I’ve always thought of a cancer vaccine as providing a "study guide" to the immune system. It points out what might be on the test, but ultimately the immune system "studies" some chapters better than others, which is exactly why the immunodominance tournament you described is such a challenge.
This approach naturally shines in earlier-stage settings where the overall tumor burden is low, clonal heterogeneity is limited, and the patient's immune system isn't yet exhausted.
But once you get to advanced metastatic disease, the sheer diversity of the mutating clones and the immunosuppressive defenses of the tumor microenvironment make the material overwhelming. At that late stage, it makes more sense to bypass the studying entirely and hand the immune system the direct answers to the test via cell therapies (TILs, TCR-T).
Of course, as you noted with the MAGE-A3 cross-reactivity trials, giving the direct answers comes with its own massive hurdles, namely manufacturing scalability, engineering necessary enhancements into the cells, related toxicities, and the lethal stakes of getting even one "answer" slightly wrong.
Great painting! What is the title? Who is it by? I tried Claude, and Google Lens, and even Bing Visual search and all 3 came up with various (wrong) guesses.
OwlPosting is a tremendous resource for learning about medical discovery - or just being humbled by how many clever ideas have been tried and failed. I still find the BioNTech vaccine for pancreatic cancer really exciting. Driving 80% post surgery recurrence down to 8/16 patients... well, that seems likely to be real, even if we completely ignore the immune response stratification.
What a thorough primer — bookmarking this. The parallel on the cell-therapy side keeps striking me: neoantigen vaccines and CAR-T are converging on the same question — not just "does it work," but "does the manufacturing model survive." For CAR-T that's now existential, with in vivo programs trying to delete the ex vivo factory entirely. Wrote about exactly that (KITE-753 vs the in vivo wave) here if you're interested: https://novc.substack.com/p/what-replaces-yescarta?r=6p7sjb
I’ve always thought of a cancer vaccine as providing a "study guide" to the immune system. It points out what might be on the test, but ultimately the immune system "studies" some chapters better than others, which is exactly why the immunodominance tournament you described is such a challenge.
This approach naturally shines in earlier-stage settings where the overall tumor burden is low, clonal heterogeneity is limited, and the patient's immune system isn't yet exhausted.
But once you get to advanced metastatic disease, the sheer diversity of the mutating clones and the immunosuppressive defenses of the tumor microenvironment make the material overwhelming. At that late stage, it makes more sense to bypass the studying entirely and hand the immune system the direct answers to the test via cell therapies (TILs, TCR-T).
Of course, as you noted with the MAGE-A3 cross-reactivity trials, giving the direct answers comes with its own massive hurdles, namely manufacturing scalability, engineering necessary enhancements into the cells, related toxicities, and the lethal stakes of getting even one "answer" slightly wrong.
Great deep dive!
Marvelous writeup! I even understood several of the words.
Fantastic article.
Great painting! What is the title? Who is it by? I tried Claude, and Google Lens, and even Bing Visual search and all 3 came up with various (wrong) guesses.
it is a diffusion model (i think)
OwlPosting is a tremendous resource for learning about medical discovery - or just being humbled by how many clever ideas have been tried and failed. I still find the BioNTech vaccine for pancreatic cancer really exciting. Driving 80% post surgery recurrence down to 8/16 patients... well, that seems likely to be real, even if we completely ignore the immune response stratification.
What a thorough primer — bookmarking this. The parallel on the cell-therapy side keeps striking me: neoantigen vaccines and CAR-T are converging on the same question — not just "does it work," but "does the manufacturing model survive." For CAR-T that's now existential, with in vivo programs trying to delete the ex vivo factory entirely. Wrote about exactly that (KITE-753 vs the in vivo wave) here if you're interested: https://novc.substack.com/p/what-replaces-yescarta?r=6p7sjb