78P A systematic review and meta-analysis of trials assessing activity of PD-1/PD-L1 immune checkpoint inhibitors (ICIs) for pre-treated advanced malignant mesothelioma (aMM)

Abstract Background aMM still represents a hard-to treat disease, due to its rarity and to the modest activity of standard chemotherapy. Recently, ICIs directed against PD-1/PD-L1 have been tested in clinical trials in chemotherapy pre-treated aMM patients, but their efficacy is still debatable. Methods We searched PubMed and proceedings of major meetings, to perform a systematic review and meta-analysis (updated at September 30th 2019) of clinical trials testing ICIs in this setting, describing activity in terms of Objective Response Rate (ORR) and Disease Control Rate (DCR). To explore the potential predictive role of PD-L1 expression, we also collected the ORR in subgroups of patients selected for PD-L1 expression (based on the highest cut-off used in each study). Results 8 studies were selected (1 phase III, 4 phase II, 2 phase IB, 1 real-world EAP data study), including 405 patients, most with pleural MM; 1 registry study was excluded due to inclusion of treatment-naive patients, 1 due to unclear inclusion criteria. 352 patients (87%) were treated with anti-PD-1 (nivolumab [N] or pembrolizumab [P]), 53 (13%) with anti-PD-L1 (avelumab [A]). Overall, ORR was 19.6% (95% CI, 16.0-23.8%) with no significant difference among drugs (N 20.0%, P 22.6%, A 9.4%; p = 0.11); DCR was 56.5% (95% CI, 51.6-61.3%) with no significant difference among drugs (N 54.0%, P 58.7%, A 58.5%; p = 0.66). When restricting the analysis to patients selected for PD-L1 expression (n evaluable=91, based on cut-offs ranging from 1% to 50% in different trials), ORR was 34.1% (95% CI, 25.2-44.3%), ranging from 18.8% to 71.4% in different trials. In unselected patients, median progression-free survival ranged from 2.5 to 6.1 months, and median overall survival ranged from 6.36 to 17.3 months. Conclusion To our knowledge, this is the first meta-analysis synthesizing the evidences of activity of PD-1/PD-L1 ICIs in pre-treated aMM. ORR and DCR in unselected patients are encouraging compared to historical results with second-line chemotherapy. Selection based on PD-L1 expression could increase the activity of immunotherapy, but trials were heterogeneous for test and cut-off. Legal entity responsible for the study The authors. Funding Has not received any funding. Disclosure M. Tagliamento: Travel / Accommodation / Expenses: Takeda, Bristol-Myers Squibb, Roche. P. Bironzo: Honoraria (self): Bristol-Myers Squibb, BI, AstraZeneca. E. Capelletto: Advisory / Consultancy: BI, AstraZeneca. S. Novello: Speaker Bureau / Expert testimony: AstraZeneca, Abbvie, Celgene, BI, Bristol-Myers Squibb, MSD, Eli Lilly, Pfizer, Roche. M. Di Maio: Honoraria (self): Bristol Myers Squibb, Merck Sharp & Dohme, Roche, AstraZeneca, Janssen, Takeda, Pfizer; Honoraria (institution): Tesaro. All other authors have declared no conflicts of interest.