Antoni Ribas, MD, PhD: The combination of nivolumab and ipilimumab for melanoma is the highest-activity immunotherapy that I’ve ever seen in this disease. To have over 50% of patients with objective responses—many of them that appear really early on, most of them that are going to be long-lasting—is spectacular. The biggest drawback is the side effects. And some patients are really hurt by this combination. But I would like to say that the 50% to 55% grade 3 to 4 toxicity tells us also that 45% to 50% of the patients do not have serious toxicity. So, roughly half of the patients go through the therapy with no serious side effects.
We always worry about the percentage of patients where we have to stop the therapy and try to reverse the course by giving corticosteroids or other agents that are immunosuppressive. It doesn’t make that much sense that we try to turn on the immune system and then we have to go back and turn it off because it’s been too toxic.
We still don’t have a balance about these. We don’t know how to select these patients. We don’t know how to tease out patients who will benefit with just nivolumab or pembrolizumab, as opposed to having the combination. We also don’t know if patients who could not respond to single-agent PD-1 could be rescued by getting the combination. We’re planning a clinical trial in SWOG to be able to test that and hope that will give us light on, do we get the difference in response rate that can be done by giving, in choosing ipilimumab later? But that’s not an approved indication. We don’t know that that will work, so we’ll have to wait for results of that trial.
I think probably the most important question we have in the melanoma field is, should we give a PD-1–blocking antibody alone or should we give ipilimumab and nivolumab up front to patients with metastatic melanoma? The clinical field has gone faster than the scientific knowledge in this case. It is one of the few times where we have a combination demonstrated that works before we figure out why it does so. If we knew exactly why the combination is better than single agent, we would be able to then think about in which patients we need the combinations as opposed to a single agent.
I think this data will be generated in the near future and they are of key importance to the field. We also don’t know the sequence of these. Can we sequence a single agent with combination immunotherapy or do we always need to start up front with the combination and then go on with nivolumab alone, which is how it’s been approved? It’s a big unmet medical need, and I hope the community is getting together to be able to tackle this important question.
The evidence from the combination of ipilimumab and nivolumab increased the sealing of activity of a PD-1–blocking antibody and tells us that we can probably do this with other combinations, that we have not reached what these therapies can do by themselves or in combination. There’s other immunotherapies that are being tested in combination with PD-1–blocking antibodies. PD-1–blocking antibodies are attacking a common negative regulator of when an immune system cell, a T cell, is trying to attack the cancer and it’s trying to kill it and the cancer protects itself.
So, we can think of combining this blockade with something that blocks upstream. CTLA4 works upstream, but also other agents or other immunotherapies could block it upstream. There’s a series of immune-activating antibodies—CD4E, to OX4E, to 4-1BB or CD137—that are in clinical trials with the idea of, putting it simply, step on the gas and release the brake at the same time. We can release two brakes at the same time. CTLA4 is the first one, but there’s also TIM-3, there’s VISTA, there’s ADAMs, there’s DIGIT, all of them getting into clinical trials.
There are other non–T-cell-centric brakes, like IDO, that are expressed in the cancer cells and in macrophages that cripple their function of T cells. Macrophages are becoming an important cell subset that is immune-suppressed in tumors. The tumor taking advantage from macrophages and blocking them with CSF1 inhibitors, receptor inhibitors, could be a way to go.
There’s a renewed interest in local therapies that may change the tumor microenvironment in one place and then get the immune system cells from there to go around the body and attack metastatic lesions in other places. There are studies of injecting oncolytic viruses, like T-VEC, into tumors in combination with PD-1–blocking antibodies. Toll-like receptors are also being tested. But then radiation therapy, where it’s an ultimate localized therapy, may also allow an immune response to be created and then amplified by blocking PD-1. They have less excitement in combination with chemotherapies because most chemotherapies are DNA toxins. And T cells need to proliferate to be able to have their effect. So, I’m not that hot on them, but people are testing them in different indications.
Combinations with targeted therapies, I think, have a lot of mileage to go and a lot to be tested. Because if you have a target that’s in the cancer and it’s not on the T cell, then you can cripple the cancer and get the immune system to attack the beta. So, there’s a lot of activity in there in many cancers.
I think the biggest challenge of combination immunotherapy is being able to select the patients, because the toxicity from the combination could be avoided if we can select the patients that benefit to one of the agents. That’s what’s limiting their immune system. It’s something where we have to go backwards from the clinic to the laboratory and be able to understand that.