Circular staplers and anastomotic leakage in colorectal surgery: meta-analysis.

Colorectal surgery has a morbidity rate ranging between 17 and 35% 1 , and anastomotic leakage (AL) represents the most frightening adverse event with an incidence rate currently ranging between 5 and 29% 2 . AL occurrence is widely known to be associated with potentially detrimental consequences, not only on the postoperative clinical course of patients, with a reported associated mortality rate of 6.8% 3–5 , but also on their long-term oncological and functional outcomes 6 , 7 . Indeed, AL onset relates to an increased risk of cancer recurrence, significantly reducing the long-term survival of patients 8 , 9 . Furthermore, from an economic point of view, AL represents an overspending for the healthcare systems with an estimated cost of approximately 20,000 Euros per patient in the USA 10 . In this context, recent decades of research specifically focused on novel surgical techniques and devices aimed at improving surgical outcomes, particularly in terms of AL rate after colorectal surgery. The introduction of stapling devices in the 1970s has played a key role in this field, progressively becoming an indispensable tool for both colorectal transection and anastomosis creation, especially in more challenging procedures such as left colorectal surgeries 11 . In this last regard, circular staplers have gained wide popularity, allowing the creation of safer anastomoses (especially in the low rectum and narrow pelvis) and increasing, at the same time, the rate of sphincter preservation procedures 12 . However, despite these clear advantages, some limitations of the circular staplers still need to be addressed. For instance, literature exposes concerns on the potential role of cross-staple lines and the so-called ‘dog-ears’ creation on AL onset after left colorectal resection 13 , 14 , and there is continuing debate on the optimal circular stapler to employ to improve postoperative surgical outcomes. This has resulted in several publications comparing the different types of devices and techniques in order to propose the most valuable methods for lowering the AL rate. Specifically, several authors focused their research on the optimal circular stapler diameter to use 15–19 , on the efficacy of colorectal anastomoses using the two- or three-row technology 20–24 , on the type of stapler (manual or powered) for fashioning the anastomosis 25–30 as well as on the comparison among different surgical techniques such as the double- (DST) and single-stapling techniques (SST) 14 , 31–34 . However, no conclusive data are currently available, and the circular stapler to employ as well as the best technique to use is still at the discretion of the operating surgeon. Based on this background, this systematic review and meta-analysis aimed to evaluate the current evidence on the incidence rate of AL in colorectal surgery according to the type of circular stapler utilized, with particular focus on the stapler diameter, number of stapler rows (two versus three rows), type of technology (powered versus manual) and type of anastomotic technique (SST versus DST), in order to give an objective and summarized overview on the current evidence presented in the literature.

The systematic review was prospectively registered in the prospective register of systemic reviews (PROSPERO), registration number: CRD42024519036. The literature search was performed according to the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) 35 . The systematic search was conducted on PubMed, Scopus and Web of Science databases from January 2012 to November 2023, addressing AL in relation to the different types of circular staplers. The search terms ‘colorectal surgery’, ‘staplers’, ‘complications’ were adapted for each database. The final search was ‘(‘colorectal surgery’ (Medical Subject Headings (MeSH) Terms) OR (‘colorectal’ (All Fields) AND ‘surgery’ (All Fields)) OR ‘colorectal surgery’ (All Fields)) AND (‘stapler’ (All Fields) OR ‘staplers’ (All Fields) OR ‘staplers’ (All Fields)) AND (‘complicances’ (All Fields) OR ‘complicate’ (All Fields) OR ‘complicated’ (All Fields) OR ‘complicates’ (All Fields) OR ‘complicating’ (All Fields) OR ‘complication’ (All Fields) OR ‘complications’ (All Fields) OR ‘complications’ (MeSH Subheading) OR ‘complications’ (All Fields))’ ( Supplementary materials , Table S1 ). Inclusion criteria were retrospective, cohort, prospective and randomized clinical trials (RCTs) studies reporting the AL rate evaluation after left colorectal resection according to the type of circular stapler employed, published in the English language and including only adult patients aged 18 years and over. Non-English articles, review articles, editorials, opinion statements, animal studies, case reports and studies on colorectal surgery for extra-colonic diseases (that is: endometriosis, gynaecologic tumours) or right/transverse colon diseases were considered as exclusion criteria. The articles were uploaded into Systematic Review Accelerator ( www.sr-accelerator.com ), which is a web-based screening tool. Abstracts and titles were screened independently on Systematic Review Accelerator by two reviewers. Two independent authors performed the data extraction. Discrepancies in the screening and data extraction were discussed and any enduring disagreement was resolved with the input of a senior author. Study characteristics, including