Computed Tomography Features of Lung Adenocarcinomas With Programmed Death Ligand 1 Expression

Abstract

IntroductionThe development of immune checkpoint inhibitors against programmed death 1 has paved the way for a new era of treatment of lung cancer. Programmed death-ligand 1 (PD-L1) is expected to predict the response of immune checkpoint inhibitors in lung cancer. Predicting PD-L1 expression using a noninvasive method before immunotherapy would, therefore, help identify patients for whom immunotherapy can be successful. Patients and MethodsA total of 394 patients with resected lung adenocarcinoma who had undergone preoperative thin-section computed tomography (CT) were analyzed for PD-L1 expression by immunohistochemistry and evaluated to determine the association between PD-L1 expression and CT characteristics, including convergence, surrounding ground glass opacity (GGO), air bronchogram, notching, pleural indentation, spiculation, and cavitation. ResultsOf the 394 patients, 78 (19.8%) were positive and 316 (80.2%) were negative for PD-L1 expression. Univariate analysis demonstrated that PD-L1+ adenocarcinoma was significantly associated with the presence of convergence (P < .01), notching (P < .01), spiculation (P < .01), and cavitation (P < .01) and the absence of surrounding GGO (P < .01) compared with PD-L1− cases. On multivariate analysis, the presence of convergence (P < .01) and cavitation (P < .01) and the absence of surrounding GGO (P = .02) and air bronchogram (P = .03) were significantly associated with PD-L1 expression. ConclusionPD-L1+ adenocarcinoma cases showed convergence and cavitation more frequently than did PD-L1− cases. In contrast, surrounding GGO and air bronchogram were observed less frequently in PD-L1+ cases than in PD-L1− cases. These results will prove helpful in identifying PD-L1–expressing adenocarcinoma by CT before immunotherapy.