Triplet demonstrates activity in metastatic colorectal cancer
Click Here to Manage Email Alerts
The combination of panitumumab, ipilimumab and nivolumab induced responses among patients with metastatic colorectal cancer, according to phase 2 study results presented at Gastrointestinal Cancers Symposium.
The study, which included patients with KRAS, NRAS and BRAF wild-type and microsatellite-stable tumors, provides evidence that combinations of immunotherapies with anti-EGFR therapy are rational and merit further investigation, according to Michael Sangmin Lee, MD, assistant professor at The University of Texas MD Anderson Cancer Center.
“Standard-of-care options for patients in the third-line setting may include trifluridine-tipiracil [Lonsurf, Taiho Oncology] or regorafenib [Stivarga, Bayer]; these have low response rates and modest PFS,” he told Healio. “Identifying an active immunotherapeutic approach in microsatellite-stable colorectal cancer is a huge area of need, as we have not seen activity of immuno-oncology agents. In our study, panitumumab [Vectibix, Amgen], ipilimumab [Yervoy, Bristol Myers Squibb] and nivolumab [Opdivo, Bristol Myers Squibb] showed a promising response rate, higher than we would have expected with panitumumab monotherapy.”
Anti-EGFR therapies, such as panitumumab, are standard for the treatment of patients with KRAS, NRAS and BRAF wild-type metastatic colorectal cancer, and researchers have found that these agents are immunomodulatory and induce immunogenic cell death, Lee added.
“Translational studies in patients whose tumors responded to anti-EGFR therapy have also shown significant increases in cytolytic activity and T-cell infiltration, but also increases in expression of immune checkpoint pathways at progression, including PD-L1 and CTLA-4,” he said. “Given this, we hypothesized that combining panitumumab with the anti-PD-1 agent nivolumab and the anti-CTLA-4 agent ipilimumab would be more active than what we would have expected with panitumumab alone.”
Lee and colleagues tested their hypothesis in a cohort of 56 patients (median age, 56 years; range, 36-74; 66% men; 77% white), six of whom were included in a safety run-in phase, which demonstrated no dose-limiting toxicities in the first 12 weeks.
Twenty-six of the patients participated in stage I of the multicenter, Simon’s two-stage trial, which demonstrated nine responses out of seven needed to merit further enrollment to stage II. Stage II included 24 patients and required 17 responses to warrant further study.
Twelve-week response rate — assessed among 49 patients, after excluding seven who were unevaluable — served as the study’s primary endpoint.
Median follow-up was 12.1 months.
Researchers observed 17 responses at week 12, for a response rate of 35% (95% CI, 21-48), meeting the prespecified criteria for further study. All these were partial responses, with an additional 43% of patients showing stable disease and 22% of patients experiencing disease progression.
The best response rate at any time, which included confirmed and unconfirmed responses, was 41%. The confirmed overall response rate was 18% (95% CI, 10-31).
Seven patients remained on treatment at the time of the analysis, some of whom are awaiting confirmation of their response, Lee said during his virtual presentation.
“As there are still subjects on trial, we will look to report on the durability of response in a future meeting,” he told Healio. “We did see that some patients, despite having a response, did not have a durable response, and it would be important to understand what mechanism drives this acquired resistance or if we are still seeing emergence of subclonal resistance mutations driving that are already known to drive acquired resistance to current standard-of-care anti-EGFR therapy.”
Median PFS was 5.7 months (95% CI, 5.5-7.9) and median OS, data for which are still immature, was 27 months (95% CI, 14.5 to not estimable).
No unexpected toxicities occurred, Lee said. The most common treatment-related grade 3 to grade 4 adverse events included hypomagnesemia (11%), rash acneiform (11%), lipase increase (9%), amylase increase (7%), aspartate aminotransferase increase (5%), alanine aminotransferase increase (5%), diarrhea (5%), hypophosphatemia (5%) and maculopapular rash (5%).
Two grade 5 toxicities occurred — myocarditis and colonic perforation — but only the former was considered possibly related to treatment.
“Although toxicity profiles were not unexpected, we did still see immune-related adverse events as you would expect to see with ipilimumab/nivolumab-based combinations, and these toxicities do need to be considered,” Lee said.
Lee told Healio that he and his team are continuing their work on translational studies to establish more biomarkers of activity in this setting.
However, in the Q&A session of the virtual presentation, some experts described the median PFS as “relatively disappointing” and the response rate as lower than expected for the combination.
“The response rate was encouraging although with a preponderance of left-side tumors [50%] as a caveat, which probably did help the response rate to be a bit higher,” Lee responded. “Our null hypothesis was based on the ASPECCT study [of single-agent panitumumab] where they saw a 22% response rate. We certainly did meet that primary endpoint, but it’s not a slam dunk.”
Perspective
Back to Top
Michael J. Hall, MD, MS
This single-arm study used a Simon two-stage phase 2 design to examine the preliminary efficacy of the combination of EGFR inhibition with panitumumab combined with dual-agent immune checkpoint blockade in RAS wild-type, microsatellite-stable colorectal cancer treated with one or two prior lines of therapy. Lee and colleagues hypothesized that targeted EGFR inhibition with panitumumab would prime immune response to nivolumab and ipilimumab in microsatellite-stable colorectal cancer through increased cytolytic activity and upregulation of T-cell infiltration. The study met its primary endpoint, and seven patients remain on study. Only hypomagnesemia (11%) and rash (11%) had treatment-related adverse event rates of grade 3 or higher that occurred at an incidence greater than 10%.
However, the median PFS of 5.7 months was similar to PFS estimates seen in comparable studies of single-agent panitumumab in the first- and second-line settings for RAS wild-type microsatellite-stable colorectal cancer and, thus, the added biologic benefit of immune checkpoint blockade in this population remains uncertain at this time.
Collapse
Lee MS, et al. Abstract 7. Presented at: Gastrointestinal Cancers Symposium (virtual meeting); Jan 15-17, 2021.
Click Here to Manage Email Alerts