Prof. Dr Eugen Musch, Department of Internal Medicine, Marienhospital Bottrop, Josef-Albers-Straße 70, D-46236 Bottrop, Germany. E-mail: email@example.com
The imbalance of pro- and anti-inflammatory cytokines plays an important role in the pathogenesis of inflammatory bowel disease. Shifting this disturbed ratio by means of TNF-antibodies or interferon has been shown to be helpful in treating Crohn's disease and multiple sclerosis, respectively.
This pilot study investigated whether interferon-β can induce clinical remission in corticoid-refractory ulcerative colitis.
Twenty-five patients with steroid-refractory active ulcerative colitis (Clinical activity index according to Rachmilewitz: 13.5 ± 5.2) were treated in an open pilot trial with 0.5 MIU human natural interferon-β (hn-IFN-β) i.v. (n=18) or 1 MIU recombinant interferon-β-1a (r-IFN-β-1-a) s.c. (n=7) daily with the goal of induction of remission. Subsequent maintenance treatment was carried out for 52.0 ± 78.8 weeks (range 4–336 weeks) with the same dose, three times per week.
Twenty-two of 25 patients (88%) went into remission during induction treatment (hn-IFN-β 16/18, r-IFN-β-1a 6/7). Mean time to response was 3.0 ± 1.3 weeks. Mean length of remission was 13.0 ± 19.7 months. Only eight of 22 patients in remission relapsed during maintenance treatment. Five of these went into remission again after increasing the dose. Adverse events consisted of slight to moderate flu-like symptoms and slight to moderate hair loss in five of 15 female patients.
Although this open pilot study included only a small number of patients, the high response rate suggests that interferon-β may be a safe and effective treatment for steroid-refractory active ulcerative colitis.
Patients with active ulcerative colitis can be brought into remission by treatment with aminosalicylates or steroids in 80–90% of cases.1–3 Some of the remaining 10–20% patients who do not respond to standard treatments respond to immunosuppressive agents such as ciclosporin A,4 azathioprine or 6-mercaptopurine,3 but the side-effects of these drugs, especially of ciclosporin, make their use problematic. However, there are still many patients who do not respond to immunosuppressive treatment or who do not tolerate these treatments, and therefore proctocolectomy is still the outcome in a relatively high proportion of cases—up to 45% after 25 years.4, 5
The aetiology of inflammatory bowel disease is still unknown.6 It has been shown that an imbalance in the immune system plays a major role in the pathophysiology of ulcerative colitis.7, 8 Some of these disturbances, such as a disturbed production of interleukin-1 receptor antagonist (IL-1ra), seem to be genetically determined.9, 10
Correcting the imbalances between pro and anti-inflammatory cytokines has been shown to be of therapeutical benefit in Crohn's disease11 and ulcerative colitis.12
With increasing knowledge about the regulation of inflammatory processes in inflammatory bowel disease, it appears possible to use specific cytokines in the treatment of inflammatory bowel disease patients.
The reported experience of previous studies on interferon-α treatment in a small number of patients with inflammatory bowel disease, however, has been somewhat conflicting. Its use was not uniformly positive in patients with Crohn's disease, but was somewhat promising in a few small series of patients with ulcerative colitis, leaving, however, some uncertainty about the degree of remission.13–20
The decision to use IFN-β for the treatment of ulcerative colitis was based on the following considerations.
First, interferon-β has been used successfully in many experimental and therapeutic trials in patients with multiple sclerosis, which is also believed to be an immune-mediated disorder that occurs in genetically susceptible people.21 Several mechanisms of action of interferon-β in multiple sclerosis, such as the up-regulation of the anti-inflammatory cytokines IL-10 and IL-1ra, and a down-regulation of the pro-inflammatory cytokines TNF and IL-2, and of the co-stimulatory molecules B7-1 and CD40 in lymphocytes, are described22–26 and therefore fit exactly with the hypothesis of cytokine imbalance being one of the major patho-mechanisms in chronic active ulcerative colitis.
With these pathophysiological factors in mind, we initiated an open pilot trial which gave interferon-β to patients who were resistant to current standard therapy, including steroids, a situation which in contrast to previous studies included patients with a very unfavourable, but with regard to pre-therapy, well-defined situation.
Second, we chose interferon-β for this study as it might be considered as the drug of choice before interferon-α in the management of inflammatory bowel disease. According to results of in vitro studies, contrary to interferon-α and -γ, interferon-β does not interfere with the arachidonic and leukotriene B4 metabolisms of the intestinal mucosa.
