Oral tacrolimus long-term therapy in patients with Crohn’s disease and steroid resistance


Professor A. Francavilla, Cattedra di Gastroenterologia, Università di Bari, Policlinico V. le Ennio, 70124 Bari, Italy. E-mail: a.francavilla@gastro.uniba.it



To report the results of a prospective, open-label, uncontrolled study in 13 patients affected by Crohn’s disease with resistance to steroids.


The patients were treated long-term with oral tacrolimus, aiming to both resolve acute attacks and maintain remission. Tacrolimus was administered at the dose of 0.1–0.2 mg.day/kg and adjusted in order to achieve levels of 5–10 ng/mL; only mesalazine was continued concomitantly. Steroids and total parenteral nutrition were tapered when appropriate.


Median treatment was 27.3 months. Only one patient dropped out due to adverse events. Crohn’s disease activity index score significantly decreased after 6 months in 11 patients; for 1 year in nine of them, and 7 years in two of them. The inflammatory bowel disease life-quality questionnaire score significantly increased over the same periods. A marked drop in hospitalizations was recorded. In three out of six patients complete closure of fistulas occurred. Tacrolimus allowed total parenteral nutrition to be withdrawn in three out of five patients. Supplementation with low-dose steroids was required in five patients. Two patients underwent surgery.


Tacrolimus therapy appears to be associated with both short- and long-term benefits, and may represent a therapeutic option in Crohn’s disease when conventional therapies fail. This study encourages its use in controlled trials.


The treatment of choice in active Crohn’s disease is based on high doses of corticosteroids. However, this therapy may sometimes be ineffective and resistance is documented.1, 2 Furthermore, the vast majority of steroid-resistant subjects experience severe adverse events.

Because of the slow onset of action, conventional immunosuppressive therapies such as azathioprine are mainly used to maintain a clinical remission.3 New drugs such as ciclosporin and tacrolimus, currently used to prevent allograft rejection after organ transplantation, have shown promising results. These two drugs, although structurally unrelated, have similar cellular effects inhibiting the production by T helper lymphocytes of interleukin 2, other cytokines and tumour necrosis factor α. They do so by blocking the transcription in the early phase of T cell gene activation and suppressing the expression of interleukin 2 and 7 receptors. Tacrolimus differs from ciclosporin in its effects on transforming growth factor β, high levels of which have been implicated in chronic renal allograft rejection.4 These peculiarities have led to the suggestion of these drugs being used in autoimmune disorders such as ankylosing spondylitis and primary sclerosing cholangitis.5[6]–7 Preliminary reports show that ciclosporin and tacrolimus, when administered intravenously for a short period, are effective in the active phase of inflammatory bowel diseases.2, 8[9][10][11]–12

In our Department, the Regional Centre for the selection and follow-up of patients undergoing liver transplantation, we have been using tacrolimus since it was approved for clinical use, and we have successively broadened our experience with the drug to other conditions. We describe here our experience with oral tacrolimus in 13 patients affected by severe Crohn’s disease who had developed resistance to steroids. This is the first time that the drug has been used in the long term to both resolve acute attacks and maintain clinical remission.



Since July 1991, a total of seven male and six female patients with steroid-resistant Crohn’s disease have been enrolled in a prospective, open-label, uncontrolled study. Resistance was clearly stated at the moment of enrolment; all patients showed Crohn’s disease activity index values indicative of an active disease despite steroid therapy being performed for a period of over 6 months.

Diagnosis of Crohn’s disease was established according to standard criteria by clinical assessment, radiology, endoscopy and histology. Pregnancy and renal and hepatic impairment were considered as exclusion criteria. Written informed consent was obtained from all participants. The study was approved by our Institutional Board Review, i.e. the Ethics Committee of the ‘Policlinico’ University Hospital in Bari, Italy.

