Stoma-free survival after anastomotic leak following rectal cancer resection: worldwide cohort of 2470 patients

Creative Commons License

Greijdanus N. G., Wienholts K., Ubels S., Talboom K., Hannink G., Wolthuis A., ...More

British Journal of Surgery, vol.110, no.12, pp.1863-1876, 2023 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 110 Issue: 12
  • Publication Date: 2023
  • Doi Number: 10.1093/bjs/znad311
  • Journal Name: British Journal of Surgery
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Academic Search Premier, Abstracts in Social Gerontology, CAB Abstracts, EMBASE, MEDLINE
  • Page Numbers: pp.1863-1876
  • Dokuz Eylül University Affiliated: Yes


Background: The optimal treatment of anastomotic leak after rectal cancer resection is unclear. This worldwide cohort study aimed to provide an overview of four treatment strategies applied. Methods: Patients from 216 centres and 45 countries with anastomotic leak after rectal cancer resection between 2014 and 2018 were included. Treatment was categorized as salvage surgery, faecal diversion with passive or active (vacuum) drainage, and no primary/secondary faecal diversion. The primary outcome was 1-year stoma-free survival. In addition, passive and active drainage were compared using propensity score matching (2: 1). Results: Of 2470 evaluable patients, 388 (16.0 per cent) underwent salvage surgery, 1524 (62.0 per cent) passive drainage, 278 (11.0 per cent) active drainage, and 280 (11.0 per cent) had no faecal diversion. One-year stoma-free survival rates were 13.7, 48.3, 48.2, and 65.4 per cent respectively. Propensity score matching resulted in 556 patients with passive and 278 with active drainage. There was no statistically significant difference between these groups in 1-year stoma-free survival (OR 0.95, 95 per cent c.i. 0.66 to 1.33), with a risk difference of -1.1 (95 per cent c.i. -9.0 to 7.0) per cent. After active drainage, more patients required secondary salvage surgery (OR 2.32, 1.49 to 3.59), prolonged hospital admission (an additional 6 (95 per cent c.i. 2 to 10) days), and ICU admission (OR 1.41, 1.02 to 1.94). Mean duration of leak healing did not differ significantly (an additional 12 (-28 to 52) days). Conclusion: Primary salvage surgery or omission of faecal diversion likely correspond to the most severe and least severe leaks respectively. In patients with diverted leaks, stoma-free survival did not differ statistically between passive and active drainage, although the increased risk of secondary salvage surgery and ICU admission suggests residual confounding.