Portal de Periódicos da CAPES

Infliximab maintenance treatment reduces hospitalizations, surgeries, and procedures in fistulizing Crohn’s disease

2005; Elsevier BV; Volume: 128; Issue: 4 Linguagem: Inglês

10.1053/j.gastro.2005.01.048

1528-0012

Gary R. Lichtenstein, Songkai Yan, Mohan Bala, Marion Blank, Bruce E. Sands,

Autoimmune and Inflammatory Disorders

Background & Aims: Infliximab is effective in closing fistulas in patients with Crohn’s disease. We examined the effect of infliximab maintenance treatment on hospitalizations, surgeries, and procedures in patients with fistulizing Crohn’s disease enrolled in the ACCENT II study. Methods: After 5 mg/kg infliximab at weeks 0, 2, and 6, a total of 282 patients were separately randomized at week 14 as responders (at least a 50% reduction from baseline in the number of draining fistulas at both weeks 10 and 14) or nonresponders to receive placebo or 5 mg/kg infliximab maintenance every 8 weeks. At week 22 and later, patients who lost response could be treated with a maintenance dose 5 mg/kg higher. Data on Crohn’s disease-related hospitalizations, surgeries, and procedures were compared between the treatment groups for responders and all randomized patients. Results: A total of 282 patients were randomized at week 14, of whom 195 were randomized as responders. Among patients randomized as responders, those who received infliximab maintenance had significantly fewer mean hospitalization days (0.5 vs. 2.5 days; P < .05), mean numbers (per 100 patients) of hospitalizations (11 vs. 31; P < .05), all surgeries and procedures (65 vs. 126; P < .05), inpatient surgeries and procedures (7 vs. 41; P < .01), and major surgeries (2 vs. 11; P < .05), compared with those who received placebo maintenance. Conclusions: In patients with fistulizing Crohn’s disease, infliximab 5 mg/kg every 8 weeks significantly reduced hospitalizations, surgeries, and procedures compared with placebo. Background & Aims: Infliximab is effective in closing fistulas in patients with Crohn’s disease. We examined the effect of infliximab maintenance treatment on hospitalizations, surgeries, and procedures in patients with fistulizing Crohn’s disease enrolled in the ACCENT II study. Methods: After 5 mg/kg infliximab at weeks 0, 2, and 6, a total of 282 patients were separately randomized at week 14 as responders (at least a 50% reduction from baseline in the number of draining fistulas at both weeks 10 and 14) or nonresponders to receive placebo or 5 mg/kg infliximab maintenance every 8 weeks. At week 22 and later, patients who lost response could be treated with a maintenance dose 5 mg/kg higher. Data on Crohn’s disease-related hospitalizations, surgeries, and procedures were compared between the treatment groups for responders and all randomized patients. Results: A total of 282 patients were randomized at week 14, of whom 195 were randomized as responders. Among patients randomized as responders, those who received infliximab maintenance had significantly fewer mean hospitalization days (0.5 vs. 2.5 days; P < .05), mean numbers (per 100 patients) of hospitalizations (11 vs. 31; P < .05), all surgeries and procedures (65 vs. 126; P < .05), inpatient surgeries and procedures (7 vs. 41; P < .01), and major surgeries (2 vs. 11; P < .05), compared with those who received placebo maintenance. Conclusions: In patients with fistulizing Crohn’s disease, infliximab 5 mg/kg every 8 weeks significantly reduced hospitalizations, surgeries, and procedures compared with placebo. Fistulas can occur in patients with Crohn’s disease and result from inflammation penetrating from affected bowel into adjacent organs, tissues, or skin. It has been estimated in population-based cohort studies that 20%–40% of individuals with Crohn’s disease will develop fistulas over the lifetime of their disease.1Steinberg D.M. Cooke W.T. Alexander-Williams J. Abscess and fistulae in Crohn’s disease.Gut. 1973; 14: 865-869Crossref PubMed Scopus (121) Google Scholar, 2Farmer R.G. Hawk W.A. Turnbull Jr, R.B. Clinical patterns in Crohn’s disease a statistical study of 615 cases.Gastroenterology. 1975; 68: 627-635Abstract Full Text PDF PubMed Scopus (573) Google Scholar Crohn’s disease is associated with significant health-care costs,3Hay J.W. Hay A.R. Inflammatory bowel disease costs-of-illness.J Clin Gastroenterol. 1992; 14: 309-317Crossref PubMed Scopus (216) Google Scholar, 4Feagan B.G. Vreeland M.G. Larson L.R. Bala M.V. Annual cost of care for Crohn’s disease a payor perspective.Am J Gastroenterol. 