Complete mesocolic excision an assessment of feasibility and outcome.

Surgery is the most important factor for radical treatment of colon cancer, and the long-term prognosis can be improved by improving the surgical treatment without increased risk of perioperative mortality. Complete mesocolic excision (CME), in which more extensive lymph node (LN) dissection is performed, has been shown in single-centre studies with historical controls to be associated with better oncological outcome. However, better evidence is needed. The main purpose of this PhD thesis was to investigate whether CME could be implemented in a colorectal surgical department in Denmark, whether more extensive dissection could demonstrate LN metastases outside the mesocolon, and to demonstrate a possible association between CME and improved oncological results without increased risk of perioperative mortality. This thesis includes five articles. Two articles (IV and V) are based on the population of patients undergoing elective resection for colon cancer in the Capital Region from June 2008 to December 2013. Two articles (II and III) are based on data from the local colon database in Hillerød, and the last article (I) is a systematic review concerning the risk of metastases from colon cancer to the central LNs in the mesocolon. Article I found a risk of metastases in central LNs to be reported in 1-22% of the cases of right-sided colon cancers, and in up to 12% of the cases with sigmoid tumours. The populations included and methods used in the studies were very heterogeneous and no definitive conclusions can be drawn. It was shown in article II that the surgical quality, i.e. quality of the specimens assessed by the pathologists, improved with implementation of CME in Hillerød. The vascular tie was higher, and the implementation was not associated with an increased risk of perioperative mortality. Article III demonstrated a risk of LN metastases in the gastrocolic ligament along the stomach for tumours located in the transverse colon, in the ascending or descending colon close to or in the flexures. It occurred in 4% of all patients and 13% of the patients with LN metastases in mesocolon. Resection of these LNs seems advisable for these tumour locations. Article IV showed no association between increased perioperative mortality and CME (n = 529) when compared with non-CME (n = 1,701). The 30-day mortality was 4.2% after CME compared with 3.7% after non-CME (p = 0.605), and the 90-day mortalities were 6.2% and 4.9% (p = 0.219) respectively. Odds ratios for 30-day and 90-day mortalities after CME were respectively 1.07 (95% confidence interval: 0.62-1.80) and 1.25 (0.77-1.94) in the multi-variable logistic regression analyses. Postoperative respiratory failure and need for vasopressors were significantly more frequent in the CME group and, besides CME itself, could be associated with the fewer laparoscopic resections and more severe preoperative comorbidity in the CME Group. Article V demonstrated an association between higher four-year disease-free survival for stage I-III tumours and CME (n = 364) when compared with non-CME (n = 1,031). Most notable was the difference for stage I and II cancers. The four-year disease-free survival for stage I was 100% in the CME group compared with 89.8% (83.1-96.6) in the non-CME group (p = 0.046). For stage II the disease-free survivals were 91.9% (87.2-96.6%) in the CME group and 77.9% (71.6-84.1%) in the non-CME group (p = 0.0033), and for stage III 73.5% (63.6-83.5) and 67.5% (61.8-73.2) (p = 0.13) respectively. In the multivariable Cox regression models, CME was a significant predictive factor for higher dis-ease-free four-year survival for stage I-III patients with hazard ratios (HR) for CME of 0.59 (0.42-0.83, p = 0.0025). For stage II the HR was 0.44 (0.23-0.86, p = 0.018) and for stage III 0.64 (0.42-1.00, p = 0.048).