We made the presupposition that by using interferon-β, the risk of induction of pro-inflammatory chemical mediators in the intestinal mucosa—which then triggers colitis—would be avoidable. The contradictory results which arose from earlier tests with interferon-α in patients with chronic bowel disease could possibly have been caused by the inexact, non-homogenous selection of patients, their pre-therapeutic condition and treatments, and the impossibility of calculating the individual activation processes which promote pro-inflammatory chemical mediators.27, 28
PATIENTS AND METHODS
Twenty-five patients entered the pilot study (15 female, 10 male) with a mean age of 37 years (range 14–72) who all demonstrated severe, active ulcerative colitis proving refractory to basic medication.
Three of the patients were suffering from proctosigmoiditis, seven from left-sided colitis and 15 from pancolitis. Diagnosis to confirm the extent of the colitis was obtained by coloscopy or colon barium enema carried out within the previous 12 months.
Clinical and endoscopic assessment of disease activity was evaluated according to the scoring systems for clinical symptoms and endoscopic findings defined by D. Rachmilewitz29 (see Table 1).
Table 1. . Scoring systems for clinical symptoms and endoscopic findings, Rachmilewitz 29
The clinical activity index (CAI), according to Rachmilewitz, averaged 13.3 ± 5.1 (mean ± s.d.).1 The mean duration of the disease was 9.1 years (range 2–19). The endoscopic activity index according to Rachmilewitz was 9.2 ± 2.3 points for all patients. This score was confirmed by rectosigmoidoscopy prior to the introduction of interferon therapy (see Table 2).
Table 2. . Patients characteristics
All patients were undergoing treatment with mesalazine/sulfasalazine at a median daily dose of 3 g (range 3–4 g), and glucocorticoids of at least 30 mg per day. The median daily 5-ASA and steroid doses in the 4-week period prior to enrolment into the study amounted to: 5-ASA 86 g (range 84–112 g), corticosteroids 1000 mg (range 840–1500 mg).
A treatment-refractory situation was defined according to the following criteria: patients failed to respond (CAI ≤ 9) to standard treatment with 5-aminosalicylic acid at daily doses of 3–4 g p.o. plus the optional additional topical administration of 5-ASA (2–4 g) and corticosteroids at daily doses of ≥ 30 mg for 14 days, mean 35.2 mg (range 30–50 mg).
Criteria for exclusion from the study were women of gestational age without safe contraception, pregnancy or lactation, and patients under immunosuppressive treatment other than glucocorticosteroids. In addition, patients with disease confined to the rectum, disease of bacterial origin, or toxic megacolon were not eligible. Other criteria for the exclusion were malignant disease, patients with psychiatric disorders, alcohol or drug abusers, a blood tranfusion within 14 days before beginning interferon medication, coagulopathy, manifest haematotoxicity, WBC < 2500/μL, thrombocytopenia > 80 000/μL, latent or manifest autoimmune disorders.
The study of the therapeutical effect of interferon-β on patients with corticoid-refractory ulcerative colitis was approved by the ethics committee of the Westfälische Wilhelms-Universität Münster and the General Medical Council Westfalen-Lippe.
All patients were informed about the experimental nature of the trial prior to the start of interferon-β therapy and gave their written informed consent to take part in the study. For patients under 18 years of age the parents gave the written informed consent.
Eighteen patients received human natural interferon-β (hn-IFN-β) (group N) and seven patients received recombinant interferon-β-1a (r-IFN-β-1a) (group R) in addition to their basic treatment. A daily dose level of 0.5 MIU of hn-IFN-β was administered intravenously in the form of a bolus injection. In order to prepare individual doses of 0.5 MIU, a vial of Fiblaferon (Biosyn-Arzneimittel GmbH, Fellbach) was dissolved in 3 mL of the original solution, filled under sterile conditions in six aliquots each comprising 0.5 mL (= 0.5 MIU) in 1 mL syringes and immediately deep-frozen. For injection, the deep-frozen solution was thawed and immediately administered.
r-IFN-β-1a (Bioferon, ASTA-Frankfurt) was available in ampoules, each comprising 1 MIU per mL solution.
r-IFN-β was administered via s.c. injection at a dose of 1 MU in addition to the basic medication.
Besides the admitted human natural interferon-β we also made use of a certain amount of recombinant interferon-β being released for this test to test its potential for use in a less expensive version of this therapy.
