Intravenous Iron for the Treatment of Anaemia in Critical ... · Anaemia in Intensive Care (IRONMAN) RCT was conducted in four centres in Western Australia over a period of two years

Intravenous Iron for the Treatment of Anaemia in Critical Illness

Edward Litton MBChB MSc FCICM

This thesis is presented for the degree of Doctor of Philosophy of the University of Western Australia

School of Medicine

January 2018

ii

Thesis declaration

I, Edward Litton certify that:

This thesis has been substantially accomplished during enrolment in the degree.

This thesis does not contain material which has been accepted for the award of any other degree

or diploma in my name, in any university or other tertiary institution. No part of this work will, in

the future, be used in a submission in my name, for any other degree or diploma in any university

or other tertiary institution without the prior approval of The University of Western Australia and

where applicable, any partner institution responsible for the joint-award of this degree. This thesis

does not contain any material previously published or written by another person, except where

due reference has been made in the text. The work(s) are not in any way a violation or

infringement of any copyright, trademark, patent, or other rights whatsoever of any person. The

research involving human data reported in this thesis was assessed and approved by The

University of Western Australia Human Research Ethics Committee. Approval #: RA/4/1/6200

Written patient consent has been received and archived for the research involving patient data

reported in this thesis. The following approvals were obtained prior to commencing the relevant

work described in this thesis: EMHS HREC REG 13-042, SCGH HREC RN 2013/065, JHC

HREC 1307, SMHS HREC ref 12/347. The work described in this thesis was partially funded

by SHRAC Research Translation Project Grant 2016 (Round 6) Ref: F-AA-12440 $200,000

Intravenous iron or placebo for anaemia in intensive care: a randomised controlled study. This

thesis contains published work and/or work prepared for publication, some of which has been

co-authored.

Signature:

Date: 04/07/2017

iii

Abstract

Both anaemia and allogeneic red blood cell transfusion are common and potentially harmful in

patients admitted to the intensive care unit (ICU). This project was designed to test the

hypothesis that intravenous iron therapy is effective in reducing allogeneic red blood cell (RBC)

transfusion requirement in critically ill patients with anaemia who are admitted to the ICU.

A systematic review and meta-analysis of 75 studies of intravenous iron administration in acute

care settings found that intravenous iron compared with oral or no iron was effective in

decreasing RBC transfusion (risk ratio [RR] 0.7, 95% CI 0.6-0.9), but was associated with an

increased risk of infection (RR 1.3, 95% CI 1.1-1.6). No high quality randomised controlled trials

(RCT) of intravenous iron therapy in critically ill patients admitted to the ICU were identified.

Having established a plausible effect size for intravenous iron therapy in reducing the risk of RBC

transfusion, a prospective observational study was conducted to inform the design of a phase II

RCT of intravenous iron therapy in critically ill patients admitted to the ICU. In this study, simple

clinical characteristics available early after ICU admission were identified that could predict the

subsequent risk RBC transfusion. In addition, standard measures of iron metabolism were found

to be of limited value in predicting subsequent RBC transfusion.

A multicentre, randomised, placebo-controlled trial of 140 participants was designed on the basis

of these data, including pre-publication of the study protocol. The Intravenous Iron or Placebo for

Anaemia in Intensive Care (IRONMAN) RCT was conducted in four centres in Western Australia

over a period of two years. The iron group received 97 red blood cell units versus 136 red blood

cell unit in the placebo group. The incidence rate ratio for RBC transfusion was 0.71 [95%

confidence interval (0.43-1.18) P=0.19]. The median haemoglobin at hospital discharge was

significantly higher in the IV iron group compared with the placebo group (107 g/L (IQR 97-115)

vs. 100 g/L (IQR 89-111), P=0.02). No immediate severe adverse reactions were reported in

participants receiving intravenous iron and no between-group difference was found in incident

infection.

iv

In a follow up study using a nested cohort design, baseline hepcidin concentration of patients

enrolled in the IRONMAN study was found to be an independent predictor of subsequent RBC

transfusion requirement. The association between intravenous iron therapy and RBC transfusion

was also modified by baseline hepcidin concentration, suggesting that baseline hepcidin

concentration may have prognostic value in identifying patients in whom intravenous iron therapy

is effective in decreasing RBC transfusion quantity.

On the basis of these results, future studies of intravenous iron therapy in critically ill patients

admitted to the ICU may benefit from targeting a cohort of patients at higher risk of subsequent

RBC transfusion and, potentially, greater response to intravenous iron therapy. Based on a mean

RBC transfusion of 1.9 units (standard deviation 3), and a mean difference of 0.5 units, found in

the IRONMAN study, a future trial of 567 participants per group would have 80% power to detect

a change in RBC units transfused of 0.5 units (alpha=0.05).

v

Table of Contents

Thesis declaration……………………………………………................................... ii

Abstract…………………………………………………………………………………. iii

Table of contents………………………………………………………………………. v

Acknowledgements…………………………………………………………………… viii

Authorship declaration: Co-authored publications……………………….………… ix

Abbreviations…………………………………………………………………………… xiv

Chapter 1. Introduction……………………………………………………………… 1

Chapter 2. The Safety and Efficacy of Intravenous Iron Therapy in Reducing

Allogeneic Blood Transfusion: A Systematic Review and Meta-analysis of

Randomised Clinical Trials…………………………………………………………....

8

2.1 Introduction…………………………………………………………………….. 9

2.2 Methods………………………………………………………………………… 10

2.3 Results………………………………………………………………………….. 12

2.4 Discussion……………………………………………………………………… 15

2.5 Conclusion……………………………………………………………………… 17

2.6 Figures………………………………………………………………………….. 18

2.7 Tables…………………………………………………………………………… 24

Chapter 3. Iron-Restricted Erythropoiesis and Risk of Red Blood Cell

Transfusion in the Intensive Care Unit: A Prospective Observational Study……

42

3.1 Introduction…………………………………………………………………….. 43

3.2 Methods………………………………………………………………………… 44

3.3 Results………………………………………………………………………….. 46

3.4 Discussion……………………………………………………………………… 48

3.5 Conclusion……………………………………………………………………… 50

3.6 Figures……………………...………………………………………………….. 51

3.7 Tables…………………………………………………………………………… 52

vi

Chapter 4. The IRONMAN trial: A Protocol For a Multicentre Randomised

Blinded Trial of Intravenous Iron in Intensive Care Unit Patients With

Anaemia…………………………………………………………………………………

56

4.1 Introduction…………………………………………………………………….. 57

4.2 Study design…………………………………………………………………… 59

4.3 Data management…………………………………………………………….. 63

4.4 Conclusion……………………………………………………………………… 64

4.5 Figures……………………...………………………………………………….. 65

4.6 Tables…………………………………………………………………………… 66

Chapter 5. Intravenous Iron or Placebo for Anaemia in Intensive Care: The

IRONMAN Multicentre Randomized Blinded Trial…………………………...…….

69

5.1 Introduction…………………………………………………………………….. 70

5.2 Methods………………………………………………………………………… 71

5.3 Results………………………………………………………………………….. 75

5.4 Discussion……………………………………………………………………… 77

5.5 Conclusion……………………………………………………………………… 81

5.6 Figures……………………...………………………………………………….. 82

5.7 Tables…………………………………………………………………………… 83

5.8 Supplementary data…………………………………………………………… 86

Chapter 6. Utility of Hepcidin in Predicting risk of Red Blood Cell Transfusion

and Response to IV Ion Therapy in Patients Admitted to the Intensive Care

Unit: A Nested Cohort Study…...……………………………………………………..

89

6.1 Introduction…………………………………………………………………….. 87

6.2 Methods………………………………………………………………………… 88

6.3 Results………………………………………………………………………….. 90

6.4 Discussion……………………………………………………………………… 92

6.5 Conclusion……………………………………………………………………… 94

vii

6.6 Figures……………………...………………………………………………….. 95

6.7 Tables…………………………………………………………………………… 97

Chapter 7. Conclusion……………………………………………………………….... 104

7.1 Thesis overview………………………………………………………………... 105

7.2 Limitations……………………………………………………………………… 106

7.3 Significance and future directions…………………………………………… 108

References…………………………………………………………………..………… 110

Appendix 1 – Human Research Ethics Committee Approvals……………………. 120

Appendix 2 - Publications, Presentations and Prizes Arising From this Thesis… 126

viii

Acknowledgements I would like to thank my supervisors Prof S Webb, A/Prof K M Ho, Prof W Erber and Dr J Allan for

the time and support provided to me in completing this work. I would also like to thank my co-

investigators and colleagues in the ICU who contributed to this research.

The IRONMAN study (Chapters 4,5 & 6) was funded by a grant from the West Australian State

Health Research Advisory Council. Study drug was supplied by Vifor Pharma according to a

Letter of Understanding reviewed and agreed upon by Royal Perth Hospital and adhering to the

principles of scientific independence in the conduct and reporting of the trial. The IRONMAN

study was endorsed by the Australian and New Zealand Intensive Care Society Clinical Trials

Group (ANZICS CTG) and is part of the Blood-Centre of Research Excellence.

The IRONMAN study was registered prospectively with the Australian New Zealand Clinical Trials

Registry (ACTRN12612001249842).

This research was not supported by an Australian Government Research Training Program

(RTP) Scholarship.

ix

Authorship Declaration: Co-authored Publications

This thesis contains work that has been published and is being prepared for publication:

Details of the work: The Safety and Efficacy of Intravenous Iron Therapy in Reducing

Allogeneic Blood Transfusion: A Systematic Review & Meta-analysis of RCTs

Litton et al BMJ 2013;347;f4822

Location in thesis: Chapter 2

Student contribution to work: Edward Litton conceived the initial idea for the systematic review,

devised the research protocol, the case report form and the database, and was one of two

authors (along with Dr Xiao) who conducted the online search and extracted data into the case

report form. Edward Litton conducted all analyses of the data (except for the meta-regression

conducted by A/Prof Ho), wrote the first draft of the manuscript and was responsible for editing

all subsequent drafts including submitted and accepted versions to the BMJ.

Co-author signatures and dates:

x

Details of the work: Iron-restricted erythropoiesis and risk of red blood cell transfusion in the

Intensive Care Unit: A prospective observational study

Litton et al. Anaesthesia and Intensive Care 2015; 43(5) 612-6


Student contribution to work: Edward Litton conceived the initial idea for the observational study,

devised the research protocol, the case report form and the database, and was one of two

authors (along with Dr Xiao) who collected the data and transcribed it into the database. Edward

Litton conducted all analyses of the data, wrote the first draft of the manuscript and was

responsible for editing all subsequent drafts including submitted and accepted versions to

Anaesthesia & Intensive Care.

Co-author signatures and dates:

xi

Details of the work: The IRONMAN trial: A protocol for a multicentre randomised placebo-

controlled trial of intravenous iron in intensive care unit patients with anaemia

Litton et al. Critical Care & Resuscitation 2015 17(2) 144-5


Student contribution to work: Dr Edward Litton conceived the initial idea for the randomised

controlled trial along with Prof Steve Webb (supervisor and co-author) and Prof Toby Richards

(co-author). Dr Edward Litton wrote the first and subsequent drafts of the protocol, the case

report form, data dictionary and managed the study databases. Dr Edward Litton, along with

other site investigators, the project manager and research coordinators, contributed to data

collection and gaining participant consent. Dr Edward Litton conducted start up meetings at all

contributing sites and chaired all meetings of the study management committee.

Dr Edward Litton wrote the first draft of the protocol manuscript and was responsible for editing all

subsequent drafts including submitted and accepted versions to Critical Care & Resuscitation. I

certify the above statement of contribution to be correct and agree to inclusion of the above work

in the PhD thesis of Dr Edward Litton

Coordinating Supervisor signature:

xii

Details of the work: Intravenous Iron or Placebo for Anaemia in Intensive Care: The IRONMAN

multicentre randomized blinded trial

Litton et al. Intensive Care Medicine 2016 42(11) 1715-1722


Student contribution to work: Dr Edward Litton conceived the initial idea for the randomised

controlled trial along with Prof Steve Webb (supervisor and co-author) and Prof Toby Richards

(co-author). Dr Edward Litton wrote the first and subsequent drafts of the protocol, the case

report form, data dictionary and managed the study databases. Dr Edward Litton, along with

other site investigators, the project manager and research coordinators, contributed to data

collection and gaining participant consent. Dr Edward Litton conducted start up meetings at all

contributing sites and chaired all meetings of the study management committee. Dr Edward

Litton was responsible for collating and cleaning the database and conducted all analyses of

the data. Dr Edward Litton wrote the first draft of the primary results manuscript and was

responsible for editing all subsequent drafts including submitted and accepted versions to

Intensive Care Medicine. I certify the above statement of contribution to be correct and agree

to inclusion of the above work in the PhD thesis of Dr Edward Litton


xiii

Details of the work: Utility of hepcidin in predicting risk of red blood cell transfusion and

response to IV iron therapy in patients admitted to the Intensive Care Unit: A nested cohort

study

(Submitted for publication)


Student contribution to work: Dr Edward Litton conceived the initial idea for the hepcidin

substudy. Dr Edward Litton wrote the first and subsequent drafts of the substudy protocol, the

case report form, data dictionary and managed the study databases. Dr Edward Litton, along

with other site investigators, the project manager and research coordinators, contributed to

data collection and gaining participant consent. Dr Edward Litton was responsible for collating

and cleaning the database and conducted all analyses of the data.

Dr Edward Litton wrote the first draft of the hepcidin sub study manuscript and was responsible

for editing all subsequent drafts including submitted and accepted versions to the target journal. I

certify the above statement of contribution to be correct and agree to inclusion of the above work

in the PhD thesis of Dr Edward Litton


xiv

Abbreviations

AE ……………… Adverse Event

AKI ……………… Acute Kidney Injury

ANZICS ……………… Australian and New Zealand Intensive Care Society

APACHE ……………… Acute Physiology and Chronic Health Evaluation

CI ……………… Confidence Interval

COAD ……………… Chronic Obstructive Airways Disease

CRF ……………… Case Report Form

CTG ……………… Clinical Trials Group

ESA ……………… Erythropoiesis-Stimulating Agent

Fe ……………... Iron

DOH ……………… Department of Health

DVT ……………… Deep Vein Thrombosis

Hb ……………… Haemoglobin

HREC ……………… Human Research Ethics Committee

ICU ……………… Intensive Care Unit

IQR …………… Interquartile Range

IRE ……………. Iron-Restricted Erythropoiesis

IV ……………… Intravenous

LOS ……………… Length Of Stay

MCV ……………… Mean Corpuscular Volume

MV ……………… Mechanical Ventilation

NaCl ……………… Sodium Chloride

NHMRC ……………… National Health & Medical Research Council

PE ……………… Pulmonary Embolism

RBC ……………… Red Blood Cell

RCT ……………… Randomised Controlled Trial

xv

RRT ……………… Renal Replacement Therapy

SAE ……………… Serious Adverse Event

SD ……………… Standard Deviation

SOFA ……………… Sequential Organ Failure Assessment Score

SIRS ……………… Systemic Inflammatory Response Syndrome

WA ……………… Western Australia

1

Chapter 1

Introduction

2

Epidemiology of RBC transfusion and anaemia in critically ill patients admitted to ICU

RBC transfusion is common in critically ill patients in Australia and worldwide, with 17-45% of all

patients admitted to ICU, and more than 70% of those staying greater than 7, days receiving one

or more RBC units 1-3. Critically ill patients who are transfused receive a mean of 4 RBC units in

ICU accounting for nearly 20% of all RBC units transfused in Australia 4. Although RBC

transfusion for major haemorrhage may be life saving, more than 75% of all RBC units transfused

in ICU are given for anaemia 1 2. IV iron is a plausible candidate intervention to improve outcomes

in critically ill patients because it may reduce RBC transfusion and anaemia.

Risks and benefits associated with RBC transfusion

In a recent systematic review, RBC transfusion in critically ill patients was an independent

predictor of death [pooled odds ratio 1.7, 95% confidence interval (CI) 1.4-1.9], nosocomial

infection, multi-organ dysfunction syndrome and acute respiratory distress syndrome 5. Although

there have been substantial changes to practice in the ensuing period, including widespread

implementation of leukoreduction, the 1999 RCT by Hebert et al. remains the highest quality

evidence to guide threshold for RBC in ICU. In this study, a liberal compared with a restrictive

transfusion policy was associated with a 4.6% absolute increase in mortality (p=0.11) 6. The

mortality difference was significant for those with an Acute Physiology and Chronic Health

Evaluation (APACHE) II score of <20 and those under 55 years of age. Proposed mechanisms by

which RBC transfusion could influence mortality and morbidity include the inherent effects of

associated with exposure to an allogeneic material and the potentially modifiable effects of RBC

storage and immunomodulation. Biochemical changes occurring during storage of RBC units

prior to transfusion include depletion of nitric oxide, adenosine triphosphate, and 2,3-

diphosphoglycerate (2,3-DPG). Together with the reduced deformability of stored RBCs, these

changes may mediate adverse effects including impaired oxygen delivery, reduced

microcirculatory flow and activation of the vascular endothelium with intravascular thrombosis,

vasoconstriction and leucocyte adhesion 7.

3

Irrespective of any causal relationship with adverse outcomes, there are additional societal risks

associated with RBC transfusion. Changing population demographics are increasing the scarcity

of this resource. As the population ages the proportion of potential donors (younger people)

decreases and the proportion of potential recipients (older people) increases. The cost of

producing and delivering RBC units for transfusion is difficult to quantify but is also increasing 8.

Although there are risks associated with RBC transfusion, these must be balanced against

potential benefits. RBC transfusion is life-saving for patients with life-threatening bleeding. In the

ICU setting, the most common indication for RBC transfusion is anaemia. For patients with

anaemia, the potential benefit of RBC transfusion relates predominantly to increasing the total

oxygen carrying capacity (DO2) of the blood. Mechanisms to compensate for the decrease in DO2

associated with progressive anaemia include increase in stroke volume, heart rate and oxygen

extraction (VO2). The threshold of anaemia below which DO2 decompensates and is insufficient

to maintain cellular oxygenation varies, depending on the severity of illness and physiological

reserve of the patient. Consequently, there is a threshold below which the benefits of RBC

transfusion will outweigh the risks but this requires individual assessment based on the potential

indication for RBC transfusion, severity of critical illness and physiological reserve.

Risks and benefits associated with anaemia

Anaemia in patients admitted to the ICU is associated with increased risk of adverse outcomes

including prolonged invasive ventilation, myocardial infarction and mortality 9. Although large

studies are lacking, Bateman et al found that a substantial proportion of patients who survive ICU

admission at a single institution were still anaemic at six months, and that this was associated

with evidence of ongoing inflammation 10. However, high quality evidence is lacking to understand

whether the association between anaemic and adverse outcomes is causal.

The needs for additional evidence is heightened by the fact that there may be benefits associated

with anaemia occurring at the onset of critical illness. Anaemia is part of a highly evolutionarily

conserved component of the acute inflammatory response. Mechanisms including the inhibition of

erythropoietin, direct inhibition of erythroid precursors by circulating cytokines and iron

sequestration decrease RBC production. At the same time, RBC destruction is increased and

4

there may be a dilutional effect related to expansion in plasma volume. Potential benefits of the

resultant anaemia include maintenance of blood fluidity in the setting of high concentrations of

circulating cytokines as well the contribution of RBC rheology to microvascular vasoregulation 9.

Physiological iron metabolism

The majority of the 3-4g of iron occurring in a healthy adult human occurs as part of haemoglobin

molecules in red blood cells. Intracellular ferritin stores the majority of the remaining iron, so that

only a very small proportion, less than one percent, is freely available in the circulation.

Regulation of iron body stores occurs entirely by modulation of uptake. Loss occurring through

shedding of mucosal linings and bleeding is not regulated.

The predominant regulator of iron metabolism is hepcidin an endogenous peptide primarily

produced by liver hepatocytes11. Hepcidin inhibits iron absorption and availability by binding to

and degrading ferroportin, the principle cellular iron efflux channel. Hepcidin release is stimulated

by iron loading and inflammation and inhibited by erythropoiesis, iron deficiency and hypoxia.

High hepcidin concentrations result in iron sequestration and may therefore have a role in

predicting whether exogenous iron, delivered as IV iron therapy is likely to made available to

developing RBCs and thus increase erythropoiesis. As such, hepcidin may be useful both as a

diagnostic guide and to monitor response to treatment with IV iron. However its role in critical

illness has not been evaluated to date12.

Iron metabolism in critical illness

Critical illness results in an acute inflammatory response that has substantial effects on iron

metabolism. Both ferritin and hepcidin release occur as part of the acute phase response. The

increased ferritin results in increased iron sequestration due to the higher affinity of circulating

iron for ferritin than transferrin and the down-regulation of macrophage ferroportin by hepcidin.

Increased hepcidin levels also decrease iron absorption. The net effect is decreased circulating

iron levels and reduced iron availability.

Rationale for IV iron therapy in critically ill patients admitted to ICU

In a recent large multicentre observational study of transfusion practice in 47 ICUs in Australia,

only 2% of RBC transfusions were administered outside of current National Health and Medical

5

Research Council (NHMRC) guidelines1. Given that transfusion in the ICU is common despite

extremely high concordance to current restrictive transfusion guidelines, and that current

transfusion thresholds exacerbate the incidence and severity of anaemia in critically ill patients,

there is an unmet need for novel interventions that both reduce the incidence of RBC transfusion

and incidence and/or severity of anaemia.

Epoetin alpha has been investigated as one such agent. Endogenous erythropoietin levels

decrease rapidly in the setting of acute inflammation associated with the onset of critical illness 13.

However, previous RCTs found that ESAs were not effective in reducing RBC transfusion in

critically ill patients and may be associated with an increased risk of thrombotic complications 14.

Iron is essential for endogenous RBC production. In critically ill patients iron-restricted

erythropoiesis may occur through multiple pathways including absolute iron deficiency, iron

sequestration or functional iron deficiency. Risk factors for both iron deficiency and critical illness

include advancing age and chronic illness. However, enteral administration of iron is ineffective

in patients who are critically ill due to gastrointestinal intolerance, decreased iron absorption from

routine use of gastric acid suppression, physiological limits to maximal enteral iron absorption

and inhibition of absorption due to high hepcidin levels that occur in critical illness. Parenteral iron

overcomes these disadvantages and has been shown to be superior to enteral iron for the

correction of iron-restricted erythropoiesis in a number of patient populations 15 16. IV iron is an

established therapy in guidelines for the management of the anaemia of chronic kidney disease17.

In a recent systematic review of 13 studies comparing IV to oral or no iron in a range of patient

populations, and including observational ICU data, IV non-dextran iron was associated with a

significant increase in short-term haemoglobin concentration 18. IV iron is also effective in

improving the anaemia of inflammation associated with iron-restricted erythropoiesis in iron-

replete patients19.

IV Iron Safety

Newer, non-dextran IV iron preparations have a favorable safety profile with a low risk of serious

side effects including anaphylaxis. Of perhaps greater concern for critically ill patients is the

possible association of IV iron with oxidative stress and infection. In an experimental canine

6

model of bacterial pneumonia and anaemia, IV iron compared to fresh RBC transfusion was

associated with an increase in non-transferrin-bound iron, shock and mortality 20. Clinical data

related to oxidative stress associated with IV iron in critically ill patients is lacking. However free

iron, either as a result of endogenous over-saturation or excessive exogenous iron, can result in

cellular injury by catalysing the production of reactive oxygen species causing lipid peroxidation

and oxidation of amino acids 21.

Iron can also potentiate bacterial growth in vitro 22. The growth of certain bacterial species that

are common in patients admitted to the ICU, including Escherichia coli, Klebsiella species and

Pseudomonas species are enhanced by unbound iron, whilst organisms including

Staphylococcus aureus are able to acquire bound iron through membrane transferrin receptors 23.

There is also clinical evidence of a potential causal relationship between IV iron therapy and

infection. Iron therapy was associated with increased malaria, non-malarial infection and mortality

in several studies of oral iron supplementation conducted in malarial regions 24. In contrast, no

increase in infection has been observed with IV iron therapy in large cohort studies of patients

undergoing long-term dialysis, or following surgery in the developed world setting 25 26.

Furthermore, a mouse model of critical illness and anaemia also found no association between IV

iron and risk of infection 27.

The available evidence suggests that the strength of the relationship between iron therapy and

infection risk is likely to be context-dependent. Relevant factors are likely to include the specific

indication, iron preparation, route and dose used, the severity of the acute illness, immune status

and co-morbidities of the cohort under investigation, and the iron scavenging capacity of the

specific pathogenic organisms to which the host is likely to be exposed. Thus, the variation in

findings relevant to critically ill patients have several potential explanations. The majority of

studies in which no association between iron and infection have been found are either

observational, for which the relationship may be obscured by residual confounding, or are smaller

RCTs that may be underpowered to detect a clinically important difference. It is also plausible

that the low free iron levels associated with newer IV iron preparations has attenuated the risk of

infection. In addition, there is evidence that allogeneic RBC transfusion increases the risk of

7

bacterial infection 28. Given that RBC transfusion may be an alternative treatment to IV iron, an

important question remains as to the risk of infection related to IV iron relative to RBC

transfusion. For critically ill patients, a group that are at increased risk of nosocomial infection, the

relationship between IV iron and infection requires consideration in the design of clinical trials in

patients admitted to the ICU 29.

Summary

IV iron therapy is an established treatment for anaemia in a wide variety of clinical settings. RBC

transfusion in critically ill patients admitted to the ICU is a major public health issue. Critically ill

patients account for a substantial proportion of the total quantity of this scarce and costly

resource. RBC transfusion is also associated with adverse outcomes. This project was designed

to test the hypothesis that intravenous iron therapy is effective in reducing allogeneic RBC

transfusion requirement in critically ill patients with anaemia who are admitted to the ICU. The

aims will be addressed in Chapters 2-6. The aim of Chapter 2 was to conduct a systematic review

and meta-analysis of the safety and efficacy of IV iron focusing primarily on RBC transfusion

infection risk. The aim of Chapter 3 was to conduct an observational study describing the

characteristics of patients with iron-restricted erythropoiesis (IRE) on admission to ICU and the

determinants of subsequent RBC transfusion risk. The aims of Chapters 4 and 5 were to describe

the study protocol and the results of a multicentre RCT of IV iron administration in critically ill

patients admitted to the ICU. Chapter 6 describes a nested sub-study, with the aim of assessing

baseline hepcidin concentration as a predictor of RBC transfusion and response to IV iron. In

Chapter 7, the results of the preceding chapters are integrated to answer the hypothesis of the

thesis. Figures and tables occur at the end of each subsequent chapter after the chapter

conclusion and are referenced throughout.

