The first two authors contributed equally to the study.
Infliximab or cyclosporine as rescue therapy in hospitalized patients with steroid-refractory ulcerative colitis: A retrospective observational study
Article first published online: 15 MAR 2011
Copyright © 2011 Crohn's & Colitis Foundation of America, Inc.
Inflammatory Bowel Diseases
Volume 18, Issue 2, pages 212–218, February 2012
How to Cite
Sjöberg, M., Walch, A., Meshkat, M., Gustavsson, A., Järnerot, G., Vogelsang, H., Hertervig, E., Novacek, G., Friis-Liby, I., Blomquist, L., Angelberger, S., Karlen, P., Grännö, C., Vilien, M., Ström, M., Verbaan, H., Hellström, P. M., Dejaco, C., Magnuson, A., Halfvarson, J., Reinisch, W. and Tysk, C. (2012), Infliximab or cyclosporine as rescue therapy in hospitalized patients with steroid-refractory ulcerative colitis: A retrospective observational study. Inflamm Bowel Dis, 18: 212–218. doi: 10.1002/ibd.21680
- Issue published online: 10 JAN 2012
- Article first published online: 15 MAR 2011
- Manuscript Accepted: 17 JAN 2011
- Manuscript Received: 14 JAN 2011
- ulcerative colitis;
- rescue therapy;
Cyclosporine (CsA) or infliximab (IFX) are used as rescue therapies in steroid-refractory, severe attacks of ulcerative colitis (UC). There are no data comparing the efficacy of these two alternatives.
Outcome of rescue therapy was retrospectively studied in two cohorts of patients hospitalized due to steroid-refractory moderate to severe UC: 1) a Swedish-Danish cohort (n = 49) treated with a single infusion of IFX; 2) an Austrian cohort (n = 43) treated with intravenous CsA. After successful rescue therapy, maintenance immunomodulator treatment was given to 27/33 (82%) of IFX patients and to 31/40 (78%) of CsA patients. Endpoints were colectomy-free survival at 3 and 12 months. Kaplan–Meier and Cox regression models were used to evaluate the association between treatment groups and colectomy.
At 15 days, colectomy-free survival in the IFX cohort was 36/49 (73%) versus 41/43 (95%) in the CsA cohort (P = 0.005), at 3 months 33/49 (67%) versus 40/43 (93%) (P = 0.002), and at 12 months 28/49 (57%) versus 33/43 (77%) (P = 0.034). After adjusting for potential confounding factors, Cox regression analysis yielded adjusted hazard ratios for risk of colectomy in IFX-treated patients of 11.2 (95% confidence interval [CI] 2.4–53.1, P = 0.002) at 3 months and of 3.0 (95% CI 1.1–8.2, P = 0.030) at 12 months in comparison with CsA-treated patients. There were no opportunistic infections or mortality.
Colectomy frequencies were significantly lower after rescue therapy with CsA than with a single infusion of IFX both at 3 and 12 months' follow-up. The superiority of CsA was seen principally during the first 15 days. (Inflamm Bowel Dis 2011;)
Approximately 15% of patients with ulcerative colitis (UC) will have a severe attack during the course of their disease. Corticosteroids remain a mainstay in the management but about 30% of the patients are unresponsive and will require colectomy short-term.1, 2 Long-term follow-up studies report even higher colectomy frequencies.3 Therefore, there has been a search for improved medical treatment.
The calcineurin inhibitor cyclosporine (CsA) was the first successful rescue therapy introduced in the 1990s. In the first randomized, controlled study, 9/11 patients responded to CsA, whereas none to placebo.4 Subsequent controlled trials5, 6 and uncontrolled series7–9 showed short-term response rates of about 65%–85%.10–12 However, the long-term efficacy has been questioned and retrospective analyses have shown increasing colectomy frequencies with time.13, 14
Infliximab (IFX) was introduced in steroid-refractory, severe attacks of UC with initially conflicting results.15, 16 However, in a subsequent randomized, controlled trial, colectomy rates at 3 months were significantly lower after IFX rescue therapy (29%) compared to placebo (67%).17 The superiority of IFX was seen mainly in patients with somewhat less severe UC (colectomy frequencies 0% versus 62.5%) but not in patients with fulminant UC (colectomy frequencies 47% versus 69%). A follow-up study showed that the results after 3 years remained in favor of IFX.18 Following uncontrolled series have confirmed the efficacy of IFX in this group of patients.19–23
There are currently no data comparing the efficacy of these two rescue therapies in steroid-refractory, severe attacks of UC. Therefore, our aim was to report a retrospective observational study comparing CsA and IFX as rescue therapies in two cohorts of patients, hospitalized due to steroid-refractory, acute attacks of UC.
