Personalized Cancer Vaccine Boosts Recurrence-Free Survival in Melanoma

ORLANDO -- Adding a personalized cancer vaccine to anti-PD-1 immunotherapy for resectable melanoma led to a 44% improvement in recurrence-free survival (RFS) versus immunotherapy alone, according to a study reported here.

Patients who received the mRNA vaccine, derived from their own tumor, had an 18-month RFS of 78.6% as compared with 62.2% for patients who received only immunotherapy. The vaccine did not add substantially to the toxicity seen with single-agent pembrolizumab (Keytruda). The combination led to a similar reduction in the risk of recurrence or death in patients with high or low tumor mutational burden (TMB), which has been shown to be a predictor of response to immunotherapy, reported Jeffrey Weber, MD, of NYU Langone Perlmutter Cancer Center in New York City, at the American Association for Cancer Research meeting.

"For the first time in a randomized study with a control arm, the addition of an mRNA neoantigen vaccine appeared to augment the benefit of PD-1 blockade, without adding significant high-grade toxicity," said Weber. "This study is extraordinarily important, because it gives hope that this novel strategy will provide clinical benefit."

The study provides the first compelling evidence that a vaccine approach might have value in treating cancer, said Timothy Yap, MBBS, PhD, of the MD Anderson Cancer Center in Houston.

"Vaccine therapy has really been disappointing in the past, but this study really provides the first evidence that vaccine therapy, using this neoantigen approach, can work," he said.

During a discussion that followed the presentation, Weber suggested that the melanoma trial is just the beginning of what could be a therapeutic development not only in melanoma but for cancer in general.

"This approach is absolutely applicable to any solid tumor that expresses neoantigens, and that's almost every solid tumor," he said. "The number of neoantigens and tumor mutations varies hugely among solid tumors. I think breast cancer has an average of maybe 40 single nucleotide variants. Melanoma is the champion with between 600 and 800 per tumor. It only takes one. There is no reason why this strategy would not work in the adjuvant mode in any solid tumor that was sensitive to PD-1 blockade."

"Heme malignancies tend to have many fewer SNPs [single nucleotide polymorphisms], but I see no reason why you couldn't pursue the strategy as long as you had a good algorithm for choosing the best neoantigens," Weber said.

Whether the vaccine approach would work with other types of cancer therapy is less certain, said co-investigator Ryan Sullivan, MD, of the Mass General Cancer Center in Boston. And the use of a vaccine alone would have a very high bar to clear, given the success of PD-1/L1 inhibitors in the adjuvant setting.

"There would have to be very compelling data supporting the ability to launch a trial where you would not be using the standard of care. Biologically, it's very possible that a vaccine itself could be helpful, but I think practically speaking, it's very difficult to study," said Sullivan.

Asked whether a vaccine "booster" at the end of treatment might be useful to prolong response to immunotherapy, Weber said, "It's a great idea."

Adjuvant therapy with an immune checkpoint inhibitor has become standard of care for patients with high-risk resectable melanoma. Whether the benefits of immunotherapy could be enhanced by a personalized cancer vaccine was unclear.

The underlying premise of a personalized cancer vaccine is to target an individual patient's unique tumor mutations. Vaccine production involves obtaining a specimen of a patient's tumor and a sample of normal tissue, performing next-generation sequencing of tumor DNA and RNA, and determining the patient's HLA type.

"You take that data and plug it into a computerized algorithm that will come up with the most logical, potentially most effective, immunogenic neo-epitopes, the peptides that potentially bind to the MHC [major histocompatibility] class I or class II, expressed by that patient," said Weber. "Since they are neoantigens, they are absolutely tumor specific. They should not be in normal tissue because the computer filters out anything but a nonsynonymous variant."

The resulting collection of neoantigens are assembled into a single messenger RNA molecule that is used to produce a microencapsulated vaccine. The same technology was used to produce mRNA vaccine for COVID-19 and was developed by vaccine manufacturer Moderna. This particular cancer vaccine was based on 34 tumor-specific neoantigens.

Investigators in the multicenter phase II mRNA-4157-P/KEYNOTE-942 trial enrolled patients with resectable stage IIIB-D/IV melanoma, who had complete resection, and were disease free at enrollment. The patients were randomized 2:1 to pembrolizumab plus the vaccine or to pembrolizumab alone. Patients in the vaccine arm received pembrolizumab for up to a year plus as many as nine doses of the mRNA vaccine, followed by additional pembrolizumab.

The primary endpoint was RFS. The primary analysis included 157 patients who had a median follow-up of 23-24 months. The data showed that patients who received the vaccine plus pembrolizumab had a hazard ratio of 0.561 versus those treated with pembrolizumab alone (95% CI 0.309-1.017, P=0.0266).

Sullivan reported findings from a biomarker analysis that included TMB, tumor inflammation score (TIS), and PD-L1 expression. The results showed that patients fared better with the vaccine regardless of whether they had high TMB, TIS, or PD-L1 expression (predictive of response to pembrolizumab) or low levels of the markers.