8

Chapter 2

The Safety and Efficacy of Intravenous Iron Therapy in Reducing

Allogeneic Blood Transfusion: A Systematic Review and Meta-

analysis of Randomised Clinical Trials

9

2.1 Introduction

Iron is essential for the production of red blood cells (RBCs) and is the most common nutritional

deficiency worldwide, both in developed and developing countries 30. Allogeneic RBC transfusion

may be lifesaving for the management of acute severe blood loss, however there are increasing

concerns about its associated serious adverse events, costs and scarcity 31. Safe and effective

strategies to reduce RBC transfusion are urgently needed.

Correction of iron deficiency anaemia using oral iron is limited by gastrointestinal absorption and

is particularly ineffective in the setting of co-existing acute or chronic medical conditions32. IV iron

therapy has an established role in the treatment of iron deficiency anaemia, supported by

laboratory results, where oral iron preparations are ineffective or cannot be used 17. Recent

advances in the understanding of iron metabolism and the association between allogeneic RBC

transfusion and adverse outcomes has increased the interest in the use of IV iron to reduce RBC

transfusion in a variety of acute clinical settings 33 34. Although older IV iron preparations were

associated with a risk of anaphylaxis, newer preparations have largely alleviated this problem 35.

Nevertheless, whether IV iron is associated with other important adverse events, in particular the

theoretical risk of infection, remains uncertain18 22 25.

This systematic review and meta-analysis, undertaken according to PRISMA guidelines36, aimed

to evaluate the safety and efficacy of IV iron focusing primarily on its effects on transfusion

requirement and risk of infection.

10

2.2 Methods

Eligibility Criteria

Published RCTs were searched to identify those in which IV iron was compared with either oral

iron or no iron supplementation. Studies were excluded if they were randomised but with a

crossover design, observational, did not report an outcome of interest, provided insufficient data

for outcomes to be reported or did not contain a group or sub-group in which the independent

effect of IV iron could be assessed. The primary outcomes of interest were: change in

haemoglobin (Hb) and risk of transfusion (efficacy), and risk of infection (safety). Secondary

outcomes of interest included adverse events, and serious adverse events as defined by the

primary studies, anaphylaxis, mortality, length of hospital stay (LOS), cost and cost-effectiveness.

Search Strategy

The primary search was conducted using MEDLINE, EMBASE and the Cochrane Central

Register of Controlled Trials for randomised trials using the terms ‘iron’ and ‘ferric compounds’

and ‘intravenous’. The search included the time period between 1966 and June 2013 and was

conducted without language restrictions. Reference lists of all included studies were searched as

well as relevant review articles and conference proceedings. The manufacturers of IV iron

formulations were also contacted to request access to unpublished trial data.

The primary search was conducted independently by two investigators. Where uncertainty

existed on study eligibility or additional data on end points were needed, the corresponding

author was contacted to request further information.

Study Selection

Potentially relevant titles were retrieved for full text review and data from eligible studies was

transcribed into a prespecified proforma. Studies that were published only in abstract form were

excluded. Disagreement on study inclusion or endpoints was resolved by a third investigator.

11

Data Analysis

The primary outcomes of interest of this review were efficacy (change in Hb and proportion of

patients requiring allogeneic RBC transfusion) and safety (all-cause infection). Where studies

included both oral iron and no iron comparison groups, IV iron was compared preferentially with

oral iron. For categorical data, the risk of an outcome was defined as the number of patients with

an event compared with the number of patients with and without an event. For continuous data,

mean, standard deviation and participant number were used. Data from studies fulfilling the

eligibility were pooled for meta-analysis using a random- effects model. Studies with zero events

in one of the study groups were included as this has been shown to provide a more valid estimate

of true treatment effect 37. Standardised mean difference and risk ratio (RR) with 95% confidence

interval for continuous and categorical outcomes were calculated respectively, and a P value of

<0.05 was taken as significant. Heterogeneity was assessed using the I2 statistic and an I2>40%

was considered as significant heterogeneity. Meta-regression was undertaken to examine the

effect of IV iron dose, baseline iron study results and ESA therapy on the associations between

IV iron and the primary outcomes.

A sensitivity analysis was conducted on the efficacy (transfusion) and safety (infection) outcomes

by excluding studies with a high risk of bias for one or more key domains using the Cochrane

Collaboration’s Tool for Assessing Risk of Bias 38. Publication bias was assessed by a funnel plot,

plotting the odds ratio for proportion transfused against the standard error of the log odds ratio.

The statistical analysis was conducted using STATA (Intercooled Version 11.2, StataCorp,

College Station, Texas, USA) and Comprehensive Meta-analysis (version 2.2.034, Biostat, USA,

2006).

12

2.3 Results

The initial electronic search returned 1815 citations. After examination of the titles and abstracts,

126 were retrieved for more detailed examination. A total of 75 studies including 10879

participants fulfilled the inclusion criteria and were included in the systematic review 11 15 16 39-108.

Of the included studies, 72 studies including 10605 participants provided sufficient quantitative

outcome data to be included in the primary meta-analyses. The flow chart of study inclusion is

presented in Figure 1.

Study Characteristics and Validity Assessment

The sample size of the included studies varied between 25 and 507 participants and involved a

wide range of clinical specialties (renal n=19, obstetric n=19, surgical n=11,

oncology/haematology n=11, cardiology n=4, gastroenterology n=4, other n=7 ). The baseline Hb

(range in mean Hb 6.0-14.5 g/dl) and iron studies (range in mean ferritin 7-761 mcg/dl) of the

included patients also varied between studies. The most common IV iron preparation used in the

included studies was iron sucrose (n=42), followed by iron gluconate (n=10) and ferric

carboxymaltose (n=10). Dextran iron was used in seven studies a further six studies used other

iron preparations. Efficacy outcomes including change in Hb and transfusion were variably

reported, as were safety outcomes including infection, mortality, serious adverse events, and

anaphylaxis. The characteristics of the included studies are shown in table 1.

Overall, the risk of bias was low for 18 studies and high for 57 studies The overall high risk of

bias was accounted for by the majority of studies not being blinded to participants or study

personnel (n=56). Additional data included in the systematic review was provided from the

authors of nine studies. The results of the validity assessment are shown in Table 2.

Quantitative Data Synthesis

Change in Hb and proportion of patients requiring allogeneic RBC transfusion

A total of 59 studies comprising 7610 participants reported the change in Hb before and after

treatment. When pooled, IV iron was associated with a significant increase in standardised mean

Hb (0.7 g/dl, 95% CI 0.5-0.8) compared with oral iron or no iron supplementation (see Figure 2).

There was significant heterogeneity between the studies (I2 87.7% p<0.01).

13

A total of 22 studies comprising 3321 participants reported on the risk of requiring allogeneic RBC

transfusion. IV iron therapy was associated with a significant reduction in the risk of requiring

allogeneic RBC transfusion, (RR 0.7, 95% CI 0.6-0.9) without significant heterogeneity (I2 9%,

p=0.3) (see Figure 3).

There was a potential interaction between use of ESA and the effect of IV iron therapy, with a

greater effect of IV iron on reducing risk of requiring transfusion when concurrent ESA was used

(slope of the regression line 0.32, 95% CI 0.02-0.63, p=0.04) (see Figure 4). Similarly, a lower

baseline ferritin level was associated with greater therapeutic effect in reducing the risk of

requiring RBC transfusion after IV iron therapy (slope of the regression line 0.002, 95% CI 0.002-

0.004, p=0.04). There was however, no interaction between baseline transferrin saturation and

risk of requiring RBC transfusion after IV iron therapy.

Effect of IV iron on all-cause infection

After excluding three studies with zero events in both intervention and comparison groups67 89 94,

a total of 24 studies (n=4400) reported data on risk of infection after receiving IV iron compared

with either oral iron on no iron supplementation. IV iron was associated with a significant increase

in risk of infection of 1.3 (95% CI 1.1-1.6) (see Figure 5) without significant heterogeneity (I2

22.7% p=0.2).

Increased risk of infection was observed both in studies comparing IV iron to oral iron and no

iron. There was no interaction between baseline ferritin, transferrin saturation, iron per dose or

ESA and risk of infection.

Effect of IV iron therapy on other safety end-points

There was no significant difference in mortality (RR 1.1, 95% CI 0.8-1.5), or serious adverse

events (RR 1.1 95%CI 0.9-1.2) with IV iron therapy in 20 and 19 trials respectively. Adverse

events were also not significantly different between IV iron therapy and oral iron or no

supplemental iron (RR 0.9, 95% CI 0.8-1.1). Of the 32 studies reporting anaphylaxis, eight cases

of anaphylaxis were reported in participants receiving IV iron (n=2186).

14

Sensitivity analysis and publication bias

Excluding studies with high risk of bias using the Cochrane Collaboration’s Tool for Assessing

Risk of Bias38, limited the meta-analysis of risk of transfusion to only five studies57 65 77 92 98

(n=901) and this did not change the direction of the association between IV iron and risk of

requiring allogeneic RBC transfusion, but this association was no longer statistically significant

(RR 0.8, 95% CI 0.6-1.1, p=0.66). The sensitivity analysis for risk of infection limited the meta-

analysis to eight studies44 45 58 62 71 81 92 98, and did not substantially change the direction and

magnitude of the association between IV iron therapy and risk of infection (RR 1.4, 95% CI 1.0-

1.8, p=0.03). There was no evidence of publication bias in reporting allogeneic RBC transfusion

requirement in the pooled studies (see Figure 6).

15

2.4 Discussion

IV iron is increasingly advocated to treat anaemia with the aim of reducing allogeneic RBC

transfusion; however, the risks and benefits of IV iron remain uncertain. This is the largest

systematic review and meta-analysis of IV iron to date specifically assessing both the safety and

efficacy of IV iron in patients from a wide spectrum of specialties. In this meta-analysis IV iron

was effective in increasing Hb and reducing the risk of requiring allogeneic RBC transfusion but

was associated with an increased risk of all-cause infection.

Allogeneic RBC transfusion is associated with increased risk of serious adverse events including

increased mortality 5. In this meta-analysis, IV iron was found to be effective in reducing the risk

of RBC transfusion. This benefit appeared consistent across different disease categories and IV

iron formulations and was present whether IV iron was compared with oral iron or no iron. These

findings are in keeping with recent advances in the understanding of iron metabolism; IV iron is

more effective than oral iron, particularly in the setting of acute or chronic inflammation by

bypassing the effects hepcidin - an inhibitor of gastrointestinal iron absorption 109. As such, IV iron

may have an important role in a patient blood management strategy designed to reduce

allogeneic RBC transfusion for many hospitalised patients.

Additionally, these data showed that the effect of IV iron on the risk of requiring allogeneic RBC

transfusion can be further enhanced by concomitant use of ESA. None of the studies included in

the transfusion meta-analysis were conducted in patients with chronic renal failure where ESA

use is a standard of care. The use of ESAs alone may induce a state of functional iron deficiency

and has been postulated as a potential mechanism for the negative results of previous studies of

ESA use in critical care 110 111. Whether the addition of IV iron to ESA in this setting is beneficial

requires further investigation.

The reduction in allogeneic RBC transfusion must be considered alongside our finding of an

increased risk of all-cause infections after IV iron therapy. Free iron has been shown to potentiate

bacterial growth in vitro22. Clinical evidence however, on the association between IV iron therapy

and infection, has been inconclusive, with no increase in infection observed with IV iron therapy

in dialysis, or postoperative surgical patients, or a mouse model of critical care anaemia 27 26 25.

16

This discrepancy may be explained by the low free iron levels associated with newer IV iron

preparations 112 29. Our finding may also be a false positive result. Infection was not a pre-defined

end-point in many pooled studies and it is possible that missing data could have created

unmeasured bias in our analysis. Furthermore, no significant association between iron dose and

risk of infection was found, and overall, serious adverse events and mortality were not

significantly increased in those receiving IV iron compared with oral or no iron. Until RCTs of IV

iron adequately powered for patient-centered outcomes are available, including standardised

definitions for infection, it may be preferable to use IV iron preparations associated with relatively

low free iron concentrations.

Although this was a large a comprehensive systematic review, several limitations bear

consideration. First, data on all outcomes were not available from each study and the doses and

preparations of IV iron used in the pooled studies varied. Nevertheless, heterogeneity in the risk

of requiring transfusion and infection were low, and the number of studies available was sufficient

to conduct a number of meta-regression analyses to assess the interaction between different

predictors and efficacy of IV iron therapy in increasing Hb concentration. Previous smaller

systematic reviews of IV iron therapy were unable to provide sufficient data to estimate important

outcomes such as infection risk or assess the effect of different predictors on efficacy 113 114.

Second, the quality of the included studies was variable and the overall risk of bias of the

included studies was high despite excluding studies that had not been published in their full form.

Finally, meta-regression of trial characteristics using participant level characteristics, such as

baseline ferritin, may lead to aggregation bias and may be underpowered to detect true

differences, such as the effect of IV iron type and dose.115 A large number of studies were

included and meta-regression was only conducted where heterogeneity was low on forest plot

based on an a priori-defined criterion of insignificant heterogeneity (I I2<40%), which reduces but

cannot completely exclude the risk of bias.

17

2.5 Conclusion

IV iron therapy is associated with a significantly reduced risk of requiring allogeneic RBC

transfusion. These findings suggest that IV iron may potentially have broad applicability to many

hospitalised patients in whom anaemia is common. This benefit is counterbalanced by a potential

increased risk of infection. Further RCTs of IV iron are required to define whether it should be

used as a first line agent in reducing allogeneic RBC transfusion in hospitalised patients. Such

trials should include well-defined infection end-points and be adequately powered for important

patient-centered endpoints including mortality and major morbidity.

18

2.6 Figures

Figure 1. Flow Diagram of Study Selection

19

Figure 2. Forrest Plot of Standardised Mean Difference in Haemoglobin with IV Iron Compared

with Oral Iron and No Iron

NOTE: Weights are from random effects analysis

.

.

Overall (I−squared = 87.7%, p = 0.000)

Hulin, S

Bencaiova, G

Li, H

Qunibi, WY

Edwards, TJ

Al−Momen, AK

Hedenus, M

Krayenbuehl, PA

Subtotal (I−squared = 89.2%, p = 0.000)

Li, H

Dangsuwan, P

Okonko, DO

Kim, YH

Pollack, A

Toblli, JE

Aggarwal, HK

Garrido−Martin, P

Madi−Jebara, SN

Provenzano, P

Auerbach, M

Sloand, JA

Warady, BA

Khalafallah, A

Singh, K

Steensma, DP

Van Iperen, CE

Agarwal, R

Anker, SD

Froessler, B

Charytan, C

Verma, S

Karkouti, K

name

McMahon, LP

Li, H

Adhikary, L

Al, RA

Spinowitz, BS

Singh, H

Grote, L

Onken, JE

IV Iron Compared With No Iron

Schroder, MD

Kulnigg, S

Kochhar, PK

Seid, MH

Kasper, SM

Henry, DH

Stoves, J

Neeru, S


Maccio, A

Breymann, C

Schindler, E

Macdougall, IC

Bayoumeu, F

Westad, S

Beck−Da−Silva, L

Bhandal, N

Coyne, DW

Auerbach, M

Shafi, D

Olijhoek, G

IV Iron Compared With Oral Iron

2005

2009

2008

2011

2009

1996

2007

2011

2008

2010

2008

2009

2001

2007

2003

2012

2004

2009

2004

2004

2004

2010

1998

2011

2000

2006

2009

2013

2005

2011

2006

year

2010

2009

2011

2005

2008

2006

2009

2013

2005

2008

2012

2008

1998

2007

2001

2012

2010

2008

1994

1996

2002

2008

2013

2006

2007

2010

2012

2001

0.65 (0.51, 0.79)

0.11 (−0.30, 0.52)

−0.02 (−0.26, 0.22)

1.71 (1.31, 2.10)

0.44 (0.19, 0.70)

0.62 (0.10, 1.14)

1.70 (1.27, 2.14)

0.24 (−0.24, 0.72)

1.54 (1.07, 2.01)

0.66 (0.49, 0.82)

1.13 (0.50, 1.76)

0.59 (−0.02, 1.19)

0.09 (−0.62, 0.80)

1.57 (0.97, 2.17)

1.60 (0.72, 2.48)

2.91 (2.01, 3.82)

1.59 (0.88, 2.31)

0.00 (−0.38, 0.38)

0.56 (0.12, 1.00)

0.43 (0.17, 0.70)

1.41 (1.06, 1.76)

1.20 (0.34, 2.06)

0.02 (−0.64, 0.68)

0.70 (0.41, 1.00)

1.01 (0.60, 1.43)

0.13 (−0.06, 0.32)

0.28 (−0.53, 1.08)

0.23 (−0.22, 0.69)

0.29 (0.08, 0.51)

0.10 (−0.18, 0.38)

0.42 (0.02, 0.83)

1.39 (1.03, 1.75)

0.28 (−0.47, 1.04)

SMD (95% CI)

0.86 (0.42, 1.31)

0.65 (0.37, 0.94)

0.76 (0.33, 1.19)

0.65 (0.23, 1.08)

0.56 (0.30, 0.83)

0.70 (0.26, 1.14)

0.35 (−0.16, 0.86)

0.77 (0.59, 0.95)

0.33 (−0.25, 0.92)

0.60 (0.29, 0.91)

2.17 (1.67, 2.67)

0.67 (0.43, 0.90)

0.00 (−0.39, 0.39)

0.96 (0.57, 1.35)

0.07 (−0.52, 0.65)

1.13 (0.68, 1.58)

0.62 (0.34, 0.90)

−0.18 (−0.50, 0.14)

0.05 (−0.17, 0.28)

1.53 (0.96, 2.11)

1.77 (0.97, 2.58)

0.01 (−0.56, 0.58)

−0.77 (−1.13, −0.41)

−1.16 (−2.21, −0.11)

0.40 (−0.20, 1.00)

0.37 (0.03, 0.71)

0.60 (0.34, 0.86)

1.00 (0.71, 1.29)

0.43 (−0.12, 0.98)

100.00

1.79

1.97

1.81

1.96

1.63

1.75

1.69

1.70

77.92

1.49

1.52

1.37

1.52

1.17

1.14

1.37

1.82

1.75

1.95

1.86

1.18

1.44

1.92

1.78

2.02

1.25

1.73

2.00

1.94

1.79

1.85

1.31

Weight

1.74

1.93

1.76

1.77

1.95

1.74

1.65

2.03

1.55

1.91

1.67

1.98

1.81

1.81

1.55

1.73

22.08

1.89

1.99

1.56

1.25

1.56

1.85

0.98

1.52

1.87

1.96

1.92

1.59

%

0.65 (0.51, 0.79)

0.11 (−0.30, 0.52)

−0.02 (−0.26, 0.22)

1.71 (1.31, 2.10)

0.44 (0.19, 0.70)

0.62 (0.10, 1.14)

1.70 (1.27, 2.14)

0.24 (−0.24, 0.72)

1.54 (1.07, 2.01)

0.66 (0.49, 0.82)

1.13 (0.50, 1.76)

0.59 (−0.02, 1.19)

0.09 (−0.62, 0.80)

1.57 (0.97, 2.17)

1.60 (0.72, 2.48)

2.91 (2.01, 3.82)

1.59 (0.88, 2.31)

0.00 (−0.38, 0.38)

0.56 (0.12, 1.00)

0.43 (0.17, 0.70)

1.41 (1.06, 1.76)

1.20 (0.34, 2.06)

0.02 (−0.64, 0.68)

0.70 (0.41, 1.00)

1.01 (0.60, 1.43)

0.13 (−0.06, 0.32)

0.28 (−0.53, 1.08)

0.23 (−0.22, 0.69)

0.29 (0.08, 0.51)

0.10 (−0.18, 0.38)

0.42 (0.02, 0.83)

1.39 (1.03, 1.75)

0.28 (−0.47, 1.04)

SMD (95% CI)

0.86 (0.42, 1.31)

0.65 (0.37, 0.94)

0.76 (0.33, 1.19)

0.65 (0.23, 1.08)

0.56 (0.30, 0.83)

0.70 (0.26, 1.14)

0.35 (−0.16, 0.86)

0.77 (0.59, 0.95)

0.33 (−0.25, 0.92)

0.60 (0.29, 0.91)

2.17 (1.67, 2.67)

0.67 (0.43, 0.90)

0.00 (−0.39, 0.39)

0.96 (0.57, 1.35)

0.07 (−0.52, 0.65)

1.13 (0.68, 1.58)

0.62 (0.34, 0.90)

−0.18 (−0.50, 0.14)

0.05 (−0.17, 0.28)

1.53 (0.96, 2.11)

1.77 (0.97, 2.58)

0.01 (−0.56, 0.58)

−0.77 (−1.13, −0.41)

−1.16 (−2.21, −0.11)

0.40 (−0.20, 1.00)

0.37 (0.03, 0.71)

0.60 (0.34, 0.86)

1.00 (0.71, 1.29)

0.43 (−0.12, 0.98)

100.00

1.79

1.97

1.81

1.96

1.63

1.75

1.69

1.70

77.92

1.49

1.52

1.37

1.52

1.17

1.14

1.37

1.82

1.75

1.95

1.86

1.18

1.44

1.92

1.78

2.02

1.25

1.73

2.00

1.94

1.79

1.85

1.31

Weight

1.74

1.93

1.76

1.77

1.95

1.74

1.65

2.03

1.55

1.91

1.67

1.98

1.81

1.81

1.55

1.73

22.08

1.89

1.99

1.56

1.25

1.56

1.85

0.98

1.52

1.87

1.96

1.92

1.59

%

0−3.82 0 3.82

20

Figure 3. Forrest Plot of Risk of Red Blood Cell Transfusion with IV Iron Compared With Oral Iron

and No Iron


.

.


Weisbach, V

name


Auerbach, M

Na, HS


Meyer, MP

Steensma, DP

Dangsuwan, P

Breymann, C

Froessler, B

Hedenus, M

Kim, YT

Auerbach, M

Garrido−Martin, P

Henry, DH

Pedrazzoli, P

Al, RA

Karkouti, K

Edwards, TJ

Madi−Jebara, SN

Serrano−Trenas, JA

Westad, S

Kochhar, PK

IV Iron Compared with No Iron

IV Iron Compared with Oral Iron

Bayoumeu, F

1999

year

2004

2011

1996

2011

2010

2008

2013

2007

2007

2010

2012

2007

2008

2005

2006

2009

2004

2011

2008

2012

2002

0.74 (0.62, 0.88)

2.40 (0.80, 7.23)

RR (95% CI)

0.64 (0.49, 0.85)

0.91 (0.39, 2.12)

0.38 (0.21, 0.68)

0.82 (0.67, 1.00)

0.20 (0.01, 3.93)

0.92 (0.56, 1.52)

0.36 (0.16, 0.82)

1.55 (0.06, 37.82)

0.32 (0.03, 3.03)

2.06 (0.20, 21.65)

0.62 (0.38, 1.01)

0.90 (0.65, 1.25)

0.73 (0.47, 1.13)

1.08 (0.55, 2.12)

0.42 (0.08, 2.08)

0.33 (0.01, 7.97)

0.48 (0.15, 1.52)

0.15 (0.01, 3.08)

0.94 (0.46, 1.93)

0.80 (0.56, 1.16)

0.43 (0.14, 1.28)

0.33 (0.01, 7.99)

0.32 (0.01, 7.48)

199/1479

6/30

Treatment

81/425

9/78

11/54

118/1054

0/21

20/164

5/22

1/227

1/101

2/33

12/30

Events,

41/116

20/54

11/63

2/73

0/45

4/21

0/34

17/80

33/100

4/59

0/50

0/24

306/1594

5/60

Control

120/385

10/79

29/54

186/1209

2/21

43/326

14/22

0/117

3/97

1/34

29/45

Events,

48/122

27/53

20/124

5/76

1/45

4/10

2/26

9/40

41/100

11/70

1/50

1/23

100.00

2.25

Weight

41.38

3.73

7.28

58.62

0.32

9.54

3.83

0.28

0.56

0.51

9.77

%

17.56

11.70

5.69

1.08

0.28

2.03

0.32

5.09

15.30

2.30

0.28

0.29

0.74 (0.62, 0.88)

2.40 (0.80, 7.23)

RR (95% CI)

0.64 (0.49, 0.85)

0.91 (0.39, 2.12)

0.38 (0.21, 0.68)

0.82 (0.67, 1.00)

0.20 (0.01, 3.93)

0.92 (0.56, 1.52)

0.36 (0.16, 0.82)

1.55 (0.06, 37.82)

0.32 (0.03, 3.03)

2.06 (0.20, 21.65)

0.62 (0.38, 1.01)

0.90 (0.65, 1.25)

0.73 (0.47, 1.13)

1.08 (0.55, 2.12)

0.42 (0.08, 2.08)

0.33 (0.01, 7.97)

0.48 (0.15, 1.52)

0.15 (0.01, 3.08)

0.94 (0.46, 1.93)

0.80 (0.56, 1.16)

0.43 (0.14, 1.28)

0.33 (0.01, 7.99)

0.32 (0.01, 7.48)

199/1479

6/30

Treatment

81/425

9/78

11/54

118/1054

0/21

20/164

5/22

1/227

1/101

2/33

12/30

Events,

41/116

20/54

11/63

2/73

0/45

4/21

0/34

17/80

33/100

4/59

0/50

0/24

1.00772 1 129

21

Figure 4. Regression of Erythroid Stimulating Agent on Log Risk Ratio of Red Blood Cell

Transfusion. Slope of regression line = 0.60 (95% confidence interval [CI] 0.09-1.11, p=0.02)

22

Figure 5. Forrest Plot of Risk of Infection with IV Iron


.

.


