Critical insights on real-life PD-L1 histopathological workflow and assessment in esophageal, esophagogastric junction, and gastric carcinoma in Italy
Alessandro Gambella, Federica Grillo, Paola Parente, Alessandro Vanoli, Alessandro Caputo, Michele Paudice, Valentina Angerilli, Federica Castri, Alessandro Pastorino, Matteo Fassan, Luca Mastracci, Luca Albarello, Maria Raffaella Ambrosio, M.R. Aprile, Isabella Banchelli, V. Barresi, Paola Billo, S. Bruno, Paola Cassoni, Bruna Cerbelli, Maria DʼArmiento, Giuseppe De Lisi, Enrico Costantino Falco, Dennis Fiorini, Marta Fortunato, Stefano Lazzi, Stefano La Rosa, G Locci, Maria Cristina Macciomei, M. Martini, Laura Melocchi, P. Morbini, Vincenzo Nardini, Iacopo Panarese, Bruna Pozzi, Andrea Remo, Lindsey Savino, Paola Spaggiari, F. Sommavilla
Abstract
BACKGROUND: Immune checkpoint inhibitors (ICIs) targeting programmed cell death protein 1/programmed death-ligand 1 (PD-1/PD-L1) have improved survival in locally advanced and metastatic esophageal, esophagogastric junction, and gastric carcinoma (GEC). Patient selection for ICI-treatment relies on PD-L1 protein expression assessment via immunohistochemistry (IHC). This study aimed to evaluate the real-world assessment of PD-L1 IHC results compared with clinical trial data. PATIENTS AND METHODS: This multicentric, real-world retrospective study analyzed PD-L1 IHC data from 28 Italian pathology centers of GEC cases diagnosed between October 2023 and September 2024. The study documented PD-L1 expression distribution via combined positive (CPS), tumor proportion (TPS), and tumor area positivity (TAP) scores, and investigated the impact of several factors on IHC results. RESULTS: We collected 1936 cases: 1802 adenocarcinomas (ADCA), 131 squamous carcinomas (SCC), and 3 carcinomas of non-specific histotype. Most institutions reported CPS and TPS data, whereas a minority used TAP. Overall, CPS, TPS, and TAP scores were in line with the data in literature and clinical trials for both ADCA and SCC, but inter-institutional heterogeneity was observed as represented by CPS ≥1 ADCA cases (range among institutions: 43.6%-100%). Inter-institutional heterogeneity was significantly associated with several variables, including (i) PD-L1 IHC case workload, with lower workload centers reporting more CPS ≥1 cases on average, and (ii) PD-L1 clone, with the 22C3 clone showing higher CPS scores than the SP263 clone. Tissue block aging was also significantly associated with a lower PD-L1 score, with a critical time window at 24-60 months. CONCLUSIONS: This study confirms the alignment of GEC PD-L1 expression in Italian real-world practice with clinical trials. Inter-institutional variability and the significant influence of preanalytical factors, particularly tissue aging and PD-L1 clone, highlight important challenges in routine PD-L1 testing. Addressing these issues is crucial to enhance the reliability of PD-L1 IHC assessment and ensure optimal patient selection for ICIs in GEC.