doi:10.1111/codi.12490
Systematic review
Treatment with corticosteroids and the risk of anastomotic leakage following lower gastrointestinal surgery: a literature survey T. F. Eriksen, C. B. Lassen and I. G€ ogenur Department of Surgery, Herlev Hospital, University of Copenhagen, Herlev, Denmark Received 1 July 2013; accepted 29 September 2013; Accepted Article online 12 November 2013
Abstract Aim Background Anastomotic leakage is a serious complication in colorectal surgery. Treatment with corticosteroids is known to impair wound healing but their effect on the healing of a colorectal anastomosis remains unclear, and studies have reported conflicting results. Objective The aim of this study was to evaluate the current evidence regarding the effect of corticosteroids on the risk of anastomotic leakage following colorectal surgery. Method Search strategy A systematic review was conducted following a search of PubMed and Embase. Selection criteria Inclusion criteria were studies published in English and involving humans. A minimum cohort of 50 patients was required and anastomoses involving the ileum, colon and rectum were included. Studies that investigated corticosteroids as a risk factor for anastomotic leakage were included regardless of the duration and the dose of corticosteroids. Data Collection and analysis A comparison was conducted between anastomotic leakage in noncorticosteroid- and cortico-
Introduction Anastomotic leakage (AL) is a major cause of morbidity and mortality after gastrointestinal surgery and is the complication of most concern in colorectal surgery. The occurrence of a leak results in increased short-term morbidity and mortality. In addition, there are also long-term effects in patients with colorectal cancer where the AL results in increased risk of local recurrence [1,2]. Several patient-related factors, as well as those related to the surgical procedure, have been invesCorrespondence to: Tina Fransg ard Eriksen, Department of Surgery, Herlev Hospital, University of Copenhagen, Herlev Ringvej 75, 2730 Herlev, Denmark. E-mail: [email protected]
O154
steroid-treated patients. The main outcome measure was the risk of anastomotic leakage. Results Twelve studies with a total of 9564 patients were included in the review. In total, 1034 patients received corticosteroids in the preoperative period, and 344 patients were diagnosed with anastomotic leakage, 70 of whom had received corticosteroids. Six of the 12 studies showed an increased risk for anastomotic leakage in the corticosteroid group. Overall, the anastomotic leakage rate was 6.77% (95% CI: 5.48–9.06) in the corticosteroid group and 3.26% (95% CI: 2.94–3.58) in the noncorticosteroid group. Conclusion Caution should be shown in patients scheduled for lower gastrointestinal surgery with anastomosis. Keywords anastomotic leakage, corticosteroids, gastrointestinal surgery colorectal anastomoses
tigated in the search for modifiable risk factors for AL. The single factor consistently shown to predict leakage is a low rectal anastomosis [3]. Much research has been carried out to determine additional risk factors for AL and those reported include radiotherapy, male sex [4,5], smoking [6], perioperative treatment with cyclooxygenase 2 (COX-2)-selective nonsteriodal anti-inflammatory drugs (NSAIDs) [7], blood transfusion, obesity, atherosclerosis [8] and inflammatory bowel disease [3]. Better understanding of the risk factors for AL will not only improve the management of high-risk patients but will also allow for preventive treatments, such as creating a protective stoma. Systemic corticosteroids are widely used in the treatment of patients with autoimmune and chronic inflammatory diseases because of
Colorectal Disease ª 2013 The Association of Coloproctology of Great Britain and Ireland. 16, O154–O160
T. F. Eriksen et al.
their potent immunomodulatory action. Unfortunately, corticosteroids have been linked to numerous, and sometimes severe, complications. Although corticosteroids are known to impair wound healing, their influence on the healing of colorectal anastomoses remains unclear and studies have reported conflicting results [6,8–18]. The aim of this systematic review was to evaluate the current evidence regarding the effect of preoperative treatment with corticosteroids on the risk of AL following colorectal surgery.
Method This review was conducted according to the Preferred Reporting Items for Systematic Reviews and MetaAnalyses (PRISMA) [19]. A systematic literature search of the PubMed and Embase databases was performed. The PubMed database search was performed by combining the following keywords, both as text words and as Medical Subject Headings (MeSH terms): (Anastomotic Leak OR Anastomotic leaks OR Anastomotic leakage OR Anastomotic leakages) AND (corticosteroid OR corticosteroids OR adrenal cortex hormones OR adrenal cortex hormone). The Embase database search was performed combining the following keywords: (Anastomotic Leak OR Anastomotic leaks OR Anastomotic leakage OR Anastomotic leakages) AND (corticosteroid OR corticosteroids OR adrenal cortex hormones OR adrenal cortex hormone) (Fig. 1). Two investigators independently reviewed the titles and abstracts of all citations identified in the literature search. Only articles published in English and involving humans were included in the review. Potentially relevant studies were retrieved and selection criteria were applied. Only studies that included patients with anastomoses involving the ileum, colon or rectum were included. Both laparoscopic and open surgeries were included. A minimum cohort of 50 patients was required for all studies considered. Studies that investigated corticosteroids as a risk factor for AL were included regardless of the duration and the dose of corticosteroids. Data extracted from the selected studies included: year of publication; gender ratio; mean age at the time of surgery; study design; American Society of Anesthesiology (ASA) score; and pathology. Details regarding the surgery (emergency or elective) and type of anastomosis (ileoileal, ileocolic, colocolic, colorectal or ileal pouchanal anastomosis (IPAA) were recorded. In addition, data on corticosteroid use (number of patients, definition of corticosteroid use, duration of corticosteroid use and mean dose of corticosteroids) were collected. If data were not shown, the authors were contacted by e-mail and asked to provide the missing information.
