Propofol-based total intravenous anesthesia versus inhalational anesthesia in oncologic surgery: a systematic review and meta-analysis of randomized clinical trials on survival and recurrence

Background The perioperative period may influence cancer progression, and propofol-based total intravenous anesthesia (TIVA) has been proposed to preserve immune function relative to inhalational anesthesia. We evaluated its effect on survival and recurrence after oncologic surgery. Methods Following PRISMA 2020, PubMed, Scopus, and Cochrane CENTRAL were searched to August 2026, for randomized controlled trials of adults undergoing oncologic surgery allocated to propofol-based TIVA or inhalational anesthesia. Primary outcomes were all-cause mortality and cancer recurrence, the latter reported as recurrence-free survival; secondary outcomes were perioperative complications and hospital length of stay. Hazard ratios (HR) were the primary effect measure and risk ratios (RR) a sensitivity analysis. Random-effects pooling used restricted maximum likelihood with Hartung-Knapp-Sidik-Jonkman sensitivity analyses. Results Nine trials (5,771 patients) were included. Mortality did not differ between propofol and inhalational anesthesia (HR 1.08, 95% CI 0.95–1.23; k = 4), nor did recurrence-free survival (HR 1.08, 0.95–1.22; k = 5); risk-ratio sensitivity analyses were concordant (RR 0.99, 0.93–1.06; RR 0.95, 0.79–1.15). Propofol was associated with fewer perioperative complications (RR 0.88, 0.77–0.99; k = 3), although this did not persist under Hartung-Knapp-Sidik-Jonkman adjustment (RR 0.88, 0.70–1.10). Hospital stay did not differ (MD -0.28 days, -0.58 to 0.02; k = 2). Heterogeneity was low throughout. Certainty of evidence was moderate for primary and low for secondary outcomes. Conclusion Current randomized evidence does not demonstrate a reduction in mortality or cancer recurrence with propofol-based TIVA, but does not establish equivalence: the recurrence-free survival interval cannot exclude a meaningful increase in risk. Larger trials with predefined oncologic endpoints are needed. Clinical trial number Not applicable.