Spartalizumab-based combination fails in BRAF-mutated advanced melanoma
Click Here to Manage Email Alerts
The addition of spartalizumab to dabrafenib and trametinib failed to prolong PFS among patients with BRAF-mutated advanced melanoma, according to results from the randomized phase 3 COMBI-i trial presented at ESMO Virtual Congress 2020.
“The emergence of BRAF and MEK inhibitor targeted therapies and immune checkpoint blockade has transformed the management of metastatic melanoma over the past few years,” Paul Nathan, MD, consultant medical oncologist at Mount Vernon Cancer Centre in Northwood, United Kingdom, said during his presentation. “However, despite these advances, many of our patients experience disease progression and we do need new treatment strategies.”
Preclinical data have suggested the addition of an anti-PD-1 antibody, such as spartalizumab (PDR001, Novartis), to the combination of the BRAF inhibitor dabrafenib (Tafinlar, Novartis) and the MEK inhibitor trametinib (Mekinist, Novartis) may enhance antitumor activity, he added.
“Phase 2 and phase 3 trials have shown that combining checkpoint inhibitors and targeted therapy may improve outcomes in BRAF V600-mutant metastatic melanoma,” Nathan said. “Also, an early clinical finding suggested that spartalizumab plus dabrafenib and trametinib may be associated with a higher percentage of patients achieving durable responses.”
Based on this rationale, Nathan and colleagues randomly assigned 532 patients with BRAF V600-mutated unresectable or metastatic melanoma to receive 150 mg dabrafenib twice daily plus 2 mg trametinib once daily with either 400 mg IV spartalizumab every 4 weeks (n = 267) or placebo (n = 265).
Investigator-assessed PFS served as the study’s primary endpoint, with a predetermined significance threshold of P < .025 based on 352 target events to ensure 80% power and an assumption of a 5-month delay in treatment effect. OS served as a key secondary endpoint, but this endpoint could only be assessed upon demonstrating significance of the primary endpoint.
The primary analysis occurred after a minimum follow-up of 24 months (median, 27.2 months; range, 24-33.6) and 312 events, leading to a statistical threshold of P = .02497.
At the time of the data cutoff in July, 37% of patients assigned spartalizumab-dabrafenib-trametinib discontinued treatment due to progressive disease compared with 48% of patients assigned dabrafenib-trametinib. However, more patients assigned the triplet had stopped treatment due to an adverse event (19.1% vs. 8.7%).
Median PFS was 16.2 months with the triplet compared with 12 months with the doublet, which did not meet the study’s threshold for significance (HR = 0.82; 95% CI, 0.65-1.02).
“We noted that the performance of patients receiving the doublet was superior to that we’ve seen historically with the COMBI-d and COMBI-v studies,” Nathan said.
Estimated PFS rates favored the triplet at 12 months (58% vs. 50%) and 24 months (44% vs. 36%) but were not statistically different.
Researchers noted “interesting trends” in favor of the immunotherapy-containing triplet on PFS subgroup analyses, Nathan said.
For instance, there was a trend toward PFS benefit with the triplet with increasing lactate dehydrogenase levels ( 1 to < 2 x the upper limit of normal [ULN], HR = 0.78; 95% CI, 0.52-1.17; 2 x ULN, HR = 0.72; 95% CI, 0.42-1.25), more sites of metastases ( 3, HR = 0.63; 95% CI, 0.46-0.87), greater bulk of disease (baseline sum of lesion diameters 66 mm, 0.52; 95% CI, 0.37-0.73) and the presence of high tumor mutational burden ( 10 mut/Mb, HR = 0.7; 95% CI, 0.47-1.06).
Although OS was not formally tested for statistical significance, median OS was not reached, with an HR of 0.78 (95% CI, 0.58-1.04).
The objective response rate (68.5% vs. 64.2%) and complete response rates (19.9% vs. 17.7%) appeared comparable between the triplet and doublet groups. The median duration of response had not been reached with the triplet and was 20.7 months with the doublet.
A greater proportion of patients assigned the triplet experienced grade 3 or worse treatment-related adverse events (54.7% vs. 33.3%) and adverse events leading to discontinuation (12.4% vs. 8%).
“We saw differences in dose modifications and discontinuations between patients receiving treatment on the triplet arm compared with the doublet, reflecting the increased toxicity of this combination,” Nathan said. “Additional analyses are ongoing to better understand these results. Further OS follow-up may provide additional insights.”
Perspective
Back to Top
Bartosz Chmielowski, MD, PhD
Unfortunately, this trial did not meet its primary endpoint based on the statistical design of the trial. How did this happen?
This is not the first clinical trial that addressed a similar question. We know the results of KEYNOTE-022, which evaluated dabrafenib and trametinib plus pembrolizumab (Keytruda, Merck) or placebo. This trial contained only 120 patients, with PFS as the endpoint. This trial also was statistically negative and did not meet its primary endpoint.
Another clinical trial addressing a similar question was IMspire150, a trial of vemurafenib [Zelboraf, Genentech/Roche] plus cobimetinib [Cotellic, Genentech/Roche] with atezolizumab [Tecentriq, Genentech/Roche] or placebo, with a run-in period of targeted therapy alone, and PFS as the endpoint. This clinical trial did meet its primary endpoint and the therapy is currently approved for patients with melanoma in the United States.
Where does the difference come from?
When we overlay the PFS curves from the KEYNOTE-022 trial with the IMspire150 trial, we can clearly see when these curves overlap and it’s very difficult to see a difference. When we add the current trial, we also see the experimental curve overlays with the other two curves, and the control curve is a little bit better than the other two [control] curves. It tells us that there is no clinical difference, no matter which of these three therapies is used. The only difference is based on the statistical design of the trial, determining whether the trial became positive or negative. It also tells us that minimal differences in the patient populations will result in significant changes in the statistical result.
We could also argue that the patient population included in this trial had a better prognosis, with most of the patients having normal lactate dehydrogenase and few metastatic sites. When we look at the subset analysis, we see that patients with more than three metastatic sites had the biggest benefit from triplet therapy.
One could also argue that not every targeted therapy is the same, as some of the control arms might do better or worse.
Are we asking the correct question with all of these clinical trials? We know that targeted therapy increases the antigen presentation in melanoma and increases T-cell infiltration in the tumor. But, we are not asking the question of whether targeted therapy would increase the efficacy of immunotherapy, rather the opposite. To answer this question, we would need a clinical trial where everyone is treated with PD-1 therapy and patients are randomly assigned to targeted therapy or placebo. This question will be answered by the STARBOARD trial, where pembrolizumab is being evaluated with encorafenib [Braftovi, Pfizer] and binimetinib [Mektovi, Pfizer] or placebo.
References:
Ascierto PA, et al. Nat Med. 2020;doi:10.1038/s41591-019-0448-9.
Gutzmer R, et al. Lancet. 2020;doi:10.1016/S0140-6736(20)30934-X.
Jonsson Comprehensive Cancer Center
Collapse
Nathan P, et al. Abstract LBA43. Presented at: European Society for Medical Oncology Virtual Congress 2020; Sept. 19-21, 2020.
Click Here to Manage Email Alerts