Effects of anastomotic technique on early postoperative outcome in open right‐sided hemicolectomy

Background Despite recent improvements in colonic cancer surgery, the rate of anastomotic leakage after right hemicolectomy is still around 6–7 per cent. This study examined whether anastomotic technique (handsewn or stapled) after open right hemicolectomy for right‐sided colonic cancer influences postoperative complications. Methods Patient data from the German Society for General and Visceral Surgery (StuDoQ) registry from 2010 to 2017 were analysed. Univariable and multivariable analyses were performed. The primary endpoint was anastomotic leakage; secondary endpoints were postoperative ileus, complications and length of postoperative hospital stay (LOS). Results A total of 4062 patients who had undergone open right hemicolectomy for colonic cancer were analysed. All patients had an ileocolic anastomosis, 2742 handsewn and 1320 stapled. Baseline characteristics were similar. No significant differences were identified in anastomotic leakage, postoperative ileus, reoperation rate, surgical‐site infection, LOS or death. The stapled group had a significantly shorter duration of surgery and fewer Clavien–Dindo grade I–II complications. In multivariable logistic regression analysis, ASA grade and BMI were found to be significantly associated with postoperative complications such as anastomotic leakage, postoperative ileus and reoperation rate. Conclusion Handsewn and stapled ileocolic anastomoses for open right‐sided colonic cancer resections are equally safe. Stapler use was associated with reduced duration of surgery and significantly fewer minor complications.


Introduction
Colorectal cancer affects more than one million patients per year worldwide, accounting for more than 500 000 deaths annually 1 . Over the past two decades improvements in adjuvant chemotherapy and surgical quality have led to better long-term survival 2 . Complication rates remain high, however, with 30-day mortality rates following colonic cancer surgery of up to 10 per cent 3 . In particular, the rate of anastomotic leakage after right hemicolectomy is surprisingly high (6⋅4-7⋅5 per cent) 4,5 compared with that for left hemicolectomy (anastomotic leak rate 1⋅9-6⋅5 per cent) 5,6 .
Anastomosis techniques after right hemicolectomy vary widely in clinical practice. Ileocolic anastomoses can be end-to-end, end-to-side, side-to-end or side-to-side 7 . They can be handsewn in one or more layers, using interrupted or continuous sutures in a variety of sizes, needle configurations and materials, or stapled using linear or circular proprietary devices. In 2011, a Cochrane systematic review 8 including 1125 patients with ileocolic anastomosis found a significant advantage for stapled anastomosis with respect to anastomotic leak rate, although studies published subsequently have conversely identified stapled anastomosis as an independent risk factor for anastomotic leakage 9 .
Patients with colonic cancer registered in StuDoQ at 17 August 2017 n = 16 151 Patients excluded n = 12 089 Missing consent or lack of essential data n = 2404 Cancer location other than right-sided n = 7294 Emergency or non-surgical treatment n = 1158 Concurrent liver resection n = 165 Creation of anastomosis ostomy n = 55 Liver resection plus ostomy n = 5 Endoluminal tumour resection n = 2 Laparoscopic resection n = 1006 Patients included in study n = 4062 The technical requirements for surgical resection of right-sided cancers have changed greatly in recent years with the introduction of complete mesocolic excision (CME) 10 , but the optimal anastomotic technique remains an unresolved issue.
Data for colonic cancer were retrieved from the Study, Documentation and Quality Centre (StuDoQ| ColonCancer) registry of the German Society for General and Visceral Surgery (DGAV) to investigate whether the anastomosis techniques influence early postoperative complications.