first author, article title, year of publication, study type, number of patients, level of evidence, and data related to the defined outcome were extracted and reported on an electronic database. The outcome of interest was the analysis of each staplers’ characteristics, including: diameter, number of rows, technology (manual versus powered) and anastomotic technique (SST versus DST) on AL occurrence. The Newcastle–Ottawa classification 36 was independently used by two authors for the quality assessment. Grading was based on a scale of zero to nine according to the following domains: S1: representativeness of the exposed cohort; S2: selection of the non-exposed cohort; S3: ascertainment of exposure; S4: demonstration that outcome of interest was not present at start of the study; C: comparability; O1: assessment of outcome; O2: sufficient length of follow-up for outcomes to occur; O3: adequacy of follow-up. The maximum attainable score was nine, representing the highest study quality. The meta-analysis was conducted using Jamovi software (v. 2.4.11.0). The analysis was carried out using the log odds ratio (ORs) as the outcome measure. The ORs and 95% confidence intervals (95% c.i.) were calculated to estimate the association between binary factors and AL. A fixed-effects model was fitted to the data. Furthermore, the Q test for heterogeneity (Cochran 1954) and the I 2 statistic were reported. Studentized residuals and Cook’s distances were used to examine whether studies may be outliers and/or influential in the context of the model. Studies with a studentized residual larger than the 100×(1–0.05/(2×k))th percentile of a standard normal distribution were considered potential outliers (that is using a Bonferroni correction with two-sided a = 0.05 for k studies included in the meta-analysis). Studies with a Cook’s distance larger than the median plus six times the interquartile range of the Cook’s distances were considered to be influential. Finally, the rank correlation test and the regression test, using the standard error of the observed outcomes as a predictor, were used to check for funnel plot asymmetry. A sensitivity analysis (excluding one study at a time) was conducted to test the stability of the pooled results.

The systematic search identified 2582 eligible studies (207 from PubMed, 2178 from Scopus and 197 from Web of Science). A total of 2553 studies were excluded for duplication or no eligibility after screening of the titles and abstracts. Of the 28 studies eligible for full-text evaluation, 21 14–34 were included in the meta-analysis. Figure 1 details the study flow chart. The quality assessment of the studies is reported in Table 1 . AL was defined according to the International Study Group of Rectal Cancer in 11 of 21 16 , 17 , 20–23 , 26 , 28 , 30 , 31 , 34 studies, while clinical and/or radiological detection was used for AL assessment in the remaining studies 14 , 15 , 18 , 19 , 24 , 25 , 27 , 29 , 32 , 33 . PRISMA flow chart Quality assessment of the included studies according to the Newcastle–Ottawa scale *one score; **two score. Overall, 24 511 patients from the 21 selected studies were included 14–34 . Table 2 reports the characteristics of the studies, including the study design, number of patients, type of anastomosis, the overall rate of AL and, when indicated, the type of stapler used. Characteristics of the studies included in the meta-analysis * P value was not reported when a single disease location/anastomosis distance from the anal verge was evaluated by the authors. †Refers to the level of the anastomosis. ‡Although there was a statistically significant difference between groups in terms of anastomosis distance from the anal verge, the type of anastomosis was not recognized as an independent risk feature for AL in the multivariable analysis. R, retrospective; AL, anastomotic leakage; DST, double-stapling technique; SST, single-stapling technique; PCS, powered circular stapler; NR, not reported; ISREC, International Study Group of Rectal Cancer; TCS, triple-row circular stapler; DCS, double-row circular stapler; i.q.r., interquartile rank; ns, not significant. Of the 21 studies, five 15–19 focused on the stapler diameter (28–29 mm versus 31 mm or 25 mm versus 28–29 mm), five 20–24 on the number of stapler rows (two rows versus three rows), six 25–30 on the comparison between manual and powered circular staplers, and five 14 , 31–34 on the comparison between the SST and DST. Five studies 15–19 were identified on the correlation between the circular stapler diameter and AL rate involving a total of 1773 patients. Two articles 17 , 19 compared the AL incidence rate between the 25 mm circular staplers and the 28–29 mm (1027 patients), while three studies 15 , 16 , 18 compared the use of the 28–29 mm circular staplers to the 31–33 mm circular staplers (746 patients). No difference was observed in terms of AL rate between the 25 mm group (1.71%; 5 of 292) and the 28–29 mm cohort (3.81%; 28 of 735) (OR −0.46, 95% c.i. −1.39 to 0.46; P = 0.2). The pooled results are shown in Fig. 2 . Forest plot showing pooled estimate of anastomotic leak rate comparing circular stapler diameter (25 mm versus 28–29 mm) According to the Q test, there was no significant amount of heterogeneity in the true outcomes (Q = 1.55, P = 0.21, I 2 = 35.71%). Conversely, the AL rate was significantly lower (2.66%; 8 of 301) in the 28–29 mm group compared with the 31–33 