Urban Renata R, Smith Alan, Agnew Kathy, Bonato Vinicius, Goff Barbara A
*Division of Gynecologic Oncology, University of Washington Medical Center, Seattle, WA; †Applied Clinical Intelligence, Bala Cynwyd, PA; and ‡Vermillion, Inc, Austin, TX.
Int J Gynecol Cancer. 2017 Feb;27(2):233-238. doi: 10.1097/IGC.0000000000000873.
This study aimed to evaluate the predictive ability of a multivariate biomarker test in combination with a symptom index (SI) to identify ovarian cancer in a cohort of women planning to undergo surgery for a pelvic mass.
This was a prospective study of patients seen at a tertiary care medical center. Following consent, patients completed an SI and preoperative serum was collected for a Food and Drug Administration-cleared multivariate biomarker test [multivariate index assay (MIA)]. Results for the SI and MIA were correlated with operative findings and surgical pathology.
Of 218 patients enrolled, 124 (56.9%) had benign disease and 94 (43.1%) had borderline tumors or carcinomas. Sixty-six patients had a primary ovarian or fallopian tube cancer. The median age of patients enrolled in this study was 54 years (interquartile range, 44-63 years), of whom 148 (67.9%) were postmenopausal. More than a third (36.3%) of patients with benign masses was accurately identified as low risk by MIA and SI. The sensitivity and negative predictive value (NPV) of the SI relative to primary ovarian cancer was 87.9% (95% CI, 77.9%-93.7%) and 91.6% (95% CI, 84.3%-95.7%), respectively. The sensitivity and NPV of CA125 was 75.4% (95% CI, 63.7%-84.2%) and 86.4% (95% CI, 79.1%-91.5%), respectively, and the sensitivity and NPV of the MIA were 93.9% (95% CI, 85.4%-97.6%) and 94.5% (95% CI, 94.5%-100%), respectively. The overall sensitivity for the combination of MIA plus SI was 100% (66/66; 95% CI, 94.5%-100%), and specificity was 36.3% (45/124; 95% CI, 28.4%-45.0%), with an NPV of 100% (95% CI, 92.1%-100%).
The addition of a patient-reported SI, which captures subjective symptoms in an objective manner, improved the sensitivity of MIA across all stages and subtypes of ovarian cancer.
本研究旨在评估一种多变量生物标志物检测结合症状指数（SI）在一组计划接受盆腔肿块手术的女性中识别卵巢癌的预测能力。
这是一项对一家三级医疗中心的患者进行的前瞻性研究。获得患者同意后，患者完成SI评估，并采集术前血清用于一项获得美国食品药品监督管理局批准的多变量生物标志物检测[多变量指数分析（MIA）]。SI和MIA的结果与手术结果及手术病理相关。
在纳入的218例患者中，124例（56.9%）患有良性疾病，94例（43.1%）患有交界性肿瘤或癌。66例患者患有原发性卵巢或输卵管癌。本研究纳入患者的中位年龄为54岁（四分位间距，44 - 63岁），其中148例（67.9%）为绝经后女性。超过三分之一（36.3%）的良性肿块患者通过MIA和SI被准确识别为低风险。相对于原发性卵巢癌，SI的敏感性和阴性预测值（NPV）分别为87.9%（95%CI，77.9% - 93.7%）和91.6%（95%CI，84.3% - 95.7%）。CA125的敏感性和NPV分别为75.4%（95%CI，63.7% - 84.2%）和86.4%（95%CI，79.1% - 91.5%），MIA的敏感性和NPV分别为93.9%（95%CI，85.4% - 97.6%）和94.5%（95%CI，94.5% - 100%）。MIA加SI联合检测的总体敏感性为100%（66/66；95%CI，94.5% - 100%），特异性为36.3%（45/124；95%CI，28.4% - 45.0%），NPV为100%（95%CI，92.1% - 100%）。
添加一种以客观方式捕捉主观症状的患者报告的SI，提高了MIA在卵巢癌所有阶段和亚型中的敏感性。