Steensma, DP

Henry, DH

Bastit, L

Kulnigg, S

Grote, L

Serrano−Trenas, JA

Hedenus, M

Schroder, MD

Friel, JKEvstatiev, R

Van Wyck, DB

Pollack, A

IV Iron Compared With No Iron

Coyne, DW

Khalafallah, A

Breymann, C

Allen, RP


Singh, K

Okonko, DO

Singh, H

Van Wyck, DB

name

Meyer, MP

Anker, SD

Qunibi, WY

IV Iron Compared With Oral Iron

Lindgren, S

Bencaiova, G

Krayenbuehl, PA


Schindler, E

2011

2007

2008

2008

2009

2011

2007

2005

19952013

2007

2001

2007

2010

2008

2011

1998

2008

2006

2005

year

1996

2009

2011

2009

2009

2011

1994

1.34 (1.10, 1.64)

2.53 (1.30, 4.91)

0.77 (0.38, 1.56)

1.10 (0.75, 1.64)

1.38 (0.58, 3.31)

3.20 (0.14, 75.55)

1.23 (0.63, 2.42)

3.09 (1.40, 6.81)

5.43 (0.28, 107.33)

1.10 (0.59, 2.04)1.41 (0.60, 3.31)

1.12 (0.65, 1.91)

0.95 (0.21, 4.32)

0.78 (0.33, 1.85)

(Excluded)

2.45 (0.85, 7.03)

1.31 (0.24, 7.12)

1.17 (0.94, 1.47)

(Excluded)

0.16 (0.01, 3.64)

0.61 (0.26, 1.43)

5.19 (0.25, 106.38)

RR (95% CI)

1.00 (0.16, 6.45)

1.06 (0.68, 1.66)

1.75 (0.80, 3.82)

3.07 (1.07, 8.80)

9.00 (1.16, 70.03)

1.82 (0.72, 4.59)

1.63 (1.16, 2.29)

(Excluded)

326/2413

28/164

9/63

43/203

18/137

1/29

16/100

18/33

2/22

9/1412/105

24/174

2/10

8/68

0/92

19/227

3/24

179/1022

0/50

0/24

9/75

2/79

Treatment

2/21

50/304

20/147

12/45

9/130

10/43

Events,

147/1391

0/30

208/2056

11/163

23/124

37/193

6/63

0/31

13/100

6/34

0/24

7/128/99

22/178

4/19

10/66

0/91

4/117

2/21

123/815

0/50

1/11

9/46

0/82

Control

2/21

24/155

8/103

4/46

1/130

6/47

Events,

85/1241

0/30

100.00

6.41

5.83

11.60

4.24

0.40

6.22

4.96

0.45

7.054.42

8.35

1.64

4.31

0.00

3.11

1.34

59.32

0.00

0.41

4.45

0.44

Weight

1.11

10.28

5.06

3.12

0.93

3.88

%

40.68

0.00

1.34 (1.10, 1.64)

2.53 (1.30, 4.91)

0.77 (0.38, 1.56)

1.10 (0.75, 1.64)

1.38 (0.58, 3.31)

3.20 (0.14, 75.55)

1.23 (0.63, 2.42)

3.09 (1.40, 6.81)

5.43 (0.28, 107.33)

1.10 (0.59, 2.04)1.41 (0.60, 3.31)

1.12 (0.65, 1.91)

0.95 (0.21, 4.32)

0.78 (0.33, 1.85)

(Excluded)

2.45 (0.85, 7.03)

1.31 (0.24, 7.12)

1.17 (0.94, 1.47)

(Excluded)

0.16 (0.01, 3.64)

0.61 (0.26, 1.43)

5.19 (0.25, 106.38)

RR (95% CI)

1.00 (0.16, 6.45)

1.06 (0.68, 1.66)

1.75 (0.80, 3.82)

3.07 (1.07, 8.80)

9.00 (1.16, 70.03)

1.82 (0.72, 4.59)

1.63 (1.16, 2.29)

(Excluded)

326/2413

28/164

9/63

43/203

18/137

1/29

16/100

18/33

2/22

9/1412/105

24/174

2/10

8/68

0/92

19/227

3/24

179/1022

0/50

0/24

9/75

2/79

Treatment

2/21

50/304

20/147

12/45

9/130

10/43

Events,

147/1391

0/30

1.00702 1 142

23

Figure 6. Funnel Plot for the Odds Ratio of Transfusion Against the Standard Error of the Log

Odds Ratio

0.5

11.

52

Stan

dard

Erro

r Log

Odd

s Ra

tio

−4 −2 0 2 4Log Odds Ratio

Funnel plot with pseudo 95% confidence limits

24

2.7 Tables

Table 1. Description of Included Studies

Name Year N Category Iron Type

Dose Schedule

Control ESA Outcomes Measures Follow Up (weeks)

Hb RBC Transfusion

Infection

Mortality

Other

1 Adhikary, L

2011

90 Renal Sucrose 200mg alternate days for 5 doses

Oral ferrous fumarate 125mg three times daily

Yes, both groups

Y Anaphylaxis

4

2 Agarwal, R

2006

75 Renal Gluconate

250mg weekly, total dose 10mg/kg

Oral iron No Y AE SAE Anaphylaxis Mortality

10

3 Aggarwal, HK

2003

40 Renal Dextran 100mg every 2 weeks

Oral iron Yes, both groups

Y SAE Anaphylaxis

12

4 Al, RA 2005

90 Obstetrics Sucrose Ganzoni formula. Max 400mg/day over 5 days. Total dose 11mg/kg

Oral iron polymaltose

No Y Y SAE Anaphylaxis Iron studies Fetal birth weight

10

5 Al-Momen, AK

1996

111

Obstetrics Sucrose Calculated from body weight

Oral ferrous sulphate

No Y AE Anaphylaxis

15

6 Allen, RP

2011

46 Restless leg syndrome

Carboxymaltose

500mg twice, 2 days apart

Placebo No Y IRLS restless leg score

4

7 Anker, SD

2009

459

Heart failure

Carboxymaltose

Ganzoni Formula 200mg weekly till replete

Placebo No Y Y Y Self-reported patient global assessment SAE Anaphylaxis

26

8 Auerbach, M

2004

157

Oncology Dextran Total dose infusion or 100mg repeated

Oral iron group and no iron group

Yes, both

Y Y Y AE Anaphylaxis

6

9 Auerbach, M

2010

243

Oncology Dextran 50mg per dose, three weekly for 12 weeks

Oral iron or no iron

Yes, both groups

Y Y Y Functional assessment AE SAE Anaphylaxis

15

10 Bastit, L*

2008

396

Oncology Sucrose 200mg every 3 weeks

Standard care, no iron

Yes, Both

Y Y Y Y QOL SAE Anaphylaxis

16

25

11 Bayoumeu, F

2002

50 Obstetrics Sucrose Lorentz Formula

Oral iron sulphate

No Y Y SAE Anaphylaxis

4

12 Beck-Da-Silva, L

2013

23 Cardiac failure

Sucrose 200mg weekly for 5 weeks

Oral ferrous sulphate 200mg three times per day

No Y VO2 max

13

13 Benacaiova, G

2009

260

Obstetrics Sucrose 200mg twice or three times


No Y Y Y Iron studies SAE

20-25

14 Bhandal, N

2006

44 Obstetrics Sucrose 200mg two days apart



5

15 Birgregard, G

2010

120

Healthy blood donors

Sucrose 200mg after each donation


No Iron studies AE SAE Anaphylaxis RLS

8-52

16 Breymann, C

2008

349

Obstetrics Carboxymaltose

15mg/kg to max of 1000mg weekly up to 3 doses mean total dose 1347mg


No Y Y Y AE SAE Anaphylaxis

12

17 Charytan , C

2005

96 Renal Sucrose 200mg weekly for 5 doses


No Y Y SAE Anaphylaxis Iron studies

5

18 Coyne, DW

2007

134

Renal Gluconate

125mg 3 times a week for 8 doses

No iron Yes, Both

Y Y Y SAE Iron studies

6

19 Dangsuwan, P

2010

44 Oncology Sucrose 200mg Oral ferrous sulphate


2

20 Edwards, TJ

2009

62 Colorectal surgery

Sucrose 300mg twice (BMI)

Placebo No Y Y LOS hospital

2

21 Evstatiev, R

2013

256

Gastrointestinal

Ferric carboxymaltose

500mg per dose, median dose 1000mg

Placebo No Y Y Y SAE 34

22 Friel, JK

1995

26 Neonatal TPN

Dextran 200mcg/kg/ day

No iron No Y Y 8

23 Froessler, B

2013

271

Obstetric Sucrose 200mg twice, minimum 24 hours apart

Oral ferrous sulphate 500mg daily

No Y Y AE 7

24 Garrido-Martin, P

2012

210

Cardiothoracic Surgery

Sucrose Three doses of 100mg each

Oral ferrous fumarate 105mg or placebo

No Y Y SAE 4

25 Grote, L*

2009

60 Restless Leg Syndrome

Sucrose 200mg, 5 doses over 3 weeks

Placebo No Y Y RLS score AE Anaphylaxis

11

26

26 Hedenus, M*

2007

67 Haematological Malignancy

Sucrose 100mg weekly for 6 weeks then fortnightly for 8 weeks

No iron Yes, Both

Y Y Y Y AE Epoetin dose

16

27 Henry, DH

2007

187

Oncology Gluconate

125mg weekly for 8 weeks


All Y Y Y Y AE SAE

12

28 Hulin, S

2005

93 Surgery - Cardiac

Sucrose 5mg/kg No iron No Y 0.7

29 Karkouti, K

2006

38 Surgery – Cardiac and orthopaedic

Sucrose 200mg daily for 3 days

Placebo Yes but ESA group excluded from analysis

Y Y 6

30 Kasper, SM

1998

128

Surgery – autologous blood

Gluconate

102.5mg Oral iron or placebo

No Y AE 3

31 Khalafallah, A*

2010

200

Obstetrics Polymaltose

Formula Oral ferrous sulphate

No Y Y SAE 24

32 Kim, YH

2009

76 Gynecology

Sucrose 200mg alternate days until replete by formula

Oral iron No Y SAE 3

33 Kim, YT

2007

75 Oncology Sucrose 200mg per dose

No iron No Y SAE Anaphylaxis

6

34 Kochhar, PK

2012

100

Obstetrics Sucrose 200mg per dose

No iron No Y SAE Anaphylaxis

6

35 Krayenbuehl, PA*

2011

90 Fatigue Sucrose 800mg over 2 weeks. Total dose 13mg/kg

Placebo No Y Y SAE AE Anaphylaxis

12

36 Kulnigg, S*

2008

200

Gastroenterology

Carboxymaltose

Ganzoni formula max per dose 1000mg or 15mg/kg


No Y Y Y AE SAE Anaphylaxis QOL

12

37 Li, H (1)

2008

46 Renal Sucrose 200mg weekly for 4 weeks then fortnightly

Oral ferrous succinate

Yes, Both

Y Y AE Anaphylaxis

8

38 Li, H (2)

2008

136

Renal Sucrose 100mg twice weekly for 8 weeks


Yes, Both

Y AE SAE

12

39 Li, H (3)

2009

194

Renal Sucrose 200mg weekly for 4 weeks then either 100 or 200mg weekly for 8 weeks


Yes, Both

Y Iron studies ESA requirement AE Cost-effectiveness

12

40 Lindgren, S*

2009

91 Gastroenterology

Sucrose Ganzoni formula, 200mg weekly till replete. Mean dose 1708 mg


No Y Iron studies AE

20

41 Maccio, A

2010

148

Oncology Gluconate

125mg weekly

Oral iron lactoferrin

Yes, Both

Y SAE 12

42 Macdoug

1996

37 Renal Dextran 250mg every 2


All Y Anaphylaxis

16

27

all, IC weeks Cost 43 Madi-Jebara, SN

2004

120

Cardiac Surgery

Sucrose Target formula 200mg/day

Placebo Yes but ESA group excluded from analysis

Y Y SAE Anaphylaxis

4

44 McMahon, LP

2010

100

Renal Sucrose 100-200mg every 2 months


No Y Y Y 52

45 Meyer, MP

1996

42 Neonates Sucrose 6mg/kg/week

Oral ferrous lactate

Yes, Both

Y Y Epoetin dose

4

46 Na, HS 2011

113

Orthopaedics

Sucrose 200mg, one to three doses total.

No iron or epoetin

Yes, Iron group only

Y Y 6

47 Neeru, S

2012

100

Obstetrics

Sucrose Formula based on target Hb 110g/l

Oral ferrous fumarate 300mg

No Y 4

48 Okonko, DO

2008

35 Cardiac Failure

Sucrose Formula No iron No Y Y Y VO2 peak SAE

18

49 Onken, JE

2013

507

Haematology

Ferric carboxymaltose

Two doses of 15mg/kg 1 week apart

Oral iron, ferrous sulphate 325mg three times daily

No Y Y SAE

6

50 Olijhoek, G

2001

110

Surgery - preoperative

Saccharate

200mg Oral iron Yes, but excluded from metaanalysis

Y SAE anaphylaxis

2

51 Pedrazzoli, P

2008

149

Oncology Gluconate

125mg weekly for 6 doses

No iron Both Y Y Y SAE Anaphylaxis

16

52 Pollack, A

2001

29 Neonates Sucrose 2mg/kg Oral, oral iron + epoetin, or IV iron + epoetin

Yes Y Y Y Y AE 3

53 Provenzano, R

2009

230

Renal Ferumoxytol

510mg twice within 1 week

Oral iron Both Y Y SAE 5

54 Qunibi, WY

2011

255

Renal Carboxymaltose

Maximum total 2000mg over 3 doses

Oral iron Both Y Y AE 8

55 Schaller, G

2005

38 Renal Sucrose 300mg Placebo Both Anaphylaxis

0.1

56 Schindler, E

1994

60 Orthopaedic

Gluconate

0.75mg/kg, 3 doses two weeks apart

Oral iron No Y Y SAE Anaphylaxis

6

57 Schroder, MD

2005

46 Gastrointestinal

Sucrose 7mg/kg initially, then 200mg 1-2 times weekly

Oral iron sulphate

No Y SAE Anaphylaxis

6

28

58 Seid, MH

2008

291


Formula – modified Ganzoni, weekly iron


6

59 Serrano-Trenas, JA*

2011

200

Orthopaedics

Sucrose 200mg, 3 doses 2 days apart

No iron No Y Y Y 4

60 Shafi, D

2012

200

Obstetrics

Sucrose Formula based on weight and target Hb

Oral ferrous ascorbate 100mg per day


6

61 Singh, H

2006

121

Renal Sucrose 1000mg total in divided doses

No iron Yes, both groups

Y Y Anaphylaxis

10

62 Singh, K

1998

100

Obstetrics Polymaltose

Formula, total dose

Oral iron fumarate

No Y Y Anaphylaxis

12

63 Sloand, JA

2004

25 Renal and Restless legs syndrome

Dextran 1000mg No iron No Y 4

64 Spinowitz, BS

2008

304

Renal Ferrumoxytol

510mg, two doses

Oral iron Both Y SAE 5

65 Steensma, DP*

2011

502

Oncology Gluconate

187.5mg every 3 weeks


Yes, both

Y Y Y Y SAE 16

66 Stoves, J

2001

45 Renal Sucrose 300mg monthly


Yes, both

Y Y 20

67 Toblli, JE

2007

40 Cardiac (CCF)

Sucrose 200mg weekly

Placebo No Y 26

68 Van Iperen, CE

2000

36 Intensive Care

Iron Sacharate

20mg per dose

No iron, IV iron, or IV iron + epoetin

Yes but ESA excluded from analysis

Y Mortality

3

69 Van Wyck, DB (1)

2007

361


Formula, mean cumulative dose 1403mg


No Y Y Y SAE 6

70 Van Wyck, DB (2)

2009

477

Gynecology

Carboxymaltose

Formula, mean dose 1568mg


No Y Y SAE QoL

6

71 Van Wyck, DB (3)

2005

188

Renal Sucrose 1000mg total, divided doses over 14 days


Yes, Both groups

Y Y Y AE 8

72 Verma, S

2011

150

Obstetrics Sucrose 600mg Oral ferrous sulphate

No Y Anaphylaxis AE

4

73 Warady, BA

2004

35 Renal Dextran Oral iron Yes Y SAE 16

74 Weisbach, V

1999

123

Surgery – preoperative Orthopaedics & Cardiothoracic

Sucrose 200mg Oral iron fumarate

No Y Y 5

75 Westad, S

2008

129

Obstetrics Sucrose 600mg total in 3 divided doses


No Y Y QoL 12

29

Table 2 Risk of Bias of Included Studies Name Randomisatio

n Sequence Generation

Allocation Concealment

Blinding of Participants or Personnel

Blinding of Outcome Assessment

Incomplete Outcome Data

Selective Reporting

Other Bias

1 Adhikary, L

Unclear Reported only as assigned to two groups

Unclear No method of allocation concealment

High Open-label study

Unclear No reporting of blinding of outcome assessment

Unclear 90 participants included in outcome analysis but number randomised not reported

Unclear Primary and secondary end-points not reported in methodology

Unclear Analysis not reported as intention-to-treat, differences in baseline Hb between groups

2 Agarwal, R Low Computer-generated randomisation schedule

Low Central randomisation



Unclear 89 participants randomised, 75 included in intention-to-treat analysis

Low Data provided on all prespecified outcomes

Low Intention to treat analysis, good baseline balance, administration of ESA prohibited

3 Aggarwal, HK

High Patients only described as divided into two groups

High Patients only described as divided into two groups



Low Follow up was complete

Low Data provided on all prespecified outcomes

Unclear All patients given stable dose of ESA, good baseline balance but whether analysis was intention to treat not reported

4 Al, RA Low Use of a computer-generated randomisation table

Low Use of consecutively numbered opaque envelopes



Low Follow up was complete

Low Data was provided on all prespecified outcomes

Low Intention to treat analysis, potential effect of prior use of oral iron explored

5 Al-Momen, AK

High Sequential selection

High Sequential selection



Unclear Participant numbers differed by 7 in the two groups despite sequential selection

Unclear List of prespecified outcomes not reported

Unclear Co-interventions and whether analysis was intention to treat not described.

6 Allen, RP Unclear Randomisation sequence generation not described

Unclear Allocation concealment not described

Low Study staff and participants blinded

Unclear Described as independently evaluated

Low Outcome data not available on only 3 participants randomised


Unclear Analysis was intention to treat but co-interventions not described.

7 Anker, SD Low Computer-generated permuted block randomisation


Low Study staff and participants blinded

Low Outcome assessment blinded

Low withdrawal of 37 participants from a total of 459


Low Analysis was intention to treat and co-interventions described

30

randomised

8 Auerbach, M

Unclear Randomisation sequence generation not described




Low Outcome data available on 155 of 157 participants randomised


High Enrolment ceased before target enrolment reached due to slow recruitment

9 Auerbach, M

Low Randomisation list created and maintained by an independent group

Low Allocation by central telephone system


Unclear Study blinded for ESA whilst ongoing but unblinded after all participants completed the study

Low Outcome data available for 238 out of 243 participants randomised

Low Prespecified outcome measures reported

Unclear Use of oral iron acceptable but not protocolised in non-IV iron group, no interaction detected between ESA and iron but power low

10 Bastit, L Low Randomisation sequence stratified by site

Low Allocation concealed using an interactive voice response system



Low Safety data analysed on all patients randomised


Low All participants received fixed dose ESA, intention to treat analysis provided

11 Bayoumeu, F

Low Randomisation table used






Low Analysis was intention to treat, co-interventions described

12 Beck-Da-Silva, L



Low Participants and study personnel blinded to allocation


High Primary outcome available for 18 out of 23 participants randomised


High Study terminated early with <30% of planned sample size recruited

13 Benacaiova, G

Low Computer-generated randomisation sequence

Low Consecutively numbered opaque envelopes



High Outcome data not available for 31 of 260 participants randomised


Low Analysis was intention to treat, co-interventions described

14 Bhandal, N


Low Consecutively numbered opaque envelopes



Low Outcome data available in 43 out of 44 participants randomised


Unclear Analysis was intention-to-treat but co-interventions were not described

15 Birgregard, G

Low Minimisation

Low Centralised

High Open-label

Unclear No reporting of

Low Outcome

Low Data was

Low Analysis was

31

method used

randomisation via web-based system

study blinding of outcome assessment

data available in 112 out of 120 participants randomised

provided on all prespecified outcomes

intention-to-treat

16 Breymann, C

Unclear Randomised in 2:1 ratio, stratified by country and severity of anaemia by method of sequence generation not reported






Unclear Analysis was intention-to-treat, co-interventions not reported

17 Charytan , C





High 83 participants out of a total of 102 randomised completed the study


Low Analysis was intention-to-treat, participants stratified according to previous ESA use

18 Coyne, DW

Low Computer-generated randomisation scheme






Low Analysis was intention-to-treat, ESA dose changes accounted for in study design

19 Dangsuwan, P

Low Random table used




Low Outcome data available on all 44 participants randomised


Low Analysis was intention-to-treat, all patients received RBC transfusion according to standardised protocol

20 Edwards, TJ

Low Computer-generated randomisation sequence used

Low Sealed, sequentially numbered opaque envelopes

Low Participants blinded, chief investigator and perioperative clinicians blinded, investigator administering infusion not blinded




Low Analysis was intention-to-treat, potential confounders collected, RBC transfusion in accordance with a strict protocol

21 Evstatiev, R

Low Randomised 1:1 according to predefined computer-generated list

Low Sequentially numbered envelopes used

Low Participants blinded


High Outcome data not available for 52 out of 256 participants randomised

Low Data provided on all prespecified outcome measures

Unclear Unclear whether full analysis participant set was intention-to-treat

32

22 Friel, JK Unclear

Randomisation sequence generation not described




Unclear Outcome data available for 26 participants, total number randomised not reported

Unclear Specific primary and secondary a priori-defined end-points not reported

Unclear Unclear whether analysis was intention to treat and co-interventions such as ESA use not reported

23 Froessler, B

Low 1:1 randomisation via telephone service

Low Telephone service used


Low Data were analysed by a statistician blinded to the treatment group

High No outcome data for 77 out of 271 participants randomised


Low Intention-to-treat analysis performed, transfused patients excluded from further Hb analysis


Low Random number list used

Low Assigned to intervention in pharmacy department

Low Blinding by placebo


High No outcome data for 51 out of 210 participants randomised

Unclear Discussion states no increase in infection but data not provided

Low Analysis was intention-to-treat

25 Grote, L Low Minimisation method used


Low Participants and study staff blinded


Low Outcome data available for all 60 participants randomised


Unclear Intention-to-treat analysis provided, co-interventions not described

26 Hedenus, M





Low Outcome data available for 60 of 67 participants randomised


Low Analysis included per-protocol and intention-to-treat, ESA dosing accounted for

27 Henry, DH





Low Safety population evaluated with 187 out of 189 participants randomised


Low Analysis included per-protocol and intention-to-treat, oral iron, ESA dosing and RBC transfusion accounted for in methodology

28 Hulin, S Unclear Randomisation sequence generation not described





Unclear Specific primary and secondary a priori-defined end-points not reported

Unclear Analysis not reported as intention-to-treat, co-interventions not described

29 Karkouti, Low Low Low Unclear Low Low Unclear

33

K Computer-generated randomisation sequence used

Sequentially numbered sealed, opaque envelopes

Participants and study staff blinded

No reporting of blinding of outcome assessment

Outcome data missing for 7 of 38 participants randomised

Data was provided on all prespecified outcomes

Analysis was intention-to-treat, transfusion guidelines provided but co-interventions not described

30 Kasper, SM

Unclear Randomisation sequence described as blocks of 5







31 Khalafallah

Low Randomised in blocks of 10

Low Assignment performed by pharmacy department





Unclear Analysis was intention-to-treat but co-interventions not described

32 Kim, YH Low Computer-generated randomisation sequence




Low Outcome data missing for 20 out of 76 participants enrolled but safety data analysed on all participants

Unclear No secondary outcomes measures reported as pre-specified


33 Kim, YT Unclear Randomisation sequence generation not described

Low Envelope procedure described



Unclear Number of participants randomised not reported

Unclear A-priori endpoints not reported

Unclear Protocol for RBC transfusion provided, analysis not reported as intention-to-treat

34 Kochhar, PK

Low Randomisation table used







35 Krayenbuehl, PA

Low Randomisation schedule generated by external provider

Low Randomisation schedule generated by external provider

Low participants and study staff blinded

Low Investigators blinded to study group

Low Outcome data available on all 90 participants randomised


Low Intention-to-treat analysis conducted

36 Kulnigg, S

Low Randomisatio

Low Central

High Open-label

Unclear No reporting of

Low Outcome

Low Data was

Low Intention-to-

34

n schedule generated by external provider

randomisation system

study blinding of outcome assessment

data available for 196 out of 200 participants randomised

provided on all prespecified outcomes, no post-hoc analyses performed

treat analysis conducted, co-interventions described

37 Li, H (1) Low Computer-generated randomisation sequence





Unclear Clearly defined a-priori endpoints not reported

Uncertain Titration of ESA allowed but not reported as an outcome, analysis not reported as intention-to-treat

38 Li, H (2) Low Computer-generated random number list used





Unclear Clearly defined a-priori secondary end-points not reported


39 Li, H (3) Low Computer-generated randomisation sequence





Unclear Clearly defined a-priori secondary end-points not reported


40 Lindgren, S

Low Minimisation method used

Low Internet used for allocation to treatment arm


Unclear final assessment done from computerized information only

Unclear 13 participants out of total of 91 randomised were withdrawn


Low Intention-to-treat analysis performed, co-intervention data collected

41 Maccio, A

Unclear Sequence generation not described

Unclear Randomisation 1:1 but allocation concealment not described





Low Study design controlled for co-interventions

42 Macdougall, IC






Unclear Principle end-points provided but without specifics, e.g. ‘iron status’

Uncertain Analysis not reported as intention-to-treat

43 Madi-Jebara, SN

Unclear Randomisation sequence generation not



Unclear No reporting of blinding of outcome

Low Outcome data available

Low Data was provided on all

Low Participants receiving RBC transfusion

35

described

assessment for all 120 study participants

prespecified outcomes

excluded from further analysis

44 McMahon, LP

Low Block randomisation




High Outcome data available for 85 out of 100 participants enrolled


Unclear 6 oral iron group patients received infrequent IV iron, analysis not reported as intention-to-treat

45 Meyer, MP






Uncertain Specific, a priori end-points not reported

Unclear Analysis not described as intention-to-treat

46 Na, HS Unclear Randomisation sequence generation not described

Low Sealed envelopes used





Low RBC transfusion guideline used

47 Neeru, S

Unclear Block randomisation but no further description of methods

Unclear No description of allocation concealment


Unclear No blinding reported of outcome assessment


Unclear RBC transfusion reported only for one group, unclear whether primary outcome prespecified

Unclear 6 participants crossed over from oral to IV iron

48 Okonko, DO

Low Computer-generated randomisation in a 2:1 ratio

Low Treatment allocation concealed from the investigators

Low Study investigators blinded

Low outcome assessment by blinded investigators

Unclear Outcome data available for 30 out of 35 participants


Low Missing data imputed but sensitivity analysis conducted without imputation

49 Olijhoek, G

Low randomisation schedule used, 1:1:1:1 ratio


Low Administration of ESA blinded

Unclear No outcome assessment blinding reported




50 Onken, JE

Low Randomised 1:1 using interactive voice system

Low Use of an interactive voice system

High Open-label

Low Composite safety events adjudicated by a blinded clinical committee

Low Outcome data available in 495 out of 507 participants



36

51 Pedrazzoli, P





High 33 participants out of a total of 149 randomised were exclude from per protocol population


Unclear Study stopped early with 149 out of 420 planned participants recruited

52 Pollack, A


Low Sequentially numbered sealed envelopes



High No outcome data for 9 out of 38 participants enrolled

Unclear Specific a priori-defined primary and secondary end-points not reported

Unclear Characteristics of participants disqualified did not differ from those who completed study but data not provided

53 Provenzano, P

Low Telephone system

Low Telephone system

Low Open-label study




Low Intention-to-treat analysis conducted, safety population included all patients receiving at least one dose of study medication

