|
|
| |||||||||||||||||||||||||||||||||||||||
Background There has been no new effective drug therapy for patients with severe ulcerative colitis since corticosteroids were introduced almost 40 years ago. In an uncontrolled study, 80 percent of 32 patients with active ulcerative colitis refractory to corticosteroid therapy had a response to cyclosporine therapy.
Methods We conducted a randomized, double-blind, controlled trial in which cyclosporine (4 mg per kilogram of body weight per day) or placebo was administered by continuous intravenous infusion to 20 patients with severe ulcerative colitis whose condition had not improved after at least seven days of intravenous corticosteroid therapy. A response to therapy was defined as an improvement in a numerical symptom score (0 indicated no symptoms, and 21 severe symptoms) leading to discharge from the hospital and treatment with oral medications. Failure to respond to therapy resulted in colectomy, but some patients in the placebo group who had no response and no urgent need for surgery were subsequently treated with cyclosporine.
Results Nine of 11 patients (82 percent) treated with cyclosporine had a response within a mean of seven days, as compared with 0 of 9 patients who received placebo (P<0.001). The mean clinical-activity score fell from 13 to 6 in the cyclosporine group, as compared with a decrease from 14 to 13 in the placebo group. All five patients in the placebo group who later received cyclosporine therapy had a response.
Conclusions Intravenous cyclosporine therapy is rapidly effective for patients with severe corticosteroid-resistant ulcerative colitis.
Cyclosporine selectively inhibits immune responses mediated by T lymphocytes and has proved effective in patients with chronic corticosteroid-resistant Crohn's disease5. In an uncontrolled study approximately 80 percent of patients with severe ulcerative colitis refractory to treatment with corticosteroids responded to cyclosporine therapy6,7. We undertook this study to evaluate further the efficacy and safety of intravenous cyclosporine therapy in patients with severe ulcerative colitis.
Methods
Patient Selection and Eligibility
From April 1991 to April 1992, we studied 20 patients ranging from 18 to 65 years of age who had severe active ulcerative colitis and were admitted to Mt. Sinai Hospital in New York or the University of Chicago Hospital. Patients were eligible to enter the study if they had had no response to intravenous corticosteroid therapy (equivalent to a daily dose of 300 mg of hydrocortisone) after seven or more days. Patients who were transferred to either institution after having no response to this therapy elsewhere were also eligible for the study. Patients with a relapse of active disease after a recent hospitalization, during which they had responded to intravenous and then oral corticosteroid therapy, were also eligible if they had no response to an additional 60 hours of intravenous corticosteroid therapy. All the patients had a score of 10 or higher on a clinical-activity index used in an earlier open trial of cyclosporine,6,7 in which we found that patients with scores below 10 after treatment could leave the hospital without surgery (Table 1).
|
Drug Administration and Monitoring
The patients assigned to receive cyclosporine were given a dose of 4 mg per kilogram of body weight per day by continuous infusion for up to 14 days. The patients assigned to the placebo group received an identical-appearing intravenous solution of cremaphor and alcohol.
At each institution one physician who was aware of the patients' treatment assignments monitored blood cyclosporine concentrations and adverse effects. The dose of cyclosporine never exceeded 4 mg per kilogram per day, but it was reduced if the serum creatinine concentration increased by 30 percent above base line, serum liver-enzyme values increased by 50 percent, or diastolic blood pressure consistently exceeded 90 mm Hg despite antihypertensive therapy. Blood cyclosporine concentrations were measured every other day. In Chicago, concentrations of 100 to 400 ng per milliliter, as determined by radioimmunoassay with a monoclonal antibody (Sandoz, Hanover, N.J.), were considered therapeutic; in New York, concentrations of 400 to 800 ng per milliliter, as determined by radioimmunoassay with a polyclonal antibody (Sandoz), were considered therapeutic. Although values obtained with the latter assay were used for dose adjustments in New York, the concentrations in those samples were also measured with the monoclonal-antibody assay to confirm the correlation and are reported here. More frequent measurements were made if the cyclosporine concentrations exceeded these ranges or if the monitoring physician, who also randomly adjusted the dosages of placebo, deemed them necessary.
