Predictive Factors for Anastomotic Leakage Following Colorectal Cancer Surgery: Where Are We and Where Are We Going?

“

TL;DR

Anastomotic leakage (AL) is a severe complication after colorectal cancer surgery, occurring in 2–19% of cases and up to 24% in distal rectal cancer, with higher reoperation rates, morbidity, and mortality. Key risk factors include male gender, obesity, smoking, low rectal anastomosis, multiple stapler firings, and lack of a protective stoma, while preoperative chemoradiotherapy shows conflicting evidence. Current prediction models and biomarkers (e.g., CRP, procalcitonin, ICG-FA) show promise but require further validation to improve clinical decision-making and reduce AL incidence.

Full summary

4,557 CHARS

**Background:** Colorectal cancer (CRC) is the third most commonly diagnosed malignancy worldwide. Anastomotic leakage (AL) is one of the most severe postoperative complications following CRC surgery, with reported incidence between 2% and 19% overall, and up to 24% in patients undergoing surgery for distal rectal cancer. AL is associated with higher reoperation rates, increased morbidity and mortality, higher local recurrence rates, and decreased long-term survival. The International Study Group of Rectal Cancer (ISREC) defined AL as a defect of continuity at the surgical anastomosis creating communication between intra-luminal and extra-luminal compartments, with three severity grades. This narrative review aims to identify risk factors for AL, present scoring systems for prediction, and highlight future directions.

**Methods:** This is a narrative review of the current literature on risk factors for AL following colorectal cancer surgery. The authors classified risk factors into preoperative, intraoperative, and perioperative categories. They also reviewed scoring systems, biomarkers, and emerging technologies for AL prediction.

**Key Results:** - **Preoperative risk factors:** Male gender is a significant risk factor (odds ratio [OR] 3.8; 95% CI 1.9–7.7; p < 0.001). Age >70 years was associated with AL in one study (p = 0.009), but most studies found no correlation. Heavy smoking (>40 pack-years) and excessive alcohol (>5 drinks/day) increase AL risk. ASA score >II and high Charlson comorbidity index increase risk. BMI >30 kg/m² is an independent risk factor. Low preoperative serum albumin (<3.5 gr/dL) is associated with increased AL risk. Tumor location in the rectum versus colon was independently predictive (OR 18.20; 95% CI 2.33–142.02; p = 0.005). Tumors ≥3 cm and advanced TNM stage increase AL risk. Preoperative chemoradiotherapy (PCRT) shows conflicting results: some studies found increased risk (HR 6.284; 95% CI 2.829–13.961; p < 0.001), while meta-analyses found no significant effect (OR 1.02; 95% CI 0.80–1.30; p = 0.88). Mechanical bowel preparation alone does not reduce AL, but combined with oral antibiotics may reduce complications. - **Intraoperative risk factors:** Low anastomoses (≤5 cm from anal verge) have significantly higher AL risk (OR 6.855; 95% CI 1.271–36.964; p = 0.025). Tumors ≤7 cm from anal verge are associated with higher AL rates (OR 3.445; 95% CI 1.631–7.279; p = 0.001). Protective diverting stoma significantly reduces AL and reoperation rates (pooled RR 0.47; 95% CI 0.33–0.68; p < 0.0001 for AL; RR 0.36; 95% CI 0.28–0.46; p < 0.00001 for reoperation). Multiple stapler firings (≥2) increase AL risk (OR 2.44; 95% CI 1.34–4.42; p = 0.003). Indocyanine green-fluorescence angiography (ICG-FA) was associated with lower AL rates (OR 0.4; 95% CI 0.2–0.8; p = 0.02), though a multicenter RCT found no significant impact. Left colic artery preservation was associated with reduced AL (p < 0.001 on multivariate analysis). Prolonged operative time (>220 min) was independently associated with AL. Pelvic drainage showed mixed results; one study found lack of drainage was an independent risk factor (OR 3.814; 95% CI 1.207–12.644; p = 0.0225). Transanal tube absence was an independent risk factor (OR 33.5; 95% CI 1.82–618; p = 0.018). - **Perioperative risk factors:** Corticosteroid use, NSAIDs, and immunosuppressive drugs increase AL risk. Perioperative blood transfusion was independently associated with AL (OR 2.83; 95% CI 1.59–5.06; p < 0.0001). Gut microbiome alterations, particularly Enterococcus faecalis and Lachnospiraceae abundance, may play a role in AL pathogenesis. - **Scoring systems and biomarkers:** The modified DULK score has 97% sensitivity, 57% specificity, and 99.5% negative predictive value. CRP levels <180 mg/L on postoperative day 4 can help exclude AL. Procalcitonin improves predictive accuracy when combined with CRP. AI-based models show promise (area under the curve up to 0.87).

**Clinical Implications:** Understanding preoperative, intraoperative, and perioperative risk factors can guide surgeons in intraoperative decision-making, particularly regarding the selective use of protective stomas in high-risk patients. ICG-FA and other perfusion assessment techniques may help reduce AL risk. Scoring systems and biomarkers (especially CRP and procalcitonin) can aid in early detection and exclusion of AL, potentially reducing morbidity and mortality. Further research is needed to validate prediction models and identify novel biomarkers.