Magnetic resonance imaging presentation of diffuse and focal adenomyosis before and after pregnancy.

RESEARCH QUESTION: Is there a change in magnetic resonance imaging (MRI) criteria of diffuse and focal phenotypes of adenomyosis before and after pregnancy? DESIGN: A retrospective, monocentric, observational study in a single academic tertiary referral centre for endometriosis diagnosis and management. Women were followed for symptomatic adenomyosis, and without a prior history of surgery who give birth after 24+0 weeks. For each patient, pelvic MRI pre- and post-pregnancy was performed by two experienced radiologists with the same image acquisition protocol. Diffuse and focal adenomyosis MRI presentation were analysed before and after pregnancy. RESULTS: Between January 2010 and September 2020, of the 139 patients analysed, 96 (69.1%) had adenomyosis at MRI distributed as follow: 22 (15.8%) presented diffuse adenomyosis, 55 (39.6%) focal adenomyosis and 19 (13.7%) both phenotypes. The frequency of isolated diffuse adenomyosis on MRI was significantly lower before versus after pregnancy (n = 22 [15.8%] versus n = 41 [29.5%], P = 0.01). The frequency of isolated focal adenomyosis was significantly higher before pregnancy than after pregnancy (n = 55 [39.6%] versus n = 34 [24.5%], P = 0.01). The mean volume of all focal adenomyosis lesions on MRI decreased significantly after pregnancy, from 6.7 ± 2.5 mm3 to 6.4 ± 2.3 mm3, P = 0.01. CONCLUSION: The current data indicate that, based on MRI, there is an increase in diffuse adenomyosis and a decrease in focal adenomyosis after pregnancy.

Histological Risk Factors for Local Recurrence of Phyllodes Tumors of the Breast.

BACKGROUND/AIM: The rate of local recurrence (LR) of phyllodes tumor (PT) varies from 4 to 18%. Several histological risk factors of LR of PT are known. The aim of this study was to estimate the LR rate of PT according to PT grade and to evaluate histological risk factors of PT LR in our retrospective cohort. PATIENTS AND METHODS: This was a two-center study, conducted from 1995 to 2019. All patients with PT diagnosed on surgical specimen were included. PT was diagnosed histologically according to the grade category defined by the 2012 World Health Organization classification as benign, borderline or malignant PT. Univariate analysis and then multivariate logistic regression analysis were performed to determine histological risk factors of LR of PT. RESULTS: A total of 224 patients with PT were included: 152 with benign, 49 with borderline and 23 with malignant PT. The median and standard deviation for the duration of follow-up was 136.60 ± 167.43 months, and 18 patients (8.04%) developed LR: 7 (4.61%), 7 and (14.29%) and 4 (17.39%) with benign, borderline and malignant PT, respectively. In univariate analysis, LR was statistically increased for histological size ≥45 mm (p=0.003), borderline/malignant TP (p=0.006) and dense stromal cellularity (p<0.001). In multivariate analysis, only histological size ≥45 mm and cellularity were statistically associated with LR (odds ratio=1.83, 95% confidence interval=1.06-9.83, p=0.04; and odds ratio=3.69, 95% confidence interval=1.11-12.28, p=0.03, respectively). CONCLUSION: Histological size ≥45 mm and dense stromal cellularity were demonstrated as histological risk factors of LR of PT. In our cohort, no association was found between LR and PT grade nor LR and surgical margins ≥10 mm.