Young-Sun Kim MD, Presenter: Nothing to Disclose

Hyo Keun Lim MD, Abstract Co-Author: Nothing to Disclose

Hyunchul Rhim MD, PhD, Abstract Co-Author: Nothing to Disclose

To generate screening MRI-based prediction model for therapeutic responses of MR-guided high-intensity focused ultrasound (MR-HIFU) ablation of uterine fibroids

A total of 160 symptomatic uterine fibroids (diameter 8.3cm, range 3.1-15.0cm) in 112 women (age 43.3, range 25-55) who were treated with MR-HIFU ablation were retrospectively analyzed. The following three parameters of screening MRI were evaluated. 1) Subcutaneous fat was measured as a thickness of the most compressed point (mm) on prone position. 2) Relative peak enhancement (%) was calculated based on time-signal intensity curve analysis of fibroid in perfusion MRI (100 dynamics, 3s time resolution), in which 0% refers the same signal intensity as in precontrast image. 3) Signal intensity was assessed as a ratio of T2 signal intensity of uterine fibroids to that of skeletal muscle. Those parameters were used to generate prediction models with regards to ablation efficiency (i.e., non-perfused volume/treatment cell volume) and ablation quality (grade 1~5, from poor to excellent), respectively, using generalized estimating equation (GEE) analysis. Then, cut-off values for successful treatment (ablation efficiency >1.0; ablation quality grade 4 or 5) were determined based on receiver operating characteristic (ROC) curve analyses.

GEE analyses produced the models of “y1=2.2637–0.0415x1–0.0011x2–0.0772x3” and “y2=6.8148–0.1070x1–0.0050x2–0.2163x3”, where y1=ablation efficiency, y2=ablation quality, x1=subcutaneous fat thickness, x2=relative peak enhancement, and x3=T2 signal intensity ratio (p-values for x1, 0.0068 and <0.0001; for x2, 0.1952 and 0.0001; for x3, <0.0001 and <0.0001, respectively). Cut-off values for successful treatments based on ROC curve analyses turned out to be 1.312 for of ablation efficiency (AUC, .7236; sensitivity, .6882; specificity, .6866) and 4.019 for ablation quality (AUC, .8794; sensitivity, .7156; specificity, .9020).

Simple equation models to predict therapeutic responses of MR- HIFU ablation of uterine fibroids in terms of ablation efficiency and quality were generated, which are easily applicable to screening MRI.

With regards to MR- HIFU ablation of uterine fibroids, there have been no screening MR criteria that comprehensively consider multiple influencing factors. These prediction models would contribute to reducing the risk of unsuccessful, thus wasteful procedures.

Kim, Y, Lim, H, Rhim, H, Screening MRI-based Prediction Model for Therapeutic Response of MR-HIFU Ablation of Uterine Fibroids.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14003971.html