2000; 95: 1955-1960Crossref PubMed Google Scholar, 5Bodger K. Cost of illness of Crohn’s disease.Pharmacoeconomics. 2002; 20: 639-652Crossref PubMed Scopus (77) Google Scholar and patients with fistulas may account disproportionately for these costs. A prospective study evaluating the natural history of Crohn’s disease showed that 33% of patients with chronically or intermittently active disease developed complications requiring hospitalization and surgery in the first year after diagnosis, 13% in the second year, and 3% in each subsequent year.6Binder V. Hendriksen C. Kreiner S. Prognosis in Crohn’s disease-based on results from a regional patient group from the county of Copenhagen.Gut. 1985; 26: 146-150Crossref PubMed Scopus (255) Google Scholar In addition, a population-based study found that 83% of patients with fistulas required surgery and that 23% of the patients required a bowel resection.7Schwartz D.A. Loftus Jr, E.V. Tremaine W.J. Panaccione R. Harmsen W.S. Zinsmeister A.R. Sandborn W.J. The natural history of fistulizing Crohn’s disease in Olmsted county, Minnesota.Gastroenterology. 2002; 122: 875-880Abstract Full Text Full Text PDF PubMed Scopus (777) Google Scholar, 8Present D.H. Crohn’s fistula current concepts in management.Gastroenterology. 2003; 124: 1629-1635Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar These aspects of care (hospitalization and surgery) account for most of the direct cost of care for patients with Crohn’s disease.3Hay J.W. Hay A.R. Inflammatory bowel disease costs-of-illness.J Clin Gastroenterol. 1992; 14: 309-317Crossref PubMed Scopus (216) Google Scholar, 4Feagan B.G. Vreeland M.G. Larson L.R. Bala M.V. Annual cost of care for Crohn’s disease a payor perspective.Am J Gastroenterol. 2000; 95: 1955-1960Crossref PubMed Google Scholar, 5Bodger K. Cost of illness of Crohn’s disease.Pharmacoeconomics. 2002; 20: 639-652Crossref PubMed Scopus (77) Google Scholar If new therapies can reduce hospitalizations and surgeries, the resulting cost savings might substantially offset the incremental cost of these therapies. A 3-dose induction regimen of infliximab given at weeks 0, 2, and 6 results in fistula closure.9Present D.H. Rutgeerts P. Targan S. Hanauer S.B. Mayer L. van Hogezand R.A. Podolsky D.K. Sands B.E. Braakman T. DeWoody K.L. Schaible T.F. van Deventer S.J. Infliximab for the treatment of fistulas in patients with Crohn’s disease.N Engl J Med. 1999; 340: 1398-1405Crossref PubMed Scopus (2496) Google Scholar Recently, the ACCENT II trial examined the effects of maintenance dosing with infliximab in patients with fistulizing Crohn’s disease. This study showed that patients who continued to receive infusions of infliximab every 8 weeks were significantly more likely to maintain fistula closure than patients who stopped treatment after induction.10Sands B.E. Anderson F.H. Bernstein C.N. Chey W.Y. Feagan B.G. Fedorak R.N. Kamm M.A. Korzenik J.R. Lashner B.A. Onken J.E. Rachmilewitz D. Rutgeerts P. Wild G. Wolf D.C. Marsters P.A. Travers S.B. Blank M.A. van Deventer S.J. Infliximab maintenance therapy for fistulizing Crohn’s disease.N Engl J Med. 2004; 350: 876-885Crossref PubMed Scopus (1909) Google Scholar We evaluated the effect of an infliximab maintenance regimen on hospitalizations, surgeries, and procedures in patients enrolled in the ACCENT II trial. ACCENT II was a phase III, multicenter, randomized, double-blind, placebo-controlled clinical trial that compared the efficacy and safety of maintenance-dose administration with infliximab with that of a 3-dose induction regimen of infliximab only in reducing the number of draining fistulas.10Sands B.E. Anderson F.H. Bernstein C.N. Chey W.Y. Feagan B.G. Fedorak R.N. Kamm M.A. Korzenik J.R. Lashner B.A. Onken J.E. Rachmilewitz D. Rutgeerts P. Wild G. Wolf D.C. Marsters P.A. Travers S.B. Blank M.A. van Deventer S.J. Infliximab maintenance therapy for fistulizing Crohn’s disease.N Engl J Med. 2004; 350: 876-885Crossref PubMed Scopus (1909) Google Scholar Eligible patients included men and women (≥18 years) with single or multiple distinctly identifiable draining fistulas, including perianal fistulas and enterocutaneous fistulas, of at least 3 months’ duration. Women with rectovaginal fistulas were included if at least 1 other enterocutaneous draining fistula was present. Setons were permitted at screening but were required to be removed by week 2. Concurrent therapies for Crohn’s disease were permitted at stable doses and included 5-aminosalicylates, oral glucocorticoids, azathioprine, 6-mercaptopurine, mycophenolate mofetil, methotrexate, and