Contrast-enhancement-defined resection outcomes with sodium fluorescein and 5-aminolevulinic acid in glioblastoma surgery: a multicenter retrospective cohort of 1389 patients.

backgroundSodium fluorescein (SF) and 5-aminolevulinic acid (5-ALA) have distinct biological mechanisms and may highlight different tumor compartments. Comparative multicenter data remain limited, and interpretation depends on the imaging endpoint used.

methodsWe performed a retrospective analysis of prospectively maintained databases from 12 European neurosurgical centers, including 1389 consecutive patients with locally documented IDH-wildtype glioblastoma undergoing fluorescence-guided surgery. Patients received SF alone (n = 650), 5-ALA alone (n = 457), or dual fluorescence (n = 282). The primary endpoint was contrast-enhancement-defined gross total resection (CE-GTR); secondary endpoints were contrast-enhancing extent of resection (CE-EOR), postoperative KPS, and overall survival (OS). Conventional multivariable models were supplemented by center-clustered generalized estimating equations (GEE), center fixed-effects logistic regression, a Cox model with center-clustered robust standard errors, and a center-stratified Cox sensitivity analysis.

resultsCE-GTR was achieved in 74.9% of the SF group, 62.6% of the 5-ALA group, and 68.4% of the dual-fluorescence group (overall p < 0.001). In conventional multivariable analysis, 5-ALA was associated with lower odds of CE-GTR than SF (OR 0.63, 95% CI 0.49-0.81; p < 0.001). The association persisted in the center-clustered GEE model (OR 0.57, 95% CI 0.36-0.90; p = 0.015) but was attenuated after inclusion of center fixed effects (OR 0.65, 95% CI 0.37-1.12; p = 0.121). In the adjusted center-clustered CE-EOR analysis, SF remained associated with a 6.7% points higher mean CE-EOR than 5-ALA (95% CI 3.3-10.0; Bonferroni-adjusted p = 0.003), whereas the other pairwise contrasts were not significant. Median OS was 18.2 months (95% CI 16.8-19.6) with SF, 17.9 months (95% CI 16.6-19.5) with 5-ALA, and 18.5 months (95% CI 17.5-20.2) with dual fluorescence. In the center-clustered Cox model, 5-ALA was associated with a lower hazard than SF (HR 0.84, 95% CI 0.74-0.97; p = 0.014), but this association was attenuated in the center-stratified model (HR 0.88, 95% CI 0.64-1.22; p = 0.451). Dual fluorescence was not associated with survival.

conclusionsSF was associated with higher rates of CE-defined GTR and greater resection of the contrast-enhancing tumor compartment than 5-ALA. However, attenuation in the center fixed-effects analysis indicates that institutional practices and case selection contributed to the observed difference. No consistent center-independent survival advantage was demonstrated. These findings do not establish the intrinsic superiority of SF, particularly because non-enhancing and supramaximal resection could not be assessed. CLINICAL TRIAL NUMBER: Not applicable.