Emerging Treatments | Personalized Cancer Vaccines

Session Focus

Gerberry framed personalized cancer vaccines as an area that has gotten hot very fast, with the Merck data and forthcoming ESMO results setting up the central question: where next.

His specific framing was whether immune sensitivity is an absolute requirement for exploring a personalized vaccine approach, or whether the greater unmet need lies in cold tumors resistant to immuno-oncology.

Why Melanoma Was the Right Starting Point

Ryan Sullivan MD explained the logic behind the initial choice, which he considered sound for several reasons.

Manufacturing time. Quality control and vaccine production take time, though less than six or seven years ago. That argues for a disease setting where half the patients will not have progressed by the time they receive the vaccine. An adjuvant setting after surgery provides that window, since even patients who recur typically do not recur within the first few months.

An approved comparator. Pembrolizumab and nivolumab were already standard of care in adjuvant melanoma, making a randomized design straightforward: vaccine plus pembrolizumab versus pembrolizumab alone.

Field history. Melanoma has long been where immunotherapy development happens, largely because cytotoxic chemotherapy does not work there.

He contrasted that with BioNTech’s approach in pancreatic cancer, which he called audacious in a good way. That strategy developed from data in a very small subset of patients showing that those who successfully generated documented immunization did better than those who did not. He described that as proof of concept both that tumor-specific T cells can be generated and that they can potentially change outcomes even in that disease.

His synthesis: the two ends of the spectrum answer two distinct questions. Can immune checkpoint inhibition be improved where checkpoints already work, and can an immune response be stimulated in a disease where that is not normally expected.

Given success at both ends, his conclusion was that everything is fair game. He noted that development still proceeds stepwise, as Keytruda did starting with lung and melanoma before expanding, but the excitement comes precisely from preconceived ideas about what would work turning out to be wrong.

Cold Tumors and Combination Partners

Xin Gao MD described the immune landscape across his own tumor types.

Bladder and kidney cancers are somewhat less immunoresponsive than melanoma but are genuinely immunoresponsive, with checkpoint inhibitors used as standard of care from localized through metastatic disease.

Prostate sits at the far other end as a cold tumor, where checkpoint inhibitors are used only in rare circumstances such as a TMB-high tumor.

On why an additional agent is needed, he gave two distinct rationales.

In non-muscle invasive bladder cancer, intravesical BCG has been standard of care for decades. It elicits an immune response and can be curative for a rare subset, but patients recur and progress. Building on a standard of care that exists for a reason, and asking how to boost that response, is the logic for substituting BCG where pembrolizumab would sit in other settings.

For cold tumors like prostate, his point was that unresponsiveness to checkpoint inhibitors does not mean unresponsiveness to immunotherapy. T cell engagers and newer immunotherapy forms have produced responses, sometimes deep responses, in trials. The combination partner may therefore need to be a different immunotherapy approach entirely.

Does Tumor Mutational Burden Matter

Gerberry pressed on the role of elevated TMB. The intent may be maximizing neoantigens available for inclusion in a vaccine, but in both adjuvant melanoma and the metastatic setting, TMB levels have not necessarily indicated activity of immune-based therapies.

Sullivan gave a carefully qualified answer. In melanoma, several factors associate with higher response rates. PD-L1 elevated tumors show higher response rates, though there is no PD-L1-based indication. Higher TMB likewise associates with higher response rates.

The complication is what follows from association. Melanoma has essentially nothing besides checkpoint inhibitors and their combinations, and despite many drugs developed, they all work similarly. They still work in a subset of PD-L1 low patients and in a subset of TMB low patients, which is what makes these markers associative rather than determinative.

Key Takeaways

1. Adjuvant melanoma solved a manufacturing problem as much as a biological one, providing a window before recurrence.

2. An existing approved standard of care made the randomized design straightforward.

3. Pancreatic data established that documented immunization correlates with better outcomes even in a difficult tumor.

4. Success at both ends of the immune spectrum suggests no disease setting should be excluded on principle.

5. Cold tumors may need a different immunotherapy partner rather than a checkpoint inhibitor.

6. TMB and PD-L1 associate with response without determining it, since therapies still work in low-expressing subsets.