The blunt point: ROI follows clinical success, and cancer vaccines have a long history of “great story, zero payoff.” Tal explains why the Moderna PCV data in melanoma is different, and why trial design is the real lever.
Q1
ROI Black Holes: We’ve moved from “Can mRNA work?” to “Can it be a business?”
Which part of the personalized cancer vaccine value chain manufacturing, logistics, or clinical trial design do you see as the biggest "black hole" for ROI today?
A:
I love this question. The return on investment is a direct function of the probability of clinical success. Throughout my professional career that probability has been zero and nobody has ever gotten a positive ROI on a cancer vaccine. My first project as a post doc with Steve Rosenberg 30 years ago was to develop phase one cancer vaccine program, and a couple of years later we even published on a nascent mRNA cancer vaccine that worked in a mouse model. While some of these mice models predicted the ability to generate reactive T cells in patients, none translated into clinical efficacy.
So why is this time different and I'm referring to the randomized phase 2 an adjuvant melanoma of Moderna's personalized cancer vaccine. I think for three reasons. First, is that we've learned that when the immune system does see an antigen in the context of cancer, it is far far more likely to be a neo epitope than it is to be encoded by a non-mutated protein. And we now have both the sequencing tools and bond for attic tools to be able to quickly analyze a patient's tumor for those new epitopes that are most likely to bind to that patient's HLA.
The second is that mRNA vaccine provide a uniquely potent T cell vaccine. It is worth noting that this is not by our clever design but rather by evolution. When you concatenate 30 or more short peptides and one long mRNA chain, what comes off the ribosome is a misfolded protein that you would expect get preferentially shuttle to the protozoal machinery and displayed for recognition because this is how the immune system has evolved T cells to recognize RNA virally infected cell cells.
Recognizing these two points and the complete lack of predictability of urine models when it comes to cancer vaccines is why we advanced Moderna's cancer vaccine to the clinic without attempting to cure mice first finally, the third factor is the clinical context. We very carefully chose the first study to be an adjuvant melanoma in combination with Keytruda. The heavier, mutation burden in melanoma suggests that it is the place where we are most likely to pick up relevant new epitopes; the proven activity of Keytruda in adjuvant melanoma combined with solid pre-clinical data suggesting that PD one blockade can enhance vaccination made this the obvious combination.
Finally, treatment in the edge of event, setting provides us with the smallest tumor burden, and the longest time window during which we can prime, boost and further stimulate an immune response. If you look at the history of developing cancer medicines, every medicine that ever worked in late metastatic disease works better the earlier you go. So coming back to your question, I think the clinical trial design is a big factor in terms of having a higher probability of success, and therefore providing a positive return on investment. By extension, I am much more hesitant to predict the utility of a personalized cancer vaccine, even if combined with a checkpoint inhibitor, in later stage metastatic, epithelial tumors. I think this has been born out, even in melanoma by some of the competitors trials.
If one desires, a positive return on investment, one should conduct a definitive experiment in the context where it is most likely to give a positive answer. That means a randomized phase 2, rather than a phase 1, when combining with another medication and studying at the adjuvant rather than the metastatic setting.
I think both the manufacturing and logistics are no longer an issue. These were significant challenges when we started back in 2015 but based on the success of mRNA vaccines for infectious disease as well as the investment that have put into improving the manufacturing footprint. At this point, this should not be an impediment for a personalized cancer vaccine to provide a solid business case for Moderna. That said this, of course, is true for Moderna and maybe Biontech, but would be a challenge for any newcomer to the field, who would need to consider the costs of CMC in their overall ROI.
Logistics are even simpler. I don't want to trivialize this as it requires very careful and close coordination between the treating centers, the sequencing and bond for Matic analysis, and the manufacturer. At a high level, in an era where we can provide personalized cell therapy, a personalized vaccine is much easier by comparison.
