Ulcerative colitis is a chronic inflammatory disease of

Transcription

1 GASTROENTEROLOGY 2009;137: CLINICAL Colectomy Rate Comparison After Treatment of Ulcerative Colitis With Placebo or Infliximab WILLIAM J. SANDBORN,* PAUL RUTGEERTS, BRIAN G. FEAGAN, WALTER REINISCH, ALLAN OLSON, JEWEL JOHANNS, # JIANDONG LU, # KEVIN HORGAN, # DANIEL RACHMILEWITZ,** STEPHEN B. HANAUER, GARY R. LICHTENSTEIN, WILLEM J. S. DE VILLIERS, DANIEL PRESENT, BRUCE E. SANDS, ## and JEAN FRÉDÉRIC COLOMBEL*** *Division of Gastroenterology and Hepatology, Mayo Clinic, Rochester, Minnesota; Department of Gastroenterology, University Hospital Gasthuisberg, Leuven, Belgium; Robarts Research Institute, The University of Western Ontario, London, Ontario, Canada; Department of Gastroenterology and Hepatology, Univ Klinik Innere Medizin IV, AKH Wien, Vienna, Austria; R. W. Johnson Pharmaceutical Research and Development, San Diego, California; # Centocor Research & Development, Inc., Malvern, Pennsylvania; **Division of Medicine, Shaare Zedak Medical Center, Jerusalem, Israel; Division of Gastroenterology, University of Chicago, Chicago, Illinois; Department of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania; Division of Digestive Diseases and Nutrition, Department of Medicine, University of Kentucky, Lexington, Kentucky; Division of Gastroenterology, Department of Medicine, Mount Sinai Medical Center, New York, New York; ## Gastrointestinal Unit, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts; and ***Department of Hepatogastroenterology, Hopital Claude Huriez and Centre d Investigation Clinique, Centre Hospitalier Universitaire de Lille, Lille, France This article has an accompanying continuing medical education activity on page Learning Objective: Upon completion of reading this article, successful learners will be able to apply the results of the study to their practice by weighing the potential benefits and the risk of infliximab in individual patients with moderate to severe ulcerative colitis. See related article, Turner D et al, on page 1081 in CGH; podcast interview: org/gastropodcast; see CME quiz on page 1520; see editorial on page BACKGROUND & AIMS: The efficacy of infliximab for treating patients with ulcerative colitis has been established. METHODS: The Active Ulcerative Colitis Trial (ACT)-1 and ACT-2 randomized, double-blind, placebo-controlled studies evaluated infliximab induction and maintenance therapy in moderately to severely active ulcerative colitis. Overall, 728 patients received placebo or infliximab (5 or 10 mg/kg) intravenously at weeks 0, 2, and 6, then every 8 weeks through week 46 (ACT-1) or 22 (ACT-2). Colectomy, hospitalization, and surgery/procedure data through 54 weeks after the first infusion were obtained from ACT-1, ACT-2, and associated data sources. In the prespecified analysis, all data were combined to ascertain time to colectomy. Kaplan Meier product-limit method was used to estimate the cumulative incidence of colectomy, and log-rank test was used to compare the combined infliximab group and placebo. RESULTS: Eighty-seven percent (630 of 728) of patients had complete colectomy follow-up; 13% (98 of 728) of patients had a median follow-up of 6.2 months. The cumulative incidence of colectomy through 54 weeks was 10% for infliximab and 17% for placebo (P.02), yielding an absolute risk reduction of 7%. Compared with placebo, fewer ulcerative colitis-related hospitalizations and surgeries/procedures per 100 patient-years of treatment occurred with infliximab therapy: 40 vs 20 (P.003) and 34 vs 21 (P.03), respectively. Serious adverse events occurring in infliximab-treated patients included serious infections, tuberculosis, histoplasmosis, listeriosis, and malignancy. CONCLUSIONS: Patients with moderately to severely active ulcerative colitis treated with infliximab were less likely to undergo colectomy through 54 weeks than those receiving placebo. To view this article s video abstract, go to the AGA s YouTube Channel. Ulcerative colitis is a chronic inflammatory disease of the colon. Epidemiologic studies in patients with ulcerative colitis have reported 10-year cumulative risks of colectomy ranging from 9% 21%. 1 3 Following colectomy with ileoanal pouch, patients are at risk for pouchitis (50% by 5 years), female infertility, nocturnal fecal incontinence, and pouch failure Aminosalicylates, corticosteroids, azathioprine, cyclosporine, and infliximab are effective for induction and/or maintenance of remission in ulcerative colitis 7,8 ; however, these therapies have not been evaluated for efficacy in reducing colectomy, except in pilot studies of cyclosporine and infliximab in hospitalized patients Medical therapies that reduce colectomy are needed. Abbreviations used in this paper: TNF-, tumor necrosis factor- ; ACT, Active Ulcerative Colitis Trial; RESULTS-UC, REMICADE Safety Under Long-term Study in Ulcerative Colitis by the AGA Institute /09/$36.00 doi: /j.gastro

