Share this article on:

Oral Tacrolimus for Steroid-dependent and Steroid-resistant Ulcerative Colitis in Children

Ziring, David A; Wu, Steven S; Mow, William S; Martín, Martín G; Mehra, Mini; Ament, Marvin E

Journal of Pediatric Gastroenterology and Nutrition: September 2007 - Volume 45 - Issue 3 - p 306–311
doi: 10.1097/MPG.0b013e31805b82e4
Original Articles: Gastroenterology

Purpose: To evaluate tacrolimus in 3 situations: for the induction of remission in children with severe steroid-resistant ulcerative colitis (UC); for steroid sparing in children with steroid-dependent UC in whom treatment with other immunosuppressants fails; and for the maintenance of remission in children with steroid-dependent and steroid-resistant UC.

Patients and Methods: We retrospectively evaluated 18 consecutive patients (13 with pancolitis) who were treated with oral tacrolimus at our institution from May 1999 to October 2005. Nine patients had steroid-resistant UC and 9 patients were steroid-dependent. We started patients initially on tacrolimus 0.2 mg/kg divided twice daily, with a goal plasma trough level of 10 to 15 ng/mL for the first 2 weeks, and then titrated doses to achieve plasma levels between 7 and 12 ng/mL after induction.

Results: Of the 18 patients in this study, 17 showed a positive response to tacrolimus therapy (ie, cessation of diarrhea and other symptoms) and 5 showed a prolonged response to tacrolimus. The mean time from initiation of tacrolimus therapy until response was 8.5 days. The mean duration of response was 260 days. Eleven of 18 patients required colectomy, including all of the patients with steroid-resistant UC, but only 2 of 9 who were steroid-dependent. The mean time from initiation of tacrolimus until colectomy was 392 days.

Conclusions: It is possible that tacrolimus may benefit selected patients with steroid-dependent UC, including those who are intolerant of 6-mercaptopurine or azathioprine. Conversely, patients with steroid-resistant UC are unlikely to sustain a prolonged clinical response to tacrolimus and seem to require colectomy eventually. Careful considerations of risk versus benefit, as well as close monitoring for adverse effects, are essential in all patients.

Division of Pediatric Gastroenterology, Hepatology, and Nutrition, Mattel Children's Hospital, David Geffen School of Medicine at the University of California Los Angeles

Received 28 June, 2006

Accepted 5 March, 2007

Address correspondence and reprint requests to Marvin E. Ament, MD, Division of Pediatric Gastroenterology, Hepatology, and Nutrition, Mattel Children's Hospital, David Geffen School of Medicine at UCLA, 12-383 MDCC, 10833 LeConte Ave, Los Angeles, CA 90095-1752 (e-mail:

Back to Top | Article Outline


Children with steroid-dependent or steroid-resistant ulcerative colitis (UC) have few medical therapeutic options. Most of these patients ultimately require surgical intervention to stop disease sequelae such as bleeding, chronic debilitating diarrhea, and hypoproteinemia, and to spare the long-term effects of steroids. To date, children with severe UC who have been treated with intravenous cyclosporine have not seen therapeutic benefits comparable to those in adult patients with similar extent of disease (1–4). Like cyclosporine, tacrolimus (FK506) is a potent immunosuppressive agent of the calcineurin inhibitor class used in liver and other solid organ transplantation. Its toxicity profile is similar to that of cyclosporine, although there seems to be significantly lower incidences of hypertension, hyperlipidemia, and renal failure with the use of tacrolimus on a long-term basis in solid organ transplantation (5).

Several studies have addressed the use of tacrolimus in pediatric UC (6–10). These investigators attempted to use tacrolimus to “bridge” therapy from steroids to immunomodulators (6-mercaptopurine [6-MP] and azathioprine) because of the delayed onset of efficacy of these drugs. However, to date, there have been no large, long-term studies of the prolonged use of tacrolimus for pediatric steroid-dependent and steroid-resistant UC.

The aim of this study was to evaluate tacrolimus as a drug in the following situations: for the induction of remission in pediatric patients with severe steroid-resistant UC; for the possibility of steroid sparing in patients who are steroid dependent but in whom treatment with other traditionally used immunosuppressants fails; and for the maintenance of remission in patients with both steroid-dependent and steroid-resistant UC.