Corticosteroids and anastomotic leakage
All patients in the included studies had an anastomosis; however, unfortunately the proportion of patients receiving a covering stoma was not available. AL was defined as a clinically apparent leakage and the following data were collected: the total number of patients with AL; the number of patients with AL in the corticosteroid group; and the number of patients with AL in the noncorticosteroid group. The rates of infectious complications were also recorded. Statistical analyses were performed only on the data extracted from the selected studies. Basic descriptive statistics (simple counts and percentages) were used to summarize the patients, study and outcome data. Overall anastomotic leak rate was reported with 95% CI. The OR was calculated for each study. Studies included in this review are not sufficiently similar to make it reasonable to combine their results and conduct a meta-analysis. This is especially because of the differences in definition of corticosteroid use and duration of treatment, which ranged from 60 to 90 min preoperatively to 44 years.
Results Patient characteristics
In total 9465 patients were included in the studies. The mean age was 59 years and men comprised 49% of the population. The ASA classification was not reported in five studies [9,12–14,18], whereas in the others it mainly ranged from ASA 1 to ASA 4 [6,8,10,11,15–17]. Three studies included only patients with inflammatory bowel disease (IBD) [9,13,14] (two of which included only patients with Crohn’s disease [9,13]), six included patients with malignant as well as benign pathology [6,8,11,16–18] and two included only patients with colorectal cancer [10,15]. One study solely included patients with IBD and familial adenomatous polyposis (FAP) [12] (Table 1). Corticosteroids
In total, 1034 patients were in the corticosteroid-treated group and 8410 patients did not receive corticosteroids. Although all the studies had different definitions of corticosteroid use, three did not report their definition [9,16,18]. Most studies defined the use of corticosteroids as treatment for longer than 1 month [11,13,14]. The duration of corticosteroid treatment ranged from 60 to 90 min preoperatively [10] and for up to 44 years before surgery [11]. Only Tresallet et al. [11] and Vignali et al. [10] reported the mean dosage of corticosteroids used (10 mg/day and 30 mg/day, respectively). Three studies divided the steroid-treated patients into high-dosage
Colorectal Disease ª 2013 The Association of Coloproctology of Great Britain and Ireland. 16, O154–O160
O155
T. F. Eriksen et al.
Corticosteroids and anastomotic leakage
Citations identified for screeninig (n = 94)
Rejected (met exclusion criteria) (n = 63)
Studies retrieved (n = 31) Duplicates removed (n = 7)
Studies reviewed (n = 24)
Rejected (did not meet inclusion criteria) (n = 10)
Studies for further review (n = 14)
Additional studies identified from bibliographies and related cirations (n = 5)
Studies for further review (n = 19)
Rejected, insufficient data (n = 7) Figure 1 Preferred reporting items for systematic reviews and meta-analyses (PRISMA) flow chart.
Studies included (n = 12)
(≥ 20 mg/day of corticosteroid) and low-dosage (< 20 mg/day of corticosteroid) groups [12–14]. Type of surgery and anastomoses
The patients in the studies underwent elective as well as emergency surgery. The procedures were performed laparoscopically and/or open [6,8–18], and the anastomoses were hand sewn as well as stapled (Table 2). The types of anastomoses were ileoileal, ileocolic, ileorectal, colocolic or colorectal, and only two studies included IPAA [12,14]. Four studies included only colocolic and/or colorectal anastomoses (Table 2) [6,8,11,15]. The surgery included right- and left-sided colectomies, anterior resection, segmental resection and subtotal- or total colectomies (Table 2). All studies included different types of surgery, except for Tresallet et al. [11] (who included only patients undergoing left-sided hemicolectomy) and Lake et al. [12] (who included only patients having a total colectomy).
O156
Anastomotic leakage
Of the 9465 patients included in this review, 344 were diagnosed with AL. The different studies reported a leakage rate ranging from 1.98 to 14.16%. Twohundred and seventy-four patients diagnosed with an AL were in the nonsteroid group and 70 patients undergoing steroid treatment had AL (Table 1). The total leak rate was 3.63%. The leak rate in the corticosteroid group was 6.19% (95% CI: 5.48–9.06) and that in the noncorticosteroid group was 3.33% (95% CI: 2.94–3.58). Tay et al. [9], El-Hussana et al. [13], Ziv et al. [14] and Lake et al. [12] included only patients diagnosed with IBD and reported leakage rates of 2.38–7.86%. Konishi et al. [15] and Vignali et al. [10] included only patients with colorectal cancer and reported leakage rates of 2.81% and 7.69%, respectively. The remaining six studies included patients with benign and malignant diagnoses and reported leakage rates of 1.98–14.16% [6,8,11,16–18] (Table 1).
Colorectal Disease ª 2013 The Association of Coloproctology of Great Britain and Ireland. 16, O154–O160
Colorectal Disease ª 2013 The Association of Coloproctology of Great Britain and Ireland. 16, O154–O160 1.07–27.24
5.41
0.95
0.12–7.49
9
11
17 374
53 553
11 2
65 M
62 B + M, not IBD
12 1
257/134
325/281
Konishi et al. [15] (n = 391)
2.08–14.33
5.46
22
28 6
41 723
69 B+M
359/405
Golub et al. [16] (n = 764)
2
2
0.13–7.69
1.00
4
26 26
* M
31/21
Vignali et al. [10] (n = 52)
0.21–8.10
1.29
2
5 3
54 46
* IBD + FAP
*
Lake et al. [12] (n = 100)
3.12–27.43
9.24
12
19 7
20 218
65 B+M
144/115
Sliecker et al. [8] (n = 259)
1.23–3.96
2.20
134
153 19
214 4909
67 B+M
2410/2713
Ziegler et al. [17] (n = 5123)
0.84–8.24
2.63
4
16 12
361 310
34 UC
359/312
Ziv et al. [14] (n = 671)
0.03–3.18
0.32
3
4 1
50 50
* Cr
*
Tay et al. [9] (n = 100)
0.53–2.46
1.15
16
29 13
154 215
37 Cr
120/249
El-Hussuna et al. [13] (n = 369)
Not available Not available
33
33 0
3 230
68 B+M
137/96
Richards et al. [6] (n = 233)
1.01–9.15
3.04
26
30 4
41 756
63 B+M
386/411
Lujan et al. [18] (n = 797)
AL, anastomotic leakage; B, benign; Cr, Crohn’s disease; FAP, familial adenomatous polyposis; IBD, inflammatory bowel disease; M, malignant; UC, ulcerative colitis. *Not reported.