54 Qunibi, WY

Low Interactive voice-response system used

Low Centralised system





Unclear Initial 2:1 randomisation ratio, changed to 1:1 due to slow recruitment

55 Schaller, G


Unclear Allocation concealment not reported


Low Laboratory personnel blinded to group assignment



Unclear Potential differential use of ESA

56 Schindler, E




Unclear No blinding of outcome assessment reported



Unclear Analysis not described as intention-to-treat

57 Schroder, MD

Low Computer-generated random number table






Unclear Differential distribution of inflammatory bowel disease type

58 Seid, MH Low Unclear High Unclear Low Low Low

37

1:1 randomisation, stratified by baseline Hb

Allocation concealment not described

Open-label study

No blinding of outcome assessment reported

All 291 participants randomised included in the intention-to-treat analysis

Data was provided on all prespecified outcomes

Analysis was intention-to-treat

59 Serrano-Trenas, JA

Low Randomisation list used with 1:1 ratio

Low Sealed, opaque envelopes

Low outcome data assessor blinded


Unclear Outcome data available for 179 out of 200 participants randomised


Low Analysis was intention-to-treat, protocol provided for RBC transfusion

60 Shafi, D

Low Computer generated randomisation sequence

Low Sequentially numbered sealed opaque envelopes



Low Outcome data available on all enrolled participants


Unclear All participants analyses in group to which randomised but RBC transfusion not described

61 Singh, H Unclear Randomisation 2:1 but sequence generation not described







62 Singh, K Unclear Randomisation sequence generation not described




Unclear 100 participants randomised but number with outcome data not reported

Unclear Primary outcome specified but not specifics of secondary outcome measures

Unclear 12 participants in oral iron group switched to IV iron

63 Sloand, JA



Low participants and study staff blinded




Unclear Plan to recruit 30 participants but study stopped after 25 recruited

64 Spinowitz, BS







Low Intention-to-treat analysis performed

65 Steensma, DP

Low 1:1:1 stratified randomisation

Low Central allocation concealment

Low Patients and investigators blinded to oral or no iron


Low Outcome data available for 490 out of 502


Unclear Stopped early due to excess serious adverse events in IV

38

participants randomised

iron arm

66 Stoves, J Low Computer-generated randomisation schedule

Low Computer-based allocation





Unclear Analysis not reported as intention-to-treat

67 Toblli, JE Low Random number table used



Low Physicians performing echocardiography blinded



Low Baseline balance and co-interventions described

68 Van Iperen, CE





Low Outcome data available for all 36 participants enrolled


Low Co-interventions described, intention-to-treat analysis performed

69 Van Wyck, DB (1)

Low Computerised random number generation

Low Interactive voice-response system



Low Outcome data for safety evaluation available for 352 out of 361 participants randomised



70 Van Wyck, DB (2)





Low Outcome data for safety evaluation available for 456 out of 477 participants randomised



71 Van Wyck, DB (3)








72 Verma, S Unclear Randomisation sequence generation not described




Unclear 150 participants included in outcome analysis but number randomised not reported

Unclear No prespecified outcome parameters other than Hb


73 Warady, BA

Low Random

Unclear Allocation

High Open-label

Unclear No blinding of

Low Outcome

Low Data was

Unclear 1 patient in

39

number table used

concealment not described

study outcome assessment reported

data available for all 35 participants randomised

provided on all prespecified outcomes

oral iron group received IV iron, analysis not reported as intention-to-treat

74 Weisbach, V

Unclear Randomisation list used

High Chronological enrollment with sequential order of trial medication




Unclear Secondary end-points not specifically reported

Unclear Analysis not reported as intention-to-treat

75 Westad, S

Low Minimisation used







Table 3. Results Name Baseline

Hb (g/L) Baseline Ferritin (mcg/L)

Baseline TSAT (%)

Hb change mean (g/l) IV iron versus comparator

Hb change (% achieved target)

Transfusion (n) IV iron versus comparator

Infection (n) Mortality (n)

1 Adhikary, L 92

98 16 10.6 vs 3.2 24.4 vs 20.0

2 Agarwal, R 106 69 18 2 vs 4

3 Aggarwal, HK

60 186 61 43 vs 27

4 Al, RA 98 45 12 vs 6 95.6 vs 62.2

5 Al-Momen, AK

76 12 52 vs 35

6 Allen, RP 32 -8.9 vs -4.0

3/24 vs 2/21

7 Anker, SD 119 54 17 11 vs 6 50 vs 28 50/304 vs 24/155

8 Auerbach, M

96 258 18 24 vs 12 68 vs 25 1/78 vs 1/791

9 Auerbach, M

93 312 19 vs 13 41/116 vs 48/122

8/117 vs 13/121

10 Bastit, L 99 279 29 86 vs 73 18/200 vs 39/196

21/200 vs 15/196

11 Bayoumeu, F

96 7 10 11.1 vs 11.0 12.5 vs 17.4

0/24 vs 1/23

12 Beck-Da-Silva, L

112 132 17 10.4 vs 16.9 20 vs 43 2/10 vs 0/7

13 Benacaiova, G

108 12.2 vs 12.4 80 vs 75 1/110 vs 1/119 9/130 vs 1/130

14 Bhandal, N

74 12 11.5 vs 11.2 0/22 vs 1/21

15 Birgregard, G

136 35

16 Breymann, C

96 39 12 33.7 vs 32.9 1/227 vs 0/117 19/227 vs 4/117

17 Charytan 98 114 16 10 vs 7 54.2 vs

40

, C 31.3 18 Coyne, DW

103 761 19 16 vs 11 46.9 vs 29.2

8/68 vs 10/66 1/68 vs 1/66

19 Dangsuwan, P

90 9 vs 4 5/22 vs 14/22

20 Edwards, TJ

135 -1.9 vs -5 2/34 vs 5/26

21 Evstatiev, R

136 75 23 27.2 vs 40.4

12/105 vs 8/99 0/105 vs 0/99

22 Friel, JK 9/14 vs 7/12

23 Froessler, B

102 9 25 vs 24 1/101 3/97 0/100 vs 0/94


139 252 -13 vs -13 20/54 vs 27/53

25 Grote, L 130 20 5.5 vs 1.5 1/29 vs 0/31

26 Hedenus, M

103 128 22 28 vs 16 93 vs 53 2/33 vs 1/34 18/33 vs 6/34 0/33 vs 4/34

27 Henry, DH

103 351 32 24 vs 16 18 vs 16 11/63 vs 20/124 9/41 vs 23/88

28 Hulin, S 107 67

29 Karkouti, K

84 220 9.5 vs 7.0 16.5 vs 12.0

2/11 vs 4/10

30 Kasper, SM

145 162 25 11.8 vs 11.8

31 Khalafallah

108 18 14 19.2 vs 12.5 84 vs 71 0/92 vs 0/92

32 Kim, YH 76 76.7 vs 11.5

33 Kim, YT 113

12/30 vs 29/45

34 Kochhar, PK

76 17 57 vs 36 0/50 vs 1/50

35 Krayenbuehl, PA

133 22 23 1 vs 0 10/43 vs 6/47

36 Kulnigg, S

88 6 56 37 vs 28 76.5 vs 68.3

18/137 vs 6/63

37 Li, H (1) 88 111 19 34.1 vs 22.1 0/26 vs 0/20

38 Li, H (2) 81 193 22 39.2 vs 18.6 88.6 vs 44.0

39 Li, H (3) 88 151 10 24.7 vs 17.8

40 Lindgren, S

104 13 7 66 vs 47 12/45 vs 3/46

41 Maccio, A 98 493 26 16 vs 18 50 vs 56

42 Macdougall, IC

73 25 45 vs 27

43 Madi-Jebara, SN

142 17/80 vs 9/40

44 McMahon, LP

119 106 21 2 vs 1 5/52 vs 1/48

45 Meyer, MP

0/21 vs 2/21 2/21 vs 2/21

46 Na, HS 121 76 21 11/54 vs 29/54

47 Neeru, S

95 20 vs 12 66.7 vs 61.4

48 Okonko, DO

123 70 21 5 vs 4 0/24 vs 1/11 1/24 vs 0/11

49 Onken, JE

106 33 22 15.7 vs 8.0 57.0 vs 29.1

0/244 vs 2/251

50 Olijhoek, G

123 2 vs -1 no ESA 15 vs 16 ESA

41

51 Pedrazzoli, P

99 341 29 76.7 vs 61.8

2/73 vs 5/76 4/73 vs 3/76

52 Pollack, A -8 v -24

2/10 vs 4/19

53 Provenzano, P

106 350 16 10 vs 5 49 vs 25 1/110 vs 3/114

54 Qunibi, WY

100 108 15 10.5 vs 7.0 60.4 vs 34.7

20/147 vs 8/103

55 Schaller, G

122 171 28

56 Schindler, E

135 105

57 Schroder, MD

97 100 6 25 vs 21 2/22 vs 0/24

58 Seid, MH 89 24 40 vs 34 91.4 vs 66.7

33/100 vs 41/100

16/100 vs 13/100

11/100 vs 10/100

59 Serrano-Trenas, JA

120

60 Shafi, D

79 8 29 vs 20

61 Singh, H 106 175 18 13 vs 6

9/75 vs 9/46

62 Singh, K 84 8 29 vs 13

0/50 vs 0/50

63 Sloand, JA

111 21 2 vs -4

64 Spinowitz, BS

100 145 11 8.2 vs 1.6

65 Steensma, DP

99 465 22 26 vs 24 70 vs 66 20/164 vs 43/326

26/164 vs 10/163

8/164 vs 9/326

66 Stoves, J 98 70 vs 59 0/22 vs 1/23 67 Toblli, JE 102 72 20 15 vs -4

68 Van Iperen, CE

100 13 vs 10 5/12 vs 12/12 2/12 vs 4/12

69 Van Wyck, DB (1)

90 25 10 96.4 vs 94.1

24/174 vs 22/178

1/174 vs 0/178

70 Van Wyck, DB (2)

94 68 6 82.0 vs 61.8

71 Van Wyck, DB (3)

101 98 16 7 vs 4 44.3 vs 28.0

2/79 vs 0/82

72 Verma, S 75 39 vs 27

73 Warady, BA

115 139 38 -1.5 vs -1.7

74 Weisbach, V

144 163 25 5 vs 6 6/30 vs 5/60

75 Westad, S

78 40 vs 46

4/59 vs 11/70

42

Chapter 3

Iron-Restricted Erythropoiesis and Risk of Red Blood Cell

Transfusion in the Intensive Care Unit: A Prospective

Observational Study

43

3.1 Introduction

Anaemia is nearly universal in ICU patients and is the most common indication for RBC

transfusion, despite high concordance with restrictive guidelines 1 2. Both anaemia and RBC

transfusion are associated with increased morbidity and mortality in critical illness 2 3 5.

Intravenous (IV) iron therapy increases haemoglobin and decrease transfusion requirement in

selected patients 116. As such, IV iron therapy may plausibly improve outcomes during critical

illness, however the optimal criteria for selecting patients who may benefit from this therapy in

ICU are uncertain.

IV iron therapy improves haemoglobin by overcoming iron-restricted erythropoiesis (IRE). This is

syndrome includes absolute iron deficiency, functional iron deficiency (the inability of iron stores

to meet elevated erythropoeitic demands) and iron sequestration (a reduction in the availability of

stored iron in the setting of inflammation) 32 110. Recent guidelines have promoted ferritin as an

essential assay in the diagnosis of IRE and decision to initiate IV iron therapy 117 118.

The hypothesis was that that IRE, diagnosed by iron studies on admission to ICU, would identify

a substantial group of patients at high risk for subsequent RBC transfusion, and hence, provide a

simple method to potentially determine likely response to IV iron therapy. The aim of this study

therefore, was to describe the characteristics of patients with IRE on admission to ICU and

determine the optimal variables to identify the group of patients at risk of subsequent RBC

transfusion.

44

3.2 Methods

The study was undertaken in the 23-bed combined medical/surgical ICU at Royal Perth Hospital,

a university-affiliated tertiary referral centre in Perth Western Australia. Data from consecutive

ICU admissions was recorded in a prespecified case report form. Patients were followed from

ICU admission till discharge from hospital, censored at 60 days post admission to ICU. The

diagnosis of IRE using a cutoff of ferritin <300mcg/L and transferrin saturation (TSAT)<20% was

based on previously published consensus statement and guidelines for the laboratory diagnosis

of functional iron deficiency 117-119. The study was approved by the clinical safety quality unit

(110623-1).

Statistical Analysis

The associations between baseline variables and IRE status were assessed using chi-square,

Student t-test and Mann-Whitney U tests for categorical, parametric and non-parametric

continuous variables, respectively. After excluding patients who received IV iron therapy,

univariate logistic analysis was used to assess the association between baseline factors and

odds of any subsequent in-hospital RBC transfusion. Baseline factors with a p value<0.25 in the

univariate logistic regression analyses were then entered in the initial multivariable logistic

regression model and eliminated in backward stepwise fashion with a significance level of

p<0.05. The utility of the final model of risk factors for predicting RBC transfusion was assessed

by receiver operator characteristic curve and the Hosmer-Lemeshow test.

A sensitivity analysis was then conducted investigating the utility of the predictive model for RBC

transfusion in patients eligible for IV iron therapy on the basis of iron study parameters from

previously published RCTs of IV iron. 116. Parameters varied between studies, however none

included patients with a ferritin >1200 mcg/L or TSAT>50%. The sensitivity analysis was

therefore limited to patients with a ferritin <1200mcg/L and TSAT≤50% to exclude patients in

whom iron over-saturation may limit the efficacy of IV iron.

Sample Size Calculation

Compared with a previous estimated prevalence of 35% for functional iron deficiency on

admission to ICU, a prevalence of 45% would require a sample size of 184 (p=0.05 and power

45

80%) 120. Assuming approximately 10% of patients without sufficient baseline results to be

included in the analysis, a sample size of 200 patients was planned for inclusion in the study.

46

3.3 Results

Between 5 March and 14 April 2012 there were 201 consecutive ICU admissions. Admission

blood results were not available in six cases, therefore, a total of 195 patients were included in

the analysis. The median age of the cohort was 53 (interquartile range (IQR) 36-65), 68% (95%

confidence interval (CI) 61-74) were male and 75%, (95%CI 69-81) received mechanical

ventilation. The participant characteristics are presented in table 1.

Characteristics of patients with iron-restricted erythropoiesis

The prevalence of iron-restricted erythropoiesis on admission to ICU, defined according to

ferritin<300mcg/L and TSAT<20%, was 26.2% (95% CI 19.9-32.4). Age, gender, acute

physiology and chronic health evaluation (APACHE) II score, sequential organ failure

assessment (SOFA) score, haemoglobin and C-reactive protein were similar between those with

and without IRE. Compared with patients without IRE, patients with IRE had significantly lower

mean corpuscular volume and mean corpuscular haemoglobin concentration (MCHC), mean

difference 2.5fL (95%CI 0.8-4.3, p<0.001) and 1.3pg (95% CI 0.6-2.0, p<0.001) respectively.

The proportion of patients with IRE subsequently receiving RBC transfusion was significantly

lower than the proportion of patient without IRE receiving RBC transfusion, as was the proportion

of patients with and without IRE discharged from hospital with a haemoglobin<100g/L, absolute

mean difference 18.9% (95%CI 4.7-33.1) and 36.6% (95%CI 20.9-52.2) respectively. Only four

patients received IV iron after admission to ICU, none fulfilled the criteria for IRE on admission to

ICU. The characteristics and outcomes of patients according to admission IRE status are

presented in table 2.

Risk factors for RBC transfusion

After excluding four patients who received IV iron therapy, 16 variables were assessed for

association with subsequent RBC transfusion on univariate analysis (see table 3). Age,

emergency ICU admission, chronic renal impairment, RBC transfusion prior to ICU admission ,

APACHE II score, SOFA score, renal replacement therapy, ICU length of stay, ICU admission

haemoglobin, c reactive protein, mean corpuscular volume, mean corpuscular haemoglobin and

47

IRE (ferritin<300 & TSAT<50%) all had p <0.25 and were therefore included in the subsequent

multivariable logistic regression for stepwise elimination.

In the final multivariable model presented in table 4, five variables were found to be significantly

associated with subsequent risk of RBC transfusion (previous RBC transfusion, SOFA score, ICU

length of stay, admission haemoglobin and c-reactive protein). IRE was not independently

associated with risk of subsequent RBC transfusion.

The receiver operator characteristic (ROC) area under the curve (AUC) for risk of transfusion

using the five variable multiple regression model was 0.93 (95%CI 0.89-0.97). Calibration, as

assessed by Hosmer-Lemeshow was 3.20 (p=0.92), see figure 1.

Predicting RBC transfusion in those who may benefit from IV iron

After excluding patients unlikely to benefit from IV iron (ferritin>1200mcg/L & TSAT>50%), the

ROC AUC for the five-variable model was 0.91 (95%CI 0.84-0.95), see table 4. Predictive

accuracy was also not substantially different when limited to a more clinically useful model

comprising only three simple variables categorized according to maximal ROC AUC (RBC

transfusion prior to ICU admission, Hb<100g/l on admission to ICU and ICU length of stay>3

days), ROC AUC 0.88 (95%CI 0.80-0.95) p=0.20, Hosmer-Lemeshow for three variable model

6.33, p=0.10, , see table 5. In this subgroup, 30 (22.2% (95% CI 15.1-29.3) and 31 (25.5% (95%

(CI17.9-33.8) patients respectively were admitted to ICU and discharged from hospital with a

haemoglobin<100g/L.

48

3.4 Discussion

In this study a diagnosis of IRE, based on a ferritin<300mcg/L and TSAT<20%, was found to be

moderately prevalent on admission to ICU (26.2% (95% CI 19.9-32.4). However IRE was not

independently associated with risk of subsequent RBC transfusion, despite those with IRE having

a significantly lower mean corpuscular volume and mean corpuscular haemoglobin. Rather than

identifying a group at high risk of RBC transfusion, IRE appeared to be protective on univariate

analysis, possibly due to the competing effect of increased ferritin as a marker of the severity of

acute inflammation and illness severity, itself likely to increase the incidence and severity of

anaemia.

The optimal criteria for determining response to IV iron in critically ill patients with IRE remain

unknown. A previous study in patients with chronic kidney impairment suggested that iron studies

are inadequate in guiding response to IV iron 121. Our findings concur, and suggest that a

diagnosis of IRE on the basis of iron studies alone is likely to have limited clinical utility in

determining response to IV iron in the critical care setting, particularly given the low incidence of

RBC transfusion and severe anaemia on hospital discharge in this group.

In contrast, this study found that a predictive model based on three simple clinical criteria (RBC

transfusion prior to ICU admission, haemoglobin<100g/L and ICU length of stay>3 days)

identified patients at high risk of subsequent in-hospital RBC transfusion and may have greater

clinical utility as a way of identifying patients who may benefit from IV iron therapy. Although

length of ICU stay is not known at baseline, previous RCTs in which trial eligibility requires the

judgment of the treating clinician that the patient is likely to require ICU level care beyond the

next calendar day demonstrate good predictive ability 122. This high-risk group was also at high

risk of discharge from hospital with severe anaemia, a condition that may then persist long after

discharge from hospital and be associated with increased morbidity and mortality 9 10.

Free iron associated with RBC transfusion may increase the risk of oxidative stress and infection

123 124. Whether IV iron therapy is associated with similar free iron release in critical illness is an

important additional consideration. A recent RCT of IV iron in trauma patients excluding patients

with a ferritin > 1000ng/ml or TSAT>50% however found no significant difference in RBC

49

transfusion requirement, infection, or mortality but generalisability may limited by the liberal

haemoglobin inclusion threshold, dose and type of IV iron 125 126.

Several limitations of this study require consideration. First, this was a single centre observational

study and the final predictive model was based on backwards stepwise elimination limited by

consideration only of statistical significance and with a high number of variables to participants.

As such, the generalisability of the model for predicting transfusion is uncertain. However, the

cohort of patients and incidence of RBC transfusion, were generally representative of Australian

tertiary ICU admissions. Second, only four patients received IV iron therapy in ICU. An

association between IV iron therapy and subsequent RBC transfusion in patients who fulfilled the

high risk criteria was therefore unable to be assessed and future RCTs are warranted to evaluate

the safety and efficacy of IV iron in critically ill patients126. Third, bleeding events were not

considered and may have a substantial effect on the outcome of interest for which the

relationship with the exposure of interest is less strong. However, previous studies suggest that

the predominant indication for RBC transfusion in the ICU is anaemia, not bleeding, potentially

mitigating the importance of this omission. Fourth, the competing risk of death was not assessed,

although mortality was low (7.7%) and would not be expected to have a major effect on the

predictive model. Finally, several potential markers of IRE including hepcidin, soluble transferrin

receptors and zinc protophoryn were also not assessed. The addition of these markers may have

helped differentiate absolute iron deficiency from functional iron deficiency although the role of

these markers in critical illness remains uncertain.

50

3.5 Conclusion

Despite moderate prevalence, IRE, diagnosed by iron studies on admission to ICU, is associated

with a low incidence of severe anaemia and subsequent RBC transfusion and not independently

associated with either on multivariable analysis. IRE is unlikely to be useful in determining the

response to IV iron. An alternative approach of early identification of patients at high risk of

subsequent in-hospital RBC transfusion using a simple predictive model and excluding iron over-

saturation may be a preferable strategy for identifying patients that may benefit from IV iron

therapy in ICU.

51

3.6 Figures Figure 1 ROC Probability of RBC Transfusion

N=177, ROC AUC 0.93 (0.89-0.97) Hosmer-Lemeshow 3.20, p=0.92 Figure 2. ROC Probability of RBC Transfusion after Excluding Patients Unlikely to Benefit from IV Iron (ferritin>1200mcg/L and/or TSAT>50%) Comparison of Full Model and Three Risk Factor Model

ROC AUC 0.91 (95%CI 0.84-0.97) versus 0.88 (95%CI 0.80-0.95) p=0.20 Hosmer-Lemeshow for three variable model 6.33, p=0.10, N=121

52

3.7 Tables Table 1 Participant Characteristics

n=195

Age

53 (36-65)

Male Gender, N (%)

132 (68)

Admission Type, N (%) Medical Elective Surgical Emergency Surgical

90 (46) 36 (18) 69 (35)

Apache II Score Admission

14 (9-18)

SOFA Score Admission 5 (3-7)

Mechanical Ventilation, N (%) 146 (75)

Renal Replacement Therapy, N (%)

22 (11)

Median (Interquartile range) unless otherwise stated APACHE Acute Physiology and Chronic Health Evaluation, SOFA Sequential Organ Failure Assessment,.

53

Table 2 Characteristics and Outcomes of Patients With and Without IRE on Admission to ICU

Iron-restricted erythropoiesis N=51

Not iron-restricted erythropoiesis N=144

P Value

Baseline Characteristics Age

44 (23-59) 54 (42-66) <0.01

Male Gender N (%)

31 (61) 101 (70) 0.22

APACHE II

14 (10-17) 14 (9-19) 0.39

SOFA

5 (3-7) 5 (3-7) 0.33

Haemoglobin g/l, mean (95% CI)

118 (112-124) 115 (111-118) 0.40

Mean Corpuscular Volume fL

89 (86-92) 91 (88-93) <0.01

Mean corpuscular haemoglobin pg, mean (95%CI)

29.7 (28.9-30.5) 31.0 (30.6-31.3) <0.001

Fe mcmol/L

5 (3-7) 13 (5-19) <0.01

Transferrin saturation, %

10 (7-14) 26 (13-50) <0.01

Ferritin mcg/L

144 (78-215) 448 (227-1165) 0.03

C Reactive Protein mg/L

65 (9-86) 63 (3-90) 0.88

Outcomes RBC transfusion N (%)

7 (14) 47 (33) 0.01

RBC units in those transfused, mean (95%CI)

3.4 (1.7-5.2) 5.2 (3.4-7.0) 0.46

ICU LOS

1 (1-4) 2 (1-4) 0.21

ICU Mortality N (%)

6 (12) 9 (6) 0.21

Haemoglobin change from ICU admission to hospital Discharge in non-transfused survivors g/l

2 (-3-6)

-7 (-10 - -4) <0.01

Discharged from hospital with Hb<100g/L, n (%)

3 (7) 54 (44) <0.001

Hospital LOS

6 (2-14) 12 (6-21) <0.01

Hospital Mortality

6 (12) 11 (8) 0.37

Median (Interquartile range) unless otherwise stated.

54

Table 3. Univariate Analysis of Association Between Baseline Variable and In-Hospital RBC Transfusion After Admission to ICU (n=191)

Odds ratio (95% CI) P Value

Age

1.03 (1.01-1.05)a 0.004

Gender

1.11 (0.56-2.21) 0.760

Emergency ICU admission

2.02 (0.79-5.20) 0.144

Chronic renal impairment+

3.88 (1.61-9.35) 0.003

Chronic acid suppression use#

1.12 (0.51-2.48) 0.771

RBC units transfused prior to ICU admission

21.32 (8.90-51.08) <0.001

APACHE II

1.07 per point (1.02-1.12)b 0.006

SOFA

1.22 (1.08-1.37)b 0.001

Mechanical ventilation

0.77 (0.38-1.57) 0.470

Renal replacement therapy

6.95 (2.62-18.44) <0.001

ICU LOS

1.27 (1.15-1.41)c <0.001

C reactive protein mg/L

1.01 (1.00-1.01)d 0.001

Hemoglobin g/L

0.93 (0.90-0.95)e <0.001

Mean corpuscular volume (fL)

0.97 (0.91-1.02)f 0.264

Mean corpuscular haemoglobin (pg)

0.92 (0.81-1.05)g 0.236

Iron-restricted erythropoiesis ^

0.34 (0.14-0.80) 0.014

+Chronic renal impairment defined according to creatinine>110 #Chronic acid suppression included all patients receiving proton pump inhibitor or H2 receptor blocker on admission to hospital ^Iron-restricted erythropoiesis defined according to ferritin<300 & transferrin saturation<20% References for odds ratios: a per year, b per point, c per day, d per mg/L, e per g/L, f per fL, g per pg

55

Table 4. Final Multivariable Model After Excluding Four Patients Who Received IV Iron Whilst in ICU, Then Stepwise Elimination of All Variables with p>0.05

Odds ratio (95% CI) Coefficient* P Value


11.11 (3.74-33.07) 2.408 <0.001

SOFA

1.23(1.03-1.48) 0.210 0.02

ICU LOS

1.25 (1.08-1.45) 0.225 0.003

C reactive protein mg/L

1.01 (1.00-1.01) 0.007 0.007

Hemoglobin g/L

0.95 (0.92-0.98) -0.157 <0.001

, *logistic regression coefficient constant 1.889 Table 5. Model for Predicting RBC Transfusion, Excluding Patients with Iron Over-Saturation and Comprising only Three Simple Variables Categorized According to Maximal ROC AUC (RBC transfusion prior to ICU admission, Hb<100g/l on admission to ICU and ICU length of stay>3 days),

Odds ratio (95% CI) Coefficient* P Value


7.44 (2.15-25.78) 2.01 0.002

ICU LOS>3 days

3.59 (1.17-11.02) 1.28 0.025

Hemoglobin<100g/L

9.50 (2.92-30.87) 2.25 <0.001

*logistic regression coefficient constant -3.13

56

Chapter 4

The IRONMAN Trial: A Protocol for a Multicentre Randomised

Blinded Trial of Intravenous Iron in Intensive Care Unit Patients

with Anaemia

57

4.1 Introduction

RBC transfusion is common in critically ill patients in Australia and worldwide, with 17-45% of all

patients admitted to an ICU, and more than 70% of those staying greater than 7 days receiving

one or more RBC units 1-3. Transfused critically ill patients receive a mean of 4 RBC units in ICU

and account for nearly 20% of all RBC transfusions in Australia 127.

