Management of colorectal anastomotic leakage: differences between salvage and anastomotic takedown

Transcription

1 The American Journal of Surgery (2012) 204, Clinical Science Management of colorectal anastomotic leakage: differences between salvage and anastomotic takedown Domenico Fraccalvieri, M.D., Sebastiano Biondo, M.D.*, Jose Saez, M.D., Monica Millan, M.D., Esther Kreisler, M.D., Thomas Golda, M.D., Ricardo Frago, M.D., Bernat Miguel, M.A. Department of Surgery, Colorectal Unit, Bellvitge University Hospital and IDIBELL, University of Barcelona, Barcelona, Spain KEYWORDS: Colorectal anastomosis; Leakage; Anastomotic takedown; Salvage; Bowel restoration Abstract BACKGROUND: The aim of this study was to evaluate and compare the morbidity associated with 2 strategies of treatment of colorectal anastomotic leakage: surgical drainage of anastomosis with loop ileostomy versus anastomotic takedown. METHODS: An observational study of patients operated on for ileocolic or colorectal anastomotic leakage between 2001 and Patients were classified into 2 groups: group 1, salvage of the anastomosis, and group 2, anastomotic takedown. Mortality and morbidity were assessed. Morbidity and mortality of bowel restoration were also evaluated. RESULTS: Thirty-nine patients were included into group 1 and 54 into group 2. Mortality was 15% for group 1 and 37% for group 2 (P.022). The rate of patients suitable for stoma reversal was 91% for loop ileostomy and 38% for end stoma (P.001). Morbidity was 18% after loop ileostomy closure and 71% after end stoma reversal (P.021). Hospitalization was 10 days and 21 days, respectively (P.009). There was no mortality. CONCLUSIONS: Salvage of anastomosis with loop ileostomy is an effective strategy to control peritoneal sepsis for colorectal anastomotic leakage Elsevier Inc. All rights reserved. Anastomotic leakage (AL) is the most feared and dreadful specific complication of colorectal surgery, leading to significant morbidity, increased mortality, and prolonged hospital stay. There is also a significant increase in the use of hospital resources and costs after AL. 1 The reported incidences vary from.5% to over 30% 2 5 depending on the inclusion criteria, the case mix, and the definition of leak. In emergency colorectal procedures, AL occurs in 2% to 16% of cases of colonic obstruction and in 6% to 19% of cases operated on for colonic peritonitis. 6 AL adversely affects the morbidity and mortality of postoperative patients with a mortality rate of 25% to 35% in large series. It may * Corresponding author: Tel.: ; fax: address: sebastianobiondo@yahoo.com also result in a poorer functional outcome and increase the risk of permanent stoma formation. 7 9 There is no universally accepted definition of colorectal AL. It may present as diffuse peritonitis requiring abdominal reoperation; as fecal discharge from the wound or drain; as a localized abscess, which may be amenable to computed tomography scan guided percutaneous drainage; or as extravasation of radiologic contrast in an otherwise asymptomatic patient, which may only require surveillance. 10 Conventional management of a clinical AL with local or diffuse peritonitis often requires taking down the anastomosis, with creation of an end colostomy or ileostomy; the distal bowel is closed and left within the abdominal cavity or is exteriorized as a mucosal fistula. However, the salvage of a leaking colorectal or coloanal anastomosis using sur /$ - see front matter 2012 Elsevier Inc. All rights reserved. doi: /j.amjsurg