mm cohort (6.07%; 27 of 445) (OR −0.92, 95% c.i. −1.74 to −0.10; P = 0.02) ( Fig. 3 ). Forest plot showing pooled estimate of anastomotic leak rate comparing circular stapler diameter (28–29 mm versus 31–33 mm) According to the Q test, the true outcomes appear to be heterogeneous (Q = 1.78, P = 0.41, I 2 = 0%). An examination of the studentized residuals revealed that one study 16 reported a value larger than ± 2.3940, being a potential outlier in the context of this model. Five studies 20–24 were identified on the correlation between the number of rows of the circular stapler and AL for a total of 19 158 patients. As shown in Fig. 4 , the AL rate was similar between the two study groups, with an AL incidence of 7.66% (1314 of 17 152) in the two-row group compared with 6.13% (123 of 2006) in the three-row group (OR −0.01, 95% c.i. −0.16 to 0.13; P = 0.85). Forest plot showing pooled estimate of anastomotic leak rate comparing the two- and three-row circular staplers According to the Q test, the true outcomes appear to be heterogeneous (Q = 19.08, P = 0.002, I 2 = 73.79%). Six studies 25–30 focused on the correlation between the powered or manual circular stapler and AL on 2180 patients. The AL rate was 7.74% (119 of 1537) in the manual group and 4.04% (26 of 643) in the powered group and the pooled results are shown in Fig. 5 (OR −0.83, 95% c.i. −1.31 to −0.35; P < 0.001). According to the Q test, there was no significant amount of heterogeneity in the true outcomes (Q = 3.87, P = 0.57, I 2 = 0%). Forest plot showing pooled estimate of anastomotic leak rate comparing the powered and manual circular staplers Five studies 14 , 31–34 were identified for the comparison between SST and DST for AL rate in 1372 patients. The SST related to a significantly lower rate of AL (5.46%; 26 of 476) than the DST (13.28%; 119 of 896). The pooled results are shown in Fig. 6 (OR 0.79, 95% c.i. 0.33 to 1.25; P < 0.001). According to the Q test, there was no significant amount of heterogeneity in the true outcomes (Q = 6.59, P = 0.16, I 2 = 32.3%). Forest plot showing pooled estimate of anastomotic leak rate comparing the single- and double-stapling techniques A sensitivity analysis was also performed for each statistically significant result (the 29 mm stapler compared with the 33 mm, powered versus manual technology and SST versus DST). As shown in Supplementary materials , Figs. S1–3 , the sensitivity analysis, conducted by excluding one study at a time, confirmed the findings of the meta-analysis, except for the comparison between the 29 and 33 mm staplers when the study of Jiang et al . 16 was excluded.

The reported incidence rate of AL after left colorectal surgery currently ranges between 5 and 28% 2 . Multiple studies have identified a multifactorial origin of this detrimental complication with the well-established influencing role of some non-modifiable patient-related risk factors, such as diabetes, smoking, advanced age, a high American Society of Anesthesiologists (ASA) score, male sex, a narrow pelvis, neoadjuvant treatment and the presence of a diverting stoma 37–39 . Recent advancements in surgical techniques and devices have significantly contributed to the amelioration of the surgical outcomes, including the rate of AL. Indeed, the introduction of the DST by Knight-Griffen 40 and Moran 41 , with the use of circular staplers for fashioning the colorectal anastomosis, significantly improved the surgical outcomes after colorectal resection, lowering the incidence rate of AL and increasing the rate of sphincter-preserving procedures 12 . This resulted in a rapid and exponential increase in the manufacturing of circular staplers as well as amelioration of this technology, also incentivized by the widespread use of the minimally invasive approach that made using surgical staplers even more essential. This reflected in the introduction of some important innovations such as the powered system, the three-row technology and a plethora of different diameters. However, despite these recent advancements, no clear evidence is present in the literature concerning the best circular stapler technology to employ in left colorectal surgery and no conclusive findings have been reported on the influencing role of these technologies on AL incidence. The aim of this meta-analysis was, thus, to evaluate the results in terms of AL using the different currently available circular staplers and to identify potential advantages of one technology over the others in terms of AL rate. Concerning the diameter of the circular staplers, there is no universal consensus on the optimal value to use when performing a colorectal anastomosis. Based on the postulation that a large anastomosis is potentially associated with a lower rate of anastomotic strictures 42 , several colorectal surgeons prefer staplers with a larger diameter. However, the use of smaller circular staplers presents some clear advantages such as easier insertion through the anus and the potentially lower rate of anus and sphincter muscle lesions 43 . In addition, their use in cases of low or very low rectal resections may decrease the incidental involvement of the vagina and sphincter muscles, reducing the risk of postoperative fistula 16 . The present meta-analysis included five studies focusing on the influence of the circular stapler diameter on AL, divided into two subgroups of analysis according to the circular stapler diameter. The