In a recent systematic review of observational studies, conducted by Marik et al., RBC

transfusion in the critically ill was an independent predictor of death (pooled odds ratio 1.7, 95%

confidence interval 1.4-1.9), nosocomial infection, multi-organ dysfunction syndrome and acute

respiratory distress syndrome 5. Allogeneic RBC transfusion is also an increasingly costly and

scarce resource 128.

Transfusion in the ICU remains common despite extremely high concordance to current

restrictive transfusion guidelines 1. More than 75% of RBC units transfused in ICU are given for

anaemia, rather than major haemorrhage 1 2, and anaemia itself is also associated with adverse

outcomes 129. There is therefore an unmet need for novel interventions that reduce the incidence

of anaemia and therefore transfusion.

Iron-restricted erythropoiesis is extremely common in critically ill patients and may occur through

absolute iron deficiency, functional iron deficiency or iron sequestration 9. Administration of iron

enterally is ineffective in patients who are critically ill due to gastrointestinal intolerance,

decreased iron absorption from routine use of gastric acid suppression, physiological limits to

maximal enteral iron absorption and inhibition of absorption due to high hepcidin levels that occur

in critical illness 119. IV iron overcomes these disadvantages and has been shown to be superior

to enteral iron for the correction of iron-restricted erythropoiesis in a number of patient

populations 19 116. Furthermore, the diagnosis and management of iron deficiency and suboptimal

iron stores in the critically ill has been identified as an important evidence gap by the National

Blood Authority 130.

Iron is essential for bacterial growth, and exogenous iron therefore associated with a theoretical

increased risk of infection, however most RCTs to date have not included infection as a

prespecified endpoint and the risk in critically ill patients remains uncertain 22 116. In an animal

58

model of sepsis, IV iron improved Hb and was not associated with increased risk of death 27. A

RCT by Pieracci et al of low dose iron sucrose in trauma patients found no significant difference

between the groups in infection 125.

The Pieracci trial also found no significant difference in Hb concentration or RBC transfusion

requirement associated with IV iron 116. In comparison, the IRONMAN trial will enroll a broader

population of critically ill patients and has been designed to optimise IV iron efficacy by only

including more severely anaemic patients (included if Hb<100g/L versus Hb<120g/L) and

administering an alternative and higher dose IV iron (500mg ferric carboxymaltose versus 100mg

iron sucrose) associated with greater erythropoeitic response 131.

The hypothesis of this study is that IV iron supplementation in critically ill patients who are

anaemic but do not have severe sepsis is effective in reducing RBC transfusion. A reduction in

mean RBC transfusion requirement may lead to a reduction in mortality and major morbidity, as

well as substantial healthcare costs savings.

59

4.2 Study Design

The IRONMAN trial is a multicentre, phase IIb, randomised, placebo-controlled parallel group trial

comparing IV iron in addition to standard care, to standard care alone, in patients admitted to the

ICU who are anaemic. The primary end-point is the mean number of RBC transfusions from

study enrolment to discharge from hospital. Secondary endpoints included the proportion of

patients transfused, ICU and hospital mortality and infection. A full list of outcome measures is

provided in table 1.

The IRONMAN trial is planned to enroll 140 participants across four centres. Adult patients within

48 hours of admission to ICU (or a high dependency area under the supervision of an Intensivist),

predicted to remain in the ICU beyond the next calendar day, with a haemoglobin (Hb) of <100g/l

in the preceding 24 hours and without exclusion criteria are eligible for enrolment after

prospective consent. A complete list of inclusion and exclusion criteria is provided in table 2.

Participants

Patients admitted to the ICU will be assessed by trained study personnel including the research

coordinators and medical staff, at each study site. Patients will be eligible for enrolment if they

fulfill all of the inclusion criteria and none of the exclusion criteria (table 2). The key inclusion

criteria were based on the findings of the prospective observational study (Chapter 3) that found

that ICU LOS and Hb<100g/L were independent predictors of RBC transfusion. In this study,

ferritin and TSAT were not predictive of future RBC transfusion. Therefore participants were

excluded on the basis of risk of iron overload (ferritin >1200mcg/L and/or TSAT>50%) rather than

included on the basis of likelihood of response to IV iron at lower ferritin and TSAT levels.

Participants will be allocated to the treatment arm using a randomly generated sequence.

Randomisation will be in variable block size and stratified by site. Allocation concealment will be

maintained by using sequentially numbered, sealed, opaque envelopes containing the numeric

code of the study arm to which the participant was randomised. The randomisation code will be

documented in the participants notes, the case report form (CRF) and provided to the ICU

nursing shift coordinator with access to the unblinding code for preparation of the intervention.

Active medication and re-dosing

60

There are substantial pharmacokinetic and pharmacodynamics differences between

commercially available IV iron preparation. These differences relate predominantly to the effect of

differences in the size of the iron core and composition of the carbohydrate shell on release and

distribution of the iron therapy 132. Given the lack of data for patients admitted to the ICU, the

choice of IV iron therapy and dosing schedule for the IRONMAN trial was based on the available

safety and efficacy extrapolated from non-critically ill patients. Although conclusive evidence is

lacking, there is some evidence from comparative studies that ferric carboxymaltose is more

effective and more cost-effective than iron sucrose, a commonly prescribed alternative iron

preparation 133 134.

Participants randomised to the intervention arm will receive 500mg of ferric carboxymaltose

(Ferrinject) as an IV infusion (figure 1). Ferric carboxymaltose is an iron-carbohydrate complex

licensed in Australia 135. It can be safely administered as a short IV infusion, without need for a

test dose, and provides controlled release of iron with a low risk of acute toxicity, infusion reaction

or immediate hypersensitivity. Previous large multicentre RCTs have used ferric carboxymaltose

in similar doses and reported efficacy with a low adverse event rate comparable to placebo 131 136.

Dosing schedule was 500mg ferric carboxymaltose prepared in 100ml sodium chloride 0.9%

infused over a total of 60 minutes. Given the lack of previous interventional studies in critically ill

patients, the choice of dose was based on the lowest dose associated with improved outcomes in

non-critically ill patients. Smaller, more frequent doses based on reassessment of iron saturation

may reduce the risk of oversaturation and may therefore also be more effective than single larger

doses.

Patients will be re-dosed with study medication if they remained in the ICU (or high dependency

unit that is under the supervision of an Intensivist) and are at least 4 days beyond their previous

dose, their Hb is less than 100g/L in the preceding 24 hours and they continue not to fulfill any of

the exclusion criteria. Initial study eligibility and re-dosing are dependent on excluding potential

iron overload (ferritin >1200ng/ml or transferrin saturations >50%). Previous RCTs demonstrating

the safety and efficacy of IV iron have used similar ferritin and transferrin saturation thresholds 136

137. Assessment for re-dosing will continue according these criteria until death, discharge from the

61

ICU or a maximum of 4 doses of study drug (total 2000mg IV iron or placebo) are administered,

whichever occurs first (figure 2).

Placebo

Participants randomised to the placebo arm will receive 100ml of sodium chloride 0.9% delivered

by an identical infusion pump over a total of 60 minutes, in addition to standard care. All aspects

of patient management other than the specific study-related procedures will be at the direction of

the treating clinician. Open-label oral or IV iron and open-label ESA will be discouraged and

considered a protocol violation in patients participating in the IRONMAN study. No formal RBC

transfusion will be specified as part of the study protocol. However, clinicians will be encouraged

to follow the critical care National Blood Authority transfusion guidelines including a restrictive

transfusion trigger 138.

Blinding

Study treatment will be blinded using an opaque sleeve covering the syringe and giving set

(figure 2). An un-blinded research nurse or pharmacist will draw up the study medication and the

bedside nurse will deliver the infusion. Outcome assessment and data collection will be

conducted by the blinded research coordinators at each study site. Analysis of the study results

will be conducted blinded by the primary investigator. The efficacy of blinding will be assessed

according to a sub-study questionnaire that asked the clinician responsible for the patient

whether they were aware of the study allocation.

Discontinuation of study treatment

Participants will be discontinued from receiving further study treatment on discharge from the ICU

or after the fourth dose of study treatment, whichever occurs first. Any participant that develops

sepsis (as defined by two or more SIRS criteria plus antibiotics started or changed by the treating

clinician for suspected or confirmed infection) will be discontinued from receiving further study

treatment for the duration of the period of sepsis. Participants that experience a suspected or

confirmed immediate hypersensitivity reaction temporarily related to delivery of the study

intervention will have delivery of the study treatment ceased immediately.

Sample Size

62

The assumption of a mean of 4 RBC transfusions in eligible patients remaining in the ICU >2

days was based on the findings of the observational study (Chapter 3), a standard deviation in

the intervention and control groups of 2 and a loss to follow up or incomplete data rate of 10%

(including those participants initially enrolled by the next of kin then declining to provide ongoing

participant consent), a study of 140 patients has 80% power to detect a decrease in mean

number of RBC transfusions of 1 unit at a significance level of 5%.


All analyses will be conducted on an intention-to-treat basis without adjustment for baseline

variables. Differences in outcome variables will be compared using t-test and Chi-Squared test as

appropriate if normally distributed and using non-parametric equivalents if not normally

distributed. Analysis will be primarily conducted using STATA version 10.2 (College Station,

Texas 77845 USA). Data will be censored at 60 days post study enrolment for Hb, RBC

transfusion and vital status.

63

4.3 Data Management

All data will be collected by trained staff at each study site using a paper source document

developed by the management committee. Data will then be entered into a secured, password-

protected, web database (www.savant.net.au). Data queries will be automatically generated via

the electronic data collection database. Randomised patients will be followed up to death or

hospital discharge. A ‘day’ in ICU is defined as commencing at midnight. A list of study data is

provided in table 3.

Safety Monitoring

A drug safety and monitoring board (DSMB) will be convened comprising of three experienced

WA researchers including two Intensivists not associated with the study, along with a senior

Emergency Medicine clinician. Serious adverse events will be reported according to the Good

Clinical Practice Guidelines and the requirements of institutions in which the study will take place.

The DSMB will receive notification of all SAEs. No interim analysis is planned. However, the

DSMB reserves the right to request an interim analysis on the basis of un-blinded SAEs. In

keeping with the advice of Cook et al, events that are part of the natural history of the primary

disease process or expected complications of critical illness will not be reported as serious

adverse events unless thought to be causally related to the study intervention or otherwise

specified in the CRF 139.

Ethical Issues

The study will not proceed at any site until approval had been gained by the responsible Human

Research Ethics Committee (HREC). Prospective informed consent will be sought from eligible

patients wherever they retain capacity. However, the proportion of critically ill patients fulfilling

these characteristics is likely to be low and would not be representative of the patient population

most likely to benefit from the intervention under investigation. For eligible patients without

capacity to provide consent at the time of study eligibility, prospective informed consent will be

obtained prior to study enrolment from the designated next-of-kin. Consent for ongoing

participation will then be sought from participants who regained capacity as soon as practicable.

64

4.4 Conclusion

RBC transfusion is associated with increased morbidity and mortality in critically ill patients. RBCs

are also an increasingly costly and scarce resource. In the ICU, RBC transfusion occurs

predominantly for anaemia, and despite high compliance with recommended transfusion

thresholds, the incidence of RBC transfusion remains high. The aim of the IRONMAN

randomised controlled trial is to determine whether IV iron administered to anaemic patients

admitted to the ICU results in a decrease in mean RBC transfusion requirement.

65

4.5 Figures

Figure 1. Active Medication Kit

Figure 2. IRONMAN Trial Dosing Flow Chart

66

4.6 Tables

Table 1. Trial End Points Primary Outcome

Mean number of RBC units transfused from study enrolment to discharge from hospital

Secondary Outcomes

Proportion of participants who receive RBC transfusion from enrolment to ICU discharge

ICU and hospital mortality

Duration of admission to ICU and hospital

Organ-failure support-free days between enrolment and ICU discharge

Proportion of patients who develop nosocomial infection in ICU including all-

cause incident infection confirmed blood stream infection and incident

infection associated with new organ failure

Number of SAEs and proportion of patients who develop a SAE

Mean number of RBC units transfused and proportion of patients transfused from study enrolment to discharge

from hospital adjusted for baseline Hb, pre-enrolment transfusion, ferritin, transferrin saturation, hepcidin, soluble

transferrin receptors, renal replacement therapy

Subgroups


from hospital in patients with baseline transferrin saturations <20%


from hospital in patients with baseline ferritin <200ng/ml


from hospital in patients receiving more than one dose of study drug

Mean Hb on discharge from ICU and hospital in patients not receiving RBC

transfusion in ICU after study enrolment

Duration from enrolment to time of first RBC transfusion in patients receiving at least one RBC unit after enrolment

67

Table 2. Trial Eligibility Criteria Inclusion Criteria

Admitted to an ICU for less than 48 hours

Anticipated to require ICU care beyond the next calendar day

Hb less than 100g/l at any time in the preceding 24 hours

Age 18 years or greater

Exclusion Criteria

Suspected or confirmed severe sepsis (two or more systemic inflammatory response syndrome criteria, suspected

or confirmed infection and one or more organ system failure)

Serum ferritin greater than 1200ng/ml or transferrin saturation greater than 50%

History of haemachromatosis or aceruloplasminaemia

Known prior administration of IV iron in the preceding 3 months

Jehovah’s Witness or other documented exclusion to receiving blood products

Receiving ESA (e.g. epoetin or darbepoeitin) in the preceding 3 months

Known hypersensitivity to IV iron

Pregnancy

Treatment intent is palliative

Death is imminent and inevitable

Weight less than 40kg

Participating in a competing study

68

Table 3. Data to be collected in the IRONMAN trial Baseline

Age and gender

Date of Hospital and ICU Admission

Number of RBC units transfused between arrival in hospital and ICU admission

First Hb and MCV on arrival to hospital

Episode of bleeding prior to ICU admission

Source of ICU admission

APACHE II score and diagnostic code on ICU admission

SOFA score and components on ICU admission

Hb closest to but prior to enrolment in study

Organ support on enrolment in study: mechanical ventilation, vasopressors, renal replacement therapy

Iron studies: serum iron, transferrin, transferrin saturation, ferritin, soluble transferrin receptors

Hepcidin and C-reactive protein

Daily

SOFA score and components

Hb

Hepcidin and C-reactive protein

Number of RBC units for which transfusion commenced during calendar day

Indication for transfusion

Organ support during calendar day: mechanical ventilation, vasopressors, renal replacement therapy

New infection

Organ failure associated with infection

New episode of bacteraemia

Redosing of study drug

Iron studies on days eligible for redosing: serum iron, transferrin, transferrin saturation, ferritin

Adverse events including anaphylaxis

Discharge

Date of discharge from ICU and hospital

Readmission to ICU

Survival status at ICU and hospital discharge

Number of RBCs transfused after ICU discharge

Received non-study-drug related iron

Hb on hospital discharge

Discharge destination

69

Chapter 5

Intravenous Iron or Placebo for Anaemia in Intensive Care: The

IRONMAN Multicentre Randomized Blinded Trial

70

5.1 Introduction

Anaemia is extremely common in patients admitted to the ICU and is the most common indication

for allogeneic RBC transfusion even when adherence with transfusion guidelines is high1 2. Both

anaemia and RBC transfusion may be harmful to critically ill patients. Anaemia is an independent

risk factor for mortality and major morbidity in patients undergoing major surgery and in general

ICU patients; RBC transfusion is associated with mortality, nosocomial infection, multi-organ

dysfunction syndrome and the acute respiratory distress syndrome (ARDS) in patients treated in

an ICU3 129 140 141.

Progressive anaemia and subsequent RBC transfusion are predictable at the time of ICU

admission142. In selected patients, novel interventions implemented shortly after ICU admission

could reduce the incidence and severity of anaemia, the need for RBC transfusion, and therefore

the burden of associated morbidity and mortality. Intravenous (IV) iron decreases both the

severity of anaemia and incidence of RBC transfusion in non-critically ill patients 116. However,

there is a theoretical risk of causing or worsening infection and older preparations are associated

with anaphylactic reactions 24 116 143. High quality safety and efficacy data for IV iron in the critical

care setting are lacking.

The multicentre Intravenous Iron or Placebo for Anaemia in Intensive Care (IRONMAN) RCT was

designed to test the hypothesis that, in critically ill patients admitted to the ICU who are anaemic,

early administration of IV ferric carboxymaltose, compared with placebo, reduces the mean

number of RBC units transfused between randomisation and hospital discharge.

71

5.2 Methods

Study Design and Oversight

Between 20 June 2013 and 6 June 2015, a randomized, placebo-controlled, blinded trial was

conducted in four ICUs in Perth, Western Australia. The study protocol was registered

prospectively on the Australian New Zealand Clinical Trials Registry

(ANZCTRN12612001249842), was approved by the ethics committee at each participating site,

and has been published previously126. Prospective consent was obtained from all participants or

their legal surrogates. The trial was overseen by an independent Data Safety Monitoring

Committee. Study drug was supplied by Vifor Pharma© which had no other role in the design or

conduct of the study or analysis and reporting of the results.

Study Population

Patients were eligible to participate if they were 18 years of age or older, within 48 hours of

admission to ICU, anticipated to require ICU care beyond the next calendar day and had a

haemoglobin (Hb) less than 100 g/L at any time in the preceding 24 hours. Exclusion criteria

included suspected or confirmed severe sepsis, a ferritin greater than 1200 ng/ml or transferrin

saturation greater than 50%. A complete list of the exclusion criteria are provided in the

supplementary appendix.

Randomization and blinding

Eligible patients were randomly assigned in a 1:1 ratio to receive either IV iron or placebo. The

randomization sequence was generated by an online resource was stratified according to study

centre144. Allocation concealment was maintained by using permuted block randomisation and

sealed, opaque, consecutively numbered envelopes at each study site that had been generated

centrally by staff unrelated to the study or ICU. Randomisation was to a study number. Study

medication was then prepared by a clinical nurse or pharmacist not involved in the care of the

patient. An opaque sleeve covering the study drug infusion syringe and giving set was used to

maintain blinding of the participants, treating, site researchers and data collectors 126. The

adequacy of blinding was assessed by conducting a blinding sub-study measuring inter-rater

72

agreement between the study intervention actually delivered and the opinion of the intervention

arm according to the attending clinician using Cohen’s Kappa.

Study Treatments

Patients randomized to the IV iron group received 500mg of ferric carboxymaltose in 100 ml of

0.9% saline delivered in two consecutive 50ml syringes. Details of the study treatment including a

photo of the blinding set up have been published previously126. Patients in the placebo group

received 100ml of 0.9% saline alone. Four days after receiving the initial or subsequent dose of

study drug, patients remaining in the ICU were assessed for repeat dosing. Participants were

eligible for re-dosing if they continued to fulfill the study eligibility criteria, including repeated

ferritin and transferrin saturation parameters and an Hb<100g/L. Assessment for suitability for re-

dosing continued until the patient was discharged from the ICU, received four doses of study drug

or died, whichever occurred first.

The IV iron formulation was chosen on the basis of data supporting superiority of ferric

carboxymaltose at fixed dose compared with an alternate IV iron formulation and low reported

side effect profile 131 135. The ferritin and transferrin saturation (TSAT) cutoffs were chosen on the

basis of the higher end of the effective reported range (ferritin <1200mn/ml) and lack of

interaction between TSAT and IV iron on RBC transfusion 116 137.

All aspects of patient management, including decision for RBC transfusion and ICU discharge,

were administered according to local practice and at the direction of the treating ICU clinician.

There were no RBC transfusion policies in any of the participating centres. Open-label IV iron and

erythropoiesis-stimulating agents were strongly discouraged and use of these agents were a

protocol violation.

Study Outcomes

The primary study outcome was number of RBC units transfused per patient between

randomisation and hospital discharge reported according to an intention-to-treat analysis.

Secondary outcomes included Hb at hospital discharge, proportion of patients receiving RBC

transfusion, ICU and hospital length of stay and mortality and infection. Infection was defined as

the commencement, escalation or change of IV antibiotics for a confirmed or strongly suspected

73

infection and was adjudicated locally by blinded clinical staff. Clinically confirmed deep vein

thrombosis (DVT) and pulmonary embolism (PE) were explicitly collected as SAEs. Bleeding

definitions are provided in the supplementary appendix. Admission diagnoses were based on

acute physiology and chronic health evaluation (APACHE) II diagnostic codes. Events were

deemed to be part of the natural history of the primary disease process or expected

complications of critical illness were not reported as SAEs unless thought to be causally related

to the study intervention.


All analyses were conducted on an intention-to-treat basis. No imputation was made for missing

data. Continuous variables were reported as mean (±SD) or median and interquartile range

(IQR), with between group differences analysed using Student’s t-test or the Wilcoxon rank-sum

test for apparently normal and non-normally distributed data respectively. Categorical variables

were reported as proportion and analysed using the Chi2 test or Fischer exact test as appropriate.

Data was censored at 60 days after enrolment for Hb level, RBC transfusion and vital status. A

two-sided P value of 0.05 or less was considered to be statistically significant. All analyses were

conducted with Stata Version 14 StataCorp College Station, TX77845, USA. No interim analyses

were planned or conducted.

Although the analyses were conducted according to a previously reported statistical analysis

plan126, the number of RBC units was not normally distributed and, in conjunction with advice

from an independent statistician (Centre for Applied Statistics, University of Western Australia),

the primary outcome has been reported as median and IQR instead of the prespecified mean and

standard deviation (SD). The data has then been analysed using negative binomial regression

with incidence-rate ratios reported. This analysis satisfied the assumptions as count data with

over-dispersion (variance greater than the mean). The sample size of 140 participants was based

on a baseline mean of four RBC transfusions in eligible patients, determined from an

observational study conducted in one of the participating study sites, with a SD in the intervention

and control groups of two and a loss to follow-up of 10%142. This provided 80% power to detect a

decrease in the mean number of RBC transfusions of 1 unit at a significance level of 5%.

74

Additional sensitivity analyses of the primary outcome variable adjusted for predefined covariates

(enrollment Hb, RBC transfusion prior to enrollment, transferrin saturation, ferritin, soluble

transferrin receptor and renal replacement therapy) were performed using negative binomial

regression for count data. The effect of IV iron on incidence-rate ratio of RBC transfusion was

performed for predefined subgroups including transferrin saturation (<20% or ≥20%) and ferritin

(<200ng/ml or ≥200ng/ml).

75

5.3 Results

The study enrolled 140 patients, with 70 assigned to IV iron and 70 to placebo. All participants

received the intervention to which they were randomly allocated and all patients were followed up

to discharge from index hospitalisation. One patient declined consent to ongoing participation at

time of ICU discharge but consented to data use. Repeat dosing of study drug occurred in 17

patients in the IV iron group (15 patients received two doses, three patients received three doses)

and 26 patients in the placebo group (23 patients received two doses, three patients received

three doses). Seven participants in the IV iron group and three participants in the placebo group

received non-study-drug IV iron either in ICU (n=1) or post-ICU discharge (n=9). There was no

missing data for the primary or prespecified secondary outcomes (figure 1). Demographic and

clinical characteristics at baseline were similar between the groups (table 1), and there was no

significant association between perceived and actual study group allocation (McNemar’s test chi2

2.37, p=0.12).

Primary Outcome

The IV iron group was transfused 97 RBC units versus 136 RBC units in the placebo group. The

number of RBC units transfused in the ICU was 79 (81%) and 121 (89%) for the IV iron and

placebo groups respectively. The median (IQR) RBC transfusion in the IV iron and placebo

groups [1 unit (0-2) vs. 1 unit (0-3) P=0.53], incidence rate ratio (IRR) [0.71 (95% confidence

interval (CI) 0.43-1.18) P=0.19]. (table 2). There was no significant between-group difference in

RBC transfusion with the use of multivariable binomial regression adjusting for predefined

baseline covariates (P=0.77), or according to a per protocol analysis (P=0.15). Between-group

RBC transfusion was also similar in the predefined subgroups (Table 3). RBC transfusion (figure

3) and median Hb (figure 4) by day whilst in ICU are provided in the supplementary appendix.

Secondary Outcomes

Overall, the median Hb at hospital discharge was significantly higher in the IV iron group

compared with the placebo group (107 g/L (IQR 97-115) vs. 100 g/L (IQR 89-111), P=0.02). The

histograms for the Hb on hospital discharge for the two groups are provided in the supplementary

appendix (figure 2). In a post-hoc analysis, the proportion of patients discharged from hospital

76

with an Hb<100g/L was significantly lower in the IV iron compared with placebo groups (21/70

(30%) vs. 33/70 (47%), p=0.04). The IV iron and placebo groups had similar median lengths of

stay in ICU and hospital, and no significant differences in ICU and hospital mortality were

observed (table 2).

Safety

There was no statistical difference between the iron and placebo groups in infection, infection

associated with organ failure, or bacteraemia. The number of serious adverse events (SAEs) did

not differ significantly between groups. There were no immediate study-drug-related adverse

events in the IV iron group and one in the placebo group where shivering post study drug

administration was thought to be possibly related to study drug (table 4).

77

5.4 Discussion

In this multicentre randomized trial of patients admitted to the ICU who were anaemic, it was

found that IV iron, compared with placebo, did not result in a significant difference in number of

RBC units transfused. IV iron did however result in a significantly higher Hb concentration at

hospital discharge. Safety outcomes, specifically mortality, infection, clinically diagnosed venous

thrombosis and immediate infusion-related adverse events were not significantly different in those

receiving IV iron compared with placebo.

Outside of the critical care setting, trials enrolling patients with similar baseline Hb and

haematinics have shown a significant decrease in RBC transfusion associated with IV iron

therapy 116. Although the point estimate for the primary outcome in our study favored IV iron, the

difference was not significant. One possible reason is that IV iron is simply ineffective in patients

admitted to the ICU due to the modulating effects of severe inflammation on the erythropoietic

response to IV iron119 145. Given that the point estimate of the primary outcome favors IV iron with

a clinically meaningful decrease in incidence rate ratio of 0.71, and the statistically significant

increase in Hb at hospital discharge associated with IV iron, this would appear unlikely. Perhaps

more likely is the effect of the mean number of RBC units transfused being substantially lower

(1.9 units in the placebo group) than anticipated. The IRONMAN study was powered to detect a

one unit reduction from a baseline of four units transfused; the observed reduction was 0.5 units.

