Individualized Management Strategy of Petroclival Meningiomas Based on a Radiographic Classification

Background Surgical management of petroclival meningiomas (PCMs) was challenging accompanied with relatively high morbidity. Treatment selection was quite surgeon dependent, and no consensus-based strategy was proposed.

Objective The study aimed to propose a radiographic classification of PCMs to facilitate treatment selection and predict neurological outcomes.

Methods Clinical and radiographic data of 337 cases of surgically treated PCMs between May 2011 and August 2015 were retrospectively reviewed, and follow-up was accomplished in 301 cases. Skull base compartments surrounding petrous apex included (1) petrous apex itself; (2) Meckel's cave, cavernous sinus, and/or middle cranial fossa; (3) sellar area and/or sphenoidal sinus; (4) contralateral clivus and/or petrous apex; (5) invasion to CPA region and lateral to internal auditory meatus; and (6) lower third clivus. All PCMs were classified into seven subtypes based on the regions potentially involved by PCMs: type I (region 1); type II (region 1 plus 2); type III (region 1 plus 3); type IV (region 1 plus 4); type V (region 1 plus 5); type VI (region 1 plus 6); and type VII (region 1 plus 2 regions or more). Surgical approach was selected based on the classification, and the association with outcome was evaluated.

Results This consecutive cohort included 221 females (73.4%) with a mean age of 50.1 years. Mean preoperative and postoperative KPS at discharge was 79.0 and 65.7, respectively, and recent KPS was 76.5. Retrosigmoid approach (n = 36), anterior transpetrosal (n = 179), presigmoid retrolabyrinthine approach (n = 59), far lateral approach (n = 13), and others (n = 14). Gross total resection was achieved in 188 cases (62.5%) that was mildly increased than our early series. Surgical mortality and morbidity was 1.5% (n = 5) and 46.3% (n = 156). Mean follow-up duration was 36.7 months, and the recurrent rate was 3.3% (10/301). Surgical morbidity was different between various subtypes, but that was significantly higher in type VII group (p = 0.031) as well as the lesion size was largest in type VII group (p = 0.026). Other adverse factors for neurological outcome included brainstem edema, subarachnoid space, extent of surgical resection, preoperative KPS score, and intraoperative findings (adhesion to neurovascular structure and/or vessel encasement). The association between subtypes and neurological outcome was significant (p = 0.009) but compromised after adjustment for surgical approach (p = 0.064).

Conclusion The radiological classification facilitated surgical approach decision, and total resection could be improved via suitable approach. Follow-up duration should be prolonged to validate the reasonability of the classification.