The role of multiparametric MRI in differentiating uterine leiomyosarcoma from benign degenerative leiomyoma and leiomyoma variants: a retrospective analysis. Issue 1 (January 2023)

• Record Type: Journal Article
• Authors: Mahmood, N.S.; Al Rashid, A.A.; Ladumor, S.B.; Mohamed, M.A.; Kambal, A.S.; Saloum, N.; Mohamed, S.E.M.K.; Al Hyassat, S.; Singh, R.
• Is Part Of: Clinical radiology. Volume 78:Issue 1(2023)
• Page Start: 47
• Page End: 54
• Publication Date: 2023-01
• Subjects: Medical radiology -- Periodicals; Radiotherapy -- Periodicals; Radiotherapy -- Periodicals; Radiology -- Periodicals; Societies, Medical -- Periodicals; Medical radiology; Radiotherapy; Electronic journals; Periodicals; 616.0757
• DOI: 10.1016/j.crad.2022.08.144
• Languages: English
• ISSNs: 0009-9260

Abstract

Abstract : Aim: To assess qualitative and quantitative magnetic resonance imaging (MRI) factors that can help distinguish leiomyosarcoma (LMS) from benign degenerative leiomyoma (BDL) and leiomyoma variants (LV) and assess the interobserver agreement for the proposed quantitative factors. Materials and methods: Retrospective analysis of all histopathology proven cases of LV, BDL, and LMS with a preoperative MRI was performed. Twenty-seven cases were included (five LMS, three LV, and 19 BDL) with each case independently read by a pair of radiologists. Lesion size, margins, presence or absence of degeneration, necrosis, and haemorrhage were assessed on MRI along with quantitative factors such as mean T2-weighted (W) and T1W signal intensity, T1W signal heterogeneity, diffusion-weighted imaging (DWI), and apparent diffusion coefficient (ADC) ratios as well as dynamic contrast enhancement (DCE) characteristics along with the presence or absence of lymphadenopathy and extra-uterine and peritoneal spread. Mean and standard deviation for quantitative variables and frequency with percentages for qualitative variables were assessed. Results: Infiltrative margins were seen exclusively in the LMS group ( n= 1), with the remaining LMS cases showing lobulate or rounded smooth margins similar to BDL or LV. A high T2W signal <25% was seen exclusively in the BDL group ( n= 8). The presence of concomitant necrosis and haemorrhage was seen exclusively in the LMS group ( n= 2). Quantitative MRI had good inter-reader correlation but was not significantly different between the LMS, BDL, and LV groups. Conclusion: LMS, BDL, and LV may have overlapping features on multiparametric MRI making differentiation difficult. Highlights: Differentiating LMS from BDL and LV is important. Infiltrative margin, extrauterine spread and lymphadenopathy was unique to LMS. Concomitant necrosis and hemorrhage was also seen only in LMS. High T2W signal less than 25% of the cross-section was seen exclusively in BDL. Quantitative parameters were not helpful to differentiate LMS from BDL or LV.