Anastomotic Leak 

Anastomotic Leak – Introduction

  • Anastomotic leak is a defect in the bowel wall at the anastomotic site leading to leakage of intestinal contents (World J Gastroenterol 2017 Sep 7;23(33):6172full-text)
  • comprehensive definitions of an anastomotic leak use a combination of factors which may include
    • clinical indicators
      •  pain
      •  peritonitis
      •  feculent or purulent drainage
    • biochemical markers
      •  fever
      •  tachycardia
      •  leukocytosis
    • intraoperative findings
      •  anastomotic disruption
      •  gross enteric leakage
    • radiologic indicators
      •  fluid collections
      •  gas collections
    •  Reference – World J Gastroenterol 2017 Sep 7;23(33):6172full-text

American Society of Anesthesiologists (ASA) Classification System

  • ASA score has been validated to assess risk of postoperative complications before any surgery
    •  ASA I – healthy, normal patient (for example, normal body mass index for age, nonsmoking, no or minimal alcohol consumption, good exercise tolerance)
    •  ASA II – patient with mild systemic disease or condition that is well-controlled and not associated with substantial functional limitations (for example, medication-controlled hypertension, obesity with body mass index (BMI) 30-40 kg/m2, smoker, or social alcohol drinker)
    •  ASA III – patient with severe disease that is not life-threatening (for example, poorly controlled diabetes or hypertension, morbid obesity, stable angina, implanted pacemaker, chronic renal failure, or a bronchospastic disease with occasional exacerbation)
    •  ASA IV – patient with severe, systemic disease (for example, poorly controlled chronic obstructive pulmonary disorder [COPD], symptomatic chronic heart failure, unstable angina, obesity with BMI ≥ 40 kg/m2, or myocardial infarction or stroke within the last 3 months)
    •  ASA V – moribund patient not expected to survive > 24 hours without surgery (for example, massive trauma, ruptured abdominal aortic aneurysm, severe intracranial hemorrhage with mass effect, or significant cardiac pathology or multiple organ/system dysfunction)
    •  ASA VI – brain-dead-declared patient whose organs are being removed in order to transplant them into another patient
    •  Reference – ASA Physical Status Classification System 2020 Dec 12

Types

  • International Study Group of Rectal Cancer proposed grading system for anastomotic leak based on clinical and imaging findings
    • grade A – radiological or clinical exam findings without associated symptoms or abnormal laboratory tests which is managed expectantly
      •  clinical exam findings may include seepage of new enteric content through drain or fistula (usually serous fluid but may include turbid or fecal contents)
      •  usually detected by routine contrast enema studies before closure of a temporary ileostomy/colostomy
    • grade B – leaks with signs and/or symptoms which requires nonoperative intervention including antibiotics and/or image-guided drainage
      •  clinical presentation may include abdominal and/or pelvic pain, and possibly abdominal distention and turbid/purulent rectal or vaginal discharge
      •  may have turbid/purulent or fecal drain contents depending on size of the leakage
      •  imaging studies (contrast enema x-ray or computed tomography [CT]) with transrectal instillation of contrast may show leakage of the endoluminally administered contrast agent into the extra-intestinal space, and potentially pelvic fluid collection (abscess)
    • grade C – leak which requires revision laparotomy to control life-threatening sepsis, including either removal of anastomosis with end colostomy, or creation of a protective ileostomy
      •  clinical presentation typically includes purulent/fecal drainage, severe abdominal pain, fever, signs of peritonitis (rebound tenderness, abdominal wall rigidity, and tachycardia), in addition to notably elevated C-reactive protein and leukocytosis
      •  imaging studies, such as CT with transrectal contrast instillation, typically show leakage at the anastomotic site and pelvic fluid collection
    •  Reference – International Study Group of Rectal Cancer proposal (Surgery 2010 Mar;147(3):339)
  • anastomotic leaks can also be classified chronologically from the time of surgery(3)
    •  early – developing < 3 days after surgery
    •  intermediate – developing 4-7 days after surgery
    •  late – developing ≥ 8 days after surgery

Epidemiology

Incidence/Prevalence

  •  incidence of anastomotic leaks varies by location and definition, with a higher incidence occurring the more distal the anastomosis is in colorectal surgeries(1,2,3)
  • reported leak rates by site of anastomosis(1)
    •  0.5%-6% with ileocolic anastomoses
    •  0%-9% with colocolonic anastomoses
    •  0%-20% with colorectal anastomoses
  • reported leak rates with rectal anastomosis(3)
    •  1.7% with high rectal anastomoses
    •  4.1% with low rectal anastomoses
    •  7.3% with ultra-low rectal anastomoses
  • following Roux-en-Y gastric bypass surgery
    •  2%-5% of patients reported to develop anastomotic leak(3)
    • reported incidence by location

Risk Factors

  • patient-related risk factors
    •  male sex(1)
    •  obesity(1)
    •  poor nutrition(3)
    •  diabetes mellitus(2)
    •  hypoalbuminemia(1)
    • American Society of Anesthesiologist (ASA) score III or IV (compared to ASA I or II)(1)
    • Crohn disease(1)
    •  rectal cancer (Patient Saf Surg 2010 Mar 25;4(1):5full-text)
    •  corticosteroid use(1)
    •  chemoradiation (World J Gastroenterol 2016 Jul 7;22(25):5718full-text)
    • nonselective nonsteroidal anti-inflammatory drug (NSAID) use(1)
      • postoperative use of nonsteroidal anti-inflammatory drugs, particularly diclofenac, associated with increased risk for anastomotic leak following colorectal surgery
        •  based on systematic review of observational studies
        • systematic review of 7 studies (6 retrospective cohorts, 1 case-control) evaluating postoperative NSAID use in 9,835 adults following colorectal surgery
          •  5 studies evaluated nonselective NSAIDs
          •  1 study evaluated both selective and nonselective NSAIDs
          •  1 study did not specify if NSAIDs were selective or nonselective
        •  compared to controls without NSAID use, increased risk for anastomotic leak with postoperative NSAID use (odds ratio [OR] 1.58, 95% CI 1.23-2.03) in analysis of 7 studies with 9,835 patients
        •  no significant differences in risk for anastomotic leak with use of the NSAID ketorolac (nonselective) in subgroup analysis of 2 studies with 988 patients
        •  Reference – Surg Endosc 2019 Mar;33(3):879
  • technical/procedural-related risk factors
    •  more distal anastomosis(2)
    •  anastomosis performed under tension(2)
    •  straight coloanal anastomoses compared to colonic J-pouch anastomoses(1)
    •  ischemia at site of anastomosis(1,3)
    •  surgery time > 2 hours (J Am Coll Surg 2009 Feb;208(2):269), commentary can be found in J Am Coll Surg 2009 Jun;208(6):1152
    •  perioperative blood transfusion (J Am Coll Surg 2009 Feb;208(2):269), commentary can be found in J Am Coll Surg 2009 Jun;208(6):1152
    •  emergency colorectal surgery (Patient Saf Surg 2010 Mar 25;4(1):5full-text)
    •  history of previous or revisional bariatric surgery (Arq Bras Cir Dig 2015;28(1):74full-text)
    • hand-sewn ileocolic anastomoses (compared to stapled ileocolic anastomoses)(1)
      • stapled ileocolic anastomosis associated with fewer leaks than hand-sewn ileocolic anastomosis (level 2 [mid-level] evidence)
        •  based on Cochrane review of trials with methodologic limitations
        •  systematic review of 7 randomized trials comparing linear cutter stapling (isoperistaltic side-to-side or functional end-to-end) to hand-sewn techniques in 1,125 adults requiring ileocolic anastomosis
        •  6 trials did not report intention-to-treat analyses, 1 additional trial did not have blinded outcome assessment
        • comparing stapled vs. hand-sewn ileocolic anastomosis
          • stapled anastomosis associated with fewer anastomotic leaks
            • in overall analysis of 7 trials
              •  odds ratio (OR) 0.48 (95% CI 0.24-0.95)
              •  NNT 23-354 with leakage in 6% of hand-sewn group
            • in subgroup analysis of 825 cancer patients from 4 trials
              •  OR 0.28 (95% CI 0.1-0.75)
              •  NNT 16-61 with leakage in 7% of hand-sewn group
          • no significant differences between groups in
            •  anastomotic leakage in subgroup analysis of 264 noncancer patients from 3 trials
            •  stricture, anastomotic hemorrhage, anastomotic time, reoperation, mortality, intra-abdominal abscess, wound infection, or length of stay
        •  Reference – Cochrane Database Syst Rev 2011 Sep 7;(9):CD004320
    • open surgery and conversion to open surgery may each increase risk of 30-day anastomotic leak compared to laparoscopic surgery in patients having colectomy for colon cancer (level 2 [mid-level] evidence)
      •  based on retrospective cohort study
      • 25,097 patients (mean age 66 years) having elective partial or total colectomy with or without proximal diverting ostomy for colon cancer were stratified by operative approach
        •  30.5% of patients had laparoscopic surgery without conversion to open surgery
        •  30.5% of patients had hand-assisted surgery defined as any minimally invasive technique with hand or open assist (including laparoscopic, robotic, hybrid, or other) without conversion to open surgery
        •  26.7% of patients had planned open surgery
        •  8% of patients had conversion to open surgery from a minimally invasive approach
        •  4.2% of patients had robotic surgery (excluding those with hand- or open-assist) without conversion to open surgery
      •  3.32% of patients developed anastomotic leak within 30 days of surgery
      • factors associated with increased risk for anastomotic leak compared to laparoscopic surgery
        •  open surgery (adjusted odds ratio [OR] 1.72, 95% CI 1.42-2.1)
        •  conversion to open surgery (adjusted OR 1.81, 95% CI 1.39-2.36)
        •  hand-assisted surgery (adjusted OR 1.3, 95% CI 1.06-1.59)
      •  no significant difference in risk of anastomotic leakage comparing robotic surgery to laparoscopic surgery
      •  Reference – World J Surg 2017 Aug;41(8):2143

