Third-Line Treatment Timing After Second-Line Discontinuation in Advanced Pancreatic Cancer: A 30-Day Landmark Analysis.

PURPOSE: Evidence for third-line systemic therapy in advanced pancreatic cancer is limited, and observational comparisons risk time-related and selection biases. We evaluated third-line initiation after second-line discontinuation using a 30-day landmark strategy, focusing on treatment timing and patient selection. METHODS: This single-center retrospective cohort included patients with advanced pancreatic cancer discontinuing second-line systemic therapy between October 2014 and October 2025. The primary analysis included ECOG performance status (ECOG PS) 0-1 patients alive and under observation at day 30. Exposure was initiation within 30 days versus no initiation by day 30, with overall survival (OS) measured from the landmark. Complementary analyses included a 45-day landmark, overlap weighting, time-dependent Cox models, and 60-day mortality after initiation. RESULTS: Among 130 patients, 53 received third-line therapy. The primary 30-day landmark cohort included 48 patients. Early initiation showed a directionally favorable but imprecise association with OS (aHR 0.40, 95% CI 0.15-1.04). Findings were similar at the 45-day landmark (aHR 0.33, 95% CI 0.11-1.02) and attenuated with overlap weighting (aHR 0.68, 95% CI 0.38-1.20). Death within 60 days after initiation occurred in 18.9% of recipients and was associated with ECOG PS ≥ 2. CONCLUSION: Among selected patients with preserved ECOG PS alive and observed at day 30, early third-line initiation showed directionally favorable but imprecise OS. This contrast should be interpreted as early versus deferred-or-no initiation within a 30-day decision window, supporting time-bound patient selection and supportive care-integrated decision-making, not definitive causal inference.