antibiotics. Patients receiving corticosteroids who showed a fistula response at week 6 were permitted to begin tapering their dose at that time. Excluded from the study were patients who had either strictures or abscesses for which surgery might be indicated and those who had been previously treated with infliximab. Patients were screened for eligibility 2 weeks before enrollment. All eligible patients received 5 mg/kg intravenous infusions of infliximab at weeks 0, 2, and 6. A fistula response9Present D.H. Rutgeerts P. Targan S. Hanauer S.B. Mayer L. van Hogezand R.A. Podolsky D.K. Sands B.E. Braakman T. DeWoody K.L. Schaible T.F. van Deventer S.J. Infliximab for the treatment of fistulas in patients with Crohn’s disease.N Engl J Med. 1999; 340: 1398-1405Crossref PubMed Scopus (2496) Google Scholar was defined as at least a 50% reduction from baseline in the number of draining fistulas at consecutive visits conducted ≥4 weeks apart. A patient was classified as a responder if a response was observed at both weeks 10 and 14. At week 14, both responders and nonresponders were randomized to receive infusions of placebo (placebo maintenance) or 5 mg/kg infliximab (infliximab maintenance) at weeks 14, 22, 30, 38, and 46. Patients were followed up to week 54. A computer-generated adaptive randomization scheme was used that included study site, number of draining fistulas at baseline (1 vs. >1), and the presence or absence of active bowel disease at baseline (present if the Crohn’s Disease Activity Index11Best W.R. Becktel J.M. Singleton J.W. Kern F. Development of a Crohn’s disease activity index. National Cooperative Crohn’s Disease Study.Gastroenterology. 1976; 70: 439-444Abstract Full Text PDF PubMed Scopus (3097) Google Scholar [CDAI] was ≥150) as stratification factors. Loss of response was defined as recurrent draining fistulas such that conditions for response were no longer met; or a prohibited change in Crohn’s disease medication, a surgery for Crohn’s disease, or discontinuation of the study medication because of lack of efficacy. Beginning at week 22, patients who had a loss of response were eligible to cross over to 5 or 10 mg/kg infliximab from their initial treatment of placebo or 5 mg/kg infliximab, respectively, and maintain dose administration every 8 weeks. Patients were to maintain concomitant medications, except for corticosteroids, at a constant dose. Safety data were collected on all randomized patients and were included in the safety analysis. The primary end point was time to loss of fistula response through week 54. Fistula examinations were conducted at weeks 0, 2, 6, 10, 14, 22, 30, 38, 46, and 54. Data on Crohn’s disease–related hospitalizations, surgeries, and procedures were collected from baseline to the end of the study or early withdrawal and were totaled for the entire period. Patient baseline demographics and disease characteristics were summarized by using descriptive statistics for all randomized patients and patients who were randomized as responders at week 14. An intention-to-treat principle was applied to the statistical analyses. The numbers of Crohn’s disease–related hospitalizations, hospitalization days, all surgeries and procedures, inpatient surgeries and procedures, and major surgeries (those that were deemed by blinded assessors to be major enough to categorize a patient as a treatment failure in the trial, excluding drainage of abscesses, seton placement, and dilation of strictures) were compared between the 2 treatment groups by using regression with a generalized linear model. The number of weeks for which a patient was in the study was used as the offset variable in this analysis. Although count data such as these are usually assumed to follow a Poisson distribution, a negative binomial distribution was used instead to correct for potential overdispersion related to the high incidence of zeros in these numbers (most patients were not hospitalized, and many did not have surgeries or procedures). PROC GENMOD in SAS version 8.02 (SAS Institute, Cary, NC) was used for this analysis. The proportion of patients who were hospitalized during the study for Crohn’s disease-related reasons was compared between the 2 treatment groups by using a χ2 test. The time (days) to first hospitalization was compared between the 2 treatment groups by using a log-rank test with Kaplan-Meier analysis in SAS PROC LIFETEST. The results for the number of hospitalizations and surgeries/procedures are presented as the mean number per 100 patients by averaging the number of