Predicting Adenomyosis Diagnosis Based on MRI

Abstract

Listen

Translate article

<h3>Study Objective</h3> Development of a prediction tool for histopathological adenomyosis diagnosis after hysterectomy based on MRI and clinical parameters. <h3>Design</h3> Single-centre retrospective cohort study. <h3>Setting</h3> Gynaecological department of regional referral hospital from 2007-2022. <h3>Patients or Participants</h3> 296 women undergoing a hysterectomy for benign pathology with preoperative pelvic MRI. <h3>Interventions</h3> MRI's were retrospectively re-assessed for all possible adenomyosis markers (junctional zone (JZ) parameters, high signal intensity foci (HSI foci), uterine size) in a blinded fashion by two researchers. Sensitivity, specificity, positive predictive value, negative predictive value, diagnostic accuracy, and odds ratio (dOR) were calculated. Threshold values of continuous variables were investigated using Receiver Operator Characteristics (ROC) curves and Area Under the Curve (AUC). A multivariate regression model for histopathological adenomyosis diagnosis was developed based on selection of MRI and clinical variables from univariate analysis with p>0.10 and factors determined to be of clinical importance. <h3>Measurements and Main Results</h3> 131 women (44.3%) had histopathological adenomyosis. In univariate analysis, patients had comparable age at hysterectomy, BMI and clinical symptoms, p>0.05. Patients with adenomyosis had more often undergone a curettage (221% vs. 8.9%, p=0.002), and had a higher mean JZ (9.35mm vs. 8.00, p=0.001), maximal JZ (15.05mm vs. 12.45mm, p=0.002), mean JZ to myometrium ratio (0.53 vs. 0.48 p=0.009) and JZ differential (8.45mm vs, 6.75mm, p=0.009). Presence of HSI foci on MRI was quasi-pathognomic for adenomyosis diagnosis (25.5% vs. 2.4%, p<0.001). A predictive model based on the parameters of: Age at MRI, History of Curettage, Dysmenorrhoea, Hypermenorrhoea, Mean JZ, JZ differential, JZ/Myometrium ratio, Presence of HSI Foci was created, with a good AUC of 0.761. <h3>Conclusion</h3> This is the first study to create a clinical diagnostic tool based on a combination of MRI and clinical parameters for adenomyosis diagnosis. As preoperative imaging-based diagnosis of adenomyosis remains challenging, this model, after sufficient external validation, could function to become a useful clinical-decision making tool in women with suspected adenomyosis.