95% CI
AL Total AL (n) AL corticosteroid (n) AL noncorticosteroid (n) Statistical analysis OR
Gender (men/ women) Mean age (years) Pathology/ diagnosis Corticosteroid treatment Corticosteroid (n) Noncorticosteroid (n)
Variables
Tresallet et al. [11] (n = 606)
Table 1 Patient characteristics, steroid treatment and anastomotic leakage in each study.
274
344 70
1034 8410
Total, all studies (n = 9465)
T. F. Eriksen et al.
Corticosteroids and anastomotic leakage
O157
O158
Tresallet et al. [11]
x
x
*Intestinal anastomoses. †Not reported.
Surgery Elective Emergency Right-sided colectomy Left-sided colectomy Anterior resection Colectomy Segmental resection Other
Type of anastomosis Ileoileal Ileocolic Ileorectal Colorectal x Colocolic Ileal pouch anal (IPAA) Handsewn x Stapled x
Variable
x
x
x
x x
x
x
x
x x x
x x
x x
x
x x
x x
Golub et al. [16]
x
Konishi et al. [15]
x x x
x x
x
x
Vignali et al. [10]
Table 2 Types of surgery and anastomosis reported in each study.
x
x x
x
Lake et al. [12]
x
x
x
x x
x x
Sliecker et al. [8]
x
x x x
x x
x x
x
Ziegler et al. [17]
x
x
x x
x
Ziv et al. [14]
x
x
x x x
Tay et al. [9]
x
x x
x
x x x
x x
x x x x x
El-Hussuna et al. [13]
x
x x
x x
x
Richards et al. [6]
†
x x
x x
*
Lujan et al. [18]
Corticosteroids and anastomotic leakage T. F. Eriksen et al.
Colorectal Disease ª 2013 The Association of Coloproctology of Great Britain and Ireland. 16, O154–O160
T. F. Eriksen et al.
Discussion There was an apparent increased risk of AL in patients undergoing elective lower gastrointestinal surgery when receiving treatment with corticosteroids before surgery. The studies reported AL rates ranging from 1.98 to 14.16%, which corresponds with the leakage rate reported previously in the literature [3]. Although corticosteroids are generally thought to impair intestinal anastomotic healing, this effect has never been proven unequivocally in clinical studies [20]. Some experimental studies have concluded that corticosteroid use impairs the healing of colonic anastomosis in a rat model [21–23], but others have reported no such effect [20]. Baca et al. [23] conducted an experimental study to investigate how long- and short-term corticosteroid administration in high and low doses modulated the course of the healing process for colonic anastomoses. The study indicated that high and low doses of methylprednisolone impaired the healing of colon anastomosis in rats, regardless of whether the methylprednisolone was administered over a long or a short preoperative period. The most prominent negative effects were associated with long-term, high-dose corticosteroid administration. The effect on the anastomosis was evaluated by measuring bursting pressure, hydroxyproline levels and histopathology findings. However, Mastboom et al. [20] did not find that short-term treatment with methylprednisolone lowered the anastomotic bursting pressures. Comparison of the hydroxyproline content yielded no significant difference between control and methylprednisolone groups. Furst et al. also conducted a study designed to determine the effects of corticosteroids on healing colonic anastomoses in a rat model. The effects were determined by measuring bursting pressure, and the study concluded that it was significantly lowered in the group undergoing shortterm treatment at 6, 8 and 20 days after surgery [21]. The definition, duration and dose of corticosteroid treatment varied widely among the studies, and several failed to report the mean dosage and duration of corticosteroid treatment. Accordingly, it is very difficult to compare the different studies. Three of the studies [12–14] divided patients receiving corticosteroids into high- or low-dose groups. In all three the reported risk of AL was higher in the high-dose corticosteroidtreated group. Hence, it seems that a higher dose of corticosteroid results in a higher risk of AL; however, because of the limited number of studies included in this review, this cannot be stated with certainty. The duration of corticosteroid treatment ranged from 60 min preoperatively to 44 years. The studies
Corticosteroids and anastomotic leakage
with the longest, as well as the shortest, duration of corticosteroid treatment (Vignali et al. [10] and Tresallet et al. [11] respectively) failed to show a significantly increased risk of AL in the corticosteroid-treated group. Sliecker et al. [8] also examined perioperative and longterm corticosteroid treatment, and found a significantly increased risk of AL in both corticosteroid-treated groups. The studies show conflicting evidence regarding the duration of use, and therefore it is not possible to clarify if the duration of corticosteroid treatment has an influence on the risk of AL. Studies included in this review are not sufficiently similar to make it reasonable to combine their results and conduct a meta-analysis. This is especially because of the differences in definition of corticosteroid use and duration of treatment Therefore, this review has obvious limitations. Furthermore, there was only one randomized controlled trial [10] to support our findings, but that study included only 52 patients and excluded those undergoing long-term treatment with corticosteroid. The remaining studies were either retrospective or prospective. Generally, the studies in our review had the limitation of including a relatively small number of patients. An exception was the study of Ziegler et al. [17], which included 5123 patients, 214 of whom were treated with steroids. Ziegler et al. defined corticosteroid use as regular administration of oral or parenteral corticosteroids for a chronic medical condition in the 30 days before surgery. This did not include shortcourse corticosteroids (i.e. treatment for ≤ 10 days). They found that 153 patients had AL: 19 (8.89%) were from the corticosteroid group compared with 134 (2.73%) from the noncorticosteroid group. With these limitations in mind it was interesting to see that the rates of AL are different, with no overlap in the CIs of the two different rates of AL found in the corticosteroid group (6.19%; 95% CI: 5.48–9.06) and the control group (3.33%; 95% CI: 2.94–3.58). However, it will never be possible to perform a randomized clinical trial focusing on complication differences in patients receiving long-term corticosteroid treatment. Our data warrant caution in the surgical treatment, with lower gastrointestinal surgery and an anastomosis, of patients receiving corticosteroids. In conclusion, treatment with corticosteroids may increase the risk of AL in patients undergoing gastrointestinal surgery; however, further investigations are needed before a final conclusion can be made. It is obviously very difficult to conduct a large randomized placebo-controlled trial of patients receiving long- or short-term corticosteroid treatment or placebo. Nationwide database studies with prospective evaluation of
Colorectal Disease ª 2013 The Association of Coloproctology of Great Britain and Ireland. 16, O154–O160
O159
Corticosteroids and anastomotic leakage
outcomes such as anastomotic leak and registration of long-term corticosteroid use will give more valid results.
