Workshop Report: Prevention of Postoperative Recurrence in Crohn's Disease

INTRODUCTION In December 2003, the Crohn's and Colitis Foundation of America convened a workshop of pediatric gastroenterologists. The workshop was designed to highlight particular problems in children and adolescents with inflammatory bowel disease in need of prospective clinical research. Among the issues discussed, the problem of recurrent Crohn's disease (CD) after surgery was identified as a particularly important area for future research. Population based studies have shown that roughly one third of patients with CD will require intestinal resection within 1 year of requiring corticosteroid therapy (1), and upward of 80% will undergo surgical resection of involved segments of bowel within 20 years of initial diagnosis (2). Twenty to 30% of these patients will experience a symptomatic recurrence of disease within the first year after surgery, with increasing likelihood in each subsequent year (3). Unfortunately, even in patients in whom all grossly evident CD is removed (a so-called curative resection), grossly evident and clinically symptomatic CD recurs in a high proportion of patients, such that subsequent repeated surgeries are required in 15% to 45% of patients at 3 years, 26% to 65% in 10 years, and 33% to 82% in 15 years (4). Each bowel resection carries with it significant morbidities that have major implications for the growing pediatric and adolescent patient. Consequently, it is reasonable to direct considerable efforts at preventing recurrence of CD after bowel resection. Data exist on aminosalicylate (5-ASA), antibiotic, and immunomodulator (6-mercaptopurine [6-MP] and azathioprine [AZA]) use for postoperative prophylactic therapy of CD in adults. No prospective studies have been performed in children. Thus, although surgery for CD and subsequent postoperative clinical recurrence is common, an optimal medical strategy for the prevention of recurrence is not clear. The aim of this workshop report is to highlight what is and is not known about preventing postoperative recurrence in children, with the hope that such a review will serve as a guide for future clinical outcome studies in children with CD who require surgery. BACKGROUND Definitions of CD Recurrence For the purposes of this review, it is important to distinguish between postoperative CD relapse and recurrence. The term postoperative “relapse” has been applied to the situation in which symptoms caused by active CD develop because of previously identified areas of CD that have been left in situ at the time of operation. Clinical practice has long recognized the need to continue maintenance medical therapy after surgery to prevent relapse when areas of known CD have not been resected. By contrast, after a “curative” resection in which all grossly evident diseased bowel is removed, “recurrence” refers to the de novo development of CD in areas of the gastrointestinal tract previously felt to be free of disease. Defining recurrence of CD in the context of a preventative study has been challenging, and there is a lack of consistency in the literature. Previous studies have often defined postoperative CD recurrence as the need for a second intestinal resection. Although this definition clearly permits a definitive diagnosis of recurrence, most would argue that it only includes the most severe cases of recurrence and thereby ignores the bulk of patients whose recurrent CD can be managed medically. A broader definition of clinical CD recurrence has been the reappearance of symptoms associated with objective signs of active disease after “curative” resection. The pediatric literature (Table 1) has generally favored this approach, although what is required to demonstrate an objective sign of disease has not been universally agreed upon. The emergence of symptoms alone is obviously inadequate for diagnosis of recurrent CD because nonspecific symptoms such as abdominal pain and diarrhea can be associated with a host of other potential conditions such as irritable bowel syndrome, superimposed infection, or partial bowel obstruction from intra-abdominal adhesions. In addition, although the use of quantitative disease activity scores such as the Pediatric Crohn's Disease Activity Index (PCDAI) can standardize observations among multicentered observers (10,11), such activity scores have not been validated in postoperative patients.TABLE 1: Definitions of postoperative recurrence in pediatric studiesClinical recurrence has therefore been difficult to define. It also can take years to develop, so that studies designed to determine a reduced frequency of clinical recurrence must be of long duration (3-5 years). These facts have led to attempts to identify surrogate markers that could predict patients at high risk for subsequent symptomatic recurrence and thereby shorten the time necessary to carry out treatment trials. Neither common laboratory tests nor radiologic studies have proven to be adequate in this regard. Abnormal laboratory findings such as elevated erythrocyte sedimentation rate or hypoalbuminemia have been shown to be poor predictors of clinical recurrence because they are both nonspecific and generally coincide with the onset of active, clinically symptomatic disease, rather than precede it (12). Radiologic tests have been shown to be too insensitive, as evidenced by subjects with clinical symptoms and endoscopic abnormalities, yet normal