SAN DIEGO – After three years of follow-up, patients with pancreatic cancer who responded to BioNTech and Roche's personalized neoantigen vaccine, autogene cevumeran, continued to live longer without their cancers progressing than patients who did not initially respond to the vaccine.
Vinod Balachandran, a surgeon scientist at Memorial Sloan Kettering Cancer Center, presented up-to-date results from the Phase I trial during the American Association for Cancer Research's annual meeting on Sunday. In the first-in-human trial, investigators administered the personalized messenger RNA-based vaccine plus Roche's Tecentriq (atezolizumzb) and chemotherapy in 16 pancreatic cancer patients after surgical resection.
As the first step of the treatment regimen, patients with resectable pancreatic cancer underwent surgery to remove their cancers. Then, before receiving their personalized vaccines, they underwent one dose of the PD-1 checkpoint inhibitor Tecentriq, and after their autogene cevumeran vaccinations, they went on to receive 12 cycles of chemotherapy followed by a final booster dose of the vaccine.
In an earlier data readout, which was published in Nature last year, BioNTech and Roche shared that 50 percent — 8 out of 16 — of these patients had an immune response to the vaccine, as determined by the presence of CD8-positive T effector cells targeting the tumor-specific neoantigens. These immune responses appeared to correlate with recurrence-free survival, too. After about a year and a half, the median recurrence-free survival was not yet reached for those eight responding patients. Among patients who did not experience an immune response to the vaccine, meanwhile, the median recurrence-free survival was 13.4 months.
Now, after a median follow-up of 3.2 years, the recurrence-free survival improvement has continued among autogene cevumeran responders, Balachandran shared during his presentation. The median recurrence-free survival among patients who responded is still not yet reached versus 13.4 months among patients who didn't respond.
Balachandran also noted that multiple of these patients' CD8-positive T cell clones had expanded after the autogene cevumeran infusion. Of 79 vaccine-induced CD8-positive T cell clones that investigators tracked with extensive peripheral drug monitoring over time, more than 80 percent of the clones had persisted in each patient's blood after three years. Additionally, 98 percent of these 79 clones hadn't been detected in blood, tumors, and nearby tissues before the vaccine.
According to Balachandran, the fact that these clones were absent pre-vaccine is an encouraging signal when it comes to eligibility for cancer vaccination. Rather than relying on patients having existing T-cell clones that the vaccination then expands, autogene cevumeran appears to generate these neoantigen-specific T cells de novo, he said. "This isn't restricting immune response to patients who have preexisting T cells," Balachandran said. "This theoretically expands vaccine-eligible cancers beyond pancreatic cancer."
Manufacturing, infusion
Autogene cevumeran, which is also known as BNT122 and RO7198457, is designed to target up to 20 tumor-specific passenger mutation-derived neoantigens. After a patient undergoes surgical resection, samples of both their tumor and blood are sent to a central facility where BioNTech performs tumor/normal DNA sequencing and RNA sequencing. Then, using a computational model, the drugmakers home in on up to 20 neoantigens and generate a bespoke mRNA vaccine to target these select neoantigens.
The choice to target up to 20 neoantigens was somewhat arbitrary. "The optimal number of neoantigens to target in a cancer is still unknown," Balachandran said during a press conference Sunday morning. "Conceptually, you want to have maximal coverage of all cancer clones, [but] whether that could be achieved with a smaller number of neoantigens or if it's beneficial to have more … is one of the questions we're hoping to learn more about as the field moves forward."
As evidenced by the range in neoantigen quantity among other firms in the personalized cancer vaccine space, the field at large is still trying to figure this out. Merck and Moderna, who are developing their own mRNA-based personalized neoantigen cancer vaccine as adjuvant treatment for melanoma and non-small cell lung cancer, for instance, have designed their vaccine to target 34 neoantigens. The Swiss biotech Nouscom has chosen 60 neoantigens for its personalized vaccine, and PDC*Line Pharma is going after around 12 neoantigens for its personalized cancer vaccine in colorectal cancer.
The best method for selecting neoantigens is also a work in progress, Balachandran noted. "This is an important challenge in the field right now, and we're not quite there yet," he said. "There's an opportunity for us as a field to learn about what makes specific vaccine neoantigens more immunogenic than others."
Next steps, lingering questions
BioNTech and Roche are now moving autogene cevumeran into a much larger randomized, Phase II clinical trial. In the trial, which started in October 2023, drugmakers have begun enrolling 260 resectable pancreatic cancer patients who will receive either the personalized vaccine plus Tecentriq and chemotherapy or standard-of-care chemo. Patient recruitment has already started at US sites and will advance to global sites next, according to BioNTech.
BioNTech and Roche are also studying the bespoke vaccine in other tumor types and settings, including in combination with Merck's checkpoint inhibitor Keytruda (pembrolizumab) as a frontline treatment for advanced melanoma patients and as an adjuvant treatment for certain circulating tumor DNA-positive colorectal cancer patients.
Although academic investigators and sponsors alike are encouraged by the durability of patients' immune responses to autogene cevumeran and the recurrence-free survival benefit, the fact that only half of pancreatic cancer patients had an immune response to the treatment in the first place remains a conundrum.
Balachandran hopes the forthcoming Phase II study will help shed light on why half of these patients didn't respond. One theory he has, based on preclinical observations, is that removing patients' spleens during pancreatic cancer surgery could reduce their immune responses to the vaccine.
"In pancreatic cancer, we do two different types of surgeries as part of standard of care, one of which involves spleen removal," he said. "It wasn't a complete split, but many of the non-responders did not have spleens at the time of vaccination."
In the upcoming trial, he noted, patients without spleens won't be eligible.
Balachandran said he and co-investigators are in the process of looking into other features of patients who do and don't respond, including whether non-responders had on-target effects from the vaccine in the first place.