Comment

The search goes on for safe and effective long term treatments that will modify the course of aggressive Crohn's disease and reduce the need for frequent surgery or long term corticosteroids. Azathioprine and 6-mercaptopurine remain the most widely used immunomodulators but there is a need for alternatives, particularly for the 10% of patients who cannot tolerate these drugs. Methotrexate has been used in Crohn's disease for many years1-3 and a controlled trial showed benefit in active corticosteroid dependent disease4 yet it is still not widely used. Likely reasons include concerns about toxicity (particularly liver fibrosis, hypersensitivity pneumonitis, and teratogenicity), the need for parenteral administration, and lack of evidence of long term benefit.

The recent publication of Feagan and colleagues5 provides important information on the long term role of methotrexate. While not obvious from the paper's title, this was in fact a placebo controlled withdrawal study in 76 patients who had achieved remission (judged by Crohn's disease activity index) after methotrexate 25 mg intramuscularly weekly for 4–6 months. Forty patients then received methotrexate 15 mg intramuscularly every week, and the remainder received placebo injections for 40 weeks. Significantly more remained in remission on methotrexate (65%) than those on no treatment (39%). This benefit of maintenance treatment was of course in a highly selected group of patients who had already demonstrated responsiveness to methotrexate and who also tolerated the drug at a higher dose (17% previously withdrawn from the 25 mg study4). Perhaps unsurprisingly, the maintenance dose was well tolerated, with only one withdrawal for nausea and none for leucopenia or abnormal liver function.

With this recent addition to the controlled evidence, it is appropriate to summarise our current knowledge of methotrexate in Crohn's disease. Liver toxicity is not a problem for cumulative doses up to 5 g,6 ,7 and regular liver biopsy is not needed. Methotrexate is given weekly, as daily dosage increases the risk of toxicity. Folic acid 1 mg daily reduces haematological toxicity. Side effects are common, particularly nausea and headache, but are usually mild and dose related, requiring drug withdrawal in only 10%.3 Cough or dyspnoea should raise suspicion of pneumonitis but this is uncommon (1–2%) and usually settles with corticosteroids if methotrexate is withdrawn promptly.

Oral absorption is not reduced in small bowel Crohn's disease8 but absorption is highly variable and is also dose dependent, with bioavailability decreasing by a mean 13.5% when the dose is increased from 7.5 mg to 15–20 mg weekly.9There is anecdotal evidence of a dose threshold for response but no evidence that doses higher than 25 mg are beneficial. Responders should switch to oral dosing, and the lowest effective dose should be continued; if patients relapse on low doses, it is logical to increase the dose or switch from oral to intramuscular treatment.

There is now strong evidence for a role for methotrexate in Crohn's disease but many questions remain. Does it act more rapidly than the purine analogues? (The only head to head study does not show this.10) Is intramuscular injection always needed initially? How long should treatment be continued in responders? The theoretical concerns about malignancy remain for all long term immunosuppression. Although overall tolerability is comparable with the purine analogues, methotrexate will remain my second choice, particularly in women, because of teratogenicity. It is however a useful alternative, and with careful monitoring and dose adjustment is worthy of more widespread use.