Factors Not Associated With Increased Risk

  • immunomodulator use, such as
    •  azathioprine(1,2)
    •  6-mercaptopurine(2)
  • single-layer compared to double-layer hand-sewn colorectal anastomoses(1)
    • single-layer suture anastomosis may shorten operating time without increasing risk for postoperative leak or complications compared to double-layer anastomosis (level 2 [mid-level] evidence)
      •  based on Cochrane review of trials with methodologic limitations
      •  systematic review of 7 randomized trials comparing single-layer to double-layer suture anastomosis in 842 patients having gastrointestinal surgery
      •  most trials did not report allocation concealment or blinding of outcome assessor; authors presented conflicting information on these quality measures for 1 trial
      • no significant differences between groups in
        •  anastomotic leak in analysis of 7 trials with 842 patients
        •  major complications in analysis of 7 trials with 842 patients
        •  mortality in analysis of 4 trials with 403 patients
        •  length of hospital stay in analysis of 3 trials with 390 patients
      •  single layer suture anastomosis associated with reduced operating time (mean difference -11.12 minutes, 95% CI -16.37 to -5.87 minutes) in analysis of 2 trials with 218 patients
      •  Reference – Cochrane Database Syst Rev 2012 Jan 18;(1):CD005477

Etiology and Pathogenesis

Clinical Presentation

  •  up to 50% of patients are reported to be asymptomatic, with leaks detected incidentally on imaging(3)
  • signs and symptoms, if present, usually develop 5-8 days postoperatively, and may include(1)
    •  abdominal pain, which can be severe(1,2,3)
    •  fever(1,2)
    •  nausea(1)
    •  tachycardia(1,2,3)
    •  tachypnea(2)
    •  oliguria(2)
    •  mental status changes(2)
    •  feculent drainage or discharge(1)
  • acute, self-limited leaks or slow leaks may present as chronic fistulas, with symptoms arising months to years after surgery(3)
    •  presentation may include pain or general feeling of malaise
    •  in patients with a history of bariatric surgery, fistulas that develop between the gastric pouch and gastric remnant may result in weight regain
  • most common signs and symptoms in retrospective cohort of 58 patients with anastomotic leak presenting at mean 8 days (range 3-65 days) postoperatively
    •  nonspecific findings (such as low-grade fever, mild tachycardia, and leukocytosis) in 80%
    •  abdominal pain in 64%
    •  fever in 52%
    •  nausea in 24%
    •  frank peritonitis in 22%
    •  Reference – Arch Surg 2009 Apr;144(4):333
  • American Association of Clinical Endocrinologists/American College of Endocrinology/Obesity Society/American Society for Metabolic and Bariatric Surgery/Obesity Medicine Association/American Society of Anesthesiologists (AACE/ACE/OS/ASMBS/OMA/ASA) recommendations for assessment of postoperative anastomotic leak
    •  signs of possible anastomotic leak include sustained resting tachycardia, hypoxia, and fever
    •  consider assessing patients with respiratory distress or failure to wean from ventilation after bariatric surgery for postoperative anastomotic leak (AACE/ACE/OS/ASMBS/OMA/ASA Grade D)
    • consider computed tomography (CT) (preferred over upper-gastrointestinal (GI) studies [water-soluble contrast followed by thin barium]) to evaluate for anastomotic leaks in suspected patients (AACE/ACE/OS/ASMBS/OMA/ASA Grade C, BEL 3)
    • consider selected diatrizoate meglumine and diatrizoate sodium upper-GI study in the absence of abnormal signs or symptoms to identify any subclinical leaks before discharge; however, routine use is not recommended (AACE/ACE/OS/ASMBS/OMA/ASA Grade C, BEL 3)
    •  consider C-reactive protein test if postoperative leak suspected after hospital discharge (AACE/ACE/OS/ASMBS/OMA/ASA Grade B, BEL 2)
    • exploratory laparotomy or laparoscopy is recommended for highly suspected anastomotic leaks (AACE/ACE/OS/ASMBS/OMA/ASA Grade A, BEL 1)
    •  Reference – AACE/ACE/OS/ASMBS/OMA/ASA clinical practice guideline on for the perioperative nutritional, metabolic, and nonsurgical support of the bariatric surgery patient – 2019 update (Obesity (Silver Spring) 2020 Apr;28(4):O1)

Diagnosis

Making the Diagnosis

  •  diagnosis can be difficult, as clinical presentation varies widely and is usually nonspecific(2)
  •  suspect the diagnosis in patients who have undergone intestinal anastomosis and who have postoperative abdominal pain, fever, nausea, tachycardia, hyperthermia, tachypnea, oliguria, or mental status changes(1,2,3)
  •  the diagnosis is typically confirmed by imaging studies, usually computed tomography with contrast, showing a large collection of free fluid, extravasation of contrast material, or a perianastomotic fluid collection(2)

Testing Overview

  • computed tomography (CT) scan with oral and/or rectal contrast or a contrast enema is commonly used to confirm diagnosis of anastomotic leak and can assist with management planning
  •  for upper gastrointestinal anastomotic sites, imaging studies such as contrast esophagrams and upper gastrointestinal x-ray series have been used, but are reported to have limited sensitivity
  • blood test findings associated with anastomotic leak include
    •  elevated C-reactive protein (CRP)
    •  elevated white blood cell count
  •  analysis of biomarkers for ischemia in drain fluid may aid early diagnosis of anastomotic bowel leak

Clinical Prediction Rules

  • modified Dutch leakage (DULK) score appears at least as effective as original DULK score, and associated with high sensitivity for clinically relevant anastomotic leak following colorectal surgery (level 2 [mid-level] evidence)
    •  based on retrospective prognostic cohort study
    • 782 patients with an anastomosis in the colon or rectum between October 2007 and November 2009 from 5 Dutch centers were evaluated by the original and modified DULK score
      • modified DULK score based on 4 parameters with a score of ≥ 1 considered positive for anastomotic leak
        •  respiratory rate ≥ 20/minute (1 point)
        •  deteriorating clinical condition (1 point)
        •  abdominal pain other than wound pain (1 point)
        •  C-reactive protein (CRP) > 250 (1 point)
      • original DULK score based on 13 parameters with a score of ≥ 4 considered positive for anastomotic leak
        •  fever > 38 degrees C (100.4 degrees F) (1 point)
        •  heart rate > 100/minute (1 point)
        •  respiratory rate > 30/minute (1 point)
        •  urinary production < 30 mL/hour or 700 mL/day (1 point)
        •  agitation or lethargic mental status (2 points)
        •  deteriorating clinical condition (2 points)
        • presence of
          •  bowel obstruction (2 points)
          •  gastric retention (2 points)
          •  fascial dehiscence (2 points)
          •  abdominal pain other than wound pain (2 points)
        •  increase of ≥ 5% in leukocyte number or CRP (1 point)
        •  increase of ≥ 5% in urea or creatinine (1 point)
        • nutritional status
          •  tube feeding (1 point)
          •  total parenteral nutrition (2 points)
    •  reference standard was clinically apparent leakage (such as fecal discharge from drains or abdominal wound) or radiological, endoscopic, or surgically proven leakage
    •  10.4% of patients had clinically relevant anastomotic leak per reference standard
    • diagnostic performance of modified DULK score with cutoff ≥ 1 point for detection of clinically relevant anastomotic leak
      •  sensitivity 97%
      •  specificity 56.8%
      •  positive predictive value 17.2%
      •  negative predictive value 99.5%
    • diagnostic performance of original DULK score with cutoff ≥ 4 points for detection of clinically relevant anastomotic leak
      •  sensitivity 97%
      •  specificity 53.6%
      •  positive predictive value 16.2%
      •  negative predictive value 99.5%
    •  Reference – Colorectal Dis 2013 Sep;15(9):e528

