CT characteristics for predicting prognosis of gastric cancer with synchronous peritoneal metastasis

Front Oncol. 2023 Jan 11:12:1061806. doi: 10.3389/fonc.2022.1061806. eCollection 2022.

Abstract

Introduction: To explore the CT characteristics for the prediction of long term survival in gastric cancer patients with synchronous peritoneal metastasis (PM).

Materials and methods: Sixty-six patients diagnosed as gastric cancer with synchronous peritoneum metastasis were enrolled in this retrospective study. Ten anatomic peritoneal regions were evaluated to check for the signs of PM on CT. One positive area equaled one score. The CT characteristic-based PM score (CT-PMS) was the sum of the total points assigned to all 10 regions, with a range of 0-10. The triple tract dilatation (TTD) sign caused by peritoneal metastasis, the presence of extensive lymph node metastasis (ELM), and the grade of ascites were recorded. The overall survival (OS) was used as the prognostic indicator. The performance of the CT characteristics was assessed by the Kaplan-Meier analysis and Cox proportional hazards model, while its reproducibility was evaluated by Kappa statistic and weighted Kappa statistic.

Results: Patients with a CT-PMS of 3-10 had significantly poorer OS (P = .02). Patients with either the presence of TTD sign, or ELM had a trend toward unfavorable OS (both P = .07), and when CT-PMS of 3-10 was detected simultaneously, the survival was further reduced (P = .00 for TTD sign; P = .01 for ELM). The grade of ascites failed to show a significant correlation with OS. The interobserver reproducibility for assessing the CT-PMS, the presence of TTD sign, the presence of ELM, and the grade of ascites had a substantial to almost perfect agreement.

Conclusion: The prognosis of gastric cancer patients with PM has a correlation with the extent of metastasis dissemination on baseline CT. A CT-PMS of 3-10 is associated with a worse prognosis than that of 0-2. The presence of TTD sign and ELM may help further select patients with extraordinarily poor prognoses.

Keywords: computed tomography; gastric neoplasm; palliative chemotherapy; peritoneal metastasis; survival analysis.

Grants and funding

This study was funded by Beijing Natural Science Foundation (No. Z180001; Z200015); National Natural Science Foundation of China (No. 91959205); Science Foundation of Peking University Cancer Hospital (JC202301); PKU-Baidu Foundation (No. 2020BD027).