Diagnostic Value of Carcinoembryonic Antigen in Ascites for Colorectal Cancer with Peritoneal Carcinomatosis.
10.4166/kjg.2018.71.6.332
- Author:
Sung Eun SONG
1
;
Paul CHOI
;
Jae Hyun KIM
;
Kyoungwon JUNG
;
Sung Eun KIM
;
Won MOON
;
Moo In PARK
;
Seun Ja PARK
Author Information
1. Department of Internal Medicine, Kosin University College of Medicine, Busan, Korea. parksj6406@daum.net
- Publication Type:Original Article
- Keywords:
Carcinoembryonic antigen;
Ascites;
Peritoneal neoplasms;
Colorectal neoplasms
- MeSH:
Ascites*;
Carcinoembryonic Antigen*;
Carcinoma*;
Colorectal Neoplasms*;
Diagnostic Tests, Routine;
Humans;
Logistic Models;
Medical Records;
Ovarian Neoplasms;
Paracentesis;
Peritoneal Neoplasms;
Retrospective Studies;
ROC Curve;
Sensitivity and Specificity
- From:The Korean Journal of Gastroenterology
2018;71(6):332-337
- CountryRepublic of Korea
- Language:English
-
Abstract:
BACKGROUND/AIMS: Diagnostic tests for carcinoembryonic antigen (CEA) in ascites have been performed in various malignant cases, but there is only few data on the applicability of CEA for colorectal cancer (CRC) patients with peritoneal carcinomatosis. We aimed to determine the usefulness of CEA in ascites (aCEA) as a diagnostic parameter for CRC with peritoneal carcinomatosis. METHODS: Between January 2000 and May 2013, the medical records of 259 patients who underwent paracentesis for the evaluation of ascites were retrospectively reviewed. CRC patients with ascites (n=82) and patients with non-malignant ascites (n=177) were evaluated. Patients who had other malignancies, including gastric or ovarian cancer, with ascites were excluded. The optimal diagnostic cut-off value of aCEA for CRC with peritoneal carcinomatosis was determined using receiver operating characteristic curve analysis. The value of aCEA for predicting the occurrence of peritoneal carcinomatosis was evaluated using a logistic regression model. RESULTS: The optimal cut-off value of aCEA to diagnose CRC with peritoneal carcinomatosis was 3.89 ng/mL, and the area under the curve for aCEA was 0.996 (sensitivity 96.3%, specificity 100%, positive predictive value 100%, negative predictive value 98.3%). Multivariate logistic regression analysis showed that aCEA was an independent factor predicting the occurrence of peritoneal carcinomatosis. CONCLUSIONS: In this study, we showed that aCEA may be a useful parameter for diagnosing CRC with peritoneal carcinomatosis, and we propose an optimal aCEA cut-off value of 3.89 ng/mL. Further study that includes patients with other malignant ascites may be necessary to validate these findings.
