Novel models for prediction of benefit and toxicity with FOLFIRINOX treatment of pancreatic cancer using clinically available parameters

BACKGROUND: Despite modern chemotherapy regimens, survival of pancreatic cancer patients remains dismal. Toxicity is a major concern and it is a challenge to upfront identify patients with the highest benefit from aggressive polychemotherapy. We aimed to evaluate ORR and side effects of the FOLFIRINOX regimen, highlighting dose modification and to explore possible prognostic response factors as a clinical tool. METHODS: This retrospective study includes 123 patients with metastatic PC that were treated with FOLFIRINOX between the years 2007 to 2016 in a single academic institution. Survival rates were analysed using the Kaplan-Meier method. Prognostic models including laboratory and clinical parameters were calculated using Cox proportional models in univariate and multivariate analyses. RESULTS: Median age at diagnosis was 64 years (47–78 years), 71 (57, 7%) were male and the majority had an ECOG performance status of 0 or 1 (63 patients; 83.7%). After a median follow up of 17.8 months, median progression free survival (PFS) and overall survival (OS) were 5.7 (4.55–6.84; 95%CI) and 11.8 months (9.35–14.24; 95%CI) respectively. Overall response rate with FOLFIRINOX was 34.9% and stable disease rate was 21.9%. Regarding Grade 3/4 side effects, 62 events, were reported in 37 patients. Looking at risk factors e.g. patient characteristics, tumor marker, inflammatory markers and body composition multivariate analyses proved CEA >4 elevation and BMI > 25 at the time point before palliative chemotherapy to be independent negative prognostic factors for OS. Grouping patients with no risk factor, one or two of these risk factors we analyzed a median OS of 17.4 moths, 9.6 months and 6.7 months (p<0.001) respectively. In addition we identified thrombocytosis and low BMI as predictors of early toxicity. CONCLUSION: This study identifies two easily available factors influencing overall survival with FOLFIRINOX therapy. By combining these two factors to create a score for OS, we propose a prognostic tool for physicians to identify patients, who are unlikely to benefit more from FOLFIRINOX or likely to experience toxicity.