- Email: [email protected]
Stapled versus hand-sewn pediatric intestinal anastomoses: A retrospective cohort study
Journal of Pediatric Surgery 53 (2018) 959–963
Contents lists available at ScienceDirect
Journal of Pediatric Surgery journal homepage: www.elsevier.com/locate/jpedsurg
Stapled versus hand-sewn pediatric intestinal anastomoses: A retrospective cohort study☆ Graeme C. Hintz a, Abdullah Alshehri b, Carolyn M. Bell a, Sonia A. Butterworth a,⁎ a b
Division of Pediatric Surgery, Department of Surgery, British Columbia Children's Hospital, Vancouver, British Columbia V6J 4K7, Canada Department of Surgery, Faculty of Medicine, King Saud University, Riyadh 11472, Saudi Arabia
a r t i c l e
i n f o
Article history: Received 13 January 2018 Accepted 1 February 2018 Keywords: Stapled intestinal anastomosis Complications Pediatric
a b s t r a c t Background: Whereas the adult literature has demonstrated the acceptable safety profile of stapled anastomoses when compared to the hand-sewn alternative, the choice of intestinal anastomosis using sutures or staples remains inadequately investigated in children. The purpose of this study is to compare the anastomotic outcomes of both techniques in children under 5 years of age. Methods: A retrospective analysis of patients undergoing intestinal anastomosis at a single tertiary centre (2012–2016) was undertaken. Demographics, diagnosis, anatomy, and complications were compared between the hand-sewn (HS) and stapled anastomosis (SA) groups. Primary outcomes were anastomotic leak and/or stricture requiring intervention. Results: There were 72 patients with 90 intestinal anastomoses (67 HS, 23 SA). Baseline demographics between the two anastomotic groups were comparable. The overall anastomotic complication rate was 23.9% (HS) and 17.4% (SA). In the ileocolic subgroup, anastomotic complications occurred in 3/7 HS vs. 0/5 SA (ns). There were no statistically significant differences in primary outcomes between HS and SA. All SA complications occurred with 3.5 or 3.8 mm staples. Conclusions: In our study population, no statistically significant difference between hand-sewn and stapled intestinal anastomosis outcomes was found. However, further investigation is warranted. Level of Evidence: 3 (Retrospective Comparative Treatment Study) © 2018 Elsevier Inc. All rights reserved.
The creation of an intestinal anastomosis in children is commonly performed. Though the choice of anastomotic technique often hinges on the caliber, quality, and disease process at hand, it also frequently is a personal choice of the surgeon, based on his or her experience [1]. In adults, the safety profile of stapled intestinal anastomoses is well documented [1-4]. Furthermore, level one evidence demonstrates that stapled ileocolic anastomoses have superior results, when compared to the hand-sewn alternative [2-4]. Stapling devices have been used with increased frequency over the past several decades for the creation of pediatric intestinal anastomoses and, with the development of endoscopic staplers, their use in even the small caliber intestine of infants has been made possible [5-9]. However, the data on the use of any stapling device in the pediatric literature is less robust than that found in the adult realm. The studies that exist are mostly limited to case reports or small retrospective case reviews
[10-16]. Most have found no difference in safety parameters between the two methods of anastomoses, and multiple have reported shorter operative times in the stapled groups [10,11]. A preliminary audit of our institution's outcomes demonstrated a significant increase in the likelihood of using a stapler for the creation of intestinal anastomoses in older children and adolescents, and a significant increase in anastomotic complications in younger patients, regardless of technique. Given this, we sought to focus the present study on a younger patient cohort, in order to improve the concordance of baseline group characteristics, as well as to focus on the group at highest risk of poor outcomes. Therefore, our study's purpose is to compare the outcomes of hand-sewn (HS) and stapled (SA) intestinal anastomoses in young children and infants, with a secondary aim to identify patient and technique factors that are associated with improved outcomes. 1. Materials and methods
☆ This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. ⁎ Corresponding author at: British Columbia Children's Hospital, 4480 Oak Street, Room K0-110 ACB, Vancouver, British Columbia C V6H 3V4, Canada. Tel.: +1 604 875 3744; fax: +1 604 875 2721. E-mail address: [email protected] (S.A. Butterworth). https://doi.org/10.1016/j.jpedsurg.2018.02.021 0022-3468/© 2018 Elsevier Inc. All rights reserved.
A retrospective review of pediatric patients under 5 years of age undergoing an intra-abdominal intestinal anastomosis between 2012 and 2016 at a single large tertiary pediatric hospital was performed. Institutional ethics approval was obtained [#H16–01016]. Demographic data including gestational age at birth, gender, the presence of any congenital
960
G.C. Hintz et al. / Journal of Pediatric Surgery 53 (2018) 959–963
Table 1 Patient Details.
Number of Anastomoses Demographics Premature Birth weighta Gender Female Male Congenital anomaly present Age at procedurea Weight at procedurea Diagnosis Imperforate anus Intestinal atresia Malrotation and volvulus Hirschsprung's disease Intussusception Meconium ileus Adhesive small bowel obstruction Intestinal duplication Gastroschisis Meckel's diverticulum Necrotizing enterocolitis Spontaneous intestinal perforation Dysmotility Familial intrahepatic cholestasis Foreign body ingestion Internal hernia Intestinal stenosis Trauma Operative/anastomotic characteristics Procedure type Bowel resection and primary anastomosis Delayed primary anastomosis/stoma closure Bishop-Koop reconstruction Procedure priority Emergent Urgent Elective Anastomotic level Small bowel Small-large bowel Large bowel Bowel health compromised Contamination present Size discrepancy present Follow-upa a b c d
Hand-Sewn
Stapled
p Value
67
23
n/a
18 (26.9%) 2936
3 (13%) 2962
0.235 0.797 0.632
33 (49.3%) 34 (50.7%) 25 (37.3%) 8.9 months 7950 g
10 (43.5%) 13 (56.5%) 4 (17.4%) 10.5 months 8200 g
24b 13b 4 7 3 3 2 3 0 2 0 2 0 1 1 0 1 0
2 2 3 2 4 2 2 0 2 0 2 0 1 0 0 1 0 1
24 (36.4%) 38 (57.6%) 4 (6%)
14 (60.9%) 9 (39.1%) 0 (0%)
8 (11.9%) 18 (26.9%) 41 (61.2%)
3 (13%) 13 (56.5%)c 7 (30.4%)c
25 (37.3%) 9 (13.6%) 33 (50%) 6 (9%) 1 (1.5%) 13 (19.4%) 12 months
12 (52.2%) 8 (34.8%)d 3 (13%) 5 (22%) 3 (13%) 6 (26%) 13 months
0.078 0.629 0.848 0.012b
0.087
0.024c
0.011d
0.106 0.020 0.498 0.739
Median. Within the hand-sewn group, observed cases of imperforate anus and intestinal atresia were significantly higher than expected. Within the stapled group, observed case of urgent priority were significantly higher, and cases of elective priority significantly lower, than expected. Within the stapled group, observed cases of ileocolic anastomoses were significantly higher than expected.
