SEARCH

SEARCH BY CITATION

Keywords:

  • Crohn's disease;
  • postoperative recurrence;
  • prevention

Abstract

  1. Top of page
  2. Abstract
  3. Search Strategy
  4. NATURAL HISTORY OF POSTOPERATIVE RECURRENCE
  5. RISK FACTORS FOR POSTOPERATIVE RECURRENCE
  6. DIAGNOSIS OF RECURRENCE AND MONITORING
  7. DRUG PREVENTION OF DISEASE RECURRENCE
  8. OPTIMAL POSTOPERATIVE TREATMENT STRATEGY
  9. REFERENCES

Despite improved immunosuppressive therapy, surgical resection is still often required for uncontrolled inflammatory disease and the stenosing and perforating complications of Crohn's disease. However, surgery is not curative. A majority of patients develop disease recurrence at or above the anastomosis. Subclinical endoscopically identifiable recurrence precedes the development of clinical symptoms; identification and treatment of early mucosal recurrence may therefore prevent clinical recurrence. Therapy to achieve mucosal healing should now be the focus of postoperative therapy. A number of clinical risk factors for the development of earlier postoperative recurrence have been identified, and reasonable evidence is now available regarding the efficacy of drug therapies in preventing recurrence. This evidence now needs to be incorporated into prospective treatment strategies. (Inflamm Bowel Dis 2011;)

Approximately 80% of patients with Crohn's disease (CD) require surgery at some point in their life1 and of these up to 70% require a second operation.2–5 The number of patients requiring surgery has not declined significantly over the past few decades despite the use of immunosuppressive therapy.6 The advent of biologic therapy has resulted in improved disease management and short-term improvement in operative rates but the long-term impact of biologic therapy on operative rates is unknown. Although surgery is effective in addressing the stenosing and perforating complications of CD, and improves quality of life,7 it is not curative.

There is currently no consensus as to the optimal approach to prevent the postoperative recurrence of CD. Attention has focused on 1) operative technique, in particular the type of surgical anastomosis; 2) assessment of patient risk of recurrence; 3) drug therapy; and 4) endoscopic monitoring to detect early mucosal disease, as endoscopically observed disease recurrence predicts clinical recurrence.8–10

This systematic review was undertaken to summarize the current evidence regarding: 1) the natural history of postoperative recurrence; 2) risk factors for recurrence; 3) diagnosis of recurrence and the role of disease monitoring; and 4) drug trials of medical prophylaxis. Our aim was to collate this information with a view to developing an optimal postoperative management strategy.

Search Strategy

  1. Top of page
  2. Abstract
  3. Search Strategy
  4. NATURAL HISTORY OF POSTOPERATIVE RECURRENCE
  5. RISK FACTORS FOR POSTOPERATIVE RECURRENCE
  6. DIAGNOSIS OF RECURRENCE AND MONITORING
  7. DRUG PREVENTION OF DISEASE RECURRENCE
  8. OPTIMAL POSTOPERATIVE TREATMENT STRATEGY
  9. REFERENCES

A systematic electronic search of the English literature for the period January 1970 to December 2010 was conducted using Medline (EBSCO host) and the Cochrane Library. The search was undertaken to identify articles on postoperative recurrence of CD pertaining to: 1) natural disease history postoperatively; 2) risk factors for recurrence; 3) diagnosis of recurrence and monitoring; and 4) drug trials of medical prophylaxis.

The search strategy employed a combination of Medical Subject (MeSH) headings and keywords as follows: “CD,” “postoperative care,” “postoperative complications,” “postoperative period,” “post,” “following,” “surgery,” “operation,” “treatment outcome,” “treatment failure,” “disease-free survival,” “medical futility,” “prognosis,” “epidemiology,” “outcome,” “course,” “complication,” “history,” “risk factors,” “smoking,” “perforating disease,” “penetrating disease,” “fistula,” “duration of disease,” “disease duration,” “previous resection,” “prior resection,” “family history,” “disease location,” “anastomosis,” “surgical adverse event,” “disease extent,” “age of onset,” “risk factor,” “gender,” “genetic,” “anastomotic configuration,” “laparoscopic,” “open,” “length of resection,” “resection length,” “microscopic disease,” “granuloma,” “myenteric plexitis,” “NOD2,” “ATG6L1,” “IRGM,” “diagnosis,” “diagnostic techniques and procedures,” “assessment,” “colonoscopy,” “endoscopy,” “fecal calprotectin,” “fecal lactoferrin,” “SICUS,” “small intestine contrast enhanced ultrasound,” “capsule endoscopy,” “MRI,” “magnetic resonance imaging,” “small bowel MRI,” “ultrasound,” mucosal healing, “serologic markers,” “anti-microbial antibiodies,” “aminosalicylate,” “5ASA,” “sulfasalazine,” “sulphasalazine,” “mesalamine,” “mesalazine,” “azathioprine,” “thiopurine,” “6-mercaptopurine,” “enteral nutrition,” “enteral nutrition,” “infliximab,” “adalimumab,” “certolizumab,” “natalizumab,” “drug therapy,” “prophylaxis.” Boolean operators (“not,” “and,” “or”) were also used in succession to narrow or widen the search. Total results from all searches were 646. Two authors screened the abstracts and identified potentially relevant articles. Additional studies were identified via a manual review of the reference list of identified studies and review articles. There were a total of 478 articles studied to construct this review.

NATURAL HISTORY OF POSTOPERATIVE RECURRENCE

  1. Top of page
  2. Abstract
  3. Search Strategy
  4. NATURAL HISTORY OF POSTOPERATIVE RECURRENCE
  5. RISK FACTORS FOR POSTOPERATIVE RECURRENCE
  6. DIAGNOSIS OF RECURRENCE AND MONITORING
  7. DRUG PREVENTION OF DISEASE RECURRENCE
  8. OPTIMAL POSTOPERATIVE TREATMENT STRATEGY
  9. REFERENCES

Recurrence of CD affects the majority of patients following surgery involving a resection.2–5, 8–12 Resection is not curative; even if all macroscopically involved intestine is removed, disease recurs, most commonly at or above the anastomosis.8–10 Recurrent disease is initially characterized by subclinical endoscopic lesions that precede the development of clinical symptoms.9 Moreover, the severity of subclinical endoscopic lesions predicts the likelihood of developing subsequent recurrent clinical disease and the requirement for reoperation.9 Endoscopically identified recurrence, which refers to new lesions characterized by ulceration or inflammation, can be visualized at the neoterminal ileum and anastomosis, using colonoscopy, within weeks to months of surgery.8, 9 Within 1 year postoperatively, endoscopic recurrence is observed in up to 90% of patients, clinical recurrence of CD is observed in up to 30% of patients, and 5%–10% will have had further surgery.8 Thereafter, the initial cumulative probability of experiencing clinical recurrence is ≈10% per year,13 so that by 5 years postoperatively up to 50% of patients will have developed clinical recurrence9, 10, 14, 15 and by 10 years of their initial operation up to 70% of patients will have required reoperation.2–5, 15–23

