The methotrexate therapeutic response in rheumatoid arthritis

Methotrexate (MTX) is used frequently as a disease modifying antirheumatic drug (DMARD) for rheumatoid arthritis (RA), and patients tend to continue taking this drug for longer periods than alternative single agents. The shape of the therapeutic response beyond one or 2 years, however, has not been fully studied. We examined the properties of the pure MTX "therapeutic segment," that period that begins with start of MTX and terminates when MTX is discontinued or another DMARD is added, by observational study.

We studied new MTX starts for the period 1988 through 1996 for 437 patients from a parent cohort of 4253 patients. Patients were drawn from 8 Arthritis, Rheumatism, and Aging Medical Information System (ARAMIS) data centers: 2 community based populations; 2 private rheumatological practices; 2 university referral practices; and 2 university clinics for underserved minority urban populations. Health Assessment Questionnaire (HAQ) Disability Index scores (0-3) were obtained prospectively each 6 months.

At MTX start, patients had relatively long average disease duration of 16.7 years, and had moderately severe disability, with an initial HAQ mean disability score of 1.48. Over the 10 year period examined in the parent cohort of 4253 patients (and thus irrespective of therapy), the prevalence of MTX use rose from 19% to 45%, while mean HAQ disability declined from 1.34 to 1.11. This correspondence is consistent with an accrual of benefits from more frequent use of MTX and other DMARD over this period. The MTX therapeutic segment revealed a distinct shape. HAQ-Disability Index values began at 1.48 at baseline and declined to a maximal improvement of 1.23 at 30 months. This long period to maximum benefit may have been partly driven by a slow titration upward to an optimal dosage. After 42 months, disability for this population began to re-progress and reached 1.39 at 84 months, still below the pretreatment baseline. Re-progression to baseline was about 8 or more years. Cumulative disability averted with MTX treatment for this population was roughly 1.30 disability-unit-years.

MTX treatment of RA in practice differs substantially from common perception and appears suboptimal by being too little, too late, and too long to treatment change. A modification of the "sawtooth strategy" in which the disease is "ratcheted down" by change of MTX therapy at 3 years or when re-progression has proceeded halfway to baseline, rather than waiting for return to baseline, is suggested by these data. Also suggested is the need for more rapid upward dosage titration and longer maintenance of an optimal or highest tolerated dosage. "Therapeutic segment" data provide insights into strategic approaches to management of RA since they allow estimation of population aggregate properties such as time to maximum benefit and the time to return to baseline.