2 October 2009 COLECTOMY RATES AFTER INFLIXIMAB TREATMENT 1251 Infliximab is an IgG 1 chimeric monoclonal antibody directed against tumor necrosis factor (TNF)-. 12 Two trials, Active Ulcerative Colitis Trial (ACT)-1 and ACT-2, demonstrated that infliximab is effective for induction and maintenance of clinical response and remission in outpatients with moderately to severely active ulcerative colitis despite concurrent medications. 13 At the time of the publication of the ACT-1 and -2 studies, 13 colectomy outcomes were unknown for many of the patients; therefore, colectomy analysis was deferred. Subsequently, protocol amendments, study extensions, and targeted poststudy follow-up allowed more complete data collection for those whose colectomy status was unknown; a prespecified statistical analysis plan then enabled us to evaluate the effect of infliximab therapy on colectomy outcomes. Materials and Methods Patients, Design of the Studies, and Sources of Data ACT-1 and -2 were randomized, double-blind, placebo-controlled trials conducted at 117 sites between March 2002 and March Patients were outpatients with moderately to severely active ulcerative colitis defined as a Mayo Clinic score 14 of 6 to 12 points and an endoscopic subscore of at least 2 points despite concurrent treatment with corticosteroids and/or azathioprine or 6-mercaptopurine and/or 5-aminosalicylate-containing medications (ACT-2 only). Patients who received intravenous corticosteroids within 2 weeks or were judged likely to require colectomy within 12 weeks were excluded. Standard chest radiographs were obtained during screening. Patients with positive tuberculin skin tests with the use of purified protein derivative were ineligible. 13 Eligible patients had an established diagnosis of ulcerative colitis. Endoscopy with biopsy was performed during screening to confirm the diagnosis of ulcerative colitis by both the physician performing the endoscopy and the pathologist reviewing the biopsy specimen. Patients with a diagnosis of indeterminate colitis, Crohn s disease, or clinical findings suggestive of Crohn s disease (ie, fistula or granuloma on biopsy) were ineligible. 13 Patients were randomized to receive intravenous infliximab (REMICADE ; Centocor Ortho Biotech, Inc, Malvern, PA) at doses of 5 or 10 mg/kg or placebo at weeks 0, 2, and 6 and then every 8 weeks through week 46 (ACT-1) or 22 (ACT-2). At week 30, patients who completed ACT-2 and in the opinion of the investigator would benefit from continued treatment were eligible to continue the same blinded treatment in a study extension. Patients were followed through 54 weeks. The primary efficacy end point in both trials was clinical response Figure 1. Enrollment and treatment of patients in the ACT-1 and -2 trials. Data sources for colectomy follow-up through 54 weeks include ACT-1, ACT-2, ACT-2 extension, RESULTS-UC, and retrospective data collection.

3 1252 SANDBORN ET AL GASTROENTEROLOGY Vol. 137, No. 4 at week 8, 13 and the incidence of colectomy was a secondary efficacy end point specified in each protocol. Patients who discontinued ACT-1 or -2 (including the extension) prior to 54 weeks were eligible to enter REMICADE Safety Under Long-term Study in Ulcerative Colitis (RESULTS-UC), a safety follow-up study (Figure 1). Data were obtained on colectomy, hospitalization, commercial infliximab use, death, serious infection, new malignancy, autoimmune diseases, and delayed hypersensitivity reactions. In response to a lower than expected participation rate in RESULTS-UC and to optimize colectomy follow-up through 54 weeks after the first infusion, the ACT-1 and -2 protocols were amended in January 2006 to include a targeted poststudy follow-up, in which data through 54 weeks on colectomy, hospitalization, and commercial infliximab use were retrospectively collected in patients who discontinued treatment and did not enter RESULTS-UC (Figure 1). Data through 54 weeks from all data sources were pooled and analyzed according to the prespecified plan. The Institutional Review Board at each site approved all protocols and amendments. All patients gave written informed consent for study procedures, and additional informed consent was sought for patients targeted for poststudy follow-up. Efficacy Evaluations Data for patients were evaluated through 54 weeks or last follow-up visit for colectomy with the time to colectomy as the primary end point in this prespecified combined analysis. Data were also analyzed for ulcerative colitis-related hospitalizations and surgeries/procedures (all surgeries and endoscopic procedures used to evaluate or treat ulcerative colitis) and colectomy. Safety Evaluations Adverse events through 54 weeks were summarized. Sera for determination of antibodies to infliximab, antinuclear antibodies, and anti-double-stranded DNA antibodies were collected at weeks 0, 30, and 54 and analyzed as previously described. 13,15 Statistical Analysis The primary end point was time to colectomy through 54 weeks after the first infusion in patients enrolled in the combined ACT-1 and -2 trials. Time to colectomy for patients who did not have a colectomy was censored at the last follow-up date through 54 weeks. Time to colectomy was summarized as the cumulative incidence of colectomy through 54 weeks estimated by the Kaplan Meier product-limit method; the placebo and the combined infliximab groups were compared using the log-rank test stratified by study. Other secondary end points were the numbers of ulcerative colitis-related hospitalizations and surgeries/procedures through 54 weeks. The numbers of ulcerative colitis-related hospitalizations and surgeries/procedures through 54 weeks were expressed as events per 100 patient-years, and the combined infliximab and placebo groups were compared using the Wilcoxon rank-sum test. To evaluate the consistency of treatment effect between infliximab and placebo, prespecified subgroup analyses included the following: study (ACT-1 or ACT-2), treatment group (infliximab 5 or 10 mg/kg groups), and 16 demographic or baseline disease characteristics (Figure 2B). The hazard ratio, 95% confidence intervals, and interaction P values were calculated based on Cox proportional hazards model stratified by study (if applicable). Analyses to identify predictors of colectomy from a pool of baseline demographic and disease characteristics and treatment were performed post hoc using Cox proportional hazards models. Demographic and baseline characteristics were compared using a 2 or Fisher exact test for categorical variables or t test for continuous variables. Safety analyses used the Fisher exact test to compare placebo and the combined infliximab