Role of PD-L1 Status in mRCC

Martin H. Voss, MD:PD-L1 testing has been investigated intensely in kidney cancer and has been in other diseases, and this actually goes beyond the application of checkpoint inhibitor therapy. There are data from the COMPARZ trial, which was a tyrosine kinase inhibitor study that compared sunitinib with pazopanib using PD-L1 immunohistochemistry and demonstrated that patients who are treated with tyrosine kinase inhibitors and have higher tumor expression of PD-L1 levels have adverse outcomes—meaning patients do worse with high PD-L1 levels—suggesting that this aspect of tumor biology may in fact add to the aggressiveness of cancer. Now several years later, we have data from the I-O trials that have helped us to understand that it not only is an adverse prognostic marker in terms of the biology of disease, but it is a predictive marker that shows favorable outcome in the immunotherapy-treated patients, which makes sense because it’s the target.

The early studies for PD-1 and PD-L1 inhibitors in this disease suggested that response rates were higher in patients who were PD-L1—expressing on tumor cells and on immune cells. It’s important to know, and has always been stressed, that lack of PD-L1 expression does not preclude patients from having benefit from these agents, so it should not lead to these patients not being considered for these trials.

The randomized phase III trial that led to the approval of nivolumab in the second-line setting, the CheckMate-025 study, compared nivolumab with everolimus in TKI-pretreated patients. And, importantly, PD-L1 expression status did not seem to affect the superiority over everolimus in these patients, meaning whether or not a patient was PD-L1—positive, they did better with immunotherapy than with the mTOR inhibitor. So, at the time, PD-L1 testing was deemed not a relevant marker to make treatment decisions. Since then, we now have the CheckMate-214 study—which, again, is the ipilimumab/nivolumab versus sunitinib trial that compared tyrosine kinase inhibitor therapy with combination I-O therapy. And, of course, PD-L1 expression status was investigated on that trial—again, we tested for expression on tumor cells, not on immune cells. And the best way to summarize the data is that we again saw an adverse signal for the TKI-treated patients, meaning the cancer-specific survival with sunitinib therapy was worse in PD-L1–expressing tumors. And on the flip side, patients who were treated with an immunotherapy combination were more likely to benefit from immunotherapy if they were high expressers.

So, for example, the CR rate in intermediate- and poor-risk patients was 9% across the board, but if we looked at patients who were PD-L1—positive and had intermediate- or poor-risk disease, the CR rate went up to 16%. If we looked at overall survival benefit of ipilimumab/nivolumab over sunitinib in PD-L1–negative versus PD-L1–positive patients, then we note—and Bob Motzer presented this at SITC in 2017—that superiority was seen in both instances, but it was much more marked in patients who were PD-L1–positive. So, the hazard ratio went down to below 0.5 in patients who were PD-L1–high expressors, comparing the I-O combination over sunitinib.

So, in my mind, it is not a marker that will help us make decisions between 2 regimens, but it’s certainly prognostic in terms of the tumor biology if we do not consider therapy. I think it is predictive for combination immunotherapy with ipilimumab/nivolumab. And knowing that a patient is PD-L1—positive up front should be a strong motivator to pursue ipilimumab/nivolumab combination therapy. That being said, as stressed before, a PD-L1–negative patient can still benefit and is still, according to the CheckMate-214 data, more likely to benefit from that combination than from sunitinib alone.

Transcript edited for clarity.