Intraoperative Leak Testing

  • intraoperative air-leak testing for assessment of anastomotic bowel integrity allows for quick identification of leaks with ability to reanastomoses, divert fecal transit, or repair(2)
    • air-leak testing is usually performed by filling pelvis with saline, occluding proximal bowel, and introducing air transanally using a syringe, bulb, or flexible endoscope(2)
      •  reported to identify leaks in up to 25% of anastomoses
      •  not all positive air-leak tests will result in clinically detected or radiographically visible anastomotic leaks
    •  injection of methylene blue through orogastric tube is another method used for air-leak testing (Arq Bras Cir Dig 2015;28(1):74full-text)
  • techniques for intraoperative leak testing which have not shown sufficient efficacy to be widely accepted include
    •  Doppler flowmetry
    •  scanning laser flowmetry
    •  fluorescence videography
    •  near-infrared spectroscopy
    •  intramucosal pH measurements
    •  intraoperative colonoscopy
    •  intraoperative visible light spectroscopy
    •  Reference – Anticancer Res 2010 Feb;30(2):601
  • intraoperative endoscopic air-leak testing may reduce postoperative anastomotic leaks compared to intraoperative visual inspection in patients having laparoscopic Roux-en-Y gastric bypass (level 2 [mid-level] evidence)
    •  based on randomized trial without allocation concealment
    •  100 patients having laparoscopic Roux-en-Y gastric bypass randomized to intraoperative air-leak testing vs. intraoperative visual inspection and followed for 30 days
    •  4 leaks were found in intraoperative endoscopic air-leak testing group and repaired intraoperatively
    • comparing intraoperative endoscopic air leak testing vs. visual inspection
      •  postoperative gastrojejunal anastomotic leak in 0% vs. 8% (p = 0.04)
      •  need for reoperation in 0% vs. 8% (p = 0.04)
      •  mean duration of surgery 194 minutes vs. 159 minutes (p < 0.001)
      •  mean length of hospital stay of 2.44 days vs. 3.46 days (p = 0.025)
    •  no significant differences in rate of bleeding of gastrojejunal anastomosis, narrow gastrojejunal anastomosis, or 30-day mortality
    •  Reference – Int J Surg 2018 Feb;50:17
  • intraoperative air-leak testing of left-sided colorectal anastomoses may be associated with reduced incidence of clinical leaks (level 2 [mid-level] evidence)
    •  based on retrospective cohort study
    •  998 left-sided colorectal anastomoses analyzed from prospective database
    •  7.9% of 825 anastomoses with air-leak testing had intraoperative air leaks
    • incidence of clinical leaks (p < 0.03 among groups)
      •  3.8% anastomoses with negative air-leak test results
      •  7.7% anastomoses with positive air-leak test results
      •  8.1% untested anastomoses
    •  Reference – Arch Surg 2009 May;144(5):407
  • intraoperative methylene blue enema following colonic anastomosis reported to detect intraoperative leak (level 3 [lacking direct] evidence)
    •  based on case series
    •  229 surgeries with colonic anastomosis and intraoperative methylene blue enema were evaluated
    •  intraoperative leak detected in 7% of anastomoses (4.5% proximal and 8% distal anastomoses)
    •  postoperative leak rate 3%
    •  no cases of postoperative leak reported in patients who had intraoperative repair following detection of intraoperative leak
    •  Reference – BMC Surg 2007 Aug 2;7:15full-text

Blood Tests

  • blood test findings associated with anastomotic leak include(1,2)
    •  increasing C-reactive protein (CRP) (may be > 250 mg/L)
    •  elevated white blood cell counts
  •  normal serum CRP on postoperative day 3-5 may help identify patients at low risk of anastomotic leak after colorectal surgery (level 2 [mid-level] evidence)
    •  based on systematic review limited by clinical heterogeneity
    •  systematic review of 7 cohort studies evaluating serum CRP for predicting anastomotic leakage in 2,483 patients who had colorectal surgery
    •  definitions of anastomotic leak and reference standards for anastomotic leak varied across studies
    •  pooled prevalence of anastomotic leakage 9.6%
    • pooled performance of serum CRP for predicting anastomotic leak
      • on postoperative day 3 with cutoff of 172 mg/L in analysis of 5 studies with 2,126 patients
        •  sensitivity 76% (95% CI 66%-84%)
        •  specificity 76% (95% CI 67%-83%)
        •  negative predictive value 97%
      •  similar performance found on postoperative day 4 with cutoff of 124 mg/L and on postoperative day 5 with cutoff of 144 mg/L
    •  Reference – Br J Surg 2014 Mar;101(4):339
  • increase in CRP > 50 mg/L between any 2 postoperative days might help rule out anastomotic leak needing radiologic or surgical intervention in adults having elective colorectal resection with anastomosis (level 2 [mid-level] evidence)
    •  based on diagnostic cohort study without blinding of test under investigation
    • 933 adults having elective colorectal resection or restoration of continuity with anastomosis were assessed for blood CRP measurement before surgery and daily for 5 days after surgery or until discharge
    • 833 patients (89%) (mean age 64 years, 54% male) with complete data followed for 30 days and included in analysis
    • anastomotic leak defined as defect in intestinal wall at site of anastomosis requiring operative or radiologic intervention performed according to physician’s discretion (reference standard)
    • 4.9% had anastomotic leak requiring operative or radiologic intervention by reference standard
    • 2.5% had anastomotic leak managed medically
    • for diagnosis of anastomotic leak requiring operative or radiologic intervention
      • increase in CRP > 50 mg/L between any 2 postoperative days had
        • sensitivity 85%
        • specificity 51%
        • positive predictive value 8%
        • negative predictive value 99%
    • similar results for diagnosis of anastomotic leak requiring medical, operative, or radiologic intervention
    • Reference – PREDICT trial (Br J Surg 2020 Dec;107(13):1832)
    • CLINICIANS’ PRACTICE POINT: This test may have marginal utility even though the negative predictive value of 99% suggests a posttest risk of 1% of having an undetected anastomotic leak, as this is not very different from the pretest risk of 4.9%. Further studies in cohorts with higher prevalence of anastomotic leak may help clarify its utility.
  • pilot study assessing monocytic human leukocyte antigen-DR (HLA-DR), white blood cells, C-reactive protein, and procalcitonin for prediction of anastomotic leak in adults having elective colorectal resection with anastomosis can be found in J Am Coll Surg 2019 Aug;229(2):200