radiographs (13). Because of this, many adult trials have adopted the use of endoscopic recurrence as a primary endpoint (12,14). Based on the work of Rutgeerts et al. (12) (Table 2), evidence of significant visual pathology on ileocolonoscopy (a score of 3 or 4, representing diffuse aphthous ileitis and diffusely inflamed mucosa) within one year of curative resection is highly predictive of ultimate clinical recurrence defined by the development of symptoms or complications of inflammation. This surrogate endpoint allows for a shorter and more discrete duration of follow-up, and interventions can be compared on the basis of their ability to prevent high-grade endoscopic findings. However, this type of endoscopic scoring has never been validated in a pediatric population. By requiring ileocolonoscopy, it also significantly increases the cost of any proposed study and, by its invasive nature, potentially limits the number of children who could be recruited into a prospective trial.TABLE 2: Endoscopic criteria for postoperative recurrence Rutgeerts scoring system to evaluate the severity of recurrent endoscopic Crohn's disease lesionsAlthough there have been a number of studies investigating noninvasive or minimally invasive tests as indicators of CD activity, none have been evaluated as potential surrogate markers for prediction of postoperative recurrence. Possible candidates for future investigation in this setting include fecal calprotectin (15), gut secretion of interleukin (IL) 1b and IL 8 (16), abnormal intestinal permeability (17), IL 6 (18), soluble IL 2 receptor (19), and fecal tumor necrosis factor-α (20). Among these candidates, fecal calprotectin appears to be particularly promising (21-23). It is a calcium binding, S100 protein found within the neutrophil cytosol that is released with cell death or activation. It is stable at room temperature for 1 week and can be measured from a random, spot stool using a commercially available enzyme linked immunoadsorbent assay that has excellent inter- and intra-observer reproducibility. Previous studies have demonstrated good correlation between fecal concentrations and both ulcerative colitis and CD activity, and increases in fecal calprotectin more than 50 mg/L have been shown to predict future relapse. Risk Factors for CD Recurrence A number of studies have assessed possible risk factors for CD recurrence in both children and adults. Although there is some disagreement among studies (Table 3A), disease location (extensive colonic or ileocolonic CD), indication for surgery (failed medical therapy or perforating CD), and the postoperative discontinuation of preoperative 6-MP/AZA appear to be the most significant risk factors for CD recurrence in children. It is not clear whether this last risk factor is caused by a bias toward treating only the sickest children with immunomodulator therapy before surgery or whether the loss of immune suppression associated with discontinuation of 6MP/AZA postoperatively leads to more rapid immune dysregulation and earlier CD recurrence (5). In adult CD populations, cigarette smoking has been identified as an additional risk factor (24,25). Type of surgical anastomosis, duration of preoperative CD, disease activity, and age or sex of the patient appear to be less important variables (Table 3B).TABLE 3A: Risk factors for postoperative recurrence: pediatric studiesTABLE 3B: Risk factors for postoperative recurrence: adult studiesChemoprevention of Recurrence On the basis of the adult literature, there are several classes of medication that have been evaluated for the prevention of postoperative recurrence. To date, none have been prospectively evaluated for this indication in a pediatric CD population. Mesalamine Mesalamine has been assessed in several trials, with mixed results. In a multicenter, randomized, double-blind, placebo controlled clinical trial involving 87 patients (29), it was found that severe endoscopic inflammation based on the Rutgeerts scoring system (endoscopic scores of 3 or 4) was significantly less common in the group treated with 3 g/day of oral mesalamine for 1 year compared with those treated with placebo. In this study, severe endoscopic lesions were seen in only 24% of the mesalamine group 1 year after surgery, compared with 56% in the controls (P < 0.004). However, rates of clinical recurrence at 1 year were not statistically different (7/44 receiving mesalamine, 10/43 receiving placebo). A combined Canadian/American trial involving 163 patients undergoing “curative resection” found that 3 g/day of mesalamine resulted in lower rates of symptomatic recurrence (defined as symptoms with either radiologic or endoscopic confirmation of disease) versus placebo (30). In the treatment group, 31% had symptomatic recurrence versus 41% in the placebo group (P = 0.03) with a relative risk of developing recurrent disease of 0.63 (95% confidence interval 0.4-0.97). Another randomized, multicenter, but open-label study enrolled 110 adults undergoing their first intestinal resection (31). Treatment with 2.4 g/day of mesalamine resulted in a cumulative 2 year endoscopic recurrence rate of 52% compared with 85% in the subjects given no postoperative treatment (P = 0.002). Symptomatic recurrences were seen in only 18% of the treated group compared with 41% of the untreated subjects (P = 0.006). Severe recurrences were also less frequent in the treated group (17 versus 38%, P = 0.021). By