Case Presentation

Introduction

Crohn's Disease Activity Index

The First RCT

Contradictory Data

A Meta-Analysis

Azathioprine for Quiescent Disease

Ulcerative Colitis

Safety

Future Therapies

Conclusions

Mr. X is a 32-year-old white male with no significant past medical history except Crohn's disease. He presented to GI clinic for evaluation. He had been diagnosed with Crohn's disease at age 23 and required urgent resection of his terminal ileum. Since then, he experienced a number of exacerbations and remissions of Crohn's disease requiring frequent and prolonged steroids use. 5-aminosalicylic acid had not been beneficial. He began taking azathioprine (Imuran) six months earlier in an effort to reduce his steroid dose. He had done well since then and had been weaned from steroids two months ago.

Upon questioning, the patient stated that he felt moderately well, occasionally had mild abdominal pain, and had arthralgias in his hands and knees. He had 3 soft stools daily but did not require opiates for diarrhea.

Medications included azathioprine 2.5 mg/kg/day, folate 1 mg qday, B12 1000 mcg IM qmonth.

On exam, the patient weighed 5% less than his baseline weight. In general he looked well. His abdomen had normal bowel sounds, soft, nontender, and had no abdominal masses. He was guaiac negative.

Laboratory data showed a hematocrit of 36.

He was judged to be in remission.

The patient was enthusiastic about being off steroids but worried about relapse.

Some of the clinical questions pertinent to this patient:

1. What is the efficacy of azathioprine in maintaining remission in Crohn's disease?
2. What toxicities are associated with azathioprine?

Additional questions pertinent to all patients with IBD:

1. What is the role of immunosuppressant medications in treating active and quiescent Crohn’s disease and active and quiescent ulcerative colitis?
2. What data is there to support the use of immunosuppressants?
3. What effects do immunosuppressants have on steroid consumption and fistulae?
4. What toxicities are associated with immunosuppressants?
5. When should immunosuppressants be used?
6. What monitoring is required when using immunosuppressants?

Inflammatory Bowel Disease (IBD) encompasses the two diseases known as Crohn's Disease (CD) and Ulcerative Colitis (UC). Remissions and recurrences characterize the nature of IBD. The cause of IBD remains unknown but the disease is characterized by inflammation of the gastrointestinal tract. Treatments for IBD have targeted the inflammatory cascade, aiming to interrupt inflammation and thus decrease disease activity. To a limited extent, topical and systemic corticosteroids and topical 5-aminosalicylic acid (5-ASA) have controlled many patients' disease. However in the course of 1 year, 50% of patients with IBD will require steroids for treatment. Of these patients, 50% will be refractory to steroids or dependent on steroids in one year.¹ Such patients are candidates for immunosuppressant therapy.

Treatment of IBD with 6-mercaptopurine (6-mercaptopurine, Purinethol) and azathioprine (azathioprine, Imuran) has been studied extensively over the past 30 years. Although initially controversial, mounting evidence has demonstrated the efficacy and safety of these agents in inducing and maintaining a remission of IBD. This paper will review the pertinent pathophysiology, pharmacokinetics, adverse effects, and therapeutics trials of these immunosuppressants as well as briefly examine use of cyclosporin and methotrexate.

The initial anti-inflammatory agent used to treat IBD was prednisolone. Corticosteroids were shown to be effective in active IBD in 1956. ² Steroids affect the immune system in many complex ways and are associated with numerous side effects including infections, glucose intolerance, depression, osteoporosis, cataracts, and weight gain. Alternative immunosuppressants could theoretically suppress IBD activity while avoiding some of the side effects associated with steroids.

