Umeå universitets logga

Surgery is still the most common treatment for rectal cancer, being the most effective and cost-efficient modality. However, it is not without risk, nor without controversies. This dissertation is an evaluation of the pros and cons of high versus low ligation, whether anastomotic leakage is still prevalent after surgery and associated with increased mortality, and if the risk of leakage could be predicted by early postoperative pain.

Study I relied upon case records and registry data to evaluate the causal effect of high ligation on the risk of anastomotic leakage after anterior resection in 722 patients with increased cardiovascular risk. When controlling for confounders, no association was found overall. However, an increased risk for leakage after high ligation was noted for the few patients who suffered from both manifest cardiovascular disease and ASA III–IV (OR 3.66, 95% CI 1.04–12.85) and when performed in a low volume hospital (OR 3.89, 95% 1.58– 9.59). Study II estimated the risk of anastomotic leakage and death after anterior resection today. Among the 6,948 patients, 10.0% suffered from leakage, in whom mortality was 3.9% versus 1.5% for patients without a leak. However, this increased mortality was driven entirely by patients in need of reintervention, who exhibited a 7.5% 90-day mortality, resulting in a significantly increased risk (OR 5.57, 95% CI 3.29–9.44), when controlling for confounders, while conservatively treated leakage was not associated with mortality. Age acted as an effect modifier, as postoperative mortality after leakage was increased in the elderly.

Study III returned to high versus low ligation as exposure, to evaluate the long-term oncological benefits of either ligation level, with cancer-specific survival as primary outcome. The final cohort of 8,287 patients who underwent abdominal surgery for rectal cancer, with curative intent, was followed for a minimum of 3.5 years. After propensity score matching, no significant differences were found between high and low ligation for any survival or recurrence analysis, nor in the unmatched cohort, when controlling for confounders. A statistically significant difference was found for lymph node harvest, which was slightly greater in high ligation (17.7 vs 16.7 lymph nodes). Finally, study IV estimated the independent predictive ability of postoperative pain, measured on the numerical rating scale (NRS), on the risk for anastomotic leakage after colorectal cancer surgery. It seems as if increased early pain is an independent predictor for leakage (OR 1.73, 95% CI 1.22– 2.46 for NRS 4–10), with increasing risk of leakage with increasing pain (OR 2.42 for NRS 8–10). In addition, increased pain was more strongly associated with more severe leakage.

In summary, the level of ligation seems to be of importance only in a select group of high-risk patients, but offers no obvious oncological advantages. The high incidence and serious sequelae of anastomotic leakage makes it one of the most important clinical challenges in colorectal surgery, with especially detrimental effects in the elderly. A better understanding of the causal pathways behind leakage, and the overall harm and benefit of ligation level and diverting stomas, might allow a better selection of treatment for future patients.