2 672 The American Journal of Surgery, Vol 204, No 5, November 2012 gical drainage and a proximal diverting stoma (avoiding anastomotic resection) has been proposed as an alternative management and reported in selected cases The aim of this observational study was to compare 2 surgical strategies to treat patients with AL after colorectal surgery: salvage of the anastomosis with derivative loop ileostomy versus anastomotic takedown. Patients and Methods Between January 2001 and May 2009 at Bellvitge University Hospital, all patients who underwent reoperation for AL after elective or emergency colorectal surgery were entered into the study. Patients were identified from a prospective database of elective and emergency colorectal procedures performed at our institution. AL was defined as generalized or localized peritonitis, the presence of pelvic abscess, or discharge of feces, pus, or gas from the abdominal drain or wound. All anastomotic leaks were confirmed by one or more of the following methods: an abdominopelvic computed tomography scan, water soluble contrast enema, digital rectal examination, and laparotomy. Patients with a preoperative diagnosis of AL not confirmed during laparotomy were excluded. None of the patients included had a protective stoma at the original surgery. According to the operation performed for the AL, patients were classified into 2 groups: patients treated by salvage of the anastomosis (with or without reanastomosis) and diverting loop ileostomy (group 1) and patients managed by anastomotic takedown with creation of an end colostomy or end ileostomy (group 2). Operations for AL were performed by a general surgeon or a colorectal surgeon (member of the colorectal surgery unit). The decision to perform one procedure or the other was left to the discretion of the surgeon on call. Mortality was defined as in-hospital death irrespective of the interval between primary operation and death. Demographic, physiological, clinical, and surgical data relating to the original operation and the surgery for leakage were collected for each patient. Age, sex, American Society of Anesthesiologists (ASA) grade, presence of comorbidities, indication for surgery, grade of peritonitis, type of resection, setting of surgery, type and location of the anastomosis, presence of leukocytosis and preoperative organ failure, length of stay and postoperative complications were the main variables evaluated. All patients fit for stoma reversal had a water-soluble contrast enema before surgery to rule out anastomotic leak or stenosis. Statistical analysis Statistical analysis was performed using the SPSS software package (SPSS for Windows, version 15.0; SPSS Inc, Figure 1 A flowchart of surgical management of patients with anastomotic leak. Chicago, IL). The level of statistical significance was set at P.05. To perform bivariable analysis, the chi-square and Fisher exact tests were used for the qualitative data; the Student t test or Mann-Whitney U tests were used for the quantitative data depending on data-application conditions. Results Ninety-three consecutive patients, 61 men (65.6%) and 32 women (34.4%), with an average age of 67.9 years (range 18 89), were included in the study. Indications for the first operation were colorectal cancer in 76 patients (81.7%) and benign disease in 17 patients (18.3%). Sixtyeight patients (73.1%) had undergone elective procedures, whereas 25 patients (26.9%) were operated on as an emergency. Fifty-three patients (57.0%) were classified as ASA grade 2, 38 patients (40.9%) as ASA 3, and 2 patients (2.2%) as ASA 4. Eighty-four patients (90.3%) had one or more associated diseases. Clinical suspicion of AL was confirmed by radiologic studies in 67 cases. At reoperation, local peritonitis was found in 15 cases (16.1%) and diffuse peritonitis in 78 patients (83.9%). Group 1 included 39 patients (41.9%), 21 patients with drainage of the anastomosis and derivative loop ileostomy and 18 patients with reanastomosis and ileostomy. Group 2 included 54 patients (58.1%). All these