PATIENTS AND METHODS
Two cohorts of patients, hospitalized for a moderate to severe attack of UC and given rescue therapy due to steroid-refractory disease, were studied retrospectively: 1) a Swedish-Danish cohort (n = 49) treated with IFX in the period 1999–2007; 2) an Austrian cohort (n = 43) treated with CsA at Vienna General Hospital between 1993 and 2005, thus, before IFX was used in the hospital for treatment of this group of patients. Twenty-four IFX patients, recruited from nine Swedish and one Danish hospital, have been reported in a randomized, placebo-controlled trial of IFX.17 The remaining 25 IFX patients were hospitalized and treated at Örebro University Hospital.
Both gastroenterologists and colorectal surgeons were involved in the care of patients with a severe attack of UC, and the optimal timing for colectomy in unresponsive patients was jointly assessed on a regular basis. Healthcare, including costs of drugs during hospitalization, is in each country financed by public funding.
Inclusion and Exclusion Criteria
Patients were eligible for the study according to the following criteria: 1) age >16 years; 2) hospitalization due to a moderate to severe attack of left-sided or extensive UC; 3) a disease course, during hospital stay, unresponsive to at least 3 days of intravenous or oral corticosteroid therapy of ≥40 mg prednisolone or other equipotent corticosteroid; 4) a minimum follow-up of 12 months.
Exclusion criteria were previous exposure to any of the study drugs, exposure to the comparative drug within 3 months after rescue therapy, chronic steroid-dependent UC defined as active UC treated with ≥10 mg prednisolone for >3 months or flare within 3 months after steroid withdrawal, and change of diagnosis to Crohn's disease during follow-up.
Endpoints were colectomy-free survival at 3 and 12 months and safety of treatment.
Diagnostic and Clinical Criteria
The diagnosis of UC was based on established clinical, endoscopic, and histopathological criteria. Extent of disease was assessed by colonoscopy and classified as left-sided or extensive, and severity of UC by modified Truelove–Witts' criteria at admission to hospital.24 Steroid-refractory disease was determined according to the “fulminant colitis index” or the Seo-index and/or on clinical grounds.25, 26 Decision for colectomy was made on clinical assessment in collaboration between gastroenterologists and colorectal surgeons.
Descriptive data are presented as median (range) or proportions. For categorical variables χ2 test or Fisher's exact test, when appropriate, was used to test differences between treatment groups. For comparison of continuous variables the Mann–Whitney U-test was used. Time to colectomy was illustrated with a Kaplan–Meier plot, and differences between groups were tested with a log-rank test. Cox regression models were used to evaluate the influence of potential confounding factors on the association between treatment groups and time to colectomy. We evaluated proportionality assumptions in the models using Schoenfeld residuals. If an association demonstrated nonproportional hazards, internal stratification by the relevant independent variables was performed. Hazard ratios (HR) were used as the relative measure of association with 95% confidence intervals (95% CI). All statistical analyses were performed with STATA software v. 11 (Tulsa, OK).
The study was approved by local Ethics Committees.