anomalies, underlying surgical diagnosis, age and weight at procedure, and the length of follow-up was collected. Characteristics of the anastomosis and surgery were also collected, including the procedure type, booking priority, the health of the bowel noted intraoperatively, whether purulent or enteric contamination was noted intraoperatively, location of the anastomosis, and whether a size discrepancy was documented between the ends of the anastomosis. Emergent priority was defined as an operative booking priority of less than 1 h, elective as any procedure scheduled in advance of patient admission, and urgent as a procedure with a booking priority between these two extremes. Intestinal compromise was defined by the documentation in the operative note of ischemia, pallor, discoloration, and/or friability of the bowel. Contamination was indicated by the presence of enteric spillage or purulent peritoneal fluid. Details regarding the stapler device or suture material used were also recorded. Our primary outcome measure was anastomotic complication, which was defined as either an anastomotic leak or stricture, or both, requiring intervention. A leak was defined as the presence of a communication between the intra- and extra-luminal compartments, as manifested by intra-abdominal enteric contents or entero-cutaneous fistula
post-operatively. Secondary outcomes included operative time, estimated blood loss (EBL), time to initial and full enteral feeding, bowel obstruction, abdominal abscess development, and length of stay (LOS). We used chi-square or Fisher-exact tests for comparison of binary outcomes, Student's t-test for normally distributed data, and Mann– Whitney U for non-parametric data. A multivariate analysis was performed, examining the influence of both the type of anastomosis (HS vs. SA), as well as the weight at surgery, on the outcome variable of anastomotic complication. Statistical analysis was carried out using IBM SPSS Statistics-version 20 (IBM, New York, USA). A p value of b 0.05 was considered to be statistically significant. Hand-sewn anastomoses were created in an end-to-end fashion using either a monofilament or braided, absorbable suture (PDS® or Vicryl® respectively). Stapled anastomoses were created in a side-toside, functional end-to-end fashion with open linear or endoscopic linear staplers, having their common opening closed with either a stapler or, in the case of limited intestinal length, with a hand-sewn technique. Prior to October 2015, staplers were Proximate® and ETS® (Ethicon, Somerville, NJ). After this date DST GIA® and Endo GIA® (Covidien,
G.C. Hintz et al. / Journal of Pediatric Surgery 53 (2018) 959–963 Table 2 Outcome comparisons between hand-sewn and staples anastomoses.
Anastomotic complication Leak Stricture Secondary outcomes Length of operationb Estimated blood loss b 10 mL 10-100 mL N 100 mL Time to initial feedsb Time to full Feedsb Bowel obstruction Abdominal abscess Length of stayb
Hand-Sewn
Stapled
p Value
16 (23.9%)a 7 (10.4%) 14 (20.9%)
4 (17.4%) 2 (8.7%) 3 (13%)
0.518 0.809 0.406
107 min
127 min
0.352 0.136
37 (55.2%) 21 (31.3%) 1 (1.5%) 3 days 5 days 17 (25.4%) 0 (0%) 7 days
8 (34.8%) 12 (52%) 1 (4.3%) 7 days 11 days 5 (7.6%) 0 (0%) 11 days
0.058 0.161 0.726 n/a 0.501
a Both a leak and a stricture occurred in 5 of the hand-sewn group, and 1 of the stapled group; each was only counted as a single anastomotic complication overall. b Median.
New Haven, CT) devices were used. For small caliber intestinal lumens, a lubricated #22 French red rubber catheter was inserted to determine if it would admit the small end of the endoscopic stapler. If admitted, the catheter was then gently flushed with warm saline within the intestinal lumen, the catheter removed, and then the stapled anastomosis completed. If the intestine could not admit the catheter, a sutured anastomosis was performed. Surgeon preference, as well as an intestinal caliber, determined the anastomotic choice. If key data points were missing on chart review, for instance anastomotic details, these cases were excluded from the trial. Cases in which less significant details lacked were still included. For instance, even if a procedure's estimated blood loss was not recorded, as long as it was not missing key anastomotic details and outcomes, available data for the patient was used in the analysis. 2. Results Over the four years reviewed, a total of 90 intestinal anastomoses were performed in 72 patients. 67 were hand-sewn, and 23 were stapled (Table 1). Both age and weight at the time of surgery were comparable between the anastomotic groups. In the HS group, the median age at the time of operation was 8.9 months, while it was 10.5 months in the SA group (p = 0.629). In the HS group, median weight was 8 kg, while it was 8.2 kg in the SA group (p = 0.848). A significant difference between anastomotic groups was seen in their diagnostic categories, with more
961
cases of imperforate anus (for colostomy closure) and intestinal atresia being performed in a hand-sewn fashion (p = 0.012). No significant differences in the other patient characteristics, including prematurity and comorbidity, were seen. There were significant differences between the HS and SA groups with respect to anatomic and surgical factors. A stapled technique was more often performed in emergent cases compared to elective cases (p = 0.024), purulent or enteric contamination (p = 0.02), and anastomoses between small and large bowel (p = 0.011). Other operative characteristics, including intestinal size discrepancy, did not differ significantly between groups. Median follow-up was around 1 year for both groups. There were a total of 20 anastomotic complications: 11 with strictures alone, 3 with leaks alone, and 6 with both a leak and a stricture. Comparing HS to SA, no significant difference in these primary outcomes was seen, with 10.4% of HS, and 8.7% of SA suffering a leak (p = 0.809), and 20.9% of HS, and 13% of SA suffering a stricture (p = 0.406) (Table 2). When the ileocolic subgroup was evaluated specifically, there was an anastomotic complication rate of 42.9% (3 of 7) in the HS group, compared to 0% (0 of 5) in the SA group. This result was not statistically significant (p = 0.091). For the four of twenty-three SA which had a complication, stapler type, stapler size, and technique were evaluated (Table 3). In two, a Proximate® linear stapler with 3.5 mm or 3.8 mm staples was used with patients weights of 20, and 22 kg. These resulted in a leak and stricture in the former, and just a stricture in the latter. In two other infants, weighing 6.04 and 6.41 kg, a SA was performed using ETS® 3.5 mm staples. These resulted in a stricture in the former, and a leak in the latter. There were no leaks or strictures in patients undergoing SA after October 2015, when our institution switched stapler vendor. In terms of secondary outcomes, no statistically significant differences were seen between the two anastomotic groups. Median time to initial and full feeds was 3 and 7 days in the hand-sewn group, respectively, and 5 and 11 days in the stapled group, respectively. This, and all other secondary outcomes were similar, including development of an intra-abdominal abscess (0% in both groups), post-operative adhesive bowel obstruction (median incidence of 16.5% overall), length of operation (median of 117 min overall), and length of stay (median of 9 days overall). In a multivariate analysis of potential risk factors for anastomotic complications, stapled technique and weight b/=10 kg were both assessed, and not found to be significant independent predictors of poor outcome(Table 4).