group. The sample sizes for ACT-1 and -2 were based on the primary end point of clinical response at week 8. To increase the power to detect a treatment difference for the colectomy end point, the 2 studies and the 2 infliximab groups were pooled. Assuming a hazard ratio of 0.5, and a 2-sided significance level of.05, at least 74 events would provide 80% power to detect a significant difference in the time to colectomy between the combined infliximab and placebo groups. All efficacy analyses used intention-to-treat methods. All reported P values are 2-sided and not adjusted for multiple testing. Results Patient Disposition and Baseline Characteristics In the ACT-1 and -2 trials, 728 patients were randomized to treatment: 244 to placebo and 242 each to infliximab 5 and 10 mg/kg. The baseline characteristics were generally similar among treatment groups (Table 1). One hundred forty-two patients from ACT-2 entered the extension trial (31 in the placebo group, and 52 and 59 in the infliximab 5- and 10-mg/kg groups, respectively), 284 patients from ACT-1 and -2 entered RESULTS-UC, and 135 patients with incomplete colectomy data through 54 weeks were targeted for poststudy follow-up, of whom only 37 patients gave consent for colectomy data collection (Figure 1). Pooling all data sources, 630 of 728 randomized patients (87%) had complete colectomy follow-up data through 54 weeks: (1) 55.1% (401 of 728) of patients from ACT-1 or ACT-2 through extension week 24; (2) 26.4% (192 of 728) of patients from RESULTS-UC; and (3) 5.1% (37 of 728) of patients through targeted poststudy fol-

4 October 2009 COLECTOMY RATES AFTER INFLIXIMAB TREATMENT 1253 Figure 2. (A) Kaplan Meier estimate of the proportion of patients free of colectomy through 54 weeks for the placebo and combined infliximab groups in ACT-1 and -2. (B) Summary of hazard ratios by subgroup for time to colectomy through 54 weeks after the first infusion in ACT-1 and -2. The vertical solid line represents the treatment effect for the overall population. CI, confidence interval; NAP, not applicable; UC, ulcerative colitis; 5-ASA, 5-aminosalicylates; 6-MP, 6-mercaptopurine. (C) Summary of hazard ratios by treatment group and by study for time to colectomy or commercial infliximab use through 54 weeks after the first infusion in ACT-1 and -2. low-up (Figure 1). The remaining 98 patients (13%) (Figure 1) had a median follow-up of 6.2 months (range, months). When measured by the total patientyears of follow-up, 91.6% of potential colectomy follow-up (646.4 patient-years of potential patientyears; Table 2) was obtained. Overall, more patients had complete colectomy follow-up in the infliximab groups as compared with placebo. Among 98 patients with incomplete colectomy follow-up (42 in the placebo group, and 24 and 32 in the infliximab 5- and 10-mg/kg groups, respectively), informed consent could not be obtained because 53 were lost to follow-up, 33 were at sites that did not participate in the targeted poststudy follow-up data collection, 11 refused to consent, and 1 patient withdrew consent. These patients had a median follow-up of 6.2 months. Treatment was discontinued prematurely by 54% (132 of 244) of patients in the placebo group and by 30% (72 of 242) and 32% (78 of 242) of patients in the infliximab 5- and 10-mg/kg groups, respectively (P.001 for all comparisons with placebo).

5 1254 SANDBORN ET AL GASTROENTEROLOGY Vol. 137, No. 4 Table 1. Demographic and Baseline Characteristics of the Pooled ACT 1 and ACT 2 Populations Infliximab Characteristic Placebo 5 mg/kg 10 mg/kg Combined P value a Patients randomized, n Male sex, n (%) 143 (58.6) 154 (63.6) 140 (57.9) 294 (60.7).63 White race, n (%) 228 (93.4) 232 (95.9) 224 (92.6) 456 (94.2).74 Age, y Weight, kg Disease duration, y Colonic area involved Total No. of patients Left side, n (%) 136 (56.7) 133 (56.1) 142 (58.9) 275 (57.5).87 Extensive, n (%) 104 (43.3) 104 (43.9) 99 (41.1) 203 (42.5) Mayo Clinic score b C-reactive protein c Total No. of patients Mean, mg/dl Median, mg/dl Elevated CRP, n (%) 146 (60.8) 154 (64.2) 145 (60.4) 299 (62.3).75 Concomitant medication, n (%) Any ulcerative colitis medication 230 (94.3) 228 (94.2) 230 (95.0) 458 (94.6) Corticosteroids 139 (57.0) 130 (53.7) 139 (57.4) 269 (55.6) mg/day 97 (39.8) 85 (35.1) 93 (38.4) 178 (36.8) Aminosalicylates 174 (71.3) 174 (71.9) 177 (73.1) 351 (72.5).73 Immunomodulators 107 (43.9) 118 (48.8) 109 (45.0) 227 (46.9).48 Azathioprine 71 (29.1) 86 (35.5) 81 (33.5) 167 (34.5) 6-mercaptopurine 36 (14.8) 32 (13.2) 28 (11.6) 60 (12.4) Refractory to corticosteroids, n (%) 74 (30.3) 71 (29.3) 72 (29.8) 143 (29.5).86 Smoking status, n (%).57 Current smoker 13 (5.3) 10 (4.1) 9 (3.7) 19 (3.9) Nonsmoker 123 (50.4) 130 (53.7) 129 (53.3) 259 (53.5) Prior smoker 108 (44.3) 102 (42.1) 104 (43.0) 206 (42.6) NOTE. Plus-minus values are means SD. Race was assigned by the local investigator. a P values for all categorical variables except smoking status are based on Fisher exact test. P values for smoking status are based on the 2 test. P values for all continuous variables except mean C-reactive protein (CRP) are based on the pooled t test. The P value for the mean CRP is based on Satterwaite t test. b The Mayo Clinic scores range from 0 to 12, with higher scores indicating more severe disease. c Elevated baseline C-reactive protein values were those of 0.6 mg/dl or more. Efficacy Time to colectomy through 54 weeks. Eighty-two patients (36 who received placebo; 28 and 18 who received infliximab 5 and 10 mg/kg, respectively) had a colectomy through 54 weeks. The determination of whether or not to proceed to colectomy was based on medical judgment; no criteria were prespecified to determine whether a patient should proceed to colectomy. The reason for colectomy was found to be exacerbation of disease or failure