Tacrolimus (FK-506) Treatment of Fulminant Colitis in a Child

Immunosuppressive drugs are increasingly being used as a therapeutic alternative to colectomy in patients with ulcerative or Crohn's colitis (1,2). A recent double-blind controlled trial demonstrated the efficacy of i.v. cyclosporine A over placebo in postponing colectomy in adult patients with severe ulcerative colitis (3). In addition, oral cyclosporine has been utilized in the pediatric population to treat fulminant colitis (4). However, cyclosporine therapy has many potential disadvantages. Use of cyclosporine in both inflammatory bowel disease and liver transplantation has been associated with hypertension, nephrotoxicity, hirsutism, and coarsening of facial features (5). In addition, most protocols for treatment of fulminant colitis with cyclosporine recommend i.v. infusion (3). Oral absorption of cyclosporine A is erratic and highly dependent on both bile flow and intestinal mucosal integrity (6,7). Also, published reports of cyclosporine use in fulminant colitis suggest that, although cyclosporine may help postpone colectomy, the rate of relapse is high once cyclosporine is discontinued, and the majority of patients ultimately undergo colectomy. The macrolide immunosuppressant tacrolimus (FK-506) has been utilized successfully in treating cyclosporine-refractory liver rejection in the liver transplant population, and oral tacrolimus is the drug of choice in patients undergoing liver bowel transplantation (8). Potential advantages of tacrolimus over cyclosporine include its greater potency and more reliable oral absorption, even in the face of bowel inflammation or cholestasis (9). In addition, some investigators suggest that initial use of tacrolimus may be superior in preventing rejection in solid organ transplant patients, though this finding has been disputed (10). The side effects of tacrolimus are similar to those of cyclosporine, but some adverse effects (including hypertension, gingival hyperplasia, and coarsening of facial features) may be less frequent with tacrolimus use. Tacrolimus has been noted to improve symptoms of inflammatory bowel disease in the majority of patients who received the drug because they had liver transplants for primary sclerosing cholangitis (11). In addition, open-label trials of tacrolimus for autoimmune diseases (including pediatric nephrotic syndrome and autoimmune chronic active hepatitis) suggest the drug is efficacious (12-14). Because of these findings, we have now begun to evaluate the effects of tacrolimus in pediatric patients with fulminant colitis. We use oral tacrolimus as a short-term induction agent in patients with fulminant colitis who are facing the prospect of colectomy; we then place patients on 6-mercaptopurine and stop tacrolimus after a 3-month period. This article reports our experience with this drug in our first pediatric patient undergoing this regimen. CASE REPORT A 6-year-old girl presented to Children's Hospital (Boston, MA, U.S.A.) with a severe exacerbation of colitis. She had been diagnosed with colitis 1½ years earlier after presenting with rectal bleeding. An upper GI examination with small bowel follow-through performed at the time of initial diagnosis demonstrated no evidence of small bowel disease. The patient responded to a short course of prednisone and was subsequently treated with olsalazine, which was continued for a year. The olsalazine was discontinued 6 months prior to this presentation at Children's Hospital; she had remained asymptomatic for 5 months. One month prior to this admission, she had presented with bloody diarrhea and olsalazine therapy had been reinstituted. When she did not improve, prednisone was begun. In spite of prednisone therapy, she continued to have abdominal cramping and blood in her stool. At the time of admission, the patient was having 15 bowel movements per day, all of which were liquid and contained large amounts of blood. Her initial hematocrit was 18%, and she was transfused with packed red blood cells. Her initial albumin was 2.0 g/dl. The sedimentation rate and white blood count were elevated, but other laboratory studies were within normal limits (Table 1). Intravenous methylprednisolone, 2 mg/kg/day, was begun. Despite 10 days of steroid therapy, the patient continued to have 15 bowel movements per day with large amounts of blood and also had persistent abdominal cramping. Sigmoidoscopy demonstrated relative sparing of her rectum, with deep linear ulcerations and nodularity of the rectosigmoid. Colonic biopsies confirmed the presence of focal colitis with some rectal sparing. The patient was therefore given the diagnosis of indeterminate colitis, probably Crohn's colitis. The patient's parents were given the option of either surgery or immunosuppression and opted for tacrolimus treatment. Our protocol was approved by our institutional review board, and informed consent was obtained. The starting tacrolimus dose was 