Clinical and genomic characterization of early-onset pancreatic cancer

4018 Background: Pancreatic cancer is a leading cause of cancer-related death worldwide. Early-onset tumours (≤ 50 years) have risen alarmingly in recent years. However, the clinical and genomic particularities of early-onset pancreatic cancer (EOPC) remain poorly defined. We aimed to characterize EOPC and study its implications for treatment and prognosis. Methods: We performed a retrospective analysis of EOPC patients and a control group of patients ≥ 70 years (defined as average-onset pancreatic cancer, AOPC) followed at a tertiary cancer centre from 2010 to 2022. We collected baseline patient characteristics, tumor molecular profiling, germline genetic alterations, survival and treatment outcomes. We used a targeted gene panel to identify somatic genomic events and classified them according to the ESMO scale for clinical actionability of molecular targets (ESCAT). Key molecular findings were validated in an external cohort. We used a propensity score weighting method and multivariate Cox regression analysis to adjust for covariates. Results: We reviewed 824 patients, 336 of whom met all inclusion criteria (EOPC N = 139, AOPC N = 197). EOPC was associated with smoking status (current, 15.9 vs 7.9%, p = 0.03), lower prevalence of diabetes (3.7 vs 39.6%, p < 0.01), better performance status (ECOG 0, 42.9 vs 19.3%, p < 0.01), higher CA19.9 levels (median 574 vs 207.6UI/L, p = 0.07) and higher albumin levels (median 4.2 vs 3.9g/L, p < 0.01). EOPC showed a non-significant higher prevalence of germline alterations (22.4 vs 14.5%, p = 0.18). After adjustment for baseline covariates, we observed no differences in survival (HR 0.94, 95% CI 0.63-1.4). We found a lower prevalence of KRAS MUT tumors in EOPC when compared with AOPC (83.1 vs 91.1%, p = 0.12) and validated this finding in an independent cohort of 803 patients (82.9 vs 94.5%, p < 0.01). Notably, EOPC were enriched in potentially actionable alterations when compared with AOPC, both in our cohort (19.1 vs 14.4%) and the validation cohort (14.4 vs 7.9%, p < 0.01). Moreover, 294 patients in our cohort (EOPC N = 130; AOPC N = 164) were diagnosed with or eventually presented metastasis. EOPC more frequently received first-line 5FU-based chemotherapy (44.1 vs 14.7%, p < 0.01) and received a greater number of treatment lines (median 2 vs 1, p < 0.01). EOPC had a longer progression-free survival on first-line chemotherapy when compared with AOPC (HR 0.61, 95% CI 0.43-0.87, p < 0.01) although there were no differences in response rate. Six EOPC patients received matched targeted therapies and 5 patients with AOPC. After adjusting for the number of treatment lines received, EOPC treated with targeted therapies exhibited longer OS compared with EOPC who did not (HR 0.34 95% CI 0.12-0.93, p = 0.04), whereas this trend was not observed in AOPC (HR 0.82 95% CI 0.32-2.11, p = 0.68). Conclusions: EOPC patients harbor unique clinical and molecular features and may particularly benefit from precision-based oncology approaches.