Long-term outcome of trigeminal neurinomas with modified classification focusing on petrous erosion

Background: The authors present a retrospective analysis of 29 consecutive patients with trigeminal neurinomas. Modified classification based on the tumor distribution over petrous ridge is suggested to select the optimal surgical approach and to predict the outcome. Methods: Preoperative computed tomography (CT) or magnetic resonance (MR) images were analyzed to measure a long diameter of tumor axis (LD), which divided into middle and posterior fossa and a width of the petrous erosion (PW) by the tumor on axial image. The tumors were classified into four groups (M, tumor confined to middle fossa; Mp, tumor mainly in middle fossa; M = P, tumor equally distributed into both middle and posterior fossa; Pm, tumor mainly in cisternal space of posterior fossa with slender, rarely round extension into Meckel's cave), and extent of removal was compared between the groups. Long-term outcomes were based on recurrence-free survival. Results: Total removal was achieved in 16 patients including stage operation (55%). Excluding three tumors, which underwent gamma knife for residual tumor, 8 of 10 subtotally removed tumors recurred at 50 months on average. The LD and PW are significantly different among the tumor types. The M = P type is the largest in both parameter, while the M type is the smallest in LD; Pm type is the smallest in PW, respectively. The tumors with equal or lager posterior fossa component (M = P and Pm type) were more difficult to remove totally by single attempt than those with smaller or nil posterior fossa mass (Mp and M type). Also, the corrected PW (PW over the tumor diameter of minor portion) shows a tendency to influence the extent of removal as the relatively narrow PW hindered the total removal. Conclusions: Aggressive attempt for total removal using skull base approach is recommended, especially for those having dominant posterior fossa component and for those having relatively narrow PW.