Azathioprine without oral ciclosporin in the long-term maintenance of remission induced by intravenous ciclosporin in severe, steroid-refractory ulcerative colitis


Correspondence to: Dr M. A. Gassull, Department of Gastroenterology, Hospital Universitari Germans Trias i Pujol, Carretera del Canyet, s/n, E-08916 Badalona, Catalonia, Spain.


Background : Intravenous ciclosporin is considered to be the only alternative to avoid surgery in severe, steroid-refractory ulcerative colitis. In responders, some authors recommend a switch to oral ciclosporin to act as a ‘bridge’ until the therapeutic action of azathioprine is achieved for maintenance treatment.

Aim : To report the short- and long-term outcome of intravenous ciclosporin-responsive ulcerative colitis patients treated with oral azathioprine without oral ciclosporin.

Methods : The records of all patients treated with intravenous ciclosporin for severe, steroid-refractory ulcerative colitis were reviewed. Responders following treatment with azathioprine but without oral ciclosporin as maintenance therapy were included. Patients with colonic cytomegalovirus infection and/or follow-up of less than 1 year were excluded.

Results : Twenty-seven patients were included. Steroids were discontinued in 24 (89%). The median follow-up was 36 months. Eighteen (75%) patients presented mild or moderate relapses, which were easily managed with salicylates or steroids. Cumulative probabilities of relapse were 42%, 72% and 77% at 1, 3 and 5 years, respectively. Eleven (40.7%) patients underwent elective colectomy. Cumulative probabilities of colectomy were 29%, 35% and 42% at 1, 3 and 5 years, respectively. No opportunistic infections were observed.

Conclusions : Oral azathioprine seems to be enough to maintain long-term remission induced by intravenous ciclosporin in patients with steroid-refractory ulcerative colitis. The ‘bridging step’ with oral ciclosporin may not be necessary in this subset of patients, although a randomized controlled trial is warranted to confirm this hypothesis.


About 15% of patients with ulcerative colitis will present with a severe acute attack requiring hospital admission and treatment with intravenous steroids.1 Unfortunately, however, about 40% of these cases do not respond to this treatment.2 Until the early 1990s, total proctocolectomy was the only therapeutic option in this subset of patients. However, in 1994, Lichtiger et al. reported positive results in the only placebo-controlled trial performed with intravenous ciclosporin in steroid-refractory ulcerative colitis.3 Since then, some uncontrolled series have confirmed that intravenous ciclosporin induces clinical response in up to 80% of these patients, thereby avoiding emergency colectomy,4–7 and even improving the quality of life in comparison with patients undergoing total colectomy.8

No studies have been performed to evaluate the best therapeutic regimen for maintenance therapy in intravenous ciclosporin responders, but some authors recommend a switch to oral ciclosporin as soon as a clinical response has been achieved.9 However, it is also recommended that oral ciclosporin should be discontinued within the first 6 months, because of its potential side-effects (mainly nephrotoxicity, which can become irreversible10, 11), and that oral azathioprine should be introduced as long-term maintenance treatment, while the patient is still on ciclosporin, and steroids are being tapered.9 In addition, when used concomitantly with steroids and azathioprine, opportunistic infections may occur.12 For this reason, the prevention of Pneumocystis carinii infections with trimethoprim–sulfamethoxazole has been proposed whilst patients are on triple immunosuppressant therapy.13 In 1996, we suggested that oral azathioprine alone could be enough to maintain remission induced by intravenous ciclosporin, avoiding the potential adverse effects of prolonged treatment with this drug and decreasing the risk of opportunistic infections.14

In this paper, we report our experience with long-term azathioprine maintenance therapy, without the ‘oral ciclosporin step’, in patients with steroid-refractory ulcerative colitis responsive to intravenous ciclosporin.

Patients and methods

The medical records of all ulcerative colitis patients in our department treated with intravenous ciclosporin because of a severe acute attack unresponsive to at least 7 days of intravenous full steroids (prednisolone, 1 mg/kg body weight) were reviewed. All patients received ciclosporin in the form of intravenous Sandimmun solution (Novartis Farmacéutica S.A., Barcelona, Spain) at a daily dose of 4 mg/kg. Patients received the daily dose divided into two single 4-h perfusions every 12 h. The ciclosporin dose was adjusted to maintain a whole-blood ciclosporin concentration within a target range of 100–200 ng/mL. For this reason, the whole-blood ciclosporin concentration (as measured by polarized fluorescence competitive immunoassay) and renal function (serum creatinine, electrolytes) were determined three times per week, and the blood pressure was carefully monitored during ciclosporin therapy. The ciclosporin dose was reduced if drug levels were greater than 200 ng/mL, or if there was an increase in serum creatinine > 30% above baseline values.