Acknowledgements The authors thank T. W. Klausen, University of Copenhagen, Herlev Hospital, for statistical advice, and Richard Shock for editing services.
Author contributions All authors have provided substantial contributions to the conception and design, analysis and interpretation of data. All authors have drafted the article and revised it critically for intellectual content, and approved the final work to be published.
Conflicts of interest None.
References 1 Bell SW, Walker KG, Rickard MJ et al. Anastomotic leakage after curative anterior resection results in a higher prevalence of local recurrence. Br J Surg 2003; 90: 1261–6. 2 Branagan G, Finnis D. Prognosis after anastomotic leakage in colorectal surgery. Dis Colon Rectum 2005; 48: 1021–6. 3 Chambers WM, Mortensen NJ. Postoperative leakage and abscess formation after colorectal surgery. Best Pract Res Clin Gastroenterol 2004; 18: 865–80. € k O, Andersson M, Ruteg 4 Matthiessen P, Hallb€ oo ard J, Sj€ odahl R. Risk factors for anastomotic leakage after anterior rection of the rectum. Colorectal Dis 2004; 6: 462–9. 5 Rullier E, Laurent C, Garrelon JL, Michel P, Saric J, Parneix M. Risk factors for anastomotic leakage after resection of rectal cancer. Br J Surg 1998; 85: 355–8. 6 Richards CH, Campbell V, Ho C et al. Smoking is a major risk factor for anastomotic leak in patients undergoing low anterior resection. Colorectal Dis 2012; 14: 628–33. 7 Klein M, G€ ogenur I, Rosenberg J. Postoperative use of non-steroidal anti-inflammatory drugs in patients with anastomotic leakage requiring reoperation after colorectal resection: cohort study based on prospective data. BMJ 2012; 345: e6166. 8 Sliecker JC, Komen N, Mannaerts GH et al. Long-term and perioperative corticosteroids in anastomotic leakage. A prospective study of 259 left-sided colorectal anastomoses. Arch Surg 2012; 147: 447–552. 9 Tay GS, Binion DG, Eastwood D, Otterson MF. Multivariate analysis suggests improved perioperative outcome in Crohn’s disease patients receiving immunomodulator therapy after segmental resection and/or strictureplasty. Surgery 2003; 134: 565–72.
O160
T. F. Eriksen et al.
10 Vignali A, Di Palo S, Orsenigo E, Ghirardelli L, Radaelli G, Staudacher C. Effect of prednisolone on local and systemic response in laparoscopic vs. open colon surgery: a randomized double-blind, placebo-controlled trial. Dis Colon Rectum 2009; 52: 1080–8. 11 Tresallet C, Royer B, Godiris-Petit G, Menegaux F. Effect of systemic corticosteroids on elective left-sided colorectal resection with colorectal anastomosis. Am J Surg 2008; 195: 447–51. 12 Lake JP, Firoozmand E, Kang JC et al. Effect of high-dose steroids on anastomotic complications after proctocolectomy with ileal pouch-anal anastomosis. J Gastrointest Surg 2004; 8: 547–51. 13 El-Hussuna A, Andersen J, Bisgaard T et al. Biologic treatment or immunomodulation is not associated with postoperative anastomotic complications in abdominal surgery for Crohn’s disease. Scand J Gastroenterol 2012; 47: 662–8. 14 Ziv Y, Church JM, Fazio VW, King T, Lavery I. Effect of systemic steroids on ileal pouch-anal anastomosis in patients with ulcerative colitis. Dis Colon Rectum 1996; 39: 504–8. 15 Konishi T, Watanabe T, Kishimoto J, Nagawa H. Risk factors for anastomotic leakage after surgery for colorectal cancer: results of prospective surveillance. J Am Coll Surg 2006; 202: 439–44. 16 Golub R, Golub RW, Cantu R, Stein DH. A multivariate analysis of factors contributing to leakage of intestinal anastomoses. J Am Coll Surg 1997; 184: 364–72. 17 Ziegler MA, Catto JA, Riggs TW, Gates ER, Grodsky MB, Wasvary J. Risk factors for anastomotic leak and mortality in diabetic patients undergoing colectomy. Arch Surg 2012; 147: 600–5. 18 Lujan JJ, Nemeth Z, Barratt-Stopper PA, Bustami R, Koshenkov VP, Rolandelli RH. Factors influencing the outcome of intestinal anastomosis. Am Surg 2011; 77: 1169– 75. 19 Moher D, Liberati A, Tetzlaff J, Altman DG; PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med 2009; 18: 264–9. 20 Mastboom WJ, Hendriks T, de Man BM, de Boer HH. Influence of methylprednisolone on the healing of intestinal anastomoses in rats. Br J Surg 1991; 78: 54–6. 21 Furst MB, Stromberg BV, Blatchford GJ, Christensen MA, Thorson AG. Colonic anastomoses: bursting strength after corticosteroid treatment. Dis Colon Rectum 1994; 37: 12– 5. 22 Polat A, Nayci A, Polat G, Aks€ oyek S. Dexamethasone downregulates endothelial expression of intercellular adhesion molecule and impairs the healing of bowel anastomoses. Eur J Surg 2002; 168: 500–6. 23 Baca B, Ozben V, Boler DE et al. Effect of corticosteroid dose and duration of administration on colonic anastomosis. Inflamm Bowel Dis 2010; 16: 2162–7.
