Azathioprine vs Conventional Management of Crohn's Disease
This study, performed in adult patients with a high risk of disabling CD, failed to show that early treatment with azathioprine within 6 months of diagnosis was more effective than conventional management for increasing the duration of remission over the next 36 months. Among the secondary efficacy measures, only the development of fewer perianal complications and less need for perianal surgery was associated with early azathioprine use.
The efficacy of azathioprine for maintenance of remission in patients with CD, albeit modest (number needed to treat of 6), is well established when used at the appropriate dosage (2 to 2.5 mg · kg · day) and duration (at least 17 weeks). However, it has been claimed that this effect may be jeopardized if prescribed too late, at a time when irreversible damage has already occurred. This allegation is mainly based on some retrospective studies that suggest a benefit of early introduction of azathioprine in reducing the rates of surgery. In the present study, the median interval between diagnosis and inclusion was only 2 months and patients with severe or complicated disease requiring biologic agents or surgery were excluded. Despite that treatment with azathioprine was initiated before any complication of CD occurred, it did not significantly change the disease course or reduce the occurrence of intestinal complications requiring surgery when compared with the conventional treatment group. It is difficult to compare our data with previous clinical trials on thiopurines or methotrexate in adults with CD because they used different end points and none included newly diagnosed patients. Still, the efficacy of azathioprine monotherapy is probably modest during the early phase of CD or confined to a subset of patients. In the SONIC trial, at week 26, only 30% of patients with a median disease duration of 2.4 years and receiving azathioprine monotherapy achieved corticosteroid-free remission and only 15% had mucosal healing. In the placebo-controlled AZTEC trial, azathioprine started within the first 8 weeks after diagnosis of CD was not associated with an increased rate of sustained corticosteroid-free remission at week 76 compared with placebo. Actually, the only prospective study thus far showing the efficacy of early treatment with thiopurine in children was small and of short duration. Differences with the AZTEC and RAPID trials may be related to different phenotypes; there is a higher proportion of extensive disease, rapid disease progression, and increased activity in pediatric cases of CD.
The only apparent benefit of early treatment with azathioprine was a reduced occurrence of active perianal lesions and less need for perianal surgery. Although their effect is slow and unpredictable, thiopurines have been shown to increase the rate of fistula closure in patients with CD, and in the long-term azathioprine responders have a decreased cumulative rate of perianal surgery compared with controls. The beneficial effect of early treatment with azathioprine on perianal disease needs to be confirmed by another study before recommending early initiation of azathioprine therapy in patients with rectal involvement and/or perianal lesions at diagnosis.
A relatively high proportion of patients (approximately one third) in the conventional arm who were considered at high risk for disabling disease still did not require immunomodulator or biologic therapy during the study. This percentage is higher than the 16% observed in the original study by Beaugerie et al, in which patients were followed up for 5 years but also included the most severe cases requiring anti-TNF therapy or surgery at diagnosis. Even though these patients may still develop significant damage over time, this observation underlines the need for better predictors of disease course in CD.
This study had several strengths. It was conducted in a well-characterized group of patients with early CD who had predictors of disabling disease. The end points were related to the whole 3-year study period rather than measured at a particular point in time. Finally, it compared 2 therapeutic strategies relevant to daily clinical practice. This study also had several limitations. First, investigators and patients were aware of the treatment assignment. We believe that our primary and secondary end points, which took into account all significant events and therapeutic changes occurring during each trimester over a 3-year period, still allowed a valid comparison of the 2 therapeutic strategies even unblinded. Second, treatment with azathioprine was associated with a high rate of adverse events, with a need for drug discontinuation in up to one-fourth of the patients. However, this occurred in both groups and azathioprine was immediately replaced by mercaptopurine or methotrexate, allowing most patients in the early azathioprine group to remain on immunomodulator therapy throughout the study. Third, azathioprine was also rapidly started in some patients in the conventional management group, and the difference in the median delay of first azathioprine prescription was only 11 months between the 2 groups. This actually reflects current clinical practice, which is to prescribe immunomodulator and/or biologic therapy promptly in patients with chronic active disease or poor response to corticosteroids. Nevertheless, this 11-month interval is probably long enough to show a difference in clinical activity over a period of 3 years in a selected group of patients with a high risk of progressing disease. It has indeed been shown that the first 3 years are probably the most important in shaping progression of CD, which then tends to decrease and stabilize afterward.
In conclusion, our study failed to show that early initiation of azathioprine therapy significantly increases the proportion of trimesters in clinical remission during the first 3 years of CD as compared with conventional management in patients at risk for disabling CD. These results do not support systematic early prescription of azathioprine in adult patients with CD at risk for a disabling course.