of medical management in 82% (67 of 82) of patients. More than 95% (79 of 82) of patients with colectomies had active ulcerative colitis based on historical, surgical, and/or pathologic evidence. Based on available data for the Mayo score before colectomy, 11% (9 of 82) of patients did not have moderately or severely active ulcerative colitis and had other medical reasons for colectomy (eg, reluctance to remain on corticosteroid therapy, pathology results indicative of active disease). The cumulative incidence of colectomy through 54 weeks was greater for the placebo group (17%) than for the combined infliximab group (10%) (P.02; Figure 2A; Table 2), yielding an absolute risk reduction of 7% with a 95% confidence interval of ( ). The hazard ratio of 0.59 represents a 41% reduction in the risk of colectomy for the combined infliximab group compared with the placebo group (Figure 2B). Patients who received either dose of infliximab had a lower risk of undergoing colectomy than those treated with placebo, but the difference was only significant for 10 mg/kg (P.007) (Table 2, Figure 2B). There appears to be a difference in the treatment effect of infliximab between studies (interaction, P.09), where the overall treatment effect was mainly driven by the difference observed in ACT-2. Efficacy of infliximab was generally consistent among demographic and baseline disease characteristics (Figure 2B). In a multivariate Cox proportional hazards model, in addition to treatment with placebo, concomitant corticosteroid use, baseline C-reactive protein concentrations of at least 2 mg/dl, North American center location,

6 October 2009 COLECTOMY RATES AFTER INFLIXIMAB TREATMENT 1255 Table 2. Duration of Follow-up, Proportion of Patients with Colectomy, Number of Ulcerative Colitis-Related Hospitalizations, and Number of Ulcerative Colitis-Related Surgeries/Procedures Through 54 Weeks Infliximab Placebo 5 mg/kg 10 mg/kg Combined n 244 n 242 n 242 N 484 Duration of follow-up through 54 weeks after the first infusion Potential duration of follow-up in patient-years a Total colectomy follow-up in patient-years (%) b (87.8) (94.6) (92.4) (93.5) Cumulative incidence of patients with colectomy within 54 weeks analysis Patients with colectomy within 54 weeks, n (%) c 36 (17) 28 (12) 18 (8) 46 (10) P value c Proportion of patients with colectomy beyond 54 weeks analysis Patients with colectomy beyond 54 weeks, n (%) 17 (7) 9 (4) 14 (6) 23 (5) Number of ulcerative colitis-related hospitalizations through 54 weeks analysis Hospitalizations, n (%) d (75) 203 (84) 205 (85) 408 (84) 1 46 (19) 31 (13) 33 (14) 64 (13) 2 9 (4) 8 (3) 3 (1) 11 (2) 2 5 (2) 0 (0) 1 (0.4) 1 (0.2) P value e Events per 100 patient-years, n Number of ulcerative colitis-related surgeries/procedures analysis Surgical procedures, n (%) d (81) 207 (86) 214 (88) 421 (87) 1 34 (14) 28 (12) 18 (7) 46 (10) 2 8 (3) 3 (1) 7 (3) 10 (2) 2 5 (2) 4 (2) 3 (1) 7 (1) P value e Events per 100 patient-years, n a Potential duration of follow-up for patients without colectomy is 54 weeks. For patients with colectomy, the potential duration of follow-up is time to colectomy from the first administration date. b Total duration of colectomy follow-up is the sum of the complete colectomy follow-up and incomplete colectomy follow-up. c Percentages were based on the Kaplan Meier estimate. P values were based on the log-rank test stratified by study (ACT-1, ACT-2). d Patients are categorized by the number of hospitalizations or surgeries/procedures experienced. e P values are based on Wilcoxon rank-sum test. disease duration less than 3 years, and a baseline Mayo score of at least 10 points were significantly associated with the risk of colectomy. Results of the full Cox proportional hazards model are presented in Table 3. Secondary end points. There was a significantly reduced number of ulcerative colitis-related hospitalizations per 100 patient-years of treatment for the 5-mg/kg (21), 10-mg/kg (19), and combined (20) infliximab groups compared with placebo (40; P.02, P.007, and P.003, respectively; Table 2). There was a significantly reduced number of ulcerative colitis-related surgeries/procedures per 100 patient-years of treatment for the infliximab 10 mg/kg (19) and combined infliximab (21) groups compared with placebo (34; P.02, P.03, respectively; Table 2). The number of ulcerative colitis-related surgeries/procedures in patients who received infliximab 5 mg/kg was 22 compared with 34 in patients who received placebo (P.15). Commercial infliximab was used by 11% (27 of 244) of patients in the placebo group compared with 6% (28 of 484) in the combined infliximab group. Based on this observation, rescue therapy with commercial infliximab after study discontinuation was evaluated as a potential confounder to the colectomy analysis (Figure 2C). Post hoc analyses of the time to colectomy or use of commercial infliximab demonstrated that infliximab 5 and 10 mg/kg significantly reduced the incidence of colectomy or commercial infliximab use compared with placebo (P.001 and P.001, respectively). These results are consistent across dose groups and studies. Safety The cumulative safety experience through 54 weeks includes data collected from all sources (Table 4). Among all patients, the mean duration of treatment was 23 weeks in the placebo group as compared with 33 weeks in the combined infliximab group (Table 4). The proportion of patients with adverse events was slightly lower in the placebo group (80%) compared with the combined infliximab group (86%) (Table 4). Treatment was discontinued because of an adverse event in 9% of patients in the placebo group compared with 7% in the combined infliximab group. Serious adverse events occurred in 23% of patients in the placebo group and 18% of patients in the combined infliximab group.