Colorectal Disease ª 2013 The Association of Coloproctology of Great Britain and Ireland. 16, O154–O160
Copyright of Colorectal Disease is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
Systematic review
Treatment with corticosteroids and the risk of anastomotic leakage following lower gastrointestinal surgery: a literature survey T. F. Eriksen, C. B. Lassen and I. G€ ogenur Department of Surgery, Herlev Hospital, University of Copenhagen, Herlev, Denmark Received 1 July 2013; accepted 29 September 2013; Accepted Article online 12 November 2013
Abstract Aim Background Anastomotic leakage is a serious complication in colorectal surgery. Treatment with corticosteroids is known to impair wound healing but their effect on the healing of a colorectal anastomosis remains unclear, and studies have reported conflicting results. Objective The aim of this study was to evaluate the current evidence regarding the effect of corticosteroids on the risk of anastomotic leakage following colorectal surgery. Method Search strategy A systematic review was conducted following a search of PubMed and Embase. Selection criteria Inclusion criteria were studies published in English and involving humans. A minimum cohort of 50 patients was required and anastomoses involving the ileum, colon and rectum were included. Studies that investigated corticosteroids as a risk factor for anastomotic leakage were included regardless of the duration and the dose of corticosteroids. Data Collection and analysis A comparison was conducted between anastomotic leakage in noncorticosteroid- and cortico-
Introduction Anastomotic leakage (AL) is a major cause of morbidity and mortality after gastrointestinal surgery and is the complication of most concern in colorectal surgery. The occurrence of a leak results in increased short-term morbidity and mortality. In addition, there are also long-term effects in patients with colorectal cancer where the AL results in increased risk of local recurrence [1,2]. Several patient-related factors, as well as those related to the surgical procedure, have been invesCorrespondence to: Tina Fransg ard Eriksen, Department of Surgery, Herlev Hospital, University of Copenhagen, Herlev Ringvej 75, 2730 Herlev, Denmark. E-mail: [email protected]
O154
steroid-treated patients. The main outcome measure was the risk of anastomotic leakage. Results Twelve studies with a total of 9564 patients were included in the review. In total, 1034 patients received corticosteroids in the preoperative period, and 344 patients were diagnosed with anastomotic leakage, 70 of whom had received corticosteroids. Six of the 12 studies showed an increased risk for anastomotic leakage in the corticosteroid group. Overall, the anastomotic leakage rate was 6.77% (95% CI: 5.48–9.06) in the corticosteroid group and 3.26% (95% CI: 2.94–3.58) in the noncorticosteroid group. Conclusion Caution should be shown in patients scheduled for lower gastrointestinal surgery with anastomosis. Keywords anastomotic leakage, corticosteroids, gastrointestinal surgery colorectal anastomoses
tigated in the search for modifiable risk factors for AL. The single factor consistently shown to predict leakage is a low rectal anastomosis [3]. Much research has been carried out to determine additional risk factors for AL and those reported include radiotherapy, male sex [4,5], smoking [6], perioperative treatment with cyclooxygenase 2 (COX-2)-selective nonsteriodal anti-inflammatory drugs (NSAIDs) [7], blood transfusion, obesity, atherosclerosis [8] and inflammatory bowel disease [3]. Better understanding of the risk factors for AL will not only improve the management of high-risk patients but will also allow for preventive treatments, such as creating a protective stoma. Systemic corticosteroids are widely used in the treatment of patients with autoimmune and chronic inflammatory diseases because of
Colorectal Disease ª 2013 The Association of Coloproctology of Great Britain and Ireland. 16, O154–O160
T. F. Eriksen et al.
their potent immunomodulatory action. Unfortunately, corticosteroids have been linked to numerous, and sometimes severe, complications. Although corticosteroids are known to impair wound healing, their influence on the healing of colorectal anastomoses remains unclear and studies have reported conflicting results [6,8–18]. The aim of this systematic review was to evaluate the current evidence regarding the effect of preoperative treatment with corticosteroids on the risk of AL following colorectal surgery.
Method This review was conducted according to the Preferred Reporting Items for Systematic Reviews and MetaAnalyses (PRISMA) [19]. A systematic literature search of the PubMed and Embase databases was performed. The PubMed database search was performed by combining the following keywords, both as text words and as Medical Subject Headings (MeSH terms): (Anastomotic Leak OR Anastomotic leaks OR Anastomotic leakage OR Anastomotic leakages) AND (corticosteroid OR corticosteroids OR adrenal cortex hormones OR adrenal cortex hormone). The Embase database search was performed combining the following keywords: (Anastomotic Leak OR Anastomotic leaks OR Anastomotic leakage OR Anastomotic leakages) AND (corticosteroid OR corticosteroids OR adrenal cortex hormones OR adrenal cortex hormone) (Fig. 1). Two investigators independently reviewed the titles and abstracts of all citations identified in the literature search. Only articles published in English and involving humans were included in the review. Potentially relevant studies were retrieved and selection criteria were applied. Only studies that included patients with anastomoses involving the ileum, colon or rectum were included. Both laparoscopic and open surgeries were included. A minimum cohort of 50 patients was required for all studies considered. Studies that investigated corticosteroids as a risk factor for AL were included regardless of the duration and the dose of corticosteroids. Data extracted from the selected studies included: year of publication; gender ratio; mean age at the time of surgery; study design; American Society of Anesthesiology (ASA) score; and pathology. Details regarding the surgery (emergency or elective) and type of anastomosis (ileoileal, ileocolic, colocolic, colorectal or ileal pouchanal anastomosis (IPAA) were recorded. In addition, data on corticosteroid use (number of patients, definition of corticosteroid use, duration of corticosteroid use and mean dose of corticosteroids) were collected. If data were not shown, the authors were contacted by e-mail and asked to provide the missing information.
Corticosteroids and anastomotic leakage
All patients in the included studies had an anastomosis; however, unfortunately the proportion of patients receiving a covering stoma was not available. AL was defined as a clinically apparent leakage and the following data were collected: the total number of patients with AL; the number of patients with AL in the corticosteroid group; and the number of patients with AL in the noncorticosteroid group. The rates of infectious complications were also recorded. Statistical analyses were performed only on the data extracted from the selected studies. Basic descriptive statistics (simple counts and percentages) were used to summarize the patients, study and outcome data. Overall anastomotic leak rate was reported with 95% CI. The OR was calculated for each study. Studies included in this review are not sufficiently similar to make it reasonable to combine their results and conduct a meta-analysis. This is especially because of the differences in definition of corticosteroid use and duration of treatment, which ranged from 60 to 90 min preoperatively to 44 years.
