HSR&D Citation Abstract
Search | Search by Center | Search by Source | Keywords in Title
Long-Term Persistence of Methotrexate Use is Lower Than Azathioprine in the Treatment of Inflammatory Bowel Disease
Chen D, Waljee AK, Govani SM, Higgins PD, Stidham RW. Long-Term Persistence of Methotrexate Use is Lower Than Azathioprine in the Treatment of Inflammatory Bowel Disease. Paper presented at: Digestive Disease Week Annual Conference; 2016 May 24; San Diego, CA.
Long-term adherence to methotrexate is reduced compared to azathioprine in the treatment of inflammatory bowel disease.
Waljee, Akbar K.
Govani, Shail M.
Higgins, Peter D. R.
Stidham, Ryan W.
1. Gastroenterology and Hepatology, University of Michigan, Ann Arbor, MI, United States.
Background: Immunomodulators are effective agents in mono- or combination therapy for inflammatory bowel disease. Differences between drug-induced bone marrow suppression and hepatitis, adverse events, and patient intolerance may impact choices between thiopurines and methotrexate. The aim of this study was to compare overall persistence of medication use between azathioprine (AZA) and methotrexate (MTX).
Methods: We conducted a retrospective record review at a single large tertiary medical center, identifying adults with verified Crohn's disease (CD) or ulcerative colitis (UC) using AZA or MTX between 2000-2014. Medication related data including medication start and stop date, and reason for discontinuation up were collected by manual review of electronic medical records. To determine medication discontinuation due to patient intolerance and adverse events, elective discontinuations (e.g. medication deemed ineffective, use of alternative therapy or surgery, financial, family planning) were censored. Kaplan-Meier curves using the life table method and Cox regression were performed using SAS 9.4.
Results: The study population and included 1452 patients; 1077 (74.2%) using AZA and 375 (25.8%) using MTX with 49.9% of all subjects being male (p = 0.677). Patient mean follow-up time was 59.0 vs. 30.0 months for AZA vs MTX users, respectively. AZA patients were significantly younger (38.4 vs 46.7yrs, p < .0001), exhibited a longer disease duration (13.4 vs 10.1yrs, p < .0001) and a higher prevalence of UC (40.7% vs. 14.6%, p < .0001) compared to MTX users. When censoring elective medication stops, medication discontinuation was more common among MTX users at 6 months (12.1% vs. 7.8%, p = 0.046), 12 months (16.1% vs. 9.2%, p < .0001), and 36 months (32.5% vs. 13.0%, p < .0001, Figure 1). Examination of the reason for medication discontinuation showed intolerance more common among MTX users, adverse events more common in AZA users, and drug-induced bone marrow suppression or hepatitis no different between groups (Table 1). Among the 47.1% of MTX users with prior thiopurine use, there was no difference in medication discontinuation (45.5% vs 54.6%, p = 0.110). After adjusting for demographics, smoking status, IBD type and duration, prior immunomodulator use, and anti-TNF use, MTX users exhibited an increased hazard of discontinuing medication for non-elective reasons compared to AZA users (HR 9.59, 95%CI 5.80, 15.83).
Conclusions: Though limited by unaccounted confounders and the lack of a reliable disease activity assessment, this single center study suggests that patient tolerance is a significant barrier to long-term methotrexate use compared to azathioprine. While comparative effectiveness studies of immunomodulars are lacking, comparative safety and tolerability data may help guide immunomodulator choice in patients with IBD.
Table 1. Reason for Non-Elective Medication Discontinuation
Patient Intolerance33.7%59.6% < .0001
Figure 1. Probability Estimate of Immunomodulator Tolerance in IBD
Kaplan-Meier curve demonstrating persistence of azathioprine and methotrexate use over the study period. Elective discontinuations, including presumed insufficient efficacy, referral to surgery without resumption, de-escalation of therapy, family planning, financial challenges, and patient decision without adverse event or symptomatic intolerance, were censored in this analysis.