Table 3 Anastomotic complication analysis. Anastomosis
Technique
Weight at OR (kg)
Diagnosis
Anastomotic level
Complication
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Handsewn Handsewn Handsewn Handsewn Handsewn Handsewn Handsewn Handsewn Handsewn Handsewn Handsewn Handsewn Handsewn Handsewn Handsewn Handsewn Stapled (3.5 Stapled (3.8 Stapled (3.8 Stapled (3.8
2.07 2.4 3.55 6 6.38 6.38 6.38 6.75 6.9 7.23 7.23 8.2 9.8 9.9 10 22 6.04 6.41 20 22
Intestinal atresia Intestinal atresia Malrotation Intestinal atresia Meconium ileus Meconium ileus Meconium ileus Hirschsprung Imperforate anus Intestinal atresia Intestinal atresia Imperforate anus Hirschsprung Imperforate anus Hirschsprung Malrotation Gastroschisis Imperforate anus Malrotation Malrotation
Small bowel Small bowel Ileocolic Ileocolic Small bowel Small bowel Small bowel Small bowel Coloanal Colonic Colonic Colonic Ileorectal Colonic Small bowel Ileocolic Small bowel Colonic Small bowel Small bowel
Stricture Stricture Stricture Stricture Leak, stricture Leak, stricture Leak, stricture Stricture Stricture Stricture Stricture Leak, stricture Stricture Leak Leak Leak, stricture Stricture Leak Stricture Leak, stricture
mm) mm) mm) mm)
962
G.C. Hintz et al. / Journal of Pediatric Surgery 53 (2018) 959–963
Table 4 Multivariate regression analysis of factors related to anastomotic leak and/or stricture.a Factor
Odds ratio
95% Confidence interval
p Value
Hand-sewn anastomosis Weight b10 kg
1.40 2.58
0.40–4.86 0.68–9.97
0.60 0.16
a
Adjusted for age and anastomotic level.
3. Discussion Stapled intestinal anastomoses have been widely studied in the adult population, and thus their safety and efficacy when compared to hand-sewn anastomoses is un-questioned [1–3]. In fact, there is evidence to suggest that, besides a decrease in operative time, a decreased rate of anastomotic leak may be seen in ileocolic anastomoses specifically, with the use of a stapled technique [4]. In children however, the evidence base is not nearly as strong. Despite stapling devices being first used in pediatric intestinal anastomoses in the late 1960s [5], and more commonly for a variety of indications through the 1990s until present day [6-9], evidence is limited to several case reports and series, and small retrospective reviews [10-16]. Overall, the evidence that is available has shown no difference between stapled and hand-sewn anastomotic techniques, and many studies mirror those in the adult population in that operative times are decreased with the use of a stapler device [10-13,15,16]. In our institution, we found stapling was more likely to be utilized in cases that were emergent in nature, or were classified as contaminated. Each of these factors could potentially increase the risk to the anastomosis. That we did not find a significant increase in anastomotic complications using SA despite it being used more often in these high-risk cases may suggest that the use of stapler devices offers improved outcomes in patients who are especially compromised. Our study failed to show a decrease in operative time when staplers were used however, unlike what several other studies have shown [4,12]. It is possible that the profile of patients undergoing SA, that is of being potentially more ill than those undergoing a hand-sewn anastomosis, contributed to the lack of this finding. We also found that the hand-sewn technique was more often used in cases of imperforate anus (colostomy closure), and intestinal atresia. This is likely due to the fact that these pathologies are treated while patients are yet infants, and therefore have a smaller intestinal caliber. Our findings are not directly comparable to some other pediatric studies, which evaluated both HS and SA techniques in that all of our patients underwent end-to-end anastomoses, whereas other pediatric case series have included anti-mesenteric stapling (e.g. patent ompaholomesenteric duct), and Roux-en-Y anastomoses [11,12]. Arguably, the end-to end anastomosis is a higher risk undertaking [17], and this may contribute to our relatively high complication rate in this group, as compared with these other studies. The technical aspects of stapled intestinal anastomoses are important considerations. The study by Kozlov et al. made use of endoscopic staplers with 2.5 mm staples, and no leaks or strictures occurred [12]. In the study by Sato et al., endoscopic staplers were also used, but with 1.0–1.5 mm staples, and again no anastomotic complications occurred [11]. In all four of the patients in our study with a complication of a stapled anastomosis, 3.5 or 3.8 mm staples were used: in one, an endoscopic stapler with three rows of 3.5 mm staples and, in three, an open linear stapler was used with two rows of 3.8 mm. All patients who had a leak from a stapled anastomosis in our study were 22 kg or less, suggesting that perhaps the staple height was too large for the patient's tissue. Intestinal thickness increases with age and does not reach maximal wall diameter until a patient's third decade of life [18]. Mean intestinal wall diameters in the small bowel are from 0.7 mm (0-4 years) to 1.1 mm (15–19 years); colonic wall thickness averages 1.0 mm (0–4 years) and 1.4 mm (15–19 years). Typically the minimum recommended thickness of tissue for 3.5 mm and 3.8 mm staples is 1.5 mm [19,20] and, given the patients' ages and tissue types (three
