Early Phase I data reported last year at the American Society of Clinical Oncology Annual Meeting and in the New England Journal of Medicine demonstrated that nivolumab, a novel anti-PD-1 antibody, was safe and active in melanoma patients. Now, with longer follow-up, researchers reported at the American Society of Clinical Oncology that they see long overall survival and durable responses in melanoma patients treated with nivolumab in the Phase I trial (Abstract CRA9006).
“The PD-1 studies are incredibly exciting,” said Lynn Schucter, MD, the C. Willard Robinson Professor of Hematology-Oncology and Attending Physician at the University of Pennsylvania Hospital, speaking during a news conference at the meeting. She emphasized the need for melanoma patients to enroll in clinical trials so that they can receive this drug and ones like it, given the apparent benefit in early-phase trials: “I think for melanoma patients, getting access to these new agents is really the most critical part of their care.”
Favorable One- and Two-Year Survival Rates
A total of 107 patients with metastatic melanoma enrolled in the trial, which was led by Mario Sznol, MD, Professor of Medical Oncology at Yale Cancer Center. To be eligible, patients had to have had at least one prior therapy but no more than four for advanced disease and no prior treatment with ipilimumab. Patients received intravenous nivolumab every two weeks at a dose of 0.1 to 10 mg/kg.
The patients had a median age of 61, and 97 percent had an ECOG performance status of 0 to 1. However, the group was heavily pretreated, with 66 percent of patients having had at least two prior therapies and 25 percent, three or more. Additionally, 78 percent of patients had visceral metastases and 36 percent had elevated lactate dehydrogenase, which is associated with a poorer outcome.
The investigators saw no dose limiting toxicities. The overall rate of serious adverse events was low at five percent, including two cases each of skin, gastrointestinal, and endocrine toxicities and one case each of hepatic and renal toxicities. Any grade adverse events occurred in 54 percent of patients, with skin toxicities the most frequent, affecting 36 percent of patients, followed by gastrointestinal (18%) endocrinopathies (13%), hepatic (7%), infusion reaction (6%), pulmonary (4%), and renal (1%) side effects.
“There was no additional safety signal seen with the additional year of follow-up,” Sznol said. The overall response rate was 31 percent across all doses, with no apparent dose response. The median duration of response was two years and 45 percent of the responders showed a response at the first tumor assessment at eight weeks, indicating that the responses can be both rapid and durable.
Moreover, 12 of 17 patients who went off drug for reasons other than disease progression continued to respond for at least 16 weeks, and eight of those were continuing response at the time of the data analysis (range of 16 to 56 weeks). “You don't need continued treatment in a subset of patients in order to maintain response,” Sznol said.
Pointing to the Kaplan-Meier curve for progression-free survival, he noted that the curve dropped rapidly at first but then flattened out. “Although the median progression-free survival of 3.7 months is not overly impressive, the one- and two-year progression-free survival rates of 36 and 27 percent are,” Sznol said.
The median overall survival for the study group is 16.8 months. The one-year estimated overall survival rate is 62 percent and the two-year rate is 43 percent. With a median follow-up of 22 months (range 14 to 51 months), 47 patients remain alive.
During the discussion that followed Sznol's talk, several audience members remarked how interesting the data were. Kevin Kim, MD, Associate Professor in the Department of Melanoma Medical Oncology at the University of Texas MD Anderson Cancer Center, asked about the optimal duration of treatment with nivolumab. Sznol said the optimal duration remains unclear, but that his group at Yale have discontinued therapy on two patients with near complete responses, one after three cycles and the other after five cycles, neither patient has relapsed.
“So at least at complete responders, you probably don't need to continue the drug,” he said. “The question is, what do you do in patients with stable disease or partial response. Do you continue to give them drug every two weeks or do you decrease the interval of dosing? I don't think we have an answer to that question.”
Another audience member asked if it was possible to biopsy lesions in responders to see if the lesions really contain residual disease. “I think the rate of complete response here is a little bit underestimated,” Sznol responded. “Of the five patients we treated at Yale who are responders, four are complete responders or near-complete responders and the fifth patient, who has a liver nodule, is PET negative.
“So there wasn't much for us to biopsy in our long-term responders. We have biopsied a few patients with progressive disease, but we haven't analyzed that tissue yet.”
Finally, one audience member asked about the likelihood that the drug works on brain metastases. Sznol noted that this trial excluded patients with active brain lesions, but accepted patients with previously treated central nervous system tumors. Therefore the answer to the question remains unknown. “But we have long-term responders who didn't develop any brain metastases, so that suggests that maybe we are controlling disease in the brain,” he said.
There are now three ongoing Phase III trials testing nivolumab in patients with metastatic melanoma. A trial testing the drug in patients with brain metastases has been proposed, but Sznol said he did not know if it was approved or would go forward.