The IFX cohort included 49 patients (30 male) with a median age of 38 (17–60) years and the CsA cohort 43 patients (21 male) with a median age of 32 (17–72) years. Demographic and clinical characteristics of the two cohorts are shown in Table 1. Levels of hemoglobin were significantly lower in the CsA cohort and there were more patients with a first attack of UC in the IFX cohort. Although not reaching statistical significance, levels of C-reactive protein (CRP) were higher among CsA-treated patients and there were more patients with extensive disease in the IFX cohort.
|Cyclosporine (n = 43)||Infliximab (n = 49)||P-valuea|
|Gender, male (%)||49 %||61 %||0.233|
|Age at diagnosis, median (min-max)||26 (13–64)||28 (15–59)||0.486|
|Maintenance therapy at admissionb|
|None||8 (19 %)||19 (39 %)||0.102|
|5-ASA||26 (62 %)||21 (43 %)|
|Azathioprine/6-MPc||8 (19 %)||9 (18 %)|
|Maintenance therapy after rescue therapy (Only colectomy-free patients at 3 months)|
|Azathioprine/6-MP||31/40 (78 %)||27/33 (82%)||0.650|
|Age at rescue therapy, median (min-max)||32 (17–72)||38 (17–60)||0.573|
|Years from diagnosis to rescue therapy, median (min-max)||5 (0–29)||3 (0–33)||0.492|
|Never||25 (58 %)||23 (48 %)||0.086|
|Smoker||7 (16 %)||3 (6 %)|
|Exsmoker||11 (26 %)||22 (46 %)|
|Type of UC|
|First attack of UC||4 (9 %)||13 (26 %)||0.034|
|Relapse of established UC||39 (91 %)||36 (74 %)|
|Extent of UC|
|Left-sided||13 (30 %)||7 (14 %)||0.064|
|Extensive||30 (70 %)||42 (86 %)|
|CRP at rescue therapyb|
|<30 mg/L||15 (35 %)||26 (54 %)||0.065|
|≥30 mg/L||28 (65 %)||22 (46 %)|
|Hemoglobin at rescue therapyb|
|<10.5 g/dL||23 (54 %)||9 (19 %)||0.001|
|≥10.5 g/dL||20 (46 %)||39 (81 %)|
|Severity of UC at admission according to Truelove-Witts' classification|
|Moderate||8 (19 %)||12 (24 %)||0.495|
|Severe||35 (81 %)||37 (76 %)|
|Fulminant colitis index at day 3–4d|
|median (min-max)||17 (6–49)||10 (3–41)||<0.001|
In the IFX group, all patients received intravenous corticosteroids corresponding to a median prednisolone dose of 64 (48–96) mg/day. In the CsA-group, prednisolone was given intravenously (n = 21) at a daily dose of 75 (50–100) mg or orally (n = 22) at a dose of 50 (40–75) mg. In both cohorts, subsequent oral prednisolone treatment was tapered and withdrawn during 6–8 weeks.
Rescue therapy with IFX consisted of a single infusion IFX 5 mg/kg. During follow-up, two patients received one more infusion of IFX 9 months after the first dose. One of them had a colectomy 1 month later and the other one 1.5 years later.
Cyclosporine was given as a continuous intravenous infusion of 4 (2–5) mg/kg for 7 (2–14) days followed by oral CsA 4 (0–14) mg/kg for 18 (1–52) weeks. The dose of CsA during the oral treatment phase was adjusted to reach serum trough levels of 200–250 ng/mL. Three patients did not receive oral CsA; one had a colectomy 3 days after rescue therapy; one had an anaphylactic reaction to intravenous CsA, and one did not respond to intravenous CsA. The two latter patients, however, escaped surgery. One patient received another rescue therapy course with CsA 6 months after the first treatment and had a colectomy 10 weeks later.
At the time of rescue therapy, 27 patients (CsA; n = 13, IFX; n = 14) had current or previous treatment with azathioprine/6-mercaptopurine (Table 1). In patients with a colectomy-free survival at 3 months, azathioprine was prescribed to 41 thiopurine-naïve patients (CsA; n = 22, IFX; n = 19), and was continued or resumed in 17 other patients (CsA; n = 9, IFX; n = 8), which gives thiopurine treatment in a total of 31/40 (78%) CsA-treated and 27/33 (82%) IFX-treated patients. All other patients were given maintenance aminosalicylate treatment except for one patient who had 18 weeks of oral CsA.