out of four were small bowel), it is expected that for the four with an anastomotic complication, the staples were too big, implicating them as a causative factor in their leaking. Since the change of staplers occurred in October of 2015, there have been no anastomotic complications. One possible reason for this is that Covidien Endo staplers have three staple heights for each load, thus covering a wider range of tissue thickness than the stapler used prior to October 2015 in our institution. In addition, the entire end of the stapler is switched after each firing (both the fixed and stepped cartridge anvils), presumably decreasing the chance of any errant staples causing a misfiring of the blade. For the stapled anastomoses that resulted in stricture in our study, another possible technical cause was that the diameter of the resultant anastomosis was insufficient. To explain, Jackson et al. reported two patients who had significant complications related to dilation in the proximal limb of a stapled anastomosis [14]. One possible cause may have been a staple line that was either too long, precipitating torsion or too small, resulting in partial obstruction and resultant proximal dilation. We would suggest it is important to ensure an anastomosis that is sufficient, but not too long in the case of luminal disparity so as to minimize risk of torsion of the anastomosis. One suggestion is that the anastomotic length be wider than the smallest lumen, but no longer than twice the diameter of the largest of the lumens, though to our knowledge no evidence exists to guide this practice. There are several limitations to our study. First, the choice of anastomotic technique was not random and, for infants in particular, SA was only regularly considered by one of the five surgeons at our institution, thus introducing significant bias. Next, there was a small incidence of primary outcomes overall, which implies our population was not adequate to detect statistically significant differences, especially in regard to our regression analysis. In addition, we grouped the patients who had both a stapled or hand-sewn closure of the common channel in the SA group, which arguably may have been better to include as a third unique anastomotic type. The finding that there were no intraabdominal abscesses was a surprise, and potentially suggests that their occurrence was not documented in a way that was captured in our data collection. Another limitation of our study was that the grade of morbidities was not classified; classification would have allowed better comparison of patients who had a complication. 4. Conclusions We present, to our knowledge, one of the largest reported pediatric studies comparing outcomes of hand-sewn and stapled intestinal anastomoses, and we found no significant differences in anastomotic outcomes between the two groups. This suggests that, when permitted by intestinal size, a stapled anastomosis is a safe choice. Ultimately, as already carried out in the adult population, a prospective, multi-centre trial would be most helpful in comparing stapled to hand-sewn anastomotic techniques in the pediatric population, and we would suggest this as a future direction. We also would propose a continued focus on young children or infants, given the higher complication rate seen in these sub-groups. Consequences of an intestinal anastomotic complication can result in substantial morbidity and even death; rigorous evaluation of patient and operative factors is key to optimizing outcomes. Acknowledgements Ms. Maryam Noparast for assistance with statistical analysis. References [1] Goulder Frances. Bowel anastomoses: The theory, the practice and the evidence base. World J Gastrointest Surg 2012;4:208–13. [2] Neutzling CB, Lustosa SA, Da Silva EM, et al. Stapled versus handsewn methods for colorectal anastomosis surgery. Cochrane Database Syst Rev 2012;15:CD003144.
G.C. Hintz et al. / Journal of Pediatric Surgery 53 (2018) 959–963 [3] Slieker JC, Daams F, Mulder IM, et al. Systematic Review of the Technique of Colorectal Anastomosis. JAMA Surg 2013;148:190–201. [4] Choy PY, Bissett IP, Docherty JG, et al. Stapled versus handsewn methods for ileocolic anastomoses. Cochrane Database Syst Rev 2011;7:CD004320. [5] Talbert JL, Seashore JH, Ravitch MM. Evaluation of a modified Duhamel operation for correction of Hirschsprung's disease. Ann Surg 1974;179(5):671. [6] Hedlund H. Colorectal resection and anal anastomosis with an intraluminal stapler in Hirschsprung disease. Pediatr Surg Int 1997;12:142–4. [7] Olguner M, Akgur FM, Ucan B, et al. Laparoscopic appendectomy in children performed using single endoscopic GIA stapler for both mesoappendix and base of appendix. J Pediatr Surg 1998;33(9):1347. [8] Valla JS, Steyaert H, Leculee R, et al. Meckel's diverticulum and laparoscopy of children. What's new? Eur J Pediatr Surg 1998;8:26–8. [9] Mattioli G, Castagnetti M, Gaudullia P, et al. Stapled restorative proctocolectomy in children with refractory ulcerative colitis. J Pediatr Surg 2005;40:1773–9. [10] Wrighton L, Curtis JL, Gollin G. Stapled intestinal anastomoses in infants. J Pediatr Surg 2008;43:2231–4. [11] Sato K, Uchida H, Tanaka Y, et al. Stapled intestinal anastomosis is a simple and reliable method for management of intestinal caliber discrepancy in children. Pediatr Surg Int 2012;28:893–8.
963
[12] Kozlov Y, Novogilov V, Podkamenev A, et al. Stapled bowel anastomoses in newborn surgery. Eur J Pediatr Surg 2013;23:63–6. [13] Mitchell IC, Barber R, Fischer AC, et al. Experience performing 64 consecutive stapled intestinal anastomoses in small children and infants. J Pediatr Surg 2011;46:128–30. [14] Jackson CC, Bettolli MM, De Carli CF, et al. Beware of stapled side-to-side bowel anastomoses in small children. Eur J Pediatr Surg 2007;17:426–7. [15] Powell RW. Stapled intestinal anastomosis in neonates and infants: use of the endoscopic intestinal stapler. J Pediatr Surg 1995;30:195–7. [16] Simmons JD, Gunter III JW, Manley JD, et al. Stapled intestinal anastomosis in neonates. Am Surg 2010;76:644–6. [17] Bagolan P, Nappo S, Trucchi A, et al. Neonatal intestinal obstruction: reducing shortterm complications by surgical refinements. Eur J Pediatr Surg 1996;6:354–7. [18] Haber HP, Stern M. Intestinal ultrasonography in children and young adults: bowel wall thickness is age dependent. J Ultrasound Med 2000;19:315–21. [19] “Proximate Linear Cutters.” Ethicon US, LLC. http://www.ethicon.com/healthcareprofessionals/products/staplers/linear-cutters/proximate-linear-cutters, Accessed date: 13 July 2017. [20] “Endo GIA Reloads with Tri-Spale Technology, Technical Brochure.” Medtronic US. http://www.medtronic.com/content/dam/covidien/library/us/en/product/surgicalstapling/endo-gia-reloads-technical-brochure.pdf, Accessed date: 13 July 2017.
