Clinical features as prognostic markers in patients with solid tumors treated with programmed cell death protein 1/programmed cell death-ligand 1 inhibitors: a retrospective cohort study.

Background: The development of immune checkpoint inhibitors (ICIs) has led to improvement of outcomes in the treatment of various cancers. But not all patients benefit from this treatment. Moreover, existing biomarkers for predicting the efficacy of ICIs therapy are imperfect. Clinical factors such as age, body mass index (BMI), and performance status (PS) score collectively influence the efficacy of immunotherapy by modulating tumor immunogenicity and host immune status. This study aims to investigate the prognostic and predictive value of clinical features in patients with solid tumors treated with programmed cell death protein 1 (PD-1)/programmed cell death-ligand 1 (PD-L1) inhibitors. Methods: In this retrospective study, data of a total of 435 patients with advanced tumors who met the inclusion and exclusion criteria and received PD-1/PD-L1 inhibitor treatment in a single academic medical center from November 2016 to October 2021 were analyzed. The clinical features of patients examined included age, BMI, gender, number of treatment lines, smoking history, history of alcohol use, PD-L1, brain metastasis, liver metastasis, PS score, efficacy evaluation, primary cancer type, and histologic type. Patients were grouped and analyzed according to their median age (62 years old) and according to the mean BMI (23.8 kg/m Results: There were 435 eligible patients evaluated in the study. Among the patients receiving PD-1/PD-L1 inhibitor treatment, 61 patients had a partial response (PR), 236 patients had stable disease (SD), and 138 patients had progressive disease (PD), as their best response. The objective response rate (ORR) was 14.0%, and the disease control rate (DCR) was 68.3%. The median progression-free survival (mPFS) was 4.0 months, and the median overall survival (mOS) was 6.8 months. Multivariate analysis confirmed that the number of treatment lines was an independent predictive factor for treatment effect (P<0.05). Multivariate regression analysis indicated that first-line PD-1/PD-L1 inhibitor treatment was associated with superior efficacy as compared to second-line treatment [hazard ratio (HR) =0.375; P<0.05]. The line of treatment was an independent factor for response to PD-1/PD-L1 inhibitor treatment (P=0.048). Liver metastases (HR =1.616; P<0.001) and PD-1/PD-L1 inhibitor as third or later line of treatment (HR =1.305; P=0.044) were both independent risk factors for worse PFS. Additionally, multivariate analysis confirmed the independent predictive role of PS (HR =2.779; P=0.001), liver metastases (HR =1.458; P=0.02), and combination therapy (HR =1.799; P=0.004) for OS after PD-1/PD-L1 inhibitor treatment. Conclusions: Preserved immune function and a favorable tumor microenvironment are conducive to maximizing the efficacy of immunotherapy. In contrast, high tumor burden, immune exhaustion induced by prior therapies, and the immunosuppressive hepatic microenvironment may collectively suppress the antitumor effects of ICI.