7 1256 SANDBORN ET AL GASTROENTEROLOGY Vol. 137, No. 4 Table 3. Factors Predictive of Colectomy Through 54 Weeks No. of patients No. of colectomies within 54 weeks Colectomy adjusted hazard ratio (95% CI) P value Sex Female ( ).89 Male Reference Center location Center located in North America ( ).05 Center located in Western Europe, the Southern Hemisphere, Reference Israel, and the Czech Republic Baseline concomitant medications Azathioprine or 6-mercaptopurine use ( ).71 Azathioprine or 6-mercaptopurine nonuse Reference Aminosalicylates use ( ).73 Aminosalicylates nonuse Reference Corticosteroids use ( ).01 Corticosteroids nonuse Reference Age, y ( ) Reference Baseline C-reactive protein concentration At least 2.0 mg/l ( ).04 Less than 2.0 mg/l Reference Missing 8 2 Weight, kg ( ) Reference Duration of disease, y ( ) Reference Involved colonic area Extensive ( ).38 Left sided Reference Missing 10 3 Baseline Mayo Clinic score points ( ) points Reference Treatment Placebo ( ).05 Infliximab Reference Infections occurred in 40% in the combined infliximab group compared with 33% in the placebo group (Table 4). Serious infections occurred in 4% in the combined infliximab group and 2% in the placebo group. One patient each who received infliximab developed tuberculosis and histoplasmosis. Herpes zoster and varicella zoster occurred in 1% and 0.4%, respectively, in the combined infliximab group, and 0.4% for both in the placebo group. As reported previously, 4 patients developed cancer 13 : 2 (prostate adenocarcinoma, rectal adenocarcinoma) and 1 (basal cell carcinoma) in patients who received infliximab 5 mg/kg and 10 mg/kg, respectively, and 1 who received placebo (basal cell carcinoma). Through 54 weeks in RESULTS-UC, 1 new cancer (squamous cell skin carcinoma) developed in a patient who received infliximab 5 mg/kg. Additionally, 1 patient in each of the 3 treatment groups developed colonic dysplasia through 54 weeks, including 1 reported previously. 13 Through 54 weeks, new autoimmune disorders occurred in 2 patients who received infliximab (lupus-like reaction and sicca syndrome) and 1 patient who received placebo (rheumatoid arthritis). As reported previously, 2 patients reported optic neuritis and 1 patient reported a multifocal motor neuropathy with conduction block syndrome; all 3 patients received infliximab. 13 Through 54 weeks, no new adverse events of neuropathy were reported. No patients died through 54 weeks. After 54 weeks, 6 patients died, 4 who received infliximab (histoplasmosis 4 weeks after the last infusion, 13 listeria encephalitis 3 years after the last infusion, prostate cancer 3.5 years after the last infusion, and natural causes 10 months after the last infusion) and 2 who received placebo (suicide and cerebrovascular accident). Discussion There has been a paucity of data published on colectomy outcomes in patients with ulcerative colitis who are treated with biologic therapy. Prior to the pub-

8 October 2009 COLECTOMY RATES AFTER INFLIXIMAB TREATMENT 1257 Table 4. Cumulative Safety Experience Through 54 Weeks Infliximab Placebo 5 mg/kg 10 mg/kg Combined ACT-1 and -2, and ACT-2 extension through 54 weeks n 244 n 242 n 242 N 484 P value a Mean duration of treatment, wk Mean duration of follow-up, wk Any adverse event, n (%) 196 (80) 208 (86) 209 (86) 417 (86).05 Adverse events occurring in 10% of any treatment group, n (%) Worsening ulcerative colitis 61 (25) 36 (15) 41 (17) 77 (16) Abdominal pain b 31 (13) 22 (9) 36 (15) 58 (12) Nausea 23 (9) 21 (9) 32 (13) 53 (11) Upper respiratory tract infection 43 (18) 39 (16) 48 (20) 87 (18) Pharyngitis 16 (7) 23 (10) 27 (11) 50 (10) Sinusitis 12 (5) 20 (8) 25 (10) 45 (9) Pain b 30 (12) 25 (10) 29 (12) 54 (11) Fatigue 19 (8) 21 (9) 29 (12) 50 (10) Arthralgia 26 (11) 40 (17) 32 (13) 72 (15) Fever 22 (9) 27 (11) 23 (10) 50 (10) Headache 45 (18) 44 (18) 46 (19) 90 (19) Anemia 25 (10) 11 (5) 14 (6) 25 (5) Adverse events of particular interest, n (%) Fungal dermatitis 8 (3) 2 (0.8) 4 (2) 6 (1) Pneumonia 0 (0) 4 (2) 6 (3) 10 (2) Varicella-zoster virus infection c 1 (0.4) 2 (0.8) 0 (0) 2 (0.4) Herpes zoster c 1 (0.4) 3 (1) 2 (0.8) 5 (1) Adverse events leading to study drug discontinuation, n (%) 23 (9) 14 (6) 20 (8) 34 (7).31 Serious adverse events, n (%) 57 (23) 43 (18) 46 (19) 89 (18).12 Infections, n (%) 80 (33) 94 (39) 100 (41) 194 (40).06 Serious infections, n (%) 6 (2) 7 (3) 12 (5) 19 (4).39 Bacterial infection 1 (0.4) 0 (0) 0 (0) 0 (0) Upper respiratory tract infection 1 (0.4) 0 (0) 0 (0) 0 (0) Pneumonia 0 (0) 2 (1) 3 (1) 5 (1) Tuberculosis 0 (0) 0 (0) 1 (0.4) 1 (0.2) Abscess 2 (1) 0 (0) 3 (1) 3 (0.6) Pharyngitis 1 (0.4) 0 (0) 1 (0.4) 1 (0.2) Gastroenteritis 0 (0) 2 (1) 1 (0.4) 3 (0.6) Earache 0 (0) 1 (0.4) 0 (0) 1 (0.2) Fever 0 (0) 1 (0.4) 1 (0.4) 2 (0.4) Vaginitis 0 (0) 0 (0) 1 (0.4) 1 (0.2) Appendicitis 0 (0) 1 (0.4) 0 (0.0) 1 (0.2) Colitis 0 (0) 0 (0) 1 (0.4) 1 (0.2) Infection 1 (0.4) 0 (0) 2 (1) 2 (0.4) Pancreatitis 0 (0) 1 (0.4) 0 (0) 1 (0.2) Pericarditis 0 (0) 0 (0) 1 (0.4) 1 (0.2) Pleurisy 0 (0) 0 (0) 1 (0.4) 1 (0.2) Pyelonephritis 0 (0) 0 (0) 1 (0.4) 1 (0.2) Sinusitis 1 (0.4) 0 (0) 0 (0) 0 (0) Possible delayed hypersensitivity reactions, n (%) 2 (1) 2 (1) 1 (0.4) 3 (0.6) 1.00 Antinuclear antibodies, n (%) d 7/195 (4) 30/212 (14) 28/202 (149) 58/414 (14).001 Antibodies to double-stranded DNA, n (%) d 0/197 (0) 23/214 (11) 18/208 (9) 41/422 (10).001 Antibodies against infliximab during the study, n (%) e 22/208 (11) 14/204 (7) 36/412 (9) RESULTS-UC through 54 weeks Patients with long-term safety follow-up, n Malignancy 0 (0) 1 (0.6) Squamous cell skin carcinoma 0 (0) 1 (0.6) Serious infections 4 (3) 1 (0.6) Infection 1 (0.9) 1 (0.6) Abscess 1 (0.9) 0 (0) Sepsis 1 (0.9) 0 (0) Pancreatitis 0 (0) 0 (0) Cholecystitis 0 (0) 0 (0) Pneumonia 1 (0.9) 0 (0) Autoimmune disorder 1 (0.9) 1 (0.6) Rheumatoid arthritis 1 (0.9) 0 (0) Sicca syndrome 0 (0) 1 (0.6)