Contents lists available at ScienceDirect
Journal of Pediatric Surgery journal homepage: www.elsevier.com/locate/jpedsurg
Stapled versus hand-sewn pediatric intestinal anastomoses: A retrospective cohort study☆ Graeme C. Hintz a, Abdullah Alshehri b, Carolyn M. Bell a, Sonia A. Butterworth a,⁎ a b
Division of Pediatric Surgery, Department of Surgery, British Columbia Children's Hospital, Vancouver, British Columbia V6J 4K7, Canada Department of Surgery, Faculty of Medicine, King Saud University, Riyadh 11472, Saudi Arabia
a r t i c l e
i n f o
Article history: Received 13 January 2018 Accepted 1 February 2018 Keywords: Stapled intestinal anastomosis Complications Pediatric
a b s t r a c t Background: Whereas the adult literature has demonstrated the acceptable safety profile of stapled anastomoses when compared to the hand-sewn alternative, the choice of intestinal anastomosis using sutures or staples remains inadequately investigated in children. The purpose of this study is to compare the anastomotic outcomes of both techniques in children under 5 years of age. Methods: A retrospective analysis of patients undergoing intestinal anastomosis at a single tertiary centre (2012–2016) was undertaken. Demographics, diagnosis, anatomy, and complications were compared between the hand-sewn (HS) and stapled anastomosis (SA) groups. Primary outcomes were anastomotic leak and/or stricture requiring intervention. Results: There were 72 patients with 90 intestinal anastomoses (67 HS, 23 SA). Baseline demographics between the two anastomotic groups were comparable. The overall anastomotic complication rate was 23.9% (HS) and 17.4% (SA). In the ileocolic subgroup, anastomotic complications occurred in 3/7 HS vs. 0/5 SA (ns). There were no statistically significant differences in primary outcomes between HS and SA. All SA complications occurred with 3.5 or 3.8 mm staples. Conclusions: In our study population, no statistically significant difference between hand-sewn and stapled intestinal anastomosis outcomes was found. However, further investigation is warranted. Level of Evidence: 3 (Retrospective Comparative Treatment Study) © 2018 Elsevier Inc. All rights reserved.
The creation of an intestinal anastomosis in children is commonly performed. Though the choice of anastomotic technique often hinges on the caliber, quality, and disease process at hand, it also frequently is a personal choice of the surgeon, based on his or her experience [1]. In adults, the safety profile of stapled intestinal anastomoses is well documented [1-4]. Furthermore, level one evidence demonstrates that stapled ileocolic anastomoses have superior results, when compared to the hand-sewn alternative [2-4]. Stapling devices have been used with increased frequency over the past several decades for the creation of pediatric intestinal anastomoses and, with the development of endoscopic staplers, their use in even the small caliber intestine of infants has been made possible [5-9]. However, the data on the use of any stapling device in the pediatric literature is less robust than that found in the adult realm. The studies that exist are mostly limited to case reports or small retrospective case reviews
[10-16]. Most have found no difference in safety parameters between the two methods of anastomoses, and multiple have reported shorter operative times in the stapled groups [10,11]. A preliminary audit of our institution's outcomes demonstrated a significant increase in the likelihood of using a stapler for the creation of intestinal anastomoses in older children and adolescents, and a significant increase in anastomotic complications in younger patients, regardless of technique. Given this, we sought to focus the present study on a younger patient cohort, in order to improve the concordance of baseline group characteristics, as well as to focus on the group at highest risk of poor outcomes. Therefore, our study's purpose is to compare the outcomes of hand-sewn (HS) and stapled (SA) intestinal anastomoses in young children and infants, with a secondary aim to identify patient and technique factors that are associated with improved outcomes. 1. Materials and methods
☆ This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. ⁎ Corresponding author at: British Columbia Children's Hospital, 4480 Oak Street, Room K0-110 ACB, Vancouver, British Columbia C V6H 3V4, Canada. Tel.: +1 604 875 3744; fax: +1 604 875 2721. E-mail address: [email protected] (S.A. Butterworth). https://doi.org/10.1016/j.jpedsurg.2018.02.021 0022-3468/© 2018 Elsevier Inc. All rights reserved.
A retrospective review of pediatric patients under 5 years of age undergoing an intra-abdominal intestinal anastomosis between 2012 and 2016 at a single large tertiary pediatric hospital was performed. Institutional ethics approval was obtained [#H16–01016]. Demographic data including gestational age at birth, gender, the presence of any congenital
960
G.C. Hintz et al. / Journal of Pediatric Surgery 53 (2018) 959–963
Table 1 Patient Details.
Number of Anastomoses Demographics Premature Birth weighta Gender Female Male Congenital anomaly present Age at procedurea Weight at procedurea Diagnosis Imperforate anus Intestinal atresia Malrotation and volvulus Hirschsprung's disease Intussusception Meconium ileus Adhesive small bowel obstruction Intestinal duplication Gastroschisis Meckel's diverticulum Necrotizing enterocolitis Spontaneous intestinal perforation Dysmotility Familial intrahepatic cholestasis Foreign body ingestion Internal hernia Intestinal stenosis Trauma Operative/anastomotic characteristics Procedure type Bowel resection and primary anastomosis Delayed primary anastomosis/stoma closure Bishop-Koop reconstruction Procedure priority Emergent Urgent Elective Anastomotic level Small bowel Small-large bowel Large bowel Bowel health compromised Contamination present Size discrepancy present Follow-upa a b c d
Hand-Sewn
Stapled
p Value
67
23
n/a
18 (26.9%) 2936
3 (13%) 2962
0.235 0.797 0.632
33 (49.3%) 34 (50.7%) 25 (37.3%) 8.9 months 7950 g
10 (43.5%) 13 (56.5%) 4 (17.4%) 10.5 months 8200 g
24b 13b 4 7 3 3 2 3 0 2 0 2 0 1 1 0 1 0
2 2 3 2 4 2 2 0 2 0 2 0 1 0 0 1 0 1
24 (36.4%) 38 (57.6%) 4 (6%)
14 (60.9%) 9 (39.1%) 0 (0%)
8 (11.9%) 18 (26.9%) 41 (61.2%)
3 (13%) 13 (56.5%)c 7 (30.4%)c
25 (37.3%) 9 (13.6%) 33 (50%) 6 (9%) 1 (1.5%) 13 (19.4%) 12 months
12 (52.2%) 8 (34.8%)d 3 (13%) 5 (22%) 3 (13%) 6 (26%) 13 months
0.078 0.629 0.848 0.012b
0.087
0.024c
0.011d
0.106 0.020 0.498 0.739
Median. Within the hand-sewn group, observed cases of imperforate anus and intestinal atresia were significantly higher than expected. Within the stapled group, observed case of urgent priority were significantly higher, and cases of elective priority significantly lower, than expected. Within the stapled group, observed cases of ileocolic anastomoses were significantly higher than expected.