Azathioprine is an antimetabolite that is converted in the body to 6-mercaptopurine. Azathioprine is 55% 6-mercaptopurine by molecular weight and 88% of azathioprine is converted to 6-mercaptopurine making a dose of 2 mg of azathioprine equivalent to 1 mg of 6-mercaptopurine. After absorption, azathioprine is converted to 6-mercaptopurine which is then converted to thioinosinic acid. 6-mercaptopurine resembles guanine, a purine used to make DNA and RNA. An altered purine analog, thioinosinic acid inhibits purine synthesis and metabolism, blocking DNA and RNA synthesis. Thioinosinic nucleotides act by inhibiting early immune response during the proliferative cycle of T or B lymphocytes^{3, 4}. Blocking the inflammatory response is thought to be the mechanism by which antimetabolites work in IBD.

In order to adequately measure the effect of a drug on CD, the severity of CD must be quantified. This has been difficult because heterogeneous nature of CD. Evaluation requires the combination of subjective interpretations by patient and physician as well as objective data. The Crohn’s Disease Activity Index (CDAI) has correlated with the physician’s assessment of the patients well-being and has been repeatedly validated and currently is the best and most universally accepted assessment of Crohn’s disease activity. This index was first published in 1979. All studies cited that were published after 1980 used the CDAI to quantify disease activity. Studies published at or before 1980 used similar indices. A familiarity with the variables measured is helpful in understanding the therapeutic goals of the following trials.

The score on the index is calculated by multiplying the numerical value for each variable by the multiplication factor shown and totaling the resulting values. A score below 150 indicates remission; a score above 450 indicates severe illness. Based on the index of Best et al.⁵

For example, our patient who had 3 soft stools/day, mild abdominal pain, felt moderately good, had arthralgias but no other complications, did not used opiates for diarrhea, had no abdominal mass, had Hct of 36, and weighed 5% less than his standard weight would have a CDAI of 109. He is in remission.

Alternatively, a hypothetical male patient who had 5 soft stools/day, moderate abdominal pain, felt poor, had arthralgias, a perianal fistula, recent fever, used opiates for diarrhea, had a questionable abdominal mass, had Hct of 30, and weighed 10% less than his standard weight would have a CDAI of 270. He is having active disease. With a knowledge of how CD activity is measured, one can understand the efficacy of therapies.

The First RCT: Willoughby et al.⁶

Azathioprine was first used for CD in 1969.⁷ Initial reports of success lead to two randomized controlled trials (RCTs) in 1971. Willoughby et al. randomized 22 hospitalized patients to received prednisolone and azathioprine or prednisolone and placebo. 12 were having active disease (Group I) and 10 were in remission but requiring steroids (Group 2). The prednisolone was begun at 60 mg/day and then was tapered to 10 mg/day over 3 weeks and if improvement was maintained, prednisolone was discontinued. In the study group, azathioprine was given in 4 mg/kg for 10 days and then 2 mg/kg thereafter. The groups were studied for 24 weeks (6 months). An index of disease activity similar to CDAI was used to evaluate efficacy. The possible index scores ranged from 0 to 38.

Ten of eleven patients (91%) receiving azathioprine were in clinical remission at the end of the study. In azathioprine patients, the mean disease activity index changed in Group I patients’ from 10.7 to 2.3 and in Group II from 3.2 to 4.4. Only 3 of the eleven patients (27%) on placebo completed the study in remission. 8 patients had a relapse of disease. The placebo group's mean disease activity index changed in Group I from 7.8 to 8.0 and in Group II from 1.8 to 8.6.

A steroid sparing effect was seen in the azathioprine group. Azathioprine patients took prednisolone an average of 14 weeks while the placebo group took prednisolone an average of 20 weeks.

Leukopenia was noted in 2 azathioprine patients. This did not require cessation of the drug.

This small study by Willoughby et al. showed three important observations that would later be studied in greater detail by other investigators:

1. Azathioprine plus prednisolone was better than prednisolone alone in active and quiescent CD.
2. Patients on azathioprine required less corticosteroids than their placebo counterparts.
3. Azathioprine causes reversible leukopenia.