Only those patients achieving clinical response (normalization of bowel movements, decrease in acute phase reactants, such as C-reactive protein, erythrocyte sedimentation rate and platelet count) after 7–14 days on intravenous ciclosporin, and followed for at least 1 year (unless colectomy was required before), are reported. Patients treated with oral ciclosporin and/or not treated with oral azathioprine after achieving a response were excluded. As rectal biopsies to rule out cytomegalovirus infection in this situation are routinely performed in our department, all cases with colonic cytomegalovirus infection during the acute attack were not included.

On the day that intravenous ciclosporin was discontinued, steroid tapering and azathioprine treatment were started simultaneously, to reach a total daily dose of 2–2.5 mg azathioprine/kg body weight. No patient received prophylactic treatment for Pneumocystis carinii pneumonia. Complete blood cell counts, erythrocyte sedimentation rate, liver function tests and clinical evaluation were performed monthly during the first year, and every 3 months thereafter. The azathioprine dose was not diminished unless adverse effects appeared. Likewise, azathioprine was maintained indefinitely unless initial treatment failure (clinical relapse before complete steroid withdrawal), steroid-refractory acute attacks or steroid-dependent disease occurred during follow-up (once steroids had been discontinued for the initial episode).

Data are shown as frequencies, or median and interquartile interval, as required.


Thirty-seven of 48 ulcerative colitis patients treated with intravenous ciclosporin for an acute, severe, steroid-refractory attack achieved a clinical response. Ten were excluded from the study because of concomitant colonic cytomegalovirus infection (n = 3), oral ciclosporin administration after intravenous ciclosporin response (n = 5), maintenance treatment with salicylates (n = 1) and follow-up after intravenous ciclosporin of less than 1 year (n = 1). As shown in Table 1, the patients included were mainly non-smokers (never smoked or former smokers), and most (78%) suffered from extensive ulcerative colitis.

Table 1.  Clinical and epidemiological characteristics of patients * ( n  = 27)
  • *

     All parameters expressed as median (interquartile interval), except for gender, smoking habits and disease extent, which are expressed as frequencies.

 Male/female (%)18 (67)/9 (33)
Smoking habits
 Never/former/current smoker (%)15 (56)/9 (33)/3 (11)
 Left-sided/extensive (%)6 (22)/21 (78)
Age (years)32 (24–45)
Time from ulcerative colitis  diagnosis (months)12 (2–78)
Prednisone dose at azathioprine  start (mg/day)50 (40–60)
Azathioprine daily dose  (mg/kg body weight)2.25 (2.1–2.42)
Follow-up (months)36 (18–55)

In 24 (88.9%) of the 27 patients who clinically responded to intravenous ciclosporin, steroids were discontinued after a median of 90 days (60–120 days) without clinical relapse. In one patient, prednisone was decreased by > 50%, but could not be withdrawn completely; this patient went into remission after a second course of intravenous ciclosporin, but was finally colectomized because of an early flare-up whilst on methotrexate. In the remaining two patients, the prednisone dose could not be decreased at all, and both underwent elective colectomy.

In 24 patients, steroids were discontinued. The median follow-up (end of the study in July 2000 or azathioprine discontinuation) was 36 months (18–55 months). During follow-up, 18 of the 24 patients (75%) suffered from mild to moderate acute attacks (one attack in five patients, two attacks in seven patients, three attacks in four patients and four attacks in two patients). The cumulative probabilities of flare-up were 42%, 54%, 72% and 77% at 1, 2, 3 and 5 years, respectively (Figure 1). Three of these 18 relapsing patients were successfully managed with topical or oral salicylates or topical steroids, whereas in 13 cases systemic steroids were required. The median time until systemic steroid re-introduction in this group was 12 months (5–35 months). In 11 of these patients, prednisone was again withdrawn completely.

Figure 1.

Cumulative probability of ulcerative colitis flare-up after complete steroid withdrawal.

Colectomy was performed in 11 of the initial 27 patients (40.7%): three due to relapse whilst steroids were being tapered off; two due to steroid dependence; one due to steroid refractoriness during the onset of a new attack more than 1 year after steroids had been stopped; and five as elective treatment (two for first relapse and three after more than one relapse). Only in five of these 11 patients was colectomy performed within the first year of follow-up. The clinical characteristics of these patients are shown in Table 2. The cumulative probabilities for colectomy were 29%, 30%, 35% and 42% at 1, 2, 3 and 5 years, respectively (Figure 2).

