| [Objective] Uterine leiomyoma is the most common benign tumor of the female reproductive tract, and abdominal hysterectomy or myomectomy is the most traditional surgical approach. In recent years, less invasive surgical and medical options have been developed and obtained preferably curative effect, including uterine arterial embolization (UAE), GnRH analogues (GnRH-a) et al. But not all of these methods are efficacious to each leiomyoma, the curative effect is related to the histological subtypes of the leiomyomas. These methods are merely efficient to the uncomplicated leiomyomas rather than to degenerated leiomyomas. Besides, these methods have side effects to the patients. A number of foreign scholars reported that MRI could correctly differentiate the subtypes of the leiomyomas, and was reliable to select leiomyomas which could be cured by conservative management. However there were no reports on the accuracy of MRI in differentiating the subtypes of the leiomyomas in our country. The present study is to evaluate the accuracy of MRI in predicting leiomyoma subtypes, and correlate signal intensity on TaWI and cellular content of the leiomyomas.[Materials and Methods] There were 35 patients enrolled in the study, of which average age was 45. MRI was performed preoperatively on a 1.5Tesla unit (Signa;GE Medical Systems). All the patients were scaned in body coil, acquiring T1WI images with conventional spin-echo technique and TaWI images with fast spin-echo technique in the axial and sagittal plane. Fat-suppressed T1WI images were performed if a leiomyoma demonstrate high signal intensity on TiWI. Respiratory compensation technique and abdomen belt were applied to diminish the artifacts. Leiomyomas which were the solitary or the biggest of multiple or whose signal intensity were not typical were included in our study. Leiomyomas were classified based on the MRI finding. The signal intensity of leiomyomas, normal uterine muscle, skeletal muscle and fat were measured on T2WI, the percent contrast in signal intensity of the leiomyomas and the uterine muscle were calculated. After observing the macroscopic appearance of the leiomyoma section, we sliced the leiomyomas serially to be embeded in paraffin and stained with hematoxylin and eosin. The degenerated leiomyomas were distinguished from the uncomplicated, and the subtypes of degeneration were diagnosed on the microscope. The sensitivity, specificity, accuracy, PPV and NPV of MRI in differentiating leiomyoma subtypes were evaluated. The signal intensity of uncomplicated leiomyomas on TaWI were correlated with whose cellular content and the correlative coefficient were calculated.[Results] Thiry-eight surgical proved leiomyomas in 35 patients were included in our study. Among them, there were 30 uncomplicated (including 9 cellular and 21 ordinary). The other 8 were degenerated, including 3 myxoid, 2 hyaline, 2 hemorrhagic and 1 cystic. Thirty-three leiomyoma subtypes were accurately identified by MRI in 38, the accuracy is 86.84%. MRI correctly identified all of the uncomplicated leimyomas, and eight of them were accurately identified as cellular leiomyoma, but two of them were misdiagnosed as cellular leiomyoma. MRI had an 80% sensitivity, 96.43% specificity, 92.11% accuracy, 88.89% PPV and 93.1% NPV for diagnosing a cellular leiomyoma, had an486.36% sensitivity, 87.5% specificity, 86.84% accuracy, 90.48% PPV and 82.35% NPV for diagnosing an ordinary leiomyoma. All of the 3 myxoid, 2 hemorrhagic and 1 cystic leiomyomas were correctly identified by MRI, 2 hyaline leiomyomas could not be distinguished from ordinary leiomyomas since they shared identical appearance. The sensitivity, specificity, accuracy, PPV and NPV for diagnosing a degenerated leiomyoma were 75%, 100%, 94.74%, 100% and 93.75%.[Conclusions] MRI is fairly accurate in differentiating betweendegenerated leiomyomas and umcomplicated. MRI can correctly identify myxoid, hemorrhagic and cystic leiomyomas of the degenerated leiomyomas. However, MRI is incapable of differen...
|
PDF Full Text Request