Imaging Studies

  • computed tomography (CT) scan with oral and/or rectal contrast or a contrast enema is commonly used to confirm diagnosis of colorectal anastomotic leak, and can assist with management planning(1,2)
    • multidirectional CT may help confirm anastomotic leak in patients having left-sided colonic anastomoses with circular stapling (level 2 [mid-level] evidence)
      •  based on retrospective diagnostic cohort study
      •  170 patients having left-sided colonic anastomoses with circular stapling were evaluated
      • anastomotic leak suspected in 17.6% (30 patients) postoperatively
        •  28 patients had multidirectional CT scan enhanced (IV contrast) or unenhanced with addition of rectal contrast
        •  2 patients had immediate surgery
      •  reference standard was surgery and clinical outcome
      •  7.6% of patients had anastomotic leak by reference standard, with 20% related mortality
      • diagnostic performance of multidirectional CT scan for detection of clinically suspected anastomotic leak
        •  sensitivity 91%
        •  specificity 100%
        •  positive predictive value 100%
        •  negative predictive value 95%
      • diagnostic performance of individual signs on multidirectional CT
        • extravasation of rectal contrast
          •  sensitivity 100%
          •  specificity 100%
        • presence of peri-anastomotic free air
          •  sensitivity 81%
          •  specificity 74%
        • presence of peri-anastomotic fluid collections
          •  sensitivity 63%
          •  specificity 88%
        • staple line integrity
          •  sensitivity 72%
          •  specificity not applicable
      •  Reference – Clin Radiol 2014 Jan;69(1):59
  • CT scan with oral, water-soluble contrast provides very high sensitivity for detection of anastomotic leaks after bariatric surgery(3)
    • other imaging studies, including contrast esophagrams and upper gastrointestinal x-ray series, have been used but are reported to have limited sensitivity(3)
      • selective upper gastrointestinal series appears to have higher sensitivity and similar specificity as routine upper gastrointestinal series for detecting anastomotic leaks in patients with laparoscopic Roux-en-Y gastric bypass (level 2 [mid-level] evidence)
        •  based on retrospective diagnostic cohort study without blinding
        • 804 patients (mean age 41 years) having laparoscopic Roux-en-Y gastric bypass were assessed with selective or routine gastrointestinal series
          •  52.5% of patients had selective upper gastrointestinal series (performed if tachycardia, fever, drainage content, or general condition was suspicious for gastrojejunostomy leak) from May 2005 to April 2010
          •  47.5% of patients had routine upper gastrointestinal series from June 2000 to April 2005
        •  reference standard was operative or clinical findings
        •  1.1% of patients had anastomotic leak per reference standard
        • diagnostic performance of selective upper gastrointestinal series for detection of anastomotic leak
          •  sensitivity 80%
          •  specificity 91%
          •  positive predictive value 80%
          •  negative predictive value 91%
        • diagnostic performance of routine upper gastrointestinal series for detection of anastomotic leak
          •  sensitivity 50%
          •  specificity 97%
          •  positive predictive value 18%
          •  negative predictive value 99%
        • no significant differences between groups in
          •  time to diagnosis
          •  overall leak rate
        •  Reference – Obes Surg 2011 Aug;21(8):1238

Drain Fluid Assessment

  • analysis of biomarkers of ischemia in drain fluid may aid early diagnosis of anastomotic bowel leak(1)
  • elevated lactate and pH levels < 7 in peritoneal drain fluid might each be associated with colorectal anastomotic leaks (level 2 [mid-level] evidence)
    •  based on systematic review of diagnostic studies limited by clinical heterogeneity
    • systematic review of 13 prospective diagnostic studies evaluating the ability of peritoneal fluid biomarkers to identify early colorectal anastomotic leaks
      •  7 studies evaluated biomarkers of ischemia (5 studies evaluated lactate, 2 studies evaluated pH) in 1,088 patients
      •  7 studies evaluated biomarkers of inflammation (cytokines) in 322 patients
    •  meta-analysis not performed due to clinical heterogeneity, including differences in patient populations, end points analyzed, and definitions of anastomotic leaks
    • ischemic biomarkers consistently showing associations with anastomotic leaks include
      •  elevated lactate in 5 studies with 245 patients
      •  pH < 7 in 2 studies with 843 patients
    •  elevated inflammatory biomarkers (interleukin 6 and tumor necrosis factor-alpha) showed conflicting associations with anastomotic leaks in 7 studies with 322 patients
    •  Reference – Int J Colorectal Dis 2017 Jul;32(7):935

Management

Management Overview

  • management is based on clinical presentation which can vary from asymptomatic to emergent sepsis(2)
    •  patients with severe sepsis who are hemodynamically unstable should be aggressively resuscitated, followed by antibiotics and revision laparotomy to control the anastomotic leak(3)
    • in stable, symptomatic patients without sepsis
      •  upper gastrointestinal anastomotic leaks can be managed by stopping oral intake and instituting parenteral or jejunal nutritional support, broad-spectrum antibiotics, and percutaneous drainage, which is reported to resolve leaks in the majority of patients(3)
      •  colorectal anastomotic leaks can be managed by stopping oral intake and instituting parenteral nutritional support, broad-spectrum antibiotics, and percutaneous drainage(2)
      •  more aggressive surgery may be necessary in patients with continuing leakage of enteric contents or lack of clinical improvement following drainage(2)
    • endoscopic therapies using self-expanding metal or covered stents, clips, glue, or suturing (alone or in combination), can be used for(2,3)
    • endoscopic vacuum therapy (EVT) can be used for
      • acute and chronic anastomotic leaks < 5 cm
      • critically ill, hemodynamically unstable patients for infectious source control (if patient does not respond to EVT therapy, surgery may still be required)
    •  patients without symptoms may be managed expectantly with close surveillance so that immediate action can be taken for sudden exacerbations (J Visc Surg 2014 Dec;151(6):441)
  • management strategy based on the International Study Group of Rectal Cancer proposed grading system for anastomotic leaks
    •  grade A leaks (radiological or clinical exam findings without associated symptoms or abnormal laboratory tests) are usually managed expectantly, but may require delay in ileostomy/colostomy closure
    •  grade B leaks (radiological or clinical exam findings with associated symptoms or abnormal laboratory tests) can be managed nonoperatively with antibiotics and/or drainage using pelvic drain or transanal lavage
    •  grade C leaks require revision laparotomy for control of life-threatening sepsis, including either removal of anastomosis with end colostomy, or creation of a protective ileostomy
    •  Reference – Surgery 2010 Mar;147(3):339

Antibiotics

  •  in symptomatic patients who are stable, antibiotics are first-line treatment and may be used alone or in combination with percutaneous drainage or surgery based on severity of leak(2)
  •  gram-negative and anaerobic coverage may be used for small abscesses not requiring percutaneous drainage(2)
  • Surgical Infection Society and the Infectious Diseases Society of America (SIS/IDSA) recommended options for empiric IV antibiotics for adults with complicated intra-abdominal infection include (SIS/IDSA Grade B, Level II)
    • piperacillin-tazobactam 3.375 g every 6 hours
    • ticarcillin-clavulanate 3.1 g every 6 hours
    • doripenem 500 mg every 8 hours
    • ertapenem 1 g every 24 hours
    • imipenem-cilastatin 500 mg every 6 hours or 1 g every 8 hours
    • meropenem 1 g every 8 hours
    • cefazolin 1-2 g every 8 hours
    • cefepime 2 g every 8-12 hours
    • cefotaxime 1-2 g every 6-8 hours
    • cefoxitin 2 g every 6 hours
    • ceftazidime 2 g every 8 hours
    • ceftriaxone 1-2 g every 12-24 hours
    • cefuroxime 1.5 g IV every 8 hours
    • tigecycline 100-mg IV initial dose, then 50 mg every 12 hours
    • ciprofloxacin 400 mg IV every 12 hours
    • levofloxacin 750 mg IV every 24 hours
    • moxifloxacin 400 mg IV every 24 hours
    • metronidazole 500 mg IV every 8-12 hours or 1,500 mg every 24 hours
    • gentamicin or tobramycin 5-7 mg/kg IV every 24 hours
    • amikacin 15-20 mg/kg every 24 hours
    • aztreonam 1-2 g every 6-8 hours
    • vancomycin 15-20 mg/kg every 8-12 hours
    •  Reference – SIS/IDSA guidelines on diagnosis and management of complicated intra-abdominal Infection in adults and children (Clin Infect Dis 2010 Jan 15;50(2):133), correction can be found in Clin Infect Dis 2010 Jun 15;50(12):1695, commentary can be found in Clin Infect Dis 2010 Sep 15;51(6):757
  • SIS/IDSA recommended options for empiric IV antibiotics for children with complicated intra-abdominal infection include
    • amikacin 15-22.5 mg/kg/day every 8-24 hours
    • ampicillin sodium 200 mg/kg/day every 6 hours
    • ampicillin/sulbactam 200 mg/kg/day of ampicillin component every 6 hours
    • aztreonam 90-120 mg/kg/day IV every 6-8 hours
    • cefepime 100 mg/kg/day IV every 12 hours
    • cefotaxime 150-200 mg/kg/day every 6-8 hours
    • cefotetan 40-80 mg/kg/day every 12 hours
    • cefoxitin 160 mg/kg/day every 4-6 hours
    • ceftazidime 150 mg/kg/day every 8 hours
    • ceftriaxone 50-75 mg/kg/day every 12-24 hours
    • cefuroxime 150 mg/kg/day every 6-8 hours
    • ciprofloxacin 20-30 mg/kg/day every 12 hours
    • clindamycin 20-40 mg/kg/day every 6-8 hours
    • ertapenem
      •  3 months to 12 years: 15 mg/kg twice daily (not to exceed 1 g/day) every 12 hours
      •  ≥ 13 years: 1 g/day every 24 hours
    • gentamicin 3-7.5 mg/kg/day every 2-4 hours
    • imipenem/cilastatin 60-100 mg/kg/day every 6 hours
    • meropenem 60 mg/kg/day 8 hours
    • metronidazole 30-40 mg/kg/day every 8 hours
    • piperacillin/tazobactam 200-300 mg/kg/day of piperacillin component every 6-8 hours
    • ticarcillin/clavulanate 200-300 mg/kg/day of ticarcillin component every 4-6 hours
    • tobramycin 3.0-7.5 mg/kg/day every 8-24 hours
    • vancomycin 40 mg/kg/day as 1-hour infusion every 6-8 hours
    •  Reference – SIS/IDSA guidelines on diagnosis and management of complicated intra-abdominal Infection in adults and children (Clin Infect Dis 2010 Jan 15;50(2):133), correction can be found in Clin Infect Dis 2010 Jun 15;50(12):1695, commentary can be found in Clin Infect Dis 2010 Sep 15;51(6):757

