Beyond independent predictors: interrelationships among preoperative characteristics associated with surgical complexity in insular gliomas

Objective Preoperative spatial characteristics of insular gliomas are commonly analyzed as separate predictors of surgical outcome, although several may describe overlapping aspects of tumor growth. This study examined the interrelationships among routinely assessed spatial characteristics and evaluated their individual and adjusted associations with achievable extent of resection. Methods This retrospective single-center study included 167 consecutive patients who underwent surgery for histopathologically confirmed insular gliomas between 2017 and 2024. Preoperative variables included tumor size, multizonal involvement, extrainsular extension, and Berger–Sanai topography. Relationships among tumor size, multizonal involvement, and extrainsular extension were assessed using Spearman correlation coefficients. Associations with partial resection (<90%) were evaluated using univariable and multivariable logistic regression. Multicollinearity was assessed using variance inflation factors and tolerance values. Postoperative WHO grade was evaluated separately in a sensitivity analysis. Results Tumor size correlated with multizonal involvement (ρ = 0.640, p < 0.001) and extrainsular extension (ρ = 0.609, p < 0.001), while multizonal involvement correlated with extrainsular extension (ρ = 0.497, p < 0.001). Multicollinearity diagnostics were acceptable (VIF 1.34–3.03). In multivariable analysis, tumor size 3–5 cm (OR 3.98, 95% CI 1.29–12.24), tumor size >5 cm (OR 17.37, 95% CI 3.36–89.87), and multizonal involvement (OR 5.16, 95% CI 2.06–12.95) remained independently associated with partial resection. Associations of extrainsular extension and Berger–Sanai zones were attenuated after adjustment. Conclusions Preoperative spatial characteristics of insular gliomas provide partially overlapping information. Tumor size and multizonal involvement retained independent associations with partial resection, supporting consideration of both tumor extent and spatial distribution when estimating achievable resection.