Back to Top | Article Outline


Eighteen consecutive pediatric patients with a diagnosis of UC (13 with pancolitis) treated with oral tacrolimus at our institution between May 1999 and October 2005 were evaluated in this study (Table 1). These 18 patients comprise our entire experience in the treatment of pediatric patients with UC with tacrolimus. This was a retrospective chart review approved by our institutional review board. The diagnosis of UC was established by clinical history, physical examination, colonoscopy with characteristic biopsy findings, and the absence of recognized infectious etiologies by biopsy and stool examination. The inclusion criteria were age of 1 to 18 years with steroid-resistant UC as defined by inability to wean from steroids secondary to recurrence of symptoms of diarrhea, urgency, and/or bloody stools; or steroid-refractory UC as defined by continued symptoms of bloody diarrhea despite 5 days of intravenous steroid treatment at a dosage of 1 to 2 mg/kg (≤60 mg/d). Exclusion criteria were the presence of infection as determined by the presence of Clostridium difficile toxin, bacterial pathogens, or ova and parasites in the stool; evidence of cytomegalovirus (CMV) colitis by flexible sigmoidoscopy in our first several years of experience with using tacrolimus in patients with UC; or an indication for emergent surgery (eg, toxic megacolon, perforation, peritonitis). We performed a systematic examination for CMV by immunostaining in the tissue of patients seen early in the experience, but never found evidence of active CMV colitis. In later patients there were no viral inclusions seen on routine histological examination. All of our patients were cyclosporine- and infliximab-naïve. Nine patients were diagnosed with steroid-resistant UC and 9 had steroid-dependent UC. All of the patients underwent comprehensive laboratory testing before the initiation of tacrolimus, including complete blood count, magnesium measurements, and tests of liver and renal function (as described later).



Back to Top | Article Outline

Dosage and Monitoring

We started treatment initially with tacrolimus 0.2 mg/kg divided twice daily into equal doses, with a goal plasma trough level of 10 to 15 ng/mL for the first 2 weeks. In those who showed a response, this was followed by titration of doses to achieve plasma levels between 7 and 12 ng/mL for maintenance therapy.

All of the patients received prophylaxis for Pneumocystis carinii pneumonia with trimethoprim-sulfamethoxazole, and magnesium supplementation if necessary. We monitored all of our inpatients with daily measurements of serum electrolytes, blood urea nitrogen, creatinine, magnesium, and tacrolimus trough levels until these laboratory values were stable.

Back to Top | Article Outline

Data Collection

Data collected before the initiation of tacrolimus treatment included the following: complete blood count with platelets, erythrocyte sedimentation rate, C-reactive protein, blood urea nitrogen, creatinine, and perinuclear anti-neutrophil cytoplasmic antibodies (pANCA) serology result (positive or negative). When treatment began, we also recorded the tacrolimus dose in milligrams per kilogram and calculated the mean tacrolimus plasma trough level (the mean of all recorded values during the time the patient was receiving the drug). We initially evaluated our patients' clinical response after 3 to 5 days of therapy. Treatment response was defined as the absence of the following symptoms: bloody diarrhea, tenesmus, loose stools, urgency, transfusion dependence, and intolerance of oral feedings. Treatment failure was defined by the continued presence of ≥1 of these symptoms. Abbreviated response was defined as a response of <3 months' duration. Patients who exhibited a clinical response continued to receive tacrolimus, with plasma levels ranging between 7 and 12 ng/mL. We calculated the time in days from the date of diagnosis until the patients started to receive tacrolimus, the time in days from the start of tacrolimus until response was seen (as defined earlier), and the duration in days that the patient continued to receive tacrolimus. Additionally, clinical data regarding medication use, symptoms, and adverse events were recorded for a follow-up period ranging from 4 months to nearly 4 years. Adverse events were recorded to evaluate possible tacrolimus-induced toxicity.

Back to Top | Article Outline


Data were plotted and mean and standard deviations (SDs) were calculated using Prism software (GraphPad Software, San Diego, CA).

