Evaluation of serum CA125, HE4 and CA724 and the risk of ovarian malignancy algorithm score in the diagnosis of high-grade serous ovarian cancer

Aim: To develop a new algorithm for the detection of high-grade serous ovarian cancer (HGSOC). Methods: Patients diagnosed with HGSOC, borderline ovarian tumours (BOTs) or benign ovarian masses (BOMs) were enrolled between February 2019 and December 2020. Patients with BOTs or BOMs were grouped as nonHGSOC. The cases were divided randomly into a training cohort (two-thirds of cases) and a validation cohort (one-third of cases). Logistic regression was used to find risk factors for HGSOC and to create a new algorithm in the training cohort. Receiver operating characteristic curves were used to compare the diagnostic value of tumour biomarkers. Sensitivity and specificity of tumour markers and the new algorithm were calculated in the training cohort and validation cohort. Results: This study found significant differences in age; BRCA1/2 mutation status; CA125, CA724 and HE4 levels; and Risk of Ovarian Malignancy Algorithm score between the two groups. Logistic regression analysis showed that CA125 and BRCA1/2 were risk factors for HGSOC. A new algorithm combining CA125 and BRCA1/2 increased the specificity of CA125 for diagnosis of HGSOC. The new algorithm had sensitivity of 81.08% and specificity of 93.10% in the training cohort. Conclusion: The new algorithm using CA125 and BRCA1/2 helped to distinguish between patients with HGSOC and patients with non-HGSOC.