NEW YORK – The optimal course of pre-surgery immunotherapy treatment for stage III melanoma patients may be contingent on the level of interferon gamma (IFN-γ) gene expression in their tumors, according to data presented Monday during the European Society for Medical Oncology (ESMO) Congress.
Christian Blank, an oncologist with the Netherlands Cancer Institute, presented results from the Phase Ib DOMINI trial, in which patients with stage III melanoma were randomized to one of four immunotherapy cohorts after baseline tumor gene expression screening using an assay from NanoString Technologies to determine whether patients had IFN-γ signature-high or IFN-γ signature-low cancers. Specifically, patients' tumor biopsies were tested for RNA expression in a group of genes that play a role in interferon signaling. (Blank didn't disclose the specific research assay used in this analysis during his presentation.)
Patients whose tumors were considered IFN-γ signature-high were randomized to receive either the PD-1 inhibitor nivolumab (Bristol Myers Squibb's Opdivo) monotherapy, or nivolumab monotherapy plus the investigational class I histone deacetylase (HDAC) inhibitor domatinostat, a product of German biotech 4SC.
Patients with IFN-γ signature-low tumors, meanwhile, were randomized to receive either the same nivolumab-domatinostat combination as the latter IFN-γ signature-high group or the combination of nivolumab, ipilimumab (BMS' CTLA4 inhibitor Yervoy), and a slightly lower dose of domatinostat.
The end goal of the ongoing study, according to Blank, is to identify patients for whom neoadjuvant immunotherapy can be deescalated and those for whom further escalation would be beneficial.
"We want to deescalate interferon signature-high patients and to escalate interferon signature-low patients, who have a high chance of relapsing," he further explained. But as a Phase I trial, the first priority and the study's primary endpoint was to assess the treatments' safety and feasibility, as determined by the percentage of patients who went on to have surgery as planned, six weeks into the study.
The trial's secondary endpoints, meanwhile, included measures of treatment efficacy, such as pathologic response and relapse-free survival rates, as well as biomarker analyses beyond the IFNγ signature, including RNA signatures associated with pathologic complete response and relapse-free survival within each arm, as determined by RNA sequencing and NanoString gene expression analysis, and changes in immune infiltrates and markers compared to baseline. Researchers also looked at the relationship between microbiome diversity in fecal samples and treatment response and toxicity. Data for some of these biomarker analyses, including the microbiome diversity, won't be available for years.
Each of the four treatment arms included 10 patients, making the trial too small — and, indeed, too early — to draw any definitive practice-changing conclusions, but the ESMO analysis was nonetheless significant as this is the first time researchers evaluated the ability of this IFN-γ signature to guide neoadjuvant immunotherapy for this melanoma patient population.
Ultimately, 100 percent of the patients in each arm were able to have surgery at six weeks, as initially planned. In terms of pathologic response, the rates were 90 percent among the IFN-γ signature-high patients who received nivolumab alone (arm A); 80 percent among IFN-γ signature-high patients who had received nivolumab-domatinostat combination (arm B); 30 percent among IFN-γ signature-low patients who had received the same combination therapy (arm C); and 40 percent among IFN-γ signature-low patients who received the combination with the addition of ipilimumab (arm D).
When the endpoint was narrowed to pathologic complete response, these rates were 70 percent, 50 percent, 10 percent, and 30 percent in arms A, B, C, and D, respectively.
While the data were not mature for Blank to report confirmed survival benefit, he did share relapse-free survival rates in the four arms at six months following surgery. These estimates were 100 percent for both arms A and B, and 90 percent and 53.3 percent, respectively, for arms C and D. When stratified only according to IFN-γ signature status, the six-month relapse-free survival rate estimates were 100 percent for IFN-γ signature-high patients and 79.4 percent for IFN-γ signature-low patients.
"You see that the interferon signature really nicely can predict patients that can be deescalated," Blank said. "The signature is a very good discriminator for identifying patients who can benefit from nivolumab alone … versus patients who are in need of combination therapy." Among the IFN-γ signature-high patients, adding domatinostat to neoadjuvant nivolumab did not bring any added benefit, according to Blank.
As a next step, investigators will test the slightly higher domatinostat dosage in the IFN-γ signature-low combination arm to shed further light on the benefit of treatment escalation. In the IFN-γ signature low group, as of now, Blank noted, "What we see … from the triple therapy is not better than the doublet … But we have had this lower dosing, so we are now testing that [higher dose] to see if we get that higher response [with adding domatinostat]. But we really don't know yet."
Future utility of IFN-γ signature
According to Blank, the IFN-γ signature that he and his team used as a biomarker to stratify patients in the DOMINI trial is rooted in an algorithm that uses interferon signature-related genes that have already been published and defined. Blank noted as well that the test process has a "very strict quality control upfront and then had correction factors." It was developed based on cohorts of stage III patients in the neoadjuvant setting, but Blank said it would also have utility for stage IV melanoma patients.
He further shared that the German firm developing domatinostat, 4SC — which is also sponsoring the DOMINI trial — will most likely incorporate the signature into a commercially available testing kit. "It works very fast," he said. "You have the data within two days."
Importantly, Blank mentioned that in the stage IV melanoma setting, previous studies have shown that adding domatinostat to treatment could make IFN-γ signature-low patients become IFN-γ signature-high patients. However, it will take time to determine whether the same phenomenon occurs in the ongoing DOMINI study.
While more research is needed to validate the signature as a predictive biomarker for response to these various combinations of nivolumab, ipilimumab, and added domatinostat in the neoadjuvant melanoma treatment setting, Blank is optimistic that the signature has other potential uses in guiding treatment.
"The interferon signature is a signature showing how your immune system reacts in a network," he said. "It's not only the T-cell reaction towards the tumor … It can be used [as a biomarker for] all these other targeted therapies and chemotherapies because [they] are antigen-freeing, and the functioning immune system will pick that up." With IFN-γ signature-low patients, however, who have immunologically colder tumors, Blank added, "This is [where] we have to work hard and help the immune system to become activated."