• oral tacrolimus;
  • refractory inflammatory bowel disease;
  • inflammatory bowel disease;
  • IBD;
  • CD;
  • Crohn's disease;
  • UC;
  • ulcerative colitis;
  • CDAI;
  • Crohn's disease activity index;
  • AZA;
  • Azathioprine;
  • 6-MP;
  • 6-mercaptopurine


  1. Top of page
  2. Abstract

Background: This study aimed to evaluate the efficacy of oral tacrolimus in patients with inflammatory bowel disease (IBD) refractory to conventional therapy, including azathioprine, 6-mercaptopurine, and infliximab.

Methods: Retrospective review of all patients with IBD treated with oral tacrolimus was undertaken. Tacrolimus was administered at an initial dose of 0.05 mg/kg twice daily, aiming for serum trough levels of 5–10 ng/mL. We evaluated clinical response, a retrospective estimated Crohn's disease activity index (CDAI) for Crohn's disease (CD), modified Truelove-Witts index for ulcerative colitis (UC), and modified pouch disease activity index (mPDAI) for pouchitis. Patients had been monitored clinically for benefit and side effects and by whole blood tacrolimus level approximately every 4 weeks for the duration of treatment. Clinical remission was defined as an estimated CDAI <150 (CD), an inactive disease score on the Truelove-Witts index (UC), and mPDAI <5 (pouchitis).

Results: Twelve patients with CD, six with UC, and one with pouchitis, all resistant to previous therapies, were treated for a median of 5 months. After 4 weeks 10 CD (83%), four UC (67%) patients, and one pouchitis patient had a clinical response. There was a median reduction of the estimated CDAI of 108 points (range 35–203; P = 0.002) and stool frequency of three per day at week 4. Remission was achieved in 42% (5/12) of CD and 50% (3/6) of UC patients at the end of follow-up. Side effects included temporary elevated creatinine (n = 1), tremor (n = 3), arthralgia (n = 1), insomnia (n = 1), and malaise (n = 1). Four patients discontinued treatment due to side effects.

Conclusion: Oral tacrolimus is well tolerated and effective in patients with refractory IBD in the short- to medium-term. Further controlled, long-term evaluation is warranted.

(Inflamm Bowel Dis 2007)

The medical management of patients with refractory inflammatory bowel disease (IBD) remains a challenge. Up to 28% of patients with IBD are intolerant of thiopurines,1 and a further 10% are unresponsive.2 Infliximab has proven to be very effective in patients with both Crohn's disease (CD)3 and ulcerative colitis (UC),4 although at 1 year approximately one-fifth of all initially treated CD and UC patients are in remission. There is still a large gap in the therapeutic armamentarium of both conditions.

Oral tacrolimus, a macrolide immunosuppressant approved for the prophylaxis of kidney and liver transplant rejection, has been shown to be effective in refractory IBD. It has a better intestinal absorption and side effect profile compared to cyclosporine.5 Earlier small uncontrolled studies showed benefit in active luminal CD and UC.6–10 One placebo-controlled trial showed significant improvement over placebo for perianal fistulizing CD, although the remission rate was not superior to placebo.11 Oral tacrolimus has also been shown to be effective in refractory UC in a placebo-controlled trial.12

We have been using oral tacrolimus to treat patients with refractory IBD unresponsive to standard therapies and present here the results of our experience.


  1. Top of page
  2. Abstract

Patients and Design

Patients were treated with tacrolimus therapy if they had a diagnosis of CD, UC, or pouchitis and were experiencing symptoms of moderate-to-severe chronic active disease, were steroid-dependent, or had failed therapy with azathioprine, 6-mercaptopurine, or infliximab. Patients with anorectal Crohn's fistulas, continuously draining despite concomitant medical therapy, were included. Patients were informed about the potential risks and benefits of tacrolimus therapy and consented to its use. In all patients disease diagnosis had been achieved according to standard criteria by prior clinical assessment, radiology, endoscopy, and histology.