Drainage

  •  percutaneous, transanal, or surgical drains may be used for International Study Group of Rectal Cancer grade B and grade C leaks depending on location of anastomosis, size of abscess, and size of defect(1,2)
  • the ability to perform image-guided percutaneous drainage is based upon having(2)
    •  safe radiographic window
    •  experienced radiologist
    •  homogeneity of fluid
    •  abscess ≥ 3 cm
  • transanal drainage may be achieved in those anastomotic leaks (usually grade B) that can be reached endoscopically, especially after failed percutaneous drainage(1,2)
    •  drainage is typically performed through the anastomotic dehiscence, with techniques ranging from simply opening the defect to allow drainage, use of a formal surgical device, or drain insertion
    •  follow-up x-ray with contrast through transanal drain may be performed to monitor abscess
    •  drain typically removed when cavity has decreased to size of drain

Endoscopic Therapies With Stents, Clips, Glue, and Suturing

  • endoscopic therapies such as stents (most common), clips, glue, and suturing can be used alone or in combination for management of anastomotic leaks in patients with(2,3)
  • self-expanding metal or covered stents can be used to help treat an anastomotic leak following percutaneous drainage of the extraluminal abscess(2)
    • stents are placed endoscopically under fluoroscopic guidance
      •  stents are usually left in place for 2-8 weeks, with longer times associated with more difficult extraction
      •  plastic stents may be more easily removed, but may have higher migration risk
      •  successful leak closure rates for combined plastic and metal stents reported to be about 88% with majority being closed with 1 treatment
      •  stent placement allows for resumption of enteral feeding, which is theorized to accelerate recovery
      •  Reference – Clin Gastroenterol Hepatol 2013 Apr;11(4):343
    • use of stents may be limited by complications, including
      • stent migration (reported in > 40%)(2,3)
        •  often requires repeat endoscopy or surgical removal
        •  endoscopic clips can be used to help reduce the risk for migration, but evidence for their efficacy is limited
      •  gastrointestinal erosion requiring surgery(3)
      • less commonly
  • glue injection, endoscopic clips, and suturing have each been used to help close anastomotic leaks, often in conjunction with stenting(3)
    • glue injected into anastomotic leak may help seal the wound and prevent continued contamination
      •  both fibrin and cyanoacrylate-based glues have been used, with no evidence to suggest superiority of either
      •  glue injection alone might be an option for small leaks, but usually performed in addition to stenting
    •  endoscopic clips have limited data to support their use, though large, over-the-scope clips are reported to be effective in conjunction with stenting for some patients
    •  suturing is rarely performed for management of anastomotic leaks, but has been used in patients with late fistulas at high risk for failure with primary operative repair
  • in patients who develop an anastomotic leak after bariatric surgery(3)
    •  primary surgical repair > 2-3 days after leak or in patients with high peritoneal contamination is unlikely to be successful
    • management typically requires endoscopic therapies, such as
      • transluminal endoscopy
        •  characterized by entering the soiled peritoneal cavity through the leak, endoscopic debridement, necrosectomy, irrigation, and drainage
        •  drainage catheters can be left in the peritoneum and traverse the leak, exiting transnasally
        •  transnasal peritoneal drain can be removed after systemic and local contamination has resolved (the leak will seal if additional/continuing contamination is avoided)
      •  endoscopic stenting, with or without additional endoscopic glue injection, clips, or suturing

Endoscopic Vacuum Therapy (EVT)

Description and Indications

  • endoscopic vacuum therapy (EVT) (also called endoluminal vacuum therapy [EVAC]) is a minimally invasive technique for management of small or large bowel anastomotic leaks
    • technique involves placing a sponge attached to a nasogastric suctioning tube at the site of the leak
    • reported to allow for healing via macrodeformation, microdeformation, perfusion changes, control of exudate, and bacterial clearance, similar to healing mechanisms associated with wound vacuums for management of skin wounds
    • may be used to treat anastomotic leaks following surgery to any part of the gastrointestinal tract
    • Reference – World J Gastrointest Endosc 2019 May 16;11(5):329full-text
  • no standardized indications for use have been established, but some indications include
  • absolute contraindications include
  • relative contraindications include
  • in patients failing to respond to EVT, surgery is necessary to repair the leak (World J Gastrointest Endosc 2019 May 16;11(5):329full-text)

Procedure

  • endoscopic vacuum therapy (EVT for small bowel anastomotic leaks
    • start anesthesia, typically with endotracheal intubation for safe airway during passage of sponge
    • perform endoscopy to identify and characterize bowel wall defect
    • examine contaminated cavity with or without fluoroscopy to determine appropriate sponge size and placement
      • standard size sponge is 3-7 cm in length and 2-3 cm in diameter (sponge size is limited by diameter of esophagus)
      • defects < 1 cm without an associated cavity can typically be managed with intraluminal sponge placement
      • larger defects may require dilation in order to access the cavity and place the sponge extraluminally
    • place sponge at anastomotic leak with endoscope
      • introduce nasogastric tube (NGT) into patient’s nares and feed tube to posterior pharynx
      • remove NGT from mouth using finger or grasper instrument
      • secure sponge to tip of NGT using silk ties or permanent 2-0 silk or nonabsorbable sutures (or larger)
      • place nonabsorbable permanent suture at distal end of tube and tie into small loop to facilitate endoscopic placement and retrieval
      • place grasper through working channel of endoscope before insertion into patient’s mouth, then grasp short suture loop with device
      • lubricate sponge and endoscope and insert into patient’s mouth, moving endoscope distally (avoid trauma to upper esophageal sphincter during insertion)
        • if perforation is < 1 cm, move endoscope to intraluminal position at perforation site
        • if perforation is ≥ 1 cm, move endoscope through perforation into cavity (may require dilation to place sponge extraluminally)
      • advance grasper while withdrawing endoscope and deposit sponge in appropriate position
      • secure NGT to nose of patient
      • connect suction tubing to vacuum therapy unit and canister
      • attach NGT with sponge to canister tubing using custom adapter
      • set vacuum therapy pressure to 125-175 mmHg at continuous moderate intensity
      • vacuum therapy may be changed to intermittent suction (5 minutes on, 2 minutes off) if patient is uncomfortable
    • Reference – World J Gastrointest Endosc 2019 May 16;11(5):329full-text
  • EVT for large bowel anastomotic leaks
    • sedation or anesthesia may not be needed for EVT procedure for large bowel anastomotic leaks but is often used
    • in most cases, a diverting stoma proximal to the leak site is necessary prior to EVT for lower gastrointestinal tract leaks, which can be reversed following healing
    • perform flexible endoscopic procedure to
      • identify and characterize bowel wall defect
      • examine contaminated cavity to determine sponge size
    • perform lavage
    • cut sponges according to the size of the cavity adjacent to the leak (for large cavities, 2 or more sponges may be used)
    • insert colonoscope transanally to cavity
    • advance flexible overtube over colonoscope to end of cavity, and retract colonoscope
    • push sponge through fixed lubricated overtube to appropriate position
    • connect evacuation tube of sponge to vacuum bottle
    • electronic vacuum pump system may be used to maintain continuous suction at 125 mmHg
    • Reference – J Gastrointest Surg 2016 Feb;20(2):328
  • endoscopic vacuum therapy associated with a high rate of complete healing of anastomotic leakage and stoma reversal in patients with anastomotic leakage after colorectal surgery (level 2 [mid-level] evidence)
    •  based on systematic review of observational studies
    • systematic review of 17 cohort studies (15 retrospective, 2 prospective) evaluating EVT for management of anastomotic leakage following colorectal surgery in 276 patients
    • EVT associated with
      • 85.3% weighted mean success rate (complete healing of anastomotic leakage), with a median duration from EVT to complete healing of 47 days (range 40-105 days)
      • 75.9% weighted mean rate of stoma reversal
      • 11.1% weighted mean complication rate
    • factors associated with significantly increased risk for treatment failure included
      • preoperative radiation therapy (p = 0.018)
      • lack of diverting stoma prior to treatment (p = 0.009)
      • development of complications (p = 0.002)
      • male sex (p = 0.014)
    • Reference – BJS Open 2019 Apr;3(2):153full-text