Methods
Informed consent and data safety procedures were approved by the Society for Technology, Methods, and Infrastructure for Networked Medical Research (www. tmf-ev.de), and publication guidelines were established by the DGAV (www.dgav.de/studoq/datenschutzkonzeptund-publikationsrichtlinien.html).
The StuDoQ|ColonCancer registry is a voluntary prospectively created database for colonic cancer surgery established by the DGAV in January 2010 (www.dgav.de/ studoq; www.en.studoq.de), designed to facilitate assessment of quality and risk factors in colonic cancer surgery in Germany. Data from participating centres are entered in pseudonymized form using a browser-based tool and subjected to automatic plausibility and cross-checking controls. Hospitals included in the study data are listed in Table S1 (supporting information).
For this study, all patients with right-sided or extended right hemicolectomy were identified from the registry and relevant demographic data, co-morbidities, and information on operations, histology and perioperative course were extracted in anonymized form for analysis. Patients undergoing emergency surgery, non-right-sided resection, laparoscopic right-sided resection, endoluminal resection, simultaneous liver metastasis resection or creation of any kind of ostomy were excluded. CME should have been  performed according to the description by Hohenberger and colleagues 11 . Extended hemicolectomy was defined as any right-sided colonic resection, including ligation of the middle colic artery and vein. Anastomotic leakage requiring intervention 12,13 , surgical-site infection (SSI) necessitating reopening of the wound 14 , Clavien-Dindo complication grade 15 , burst abdomen, reoperation and in-hospital mortality were evaluated, along with any need for unplanned postoperative ventilation for more than 48 h, pneumonia, length of postoperative hospital stay (LOS) and readmission. Overall postoperative morbidity was summarized according to the Clavien-Dindo classification: grade 0, none; grade I-II, minor; grade IIIa-IV, major; grade V, death.
Patients were grouped according to the type of anastomosis (handsewn or stapled using any type of MTL30 is a new, validated, endpoint parameter specific for the German health sector; it combines mortality, transfer to a higher-level hospital owing to complications, and length of stay beyond 30 days after surgery.

Statistical analysis
Statistical analyses are two-sided, with a significance level of 0⋅050. Continuous variables are expressed as mean(s.d.) values, and categorical parameters as absolute frequency and percentage. Univariable analysis was performed using χ 2 and Mann-Whitney U tests, as appropriate. Multivariable analysis was by Cox regression. All variables with P < 0⋅100 in univariable analysis were included in the multivariable analysis.

Discussion
The optimal anastomosis technique after right hemicolectomy is still a matter for debate 8 . Most trials 17 -20 are too small to provide definitive conclusions regarding the various techniques, or focus on Crohn's disease where the patient's postoperative course may be influenced by the underlying disease.
A large multicentre study 21 analysing 999 patients who underwent ileocolic anastomosis between 2002 and 2007, mainly for colonic cancer (95⋅8 per cent) found that patients with a handsewn anastomosis had a significantly higher leak rate (4⋅9 per cent) than patients with a stapled anastomosis (2⋅5 per cent). A meta-analysis 8 also found a lower anastomotic leak rate for stapled anastomoses, but no differences in other outcomes. Conversely, stapled anastomosis was identified as an independent risk factor for anastomotic leakage in a study from Denmark 5 . In a recent multinational snapshot audit 22 of anastomosis following right-sided colonic resection for colonic cancer or inflammatory bowel disease, the use of staplers was identified as an independent risk factor for anastomotic leakage.
The present study offers the first registry-based analysis of the impact of anastomosis technique (stapled versus handsewn) following oncological right hemicolectomy. The two techniques appear equally safe in terms of anastomotic leakage, reoperation rate, postoperative ileus, SSI and other surgical complications.
The primary endpoint of anastomotic leak rate in this study of 3⋅6 per cent was below the 6⋅4-7⋅5 per cent reported by Bakker and colleagues 5 and Krarup and co-workers 4 . This lower rate may be attributable to the fact that patients were registered at specialized cancer centres, with some bias for this population. The German Cancer Society certificates specialized cancer centres on the basis of several parameters, of which anastomotic leakage is one. Hospitals can use their own submitted data set to StuDoQ for quality control and certification, thereby perhaps creating a documentation bias.
The 13-day LOS reported in this study is considerably longer than that of studies on colectomy from the American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) registry 23 . This may be due to less economic pressure to reduce LOS, slower implementation of enhanced recovery regimens in clinical practice, and the higher rate of CME. The rate of postoperative ileus in the present study was surprisingly low compared with that of the ACS NSQIP registry 23,24 (3⋅9 versus 12⋅7 per cent respectively). This difference could also be a documentation bias, as postoperative ileus has not been defined precisely. Additionally, the rate of CME was higher in the handsewn group, which may account for the higher rate of minor (grade I-II) complications, as it has been shown 10 that CME resection is associated with more, and more severe, complications than non-CME resection.
Limitations of the present study include the lack of cost analysis and information missing in the StuDoQ registry, such as end-to-end or side-to-side connections and details of handsewn techniques and materials used. Although there was no difference in the short term, it is unclear whether the two methods differ in the long term with regard to stenosis rate, oncological outcome or quality of life.
Stapled and handsewn techniques for creating ileocolic anastomoses after open oncological right hemicolectomy seem equally safe and effective in the short term. Stapling was associated with reduced duration of surgery and fewer minor (grade I-II) complications.