Page 57 - Demo

FOLFIRINOX reintroduction in pancreatic cancer553approach (ASCO 2022) (14). In a retrospective study, maintenance fluorouracil monotherapy after FOLFIRINOX was associated with an OS of 18.3 months without OS or PFS difference between fluorouracil or FOLFIRI maintenance (15). The OS of patients retreated with FOLFIRINOX without maintenance therapy in our cohort does not seem detrimental compared to the survival rates in the maintenance studies, although these data cannot be directly compared. To specify maintenance treatment to tumor biology, olaparib maintenance can be considered for patients with a germline BRCA1/2 mutation, but unfortunately without a significant OS benefit (16). Patients (n=3) who received olaparib%u00a0in%u00a0our%u00a0study%u00a0were%u00a0not%u00a0included%u00a0in%u00a0the%u00a0FOLFIRINOX%u00a0reintroduction%u00a0cohort. The optimal approach after first line FOLFIRINOX is not known. A second line cohort study reported an OS of 7.5 months after start of second line gemcitabine + nab-paclitaxel after 5FU based combinations (17). An OS of 11.5 and 12.4months from diagnosis was reported for gemcitabine based second line therapy after FOLFIRINOX failure (18). In a previous Dutch NCR cohort the median OS was 11.2 months for patients who were treated with all types of systemic second line therapy (8). A comparison with our data is not possible due to the fact that second line studies focused on patient with%u00a0 FOLFIRINOX%u00a0 resistant%u00a0 disease.Our study has several limitations. First, due to the design of a non-randomized retrospective cohort study, patient and treatment selection bias influences a comparison between different treatments after first line FOLFIRINOX. In the landmark study of Conroy et al. the median number of FOLFIRINOX cycles administered was 10 (range 1 - 47) compared to 5 cycles (range 1 - 32) of first line FOLFIRINOX in our cohort (6). A substantial number of patients in our cohort (n=226, 16,7%) was only treated with 1 cycle of first line FOLFIRINOX, suggesting not all patients in our cohort would have been eligible for the landmark study. Patients who are eligible for FOLFIRINOX reintroduction are a selection of patients with clinical benefit on first line FOLFIRINOX, limited residual toxicity and a sufficient clinical condition. For example, patients in our cohort who could be treated with FOLFIRINOX reintroduction or other subsequent therapies were treated with more cycles of FOLFIRINOX in first line compared to all first-line FOLFIRINOX patients. In addition, patients treated with FOLFIRINOX reintroduction less frequently presented with metastatic disease at diagnosis (57 vs. 73% all first line FOLFIRINOX patients). One can envision that this difference translates into more patients being in a sufficient condition to be retreated with FOLFIRINOX. Second, information on WHO performance status and treatment response was missing in many patients