anomalies, underlying surgical diagnosis, age and weight at procedure, and the length of follow-up was collected. Characteristics of the anastomosis and surgery were also collected, including the procedure type, booking priority, the health of the bowel noted intraoperatively, whether purulent or enteric contamination was noted intraoperatively, location of the anastomosis, and whether a size discrepancy was documented between the ends of the anastomosis. Emergent priority was defined as an operative booking priority of less than 1 h, elective as any procedure scheduled in advance of patient admission, and urgent as a procedure with a booking priority between these two extremes. Intestinal compromise was defined by the documentation in the operative note of ischemia, pallor, discoloration, and/or friability of the bowel. Contamination was indicated by the presence of enteric spillage or purulent peritoneal fluid. Details regarding the stapler device or suture material used were also recorded. Our primary outcome measure was anastomotic complication, which was defined as either an anastomotic leak or stricture, or both, requiring intervention. A leak was defined as the presence of a communication between the intra- and extra-luminal compartments, as manifested by intra-abdominal enteric contents or entero-cutaneous fistula
post-operatively. Secondary outcomes included operative time, estimated blood loss (EBL), time to initial and full enteral feeding, bowel obstruction, abdominal abscess development, and length of stay (LOS). We used chi-square or Fisher-exact tests for comparison of binary outcomes, Student's t-test for normally distributed data, and Mann– Whitney U for non-parametric data. A multivariate analysis was performed, examining the influence of both the type of anastomosis (HS vs. SA), as well as the weight at surgery, on the outcome variable of anastomotic complication. Statistical analysis was carried out using IBM SPSS Statistics-version 20 (IBM, New York, USA). A p value of b 0.05 was considered to be statistically significant. Hand-sewn anastomoses were created in an end-to-end fashion using either a monofilament or braided, absorbable suture (PDS® or Vicryl® respectively). Stapled anastomoses were created in a side-toside, functional end-to-end fashion with open linear or endoscopic linear staplers, having their common opening closed with either a stapler or, in the case of limited intestinal length, with a hand-sewn technique. Prior to October 2015, staplers were Proximate® and ETS® (Ethicon, Somerville, NJ). After this date DST GIA® and Endo GIA® (Covidien,
G.C. Hintz et al. / Journal of Pediatric Surgery 53 (2018) 959–963 Table 2 Outcome comparisons between hand-sewn and staples anastomoses.
Anastomotic complication Leak Stricture Secondary outcomes Length of operationb Estimated blood loss b 10 mL 10-100 mL N 100 mL Time to initial feedsb Time to full Feedsb Bowel obstruction Abdominal abscess Length of stayb
Hand-Sewn
Stapled
p Value
16 (23.9%)a 7 (10.4%) 14 (20.9%)
4 (17.4%) 2 (8.7%) 3 (13%)
0.518 0.809 0.406
107 min
127 min
0.352 0.136
37 (55.2%) 21 (31.3%) 1 (1.5%) 3 days 5 days 17 (25.4%) 0 (0%) 7 days
8 (34.8%) 12 (52%) 1 (4.3%) 7 days 11 days 5 (7.6%) 0 (0%) 11 days
0.058 0.161 0.726 n/a 0.501
a Both a leak and a stricture occurred in 5 of the hand-sewn group, and 1 of the stapled group; each was only counted as a single anastomotic complication overall. b Median.
New Haven, CT) devices were used. For small caliber intestinal lumens, a lubricated #22 French red rubber catheter was inserted to determine if it would admit the small end of the endoscopic stapler. If admitted, the catheter was then gently flushed with warm saline within the intestinal lumen, the catheter removed, and then the stapled anastomosis completed. If the intestine could not admit the catheter, a sutured anastomosis was performed. Surgeon preference, as well as an intestinal caliber, determined the anastomotic choice. If key data points were missing on chart review, for instance anastomotic details, these cases were excluded from the trial. Cases in which less significant details lacked were still included. For instance, even if a procedure's estimated blood loss was not recorded, as long as it was not missing key anastomotic details and outcomes, available data for the patient was used in the analysis. 2. Results Over the four years reviewed, a total of 90 intestinal anastomoses were performed in 72 patients. 67 were hand-sewn, and 23 were stapled (Table 1). Both age and weight at the time of surgery were comparable between the anastomotic groups. In the HS group, the median age at the time of operation was 8.9 months, while it was 10.5 months in the SA group (p = 0.629). In the HS group, median weight was 8 kg, while it was 8.2 kg in the SA group (p = 0.848). A significant difference between anastomotic groups was seen in their diagnostic categories, with more
961
cases of imperforate anus (for colostomy closure) and intestinal atresia being performed in a hand-sewn fashion (p = 0.012). No significant differences in the other patient characteristics, including prematurity and comorbidity, were seen. There were significant differences between the HS and SA groups with respect to anatomic and surgical factors. A stapled technique was more often performed in emergent cases compared to elective cases (p = 0.024), purulent or enteric contamination (p = 0.02), and anastomoses between small and large bowel (p = 0.011). Other operative characteristics, including intestinal size discrepancy, did not differ significantly between groups. Median follow-up was around 1 year for both groups. There were a total of 20 anastomotic complications: 11 with strictures alone, 3 with leaks alone, and 6 with both a leak and a stricture. Comparing HS to SA, no significant difference in these primary outcomes was seen, with 10.4% of HS, and 8.7% of SA suffering a leak (p = 0.809), and 20.9% of HS, and 13% of SA suffering a stricture (p = 0.406) (Table 2). When the ileocolic subgroup was evaluated specifically, there was an anastomotic complication rate of 42.9% (3 of 7) in the HS group, compared to 0% (0 of 5) in the SA group. This result was not statistically significant (p = 0.091). For the four of twenty-three SA which had a complication, stapler type, stapler size, and technique were evaluated (Table 3). In two, a Proximate® linear stapler with 3.5 mm or 3.8 mm staples was used with patients weights of 20, and 22 kg. These resulted in a leak and stricture in the former, and just a stricture in the latter. In two other infants, weighing 6.04 and 6.41 kg, a SA was performed using ETS® 3.5 mm staples. These resulted in a stricture in the former, and a leak in the latter. There were no leaks or strictures in patients undergoing SA after October 2015, when our institution switched stapler vendor. In terms of secondary outcomes, no statistically significant differences were seen between the two anastomotic groups. Median time to initial and full feeds was 3 and 7 days in the hand-sewn group, respectively, and 5 and 11 days in the stapled group, respectively. This, and all other secondary outcomes were similar, including development of an intra-abdominal abscess (0% in both groups), post-operative adhesive bowel obstruction (median incidence of 16.5% overall), length of operation (median of 117 min overall), and length of stay (median of 9 days overall). In a multivariate analysis of potential risk factors for anastomotic complications, stapled technique and weight b/=10 kg were both assessed, and not found to be significant independent predictors of poor outcome(Table 4).