Interferon-β was administered daily over an induction period not exceeding 8 weeks. Maintenance therapy was given at the same dose level three times per week. The initial dose level of corticosteroid therapy was reduced by 10 mg of corticosteroid equivalent dose per week to a daily dose of 20 mg, depending on the falls in CAI score, and thereafter by 5 mg of prednisolone equivalent dose per week until a complete withdrawal of glucocorticoids was reached. Neither the dosage nor the method of application of 5-ASA was changed in any of the patients.
Evaluation of therapeutic response
The effect of the interferon-β was assessed exclusively on the basis of clinical activity. The activity index according to Rachmilewitz (CAI) was used at the measurement parameter. This was checked on a weekly basis during the pre-treatment phase with corticosteroid + 5-ASA basic medication and during subsequent therapy with identical basic medication + interferon-β. Control of the endoscopic activity index was not essential in the pilot study. The criterion for clinical remission was a fall in CAI to < 4 points.
Any side-effects arising from the interferon-β therapy were assessed according to the WHO recommendation for the evaluation of active and subjective toxicity.
Age, duration of colitis, clinical and endoscopic pre-treatment activity indices, mean duration of induction-, re-induction-, maintenance therapy, time of relapse as well as activity indices at the time of relapse are presented as mean ± S.E.M.
CAI-indices during the induction period are presented as diagrams of weekly mean CAIs ± S.E.M. for each treatment group (n-IFN-β and r-IFN-β). The Mann–Whitney U-test was used to evaluate comparative statistical significance.
These statistical calculations were performed using the SPSS/PC software package.
Induction of remission by IFN-β
Twenty-five patients were treated with interferon-β (IFN-β). Eighteen of them (group N) received 0.5 MIU human natural interferon-β (hn-IFN-β) i.v. daily and seven (group R) received 1.0 MIU recombinant interferon-β (r-IFN-β-1a) s.c. daily for up to 8 weeks.
Twenty-two of these 25 patients (88%) went into complete remission (CAI ≤ 4) (Table 2). Mean CAI scores decreased in these patients from 13.4 ± 5.1 (mean ± s.d.) to 1.1 ± 1.2.
Regarding the different types of interferon-β, 16/18 (89%) went into remission in group N and 6/7 (86%) went into remission in group R. Mean CAI scores decreased from 12.8 ± 4.7 to 1.0 ± 1.1 in group N and from 14.9 ± 5.7 to 1.2 ± 1.6 in group R.
The time between the start of the IFN-β treatment and complete remission was 3.0 ± 1.3 weeks in all patients. In the subgroups this time was 2.8 ± 1.3 weeks and 3.7 ± 1.2 weeks, respectively. However, a strong decrease in CAI was noted as early as 1 week after initiating treatment in most patients (Figure 1).
Within 2 weeks of obtaining complete remission, all responding patients were taken off the corticosteroids.
Maintenance of remission by IFN-β
After induction of remission, all 22 patients were treated three times a week with the same dose of IFN-β which they had received during the daily induction. For all patients the duration of the maintenance treatment was 52.0 ± 78.8 weeks (range 4–336). In the subgroups the duration of maintenance treatment was 60 ± 90 weeks (range 4–336) in group N and 32 ± 19 weeks (range 4–64) in group R. Relapses occurred after 78 ± 86 weeks (4–288) of follow-up. (Table 3)
Table 3. . Results of therapies with interferon-β
Relapses during maintenance and reinduction of remission by IFN-β
Eight of 22 patients (36%) (group N: 5; group R: 3) relapsed during the maintenance treatment. They did so within 67 ± 87 weeks (range 12–288). (group N: 79.2 ± 104.8 weeks (range 12–292) group R: 48.4 ± 34.4 weeks (range 24–96) during maintenance treatment. They developed mean CAIs of 8.5 ± 0.9 (group N: 8.6 ± 1.0; group R: 8.3 ± 1.2).
Five of these eight patients (63%) were re-treated by increasing their interferon-β dose to daily dosages (Table 3). Four (group N: 3/3; group R: 1/2) of these five were once again brought into remission and a second maintenance phase.
Withdrawal of IFN-β during maintenance treatment
In 10 patients interferon-β was discontinued at the patients request, following complete remission. These patients were subsequently treated with 5-aminosalicylic acid monotherapy. Six of these patients relapsed after 62.0 ± 106.4 weeks (range 2–292). No exacerbation of ulcerative colitis occurred warranting surgery or colectomy throughout the entire study.
Statistical comparisons between relapse time of patients on maintenance therapy (78 ± 86 weeks) and the group of patients not using maintenance therapy with interferon-β (62 ± 106 weeks) was analysed by calculating the mean difference with its 95% confidence interval according to the Mann–Whitney U-test. There was a benefit, of slight statistical significance (P < 0.055), in favour of the group continuing maintenance therapy with interferon-β.