events among patients and then multiplying by 100. Thus, multiple hospitalizations or surgeries experienced by any given patient are included in these results. Logistic regression analysis was used to explore factors (C-reactive protein level, CDAI, baseline number of draining fistulas, baseline dose of concomitant corticosteroids, baseline dose of azathioprine/6-mercaptopurine, and treatment group) that may have been associated with hospitalization. In these analyses, all original data were used, including data after crossover to 5 mg/kg infliximab every 8 weeks in patients who were originally in the placebo maintenance group and after crossover to 10 mg/kg infliximab every 8 weeks in patients who were originally in the 5 mg/kg infliximab maintenance group. Most of these analyses were performed post hoc without multiplicity adjustment. A summary of the baseline characteristics of all randomized patients and patients randomized as responders at week 14 is presented in Table 1. The baseline patient demographics and disease characteristics were similar (P > .05) between the 2 treatment groups. The overall patient population was predominantly white (>96%), with an average age of approximately 38 years and a balanced sex ratio. The mean duration of Crohn’s disease was 12.5 years, and more than 50% of the patients had ≥2 fistulas. Approximately half of the patients had both ileal and colonic involvement, and slightly more than half had previous segmental resections. At baseline, approximately 28% of patients were taking corticosteroids, approximately 34% were taking immunomodulators, and approximately 46% were taking aminosalicylates as concomitant medications.Table 1Baseline Patient CharacteristicsVariableAll randomized patients (n = 282)Patients randomized as responders (n = 195)Demographics Age, y (mean ± SD)39 ± 1238 ± 11 No. female (%)138 (48.9%)94 (48.2%) White race (%)273 (96.8%)186 (95.4%)Disease characteristics Mean Crohn’s disease duration, y (median)12.5 (11.5)12.4 (11.2) Patients with >1 draining fistula (%)159 (56.4%)115 (59.0%) CDAI, median (mean ± SD)174 (193 ± 102)176 (191 ± 102)Involved intestinal areas, n (%) Ileum48 (17.0%)34 (17.4%) Colon92 (32.6%)64 (32.8%) Ileum and colon142 (50.4%)97 (49.7%) Extraintestinal manifestations (%)99 (35.5%)67 (34.7%) Previous segmental resections (%)156 (55.3%)109 (55.9%)Medications (%) Corticosteroids81 (28.7%)55 (28.2%) Immunomodulatory agents97 (34.4%)67 (34.4%) Aminosalicylates132 (46.8%)90 (46.2%) Open table in a new tab The results of the primary end-point analysis were presented previously.10Sands B.E. Anderson F.H. Bernstein C.N. Chey W.Y. Feagan B.G. Fedorak R.N. Kamm M.A. Korzenik J.R. Lashner B.A. Onken J.E. Rachmilewitz D. Rutgeerts P. Wild G. Wolf D.C. Marsters P.A. Travers S.B. Blank M.A. van Deventer S.J. Infliximab maintenance therapy for fistulizing Crohn’s disease.N Engl J Med. 2004; 350: 876-885Crossref PubMed Scopus (1909) Google Scholar The time to loss of response was significantly longer for patients receiving 5 mg/kg infliximab maintenance treatment than for patients receiving placebo maintenance treatment (P < .001). The median time from randomization at week 14 to loss of response was 14 weeks for the placebo maintenance group and was >40 weeks (never reached by the end of the study) for the 5 mg/kg infliximab maintenance group. As shown in Table 2, there was a significant difference (P < .05) between the 2 treatment groups in the mean number of Crohn’s disease-related hospitalizations for all randomized patients and patients randomized as responders. The 5 mg/kg infliximab maintenance group had a reduction of >50% in hospitalizations compared with the placebo maintenance group (14 vs. 31 and 11 vs. 31 hospitalizations per 100 patients, respectively, for all randomized patients and patients randomized as responders). In addition, the mean number of hospitalization days per patient in the 5 mg/kg infliximab maintenance group was reduced compared with that in the placebo maintenance group by 67% and 80% (0.8 vs. 2.4 days and 0.5 vs. 2.5 days per patient, respectively), for all randomized patients and patients randomized as responders. The difference in this average duration of hospitalizations was also significant (P < .05) between the 2 treatment groups for patients randomized as responders and had a trend toward significance (P = .110) for all randomized patients. Among patients randomized as nonresponders, the 5 mg/kg infliximab maintenance group also had lower mean numbers of hospitalizations (19 vs. 32 per 100 patients) and