Development and Validation of a Prognostic Nomogram for Colorectal Cancer Patients With Synchronous Peritoneal Metastasis

• Main Authors: Zifeng Yang, Yong Li, Xiusen Qin, Zejian Lv, Huaiming Wang, Deqing Wu, Zixu Yuan, Hui Wang
• Format: Article
• Language: English
• Published: Frontiers Media S.A. 2021-07-01
• Series: Frontiers in Oncology
• Subjects: colorectal cancer; nomogram; prognosis; peritoneal metastasis (PM); synchronous peritoneal metastasis
• Online Access: https://www.frontiersin.org/articles/10.3389/fonc.2021.615321/full

PurposeSynchronous peritoneal metastasis (S-PM) is considered a poor prognostic factor for colorectal cancer (CRC) and there is no nomogram to predict the survival of these patients. In this study, we aimed to use a multicenter data to identify the factors associated with S-PM of CRC to construct a nomogram for predicting the overall survival (OS) of these patients.MethodsCRC patients with S-PM from two medical centers were enrolled between September 2007 and June 2017. Multivariate analysis was used to identify independent factors associated with OS for the nomogram to predict the 1-, 2-, and 3-year OS rates in the development group. The concordance index (C-index), calibration plot, relative operating characteristic (ROC) curve with area under the curve (AUC) were calculated to evaluate the performance of the nomogram in both the development and an external validation group.Results277 CRC patients with S-PM in the development group and 68 patients in the validation group were eligible for this study. In multivariate analysis of development group, age, carbohydrate antigen 19-9 (CA19-9), carbohydrate antigen 125 (CA125), cytoreductive surgery (CRS), hyperthermic intraperitoneal chemotherapy (HIPEC), and chemotherapy were independent variables for OS, based on which the nomogram was built. The C-index of the nomogram in the development and validation group was 0.701 (95% Cl, 0.666–0.736) and 0.716 (95% Cl, 0.622–0.810); demonstrating good discriminative ability. The calibration plots showed satisfactory consistency between actual observation and nomogram-predicted OS probabilities in the development and external validation group. The nomogram showed good predictive accuracy for 1-, 2-, and 3-year OS rates in both groups with AUC >0.70. An online dynamic webserver was also developed for increasing the ease of the nomogram.ConclusionsWe developed and validated a predictive nomogram with good discriminative and high accuracy to predict the OS in CRC patients with S-PM.