Preinterventional quantitative magnetic resonance imaging predicts uterus and leiomyoma size reduction after uterine artery embolization

Abstract

To investigate the relationship between magnetic resonance imaging (MRI) measures and uterus and leiomyoma size reductions after uterine artery embolization (UAE). Fifty-two women with leiomyomas underwent selective UAEs. Uterine and dominant leiomyoma sizes were measured with preinterventional MRI and a 6-month follow-up MRI. Four MRI measures of the dominant leiomyoma were recorded: T1 time; T2 time; leiomyoma-to-skeletal muscle T2 SI-ratio; and percentage of contrast enhancement. To evaluate the predictive value of MRI measures we used Spearman rank correlation, area under the receiver operating characteristic (ROC) curve (A(z)), and values for diagnostic performance. Uterus and dominant leiomyoma size reductions were highly variable. Leiomyoma size reductions of >or=75% were accurately predicted with leiomyoma-to-skeletal muscle T2 SI-ratio (ROC curve A(z) = 0.930; 95% confidence interval [CI]: 0.853, 1.000). Leiomyoma size reductions >or=75% were predicted by leiomyoma-to-skeletal muscle T2 SI-ratio >or=3.5 and T1-time >or=750 msec with 100% and 86% sensitivities and 67% and 72% specificities, respectively. Uterus size reduction >or=50% were identified by dominant leiomyoma-to-skeletal muscle T2 SI-ratio >or=2.5. Uterus and dominant leiomyoma size reductions after UAE were predicted with preoperative MRI measures of the dominant leiomyoma.