Colectomy at 3 and 12 Months
Colectomy frequencies were significantly lower both at 3 and 12 months' follow-up in patients given CsA compared to those given IFX. At 3 months, 40/43 (93%) of CsA-treated patients had a colectomy-free survival compared to 33/49 (67%) given IFX (P = 0.002). At 12 months, 33/43 (77%) in the CsA group escaped colectomy compared to 28/49 (57%) in the IFX group (P = 0.034) (Fig. 1). The superiority of CsA treatment was seen principally during the first 15 days (Fig. 1). During this early period, 2/43 (5%) CsA patients versus 13/49 (27%) IFX patients were operated on (P = 0.005). In the period from day 16 to 12 months, colectomy frequencies were similar; 8/41 (20%) for CsA- versus 8/36 (22%) of IFX-treated cases. Of all colectomies performed, 2/10 (20%) were carried out during the first 2 weeks in the CsA cohort compared to 13/21 (62%) in the IFX cohort.
By means of Cox regression analysis, unadjusted HRs for risk of colectomy at 3 months in IFX-treated patients compared to CsA-treated patients was 5.4 (95% CI 1.6–18.7), P = 0.007 and at 12 months 2.2 (95% CI 1.0–4.7), P = 0.039 (Tables 2, 3). The difference between the two treatment groups remained after adjusting for potential confounding factors as listed in Table 1, with adjusted HRs of 11.2 (95% CI 2.4–53.1), P = 0.002, and 3.0 (95% CI 1.1–8.2), P = 0.030, respectively (Tables 2, 3).
|Colectomy 3 Months (%)||Unadjusted||Adjusteda|
|HR (95 % CI)||P-value||HR (95 % CI)||P-value|
|Cyclosporine (n = 43)||7||1.0||1.0|
|Infliximab (n = 49)||33||5.4 (1.6–18.7)||0.007||11.2 (2.4–53.1)||0.002|
|Colectomy 12 Months (%)||Unadjusted||Adjusteda|
|HR (95 % CI)||P-value||HR (95 % CI)||P-value|
|Cyclosporine (n = 43)||23||1.0||1.0|
|Infliximab (n = 49)||43||2.2 (1.0–4.7)||0.039||3.0 (1.1–8.2)||0.030|
The mode of corticosteroid administration in the CsA cohort prior to rescue therapy did not influence subsequent colectomy frequencies. At 3 months, 1/21 given intravenous and 2/22 given oral corticosteroids had a colectomy, and at 12 months the corresponding figures were 4/21 versus 6/22.
Compared to patients with ongoing or past thiopurine therapy, thiopurine-naïve patients treated with azathioprine had a lower colectomy frequency both at 3 months (11% versus 30%, P = 0.06) and at 12 months (22% versus 44%, P = 0.04). This was seen in both cohorts and irrespective of given rescue therapy (data not shown).
There were no opportunistic infections or mortality in either group. Adverse events such as muscle cramps, tremor, paresthesia, or edema were reported by 15 patients treated with CsA. Mild nephrotoxicity occurred in one patient, and another patient had an anaphylactic reaction to intravenous CsA and was well after accurate treatment. Fourteen patients treated with IFX reported mild, nonspecific side effects such as arthralgia, exanthema, depression, perspiration, or itching. Postoperative small bowel obstruction, not requiring a reoperation, occurred in one IFX-treated patient 14 weeks after colectomy. Septicemia and pneumothorax related to central venous lines were seen in two patients, who recovered after appropriate treatment.
There are currently no controlled data comparing the efficacy of CsA and IFX as rescue therapy in steroid-refractory, severe attacks of UC. In this observational study, colectomy frequencies both at 3 and 12 months were significantly lower in patients treated with CsA than those given IFX, and the difference remained after adjusting for potential confounding factors. The superiority of CsA was seen exclusively during the first 15 days, whereas colectomy frequencies were similar in the two cohorts in the subsequent period up to 12 months (Fig. 1).