results showed a higher rate of AL when the 31–33 mm stapler was used in comparison to the 28–29 mm (6.07% versus 2.66%; P = 0.02), while no difference was documented between the 25 mm and 28–29 mm groups (1.71% versus 3.81%; P = 0.2). This data would support the idea that the use of larger staplers is associated with a higher risk of AL. The majority of the currently available circular staplers have two rows of staples aimed to guarantee adequate tissue apposition avoiding damage and ischaemia of the anastomosis. More recently, technological improvement has led to the introduction of Tri-Staple technology ᵀᴹ (tri-EEA™) (Medtronic, Mansfield, MA, USA) based on the use of three lines of staples conceived for a higher resistance of the sutures, as suggested by preclinical animal models and artificial intelligence studies 20 , 23 . However, given the recent introduction of such a technology, the real effect on clinical outcomes is still to be demonstrated. According to the authors' results on pooled data of 19 158 patients, no difference was observed in terms of AL rate between the two- and three-row groups (6.13% versus 7.66%; P = 0.85). Nevertheless, these results are deeply influenced by the study of Wang et al., 24 which included more than 80% of the study population. The study of Wang et al. presents some drawbacks that need to be highlighted. First, it was a multicentric retrospective study, involving more than 400 hospitals across the USA, which implies significant variability in the surgical technique employed. Furthermore, no stratification for hospital volume was provided, leading to a potential bias related to the different surgical experiences in the different hospitals involved. Another technological solution proposed alternatively to the manual circular stapler is the powered one, introduced in 2019 by Ethicon (Echelon Circular™ Powered Stapler (ECP), Ethicon Endo-Surgery, Inc., Cincinnati, OH, USA). The rationale behind this technology is the potential capability of reducing the surgeon’s strength grip as well as the stapler head movements, ensuring a more stable suture line. An ex vivo preclinical model confirmed this theory, showing a more stable stapler line and a higher leak pressure 44 . In the present meta-analysis the authors examined six studies with a pooled population of 2180 patients and the powered device was confirmed to significantly lower the rate of AL in comparison to the manual technology (4.04% versus 7.74%; P < 0.001). In addition, the authors compared the DST to the SST. Five studies 14 , 31–34 were retrieved (pooled population of 1372 patients) with an AL rate lower in the SST group compared with the DST (5.46% versus 13.28%; P : < 0.001). One study 45 showed, in an animal model, how the bursting pressure in the DST was significantly lower than in the SST. This technical issue is well known and feared among colorectal surgeons and different invagination techniques of the ‘dog-ears’ have been proposed in order to avoid the creation of these weak spots 14 , 46 , 47 . It could be hypothesized that the reduction of the AL leak rate in the SST group may be related to the capability of the SST to overcome the ‘dog ears’ problem of the DST, creating a perfectly circular anastomosis without any point of weakness, characterized by an inadequate blood supply. However, the authors of the studies included in this meta-analysis used different solutions to overcome this same technical drawback 33 . Some authors 14 , 32 invaginated the extremities of the linear transection line of the rectum before fashioning the colorectal anastomosis, transforming the original DST to SST. On the other hand, another proposed SST is a transanal approach avoiding transecting the rectum with a linear stapler, creating a perfectly circular anastomosis 34 . The sensitivity analysis conducted for each statistically significant result of the meta-analysis further validated the findings, except in the comparison between the 29 and 33 mm staplers when the study by Jiang et al. 16 was excluded. This outcome may be attributable to the limited number of studies comparing the 29 and 33 mm staplers, which may have adversely affected the sensitivity analysis. This study includes a large number of studies and patients, addressing the key aspects regarding the use of circular staplers in colorectal surgery. However, the main limitations of this meta-analysis are related to the retrospective design of all the studies included. Indeed, although all the studies were of high quality according to the Newcastle–Ottawa scale, variations in the definition of AL, stapler technologies and study cohorts’ features may have contributed to the heterogeneity of the analysed population. Furthermore, the limited number of studies available on the addressed topics (especially for the evaluation of the role of larger circular staplers) represents an additional limitation for the generalization of the results. Thus, the need for future RCTs for each of the subanalyses conducted, also evaluating the influence of combined circular stapler features on AL rate, is necessary in order to further corroborate these results. In conclusion, there is a higher rate of AL using a larger circular stapler (31–33 mm), while the use of powered circular staplers and the SST seem to be associated with a lower incidence of AL. Of note, the use of three-row circular staplers did not demonstrate any advantages on AL occurrence.