Back to Top | Article Outline


The mean age of our patient population was 11.5 ± 4.3 years (range, 1–16 y), including 8 male and 10 female patients. Their mean white blood cell count at the time of initiation of therapy was 11,700 ± 6670 cells/μL (range, 5900–34,000). Mean hemoglobin and hematocrit levels were 9.7 ± 2.3 g/dL (range, 6.9–15.3) and 30.8% ± 5.6% (range, 22.8%–43.8%). This anemia in the majority of patients was likely due to anemia of chronic disease because the mean corpuscular volume was 82.9 ± 6.9 fL (range, 64.7–91.4). Mean platelet count was 349,000/μL ± 171,000/μL (range, 195,000–819,000); this relative thrombocytosis probably represents a systemic acute phase reactant in this population. The mean erythrocyte sedimentation rate and C-reactive protein level at the start of therapy were elevated at 22 ± 21 mm/h (range, 1–63) and 2 ± 1.4 mg/dL (range, 0.6–4.1), respectively, again reflecting the presence of systemic inflammation in most patients. In those patients in whom serum antinuclear cytoplasmic antibodies were measured (n = 13), 7 were positive and 6 were negative. The mean tacrolimus dose used was 0.2 ± 0.07 mg/kg (range, 0.08–0.38), which resulted in a mean plasma trough value of 11.6 ± 2.9 ng/mL (range, 3.6–16.5). The mean time from diagnosis to the start of tacrolimus therapy was 82.5 ± 347 days (range, 7–1136).

Of the 18 patients in this study, 17 (94%) responded positively to tacrolimus therapy; only 1 of the patients with steroid-resistant UC did not show a response (Fig. 1). Among the 15 responders for whom precise dates are available, the mean time in days from initiation of tacrolimus therapy until response was noted was 8.5 ± 6.7 days (range, 4–25; Fig. 2). The mean duration of response was 168.5 ± 397 days (range, 15–1446), including patients who had to discontinue tacrolimus (n = 12) and those who continued to receive tacrolimus as of the last follow-up date (n = 3). The remaining 2 responders continued to receive tacrolimus with a follow-up of >800 days, although the dates of starting and responding to therapy were not precisely known.

FIG. 1

FIG. 1

FIG. 2

FIG. 2

Of the 9 patients with steroid-resistant UC, 8 (89%) initially showed a response to tacrolimus, although ultimately all of them eventually proceeded to undergo colectomy (Fig. 3). Mean length of response was 415 ± 510 days (range, 15–1446), although 3 patients had a response of <3 months before they required surgery. One patient had a response that lasted 47 days before other adverse events required a change in therapy. One patient did not show a response to tacrolimus after 15 days and, in fact, had a seizure (believed to be related to tacrolimus toxicity) during titration of tacrolimus therapy. In this case tacrolimus was discontinued immediately. These data indicate that although tacrolimus is able to induce prolonged remission, this response varies widely among individual patients.

FIG. 3

FIG. 3

Patients with steroid-dependent UC had a higher likelihood of achieving prolonged remission with tacrolimus. All 9 of these patients had been initiated on 6-MP or azathioprine therapy for attempted steroid sparing. Three of these patients were unable to tolerate 6-MP or azathioprine therapy because of hepatic toxicity or other medication-related side effect (eg, pancreatitis, headache, nausea); the remainder had been unable to wean off steroids while receiving 6-MP. All 9 patients with steroid-dependent UC responded to tacrolimus initially. Two of 9 responders (22%) had relapse after 138 and 534 days and required colectomy. Of the remaining 7 with a prolonged response to tacrolimus (78%), all of them were successfully weaned off steroids; 5 of 9 are receiving tacrolimus alone (56%) and are in remission for a median of 323 days. The other 2 patients (22%) were successfully transitioned to 6-MP maintenance therapy alone after 47 and 149 days on tacrolimus, respectively. The patients who have successfully continued to receive tacrolimus the longest have been in remission for >800 days.

There were no statistically significant differences noted between patients who had a prolonged response to tacrolimus and those who ultimately required colectomy with regard to sex, age, initial hemoglobin, erythrocyte sedimentation rate, platelet count, or mean trough tacrolimus level (data not shown). Similarly, there were no significant differences in laboratory values between the steroid-dependent and steroid-resistant groups. Inflammatory bowel disease serological examinations were performed in 7 patients with steroid-resistant UC and 6 with steroid-dependent UC. Of the 5 patients who have maintained a response to tacrolimus, 4 had serologies tested, with 2 being pANCA-positive and 2 pANCA-negative. Of the 9 in whom tacrolimus therapy failed and eventually underwent colectomy, 8 had serologies tested: 5 were positive and 3 were negative. These data do not imply any correlation between pANCA status and outcome.

