Abstract
BACKGROUND: Patients with pancreatic cancer and liver metastases (PCLM) are typically deemed ineligible for curative surgery, with chemotherapy being the standard care. However, surgical resection may benefit select patients. This study investigated whether integrating surgery with chemotherapy improves overall survival (OS) compared to chemotherapy alone in PCLM. METHODS: We conducted a retrospective cohort study of 24,802 patients with stage IV pancreatic cancer from the National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) program (2010-2021). A 1:4 propensity score matching (PSM) was applied to balance 15 baseline covariates. Survival differences between groups were assessed via Kaplan-Meier and multivariate Cox analyses. RESULTS: A total of 24,802 patients were included in this study, constituting a cohort with a biased distribution of age, sex, and disease stage. Among them, only 686 (2.8%) underwent surgery combined with chemotherapy, while 24,116 (97.2%) received chemotherapy alone. Univariate analysis revealed that patients younger than 65 years of age presented a reduced risk of mortality [hazard ratio (HR) =1.3]. Similarly, an earlier disease stage and a lower burden of metastatic disease were associated with a more favorable prognosis. According to the multivariate Cox proportional hazards model, primary tumor location emerged as an independent predictor of survival. Specifically, patients with tumors in the pancreatic body (HR =0.5) or tail (HR =0.4) demonstrated a significantly lower mortality risk than did those with tumors in the pancreatic head. Furthermore, Kaplan-Meier analysis indicated that patients who underwent surgery combined with chemotherapy had a substantially prolonged survival duration relative to those receiving chemotherapy alone (median OS: 18 vs. 6 months; P<0.001). Subgroup analysis on the basis of the site of metastasis revealed differential impacts on survival, whereas osseous metastases had a modest effect on OS, and both hepatic and pulmonary metastases were significantly correlated with a poorer prognosis. PSM successfully matched 645 surgery patients (94.0% matching rate) with 2,580 nonsurgery patients. After PSM, the median OS remained significantly longer in the surgery group (17 vs. 8 months, P<0.001), but the survival difference was attenuated by 25.0% (from 12 to 9 months), with the HR ranging from 0.4 to 0.46 [95% confidence interval (CI): 0.42-0.51]. Only 8 of 15 covariates (53.3%) achieved good balance [standardized mean difference (SMD) <0.1] after matching, indicating residual confounding. CONCLUSIONS: In this retrospective analysis, selected patients receiving surgery plus chemotherapy showed significantly longer OS than those receiving chemotherapy alone. However, the attenuation of survival benefit after PSM and residual imbalances in key prognostic factors suggest that the observed advantage may largely reflect patient selection rather than a true treatment effect. Prospective studies with detailed data on performance status and metastatic burden are warranted to define the role of surgery in this setting.