Postprocedural Care

Complications

  • complications may include
    • pain, nausea, and vomiting from nasogastric tube
    • sponge dislocation
    • anastomotic strictures
    • bleeding
      • minor bleeding can occur after sponge exchange due to growth of granulation tissue into sponge (may be mitigated through more frequent sponge exchanges)
      • major bleeding is less common, and typically reported in upper gastrointestinal leaks (perform triple phase computed tomography to direct possible management)
    • Reference – World J Gastrointest Endosc 2019 May 16;11(5):329full-text
  • 20% complication rate reported with EVT, mainly from
  • endoscopic vacuum therapy associated with high anastomotic salvage rate in patients with colorectal anastomotic leaks (level 3 [lacking direct] evidence)
    •  based on systematic review of observational studies
    • systematic review of 36 cohort studies and case series (5 prospective cohorts, 4 retrospective cohorts, 25 case series, and 2 case reports) evaluating endoscopic salvage techniques in 388 stable patients with acute anastomotic leaks
    • meta-analysis not performed due to heterogeneity of study types and available data
    • 13 studies evaluated EVT in 197 patients
      • mean 88.8% (range 66.6%-100%) anastomotic salvage rate
      • very few complications (predominantly pain and stenosis) reported
    • Reference – Surg Endosc 2019 Apr;33(4):1049full-text

Revision Laparotomy

  • revision laparotomy is indicated in addition to antibiotics in patients with peritonitis and sepsis (following aggressive resuscitation if needed) and in those refractory to initial conservative management with antibiotics and drainage(2)
    •  goals include source control with washout and fecal diversion
    • fecal diversion may include
      •  taking down anastomosis with construction of an end colostomy or ileostomy
      •  leaving anastomosis in place and creating a proximal diversion with loop ileostomy
      •  repairing or revising anastomosis and constructing a proximal diversion
    •  minimally invasive surgery reported to have low morbidity and mortality compared to open technique in peritoneal lavage and ileostomy construction
  • salvage of anastomosis with loop ileostomy associated with decreased mortality compared to anastomotic takedown in patients needing surgical intervention after anastomotic colorectal leak (level 2 [mid-level] evidence)
    •  based on retrospective cohort study
    • 93 patients (mean age 67.9 years) with anastomotic colorectal leak treated with salvage of the anastomosis (with loop ileostomy) or anastomotic takedown as per surgeon’s preference were evaluated
      •  42% of patients had salvage of the anastomosis and loop ileostomy (either drainage of anastomosis and derivative loop ileostomy or reanastomosis and ileostomy)
      •  58% of patients had anastomotic takedown with construction of an end colostomy or end ileostomy
    • anastomotic leak defined as ≥ 1 of following
      •  generalized or localized peritonitis
      •  pelvic abscess
      •  discharge of feces, pus, or gas from drain or wound
    • comparing loop ileostomy vs. anastomotic takedown
      •  mortality 15% vs. 37% (p = 0.02)
      •  overall morbidity (composite of all complications) 87.2% vs. 75.9% (not significant)
    •  Reference – Am J Surg 2012 Nov;204(5):671

Complications and Prognosis

Complications

  • anastomotic leak can result in severe, life-threatening complications, including
    •  sepsis(1,2,3)
    •  peritonitis(2)
    •  local cancer recurrence in patients with surgery for colorectal cancer (cancer cells in resected bowel may implant extraluminally via the leak)(2)
  • other complications may include
    •  fistula formation(3)
    •  abscess(3)
    •  stricture formation(2)
  • anastomotic leak after Roux-en-Y gastric bypass associated with increased mortality and morbidity
    •  based on retrospective cohort study
    •  840 patients (median age 45 years) having Roux-en-Y gastric bypass were analyzed with follow-up of 11 months
    •  4.3% developed anastomotic leaks
    •  35 patients died during follow-up
    • comparing patients with anastomotic leak vs. no anastomotic leak
      •  mortality 14% vs. 4% (p = 0.01)
      •  overall complications 61% vs. 29% (p < 0.001)
      •  duration of hospital stay 24.5 days vs. 4.5 days (p < 0.001)
    • anastomotic leak associated with increased risk of complications including
      •  sepsis (odds ratio [OR] 27, 95% CI 2-472)
      •  renal failure (OR 16, 95% CI 3-99)
      •  small bowel obstruction (OR 11, 95% CI 2-68)
      •  internal hernia (OR 10, 95% CI 2-51)
      •  venous thromboembolism (OR 9, 95% CI 3-27)
      •  incisional hernia (OR 5, 95% CI 2-13)
    •  Reference – Arch Surg 2007 Oct;142(10):954

Prognosis

  •  most early leaks (occurring within 3 days of surgery) in stable patients without sepsis can be successfully managed with conservative treatment(3)
  • mortality
    •  reported in about 0.5%-29% of patients with anastomotic leak(1,3)
    •  0.4% mortality due to anastomotic leak in systematic review of 39 studies with 24,232 patients (8.6% of whom developed anastomotic leak) following rectal carcinoma resection (Int Surg 2014 Mar-Apr;99(2):112full-text)
    • anastomotic leak associated with increased risk for local recurrence and reduced overall and cancer-specific survival in patients with curative anterior resection for rectal cancer
      •  based on systematic review of observational studies
      •  systematic review of 14 cohort studies (7 prospective, 7 retrospective) evaluating effects of anastomotic leak in 11,353 patients after curative anterior resection for rectal cancer
      • anastomotic leak associated with
        •  increased risk for local recurrence (hazard ratio [HR] 1.71, 95% CI 1.22-2.38) in analysis of 9 studies with 6,638 patients
        •  decreased overall survival (HR 1.67, 95% CI 1.19-2.35) in analysis of 10 studies with 9,309 patients, results limited by significant heterogeneity
        •  decreased cancer-specific survival (HR 1.3, 95% CI 1.08-1.56) in analysis of 7 studies with 6,516 patients
      •  no significant differences in risk for distant recurrence in analysis of 3 studies with 2,397 patients
      •  Reference – World J Surg 2017 Jan;41(1):277full-text
    • open gastric bypass surgery and jejunojejunostomy leaks each associated with increased mortality in patients with anastomotic leak following Roux-en-Y gastric bypass surgery
      •  based on retrospective cohort study
      • 3,828 patients with Roux-en-Y gastric bypass procedure were evaluated
        •  61.1% had open gastric bypass
        •  28.2% had laparoscopic gastric bypass
        •  10.7% had revision procedure
      •  3.9% of patients developed anastomotic leak, with leak-related mortality 0.6%
      • comparing mortality rates in patients who developed leaks
        •  24.6% with open surgery vs. 8.9% with laparoscopic surgery (p = 0.025)
        •  40% with jejunojejunostomy leaks vs. 9% with gastrojejunostomy leaks (p = 0.005)
      •  Reference – J Gastrointest Surg 2007 Jun;11(6):708
  • anastomotic leak following curative resection for colon cancer associated with delay in or decreased likelihood of receiving chemotherapy, and increased risk of disease recurrence and all-cause mortality at 5 years
    •  based on retrospective cohort study
    •  9,333 patients who had surgery for colon cancer with primary anastomosis without protecting ostomy were analyzed
    •  6.4% had anastomotic leak
    •  744 patients died within 120 days
    • 8,589 patients who survived > 120 days were followed for median 5.3 years
      •  14.9% had distant recurrence
      •  10% had local recurrence
    • compared to no anastomotic leak, anastomotic leak associated with
      •  greater delay to initial administration of chemotherapy (p < 0.001)
      •  decreased likelihood of receiving adjuvant chemotherapy (adjusted hazard ratio [HR] 0.58, 95% CI 0.45-0.74)
      •  increased risk of distant recurrence (adjusted HR 1.42, 95% CI 1.13-1.78)
      •  increased all-cause mortality (adjusted HR 1.2, 95% CI 1.01-1.44)
    •  Reference – Ann Surg 2014 May;259(5):930, commentary can be found in Ann Surg 2016 Feb;263(2):e17

