Higher Resolution Magnetic Resonance Elastography for the Evaluation of Intratumoral Heterogeneity in Meningiomas

Introduction: Magnetic resonance elastography (MRE) is an MRI-based modality that analyzes the movement of shear waves thorough tissue to determine stiffness, offering a way to “palpate” with imaging. A prior study on meningiomas showed prospective MRE measurements correlated well with surgeons' intraoperative observations regarding tumor consistency. However, this study reported the overall tumor consistency based on a lower resolution MRE technique and did not detect intratumoral differences. The purpose of our study was to evaluate a higher-resolution MRE technique to investigate whether intratumoral variations in softness or hardness could be detected with MRE.

Methods: A total of 15 meningiomas in 14 patients underwent MRE before surgery. MRE data were collected with a spin-echo EPI pulse sequence on a 3T MR imager. Shear waves at 60 Hz were introduced with a soft pillow-like driver placed under the head. The curl of the wave images was calculated and stiffness was determined with a direction-inversion algorithm. If the tumor had intratumoral regions with distinctly different stiffness values, it was considered heterogeneous. Intratumoral portions were considered hard if there was a significant area ≥ 6 kPa. Intraoperative observations regarding softness or hardness were graded on a 5-point scale based on the degree of suction, ultrasonic aspiration, or scissors and cautery. A durometer was used to semiquantitatively measure the hardness of surgical specimens. Statistics included chi-squared, sensitivity, specificity, positive and negative predicative values (PPV and NPV), and Spearman rank correlation coefficient.

Results: On MRE and at surgery respectively, 8 (53%) versus 6 (40%) tumors were homogenous; 7 (47%) versus 9 (60%) tumors were heterogeneous; 6 (40%) versus 10 (67%) tumors had hard portions; and 14 (93%) versus 12 (80%) tumors had soft portions. MRE sensitivity, specificity, PPV and NPV were as follows: for heterogeneity, 78, 100, 100, and 75%; for hardness, 60, 100, 100, and 56%; and for softness, 100, 33, 86, and 100%. Overall, 10 (67%) tumors matched well between MRE and intraoperative consistency. In these 10 tumors, MRE measurements correlated well with intraoperative observations (p = 0.018) and durometer readings (p = 0.046). The durometer also correlated well with surgeon findings (p = 0.002). A tumor size ≤ 3.5 or very vascular tumors were more likely to be inconsistent between MRE and intraoperative findings (p < 0.05).

Conclusion: Regarding intratumoral stiffness, MRE was excellent at ruling-in whether a tumor would be heterogeneous (100% specificity) with hard portions (100% PPV), but was less effective in ruling-out heterogeneity (78% sensitivity) and hard portions (56% NPV). In some tumors, MRE showed as entirely soft, portions of the tumor were hard (33% specificity). MRE was inconsistent in tumors that were more vascular or < 3.5 cm. In the 10 tumors that MRE correctly evaluated, there was good correlation between intratumoral MRE measurements with intraoperative observations and durometer readings.