One- and two-year overall survival following stereotactic body radiotherapy versus conventional radiotherapy for pancreatic cancer: a systematic review and meta-analysis

Background: Pancreatic cancer remains one of the most aggressive malignancies with a dismal prognosis. Both stereotactic body radiotherapy (SBRT) and conventional radiotherapy (CRT) are used in pancreatic cancer treatment, however, there is a lack of robust evidence comparing their efficacy, particularly in terms of 1-year and 2-year overall survival (OS). Methods: We systematically searched PubMed, Embase, Web of Science, and the Cochrane Library (up to March 2025) for head-to-head comparative studies evaluating SBRT versus CRT in pancreatic cancer under the PRISMA 2020 guideline. Random-effects meta-analyses were used to pool 1- and 2-year OS rates, with heterogeneity test assessed. Prespecified subgroup and sensitivity analyses examined the influence of study design, treatment intent, chemotherapy use, and outlier impact. Publication bias was evaluated using funnel plots and Egger’s test. Study quality was assessed using the Newcastle–Ottawa Scale, and the overall review process was appraised using the AMSTAR 2 tool. Results: Twelve retrospective studies comprising 6,566 patients were included. SBRT significantly improved 1-year OS compared to CRT (logit OR = 0.43, 95% CI: 0.26–0.60; I 2 = 22.5%), corresponding to an OR of 1.54. Subgroup analyses stratified by disease stage, chemotherapy regimen, and study design consistently favored SBRT. The prediction interval (OR: 1.14–13.50) supported the robustness of the pooled effect. Cumulative meta-analysis demonstrated a progressively increasing survival advantage of SBRT over CRT from 2015 to 2023. For 2-year OS, SBRT also showed a significant benefit over CRT (logit OR = 0.33, 95% CI: 0.02–0.64; p = 0.039), equivalent to an OR of 1.39. No evidence of publication bias was found (Egger’s test p = 0.811). Conclusion: SBRT is associated with superior 1- and 2-year OS versus CRT in pancreatic cancer. These findings warrant further prospective validation to optimize integration in contemporary treatment algorithms.