Magnetic Resonance Elastography in Vestibular Schwannoma

Introduction: Magnetic resonance elastography(MRE) is a MRI-based imaging sequence that uses mechanical waves to quantify the stiffness of tissue in kilopascal(kPa). More recently, a form of MRE has been developed called slip-interface imaging(SII) that uses octahedral shear-strain(OSS)to evaluate the surgical plane between extra-axial brain tumors and adjacent neural structures. The feasibility of MRE for evaluating the extent of resection and facial nerve outcomes in vestibular schwannoma (VS) has never been evaluated.

Methods: After institutional review board approval and written consent, 11 patients with a VS ≥2.0 cm in maximum diameter were prospectively evaluated by MRE prior to surgery from February 2014 to July 2015. Mean patient age was 54.3 ± 14.5(30–75) years and 7(64%) were female. MRE data were collected with a spin-echo EPI-MRE pulse sequence on a 3T MR scanner. Shear waves at 60 Hz were introduced with a soft pillow-like driver placed under the head. Retrospectively collected data included patient characteristics; surgeon impression of tumor consistency, adherence, and vascularity; MRE stiffness measured in kilopascal(kPa); SII measured by OSS; extent of resection; House-Brackmann scores preoperatively, postoperatively, and at latest follow-up; and latest-follow-up in months. Tumor consistency and adhesion were graded on a 3-point scale. Cohen κ coefficients were considered 0.20, poor agreement; 0.21–0.40, fair agreement; 0.41–0.60, moderate agreement; and >0.60, good agreement. P-values <0.05 were considered significant.

Results: Mean tumor size was 3.2 ± 0.67(2–4)cm. Six(55%) patients underwent a retrosigmoid craniotomy while 5(45%) underwent translabyrinthine craniectomy. Mean follow-up was 13.4 ± 8.9(3–28) months.

With regard to tumor stiffness, surgeons categorized 5 as soft, 4 as intermediate, and 2 as firm. Mean MRE stiffness increased for each group, 3.09 ± 0.58(2.39–3.95), 3.17 ± 0.48(2.58–3.74), and 3.5 ± 0.39 (3.26, 3.80) kPa respectively, but the differences were not statistically significant (p = 0.633). Two tumors were soft at surgery, but had MRE values >3.30 kPa. Both were noted to be vascular at surgery, a feature which has been noted to cause poor correlation in liver and meningioma MRE.

With regard to tumor adherence, surgeons categorized 5 as complete, 3 as partial and 3 as no separation. SII categorized 7 tumors as complete, 3 as partial and 1 as no separation (κ=0.70 95%CI 0.35–1.00). Both tumors with discrepancy between SII and surgical findings were read as non-adherent on SII, but were completely adherent at surgery. Both had cyst in the posterior portion of the tumor which may lead to a false positive for non-adherence from intratumoral movement.

Surgical consistency, adhesion, SII, MRE, and tumor size were evaluated for extent of resection and for House-Brackman scores ≥ 3 in the 9 tumors that had good correlation with either stiffness or adhesion. Both mean MRE stiffness [3.34 ± 0.37(2.85 – 3.80)v2.64 ± 0.28(2.39–2.85), p = 0.025] and tumor size[3.53 ± 0.47(2.8–4.0)v2.5 ± 0.50(2.0–3.0)p = 0.011] were higher in patients that had near-total/subtotal resection compared with those that had gross-total resection (p = 0.024).

Conclusion: MRE is feasible in VS and has the potential to provide information regarding tumor consistency and adhesion. Further study is needed to determine clinical utility and improve accuracy.