Clinical features are summarized in Table 1. All patients were treated with tacrolimus and 5-aminosalicylates (5-ASA), whilst other therapies were stopped, with the exception of total parenteral nutrition in five subjects (patients 3, 8, 9, 11 and 12). Only at a later stage were steroids and other drugs gradually brought in when necessary, i.e. when tacrolimus and 5-ASA alone were unable to maintain clinical remission.

Table 1.  Clinical characteristics of 13 patients with Crohn’s disease treated with tacrolimus Thumbnail image of

Tacrolimus treatment modalities

Tacrolimus capsules (1 mg or 5 mg: Prograf, Fujisawa, USA, Inc., Deerfield, IL) were initially given at a dose of 0.1–0.2 mg.day/kg, aiming for a serum trough level of 5–10 ng/mL. Blood assays of the drug were determined by competitive binding microparticle enzyme immunoassay (Abbott IM) twice a week for 2 weeks, weekly for a further 2 weeks and then once a month for the remaining period of therapy.13 Patients were monitored for adverse events such as nephrotoxicity, hypertension, infections, headache, tremor, paresthesias, sore muscles, insomnia, diarrhoea, nausea, vomiting, hyperkalemia, hyperglycaemia and alopecia.

Clinical follow-up

Subjective clinical conditions were assessed using the inflammatory bowel disease life-quality questionnaire (IBDQ) at entry and then every 6 months thereafter.14 Each patient’s Crohn’s disease activity index (CDAI) was also assessed at the same time. Concomitant values of IBDQ > 180 and CDAI < 150 were considered as numerical parameters of remission. A clinically meaningful improvement in the Crohn’s disease activity index and inflammatory bowel disease life-quality questionnaire scores was considered to be a reduction of at least 50% in the basal value. The number of hospital admissions 12 months before therapy and in the first year of treatment was used as a further parameter. Additional investigations (i.e. laboratory, endoscopic and radiological examinations) were performed when appropriate. The effect of this therapy on fistulas was evaluated by physical examination associated with ultrasound scan (when external) and serial small bowel X-ray (when internal). Serum creatinine levels were monitored monthly in order to assess renal function.


One-way analysis of variance was employed to compare the baseline values of CDAI, IBDQ, erythrocytic sedimentation ratio (ESR), C reactive protein (CRP), haemoglobin and platelet count, before treatment and at the 6th and 12th month of therapy. Where this evaluation rejected the hypothesis of mean equality between the groups of values, a further evaluation was performed using the Student–Neuman–Keuls test. The criterion for this comparison was a significance level < 0.05.


The median period of treatment with tacrolimus was 27.3 months (range 2–93 months).


After the introduction of tacrolimus, all patients with steroid or ciclosporin adverse events experienced their complete reversal. Tacrolimus side-effects were observed in six out of 13 patients (Table 2). Only one (patient 4) dropped out after 2 months of therapy because of severe tremor. He later underwent surgery for intestinal occlusion. No significant change was observed in serum creatinine during tacrolimus therapy.

Table 2.  Tacrolimus treatment details and clinical outcome Thumbnail image of

Crohn’s disease activity index and inflammatory bowel disease life-quality questionnaire changes

Figure 1(A) reports Crohn’s disease activity index values before the treatment and at the 6th (11 patients) and 12th (nine patients) month of therapy. Statistical analysis demonstrated a significant decrease (P < 0.001) as follows: baseline > 6th month=1 year. Moreover, in a 7-year follow-up of patients 6 and 7, Crohn’s disease activity index values remained unchanged after a significant fall at the 6th month. In the same graph (Figure 1B), inflammatory bowel disease life-quality questionnaire values before treatment and at the 6th (11 patients) and 12th (nine patients) months of therapy are shown. As in the case of Crohn’s disease activity index, statistical analysis demonstrated a significant increase (P < 0.001) as follows: baseline < 6th month=1 year.

Figure 1.

