Impact of primary tumor resection on the survival of patients with unresectable colon cancer liver metastasis at different colonic subsites: a propensity score matching analysis

Abstract

Objective

To investigate the effect of primary tumor resection (PTR) on the prognosis of patients with unresectable colon cancer liver metastasis (UCCLM) at seven colonic subsites using the Surveillance, Epidemiology, and End Results (SEER) database.

Methods

Propensity score matching (PSM) was performed to balance selection bias using all available variables that could be of potential relevance. After matching, the groups were redefined in a 1:1 ratio using the nearest method. Cancer-specific survival (CSS) was compared among the patients of PTR and non-PTR groups. Cox regression models were used to identify the prognostic factors for CSS.

Results

CSS was significantly different between all groups. Cox regression analysis showed that PTR was an independent prognostic factor for all groups. After PSM, PTR significantly prolonged CSS for all groups. Subgroup analysis showed that PTR did not improve the prognosis of N2 stage patients in the cecum, ascending colon, and descending colon groups; T1 + T2 stage patients in the hepatic flexure group; and patients with a tumor size ≤5 cm in the splenic flexure group. Segmental colectomy could prolong CSS of patients in the cecum, ascending colon, transverse colon, splenic flexure, and sigmoid colon groups, while extended colectomy could prolong CSS of patients in the hepatic flexure and descending colon groups.

Conclusion

At different colonic subsites, UCCLM patients had different CSS. PTR could improve their prognosis, however, N stage, T stage, and tumor size are important reference indicators. In addition to patients in the hepatic flexure and descending colon groups, we suggested that patients in other groups should choose segmental colectomy.

Keywords:

• Colon cancer liver metastasis
• primary tumor resection
• colonic subsites
• cancer-specific survival
• SEER database

Acknowledgments

The authors acknowledge the SEER database from which clinical data was collected. The authors also thank Dr. Liangliang Cai of the Guangxi University of Chinese Medicine for guidance in statistical methods and application.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The study was supported by the National Natural Science Foundation of China [81560389 and 81960436], Project of the Jiangxi Provincial Department of Science and Technology [20202BBGL73037], and Double Thousand Talents Project of Jiangxi Province [jxsq2019201100].