The State of the Industry: Cell Therapy Manufacturing and Gene Therapy Post-Conference Report

Fabio D’Agostino, Biopharmaceutical Bioprocessing Technology Centre, Newcastle University, reported from Cell Therapy Manufacturing and Gene Therapy Congress in February:

Manufacturing biological medicinal products by the tonne might sound easy today but, as a recent Genentech infographic explained, the industry struggled to produce even grams of antibodies 30 years ago.

The dream of a second biotech revolution triggered by advancement in cell and gene therapies is driving big and the smaller pharma to find solutions to current manufacturing challenges that affect this new class of medicinal products.

Christopher Bravery (Director, Consulting on Advanced Biologicals Ltd., Member, ISCT EU Legal and Regulatory Affairs Committee, UK) opened the Cell Therapy Manufacturing and Gene Therapy Congress and introduced the first plenary talks from Johnathan Appleby (MDL and CSO, Rare Diseases Gene Therapy, GSK) ,Giovanni Milazzo (Head Regulatory Affairs Advanced Medicines, Chiesi Farmaceutici) and Diego Ardigo’ (Project Leader, Cell and Gene Therapy Products,Chiesi Farmaceutici).  Very different companies but with a very similar challenge: turning data from research in academia and clinical environment into a MAA and therefore a commercially valuable cell medicinal product.

EXPLORE: Join 600+ bioprocessing decision makers, at BioProcess
International European Summit in April

The Challenges GSK Faced in Manufacturing a Gene Therapy

GSK entered a collaboration with The San Raffaele Telethon Institute for Gene Therapy in Italy in 2010 to manufacture and commercialize a gene therapy for the treatment of severe combined immunodeficiency due to adenosine deaminase deficiency (ADA-SCID). Current treatment relies on allogeneic haematological stem cell transplantation (HSCT) from a HLA-matched donor but most patients lack a suitable donor. GSK’s product, referred to as GSK2696273, works in two stages: cells will engraft and then differentiate mainly into T and B cells containing adenosine deaminase gene (ADA).

Johnathan Appleby talked us through the three main challenges they faced:

  • Quality of clinical data available: from the clinical data available, they had to decide what could be used for the MAA (there were also some cases of compassionate use).
  • Pre-clinical model: In terms of safety studies, the struggle was about finding a relevant animal model where these cells could engraft for long enough to test the hazard. Regulators agreed that they did not need to do it after several failed attempts and only because they already had the clinical evidence to mitigate the risk.
  • Manufacturing changes and comparability studies: When GSK entered this collaboration, the product had already been used in 17 patients. They had to make some changes to the process for making the viral vector and to the process for making the gene modified cells. They compared in vitro the vectors and the gene modified cells manufactured with the old and the new process. This extensive comparison should demonstrate comparability without having to run another clinical trial. When considering whether a change requires comparability data the same principles apply for traditional biological products and ATMPs, so a pragmatic approach is to consider the risk to the patient. If this is low, then perhaps it is acceptable to push the clinical demonstration of comparability out as a post approval commitment.

Appleby also explained that although the mechanism of action is mainly provided by the gene modified stem cells, these need to be supported by non-gene modified cells which will improve engraftment and time to recovery. The active ingredient has been defined as “an autologous CD34+ enriched cell fraction that contains CD34+ cells transduced with retroviral vector containing the adenosine deaminase gene”.

Later on during the panel discussion, he commented on the difficulty in exploring dose ranges, which is a requirement for the MAA. Given the limited number of patients for rare diseases, it was difficult to draw significant conclusions in terms of dose. As differences in the amount of cells and vector copy number were present in the manufactured product, it was possible to investigate indications of dose response relative to T-cells recovery and measurement of engraftment. With the small data set available, there were some trends emerging that could be confirmed after approval. Interaction with regulatory authorities was key and Appleby emphasised that this is not an industry where we can rely on “as we did last time”.

The Hurdles Faced Taking Holoclar to Market

Giovanni Milazzo and Diego Ardigo’ from Chiesi Farmaceutici described the “hurdles to jump” when in 2008 they decided to “re-engineer” 20 years of research and clinical data from 219 patients treated in 21 clinical centres in Italy into GxP data to submit an MAA for Holoclar (a medicinal product comprising ex vivo expanded autologous human corneal epithelial cells containing stem cells).

Giovanni explained their strategy was mainly based on a gap analysis:

1) What are the gaps that we need to fill?
2) How can we fill those gaps?
3) Plan and risk-based assessment to decide which gaps needed to be filled prior to the MAA and what could be instead accepted as post-approval mitigation action?

In terms of the manufacturing process, they had to fill three gaps:

  • Raw material selection: They selected TSE compliant raw materials and materials compliant with Eu. Pharmacopeia, secured technical agreement with suppliers, switch to irradiated serum and validate the lethal irradiation of J-3T3 feeder layer.
  • Release specification for drug substance and drug product: The biggest challenge was to validate some additional analytical methods with autologous product. Regulators agreed that this validation study can be done with starting material from cadaveric corneas as opposed to healthy donors.
  • Stability studies: The introduction of an intermediate cell bank allows more flexibility for the scheduling of the recipient patient as the manufacturing process can be halted. However, the price to pay was a double stability study on the cryopreserved intermediate cell bank and during transportation and use of the product.