Results Patient characteristics
In total 9465 patients were included in the studies. The mean age was 59 years and men comprised 49% of the population. The ASA classification was not reported in five studies [9,12–14,18], whereas in the others it mainly ranged from ASA 1 to ASA 4 [6,8,10,11,15–17]. Three studies included only patients with inflammatory bowel disease (IBD) [9,13,14] (two of which included only patients with Crohn’s disease [9,13]), six included patients with malignant as well as benign pathology [6,8,11,16–18] and two included only patients with colorectal cancer [10,15]. One study solely included patients with IBD and familial adenomatous polyposis (FAP) [12] (Table 1). Corticosteroids
In total, 1034 patients were in the corticosteroid-treated group and 8410 patients did not receive corticosteroids. Although all the studies had different definitions of corticosteroid use, three did not report their definition [9,16,18]. Most studies defined the use of corticosteroids as treatment for longer than 1 month [11,13,14]. The duration of corticosteroid treatment ranged from 60 to 90 min preoperatively [10] and for up to 44 years before surgery [11]. Only Tresallet et al. [11] and Vignali et al. [10] reported the mean dosage of corticosteroids used (10 mg/day and 30 mg/day, respectively). Three studies divided the steroid-treated patients into high-dosage
Colorectal Disease ª 2013 The Association of Coloproctology of Great Britain and Ireland. 16, O154–O160
O155
T. F. Eriksen et al.
Corticosteroids and anastomotic leakage
Citations identified for screeninig (n = 94)
Rejected (met exclusion criteria) (n = 63)
Studies retrieved (n = 31) Duplicates removed (n = 7)
Studies reviewed (n = 24)
Rejected (did not meet inclusion criteria) (n = 10)
Studies for further review (n = 14)
Additional studies identified from bibliographies and related cirations (n = 5)
Studies for further review (n = 19)
Rejected, insufficient data (n = 7) Figure 1 Preferred reporting items for systematic reviews and meta-analyses (PRISMA) flow chart.
Studies included (n = 12)
(≥ 20 mg/day of corticosteroid) and low-dosage (< 20 mg/day of corticosteroid) groups [12–14]. Type of surgery and anastomoses
The patients in the studies underwent elective as well as emergency surgery. The procedures were performed laparoscopically and/or open [6,8–18], and the anastomoses were hand sewn as well as stapled (Table 2). The types of anastomoses were ileoileal, ileocolic, ileorectal, colocolic or colorectal, and only two studies included IPAA [12,14]. Four studies included only colocolic and/or colorectal anastomoses (Table 2) [6,8,11,15]. The surgery included right- and left-sided colectomies, anterior resection, segmental resection and subtotal- or total colectomies (Table 2). All studies included different types of surgery, except for Tresallet et al. [11] (who included only patients undergoing left-sided hemicolectomy) and Lake et al. [12] (who included only patients having a total colectomy).
O156
Anastomotic leakage
Of the 9465 patients included in this review, 344 were diagnosed with AL. The different studies reported a leakage rate ranging from 1.98 to 14.16%. Twohundred and seventy-four patients diagnosed with an AL were in the nonsteroid group and 70 patients undergoing steroid treatment had AL (Table 1). The total leak rate was 3.63%. The leak rate in the corticosteroid group was 6.19% (95% CI: 5.48–9.06) and that in the noncorticosteroid group was 3.33% (95% CI: 2.94–3.58). Tay et al. [9], El-Hussana et al. [13], Ziv et al. [14] and Lake et al. [12] included only patients diagnosed with IBD and reported leakage rates of 2.38–7.86%. Konishi et al. [15] and Vignali et al. [10] included only patients with colorectal cancer and reported leakage rates of 2.81% and 7.69%, respectively. The remaining six studies included patients with benign and malignant diagnoses and reported leakage rates of 1.98–14.16% [6,8,11,16–18] (Table 1).
Colorectal Disease ª 2013 The Association of Coloproctology of Great Britain and Ireland. 16, O154–O160
Colorectal Disease ª 2013 The Association of Coloproctology of Great Britain and Ireland. 16, O154–O160 1.07–27.24
5.41
0.95
0.12–7.49
9
11
17 374
53 553
11 2
65 M
62 B + M, not IBD
12 1
257/134
325/281
Konishi et al. [15] (n = 391)
2.08–14.33
5.46
22
28 6
41 723
69 B+M
359/405
Golub et al. [16] (n = 764)
2
2
0.13–7.69
1.00
4
26 26
* M
31/21
Vignali et al. [10] (n = 52)
0.21–8.10
1.29
2
5 3
54 46
* IBD + FAP
*
Lake et al. [12] (n = 100)
3.12–27.43
9.24
12
19 7
20 218
65 B+M
144/115
Sliecker et al. [8] (n = 259)
1.23–3.96
2.20
134
153 19
214 4909
67 B+M
2410/2713
Ziegler et al. [17] (n = 5123)
0.84–8.24
2.63
4
16 12
361 310
34 UC
359/312
Ziv et al. [14] (n = 671)
0.03–3.18
0.32
3
4 1
50 50
* Cr
*
Tay et al. [9] (n = 100)
0.53–2.46
1.15
16
29 13
154 215
37 Cr
120/249
El-Hussuna et al. [13] (n = 369)
Not available Not available
33
33 0
3 230
68 B+M
137/96
Richards et al. [6] (n = 233)
1.01–9.15
3.04
26
30 4
41 756
63 B+M
386/411
Lujan et al. [18] (n = 797)
AL, anastomotic leakage; B, benign; Cr, Crohn’s disease; FAP, familial adenomatous polyposis; IBD, inflammatory bowel disease; M, malignant; UC, ulcerative colitis. *Not reported.