hospitalization days (1.6 vs. 2.3 days per patient) than the placebo maintenance group, but the differences were not significant (P > .05). An additional analysis showed that the reduction of the number of hospitalizations in the infliximab maintenance group was consistent between the United States and non-US countries.Table 2Number of Crohn’s Disease-Related Hospitalizations, Hospitalization Days, Surgeries and Procedures by Treatment GroupVariableAll randomized patients (n = 282)Patients randomized as responders (n = 195)Placebo maintenance (n = 143)5 mg/kg infliximab maintenance (n = 139)aP values are for comparisons between the 2 treatment groups.Placebo maintenance (n = 99)5 mg/kg infliximab maintenance (n = 96)aP values are for comparisons between the 2 treatment groups.HospitalizationsbEach entry is the mean number per 100 patients with the total number in parentheses.31 (45)14 (19) P < .0531 (31)11 (11) P < .05Hospitalization dayscEach entry is the mean number of days hospitalized per patient.2.40.8 P = .1102.50.5 P < .05All surgeries and proceduresbEach entry is the mean number per 100 patients with the total number in parentheses.118 (169)60 (83) P < .01126 (125)65 (62) P < .05Inpatient surgeries and proceduresbEach entry is the mean number per 100 patients with the total number in parentheses.45 (65)10 (14) P < .00141 (41)7 (7) P < .01Major surgeriesbEach entry is the mean number per 100 patients with the total number in parentheses.13 (18)2 (3) P < .0511 (11)2 (2) P < .05a P values are for comparisons between the 2 treatment groups.b Each entry is the mean number per 100 patients with the total number in parentheses.c Each entry is the mean number of days hospitalized per patient. Open table in a new tab Figure 1 illustrates the cumulative number of hospitalizations during the study in the 2 treatment groups for patients randomized as responders. The first hospitalization occurred on day 5 in the 5 mg/kg infliximab maintenance group of the study and on day 61 in the placebo maintenance group. However, the rate of increase in the total number of hospitalizations was more gradual in the 5 mg/kg infliximab maintenance group than in the placebo maintenance group. The cumulative number of hospitalizations in the 2 treatment groups diverged after week 14, when patients were randomized to the placebo or infliximab maintenance groups. A similar pattern was observed in the cumulative number of hospitalizations for all randomized patients, with a total number of hospitalizations of 19 and 45, respectively, in the 5 mg/kg infliximab and placebo maintenance groups. A separate time-to-event analysis showed that the time (number of days) to first hospitalization was significantly longer (P < .05) in the 5 mg/kg infliximab maintenance group than in the placebo maintenance group for patients randomized as responders (Figure 2 shows the Kaplan-Meier curves for this analysis).Figure 2Kaplan-Meier curves for time to first hospitalization for patients randomized as responders. The figure provides the percentage of patients who had not been hospitalized by time in each treatment group.View Large Image Figure ViewerDownload (PPT) Figure 3 shows that the proportion of patients hospitalized in the 5 mg/kg infliximab maintenance group was less than half that in the placebo maintenance group (approximately 8% vs. approximately 18.5%). This difference was significant (P < .05) for all randomized patients and patients randomized as responders. In logistic regression analysis on the placebo maintenance group of all randomized patients, baseline CDAI was the only variable that predicted hospitalization (P < .05; higher CDAI was associated with higher likelihood of hospitalization). When patients in the 5 mg/kg infliximab maintenance group and the treatment group variable were also included in this analysis, the specific treatment group became the only predictor of hospitalization (P < .05; patients in the placebo maintenance group had a higher likelihood of hospitalization than those in the 5 mg/kg infliximab maintenance group). The 5 mg/kg infliximab maintenance group had an approximately 50% reduction in the mean number of all surgeries and procedures, compared with the placebo maintenance group (60 vs. 118 and 65 vs. 126 surgeries and procedures per 100 patients, respectively, for all randomized patients and patients randomized as responders; Table 2). This difference was highly significant (P < .01) for all randomized patients and was significant (P < .05) for patients randomized as responders. The mean number of inpatient