J Med Assoc Thai 2016; 99 (12):1263

Views: 1,530 | Downloads: 60 | Responses: 0

PDF XML Respond to this article Print Alert & updates Request permissions Email to a friend


Diagnostic Accuracy and Optimal Cutoff Value of Serum HE4 to Predict Ovarian Cancer in Thai Women with Pelvic Masses
Yanaranop M , Tiyayon J , Nakrangsee S , Thinkhamrop B Mail

Background: Serum human epididymis protein 4 (HE4) is highly expressed in women with ovarian cancers (OCs), but data about its clinical application of HE4 for Thai women is limited.

Objective: To evaluate the diagnostic accuracy and optimal cutoff for HE4 in distinguishing benign lesions, borderline ovarian tumor (BOTs), and OCs compared with CA125 in Thai women at Rajavithi Hospital.

Material and Method: The cross-sectional study was conducted in Thai women aged older than 18 years old with pelvic masses whom underwent elective surgery at Rajavithi Hospital between 2012 and 2013. Preoperative serum HE4 and CA125 levels were measured and pathologic specimens were reviewed.

Results: Of the 518 participants evaluated, 316 had benign lesions, 43 had BOTs, and 159 had OCs. Between non-cancers and OCs, area under receiver operating characteristic curve (ROC-AUC) for HE4 hardly differed from CA125 (0.85 vs. 0.83, p = 0.402) but was significantly lower in postmenopausal women (0.79 vs. 0.86, p = 0.049). The optimal cutoff value of HE4 was 72 pM/L for all menopausal status. Lower HE4 was seen in 30.8% of mucinous carcinoma and 31.7% of clear cell carcinoma. The HE4 ROC-AUC was significantly higher than CA125 ROC-AUC in distinguishing benign diseases and BOTs (0.71 vs. 0.53, p<0.001), HE4 in 70% of BOTs was 51 to 95 pM/L.

Conclusion: Although the 72 pM/L cutoff for HE4 was appropriate in distinguishing between non-cancers and OCs for both pre- and postmenopausal women, the limitation for postmenopausal women, mucinous carcinomas, and clear cell carcinomas require to be complemented with CA125.

Keywords: Pelvic mass, Ovarian cancer, Human epididymis protein 4, Cutoff value


Download: PDF