Indocyanine green fluorescence imaging in minimally invasive liver resection for colorectal liver metastases: lesion detection and surgical outcomes—a systematic review and meta-analysis

Background: Indocyanine green (ICG) fluorescence imaging is used during minimally invasive liver resection for colorectal liver metastases (CRLM), but disease-specific evidence is limited and has often mixed patient-level and lesion-level outcomes. Methods: Six databases were searched to May 14, 2026. We prespecified lesion-detection and surgical-outcome modules. Primary quantitative analyses required fully separable CRLM numerators and denominators. Patient-level and lesion-level R0 outcomes were analyzed separately. Random-effects logit models used Paule–Mandel variance and Hartung–Knapp 95% confidence intervals (CIs). Diagnostic studies were assessed with QUADAS-2 and surgical cohorts with the Newcastle–Ottawa Scale. Results: Eighteen eligible reports represented 17 unique cohorts; one earlier report was excluded from quantitative synthesis because its cohort was nested in a later study. Patient-level R0 resection was reported in four studies (146/155 patients) and pooled at 92.5% (95% CI 62.7–98.9%; I2 = 71.3%). Lesion-level R0 resection was reported in four studies (424/455 lesions) and pooled at 93.0% (95% CI 83.2–97.3%; I2 = 25.3%). Major morbidity (Clavien–Dindo grade III or higher) occurred in 20/354 patients across four CRLM-specific cohorts, with a pooled proportion of 6.2% (95% CI 3.1–11.8%; I2 = 0%). Six studies reported fluorescence-positive proportions for CRLM lesions ranging from 59.6 to 95.6%; these were not pooled because reference standards and outcome definitions differed. The only controlled comparison reported higher lesion-level R0 with ICG (odds ratio 5.65, 95% CI 1.02–31.48), but was small and retrospective. Conclusions: ICG fluorescence imaging is feasible for lesion visualization and margin-oriented navigation in selected CRLM patients at experienced centers. Evidence remains insufficient to establish comparative benefit over conventional minimally invasive resection. Adequately powered randomized trials with standardized ICG and intraoperative-ultrasonography protocols and patient-centered oncologic outcomes are essential. Trial Registration: PROSPERO registration: CRD420261394514 © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2026.