PD-L1 Expression in Mismatch Repair-deficient Endometrial Carcinoma and Tumor-associated Immune Cells: Differences Between MLH1 Methylated and Nonmethylated Subgroups.

Abstract

Mismatch repair (MMR)-deficient endometrial carcinomas show increased programmed cell death-ligand 1 (PD-L1) expression compared with MMR-intact endometrial carcinomas, but there are limited data regarding PD-L1 expression between sporadic and inherited carcinomas exhibiting MMR loss. Most of the studies investigating PD-L1 expression in endometrial carcinoma have used tissue microarrays and did not examine all tumor blocks. In this study, we analyzed the expression of PD-L1 in resection specimens of 176 consecutive endometrial carcinomas using all tumor blocks; we compared PD-L1 expression in MMR-deficient endometrial carcinomas, including the MLH1 and PMS2-loss subgroup, and the other MMR-loss subgroups (MSH2 and MSH6, isolated PMS2, and isolated MSH6), with the MMR-intact subgroup. MLH1 methylation was performed in tumors with MLH1 and PMS2 loss. Tumor cell (TC) and tumor-associated immune cell (IC) PD-L1 positivity with a 1% cutoff was observed in 21% (n=37) and 66.5% (n=117) of cases, respectively, and with a 5% cutoff in 5.1% (n=9) and 39.8% (n=70) of cases, respectively. MMR protein deficiency was a statistically significant parameter associated with IC PD-L1 positivity, with 1% and 5% cutoffs on multivariate analysis [odds ratio (OR)=5.236, 95% confidence interval (CI)=2.075-13.211, P=0.001, and OR=3.702, 95% CI=1.759-7.791, P=0.001, respectively]. The multivariate analysis showed that IC PD-L1 positivity, using both 1% and 5% cutoffs, was significantly associated with the MLH1 and PMS2 loss compared with the MMR protein-intact subgroup (MLH1 and PMS2 loss for 1% cutoff: OR=5.104, 95% CI=1.876-13.881, P=0.001, and for 5% cutoff: OR=3.322, 95% CI=1.540-7.166, P=0.002). Squamous differentiation was an independent predictor for TC PD-L1 positivity, with a 5% cutoff (OR=6.102, 95% CI=1.280-10.096, P=0.026). Larger tumor size was an independent predictive factor for IC PD-L1 positivity with a 1% cutoff (OR=6.757, 95% CI=1.569-29.109, P=0.010). Overall, 48 (92.3%) of 52 MLH1 methylated tumors showed IC PD-L1 positivity with 1% cutoff, and 34 (65.4%) of 52 MLH1 methylated tumors showed IC PD-L1 positivity with 5% cutoff. Our results show a higher rate of IC PD-L1 positivity than in previous studies. This is likely due in part to the use of all tumor blocks. MLH1 and PMS2 loss was an independent predictive factor for IC PD-L1 positivity, with both 1% and 5% cutoffs. Using univariate analysis, we observed decreased disease-free survival for IC PD-L1 positivity ≥5%. Our study results should now be tested and proven in larger cohorts, with longer follow-up data.