9 1258 SANDBORN ET AL GASTROENTEROLOGY Vol. 137, No. 4 Table 4. Continued Infliximab Placebo 5 mg/kg 10 mg/kg Combined ACT-1 and -2, and ACT-2 extension through 54 weeks n 244 n 242 n 242 N 484 Patients with incomplete colectomy data through 54 weeks Patients with long-term safety follow-up, n Mean duration of follow-up, wk Patients with 1 or more serious adverse events, n (%) f 6 (43) 5 (33) 1 (13) 6 (26) Ulcerative colitis 6 (43) 5 (33) 1 (13) 6 (26) Fever 1 (7) 0 (0) 0 (0) 0 (0) P value a NOTE. The composite of ACT 1 through week 54, ACT 2 through week 30, and safety data collected and verified from the ACT 2 study extension through week 24 are summarized in this Table. In addition, safety data are also included from RESULTS-UC and poststudy safety experience through 54 weeks after the first study medication infusion in ACT 1 or ACT 2 and are summarized for infliximab, irrespective of dose group, and placebo. Em dashes denote not applicable. Safety data through week 54 of ACT 1 and week 30 of ACT 2 have been reported previously 13 but are also included here to represent the complete safety experience from these studies. a P values are based on Fisher exact test for the comparison of the combined infliximab group with the placebo group. b Abdominal pain includes adverse event terms of abdominal cramps, abdominal pain, abdominal discomfort, epigastric pain, or gas cramps. Pain includes adverse event terms of bilateral knee and hip pain, right knee pain, or musculoskeletal aches. c Varicella-zoster virus infection includes adverse event terms of varicella and varicella zoster. Herpes zoster includes adverse event terms of shingles or infection by herpes zoster, herpes zoster left leg, and herpes zoster infection. d Denominators represent patients with negative findings at baseline. Antinuclear antibody test results were considered positive if titer was at least 1:320. Samples positive at any time for antinuclear antibodies (titer at least 1:40) were tested for anti-double-stranded DNA antibodies. Anti-double-stranded DNA test results were positive if titer was at least 1:10 (by Crithidia assay) and at least 5.4 IU (by Farr radioimmunoassay). e Denominator represents patients with appropriate samples. Patients with appropriate samples either had antibodies to infliximab at some time point following their first infusion or had 1 or more samples obtained after their last infusion. f Adverse events included here are infliximab-related adverse events or those requiring hospitalization for treatment of UC (including colectomy). lication of the ACT trial results, 2 placebo-controlled pilot studies suggested that infliximab might reduce short-term colectomy rates in hospitalized patients with corticosteroid-refractory ulcerative colitis: in 1 study, 3 of 3 patients in the placebo group and 4 of 8 patients in the infliximab group underwent colectomy during shortterm follow-up; 10 in a second study, 14 of 21 patients in the placebo group and 7 of 24 patients in the infliximab group underwent colectomy by week To evaluate the role of anti-tnf- therapy in preventing colectomy, a prespecified analysis was planned, and a comprehensive program was developed to obtain colectomy outcome data from the large cohort of 728 patients in the ACT trials. In the ACT trials, treatment with infliximab was associated with an absolute risk reduction of 7% in the incidence of colectomy through 54 weeks in outpatients with moderately to severely active ulcerative colitis despite concurrent medications. Of note, patients who received either dose of infliximab had a lower risk of undergoing colectomy than those treated with placebo (36 of 244 patients, 17%); this difference was significant in the 10-mg/kg group (18 of 242 patients, 8%; P.007) but not in the 5-mg/kg group (28 of 242 patients, 12%; P.166). A low colectomy rate was expected in the ACT trials (eligible patients were outpatients who had not received intravenous corticosteroids within 2 weeks and were judged unlikely to require colectomy within 12 weeks). Thus, the overall treatment effect of infliximab associated with an absolute risk reduction for colectomy of 7% needs to be considered in the context of this clinical outcome, which occurs infrequently. Whereas the colectomy rate in patients with moderateto-severe ulcerative colitis was reduced with infliximab therapy in this study, more evidence-based use of infliximab is needed to evaluate the short- and long-term clinical outcomes. Recently, a long-term, retrospective review of colectomy rates in 314 patients with acute severe ulcerative colitis showed that urgent colectomy was avoided in steroid-refractory patients, but the longterm risk of elective colectomy was not reduced. 16 The results of a small, single-center, retrospective uncontrolled analysis of 15 ulcerative colitis patients who received a single infusion of infliximab 5 mg/kg showed that infliximab prevented colectomy in patients with more acute disease (ie, those receiving azathioprine/6- mercaptopurine for less than 8 weeks) compared with those receiving long-term maintenance immunomodulators. 17 Our results show that induction followed by maintenance therapy with infliximab reduced the rate of colectomy through week 54, which may or may not have been elective, and duration of disease ( 5 years vs 5 to 15 years vs 15 years) did not have a significant impact on colectomy rates. Also, previous analyses showed that duration of disease did not affect rates of clinical response, clinical remission, or mucosal healing ( 3 years vs 3 years). 18 Hospitalization accounts for approximately 50% of all lifetime costs associated with ulcerative colitis The