The study was underpowered to detect such a difference leading to the possibility of a type II

error (see supplementary appendix for a power calculation for a future trial of IV iron).

Whilst our study attempted to identify a cohort of patients at high risk of progressive anaemia and

subsequent RBC transfusion, characteristics associated with an erythropoiesis response to IV

iron in the critical care setting are poorly understood and require further consideration. For

example, the relative efficacy of IV iron in patients with anaemia at least partly due to absolute

iron deficiency, compared with anaemia of inflammation alone, remains uncertain, and

measurement of hepcidin may be of value 119. Future trials of IV iron in critical illness should

consider adopting a lower Hb threshold for enrolment, only enrolling patients with a longer

predicted length of stay, and targeting the intervention at those most likely to mount an

78

erythropoeitic response. This would have the simultaneous effect of identifying a population at

higher risk of RBC transfusion and prolonged ICU stay and greater risk of adverse outcomes.

Pieracci et al conducted an RCT of IV iron sucrose in trauma patients admitted to the ICU and

found no difference in Hb concentration125. In contrast, the IRONMAN study found that IV iron

resulted in a statistically significant increase in Hb at hospital discharge and a greater proportion

of patients discharged with an Hb>100g/l, although the clinical significance of these findings is

uncertain.

Compared with Pieracci et al, our study used a higher dose of iron, and an alternative preparation

previously shown to be associated with greater erythropoietic response 131. The IRONMAN study

also enrolled patients at higher risk of RBC transfusion (Hb threshold for enrollment 100 g/l vs.

120 g/l) and included a broader range of critically ill patients, potentially at greater risk of

preexisting iron deficiency.

It is plausible that a higher Hb during recovery from critical illness may be of clinical benefit,

including more rapid functional recovery and decreased LOS. Although the IRONMAN study did

not find a significant decrease in hospital LOS associated with IV iron, the median duration from

initiation of IV iron to hospital discharge was 11 days, whereas maximal therapeutic effect may

not occur for three to four weeks. Whether the observed difference was greater post-discharge,

and the clinical benefits of a higher Hb in a cohort of patients with a longer estimated LOS require

further consideration.

Bateman et al found that moderately severe anaemia at the time of ICU discharge was

associated with a markedly reduced health-related quality of life score at three and six months

compared with a non-selected ICU cohort, and that over half remained anaemic at six months10.

Postoperative rehabilitation studies suggest that anaemia is associated with fatigue, reduced

exercise capacity, muscle strength and performance in activities of daily living and may impair

recovery 146. Furthermore, Froessler et al, found that IV iron prior to major abdominal surgery

was associated with a significant decrease in hospital length of stay and a significant increase in

Hb at four weeks, suggesting a role for IV iron in enhancing recovery147.

79

The IRONMAN study found no association between IV iron and infection. New infection was

defined in terms of the commencement, escalation or change of antibiotics. This definition was

pragmatic, reflective of clinical practice, and assessed by blinded clinicians. Future studies may

consider blinded adjudication by independent experts and powering the study to exclude a

clinically important difference in infection.

The formulation and dosing of IV iron in the IRONMAN study resulted in no immediate adverse

events. Given the lack of data in for IV iron use in ICU, a cautious dosing approach was chosen

and it is plausible that in future studies, a higher, weight-based dosing and/or continued dosing

after ICU discharge, may result in a greater response to IV iron. The comparative efficacy of other

IV iron preparations in this context remains uncertain.

Strengths

The IRONMAN study has a number of strengths including a pragmatic design, effective blinding,

administration of the study drug to all participants according to the assigned study group,

complete follow up to discharge from index hospitalisation and the use of a restrictive RBC

transfusion approach

Limitations

The data distribution for the primary outcome required a change to the planned statistical

analysis, adding to the possibility of a type II error. Baseline transfusion was lower than planned,

reducing the power of our study to detect a difference in RBC units. A small proportion of patients

received non-study IV iron; however, the number were not significantly different between groups

and did not change the findings when the groups were analysed per protocol. The significant

increase in Hb at discharge was a secondary outcome and there is a risk that this is a chance

finding due to multiple testing. However, the point estimate for RBC transfusion also favors IV

iron, so a false positive result is considered less likely. Fewer patients required transfusion for

major haemorrhage in the IV iron compared with placebo groups, although the difference was not

statistically significant. Although a differential effect of mortality or hospital LOS may affect

interpretation of the primary end-point, neither were significantly different between groups and so

this is considered unlikely. Finally, threshold for RBC transfusion was at the discretion of the

80

treating clinician and not specified as part of the study. Treating clinicians were however blinded

to the study allocation and median Hb prior to transfusion was within published guidelines and not

significantly different between groups138.

81

5.5 Conclusion

In patients admitted to the ICU who were anaemic, IV iron compared with placebo, did not result

in a significant difference in RBC transfusion at hospital discharge. Patients who received IV iron

had a significantly higher Hb at hospital discharge.

82

5.6 Figures

Figure 1. Participant Flow

83

5.7 Tables Table 1 Baseline Characteristics Characteristic*

IV Iron (n=70)

Placebo (n=70)

Age – years

58.5 (18.8) 56.0 (21.1)

Male gender – no. (%)

44 (63) 52 (74)

APACHE II score

12.2 (5.7) 13.8 (6.1)

SOFA score

6.1 (2.5) 6.6 (3.3)

ICU admission source – no (%) Emergency department Hospital ward Operating theater Other hospital

14 (20) 5 (7) 50 (71) 1 (1)

13 (19) 4 (6) 50 (71) 3 (4)

ICU admission type – no (%) Surgical Medical

61 (87) 9 (13)

60 (86) 10(14)

ICU admission subtype – no (%) Surgical subgroups - General surgical - Cardiothoracic - Trauma Neurosurgical Medical subgroups - Congestive cardiac failure - Cardiac ischaemia - Cardiogenic shock - Pulmonary embolism - Gastrointestinal bleeding - Acute kidney injury - Metabolic - Neurological (undefined) - Overdose - COAD - Respiratory (undefined)

9 (13) 30 (43) 20 (29) 2 (3) 2 (3) 1 (1) 0 (0) 2 (3) 2 (3) 1 (1) 1 (1) 0 (0) 0 (0) 0 (0) 0 (0)

13 (19) 19 (27) 25 (36) 3 (4) 3 (4) 1 (1) 1 (1) 0 (0) 0 (0) 1 (1) 0 (0) 1 (1) 1 (1) 1 (1) 1 (1)

Mechanical ventilation – no. (%)

45 (64) 48 (69)

Vasoactive infusion – no. (%)

51 (73) 48 (69)

Renal replacement therapy – no. (%)

3 (4) 5 (7)

Haemoglobin – median g/l (IQR)

89 (81-94) 87 (79-95)

Ferritin –ng/ml+

317 (218) 365 (436)

Transferrin saturation - %

13 (10) 14 (12)

C reactive protein – mg/l 111 (83)

122 (85)

RBC transfusion – median units (IQR)

0.5 (0-4) 1.5 (0-4)

RBC transfusion prior to randomisation –no (%)

13 (19) 18 (26)

Time from ICU admission to initiation of study – hours

28 (13) 31 (13)

*Mean and standard deviation (SD) unless otherwise reported. +ng/ml has a conversion factor of 1 to the standard international units mcg/ml

84

Table 2. Study Outcomes Variable* IV Iron

(n=70)

Placebo (n=70)

P Value Risk ratio^

or median difference$ for IV iron compared with placebo (95% CI)

Primary outcome – total RBC units/participants

97/70 136/70

RBC units 1 (0-2) 1 (0-3) 0.53

0.71^ (0.43-1.18)

Received RBC transfusion – participants transfused/total participants (%)

38/70 (54)

39/70 (56) 0.87 0.97^ (0.72-1.31)

RBC units per transfused patient

2 (1-3) 2 (1-5) 0.25 0.73^ (0.50-1.06)

RBC units transfused in ICU – RBC units ICU/ total RBC units (%)

79/97 (81) 121/136 (89) 0.10

Indication for RBC transfusion in ICU – no participants (% total participants transfused in ICU) Major bleeding Minor bleeding Anaemia Low cardiac output Other

1 (3) 7 (21) 28 (85) 2 (6) 0 (0)

3 (8) 8 (21) 31 (82) 3 (8) 1 (3)

0.62 0.79 0.61 1.0 1.0

0.33^ (0.04-3.13) 0.88^ (0.34-2.28) 0.90^ (0.61-1.33) 0.67^ (0.11-3.87)

Hb prior to transfusion g/L

76 (71-81) 75 (69-84) 0.74 1$ (-3.13-5.13)

Hb at hospital discharge g/L

107 (97-115) 100 (89-111) 0.02 7$ (1.89-12.11)

Duration from study drug to first RBC transfusion – days

2 (1-3) 1 (1-2) 0.22 1$ (0.29-1.71)

Duration from study drug to determination of Hb at hospital discharge – days

11 (7-24) 15 (8-24) 0.51 -4$ (-8.98-1.98)

ICU mortality – no/total (%)

5/70 (7) 3/70 (4)

0.47 1.67^ (0.41-6.71)

Hospital mortality – no/total (%)

7/70 (10)

6/70 (9)

0.77 1.17^ (0.41-3.30)

Duration of stay ICU - days Hospital - days

6 (5-9) 15 (11-28)

6 (5-9) 18 (11-25)

0.70 0.75

0$ (-1.07-1.07) -3$ (-7.95-1.95)

ICU organ failure support-free days

2 (1-3) 2 (1-3) 0.89 0$ (-0.68-0.68)

*Median and interquartile range (IQR) unless otherwise reported.

85

Table 3. Subgroup Analysis – Effect of IV Iron on Incidence-Rate Ratio for RBC Transfusion* Incidence-rate ratio

(95% Confidence interval)

P value P value for interaction

Transferrin saturation≤20% Yes (n=113) 0.73 (0.42-1.29)

0.29 0.92

No (n=27)

0.78 (0.31-1.94) 0.60

Ferritin≤200ng/ml Yes (n=54)

0.65 (0.25-1.70) 0.38 0.75

No (n=86)

0.77 (0.43-1.36) 0.36

*Negative binomial univariate regression. An incidence-rate ratio of less than one favors intravenous iron. Table 4. Safety Variable* IV Iron

(n=70)

Placebo (n=70)

P Value Relative risk (95% CI)

Nosocomial infection – no. (%)

20 (28.6) 16 (22.9) 0.44 1.25 (0.71-2.21)

Nosocomial infection associated with organ failure – no. (%)

2 (2.9) 0 (0) 0.50

Bacteraemia – no. (%)

2 (2.9) 1 (1.4) 1.0 2.0 (0.19-21.56)

Immediate study-drug related AEs – no. (%)

0 (0) 1 (1.4) 1.0

SAEs – no. (%)

4 (6) DVT=2 PE=2

4 (6) DVT=1 PE=3

1.0 1.0 (0.26-3.84)

*Mean and standard deviation (SD) unless otherwise reported.

86

5.8 Supplementary data Additional IRONMAN RCT investigators Dr Andy Chapman, Ms Janet Ferrier, Dr Stuart Baker, Prof Wendy Erber, A/Prof Craig French, Dr David Hawkins, Ms Alisa Higgins, Dr Axel Hoffmann, Dr Bart De Keulenaer, Mr Shannon Farmer, Dr Julie McMorrow, Prof John Olynyk, Ms Anne-Marie Palermo, Mr Toby Richards, Ms Brigit Roberts, Dr Simon Towler, Prof Steve Webb Trial Eligibility Criteria

Inclusion criteria

1. Admitted to an ICU for less than 48 hours

2. Anticipated to require ICU care beyond the next calendar day

3. Hb less than 100 g/L at any time during the preceding 24 hours

4. Age 18 years or greater

Exclusion criteria

1. Suspected or confirmed severe sepsis (two or more Systemic Inflammatory Response

Syndrome (SIRS) criteria, suspected or confirmed infection, and one or more organ system

failure)

2. Serum ferritin greater than 1200ng/ml or transferrin saturation greater than 50%

3. History of haemochromatosis or aceruloplasminaemia

4. Known prior administration of IV iron in the preceding 3 months

5. Jehovah’s Witness or other documented exclusion to receiving blood products

6. Receiving ESA (e.g. epoietin or darbepoeitin) in the 3 months prior to ICU admission

7. Known hypersensitivity to intravenous iron

8. Pregnancy

9. Treatment intent is palliative

10. Death is deemed imminent and inevitable

11. Weight less than 40kg

12. Participating in competing study

Minor and Major Bleeding Definitions

Minor bleeding = overt or suspected bleeding or bleeding apparent on imaging studies without

haemodynamic compromise (SBP<90mmHg) and not requiring transfusion or fluid resuscitation,

specific diagnostic tests or interventions or initiation or escalation in vasopressor requirement

Major bleeding = overt or suspected bleeding or bleeding apparent on imaging studies with

haemodynamic compromise (SBP<90mmhg) or requiring transfusion or fluid resuscitation,

specific diagnostic tests or interventions or initiation or escalation in vasopressor requirement

87

Power calculation for a future trial of IV iron

On the basis of the results of this study (baseline mean RBC transfusion of 1.9 units, standard

deviation 3, mean difference 0.5 RBC units), a future trial of 567 participants per group would

have 80% power to detect a change in RBC units transfused of 0.5 (alpha=0.05). The sample

size calculation would then have to be inflated to account for non-normal distribution (20%),

potential decrease in baseline RBC use over time (5%) and loss to follow up (10%). This would

give a final trial sample size of approximately 1572 participants.

Figure 2 Histogram of Hb at Hospital Discharge for IV Iron and Placebo Groups

88

Figure 3. Total RBC units by Study Day for Patients Remaining in ICU

Figure 4. Median Hb by Study Day for Patients Remaining in ICU

89

Chapter 6

Utility of Hepcidin in Predicting Risk of Red Blood Cell Transfusion

and Response to IV Iron Therapy in Patients Admitted to the

Intensive Care Unit: A Nested Cohort Study

90

6.1 Introduction

RBC transfusion occurs in approximately one third of patients admitted to the ICU, and is

associated with increased morbidity and mortality 148. IV iron reduces RBC transfusion

requirement in the acute care setting and has recently been shown to increase haemoglobin in

patients admitted to the ICU 116 149.

Anaemia is the most common indication for RBC transfusion in critically patients. However, acute

inflammation makes diagnosis of iron deficiency unreliable. In contrast, hepcidin may be down-

regulated by iron insufficiency even in the presence of inflammation. Whether hepcidin

concentration can predict risk of RBC transfusion and response to IV iron therapy in critically ill

patients, in whom iron-restricted erythropoiesis and inflammation frequently coexist, remains

uncertain 150.

The primary aim of this nested cohort study, was to assess the association between hepcidin

concentration in patients within 48 hours of ICU admission and subsequent RBC transfusion

requirement during the index hospitalisation, and to determine whether hepcidin concentration

could be used to identify a group of patients in whom IV iron therapy compared to placebo, would

decrease RBC transfusion 149.

91

6.2 Methods

The study received Human Research Ethics Committee approval at all sites prior to

commencement and prospective consent was obtained from all participants or their legal

surrogates. The protocol and primary results of the IRONMAN RCT have been presented in

Chapters 4 and 5 126 149. Briefly, the IRONMAN RCT enrolled adult patients in four centres in

Perth, Western Australia between 20 June 2013 and 6 June 2015 who were within 48 hours of

admission to ICU, had a haemoglobin (Hb) of less than 100 g/L, and were anticipated to require

ICU care beyond the next calendar day. Exclusion criteria included suspected or confirmed

severe sepsis, a ferritin greater than 1200 ng/ml or transferrin saturation greater than 50%.

Participants were randomised in a 1:1 ratio to receive either 500mg IV ferric carboxymaltose or

placebo and were followed up to hospital discharge.

Baseline blood was collected at the time of enrolment, prior to study drug administration.

Hepcidin-25 was isolated from for quantitation by liquid chromatography-quadrupole time-of-flight

mass spectrometry (LC-qTOF-MS), using a Waters Synapt G2S (Waters, Manchester, UK) as

previously described 151 152. Briefly, the hepcidin was isolated by solid phase extraction (SPE)

following the initial addition of a synthetic human hepcidin (13C18,15N3) peptide internal standard

(Peptides International, Kentucky, USA), and removal of the more abundant polypeptides by

organic solvent precipitation and centrifugation. The accurate mass measurement of the

precursor hepcidin-25 [M+5H]5+ ion was confirmed against a hepcidin-25 standard (Peptides

International, Kentucky, USA); and further by MS/MS. Quantitation was by reference to a human

hepcidin-25(13C18,15N3) calibration, prepared in human serum.


Continuous variables were reported as mean (±SD) or median and interquartile range (IQR), with

between group differences analysed using Student’s t-test or the Wilcoxon rank-sum test for

apparently normal and non-normally distributed data respectively. Categorical variables were

reported as proportion and analysed using the Chi2 test or Fischer exact test as appropriate. Data

was censored at 60 days after enrolment for RBC transfusion Hb concentration and vital status.

92

The relationship between hepcidin concentration, and other baseline risk factors, and subsequent

RBC transfusion quantity was assessed using negative binomial univariate and multivariate

analyses. Variables with a P value of <0.3 on univariate analysis were included in a multivariable

analysis with stepwise removal of variables with a P value >0.05. Interaction was assessed using

multivariable fractional polynomials to account for potential non-linear relationships. Significant

interactions (P value of <0.05) were examined graphically and a final model then produced. The

relative prognostic value with and without baseline hepcidin concentration included was assessed

using Akaike information criterion (AIC).

The methodology used in this study to examine the predictive value of hepcidin concentration

was similar to that used in a previously published RCT of IV iron in patients with chemotherapy-

induced anaemia 153. After first stratifying patients by tertile of baseline hepcidin concentration,

the incident risk ratio for RBC transfusion (as the primary outcome measure of response to IV

iron) was compared between those who were randomised to receive IV iron versus those who

received placebo. The relationship between IV iron therapy and RBC transfusion quantity across

the range of hepcidin values was further explored by locally-weighted scatterplot smoothing

(LOWESS) 154.

Data was censored at 60 days after enrolment for Hb level, RBC transfusion and vital status. A

two-sided P value of 0.05 or less was considered to be statistically significant. In order to assess

whether the association between hepcidin and RBC transfusion varied according to baseline Hb

and/or use of IV iron, interaction terms were added to the multivariate analysis. The performance

of the model with and without hepcidin as a predictor was compared using Akaike information

criterion. Assessment of the predictors of Hb at hospital discharge were determined in a similar

way, using linear regression. All analyses were conducted with Stata Version 14 StataCorp

College Station, TX77845, USA.

93

6.3 Results

Baseline hepcidin levels were available for 133 (95%) out of the 140 participants enrolled in the

IRONMAN RCT. The flow of participants is presented in figure 1. The mean time from ICU

admission to collection was 29 hours [Standard Deviation (SD) 13] and median hepcidin value

was 34.9 µg/L [interquartile range (IQR) 17.3-69.2, range 0-163.5]. The baseline characteristics

of the population are provided in table 1. There was no significant correlation between hepcidin

concentration and other baseline values including C reactive protein and standard iron indices

(Hb -0.09, P=0.308. Ferritin 0.13, P=128. Transferrin saturation -0.15, P=0.09. Soluble transferrin

receptor -0.15, P=0.10. C reactive protein -0.09, P= 0.308).

Prediction of RBC transfusion quantity

The complete list of variables assessed on univariate analysis and those added to the initial

multivariable model are provided in table 3. ICU admission related to trauma, baseline Hb,

transferrin saturation and hepcidin concentration were found to be significant independent

predictors of risk of RBC transfusion and retained in the final multivariable model. There was a

significant interaction between Hb and hepcidin concentrations in predicting the risk of RBC

transfusion (likelihood ratio test for significance of interaction p=0.0462), see figure 2. For

patients with a Hb ≥80g/L, increasing each 10 µg/ml increase in hepcidin concentration was

associated with a risk ratio of RBC transfusion of 1.09 (95%CI 1.01-1.18, P=0.034). However, for

patients with a Hb <80g/L there was no significant association between hepcidin concentration

and risk of RBC transfusion [IRR 0.95 (95%CI 0.84-1.07), p=0.387]. The variables included in the

final model are provided in table 4. Akaike information criterion (AIC) with hepcidin in the model

was 415.14 versus 468.16 with hepcidin removed.

Hepcidin and prediction of response to IV iron

The association between IV iron therapy and both RBC transfusion and Hb at hospital discharge

by tertile of hepcidin concentration are provided in table 2. For the 88 patients in the lower two

tertiles of hepcidin values, there was a significant decrease in risk of RBC transfusion associated

with IV iron therapy (38 RBC units, n=44) compared with placebo (79 RBC units, n=44), (incident

rate ratio [IRR 0.475 (95%CI 0.26-0.85), p=0.013]. However, no significant association was

94

found between IV iron therapy and risk of transfusion for the 45 patients in the highest tertile of

hepcidin value [IRR 1.33 (95%CI 0.57-3.08), p=0.518]. The LOWESS plot describing the

association between hepcidin concentration and RBC transfusion quantity in patients who

received IV iron compared with placebo is provided in figure 3.

IV iron therapy compared with placebo was not associated with a significant increase in Hb for

those in the lower two tertiles of hepcidin values [mean increase Hb 3g/L (95%CI -3-10),

p=0.361]. However, IV iron therapy was associated with a significant increase in Hb at hospital

discharge for patients in the highest tertile of hepcidin value [mean increase Hb 9 g/L (95%CI 2-

14), p=0.01]. There was no significant difference in iron, transferrin saturation, ferritin or

transferrin receptor levels associated with lower versus higher hepcidin tertile (table 5).

95

6.4 Discussion

In this study, elevated hepcidin levels, measured within 48 hours of ICU admission in critically ill

patients who were also anaemic, was an independent predictor of subsequent RBC transfusion

prior to hospital discharge. This effect was modified by Hb levels, and strongest above a cut-off

Hb of 80g/L, likely due to the high risk of RBC transfusion below this level. Importantly, IV iron

therapy compared to placebo was associated with a significant decrease in RBC transfusion

requirement when excluding patients in the upper tertile of hepcidin levels.

Lasocki et al found that hepcidin levels may be suppressed in critically ill patients with anaemia,

even in the setting of inflammation 150. Steesma et al conducted a secondary analysis of a RCT

and found that hepcidin levels predicted response to IV iron therapy in patients with

chemotherapy-induced anaemia also receiving an erythroid-stimulating agent (ESA) 153. These

results, together with our findings in which no ESA therapy was used, suggest that in the

presence of inflammation, measurement of hepcidin is useful in identifying patients in whom IV

iron therapy is likely to reduce RBC transfusion requirement.

Hepcidin levels have been shown to be the predominant predictor of erythrocyte iron

incorporation in African children with anaemia 155. Amongst adult patients admitted to the ICU,

Tacke et al have demonstrated an association between markers of increased iron availability and

mortality, predominantly amongst patients with sepsis 156. It is plausible that targeting IV iron

therapy to critically ill patients with lower hepcidin concentration, has the benefit not only of

significantly reducing RBC transfusion requirement, but may also reduce any potential risk of

initiating or exacerbating infection related to free iron.

Hepcidin synthesis is finely regulated including induction by both inflammation and iron overload

119. In our study, the median CRP concentration was 110 mg/L in patients in the lower two tertiles

of hepcidin concentration and there was no significant difference in iron indices between levels of

hepcidin. These results suggest that IV iron may be effective in reducing RBC transfusion

requirement even in the setting of substantial inflammation in the majority of patients admitted to

the ICU with an Hb of <100g/L and in whom sepsis has been excluded. The role of hepcidin

antagonists in patients with elevated hepcidin levels requires further consideration157.

96

In order for future studies to examine the utility of hepcidin for critically ill patients admitted to the

ICU, several obstacles must be overcome. First, substantial variation currently exists in hepcidin

assays and reference ranges, making comparison between studies difficult 157. Second, point of

care testing is necessary, to allow access to test results in a clinically useful timeframe. Finally, a

greater understanding is required in the changes in hepcidin levels over time and interaction with

key interventions including RBC transfusion and response to IV iron therapy.

Limitations

This study included only patients with an Hb<100g/L and did not include patients with sepsis at

the time of enrolment. The utility of hepcidin in these groups remains uncertain. Although soluble

transferrin receptor levels were assessed and found not to be predictive of RBC transfusion

requirement, other assay with potential diagnostic benefit including zinc protophoryn were not.

However, given the central role of hepcidin in iron metabolism, it is unlikely that other assays

provide superior diagnostic utility.

97

6.5 Conclusion

In critically ill patients with anaemia admitted to an ICU baseline hepcidin concentration predicts

RBC transfusion requirement and is able to identify a group of patients in whom IV iron compared

with placebo is associated with a significant decrease in RBC transfusion requirement.