Table 3 Anastomotic complication analysis. Anastomosis
Technique
Weight at OR (kg)
Diagnosis
Anastomotic level
Complication
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Handsewn Handsewn Handsewn Handsewn Handsewn Handsewn Handsewn Handsewn Handsewn Handsewn Handsewn Handsewn Handsewn Handsewn Handsewn Handsewn Stapled (3.5 Stapled (3.8 Stapled (3.8 Stapled (3.8
2.07 2.4 3.55 6 6.38 6.38 6.38 6.75 6.9 7.23 7.23 8.2 9.8 9.9 10 22 6.04 6.41 20 22
Intestinal atresia Intestinal atresia Malrotation Intestinal atresia Meconium ileus Meconium ileus Meconium ileus Hirschsprung Imperforate anus Intestinal atresia Intestinal atresia Imperforate anus Hirschsprung Imperforate anus Hirschsprung Malrotation Gastroschisis Imperforate anus Malrotation Malrotation
Small bowel Small bowel Ileocolic Ileocolic Small bowel Small bowel Small bowel Small bowel Coloanal Colonic Colonic Colonic Ileorectal Colonic Small bowel Ileocolic Small bowel Colonic Small bowel Small bowel
Stricture Stricture Stricture Stricture Leak, stricture Leak, stricture Leak, stricture Stricture Stricture Stricture Stricture Leak, stricture Stricture Leak Leak Leak, stricture Stricture Leak Stricture Leak, stricture
mm) mm) mm) mm)
962
G.C. Hintz et al. / Journal of Pediatric Surgery 53 (2018) 959–963
Table 4 Multivariate regression analysis of factors related to anastomotic leak and/or stricture.a Factor
Odds ratio
95% Confidence interval
p Value
Hand-sewn anastomosis Weight b10 kg
1.40 2.58
0.40–4.86 0.68–9.97
0.60 0.16
a
Adjusted for age and anastomotic level.
3. Discussion Stapled intestinal anastomoses have been widely studied in the adult population, and thus their safety and efficacy when compared to hand-sewn anastomoses is un-questioned [1–3]. In fact, there is evidence to suggest that, besides a decrease in operative time, a decreased rate of anastomotic leak may be seen in ileocolic anastomoses specifically, with the use of a stapled technique [4]. In children however, the evidence base is not nearly as strong. Despite stapling devices being first used in pediatric intestinal anastomoses in the late 1960s [5], and more commonly for a variety of indications through the 1990s until present day [6-9], evidence is limited to several case reports and series, and small retrospective reviews [10-16]. Overall, the evidence that is available has shown no difference between stapled and hand-sewn anastomotic techniques, and many studies mirror those in the adult population in that operative times are decreased with the use of a stapler device [10-13,15,16]. In our institution, we found stapling was more likely to be utilized in cases that were emergent in nature, or were classified as contaminated. Each of these factors could potentially increase the risk to the anastomosis. That we did not find a significant increase in anastomotic complications using SA despite it being used more often in these high-risk cases may suggest that the use of stapler devices offers improved outcomes in patients who are especially compromised. Our study failed to show a decrease in operative time when staplers were used however, unlike what several other studies have shown [4,12]. It is possible that the profile of patients undergoing SA, that is of being potentially more ill than those undergoing a hand-sewn anastomosis, contributed to the lack of this finding. We also found that the hand-sewn technique was more often used in cases of imperforate anus (colostomy closure), and intestinal atresia. This is likely due to the fact that these pathologies are treated while patients are yet infants, and therefore have a smaller intestinal caliber. Our findings are not directly comparable to some other pediatric studies, which evaluated both HS and SA techniques in that all of our patients underwent end-to-end anastomoses, whereas other pediatric case series have included anti-mesenteric stapling (e.g. patent ompaholomesenteric duct), and Roux-en-Y anastomoses [11,12]. Arguably, the end-to end anastomosis is a higher risk undertaking [17], and this may contribute to our relatively high complication rate in this group, as compared with these other studies. The technical aspects of stapled intestinal anastomoses are important considerations. The study by Kozlov et al. made use of endoscopic staplers with 2.5 mm staples, and no leaks or strictures occurred [12]. In the study by Sato et al., endoscopic staplers were also used, but with 1.0–1.5 mm staples, and again no anastomotic complications occurred [11]. In all four of the patients in our study with a complication of a stapled anastomosis, 3.5 or 3.8 mm staples were used: in one, an endoscopic stapler with three rows of 3.5 mm staples and, in three, an open linear stapler was used with two rows of 3.8 mm. All patients who had a leak from a stapled anastomosis in our study were 22 kg or less, suggesting that perhaps the staple height was too large for the patient's tissue. Intestinal thickness increases with age and does not reach maximal wall diameter until a patient's third decade of life [18]. Mean intestinal wall diameters in the small bowel are from 0.7 mm (0-4 years) to 1.1 mm (15–19 years); colonic wall thickness averages 1.0 mm (0–4 years) and 1.4 mm (15–19 years). Typically the minimum recommended thickness of tissue for 3.5 mm and 3.8 mm staples is 1.5 mm [19,20] and, given the patients' ages and tissue types (three
