Personalized DNA Vaccine Tamps Down HCC

PD-1 inhibitors have dramatically increased survival in many cancer types, including melanoma: Approximately 30% to 40% of patients achieve objective responses to pembrolizumab (Keytruda; Merck) or nivolumab (Opdivo; Bristol Myers Squibb) as monotherapy. By contrast, the response rate has been more limited, just 14% to 17%, in hepatocellular carcinoma (HCC), the fourth most common cause of cancer-related death.

But according to research reported at the 2021 Society for Immunotherapy of Cancer (SITC) Annual Meeting in Washington, DC, held November 10–14, the personalized DNA vaccine GNOS-PV02 (Geneos Therapeutics), combined with pembrolizumab and INO-9012 (plasmid-encoded IL12), could up that response rate to 25%—without causing serious adverse events.

Starting in mice, researchers generated “very robust CD4⁺ and CD8⁺ responses against a range of different antigens,” said Mark Yarchoan, MD, of Johns Hopkins Medicine in Baltimore, MD, who presented the findings. “Other vaccine platforms that we've seen have mostly generated CD4⁺ but not CD8⁺ responses.”

Skewed responses toward CD4⁺ helper T cells do occur, but the reason why remains “an immunologic mystery,” said Lisa Butterfield, PhD, of the Parker Institute for Cancer Immunotherapy in San Francisco, CA. This phenomenon may be more common for RNA- or protein-based platforms, she added, and perhaps less so for DNA vaccines such as GNOS-PV02.

Encouraged by their preclinical findings, Yarchoan and his colleagues launched GT-30, a phase I/II trial of GNOS-PV02 in patients with advanced HCC, to see whether this vaccine might prompt the immune system to respond to a PD-1 inhibitor and elicit T-cell responses. Blood and tissue samples were sequenced “to figure out which mutations are in the tumor and which ones are actually expressed,” explained Yarchoan, who offered preliminary data on the study's first 12 patients.

GT-30′s participants all received first-line TKI treatment while their personalized DNA vaccines—encoding up to 40 neoantigens unique to each patient—were manufactured over 6 to 8 weeks. Upon disease progression, or if patients were unable to tolerate TKI therapy, they were given GNOS-PV02 with INO-9012 and pembrolizumab every 3 weeks; after 4 cycles, GNOS-PV02 and INO-9012 were administered every 9 weeks.

The personalized vaccine approach successfully targeted all neoantigens present in 83% of patients, Yarchoan said. Five experienced stable disease and three others had partial responses, which “appear to be quite durable.” There were no dose-limiting toxicities with this trio of therapies, and the most common side effects were low-grade fatigue and injection site reactions.

Comparing one responder's blood and tumor tissue samples from before and after treatment, the researchers found that a number of new T-cell receptor (TCR) clones emerged that “were not present at baseline,” said Yarchoan, “and presumably are vaccine-induced.” Many of these TCR clones also infiltrated the tumor microenvironment by 9 weeks, potentially mediating regression, he added.

For SITC attendees, a key question, which remains unanswered, was whether there is an optimal number of epitopes for GNOS-PV02. In the cancer vaccine field, “there are competing immunologic theories about too many antigens distracting the resultant immune response, versus including as many candidates as might be processed and presented on the surface of tumor cells, to stimulate a really broad T-cell repertoire,” Butterfield noted. “Where the optimal situation lies may be patient-specific.”

“Personally, my opinion is that more is better,” said Yarchoan. “Our ability to predict which antigens are going to be the most appealing to the immune system is still quite limited. I think the more neoantigens you put into a vaccine, the more shots you have on goal.” –Suzanne Rose and Alissa Poh

For more news on cancer research, visit Cancer Discovery online at http://cancerdiscovery.aacrjournals.org/CDNews.