(A) Crohn’s disease activity index (CDAI) before tacrolimus (13 patients), after 6 months (11 patients) and after 1 year (9 patients). Values are expressed as mean ± s.d. One-way analysis of variance shows P < 0.001. The Student–Neumann–Keuls test indicates: baseline > 6th month=12th month. (B) Inflammatory bowel disease questionnaire (IBDQ) before tacrolimus (13 patients), after 6 months (11 patients) and after 1 year (nine patients). Score values are expressed as mean ± s.d. One-way analysis of variance shows P < 0.001. Student–Neumann–Keuls test indicates: baseline < 6th month=12th month.

Hospital admissions and laboratory tests

In the year before tacrolimus therapy, all patients underwent at least one hospital admission (range 1–4). In the first year of tacrolimus, six out of 13 patients did not require any hospitalization. Moreover, the total (eight vs. 22) and median (0.36 vs. 1.69) numbers of hospital admissions were markedly decreased by tacrolimus therapy. Significant changes were observed in the values (mean ± S.D) of ESR (P < 0.01), CRP (P < 0.01) and haemoglobin (P < 0.05), when basal values were compared with the results of both the 6th and 12th months. Moreover, a reduction in platelet count was seen after tacrolimus without significance at statistical analysis.

Clinical outcome

The disease was complicated by fistulas in six patients. In three (50%), complete healing was obtained by tacrolimus therapy (patient 2: caecum; patient 9: perineum; patient 11: enterocolic). Moreover, patient 8 experienced a dramatic improvement after 2 months of therapy, with a 10-fold decrease in the diameter of a huge perineal fistula, whilst in patients 5 and 7, no changes in fistula drainage were seen (Table 2).

Five patients needed total parenteral nutrition at enrolment. Nutritional support was withdrawn in three out of five patients (numbers 3, 8 and 9) during tacrolimus therapy.

Along with tacrolimus, five patients (numbers 5, 6, 7, 11 and 12) required steroid supplementation because of clinical relapses arising within 6 months. Clinical remission with Crohn’s disease activity index normalization was obtained in only one out of five subjects (patient 7). In three patients, the Crohn’s disease activity index decreased without reaching remission values. Only in one (patient 5), was no change seen. Two patients in this sub-group (patients 5 and 6) needed surgery. The first underwent resection of the terminal ileum and right colon following intestinal occlusion due to ileal stenosis. Patient 6 achieved remission 30 days after starting tacrolimus, with a significant improvement in dermatological lesions and articular symptoms. However, after 4 months, he relapsed requiring a mild steroid supplementation. Medical therapy was able to maintain remission and a low Crohn’s disease activity index value for 84 months. Nevertheless, a sudden occlusive episode, due to the stenosis of the ascending colon, required a right hemicolectomy. After surgery, immunosuppressive therapy was unnecessary.

Finally, no relationship was seen between the duration of corticosteroid use prior to therapy and the efficacy of tacrolimus.


We report our experience with tacrolimus in the treatment of Crohn’s disease complicated by resistance to corticosteroids. This complication effects 15–20% of patients.15 When it occurs, current thinking is that bowel resection is the only chance and that relapses may appear even after surgery.16, 17 Therefore, the need for an effective medical therapy other than steroids has been emphasized.

Ciclosporine has been employed in inflammatory bowel disease alone or in combination with low doses of steroids, 6-mercaptopurine and azathioprine. Although the efficacy of intravenously administering ciclosporin has been strongly proven in the active phase, few reports address its use in the long term.2, 8[9]–10, 18[19][20][21][22]–23 Cohen et al. have published a 5-year experience with this drug in ulcerative colitis, demonstrating its usefulness in allowing most severe patients to avoid or plan surgical operations.24 Nevertheless, restraining factors for ciclosporin have been described (i.e. infectious complications and severe adverse events, in particular nephrotoxicity).10, 19, 25 Moreover, relapses resulting from lowered serum levels due to incomplete intestinal absorption are frequent when the drug is administered orally. Indeed, three controlled trials with oral ciclosporin do not show any important advantage when the drug is employed at a dose of 5 mg/kg, whilst one set of evidence suggests that an increase in this dose (7.6 mg/kg) may be of benefit.20[21][22]–23 Intestinal absorption is improved by the use of a new formulation (Neoral) and future experiences could give a further boost to the oral administration of this drug in the long term.

