1. Cell Therapy… how to move from Heroic Medicine to Clinical Reality
Mark Briggs, PhD
Cell Therapy Technologies
GE Healthcare Life Sciences

2. An increasing percentage of healthcare dollars will be spent on new medicines
Key Trends
Increasing personalization of medicine
Spending on specialty medicines to reach $ billion in 2017, with total aggregate spending on medicines to exceed $1.2 Trillion (IMS Health)
Cancer cellular immunotherapy expected to reach $20 Billion by 2020
Patients treated by therapy type
2005 ($650B)*
2025 ($1,650B)
What market looked like 9 years ago and the relative distribution of patients treated by therapy type
In 2014, small and large molecule therapies still dominate….
In 2025, we can predict that the % patients treated with cell & gene therapies would increase, with approvals, and this number would translate into 1000’s, possible 100K’s of dose…
The point here is that we (the industry) will need an infrastructure to support the manufacture of these therapies….a lot of work ahead of us..
In 2015, 2% translates into 1000 – 100,000’s of doses
How big are we going to be in 2025?
Industry needs to deliver this large #of dose
2015 ($1,100B)
Creating a need for a strong cell therapy manufacturing ecosystem
Source: Estimates based on data from ClinicalTrials.gov, Future Medicine, IMS Reports and internal research.
* Prescribed medication healthcare costs

3. Fundamental changes required …
“The spiralling cost of new drugs mandates a fundamentally different approach to keep life-saving therapies affordable for cancer patients. We call here for the formation of new relationships between academic drug discovery centers and commercial partners, which can accelerate the development of truly transformative drugs at sustainable prices.”
Workman et al. Feb 2017

4. Therapeutic evolution has been driven by biological discovery
Present Day
Pre- 1900’s
Small Molecules
Biologics (e.g. monoclonal antibodies)
Engineered Nucleic Acids (e.g. RNA)
Natural Products, Blood Products
Replacement Therapies (e.g. Insulin)
Viruses, Viral-like particles, Gene
Therapies
Cells, tissues, & organs
Combination therapies & regenerative medicines
NOTE: Not drawn to scale

5. Cell Therapy: The administration of live whole cells or maturation of a specific cell population in a patient for the treatment of a disease.
Source:
What is Cell Therapy?
It is not……
Regenerative medicine: A branch of translational research in tissue engineering and molecular biology which deals with the "process of replacing, engineering or regenerating human cells, tissues or organs to restore or establish normal function".
en.wikipedia.org/wiki/Regenerative_medicine

6. So, is cell therapy really new?
Late 19th Century - Charles-Édouard Brown-Séquard injected animal testicle extracts in an attempt to stop the effects of aging.
1931 - Paul Niehans injected parathyroid cells from a calf to save the life of a patient who was in critical condition. The patient recovered and lived on for more than 25 years.
E. Donnall Thomas - first human bone marrow transplant between identical twins, one of whom had leukaemia.
Source: Wikipedia.org

7. Why the renewed interest in cell therapy?
2012
Diagnosed with ALL 2010
Emily was diagnosed with acute lymphoblastic leukemia (ALL) at age five in May 2010 and relapsed twice. After the second relapse, the Whiteheads were told they were out of options to treat her cancer. Not willing to give up, the Whiteheads pursued a radical new treatment called T-cell therapy at Children’s Hospital of Philadelphia (CHOP).

8. Autologous Cell Therapy: Personalised Cures
1. Cell Extraction
5. Therapeutic Delivery
Hospital/Clinic
2. Selection/
Activation/ Modification
4. Therapy Dose Creation
Centralized Manufacturing Facility
3. Cell Expansion 8

9. Allo vs. Auto Cell Therapy Production: Business models require different approaches
Allogeneic Universal Donor
Autologou s Patient Specific
Scale Up –
Large Lot Sizes
Scale Out –
Massive Parallelism
High start-up costs
Master & working cell banks
Production facilities and equipment
Comparability
Process control
At-scale Risk
Large batch failures
Distribution
Logistics
Inventory management
Precious starting material
One “shot on goal”
Non-destructive testing
Chain of custody
Patient heterogeneity
Process variability
Unpredictability
High testing costs
Massively-parallel processes
Integration
Scalability

10. Juno ditches lead CAR-T program following patient deaths
Juno Therapeutics has halted development of its lead CAR-T therapy months after patient deaths hindered the odds of success. The decision sees Juno switch its focus to JCAR017, a follow-up program that uses a defined CD4-CD8 composition company execs thinks will improve safety and efficacy. JCAR015 has appeared destined for the chop since patients died of cerebral edemas either side of a FDA clinical hold last year. But Juno held off on publicly ditching development of the CAR-T in adults with acute lymphoblastic leukemia (ALL) until this week. That delay enabled Juno to dig into the causes of the safety failings — which it thinks stem from rapid early CAR-T cell expansion — and set JCAR015 back to the extent that further work on a flawed candidate was deemed pointless.
Source: fiercebiotech.com 02 Mar 2017

11. Gartner ‘Hype Cycle’
By Jeremy Kemp at English Wikipedia, CC BY-SA 3.0,

12. Process development requires deep process understanding and control
“Apprentice” model
Process expertise
Cell “look and feel”
Focus on efficacy and safety
“Scientific Management” model
Managed processes
Sample measurement
Lean & Six Sigma reproducibility
Focus on productivity & risk

13. What about production?
Generic Workflow
Repurposing bioprocess tools, blood processing tools, and/or basic research platforms
Complex non-turnkey systems, limited standardisation
Poor interconnectivity of workflow components
Poor IT connectivity across workflow
Scalability  regulatory implications
Immature/incomplete supply chain
Etc.
Scale manufacturing process in an economically viable way for large patient populations

14. Therapeutic cell production today: Manual, disconnected & open process
Currently the process is very manual from connecting various tube sizes, moving products between systems and areas, measuring out small doses of media… There are many areas where a process break can occur, and the impact is huge.
Images: gmpbio.org biosynergyeurope.com MIT Technology Review 118 (4) (2015)

15. Unit Operations in the T-cell therapy workflow
COLLECTION
MODIFICATION
EXPANSION
HARVESTING
THAWING & PREP
Biosafe Sepax
Media
Biosafe Sefia
SCT reagents
Xuri Bioreactor,
Media & Reagents
Biosafe Sefia
Biosafe
Smart-Max

16. Aseptic physical connectivity – trivial?

17. Coping with variability?D E
Predictable Process A A C D E
Shorter – issue? A A C D E
Longer – issue?
Aborted –
Impact on patient? A A C A A C D E
Ideal
Time
Planned
availability

18. In silico process learning and monitoring
What are the required resources to process X patient samples per year that meet specified targets (cycle time)? 1
Ask the Question(s)
Process Modeling
Run Model
Analyze Results
Decision 2
Generic T-Cell Therapy Workflow 3
Key Assumptions:
Process steps, resource requirements and task cycle times are defined
# of samples per day, per week, per year
Define constraint(s)
Daily shifts are defined for the Operators and Verifiers
Task times have no variability
Etc. 4
Iterate 5
Hypothetical process simulation model

19. Smart cell processing: The Vision
January 14, 2018
Smart cell processing: The Vision
Patient
Product
HW & SW agnostic connectivity
Context aware data aggregation and feature extraction
Unit operation analytics based on domain expertise
Industrial Data Lake for Big Data storage
Process wide Big Data analytics
Prescriptive insights for process outcomes
Raw material
Subsystem
Process
Subsystem
Process
Subsystem
Process
Subsystem
Process
Subsystem
Process
Subsystem
Process
Subsystem
Smart Bioprocessing Analytics subsystem(s)
Industrial Data Lake
HW & SW Agnostic connectivity: Not specific to GE equipment or software
Feature extraction: Describe continuous curves (e.g. Chromatograms) by discrete points capturing the shape of curves
Descriptive statistics: What happened?
Predictive statistics: What will happen?
Prescriptive statistics: Make it happen the way you want
Process wide Critical Quality Attribute profile

20. Summary
Clinical development and industrialisation is occurring in lock-step but have very different metrics
Desperate need for process development understanding
Improved process measurement and management
Process risk can be minimized by:
Simplifying processes
Designing new unit operations
Developing a robust supply chain
Connecting
Digitising
Digital integration can be a powerful tool for process and facility design and optimization
Smart analytics will lead to better and safer medicines

21. So, industrialising cellular therapies…
Is like: “Building the plane whilst flying it” (Phil Vanek)
You must:
“Keep the end in mind” (Marc Turner)
Image: EDS.com
And it is :
“All about knowledge management – process development requires innovator and manufacturer collaboration” (Anne Black)
Image: changingwinds.wordpress.com
Image: mediafactory.org.au

22. “I find out what the world needs, then I proceed to invent it
“I find out what the world needs, then I proceed to invent it.” -Thomas Edison 22