out of four were small bowel), it is expected that for the four with an anastomotic complication, the staples were too big, implicating them as a causative factor in their leaking. Since the change of staplers occurred in October of 2015, there have been no anastomotic complications. One possible reason for this is that Covidien Endo staplers have three staple heights for each load, thus covering a wider range of tissue thickness than the stapler used prior to October 2015 in our institution. In addition, the entire end of the stapler is switched after each firing (both the fixed and stepped cartridge anvils), presumably decreasing the chance of any errant staples causing a misfiring of the blade. For the stapled anastomoses that resulted in stricture in our study, another possible technical cause was that the diameter of the resultant anastomosis was insufficient. To explain, Jackson et al. reported two patients who had significant complications related to dilation in the proximal limb of a stapled anastomosis [14]. One possible cause may have been a staple line that was either too long, precipitating torsion or too small, resulting in partial obstruction and resultant proximal dilation. We would suggest it is important to ensure an anastomosis that is sufficient, but not too long in the case of luminal disparity so as to minimize risk of torsion of the anastomosis. One suggestion is that the anastomotic length be wider than the smallest lumen, but no longer than twice the diameter of the largest of the lumens, though to our knowledge no evidence exists to guide this practice. There are several limitations to our study. First, the choice of anastomotic technique was not random and, for infants in particular, SA was only regularly considered by one of the five surgeons at our institution, thus introducing significant bias. Next, there was a small incidence of primary outcomes overall, which implies our population was not adequate to detect statistically significant differences, especially in regard to our regression analysis. In addition, we grouped the patients who had both a stapled or hand-sewn closure of the common channel in the SA group, which arguably may have been better to include as a third unique anastomotic type. The finding that there were no intraabdominal abscesses was a surprise, and potentially suggests that their occurrence was not documented in a way that was captured in our data collection. Another limitation of our study was that the grade of morbidities was not classified; classification would have allowed better comparison of patients who had a complication. 4. Conclusions We present, to our knowledge, one of the largest reported pediatric studies comparing outcomes of hand-sewn and stapled intestinal anastomoses, and we found no significant differences in anastomotic outcomes between the two groups. This suggests that, when permitted by intestinal size, a stapled anastomosis is a safe choice. Ultimately, as already carried out in the adult population, a prospective, multi-centre trial would be most helpful in comparing stapled to hand-sewn anastomotic techniques in the pediatric population, and we would suggest this as a future direction. We also would propose a continued focus on young children or infants, given the higher complication rate seen in these sub-groups. Consequences of an intestinal anastomotic complication can result in substantial morbidity and even death; rigorous evaluation of patient and operative factors is key to optimizing outcomes. Acknowledgements Ms. Maryam Noparast for assistance with statistical analysis. References [1] Goulder Frances. Bowel anastomoses: The theory, the practice and the evidence base. World J Gastrointest Surg 2012;4:208–13. [2] Neutzling CB, Lustosa SA, Da Silva EM, et al. Stapled versus handsewn methods for colorectal anastomosis surgery. Cochrane Database Syst Rev 2012;15:CD003144.
G.C. Hintz et al. / Journal of Pediatric Surgery 53 (2018) 959–963 [3] Slieker JC, Daams F, Mulder IM, et al. Systematic Review of the Technique of Colorectal Anastomosis. JAMA Surg 2013;148:190–201. [4] Choy PY, Bissett IP, Docherty JG, et al. Stapled versus handsewn methods for ileocolic anastomoses. Cochrane Database Syst Rev 2011;7:CD004320. [5] Talbert JL, Seashore JH, Ravitch MM. Evaluation of a modified Duhamel operation for correction of Hirschsprung's disease. Ann Surg 1974;179(5):671. [6] Hedlund H. Colorectal resection and anal anastomosis with an intraluminal stapler in Hirschsprung disease. Pediatr Surg Int 1997;12:142–4. [7] Olguner M, Akgur FM, Ucan B, et al. Laparoscopic appendectomy in children performed using single endoscopic GIA stapler for both mesoappendix and base of appendix. J Pediatr Surg 1998;33(9):1347. [8] Valla JS, Steyaert H, Leculee R, et al. Meckel's diverticulum and laparoscopy of children. What's new? Eur J Pediatr Surg 1998;8:26–8. [9] Mattioli G, Castagnetti M, Gaudullia P, et al. Stapled restorative proctocolectomy in children with refractory ulcerative colitis. J Pediatr Surg 2005;40:1773–9. [10] Wrighton L, Curtis JL, Gollin G. Stapled intestinal anastomoses in infants. J Pediatr Surg 2008;43:2231–4. [11] Sato K, Uchida H, Tanaka Y, et al. Stapled intestinal anastomosis is a simple and reliable method for management of intestinal caliber discrepancy in children. Pediatr Surg Int 2012;28:893–8.
963
[12] Kozlov Y, Novogilov V, Podkamenev A, et al. Stapled bowel anastomoses in newborn surgery. Eur J Pediatr Surg 2013;23:63–6. [13] Mitchell IC, Barber R, Fischer AC, et al. Experience performing 64 consecutive stapled intestinal anastomoses in small children and infants. J Pediatr Surg 2011;46:128–30. [14] Jackson CC, Bettolli MM, De Carli CF, et al. Beware of stapled side-to-side bowel anastomoses in small children. Eur J Pediatr Surg 2007;17:426–7. [15] Powell RW. Stapled intestinal anastomosis in neonates and infants: use of the endoscopic intestinal stapler. J Pediatr Surg 1995;30:195–7. [16] Simmons JD, Gunter III JW, Manley JD, et al. Stapled intestinal anastomosis in neonates. Am Surg 2010;76:644–6. [17] Bagolan P, Nappo S, Trucchi A, et al. Neonatal intestinal obstruction: reducing shortterm complications by surgical refinements. Eur J Pediatr Surg 1996;6:354–7. [18] Haber HP, Stern M. Intestinal ultrasonography in children and young adults: bowel wall thickness is age dependent. J Ultrasound Med 2000;19:315–21. [19] “Proximate Linear Cutters.” Ethicon US, LLC. http://www.ethicon.com/healthcareprofessionals/products/staplers/linear-cutters/proximate-linear-cutters, Accessed date: 13 July 2017. [20] “Endo GIA Reloads with Tri-Spale Technology, Technical Brochure.” Medtronic US. http://www.medtronic.com/content/dam/covidien/library/us/en/product/surgicalstapling/endo-gia-reloads-technical-brochure.pdf, Accessed date: 13 July 2017.