Laparoscopic versus open liver resection for colorectal liver metastasis: An individual participant data meta-analysis of randomized controlled trials.

Abstract

3560 Background: Laparoscopic surgery for colorectal liver metastases (CRLM) is associated with lower physical impact, shorter length of stay and less postoperative morbidity than open surgery. To analyze oncological and procedural outcomes, a European consortium including principal investigators of all 4 completed RCTs on laparoscopic (LLR) vs open (OLR) liver resection performed an individual participant data meta-analysis (IPDMA), including updated survival data. Methods: This was an IPDMA with a primary endpoint of postoperative morbidity. Secondary endpoints included overall survival (OS), disease-free survival (DFS) and resection margin status. A generalized linear mixed model was used to compare OLR and LLR , with trial as a fixed-effect. Logistic regression analysis was performed for dichotomous variables and negative binominal regression analysis for continuous variables. Survival analyses were performed using Cox-regression. Results: A total of 761 patients with CRLM were randomly allocated to LLR (n = 384) or OLR (n = 377). Preoperative chemotherapy was administered as recommended by the local multidisciplinary team (40% vs 46%). Whilst LLR was associated with significantly less postoperative morbidity overall (19% vs 27%, adjusted OR 0.62 [95%CI 0.44 to 0.88]), this was not observed in the subgroup of patients receiving neoadjuvant chemotherapy (25% vs 24%, adjusted OR 1.06 [95%CI 0.63 to 1.79]), p-value for interaction < 0.001). Hospital stay was shorter after LLR (median 4 vs 5 days, adjusted percentage difference: -27 [95%CI -37 to -15]) All cause 90-day mortality was not significantly different (1.9% vs 0.8%, adjusted OR: 3.2 [95%CI 0.65 to 16.00]). At 5-year follow-up OS and DFS were not significantly different (adjusted HR 0.97 [95%CI 0.79 to 1.20] and 1.05 [95%CI 0.86 to 1.27], respectively). In patients who received preoperative chemotherapy, a trend towards fewer R0 resections in LLR compared to OLR was noted (84% vs 90%, adjusted OR 2.00 [95%CI 0.99 to 4.03], p-value for interaction = 0.053). In patients who did not receive preoperative chemotherapy there was no difference in R0 resections (90% vs 86%, adjusted OR 0.76 [95%CI 0.41 to 1.40]). Liver specific recurrence was not different between LLR and OLR (37% vs 37%, adjusted OR 1.02 [95%CI 0.76 to 1.34), neither was time to adjuvant chemotherapy (45 days vs 49 days, adjusted OR 1.12 [95%CI 0.91 to 1.38]) nor median number of adjuvant courses (8 vs 8, adjusted percentage difference 0.07 [95%CI -0.11 to 0.24]). Conclusions: This IPDMA of 761 patients in 4 RCTs across Europe confirms that laparoscopic resection for CRLM is superior to open resection with regards to short term outcomes, with no differences observed on long term oncological outcomes. However, in patients who received preoperative chemotherapy, the benefit of the laparoscopic approach is questionable.