Treatment Paradigm in Advanced Melanoma Poised for Change… Again

Get Permission

Caroline Robert, MD, PhD

Axel Hauschild, MD, PhD

Antoni Ribas, MD, PhD

The most potent biomarker we found was the presence of T cells in and just outside the tumor. But at this point, I am not going to stop treatment if I don’t see T cells, and I won’t stop if PD-L1 is not expressed.

—Caroline Robert, MD, PhD

In the treatment of advanced/metastatic melanoma, recent debate has focused on the choice of initial therapy: ipilimumab (Yervoy) or, for patients with BRAF-mutant cancer, a BRAF/MEK inhibitor. This issue is now taking a back seat to the emerging conversation about the positioning of antibodies that target programmed cell death protein-1 (PD-1) and its ligand, PD-L1.

At the 3rd Annual World Cutaneous Malignancies Congress in Los Angeles, melanoma researchers weighed in on this topic and made predictions for the future. Antoni Ribas, MD, PhD, Professor of Medicine, Surgery, and Molecular and Medical Pharmacology, and Director of the Tumor Immunology Program at the Jonsson Comprehensive Cancer Center at the University of California, Los Angeles, predicted that the current treatment algorithm for advanced/metastatic melanoma is poised for change.

Management of Cancer in the Anti–PD-1/PD-L1 Era

Today, patients are initially evaluated for BRAF status, burden of disease, and the growth kinetics of the tumor. Then they are offered BRAF/MEK inhibitors, the anti–CTLA-4 (anti-cytotoxic T-lymphocyte–associated protein 4) agent ipilimumab, or interleukin-2.

“However, this is not how we are going to treat melanoma in 2 years,” Dr. ­Ribas predicted. “We will be evaluating the interaction between the cancer cell and the immune system, and that will tell us what to do.”

The mechanism of action of the PD-1/PD-L1 blockers—which currently include nivolumab (Opdivo), pembrolizumab (Keytruda), MPDL3280A, MEDI4736, and MSB0010718C—may allow for the prediction of response. CD8-positive T cells observed in the tumor at baseline trigger the expression of PD-1 and PD-L1, and their interaction limits the response of the immune system and can be unblocked by anti–PD-1/PD-L1 agents.1

Caroline Robert, MD, PhD, Head of the Dermatology Unit at the Institut Gustave-Roussy in Paris, France, indicated she had collaborated with Dr. Ribas on this finding and noted, “The most potent biomarker we found was the presence of T cells in and just outside the tumor. But at this point, I am not going to stop treatment if I don’t see T cells, and I won’t stop if PD-L1 is not expressed.” Dr. Ribas added, “We need to learn more about this. It sounds complicated now, but this will become a simpler thing we can all detect.”

Only patients with this characteristic immune profile—demonstrating the presence of CD8, PD-1/PD-L1—will receive an anti–PD-1 or PD-L1 antibody up front. Other patients will most likely be treated with combinations or novel approaches, especially strategies that “can bring T cells into the tumor,” he said.

Dr. Robert agreed. “In an ideal world, we would biopsy the tumor, look for the biomarker, and be able to tell whether the levels are higher enough for a PD-1 blocker or whether the patient should get anti–CTLA-4 up front.”

Symptom Load May Still Matter

Axel Hauschild, MD, PhD, Head of the Skin Cancer Working Group at the University Hospital Schleswig-Holstein in Germany and Co-Chair of the Global Melanoma Task Force, agreed that the availability of anti–PD-1/PD-L1 antibodies will turn the current treatment algorithm on its head but maintains that symptomatic load will still be relevant.

He predicted that all symptomatic patients, regardless of BRAF status, will receive immunotherapy first—ipilimumab, alone or in combination with an anti–PD-1/PD-L1 antibody. BRAF status will become important when the disease progresses. Retreatment with combined immunotherapy or with BRAF inhibitors will also be part of the future algorithm, according to Dr. Hauschild. This strategy is not yet evidence-based and must prove its superiority to the selection of treatment by BRAF status, he added.

The future approach will be greatly informed by results of trials comparing BRAF/MEK inhibition followed by anti–PD-1/PD-L1 plus ipilimumab, vs the opposite strategy, and those evaluating the combination of anti–PD-1/PD-L1 plus ipilimumab.

Nivolumab Survival Data

Dr. Robert reviewed the encouraging findings for nivolumab that are contributing to the excitement over anti–PD-1/PD-L1 antibodies. In the phase III CA209-037 trial, response rates were 32% to nivolumab and 11% to dacarbazine, and overall survival, which she and her colleagues recently reported in The New England Journal of Medicine,2 was 72.9% at 1 year, vs 42.1%, representing a 58% reduction in mortality risk (P < .0001).

In contrast to ipilimumab, she said, “The drugs also have a very good risk/benefit ratio,” with most toxicity limited to immune-related side effects.

She agreed that the combination of the two immunotherapy classes is exciting and logical. The rationale for the combination is that CTLA-4 and PD-1 are nonredundant immune checkpoints in T-cell differentiation and function, and antitumor synergy has been demonstrated in animal models, she explained.

The phase I CA209-004 study evaluated several different dosing combinations and schedules for the two drugs.3 In the combined concurrent cohorts (which included maintenance nivolu­mab), overall survival at 1 year was 85%, and 2-year survival was 79%. In one of these individual cohorts, 1-year survival reached 94%, Dr. Robert reported.

Clinical activity was seen in patients with and without PD-L1 expression and regardless of BRAF status. In patients with residual ipilimumab in plasma, the response appeared to be enhanced, suggesting that “maybe we don’t need a lot of ipilimumab,” she added.

Toxicity May Be an Issue

The combination was, however, difficult to tolerate. In the dose-finding study, 64% of patients reported grade 3/4 events with the concurrent treatment, primarily gastrointestinal (14%) and hepatic (14%). Toxicity was much less with sequential treatment.

“These [side effects] occur early and are not trivial,” Dr. Robert noted. “We will need to see if the efficacy of this combination is worth the toxicity.”

She suggested that the viability of this combination “will depend on the results of the studies…. Maybe we will use the combination for patients with a good performance status. If the patient can overcome the toxicity and have a good response, of course it will be worth it.”

When asked what magnitude of benefit will justify the toxicity, Dr. Robert responded, “Survival of 80% without grade 5 adverse events.”

Dr. Hauschild indicated the gamble may be worth it for many patients. “Why not argue that if the combination shows a strong overall survival improvement, vs the single agent, toxicity doesn’t much matter, if the patient just needs four infusions to get the job done. You could go ahead with a single agent and get a median survival time of 20 months or go with the combination and get 30 months. It’s a dramatic difference.” ■

Disclosure: Dr. Robert has been a consultant for BMS, MSD, GSK, Roche, Amgen, and Novartis. Drs. Ribas and Hauschild reported no potential conflicts of interest.


1. Tumeh PC, et al: Nature 515:568-571, 2014.

2. Robert C, et al: N Engl J Med. November 16, 2014 (early release online).

2. Kluger H, et al: 2014 ESMO Congress. Abstract 10850. Presented September 27, 2014.