Two months after the promising study by Willoughby was published, Rhodes et al.⁸ published a randomized placebo-controlled cross-over trial of 16 patients which showed that only 2 patients improved on azathioprine and 14 were either unchanged or worse while taking the drug. Patients took azathioprine and placebo for 2 months in each phase of the crossover trial. They concluded that azathioprine was ineffective and should only be used after all other measures had failed.

A few small RCTs and more uncontrolled observational studies were performed between the 1974 and 1994 which showed mixed results.

However no studies were of sufficient size to definitively decide the debate.

Although the medical community had 25 years of experience with azathioprine and 6-mercaptopurine, their role as therapeutic agents remained controversial because they were slow to act and potentially caused serious side effects. None of the smaller studies were able to decide the issue of efficacy. A large randomized controlled trial is the best way to evaluate questions regarding medical therapy but the time and number of patients required can be prohibitive. A meta-analysis of randomized controlled trials attempts to combine the results from previously published studies to increase the statistical power of the results. Pearson et al. performed such a meta-analysis of all published randomized controlled trials and published their results in Annals of Internal Medicine in 1995.

To qualify for analysis, a study had to be a double-blinded, placebo-controlled, randomized trial in patients > 18 year old and published in English or French. Studies were found by doing a MEDLINE search, a manual search of the references in RCTs and review articles, and a manual review of major gastroenterology conferences to ensure that no study was missed. Studies of both active and quiescent disease were evaluated. Each study was reviewed independently by three authors and methods and results were recorded. All results were reported on an intention-to-treat basis. Studies were also assessed for quality and given a quality score by two independent authors.

The data was analyzed with a logistic regression model that allowed for complex situations, including incorporation of differences between the trials in duration, drug dose, and cumulative dose. Some of the studies were a source of heterogeneity. This modeling accounted for differences between studies on the basis of variables that could be measured, quantified and reported. The models showed that the data was sufficiently homogenous to safely combine the data and draw conclusions.

A sensitivity analysis was also performed by excluding each study separately and analyzing the data to ensure that no single study was solely responsible for statistical significance.

Four studies focused on active disease only and three additional studies had active-disease arms in multi-arm studies. Patients could use 5-aminosalicylic acid while being in the study. Steroids were used in 6 of 7 studies and their consumption was reported in detail in four studies. Activity of disease was defined by CDAI > 150 or a similar index. 367 patients were enrolled. 177 received azathioprine or 6-mercaptopurine and 190 received placebo. One of the seven studies used 6-mercaptopurine instead of azathioprine. The results of the seven studies and their pooled data are shown in the table below.

The pooled data showed that 56% of patients responded to azathioprine or 6-mercaptopurine compared with 32% of patients taking placebo. In addition, the pooled odds ratio was statistically significant: patients were 3 times more likely to respond to treatment than placebo.

The odds ratio of response in randomized controlled studies for active disease is also represented in the figure at right.

The number needed to treat to induce one remission was 4.01

The time of peak response in the studies ranged from 9 weeks to 26 weeks. At 17 weeks the pooled odds ratio for response to treatment became statistically significant: 1.95 (CI 1.10-3.46). At greater than 17 weeks the odds ratio increased to 19.20 (CI 6.27-58.8). Thus the authors concluded that 17 weeks was the minimum period for an adequate trial of azathioprine or 6-mercaptopurine.

Cumulative dose was found to be directly proportional to the likelihood of response. The odds ratio of response increased by 1.32 with each 100 mg/kg increase in cumulative dose. These findings are illustrated in the figures below. When cumulative dose and duration of therapy were considered simultaneously, only duration was statistically significant.

Two studies focused on quiescent disease only and four additional studies had quiescent-disease arms in multi-arm studies. 319 patients were enrolled. These groups of patients included a heterogeneous collection of patients: some in remission on azathioprine, some with stable symptoms but dependent of corticosteroids, some recently put into remission with corticosteroids and azathioprine. 136 received azathioprine and 183 received placebo. No studies used 6-mercaptopurine.

