Dr. James P. Allison speaks during the Education Session “On the Shoulders of Giants: Historical Approaches to Immunotherapy in Solid Tumors.”
James P. Allison, PhD, of The University of Texas MD Anderson Cancer Center, and Suzanne L. Topalian, MD, of the Bloomberg-Kimmel Institute for Cancer Immunotherapy, highlighted the transformative nature of immune checkpoint inhibitors on cancer treatment and outlined a roadmap for future research.
“We have a good chance of curing some forms of cancer in the next few years,” Dr. Allison, who was instrumental in the development of this class of agents, said.
“[Dr. Allison’s] development of the concept of immune checkpoint blockade has transformed cancer therapy and saved thousands of lives,” Dr. Topalian, who chaired the session, said.
To date, the U.S. Food and Drug Administration (FDA) has approved six immune checkpoint inhibitors for more than nine cancers, including one approved on the basis of a tumor’s genetic makeup rather than the tumor type.
“It is important to remember how we got here—by understanding the fundamental mechanisms of T-cell activation and regulation,” Dr. Allison said.
Indeed, he was the first to identify not only the costimulatory proteins required to trigger the T-cell proliferation that begins the immune response, but, in the mid-1990s, also identified an “off” signal driven by the CTLA-4 protein receptor. It is expressed by T cells and overrides stimulatory molecules, effectively applying the brakes on any T-cell action.
The hypothesis then, he said, was that because tumors do not co-stimulate T cells, they can grow undetected by the immune system until significant necrotic cell death causes inflammation, triggering a T-cell response. However, it also triggers the “stop” program. “If the T cells can’t eliminate the tumor cells before CTLA-4 takes over, then the tumor wins,” Dr. Allison said. Suspending the brakes and blocking the CTLA-4 antigen “would give the T cells a better chance.”
The results in early mouse models injected with a CTLA-4 antibody were “astounding,” he said, with numerous types of cancers effectively cured.
“This was compelling for a number of reasons,” Dr. Allison said. “Because we were treating the immune system, not the cancer cells, it could work in any type of cancer.” Indeed, in the mouse models that followed, he said, “We found no tumors that we couldn’t eliminate.”
Dr. Suzanne L. Topalian speaks during the Education Session “On the Shoulders of Giants: Historical Approaches to Immunotherapy in Solid Tumors.”
One unique aspect of checkpoint inhibitors is that their benefits continue for years as a result of the continued antitumor activity of T cells. For instance, 10 years after enrolling in the original clinical trial for ipilimumab, approximately 20% of patients are still alive. “Virtually no one is dying from their disease,” Dr. Allison said. “It gives us hope that we can really cure at least a fraction of patients.”
PD-1 Checkpoint Inhibitors
In 2001, Dr. Allison and others identified the PD-1/PD-L1 immune checkpoint pathway. T cells express the former; cancer cells express the latter. “It is a central checkpoint that dams the immune system response against cancer,” Dr. Topalian said.
Clinical trials for the first approved PD-1 inhibitor, nivolumab, showed that blocking PD-1 could cause regression of several cancers. This, Dr. Topalian said, “became a launching pad for an avalanche of activity in this area.” Since nivolumab’s approval in 2014, one other PD-1 inhibitor, pembrolizumab, and three PD-L1 inhibitors, atezolizumab, avelumab, and durvalumab, have been approved.
Like ipilimumab, patients experience long-lasting responses to PD-1 and PD-L1 inhibitors, with one-third of patients with metastatic melanoma who received nivolumab still alive after 5 years. In lung cancer, data show that 16% of patients were still alive after 5 years. Historically, Dr. Topalian said, only about 4% would be expected to be alive.
PD-1 inhibitors have now been approved for several indications, demonstrating “that anti–PD-1 therapy can be a common denominator for anticancer therapy, with one treatment approach applied to many disease settings,” she said.
Both speakers highlighted several critical issues in the further development of immune checkpoint targeting.