The present results are consistent with previous reports of CsA. Three randomized, controlled trials have demonstrated efficacy of CsA in severe UC. Lichtiger et al4 reported a short-term response of CsA in 9/11 (82%) patients, whereas 0/9 patients responded to placebo. Five of the latter were switched to open-label CsA and all responded. In order to reduce the risk of side effects, two doses of CsA, 2 mg/kg and 4 mg/kg, were compared in 73 patients. A response of about 85% was found in both groups and colectomy frequencies at 14 days were 9% and 13%, respectively.6 In another trial comparing CsA 4 mg/kg with methylprednisone 40 mg/day in 30 patients, 64% of patients who received CsA and 53% who received corticosteroids had a response to therapy after 8 days.5 A review of uncontrolled series confirmed short-term responses in 67%–86% of patients.10
Predictors of colectomy in patients treated with CsA for a severe attack of UC were retrospectively studied by Cacheux et al.8 A combination of clinical (body temperature >37.5, heart rate >90 bpm), biological (CRP >45 mg/L), and endoscopic criteria allowed the classification of the patients into two different groups with 80% versus 20% risk of colectomy at 6 months.
However, data on long-term prognosis and colectomy risk in patients treated with CsA are divergent. Cacheux et al8 reported that 56% of 135 patients underwent colectomy after 3 years, whereas others reported a colectomy frequency after 5.5 years of 42%.9 After 7 years, colectomy frequencies of 58%13, 65%7, and 88%14 were reported. In contrast, Lichtiger11 reported in a review a colectomy-free survival of 49% of 156 patients after a mean follow-up of 12.6 years.
Seven randomized, controlled studies have assessed IFX in UC.15–17, 27–29 In five studies IFX was compared to placebo15–17, 29 and in two other studies to corticosteroids.27, 28 In a small study by Sands et al,16 4/8 (50%) patients with steroid-refractory UC were considered treatment success 2 weeks after a single infusion of IFX compared to 0/3 patients receiving placebo. In the study by Probert et al,15 43 patients with UC, unresponsive to corticosteroids and not in need of urgent colectomy, were randomized to receive either IFX 5 mg/kg or placebo at weeks 0 and 2. No difference in clinical remission frequency was seen at 6 weeks (39% versus 30%). In a Swedish-Danish trial of patients with a steroid-refractory acute attack of UC, 45 patients were randomized to receive a single infusion of IFX 5 mg/kg or placebo.17 At 3 months, 71% given IFX avoided colectomy compared to 33% given placebo. Other series report colectomy-free survival of 67%–85% during the first few months.20–22
The long-term prognosis and colectomy risk after rescue therapy with IFX is not well described. Colectomy-free survival in the Swedish-Danish trial after 3 years was 50% in IFX-treated patients and 24% in those given placebo (P = 0.012).18 The Active Ulcerative Colitis Trial (ACT)-1 and ACT-2 trials29 included patients with chronic active UC.30 Patients were randomized to receive IFX infusions, 5 or 10 mg/kg, at weeks 0, 2, and 6 and then every 8 weeks or placebo. The cumulative colectomy frequencies through 54 weeks were low and reached 10% for patients given IFX and 17% for those given placebo, which underscores that these patients had less severe UC than patients in the present study.30 A number of retrospective series have since been reported. In the largest series, Kohn et al21 reported colectomy-free survival in 58/83 (70%) patients treated with IFX for steroid-refractory, severe UC during a median follow-up period of 23 months. In a Scottish study, 13/39 patients underwent early colectomy after IFX rescue therapy, and another two patients were operated on later during a median follow-up of 203 days reaching a total colectomy frequency of 38%.22 Others reported that out of a subgroup of 14 patients, treated as in-patients with IFX owing to severe, steroid-refractory UC, eight (57%) patients had a colectomy during a follow-up of 274 days.20
A majority of our patients, unresponsive to a single infusion of IFX, underwent early colectomy during the first 15 days. The difference in treatment outcome, both at 3 and 12 months, derives entirely from this period. It seems as if CsA has a more rapid onset of action compared to IFX. However, the optimal dose and number of IFX infusions in this setting is not well studied. Apparently, a three-dose induction therapy at weeks 0, 2, 6 would not have given a better result, as most patients unresponsive to IFX were operated on early and prior to the time of a scheduled second infusion at week 2. Whether a higher dose of IFX would have given a better outcome can only be speculated on. Achieving adequate trough levels of IFX is important, as was recently shown in a maintenance study of IFX in UC.31 A detectable trough serum IFX was found to be an important prognostic factor of treatment outcome and predicted clinical remission, endoscopic improvement and remission, and a lower risk for colectomy.