The results of the six studies and their pooled data are shown in the table below.

The pooled data showed that 67% of patients on azathioprine remained in remission compared with 52% of patients taking placebo. The pooled odds ratio was statistically significant: patients were 2.27 times more likely to remain in remission while taking azathioprine than taking placebo. The NCCDS was a source of heterogeneity among these studies because of the high number of placebo-taking patients who responded. In these studies the patients enrolled in the maintenance arm were those who entered remission taking either the placebo or azathioprine; thus those who entered remission on placebo are those who would enter remission spontaneously and therefore have less severe disease. The sensitivity analysis of excluding the results of each study in turn still showed a result favoring therapy that was statistically significant. When this study (and all studies in turn) were excluded from analysis in the sensitivity analysis, the results were still statistically significant.

The odds ratio of response in randomized controlled studies for quiescent disease is also represented in the figure at left.
NNT was 6.91.

The dose of azathioprine and duration of therapy were both analyzed and each was significant. The cumulative dose combines the two variables. Analysis of cumulative dose and odds of favorable response showed an increase in odds ratio of 1.19 (CI 1.13-1.24) per 100 mg/kg increase in cumulative dose. This is graphed in the figure below.

Five studies reported fistula response in placebo or therapy groups. Fistula response was defined as complete healing or decreased discharge. Patients who developed fistula during the study were included in the unhealed fistula group. Seventy patients had fistulae. 22 of 41 patients receiving azathioprine or 6-mercaptopurine responded (54% (CI 37-69%)). 6 of 29 patients’ fistulae responded to placebo (21% (8-40%)). The pooled odds ratio was 4.44 (CI 1.50-13.20) favoring fistula healing.

The reduction of corticosteroid is another important outcome for patients who had become steroid-dependent. Five studies reported the amount of steroid consumption. The meta-analysis defined a reduction in steroid use as either a successful taper off steroids or a dose reduction to less than 10mg/d with control of symptoms. Two-hundred-twenty-six patients were on steroids during the trials. 76 of 117 (65% (CI 56-74%)) receiving antimetabolites were able to decrease steroid use. 39 or 109 patients receiving placebo were able to do the same (36% (CI 27-45%)). The pooled odds ratio of 3.69 (CI 2.12-6.42) supports significant steroid-sparing effect.

The following conclusions can be made from the meta-analysis:

1. Azathioprine and 6-mercaptopurine are more effective than placebo for active CD with an odds ratio of 3.09 favoring therapy.
2. At a minimum, 17 weeks of treatment are required for an adequate trial of azathioprine or 6-mercaptopurine in active disease.
3. The response to azathioprine or 6-mercaptopurine is directly correlated with cumulative dose in active CD thus reflecting the slow onset of action of antimetabolites.
4. Azathioprine is more effective than placebo in quiescent disease with an odds ratio of 2.27 favoring therapy.
5. Cumulative dose was also proportional to the favorable response of maintaining remission.
6. Azathioprine and 6-mercaptopurine promote fistula healing.
7. Azathioprine and 6-mercaptopurine show a steroid-sparing effect.

Immunosuppressants in ulcerative colitis has been studied less than with Crohn’s disease because of a known cure for UC: total colectomy. However, there are some patients who wish to avoid surgery and antimetabolites have been studied in these subjects. The most widely cited publication of in these patients is by Hawthorne et al. in the British Medical Journal.¹⁶

Sixty-seven UC patients in remission who had been taking azathioprine for six months or more were randomized to receive either additional azathioprine (median dose was 100 mg/day, range 50-200 mg/day) or placebo. Their primary outcome measured was rate of relapse over one year. 67 or 79 patients took 5-aminosalicylic acid compounds. A separate group of 12 patients who were chronically steroid dependent was randomized and analyzed separately from the larger group that was free of steroid use. The treatment and control groups were similar with regard to age, disease activity, location of disease, use of 5-aminosalicylic acid, duration of disease before trial, duration of remission before trial (median 12 months), and duration of azathioprine use before trial.

