Meta analysis of diseased bowel resection versus diversion enterostomy in the treatment of late severe complications of chronic radiation-induced rectal injury / 中华胃肠外科杂志

ABSTRACT

Objective: To investigate the efficacy and safety of diseased bowel resection and diversion enterostomy in the treatment of late severe complications of chronic radiation-induced late rectal injury (RLRI). Methods: Studies about comparison of diseased bowel resection and diversion enterostomy in the treatment of late severe complications of chronic RLRI were screened and retrieved from databases, including PubMed, EMBASE, Scopus, Web of Science, Cochrane Library, CNKI, VIP, CBM and Wanfang. The following terms in Chinese were used to search [Title/Abstract] radiation-induced intestinal injury, radiation proctitis, surgery. The following English terms were used to search Radiation-induced intestinal injury, Bowel injury from radiation, Radiation proctitis, Surgery, Colostomy. Literature inclusion criteria (1) studies with control groups, published at home and abroad publicly, about the postoperative effects of diseased bowel resection vs. diversion enterostomy on RLRI patients with late severe complications; (2) the period of the study performed in the literatures must be clear; (3) patients at the preoperative diagnosis for RLRI with refractory bleeding, narrow, obstruction, perforation or fistula, etc.; (4) diseased bowel resection included Hartmann, Dixon, Bacon and Parks; diversion enterostomy included colostomy and ileostomy; (5) if the studies were published by the same institution or authors at the same time, the study with the biggest sample size was chosen; studies conducted in different time with different subjects were simultaneously included; (6) at least one prognostic indicator of the following parameters should be included the improvement of symptoms, postoperative complications, mortality, and reversed stomas rate. The stoma reduction rate was defined as the ratio of successful closure of colostomy after diseased bowel resection and diversion enterostomy. The method of direct calculation or the method of convert into direct calculation were used for stoma reduction rate. Exclusion criteria (1) a single-arm study without control group; (2) RLRI patients did not undergo diseased bowel resection or diversion enterostomy at the first time; (3) RLRI patients with distant metastasis; (4) the statistical method in the study was not appropriate; (5) the information was not complete, such as a lack of prognosis in the observational indexes. After screening literatures according to criteria, data retrieval and quality evaluation were carried out. Review Manager 5.3 software was used for Meta-analysis. Sensitivity analysis was used to exam the stability of results. Funnel diagram was used to analyze the bias of publication. Results: A total of 11 literatures were enrolled, including 426 RLRI patients with late severe complications, of whom 174 underwent diseased bowel resection (resection group) and 252 underwent diversion enterostomy (diversion group), respectively. Compared with diversion group, although resection group had a higher morbidity of complication (35.1% vs. 15.9%, OR=2.67, 95% CI 1.58 to 4.53, P<0.001), but it was more advantageous in symptom improvement (94.2% vs. 64.1%, OR=6.19, 95% CI 2.47 to 15.52, P<0.001) and stoma reductions (62.8% vs. 5.1%, OR=15.17, 95% CI 1.21 to 189.74, P=0.030), and the differences were significant (both P<0.05). No significant difference in postoperative mortality was found between the two groups (10.1% vs. 18.8%, OR=0.74, 95% CI 0.21 to 2.59, P=0.640). There were no obvious changes between the two groups after sensitivity analysis for the prognostic indicators (the symptoms improved, postoperative complications, mortality, and reversed stomas rate) compared with the meta-analysis results before exclusion, suggesting that the results were robust and credible. Funnel diagram analysis suggested a small published bias. Conclusions: Chronic RLRI patients with late severe complications undergoing diseased bowel resection have higher risk of complication, while their long-term mortality is comparable to those undergoing diversion enterostomy. Diseased bowel resection is better in postoperative improvement of symptoms and stoma reduction rate.