patients underwent anastomotic takedown, 32 with an end colostomy and 22 with an end ileostomy (Fig. 1). Table 1 shows the number of patients according to the type of anastomosis performed in the original operation. Comparative analysis between the 2 groups showed no differences in terms of patients characteristics. Significant differences were found regarding the type of surgeon and type of operation performed (Tables 2 and 3). The overall morbidity rate was 80.6% (75/93 patients). Table 4 shows postoperative complications. There were no differences between the groups. Patients with an end stoma

3 D. Fraccalvieri et al. Treatment of colorectal anastomotic leakage 673 Table 1 Site of anastomosis and treatment All patients Group 1 Group Intraperitoneal 79 (84.9) 32 (82.1) 47 (87.0) Proximal 45 (48.4) 23 (59.0) 22 (40.7) Distal 34 (36.6) 9 (23.1) 25 (46.3) Extraperitoneal 14 (15.1) Low rectum 14 (15.1) 7 (17.9) 7 (13) after anastomotic takedown needed further surgery more frequently than those with salvage of the anastomosis and loop ileostomy (18.5% vs 7.7%) without statistical differences. The overall mortality rate was 28.0% (26/93). Six of 39 patients in group 1 died (15.4%), whereas in group 2 the death rate was 37% (20/54 patients) (P.022). The length of hospital stay was days (standard deviation [SD] Table 3 Patients characteristics at the reoperation and their differences according to the anastomotic leakage treatment All patients Group 1 Group P value Hinchey score.329 Hinchey I II 15 (16.1) 8 (20.5) 7 (13.0) Hinchey III-IV 78 (83.9) 31 (79.5) 47 (87.0) Blood white cells , (47.3) 23 (59.0) 21 (38.9) 10, (52.7) 16 (41.0) 33 (61.1) Renal failure 29 (31.2) 11 (28.2) 18 (33.3).598 Respiratory 17 (18.3) 5 (12.8) 12 (22.2).247 failure Hemodynamic 18 (19.4) 6 (15.4) 12 (22.2).410 failure Type of surgeon.001 Colorectal 48 (51.6) 28 (71.8) 20 (37.0) General 45 (48.4) 11 (28.2) 34 (63.0) Table 2 Patients characteristics at the first operation and their differences according to the anastomotic leakage treatment All patients Group 1 Group P value Age (y) (48.4) 22 (56.4) 23 (42.6) (51.6) 17 (43.6) 31 (57.4) Sex.113 Male 61 (65.6) 22 (56.4) 39 (72.2) Female 32 (34.4) 17 (43.6) 15 (27.8) ASA grade.239 ASA I II 53 (57.0) 25 (64.1) 28 (51.9) ASA III IV 40 (43.0) 14 (35.9) 26 (48.1) Cardiac disease 19 (20.4) 6 (15.4) 13 (24.1).305 Coped 18 (19.4) 6 (15.4) 12 (22.2).410 Immunosuppression 4 (4.3) 2 (5.1) 2 (3.7) 1.000* Steroids 8 (8.6) 4 (10.3) 4 (7.4).716* Obesity 6 (6.5) 2 (5.1) 4 (7.4) 1.000* (BMI 35) Anemia (Hb 11 7 (7.5) 2 (5.1) 5 (9.3).695* mg/dl) Malignant disease 76 (81.7) 33 (84.6) 43 (79.6).539 Emergency surgery 25 (26.9) 7 (17.9) 18 (33.3).099 Site of anastomosis.072 Proximal 45 (48.4) 23 (59.0) 22 (40.7) Distal 34 (36.6) 9 (23.1) 25 (46.3) Low rectum 14 (15.1) 7 (17.9) 7 (13.0) ASA American Society of Anaesthesiologist classification; COPD chronic obstructive pulmonary disease; BMI Body Mass Index. *Fisher exact test ) in group 1 and days (SD 30.24) in group 2 without significant differences (P.578). Among the 33 patients who survived after salvage treatment by proximal loop ileostomy with reanastomosis or Table 4 Postoperative outcome Group 1 Group 2 (n 39) (n 54) P value Mortality 6 (15.4) 20 (37.0).022 Overall morbidity 34 (87.2) 41 (75.9).175 Wound infection (44).371 Evisceration * Anastomotic stenosis 1 NA Anastomotic dehiscence* 1 NA Intraperitoneal hemorrhage Abdominal abscess Reoperation Bowel ischemia * Pancreatitis * Myocardial infarction * Pulmonary complications Progressive septic shock Gastrointestinal bleeding * Renal failure * Cerebrovascular accident * Cardiac dysrhythmia Urinary tract infection * Some patient presented one or more complications. *One of 18 cases of drainage with loop ileostomy and reanastomosis (Fisher exact test).