Table 2.  Clinical characteristics of colectomized patients
PatientNumber of flare-upsSteroid re-introductionTime between beginning of azathioprine and colectomy (months)
  • *

     Steroids could not be tapered off after stopping intravenous ciclosporin.

 1Not withdrawn*1
 2Not withdrawn*4
 3Not withdrawn*4
Figure 2.

Cumulative probability of colectomy during follow-up.

No patient developed opportunistic infections during follow-up. A reduction in azathioprine dose due to toxicity was necessary in five patients (due to neutropenia below 1000 cells/mm3 in two, lymphopenia below 600 cells/mm3 in one and hepatotoxicity in two). Thirteen patients (48%) were maintained on azathioprine treatment, whereas it was discontinued in 14 patients during follow-up. The reasons for azathioprine discontinuation were: early treatment failure (within the first 6 months of treatment) in four patients (15%), severe acute attack > 6 months after starting azathioprine in eight patients (30%), fever of unknown origin (with negative complete study, and resolution after azathioprine discontinuation) in one patient and maintenance of remission for more than 3 years in the remaining patient.


Controlled and uncontrolled data support intravenous ciclosporin as an effective treatment for acute, severe, steroid-refractory ulcerative colitis. The effectiveness of this treatment in our series (75% of patients) compares well with the efficacy rate of 62–86% reported in previous studies.3, 5–7 Although azathioprine/6-mercaptopurine is widely used in clinical practice as long-term maintenance treatment in ulcerative colitis patients responsive to intravenous ciclosporin, to our knowledge, no single study has evaluated which is the best strategy for shifting from intravenous ciclosporin to azathioprine. As treatment with azathioprine may take 3 months for full effectiveness to be achieved, it is reasonable to use oral ciclosporin as a ‘bridging treatment’ until azathioprine becomes operative, as recommended by some authors.9 Arts et al. recently reported, in abstract form, the long-term outcome of 86 ulcerative colitis patients treated with intravenous ciclosporin.15 The response rate was 84%, and almost all responders were subsequently treated with oral ciclosporin (96%) and azathioprine (89%). Three patients (3.5%) died because of opportunistic infections (Pneumocystis carinii and Aspergillus pneumonia) whilst on oral ciclosporin.

Our data show that oral azathioprine seems to be enough to maintain remission induced by intravenous ciclosporin. Almost 90% of our patients were steroid-free and in clinical remission 4 months after intravenous ciclosporin was discontinued, and no patient developed opportunistic infections during follow-up. In a single case, azathioprine had to be withdrawn for adverse effects, whereas the remaining adverse events of treatment were easily managed by a transient or permanent decrease in azathioprine dose.

The long-term outcome of our series seems to be similar to that reported using oral ciclosporin as ‘bridging treatment’ to azathioprine therapy. Although a large proportion (75%) of our patients relapsed whilst on azathioprine treatment, the median time to the first flare-up was 12 months. Some patients were managed with topical treatment or oral salicylates and, when systemic steroids were required, they could be withdrawn in most cases. Indeed, the cumulative probability for a flare-up in our patients after steroids had been completely withdrawn was 42% at 1 year. These data are seldom reported in the literature on the long-term treatment of steroid-refractory ulcerative colitis patients.

The colectomy rates at 1 and 2 years (29% and 35%, respectively) in the present series of steroid-refractory patients are similar to that reported in the above-mentioned study by Arts et al. (25% at 178 ± 141 days).15 Actis et al. also reported the clinical outcome of a series of 19 ulcerative colitis patients with steroid-resistant episodes responding to intravenous ciclosporin and further treated with oral ciclosporin (3–6 months) and azathioprine as maintenance treatment.16 After a mean of 18 months on azathioprine, the colectomy rate was 26%. To our knowledge, this paper reports the longest follow-up in steroid-refractory ulcerative colitis patients responsive to intravenous ciclosporin.

In summary, our results are similar to those reported previously in terms of relapse and colectomy rates. Therefore, oral ciclosporin as ‘bridge’ therapy to the therapeutic effect of azathioprine may not be necessary. Moreover, the potential toxicity of ciclosporin (mainly renal damage) and the occurrence of opportunistic infections due to the combination of three immunosuppressants could be diminished using this regimen. Therefore, in patients with acute, severe, steroid-refractory ulcerative colitis responding to intravenous ciclosporin, azathioprine without oral ciclosporin seems to be a safe and effective therapeutic regimen, although this hypothesis must be confirmed by a well-designed randomized controlled trial.