Six of the 18 patients in the study experienced adverse events that required discontinuation of therapy. Adverse events included neurotoxicity as evidenced by headache (n = 1), profound tremor (n = 3) or convulsion (n = 1; at that time, the patient had a plasma tacrolimus trough level of 21.9 ng/mL), and Epstein-Barr virus–associated lymphoproliferative disease (n = 1). Patients who had adverse events were evenly distributed between the steroid-resistant and steroid-dependent groups. All of them with 1 exception (a patient who was transitioned to 6-MP) then went on to undergo colectomy. Hypomagnesemia occurred in 7 patients during tacrolimus therapy. We routinely check serum magnesium levels with each blood draw in all of the patients receiving tacrolimus and treat hypomagnesemia with oral magnesium supplementation; none of these patients were symptomatic.

Back to Top | Article Outline


Patients with severe UC are often hospitalized and critically ill, and in many cases proceed rapidly to treatment with colectomy, when severe colitis either fails to respond to initial therapy or relapses with conventional treatment. Colectomy may also become necessary to avoid the adverse effects of long-term corticosteroid therapy. However, many pediatric patients with medically resistant UC and their families resist the proposition of curative colectomy as a result of fear of surgery or of postsurgical quality of life. These patients are often willing to risk significant side effects from more potent immunomodulators in exchange for the avoidance of major surgery.

Cyclosporine has been used to rapidly induce remission in adults with severe acute UC, and approximately one third of these patients have maintained long-term clinical remission with this drug. Rowe et al (4) showed in their experience with 36 adults that after treatment with intravenous cyclosporine followed by oral cyclosporine, 48% of initial responders had a sustained response at 9 months follow-up, whereas 52% of responders required colectomy. However, children with severe UC have not fared as well with the same regimen: In a study of 14 children who received intravenous and then oral cyclosporine, 11 (79%) required colectomy at 1 year follow-up (2,3). This failure of cyclosporine to maintain long-term remission in pediatric patients with UC has motivated the search for alternative medical treatments.

We found that the use of tacrolimus in pediatric patients with steroid-resistant or steroid-dependent UC provided a rapid clinical response. This rapid response to induction therapy is similar to that provided by infliximab in pediatric patients with UC (11). Conversely, tacrolimus demonstrated limited efficacy (28%) in providing a prolonged clinical response; overall, the majority of patients with UC in this study, including all in the steroid-resistant category, underwent curative colectomy or are now considering colectomy. Of note, we observed that prolonged clinical response occurred only in patients with steroid-dependent UC, more than half of whom remain in remission receiving tacrolimus alone with a mean follow-up of 8 months. Continued follow-up will be necessary to determine the duration of remission over the long term.

There is extensive experience with the use of tacrolimus in hundreds of pediatric patients who have undergone transplantation, with exposure to the drug now exceeding 1 decade in our center and others (12). We have used tacrolimus extensively in patients undergoing liver as well as combined liver/small bowel transplantation, which constitute a diverse pediatric population defined by age, ethnicity, and diagnosis (13–15). It is known that the use of tacrolimus carries a risk of nephrotoxicity over the long term similar to that of cyclosporine. Tacrolimus does have some advantages over cyclosporine in the pediatric population because of the ease of administration, adequate bioavailability, and the absence of side effects such as hirsutism. However, it does have a number of common and generally reversible side effects, including renal toxicity, bone marrow suppression, and neurological symptoms; Epstein-Barr virus–related lymphoproliferative disease is a more rare but more serious consequence. For these reasons, our practice is to administer tacrolimus according to therapeutic target plasma levels, aiming to avoid trough tacrolimus levels in excess of 16 to 18 ng/mL. Even with this close monitoring, one third of our patients required discontinuation of tacrolimus because of adverse effects, illustrating that intolerance to this drug is a significant factor limiting its use.

We aimed to achieve plasma trough levels of tacrolimus initially in the range of 10 to 15 ng/mL for the first 2 weeks, and then lower levels in those patients who showed a response to the drug (7–12 ng/mL) because we have found that in the months after transplantation, keeping the tacrolimus trough level in this range effectively prevents transplant rejection. Although the immunopathology of UC is different from the immune reaction to an orthotopic transplant, we know that the effector arm in UC is largely driven by T cells. Tacrolimus suppresses T cell activity in the setting of transplantation as well as in diseases in which immune homeostasis is disturbed so that there is an overactivity of T cells, such as UC. However, whereas in transplantation over time the immune system may develop a measure of tolerance for the transplanted organ allowing for a decrease in the goal tacrolimus trough level, in UC immune activity does not wane, necessitating a steadily maintained plasma tacrolimus trough level of 7 to 12 ng/mL. We calculated the mean tacrolimus trough level of each patient receiving the drug over time because these levels sometimes varied widely, with occasional toxic values and occasional subtherapeutic values.

