BACKGROUND: Pleural effusion (PE) is a common metastatic site of NSCLC, associated with poor outcomes. As very few data are available about immune checkpoint inhibitors (ICI) and PE, we aimed to assess the clinical outcome of PE in NSCLC treated with ICI. METHOD: Multicenter international retrospective study of patients with metastatic NSCLC treated with ICI, between 2012 and 2019. Stratification according to the presence of PE at ICI baseline or appearance under ICI treatment (PE group) versus no history of PE (non-PE group). Primary endpoints were overall survival (OS) and early death rate (EDR, OS ≤ 3 months). RESULTS: A total of 538 patients were included: 196 in the PE group and 342 in the non-PE group. In the PE group, median age was 64, 31.6 % were female, 77.6 % had non-squamous histology, PD-L1 was ≥50 % in 38.6 % of cases (95 missing). PE was more likely associated with >2 metastatic sites (70.4 % vs. 50 %) and worse performance status (PS ≥ 2, 30.8 % vs 23.1 %). Globally, the overall median OS was 9.7 months [95 %CI: 8.1-11.8]; 6.3 [95 % CI: 4.0-8.6] in PE vs. 11.4 [95 %CI: 9.7-13.8] in the non-PE respectively, P = 0.002. Overall the EDR was 31.4 %; higher in the PE group (38.3 % vs. 27.5 %; OR 1.63, 95 %CI: 1.13-2.37, P = 0.01). In the PE PD-L1≥50 % group, EDR was 33.3 %. In multivariate analysis, after adjustment on PS, liver/intracranial/bone metastasis, ICI line and dNLR, PE remained an independent prognostic factor for OS [HR: 1.38, 95 %CI: 1.09-1.74, P = 0.007]. In the PE group, PE appeared under ICI for 31 patients (16.4 %). We observed lower EDR in this group compared to patients whom PE was already present (29.0 % vs 40.5 %, P = 0.2). CONCLUSION: PE is associated with worse immunotherapy outcomes in NSCLC treated with ICI, including in patients with ≥50 % PD-L1 tumors. Thus, in these patients, combination strategies should be explored.
BACKGROUND:Pleural effusion (PE) is a common metastatic site of NSCLC, associated with poor outcomes. As very few data are available about immune checkpoint inhibitors (ICI) and PE, we aimed to assess the clinical outcome of PE in NSCLC treated with ICI. METHOD: Multicenter international retrospective study of patients with metastatic NSCLC treated with ICI, between 2012 and 2019. Stratification according to the presence of PE at ICI baseline or appearance under ICI treatment (PE group) versus no history of PE (non-PE group). Primary endpoints were overall survival (OS) and early death rate (EDR, OS ≤ 3 months). RESULTS: A total of 538 patients were included: 196 in the PE group and 342 in the non-PE group. In the PE group, median age was 64, 31.6 % were female, 77.6 % had non-squamous histology, PD-L1 was ≥50 % in 38.6 % of cases (95 missing). PE was more likely associated with >2 metastatic sites (70.4 % vs. 50 %) and worse performance status (PS ≥ 2, 30.8 % vs 23.1 %). Globally, the overall median OS was 9.7 months [95 %CI: 8.1-11.8]; 6.3 [95 % CI: 4.0-8.6] in PE vs. 11.4 [95 %CI: 9.7-13.8] in the non-PE respectively, P = 0.002. Overall the EDR was 31.4 %; higher in the PE group (38.3 % vs. 27.5 %; OR 1.63, 95 %CI: 1.13-2.37, P = 0.01). In the PE PD-L1≥50 % group, EDR was 33.3 %. In multivariate analysis, after adjustment on PS, liver/intracranial/bone metastasis, ICI line and dNLR, PE remained an independent prognostic factor for OS [HR: 1.38, 95 %CI: 1.09-1.74, P = 0.007]. In the PE group, PE appeared under ICI for 31 patients (16.4 %). We observed lower EDR in this group compared to patients whom PE was already present (29.0 % vs 40.5 %, P = 0.2). CONCLUSION: PE is associated with worse immunotherapy outcomes in NSCLC treated with ICI, including in patients with ≥50 % PD-L1 tumors. Thus, in these patients, combination strategies should be explored.