During treatment patients were monitored at ≈4-week intervals. At each visit patients' clinical symptoms were recorded and measurement made of full blood count, liver and renal function tests, and C-reactive protein. Patients were closely monitored for side effects.

This retrospective, observational, single-center review included all patients treated. The case notes of all patients treated with oral tacrolimus at one specialist gastroenterological institution treating secondary and tertiary care patients were reviewed.

An estimated Crohn's disease activity index (CDAI)13 for the CD patients, modified Truelove-Witts score14 (inactive, mild, moderate, or severe disease) for the UC patients, and modified pouch disease activity index (mPDAI)15 for the pouchitis patient were retrospectively calculated. Assessment was made of clinical response, remission, and adverse effects.

Study Medication

All patients were started at an initial tacrolimus dose of 0.1 mg/kg/day in two divided doses. Trough whole blood levels were measured on each visit and the dose adjusted for each patient to achieve a trough level of 5–10 ng/mL.

Definition of Response

Clinical remission was retrospectively defined as an estimated CDAI less than 150 in patients with CD, an inactive disease score on the Truelove-Witts criteria for UC, and an mPDAI less than 5 for pouchitis. The latter has a potential maximum (worst) score of 12 and a potential minimum (best) score of 0.

In patients with fistulizing CD the clinical response was defined as one of: clinical remission, defined as the absence of draining fistulas; partial response, defined as the reduction of at least 50% in the number of draining fistulas as well as a decrease in size, drainage, or discomfort associated with the fistulas; and any other outcome, defined as no response.16


  1. Top of page
  2. Abstract

Seven male and 12 female patients (mean age 35, range 18–75 years) with well-documented IBD were treated. Twelve patients had CD, six had UC, and one had pouchitis. Patient details are shown in Table 1. Of the six patients with UC, three had extensive colitis to the transverse colon, two had left-sided colitis, and 1 had total colitis. Of the 12 patients with CD, seven had concurrent luminal and fistulizing disease (seven perianal).

Table 1. Baseline Characteristics of Patients
PatientAge, genderDiseaseDuration of disease (yr)Site of diseasePerianal diseasePrevious operation(s)
  1. F, female; M, male; UC, ulcerative colitis; CD, Crohn's disease; AP, abdominoperineal; TI, terminal ileum.

123 FUC5Extensive  
231 FUC8Left-sided  
366 FUC10Extensive  
437 FUC5Extensive  
575 MUC8Left-sided  
624 MUC5Total  
740 FUC10Pouch Ileal pouch-anal anastomosis
826 FCD9IleocolonicFistula 
918 MCD7Ileocolonic  
1040 FCD17Ileum Right hemicolectomy
1121 FCD4IleocolonicFistulaIleostomy
1221 FCD6PerianalFistulaColostomy, protectomy
1321 MCD11ColonicFistulaColectomy
1418 MCD9IleocolonicFistula 
1558 FCD41IleumFistulaSmall bowel resections
1658 FCD15Ileocolonic Right hemicolectomy, strictuloplasty
1726 FCD9Ileocolonic  
1833 FCD14ColonicFistulaAP resection End colostomy
1935 MCD12Ileocolonic TI resection

Mean disease duration was 10.8 years (range 4–41). Patients were either steroid-dependent or steroid-resistant and had failed (intolerant or unresponsive) previous treatment with immunomodulators including azathioprine, 6-mercaptopurine (6-MP), or methotrexate. Most patients had received multiple IBD medications, as shown in Table 2, prior to treatment with tacrolimus. In all, 89% (17/19) of patients had failed azathioprine; 32% (6/19) had failed 6-MP; 47% (9/19) had failed infliximab; 16% (3/19) had failed adalimumab; and 5% (1/19) had failed Adacolumn phoresis.

Table 2. Previous and Current Treatment
PatientPrevious treatmentConcurrent treatmentConcurrent prednisolonePrevious IV CyA
  1. 5-ASA, 5-aminosalicylic acid; AZA, Azathioprine; 6-MP, 6-mercaptopurine; MTX, methotrexate; CyA, Cyclosporine.