surgeries and procedures in the 5 mg/kg infliximab maintenance group was significantly reduced compared with that in the placebo maintenance group: by >70% (10 vs. 45 per 100 patients; P < .001) and >80% (7 vs. 41 per 100 patients; P < .01) for all randomized patients and patients randomized as responders, respectively. These inpatient surgeries and procedures included all but 2 major surgeries (i.e., fistula excision and anal fistulotomy; see Table 3), other surgeries (predominantly incision and drainage of abscesses: 9 in the 5 mg/kg infliximab maintenance group vs. 21 in the placebo maintenance group), and procedures. When only major surgeries were considered, the mean number of major surgeries per 100 patients in the 5 mg/kg infliximab maintenance group was reduced by >80% when compared with that in the placebo maintenance group (2 vs. 13 and 2 vs. 11, respectively, for all randomized patients and for patients randomized as responders; P < .05). Among patients randomized as nonresponders, the 5 mg/kg infliximab maintenance group also had lower mean numbers (per 100 patients) of all surgeries and procedures (49 vs. 100), inpatient surgeries and procedures (16 vs. 55), and major surgeries (2 vs. 16) than the placebo maintenance group, but the differences were not significant (P > .05).Table 3Listing of Major Surgeries by Indication for All Randomized PatientsaText for surgery and indication is as recorded in the case report forms.TreatmentPatientSurgeryIndication5 mg/kg maintenanceAFistula excisionbOutpatient surgeries.During surgery for presumed abscessBResection gastrocutaneous fistula and enterovaginal fistula, ileocecal resectionChronic fistulaCSmall-bowel resectionSmall-bowel obstructionPlacebo maintenanceDAnal fistulotomybOutpatient surgeries.Attempted seton placement to abscessESmall-bowel resectionStrictureFIleocolonic resectionExacerbation of Crohn’s diseaseGRevision ileostomyRecurrent Crohn’s diseaseHColostomy—permanentWorsening of fistulizing Crohn’s diseaseFistuloectomyWorsening of fistulizing Crohn’s diseaseResection of rectum—abdominoperinealWorsening of fistulizing Crohn’s diseaseILaparoscopically assisted ileal cecectomy with primary hand sewn anastomosisExacerbation of Crohn’sJIleostomyIschemia in previous stomaKIntestinal anastomosisActive Crohn’s diseaseLFistula surgically removedDraining abdominal fistulaSmall-bowel resectionCrohn’s diseaseMResection of rectum—otherFistula not closedNAbdominal perianal resectionRefractory Crohn’sResetting ileostomyRefractory Crohn’sSmall-bowel resectionRefractory Crohn’sSmall-bowel stricturoplastyRefractory Crohn’sORelocation of ileostomyIncrease in severity and number of fistulasa Text for surgery and indication is as recorded in the case report forms.b Outpatient surgeries. Open table in a new tab Table 3 presents a listing of all major surgeries and their indications (reasons) among all randomized patients. The most frequently occurring major surgeries were resection of the bowel (8 in the placebo maintenance group and 1 in the 5 mg/kg infliximab maintenance group), fistula-related surgeries (fistula resection or fistulotomy: 3 and 2, respectively), and ostomy placement or revision (5 and 0, respectively). All of the major surgeries were performed in the inpatient setting except for 2 (fistula excision for patient A and anal fistulotomy for patient D). Table 4 summarizes the categories of all nonmajor surgeries and procedures among all randomized patients and the actual number in each category, by treatment group.Table 4Number of Nonmajor Surgeries and Procedures, by Treatment Group (All Randomized Patients)Procedure categoryPlacebo maintenance (n = 143)5 mg/kg infliximab maintenance (n = 139)Endoscopy (colonoscopy or sigmoidoscopy)2510Incision and drainage of abscesses4940 Open incision and drainage2210 Seton placement only2730Imaging procedures5924 Ultrasonography2821 CT scan160 MRI73 Radiographic imaging80Rectal dilation60Laparotomy31Total parenteral nutrition42Other53CT, computed tomography; MRI, magnetic resonance imaging. Open table in a new tab CT, computed tomography; MRI, magnetic resonance imaging. Similar to the cumulative number of hospitalizations, the cumulative number of surgeries and procedures for patients randomized as responders during the study increased at a relatively slow rate in the 5 mg/kg infliximab maintenance group but much more markedly in the placebo maintenance group (Figure 4). From the beginning of the enrollment to approximately week 10, the total number of surgeries and procedures in the 5 mg/kg infliximab