Accuracy of findings in the diagnosis of uterine adenomyosis on ultrasound.

PURPOSE: MRI is the current imaging gold standard to diagnose adenomyosis, but access is often limited by high costs and availability. Transvaginal ultrasound provides a cost-effective, accurate and readily available alternative. The objective of our study was to determine the diagnostic accuracy of commonly described sonographic findings in predicting uterine adenomyosis.

METHODS

This retrospective study evaluated 649 MRI studies performed to investigate adenomyosis with a preceding transvaginal ultrasound within 12 months between 2013 and 2018. Two blinded reviewers assessed the presence or absence of six sonographic features: bulky uterus, heterogeneous myometrium, streaky myometrium, myometrial cysts, endometrial-myometrial interface ill-definition, and echogenic linear striations. The sensitivity, specificity, positive and negative predictive values of these features were calculated individually and in combination when compared to MRI as the standard of reference.

RESULTS

Adenomyosis was found in 315 (48.5%) cases on MRI. Ultrasound had a high specificity of 91.8% (95% CI 88.4 to 94.6%) but was less sensitive (36.8% (95% CI 31.5 to 42.4%)) for detecting adenomyosis. Comorbid fibroids or focal adenomyosis did not affect diagnostic accuracy. All six variables were significantly more common in patients with adenomyosis compared to those without. Individually, 'bulky uterus' and 'heterogenous myometrium' each demonstrated a mean sensitivity and specificity > 50%. The best dual combined variables were 'bulky uterus' + 'ill definition of the endometrial-myometrial interface' (sensitivity 39%, specificity 91%). The best triple combined variables were 'bulky uterus', 'heterogeneous myometrium' + 'ill definition of the endometrial-myometrial interface' (sensitivity 38%, specificity 93%).

CONCLUSION

Transvaginal ultrasound is highly specific for diagnosing uterine adenomyosis, providing a cost-effective and readily available alternative to MRI.