Evaluation in the predictive value of serum human epididymal protein 4 (HE4), cancer antigen 125 (CA 125) and a combination of both in detecting ovarian malignancy.

Background This study was conducted to evaluate the performance of human epididymal protein 4 (HE4), cancer antigen 125 (CA 125) and a combination of both via the Risk of Ovarian Malignancy Algorithm (ROMA) in detecting ovarian malignancy. Methods This was a diagnostic study enrolling 129 patients with pelvic mass(es) suspected of originating in the ovary who had been scheduled for surgery or radiological-guided biopsy. Serum HE4 and CA 125 levels were measured. HE4, CA 125 and ROMA were evaluated for sensitivity, specificity, positive predictive value and negative predictive value. The receiver operating characteristic (ROC) plots were graphed and area under the curve (AUC) values were calculated to investigate the accuracy of each marker for predicting ovarian malignancy. Results Overall, CA 125 remained significantly more sensitive (88.9% vs. 51.9%, p = 0.006) but less specific (56.9% vs. 95.1%, p < 0.001) than HE4. HE4 was superior to CA 125 in specificity (97.7% vs. 54.5%, p < 0.001) for premenopausal women. ROMA was non-significantly more sensitive (100.0% vs. 92.3%, p = 1.000) than CA 125 but both were equally specific (71.4%) for the postmenopausal group. In the premenopausal group, the AUC of serum HE4 was higher than serum CA 125 (0.851 vs. 0.817) but was equivalent to ROMA (0.851 vs. 0.859). In the postmenopausal group, ROMA exhibited an excellent AUC value as compared to CA 125 and HE4 (AUC of 0.907 vs. 0.874 vs. 0.863, respectively). Conclusion HE4 is useful in ruling out ovarian malignancy among premenopausal women. For postmenopausal women, ROMA appears to be an all-rounder with overall good sensitivity and specificity.

Cardiovascular risk factors in pre-pubertal Malays: effects of diabetic parentage.

Diabetes prevalence is increasing rapidly in Asian populations but the influence of a family history of diabetes on cardiovascular risk is unknown. To assess this relationship, 120 urban-dwelling Malays were recruited to a cross-sectional case-control study. Sixty were pre-pubertal children, 30 of diabetic parentage (Group 1) and 30 with no diabetes family history (Group 2). Group 1 and 2 subjects were the offspring of adults with (Group 3) or without (Group 4) type 2 diabetes. Subjects were assessed for clinical and biochemical variables defining cardiovascular risk. Principal component analysis assessed clustering of variables in the children. Group 1 subjects had a higher mean waist:hip ratio, diastolic blood pressure and HbA(1c) than those in Group 2, and a lower HDL:total cholesterol ratio (P<0.03). Although there were no correlations between Group 1 and 3 subjects for cardiovascular risk variables, significant associations were found in Groups 2 and 4, especially HbA(1c) and insulin sensitivity (P< or =0.004). Of five separate clusters of variables (factors) identified amongst the children, the strongest comprised diabetic parentage, HbA(1c), insulin sensitivity and blood pressure. Features of the metabolic syndrome are becoming evident in the young non-obese children of diabetic Malays, suggesting that lifestyle factors merit particular attention in this group.