The treatment of inflammatory arthritis with methotrexate in clinical practice: treatment duration and incidence of adverse drug reactions

Objective. To identify the proportion of patients with inflammatory arthritis who remain on methotrexate in the medium to long term and the incidence of side-effects in clinical practice. Method. The study population comprised all patients with inflammatory arthritis treated with methotrexate and monitored in clinics under the auspices of Staffordshire Rheumatology Centre. Two clinical auditors collected data retrospectively from the computer database used to support monitoring of patients on disease-modifying anti-rheumatic drugs. Information was collected on duration of treatments and reasons for stopping treatment. For patients identified as having potentially serious side-effects or who died whilst taking methotrexate, further information on their outcome was collected from patients' medical notes and where applicable post mortem reports and death registers. Results. Between 1986 and 1999, 673 patients were treated with methotrexate, of whom 551 had a diagnosis of rheumatoid arthritis. From the Kaplan-Meier analysis, the probability of patients remaining on treatment 5 yr after starting methotrexate was 0.74. Three hundred and sixteen patients stopped methotrexate between 1986 and 1999. In 117 patients, the methotrexate was restarted. Seventy-two patients (10.7% of all patients) stopped because of inefficacy or patient choice or situation. Thirty-seven patients (5.5%) stopped methotrexate due to abnormal haematology (usually low neutrophils). Thirty-seven patients (5.5%) stopped methotrexate due to abnormalities in liver function tests. Life-threatening side-effects were identified in 12 patients (1.8%). These included six pneumonitis, five cytopenias and one disseminated varicella zoster. Two of these patients (0.3%) died, one from pneumonitis and one from disseminated varicella zoster. A total of 25 patients (3.7%) died while taking methotrexate and four died (0.6%) within 3 months of stopping methotrexate. One death (0.15%) was directly attributable to methotrexate (methotrexate pneumonitis). Conclusion. This study has shown that methotrexate is well tolerated in clinical practice in the medium to long term. It has produced accurate data on the incidence of adverse effects of methotrexate in a local population in a non-research setting. It has identified the incidence of life-threatening side-effects to be 1.7% with one death (0.15%) directly due to methotrexate. This information should prove useful when recommending such treatment to patients with inflammatory arthritis.