contrast, two additional studies revealed no significant differences after treatment with mesalamine. A European study (32) revealed a trend toward lower clinical recurrence rates (based on CDAI score) at 18 months in a group treated with 4 g/day of mesalamine versus placebo (24.5% vs. 31.4%, P = 0.1) and in particular, patients with isolated small bowel disease. However, endoscopic recurrence was seen in similar proportions at 6 weeks postoperatively (30% mesalamine, 27% placebo) and, in fact, was slightly higher in the mesalamine treated group at 18 months (66% vs. 50%). In this series of patients, endoscopic findings were not predictive of clinical recurrence at 18 months. Similarly, a multicenter clinical trial evaluated the short-term effect of mesalamine (3 g/day) versus placebo on endoscopic recurrence in 106 adults (33). In this trial, the frequency and severity of endoscopic lesions, as assessed by the Rutgeerts score, were comparable in the two treatment groups (50% mesalamine, 63% placebo, P = 0.16) 3 months after resection. In summary, mesalamine has had variable success in the prevention of postoperative recurrence of CD. However, there is considerable variability in the treatment protocols and a number of methodologic flaws in the treatment trials. Dosing ranged from 2.4 g/day to 4 g/day. Outcome measures ranged from endoscopic recurrence based on the Rutgeerts classification to clinical recurrence (with or without documented disease). Accordingly, a wide variance of results was seen. Overall, mesalamine appears to offer limited benefit over placebo for the prevention of CD recurrence in adults, with an estimated risk reduction of no more than 10% (34,35). Additional prospective studies will be necessary before it can be determined whether this agent should be considered a useful treatment for prevention of postoperative CD recurrence in children. Antibiotics In a trial of 60 patients undergoing curative resection, a 3 month trial of metronidazole was shown to reduce the rates of endoscopic recurrence (using the Rutgeerts criteria) (36). In the treatment group, 3 of 23 patients had Rutgeerts 3 to 4 lesions at 3 months versus 12 of 28 in the placebo group (P = 0.02). This effect may have predicted prolonged disease-free survival, with lower proportions of recurrence in the treatment group at 1, 2, and 3 years, although these differences did not meet statistical significance. The most common side effects of treatment were metallic taste, gastrointestinal intolerance, and paresthesias. In a subsequent study, ornidazole (1 g/day) was used in a similar treatment trial (37). Endoscopic lesions, assessed 12 months after surgery, were seen in 79% of those treated with placebo compared with 54% of those treated with ornidazole (P = 0.04). Clinical recurrence was also reduced at 1 year in the ornidazole treated subjects (8% versus 37%, P = 0.002). 6-Mercaptopurine/azathioprine There are relatively few data regarding the use of 6-MP/AZA for postoperative recurrence. A randomized multicenter trial in which adult patients received either 6-MP (50 mg/day), mesalamine (3 g/day), or placebo revealed numerically lower rates of clinical (50%, 58%, 77% respectively), severe endoscopic (Rutgeerts grade 3-4) (16%, 48%, 42%), and radiologic (33%, 46%, 49%) recurrence at 24 months in patients given 6-MP compared with those given mesalamine or placebo (13). 6-MP was statistically superior to placebo for the reduction of the rate of clinical recurrence (P = 0.045). However, numerous flaws in the study significantly limit the interpretation of data from this trial. The study dose of 6-MP was low compared with the dose usually accepted as effective in adults and children (1.5 mg/kg/day) for the control of active CD and was performed without knowledge of a subject's thiopurine metabolism or metabolite levels. In addition, definitions of clinical recurrence were poorly validated, resulting, in some circumstances, in rates of clinical recurrence that exceeded the rates of endoscopic recurrence. An open label trial in adults evaluating AZA (2 mg/kg/day) versus mesalamine (3 g/day) after “conservative,” not “curative,” surgery (strictureplasty ± limited resection) failed to show a significant difference in the rate of clinical CD relapse at 24 months (AZA 17%, mesalamine 28%) (38). The open label study design potentially biases the results. More importantly, however, because the surgeries performed in these subjects did not remove all grossly identifiable CD, this report investigates prevention of relapse rather than recurrence. As such, its results are not directly comparable with any of the other studies discussed in this review. A recent retrospective paper in the Spanish literature details open label use of AZA (2.0-2.5 mg/kg/day) and mesalamine after intestinal resection (39). Among the 33 adult subjects treated with AZA, endoscopic recurrence was found in 8.6% compared with 87.5% in the 16 subjects treated with mesalamine (P < 0.001). Clinical recurrence was seen in none of the AZA group and 31% of the mesalamine group (P = 0.004). However, the retrospective and open label study design and lack of consistent clinical and endoscopic evaluations of the subjects suggests that the results of this study could have been influenced by significant investigator bias. Another retrospective study evaluated whether the use of immunosuppressive therapy reduced CD recurrence after a second resection (40). Among