maintenance group was slightly higher than in the placebo maintenance group (each increased from slightly more than 10 to slightly more than 20). However, the separation became apparent starting at week 14, when patients were randomized to placebo or infliximab maintenance groups: the number in the placebo maintenance group exceeded that in the 5 mg/kg infliximab maintenance group. A similar pattern was observed for all randomized patients. As previously described,10Sands B.E. Anderson F.H. Bernstein C.N. Chey W.Y. Feagan B.G. Fedorak R.N. Kamm M.A. Korzenik J.R. Lashner B.A. Onken J.E. Rachmilewitz D. Rutgeerts P. Wild G. Wolf D.C. Marsters P.A. Travers S.B. Blank M.A. van Deventer S.J. Infliximab maintenance therapy for fistulizing Crohn’s disease.N Engl J Med. 2004; 350: 876-885Crossref PubMed Scopus (1909) Google Scholar therapy with infliximab was safe and well tolerated in patients with fistulizing Crohn’s disease. The incidence of adverse events was similar in the placebo and infliximab maintenance groups. Few patients discontinued treatment because of adverse events. The most commonly reported serious adverse events were related to the gastrointestinal system. Worsening of Crohn’s disease was the most frequently reported event; it occurred in 6% of all randomized patients.10Sands B.E. Anderson F.H. Bernstein C.N. Chey W.Y. Feagan B.G. Fedorak R.N. Kamm M.A. Korzenik J.R. Lashner B.A. Onken J.E. Rachmilewitz D. Rutgeerts P. Wild G. Wolf D.C. Marsters P.A. Travers S.B. Blank M.A. van Deventer S.J. Infliximab maintenance therapy for fistulizing Crohn’s disease.N Engl J Med. 2004; 350: 876-885Crossref PubMed Scopus (1909) Google Scholar The subset of patients with Crohn’s disease who have fistulas may be distinct and present unique therapeutic challenges. The proportion of patients with fistulas seems to gradually increase during follow-up, such that nearly half of patients develop a fistula over the course of their disease. Fistulas are associated with a decreased quality of life for patients and represent a distinct therapeutic challenge, often requiring surgical intervention. Our data are the first from a large-scale, double-blind, randomized, well-controlled trial to show that medical therapy in fistulizing Crohn’s disease may reduce the need for hospitalizations, surgeries, and procedures, with their attendant morbidity and cost. Specifically, maintenance treatment with infliximab at 5 mg/kg given every 8 weeks was associated with a >50% reduction in the mean number of Crohn’s disease-related hospitalizations compared with placebo maintenance treatment. The mean number of days hospitalized for patients receiving 5 mg/kg infliximab maintenance treatment was reduced by at least 67% compared with those receiving placebo maintenance treatment. Moreover, for patients receiving 5 mg/kg infliximab maintenance treatment, the likelihood of being hospitalized was reduced by >50%, and the time to first hospitalization was significantly longer compared with placebo maintenance treatment. Patients randomized as responders seemed to have lower mean numbers of hospitalizations and hospitalization days than all randomized patients. Infliximab maintenance treatment was also associated with a >70% reduction in the mean numbers of Crohn’s disease-related inpatient surgeries and procedures and major surgeries and had an approximately 50% reduction in all surgeries and procedures, as compared with the placebo maintenance treatment. In general, the mean numbers of inpatient surgeries and procedures and major surgeries were somewhat lower in patients randomized as responders than in all randomized patients. This indicates that the benefits in nonresponders were not as substantial as in responders. It should also be noted that because the analyses included data from patients in the placebo maintenance group who crossed over to receive 5 mg/kg infliximab every 8 weeks, the benefits of infliximab maintenance treatment might have been underestimated. Crohn’s disease is associated with significant health-care costs,3Hay J.W. Hay A.R. Inflammatory bowel disease costs-of-illness.J Clin Gastroenterol. 1992; 14: 309-317Crossref PubMed Scopus (216) Google Scholar, 4Feagan B.G. Vreeland M.G. Larson L.R. Bala M.V. Annual cost of care for Crohn’s disease a payor perspective.Am J Gastroenterol. 2000; 95: 1955-1960Crossref PubMed Google Scholar, 5Bodger K. Cost of illness of Crohn’s disease.Pharmacoeconomics. 