Identification of markers associated with toxicity of FOLFIRINOX in patients over 70 years with digestive cancers.

e16305 Background: mFOLFIRINOX (mFFX) is a chemotherapy regimen used for gastro-intestinal (GI) cancers, in colorectal (CCR) and pancreatic cancers (PDAC) but limited by grade ≥ 3 toxicities in 50% of cases. Little data is available regarding its use in elderly patients, rarely included in phase 3 trials, although most of patients in real-life setting have median age at diagnosis > 70 y-o for both CCR and PDAC. Methods: Using the ConSore (Continuum Soins Research) software based on of medical records data mining, we identified consecutive patients > 70 y-o, treated with mFFX in Paoli-Calmettes Institute, France, from 09/2011 to 07/2022. We recorded geriatric data, adverse events (AE), and dose reduction during mFFX. We analyzed factors associated mFFX’s toxicity in elderly population with GI cancers. Results: We included 161 patients. Median age was 74 y-o range 71-83, including 43% >75 y-o. 82% had ECOG-PS 0-1 at diagnosis. 91% had PDAC and 9% had CCR, treated for a metastatic disease in 48%. Median number of mFFX cycles was 5 range 1-15. G8-ONCODAGE was calculated ≤14/17 for 77%. Multidimensional geriatric assessment (MGA) was performed in 29%, but 66% without MGA had a G8 ≤14. Sarcopenia (assessed by muscle mass in L3 CT-scan) was reported in 86% and malnutrition was moderate (57%) or severe (37%). Primary dose reduction was applied to 68%, dose reduction during mFFX in 69% and 32% discontinued 1 drug during mFFX’s courses. 53% experienced a grade ≥ 3 digestive AE or unplanned hospitalization, including 3 toxic deaths. Infections (23%), diarrhea (17% G3), peripheral neuropathy (16% G3), anorexia (15% G3), thrombopenia (12% G3) and vomiting (9% G3) were the most frequent AEs.Febrile neutropenia was reported in only 4%, due to systematic use of G-CSF. Median overall survival (mOS) was 15 months (51 months for CCR, 15 months for PDAC), and was 11 months for metastatic patients and 22 months for locally advanced disease. In multivariate analysis after backward selection, factors associated with grade > II digestive toxicity and/or hospitalization were MGA for patients with G8 ≤14 (OR 0.38 0.15; 0.92 ; p = 0.036), polypharmacy (OR 2.70 1.07; 7.22 ; p = 0.039), psychiatric comorbidity (OR 8.38 1.57; 75.27 , p = 0.026) and past history of another cancer (OR 0.27 0.07; 0.92, p = 0.04). A trend was found for sarcopenia (OR 2.70 0.79; 10.13, p = 0.12) and cardiovascular comorbidity (OR 0.43 0.17; 1.01 p = 0.058). Conclusions: Our data reported that mFFX is toxic and associated with similar survival in a highly selected > 70y-o patients with GI cancers, compared to their youngest counterpart included in pivotal trial. However, alternative regimens/strategies should be developed to enhance efficacy and reduce AEs. Particular attention to vulnerabilities detected using G8 and managed through MGA is of importance to reduce toxicity.