Bili Eleni, Dampala Kaliopi, Iakovou Ioannis, Tsolakidis Dimitrios, Giannakou Anastasia, Tarlatzis Basil C
First Department of Obstetrics and Gynecology, Medical School, Aristotle University of Thessaloniki, Papageorgiou General Hospital, Thessaloniki, Greece.
Third Department of Nuclear Medicine, Medical School, Aristotle University of Thessaloniki, Papageorgiou General Hospital, Thessaloniki, Greece.
Eur J Obstet Gynecol Reprod Biol. 2014 Aug;179:32-5. doi: 10.1016/j.ejogrb.2014.05.006. Epub 2014 May 20.
The aim of this study was to determine the performance of prostate specific antigen (PSA) and ultrasound parameters, such as ovarian volume and outline, in the diagnosis of polycystic ovary syndrome (PCOS).
This prospective, observational, case-controlled study included 43 women with PCOS, and 40 controls. Between day 3 and 5 of the menstrual cycle, fasting serum samples were collected and transvaginal ultrasound was performed. The diagnostic performance of each parameter [total PSA (tPSA), total-to-free PSA ratio (tPSA:fPSA), ovarian volume, ovarian outline] was estimated by means of receiver operating characteristic (ROC) analysis, along with area under the curve (AUC), threshold, sensitivity, specificity as well as positive (+) and negative (-) likelihood ratios (LRs). Multivariate logistical regression models, using ovarian volume and ovarian outline, were constructed.
The tPSA and tPSA:fPSA ratio resulted in AUC of 0.74 and 0.70, respectively, with moderate specificity/sensitivity and insufficient LR+/- values. In the multivariate logistic regression model, the combination of ovarian volume and outline had a sensitivity of 97.7% and a specificity of 97.5% in the diagnosis of PCOS, with +LR and -LR values of 39.1 and 0.02, respectively.
In women with PCOS, tPSA and tPSA:fPSA ratio have similar diagnostic performance. The use of a multivariate logistic regression model, incorporating ovarian volume and outline, offers very good diagnostic accuracy in distinguishing women with PCOS patients from controls.
本研究旨在确定前列腺特异性抗原（PSA）和超声参数（如卵巢体积和轮廓）在多囊卵巢综合征（PCOS）诊断中的表现。
这项前瞻性、观察性、病例对照研究纳入了43例PCOS女性患者和40例对照者。在月经周期的第3至5天，采集空腹血清样本并进行经阴道超声检查。通过受试者操作特征（ROC）分析评估每个参数[总PSA（tPSA）、总PSA与游离PSA比值（tPSA:fPSA）、卵巢体积、卵巢轮廓]的诊断性能，同时计算曲线下面积（AUC）、阈值、敏感性、特异性以及阳性（+）和阴性（-）似然比（LR）。构建了使用卵巢体积和卵巢轮廓的多变量逻辑回归模型。
tPSA和tPSA:fPSA比值的AUC分别为0.74和0.70，特异性/敏感性中等，LR+/-值不足。在多变量逻辑回归模型中，卵巢体积和轮廓的组合在PCOS诊断中的敏感性为97.7%，特异性为97.5%，+LR和-LR值分别为39.1和0.02。
在PCOS女性中，tPSA和tPSA:fPSA比值具有相似的诊断性能。使用包含卵巢体积和轮廓的多变量逻辑回归模型，在区分PCOS患者与对照者方面具有非常好的诊断准确性。