15-ASA, AZA, Adacolumn6 MP, PentasaYes 
25-ASA, AZA6 MPNo 
35 ASA, AZA, 6MPColazideNoOnce
45-ASA, AZA, ProbioticsAZANoOnce
5AZA, 6MPColazideYes 
6AZAColazide, probioticsNo 
75-ASA, AZA, metronidazole, ciprofloxacin, Probiotics No 
8AZA, 6MP, MTX, mycophenolate, thalidomidePentasaYes 
95-ASA, AZA,AZA, probioticsNo 
10AZA, 6MP, MTXPentasaNo 
11AZA, 6MP, MTX, Infliximab, adalimumabThalidomideNo 
125-ASA, infliximab,AZA, ThalidomideNo 
13AZA, infliximab, thalidomide, adalimumabAZA, antibioticsYes 
14AZA, infliximab No 
15AZA, 6MP, infliximab No 
166 MP, MTX, infliximab Yes 
175-ASA, AZA, MTX, infliximab, adalimumabAZAYes 
185-ASA, AZA, MTX, infliximab Yes 
19AZA, 6MP, MTX, infliximab Yes 

During initial treatment with tacrolimus, seven patients (37%) were on a concurrent thiopurine analog (azathioprine or 6-mercaptopurine). Two patients (11%) with perianal fistulas were taking thalidomide. Eight patients (42%) were still on concurrent prednisolone. The duration of tacrolimus therapy averaged 5 months (range 0.5–16).

Clinical Outcome

Crohn's Disease

At week 4 there was a decrease in the estimated CDAI for all patients, with a median reduction of 108 points (range 35–203; P = 0.002). The initial median CDAI was 281 (range 155–496); this decreased to a median of 153 (range 90–400) with treatment.

Stool frequency was improved in nine patients and remained unchanged in three patients. The mean stool frequency reduction was 3 per day, from a median of 6 per day to 3 per day. At the end of a mean follow-up of 5 months in this group, of the 12 patients five achieved remission. In a further two patients (Patients 12 and 17) there was a clinical response with estimated CDAI reduction.

Of the remaining patients, two (Patients 13 and 19) had a relapse requiring a course of steroids. The remaining three patients (Patients 15, 16, and 18) continued to have severe activity and were referred for surgery; of these patients 15 and 16 stopped tacrolimus after 5 days and 2 weeks, respectively, due to significant adverse effects. Patient 15 had a colectomy for enterocutaneous fistula and Patient 18 had a complex ileocolic fistula.

Perianal Fistula(e)

Of seven patients with fistulas, one achieved complete fistula healing. In two patients there was partial response with a reduction in fistula size and drainage. The remaining four patients showed no response in fistula healing. The mean follow-up was 5 months.

Ulcerative Colitis

The mean follow-up for the six patients with UC was 8 months. After 4 weeks of treatment three were in remission and one had a clinical response (estimated Truelove-Witts index score changing from severe to mild disease). Although the estimated Truelove-Witts score was unchanged in Patients 5 and 6, the amount and frequency of bloody diarrhea had decreased considerably. Both patients were steroid-dependent at the start of tacrolimus therapy which partly explained the mild disease activity score prior to treatment.

Three of the six patients were in remission at the end of follow-up. Steroid tapering was complete in these three patients. Two patients had a relapse during treatment. Patient 1 relapsed at 4 months and Patient 4 at 13 months. Both patients required intravenous steroids and continued to have severe disease activity. Tacrolimus was discontinued in Patient 5 due to side effects at week 4.


Patient 7 responded well to oral tacrolimus at week 4, with a reduction in stool frequency from 25 times per day to 12 times per day. There was a reduction in the mPDAI from 8 to 4. Short-term and medium-term results of all patients are shown in Table 3 and 4.

Table 3. Short-term Response to Oral Tacrolimus (at 4 Weeks)
PatientDisease activityStool frequencyBleeding
  • *

    Modified Truelove-Witts index

  • Modified Pouch Disease Activity Index

  • Crohn's Disease Activity Index.