4 674 The American Journal of Surgery, Vol 204, No 5, November 2012 Table 5 preservation of the leaking anastomosis, 30 (90.9%) have been judged suitable for stoma closure. Seventeen patients (56.6%) have just been operated on, and 13 (43.3%) are on the waiting list for surgery. Only 3 patients (9.1%) did not have stoma reversal caused by metastatic disease, advanced age, and prostate cancer with high surgical risk, respectively. Only 13 of the 34 patients (38.2%) in group 2 were judged suitable for the restoration of bowel continuity, and, to date, 7 patients (53.8%) have had surgery, and 6 (46.1%) are on the waiting list for surgery. Twenty-one patients (61.8%) did not undergo reversal of the end stoma because of the following different reasons: advanced age (n 1), high risk because of severe comorbidity (n 12), cancer progression (n 4), fear of complex surgery and possible complications (n 3), and death from other diseases (n 1). The difference found in the rate of restoration of bowel continuity between the 2 groups was statistically significant (P.001). Postoperative morbidity after closure of loop ileostomy was significantly lower than after surgery for end stoma reversal with rates of 17.6% and 71.4%, respectively (P.021). There were no cases of anastomotic leak or fistula in either group. The average length of stay was days (SD 7.41) for closure of loop ileostomy and 21 days (SD 12.21) for restorative surgery in patients with an end stoma (P.009) (Table 5). No overall mortality was observed. Comments Postoperative complications after stoma reversal Loop ileostomy (n 17) End stoma (n 7) P value Mortality Overall morbidity 3 (17.6) 5 (71).021* Anastomotic leak Need of additional surgery 1.292* Wound hemorrhage 1.292* Wound infection * Ileus * Pneumonia 1.292* Renal failure * Length of stay *Fisher exact test. Mann-Whitney U test. There is a large variability in the literature about the incidence of AL. Many reports include patients with and without a protective stoma and mix patients with symptomatic and asymptomatic anastomotic leaks; therefore, it becomes difficult to interpret the results. 2 5,10 In the present study, the authors only included patients with clinical AL confirmed by a computed tomography scan, water-soluble contrast enema, or laparotomy. Although many reports on anastomotic leaks examine etiology and risk factors, there is comparatively less information on the outcome after leaks. The main weakness of this study is the lack of randomization between the 2 surgical approaches. However, the groups were comparable with respect to all variables at the first surgery. At reoperation, the only factor with significant difference between the groups was the specialization of the surgeon. The use of a protective loop ileostomy in elective surgery to minimize the septic consequences of AL promoted our interest in evaluating the efficacy of proximal diversion by loop ileostomy and surgical drainage for the control of abdominal sepsis in postoperative local or diffuse peritonitis because of ileocolic or colorectal anastomotic disruption. We compared this more conservative surgical strategy with the traditional management of AL in colorectal surgery (resection of the anastomosis and creation of an end stoma). Proximal diversion using a loop ileostomy without resection of the leaking anastomosis has been proposed as an alternative strategy to treat the dehiscence of pelvic extraperitoneal anastomoses for low rectal resection in selected cases Recently, Hedrick et al 15 used this procedure in both intraperitoneal and extraperitoneal colon and rectal anastomosis. Of 27 patients operated on for anastomotic leak, 15 patients were successfully managed with diverting loop ileostomy and surgical drainage without resection of the anastomosis (in 12 cases, the anastomosis was intraperitoneal). In the present series of 93 patients operated on in emergency surgery for AL, the original anastomosis was intraperitoneal in 79 patients. Thirty-nine patients were treated with a more conservative surgery using a diverting loop ileostomy and surgical drains placed in close proximity of the anastomosis. The mortality rate was significantly lower among patients treated by anastomotic salvage and loop ileostomy than in the group treated by anastomotic takedown. Even though postoperative morbidity was similar in both groups, patients in group 2 needed reoperations more frequently. At our institution, colorectal emergency surgery may be performed by both colorectal surgeons and general surgeons depending on the organization chart of the emergency department. Fifty-two percent of our patients were operated on by a colorectal surgeon, and the decision to perform one procedure or the other was left to the discretion of the surgeon on call. Diverting loop ileostomy and surgical drainage of anastomosis were performed mainly by colorectal surgeons, whereas anastomotic takedown with end stoma was performed more often by a general surgeon. Recently, the authors evaluated the impact of surgical specialization on colorectal emergency surgery in a large series of patients. 16 The study showed that procedures performed by a colorectal surgeon are associated with better results in terms of postoperative mortality and morbidity. The choice of a more conservative management of AL by drainage and loop ileostomy could reflect more experience but also more

