Infliximab for induction and maintenance therapy for ulcerative colitis

Article Outline

Rutgeerts P, Sandborn WJ, Feagan BG, Reinisch W, Olson A, Johanns J, Travers S, Rachmilewitz D, Hanauer SB, Lichtenstein GR, de Villiers WJ, Present D, Sand BE, Colombel JF (University Hospital Gasthuisberg, Leuven, Belgium). Infliximab for induction and maintenance therapy for ulcerative colitis. N Engl J Med 2005;353:2462–2476.

Infliximab has become well established as therapy for active Crohn’s disease since the first report of its use in this journal over 10 years ago (Gastroenterology 1995;109:129–135). The cumulative reported experience from clinical trials and practice supports its efficacy in induction and maintenance of remission in active Crohn’s disease and its extraintestinal manifestations (Am J Gastroenterol 2002;97:2962–2972; Cochrane Database Syst Rev 2004;CD003574). The adverse effects of infliximab, and the economic implications of its use in Crohn’s disease, have been subject to extensive review (Gastroenterology 2004;126:19–31; Eur J Gastroenterol Hepatol 2005;17:1047–1052).

In ulcerative colitis (UC), however, studies of infliximab to date have been small and associated with mixed efficacy. Over the past 5 years, a number of small case series reported high response rates (>70%) in patients with active UC resistant to or dependent on oral steroids (Inflamm Bowel Dis 2001;[Suppl 1]:S30–33; Aliment Pharmacol Ther 2003;18:175–181; Dig Liver Dis 2002;34:631–634; Am J Gastroenterol 2002;97:2577–2584; Eur Rev Med Pharmacol Sci 2004;8:231–233), whereas 3 small controlled trials reported mixed results, with response rates with infliximab superior to placebo, similar to placebo, or similar to steroids (Inflamm Bowel Dis 2001;7:83–88; Gut 2003;52:998–1002; Eur J Gastroenterol Hepatol 2004;16:1167–1171). As a consequence, infliximab was used infrequently in UC until a recent publication reported its use as “rescue therapy” in severe disease refractory to intravenous steroids (Gastroenterology 2005;128:1805–1811). The results of that Swedish study, which reported that infliximab halved the short-term colectomy rate in severe steroid-refractory disease, has rekindled general interest in anti–tumor necrosis factor (TNF) therapy for UC.

Soon after came the Active Colitis Trials (ACT) 1 and 2, which assessed the efficacy of infliximab for induction and maintenance therapy in active UC (N Engl J Med 2005;353:2462–2476), the results of which are reviewed herein. This paper combines 2 randomized, double-blind, placebo-controlled trials undertaken at 117 centers worldwide, and enrolled 728 patients with active UC. The trials were funded by the manufacturers of infliximab (Centocor, Horsham, PA). Each trial recruited patients with moderate-to-severe UC, confirmed clinically and endoscopically using the Mayo score (N Engl J Med 1987;317:1625–1629), who were then randomized to receive placebo, infliximab 5 mg/kg, or infliximab 10 mg/kg. Participants were allowed concurrent steroid usage, as well as azathioprine/6-mercaptopurine use (in ACT-1) and 5-aminosalicylate use (in ACT-2). Topical therapy was not permitted within 2 weeks of enrollment. Infliximab was administered at 0, 2, and 6 weeks and every 8 weeks thereafter until week 22 (ACT-1) or week 46 (ACT-2). All concurrent medications were continued apart from steroids, which were tapered after week 8 of the trial. Follow-up assessment included endoscopy at week 8, 30 (ACT-1), and 54 (ACT-2). The primary endpoint was clinical response at week 8, based on an intention-to-treat analysis. Any patients who took prohibited medications, discontinued the study owing to lack of efficacy, or underwent surgery were not considered to have had a clinical response.

All patients randomized had similar characteristics, with a number of patients recruited having left-sided disease (54%), elevated C-reactive protein (62%), and on steroids (56%) in both trials. At randomization, most patients on steroids (60%) were taking <20 mg/d, and 72% were taking 5-aminosalicylates, but fewer than half (46%) were on immunomodulators. Almost twice as many patients in the placebo groups as the infliximab groups in both trials discontinued the study early, mostly owing to worsening of their colitis (74/121 in the placebo arm vs 94/243 in the infliximab arms in ACT-1, and 56/123 in the placebo arm vs 50/241 in the infliximab arms in ACT-2).

