Prospective RegIstration Of the peRformance of the IoTa ADNEX Model in Dutch gYnaecological Practice

Rationale: Correct pre-operative classification of ovarian tumors is critical for optimal treatment of both women with benign and malignant ovarian tumors. The International Ovarian Tumor Analysis (IOTA) consortium developed several risk-assessment strategies, including the Assessment of Different NEoplasias in the adneXa (ADNEX) model. This is a polytomous model that estimates the risk of malignancy in women with an ovarian tumor and can help differentiate benign from malignant ovarian tumors. Recent validation studies have shown superior diagnostic accuracy of the ADNEX model compared with previously developed risk-assessment strategies. However, the diagnostic accuracy of this model in daily clinical practice is still unknown. Furthermore, expertise is necessary to use this model before the results are reliable. With the recently published Dutch guideline on risk-assessment strategies of ovarian tumors, in which it is advised to use the ADNEX model, clinical data on diagnostic performance of the ADNEX model in daily Dutch gynaecological practice are necessary. Lower sensitivity and/or specificity of the model will have significant influence on the referral practice in The Netherlands.

Objective: To determine the diagnostic accuracy of the IOTA ADNEX model in the discrimination between benign and malignant ovarian tumors in daily gynaecological practice in The Netherlands.

Study design: Multicentre prospective cohort study. Hospitals from three oncologic networks in the Netherlands will participate in this trial. Ultrasound data will be prospectively recorded in a database for a duration of five years by means of an electronic case report form (eCRF). With these data, the diagnostic accuracy of the IOTA ADNEX model will be calculated.

Study population: All women ≥ 18 years with an ovarian tumor for which ultrasound assessment in accordance with the IOTA definitions has been performed and the ADNEX model has been applied. Women with a tumor that meets the IOTA criteria of a benign 'easy descriptor' will be excluded.

Data collection: All patients eligible for this study will be asked for informed consent and clinical data will be collected through eCRFs. Data will be collected in a cloud-based database by the treating physician.

Main study parameters/endpoints: The sensitivity, specificity and area under the curve of the IOTA ADNEX model.

Secondary analysis: Secondary endpoints include the influence of practice variation (eg type of center, volume per center, prevalence of malignancy, availability of consultation time specifically aimed at performing ultrasounds according to IOTA criteria) and the experience of the ultrasound performer on the diagnostic accuracy of the IOTA ADNEX model.

Inclusion criteria

• Age ≥ 18 years
• Presence of a complex ovarian tumor
• Ultrasound assessment in accordance with IOTA definitions and application of the ADNEX model
• Understanding of Dutch language

Exclusion criteria

• Women with an ovarian tumor that meets the IOTA criteria of a benign easy descriptor:

• Unilocular tumor with ground glass echogenicity in premenopausal woman suggestive of endometrioma
• Unilocular tumor with mixed echogenicity and acoustic shadows in premenopausal women (suggestive of benign cystic teratoma)
• Unilocular tumor with regular walls and maximal diameter < 10cm (suggestive of simple cyst or cystadenoma)
• Remaining unilocular tumors with regular walls
• Histologically proven malignancy prior to ultrasound