One problem, Dr. Allison said, is that “the clinical studies are outstripping our knowledge.” He noted that there are more than 250 presentations at the 2017 ASCO Annual Meeting this year involving checkpoint inhibitors, with more than 1,000 studies underway. Mechanistic insight is needed to rationally select the most appropriate combinations, he said.
He also stressed the need for reverse translation—using clinical outcomes to drive basic science research, something that is necessary, but underappreciated, and “grossly” underfunded.
Addressing Side Effects of Checkpoint Inhibitors
The side effect profile of this class is also a concern, Dr. Topalian said, particularly immune-related effects.
Although the therapies release the brakes on antitumor immune response, this can sometimes have a spillover effect, leading to an inflammatory response in normal tissue. The incidence of such effects, she said, is much higher when two checkpoint inhibitors are used together. “This makes it even more important than ever to find biomarkers to tell us which patients should receive the combination and which don’t need to.”
The underlying biologic processes that lead to the side effects seen with checkpoint inhibitors is a very active area of research, she said, with some clues emerging from the clinical setting.
For instance, it appears that the drugs may exacerbate pre-existing yet subclinical autoimmune disease. She described a patient with metastatic kidney cancer treated with an anti–PD-1 agent who developed severe myasthenia gravis soon after starting treatment. An assay of his pretreatment serum showed high levels of acetylcholine antibodies, confirming that the myasthenia gravis existed prior to therapy.
An organ-specific injury may also trigger the development of an autoimmune side effect, she said, describing the case of a patient with metastatic melanoma who had an excellent response to a PD-1 inhibitor, but then experienced renal failure 8 months later after receiving an intravenous contrast dye for a routine CT scan. A kidney biopsy showed intense T- and B-cell infiltrates, suggesting activation of cellular and serologic autoimmunity in reaction to the dye.
In some cases, Dr. Topalian said, immune-related side effects can result from cross-reactivity of an immune attack against a tumor-associated antigen. This occurred in a patient with metastatic melanoma who had a complete regression that lasted for years after stopping treatment. However, he simultaneously developed vitiligo, a patchy blanching of the skin. A biopsy of the blanched skin showed that his immune system had destroyed the skin’s melanocytes, which express several proteins similar to those expressed by melanoma cells.
Interestingly, she said, patients who experience immune-related side effects are more likely to exhibit tumor regression. In one study of patients with advanced melanoma treated with an anti–PD-1 therapy, the overall response rate was about 31%, but patients with three or more immune-related side effects had a response rate of more than 80%. “So there is some linkage between antitumor immunity and autoimmunity with these drugs,” Dr. Topalian said.
The Search for Biomarkers
Dr. Allison and Dr. Topalian stressed the importance of identifying biomarkers for efficacy. One such biomarker already in use is PD-L1, with studies finding high expression associated with a greater likelihood of response. However, the response differs across tumors, with some patients with PD-L1–negative tumors demonstrating a response in certain tumor types. This is the reasoning behind the approval of pembrolizumab in non–small cell lung cancer as a first-line treatment for those with high PD-L1 expressing tumors and as a second-line treatment for those with lower expression, Dr. Topalian said.
There is the possibility that cancer genetics could guide immunotherapy, she said. It appears that mutational load correlates with responsiveness to immune checkpoint blockade. Indeed, cancers that are most responsive to PD-1/PD-L1 inhibitors tend to have the greatest number of mutations. However, she said, there are exceptions on both sides.
Tumors with microsatellite instability resulting from mismatch repair deficiency have a particularly strong response to PD-1 inhibitors regardless of cancer type. This finding led the FDA to approve pembrolizumab in May as a second-line therapy for patients with nonresectable or metastatic microsatellite instability–high solid tumors or those with mismatch repair deficiency, regardless of tumor type.
Other areas of intense research include the role of viruses in checkpoint inhibitor response; the effect of the T-cell repertoire and functional state on response; and the effect of other cells in the tumor environment.
“All of this work is going to keep us busy for some time ahead,” Dr. Topalian concluded.
–Debra Gordon, MS