the 18 subjects on immunosuppressives, 14 received AZA, 1 6-MP, and 2 methotrexate (no doses reported). Their recurrence rates were compared with 12 controls (5 on salicylates, 7 on no medication). Clinical recurrence rate at 3 years was significantly decreased in the immunosuppressive group compared with controls (25% vs 60%, P < 0.05). Frequency of third operation for CD was also reduced in the immunosuppressive group at last follow-up (14% vs 58%, P < 0.02). Again, the retrospective study design and lack of consistent outcome measures are problematic. Finally, a retrospective analysis of 6-MP use after surgery in children demonstrated decreased risk of clinical recurrence (defined using the PCDAI) in children who received 6-MP (in addition to 5-ASA) in the perioperative period (7). In this report, three of five patients who received only 5-ASA had recurrence versus zero of five who were treated with both 5-ASA and 6-MP (1 mg/kg daily for 2 years). Mean time to recurrence in the three subjects receiving only 5-ASA was 3.7 ± 1.2 months, compared with the mean duration of follow-up in the 6-MP + 5-ASA treated subjects of 32.6 ± 18.4 months. Patients were otherwise similar with respect to type of operation and other concomitant medications. This study too is flawed, primarily because of the very small sample size, retrospective study design, and poorly documented outcome measures. In summary, despite widespread use of immunomodulator therapy as postoperative prophylaxis, the data supporting such therapy is scant, mostly retrospective, and otherwise methodologically flawed. A well-designed, prospective trial looking at the efficacy of immunomodulatory therapy to prevent postoperative recurrence has not yet been performed. Other Agents Fish Oil A single randomized controlled trial in adults has evaluated the use of a fish oil supplement as prophylactic treatment to prevent CD relapse rather than postoperative recurrence (41). Relapse was measured by change in CDAI at the end of 1 year of follow-up. The trial demonstrated that fish oil therapy was effective, with clinical relapse rates of 28% in the fish oil group compared with 69% in the placebo group (P < 0.001). The scientific rationale for this therapy comes from the delineation of the role that eicosanoids as of inflammation. 6 serve as for the of and By with 6 the 3 found in fish are to of these inflammatory by of the less Other potentially important of for fish oil include and It should be however, that the study evaluated CD relapse in adult fish oil be used to prevent postoperative recurrence in children or adults to be in as to prevent CD recurrence after intestinal resection include and only have been demonstrated for these Agents A number of additional have been in small trials. appears to offer no benefit in preventing postoperative recurrence Similarly, rates of recurrent endoscopic lesions were in adults treated with versus those receiving placebo the documented benefit of for maintenance of CD in the trial no studies have assessed the efficacy of in the prevention of postoperative CD recurrence. It is clear that the optimal treatment for prevention of postoperative recurrence of CD has yet to be It is also however, that the postoperative of pediatric CD, particularly in the age of immunomodulatory and to be more clearly may for investigating factors associated with the need for surgery and subsequent postoperative recurrence. In addition, of the of immunomodulatory and on postoperative must be need to be on the measures that can be used to and predict CD recurrence. measures of CD activity in children such as the must be validated in the postoperative setting because at the effect of a “curative” resection on the is the predictive of the scoring system for endoscopic recurrence in a pediatric should also be However, given the invasive of patients, and review may be to in such a To future in children and the need for noninvasive predictors of CD recurrence need to be identified and A study the of a surrogate such as fecal calprotectin as an of increasing postoperative CD activity and whether with endoscopic recurrence and predict clinical CD recurrence is The of such a noninvasive could subsequent treatment trials. Similarly, the development of such as may also reduce the need for confirmation of recurrence, in to to pediatric patients into prevention trials. this has been prospective pediatric treatment trials designed to reduce the risk of CD recurrence can be These studies will need to which children are at when they should be and which agent would be most must the role and effective dose for single mesalamine, 6-MP and AZA, and for postoperative such as fish or other also need to be Finally, studies should also the of in a postoperative treatment will require multicenter because no single pediatric will of subjects to a treatment trial. be to require at subjects placebo rate and reduction in CD and more than subjects placebo rate and 10% reduction in The number of subjects required to a prospective prevention study may require a with from the of and Crohn's and Colitis Foundation of America will be for such a study to be given the associated with a trial design that could require both endoscopic and noninvasive clinical and of at one or more postoperative time and clinical to determine and effective must therefore a high this to the children and who to for the

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.