ROMA and IOTA Score Comparison in Predicting Ovarian Tumour Malignancy

Abstract

Ovarian cancer is the sixth most common cancer found among women. Incidence of ovarian cancer in Dr. M. Djamil Central Hospital had increased from 103 cases to 156 cases in 2011-2012. The Risk Of Ovarian Malignancy Algorithm (ROMA) consists of qualitative serum test where several variables including HE4 serum, CA125 serum, and menopausal status are turned into numerical score (Hye Yon Cho et al 2015). The latest diagnostic modalities developed by International Ovarian Tumour Analysis (IOTA) classifies tumour into four sub groups, based on ultrasound characteristics: unilocular, multilocular, adnexal mass with solid component without papilloma projection, adnexal mass with one or more papilloma projection. Based on this knowledge, author would like to investigate accuracy of ROMA and IOTA score in predicting malignancy status of early ovarian tumour.

Objective: Determining ROMA and IOTA so thus can be utilised to predict wether the ovarian tumour is benign or malignant.

Method: This was analytical study, with cross sectional design and took place in Obstetrical and Gynaecological Outpatient Clinic of Dr. M. Djamil Central Hospital from August 2016 to January 2019. Sample was recruited using consecutive sampling method. Sixty one participants were recruited with CI 99%. Patients who met inclusion criteria would undergo Ca125 serum test and IOTA score. HE4 level was determined after surgery  and the sample of a tumour was sent to pathology anatomy laboratory of Medical Faculty of Andalas University to investigate the malignancy status. Statistical analysis involved univariate and bivariate test. The variables included frequency distribution, ROMA, and IOTA score. Bivariate analysis included specificity, sensitivity, positive predictive value, negative predictive value, positive likelihood ratio, negative likelihood ratio.

Result: Thirty one patients (50,8%) from peri menopausal group had HE4 level ≤70 pmol/L and 30 patients (49,2%) from post menopause group had HE4 level ≤140 pmol/L. Eleven patients (18%) had CA125 level <35 and 50 patients (82%) had CA125 level ≥35. ROMA index showed low risk in 17 patients (27,9%) and high risk 44 patients (72,1%). IOTA with M Rules was 51 patients (83,6%) and B rules was 10 patients (16,4%). Diagnostic value for ROMA score had sensitivity of 71,9%; RMI of 84,2%; and IOTA 82,5%. ROMA and IOTA accuracy was 68,9% and 92,2%, respectively.

Conclusion: IOTA has better sensitivity in predicting ovarian tumour compared with ROMA. ROMA has higher positive predictive value compared with IOTA positive predictive value. ROMA negative predictive value is 5,9%, meanwhile for IOTA it is hard to determine

Keywords: ovarian cancer, ROMA, IOTA, Ca-125, HE4