Discussion
This study, performed in adult patients with a high risk of disabling CD, failed to show that early treatment with azathioprine within 6 months of diagnosis was more effective than conventional management for increasing the duration of remission over the next 36 months. Among the secondary efficacy measures, only the development of fewer perianal complications and less need for perianal surgery was associated with early azathioprine use.
The efficacy of azathioprine for maintenance of remission in patients with CD, albeit modest (number needed to treat of 6), is well established when used at the appropriate dosage (2 to 2.5 mg · kg · day) and duration (at least 17 weeks). However, it has been claimed that this effect may be jeopardized if prescribed too late, at a time when irreversible damage has already occurred. This allegation is mainly based on some retrospective studies that suggest a benefit of early introduction of azathioprine in reducing the rates of surgery. In the present study, the median interval between diagnosis and inclusion was only 2 months and patients with severe or complicated disease requiring biologic agents or surgery were excluded. Despite that treatment with azathioprine was initiated before any complication of CD occurred, it did not significantly change the disease course or reduce the occurrence of intestinal complications requiring surgery when compared with the conventional treatment group. It is difficult to compare our data with previous clinical trials on thiopurines or methotrexate in adults with CD because they used different end points and none included newly diagnosed patients. Still, the efficacy of azathioprine monotherapy is probably modest during the early phase of CD or confined to a subset of patients. In the SONIC trial, at week 26, only 30% of patients with a median disease duration of 2.4 years and receiving azathioprine monotherapy achieved corticosteroid-free remission and only 15% had mucosal healing. In the placebo-controlled AZTEC trial, azathioprine started within the first 8 weeks after diagnosis of CD was not associated with an increased rate of sustained corticosteroid-free remission at week 76 compared with placebo. Actually, the only prospective study thus far showing the efficacy of early treatment with thiopurine in children was small and of short duration. Differences with the AZTEC and RAPID trials may be related to different phenotypes; there is a higher proportion of extensive disease, rapid disease progression, and increased activity in pediatric cases of CD.
The only apparent benefit of early treatment with azathioprine was a reduced occurrence of active perianal lesions and less need for perianal surgery. Although their effect is slow and unpredictable, thiopurines have been shown to increase the rate of fistula closure in patients with CD, and in the long-term azathioprine responders have a decreased cumulative rate of perianal surgery compared with controls. The beneficial effect of early treatment with azathioprine on perianal disease needs to be confirmed by another study before recommending early initiation of azathioprine therapy in patients with rectal involvement and/or perianal lesions at diagnosis.
A relatively high proportion of patients (approximately one third) in the conventional arm who were considered at high risk for disabling disease still did not require immunomodulator or biologic therapy during the study. This percentage is higher than the 16% observed in the original study by Beaugerie et al, in which patients were followed up for 5 years but also included the most severe cases requiring anti-TNF therapy or surgery at diagnosis. Even though these patients may still develop significant damage over time, this observation underlines the need for better predictors of disease course in CD.
This study had several strengths. It was conducted in a well-characterized group of patients with early CD who had predictors of disabling disease. The end points were related to the whole 3-year study period rather than measured at a particular point in time. Finally, it compared 2 therapeutic strategies relevant to daily clinical practice. This study also had several limitations. First, investigators and patients were aware of the treatment assignment. We believe that our primary and secondary end points, which took into account all significant events and therapeutic changes occurring during each trimester over a 3-year period, still allowed a valid comparison of the 2 therapeutic strategies even unblinded. Second, treatment with azathioprine was associated with a high rate of adverse events, with a need for drug discontinuation in up to one-fourth of the patients. However, this occurred in both groups and azathioprine was immediately replaced by mercaptopurine or methotrexate, allowing most patients in the early azathioprine group to remain on immunomodulator therapy throughout the study. Third, azathioprine was also rapidly started in some patients in the conventional management group, and the difference in the median delay of first azathioprine prescription was only 11 months between the 2 groups. This actually reflects current clinical practice, which is to prescribe immunomodulator and/or biologic therapy promptly in patients with chronic active disease or poor response to corticosteroids. Nevertheless, this 11-month interval is probably long enough to show a difference in clinical activity over a period of 3 years in a selected group of patients with a high risk of progressing disease. It has indeed been shown that the first 3 years are probably the most important in shaping progression of CD, which then tends to decrease and stabilize afterward.
In conclusion, our study failed to show that early initiation of azathioprine therapy significantly increases the proportion of trimesters in clinical remission during the first 3 years of CD as compared with conventional management in patients at risk for disabling CD. These results do not support systematic early prescription of azathioprine in adult patients with CD at risk for a disabling course.
SHARE