At week 8, 37.2% of patients in the placebo group, 69.4% of patients in the infliximab 5 mg/kg group, and 61.5% of patients in the 10 mg/kg group had a clinical response in ACT-1 (P < .001). By week 54 (ACT-1), 19.8% of patients in the placebo group, 45.5% of patients in the infliximab 5 mg/kg group, and 44.3% of patients in the 10 mg/kg group were judged to have had a clinical response (P < .001). Similar results were seen in ACT-2, up to 30 weeks. Clinical remission rates at week 8 were 14.9% with placebo, 38.8% with infliximab 5 mg/kg, and 32% with infliximab 10 mg/kg (P < .002). By week 54 (ACT-1), 16.5% of patients receiving placebo, 34.7% on infliximab 5 mg/kg, and 34.4% on infliximab <10 mg/kg remained in remission (P .001). Similar results were seen in ACT-2, although the placebo remission rate was lower at week 8. There was significantly superior mucosal healing, as measured by Mayo score, at weeks 8, 30, and 54 at both infliximab doses in both trials. In addition, the percentage of patients in remission and off steroids at week 30 and week 54 was significantly higher (25% vs 10%) with infliximab 5 mg/kg than with placebo in ACT-1, and with both doses of infliximab in ACT-2. Overall, there were no major differences in efficacy between the 2 doses of infliximab.

There was no difference in the percentage of patients who experienced any adverse events between the placebo and treatment groups in either trial. There were a significantly increased number of infections requiring antimicrobial therapy in the infliximab groups in ACT-1 (P = .01), but not in ACT-2. There was no difference between placebo and infliximab in the occurrence of acute or delayed infusion reactions. Three patients, all infliximab recipients, developed neurologic events, and 1 infliximab recipient developed tuberculosis. Finally, approximately 6% of infliximab recipients in both trials developed antibodies against the drug. Patients with anti-infliximab antibodies were more likely to develop infusion reactions (35.7% in ACT-1 and 50% in ACT-2) than patients without antibodies (9.8% in ACT-1 and 9.7% in ACT-2). However, they were also more likely to have a better (21.4% vs 8.3% in ACT-1) or similar (57.9% vs 57% in ACT-2) clinical response compared with patients with no antibodies.

Comment

Gastroenterologists have waited years for a large well-designed study testing infliximab in UC, and then 2 arrive in 1 paper. These trials are well-designed and reported, and provide both clinical and endoscopic evidence for the efficacy of anti-TNF-α therapy in moderate to severe UC. The participants represent the type of patients likely to be selected for additional therapy in clinical practice, and who might otherwise remain steroid dependent, or be advised to consider surgery. The data demonstrate that approximately 65% of patients treated with infliximab will have a clinical response, and approximately 35% of patients will go into remission, which persists for up to 1 year. However, although there were significantly more steroid-free remissions in the infliximab group compared with placebo, disappointingly only 22% of patients on steroids at initiation of therapy managed to maintain remission and discontinue steroids at 1 year; in other words, 4 out of 5 steroid-dependent UC patients remained on steroids 1 year after starting infliximab therapy. Whether this reflects poor steroid-sparing efficacy of infliximab or the relatively high numbers of UC patients who went straight from aminosalicylate and steroid therapy to infliximab without commencing azathioprine or 6-mercaptopurine is unclear.

One major concern raised by the safety data is the higher number of serious infections, lupus-like reactions, and neurologic diseases among patients treated with infliximab compared with placebo. There were 8 cases of pneumonia among the infliximab groups compared to none in the placebo group, 3 of which were deemed serious, including 1 case of tuberculosis and a fatal cases of histoplasmosis. In addition, 3 patients, all of whom received infliximab, had neurologic events, including 2 cases of optic neuritis and a multifocal neuropathy, although it is unclear from the discussion as to how they fared. Whereas 1 patient who received placebo developed basal cell carcinoma, 4 patients receiving infliximab developed carcinoma or dysplasia, including prostatic adenocarcinoma, rectal adenocarcinoma, basal cell carcinoma, and colonic dysplasia. The significance of these findings is unclear, given the small numbers affected. Multivariate analysis was not performed to determine the factors associated with these adverse outcomes, in particular whether other immunosuppressants played a role in these infections (Clin Gastroenterol Hepatol. 2006;4:621–630). Despite concerns over lymphoma risk, a recent matched pair study in Crohn’s disease reported no increased frequency of cancer in infliximab recipients when compared with nonrecipients (Gut 2006;55:228–233).