23. GE and GE monogram are trademarks of General Electric Company
GE and GE monogram are trademarks of General Electric Company. Imagination at work, Predix, Xuri, are trademarks of General Electric Company or one of its subsidiaries.
©2017 General Electric Company—All rights reserved. First published March 2017
Xuri Cell Expansion Systems W5 and W25, Predix are not a medical devices nor CE marked and should not be used in diagnostic processes.  Drug manufacturers & clinicians are responsible for obtaining the appropriate IND/BLA/NDA approvals for clinical applications.
GE Healthcare UK Limited, Amersham Place, Little Chalfont, Buckinghamshire, HP7 9NA, UK 23

24. Doctors at GOSH baffled when a routine cancer treatment failed
8 Young patients – 4 died
Investigation by Michael Watts (UCL)
GOSH had followed current best practice & established protocol (10 years).
Standard viability check upon thaw
"They looked absolutely superb. They were in good shape"
Not routinely functionally tested - weeks vs hours
"We were astonished when we got the function test results. The cells didn't grow properly.”
Sophie Ryan-Palmer died a month after receiving the treatment
Source: bbc.co.uk/news/health Accessed 09 June 2016

25. The wheels keep on turning…
Cellectis Gets U.S. Go-Ahead to Test 'Off-the-Shelf' Cell Therapy Feb 2017: Cellectis has won U.S. regulatory approval to run an early clinical trial using its gene edited cell therapy product UCART123 for blood cancers. Its first such "off-the-shelf" cell therapy UCART19, which is being developed with Servier and Pfizer, is now being tested in Phase I trials in Britain for acute lymphoblastic leukaemia and chronic lymphocytic leukaemia. It has already rescued two babies treated at London's Great Ormond Street Hospital from previously incurable cancer.
Celyad Registers First Hematological Patient In CAR-T NKR-2 THINK Trial Feb 2017: Celyad’s lead candidate, CAR-T NKR-2, is a CAR-T-Cell engineered to express the human NK receptor, NKG2D, which is an activating receptor that triggers cell killing through the binding of NKG2D to any of eight naturally occurring ligands that are known to be overexpressed on more than 80% of tumors. The platform targets a wide range of solid and hematological tumors. Unlike traditional CAR-T cell therapy, which target only one tumor antigen, Natural Killer (NK) cell receptors enable a single receptor to recognize multiple tumor antigens

26. CAR T Cells More Powerful When Built With CRISPR Feb 2017: In a study published in Nature, researchers from Memorial Sloan Kettering Cancer Center reported that directing a CD19-specific CAR to a precise location in the genome of T cells using the CRISPR technology resulted in uniform CAR expression in human peripheral blood T cells and enhanced T-cell potency. The edited cells vastly outperformed CAR T cells generated using γ-retroviral vectors or other randomly integrating vectors in a mouse model of acute lymphoblastic leukaemia.
Draper Looks to Ultrasound and Geometry to Reduce CAR-T Manufacturing Costs Feb 2017: The Charles Stark Draper Lab has developed two microfluidic prototypes to help make the manufacturing of cell therapies cheaper and more efficient. The devices address two key steps of the CAR T-cell therapy manufacturing process: the first uses an innovative acoustics approach to enrich the T-cells in a blood sample, and the second provides a new approach to gene transfer.

27. Patients Treated Per Batch
Whilst we can learn from bioprocess manufacturing, we are forging a new path
Similarities to bioprocess manufacturing
Requires cell culture technology
Regulated as a drug
Manufactured in sterile facilities
Scaled production with automation
Unique to Cell and Gene Therapy (Immunotherapy)
Mass customization/parallel processing
Critical custody management
Care pathways connected throughout process
Flexible factories (multiple products through same/similar line)
May persist - potentially curative
Complex distribution driven by product shelf-life
$$$
Manufacturing Cost per Dose
Biologics
Viruses
Cells
Small Molecules $
Patients Treated Per Batch
Cell manufacturing shares many features with the biologics industry. But the complexity of the cell versus the complexity of a protein or antibody, introduces some real production challenges. Cell viability, patient custody, cell characterization all are hurdles to be addressed.
 Scale up
Scale out 

28. The industry needs tool providers to provide de-risking technologies
Unit Operations
Physical Integration
Digital Integration
Scalable
Easy to use
Smart - sensors embedded
Modular
Completely closed
Low dead-space
Exchangeable unit operations
Manufacturer independent
Facility independent
Plug & play
Unit operation agnostic
Sensor enable everything
Data dumps
Connectivity through private clouds
Data lakes for analytics
Same as last slide.