J Neurol Surg B Skull Base 2014; 75 - A089
DOI: 10.1055/s-0034-1370495

Tuberculum Sellae Meningiomas - Radiological Classification and Its Clinical Significance

Silky Chotai 1, Liu Yi 1, Pan Jun 1, Songtao Qi 1
  • 1Columbus, USA

Background: The aim of this study is to propose a systematic classification for TSM based on their origin and growth pattern.

Methods: A retrospective review of medical records was conducted for 106 patients with TSM who underwent surgical resection at our institution between January 2000 and December 2008. Based on the growth pattern of TSM, we classified the cases into type A (planum sphenoidale), B (tuberuculum sellae) and C1 and C2 (diapharmatic sellae). Type A: tumor located at planum sphenoidale, rarely involves the optic pathway or pituitary stalk (PS); type B located at the tuberculum sellae, mainly involves the optic pathway but rarely involves PS; type C located at the diaphragma sellae, which can involves both the optic pathway and PS. Type C was then divided into C1 and C2, C1- tumor pushes the chiasm in pre-fixed position, C2- optic chiasm is pushed in post-fixed position with expansion of pre-chiasmatic space. The factors that might influence the outcome and recurrence in each tumor type were analyzed.

Results: There were 32 men and 74 women with mean age 48.5 ± 13.1 years (range, 20 to 78 years). The most common presenting symptom was visual impairment followed by headache. The rate of visual impairment as the presenting feature and optic pathway involvement on MRI was significantly higher in type C2 TSM and lowest in type A TSM (97.8% vs13.6%, p < 0.001). The h-p axis impairment and involvement on MRI was significantly higher in type C1 (100% and 90.9%, p < 0.001). The total resection (Simpson I and III) was achieved in 79.2% (84/106) patients. The rate of total resection was significantly higher in type A (86.4%, 19/22, p < 0.03). The mean follow-up time was 70.4 months (median 86 months, range 64.5–76.3 months). The surgical approach employed was the only significant predictor of postoperative visual impairment (p = 0.001, OR: 8.201, CI: 2.453–27.420). The tumor type (p = 0.036, OR: 0.244, CI: 0.065–0.912) and surgical approach (p = 0.032, OR: 0.013, CI: 0.000–0.684) employed were significant predictors of postoperative h-p function impairment. The rate of recurrence was 14% (14/100). The subtotal resection was the only significant predictor of recurrence (p = 0.008, OR: 0.086, CI: 0.014–0.529). The tumor type C1 was most likely to recur (p = 0.001, OR–15.6, CI- 2.9–82.8). The mean PFS time and rate was 102.9 ± 3.2 months and 86%. The PFS was highest in type A TSM 95% and 103.9 ± 3.9 months.

Conclusion: The optimal selection of surgical approach, based on the tumor type, is important to achieve favorable postoperative outcomes. The proposed classification is useful to systematize the most appropriate surgical approach for each tumor type that might enable complete tumor resection. The significantly high rate of complete resection was achieved in type A followed by B and C2. The type C1 had significantly low rate of total resection and high rate of recurrence.