Abstract 1143P
Background
Approximately 40-50% of pts with UM will ultimately develop metastatic disease. There is currently no standard approach for mUM. Recently 2 phase II studies testing N+I have reported overall survivals exceeding 12 m without a great improvement in ORR (Piulats JM at ESMO 2018; Pelster M at ASCO 2019).
Methods
We performed a cohort study involving pts with mUM recruited in the two clinical trials testing N+I. The study used inverse probability of treatment propensity-score matching (PSM) using individualized patient data collected in the PUMMA study. This includes individualized patient data from 29 trials, and 912 patients, with the objective to determine benchmarks for OS.
Results
Raw and accelerated failure time models were fitted to assess the treatment effect on OS. Adjusting confounders were age, sex, LDH, M1 size and localization. In the primary analysis 87 pts were treated with N+I and 516 pts were treated with other systemic treatments. Because it is reported in other diseases that liver M1 perform worst with ICIs, a possible interaction between treatment and M1 localization was tested. In pts with extrahepatic metastasis, expected survival was > 2.6 times higher in N+I treated pts (HR 0.38; 95%CI 0.15-0.94). In pts with liver-only metastases no differences were observed (HR 1.02; 95%CI 0.65-1.60), neither in pts with liver+extra-liver disease (HR 0.81; 95%CI 0.44-1.49). After applying PSM 85(N+I):170(PUMMA) pts were matched by age, sex, ECOG, LDH, and M1 localization. A statistically significant interaction between treatment and M1 localization was tested. In pts with only extrahepatic metastasis, expected survival was > 3.2 times higher in N+I treated pts (HR 0.28; 95%CI 0.11-0.69). In pts with liver-only (HR 1.02; 95%CI 0.59-1.56) or liver+extra-liver (HR 1.05; 95%CI 0.54-2.03) no differences were observed in expected survival between treatments.
Conclusions
Up to 20% of pts with metastatic uveal melanoma present only extrahepatic disease and clearly benefits from N+I treatment. Biological samples from both studies are being analyzed to explain this observation. Liver metastases in mUM still remains a challenge and patients should be included in clinical trials.
Clinical trial identification
NCT02626962 - Spanish Melanoma Group (GEM) NCT01585194 - MD Anderson CRD42014006965 - Meta-analysis registration number - IRCI.
Editorial acknowledgement
Legal entity responsible for the study
Spanish Melanoma Group - MD Anderson - IRCI.
Funding
BMS.
Disclosure
J.M. Piulats Rodriguez: Advisory/Consultancy, Research grant/Funding (self): BMS. S. Patel: Advisory/Consultancy, Research grant/Funding (self): BMS. E. Espinosa, A. Berrocal, L. de la Cruz Merino, A.M. Joshua: Advisory/Consultancy: BMS. All other authors have declared no conflicts of interest.