P2.04-017 Prognostic Relevance of PD-1/PD-L1 Pathway in Thymic Malignancies with Combined Immunohystochemical and Biomolecular Approach: Topic: Thymic Malignancies Clinical and Translational

Recent studies on cancer active immunotherapy showed the effectiveness of blockade of PD-1/PD-L1 pathway in several tumors. The present study evaluated PD-1 and PD-L1 protein expression in thymic epithelial tumors (TETs) in order to improve our knowledge about immune system influence in TETs pathogenesis and natural history and to encourage new therapeutic approach with immunomodulating agents. However, considering the potential bias related to immunohistochemical technique (IHC), we aimed to confirm the PD-1 and PD-L1 immunohistochemical analysis results with a molecular assay. PD1 and PDL-1 immunohistochemical staining, using BMS PD-L1 (clone 28-8) assay from Dako, and mRNA expression by RT-PCR were evaluated on 68 tissue microarray (TMA) samples (63 thymomas and 5 thymic carcinomas) of patients who underwent surgery for TETs in our institution between 1993 and 2013. M/F ratio was 33/35, median age was 60.5 years (range 21-81). 20 patients presented with Myasthenia Gravis, while 4 experienced other syndromes. All the patients underwent radical surgery (thymectomy) and in the half cases patients underwent a previous biopsy. Out of the 63 thymomas, 27% were AB, 6% B2, 18% A, 6% B1 and 19% B3, 5% were mixed B1-B2 and 19% mixed B2-B3, according to WHO classification. Approximately 56% of the patients had large tumors (>5 cm). According to Masaoka-Koga staging, 41% patients presented pathologic stage IIA, whereas 11%, 29%, 10%, 5%, 4% were found to have stages I, IIB, III, IVA and IVB, respectively. PD-L1 mRNA tumor expression was significantly associated with worse prognosis in patients with a tumor diameter >5 cm (HR: 5.40 CI 95% 1.9-78.1, p=0,0083). In particular, median OS was 82 months in patients with simultaneous PD-L1 immunostaining (>1%) and PD-L1 mRNA expression compared to not reached median OS of patients with simultaneous PD-L1 negative IHC and not expressed PD-L1 mRNA (p=0.0178). Furthermore in this patients’ subgroup the age > 60 years was associated with worse prognosis (102 vs. 197 months, p=0,0395). Finally, the PDL-1 IHC positivity resulted significantly associated with paraneoplastic myasthenia gravis. two previous studies investigated immunohistochemical PD-L1 expression in TETs tissue microarray with controversial results. Different techniques of expression analysis could be applied in order to provide better detection of PD-1/PDL-1. Our data support the potential role of PD-1/PD-L1 pathway in TETs progression, also suggesting that it may be useful in order to define high-risk patients after curative resection and to plan future trials with anti-PD-1 agents. With the support of BMS.