Results were reported on an intention-to-treat basis.

In the group who were chronically dependent on steroids, 5 of 7 (71%) of patients taking azathioprine relapsed. Of the 5 patients taking placebo, 2 relapsed (40%).

In the larger group that was not dependent on steroids, 33 were randomized to azathioprine and 34 to placebo. 12 of the 33 taking azathioprine relapsed (36%). In comparison, 20 of the 34 patients taking placebo (59%) relapsed. This data is presented in the Kaplan-Meier survival plot of the rate of relapse in the accompanying figure. This was statistically significant (p=0.039). In this group the NNT was 4.45.

Adverse effects reported in patients taking azathioprine were limited to 2 patients with bone marrow depression, of whom both resolved with temporarily stopping drug and reinstating it at a lower dose with the leukocyte count recovered.

The study by Hawthorne concluded the following:

1. Chronically steroid-dependent patient did not receive any benefit from azathioprine in a small group of patients
2. Because the majority of patients in the placebo and therapy group took 5-asa, the effect of azathioprine is additive to the benefit already obtained by 5-asa
3. UC relapses are half as likely in patients taking azathioprine and 5-aminosalicylic acid compared to 5-aminosalicylic acid alone
4. Azathioprine was well tolerated by most patients
5. Leukopenia was seen in the therapy group but this was reversible in all cases

Azathioprine and 6-mercaptopurine are antimetabolite immunosuppressants. As such, there was serious concern regarding their safety for short- and long-term use. To illustrate the point, azathioprine and other immunosuppressants had been used to suppress rejection in kidney transplant recipients before they were used in IBD. A high incidence of malignancies was observed in this population. Indeed, renal transplants recipients’ risk of non-Hodgkin’s lymphoma was 49.4 times that of the normal population.¹⁷ In addition to concern over malignancy, possible immunosuppression, infection, leukopenia, and direct drug toxicity made early investigators cautious when prescribing the drugs.

In 1989 Present et al. published the largest study of toxicity in IBD patients taking 6-mercaptopurine for protracted periods. This New York group had been treating patients with 6-mercaptopurine since in 1967. They reviewed all the records 396 patients with IBD who had been treated with 6-mercaptopurine between 1967 and 1985. All adverse outcomes were reported in detail in their published article. A summary of their findings is presented below.

Of the 396 patients, 120 had UC and 276 had CD. Their age range at initiation of therapy was 8 to 70 years (mean 32 years). The majors indications for initiating therapy were chronic steroid use (>6 months), inability to withdraw steroids after two attempts, fistulae, or intractable clinical symptoms despite adequate trials of conventional therapy. The mean duration of treatment was 33.6 months (range, 1 to 162 months). 40 patients were lost to follow up.

Some patients were treated with 1.5 mg/kg/day of 6-mercaptopurine. Others were begun at 50 mg daily regardless of weight and the dosage increased if no clinical improvement was noted or decreased if the leukocyte count fell. Patients were followed weekly for the first month, biweekly the second month, and every third to fourth week for the next 10 months. They had a CBC every month and a CMP every 3 to 4 months. After one year patients were seen 3 to 4 times a year and continued to have the CBC each month and CMP 2 to 4 times a year.

The authors categorized each adverse effect as one of the following:

1. Direct toxicity of the drug
2. Neoplasm
3. Infection

Pancreatitis was seen in 13 (3.3%) patients. Twelve of 13 occurred within the first month of treatment (mean 23 days). Symptoms of nausea, emesis, back pain, fever, and elevated amylase were observed. Discontinuation of the drug was followed by resolution of symptoms within 1 week. Seven of thirteen patients were rechallenged with 6-mercaptopurine and all seven had a prompt recurrence of pancreatitis. Attempts to desensitize 3 patients with incrementally increasing doses failed. No patients had chronic pancreatitis or died.