95% CI
AL Total AL (n) AL corticosteroid (n) AL noncorticosteroid (n) Statistical analysis OR
Gender (men/ women) Mean age (years) Pathology/ diagnosis Corticosteroid treatment Corticosteroid (n) Noncorticosteroid (n)
Variables
Tresallet et al. [11] (n = 606)
Table 1 Patient characteristics, steroid treatment and anastomotic leakage in each study.
274
344 70
1034 8410
Total, all studies (n = 9465)
T. F. Eriksen et al.
Corticosteroids and anastomotic leakage
O157
O158
Tresallet et al. [11]
x
x
*Intestinal anastomoses. †Not reported.
Surgery Elective Emergency Right-sided colectomy Left-sided colectomy Anterior resection Colectomy Segmental resection Other
Type of anastomosis Ileoileal Ileocolic Ileorectal Colorectal x Colocolic Ileal pouch anal (IPAA) Handsewn x Stapled x
Variable
x
x
x
x x
x
x
x
x x x
x x
x x
x
x x
x x
Golub et al. [16]
x
Konishi et al. [15]
x x x
x x
x
x
Vignali et al. [10]
Table 2 Types of surgery and anastomosis reported in each study.
x
x x
x
Lake et al. [12]
x
x
x
x x
x x
Sliecker et al. [8]
x
x x x
x x
x x
x
Ziegler et al. [17]
x
x
x x
x
Ziv et al. [14]
x
x
x x x
Tay et al. [9]
x
x x
x
x x x
x x
x x x x x
El-Hussuna et al. [13]
x
x x
x x
x
Richards et al. [6]
†
x x
x x
*
Lujan et al. [18]
Corticosteroids and anastomotic leakage T. F. Eriksen et al.
Colorectal Disease ª 2013 The Association of Coloproctology of Great Britain and Ireland. 16, O154–O160
T. F. Eriksen et al.
Discussion There was an apparent increased risk of AL in patients undergoing elective lower gastrointestinal surgery when receiving treatment with corticosteroids before surgery. The studies reported AL rates ranging from 1.98 to 14.16%, which corresponds with the leakage rate reported previously in the literature [3]. Although corticosteroids are generally thought to impair intestinal anastomotic healing, this effect has never been proven unequivocally in clinical studies [20]. Some experimental studies have concluded that corticosteroid use impairs the healing of colonic anastomosis in a rat model [21–23], but others have reported no such effect [20]. Baca et al. [23] conducted an experimental study to investigate how long- and short-term corticosteroid administration in high and low doses modulated the course of the healing process for colonic anastomoses. The study indicated that high and low doses of methylprednisolone impaired the healing of colon anastomosis in rats, regardless of whether the methylprednisolone was administered over a long or a short preoperative period. The most prominent negative effects were associated with long-term, high-dose corticosteroid administration. The effect on the anastomosis was evaluated by measuring bursting pressure, hydroxyproline levels and histopathology findings. However, Mastboom et al. [20] did not find that short-term treatment with methylprednisolone lowered the anastomotic bursting pressures. Comparison of the hydroxyproline content yielded no significant difference between control and methylprednisolone groups. Furst et al. also conducted a study designed to determine the effects of corticosteroids on healing colonic anastomoses in a rat model. The effects were determined by measuring bursting pressure, and the study concluded that it was significantly lowered in the group undergoing shortterm treatment at 6, 8 and 20 days after surgery [21]. The definition, duration and dose of corticosteroid treatment varied widely among the studies, and several failed to report the mean dosage and duration of corticosteroid treatment. Accordingly, it is very difficult to compare the different studies. Three of the studies [12–14] divided patients receiving corticosteroids into high- or low-dose groups. In all three the reported risk of AL was higher in the high-dose corticosteroidtreated group. Hence, it seems that a higher dose of corticosteroid results in a higher risk of AL; however, because of the limited number of studies included in this review, this cannot be stated with certainty. The duration of corticosteroid treatment ranged from 60 min preoperatively to 44 years. The studies
Corticosteroids and anastomotic leakage
with the longest, as well as the shortest, duration of corticosteroid treatment (Vignali et al. [10] and Tresallet et al. [11] respectively) failed to show a significantly increased risk of AL in the corticosteroid-treated group. Sliecker et al. [8] also examined perioperative and longterm corticosteroid treatment, and found a significantly increased risk of AL in both corticosteroid-treated groups. The studies show conflicting evidence regarding the duration of use, and therefore it is not possible to clarify if the duration of corticosteroid treatment has an influence on the risk of AL. Studies included in this review are not sufficiently similar to make it reasonable to combine their results and conduct a meta-analysis. This is especially because of the differences in definition of corticosteroid use and duration of treatment Therefore, this review has obvious limitations. Furthermore, there was only one randomized controlled trial [10] to support our findings, but that study included only 52 patients and excluded those undergoing long-term treatment with corticosteroid. The remaining studies were either retrospective or prospective. Generally, the studies in our review had the limitation of including a relatively small number of patients. An exception was the study of Ziegler et al. [17], which included 5123 patients, 214 of whom were treated with steroids. Ziegler et al. defined corticosteroid use as regular administration of oral or parenteral corticosteroids for a chronic medical condition in the 30 days before surgery. This did not include shortcourse corticosteroids (i.e. treatment for ≤ 10 days). They found that 153 patients had AL: 19 (8.89%) were from the corticosteroid group compared with 134 (2.73%) from the noncorticosteroid group. With these limitations in mind it was interesting to see that the rates of AL are different, with no overlap in the CIs of the two different rates of AL found in the corticosteroid group (6.19%; 95% CI: 5.48–9.06) and the control group (3.33%; 95% CI: 2.94–3.58). However, it will never be possible to perform a randomized clinical trial focusing on complication differences in patients receiving long-term corticosteroid treatment. Our data warrant caution in the surgical treatment, with lower gastrointestinal surgery and an anastomosis, of patients receiving corticosteroids. In conclusion, treatment with corticosteroids may increase the risk of AL in patients undergoing gastrointestinal surgery; however, further investigations are needed before a final conclusion can be made. It is obviously very difficult to conduct a large randomized placebo-controlled trial of patients receiving long- or short-term corticosteroid treatment or placebo. Nationwide database studies with prospective evaluation of
Colorectal Disease ª 2013 The Association of Coloproctology of Great Britain and Ireland. 16, O154–O160
O159
Corticosteroids and anastomotic leakage
outcomes such as anastomotic leak and registration of long-term corticosteroid use will give more valid results.
