Abstract
Classification of the extent of resection into gross and subtotal resection (GTR and STR) after meningioma surgery is derived from the Simpson grading. Although utilized to indicate adjuvant treatment or study inclusion, conflicting definitions of STR in terms of designation of Simpson grade III resections exist. Correlations of Simpson grading and dichotomized scales (Simpson grades I-II vs ≥ III and grade I-III vs ≥ IV) with postoperative recurrence/progression were compared using Cox regression models. Predictive values were further compared by time-dependent receiver operating curve (tdROC) analyses. In 939 patients (28% males, 72% females) harboring WHO grade I (88%) and II/III (12%) meningiomas, Simpson grade I, II, III, IV, and V resections were achieved in 29%, 48%, 11%, 11%, and < .5%, respectively. Recurrence/progression was observed in 112 individuals (12%) and correlated with Simpson grading (p = .003). The risk of recurrence/progression was increased after STR in both dichotomized scales but higher when subsuming Simpson grade ≥ IV than grade ≥ III resections (HR: 2.49, 95%CI 1.50-4.12; p < .001 vs HR: 1.67, 95%CI 1.12-2.50; p = .012). tdROC analyses showed moderate predictive values for the Simpson grading and significantly (p < .05) lower values for both dichotomized scales. AUC values differed less between the Simpson grading and the dichotomization into grade I-III vs ≥ IV than grade I-II vs ≥ III resections. Dichotomization of the extent of resection is associated with a loss of the prognostic value. The value for the prediction of progression/recurrence is higher when dichotomizing into Simpson grade I-III vs ≥ IV than into grade I-II vs ≥ III resections.