A novel diagnostic index combining HE4, CA125 and age may improve triage of women with suspected ovarian cancer - An international multicenter study in women with an ovarian mass.

AIM: To develop and validate a biomarker-based index to optimize referral and diagnosis of patients with suspected ovarian cancer. Furthermore, to compare this new index with the Risk of Malignancy Index (RMI) and Risk of Ovarian Malignancy Algorithm (ROMA). PATIENTS AND METHODS: A training study, consisting of patients with benign ovarian disease (n=809) and ovarian cancer (n=246), was used to develop the Copenhagen Index (CPH-I) utilizing the variables serum HE4, serum CA125 and patient age. Eight international studies provided the validation population; comprising 1060 patients with benign ovarian masses and 550 patients with ovarian cancer. RESULTS: Overall, 2665 patients were included. CPH-I was highly significant in discriminating benign from malignant ovarian disease. At the defined cut-off of 0.070 for CPH-I the sensitivity and specificity were 95.0% and 78.4% respectively in the training cohort and 82.0% and 88.4% in the validation cohort. Comparison of CPH-I, ROMA and RMI demonstrated area-under-curve (AUC) at 0.960, 0.954 and 0.959 respectively in the training study and 0.951, 0.953 and 0.935 respectively in the validation study. Using a sensitivity of 95.0%, the specificities for CPH-I, ROMA and RMI in the training cohort were 78.4%, 71.7% and 81.5% respectively, and in the validation cohort 67.3%, 70.7% and 69.5% respectively. CONCLUSION: All three indices perform well at the clinically relevant sensitivity of 95%, but CPH-I, unlike RMI and ROMA, is independent of ultrasound and menopausal status, and may provide a simple index to optimize referral of women with suspected ovarian cancer.

The use of HE4 in the prediction of ovarian cancer in Asian women with a pelvic mass.

OBJECTIVE: The purpose of this study was to evaluate the performance of human epididymis protein 4 (HE4) and the Risk of Ovarian Malignancy Algorithm (ROMA) for distinguishing between benign and malignant pelvis masses in Asian women. METHODS: This was a prospective, multicenter (n=6) study with patients from six Asian countries. Patients had a pelvic mass on imaging and were scheduled to undergo surgery. Serum CA125 and HE4 were measured on preoperative samples. CA125, HE4, and ROMA were evaluated for sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV). RESULTS: A total of 414 women with an adnexal mass were evaluated, of which 65 had epithelial ovarian (EOC) cancer, 16 had borderline tumors and 11 had other malignant diseases. Compared to CA125, HE4 had lower sensitivity (56.9% vs 90.8%) and NPV (91.8% vs 97.3%), but improved specificity (96.9% vs 67.1%) and PPV (78.7% vs 35.8%) for differentiating between benign pelvic mass and EOC. ROMA had similar sensitivity (89.2% vs 90.8%) and NPV (97.6% vs 97.3%) as CA125, but showed improved specificity (87.3% vs 67.1%) and PPV (58.6% vs 35.8%). ROMA accurately predicted 87.3% of benign cases as low risk, and 82.6% of stage I/II EOC and 89.2% of all EOC as high risk. CONCLUSION: ROMA showed similar sensitivity as CA125 but improved specificity and PPV, especially in premenopausal women. Using ROMA may help predict if a pelvic mass is benign or malignant and facilitate subsequent management planning.

Distinguishing benign from malignant pelvic mass utilizing an algorithm with HE4, menopausal status, and ultrasound findings.

OBJECTIVE: The purpose of this study was to develop a risk prediction score for distinguishing benign ovarian mass from malignant tumors using CA-125, human epididymis protein 4 (HE4), ultrasound findings, and menopausal status. The risk prediction score was compared to the risk of malignancy index and risk of ovarian malignancy algorithm (ROMA). METHODS: This was a prospective, multicenter (n=6) study with patients from six Asian countries. Patients had a pelvic mass upon imaging and were scheduled to undergo surgery. Serum CA-125 and HE4 were measured on preoperative samples, and ultrasound findings were recorded. Regression analysis was performed and a risk prediction model was developed based on the significant factors. A bootstrap technique was applied to assess the validity of the HE4 model. RESULTS: A total of 414 women with a pelvic mass were enrolled in the study, of which 328 had documented ultrasound findings. The risk prediction model that contained HE4, menopausal status, and ultrasound findings exhibited the best performance compared to models with CA-125 alone, or a combination of CA-125 and HE4. This model classified 77.2% of women with ovarian cancer as medium or high risk, and 86% of women with benign disease as very-low, low, or medium-low risk. This model exhibited better sensitivity than ROMA, but ROMA exhibited better specificity. Both models performed better than CA-125 alone. CONCLUSION: Combining ultrasound with HE4 can improve the sensitivity for detecting ovarian cancer compared to other algorithms.