Acknowledgements The authors thank T. W. Klausen, University of Copenhagen, Herlev Hospital, for statistical advice, and Richard Shock for editing services.
Author contributions All authors have provided substantial contributions to the conception and design, analysis and interpretation of data. All authors have drafted the article and revised it critically for intellectual content, and approved the final work to be published.
Conflicts of interest None.
References 1 Bell SW, Walker KG, Rickard MJ et al. Anastomotic leakage after curative anterior resection results in a higher prevalence of local recurrence. Br J Surg 2003; 90: 1261–6. 2 Branagan G, Finnis D. Prognosis after anastomotic leakage in colorectal surgery. Dis Colon Rectum 2005; 48: 1021–6. 3 Chambers WM, Mortensen NJ. Postoperative leakage and abscess formation after colorectal surgery. Best Pract Res Clin Gastroenterol 2004; 18: 865–80. € k O, Andersson M, Ruteg 4 Matthiessen P, Hallb€ oo ard J, Sj€ odahl R. Risk factors for anastomotic leakage after anterior rection of the rectum. Colorectal Dis 2004; 6: 462–9. 5 Rullier E, Laurent C, Garrelon JL, Michel P, Saric J, Parneix M. Risk factors for anastomotic leakage after resection of rectal cancer. Br J Surg 1998; 85: 355–8. 6 Richards CH, Campbell V, Ho C et al. Smoking is a major risk factor for anastomotic leak in patients undergoing low anterior resection. Colorectal Dis 2012; 14: 628–33. 7 Klein M, G€ ogenur I, Rosenberg J. Postoperative use of non-steroidal anti-inflammatory drugs in patients with anastomotic leakage requiring reoperation after colorectal resection: cohort study based on prospective data. BMJ 2012; 345: e6166. 8 Sliecker JC, Komen N, Mannaerts GH et al. Long-term and perioperative corticosteroids in anastomotic leakage. A prospective study of 259 left-sided colorectal anastomoses. Arch Surg 2012; 147: 447–552. 9 Tay GS, Binion DG, Eastwood D, Otterson MF. Multivariate analysis suggests improved perioperative outcome in Crohn’s disease patients receiving immunomodulator therapy after segmental resection and/or strictureplasty. Surgery 2003; 134: 565–72.
O160
T. F. Eriksen et al.
10 Vignali A, Di Palo S, Orsenigo E, Ghirardelli L, Radaelli G, Staudacher C. Effect of prednisolone on local and systemic response in laparoscopic vs. open colon surgery: a randomized double-blind, placebo-controlled trial. Dis Colon Rectum 2009; 52: 1080–8. 11 Tresallet C, Royer B, Godiris-Petit G, Menegaux F. Effect of systemic corticosteroids on elective left-sided colorectal resection with colorectal anastomosis. Am J Surg 2008; 195: 447–51. 12 Lake JP, Firoozmand E, Kang JC et al. Effect of high-dose steroids on anastomotic complications after proctocolectomy with ileal pouch-anal anastomosis. J Gastrointest Surg 2004; 8: 547–51. 13 El-Hussuna A, Andersen J, Bisgaard T et al. Biologic treatment or immunomodulation is not associated with postoperative anastomotic complications in abdominal surgery for Crohn’s disease. Scand J Gastroenterol 2012; 47: 662–8. 14 Ziv Y, Church JM, Fazio VW, King T, Lavery I. Effect of systemic steroids on ileal pouch-anal anastomosis in patients with ulcerative colitis. Dis Colon Rectum 1996; 39: 504–8. 15 Konishi T, Watanabe T, Kishimoto J, Nagawa H. Risk factors for anastomotic leakage after surgery for colorectal cancer: results of prospective surveillance. J Am Coll Surg 2006; 202: 439–44. 16 Golub R, Golub RW, Cantu R, Stein DH. A multivariate analysis of factors contributing to leakage of intestinal anastomoses. J Am Coll Surg 1997; 184: 364–72. 17 Ziegler MA, Catto JA, Riggs TW, Gates ER, Grodsky MB, Wasvary J. Risk factors for anastomotic leak and mortality in diabetic patients undergoing colectomy. Arch Surg 2012; 147: 600–5. 18 Lujan JJ, Nemeth Z, Barratt-Stopper PA, Bustami R, Koshenkov VP, Rolandelli RH. Factors influencing the outcome of intestinal anastomosis. Am Surg 2011; 77: 1169– 75. 19 Moher D, Liberati A, Tetzlaff J, Altman DG; PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med 2009; 18: 264–9. 20 Mastboom WJ, Hendriks T, de Man BM, de Boer HH. Influence of methylprednisolone on the healing of intestinal anastomoses in rats. Br J Surg 1991; 78: 54–6. 21 Furst MB, Stromberg BV, Blatchford GJ, Christensen MA, Thorson AG. Colonic anastomoses: bursting strength after corticosteroid treatment. Dis Colon Rectum 1994; 37: 12– 5. 22 Polat A, Nayci A, Polat G, Aks€ oyek S. Dexamethasone downregulates endothelial expression of intercellular adhesion molecule and impairs the healing of bowel anastomoses. Eur J Surg 2002; 168: 500–6. 23 Baca B, Ozben V, Boler DE et al. Effect of corticosteroid dose and duration of administration on colonic anastomosis. Inflamm Bowel Dis 2010; 16: 2162–7.
Colorectal Disease ª 2013 The Association of Coloproctology of Great Britain and Ireland. 16, O154–O160
Copyright of Colorectal Disease is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
![[NSAIDs and risk of anastomotic leakage after gastrointestinal surgery].](/assets/img/hold.png)