Prevention

Surgical Techniques Associated With Reduced Risk for Anastomotic Leak

  • mechanical bowel preparation plus oral antibiotics
    • oral antibiotics plus mechanical bowel preparation, but not mechanical bowel preparation alone, may reduce risk for anastomotic leak and surgical site infection (SSI) after elective colorectal surgery (level 2 [mid-level] evidence)
      •  based on retrospective cohort study
      • 32,359 patients (mean age 61 years, 52% female) having elective colorectal resection between 2012 and 2014 were evaluated at 30 days postoperatively
        •  32.9% had mechanical bowel preparation plus oral antibiotics
        •  36.6% had mechanical bowel preparation alone
        •  3.8% had oral antibiotics alone
        •  26.7% had no bowel preparation
      •  9% of patients (2,896) developed SSI
      •  27,698 patients included in propensity-adjusted analyses for outcomes
      • compared to no bowel preparation
        • reduced risk of anastomotic leak with
          •  mechanical bowel preparation plus oral antibiotics (adjusted odds ratio [OR] 0.53, 95% CI 0.43-0.64)
          •  oral antibiotics alone (adjusted OR 0.37, 95% CI 0.21-0.67)
        • reduced risk of SSI with
          •  mechanical bowel preparation plus oral antibiotics (adjusted OR 0.45, 95% CI 0.4-0.5)
          •  oral antibiotics alone (adjusted OR 0.49, 95% CI 0.38-0.64)
        • mechanical bowel preparation plus oral antibiotics associated with reduced risk of
          •  postoperative ileus (adjusted OR 0.83, 95% CI 0.43-0.64)
          •  hospital readmission (adjusted OR 0.87, 95% CI 0.79-0.97)
      • no significant differences in
        •  rate of anastomotic leak comparing mechanical bowel preparation alone to no bowel preparation
        •  rate of SSI comparing mechanical bowel preparation alone to no bowel preparation
        •  risk of cardiac or renal complications with any type of bowel preparation compared to no bowel preparation
      •  Reference – Ann Surg 2018 Apr;267(4):734
  • protective ileostomy may reduce risk of anastomotic leak in patients having laparoscopic surgery for rectal cancer (level 2 [mid-level] evidence)
    •  based on systematic review of observational studies and with incomplete assessment of trial quality
    • systematic review of 1 randomized trial and 6 cohort studies comparing protective ileostomy vs. no ileostomy in 1,092 patients having laparoscopic surgery for rectal cancer
    • mean patient age ranged from 52 to 71 years
    • trial quality assessment did not include allocation concealment
    • compared to no ileostomy, protective ileostomy associated with reduced risk of anastomotic leak (risk ratio 0.47, 95% CI 0.3-0.73) in analysis of all studies
    • no significant differences in
      • postoperative hospital stay in analysis of 4 studies with 795 patients, results limited by significant heterogeneity
      • risk of reoperation in analysis of 3 studies with 609 patients
      • risk of wound infection after surgery in analysis of 3 studies with 304 patients
      • operative time in analysis of 3 studies with 679 patients, results limited by significant heterogeneity
    • Reference – World J Surg Oncol 2021 Nov 4;19(1):318full-text
  • defunctioning stoma may reduce risk of anastomotic leak and reoperation after anterior resection in patients with rectal cancer (level 2 [mid-level] evidence)
    •  based on Cochrane review of trials with methodologic limitations
    •  Cochrane review of 6 randomized trials comparing use of defunctioning stoma vs. no stoma in 648 patients with rectal cancer having low anterior resection
    •  defunctioning stomas are temporary stomas created to protect a more distal anastomosis at particular risk of leakage or breakdown in order to improve healing
    • all trials had ≥ 1 methodologic limitation including
      •  unclear randomization
      •  no blinding
      •  no intention-to-treat analysis
    • defunctioning stoma associated with lower risk of
      • anastomotic leak in analysis of all trials
        •  risk ratio (RR) 0.33 (95% CI 0.21-0.53)
        •  NNT 7-11 with leak in 20% of controls
      • urgent reoperation in analysis of all trials
        •  RR 0.23 (95% CI 0.12-0.42)
        •  NNT 6-10 with urgent reoperation in 20% of controls
    •  no significant difference in mortality in analysis of all trials
    • Reference – Cochrane Database Syst Rev 2010 May 12;(5):CD006878
    •  similar results in systematic review of 25 studies evaluating defunctioning stoma after low anterior resection for rectal cancer (Br J Surg 2009 May;96(5):462), commentary can be found in Br J Surg 2009 Nov;96(11):1374
  • transanal tubes may help prevent anastomotic leak in patients having rectal cancer surgery with anastomosis (level 2 [mid-level] evidence)
    •  based on systematic review of trials with methodologic limitations
    •  systematic review of 7 trials (1 randomized trial, 6 nonrandomized trials) evaluating transanal tubes in 1,609 adults having rectal cancer surgery with anastomosis
    •  transanal tubes differed in types, materials, sizes, position, and indwelling time
    • all trials had ≥ 1 of these methodologic limitations
      •  lack of randomization
      •  unclear randomization sequence generation
      •  lack of allocation concealment
      •  lack of blinding
      •  small sample size
    • compared to no transanal tubes, use of transanal tubes associated with reduced
      •  anastomotic leak (relative risk 0.38, 95% CI 0.25-0.58) in analysis of 7 trials with 1,609 patients
      •  reoperation (relative risk 0.31, 95% CI 0.19-0.53) in analysis of 7 trials with 1,609 patients
    •  no significant differences in mortality in analysis of 6 trials with 1,404 patients
    •  Reference – World J Surg 2017 Jan;41(1):267
  • robotic bariatric surgery may decrease risk of anastomotic leak compared to laparoscopic bariatric surgery (level 2 [mid-level] evidence)
    •  based on systematic review of mostly observational studies
    •  systematic review of 34 studies (2 randomized trials and 32 observational studies) comparing robotic vs. laparoscopic bariatric surgery in patients with obesity
    •  bariatric surgery types included Roux-en-Y gastric bypass, sleeve gastrectomy, and gastric banding
    • robotic bariatric surgery associated with
      •  decreased risk of anastomotic leak (odds ratio 0.5, 95% CI 0.3-0.81) in analysis of 19 studies, results limited by significant heterogeneity
      •  increased operative time (p < 0.0005) in analysis of 19 studies
    •  Reference – Obes Surg 2016 Dec;26(12):3031
  • end-to-side anastomosis associated with lower anastomotic leak rate than end-to-end anastomosis after anterior resection of rectal cancer (level 2 [mid-level] evidence)
    •  based on randomized trial with allocation concealment not stated
    •  77 patients having anterior resection of rectal cancer (T1-T2 ≤ 15 cm from anal verge) were randomized to end-to-side anastomosis vs. end-to-end anastomosis using left colon
    •  anastomotic leak in 5% with end-to-side anastomosis vs. 29.2% with end-to-end anastomosis (p = 0.005, NNT 5)
    •  Reference – J Surg Oncol 2009 Jan 1;99(1):75

Surgical Techniques Not Associated With Reduced Risk for Anastomotic Leak

  •  omentoplasty(1)
  • mechanical bowel preparation alone (without oral antibiotics)
    • mechanical bowel preparation alone (without oral antibiotics) may not reduce rate of anastomotic leak, peritonitis, or wound infection in patients having elective colorectal surgery (level 2 [mid-level] evidence)
      •  based on 1 Cochrane review and 1 systematic review with wide confidence intervals that cannot exclude clinically important differences
      •  systematic review of 18 randomized trials evaluating mechanical bowel preparation (without oral antibiotics) in 5,805 patients having elective colorectal surgery
      •  wide confidence intervals includes possibility of both benefit and harm
      • no significant difference comparing mechanical bowel preparation to no preparation in
        •  overall anastomotic leak in analysis of 13 trials with 4,533 patients
        •  anastomotic leak after low anterior resection in subgroup analysis of 7 trials with 846 patients
        •  anastomotic leak after colonic surgery in subgroup analysis of 8 trials with 3,147 patients
        •  peritonitis in analysis of 10 trials with 3,983 patients
        •  reoperation in analysis of 11 trials with 4,319 patients
        •  wound infection in analysis of 13 trials with 4,595 patients
        •  infectious extra-abdominal complications in analysis of 6 trials with 3,575 patients
      •  similar results in sensitivity analyses of trials with adequate randomization or in patients with anastomosis
      •  no significant difference in any outcomes comparing mechanical bowel preparation to rectal enema in 5 trials with 1,210 patients
      • Reference – Cochrane Database Syst Rev 2011 Sep 7;(9):CD001544full-text
      •  consistent results found in systematic review of 18 trials evaluating mechanical bowel preparation without the use of oral antibiotics (Dis Colon Rectum 2015 Jul;58(7):698), commentary can be found in Dis Colon Rectum 2016 Aug;59(8):e421
    •  commentary on mechanical bowel preparation outcomes without oral antibiotics can be found in Dis Colon Rectum 2016 Aug;59(8):e421
  • C-seal biodegradable intraluminal sheath does not prevent anastomotic leak in patients having colorectal resection with stapled anastomosis (level 1 [likely reliable] evidence)
    •  based on randomized trial
    • 485 adults having elective colorectal resection with stapled circular anastomosis < 15 cm from anal verge were randomized to biodegradable intraluminal soft sheath (C-seal) vs. no C-seal (control)
      •  biodegradable C-seal was stapled to afferent bowel loop proximal to anastomosis for preventing intestinal leakage in case of anastomotic dehiscence
      •  trial protocol amended to include mechanical oral bowel preparation before use of C-seal following safety analysis in first 46 patients
    •  primary endpoint was anastomotic leak requiring invasive treatment within 30 days after primary surgery
    •  trial terminated early for futility at first interim analysis using prespecified stopping rule
    •  402 patients (median age 65 years) included in analysis
    • comparing C-seal vs. control
      •  primary endpoint in 10.4% vs. 5% (p = 0.06)
      •  dismantled anastomoses in 5% vs. 1.5% (p = 0.09)
      •  defunctioning stomas in 50% vs. 50.5% (not significant)
      •  median interval between surgery and anastomotic leak 6 days vs. 4.5 days (not significant)
      •  median duration of hospital stay 14 days vs. 13 days (not significant)
    •  Reference – C-seal trial (Br J Surg 2017 Jul;104(8):1010)
  • reinforcement of stapled circular anastomosis > 5 cm from anal verge with bioabsorbable device may not reduce anastomotic complications in patients having colorectal surgery (level 2 [mid-level] evidence)
    •  based on single-blind randomized trial
    •  302 patients (median age 66 years) having elective colorectal surgery with stapled circular anastomosis > 5 cm from anal verge were randomized to staple line reinforcement with bioabsorbable device vs. no reinforcement and followed for 12 months
    •  93% of patients received assigned intervention of bioabsorbable staple line reinforcement and were included in analyses
    • no significant differences between groups in
      •  anastomotic complications, including leak, hemorrhage, and stenosis
      •  length of hospital stay
      •  reoperation rate
    •  Reference – Dis Colon Rectum 2014 Oct;57(10):1195
  • inconsistent evidence for reduced risk of anastomotic bowel leaks for
    •  application of medical adhesives including cyanoacrylate and fibrin adhesives (Pol Przegl Chir 2017 Apr 30;89(2):49)
    •  gastric tube placement in Roux-en-Y gastric bypass (Pol Przegl Chir 2017 Apr 30;89(2):49)
    • prophylactic drains
      • prophylactic anastomotic drainage does not appear to prevent colorectal surgery complications compared to nondrainage regimens (level 2 [mid-level] evidence)
        •  based on Cochrane review without individual trial quality assessment reported
        •  systematic review of 6 randomized trials comparing prophylactic anastomotic drainage vs. no drainage in 1,140 patients receiving colorectal anastomoses
        • no significant differences comparing drainage and nondrainage regimens in
          •  mortality (3%-4%)
          •  clinical anastomotic dehiscence (1%-2%)
          •  radiological anastomotic dehiscence (3%-4%)
          •  wound infection (5%)
          •  reintervention (5%-6%)
          •  extra-abdominal complications (6%-7%)
        •  Reference – Cochrane Database Syst Rev 2004 Oct 18;(4):CD002100full-text (review updated 2008 Aug 5), also published in Colorectal Dis 2006 May;8(4):259
      • prophylactic pelvic drainage might reduce risk for anastomotic leak in patients having anterior rectal resection with extraperitoneal colorectal anastomoses (level 2 [mid-level] evidence)
        •  based on systematic review of mostly observational studies
        •  systematic review of 8 studies (3 randomized trials and 5 observational studies) evaluating prophylactic pelvic drainage in 2,277 patients having anterior rectal resection with extraperitoneal anastomoses
        • compared to no drainage, pelvic drainage associated with
          •  reduced anastomotic leak (odds ratio [OR] 0.51, 95% CI 0.36-0.73) in analysis of 3 randomized trials and 5 observational studies with 2,277 patients
          •  no significant difference in anastomotic leak in analysis of 3 randomized trials with 291 patients (OR 0.98, 95% CI 0.49-1.99)
        •  Reference – Colorectal Dis 2014 Feb;16(2):O35

Guidelines and Resources

Guidelines

International Guidelines

  •  Enhanced Recovery After Surgery Society (ERAS) guideline on perioperative care in elective colonic surgery can be found in World J Surg 2019 Mar;43(3):659.

United States Guidelines

  •  American Association of Clinical Endocrinologists/American College of Endocrinology/The Obesity Society/American Society for Metabolic and Bariatric Surgery/Obesity Medicine Association/American Society of Anesthesiologists (AACE/ACE/TOS/ASMBS/OMA/ASA) clinical practice guideline on perioperative nutritional, metabolic, and nonsurgical support of bariatric surgery patient can be found in Obesity (Silver Spring) 2020 Apr;28(4):O1.
  • American Society of Colon and Rectal Surgeons/Society of American Gastrointestinal and Endoscopic Surgeons (ASCRS/SAGES) guideline on enhanced recovery after colon and rectal surgery can be found in Surg Endosc 2023 Jan;37(1):5full-text.
  •  Surgical Infection Society/Infectious Diseases Society of America (SIS/IDSA) guideline on diagnosis and management of complicated intra-abdominal infection in adults and children can be found in Clin Infect Dis 2010 Jan 15;50(2):133, correction can be found in Clin Infect Dis 2010 Jun 15;50(12):1695, commentary can be found in Clin Infect Dis 2010 Sep 15;51(6):757.

United Kingdom Guidelines

  • National Institute for Health and Care Excellence (NICE) medical technologies guidance on Endo-SPONGE for treating low rectal anastomotic leak can be found at NICE 2021 Dec 16:MTG63PDF.

Review Articles

  •  To search MEDLINE for (Anastomotic Leak) with targeted search (Clinical Queries), click therapydiagnosis, or prognosis.

Patient Information

  •  DynaMed Editors have not identified patient education materials that meet our criteria for inclusion (freely accessible, nonpromotional, topic-specific). We will continue to search for acceptable materials and welcome your suggestions.

References

General References Used

The references listed below are used in this DynaMed topic primarily to support background information and for guidance where evidence summaries are not felt to be necessary. Most references are incorporated within the text along with the evidence summaries.

  1. An V, Chandra R, Lawrence M. Anastomotic Failure in Colorectal Surgery: Where Are We at? Indian J Surg. 2018 Apr;80(2):163-170full-text .
  2. Thomas MS, Margolin DA. Management of Colorectal Anastomotic Leak. Clin Colon Rectal Surg. 2016 Jun;29(2):138-44full-text.
  3. Bhayani NH, Swanström LL. Endoscopic therapies for leaks and fistulas after bariatric surgery. Surg Innov. 2014 Feb;21(1):90-7.
15585

Sign up to receive the trending updates and tons of Health Tips

Join SeekhealthZ and never miss the latest health information

15856