Leaving safety concerns aside for the moment, how do these results compare with other therapies for moderate to severe UC? Direct comparisons are difficult; the patients in this trial had active disease not sufficient to require intravenous steroids, and only half were on azathioprine or 6-mercaptopurine. Cyclosporine has been tested in several trials for severe UC, although mostly in patients who have had no response to intravenous steroids. The data from 2 randomized controlled trials demonstrated an early response rate of 60%–82%, and a colectomy rate of 27%–33% when used in steroid-refractory patients (Gastroenterology 2001;120:1323–1329; N Engl J Med 1994;330:1841–1845). Follow-up studies have reported a 3-year colectomy rate of 45% in both the academic and community setting (Aliment Pharmacol Ther 1998;12:973–978; Am J Gastroenterol 1995;90:2093–2096). A small study examining quality-of-life in patients treated with cyclosporine concluded that this therapy was associated with improved quality-of-life outcomes when compared with patients who had undergone a colectomy (Inflamm Bowel Dis 1999;5:1–10). In contrast, cyclosporine therapy has been associated with mortality rates of up to 3.5% (Inflamm Bowel Dis 2004;10:73–78), mostly owing to opportunistic infections.

Of note, both ACT trials excluded patients with fulminant UC, and although the recent “rescue” trial by Jarnerot et al. suggested a beneficial role for infliximab in the setting of fulminant UC refractory to intravenous steroids. Whether gastroenterologists choose infliximab, cyclosporine, or colectomy in that setting remains largely based on their individual and institutional comfort levels, expertise, and safety experiences with any of these 3 therapeutic options. Ultimately, a head-to-head randomized controlled trial is required to compare the merits of intravenous cyclosporine versus infliximab in the setting of severe or fulminant UC refractory to intravenous steroids.

Another potential option for steroid-resistant/-dependent UC is methotrexate, which was associated with remission rates in excess of 60% in open-label trials of steroid-dependent patients, including those who had failed azathioprine (Aliment Pharmacol Ther 2002;16:1751–1759; Eur J Gastroenterol Hepatol 2000;12:1227–1233; Ann Intern Med 1989;110:353–356). However, the only double-blind, randomized controlled trial to date used a low weekly methotrexate dose of 12.5 mg, compared with 25 mg, and reported no difference in remission rates compared to placebo (Gastroenterology 1996;110:1416–1421). Finally, as reviewed by Ghosh et al recently, 6-thioguanines such as azathioprine and 6-mercaptopurine have had conflicting results in reported trials to date (Gut 2006;55:6–8). Two recent trials from Europe have reported remission rates of 53%–77% when compared with 5-aminosalicylates in steroid-dependent patients, with >60% of patients maintaining this remission in follow-up (Eur J Gastroenterol Hepatol 2000;12:1227–1233; Gut 2006;55:47–53).

The results of the ACT studies have led to a rethinking of our understanding of the pathogenesis of UC. Traditionally, UC has been thought of as a Th2-lymphocyte driven condition with a different cytokine profile to Crohn’s disease (Clin Exp Immunol 1998;111:532–535; J Immunol 1996;157:1261–1270). However, both diseases are characterized by defective apoptosis of T lymphocytes, leading to their accumulation in the lamina propria, albeit via different molecular mechanisms (Gastroenterology 1999;116:557–565; Gut 2004;53:1624–1631). Infliximab targets such lingering T cells by inducing apoptosis in TNF-α–expressing lymphocytes (Gastroenterology 2003;124:1774–1785). This observation, rather than binding to TNF-α per se, probably explains why infliximab reduces inflammation in UC. It is interesting to note that sulphasalazine, methotrexate and azathioprine/6-mercaptopurine can also induce T-cell apoptosis (Gut 2004;53:1556–1558).

In conclusion, the results in the ACT studies suggest that infliximab has equivalent efficacy in moderate-to-severe UC refractory or dependent on oral steroids as it does in Crohn’s disease. Of note, fewer than half the patients in these trials were receiving azathioprine or 6-mercaptopurine and subsequent analysis revealed that being on immunomodulators had no benefit in terms of clinical response to infliximab. This has led some to recommend infliximab earlier in the therapeutic pyramid for UC. However, given the relatively disappointing ability of infliximab to achieve steroid-free remission among steroid-dependent UC patients, as well as concerns for increased serious infections, one of which was fatal, as well as neurologic and lupus-like complications compared with placebo, one has to question whether it is wise to consider infliximab ahead of azathioprine/6-mercaptopurine in UC patients refractory to optimal doses of aminosalicylates or who are steroid dependent. Although the ACT trials suggest that infliximab may have a role as an equally rapid alternative to cyclosporine in the setting of severe UC, large randomized controlled trials are required to compare both these options in severe or fulminant UC refractory to intravenous steroids. Although the colectomy-sparing effect of such therapies, either alone or possibly in combination, may hopefully provide patients with a sustainable alternative to surgery, the cost, however, is lifetime surveillance of a colon at potentially higher risk of cancer.