Immune checkpoint inhibitors (CKIs) are therapeutic weapons in several advanced malignancies. Performance status is a validated prognostic variable in cancer patients; it possibly affects the efficiency of the immune system. We performed a systematic review and meta-analysis to investigate the predictive role of PS toward treatment with CKIs in cancer patients. Following PRISMA guidelines, an electronic search from PubMed, The Cochrane Library and Embase was performed, from the inception of each database to May 31, 2018. Inclusion criteria were (1) randomized trials comparing CKI with standard therapy for the treatment of patients with solid tumors; (2) information on overall survival (OS) according to PS; (3) full text available; and (4) reported in English language. Data were pooled using HRs for OS according to random effect model. The effect of experimental versus control arms was evaluated in PS = 0 and 1-2 subgroups, and the heterogeneity between the two estimates was assessed using an interaction test. The OS differences between PS = 0 and PS = 1-2 strata were evaluated in all studies and according to predefined subgroups. Eighteen studies were eligible, with 11,354 patients [PS = 0 group 5217 patients (46%); PS = 1-2 group 6137 patients (54%)]. The pooled HR for OS was 0.78 (95% CI 0.69-0.89) in PS = 0 patients. In PS = 1-2 patients, the pooled OS HR was 0.78 (95% CI 0.71-0.86). The OS difference between PS = 0 and PS = 1-2 patients treated with CKI was not significant (P = 0.99). CKI improves survival irrespective of patients' PS. PS should not guide treatment choice for anticancer immunotherapy.
Keywords: Cancer patients; ECOG PS; Immune checkpoint inhibitors; Meta-analysis; Performance status.