The importance of optimal maintenance therapy after successful rescue therapy is emphasized by studies showing lower colectomy frequencies in patients who received thiopurines after CsA.7, 9, 14 The use of long-term immunomodulator therapy was equal in the two cohorts studied herein, which may explain that colectomy frequencies beyond 3 months were comparable. Consistent with the findings by Moskovitz et al,14 thiopurine-naïve patients in our series, who commenced azathioprine treatment, had lower colectomy frequencies both at 3 months and at 12 months compared to patients with prior thiopurine therapy. The difference was statistically significant at 12 months and with borderline significance at 3 months. This finding was seen in both cohorts and irrespective of rescue therapy given. Whether scheduled maintenance treatment with IFX, as shown in the ACT-1 and ACT-2 trials, would have led to a better outcome can only be speculated on. Maintenance IFX therapy has not been studied in this setting, and has furthermore not strictly been compared to immunomodulator therapy.
The limitation of our study is the retrospective, uncontrolled study design using two different cohorts that may be heterogeneous. The patients were comparable regarding demographic characteristics at baseline (Table 1). Age at diagnosis, duration of disease, gender distribution, smoking habits, use of immunomodulators prior to hospitalization, and distribution of severity of UC according to Truelove–Witt's index did not differ between the two groups. The use of maintenance immunomodulator therapy after successful rescue therapy was similar in the two groups as well. However, the two groups differed with respect to disease characteristics such as CRP levels, occurrence of anemia, fulminant colitis index, extent of disease, and distribution of first attack or relapsing UC (Table 1). The CsA cohort had more severe disease activity as judged on higher CRP value, lower hemoglobin concentration, and higher fulminant colitis index after 3–4 days of steroid therapy, indicating a more steroid-refractory course. On the contrary, there were more patients with a first attack of UC and extensive disease in the IFX group compared to those given CsA. However, after taking these variables into account in multivariate Cox regression analysis, adjusted HRs for colectomy after IFX rescue therapy remained significantly higher both at 3 and 12 months compared to CsA treatment (Tables 2, 3). Because of the limited number of subjects compared to number of variables in the statistical models, we also performed sensitivity analyses by excluding variables by stepwise Cox regression. The results at 3 and 12 months were robust and the analyses still showed statistically significant differences between the cohorts. Hence, in spite of higher clinical disease activity and inferior response to initial steroid therapy in the CsA cohort, rescue therapy with CsA was superior to IFX.
The organization of care of patients with a severe attack of UC is quite similar in the participating hospitals, including a close collaboration between gastroenterologists and colorectal surgeons. However, we cannot rule out that the indication and decision level for colectomy may vary between the different hospitals and individual physicians.
Both CsA and IFX may be associated with serious side effects.32 In this study, there was no mortality, opportunistic infections, or other severe adverse events except for mild nephrotoxicity and one anaphylactic reaction to CsA. The study, however, was not designed for detecting differences in this respect.
In summary, this retrospective, observational study demonstrated significantly lower colectomy frequencies both at 3 and 12 months after rescue therapy with CsA compared to IFX. The advantage of CsA was seen principally during the first 15 days, and the risk of colectomy thereafter was similar in the two cohorts. This finding requires further study and the results of a randomized, controlled trial are awaited. Furthermore, the dosing of IFX as rescue therapy and the optimal subsequent maintenance therapy needs to be defined.
- 3Long-term colectomy rate after intensive intravenous corticosteroid therapy for ulcerative colitis prior to the immunosuppressive treatment era. Am J Gastroenterol. 2007; 102: 2513–2519., , , et al.Direct Link:
- 8Predictive factors of response to cyclosporine in steroid-refractory ulcerative colitis. Am J Gastroenterol. 2008; 103: 637–642., , , et al.Direct Link:
- 9Intravenous cyclosporin in ulcerative colitis: a five-year experience. Am J Gastroenterol. 1999; 94: 1587–1592., , .Direct Link: