Abstract | OBJECTIVES: METHODS: Frozen serum samples collected from 484 patients were divided into three groups based on histopathological results: OC (n = 119), borderline ovarian tumors (BR) (n = 48), and benign ovarian tumors (BN) (n = 317). Diagnostic accuracy was calculated with an area under a receiver operating characteristic (AUC) curve. RESULTS: TFPI2 achieved the highest discrimination between the OC + BR group versus the BN group (AUC 0.8076). ROMA values best discriminated patients with OC from those with BN (AUC, 0.8966), which was equivalent to TFPI2 (AUC, 0.8937). For discriminating the OC group from the BR + BN group, the highest AUC value was achieved by ROMA values (AUC, 0.8884), and TFPI2 also showed comparable diagnostic accuracy (AUC, 0.8845). Combining TFPI2 with ROMA had the highest AUC (0.8420-0.9357). CONCLUSION: TFPI2 may be a clinically useful single marker comparable to conventional ROMA values for discriminating between benign and malignant ovarian tumors.
|
Authors | Hiroshi Kobayashi, Yuki Yamada, Ryuji Kawaguchi, Norihisa Ootake, Shohei Myoba, Fuminori Kimura |
Journal | The journal of obstetrics and gynaecology research
(J Obstet Gynaecol Res)
Vol. 48
Issue 9
Pg. 2442-2451
(Sep 2022)
ISSN: 1447-0756 [Electronic] Australia |
PMID | 35778814
(Publication Type: Journal Article)
|
Copyright | © 2022 Japan Society of Obstetrics and Gynecology. |
Chemical References |
- Biomarkers, Tumor
- CA-125 Antigen
- Lipoproteins
- lipoprotein-associated coagulation inhibitor
|
Topics |
- Algorithms
- Biomarkers, Tumor
- CA-125 Antigen
- Carcinoma, Ovarian Epithelial
(diagnosis)
- Female
- Humans
- Lipoproteins
- Ovarian Neoplasms
(diagnosis)
- ROC Curve
|