In an interview with Targeted Oncology, Jeffrey S.Weber, MD, PhD, discussed the 3-year efficacy and biomarker results from the phase III CheckMate 238 trial, which he presented at the 2019 European Society of Medical Oncology Congress.
Jeffrey S. Weber, MD, PhD
Three-year data from the phase II CheckMate 238 trial shows continued superior efficacy with nivolumab (Opdivo) versus ipilimumab (Yervoy) in patients with resected stage III or IV melanoma, according to research presented at the 2019 European Society for Medical Oncology (ESMO) Congress.
Patients in the trial who were followed for 36 months showed a relapse-free survival (RFS) rate of 58% in the nivolumab arm and 45% in the ipilimumab arm (HR, 0.68; 95% CI, (0.560.82);P< 0.0001). The nivolumab group also did better than the ipilimumab in terms of distant metastases-free survival (DMFS), with a HR of 0.78.
Multiple subgroups were analyzed in the study, including, patients with stage IIIB, stage IIIC, stage IV, a PD-L1 ≥ 5%, a PD-L1 ≤ 5%,BRAF-mutant melanoma, orBRAFwild-type disease. Across all groups, except patients with stage IV melanoma, the nivolumab arm was superior to the ipilimumab arm in terms of both RFS and DMFS. In patients with stage IV melanoma, treatment with nivolumab was superior in terms of RFS, but no DMFS data was reported.
The 3 tumor-related biomarkers that were assessed in Checkmate 238 were tumor mutational burden (TMB), interferon-gamma gene expression signature, and CD8+ T-cell infiltration. Based on high versus low values, the RFS for TMB in the nivolumab arm was 30.8 compared to 24.1, at 24 months of follow-up. With the interferon-gamma gene expression signature, the RFS was not reached (NR) and 30.8. Finally, for CD8+ T-cell infiltration, the RFS was 30.8 versus 24.9. In the ipilimumab arm, the high and low values were NR versus 15.9, NR versus 18.3, and NR versus 13.8.
In previous data, nivolumab demonstrated activity in patients with metastatic melanoma. Based on these signals, oncologist wondered if nivolumab could have good activity in the adjuvant setting, said Jeffrey S. Weber, MD, PhD.
“I ceased upon that idea early and did a pilot adjuvant trial of 33 patients with resected stage IV melanoma who received a peptide vaccine with nivolumab. Based on those encouraging results, which we have yet to reach with 6 or 7 years of follow-up on median RFS, a large randomized study called Checkmate 238 was launched,” Weber stated.
Patients in CheckMate 238 were randomized 1:1 to receive either nivolumab or ipilimumab and evaluated RFS. The secondary endpoint was overall survival (OS), which has yet to be reached in the trial. The study also included exploratory endpoints, such as DMFS for patients with stage III melanoma and biomarkers for efficacy.
In an interview withTargeted Oncology, Weber, medical oncologist, and deputy director, Perlmutter Cancer Center, NYU Langone Health, discussed the 3-year efficacy and biomarker results from the phase III CheckMate 238 trial, which he presented at the 2019 ESMO Congress.
TARGETED ONCOLOGY: What was the rationale for the CheckMate 238 trial?
Weber: PD-1 antibodies like nivolumab and pembrolizumab showed clear activity in patients with metastatic melanoma and a variety of other cancers, back in 2010 through 2012. Ultimately, these drugs were approved in 2014 for metastatic melanoma. The natural progression was, if a drug has good activity, in the 40+% response range in untreated patients with metastatic melanoma, would it be an appropriate adjuvant therapy?
I ceased upon that idea early and did a pilot adjuvant trial of 33 patients with resected stage IIIC/IV melanoma who received a peptide vaccine with nivolumab. Based on those very encouraging results, where we have yet to achieve median RFS with 6 or 7 years of follow-up, a large randomized study called Checkmate 238 was launched. In only 8 months, it accrued 906 patients. It was a randomized study, phase III, double-blinded of adjuvant ipilimumab, which was then the approved therapy in the United States in 2015, versus nivolumab as the experimental arm. There was a straight 1:1 randomization stratified by stage and PD-L1 status of 5% or more. Eligible patients had to have a performance status of 0 or 1, no prior systemic therapy, and had to have completely resected stage IIIB/C or stage IV melanoma by the American Joint Committee on Cancer 7th criteria.
TARGETED ONCOLOGY: What data did you present at ESMO?
Weber: The updated data that I presented here at ESMO 2019 suggests that with 36 months of follow-up in all patients, there's still an excellent benefit for nivolumab versus ipilimumab as the active comparator. ThePvalue for the difference remains a healthy 0.001. The hazard ratio remains an excellent 0.68. There appears to be a plateau coming because if you look at the first year, 30% of patients dropped out and relapsed, then 8% in the second year, and now only 4% in the third year. If you do the arithmetic, we're beginning to see a plateau both in the ipilimumab and nivolumab arms. Interestingly, the curves are also widening out. We saw about a 10% difference the first year, 11% or 12% in the second year, and now a 13% difference in RFS, again suggesting the continued benefit for nivolumab compared with ipilimumab in terms of RFS.