2002; 20: 639-652Crossref PubMed Scopus (77) Google Scholar, 12Silverstein M.D. Loftus E.V. Sandborn W.J. Tremaine W.J. Feagan B.G. Nietert P.J. Harmsen W.S. Zinsmeister A.R. Clinical course and costs of care for Crohn’s disease Markov model analysis of a population-based cohort.Gastroenterology. 1999; 117: 49-57Abstract Full Text Full Text PDF PubMed Scopus (317) Google Scholar and the cost associated with Crohn’s disease-related hospitalizations is substantial.13Cohen R.D. Larson L.R. Roth J.M. Becker R.V. Mummert L.L. The cost of hospitalization in Crohn’s disease.Am J Gastroenterol. 2000; 95: 524-530Crossref PubMed Google Scholar On the basis of a review of the literature, Bodger5Bodger K. Cost of illness of Crohn’s disease.Pharmacoeconomics. 2002; 20: 639-652Crossref PubMed Scopus (77) Google Scholar found that average annual direct costs for a Crohn’s disease patient in the United States were between US$6561 (1990 values) and US$12,417 (1994 values), and more than half of the average costs were related to hospitalization. Previous research has attempted to examine the effect of treatment on health-care resource utilization associated with treating Crohn’s disease. For example, Rubenstein et al.14Rubenstein J.H. Chong R.Y. Cohen R. Infliximab decreases resource use among patients with Crohn’s disease.J Clin Gastroenterol. 2002; 35: 151-156Crossref PubMed Scopus (104) Google Scholar performed a chart review of 79 Crohn’s disease patients (37 of whom had fistulous disease) who had been followed up at their institution for at least 3 years before and 1 year after initiation of infliximab therapy. In this study, decreases of 59%, 59%, and 66% were observed in the annual rate of hospitalization, gastrointestinal surgeries, and all surgeries, respectively, after initiation of infliximab treatment, compared with the period before treatment initiation. Similarly, in the ACCENT I trial, which enrolled patients who had moderate to severe luminal Crohn’s disease with no evidence of draining fistulas, the mean number of hospitalizations was lower in patients receiving maintenance treatment with infliximab compared with those receiving episodic treatment.15Rutgeerts P. Feagan B.G. Lichtenstein G.R. Mayer L.F. Schreiber S. Colombel J.F. Rachmilelwitz D. Wolf D.C. Olson A. Bao W. Hanauer S.B. Comparison of scheduled and episodic treatment strategies of infliximab in Crohn’s disease.Gastroenterology. 2004; 126: 402-413Abstract Full Text Full Text PDF PubMed Scopus (871) Google Scholar Few studies have examined the cost-effectiveness of infliximab and other therapies in fistulizing Crohn’s disease. Arseneau et al16Arseneau K.O. Cohn S.M. Cominelli F. Connors Jr, A.F. Cost-utility of initial medical management for Crohn’s disease perianal fistulae.Gastroenterology. 2001; 120: 1640-1656Abstract Full Text Full Text PDF PubMed Scopus (87) Google Scholar studied the incremental cost-utility of 3 treatment strategies of infliximab infusions with or without 6-mercaptopurine plus metronidazole, compared with treatment with 6-mercaptopurine plus metronidazole only, in patients with fistulizing Crohn’s disease. They concluded that the incremental benefit of infliximab may not justify its high cost. This analysis was based on modeling, and as the authors pointed out, prospective studies directly comparing infliximab with 6-mercaptopurine plus metronidazole are needed to provide further information. The data presented here are unique in that this is the first report from a randomized, controlled trial of the effect of infliximab maintenance treatment on Crohn’s disease-related hospitalizations and surgeries. It is noteworthy that the number of hospitalizations, surgeries, and procedures remained relatively stable in the 5 mg/kg infliximab maintenance group but had a more marked increase in the group randomized to placebo maintenance after week 14. This can be reasonably ascribed to the lack of infliximab maintenance treatment in this group after the induction treatment, which resulted in the clinical worsening of more patients who required hospitalizations, surgeries, and procedures when compared with the 5 mg/kg infliximab maintenance group. Furthermore, more bowel resections were observed in the placebo maintenance group compared with the infliximab maintenance group. In conclusion, infliximab maintenance therapy every 8 weeks significantly reduces hospitalizations, surgeries, and procedures in patients with fistulizing Crohn’s disease. Formal cost-effectiveness studies are warranted to establish whether the savings derived from infliximab maintenance therapy offset the cost of the drug and adverse effects that might arise in the course of therapy.