CHICAGO – Vemurafenib has a persistent survival benefit when given as first-line therapy for advanced BRAF-mutant melanoma, but the magnitude of benefit diminishes over time, according to an update of the BRIM-3 randomized trial.
With a median follow-up of about a year, vemurafenib was associated with a 30% relative reduction in the risk of death, relative to dacarbazine, presenting author Dr. Paul Chapman reported at the annual meeting of the American Society of Clinical Oncology. This is roughly half of the 63% relative risk reduction seen at a median follow-up of about 3 months, as reported at last year’s meeting.
Dr. Paul Chapman
The safety profile of vemurafenib was largely consistent with the earlier experience; the main high-grade adverse effects were skin toxicity (seen in about a third of patients) and liver toxicity (seen in about a tenth). But a few patients have developed second primary melanomas.
The observed temporal trend for overall survival benefit suggests "perhaps that most of the benefit in overall survival may be due to decreasing early deaths," said Dr. Chapman, an oncologist at Memorial Sloan-Kettering Cancer Center in New York.
Discussant Dr. Michael B. Atkins of the Lombardi Comprehensive Cancer Center at Georgetown University in Washington maintained that "the mature overall survival data [confirm] that vemurafenib is better than dacarbazine," and noted that it was "reassuring" that findings were similar, whether patients who crossed over to the vemurafenib arm were censored or not, "probably the most stringent test of survival."
However, he called attention to the fact that the median survival of 13-plus months with vemurafenib seen in BRIM-3 falls short of the 16-plus months seen in BRIM-2, a phase II trial in patients with previously treated advanced melanoma. "Did first-line patients have more aggressive disease and therefore fall off early?" he wondered. "Alternatively, was there less continuation of vemurafenib after progressive disease in the BRIM-3 trial relative to the BRIM-2 trial? If that was the case, some sort of confirmatory analysis should be done to see whether that can impact overall survival."
The decrease in vemurafenib efficacy seen over time might have several explanations, Dr. Atkins proposed. "Is it that the early, more robust hazard ratio was a manifestation of the so-called Lazarus effect, that these targeted therapies can cause patients who are really in bad shape and would have a short survival to have their survival prolonged, even though it was less than the survival of patients with less-aggressive tumors?" he wondered, echoing Dr. Chapman’s suggestion. Alternatively, maturation of the survival data might have contributed to the change in efficacy in patients with stage IIIC, M1a, and M1b disease who did not appear to benefit from vemurafenib.
In the larger context, the findings raise an important question pertaining to the clinical trials process generally, according to Dr. Atkins: "What does this say about the early analysis of trials like this, and reporting forever more that the hazard ratio is 0.35 or so, rather than what the more accurate 0.70 is?"
A Genentech spokesperson said that "due to differences in the study populations and trial designs, the results from these trials cannot be directly compared. BRIM3 was a global, randomized, Phase III trial comparing Zelboraf (vemurafenib) to DTIC chemotherapy in people with newly diagnosed BRAF V600 mutation-positive metastatic melanoma, while BRIM2 was a Phase II study that evaluated patients with previously treated disease."
The 675 patients in BRIM-3 had stage IIIC or IV melanoma harboring the BRAF V600E mutation. They were assigned evenly to intravenous dacarbazine or oral vemurafenib, the only targeted agent approved specifically for treating BRAF-mutant melanoma.
Large proportions of patients in both the vemurafenib and dacarbazine arms (36% and 44%, respectively) received other anticancer therapies after coming off the trial, Dr. Chapman reported. About a fifth of all patients in each arm went on to receive ipilimumab, and a fourth in the dacarbazine arm crossed over to vemurafenib.
Updated results – after a median follow-up of 9.5 months in the dacarbazine arm and 12.5 months in the vemurafenib arm – showed persistence of a progression-free survival benefit in favor of vemurafenib with censoring of patients at crossover (6.9 vs. 1.6 months; hazard ratio, 0.38; P less than .001).
There was also still an improvement in overall survival, regardless of whether patients were censored at crossover (13.6 vs. 9.7 months; HR, 0.70; P less than .001) or not (13.6 vs. 10.3 months; HR, 0.76; P less than .01).
The leading grade 3 or worse adverse events with vemurafenib remained cutaneous squamous cell carcinomas (seen in 19% of patients), keratoacanthomas (10%), and elevations of liver function tests (10%). In addition, eight patients developed new primary melanomas.