In summary, tacrolimus therapy in UC may be useful only under a limited set of clinical circumstances. In chronically steroid-dependent patients who have characteristic side effects (eg, cushingoid habitus, hypertension, osteoporosis), tacrolimus may allow for steroid withdrawal in many cases, and in some patients it may offer the possibility of prolonged remission in place of or as a bridge to traditional maintenance therapy with 6-MP or azathioprine. In patients with steroid-resistant colitis, tacrolimus is capable of inducing short- to medium-term remission; however, no long-term benefit was demonstrated, with all of the patients eventually proceeding to treatment with colectomy. Adverse effects are not inconsequential, and require a careful consideration of the risks versus benefits of tacrolimus in each case, as well as close monitoring by doctors experienced in its pediatric use. Nevertheless, in selected patients with UC, tacrolimus may offer benefits beyond those of conventional therapies.

Back to Top | Article Outline


1. Benkov KJ, Rosh JR, Schwersenz AH, et al. Cyclosporine as an alternative to surgery in children with inflammatory bowel disease. J Pediatr Gastroenterol Nutr 1994; 19:290–294.
2. Treem WR, Davis PM, Hyams JS. Cyclosporine treatment of severe ulcerative colitis in children. J Pediatr 1991; 119:994–997.
3. Treem WR, Cohen J, Davis PM, et al. Cyclosporine for the treatment of fulminant ulcerative colitis in children: immediate response, long-term results, and impact on surgery. Dis Colon Rectum 1995; 38:474–479.
4. Rowe FA, Walker JH, Karp LC, et al. Factors predictive of response to cyclosporin treatment for severe, steroid-resistant ulcerative colitis. Am J Gastroenterol 2000; 95:2000–2008.
5. Lucey MR, Abdelmalek MF, Gagliardi R, et al. A comparison of tacrolimus and cyclosporine in liver transplantation: effects on renal function and cardiovascular risk status. Am J Transplant 2005; 5:1111–1119.
6. Fellermann K, Ludwig D, Stahl M, et al. Steroid-unresponsive acute attacks of inflammatory bowel disease: immunomodulation by tacrolimus (FK506). Am J Gastroenterol 1998; 93:1860–1866.
7. Fellermann K, Tanko Z, Herrlinger KR, et al. Response of refractory colitis to intravenous or oral tacrolimus (FK506). Inflamm Bowel Dis 2002; 8:317–324.
8. Bousvaros A, Kirschner BS, Werlin SL, et al. Oral tacrolimus treatment of severe colitis in children. J Pediatr 2000; 137:794–799.
9. Baumgart DC, Wiedenmann B, Dignass AU. Rescue therapy with tacrolimus is effective in patients with severe and refractory inflammatory bowel disease. Aliment Pharmacol Ther 2003; 17:1273–1281.
10. Hogenauer C, Wenzl HH, Hinterleitner TA, et al. Effect of oral tacrolimus (FK 506) on steroid-refractory moderate/severe ulcerative colitis. Aliment Pharmacol Ther 2003; 18:415–423.
11. Mamula P, Markowitz JE, Cohen LJ, et al. Infliximab in pediatric ulcerative colitis: two-year follow-up. J Pediatr Gastroenterol Nutr 2004; 38:298–301.
12. Reding R. Tacrolimus in pediatric liver transplantation. Pediatr Transplant 2002; 6:447–451.
13. Farmer DG, McDiarmid SV, Yersiz H, et al. Outcomes after intestinal transplantation: a single-center experience over a decade. Transplant Proc 2002; 34:896–897.
14. Younes BS, McDiarmid SV, Martin MG, et al. The effect of immunosuppression on posttransplant lymphoproliferative disease in pediatric liver transplant patients. Transplantation 2000; 70:94–99.
15. MacFarlane GD, Venkataramanan R, McDiarmid SV, et al. Therapeutic drug monitoring of tacrolimus in pediatric liver transplant patients. Pediatr Transplant 2001; 5:119–124.

Steroid-dependent ulcerative colitis; Steroid-resistant ulcerative colitis; Tacrolimus; Ulcerative colitis

© 2007 Lippincott Williams & Wilkins, Inc.