2 mg every 12 h (0.12 mg/kg/dose b.i.d.). Because of the immunosuppression used, the patient was also given dapsone, 1 mg/kg/day, for Pneumocystis prophylaxis (she was allergic to sulfa antibiotics). The whole-blood tacrolimus levels were measured using the Abbott Tacrolimus Assay, and the dose was decreased to maintain a level of ≈10-15 ng/ml. After 1 week of treatment with tacrolimus, the patient's stool output decreased from 600 to 250 cc of bloody stool per day. Her transfusion requirement also decreased from 1 U/3 days to 1 U/10 days. By 2 weeks of therapy, the stool output had decreased to 150 cc/day, the stool was more formed, and the patient required no further transfusions. Bowel movement frequency had decreased to five bloody stools per day, and albumin infusions were no longer necessary. After 3 weeks of tacrolimus therapy, the patient was having three or four daily bowel movements with ≈50-60 cc of semiformed stool containing small amounts of blood. Enteral feedings were begun during the 3rd week of hospitalization, and the patient was ultimately discharged home after a 1-month hospitalization. Clinical and laboratory data are summarized in Table 1. In addition to standard clinical and laboratory parameters, we measured serum interleukin-6 levels during the course of therapy. Interleukin-6 is produced by gut mononuclear cells and is elevated in active Crohn's disease. Serum interleukin-6 levels peaked 5 days after the start of tacrolimus therapy at 15.1 pg/ml and ultimately declined to <2 pg/ml at the time tacrolimus was discontinued. We also measured peripheral blood activated T lymphocytes by flow cytometry; CD3+CD25+ T cells were 12% of the peripheral blood mononuclear cell (PBMC) population 3 weeks after the start of therapy and declined to 7% after 2 months of tacrolimus treatment. The patient tolerated the medication well. She had mild hypertension, which resolved when her steroid dose was consolidated to once daily. Her serum creatinine was 0.2 mg/dl at the start of therapy; during her 4-week hospitalization, it remained 0.2 mg/dl. The patient developed a mild hand tremor after 3 weeks of therapy, which resolved at 8 weeks. A steroid taper was begun after 4 weeks of therapy, and 6-mercaptopurine (1.5 mg/kg/day) was begun at 6 weeks of treatment. After 90 days of tacrolimus treatment, the tacrolimus and dapsone were stopped. Steroids were weaned to alternate day by 4 months and stopped at 6 months. After 9 months, the patient remains in remission on 6-mercaptopurine (1.5 mg/kg/day) and olsalazine (250 mg twice a day). DISCUSSION It is now acknowledged that inflammatory bowel disease is characterized by activation of the gut mucosal immune system. Studies of both ulcerative colitis and Crohn's disease have demonstrated increased levels of interleukin-1, interleukin-2, and interleukin-6 in bowel mucosa. In addition, activated T cells expressing high-affinity interleukin-2 receptors have been identified in the inflamed bowel (15,16). These findings have led to the development of more aggressive immunomodulatory therapy for patients with steroid-refractory colitis. Most recently, the medication cyclosporine has been studied as a medical alternative for patients with severe colitis who wish to postpone colectomy. Cyclosporine is usually administered as a continuous i.v. infusion for colitis (3). In one reported series of pediatric patients with severe colitis, Treem et al. utilized oral cyclosporine at a dose of 4 mg/kg/day (4). Though five of six patients were discharged without colectomy, four ultimately required colectomy. We are evaluating a different immunosuppressive strategy for pediatric fulminant colitis. This regimen involves the use of tacrolimus as an initial induction agent to treat fulminant colitis, followed by a gradual transition to 6-mercaptopurine for maintenance of remission. It is hoped that by use of this regimen, we can minimize the toxicities associated with long-term therapy with tacrolimus or cyclosporine. In addition, the tacrolimus offers the potential advantages of more reliable oral absorption in the face of bowel inflammation, decreased need for i.v. access in children, and potentially greater efficacy. Although we have treated only one patient thus far, based on our results with this patient we plan to use this regimen to treat additional children with fulminant colitis in the future. In addition to conventional clinical and laboratory measures used to monitor disease activity and response to therapy, we followed interleukin-6 levels during the patient's clinical course. Serum interleukin-6 (IL-6) is a cytokine produced by activated macrophages and T lymphocytes; it is important in promoting mucosal inflammation and in stimulating production of acute-phase reactants by the liver. Mahida et al. and Hyams et al. noted increased serum IL-6 in patients with Crohn's disease and a correlation between serum IL-6 and severity of disease activity (17,18). Tacrolimus