Minimally invasive vs. open segmental resection of the splenic flexure for cancer: a nationwide study of the Italian Society of Surgical Oncology-Colorectal Cancer Network (SICO-CNN)

Background: Evidence on the efficacy of minimally invasive (MI) segmental resection of splenic flexure cancer (SFC) is not available, mostly due to the rarity of this tumor. This study aimed to determine the survival outcomes of MI and open treatment, and to investigate whether MI is noninferior to open procedure regarding short-term outcomes. Methods: This nationwide retrospective cohort study included all consecutive SFC segmental resections performed in 30 referral centers between 2006 and 2016. The primary endpoint assessing efficacy was the overall survival (OS). The secondary endpoints included cancer-specific mortality (CSM), recurrence rate (RR), short-term clinical outcomes (a composite of Clavien-Dindo > 2 complications and 30-day mortality), and pathological outcomes (a composite of lymph nodes removed ≧12, and proximal and distal free resection margins length ≧ 5 cm). For these composites, a 6% noninferiority margin was chosen based on clinical relevance estimate. Results: A total of 606 patients underwent either an open (208, 34.3%) or a MI (398, 65.7%) SFC segmental resection. At univariable analysis, OS and CSM were improved in the MI group (log-rank test p = 0.004 and Gray's tests p = 0.004, respectively), while recurrences were comparable (Gray's tests p = 0.434). Cox multivariable analysis did not support that OS and CSM were better in the MI group (p = 0.109 and p = 0.163, respectively). Successful pathological outcome, observed in 53.2% of open and 58.3% of MI resections, supported noninferiority (difference 5.1%; 1-sided 95%CI - 4.7% to ∞). Successful short-term clinical outcome was documented in 93.3% of Open and 93.0% of MI procedures, and supported noninferiority as well (difference - 0.3%; 1-sided 95%CI - 5.0% to ∞). Conclusions: Among patients with SFC, the minimally invasive approach met the criterion for noninferiority for postoperative complications and pathological outcomes, and was found to provide results of OS, CSM, and RR comparable to those of open resection.