Bone Marrow Depression

Leukopenia occurred in almost all patients at some time. Leukocyte counts of less than 2.5 x 10⁹ and fever was seen in eight (2%) patients. These patients required hospitalization. Three of eight patients had positive blood cultures for bacteria and one grew cytomegalovirus and had a prolonged and complicated hospital stay. Leukocyte counts returned to normal after discontinuation of the drug in all patients and no patients died. Five of eight patients had this complication during the first month of treatment.

Thrombocytopenia and anemia also occurred but less frequently than leukopenia.

Eight (2%) patients had allergic-type reactions within the first 3 weeks of therapy. Symptoms included fever, rash, joint pain, and abdominal pain without elevation of amylase. All symptoms resolved after discontinuing the drug. All 8 patients had similar reaction when rechallenged with 6-mercaptopurine.

Elevation of liver function tests (LFT) was noted in one patient. LFTs returned to normal upon discontinuation of 6-mercaptopurine.

One patient required 5 month hospitalization for disseminated cytomegalovirus. She had fever 7 days at home before seeking medical help. On presentation she was leukopenic. She required 5 surgeries during her hospital stay, including an ileotransverse colectomy and drainage of a pyogenic liver abscess. Other organisms besides cytomegalovirus were cultured from various sites during patient’s stay. Ultimately she recovered and had done well after 7 years of follow up.

Liver Abscess

Two patients, including the one mentioned above, had infected liver abscesses. Both were drained surgically and both patients recovered.

Five (1.2%) patients had pneumonia while they took 6-mercaptopurine. Two recovered after withdrawal of 6-mercaptopurine and oral antibiotics. The other three required hospitalization. One had been on 6-mercaptopurine for 10 months and recently on steroids. Another had been on 6-mercaptopurine for 4 years. The third had taken 6-mercaptopurine intermittently for years. All patients recovered with antibiotics and cessation of 6-mercaptopurine.

Eight (2%) patients had a typical shingles eruption lasting a few weeks before resolution. One patient had severe pain that persisted for 3.5 months. One patient developed headache, fever, memory loss and disorientation. The patient’s serologic titers were consistent with herpes encephalitis and lumbar puncture showed increased leukocytes with normal protein and glucose. 6-mercaptopurine was discontinued on admission. This patient improved after 48 hours and had no residual neurologic deficit at 7 years follow up.

One 20-year-old woman on long-term steroid therapy developed fever after 2 months of 6-mercaptopurine treatment. Blood cultures were positive for Staphylococcus aureus. She had bilateral hip pain and although aspiration of the joints was negative for pus, she received treatment for septic arthritis. Much later she required bilateral hip replacement.

One 23 year-old with CD developed fever while leukopenic. Work up for source of infection was negative. She was treated empirically for 25 days with antibiotics before he recovered.

Ten (2.5%) patients developed hepatitis while taking 6-mercaptopurine. Serologic tests diagnosed hepatitis A in one patient and hepatitis B in another. Hepatitis C had not been discovered until 1989, four years after the end of this study’s end. All patients with hepatitis had complete resolution except the patient with hepatitis B. This patient continued to have elevated LFTs for 13 years of follow up but had no signs of cirrhosis.

One 33-year-old woman developed fever and elevated LFTs after a trip to Australia. She was hospitalized and 6-mercaptopurine was withdrawn. Liver biopsy showed nonspecific hepatitis. She improved and the diagnosis was made much later by serologic testing . She restarted 6-mercaptopurine and took the drug for an additional 9 years without adverse effect.