98

6.6 Figures

Figure 1 Derivation of the Cohort

Figure 2 Effect of Hb Level on Association Between Baseline Hepcidin Concentration and Log

Prediction of RBC Transfusion Quantity

99

Figure 3. LOWESS Plot – Association Between Hepcidin Concentration and Subsequent RBC

Transfusion Quantity for IV iron and Placebo

100

6.7 Tables

Table 1 Baseline Characteristics

Characteristic*

Outcome

(n=133)

Age – years 62 (41-73)

Male gender – no. (%) 91 (68)

ICU admission source – no (%)

Emergency department

Hospital ward

Operating theater

Other hospital

24 (18)

8 (6)

97 (73)

4 (3)

ICU admission type – no (%)

Medical

General surgical

Cardiothoracic

Trauma

Neurosurgical

17 (13)

20 (15)

49 (37)

42 (32)

5 (4)

APACHE II score 12 (9-17)

SOFA Score 6 (4-9)

Prior RBC transfusion–no (%) 30 (23)

Haemoglobin – g/l 88 (81-94)

Mean corpuscular volume 91 (88-94)

C Reactive protein 110 (48-170)

Iron – 3 (2-6)

Ferritin –ng/ml+ 260 (161-437)

Transferrin 17 (15-20)

Transferrin saturation - % 9 (6-16)

Soluble transferrin receptor – mg/L 1.81 (1.28-2.44)

Hepcidin – µg/mL

Tertile 1 – (0-20.08)

Tertile 2 – (20.09-53.00)

Tertile 3 – (53.01- 163.46)

34.9 (17.3-69.2)

10.6 (4.2-15.6)

34.9 (27.1-48.5)

81.2 (69.2-97.9)

*Median and interquartile range (IQR) unless otherwise reported. ICU intensive care unit, +ng/ml

has a conversion factor of 1 to the standard international units mcg/ml

101

Table 2 Tertiles of Hepcidin and RBC transfusion

IV iron No IV iron IRR or mean

difference*

(95% CI)

P value

Median RBC transfusion

Hepcidin 1st tertile (0-20.1µg)

number units/number patients

median (IQR)

23/22

1 (0-2)

35/21

0 (0-2)

0.63 (0.26-1.50)

0.293

Hepcidin 2nd tertile (20.1-53.0µg)


median (IQR)

15/22

0 (0-1)

45/23

1 (0-3)

0.35 (0.16-0.77)

0.009

Hepcidin 3rd tertile (53.0-163.5µg)


median (IQR)

43/22

1 (0-3)

34/23

1 (0-3)

1.32 (0.57-3.08)

0.518

Mean Hb at hospital discharge

Hepcidin 1st tertile (0-20.1µg) 102 (16) 96 (16) 7 (-3-17) 0.181

Hepcidin 2nd tertile (20.1-53.0µg) 107 (14) 107 (17) 0 (-9-9) 0.972

Hepcidin 3rd tertile (53.0-163.5µg) 110 (10) 101 (11) 9 (2-14) 0.010

102

Table 3. Univariate analysis of variables associated with risk of RBC transfusion

Characteristic*

(n=133)

Coefficient (95%CI)

P Value

Age -0.013 (-0.026 - -0.002) 0.027

Male gender 0.148 (-0.401 – 0.697) 0.530

ICU admission type – trauma vs non trauma

0.997 (0.508 – 1.487)

<0.001

APACHE II score -0.001 (-0.40 – 0.038) 0.954

SOFA Score 0.051 (-0.021 – 0.124) 0.165

Renal replacement therapy 0.382 (-0.665 – 1.429) 0.475

Prior RBC transfusion 0.571 (-0.008 – 1.149) 0.053

Haemoglobin -0.016 (-0.038 – 0.006) 0.145

Mean corpuscular volume

0.003 (-0.039 – 0.045) 0.901

C Reactive protein 0 (-0.003 – 0.002) 0.823

Iron 0.061 (0.007 – 0.115) 0.027

Ferritin 0.001 (-0.0 – 0.002) 0.257

Transferrin -0.043 (-0.085 - -0.0) 0.049

Transferrin saturation 0.031 (0.010 – 0.051) 0.003

Soluble transferrin receptor -0.013 (-0.163 – 0.136) 0.860

Thomas plot (soluble transferrin receptor/log

ferritin)

-0.183 (-1.028-0.663) 0.672

Hepcidin (mcg/L) 0.003 (-0.003 – 0.010) 0.286

Received IV iron -0.338 0.188

103

Table 4 Final Multivariate Model – Independent predictors of RBC transfusion

Characteristic*

(n=133)

Coefficient (95%CI)

Risk ratio (95%CI) P Value

ICU admission type

– trauma vs non trauma

0.833 (0.382 – 1.285)

2.30 (1.46-3.61)

<0.001

Haemoglobin >80 g/L

– yes vs no

-0.99 (-1.493 to -0.493)

0.37 (0.22-0.61)

<0.001

Transferrin saturation

– per 10% increase

0.237 (0.082– 0.391)

1.27 (1.09-1.48)

0.003

Hepcidin

– per 10 µg/ml increase

0.086 (0.030 – 0.142)

1.09 (1.03-1.15)

0.002

Constant for model 0.088 (95%CI -0.398-0.575)

Table 5. Iron indices according to hepcidin levels

lowest two hepcidin tertiles Highest hepcidin tertile P value

Iron 3 (2-7) 3 (2-6) 0.363

Transferrin saturation 9 (6-17) 8 (6-16) 0.649

Ferritin 247 (152-424) 263 (177-463) 0.291

Soluble transferrin receptors 1.82 (1.34-2.54) 1.63 (1.24-2.13) 0.225

C reactive Protein 110 (63-170) 70 (36-150) 0.100

Median and interquartile range unless otherwise specified

104

Chapter 7

Conclusion

105

7.1 Thesis overview

This project was designed to test the hypothesis that intravenous iron therapy is effective in

reducing allogeneic RBC transfusion requirement in critically ill patients with anaemia who are

admitted to the ICU. In order to test this hypothesis, the project included a systematic review, a

prospective observational study, a multicentre RCT and a nested cohort study within the RCT.

The main findings of the project are that: 1. IV iron is effective in decreasing RBC transfusion

requirement in a number of acute care settings outside of the ICU, but may increase the risk of

infection, 2. Simple clinical characteristics available early in the ICU admission period can identify

patients at high risk of subsequent RBC transfusion and that, in this setting, standard measures

of iron metabolism may be more useful in excluding iron overload than diagnosing IRE likely to

result in RBC transfusion, 3. IV iron therapy is biologically active in critically ill patients admitted to

the ICU as evidenced by a significant increase in Hb at hospital discharge, but further, larger,

and/or more targeted trials are required to determine whether this is associated with change in

RBC transfusion requirement, incident infection or patient-centered outcome, 4. Baseline

hepcidin concentration may be useful in identifying patients admitted to the ICU in whom IV iron

therapy will reduce RBC transfusion.

106

7.2 Limitations

There are several limitations to this project. First, the central question of whether IV iron in

critically ill patients with anaemia reduces RBC transfusion requirement has not been definitively

answered, primarily because the IRONMAN study was underpowered. The results of the

IRONMAN study suggest biological activity of IV iron in this patient group. However, although the

effect size was within the plausible range estimated from the systematic review, the baseline

RBC transfusion rate was lower than anticipated and the primary outcome was not significant. In

the systematic review, IV iron was associated with a significant increased risk of infection. This

finding may be of particular relevance in critically ill patients. The IRONMAN study did not find a

significant increase in infection risk associated with IV iron. However, the confidence interval

around the point estimate was wide and the sample size was insufficient to exclude a clinically

important difference in either infection risk or severity.

Second, the project tested a specific formulation, dose and timing of IV iron therapy in critically ill

patients and it is uncertain whether the results can be generalised to other dosing schedules or IV

iron formulations. In the IRONMAN study, the majority of RBC transfusion occurred during the

ICU stay rather than after discharge to the hospital ward. However, the results of the nested

hepcidin sub-study suggest that the benefit of IV iron therapy may be modulated by hepcidin

concentration. It is plausible that hepcidin concentration decreases as the acute inflammatory

response diminishes during recovery from critical illness. The timing of IV iron therapy in critically

ill patients may therefore have a substantial impact on the effect and this was not tested in the

IRONMAN study.

Third, although the project was designed to investigate whether IV iron reduced the requirement

for RBC transfusion, data collection in the IRONMAN study was censored at hospital discharge

and a substantial proportion of patients remained anaemic at last follow up. This project did not

address whether IV iron therapy administered early in the ICU admission results in ongoing

effects beyond hospital discharge.

Finally, this project did not directly examine the potential cost effectiveness of IV iron in critically ill

patients with anaemia. Although this was not an aim of the project the issue does bear

107

consideration given that it may be reasonable to introduce into widespread practice a therapy that

is more cost effective if harm can first be excluded.

108

7.3 Significance & Future research

IV iron therapy is increasingly being incorporated in Patient Blood Management Guidelines and

International Consensus Statements on the basis of increasing evidence of allogeneic blood-

sparing effects in non-critically ill populations 158 159. Both anaemia and RBC transfusion are

common in ICU and associated with increased morbidity and mortality. Whether IV iron therapy

has a role in critically ill patients is therefore of substantial public health interest. However, critical

illness is also associated with substantial changes to iron metabolism and the generalisability of

findings of the safety and efficacy of IV iron therapy in non-critically ill patients remains uncertain.

Although this project was not designed to, and did not find, definitive evidence of improvement in

patient-centered outcomes associated with IV iron therapy in critically ill patients admitted to the

ICU, the findings suggest biological activity and provides a pathway and rationale for further

evaluation.

There are several areas of further research to highlight. First, there is a strong rationale to

conduct a multicentre RCT powered for difference in RBC transfusion quantity based on the

baseline event rate found in the IRONMAN RCT and with the eligibility criteria modified to

increase RBC transfusion risk for enrolled patients, for example, by lowering the threshold for

inclusion to an Hb less than 90 g/L rather than less than 100g/L and/or extending the recruitment

window. This RCT should also be powered to exclude a clinically important increase in infection

associated with IV iron. If IV iron was found to decrease RBC transfusion without increase in

infection or other signal of harm, it is reasonable to consider that this therapy may be introduced

on the basis of cost and scarcity associated with allogeneic RBC transfusion alone. Whether or

not prior dose-finding studies are necessary is debatable. Given the biological effect

demonstrated in the IRONMAN RCT and potential risk for oversaturation, a conservative

approach, continuing the same regime may be warranted. In addition, given the potential

synergistic effect of ESA therapy in addition to IV iron demonstrated in the systematic review

(Chapter 2), further consideration should also be given to studying ESA in combination with IV

iron as part of either a two or three-armed RCT (IV iron + placebo vs. IV iron + ESA vs. placebo +

placebo).

109

A larger RCT of IV iron in critically ill patients could also evaluate longer-term functional outcomes

based on the significant increase in Hb found at hospital discharge. There is an unmet need to

identify treatments associated with improvements in functional outcomes of patients admitted to

the ICU. IV iron therapy has potential beneficial effects both on physical and cognitive function

and as such, is a plausible candidate intervention to improve functional outcomes in survivors of

critical illness.

The role of hepcidin in predicting RBC transfusion requirement and response to IV iron is

currently limited by the lack of a widely available, validated, point of care test as well as a detailed

understanding of the effects of interventions, including IV iron, on hepcidin concentration over

time. The development of a validated point of care test would allow hepcidin concentration to be

included as part of screening for future pragmatic RCTs of IV iron therapy, and, if successful,

incorporated into clinical practice.

Finally, the association between iron and infection in critical illness also requires further

consideration. Although a more recent systematic review and meta-analysis did not find a

significant increase in infection risk associated with IV iron, the direction of the estimate favored

harm and the confidence intervals included the point estimate of increased risk found in the

systematic review conducted for this project160. These findings have substantial implications for

the use of IV iron therapy in a number of clinical settings and, together with the lack of significant

decrease in RBC transfusion quantity found in the IRONMAN RCT, suggest that IV iron therapy

in the ICU should only be considered on a case by case basis in patients at low risk of developing

nosocomial infection. Further observational research quantifying free iron concentrations over

time and association both with iron therapy and infective outcomes are warranted.

110

References

1. Westbrook A, Pettila V, Nichol A, et al. Transfusion practice and guidelines in Australian and New Zealand intensive care units. ICM 2010;36(7):1138-46. doi: 10.1007/s00134-010-1867-8

2. Vincent JL, Baron JF, Reinhart K, et al. Anemia and blood transfusion in critically ill patients. JAMA 2002;288(12):1499-507.

3. Corwin HL, Gettinger A, Pearl RG, et al. The CRIT Study: Anemia and blood transfusion in the critically ill--current clinical practice in the United States. CCM 2004;32(1):39-52. doi: 10.1097/01.CCM.0000104112.34142.79

4. Australian Red Cross Annual Report. http://wwwredcrossorgau/annual-reportsaspx 2011 5. Marik PE, Corwin HL. Efficacy of red blood cell transfusion in the critically ill: a systematic

review of the literature. CCM 2008;36(9):2667-74. doi: 10.1097/CCM.0b013e3181844677 6. Hebert PC, Wells G, Blajchman MA, et al. A multicenter, randomized, controlled clinical trial of

transfusion requirements in critical care. Transfusion Requirements in Critical Care Investigators, Canadian Critical Care Trials Group. NEJM 1999;340(6):409-17. doi: 10.1056/NEJM199902113400601

7. Tinmouth A, Fergusson D, Yee IC, et al. Clinical consequences of red cell storage in the critically ill. Transfusion 2006;46(11):2014-27. doi: 10.1111/j.1537-2995.2006.01026.x

8. Leahy MF, Mukhtar SA. From blood transfusion to patient blood management: a new paradigm for patient care and cost assessment of blood transfusion practice. IMJ 2012;42(3):332-8. doi: 10.1111/j.1445-5994.2012.02717.x

9. Hayden SJ, Albert TJ, Watkins TR, et al. Anemia in critical illness: insights into etiology, consequences, and management. AJRCCM 2012;185(10):1049-57. doi: 10.1164/rccm.201110-1915CI

10. Bateman AP, McArdle F, Walsh TS. Time course of anemia during six months follow up following intensive care discharge and factors associated with impaired recovery of erythropoiesis. CCM 2009;37(6):1906-12. doi: 10.1097/CCM.0b013e3181a000cf

11. Dangsuwan P, Manchana T. Blood transfusion reduction with intravenous iron in gynecologic cancer patients receiving chemotherapy. GO 2010;116(3):522-5. doi: 10.1016/j.ygyno.2009.12.004 [published Online First: 2010/01/07]

12. Gaffney-Stomberg E, McClung JP. Inflammation and diminished iron status: mechanisms and functional outcomes. COCNMC 2012;15(6):605-13. doi: 10.1097/MCO.0b013e328357f63b

13. van Iperen CE, Gaillard CA, Kraaijenhagen RJ, et al. Response of erythropoiesis and iron metabolism to recombinant human erythropoietin in intensive care unit patients. CCM 2000;28(8):2773-8.

14. Corwin HL, Gettinger A, Fabian TC, et al. Efficacy and safety of epoetin alfa in critically ill patients. NEJM 2007;357(10):965-76. doi: 10.1056/NEJMoa071533

15. Henry DH, Dahl NV, Auerbach M, et al. Intravenous ferric gluconate significantly improves response to epoetin alfa versus oral iron or no iron in anemic patients with cancer receiving chemotherapy. TO 2007;12(2):231-42. doi: 10.1634/theoncologist.12-2-231

16. Agarwal R, Rizkala AR, Bastani B, et al. A randomized controlled trial of oral versus intravenous iron in chronic kidney disease. AJN 2006;26(5):445-54. doi: 10.1159/000096174 [published Online First: 2006/10/13]

17. II. Clinical practice guidelines and clinical practice recommendations for anemia in chronic kidney disease in adults. AJKD 2006;47(5 Suppl 3):S16-85. doi: 10.1053/j.ajkd.2006.03.012 [published Online First: 2006/05/09]

18. Notebaert E, Chauny JM, Albert M, et al. Short-term benefits and risks of intravenous iron: a systematic review and meta-analysis. Transfusion 2007;47(10):1905-18. doi: 10.1111/j.1537-2995.2007.01415.x

19. Auerbach M, Coyne D, Ballard H. Intravenous iron: from anathema to standard of care. AJH 2008;83(7):580-8. doi: 10.1002/ajh.21154

111

20. Suffredini DA, Xu W, Sun J, et al. Parenteral irons versus transfused red blood cells for treatment of anemia during canine experimental bacterial pneumonia. Transfusion 2017;57(10):2338-47. doi: 10.1111/trf.14214

21. Fleming RE, Ponka P. Iron overload in human disease. NEJM 2012;366(4):348-59. doi: 10.1056/NEJMra1004967

22. Fishbane S. Review of issues relating to iron and infection. AJKD 1999;34(4 Suppl 2):S47-52. doi: 10.1053/AJKD034s00047

23. Maynor L, Brophy DF. Risk of infection with intravenous iron therapy. AP 2007;41(9):1476-80. doi: 10.1345/aph.1K187

24. Oppenheimer SJ. Iron and its relation to immunity and infectious disease. JN 2001;131(2S-2):616S-33S; discussion 33S-35S.

25. Maynor L, Brophy DF. Risk of infection with intravenous iron therapy. AP 2007;41(9):1476-80. doi: 10.1345/aph.1K187 [published Online First: 2007/08/02]

26. Torres S, Kuo YH, Morris K, et al. Intravenous iron following cardiac surgery does not increase the infection rate. SI 2006;7(4):361-6. doi: 10.1089/sur.2006.7.361

27. Heming N, Letteron P, Driss F, et al. Efficacy and toxicity of intravenous iron in a mouse model of critical care anemia*. CCM 2012;40(7):2141-8. doi: 10.1097/CCM.0b013e31824e6713

28. Hill GE, Frawley WH, Griffith KE, et al. Allogeneic blood transfusion increases the risk of postoperative bacterial infection: a meta-analysis. JT 2003;54(5):908-14. doi: 10.1097/01.TA.0000022460.21283.53

29. Locatelli F, Del Vecchio L. New erythropoiesis-stimulating agents and new iron formulations. CN 2011;171:255-60. doi: 10.1159/000327328

30. Mclean E. Worldwide prevalence of anaemia 1993-2005. World Health Organisation 2008 31. Isbister JP, Shander A, Spahn DR, et al. Adverse blood transfusion outcomes: establishing

causation. TMR 2011;25(2):89-101. doi: 10.1016/j.tmrv.2010.11.001 32. Goodnough LT. Iron deficiency syndromes and iron-restricted erythropoiesis (CME).

Transfusion 2012;52(7):1584-92. doi: 10.1111/j.1537-2995.2011.03495.x 33. Hofmann A, Farmer S, Towler SC. Strategies to preempt and reduce the use of blood

products: an Australian perspective. COA 2012;25(1):66-73. doi: 10.1097/ACO.0b013e32834eb726

34. Shander A, Spence RK, Auerbach M. Can intravenous iron therapy meet the unmet needs created by the new restrictions on erythropoietic stimulating agents? Transfusion 2010;50(3):719-32. doi: 10.1111/j.1537-2995.2009.02492.x

35. Lyseng-Williamson KA, Keating GM. Ferric carboxymaltose: a review of its use in iron-deficiency anaemia. Drugs 2009;69(6):739-56. doi: 10.2165/00003495-200969060-00007

36. Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. AIM 2009;151(4):264-9, W64.

37. Friedrich JO, Adhikari NK, Beyene J. Inclusion of zero total event trials in meta-analyses maintains analytic consistency and incorporates all available data. BMC MRM 2007;7:5. doi: 10.1186/1471-2288-7-5

38. Higgins JP, Altman DG, Gotzsche PC, et al. The Cochrane Collaboration's tool for assessing risk of bias in randomised trials. BMJ 2011;343:d5928. doi: 10.1136/bmj.d5928

39. Neeru S, Nair NS, Rai L. Iron sucrose versus oral iron therapy in pregnancy anemia. IJCM 2012;37(4):214-8. doi: 10.4103/0970-0218.103467 [published Online First: 2013/01/08]

40. Adhikary L, Acharya S. Efficacy of IV iron compared to oral iron for increment of haemoglobin level in anemic chronic kidney disease patients on erythropoietin therapy. JNMA 2011;51(183):133-6. [published Online First: 2012/08/28]

41. Aggarwal HK, Nand N, Singh S, et al. Comparison of oral versus intravenous iron therapy in predialysis patients of chronic renal failure receiving recombinant human erythropoietin. JAPI 2003;51:170-4. [published Online First: 2003/05/03]

42. Al RA, Unlubilgin E, Kandemir O, et al. Intravenous versus oral iron for treatment of anemia in pregnancy: a randomized trial. OG 2005;106(6):1335-40. doi: 10.1097/01.AOG.0000185260.82466.b4 [published Online First: 2005/12/02]

112

43. al-Momen AK, al-Meshari A, al-Nuaim L, et al. Intravenous iron sucrose complex in the treatment of iron deficiency anemia during pregnancy. EJOGRB 1996;69(2):121-4. [published Online First: 1996/11/01]

44. Allen RP, Adler CH, Du W, et al. Clinical efficacy and safety of IV ferric carboxymaltose (FCM) treatment of RLS: a multi-centred, placebo-controlled preliminary clinical trial. SM 2011;12(9):906-13. doi: 10.1016/j.sleep.2011.06.009

45. Anker SD, Comin Colet J, Filippatos G, et al. Ferric carboxymaltose in patients with heart failure and iron deficiency. NEJM 2009;361(25):2436-48. doi: 10.1056/NEJMoa0908355 [published Online First: 2009/11/19]

46. Auerbach M, Ballard H, Trout JR, et al. Intravenous iron optimizes the response to recombinant human erythropoietin in cancer patients with chemotherapy-related anemia: a multicenter, open-label, randomized trial. JCO 2004;22(7):1301-7. doi: 10.1200/JCO.2004.08.119 [published Online First: 2004/03/31]

47. Auerbach M, Silberstein PT, Webb RT, et al. Darbepoetin alfa 300 or 500 mug once every 3 weeks with or without intravenous iron in patients with chemotherapy-induced anemia. AJH 2010;85(9):655-63. doi: 10.1002/ajh.21779 [published Online First: 2010/07/28]

48. Bastit L, Vandebroek A, Altintas S, et al. Randomized, multicenter, controlled trial comparing the efficacy and safety of darbepoetin alpha administered every 3 weeks with or without intravenous iron in patients with chemotherapy-induced anemia. JCO 2008;26(10):1611-8. doi: 10.1200/JCO.2006.10.4620 [published Online First: 2008/04/01]

49. Bayoumeu F, Subiran-Buisset C, Baka NE, et al. Iron therapy in iron deficiency anemia in pregnancy: intravenous route versus oral route. AJOG 2002;186(3):518-22. [published Online First: 2002/03/21]

50. Beck-da-Silva L, Piardi D, Soder S, et al. IRON-HF study: A randomized trial to assess the effects of iron in heart failure patients with anemia. IJC 2013 doi: 10.1016/j.ijcard.2013.04.181 [published Online First: 2013/05/18]

51. Bencaiova G, von Mandach U, Zimmermann R. Iron prophylaxis in pregnancy: intravenous route versus oral route. EJOGRB 2009;144(2):135-9. doi: 10.1016/j.ejogrb.2009.03.006 [published Online First: 2009/05/02]

52. Bhandal N, Russell R. Intravenous versus oral iron therapy for postpartum anaemia. BJOG 2006;113(11):1248-52. doi: 10.1111/j.1471-0528.2006.01062.x

53. Birgegard G, Schneider K, Ulfberg J. High incidence of iron depletion and restless leg syndrome (RLS) in regular blood donors: intravenous iron sucrose substitution more effective than oral iron. VS 2010;99(4):354-61. doi: 10.1111/j.1423-0410.2010.01368.x [published Online First: 2010/07/06]

54. Breymann C, Gliga F, Bejenariu C, et al. Comparative efficacy and safety of intravenous ferric carboxymaltose in the treatment of postpartum iron deficiency anemia. IJGO 2008;101(1):67-73. doi: 10.1016/j.ijgo.2007.10.009 [published Online First: 2008/02/01]

55. Charytan C, Qunibi W, Bailie GR. Comparison of intravenous iron sucrose to oral iron in the treatment of anemic patients with chronic kidney disease not on dialysis. NCP 2005;100(3):c55-62. doi: 10.1159/000085049 [published Online First: 2005/04/13]

56. Coyne DW, Kapoian T, Suki W, et al. Ferric gluconate is highly efficacious in anemic hemodialysis patients with high serum ferritin and low transferrin saturation: results of the Dialysis Patients' Response to IV Iron with Elevated Ferritin (DRIVE) Study JASN 2007;18(3):975-84. doi: 10.1681/asn.2006091034 [published Online First: 2007/02/03]

57. Edwards TJ, Noble EJ, Durran A, et al. Randomized clinical trial of preoperative intravenous iron sucrose to reduce blood transfusion in anaemic patients after colorectal cancer surgery. BJS 2009;96(10):1122-8. doi: 10.1002/bjs.6688

58. Evstatiev R, Alexeeva O, Bokemeyer B, et al. Ferric carboxymaltose prevents recurrence of anemia in patients with inflammatory bowel disease. CGH 2013;11(3):269-77. doi: 10.1016/j.cgh.2012.10.013 [published Online First: 2012/10/20]

59. Friel JK, Andrews WL, Hall MS, et al. Intravenous iron administration to very-low-birth-weight newborns receiving total and partial parenteral nutrition. JPEN 1995;19(2):114-8. [published Online First: 1995/03/01]

113

60. Froessler B, Cocchiaro C, Saadat-Gilani K, et al. Intravenous iron sucrose versus oral iron ferrous sulfate for antenatal and postpartum iron deficiency anemia: a randomized trial. JMFN 2013;26(7):654-9. doi: 10.3109/14767058.2012.746299 [published Online First: 2012/11/08]

61. Garrido-Martin P, Nassar-Mansur MI, de la Llana-Ducros R, et al. The effect of intravenous and oral iron administration on perioperative anaemia and transfusion requirements in patients undergoing elective cardiac surgery: a randomized clinical trial. ICTS 2012;15(6):1013-8. doi: 10.1093/icvts/ivs344 [published Online First: 2012/09/04]

62. Grote L, Leissner L, Hedner J, et al. A randomized, double-blind, placebo controlled, multi-center study of intravenous iron sucrose and placebo in the treatment of restless legs syndrome. MO 2009;24(10):1445-52. doi: 10.1002/mds.22562 [published Online First: 2009/06/03]

63. Hedenus M, Birgegard G, Nasman P, et al. Addition of intravenous iron to epoetin beta increases hemoglobin response and decreases epoetin dose requirement in anemic patients with lymphoproliferative malignancies: a randomized multicenter study. Leukemia 2007;21(4):627-32. doi: 10.1038/sj.leu.2404562 [published Online First: 2007/01/26]

64. Hulin S, Durandy Y. [Post-haemodilution anaemia in paediatric cardiac surgery: benefit of intravenous iron therapy]. AFAR 2005;24(10):1262-5. doi: 10.1016/j.annfar.2005.05.023 [published Online First: 2005/07/12]

65. Karkouti K, McCluskey SA, Ghannam M, et al. Intravenous iron and recombinant erythropoietin for the treatment of postoperative anemia. CJA 2006;53(1):11-9. doi: 10.1007/bf03021522 [published Online First: 2005/12/24]

66. Kasper SM, Lazansky H, Stark C, et al. Efficacy of oral iron supplementation is not enhanced by additional intravenous iron during autologous blood donation. Transfusion 1998;38(8):764-70. [published Online First: 1998/08/26]

67. Khalafallah A, Dennis A, Bates J, et al. A prospective randomized, controlled trial of intravenous versus oral iron for moderate iron deficiency anaemia of pregnancy. JIM 2010;268(3):286-95. doi: 10.1111/j.1365-2796.2010.02251.x [published Online First: 2010/06/16]

68. Kim YH, Chung HH, Kang SB, et al. Safety and usefulness of intravenous iron sucrose in the management of preoperative anemia in patients with menorrhagia: a phase IV, open-label, prospective, randomized study. AH 2009;121(1):37-41. doi: 10.1159/000210062 [published Online First: 2009/04/01]

69. Kim YT, Kim SW, Yoon BS, et al. Effect of intravenously administered iron sucrose on the prevention of anemia in the cervical cancer patients treated with concurrent chemoradiotherapy. GO 2007;105(1):199-204. doi: 10.1016/j.ygyno.2006.11.014 [published Online First: 2007/01/20]

70. Kochhar PK, Kaundal A, Ghosh P. Intravenous iron sucrose versus oral iron in treatment of iron deficiency anemia in pregnancy: a randomized clinical trial. JOGR 2013;39(2):504-10. doi: 10.1111/j.1447-0756.2012.01982.x [published Online First: 2012/08/29]

71. Krayenbuehl PA, Battegay E, Breymann C, et al. Intravenous iron for the treatment of fatigue in nonanemic, premenopausal women with low serum ferritin concentration. Blood 2011;118(12):3222-7. doi: 10.1182/blood-2011-04-346304 [published Online First: 2011/06/28]

72. Kulnigg S, Stoinov S, Simanenkov V, et al. A novel intravenous iron formulation for treatment of anemia in inflammatory bowel disease: the ferric carboxymaltose (FERINJECT) randomized controlled trial. AJG 2008;103(5):1182-92. doi: 10.1111/j.1572-0241.2007.01744.x [published Online First: 2008/03/29]

73. Abou Turk C, Williams AL, Lasky RE. A randomized clinical trial evaluating silicone earplugs for very low birth weight newborns in intensive care. JP 2009;29(5):358-63. doi: 10.1038/jp.2008.236

74. Lindgren S, Wikman O, Befrits R, et al. Intravenous iron sucrose is superior to oral iron sulphate for correcting anaemia and restoring iron stores in IBD patients: A randomized, controlled, evaluator-blind, multicentre study. SJG 2009;44(7):838-45. doi: 10.1080/00365520902839667 [published Online First: 2009/03/31]

114

75. Maccio A, Madeddu C, Gramignano G, et al. Efficacy and safety of oral lactoferrin supplementation in combination with rHuEPO-beta for the treatment of anemia in advanced cancer patients undergoing chemotherapy: open-label, randomized controlled study. TO 2010;15(8):894-902. doi: 10.1634/theoncologist.2010-0020

76. Macdougall IC, Tucker B, Thompson J, et al. A randomized controlled study of iron supplementation in patients treated with erythropoietin. KI 1996;50(5):1694-9.