10 October 2009 COLECTOMY RATES AFTER INFLIXIMAB TREATMENT 1259 rates of hospitalization for ulcerative colitis in population-based studies in Canada and the United States range from 8 to 13 per 100,000 individuals In the ACT trials, infliximab reduced the rates of ulcerative colitisrelated hospitalizations (40 vs 20, respectively) and surgeries/procedures (34 vs 21, respectively) per 100 patientyears of treatment in outpatients with moderately to severely active ulcerative colitis. Pharmacoeconomic and health utility studies are needed to understand better the costs and overall outcomes of infliximab in the management of ulcerative colitis. The adverse events observed through 54 weeks in the ACT trials are comparable with those reported with other indications for infliximab. 26 Pneumonia occurred with greater frequency with infliximab compared with placebo. In our safety analysis through 54 weeks, serious adverse events occurred in 23% of patients in the placebo group and 18% of patients in the combined infliximab group. Review of data available from serious adverse event reports indicates no notable complications in patients undergoing colectomy, and no pattern was observed between treatment groups. Anti-TNF- therapies including infliximab have been associated with the potential to increase the risk for infection (some fatal) including sepsis, pneumonia, reactivation of latent tuberculosis, histoplasmosis, and other opportunistic infections; neurologic events (eg, new onset or exacerbation of demyelinating disease); autoimmune disease; and hepatitis B reactivation. There is a possible association for increased risk for malignancy with anti-tnf- therapies, including lymphoma (including recent reports of hepatosplenic T-cell lymphoma), lung cancer, and other malignancies; however, long-term follow-up is needed to characterize better this risk. 26,27 Taken together, the occurrence of herpes zoster, pulmonary tuberculosis, fatal histoplasmosis, and listeriosis, and malignancies seen in patients treated with infliximab in ACT-1 and ACT-2 highlights the importance of vigilance for opportunistic infections and other adverse events possibly related to immunosuppressive therapy. 28 Studies of longer duration are needed to assess better the risk-benefit profile of infliximab in patients with ulcerative colitis. Several limitations of this study warrant mention. Whereas significant improvement in colectomy rate for patients receiving the 10-mg/kg dose of infliximab as opposed to the lower dose and placebo was observed, it is not possible to make a definitive conclusion regarding the dose of infliximab that would have a better colectomysparing effect because the differences observed could have resulted from factors including the low incidence of colectomy within each dose analysis and the differential use of commercial infliximab. The prespecified primary analysis that pooled study populations and combined infliximab dose groups provided the best estimate of the treatment effect of infliximab in reducing the incidence of colectomy. Overall, the greater use of commercial infliximab by patients in the placebo group (27 of 244 patients, 11%) compared with that in the combined infliximab group (28 of 484 patients, 6%) was a potential confounding factor in the original analyses necessitating the exploratory analyses of time to colectomy or use of commercial infliximab. Our study is also limited by the colectomy follow-up differential between the placebo and infliximab groups and the overall proportion of patients with incomplete colectomy follow-up (13%) because informed consent could not be obtained to collect colectomy follow-up data. A general 5-and-20 rule for loss of follow-up in clinical trials suggests that a 5% loss probably leads to little bias and that greater than 20% loss potentially compromises the validity of a trial. Although our study falls within these limits, the expectation for loss of follow-up may trend high because of the length of follow-up (54 weeks) and low expected outcome event rate for colectomy. 29 In conclusion, treatment with infliximab at weeks 0, 2, and 6 and then every 8 weeks reduced the incidence of colectomy through 54 weeks by 41% in outpatients with moderately to severely active ulcerative colitis despite concurrent medications. Supplementary Data Note: To access the supplementary material accompanying this article, visit the online version of Gastroenterology at and at doi: /j.gastro References 1. Hoie O, Wolters FL, Riis L, et al. Low colectomy rates in ulcerative colitis in an unselected European cohort followed for 10 years. Gastroenterology 2007;132: Jess T, Loftus EV Jr, Velayos FS, et al. Risk of intestinal cancer in inflammatory bowel disease: a population-based study from Olmsted County, Minnesota (comment). Gastroenterology 2006;130: Winther KV, Jess T, Langholz E, et al. Long-term risk of cancer in ulcerative colitis: a population-based cohort study from Copenhagen County. Clin Gastroenterol Hepatol 2004;2: Penna C, Dozois R, Tremaine W, et al. Pouchitis after ileal pouchanal anastomosis for ulcerative colitis occurs with increased frequency in patients with associated primary sclerosing cholangitis. Gut 1996;38: Waljee A, Waljee J, Morris AM, et al. Three-fold increased risk of infertility: a meta-analysis of infertility after ileal pouch anal anastomosis in ulcerative colitis. Gut 2006;55: Hahnloser D, Pemberton JH, Wolff BG, et al. The effect of ageing on function and quality of life in ileal pouch patients: a single cohort experience of 409 patients with chronic ulcerative colitis. Ann Surg 2004;240: Kornbluth A, Sachar DB, Practice Parameters Committee of the American College of Gastroenterology. Ulcerative colitis practice guidelines in adults (update): American College of Gastroenterology, Practice Parameters Committee. Am J Gastroenterol 2004;99: Carter MJ, Lobo AJ, Travis SP. Guidelines for the management of inflammatory bowel disease in adults (comment). Gut 2004; 53(Suppl 5):V1 V16.

