Association of PD-L1, PD-L2, and Immune Response Markers in Matched Renal Clear Cell Carcinoma Primary and Metastatic Tissue Specimens

Arnab Basu, MD, MPH; Jennifer Holmes Yearley, DVM, PhD, DAVCP; Lakshmanan Annamalai, DVM, PhD; Christopher Pryzbycin, MD; Brian Rini, MD


Am J Clin Pathol. 2019;151(2):217-225. 

In This Article


Despite exploration in clinical trials, PD-L1 expression has not been a reliable marker for response to checkpoint inhibition therapy in patients with RCC. This is somewhat in contrast to other tumor types such as non–small cell lung cancer. We investigated if a discordance in the expression of PD-L1 immune checkpoint protein and an analogous ligand, PD-L2, exists in renal tumor primaries vs metastatic sites. Our study detected that a large minority of patients have clinically significant discordance in the expression of these immune checkpoint proteins, including PD-L2. This implies that the traditional assessment of PD-L1 positivity of only the primary tumor may not comprehensively address the spectrum of expression throughout tumor sites.

Another possibility for the heterogeneity of response with PD-L1 expression may be a variation in ascribing clinically significant IHC staining for immune checkpoint proteins, which vary between and within tumor types and is an evolving concept. In the current study, 74% patients expressed PD-L1 at some quantifiable level in the primary sites, and 82% expressed any PD-L1 in metastatic deposits. In contrast, only one-fourth of primary and metastatic kidney cancer specimens express a "clinically significant" degree of PD-1, PD-L1, or PD-L2, similar to the reported prevalence of PD-L1 in other studies,[2] implying that most patients have some PD-L1 expression, even though it is not considered clinically relevant.

Another finding of this study is a positive correlation of PD-L2 expression with PD-1 and PD-L1 in metastatic tumors compared with primary tumors. This phenomenon is not well reported and is hypothesis generating. PD-L2 has overlapping roles with PD-L1 and has not been examined prospectively in the clinical trials as a predictive marker. Some other proposed predictors of increased PD-L1 expression in tumor-infiltrating cells have been a higher nuclear grade, higher TNM stage, or sarcomatoid differentiation.[5,17] In our study, an increasing nuclear grade was associated with increased expression of PD-1 and PD-L1 in primary tumors and metastatic sites, consistent with previous data. However, we did not find the TNM stage or sarcomatoid differentiation to be significant predictors of PD-L1 expression. In the absence of readily available techniques to ascertain PD-L1 expression in metastatic sites, a high nuclear grade could provide additional information and merits investigation as a marker of response to immune therapy.

Our analysis has several limitations. Because of the small sample size and multiple comparisons, the strength of observed associations is limited. Also, our analysis relies on original historical pathology review for assessment of tumor characteristics and evaluation by a single pathologist for assessment of semiquantitative scoring. The study was also designed as a correlative study to examine if discordance in immune markers exist, but not its prognostic role, since survival data on these patients were not mature and would be subject to the biases inherent of a retrospective analysis should they have been analyzed. These results provide a scientific rationale to analyze immune checkpoint protein expression in metastatic sites along with primary tumors, as well as the role of PD-L2 in conjunction with PD-L1 in future prospective trials.

Our study characterizes expression of immune markers potentially relevant to response to checkpoint inhibition in metastatic RCC. Additional analysis on a larger number of samples is needed to validate these findings. Discrepancies in immune checkpoint proteins between matched primary and metastatic tumors may have some clinically significant implications and should be examined further.