Abstract
OBJECTIVES: We sought to compare the epidemiological and clinical features of variant leiomyoma, smooth muscle tumors of uncertain malignant potential (STUMP), and sarcoma to classical leiomyoma and assess the radio-histological correlation. METHODS: This retrospective, observational study was conducted in Royal Hospital, Muscat, Oman. We included all patients who underwent surgical treatment for uterine fibroids from 1 January 2011 to 31 December 2016. Data was retrospectively collected from electronic patient records, pictures archiving system, and histopathology reports, and analyzed using SPSS version 29 and MedCalc Statistical Software. RESULTS: Out of the 14269 patients who attended the gynecology outpatient department, 19.5% were clinically diagnosed with uterine fibroids. Surgical procedures were performed in 545 (19.6%) women. Diagnosis from histopathology reports was classical leiomyoma in 508 (93.2%) cases, variant leiomyoma in 35 (6.4%), and sarcomas in two (0.4%); there were no cases of STUMP. Magnetic resonance imaging (MRI) correctly identified 125 (89.9%) cases of classical leiomyoma and 2/11 (18.2%) cases of variants. The sensitivity and specificity of MRI to predict non-classical varieties was 22.22% and 88.65%, respectively. MRI had a strong negative predictive value (94.7%) for leiomyoma variants. The area under the curve for classical and variant leiomyoma was 0.63 (0.55-0.71) and 0.55 (0.47-0.64). The accuracy of MRI in predicting classical leiomyoma was 86.00 (79.40-91.12) and 84.67 (77.87-90.03) the non-classical varieties. CONCLUSIONS: Epidemiological and clinical features were not helpful in distinguishing between classical leiomyoma and non-classical varieties. MRI had weak discriminatory power to distinguish between classical and non-classical varieties.