Dr. Robert Dreicer
Question: In urothelial carcinoma, is there a difference in efficacy or safety among agents that target PD-1 versus agents that target PD-L1?
Answer: Given the astonishing speed in which anti–PD-1/–PD-L1 agents have become an integral part of our therapeutic armamentarium in an ever-growing number of neoplasms, including advanced urothelial cancer, it should come as little surprise that the answer to this important question is, “We just don’t know.”
Although the PD-L1 inhibitor atezolizumab received U.S. Food and Drug Administration approval in May 2016 for the management of advanced urothelial cancer following platinum-based chemotherapy, the study that led to this approval included only 310 patients.1 In an update of this now-published experience presented during the 2016 ASCO Annual Meeting, the objective response in all patients was 16%, and at a median of nearly 18 months, 35 of 49 patients (71%) had ongoing objective responses.
Although higher levels of PD-L1 expression on immune cells predicted a higher likelihood of response (i.e., 28% of patients with immune cells 2/3), 9% of patients without evidence of PD-L1 expression still achieved an objective response. Therefore, although approved with a companion diagnostic to test for PD-L1 expression, testing is not required for atezolizumab use. The most common side effects were fatigue and gastrointestinal disorders, with a low rate of serious immune adverse events.2
Massard et al. presented phase I/II study data of durvalumab, a PD-L1 inhibitor, in 61 patients who were heavily pretreated with chemotherapy. In this experience, responses were essentially limited to patients expressing PD-L1 on either tumor or immune cells, in contrast to the atezolizumab experience. Toxicity was similar to that reported for both nivolumab and atezolizumab.3
Sharma et al. presented data from the phase I/II CheckMate 032 study, which evaluated the role of nivolumab, an anti–PD-1 inhibitor. In 78 patients who were heavily pretreated (unselected by PD-L1 expression), 24.4 % obtained an objective response, with 52.0% of all patients alive at 1 year, and nivolumab had a reasonable safety profile in these patients.4
Reviewing the relatively limited data to date, it is clear we have much work to do. The basic issues of the importance of PD-L1 expression on tumor cells versus immune cells, the timing of the assessment (i.e., fresh vs. archival material), and the diversity of assays used remain undefined. At a very high level, although the toxicity profile of these agents appears similar, clearly no conclusions can yet be drawn.
At the time of writing, the issue is relatively straightforward as atezolizumab is currently the only U.S. Food and Drug Administration–approved agent for this indication. A large number of phase II and phase III studies are ongoing, which will provide additional insight into the clinical utility and appropriate setting for these agents. I do not currently envision that we will have comparative data from trials in advanced urothelial cancer to address the therapeutic question of targeting PD-1 versus PD-L1, and I suspect this particular issue may ultimately not be a critical one.