One of the biggest problem in generation of commercially viable iPS cell-based therapeutic products is a long cell culture time and high cost of production, which comes with it. Last week, Mahendra Rao and Anthony Atala published a proposal for reducing cost of iPS cell-based therapeutic products. I found this proposal very interesting. Many strategies from their proposed list are based on consensus, free sharing, harmonization and improvement of regulation. Here is some of their cost-reducing strategies:
– Define iPSC as input material
– Develop a DMF that is accessible to investigators
– Use a modular manufacturing protocol and manufacture multiple products from same MCB
– Widen approved use based on function instead of treatment of a particular disease
– Simplify freedom to operate rules and patent licensing
– Harmonize rules internationally and consider common approval
The concept of iPS cell line as input material is the most attractive for allogeneic model (emphasis mine):
The FDA could regulate the manufacture of such input material and even set standards for use. Therapeutic product manufacturers would then choose which line they would use for their product and would have a straightforward change process for replacing one “input” line with another, much as is done today when one changes suppliers for any other material. Suppliers would compete to develop the best quality line or to customize a line for different therapeutic products.
Now, free sharing. The authors propose to have standard protocol for development of iPS cell input material (aka Master Cell Bank in allo- world), which freely shared between manufacturers or exist as commonly used Drug Master File and could be referenced. If this step is standardized, freely shared and regulated, there will be no need for “repeat testing and validation of material suppliers”. The end product (differentiated cells) will be always different between developers and protected by IP. Even at step of “end product”, the authors propose a consensus on functional assay, which will determine the product:
One can imagine developing comparability assays by consensus, such that a product could be defined, not by process, but by function. This has the prospect of enormously reducing costs because no retool- ing to a specific identical process would be required, and no ex- change of competitive data would be required. One would simply agree that these tests, validated against a standard control, would provide the low bar of acceptance for safety and functional utility. Groups could still compete regarding why their product is better using criteria other than safety and minimal efficacy.
In autologous model, the cost could be reduced by so-called “pooled trial”. Since all products could be standardized by functional assay (and, perhaps trials design could be harmonized), the clinical data could be pooled and analyzed once to benefit all developers in the same time:
Samples that are functionally equivalent but manufactured by different groups in different facilities and at different sites are included in the same trial. Because the number of cases will be larger, it is likely that more specific and better information will be available. This will require redesigning trials and their interpretation slightly, but we would argue that this has already occurred for approval of products such as cord blood and no reason exists why this could not be applied to IPSC-derived products. It does, however, require groups to work together and consult with the regulatory authorities to ensure that no unanticipated complications or regulatory infractions will occur.
Finally, Rao and Atala are proposing to alter regulatory system and allow approval not for single indication, but for group of conditions, where the same function of iPS cell-based therapeutic is required (for example function of hematopoiesis recovery or neurotrophic function).
It would reduce the number of trials needed and still provide the option that data would be collected for autologous products, and the use of such products could be reimbursed because the therapy would not be considered experimental.
The authors give example of cord blood BLA process as consolidation of the efforts among all stakeholders:
It will require a concerted effort by all stakeholders to develop a formal, ultimately novel or modified, BLA process for autologous manufactured therapy. Models of this exist, just as does experience. In particular, we believe that the cord blood BLA process offers a good starting point and an example of what can be accomplished when all stakeholders work together to ensure approval.
Actually, I don’t think it’s a good example, since cord blood BLA was not done by consensus of stakeholders, but enforced by FDA and it caused dramatic increase of the cost per licensed unit.
Most of these strategies sound very nice, but are they realistic? I’d remain skeptical about “free sharing”, stakeholders consensus and tweaking regulatory systems. Commercial interest of developers may preclude all those. But some of these strategies could be implemented successfully in some countries. There are some other strategies for cost reduction, widely discussed among iPS cell product developers. For example, Masayo Takahashi mentioned at least two ways to reduce the cost: (1) direct differentiation instead of iPS cell line generation and (2) switching from “auto” to “allo” model.
What do you think about Rao/ Atala proposal? What do you think the best strategy to reduce cost of iPS cell therapeutic product?