Long-term outcomes of patients undergoing hepatectomy for bilateral multiple colorectal liver metastases-a propensity score matching analysis

Abstract

Objective: Liver is the most common site of distant metastasis in colorectal cancer patients. Currently, surgical resection of colorectal liver metastasis (CRLM) still remains the most curative therapeutic option which is associated with long-term survival. However, the outcome of CRLM patients with bilobar multiple lesions has been reported to be extremely poor due to the complex techniques of the surgery and the difficulties to achieve a negative resection margin. In this study, postoperative long-term outcome in patients with bilobar versus unilobar multiple CRLM undergoing surgical resection were compared and the prognostic factors of CRLM were analyzed. Methods: A retrospective cohort study was performed. The clinicopathological data were collected retrospectively from patients with multiple CRLM who received liver resection between January 2002 and November 2018 at our department. Inclusion criteria: (1) All CRLM lesions were confirmed by preoperative enhanced CT or MRI and enhanced ultrasonography. (2) All CRLM lesions were resectable either initially or converted by systemic treatments. The CRLM patients were considered as resectable, if their extrahepatic diseases were able to be completely removed. (3) Sufficient remnant liver volume was required to maintain normal liver function, which was defined by the ratio of remnant liver volume to total liver volume (RLV-TLV), of greater than 30% in general or 40% for the patients undergoing chemotherapy. (4) Medical records and follow-up information were intact. Those undergoing multiple operations after recurrence, with R2 resection, or with a single CRLM lesion were excluded. Patients were divided into bilobar and unilobar group according to tumor distribution. One-to-one propensity score matching (PSM) was performed to balance the covariates between the bilobar group and unilobar group. After PSM, the differences in long-term outcomes between the two groups were compared. Results: A total of 491 patients met the inclusion criteria, 344 (69.6%) with bilobar and 147 (30.4%) with unilobar CRLM. In the propensity-score-matched population (bilobar, 143; unilobar, 143), baseline characteristics were similar between the two groups. The 1-, 3-, and 5-year overall survival rates in the bilobar group were 91.6%, 52.1%, and 35.3% respectively, compared with 93.7%, 56.8%, and 43.8% in the unilobar group, and the difference was not statistically significant (P=0.204). The 1-, 3-, and 5-year recurrence-free survival rates in the bilobar group were 45.7%, 33.7%, and 33.7% respectively, compared with 62.5%, 44.1%, and 42.1% in the unilobar group, and the difference was not statistically significant (P=0.075). No significant difference was found in liver-only recurrence (45.6% in bilobar vs. 53.3% in unilobar, P=0.543). Univariate analysis showed that N stage of primary tumor, diameter of the largest liver metastases, carcinoembyonic antigen level, RAS gene status and clinical risk score (CRS) were significantly associated with the prognosis of CRLM (all P<0.05). Multivariate analysis indicated that diameter of largest liver metastases > 5 cm (HR=1.888, 95% CI: 1.251-2.848, P=0.002), CRS≥3 (HR=1.552,95% CI:1.050-2.294, P=0.027) and RAS gene mutation (HR=1.561, 95% CI: 1.102-2.212, P=0.012) were independent risk factors of poor overall survival after hepatectomy. Conclusions: Tumor distribution may not affect the prognosis of multiple CRLM after resection. Surgical removal in patients with bilobar multiple CRLM provides comparable long-term survival to unilobar multiple CRLM.