The natural history of CD postoperatively differs depending on whether the bowel has been anastomosed as opposed to whether it has been defunctioned with the creation of a stoma. The latter is often undertaken in the setting of severe perforating complications of CD. An ileostomy may involve the creation of either a “loop” ileostomy, with the bowel remaining in continuity but with the fecal stream diverted to the stoma, or an end ileostomy in which the distal bowel is completely disconnected from the proximal bowel. When the fecal stream is diverted with a covering proximal “loop” ileostomy the ileocolonic anastomosis is protected from the development of mucosal lesions.24 Lower postoperative recurrence rates in the range of 20%–30% over 10 years are found among patients with an ileostomy, particularly if they have undergone colectomy with an end-ileostomy.25 However, on restoration of continuity of the bowel, with reexposure of the distal bowel to the fecal stream, inflammation recurs at or above the anastomosis in as little as 1 week.24

RISK FACTORS FOR POSTOPERATIVE RECURRENCE

  1. Top of page
  2. Abstract
  3. Search Strategy
  4. NATURAL HISTORY OF POSTOPERATIVE RECURRENCE
  5. RISK FACTORS FOR POSTOPERATIVE RECURRENCE
  6. DIAGNOSIS OF RECURRENCE AND MONITORING
  7. DRUG PREVENTION OF DISEASE RECURRENCE
  8. OPTIMAL POSTOPERATIVE TREATMENT STRATEGY
  9. REFERENCES

Several clinical factors have been evaluated for their potential value in predicting the disease course postoperatively.4, 26–38 Using clinical risk factors to identify patients at greatest risk of recurrence may be helpful in targeting those who would most benefit from monitoring and prophylaxis. A list of patient-related, disease-related, and operative factors that have been evaluated within this review as potential risk factors for postoperative recurrence is shown in Table 1.

Table 1. Factors That Have Been Evaluated as Potential Risk Factors for Postoperative Recurrence
Patient-related factors
 Smoking
 Gender
 Family history of inflammatory bowel disease
Disease-related factors
 Age at disease onset
 Duration of disease prior to operation
 History of prior resection
 Disease location
 Disease behavior - perforating disease
 Disease extent/length of resection
 Genetic factors: NOD2/CARD15
 Serology
 Granulomas
 Myenteric plexitis
Operation-related factors
 Laparoscopy versus open surgery
 Type of anastomosis
 Resection margins
 Strictureplasty
 Blood transfusions
 Postoperative complications

Patient-related Factors

Smoking

Smoking is the only modifiable patient-related factor that has consistently emerged as an independent risk factor for postoperative recurrence. The risk of reoperation appears to increase with the number of cigarettes smoked per day.33, 39–42 In a meta-analysis by Reese et al,43 which included 16 studies involving 2962 patients, smoking was found to approximately double the risk of clinical and surgical recurrence. At 10 years postoperative there was no significant difference in reoperation rate (odds ratio [OR] = 0.30; 95% confidence interval [CI] = 0.09, 1.07; P = 0.10) or postoperative acute relapses (OR = 1.54; 95% CI = 0.78, 3.02; P = 0.21) between exsmokers and nonsmokers, indicating that smoking cessation reduces clinical and surgical recurrence rates. Smoking cessation has also been demonstrated to reduce recurrence rates of CD in clinical drug trials for active disease.44, 45

Smoking has also been shown to increase the risk of endoscopic recurrence, with macroscopic lesions found in the neoterminal ileum of 70% of smokers 1 year after surgery compared with 35% of nonsmokers and 27% of exsmokers.46 Endoscopic recurrence rates appear to be equivalent, or similar, for exsmokers and nonsmokers.34, 47, 48

Gender

Gender does not appear to be a risk factor for recurrence. While some early studies suggested a higher recurrence rate among women than men,4, 49, 50 more recent studies suggest that they are equivalent.2, 9, 35, 51, 52–56

Family History of Inflammatory Bowel Disease (IBD)

Studies that have reviewed the impact of a family history of CD or IBD on postoperative recurrence have shown mixed results. A study by Ryan et al44 showed a trend toward a higher rate of postoperative recurrence among those with a family history of CD. Unkart et al33 found that a family history of IBD was associated with an increased risk of reoperation (hazard ratio [HR] 2.24, 1.16–4.30, P = 0.016). In contrast, there was no association between family history and postoperative recurrence in studies by Chardavoyne et al2 and Kurer et al.57

Disease-related Factors

Age at Onset of Disease

A number of studies have indicated that a young age of onset is a predictor of a more aggressive phenotype in CD.58, 59 However, the impact of a younger age of disease onset on postoperative recurrence has been more difficult to interpret because of the potential confounding effect of longer follow-up among younger patients. Some studies indicate that younger age of disease onset is associated with an increased incidence of recurrence, and in particular reoperation,36, 37, 44, 49, 55, 60 whereas other studies have not shown an association.2, 9, 38, 52, 53, 56

Duration of Disease Prior to Operation

Several studies have suggested that a “short” duration of disease prior to operation is associated with an increased risk of postoperative recurrence.2, 31, 35, 38, 50, 60, 61 However, this trend has not been observed in a number of other studies.9, 36, 51, 52, 62, 63 The lack of consistency between studies may relate to variation in the definition of a “short” duration, making comparisons between studies difficult.

In a prospective study by Poggioli et al,38 a disease duration of less than 6 years was associated with an increased incidence of postoperative recurrence compared with those with a disease duration of greater than 6 years. Other studies have similarly found that a disease duration of less than 10 years is associated with an increased risk of postoperative recurrence.2, 35 The association between short duration of disease and postoperative recurrence was particularly strong in a study by Yamamoto et al22 in which arbitrary time intervals between disease onset and primary surgery of less than 1 year, between 1 and 10 years, and greater than 10 years were defined. These latter studies suggest that patients with a short duration of disease prior to surgery have a more aggressive disease course.

History of Prior Resection

A number of observational studies and one randomized controlled trial have shown that a history of prior resection is a risk factor associated with postoperative recurrence.16, 17, 64, 65 The impact of previous resection on recurrence was evaluated in a randomized controlled trial comparing azathioprine (2 mg/[kg/day]) or mesalamine (3 g/day) for 24 months postoperatively.66 Although there was no difference in reoperation rates between the two groups overall, azathioprine was more effective than mesalamine in preventing clinical recurrence in those with a previous intestinal resection (OR, 4.83; 95% CI, 1.47–15.8), suggesting that those with a history of prior resection require more intensive postoperative prophylaxis to prevent clinical recurrence.

Disease Location Prior to Resection

The impact of anatomical location of disease prior to resection has been explored in a number of studies, with mixed results.2, 35, 38, 51, 54–56, 63 Small bowel disease55, 67–69 and ileocolonic54, 56 disease have been reported to be associated with an increased risk of recurrence in the majority of these studies. In a European, population-based, inception cohort upper gastrointestinal tract disease phenotype at diagnosis was associated with an increased risk of both surgical and nonsurgical recurrence.70 In a Japanese study by Keh et al,71 when compared with ileocecal disease, jejunal disease location was found to be associated with a high rate of early disease recurrence, but there was no significant difference in long-term recurrence rates between the two locations. Only one study found that those with predominantly large bowel involvement (n = 56) had an increased rate of reoperation (45%) compared with a rate of 32% in those with small bowel involvement (n = 94) and 35% in those with both small and large bowel involvement (n = 37).2 Overall, there appears to be no consistent relationship between disease location and rate of recurrence, although disease involving the proximal small bowel is widely considered to represent a high risk clinical situation because of the potential for developing short bowel syndrome following resection.