77. Madi-Jebara SN, Sleilaty GS, Achouh PE, et al. Postoperative intravenous iron used alone or in combination with low-dose erythropoietin is not effective for correction of anemia after cardiac surgery. JCVA 2004;18(1):59-63. [published Online First: 2004/02/20]

78. McMahon LP, Kent AB, Kerr PG, et al. Maintenance of elevated versus physiological iron indices in non-anaemic patients with chronic kidney disease: a randomized controlled trial. NDT 2010;25(3):920-6. doi: 10.1093/ndt/gfp584 [published Online First: 2009/11/13]

79. Meyer MP, Haworth C, Meyer JH, et al. A comparison of oral and intravenous iron supplementation in preterm infants receiving recombinant erythropoietin. JP 1996;129(2):258-63. [published Online First: 1996/08/01]

80. Na HS, Shin SY, Hwang JY, et al. Effects of intravenous iron combined with low-dose recombinant human erythropoietin on transfusion requirements in iron-deficient patients undergoing bilateral total knee replacement arthroplasty. Transfusion 2011;51(1):118-24. doi: 10.1111/j.1537-2995.2010.02783.x [published Online First: 2010/07/14]

81. Okonko DO, Grzeslo A, Witkowski T, et al. Effect of intravenous iron sucrose on exercise tolerance in anemic and nonanemic patients with symptomatic chronic heart failure and iron deficiency FERRIC-HF: a randomized, controlled, observer-blinded trial. JACC 2008;51(2):103-12. doi: 10.1016/j.jacc.2007.09.036 [published Online First: 2008/01/15]

82. Olijhoek G, Megens JG, Musto P, et al. Role of oral versus IV iron supplementation in the erythropoietic response to rHuEPO: a randomized, placebo-controlled trial. Transfusion 2001;41(7):957-63. [published Online First: 2001/07/14]

83. Onken JE, Bregman DB, Harrington RA, et al. A multicenter, randomized, active-controlled study to investigate the efficacy and safety of intravenous ferric carboxymaltose in patients with iron deficiency anemia. Transfusion 2013 doi: 10.1111/trf.12289 [published Online First: 2013/06/19]

84. Pedrazzoli P, Farris A, Del Prete S, et al. Randomized trial of intravenous iron supplementation in patients with chemotherapy-related anemia without iron deficiency treated with darbepoetin alpha. JCO 2008;26(10):1619-25. doi: 10.1200/jco.2007.12.2051 [published Online First: 2008/04/01]

85. Pollak A, Hayde M, Hayn M, et al. Effect of intravenous iron supplementation on erythropoiesis in erythropoietin-treated premature infants. Pediatrics 2001;107(1):78-85. [published Online First: 2001/01/03]

86. Provenzano R, Schiller B, Rao M, et al. Ferumoxytol as an intravenous iron replacement therapy in hemodialysis patients. CJASN 2009;4(2):386-93. doi: 10.2215/CJN.02840608 [published Online First: 2009/01/30]

87. Qunibi WY, Martinez C, Smith M, et al. A randomized controlled trial comparing intravenous ferric carboxymaltose with oral iron for treatment of iron deficiency anaemia of non-dialysis-dependent chronic kidney disease patients. NDT 2011;26(5):1599-607. doi: 10.1093/ndt/gfq613 [published Online First: 2010/10/12]

88. Schaller G, Scheiber-Mojdehkar B, Wolzt M, et al. Intravenous iron increases labile serum iron but does not impair forearm blood flow reactivity in dialysis patients. KI 2005;68(6):2814-22. doi: 10.1111/j.1523-1755.2005.00754.x [published Online First: 2005/12/01]

89. Schindler E, Scholz S, Boldt J, et al. [Effectiveness of oral versus parenteral iron substitution in autologous blood donors]. ITM 1994;21(4):236-41. [published Online First: 1994/08/01]

90. Schroder O, Mickisch O, Seidler U, et al. Intravenous iron sucrose versus oral iron supplementation for the treatment of iron deficiency anemia in patients with inflammatory bowel disease--a randomized, controlled, open-label, multicenter study. AJG 2005;100(11):2503-9. doi: 10.1111/j.1572-0241.2005.00250.x [published Online First: 2005/11/11]

115

91. Seid MH, Derman RJ, Baker JB, et al. Ferric carboxymaltose injection in the treatment of postpartum iron deficiency anemia: a randomized controlled clinical trial. AJOG 2008;199(4):435 e1-7. doi: 10.1016/j.ajog.2008.07.046 [published Online First: 2008/10/22]

92. Serrano-Trenas JA, Ugalde PF, Cabello LM, et al. Role of perioperative intravenous iron therapy in elderly hip fracture patients: a single-center randomized controlled trial. Transfusion 2011;51(1):97-104. doi: 10.1111/j.1537-2995.2010.02769.x

93. Shafi D, Purandare SV, Sathe AV. Iron deficiency anemia in pregnancy: intravenous versus oral route. JOGI 2012;62(3):317-21. doi: 10.1007/s13224-012-0222-0

94. Singh H, Reed J, Noble S, et al. Effect of intravenous iron sucrose in peritoneal dialysis patients who receive erythropoiesis-stimulating agents for anemia: a randomized, controlled trial. CJASN 2006;1(3):475-82. doi: 10.2215/cjn.01541005

95. Singh K, Fong YF, Kuperan P. A comparison between intravenous iron polymaltose complex (Ferrum Hausmann) and oral ferrous fumarate in the treatment of iron deficiency anaemia in pregnancy. EJH 1998;60(2):119-24. [published Online First: 1998/03/21]

96. Sloand JA, Shelly MA, Feigin A, et al. A double-blind, placebo-controlled trial of intravenous iron dextran therapy in patients with ESRD and restless legs syndrome. AJKD 2004;43(4):663-70. [published Online First: 2004/03/26]

97. Spinowitz BS, Kausz AT, Baptista J, et al. Ferumoxytol for treating iron deficiency anemia in CKD. JASN 2008;19(8):1599-605. doi: 10.1681/ASN.2007101156 [published Online First: 2008/06/06]

98. Steensma DP, Sloan JA, Dakhil SR, et al. Phase III, randomized study of the effects of parenteral iron, oral iron, or no iron supplementation on the erythropoietic response to darbepoetin alfa for patients with chemotherapy-associated anemia. JCO 2011;29(1):97-105. doi: 10.1200/jco.2010.30.3644 [published Online First: 2010/11/26]

99. Stoves J, Inglis H, Newstead CG. A randomized study of oral vs intravenous iron supplementation in patients with progressive renal insufficiency treated with erythropoietin. NDT 2001;16(5):967-74. [published Online First: 2001/05/01]

100. Toblli JE, Lombrana A, Duarte P, et al. Intravenous iron reduces NT-pro-brain natriuretic peptide in anemic patients with chronic heart failure and renal insufficiency. JACC 2007;50(17):1657-65. doi: 10.1016/j.jacc.2007.07.029 [published Online First: 2007/10/24]

101. van Iperen CE, Gaillard CA, Kraaijenhagen RJ, et al. Response of erythropoiesis and iron metabolism to recombinant human erythropoietin in intensive care unit patients. CCM 2000;28(8):2773-8. [published Online First: 2000/08/31]

102. Van Wyck DB, Martens MG, Seid MH, et al. Intravenous ferric carboxymaltose compared with oral iron in the treatment of postpartum anemia: a randomized controlled trial. OG 2007;110(2 Pt 1):267-78. doi: 10.1097/01.AOG.0000275286.03283.18

103. Van Wyck DB, Mangione A, Morrison J, et al. Large-dose intravenous ferric carboxymaltose injection for iron deficiency anemia in heavy uterine bleeding: a randomized, controlled trial. Transfusion 2009;49(12):2719-28. doi: 10.1111/j.1537-2995.2009.02327.x

104. Van Wyck DB, Roppolo M, Martinez CO, et al. A randomized, controlled trial comparing IV iron sucrose to oral iron in anemic patients with nondialysis-dependent CKD. KI 2005;68(6):2846-56. doi: 10.1111/j.1523-1755.2005.00758.x [published Online First: 2005/12/01]

105. Verma S. Intravenous iron therapy versus oral iron in postpartum patients in rural areas. JSAFOG 2011;3(2):67-70.

106. Warady BA, Kausz A, Lerner G, et al. Iron therapy in the pediatric hemodialysis population. PN 2004;19(6):655-61. doi: 10.1007/s00467-004-1457-5 [published Online First: 2004/04/06]

107. Weisbach V, Skoda P, Rippel R, et al. Oral or intravenous iron as an adjuvant to autologous blood donation in elective surgery: a randomized, controlled study. Transfusion 1999;39(5):465-72. [published Online First: 1999/05/21]

108. Westad S, Backe B, Salvesen KA, et al. A 12-week randomised study comparing intravenous iron sucrose versus oral ferrous sulphate for treatment of postpartum anemia.

116

AOGS 2008;87(9):916-23. doi: 10.1080/00016340802317802 [published Online First: 2008/08/23]

109. Fleming RE, Ponka P. Iron overload in human disease. NEJM 2012;366(4):348-59. doi: 10.1056/NEJMra1004967 [published Online First: 2012/01/27]

110. Goodnough LT, Nemeth E, Ganz T. Detection, evaluation, and management of iron-restricted erythropoiesis. Blood 2010;116(23):4754-61. doi: 10.1182/blood-2010-05-286260

111. Corwin HL, Gettinger A, Fabian TC, et al. Efficacy and safety of epoetin alfa in critically ill patients. NEJM 2007;357(10):965-76. doi: 10.1056/NEJMoa071533 [published Online First: 2007/09/07]

112. Ang O, Gungor M, Aricioglu F, et al. The effect of parenteral iron administration on the development of Staphylococcus aureus-induced experimental pyelonephritis in rats. IJEP 1990;71(4):507-11. [published Online First: 1990/08/01]

113. Moore RA, Gaskell H, Rose P, et al. Meta-analysis of efficacy and safety of intravenous ferric carboxymaltose (Ferinject) from clinical trial reports and published trial data. BMC BD 2011;11:4. doi: 10.1186/1471-2326-11-4 [published Online First: 2011/09/29]

114. Gafter-Gvili A, Rozen-Zvi B, Vidal L, et al. Intravenous iron supplementation for the treatment of chemotherapy-induced anaemia - systematic review and meta-analysis of randomised controlled trials. AO 2013;52(1):18-29. doi: 10.3109/0284186X.2012.702921 [published Online First: 2012/08/11]

115. Stewart GB, Altman DG, Askie LM, et al. Statistical analysis of individual participant data meta-analyses: a comparison of methods and recommendations for practice. PO 2012;7(10):e46042. doi: 10.1371/journal.pone.0046042 [published Online First: 2012/10/12]

116. Litton E, Xiao J, Ho KM. Safety and efficacy of intravenous iron therapy in reducing requirement for allogeneic blood transfusion: systematic review and meta-analysis of randomised clinical trials. BMJ 2013;347:f4822. doi: 10.1136/bmj.f4822

117. Beris P, Munoz M, Garcia-Erce JA, et al. Perioperative anaemia management: consensus statement on the role of intravenous iron. BJA 2008;100(5):599-604. doi: 10.1093/bja/aen054

118. Thomas DW, Hinchliffe RF, Briggs C, et al. Guideline for the laboratory diagnosis of functional iron deficiency. BJH 2013;161(5):639-48. doi: 10.1111/bjh.12311

119. Lasocki S, Longrois D, Montravers P, et al. Hepcidin and anemia of the critically ill patient: bench to bedside. Anesthesiology 2011;114(3):688-94. doi: 10.1097/ALN.0b013e3182065c57

120. Patteril MV, Davey-Quinn AP, Gedney JA, et al. Functional iron deficiency, infection and systemic inflammatory response syndrome in critical illness. AIC 2001;29(5):473-8.

121. Ferrari P, Kulkarni H, Dheda S, et al. Serum iron markers are inadequate for guiding iron repletion in chronic kidney disease. CJASN 2011;6(1):77-83. doi: 10.2215/CJN.04190510

122. Investigators N-SS, Finfer S, Chittock DR, et al. Intensive versus conventional glucose control in critically ill patients. NEJM 2009;360(13):1283-97. doi: 10.1056/NEJMoa0810625

123. Brissot P, Ropert M, Le Lan C, et al. Non-transferrin bound iron: a key role in iron overload and iron toxicity. BEBA 2012;1820(3):403-10. doi: 10.1016/j.bbagen.2011.07.014

124. Hod EA, Brittenham GM, Billote GB, et al. Transfusion of human volunteers with older, stored red blood cells produces extravascular hemolysis and circulating non-transferrin-bound iron. Blood 2011;118(25):6675-82. doi: 10.1182/blood-2011-08-371849

125. Pieracci FM, Stovall RT, Jaouen B, et al. A Multicenter, Randomized Clinical Trial of IV Iron Supplementation for Anemia of Traumatic Critical Illness. CCM 2014 doi: 10.1097/CCM.0000000000000408

126. Litton E, Baker S, Erber W, et al. The IRONMAN trial: a protocol for a multicentre randomised placebo-controlled trial of intravenous iron in intensive care unit patients with anaemia. CCR 2014;16(4):285-90.

117

127. McQuilten ZK, Schembri N, Polizzotto MN, et al. Hospital blood bank information systems accurately reflect patient transfusion: results of a validation study. Transfusion 2011;51(5):943-8. doi: 10.1111/j.1537-2995.2010.02931.x

128. Shander A, Hofmann A, Gombotz H, et al. Estimating the cost of blood: past, present, and future directions. BPRCA 2007;21(2):271-89.

129. Musallam KM, Tamim HM, Richards T, et al. Preoperative anaemia and postoperative outcomes in non-cardiac surgery: a retrospective cohort study. Lancet 2011;378(9800):1396-407. doi: 10.1016/S0140-6736(11)61381-0

130. Authority NB. Patient Blood Management Guidelines:Module 4 Critical Care. ISBN 978-0-9872519-9-2 2012;http://www.nba.gov.au

131. Evstatiev R, Marteau P, Iqbal T, et al. FERGIcor, a randomized controlled trial on ferric carboxymaltose for iron deficiency anemia in inflammatory bowel disease. Gastroenterology 2011;141(3):846-53 e1-2. doi: 10.1053/j.gastro.2011.06.005

132. Munoz M, Martin-Montanez E. Ferric carboxymaltose for the treatment of iron-deficiency anemia. [corrected]. EOP 2012;13(6):907-21. doi: 10.1517/14656566.2012.669373

133. Sharma N, Thiek JL, Natung T, et al. Comparative Study of Efficacy and Safety of Ferric Carboxymaltose Versus Iron Sucrose in Post-partum Anaemia. JOGI 2017;67(4):253-57. doi: 10.1007/s13224-017-0971-x

134. Calvet X, Gene E, AngelRuiz M, et al. Cost-minimization analysis favours intravenous ferric carboxymaltose over ferric sucrose or oral iron as preoperative treatment in patients with colon cancer and iron deficiency anaemia. THC 2016;24(1):111-20. doi: 10.3233/THC-151074

135. MIMS. Ferinject Product Information. MIMS Australia Pty Ltd & CMPMedica Australia Pty Ltd. New South Wales Australia. http: //www-mimsonline-com-au 2012, Accessed: 07/04/2012 2012

136. Anker SD, Comin Colet J, Filippatos G, et al. Ferric carboxymaltose in patients with heart failure and iron deficiency. NEJM 2009;361(25):2436-48. doi: 10.1056/NEJMoa0908355

137. Coyne DW, Kapoian T, Suki W, et al. Ferric gluconate is highly efficacious in anemic hemodialysis patients with high serum ferritin and low transferrin saturation: results of the Dialysis Patients' Response to IV Iron with Elevated Ferritin (DRIVE) Study. JASN 2007;18(3):975-84. doi: 10.1681/ASN.2006091034

138. Patient Blood Management Guidelines: Module 4 Critical Care. NBA ;ISBN 978-0-9872519-9-2

139. Cook D, Lauzier F, Rocha MG, et al. Serious adverse events in academic critical care research. CMAJ : 2008;178(9):1181-4. doi: 10.1503/cmaj.071366

140. Engoren M, Schwann TA, Habib RH, et al. The independent effects of anemia and transfusion on mortality after coronary artery bypass. ATS 2014;97(2):514-20. doi: 10.1016/j.athoracsur.2013.09.019

141. Holst LB, Petersen MW, Haase N, et al. Restrictive versus liberal transfusion strategy for red blood cell transfusion: systematic review of randomised trials with meta-analysis and trial sequential analysis. BMJ 2015;350:h1354. doi: 10.1136/bmj.h1354

142. Litton E, Xiao J, Allen CT, et al. Iron-restricted erythropoiesis and risk of red blood cell transfusion in the intensive care unit: a prospective observational study. AIC 2015;43(5):612-6.

143. Drakesmith H, Prentice AM. Hepcidin and the iron-infection axis. Science 2012;338(6108):768-72. doi: 10.1126/science.1224577

144. Urbaniak GCP, S. Research Randomizer (version 4.0) [Computer software]. http:wwwrandomizerorg/, 2013.

145. Piagnerelli M, Cotton F, Herpain A, et al. Time course of iron metabolism in critically ill patients. ACB 2013;68(1):22-7. doi: 10.2143/ACB.68.1.2062715

146. Carson JL, Terrin ML, Jay M. Anemia and postoperative rehabilitation. CJA 2003;50(6 Suppl):S60-4.

147. Froessler B, Palm P, Weber I, et al. The Important Role for Intravenous Iron in Perioperative Patient Blood Management in Major Abdominal Surgery: A Randomized Controlled Trial. AS 2016 doi: 10.1097/SLA.0000000000001646

118

148. Lelubre C, Vincent JL. Red blood cell transfusion in the critically ill patient. AIC 2011;1:43. doi: 10.1186/2110-5820-1-43

149. Investigators I, Litton E, Baker S, et al. Intravenous iron or placebo for anaemia in intensive care: the IRONMAN multicentre randomized blinded trial : A randomized trial of IV iron in critical illness. ICM 2016;42(11):1715-22. doi: 10.1007/s00134-016-4465-6

150. Lasocki S, Baron G, Driss F, et al. Diagnostic accuracy of serum hepcidin for iron deficiency in critically ill patients with anemia. ICM 2010;36(6):1044-8. doi: 10.1007/s00134-010-1794-8

151. Gummer J, Trengove R, Pascoe EM, et al. Association between Serum Hepcidin-25 and Primary Resistance to Erythropoiesis Stimulating Agents in Chronic Kidney Disease: A Secondary Analysis of the HERO Trial. Nephrology 2016 doi: 10.1111/nep.12815

152. van Rijnsoever M, Galhenage S, Mollison L, et al. Dysregulated Erythropoietin, Hepcidin, and Bone Marrow Iron Metabolism Contribute to Interferon-Induced Anemia in Hepatitis C. JICR 2016;36(11):630-34. doi: 10.1089/jir.2016.0043

153. Steensma DP, Sasu BJ, Sloan JA, et al. Serum hepcidin levels predict response to intravenous iron and darbepoetin in chemotherapy-associated anemia. Blood 2015;125(23):3669-71. doi: 10.1182/blood-2015-03-636407

154. May S, Bigelow C. Modeling nonlinear dose-response relationships in epidemiologic studies: statistical approaches and practical challenges. DR 2006;3(4):474-90. doi: 10.2203/dose-response.003.04.004

155. Prentice AM, Doherty CP, Abrams SA, et al. Hepcidin is the major predictor of erythrocyte iron incorporation in anemic African children. Blood 2012;119(8):1922-8. doi: 10.1182/blood-2011-11-391219

156. Tacke F, Nuraldeen R, Koch A, et al. Iron Parameters Determine the Prognosis of Critically Ill Patients. CCM 2016;44(6):1049-58. doi: 10.1097/CCM.0000000000001607

157. Girelli D, Nemeth E, Swinkels DW. Hepcidin in the diagnosis of iron disorders. Blood 2016;127(23):2809-13. doi: 10.1182/blood-2015-12-639112

158. Patient Blood Management Guidelines: Module 2 Perioperative. NBA 159. Munoz M, Acheson AG, Auerbach M, et al. International consensus statement on the peri-

operative management of anaemia and iron deficiency. Anaesthesia 2017;72(2):233-47. doi: 10.1111/anae.13773

160. Avni T, Bieber A, Grossman A, et al. The safety of intravenous iron preparations: systematic review and meta-analysis. MCP 2015;90(1):12-23. doi: 10.1016/j.mayocp.2014.10.007

119

Appendix 1 – Human Research Ethics Approvals

120

121

122

123

124

125

Appendix 2 – Publications, Presentations and Prizes Publications

1. Litton et al Safety and efficacy of intravenous iron therapy in reducing requirement for

allogeneic blood transfusion: systematic review and meta-analysis of randomised clinical trials

BMJ 2013;347;f4822 Impact Factor 19.7 (Google Scholar 136 Citations), (Other: NEJM Journal

Watch 31/10/2013 http://www.jwatch.org/na32026/2013/10/31/intravenous-vs-oral-iron-patients-

with-anemia, BMJ Rapid Responses http://www.bmj.com/content/347/bmj.f4822/rapid-

responses, Reactions Weekly, Balancing the risks and benefits of IV iron

https://link.springer.com/article/10.1007/s40278-013-5442-2)

2. Litton et al The IRONMAN trial: a protocol for a multicentre randomised placebo-controlled trail

of intravenous iron in intensive care unit patients with anaemia Critical Care & Resuscitation 2014

16(4) 285-90 Impact Factor 3.2 (Google Scholar 14 Citations)

3. Litton et al Iron-restricted erythropoiesis and risk of red blood cell transfusion in the intensive

care unit: a prospective observational study Anaesthesia and Intensive Care 2015; 43(5) 612-6

Impact Factor 1.28 (Google Scholar 3 Citations)

4. Litton RE: The IRONMAN trial: a protocol for a multicentre randomised placebo-controlled trail

of intravenous iron in intensive care unit patients with anaemia Critical Care & Resuscitation 2015

17(2) 144-5 Impact Factor 3.2

5. Litton et al Intravenous iron or placebo for anaemia in intensive care: the IRONMAN

multicentre randomized blinded trial: A randomized trail of IV iron in critical illness Intensive Care

Medicine 2016 42(11) 1715-1722 Impact Factor 10.2 (Google Scholar 3 Citations)

6. Litton et al Intravenous iron or placebo for anaemia in intensive care: the IRONMAN

multicentre randomized blinded trial Critical Care Medicine 2016 44(12) 225

Presentations

1. https://www.youtube.com/watch?v=_m-H9mjplQg The Safety and Efficacy of Intravenous Iron

Therapy in Reducing Requirement for Allogeneic Blood Transfusion: A Systematic Review and

Meta-Analysis of Randomised Clinical Trials BMJ Youtube Channel 15/08/2013

126

2. Best of ANZICS Transfusion in ICU Invited Speaker 7-9/06/2013 Delhi and Mumbai, India

3. Australia and New Zealand Intensive Care Society (ANZICS) Clinical Trials Group (CTG)

Annual Meeting Intravenous Iron for Anaemia in Intensive Care: the IRONMAN Study 07/03/2014

Noosa, Australia

4. Haematology Society of Australia and New Zealand, the Australian & New Zealand Society of

Blood Transfusion and the Thrombosis and Haemostasis Society of Australia and New Zealand

(HAA) The role of intravenous iron in critical illness Invited Speaker 19/10/2014, Perth, Australia

5. Social Media and Critical Care Gold Iron, is it fools gold? Invited Speaker 20/03/2014

Southport, Australia

6. ANZICS Annual Scientific Meeting Iron-restricted erythropoiesis in ICU: A Prospective

Observational Study 11/10/2014, Melbourne, Australia

7. Indian Ocean Rim Laboratory Haematology Congress IRONMAN – IV Iron in ICU Invited

Speaker 15/10/2015 Fremantle, Australia

8. ANZICS CTG Spring Forum IRONMAN RCT Study Results 28/10/2015 Auckland, New

Zealand

9. Australia and New Zealand Intensive Care Society Clinical Trials Group Annual Meeting

IRONMAN Updated Results 8/03/2016 Noosa, Australia

10. European Society of Intensive Care Medicine (ESICM) LIVES2016 The IRONMAN Study

Invited Speaker 2/10/2016 Milan, Italy

11. ESICM Webinar Intravenous Iron or Placebo for Anaemia in Intensive Care: The IRONMAN

Study 10/11/2016

Prizes

Society of Critical Care Medicine Research Snapshot Bronze Award Intravenous Iron or Placebo

for Anaemia in ICU: The IRONMAN Multicentre Randomized Blinded Trial 23/1/2017

Raine Research Prize for best publication by a WA researcher/clinician The IRONMAN

Multicentre Randomized Blinded Trial

Documents

Intravenous Iron for the Treatment of Anaemia in Critical ... · Anaemia in Intensive Care (IRONMAN) RCT was conducted in four centres in Western Australia over a period of two years