Table 4. Medium-term Response to Oral Tacrolimus
PatientDiseaseTreatment (month)Medium term outcomeFistulaRelapse (month)Prednisolone (mg/day)Steroid taperingAdverse effects
  1. UC, ulcerative colitis; CD, Crohn's disease; IV, intravenous.

1UC5Severe disease 4IV steroids  
2UC6Remission   complete 
3UC6Remission   complete 
4UC16Severe disease 13IV steroids  
5UC2Mild disease  6 mg Tremor, insomia, arthalgia
6UC12Remission   complete 
8CD12RemissionRemission3 complete 
10CD4Remission    Tremor
11CD7RemissionPartial response2 completeRenal impairment
12CD6Mild activityNo response    
13CD6Mild activityNo response550 mgnone 
14CD3RemissionPartial response    
15CD0.25Severe activityNo response   Malaise
16CD0.5Severe activity 215 mgnoneTremor, parasthesia
17CD2Mild activity  7.5 mg  
18CD2Severe activityNo response 25 mg  
19CD4Severe activity 4IV steroids  
Tacrolimus Monitoring

The mean trough tacrolimus whole blood level was 6.6 ng/mL (range 1.0–17.4). Mean follow-up for treatment was 6 months. At the end of follow-up 13 patients (68%) were still on treatment. Individualized data are shown in Table 5.

Table 5. Tacrolimus Dose and Duration
PatientOngoing tacrolimus therapyDuration of tacrolimus therapy (months)Mean tacrolimus level (ng/mL)Duration of follow-up (months)
1No517.4 (4.1–30.7)6
2Yes61.7 (1.0–3.2)6
3Yes66.5 (6.5)6
4Yes164.7 (2.3–9.9)16
5No213.7 (13.7)3
6Yes1211.3 (7.0–18.3)12
8Yes124.2 (3.1–5.4)12
9Yes55.3 (3.1–10.4)5
10No47.7 (7.7)6
11Yes79.9 (4.5–20.8)7
12Yes62.8 (2.1–3.5)6
13Yes610.8 (10.6–10.9)6
14Yes36.9 (6.4–7.3)3
17Yes25.1 (4.0–6.1)2
Side Effects

Adverse events were infrequent. The drug was well tolerated in 74% of the patients (14/19). Four patients stopped treatment due to side effects: three patients had severe tremor (one of whom also had insomnia and arthralgia) and one patient described severe malaise.

One patient had a temporary rise of serum creatinine to 162 μmol/L (normal range 70–150 μmol/L). The patient's trough tacrolimus level was high at 20.8 ng/mL. Tacrolimus level reduced to 5.5 ng/mL and serum creatinine normalized to 60 μmol/L within 2 days with dose reduction. The medication was continued in this patient and no other change in renal function was observed throughout the study period.


  1. Top of page
  2. Abstract

The clinical management of IBD involves the treatment of acute disease and the maintenance of remission. For both these aspects of disease management there remains a need for effective therapies for patients who have failed standard immunosuppressive therapy and treatment with biological agents.

Tacrolimus has been extensively used in solid organ transplant settings and was incidentally discovered to improve the outcome of IBD in liver transplant patients with primary sclerosing cholangitis.17, 18 It is a macrolide antibiotic isolated from Streptomyces tsukibaensis. Although structurally different, it has a similar mechanism of action to cyclosporine, acting as a calcineurin inhibitor, leading to reduced production of Interleukin-2.19 Compared to cyclosporine, it has the advantages of more predictable absorption and a lower rate of side effects, including seizures and gingival hyperplasia.20

In this series all patients with CD benefited from treatment, with a reduction in the estimated CDAI score at 4 weeks, except for two patients who had to discontinue treatment due to side effects. Almost half the patients were in remission at the end of follow-up. Two-thirds of patients with UC experienced a decline in the Truelove-Witts index score within 4 weeks. A benefit was seen in one patient with pouchitis.

