Upload
others
View
5
Download
0
Embed Size (px)
Citation preview
Improving Manufacturing in the
Pharmaceutical Industry through the
Implementation of Continuous Processing
Sarah Hunter, Process Engineer, R&D
Batch Processes have and will continue to fulfill an important role in the manufacture of Active Drug Substance; they allow multiple recipes to be executed in the same core equipment. However, the compromise is, batch fails to deliver the precision, speed of manufacture and intensification of a flow process. Here I present how, at GSK, we are developing flow processes for organometallic chemistry. Using process understanding generated using batch methods, and a combination of insilico simulation and physical demonstration we are able to redefine conventional operating parameters for this type of chemistry and industrialise processes which are not feasible in batch due to risk of scale. We believe this will allow us to deliver more medicines of value to patients, both in affordability and function
1
Flow Chemistry at GSK
Where We Are
2
We want a Manufacturing solution with these attributes...
2003 Started Programme
2016 Gain First Commercial Credibility for 3
Stage Process.
Clinical Campaign of Organometallic Flow
Process Case Study.
2017-2021 Design/Build the Space to house the
next Phase of Projects
Route
selection
Training
New technology
development and
implementation
Process
understanding
Kinetic
experiments
Calorimetry
design
Reactor
library
industrialisation
Analytical
techniques
New
simulation
platform
New
guideline
Strategic
partners
Modular
technology
Process
Simulation
Flow Chemistry at GSK
Pharma has been working on a knowledge and trial and error principal. Development has
been governed by statistical models and small scale increment changes until manufacturing
scale.
GSK is partially converting its API batch portfolio to continuous with a great ambition to reach
up to 50% by 2020.
To support this strategy, simulation has to be integrated at the centre of process development
to;
Accelerate process development
Provide more robustness in the development phase
Ensure and understanding from scale to scale
Batch is following the transition and integrate simulation as part of the workflow.
Where We Aim to Be
3
Organometallic Chemistry
Organometallic chemistry is common place in pharmaceutical synthesis and often presents a challenging
process to execute in batch.
Batch Process
4
E-1
Organometallic Reagent Substrate
Batch is….
Versatile.
Available across
scales.
Familiar to
chemists
Batch is also….
Imprecise.
Hard to scale up.
Labour intensive and
Slow.
Organometallic Chemistry
Flow Process
5
Low Inventory System
Limited
Intermediate
Hold Times
A team at GSK have been developing process workflow for effective design and scale up of
continuous organometallic processes which is currently being applied to a number of
processes in active development.
Minimised
Reagent
Handling
Effective Mass Transfer
How do We Allow the Transition to Happen From Laboratories to Manufacture
6 6
Process Understanding
Industrialisation
Demonstration
Case Study P38 Kinase Inhibition of Inflammatory Pathways
– Rheumatoid Arthritis – Candidate selected 2003
– Atherosclerosis
– Depression
– Neuropathic pain
– Acute Coronary Syndrome - PhIII
– COPD
– Focal Segmental Glomerular Sclerosis
– Stage 1 – Flow Synthesis
Case Study Flow Stage Development
– Evaluation of solubilities, rates and
stabilities required
– Kinetic model to define reactor
characteristics
Flow Batch
0.00
10.00
20.00
30.00
40.00
50.00
60.00
70.00
80.00
90.00
100.00
0 500 1000 1500
%
Time min
Aryl Grignard Stability with Temperature
30°C
40°C
50°C
0.9eq ofiPrMgCl
Process Understanding
Kinetics & Calorimetry
Calorimetry
Deprotonation
Exchange
Kinetics
Main Reaction
Degradation – Batch Time Course
Simulation PFR Model
10
Matching ;
– Temperature profile
– Residence times
– Reagent concentrations & flow rates
Deprotonated
Aryl Grignard
Methyl Grignard
Starting Material
Define process parameter ranges and
the equipment performance;
– Heat Transfer Coefficient
– Surface Area to Volume Ratio
– Micro mixing
– Bodenstein Number
The simulation demonstrates process robustness low sensitivity to parameter inputs allowing
the simplest design concept.
Plug flow reactor with static mixer;
User Requirement Specification
Continuous Process
11
L1 L2
Mixing Zone
τmix<0.5s
Residence Time Zone
τ<0.5s
U ≈ 0 W/m2.K (adiabatic)
Demonstration PFR Model
12
1
2
HXN
3
Demonstration
MeMgCl
feed
Temp
probe
To iPrMgCl
“reactor” – Heat Exchanger
solution feed Temp
probe
Continuous Process
Demonstration
Process Control
– Initial intent: to flow directly into a quench pot containing the trimethyl borate
– Extended addition time gave higher levels of the borinic acid
Work Up
18
Addition time (min) Borinic acid level
HPLC (%a/a)
0.5 1.3
3.5 4.2
8 7.9
Demonstration
Work Up
19
Demonstration
• If k1 and k3 fast but k2 slow?…..
• Rate of addition relative to progression of the “ate” complex controls borinic
levels
• Rapid trimethyl borate addition to the Grignard solution preferable
Simulation Design Space and QbD
1.051.15
1.251.35
1.45
90
92
94
96
98
100
0.00.1
0.2
0.3
0.4
Methyl Grignard charge [eq]
yield [%]
residence time [s]
98-100
96-98
94-96
92-94
90-92
1.05
1.2
1.35
1.5
0
0.01
0.02
0.03
0.04
0.00 0.10 0.20 0.31 0.41Methyl Grignard
charge [eq]Residence time [s]
0.03-0.04
0.02-0.03
0.01-0.02
0-0.01
Impurities [%]
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
1 1.1 1.2 1.3 1.4 1.5
yield
des Bromo
retro Michael
Hexyl Imp
iPr Grignard concentration [mol/L]
Gellingobserved
0.0
0.1
0.2
0.3
0.4
0.5
20 25 30 35 40
yield
des Bromo
retro Michael
Hexyl Imp
CSTR temperature [ C]
Gellingobserved
The simulation is a critical part of
development facilitating;
– Process design
– Risk evaluation
– Specification of deign space
Industrialisation
Scale Independent Design
21
Mixing
Zone
Residence
Time Zone
Scale Up
ν1
ν1
Re
Re
COV
COV
Mixing
Zone
Residence
Time Zone
U
U
Bo
Bo
Industrialisation In Practice
Processes have been demonstrated at production rates of;
• Lab scale ~100g/hr
Isolated Yield 70% th
Isolated Quality 99.7% a/a
• Pilot plant scale ~500g/hr.
Operating as a flow process
• Enabling operating conditions which are not feasible as a
batch operation.
• Eliminating a potential critical quality attribute.
A recent pilot plant campaign
produced 88.2 kg of clinical material.
Yields averaged 84.95% th.
Acknowledgements
Continuous Primary Team
Flavien Susanne, Andrew Rutter, Charles Wade
Project Teams
Robert Whitten, Thoralf Hartwig, Mark Schilling, Lindon Francis, Kathryn Southgate
David Pascoe, Mike Webb, Jason Cooke, Sophie Duffield, Peter Skett, Laura Palmer
23