Twelve neoplasms were seen in patients. Of these twelve patients, three died. Five were taking 6-mercaptopurine at the time of diagnosis. The time from initiation of drug to diagnosis of tumor ranged from 3 months 113 months (average 44 months). Two patients had obvious confounding factors:

1. The patient diagnosed with carcinoma of the colon had had UC for 38 years and refused to have colectomy when it had been advised years before. He was diagnosed with colon cancer 3 months after initiation of 6-mercaptopurine although the colon cancer certainly preceded treatment with 6-mercaptopurine.
2. The patient diagnosed with breast cancer had a mother and a sister with cancers of the breast.

These 12 patients are summarized in the table below.

The high incidence of lymphoma seen in renal transplant recipients was not seen in IBD patients.

Since this study was published, another case-control study published by Connell et al.¹⁹ examined the risk of neoplasm in 755 IBD patients taking azathioprine compared with the general population. At an average follow up of 9 years, no patients developed lymphoma. A higher incidence of colon and anal cancers was observed. This is already a recognized complication of chronic IBD.

The following points highlight the toxicity observed in 396 patients who had IBD and were treated with 6-mercaptopurine:

1. 17% of patients suffered an adverse event/effect
2. Most adverse effects were reversible
3. The one case of lymphoma may be attributable to antimetabolite therapy, however the risk of lymphoma is very small, if present
4. Other neoplasms were not significantly different from IBD patients not taking immunosuppressants
5. Patients had both common and uncommon infections
6. Infections were not related to leukopenia in most cases
7. No infections were seen in the treatment group that had not been observed in IBD patients not taking immunosuppressants
8. Many patients suffering infections also took steorids
9. Direct toxicities occurred in 7.6% patients and are absolute contraindications for antimetabolite therapy

The slow onset of action of antimetabolites prompted investigation with more potent immunosuppressants, including cyclosporin.²⁰ The primary indication for use is acute, severe ulcerative colitis. Cyclosporin is given to avoid colectomy.

Initial short-term studies showed a dramatic decrease in the number of patients requiring colectomy with cyclosporin. However, in longer-term studies, the same patients who avoided colectomy with IV cyclosporin were taken for surgical colectomy a few weeks to months later when cyclosporin was tapered. Cyclosporin also causes renal toxicity.

The appropriate use of cyclosporin is yet to be defined.

A few studies have evaluated methotrexate in patients with chronic IBD unresponsive to traditional therapy including antimetabolites.

Feagan et al.²¹ studied 141 chronically active CD patients over 16 weeks in a RCT comparing methotrexate to placebo. 39% of patients in the methotrexate group were in remission at 16 weeks. Only 19% of placebo patients entered remission. Methotrexate also had a steroid-sparing effect.

More studies are needed to confirm the hopeful findings of this study.

1. A significant number of patients with IBD do not achieve remission with 5-ASA compounds and/or steroids
2. Such patients are candidate for immusuppressants
3. Antimetabolites are efficacious in active and quiescent Crohn’s disease
4. Antimetabolites are steroid-sparing agents
5. Antimetabolites promote fistula healing
6. Antimetabolites are effective in maintaining remissions in ulcerative colitis
7. Antimetabolites are generally well tolerated
8. The increased risk of neoplasm appears to be small, if present at all
9. Leukopenia is common but reversible with dosage adjustments
10. Infectious complications do not appear to be increased in IBD patients taking antimetabolites
11. Direct toxicities are absolute contraindications to therapy
12. Studies evaluation the cost-effectiveness are needed
13. More studies are needed regarding the longer use of immunosuppressant and their impact on mortality