The side-effects observed were the well-known side-effects of interferons. They were of minor significance at the low doses of interferon-β we employed (see Table 4).
Table 4. . Interferon-β in therapy refractory ulcerative colitis. Side-effects
Due to a recurring high temperature of 40 °C, treatment had to be withdrawn after one month from one patient who had experienced a complete response during induction therapy. The colitis recurred in this patient 2 weeks after the withdrawal of hn-IFN-β. Five patients complained of unpleasant hair loss (WHO degree 1–2), which, however, did not lead to a discontinuation of the interferon-β treatment. In two patients interferon treatment triggered slight nausea for up to 2 weeks in the early stages of treatment (WHO-degree 1 affecting five patients receiving hn-IFN-β and two patients receiving r-IFN-β-1a).
One patient experienced a moderate hepatotoxic reaction with an increase in aminotranferase activities AST 32 U/L, ALT 55 U/L. Owing to the low dose of the interferon therapy, no relevant haematotoxicity was observable. None of our patients showed any manifestation of thyroid or other endocrine disorder or any induction of autoimmune complications, even following long-term maintenance therapy with either n-IFN-β or r-IFN-β.
Natural and recombinant interferon-β were both found to be a safe and effective treatments for steroid-refractory active ulcerative colitis. No major difference was found between the two forms of interferon-β regarding efficacy and side-effects.
A response rate of 88% was quite impressive in a group of patients with steroid refractory active ulcerative colitis. Since this is the first study to use interferon-β for the treatment of patients with inflammatory bowel disease, we did not include a placebo group, nor did we do a dose-finding experiment. Therefore, the exact effect of the interferon-β cannot be determined. However, the placebo response is quite low in patients refractory to standard treatment, as was shown recently by Ilnyckyi et al. who quantified the placebo response in 38 placebo-controlled treatment studies of active ulcerative colitis.30 They found that the clinical remission rate was 9.1% (confidence interval: 6.6–11.6). These rates are markedly lower than the remission rate of 88% (CI: 35.9–99.6) observed in our patients with ulcerative colitis after treatment with interferon-β. Since our patients had been pre-treated with other drugs without effect, it can be assumed that our study population had been more severely ill than the patients of most of the placebo-controlled trials for ulcerative colitis analysed by IIinyckyi et al.30 This is supported by the high rate of relapses observed after discontinuation of interferon-β. This suggests that the placebo response of our patients would not be higher than that described in those trials. However, due to the unmasked nature of the trial and the small number of patients, our findings will have to be confirmed by a larger placebo-controlled trial, which is in progress at the moment.
Treatment with interferon-β is quite expensive. It is not therefore possible to treat large numbers of patients with this drug. On the other hand, steroid refractory active disease is fortunately not so common. IFN-β acts much faster (Figure 1) than azathioprine and 6-mercaptopurine which have also been used in the treatment of active steroid-refractory disease. Although clinical evidence is sparse,31 Ciclosporin A is effective in ulcerative colitis where it may manifest its effectiveness quite promptly, has many more side-effects than interferon-β and cannot be tolerated by all patients. Therefore despite its costs, interferon-β may become a valuable drug for patients with active steroid-refractory colitis.
To our knowledge there have only been two other trials which used interferon on patients with ulcerative colitis.32, 33 Both used interferon-α. Sümer et al.32 described that 23 (82%) of their 28 patients responded to therapy with a fast improvement (within 15 days) and were in complete clinical and endoscopic remission after 6 months of therapy, which is quite similar to the remission rate of 88% observed in our study.32 Rüther et al.33 treated only four patients with ulcerative colitis, so that meaningful statistics cannot be calculated.
The mechanism of action of interferon-β is unclear. Since it up-regulates some anti-inflammatory mediators and down-regulates some pro-inflammatory mediators, it may be the anti-inflammatory actions which are responsible for the therapeutic effect.23–27 On the other hand, it is also possible that it acts by inhibiting a still unknown infectious agent like a virus, as suggested by Rüther et al.33
Only a small number of patients were included in this open pilot study, but it may be concluded that interferon-β could be used as a safe and effective alternative to induce and maintain remission in patients with steroid refractory active ulcerative colitis. A random, controlled, dosage-finding trial is currently under way.
This study was supported by a grant from Asta-Medica AG, Bioferon-AG, Dr Rentschler GmbH, and Biosyn-AG.