Concurrent Therapy
All patients received 100 mg of hydrocortisone intravenously every eight hours and hydrocortisone enemas (100 mg in a total volume of 60 ml) nightly if the drug could be retained. Patients receiving mesalamine enemas before entry continued to receive them if the drug could be retained. Likewise, oral sulfasalazine, olsalazine, or mesalamine was continued in the same doses in patients already taking these medications. Therapy with topical or oral mesalamine or related drugs was not initiated during the study.
Patients who were already taking antibiotics continued to receive them if clinically indicated. Antibiotic therapy was not initiated during the study unless indicated -- for example, for pulmonary or urinary tract infection.
The patients were treated with loperamide or codeine in an attempt to control diarrhea; the use of these drugs was accounted for in the clinical-activity score. The dosage of these drugs was not increased once the patient entered the trial. Antihypertensive drugs were continued or initiated, as indicated, if the diastolic blood pressure consistently exceeded 90 mm Hg. Acetaminophen, H2-receptor antagonists, or aluminum-based antacids were given as needed. Three patients were receiving total parenteral nutrition when they entered the study, but it was not initiated in any patient during the study. Oral intake of food was allowed at the discretion of the gastroenterologist, who was unaware of the patients' treatment assignments.
Patient Monitoring and Study End Points
The patients were evaluated daily by a gastroenterologist who was aware of their treatment assignments and by at least one gastroenterologist who was not aware of their treatment assignments or the results of any laboratory studies. The latter gastroenterologist determined the need for concomitant medications but did not ask the patients about adverse drug effects or review blood-pressure readings.
Clinical-activity scores (Table 1) were determined daily by the physicians after they had reviewed with the patient his or her daily log of the number of bowel movements (trips to the bathroom), the number of nocturnal bowel movements (ones that awakened the patient from sleep), and the incidence of blood in the stool, incontinence, and abdominal pain or cramping. The physicians also determined whether abdominal tenderness was present, decided whether antidiarrheal drugs were needed, and asked about the patient's well-being. The maximal clinical-activity score was 21. If there was disagreement in scoring between physicians, the admitting gastroenterologist's score was used. All the patients were also evaluated daily by a surgeon who was unaware of their treatment assignments. The surgeon could recommend at any time that a patient discontinue the treatment and undergo colectomy, on the basis of clinical findings.
Vital signs were assessed four times daily. All patients had frequent abdominal x-ray films obtained to exclude perforation or megacolon. A complete blood count and biochemical studies, including tests of renal and hepatic function, were done every other day. Sigmoidoscopy or colonoscopy was not performed during the study.
The primary end point was the clinical-activity score. A score of less than 10 on two consecutive days was considered to indicate a positive response to therapy. The score on the second of these two days was considered the final score. In patients who had a response, therapy was changed to 60 mg of oral prednisone daily and either oral cyclosporine (6 to 8 mg per kilogram per day) or oral placebo. If the response was maintained for an additional two days, the patient was allowed to go home while continuing to take these medications.
Patients whose clinical-activity scores did not fall below 10 for 2 consecutive days after 14 days of treatment or whose condition worsened were considered to have no response to treatment. Depending on the severity of their colitis as determined by the gastroenterologist and surgeon, they either underwent colectomy or were offered open-label cyclosporine therapy, administered by continuous intravenous infusion in a dose of 4 mg per kilogram per day for a maximum of 14 days (after they had withdrawn from the trial; the treatment code was not broken).
Statistical Analysis
Quantitative variables were compared with two-tailed Student's t-tests. Qualitative variables and differences between centers were compared with chi-square analysis with Yates' correction. All patients were assessed on an intention-to-treat basis.
The trial was terminated after 20 patients had been studied, when the physician who was aware of their treatment assignments noted a significant difference between the two groups. This was confirmed by the study monitor and two independent reviewers.
Results
Characteristics of the Patients
Among the 20 patients, 11 received cyclosporine and 9 placebo. The mean age, duration and extent of disease, initial clinical-activity score, duration of corticosteroid therapy before the study, number of patients requiring parenteral nutrition, and number requiring concomitant medications before and during the trial were similar in the two groups (Table 2). Fourteen patients were studied in New York, and six in Chicago.
|
|
No patient in the placebo group had a decline in the clinical-activity score to below 10 on two consecutive days. Four of the nine patients (44 percent) underwent colectomy. One underwent colectomy on day 3 because of toxic megacolon. Another underwent colectomy after clinical deterioration in her condition was noted; this patient later died of gram-negative sepsis with superimposed cytomegalovirus infection. Two patients had surgery for refractory symptoms. The condition of the other five patients (56 percent) was stable, and they were therefore given open-label intravenous cyclosporine after the study period.
During this period of open-label cyclosporine therapy, the evaluating physicians remained unaware of the patients' initial treatment assignments and no other treatment was introduced. The condition of all five patients who had received placebo earlier improved, with a decrease in their mean clinical-activity score from 11 (range, 11 to 13) to 7 (range, 2 to 9). The mean time to a response in this subgroup was seven days (range, four to eight).
No characteristics were identified that predicted a response to therapy in either the initial or the crossover phase of the study. All 14 patients with a response, except the 1 who chose to undergo colectomy, were treated with oral cyclosporine and discharged from the hospital 48 hours later.
Blood Cyclosporine Concentrations
Among the 11 patients who initially received cyclosporine, the mean blood concentration was 482 ng per milliliter (range, 339 to 653) in the 9 who had a response to therapy and 484 ng per milliliter in 1 who had no response (the other patient who had no response received cyclosporine for only 12 hours). In the five patients who received cyclosporine after receiving placebo, the mean blood cyclosporine concentration was 524 ng per milliliter (range, 375 to 620). There was no correlation between blood concentrations and the rapidity of response. The dosage was decreased in five patients because of elevated blood cyclosporine concentrations.
Adverse Effects
Four of the 11 patients initially treated with cyclosporine had paresthesias (36 percent), as compared with none of the patients in the placebo group. Hypertension, defined as a systolic blood pressure of more than 140 mm Hg or a diastolic blood pressure of more than 90 mm Hg for two consecutive days, was noted in four patients in the cyclosporine group (36 percent), two of whom required treatment. Hypertension developed in one patient in the placebo group (11 percent). One patient in each group reported nausea and vomiting. None of the patients had nephrotoxicity or hepatotoxicity. As noted above, one patient treated with cyclosporine had a grand mal seizure after the initiation of therapy but had no more seizures after cyclosporine was discontinued. Headaches occurred as the only side effect in two of the patients who received cyclosporine after receiving placebo.
Discussion
For many years the standard treatment for severe ulcerative colitis has been intravenous corticosteroids followed by colectomy if there is no response9. No substantial improvement in response rates has been reported with any new treatment regimen in the past 20 years10. The recognition that cyclosporine has immunosuppressive properties has led to its use in patients with severe ulcerative colitis, with variable results. A review of uncontrolled studies found that 80 percent of patients had a response to intravenous cyclosporine, whereas only 60 percent had a response to oral cyclosporine11. In our study, 82 percent of the patients treated with cyclosporine initially and 100 percent of those who had no response to placebo and who were then treated with cyclosporine had a response to the therapy. Our rationale for administering cyclosporine intravenously was based on studies documenting poor absorption of orally administered cyclosporine in patients with diarrhea and various inflammatory diseases of the bowel12. In addition, when cyclosporine is administered by continuous intravenous infusion, blood concentrations of the drug are constant6,7. Furthermore, colonic-tissue concentrations of cyclosporine are 10 times higher in normal subjects given cyclosporine parenterally than in those given the drug orally13. Although there is no proved correlation between blood and colonic-tissue concentrations of cyclosporine, it is probably the colonic concentration of cyclosporine that determines the clinical response14.
We believe that the intravenous route in this therapeutic dose range is more appropriate for short-term use in patients with severe ulcerative colitis than the lower therapeutic dose range and oral route used in patients with other more chronic autoimmune diseases15. The starting dose used in this study -- 4 mg per kilogram per day -- was chosen because of its efficacy and low level of toxicity in our open-label trial6,7. However, Actis et al.16. recently reported success in treating three patients with severe, ulcerative colitis refractory to treatment with steroids with an intravenous dose of 1 to 2 mg of cyclosporine per kilogram per day.
In our study the evaluation of the efficacy of the drug was based on a clinical-activity index that included both objective elements (number of stools, number of nocturnal bowel movements, frequency of visible blood in stool, frequency of incontinence, and the need for antidiarrheal drugs) and subjective elements (degree of pain, degree of cramping, assessment of general well-being, and presence of abdominal tenderness). When two or more physicians calculated the index, there was never a difference of more than one point in their scores. The similar rates of improvement in both the open-label and the double-blind trial suggest the validity of this index.
The decision to terminate a clinical trial requires consideration of three issues: ethics, clinical practice, and biostatistics17. After 20 patients had been enrolled, the study monitor and two independent reviewers found a significant difference in the response rates between the cyclosporine and placebo groups. All three agreed it was unethical to continue to enter patients into the placebo arm. From a clinical standpoint, it was clear that the use of cyclosporine obviated the need for colectomy and that the patients in the placebo arm fared poorly. Finally, from a biostatistical standpoint, although an interval analysis had not been planned, the finding of a P value of less than 0.001 made it extremely unlikely that entering additional patients would change the results18.
The response to cyclosporine was rapid; the mean time to a response was seven days in the cyclosporine group and in the crossover cyclosporine group. Although no predictive factors were identified, a decrease in bleeding was the first sign of clinical improvement. The response time was shorter than in other studies, possibly reflecting the use of continuous intravenous infusions.
The greatest concern associated with long-term cyclosporine treatment is nephrotoxicity19,20. In this trial there was no renal toxicity because of the strict exclusion criteria, the relatively young age of the patients, the short duration of treatment, and the frequent measurement of blood cyclosporine concentrations.
Twenty percent of recipients of kidney and liver transplants who are treated with cyclosporine have neurologic side effects. These include paresthesias, which were noted in four study patients, and seizures, which occurred in one patient. The seizures may be triggered by hypomagnesemia or hypocholesterolemia, both of which allow easier diffusion of cyclosporine across the blood-brain barrier21. The patient who had a seizure also had hypocholesterolemia.
Several patients became hypertensive and responded to treatment with calcium-channel blockers. There were no infectious complications in the cyclosporine-treated patients.
Almost 60 percent of the patients in our open trial maintained their clinical response six months after discharge, and many were subsequently able to discontinue corticosteroid therapy and had endoscopic evidence of healing7. Patients in this controlled trial continue to be monitored after discharge from the hospital to verify our initial long-term observations.
Since patients with ulcerative colitis have a high incidence of colon cancer, there is concern that immunosuppression may increase the risk of this complication22. No colon cancers have been reported in patients with ulcerative colitis who were treated with cyclosporine for up to 8 months and in whom the mean follow-up was 4.5 years (unpublished data). We conclude that intravenous cyclosporine therapy is effective and safe in severely ill patients with ulcerative colitis that has not responded to intravenous corticosteroid therapy.
We are indebted to Mr. Lawrence Kupferschmidt, Dr. Thomas Twaddell, Dr. David Sachar, Dr. Thomas Chalmers, and Dr. John Singleton for their help and to Ms. Carol Chiofolo-Stamaty and Ms. Debbie K. James for their assistance in carrying out this study.
Source Information
From the Department of Medicine, Division of Gastroenterology (S.L., D.H.P., A.K.), and the Department of Surgery (I.G., J.B.), Mt. Sinai School of Medicine, New York; and the Department of Medicine, Division of Gastroenterology (G.G., S.H.), and the Department of Surgery (F.M.), University of Chicago Medical Center, Chicago.
Address reprint requests to Dr. Lichtiger at 1185 Park Ave., New York, NY 10128.
References
| |||||||||||||||||||||||||||||||||||||||
Related Letters:
Lithium in the Treatment of Mood Disorders
Stern R., Price L. H., Heninger G. R.
Extract |
Full Text
N Engl J Med 1995;
332:127-128, Jan 12, 1995.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
