Expression And Clinical Significance Of PD-L1 And PD-1 In Gastric Carcinoma

Objective: To investigate the correlation between the expression of costimulatory molecule PD-L1 and its receptor PD-1 in gastric carcinoma as well as clinicopathologic features and prognosis of gastric cancer patients. Methods: 218 patients who received radical gastrectomy and pathologically diagnosed with gastric adenocarcinoma at the First Affiliated Hospital of Anhui Medical University between 2007 and 2009 were retrospectively enrolled in this study, combination with 113 matched lymph nodes and 20 cases of adjacent normal tissues controls were evaluated for PD-L1 and PD-1 protein expression by immunohistochemistry PV-6000 two-step method. PD-L1 were quantified using a visual grading system based on the extent of staining. For PD-1, the expression was determined by counting immunolabeling cells at high magnification. SPSS17.0 statistical software was employed to analyze the expression of PD-L1 and PD-1 compared with clinicopathologic features and overall survival of patients. Results: 1. 66.97% gastric cancer tissues showed positive PD-L1 expression, mainly located in cytoplasm and cell membrane of gastric cancer cells and interstitial lymphocyte, where the positive rate of cancer cells was 66.05%, and the positive rate of of interstitial lymphocytes was 33.94%. 40.83% cases showed the positive PD-1 expression, mainly located in the cytoplasm and cell membrane of the tumor infiltrating lymphocytes(TIL). The adjacent tissue(20 cases) had no PD-L1 and PD-1 expression. PD-L1 and PD-1 protein is also expressed on lymph node metastasis, which were highly correlated between the primary tumors and metastatic nodes, with concordance rates of 76.1% and 86.7% for PD-L1 and PD-1 respectively. 2. PD-L1 expression among tumor stroma cells was significantly correlated to PD-1 expression(P=0.006), whereas there was no significantly correlation between PD-L1 expression in cancer cells and PD-1 expression in TIL(P>0.05). 3. PD-L1 expression in gastric carcinoma ware closely related to p TNM stage(P<0.01), infiltration levels(P<0.01), lymph node metastasis(P<0.01) and vascular invasion(P<0.05). PD-L1 expression in tumor stroma cells were correlated only with lymph node metastasis(P<0.05). PD-1 expression in gastric carcinoma and clinicopathological characteristics were not significantly correlated. 4. Univariate COX regression analysis showed later p TNM staging, larger tumor size, lymph node metastasis and PD-L1 overexpression of gastric cancer were poor prognostic factors. Meanwhile, multivariate COX regression analysis showed that p TNM stage, tumor size, lymph node metastasis and PD-L1 positive expression in cancer cell was an independent prognostic factor. Kaplan-Meier survival curve analysis showed a significant correlation between postoperative survival of patients and PD-L1 expression both in gastric cancer cells and stroma cells. The patients with positive PD-L1 expression have shorter postopertive survival than the negative PD-L1 expression ones, and log-rank test showed that the difference was statistically significant(P<0.01). PD-1 positive expression in gastric carcinoma was not significantly correlated with postoperative survival. Conclusion: 1.PD-L1 and PD-1 were overexpressed in gastric cancer tissues compared to adjacent normal tissues. PD-L1 expression among tumor stroma cells was significantly correlated to PD-1 expression, whereas PD-L1 expression in cancer cells had no significantly correlation with PD-1 expression. 2.PD-L1 and PD-1 protein is expressed on lymph node metastasis, which were highly correlated between the primary tumors and metastatic nodes. 3.PD-L1 expression in gastric carcinoma ware closely related to several clinicopathologic features of patients. PD-L1 expression in tumor stroma cells were correlated only with lymph node metastasis. PD-1 expression in gastric carcinoma and clinicopathological features were not significantly correlated. 4.PD-L1 are negative prognostic factors for gastric cancer patients, suggesting that it might be used as a new biomarker to predict the disease progression and prognosis.