The programmed death receptor-1 (PD-1) ligand, PD-L1, has become a viable target for immunotherapy in cancer, with multiple antibodies now in development.
Michael A. Postow, MD
The programmed death receptor-1 (PD-1) ligand, PD-L1, has become a viable target for immunotherapy in cancer, with multiple antibodies now in development.1Because they have not been givenformalnomenclature yet, these include 3 alphanumerically defined agents, MEDI 4736, MSB 0010718C, and MPDL3280A.2-4As these drugs enter clinical use, the question arises whether these 2 classes of agents (anti-PD-1 and anti-PD-L1) should be used in combination or sequentially.
The PD-1 pathway provides an important circuit for immuneregulation, with the PD-1 receptor expressed on T cells, and the PD-L1 ligand expressed in multiple tissues and cell types.1PD-L1 functions to provide a level of immune inhibition to prevent uncontrolled, destructive, T-cell-mediated responses and autoimmunity.1
Michael A. Postow, MD, an oncologist at Memorial Sloan Kettering Cancer Center, explained that tumors can exploit the PD-1 axis to evade immune detection and T-cell attack, thus providing a rationale for use of both classes of anti-PD-1 agents.
“It’s certainly very complex, but the belief is that the tumor expresses PD-L1, the ligand, and T cells express PD-1. PD-L1 is believed to be expressed on the plasma membrane of tumor cells, and utilizing an anti-PD-1 agent thus blocks both the ability of the T-cell to recognize the ligand, as well as the ligand’s ability to interact with its receptor,” he said.
The imminent availability of these 2 different classes of anti-PD-1 agents also allows for their potential use in combination or in sequence, whereby progression of a patient on one agent (eg; an anti-PD-1) would allow for the use of another (ie, an anti-PD-L1), or both together, although Postow pointed out that the latter has not yet been explored.
“That’s probably something of interest in the future, however, to my knowledge there are no studies where both anti-PD-1 and anti-PD-L1 are being used at the same time” he said.
With regard to using these 2 types of immunotherapies in a sequential manner, Postow noted, “Sequentially is a great questionone where we have zero information on whether or not that’s effective—but I do think it’s incredibly relevant whether or not a patient who progressed after anti-PD-1 could benefit from anti-PD-L1. There is a mechanistic rationale to believe a patient could be helped with PD-L1 after PD-1, and hopefully it will be possible in the future to explore that possibility, but there’s nothing yet [in terms of clinical trials] that’s proven that that’s effective.”
As with other types of anticancer therapy, biomarkers ultimately might help to define patient populations for whom a combination and/or sequential approach might be most appropriate, and in this regard, tumor expression of the ligand, PD-L1 is a logical candidate.
“Many tumors are believed to express the ligand, and expression rates in melanoma are approximately 30%, but all of this expression depends on how you define positive expression,” Postow said. He noted that most conventional immunohistochemistry-based tests have used 5% expression as a reasonable cutoff, although a specific expression level has not been firmly established.
“This is the commonly used cut point” he says, noting that thus far expression of the PD-L1 ligand has only been examined using protein expression.
“It’s possible that other cut points might give you different results in terms of the outcomes studies, but it’s not set in stone 100% that this is the right cut point,” Postow said.
Despite the lack of firm consensus on expression level, however, Postow emphasized that a lack of PD-L1 expression does not preclude the use of either type of immunotherapy (ie, anti-PD-1 or anti-PD-L1).
“Frankly, even without the tumor expression of PD-L1, it would be reasonable to try [an immunotherapy] approach, because we know that patients who do not have PD-L1 in the tumor can also benefit from these drugs. The absence of PD-L1 in the tumor does not exclude the possibility of benefit from a PD-1 or PD-L1 drug,” Postow said.
Notably, in terms of ligand expression as a biomarker, Postow also emphasized that this has not been studied to the point where it can be used to guide clinical decision-making. For example, in the case of advanced melanoma, all patients could potentially be candidates for immunotherapy.
“Any patient, really, that’s a candidate for a PD-1, PD-L1 trialwe would encourage their trial participation, regardless of the PD-L1 expression set,” Postow said. At this point in development, he also does not anticipate much difference between anti-PD-1 and anti-PD-L1 agents regarding efficacy or safety.
“I imagine they should be generally similar. There are some subtle differences that may ultimately result in differing side effects or efficacy, but at this point, this remains unknown," Postow said.
While trials of anti-PD-1 agents (eg; nivolumab, pembrolizumab) have been published,4-6trials of anti-PD-L1 agents are still ongoing.
At ASCO 2013, Roy Herbst and coworkers from Yale Cancer Center presented data for the anti-PD-L1 agent MPDL3280A in patients with advanced solid tumors (n = 122 evaluable). The results showed the treatment to be well tolerated, with responses observed in multiple tumor types, including melanoma, colorectal cancer, renal cell carcinoma, and lung cancer.
The overall response rate (ORR) reported was 21% in the study, and PD-L1 expression appeared to be associated with responses, with a respective ORR of 39% and 13% in patients with, and without, positive PD-L1 status, while the rate of progressive disease was 12% and 59%, respectively.2
At ASCO 2014, Neil H. Segal and coworkers from Memorial Sloan Kettering also presented data from a multiarm expansion study of the anti-PD-L1 monoclonal antibody MEDI4736.4NSCLC, melanoma, gastroesophageal cancer, hepatocellular carcinoma, pancreatic cancer, and head and neck cancer were among the tumor types examined in the study (n = 151 patients).
Treatment-related adverse events (AEs) overall, and those of grade 3 or higher, occurred in 33% and 7%, respectively, but none led to drug discontinuation. At a median 6 weeks of follow up, tumor shrinkage by Response Evaluation Criteria In Solid Tumors (RECIST) was detectable in multiple tumor types.4
At ASCO 2014, pharmacokinetic and pharmacodynamic data for the anti-PD-L1 monoclonal antibody MSB0010718C were presented.3The phase I dose-escalation study enrolled 27 patients with advanced malignancies, and results showed this agent could be safely administered at doses up to 20 mg/kg with intravenous dosing every 2 weeks.