We also updated the DMFS survival data, and that is considered to be a surrogate marker for OS. There, thePvalue for the difference showing benefit for nivolumab versus ipilimumab remains at 0.044 with a HR of 0.78. Again, there's continued benefit in terms of DMFS for nivolumab compared to ipilimumab.
Finally, for the first time, we had biomarker data from a significant portion of patients on the CheckMate 238 trial. My colleagues at Bristol-Myers Squibb and their biomarker group looked at 3 tumor-related biomarkers, TMB by whole exome sequencing, a signature of 10 gamma interferon inducible genes by shot gun sequencing, CD8+ T cell infiltrate by immunohistochemistry, and, for the first time, a peripheral blood marker consisting of monocytic myeloid derived suppressor cells in the peripheral blood by flow cytometry.
The data for the tumor markers are above the median and below the median for the peripheral blood suppressive marker. The was clear association with outcome for the both the nivolumab and the ipilimumab arms, with the most potent ones with the best hazard ratios being the gamma interferon gene signature and the TMB. However, if you look at combinations of what should be independent biomarkers, like gamma interferon signature/tumor burden and gamma interferon signature/peripheral blood/monocytic myeloid derived suppressor cells, the hazard ratio for the favorable category, meaning elevation in TMB above the median, elevation of the gamma interferon signature above the median, or the myeloid derived suppressor cells below the median, you had an even better hazard ration for outcome.
This time we compared the favorable phenotype to the unfavorable phenotype, and in some cases, by doing a combination, you had a hazard ratio of 0.2. If you looked at the RFS curves for elevated TMB, elevated gamma signature in the tumor, at 3 years, those with elevations of both had an 85% RFS versus only about 25% if both were below the median. This suggests that maybe we have a composite set of biomarkers that may help direct us to the best therapy for patients with high-risk resected melanoma.
The updated results of the CheckMate 238 study at 36 months of follow-up even more strongly suggest that a PD-1 antibody like nivolumab should be your first line choice of therapy. Since the nivolumab data have the longest follow-up and the most reliable data, it certainly reinforces the notion that nivolumab should be the first-choice adjuvant therapy for patients with high-risk resected melanoma given for 1 year.
TARGETED ONCOLOGY: Were there any immune-related adverse events seen in this trial?
Weber: In the updated presentation, we did not update the toxicity data since the original presentation that I made at ESMO 2 years ago included all patients who were followed for up to 100 days after treatment, which is the maximum amount of time that we assess the toxicities.
Just to recap, 8% of patients had to stop treatment in the nivolumab arm because of toxicity and between 10% and 15% of patients has grade 3/4 immune-related adverse events that were of significance. It was similar to what we've seen with pembrolizumab and similar to what you would see in the metastatic mode when using a single agent PD-1 drug, like nivolumab or pembrolizumab.
TARGETED ONCOLOGY: What are the next steps with this research?
Weber: The next steps were suggested by the presentation given by Dirk Schadendorf, MD on the results of the German Oncology Group’s randomized adjuvant phase II trial of ipilimumab /nivolumab versus nivolumab versus placebo, showing very impressive benefit albeit with modest number, in the 50s per arm. [The study suggests that] ipilimumab plus nivolumab in standard doses is superior to nivolumab alone and, of course, to placebo. This suggests that the results of the CheckMate 915, which will come up in a year or a year in a half, may well be positive. I certainly hope that is the case.
The important thing about the CheckMate 238 trial is that in another year, we'll have the 48-month follow-up, we'll see a real plateau, and we'll also have survival data that I hope will show a significant augmentation of survival in a PD-1 adjuvant study compared to an active comparator arm for the first time.
TARGETED ONCOLOGY: Are there any other checkpoint inhibitors being explored in melanoma?
Weber: In melanoma, ipilimumab, nivolumab, and pembrolizumab are the 3 approved checkpoint inhibitors. At this ESMO, we did here about an unsuccessful study of the PD-L1 antibody atezolizumab (Tecentriq) when it was combined with cobimetinib (Cotellic), the MEK inhibitor. I don't think there's going to be huge impetus to try to get only PD-1 or other checkpoint inhibitors approved in melanoma anytime soon. We do have new phase III studies that will mature in the next year or so of lag-3 plus nivolumab versus nivolumab, we'll have the pegylated IL2 molecule, NKTR-214 with nivolumab versus nivolumab. We'll have a bunch of new combination coming up, but as single agents, I think nivolumab and pembrolizumab with be it in melanoma of the time being.
Weber JS, Vecchio MD, Mandala M, et al. Adjuvant nivolumab (NIVO) versus ipilimumab (IPI) in resected stage III/IV melanoma: 3-year efficacy and biomarker results from the phase III CheckMate 238 trial. Presented at ESMO 2019; September 27-October 1, 2019; Barcelona, Spain. Abstract 13100.