Diffuse high intensity PD-L1 staining in thymic epithelial tumors

Abstract

Introduction: Blockade of the immune checkpoint programmed death receptor ligand-1 (PD-L1)/PD-1 pathway has well-established clinical activity across many tumor types. PD-L1 protein expression by immunohistochemistry is emerging as a predictive biomarker of response to these therapies. Here, we examine PD-L1 expression in a thymic epithelial tumor (TET) tissue microarray (TMA).

Methods: The TMA contained 69 TETs and 17 thymic controls, with each case represented by triplicate cores. The TMA was stained with rabbit monoclonal antibody (clone 15; Sino Biological, Beijing, China) to human PD-L1. PD-L1 staining was scored based on intensity as follows: 0 = none, 1 = equivocal/uninterpretable, 2 = weak, and 3 = intermediate-strong. Those cases with all cores scoring three in the epithelial component were categorized as PD-L1 high and the remaining as PD-L1 low.

Results: PD-L1 high scores were more frequent in TETs than in controls (68.1% versus 17.6%; p = 0.0036). PD-L1 scores and histology were significantly correlated, with higher intensity staining in World Health Organization (WHO). B2/B3/C TETs. Only 14.8% of TETs had PD-L1 staining of associated lymphocytes. In an adjusted analysis (age/sex), PD-L1 high TETs had a significantly worse overall survival (hazard ratio: 5.40, 95% confidence interval: 1.13-25.89; p = 0.035) and a trend for worse event-free survival (hazard ratio: 2.94, 95% confidence interval: 0.94-9.24; p = 0.064).

Conclusions: PD-L1 expression was present in all cases of TETs within the epithelial component but only in a minority in the lymphocytic component. TETs stained more intensely for PD-L1 than in controls, and PD-L1 high TETs were associated with more aggressive histology and worse prognosis. This study lends rationale to a clinical trial with anti-PD-1/PD-L1 therapy in this rare tumor type.

Conflict of interest statement

Conflicts of Interests and Sources of Funding: No COI related to this submitted work for the following authors: Sukhmani K. Padda, Jonathan W. Riess, Erich J. Schwartz, Lu Tian, Holbrook E. Kohrt, Joel W. Neal, Robert B. West, Heather A. Wakelee

Heather A. Wakelee: No disclosures in relation to this submitted work. Disclosures outside the submitted work include consultancy to Peregrine [money paid to author] and grants from Celgene, Genentech/Roche, Lilly, Pfizer, MedImmune, AstraZeneca, Novartis, BMS, Clovis, Exelixis, Regeneron, Xcovery [money paid to institution].

Joel W. Neal: No disclosures in relation to this submitted work. Disclosures outside of the submitted work include grants paid to author and/or institution [Merck, ArQule, Genentech/Roche].

Figures

Figure 1

Magnification ×300. Panel A is a type B2 thymoma, Panel B is normal thymus, and Panel C is a micronodular thymoma with lymphoid stroma. Stains include H&E (left column), CK5/6 (middle column) highlighting thymic epithelial cells, and PD-L1 (right column). In panel A and B, only the epithelial cells stain for PD-L1. Panel C shows PD-L1 staining of both thymic epithelial cells and lymphocytes. Panels A, B, and C show strong (IHC 3), weak (IHC 2), and intermediate (IHC 3) PD-L1 staining, respectively.

Figure 2

PD-L1 Intensity and Outcomes (a) Overall Survival (OS) with stratification by age high TETs demonstrate no difference in EFS and OS in an unadjusted analysis. When adjusted for age and gender, PD-L1 high TETs had a statistically significant worse OS, with age being the most important adjustment. PD-L1 high TETs also had a trend for worse EFS when adjusted for age and gender, with adjustment favoring females. OS events: PD-L1high, 10 events of 47 and PD-L1low, 5 events of 22. EFS Events: PD-L1high, 18 events of 44 and PD-L1low, 5 events of 20.