blocks cytokine synthesis by lymphocytes through inhibition of transcription of cytokine messenger RNA (9). Therefore, declining serum IL-6 levels may reflect the effects of tacrolimus on systemic and mucosal lymphocytes. In addition, since IL-6 is important in promoting lymphocyte activation and differentiation, and activated T lymphocytes are increased in the peripheral blood of patients with Crohn's disease, we measured the percentage of activated T cells at two points during therapy (19). Though it was gratifying to see decreases in both IL-6 and activated T cells during treatment, these tests did not influence our decision making during therapy; the successful response was ultimately determined by standard clinical parameters. The toxicity of tacrolimus in pediatric and adult solid organ transplant recipients had been cited as equivalent to that of cyclosporine (11,20,21). Side effects include hypertension, nausea, transaminase elevation, opportunistic infections, nephrotoxicity, diabetes mellitus, and seizures. Acute nephrotoxicity and neurotoxicity appear related to the i.v. use of tacrolimus and to levels exceeding the therapeutic index. Some side effects, including hypertension, hypercholesterolemia, hirsutism, and coarsening of facial features, occur less frequently with tacrolimus. Opportunistic infections are a potential risk: to prevent Pneumocystis pneumonia, we utilized prophylactic dapsone. We also obtained Epstein-Barr virus serologies prior to the institution of therapy. Epstein-Barr virus-associated lymphoproliferative disease (LPD), which can occur with both cyclosporine and tacrolimus use, is of particular concern given its potentially life-threatening nature (22). Though an initial study of pediatric solid organ transplant recipients suggested equivalent rates of LPD in patients receiving cyclosporine A and tacrolimus (13), McDiarmid et al. (23) reported a higher incidence of LPD in patients receiving tacrolimus. Of note, the patients in the McDiarmid et al. study had all received aggressive antirejection immunosuppression, including OKT3 and high-dose pulse steroids; therefore, LPD may have been a function of the extent of immunosuppression and organ rejection rather than the tacrolimus therapy. In addition, tacrolimus has been utilized to treat other autoimmune diseases, including chronic active hepatitis and pediatric nephrotic syndrome, without any reported cases of LPD. As discussed, the potential side effects of long-term immunosuppressive therapy with cyclosporine or tacrolimus are considerable. In addition, the degree of toxicity and immunosuppression may be dose- and level-dependent. We initially aimed for a whole-blood trough level of 15 ng/ml, but when clinical improvement occurred we reduced the target level to 5-7 ng/ml. It is of interest that, as the patient's colitis improved and oral intake was initiated, the tacrolimus level dropped from 15 to 5 ng/ml without any dose adjustments. The lower levels were maintained for the last 2 months of therapy with no evidence of worsening disease activity. Our patient's response time to tacrolimus was longer than that reported in cyclosporine-treated patients, who have a mean response time of 7 days (3). It is unclear whether the slower response time was due to the severity of this particular patient's disease, or whether tacrolimus will ultimately be found to have a slower onset of action than cyclosporine in comparative studies. The facts that this patient had probable Crohn's colitis, which could have necessitated a permanent ileostomy and that she tolerated the drug without significant toxicity encouraged the patient's physicians and family to continue with therapy while the patient demonstrated slow, steady improvement. The patient demonstrated a significant improvement by 14 days and now passes one or two guaiac-negative bowel movements per day. In conclusion, oral tacrolimus offers a potential alternative to cyclosporine in patients with steroid-resistant fulminant colitis and offers the potential advantages of more reliable oral absorption and greater ease of administration. We are currently enrolling similar patients in a pilot study to examine the efficacy and toxicity of tacrolimus in the treatment of fulminant colitis. Ultimately, a randomized trial between tacrolimus and cyclosporine may best determine which agent has greater efficacy and less toxicity. Note added in proof: The patient described herein remains in clinical remission as of May, 1996, 18 mos. after the initial presentation. A second patient with fulminant ulcerative colitis has also been treated in a successfully similar manner. Acknowledgment: We thank Karen Keough for performing the interleukin-6 assays and the Clinical Immunology Laboratory at Children's Hospital for performing the peripheral blood flow cytometry. Tacrolimus (FK-506) was supplied by Fujisawa Pharmaceuticals. Dr. Bousvaros was supported by NIH GCRC CAP Award 2MO1 RR 02172.