In our series, three out of 13 patients, previously treated with oral ciclosporin, stopped the drug because of the onset and persistence of unpleasant side-effects (hirsutism and gingival hyperplasia), combined with of an unsatisfactory therapeutical response.

Tacrolimus, compared to ciclosporin, has the advantage of being well absorbed by the gastroduodenal tract and therefore can be used orally in the presence of a diseased small bowel.26 The efficacy of this drug is indirectly confirmed by its use in patients with inflammatory bowel diseases who had undergone liver transplantation, with the dual aim of avoiding rejection and treating intestinal disorder.27 Until now, the use of tacrolimus in severe or complicated Crohn’s disease has been limited to a few cases and for a short period, as a ‘bridge’ to other immunosuppressive therapies.11, 12 Among them, azathioprine plays a central role even though it is in most cases ineffective in the active phase because of the slow onset of action, and is therefore used only to maintain clinical remission.3

Recently, Fellermann et al. described a way of treating steroid-unresponsive acute attacks of inflammatory bowel diseases by intravenous tacrolimus and maintaining remission by oral administration of the same drug.28 Their results have been promising in 11 patients (six with ulcerative colitis and five with Crohn’s disease) who had been followed up for a mean period of 7 months. Our experience with tacrolimus in Crohn’s disease is longer, with a median follow-up period lasting more than 2 years. Even though adverse events occurred frequently, they were controlled by therapeutic dose arrangements in all but one patient. Moreover, none of the patients showed changes in blood creatinine, suggesting an absence of long-term nephrotoxic effects in our series.

For a complete evaluation of the effectiveness of tacrolimus therapy, we believe that it is of relevance to discriminate short- and long-term therapeutical effects. After only a few weeks of therapy we observed the following benefits:

1 complete reversal of adverse events due to previous treatments (steroids and ciclosporin);

2 closure of fistulas in three out of six patients and significant improvement in another, with a 10-fold reduction in a huge external opening;29

3 possibility of withdrawing total parenteral nutrition in three out of six patients.

With regard to the advantage of using tacrolimus long-term, we highlight the following findings:

1 a significant change in Crohn’s disease activity index and inflammatory bowel disease life-quality questionnaire values after 6 and 12 months of therapy;

2 the possibility of controlling adverse events;

3 the possibility of delaying the need for surgery and providing the time for ‘elective’ operations;

4 the possibility of withdrawing steroid therapy (indeed, a supplementation with these drugs was required in five out of 13 patients and was ineffective in all but one, in whom we observed the well-known dose-sparing described in transplanted patients).30

Our experience confirms that tacrolimus may represent a pertinent therapeutical option in Crohn’s disease when conventional therapies fail to be effective. The possibility of using tacrolimus in both the short and long-term could attribute it a singular role in the treatment of Crohn’s disease despite the recent introduction and proven effectiveness of a new specific drug, Infliximab (monoclonal antibody anti-tumour necrosis factor alpha).31 Our wide experience with tacrolimus in transplanted patients gave us the opportunity to elucidate some principles of correct administration and we came to the conclusion that the management of tacrolimus is more feasible in non-transplanted patients. In fact, in Crohn’s disease we have never seen the complications of tacrolimus therapy which are observed in transplanted patients (viral infections and additive effects with other drugs).32

In conclusion, good compliance results from oral administration and the dual possibility of obtaining both the resolution of acute attacks and the maintenance of remission. This encourages the trial of tacrolimus in more significant groups of patients in order to confirm the results we achieved in controlled trials and to broaden the therapeutic indications for tacrolimus beyond the end-stage of disease as an extreme attempt to avoid surgery.