The clinical effect appeared to be rapid. This finding has been supported by earlier experience.8 The initial benefit appeared to be maintained at medium-term follow-up, with half of the patients achieving remission at 8 months. Oral tacrolimus appeared to provide a steroid-sparing effect, with successful steroid tapering in five patients. The need for surgery was also averted in a small number of patients.

This study was a retrospective one, and as such the precise definition of disease activity is limited. Patients often had a clear clinical response to tacrolimus therapy, including a reduction in pain, bowel frequency, and need for steroids. We used retrospectively calculated disease activity scores; while these have not been validated in such a setting, in this study they reflected intraindividual variations in disease activity well and provided an additional quantitative measure of drug response.

Our positive findings are in agreement with previous work. The largest study, by Fellerman et al,8 reported 38 patients with refractory UC or indeterminate colitis who were treated with oral tacrolimus. Eighteen of 38 patients improved within 14 days, and a complete remission was achieved in 13 patients after 1 month. The overall colectomy rate was 34% over 16 months.

Baumgart et al10 demonstrated a favorable outcome in their retrospective series of 31 patients. Twenty-eight patients (90.3%) responded and 20 patients (64.5%) went into remission at 1 year follow-up. The response to tacrolimus was superior in UC compared to CD patients (74% versus 17%, respectively).

A recent prospective controlled study compared the role of tacrolimus in the induction and maintenance of remission in 37 moderate to severe CD patients with azathioprine. At 2 years there was no difference in remission rates between the two groups (tacrolimus versus azathioprine: 72.2% versus 73.6%).21 Tacrolimus would therefore appear to be an effective alternative for patients who have failed or are intolerant of azathioprine.

There are limited data on the long-term outcome of patients on tacrolimus. The only long-term study reviewed 18 children with steroid-dependent and steroid-resistant UC over a 5-year treatment period.22 Thirteen patients required colectomy and the mean time from tacrolimus therapy to colectomy was 392 days. This retrospective study suggested that tacrolimus in pediatric patients was unlikely to provide prolonged clinical response.

Our experience with fistula healing should be considered in the context of previous studies, and in particular in relation to the duration of treatment and follow-up (mean 5 months). The only controlled trial showed a significantly positive effect on improvement compared to placebo, but did not show fistula healing when patients were treated for 10 weeks.11 A more recent open study by Gonzalez-Lama et al23 of 10 patients showed substantial healing, with 40% achieving a complete clinical response and 50% a partial response, but patients were treated for 6–24 months. Longer-term therapy may therefore be required to achieve healing using this medication.

Lowry et al24 observed a benefit of a combination therapy with tacrolimus and purine analogs in perianal fistulizing CD. Sixty-four percent of their patients had complete fistula closure within 3 months, while the remaining patients experienced a partial response.

The dose of tacrolimus may be an important determinant of induction and maintenance of remission. A recent placebo-controlled trial of 60 patients showed that patients in whom therapy was aimed at achieving a higher tacrolimus trough level had better response rates. Twenty patients were randomized to each treatment: a low trough level (5–10 ng/mL whole blood), high trough level (10–15 ng/mL whole blood), or placebo. At the end of 2 weeks treatment, there was a 68.4% improvement rate in the high trough level group, 38.1% in the low trough level group, and 10% in the placebo group.12 The mean trough level in the current study was 6.6 ng/mL, which was at the lower end of therapeutic range. This may affect response. There may be room for an increased dose in our group of patients within the therapeutic window.

Tacrolimus has been commonly reported to be associated with tremor, hyperglycemia, nephrotoxicity, paresthesia, hypertension, insomnia, and gastrointestinal disturbances. Adverse effects, however, were infrequent in our patient group.

In conclusion, this retrospective study shows oral tacrolimus to be a useful therapy in patients with disease refractory to standard therapies. A high response rate and an impressive remission rate were observed in the short- to medium-term, with minimal adverse effects. Further controlled, long-term evaluation is warranted.


  1. Top of page
  2. Abstract