKUS + No HandiHaler/Q1 HandiHaler/Q1








11

CONFIDENTIAL 2017N324761_00206215

PPD

PPDPPD

PPD

PPD


Sub Study 1: ELLIPTA vs DISKUS + HandiHalerCountry: NetherlandsInvestigator ID:Investigator Name:Investigator at centre: Randomisation Randomised Treatment Subject Number Sequence Actual Treatment Sequence Deviation --------------------------------------------------------------------- DISKUS + DISKUS + No HandiHaler/ELLIPTA/Q1 HandiHaler/ELLIPTA/Q1










12

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

PPD

PPDPPD












13

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

PPDPPD

PPD








ELLIPTA/DISKUS + ELLIPTA/DISKUS + Yes HandiHaler/Q2 HandiHaler/Q1




14

CONFIDENTIAL 2017N324761_00206215

PPD

PPDPPD

PPD

PPD


Sub Study 1: ELLIPTA vs DISKUS + HandiHalerCountry: NetherlandsInvestigator ID:Investigator Name:Investigator at centre: Randomisation Randomised Treatment Subject Number Sequence Actual Treatment Sequence Deviation ---------------------------------------------------------------------- DISKUS + DISKUS + No HandiHaler/ELLIPTA/Q2 HandiHaler/ELLIPTA/Q2










15

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

PPD

PPD

PPD


Sub Study 1: ELLIPTA vs DISKUS + HandiHalerCountry: NetherlandsInvestigator ID:Investigator Name:Investigator at centre: Randomisation Randomised Treatment Subject Number Sequence Actual Treatment Sequence Deviation ----------------------------------------------------------------------- DISKUS + DISKUS + No HandiHaler/ELLIPTA/Q2 HandiHaler/ELLIPTA/Q2










16

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

PPDPPD

PPD




17

CONFIDENTIAL 2017N324761_00206215

PPD

PPDPPD

PPD

PPD


Sub Study 1: ELLIPTA vs DISKUS + HandiHalerCountry: United KingdomInvestigator ID:Investigator Name:Investigator at centre: Randomisation Randomised Treatment Subject Number Sequence Actual Treatment Sequence Deviation ----------------------------------------------------------------------- DISKUS + DISKUS + No HandiHaler/ELLIPTA/Q1 HandiHaler/ELLIPTA/Q1





18

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

PPD

PPD

PPD


Sub Study 1: ELLIPTA vs DISKUS + HandiHalerCountry: United KingdomInvestigator ID:Investigator Name:Investigator at centre: Randomisation Randomised Treatment Subject Number Sequence Actual Treatment Sequence Deviation ----------------------------------------------------------------------- ELLIPTA/DISKUS + ELLIPTA/DISKUS + No HandiHaler/Q2 HandiHaler/Q2










19

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

PPD

PPD

PPD


Sub Study 1: ELLIPTA vs DISKUS + HandiHalerCountry: United KingdomInvestigator ID:Investigator Name:Investigator at centre: Randomisation Randomised Treatment Subject Number Sequence Actual Treatment Sequence Deviation ----------------------------------------------------------------------- ELLIPTA/DISKUS + ELLIPTA/DISKUS + No HandiHaler/Q2 HandiHaler/Q2






20

CONFIDENTIAL 2017N324761_00206215

PPDPPD

PPD

PPD

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerCountry: NetherlandsInvestigator ID:Investigator Name:Investigator at centre: Randomisation Randomised Treatment Subject Number Sequence Actual Treatment Sequence Deviation ----------------------------------------------------------------------- Turbuhaler + Turbuhaler + No HandiHaler/ELLIPTA/Q4 HandiHaler/ELLIPTA/Q4

Turbuhaler + Turbuhaler + No HandiHaler/ELLIPTA/Q4 HandiHaler/ELLIPTA/Q4


ELLIPTA/Turbuhaler + ELLIPTA/Turbuhaler + No HandiHaler/Q3 HandiHaler/Q3







21

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

PPD

PPDPPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerCountry: NetherlandsInvestigator ID:Investigator Name:Investigator at centre: Randomisation Randomised Treatment Subject Number Sequence Actual Treatment Sequence Deviation ----------------------------------------------------------------------- ELLIPTA/Turbuhaler + ELLIPTA/Turbuhaler + No HandiHaler/Q3 HandiHaler/Q3






22

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

PPD

PPD

PPD












23

CONFIDENTIAL 2017N324761_00206215

PPD

PPDPPD

PPD

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerCountry: NetherlandsInvestigator ID:Investigator Name:Investigator at centre: Randomisation Randomised Treatment Subject Number Sequence Actual Treatment Sequence Deviation ---------------------------------------------------------------------- Turbuhaler + Turbuhaler + No HandiHaler/ELLIPTA/Q4 HandiHaler/ELLIPTA/Q4








24

CONFIDENTIAL 2017N324761_00206215

PPD

PPDPPD

PPD

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerCountry: NetherlandsInvestigator ID:Investigator Name:Investigator at centre: Randomisation Randomised Treatment Subject Number Sequence Actual Treatment Sequence Deviation ---------------------------------------------------------------------- ELLIPTA/Turbuhaler + ELLIPTA/Turbuhaler + No HandiHaler/Q4 HandiHaler/Q4










25

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

PPD

PPD

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerCountry: NetherlandsInvestigator ID:Investigator Name:Investigator at centre: Randomisation Randomised Treatment Subject Number Sequence Actual Treatment Sequence Deviation ---------------------------------------------------------------------- ELLIPTA/Turbuhaler + ELLIPTA/Turbuhaler + No HandiHaler/Q4 HandiHaler/Q4










26

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

PPD

PPD

PPD












27

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

PPD

PPD

PPD






28

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

PPD

PPD

PPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerCountry: United KingdomInvestigator ID:Investigator Name:Investigator at centre: Randomisation Randomised Treatment Subject Number Sequence Actual Treatment Sequence Deviation --------------------------------------------------------------------- Turbuhaler + Turbuhaler + No HandiHaler/ELLIPTA/Q3 HandiHaler/ELLIPTA/Q3










29

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

PPD

PPDPPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerCountry: United KingdomInvestigator ID:Investigator Name:Investigator at centre: Randomisation Randomised Treatment Subject Number Sequence Actual Treatment Sequence Deviation ----------------------------------------------------------------------- ELLIPTA/Turbuhaler + ELLIPTA/Turbuhaler + No HandiHaler/Q4 HandiHaler/Q4










30

CONFIDENTIAL 2017N324761_00206215

PPDPPD

PPD

PPD

PPD

PPD

PPDPPD


Sub Study 2: ELLIPTA vs Turbuhaler + HandiHalerCountry: United KingdomInvestigator ID:Investigator Name:Investigator at centre: Randomisation Randomised Treatment Subject Number Sequence Actual Treatment Sequence Deviation ----------------------------------------------------------------------- Turbuhaler + Turbuhaler + No HandiHaler/ELLIPTA/Q4 HandiHaler/ELLIPTA/Q4


31

CONFIDENTIAL 2017N324761_00206215

PPD

PPD

PPD

PPDPPD

Protocol: 206215 Page 1 of 1Population: Intent-to-Treat Listing 32 Listing of All Cause Deaths

No data to report

514

CONFIDENTIAL 2017N324761_00206215

PPD