11 1260 SANDBORN ET AL GASTROENTEROLOGY Vol. 137, No Lichtiger S, Present DH, Kornbluth A, et al. Cyclosporine in severe ulcerative colitis refractory to steroid therapy. N Engl J Med 1994;330: Sands BE, Tremaine WJ, Sandborn WJ, et al. Infliximab in the treatment of severe, steroid-refractory ulcerative colitis: a pilot study. Inflamm Bowel Dis 2001;7: Jarnerot G, Hertervig E, Friis-Liby I, et al. Infliximab as rescue therapy in severe to moderately severe ulcerative colitis: a randomized, placebo-controlled study. Gastroenterology 2005;128: Knight DM, Trinh H, Le J, et al. Construction and initial characterization of a mouse-human chimeric anti-tnf antibody. Mol Immunol 1993;30: Rutgeerts P, Sandborn WJ, Feagan BG, et al. Infliximab induction and maintenance therapy for ulcerative colitis. N Engl J Med 2005;353: Schroeder KW, Tremaine WJ, Ilstrup DM. Coated oral 5-aminosalicylic acid therapy for mildly to moderately active ulcerative colitis. A randomized study. N Engl J Med 1987;317: Hanauer SB, Wagner CL, Bala M, et al. Incidence and importance of antibody responses to infliximab after maintenance or episodic treatment in Crohn s disease. Clin Gastroenterol Hepatol 2004;2: Aratari A, Papi C, Clemente V, et al. Colectomy rate in acute severe ulcerative colitis in the infliximab era. Dig Liver Dis 2008; 40: Willert RP, Lawrance IC. Use of infliximab in the prevention and delay of colectomy in severe steroid dependent and refractory ulcerative colitis. World J Gastroenterol 2008;14: Reinisch W, Sandborn WJ, Rutgeerts P, et al. Infliximab treatment for ulcerative colitis: comparable clinical response, clinical remission, and mucosal healing in patients with disease duration 3 years vs 3 years. Poster presented at: Digestive Disease Week; May 20, 2008; San Diego, CA. 19. Hay AR, Hay JW. Inflammatory bowel disease: medical cost algorithms. J Clin Gastroenterol 1992;14: Hay JW, Hay AR. Inflammatory bowel disease: costs-of-illness. J Clin Gastroenterol 1992;14: Blomqvist P, Ekbom A. Inflammatory bowel diseases: health care and costs in Sweden in Scand J Gastroenterol 1997;32: Odes S, Vardi H, Friger M, et al. Cost analysis and cost determinants in a European inflammatory bowel disease inception cohort with 10 years of follow-up evaluation. Gastroenterology 2006; 131: Nguyen GC, Tuskey A, Dassopoulos T, et al. Rising hospitalization rates for inflammatory bowel disease in the United States between 1998 and Inflamm Bowel Dis 2007;13: Bernstein CN, Nabalamba A. Hospitalization, surgery, and readmission rates of IBD in Canada: a population-based study. Am J Gastroenterol 2006;101: Bewtra M, Su C, Lewis JD. Trends in hospitalization rates for inflammatory bowel disease in the United States. Clin Gastroenterol Hepatol 2007;5: Prescribing information for Remicade (infliximab) [package insert]. Malvern, PA: Centocor Ortho Biotech, Inc, Biancone L, Calabrese E, Petruzziello C, et al. Treatment with biologic therapies and the risk of cancer in patients with IBD. Nat Clin Pract Gastroenterol Hepatol 2007;4: D Haens G. Risks and benefits of biologic therapy for inflammatory bowel diseases. Gut 2007;56: Schultz KF, Grimes DA. Sample size slippages in randomised trials: exclusions and the lost and wayward. Lancet 2002;359: Received April 17, Accepted June 30, Reprint requests Address requests for reprints to: William J. Sandborn, MD, Mayo Clinic, 200 First Street SW, Rochester, Minnesota sandborn.william@mayo.edu; fax: (507) Acknowledgments The authors thank James Barrett and Mary Whitman, PhD, employees of the Medical Affairs Publications Group, Centocor Ortho Biotech, Inc, for editorial and writing support and Prasheen Agarwal, PhD, an employee of Centocor Research and Development, Inc, for statistical support. Conflicts of interest The authors disclose the following: William J. Sandborn, Paul Rutgeerts, Brian G. Feagan, Walter Reinisch, Stephen B. Hanauer, Gary R. Lichtenstein, Willem J. S. de Villiers, Bruce E. Sands, and Jean Frédéric Colombel have served as consultants for and received honoraria and research grants from Centocor Ortho Biotech, Inc. Daniel Present has served as a consultant for and received a research grant from Centocor Research and Development, Inc. Jewel Johanns and Jiandong Lu are employees of Centocor Clinical Research and Development, Inc., a subsidiary of Johnson & Johnson, and own stock in Johnson & Johnson. Allan Olson is a former employee of Centocor Clinical Research and Development, Inc., is currently employed at R. W. Johnson Pharmaceutical Research and Development, and owns stock in Johnson & Johnson. Kevin Horgan is a former employee of Centocor Clinical Research and Development, Inc. Funding Supported by a research grant from Centocor Research and Development, Inc, Malvern, Pennsylvania, and Schering Plough, Kenilworth, New Jersey. Supported by a grant (1-UL1-RR ) from the National Center for Research Resources, a component of the National Institutes of Health (NIH) and the NIH Roadmap for Medical Research. Some of the results presented in this article were published as an abstract and presented at The American College of Gastroenterology 2007 annual meeting in Philadelphia, Pennsylvania (Am J Gastroenterol 2007;102[Suppl 2]:Abs984); United European Gastroenterology Week 2007 annual meeting in Paris, France (Gut 2007;39:A26); and 2007 CCFA National Research and Clinical Conference, 6th Annual Advances in the Inflammatory Bowel Diseases in Aventura, Florida (Inflamm Bowel Dis 2007;14[Suppl 1]:AbsO-006). ClinicalTrials.gov numbers, NCT , NCT , NCT