1. Sandborn WJ. Azathioprine: State of the Art in Inflammatory Bowel Disease. Scand J of Gastroenterology, 1998; 33 Suppliment 225:92-9.
2. Margolin ML, Krumholz MP, Fochios SE, Korelitz BI. Clinical trials in ulcerative colitis. II. Historical review. Am J Gastroenterology, 1988; 83:227-243.
3. Chan G, Canafax D, Johnson C. The therapeutic use of azathioprine in renal transplantation. Pharmacotherapy, 1987; 7: 165-177.
4. Tidd DM, Paterson ARP. A biochemical mechanism for the delayed cytotoxic reaction of 6-mercaptopurine. Cancer Res, 1974; 34:738-746.
5. Best WR, Becktel JM, Singleton JW, Kern F. Development of a Crohn’s disease activity index: National Cooperative Crohn’s Disease Study. Gastroenterology, 1976; 70:439-44.
6. Willoughby J, Kuman P, Beckett J, Dawson A. Controlled Trial of Azathioprine in Crohn’s Disease. Lancet, 1971; 2:944-7.
7. Brooke BN, Hoffman DC, Swarbrick ET. Azathioprine for Crohn’s disease. Lancet, 1969; 2:612-4.
8. Rhodes J, Bainton D, Beck P, Campbell H. Controlled trial of azathioprine in Crohn’s disease. Lancet, 1971; 2:1273-6.
9. Klein M, Binder HJ, Mitchell M, Aaronson R, Spiro H. Treatment of Crohn’s disease with azathioprine: a controlled evaluation. Gastroenterology. 1974; 77:847-69.
10. O’Donoghue DP, Dawson AM, Powell-Tuck J, Bown RL, Lennard-Jones JE. Double-blind withdrawal trial of azathioprine as maintenance treatment for Crohn’s disease. Lancet, 1978; 2:955-7.
11. Summers RW, Switz DM, Sessions JT, Becktel JM, Best WR, Kern F, et al. National Cooperative Crohn’s Disease Study: results of drug treatment. Gastroenterology. 1979; 77:847-69.
12. Present DH, Korelitz BI, Wisch N, Glass JL, Sachar DB, Pasternack BS. Treatment of Crohn’s disease with 6-mercaptopurine. A long-term randomized double-blind study. New England Journal of Medicine, 1980; 302:981-7.
13. Ewe K, Press AG, Singe CC, Stufler M, Ueberschaer B, Hommel G, et al. Azathioprine combined with prednisolone or monotherapy with prednisolone in active Crohn’s disease. Gastroenterology, 1993; 105:367-72.
14. Candy S, Wright JP, Gerber M, Adams G, Gerig M, Goodman R. A double blind controlled study of azathioprine in the treatment and maintenance of remission in Crohn’s disease [abstract]. Gastroenterology, 1994; 106 (4 part 2):A659.
15. Pearson DC, May GR, Fick GH, Sutehrland LR. Azathioprine and 6-Mercaptopurine in Crohn Disease, A Meta-Analysis. Annals of Internal Medicine, 1995; 122:132-142.
16. Hawthorne AB, Logan RFA, Hawkey CJ, et al. Randomized controlled trial of azathioprine withdrawal in ulcerative colitis. British Medical Journal, 1992; 302:20-2.
17. Kinlen LJ. Incidence of Cancer in Rheumatoid Arthritis and Other Disorders after Immunosuppressive Treatment. Am J. Medicine, 1985; 78(suppl 1A):44-49.
18. Present DH, Meltzer SJ, Krumholz MP, Wolke A, Korelitz BI. 6-Mercaptopurine in the Management of Inflammatory Bowel Disease: Short- and Long-Term Toxicity. Annals of Internal Medicine, 1989; 111:641-649.
19. Connell WC, Kamm MA, Dickson M, Balkwill AM, Ritchie JK, Jones JE. Long-term Neoplasia Risk after Azathioprine treatment in inflammatory bowel disease. Lancet, 1994; 343:1249-1252
20. Hanauer SB. Inflammatory Bowel Disease: Drug Therapy. New England Journal of Medicine, 1996; 334:841-848.
21. Feagan, Rochon J, Fedorak RN, Irvine J, Sutherland L, et al. Methotrexate for the treatment of Crohn’s Disease. New England Journal of Medicine, 1995; 332:292-297.