(Image courtesy of Giovanni Milazzo)

Diego Ardigo explained how from a retrospective clinical trial, they were able to answer many of the questions that you would normally have at the beginning of the clinical development of the medicinal product. This was mainly thanks to the excellent work done by two clinical centres in Italy where data was collected in a very similar way to how it is normally done in standard clinical trials. However, there were still two gaps from a clinical point of view:

  • Lack of prospective clinical data to confirm long-term safety and efficacy and positive post-corneal transplantation outcome
  • No patients were treated outside of Italy

For these reasons, they applied and received conditional approval. The two post approval conditions to fulfill in order to get the full MA are:

  1. Prospective confirmatory study with longer-term follow up, including post-keratoplasty (Holocore)
  2. Real world evidence collected in an observational registry (Holosight)

Ardigo wrapped up his talk by describing the logistic hurdles in reaching all 28 Member States with a 36h shelf life and a variable manufacturing time based on cell growth.

Accelerated Market Approval for ATMPs

An interesting panel discussion covered an option available to companies seeking market approval for ATMPs: the accelerated assessment. All MAAs must reach a positive or negative opinion within 210 days’ active assessment time. To be accelerated this is squeezed to 150 days, meaning that when the clock-stops on day 120 there is no time for significant issues or major objections to be addressed. Given the complexity of these products it is very difficult to stay on the accelerated assessment path.

Christopher Bravery put forward the question: is Conditional Market Authorization good just for the patients or also for the company?

Ardigo commented that from a regulatory point of view, conditional approval is also a great tool for the company, but the real problem is then market access for that ATMP. Payers normally consider the incomplete set of data a significant limitation on the completeness of the price and reimbursement dossier. As long as a mechanism for conditional reimbursement is not available, balance between regulatory advantage and price discussion consequences should be well assessed upfront.

Pricing Cell & Gene Therapies

The topic of reimbursement for this new class of medicinal products was the centre of another panel discussion. Innovation is arguably not just needed in manufacturing, but also in the way these new drugs are evaluated and reimbursed. Alec Orphanidis (SVP Global Commercial Operations at uniQure) explained why an increasing number of experts working in this industry believe that traditional methods of assessing the value of pharmaceutical products should not be applied to cell and gene therapies, or to drugs for rare and very rare diseases. As these products are meant to be administered only once, with multiple years of expected benefit, this is more similar to surgical intervention than to pharmaceutical products.

Standard HTA assessment cannot be adapted to this treatment model as at the time of the reimbursement request we will not always have very long term data that could prove cost effectiveness. In addition, with respect to rare diseases, these indications by definition only involve limited numbers of patients. The drugs used to treat these diseases (whether they are cell or gene therapies, or more traditional products) should be assessed by different criteria than traditional molecules, because statistical significance might not be reached when it comes to the economic value calculations. Finally, this problem is compounded by drugs for “ultra-rare” diseases, given that there are far fewer patients than for rare diseases.  So, the challenges become even greater.

This is not good news for many companies who, unlike GSK and Chiesi, are now approaching clinical development. Early stages are normally characterized by high uncertainty and rather limited capital. Ohad Karnieli (former VP Manufacturing and Technology at Pluristem, now founder and CEO of Karnieli Ltd) shared many of the lessons learned from his experience at Pluristem and how to tackle many critical points that companies normally face in this industry.

One of his key recommendations was to start with your Target Product Profile (TPP) and use it as a “living document” to guide your product and process development. This will enable you to solve critical points in manufacturing very early during product development to maximize the chances of success. For example, if the expected batch size required for commercial use is not achievable with available 2D culture systems, it is advisable to start with a bioreactor (microcarriers based or packed-bed) and less starting materials. Moving to 3D culture systems only after phase II can be very risky as the two products might not be comparable. However, most current methods are not sustainable for commercial productions and new ways to process cells are needed (but more on this in the next blog).

Day 1 closed with a Christopher Bravery’s plenary talk on justification of specifications. Bravery stressed the importance of product and process characterization to define adequate justifications and how this is a “vital” requirement for MAA/BLA. As described in ICH Q6B, specifications for Drug Substance and Drug Product are set to control and confirm the quality of the product for a specific qualified/validated process and this is part of an overall control strategy. As the product progresses through clinical development, more characterization data becomes available and justifications of specifications should become more comprehensive.  For market approval, specifications need detailed justifications and most likely will require data beyond the averages from batch records. Where many developers seem to struggle most is potency, which is arguably the most important of the product specifications.

READ MORE: Part Two of Fabio’s report

Fabio D’Agostino is a passionate life sciences professional with experience in both the medical device and biopharmaceutical industry. An active member of the PDA Cell and Gene Task Force, he has contributed to a number of conferences in the cell and gene therapy industries. He was also instrumental in the launch of the new journal: Cell and Gene Therapy Insights.

After graduating with Honours from the Polytechnic University of Turin (Italy) with a BSc and a Master’s in Biomedical Engineering, he started his career at LivaNova (formerly Sorin Group) before moving to Newcastle University to take an Engineering Doctorate in Biopharmaceutical Process Development. He currently holds a research position at the Institute of Genetic Medicine (Newcastle University) where he is responsible for the development of an innovative platform for modular tissue engineering.

Stay up to date with the latest news and opinions on cell and gene therapies at

Leave a Reply