OP11.03: Diagnostic accuracy of two‐dimensional transvaginal ultrasound in the diagnosis of adenomyosis

To evaluate the diagnostic accuracy of two-dimensional transvaginal ultrasound (2D TVS) in women before transcervical resection of the endometrium (TCRE) (group A) considered having discreet adenomyosis and to compare to women planned for hysterectomy (group B) considered having severe adenomyosis. Consecutive premenopausal women (n = 137) with abnormal uterine bleeding had a 2D TVS examination before surgery. Sonographic signs of adenomyosis were evaluated and an overall subjective diagnosis of adenomyosis was listed as; yes, no or indefinite. The reference standard was standardised histologic verification of hysterectomy specimens (group B) and pathology of deeply resected chips covering the whole uterine cavity and adjacent myometrium (group A). These chips were orientated, stretch out and pinned separately. All analyses were made by one pathologist, blinded to the sonographic data. Indefinite diagnoses (n = 27) with 2D TVS were excluded for the final analysis. In 110 women the sensitivity, specificity and AUC were 76.7%, 60.0% and 0.68 (95% CI 0.59–0.78) respectively. The prevalence of adenomyosis was 18.8% in group A and 39.1% in Group B. Group A (n = 64) had a sensitivity of 72.2%, specificity of 71.4% and AUC of 0.69 (95% CI 0.56–0.82). Group B (n = 46) had a sensitivity of 83.3%, specificity of 53.8% and AUC of 0.72 (95% CI 0.58–0.86). There was no statistic significant difference between the AUC in two groups (A and B). 2D TVS may be useful for the diagnosis of adenomyosis in women for TCRE with presumed discrete symptomatic adenomyosis. However, other larger studies with standardised histologic verification in non-hysterectomies women are needed to evaluate and specify diagnostic criteria's for discrete adenomyosis.