Early palliative care, intensity of end-of-life care, and their impact on overall survival in cancer patients: A real-world study.

12065 Background: Early palliative care (EPC) and reduced aggressive end-of-life (EOL) interventions improve quality of life in advanced cancer. While trials suggest a potential survival benefit, the real-world impact on overall survival (OS) remains underexplored. This study examines the association between palliative care (PC) timing, EOL care intensity, and OS in a large real-world cohort. Methods: We conducted a population-based study using the French health claims database (SNDS) for patients with metastatic solid tumors diagnosed within 2 years who died in 2022. Patients with hematologic malignancies or PC initiated before metastasis were excluded. Patients were stratified by PC referral timing: no PC, EPC (≤2 months post-diagnosis), and late PC (LPC) ( > 2 months post-diagnosis). EOL care intensity was assessed using Earle criteria (care structure/processes) and MIEOL criteria (physically invasive interventions). The primary endpoint was OS (time from diagnosis to death). Results: Among 159,288 decedents, 85,192 met inclusion criteria (no PC = 27,325; EPC = 29,016; LPC = 28,851). Digestive (n = 29,668) and pulmonary (n = 17,065) cancers predominated. Patients with EPC/no PC had higher poor-prognosis adapted Charlson scores than LPC (23.6% and 23.4% vs. 18%, p < 0.001). EOL aggressiveness was greater in no PC compared to EPC/LPC (Earle ≥1: 51.4% vs. 23.4% vs. 21.6%; MIEOL ≥1: 25.7% vs. 16.9% vs. 17.6%; p < 0.001). Mean OS differed significantly (no PC: 167 ± 190 ; EPC: 74 ± 100 days days; LPC: 344 ± 182 days; p < 0.001). LPC was associated with the longest OS. Patients meeting Earle and MIEOL criteria were significantly associated with shorter survival (p < 0.001). Conclusions: This real-world analysis revealed no survival advantage with early PC initiation, likely reflecting clinical fragility in these patients. Early referral often signifies advanced disease rather than a direct survival benefit of PC itself. LPC was linked to the longest survival, emphasizing the need for graduated PC from metastatic diagnosis. No PC was associated with higher EOL care intensity, which did not improve survival, highlighting possible unreasonable overtreatment.