Optimal surgical sequence for colorectal cancer liver metastases patients receiving colorectal cancer resection with simultaneous liver metastasis resection: A multicentre retrospective propensity score matching study

Background: There is little evidence regarding the optimal surgical sequence for colorectal cancer liver metastasis (CRLM) patients undergoing colorectal resection with simultaneous liver metastasis resection. Methods: CRLM patients from five centers were retrospectively evaluated. The short-term outcomes included intraoperative and postoperative outcomes. Postoperative complications were measured according to the Clavien-Dindo classification. Grade I to II complications were defined as minor postoperative complications. The long-term outcomes were progression-free survival (PFS) and overall survival (OS). Propensity score matching (PSM) and inverse probability of treatment weighting (IPTW) were performed to overcome the selection bias between colorectal resection first and liver resection first. Results: A total of 1255 CRLM patients were included. In the multivariable logistic regression analysis, a body mass index (BMI) < 24 kg/m2, primary site in the left hemicolon, non-bilobar distribution of liver metastases and no preoperative chemotherapy were significantly associated with the likelihood of colorectal resection first. After 1:1 PSM, there was no significant difference between the colorectal resection first group and the liver resection first group. Compared with patients with colorectal resection first, patients with liver resection first had a comparable postoperative infection rate (15.0% vs. 16.0%, P = 0.735), a longer operation time (305.0 [231.3-416.0] vs. 300.0 [225.0-374.0], P = 0.033), more intraoperative blood loss (200.0 [150.0-400.0] vs. 100.0 [100.0-300.0], P < 0.001), a higher postoperative minor complication rate (28.7% vs. 20.7%, P = 0.023) and a higher postoperative ICU rate (14.7% vs. 8.7%, P = 0.022). IPTW-adjusted Kaplan-Meier analysis showed that patients who underwent colorectal resection first had a similar PFS (P = 0.702, median: 20.6 months vs. 16.6 months) and unfavourable OS (P = 0.014, median: 48.5 months vs. 67.0 months) compared with patients who underwent liver resection first. In the IPTW-adjusted Cox proportional hazards regression analysis, colorectal resection first was an unfavourable risk factor for OS (hazard ratio [HR] = 1.301, 95% CI 1.048-1.616, P = 0.017) and was not an independent predictor for PFS (HR = 0.986, 95% CI 0.831-1.170, P = 0.874). IPTWadjusted Cox proportional hazards regression analysis, including postoperative complications, operation time, intraoperative blood loss and postoperative chemotherapy, produced consistent results. Conclusion: Although violating the "sterility principle", patients who underwent colorectal resection first did not have an increased postoperative infection rate and had some better short-term outcomes and comparable PFS than those who underwent liver resection first.