5 D. Fraccalvieri et al. Treatment of colorectal anastomotic leakage 675 Figure 2 An algorithm for the surgical management of colorectal anastomotic leak. judgment and confidence in the case of a colorectal surgeon rather than simply technical skill. Many authors recommend anastomotic resection because of the high risk of ongoing sepsis caused by the stool contained in the proximal colon when a leaking anastomosis is left in place. We consider that intraoperative colonic lavage performed through the distal opening of the loop ileostomy decreases the fecal load of the proximal colon and may be a useful method to reduce the risk of further sepsis, abscess formation, or local peritonitis. Restoration of bowel continuity after end stoma creation is associated with high morbidity rates, AL rates of 4% to 16%, and mortality rates of up to 4%. 17 Major complex surgery is often required with a full midline laparotomy, laborious takedown of adhesions, and difficult pelvic dissection to identify the rectal stump; in these circumstances, the risk of damage of pelvic vessels, ureters, or hypogastric nerves is considerable. In many cases, partial or total resection of the rectum is necessary, and a diverting loop ileostomy is performed to protect the low rectal or coloanal anastomosis. Therefore, because of the high risk of postoperative complications, restoration of intestinal continuity is never accomplished in a significant number of patients, and only those in a good general condition are selected for stoma reversal. 18 Loop ileostomy closure is usually a quite safe procedure, technically straightforward in most cases, and quick and often feasible through a small peristomal incision, with a morbidity rate between 10% and 30% and mortality between 0% and 2% Other authors have published good long-term results after proximal diversion without anastomotic takedown in the management of colorectal AL. Parc et al 23 reported significant differences in the rate of stoma reversal between patients with a diverting loop stoma (100%, 9 patients) versus 58% of those treated with Hartmann s procedure. In the study of Hedrick et al, 15 63% of patients managed with surgical drainage and proximal diversion had restoration of intestinal continuity compared with only 33% of the patients who had an end stoma. In the present experience, 91% of patients who were treated with salvage of the anastomosis and loop ileostomy have been considered suitable for stoma closure. By contrast, only 38% of patients with an end colostomy or end ileostomy were selected for restoration of bowel continuity. The morbidity after loop ileostomy closure in the present series was 17.6%, which is similar to other reports, and significantly lower than in the end stoma reversal group. No death was observed in either group. The length of stay was approximately twice as long for the reversal of end colostomy or end ileostomy compared with loop ileostomy. We think these results should be taken into account because of its considerable effect on the use of hospital resources and on overall health costs. Although we have not performed a cost analysis, reducing the length of stay would have considerable repercussion on the use of hospital resources and on overall health costs. Given the good results of anastomotic salvage and diverting loop ileostomy, the authors propose an algorithm of management of both intraperitoneal and extraperitoneal AL in colorectal surgery (Fig. 2). Patients who have critical hemodynamic conditions during surgery may benefit from a quick and effective procedure, so drainage of the leaking anastomosis with loop ileostomy may be proposed as a good surgical option to manage a minor dehiscence of an intraperitoneal anastomosis. In a hemodynamically stable patient, we recommend performing a new anastomosis covered by a loop ileostomy. In cases of low rectal anastomosis, we begin by checking the suture by digital examination. In the presence of major disruption (more than half of the circumference) or isch-