Disease Behavior: Perforating Disease

Perforating disease, a phenotype characterized by abscess, fistula, or free perforation, appears to be an independent risk factor for postoperative recurrence.27, 30, 72–75 Although previous studies found that recurrence rates were similar for patients with perforating and nonperforating disease,26–28, 30, 60, 79, 80 a meta-analysis of 13 studies by Simillis et al73 found that perforating disease was associated with a increased rate of both clinical and surgical recurrence (HR: 1.50, 95% CI: 1.16–1.93) compared with nonperforating indications for surgery. However, there was significant heterogeneity among the 13 studies reviewed.73 Nonetheless, an increased risk of both clinical and surgical recurrence in patients with perforating disease has been confirmed in more recent studies.74, 75

A study by Sachar et al74 found that perforating disease can be underappreciated preoperatively. Of 22 patients with perforating disease 12 (55%) developed clinical recurrence within less than 3 years. The probability of chance accounting for this difference between 0 and 55% between those with and without perforating disease was 0.002 by the Fisher Exact Test. All of those found intraoperatively to have B3 disease had been preoperatively defined as having Montreal Classification76 B2 disease, whereas perforating disease was only identified intra- or postoperatively.74 With the increasing use of magnetic resonance imaging (MRI) and computed tomography (CT) scanning preoperatively, local perforating disease may now be increasingly recognized. Whether subclinical perforating disease carries the same increased relative risk as overt free perforation is unknown.

A number of studies have suggested that the indication for reoperation (perforating vs. nonperforating) is often the same as the index operation.27, 38, 77–79 However, concordance between the indication for initial operation and reoperation has not been observed in all studies.80

Disease Extent and Length of Resection

Disease extent influences the length of resection but the impact that disease extent and resection length have on the postoperative recurrence rate is less clear. Some studies have suggested that a longer resection length, especially at the initial operation, is associated with an increased rate of postoperative recurrence.81–84 Other studies indicate that resection length makes no difference to postoperative recurrence rates.3, 51, 60, 85

Drug Therapy Prior to Resection

It has been suggested that the deliberate use of antiinflammatory drugs immediately preoperatively may decrease the inflammatory burden and lead to a decreased length of surgical resection, or a reduction in the recurrence rates postoperatively. One current study which may address this question is the LIR!C-trial, a controlled trial comparing infliximab treatment or laparoscopic ileocolic resection among patients with recurrent distal ileal CD.86

Genetic Factors

Of the large number of IBD susceptibility loci found thus far in genome-wide association studies, the NOD2/CARD15 mutations have been most studied, and may have the greatest impact on progression to initial surgery and risk of postoperative recurrence. NOD2/CARD15 mutations have been associated with fibrostenosing phenotype, aggressive disease, and ileocecal resection.87 In a study of 170 patients of which 70 had had prior resections, Alvarez-Lobos et al88 found that patients carrying NOD2/CARD15 variants had a higher risk of surgical recurrence (OR, 3.29; 95% CI, 1.13–9.56), and reoperation was needed at an earlier time (P = 0.03). In other studies NOD2 risk variants, independently, had no effect on need for earlier reoperation.89–91 However, Gorbe et al89, 90 found that patients with NOD2 risk alleles who smoked exhibited a 3-fold higher likelihood for an earlier second ileocolic resection (HR = 3.3, P = 0.022) and had a shorter median time to second ileocolic resection compared with those with NOD2 risk alleles who were nonsmokers (8 and 19 years, respectively; P = 0.007). Two studies have shown that particular frameshift mutations in L1007fs of NOD2/CARD15 have been associated with reduced postoperative disease-free interval and time to reoperation.92, 93

Serology

The prognostic value of serologic markers in the postoperative setting is yet to be fully elucidated. Serologic markers that may be of prognostic significance include antibody responses to microbial antigens such as anti-Saccharomyces cerevisiae antibodies (ASCA), E. coli (Omp-C), Pseudomonas (I2), and flagellin (cBIR)94 and the anti-glycan antibodies.

The stability of anti-CBir1 expression in the postoperative setting was originally investigated by Targan et al,95 who found that there was little change in anti-CBir1 expression before and 6 months after surgery when patients were in clinical and endoscopic remission. Stability in anti-CBir1 expression was also observed in patients before and after achieving endoscopic and clinical remission with 4 months of treatment with infliximab.95

An early study of pediatric patients demonstrated that the ASCA titer decreased postoperatively96; however, this finding has not been replicated to date. Recent studies into the stability of ASCA have demonstrated mixed results. Eser et al97 found that ASCA remained stable following surgery and did not qualitatively or quantitatively predict the risk of reoperation. Similar stability of the anti-glycan antibodies was observed in a cohort study by Rieder et al98 in which positivity for Anti-L or the presence of more than three serological markers predicted progression to further surgery or a complication. In contrast, McGovern et al99 found that a higher ASCA titer was associated with endoscopic recurrence (P = 0.05). Stone et al91 found that positivity for ASCA IgA was associated with earlier second resection (HR = 2.6, P = 0.04).

Granulomas

Data regarding the predictive value of granulomas within the resection specimen are conflicting. An increase in clinical and surgical recurrence rates in patients with granulomas in their resection specimen was observed by Anseline et al56 and Cullen et al.100 In contrast, other authors have found that granulomas in resected specimens was associated with decreased recurrence rates.101, 102 More recently, a meta-analysis of 21 studies that included 2236 patients found a significantly higher recurrence rate and reoperation rate among the granulomatous group compared with the nongranulomatous group (OR: 1.37, 95% CI: 1.02–1.84 and OR: 2.38, 95% CI:1.43–3.95, respectively).103

Myenteric Plexitis

Three studies have independently found myenteric plexitis to be a predictor of postoperative recurrence. In a study by Ferrante et al104 of the resection specimens of 59 patients, those with myenteric plexitis in the resection specimen had a higher rate of endoscopic recurrence at 3 months (75% vs. 41%) and 12 months (93% vs. 59%) postoperatively. The severity of the plexitis correlated with the severity of endoscopic recurrence at both timepoints.

In a subsequent study of 164 patients who underwent CD resections, submucosal plexitis (HR = 1.87; 95% CI 1.00–3.46; P = 0.048) was found to be a predictor of clinical recurrence.105 In addition to finding that myenteric plexitis was associated with an increased rate of clinical recurrence, Ng et al106 also found that those with myenteric plexitis were more likely to have had multiple previous resections.

Operation-related Factors

Laparoscopic versus Open Surgery

Where technically feasible, laparoscopic surgery has overtaken open resection because of improved postoperative recovery, hospital stay, and cosmetic results.107 The early adverse outcomes of both surgical approaches within a year postoperatively are comparable.108, 109 In contrast to a single retrospective study which showed an increased postoperative recurrence rate with open surgery,110 three long-term retrospective111–113 and two randomized controlled trials114–116 have shown no difference in early or long-term recurrence rates between laparoscopic and open surgical approaches.117 Although laparoscopic surgery is associated with a reduced systemic inflammatory response compared to open surgery, the conserved immune response with laparoscopic surgery does not seem to reduce recurrence rates.118, 119

Type of Anastomosis

Following resectional surgery, endoscopic recurrence usually develops at, or proximal to, the anastomosis.8 While the cause of recurrence at the anastomosis is likely to be multifactorial, it has been established that that the fecal stream plays a role.53 As a consequence, luminal diameter, which varies depending on anastomotic configuration, may also be implicated in the development of clinical and surgical recurrence by causing relative obstruction, fecal stasis, and bacterial overgrowth.15, 120, 121 This has led to many authors hypothesizing that a stapled side-to-side anastomosis may prevent early stenosis, colonic reflux, fecal stasis, and secondary ischemia64, 68, 122–124 due to its wide lumen configuration.125

The impact of anastomotic configuration on postoperative recurrence rate has been evaluated in several studies.57, 65, 68, 73, 122–124, 126–129 A meta-analysis by Simillis et al73 concluded that those who underwent a side-to-side anastomosis, compared to end-to-end, had fewer postoperative complications but overall a similar postoperative recurrence rate. In contrast, a study of 141 patients by Scarpa et al68 found that reoperation rates were lower among those with a side-to-side anastomosis, an effect which was independent of whether the anastomosis was stapled or hand-sewn.

More recently a randomized controlled trial, the Canadian and American Surgical Crohn's Disease Trial (CAST), found no difference at 12 months postoperatively in endoscopic or clinical recurrence rates between those who underwent side-to side versus end-to-side anastomosis.65

Resection Margins

Initial studies suggested that radical resection, allowing wide resection margins of macroscopically normal bowel to ensure clearance of all macroscopic, and possibly microscopic, disease was associated with a reduced rate of postoperative recurrence.36, 130–134 Subsequent studies have found that nonradical resection, where there may be residual microscopic disease due to limited resection margins, results in the same postoperative recurrence rate as more radical resection.2, 52, 56, 61, 81, 135–141

In a controlled trial by Fazio et al,137 152 patients were randomized to either limited (2 cm) or extended (12 cm) macroscopically disease-free proximal resection margins. Patients were further categorized on the basis of whether there was any residual microscopic disease. Clinical recurrence (33.3 vs. 28.8%) and surgical recurrence rates (25.3 vs. 17.9%) were not significantly different between the two groups. Nor was there any significant difference in recurrence rates among those with residual microscopic disease. These data have given rise to the current practice of resecting only macroscopically involved bowel, in order to minimize the risk of intestinal failure due to short gut syndrome.

Strictureplasty

In the setting of extensive discontinuous disease a balance needs to be struck between addressing the stenosing and perforating complications of CD and preserving bowel length to avoid the development of short bowel syndrome. A key surgical approach employed to address multifocal short-segment stenosing disease has been the use of reshaping, rather than resecting, strictures, i.e., “strictureplasty.”22, 142–150 Despite the presence of residual macroscopic disease following strictureplasty, long-term postoperative recurrence rates have been reported to be as low as 2.8%–3.7%.144, 147 A meta-analysis by Yamamoto et al151 has shown that in up to 90% of patient recurrence occurs at nonstricureplasty sites.

Blood Transfusions

Studies evaluating whether the immunosuppressive effect of perioperative blood transfusion modifies the natural history of postoperative recurrence have produced mixed results.15, 152–158 A subsequent pooled analysis by Hollaar et al159 of 622 patients who had undergone resectional surgery for CD found no difference in 5-year recurrence rates between those who had received transfusion (26.9%) and those who had not (25.25%) (P = 0.456).

Postoperative Complications

There has been no consistent association between postoperative complications and increased risk of postoperative recurrence.15 A high rate of postoperative recurrence among those with postoperative complications was reported in two studies,68, 160 whereas there was no association found by Poggioli et al.38

DIAGNOSIS OF RECURRENCE AND MONITORING

  1. Top of page
  2. Abstract
  3. Search Strategy
  4. NATURAL HISTORY OF POSTOPERATIVE RECURRENCE
  5. RISK FACTORS FOR POSTOPERATIVE RECURRENCE
  6. DIAGNOSIS OF RECURRENCE AND MONITORING
  7. DRUG PREVENTION OF DISEASE RECURRENCE
  8. OPTIMAL POSTOPERATIVE TREATMENT STRATEGY
  9. REFERENCES

Clinical versus Endoscopic

The optimal monitoring for postoperative recurrence is yet to be established. The postoperative period is often associated with development of abdominal pain, which may reflect adhesion-related obstruction, calculi or dysmotility,161 and diarrhea, which may reflect bile salt dysregulation, bacterial overgrowth, or short gut syndrome rather than clinical disease recurrence. One study of 110 postoperative patients found that compared with endoscopy, the Crohn's Disease Activity Index (CDAI) could discriminate between those with and without recurrence in only 65% of cases, suggesting that CDAI alone was inadequate for the detection of postoperative recurrence of CD.162 In support of this initial study, a recent small randomized controlled by Regueiro et al163 reported that there was a poor correlation between postoperative recurrence and clinical symptoms, CDAI, and C-reactive protein (CRP), which are the conventional clinical markers of disease activity in CD.

The utility of the CDAI in determining the presence or absence of disease activity in the postoperative setting was assessed in a substudy of the CAST trial by Walters et al.164 With a cutoff of 150 the CDAI had a sensitivity of 70% and specificity of 81% for predicting the presence of recurrent endoscopic disease. The authors concluded that a combination of symptom assessment plus endoscopic evidence of recurrence should remain the gold standard definition for assessing outcome in postoperative CD trials, mainly on the basis that endoscopic recurrence precedes clinical recurrence.8, 9

Ileocolonoscopy

Of all the diagnostic modalities available, ileocolonoscopy is the most accurate in detecting mucosal lesions. It remains the gold standard for the assessment of postoperative recurrence.1 New lesions can be visualized at colonoscopy within weeks to months of surgery.8, 9, 21, 165–167 The aphthous ulcer is the earliest lesion of postoperative recurrence. In a seminal paper by Rutgeerts et al,8 ileocolonoscopy performed within a year of surgery detected recurrent mucosal lesions in 73% of patients of whom 20% had clinical recurrence. The rate of clinical recurrence at 3 years was 34%.

Based on these observations of postoperative recurrence, the Rutgeerts index was devised to score the severity of endoscopic lesions at the anastomosis and neoterminal ileum (Table 2).9 The extent and severity of endoscopic mucosal lesions were found to be predictive of subsequent clinical disease course.9 Patients with more severe endoscopic recurrence (Rutgeerts score greater than or equal to “i2,” that is, more than five aphthous lesions at the anastomosis) were found to have a higher risk of clinical recurrence at 4 years postoperatively (100% vs. 9%) compared with those with those less severe endoscopic mucosal lesions (Rutgeerts score less than i2).

Table 2. Rutgeerts Endoscopic Recurrence Score9
Endoscopic ScoreEndoscopic Findings
  1. Remission: endoscopic score of i0 or i1.

  2. Recurrence: endoscopic score of i2–i4.

  3. Severe recurrence: endoscopic score of i3–i4.

i0No lesions
i15 or fewer aphthous lesions
i2>5 aphthous lesions with normal mucosa between the lesions, or skip areas of larger lesions, or lesions confined to the ileocolonic anastomosis
i3Diffuse aphthous ileitis with diffusely inflamed mucosa
i4Diffuse inflammation with already larger ulcers, nodules, and/or narrowing

These observations by Rutgeerts have led to the recommendation to perform ileocolonoscopy within a year postoperatively to help guide therapeutic decision-making.1, 168 Trials of postoperative drug prophylaxis have employed the Rutgeerts index to score endoscopic recurrence as an efficacy endpoint.169, 170, 163

Invasive versus Noninvasive Assessment

Ileocolonoscopy is an invasive test. It is associated with a perforation risk. Furthermore, with certain surgical anastomotic configurations the anastomosis may not be readily endoscopically accessible or traversable. Noninvasive methods of assessing postoperative recurrence overcome the inconvenience of bowel preparation and the risks associated with colonoscopy. Small bowel barium follow-through and CT enteroclysis are limited by radiation exposure. More recent noninvasive techniques require further prospective validation.

Ultrasound
Abdominal ultrasound

Transabdominal ultrasonography has been investigated as a noninvasive modality to assess postoperative recurrence.171–174 A study by Rispo et al171 found that a bowel wall thickness of greater than 5 mm was predictive of severe endoscopic recurrence. When compared to ileocolonoscopy, bowel ultrasonography was able to detect postoperative recurrence with a sensitivity of 79% and specificity of 95%. The advantages of abdominal ultrasound are that it is well tolerated, cost-effective, and avoids contrast and radiation exposure. The main limitations are operator-dependence175 and difficulty in assessing the bowel wall in the setting of gas within adjacent intestinal loops.

Small Intestine Contrast Ultrasonography (SICUS)

SICUS has been used to evaluate postoperative recurrence by assessment of bowel wall thickness, dilation, and stricturing.176, 177 Use of polyethylene glycol contrast allows dissociation of intestinal loops, allowing measurement of bowel wall thickness and luminal diameter, thereby overcoming the limitations created by bowel gas in standard bowel ultrasonography.178 SICUS therefore appears to be more sensitive, specific, and accurate compared with standard ultrasonography for assessing small bowel CD lesions.175 Compared to endoscopy, SICUS is able to detect postoperative recurrence with a diagnostic accuracy of 87.5%.176

In a study by Biancone et al179, 180 comparing SICUS with ileocolonoscopy and capsule endoscopy, although SICUS was useful in assessing postoperative recurrence, it had a low diagnostic specificity. There was a poor correlation between wall thickness at the ileocolic anastomosis evaluated by SICUS and the Rutgeerts index. In contrast to this finding, Pallotta et al181 found a linear relationship between the Rutgeerts index and bowel wall thickness; 1 year or more after surgery an ileocolonic anastomosis wall thickness greater than 3.5 mm was always associated with the presence of endoscopic lesions. As with standard bowel ultrasonography, the main limitation of SICUS is that it is operator-dependent.175

CT Colonography

In a comparison of CT colonography with colonoscopy in 16 patients who had undergone with previous ileocolonic resection,182 endoscopic recurrence was detected by CT colonography with a 73% sensitivity, 100% specificity, 100% positive predictive value, 20% negative predictive value, and 75% accuracy. CT colonography detected luminal narrowing in seven of eight patients with a nontraversable stricture found at colonoscopy. False-negative findings were related to patients with mucosal lesions but without luminal narrowing. CT colonography is limited by radiation exposure. It may be useful to detect structural change, but has not been proven to detect early mucosal disease.

Small Bowel MRI

Small bowel MRI is noninvasive, allows visualization of the entire gut, and demonstrates mucosal enhancement at sites of inflammation.183 Magnetic resonance enterography (MRE) has been compared with ileocolonoscopy in a study of 30 CD patients with suspected postoperative recurrence.184 This study also aimed to devise an MRE score to grade the severity of recurrence. Compared with ileocolonoscopy, MRE had a sensitivity of 100% and specificity of 89% in detecting postoperative recurrence. The MR score correlated well with the Rutgeerts score, and exhibited reasonable interobserver agreement of 78% with a kappa coefficient of 0.673. In a separate study the same MRE scoring system and ileocolonoscopy were found to be of similar value in predicting postoperative clinical recurrence.185

Fecal Inflammatory Markers
Fecal calprotectin and fecal lactoferrin

Previous studies have demonstrated a correlation between the presence of endoscopically active CD and elevated fecal calprotectin and fecal lactoferrin.186 The utility of fecal calprotectin and fecal lactoferrin was initially examined in the postoperative setting in a cross-sectional study of 63 patients with CD who had previously undergone ileocolonic resection.187 Fecal concentrations of fecal calprotectin and fecal lactoferrin were elevated at long-term follow-up after resection of the diseased bowel, even in those patients who remained in clinical remission. These results suggested that either fecal calprotectin and fecal lactoferrin no longer correlated with disease activity after bowel resection, that not all active disease had been resected, or that microscopic inflammation persisted.

The role of fecal calprotectin as a predictive marker for endoscopic recurrence has been examined in a prospective, longitudinal study of 50 patients postileocecal resection who underwent assessment with fecal calprotectin and abdominal ultrasound at 3 months followed by colonoscopy at 1 year.188 Although at 3 months postoperatively ultrasound had a greater specificity than fecal calprotectin (90% compared to 75%), fecal calprotectin values greater than 200 mg/L showed a higher sensitivity than ultrasound (63% compared to 26%). A fecal calprotectin value greater than 200 mg may be an indication for colonoscopy among patients with negative ultrasound to identify recurrent mucosal disease.188

More recently, Lamb et al189 undertook a longitudinal study of 13 patients followed prospectively over 1 year with repetitive fecal sampling to assess whether fecal calprotectin and fecal lactoferrin correlated with clinical recurrence. Fecal calprotectin and fecal lactoferrin were found to normalize within 2 months of surgery among those with an uncomplicated disease course. Both inflammatory markers were found to correlate closely with the Harvey–Bradshaw Index (P < 0.001), with increased levels measured among those with severe clinical recurrence more than 2 months after surgery, and low levels measured among those with clinically inactive disease. Among those with mild to moderate clinical recurrence (Harvey–Bradshaw Index score 4–5) the levels of fecal calprotectin and fecal lactoferrin varied but permitted distinction between active and inactive disease.

In a cross-sectional study of 104 patients by the same authors there was greater correlation between fecal calprotectin and fecal lactoferrin with the Harvey–Bradshaw Index than CRP or platelet count.189 Of the small numbers of patients who underwent colonoscopy the markers did not correlate with endoscopic activity, indicating that further assessment of the relationship between these fecal markers, mucosal inflammation, and endoscopic changes is required.

Fecal alpha 1-antitrypsin clearance (α1AAT-Cl)

Fecal α1AAT-Cl is a marker of active inflammation that has been found to be elevated in CD.190–192 A small, prospective study evaluated the usefulness of fecal α1AAT-Cl compared with small bowel follow-through in the postoperative setting.193 Of the 11 patients studied longitudinally at 3, 6, and 12 months postileocolonic resection, fecal α1AAT-Cl was found to be above the upper normal limit in all five patients with recurrence. While in this small study fecal α1AAT-Cl appeared to be a sensitive marker of asymptomatic recurrence of CD, its use as a noninvasive marker of postoperative recurrence is limited by the lack of comparison to colonoscopy, its lack of specificity, and technical difficulties related to performing the assay.

Capsule Endoscopy

A number of small studies have suggested that capsule endoscopy may be a noninvasive alternative to ileocolonoscopy in detecting postoperative recurrence179, 194, 195 Bourreille et al194 found that capsule endoscopy detected lesions beyond the reach of the colonoscope in more than two-thirds of patients. However, capsule endoscopy was less sensitive in detecting ileal recurrence than ileocolonoscopy. In contrast, Pons-Beltran et al195 found that capsule endoscopy had a higher sensitivity in detecting CD recurrence in the neoterminal ileum than ileocolonoscopy (62% vs. 25%). Moreover, capsule endoscopy revealed lesions in the jejunum of 13 patients, including one with CD of the duodenum.

In a study of 22 patients, Biancone et al180 compared capsule endoscopy to SICUS and ileocolonoscopy at 1 year postoperatively. Capsule endoscopy was not performed in 5 of 22 patients due to the presence of narrowing. In the 17 patients who had all three assessments, capsule endoscopy identified the presence or absence of endoscopic recurrence in all patients (16 true-positive and 1 true-negative).

Capsule endoscopy requires no sedation, is well tolerated by patients, and visualizes the ileum and proximal small bowel. Whether the proximal lesions detected by capsule endoscopy represent true disease recurrence or a variant of normal is unknown.196 Small bowel lesions present at the time of surgery appear to be not significant or predictive in the short term.197–199

Scintigraphy

Granulocyte scintigraphy has been found in a small study to detect postoperative recurrence.200 However, it is limited by its low specificity, high cost, and lack of availability; hence, it cannot be recommended for routine use as a noninvasive method of assessing postoperative recurrence.

DRUG PREVENTION OF DISEASE RECURRENCE

  1. Top of page
  2. Abstract
  3. Search Strategy
  4. NATURAL HISTORY OF POSTOPERATIVE RECURRENCE
  5. RISK FACTORS FOR POSTOPERATIVE RECURRENCE
  6. DIAGNOSIS OF RECURRENCE AND MONITORING
  7. DRUG PREVENTION OF DISEASE RECURRENCE
  8. OPTIMAL POSTOPERATIVE TREATMENT STRATEGY
  9. REFERENCES

Immediate Therapy versus Tailored Treatment According to Endoscopic Findings

The two main approaches to the use of postoperative drug therapy can be summarized as either immediate prophylaxis in all patients versus diagnostic monitoring with drug therapy tailored to endoscopic findings. The latter approach is based on the premise that endoscopically identifiable mucosal recurrence precedes more advanced mucosal lesions8 and is a predictor of subsequent clinical symptoms.9

The uptake of immediate drug therapy has been variable in the past64, 201, 202 due to the limited efficacy, cost, and lack of acceptance of potential adverse drug effects in patients who have been rendered asymptomatic after resection of macroscopic disease. There have also been no long-term data to indicate that prophylactic medications reduce the rate of reoperation or hospitalizations. No prospective studies have addressed the timing and duration of therapy.

Only two retrospective studies have addressed the value of postoperative endoscopic monitoring and adjustment of treatment according to the endoscopic findings. In a study by Baudry et al,203 published in abstract form, 90 patients with an endoscopically accessible anastomosis underwent colonoscopy within 12 months of surgery, while 42 controls did not. Those with severe endoscopic recurrence (Rutgeert's score i3–i4) were treated with immunosuppressive therapy. At 5 years postoperatively clinical recurrence was greater in those who did not undergo colonoscopy (52%) compared with those who underwent colonoscopy (26%) (P = 0.01; logrank test).

In contrast, in a study by Bordeianou et al204 three strategies were tested in 199 patients: immediate postoperative treatment, treatment tailored according to endoscopic findings, and no treatment. There was no significant difference in outcome between these three approaches.

In a study of 142 patients from our own institution there was no clear benefit from postoperative colonoscopy within a year of surgery.201 However, this may have related to a lack of standardized response to endoscopic findings. In our retrospective cohort step-up medication at most consisted of modest immunosuppressive therapy, as patients had been treated prior to the recognition of the value of postoperative anti-tumor necrosis factor (TNF) therapy. To achieve improved outcomes colonoscopy may need to be performed at a standardized time postoperatively, with a consistent approach to the findings, including more intense treatment such as anti-TNF therapy.

Prospective data are needed to determine whether every patient should receive postoperative drug therapy or drug therapy tailored to endoscopic findings in the first year postoperatively.1, 25, 161

Should Preoperative Medication Use Influence Choice of Postoperative Medication?

An unresolved issue in prevention of postoperative recurrence is whether the type of preoperative medication should influence the choice of postoperative prophylaxis. The failure of a drug to control disease preoperatively may be associated with ongoing failure of the drug as postoperative prophylaxis.205 Alternatively, preoperative drug failure may relate to the drug being instituted too late in the disease course, or when the disease burden was too great.

Compliance

In a multivariate analysis of the CAST trial compliance with postoperative medication was found to be associated with a reduced incidence of clinical recurrence.65 The question of whether patients considered medications worthwhile for postoperative prevention was addressed by Kennedy et al.206 The attitude to postoperative maintenance therapy varied, with no clear clinical or demographic predictors.206 Fifty percent of participants considered a 5% reduction in the absolute risk of recurrence with azathioprine relative to 5-ASA worthwhile.

5-Aminosalicylate (5-ASA)

Several randomized controlled trials have evaluated 5-ASA in the postoperative setting, with variable results.98, 167, 207–210 The heterogeneity among these studies is likely to be due to differences in: 5-ASA preparations, release mechanisms and dosages, duration of follow-up, and definition of recurrence.211 A benefit from 5-ASAs as maintenance therapy postsurgical induction of remission was suggested by a meta-analysis by Camma et al,212 who found that 5-ASA reduced the risk of postoperative recurrence by ≈13%. More recently, a meta-analysis by Ford et al213 found that although there appeared no benefit from sulfasalazine in preventing postoperative recurrence, in contrast mesalazine was more effective than placebo or no therapy. A meta-analysis by Doherty et al214 found that mesalamine decreased clinical but not endoscopic recurrence and was inferior to azathioprine or mercaptopurine for all outcomes. A recent controlled trial by Reinisch et al215 comparing mesalazine and azathioprine for the prevention of clinical recurrence among patients with established moderate or severe endoscopic recurrence demonstrated no differences in clinical or endoscopic outcomes at 1 year. A Cochrane review indicated that the number needed to treat to prevent clinical recurrence was 12 and to prevent severe endoscopic recurrence was eight.216 The weight of evidence therefore suggests that 5-ASA is of uncertain benefit in preventing disease recurrence, and if used at all should be restricted to those at low risk of postoperative recurrence.217–220

Imidazole Antibiotics

Imidazole antibiotics (metronidazole and ornidazole) have been demonstrated in a recent meta-analysis of two randomized controlled trials to reduce the risk of endoscopic (RR 0.44; 95% CI 0.26–0.74, NNT = 4) and clinical recurrence (RR 0.23; 95% CI 0.09–0.57, NNT = 4) relative to placebo.221, 222 However, both these agents were associated with higher risk of adverse events (RR 2.39, 95% CI 1.5–3.7) and patient withdrawal.223 When compared with placebo, metronidazole (20 mg/kg/day) for 3 months postoperatively reduced both severe endoscopic recurrence and clinical recurrence at 1 year.220 The reduction in clinical recurrence rates was sustained for 3 years. In the second trial, ornidazole (1 g/day) given for 1 year postoperatively significantly reduced endoscopic recurrence and clinical recurrence compared with placebo at 1 year222; however, there was no difference in clinical recurrence rates between the two groups after long-term follow-up at 3 years. At the dosages used side effects were experienced in about half the patients in both trials, resulting in discontinuation of the drug in 13% and 21% of patients, respectively. Fewer side effects may be experienced if lower doses of imidazole are used.224 Imidazole antibiotics prevent early endoscopic recurrence and delay clinical recurrence but their long-term use is limited by drug effects and toxicity. Long-term use of low-dose therapy has not been formally evaluated.

Thiopurines

Several studies have evaluated the impact of thiopurines in the postoperative setting.66, 224–233 Of these a number of studies have compared thiopurines to aminosalicylates as immediate postoperative prophylaxis.66, 230, 232–234 In spite of the relatively small differences observed in these studies between thiopurines and aminosalicylates, a recent Cochrane review found thiopurines to be more effective overall than aminosalicylates in preventing endoscopic, severe endoscopic, and clinical recurrence.216

The long-term impact of immediate thiopurine prophylaxis on endoscopic and clinical recurrence rates was reviewed in a prospective observational study by Domenech et al.170 Clinical recurrence was observed in 23% of patients during follow-up. The cumulative probability of endoscopic recurrence was 44%, 53%, 69%, and 82%, at 1, 2, 3, and 5 years, respectively, suggesting that despite the reduction of early recurrence with azathioprine endoscopic and clinical recurrence develop over time.

The effect of combining long-term azathioprine with short-term metronidazole on recurrence was examined by D'Haens et al224 in a controlled trial of 81 patients at high risk of postoperative recurrence. All patients received metronidazole 250 mg three times daily for 3 months. In addition, patients were randomized to receive either azathioprine or placebo for 12 months. At 12 months the endoscopic recurrence rate was significantly lower in those receiving concurrent azathioprine (44%) compared to placebo (69%).

A meta-analysis by Peyrin-Biroulet et al,234 including four randomized controlled trials, recently evaluated the efficacy and safety of thiopurines as postoperative prophylaxis compared to control arms (placebo, with or without antibiotics, and aminosalicylates). Thiopurines were found to be more effective than control medications in preventing clinical recurrence at 1 and 2 years. Thiopurines were also found to be more effective than control medications in preventing endoscopic recurrence (Rutgeerts score i2–i4) at 1 year but were not effective in the prevention of very severe (i3–4) recurrence. However, thiopurines were also more likely than placebo to cause adverse events leading to drug withdrawal.

A retrospective review of 326 patients who underwent intestinal resection was undertaken by Papay et al233 to study the impact of thiopurines on surgical recurrence. Of the 326 patients, 46 required reoperation. A significant reduction in reoperation rate was observed among those who had 3 or more years of postoperative thiopurine therapy (27%) compared with those with 3 months or less (55%) of postoperative immunosuppressive therapy (P < 0.004). The benefit conferred by thiopurines in reducing reoperation rates was therefore only achieved with long-term postoperative therapy.

Overall, thiopurines appear to have the most impact when combined with metronidazole; they are also effective when used alone. Their benefit in preventing postoperative recurrence is moderate.

Anti-TNF Therapy

Anti-TNF therapy has recently been shown to provide the greatest benefit in preventing postoperative recurrence and treating endoscopic recurrence.

In a prospective nonrandomized study, endoscopic or clinical recurrence occurred in none of seven patients on infliximab in combination with methotrexate compared with 12 of 16 patients on mesalazine who developed endoscopic recurrence after 2 years.236, 237

A retrospective study by Okamoto et al238 compared the clinical value of infliximab in postoperative patients and nonoperated disease controls. Infliximab was effective in maintaining remission in 65% of postoperative patients compared with 85% of nonoperated controls. Remission rates were higher among those who received infliximab within a year postoperatively, who had undergone fewer resections, and who had been treated with scheduled maintenance therapy. Another retrospective Japanese study comparing the outcome of infliximab versus noninfliximab treated postoperative patients found that the former group, especially those who received infliximab within a year of surgery, had a longer time to reoperation and less severe endoscopic recurrence compared with non-infliximab-treated patients.

The efficacy of infliximab in treating early endoscopic recurrence following resectional surgery was examined by Yamamoto et al.239 Twenty-six patients who had maintained clinical remission on 3 g of mesalazine per day but had developed endoscopic recurrence at their 5-month colonoscopy were then nonrandomly assigned to ongoing treatment with mesalamine, or commencing azathioprine 50 mg/day or infliximab eight weekly (without three induction doses). After 6 months follow-up, clinical recurrence was observed in none of 8% of the infliximab-treated patients, three of eight (38%) of the azathioprine-treated patients, and 7 of 10 (70%) mesalazine-treated patients. Improvement in endoscopic inflammation was observed in 75% of the infliximab group, 38% of the azathioprine group, and none of the mesalazine-treated patients (P = 0.006).

A controlled trial has confirmed the efficacy of infliximab in preventing postoperative recurrence in CD.163 Twenty-four patients were randomized to receive either infliximab or placebo immediately after ileocolonic resection of all macroscopic disease. Among the infliximab-treated group there were more active smokers (45.5% vs. 7.7%), less immunomodulator use (36.4% vs. 53.8%), and higher baseline inflammatory markers compared with the placebo-treated group. Three of the infliximab-treated patients (30%) and five (38.5%) placebo-treated patients had received previous infliximab. At 1 year endoscopic recurrence (Rutgeerts score i2–i4) occurred in 1 of 11 (9.1%) of infliximab-treated patients compared to 11 of 13 (84.6%) placebo-treated patients (P = 0.0006). A marked reduction in clinical (0% vs. 38.5%) (P = 0.046) and histologic (27%) versus (85%) recurrence was also found in infliximab compared with placebo-treated patients.

Following the primary endpoint at 12 months postoperatively, patients were offered open-label infliximab and followed for another year.240 Of seven initial placebo-treated patients five (71%) were in remission at 2 years. Three of the initial-infliximab treated patients who stopped infliximab all developed endoscopic recurrence at 2 years.241 These 24-month data suggest that infliximab prevents clinical recurrence for at least 2 years.240, 241 Second, infliximab effectively heals endoscopically visible disease but its effect is maximal when started immediately postoperatively. Finally, disease often recurs if infliximab treatment is ceased after 1 year of postoperative therapy.

The impact of stopping infliximab after 3 years of therapy was explored by Sorrentino et al.242 Twelve patients treated immediately postoperatively with 5 mg/kg infliximab maintenance therapy who did not have endoscopic or clinical recurrence after 3 years underwent colonoscopy 4 months after stopping infliximab. Discontinuation of infliximab resulted in endoscopic recurrence after 4 months in 10 of 12 patients (83%). Mucosal integrity was restored and maintained for 1 year after retreating all 10 patients with eight weekly infliximab at a lower dose of 3 mg/kg. These data confirm the observation that endoscopic disease often recurs after infliximab is stopped. Infliximab therapy at a lower dose of 3 mg/kg appears to be effective for at least 1 year.

The efficacy of adalimumab in preventing postoperative recurrence was evaluated in a single nonrandomized study of 20 patients.243 A majority of the patients had risk factors for postoperative recurrence: 60% were smokers, 35% had had a prior resection, and 65% had perforating disease. At 12 months postoperatively, following a standard induction and maintenance regimen, endoscopic recurrence (Rutgeerts score of i1 or greater) was seen in two patients (10%). None of the patients had clinical recurrence. Histologic recurrence was seen in 45% of patients. These preliminary data suggest that adalimumab may also be effective in preventing postoperative recurrence.

Further trials of anti-TNF therapy in the postoperative setting are currently under way. Questions that remain to be answered include: 1) whether anti-TNF therapy can be used effectively as part of a step-up regimen among patients who are intolerant to immunosuppressive therapy or develop postoperative recurrence despite immunosuppressive therapy; 2) whether all anti-TNF therapies are equally effective in preventing and treating postoperative recurrence; 3) whether there is a benefit of combined anti-TNF and immunosuppressive therapy over anti-TNF monotherapy; and 4) at what point anti-TNF therapy can be stopped in the postoperative setting.

Pilot Studies of New Therapies

Local Injection of Infliximab

The use of local injections of infliximab to treat postoperative recurrence has been evaluated in a pilot open label study of eight patients with localized endoscopic recurrence (less than 5 cm length) who had remained clinically well (CDAI less than 150).244 After a 14–21-month follow-up there was no significant reduction in median endoscopic or histologic score, although the number or extent of lesions was decreased in seven patients. Further studies are required to assess whether this novel approach will be of benefit in altering the disease course, particularly among those with a Rutgeerts score of i3 or greater who are at greatest risk of progressive disease.

Fish Oils

Preliminary evidence suggests that fish oils may be effective in prevention of postoperative recurrence. In a controlled trial by Belluzzi et al,245–248 presented in abstract form, following ileal resection 50 patients were randomized to omega-3 fatty acids or placebo. At 1 year postoperatively significantly fewer patients in the omega-3 group (34%) had severe endoscopic recurrence than those receiving placebo (62%). These findings need to be reproduced in larger prospective studies.

Granulocyte Colonic-Stimulating Factor (G-CSF)

The ability of recombinant human granulocyte colonic-stimulating factor (rhG-CSF) to achieve mucosal healing among patients with established endoscopic recurrence has been undertaken in a small open-label prospective study.249 Five patients with severe endoscopic ileitis within 1 year of ileocolonic resection were treated with 300 mg rhG-CSF subcutaneously, three times weekly for a total of 3 months. Two of the patients achieved complete mucosal healing. None of the patients experienced severe adverse effects.

Enteral Feeding

The effectiveness of long-term enteral feeding in postoperative prevention has been evaluated in a single prospective controlled nonrandomized study by Yamamoto et al.250 The self-administration of nocturnal enteral supplementation was associated with a significant reduction in endoscopic and clinical recurrence rates at 12 months.

Unproven Therapies

Steroids

Corticosteroids have been demonstrated to be ineffective in preventing postoperative recurrence in several studies. In two placebo-controlled trials of budesonide 3 mg and 6 mg/day there was no improvement in endoscopic or clinical recurrence at 1 year postoperatively.251, 252 Moreover, steroids have been associated with increased perioperative morbidity.253 Half the patients taking steroids postoperatively become either resistant or dependent on steroids at 1 year.254

Probiotics

Probiotics are thought to mediate their therapeutic effect via exclusion of pathogens, maintenance of epithelial barrier function, and induction of adaptive immunity. In spite of potential theoretical benefit for their use, there is insufficient evidence that they are of benefit in preventing postoperative recurrence. Probiotics which have been assessed in five postoperative controlled trials include: Lactobacillus johnsonii (LA1),255, 256Lactobacillus rhamnosus strain GG (LGG),257 Synbiotic 2000,258 and VSL3#.259 A Cochrane review and meta-analysis of these five studies found that probiotics were ineffective in preventing endoscopic, severe endoscopic, or clinical recurrence postoperatively.216, 223

Interleukin-10 (IL-10)

The efficacy of the antiinflammatory cytokine IL-10 in preventing postoperative recurrence was evaluated in 65 postoperative patients in a single placebo-controlled trial.260 At 12 weeks postoperatively there was no significant difference in endoscopic (46% IL-10 vs. 52% placebo) or severe endoscopic recurrence, or clinical recurrence, between the two groups.

Cost-effectiveness of Drug Prophylaxis

There are limited data on the comparative cost-effectiveness of postoperative drug prophylaxis regimens. A decision analysis model presented in abstract form comparing five strategies for prevention of postop recurrence: 1) no treatment; 2) azathioprine; 3) antibiotics (metronidazole); 4) upfront infliximab; and 5) tailored infliximab found that tailoring infliximab use to high-risk patients yields a more acceptable incremental cost-effectiveness ratio compared with upfront infliximab.261 Nonetheless, antibiotics appeared to be the most cost-effective option for patients able to tolerate treatment.

OPTIMAL POSTOPERATIVE TREATMENT STRATEGY

  1. Top of page
  2. Abstract
  3. Search Strategy
  4. NATURAL HISTORY OF POSTOPERATIVE RECURRENCE
  5. RISK FACTORS FOR POSTOPERATIVE RECURRENCE
  6. DIAGNOSIS OF RECURRENCE AND MONITORING
  7. DRUG PREVENTION OF DISEASE RECURRENCE
  8. OPTIMAL POSTOPERATIVE TREATMENT STRATEGY
  9. REFERENCES

Prospective treatment strategies for postoperative prevention are lacking. Although it is not currently possible to identify all patients who will ultimately develop recurrence and the optimal medication regimen to prevent postoperative recurrence is yet to be established, there is sufficient evidence regarding drug prophylaxis to propose a treatment strategy.

An optimal treatment strategy may be based on a patient's clinical risk factors with treatment then adjusted according to endoscopic findings in the first year after surgery; however, this remains to be proven.

Given the current level of evidence about the preventive value of postoperative drug therapy, we consider that there are potentially three valid approaches to this clinical situation:

  • 1
    Postoperative drug treatment stratified according to the risk of disease recurrence. Those at higher risk (smokers, perforating disease, recurrent operations), who comprise a majority of patients, would receive either a thiopurine or anti-TNF therapy immediately postoperatively. Low-risk patients would receive no drug therapy, or limited therapy with an imidazole antibiotic. Such an approach may also include endoscopic assessment sometime in the first postoperative year, with an intensification of treatment for more advanced mucosal disease identified colonoscopically (Fig. 1).
  • 2
    Place all high-risk patients on anti-TNF therapy postoperatively (Fig. 2). Such an approach would be effective in reducing recurrence, but would need to be balanced by the incidence of adverse events and cost. Such an approach would not be funded by government or health funds in most countries.
  • 3
    Treat no patient with immediate postoperative drug therapy, but restrict the commencement of drug therapy to those patients who develop mucosal recurrence within some months on colonoscopy (Fig. 3).
thumbnail image

Figure 1. Immediate postoperative prophylaxis with step-up therapy.

Download figure to PowerPoint

thumbnail image

Figure 2. Immediate postoperative prophylaxis with anti-TNF therapy for all high-risk patients.

Download figure to PowerPoint

thumbnail image

Figure 3. Postoperative therapy tailored to postoperative endoscopic findings.

Download figure to PowerPoint

Patients undergoing surgery for CD who are left with residual disease postoperatively or have more proximal disease, or in whom the anastomosis is unreachable via standard colonoscopy, require a different approach. In the event of any clinical factors placing these patients at high risk of recurrence (smokers, perforating disease, extensive small bowel disease) consideration should be given to routine prophylactic drug therapy with either a thiopurine or an anti-TNF antibody. Postoperative monitoring of such patients on or off treatment is likely to require a combination of noninvasive imaging, biochemical markers such as CRP, and fecal inflammatory markers.

Further studies are currently in progress that will provide further evidence on the value of thiopurines and anti-TNF therapy postoperatively, as well as the value of endoscopic evaluation and drug treatment according to endoscopic findings.

REFERENCES

  1. Top of page
  2. Abstract
  3. Search Strategy
  4. NATURAL HISTORY OF POSTOPERATIVE RECURRENCE
  5. RISK FACTORS FOR POSTOPERATIVE RECURRENCE
  6. DIAGNOSIS OF RECURRENCE AND MONITORING
  7. DRUG PREVENTION OF DISEASE RECURRENCE
  8. OPTIMAL POSTOPERATIVE TREATMENT STRATEGY
  9. REFERENCES