6 676 The American Journal of Surgery, Vol 204, No 5, November 2012 emia, the anastomosis should be resected with the creation of an end stoma. In the presence of minor dehiscence or when firm adhesions make the defect not visible, we recommend performing a loop ileostomy with drainage of the leaking anastomosis left in situ. In low rectal anastomoses, we consider it sensible and reasonable to avoid long pelvic dissections in an emergency reoperation. Performing a total revision of an extraperitoneal rectal anastomosis could lead to a coloanal anastomosis with a higher risk of subsequent problems. Furthermore, in our experience, this management had good short-term results. The present study suggests that the use of a diverting loop ileostomy, associated with salvage of the anastomosis, is an effective alternative strategy to control peritoneal sepsis caused by leakage of both extraperitoneal and intraperitoneal colorectal anastomosis. In our experience, this conservative procedure is safe and involves less mortality compared with anastomotic takedown and end stoma. Furthermore, because a loop ileostomy can be reversed in a minimally invasive manner, a high rate of patients who suffer an AL can undergo restoration of bowel continuity with low morbidity and mortality. Despite our encouraging results, further randomized studies are necessary to corroborate our conclusions. References 1. Frye J, Bokey EL, Chapuis PH, et al. Anastomotic leakage after resection of colorectal cancer generates prodigious use of hospital resources. Colorectal Dis 2009;11: Walker KG, Bell SW, Rickard MJ, et al. Anastomotic leakage is predictive of diminished survival after potentially curative resection for colorectal cancer. Ann Surg 2004;240: Biondo S, Kreisler E, Millan M, et al. Differences in postoperative and long-term outcomes between obstructive and perforated colonic cancer. Am J Surg 2008;195: Vignali A, Fazio VW, Lavery IC, et al. Factors associated with the occurrence of leaks in stapled rectal anastomoses: a review of 1,014 patients. J Am Coll Surg 1997;185: Rullier E, Laurent C, Garrelon JL, et al. Risk factors for Anastomotic leakage after resection of rectal cancer. Br J Surg 1998;85: Biondo S, Pares D, Kreisler E, et al. Anastomotic dehiscence after resection and primary anastomosis in left-sided colonic emergencies. Dis Colon Rectum 2005;48: Merad F, Hay JM, Fingerhut A, et al. Omentoplasty in the prevention of Anastomotic leakage after colonic or rectal resection. Ann Surg 1998;227: Alves A, Panis Y, Pocard M, et al. Management of Anastomotic leakage after nondiverted large bowel resection. J Am Coll Surg 1999;198: Law WL, Choi HK, Lee YM, et al. Anastomotic leakage is associated with poor long-term outcome in patients after curative colorectal resection for malignancy. J Gastrointest Surg 2007;11: Bruce J, Krukowski ZH, Al-Khairy G, et al. Systematic review of the definition and measurement of anastomotic leak after gastrointestinal surgery. Br J Surg 2001;88: Moran B, Heald R. Risk factors for and management of anastomotic leakage in rectal surgery. Colorectal Dis 2001;3: Watson A, Krukowsky Z, Munro A. Salvage of large bowel anastomotic leak. Br J Surg 1999;86: Pera M, Delgado S, García Valdecasas JC, et al. The management of leaking rectal anastomosis by minimally invasive techniques. Surg Endosc 2002;16: Eckmann C, Kujath P, Schiedeck TH, et al. Anastomotic leakage following low anterior resection: results of a standardized diagnostic and therapeutic approach. Int J Colorectal Dis 2004;19: Hedrick TL, Sawyer RG, Foley EF, et al. Anastomotic leak and loop ileostomy: friend or foe? Dis Colon Rectum 2006;49: Biondo S, Kreisler E, Millan M, et al. Impact of surgical specialization on emergency colorectal surgery outcomes. Arch Surg 2010:145: Banerjee S, Leather AJM, Rennie JA, et al. Feasibility and morbidity of reversal of Hartmann s. Colorectal Dis 2005;7: Aydin HN, Remzi FH, Tekkis PP, et al. s reversal is associated with high postoperative adverse events. Dis Colon Rectum 2005;48: Bakx R, Bush OR, Bemelman WA, et al. Morbidity of temporary loop ileostomies. Dig Surg 2004;21: Giannakopoulos GF, Veenhof AA, van der Peet DL, et al. Morbidity and complications of protective loop ileostomy. Colorectal Dis 2009; 11: Wong KS, Remzi FH, Gorgun E, et al. Loop ileostomy closure after restorative proctocolectomy: outcome in 1504 patients. Dis Colon Rectum 2005;48: Saha AK, Tapping CR, Foley GT, et al. Morbidity and mortality after closure of loop ileostomy. Colorectal Dis 2009;11: Parc Y, Frileux P, Schmitt G, et al. Management of postoperative peritonitis after anterior resection: experience from a referral intensive care unit. Dis Colon Rectum 2000;43: Edwards DP, Leppington-Clarke A, Sexton R, et al. Stoma related complications are more frequent after transverse colostomy than loop ileostomy: a prospective randomized clinical trial. Br J Surg 2001;88: Güenaga KF, Lustosa SA, Saad SS, et al. Ileostomy or colostomy for temporary decompression of colorectal anastomosis. Cochrane Database Syst Rev 2007;24:CD Perez RO, Habr-Gama A, Seid VE, et al. Loop ileostomy morbidity: timing of closure matters. Dis Colon Rectum 2006;49: Thalheimer A, Beuter M, Kortuem M, et al. Morbidity of temporary loop ileostomy in patients with colorectal cancer. Dis Colon Rectum 2006;49: