DMARD Discontinuation in RA: Feasible, but Questionable

November 11, 2014
Lois Wingerson

(ACR 2014) Encouraging real-life data to be presented at the ACR annual meeting suggest that at least some patients can safely discontinue biologics. But with the true meaning of remission still uncertain, it's not yet clear who they are.

The concept of “remission” in rheumatoid arthritis (RA) is clearly defined in numerous indexes (DAS28, the ACR20/50/70 scores and more). But in real-life clinical care, the concept of “remission” is subjective. Patients judged to be in remission often have symptoms and continuing inflammation.

Exactly when is it safe, then, to de-escalate treatment? Are there ways to predict which patients are likely to maintain remission?  What is the best way to measure it? With new rheumatoid arthritis guidelines scheduled for release, real-life answers for all of these questions are high on the agenda at the American College of Rheumatology annual meeting this month in Boston.

Among the new results scheduled for presentation:

Reducing and even discontinuing DMARDs is feasible for at least some patients who have been in clinical remission for at least 6 months. (Abstract 940) Results from the randomized multicenter RETRO study under way in Germany show that halving the dose of DMARDs for 6 months and then stopping altogether seems to be just as effective as simply halving the dose for a full year. Among 101 patients in the 3-arm study, reducing or discontinuing treatment did come with an increased risk of new flares, but neither strategy was more risky than the other. The only factor that predicted recurrence was the presence of anti-citrullinated protein antibodies (ACPA).

Most patients can succeed at de-escalation (as defined clinically), with considerable cost savings. (Abstract 941) In what they call the largest observational study of clinical and cost outcomes from de-escalating biologics in RA, a team from the Geisinger Health System in Danville PA assessed electronic health records of all 940 RA patients treated with biologics in 2013. De-escalation was successful (as defined by the Clinical Disease Activity Index or CDAI) for 85% of patients, at projected annual savings of more than $1.2 million for those able to stay at the reduced dosage for at least a year.

But how well such purely clinical measures of remission reflect actual relief from the disease continues to be a hot topic. For instance, using CDAI as in the previous study may not indicate real-life, functional remission at all.

Because it does not include questions about function, the CDAI may not reveal whether patients are actually well enough to get back to normal life, judging from a study based on data from an RA registry at the University of Pittsburgh. (Abstract 886) Extracting data on 511 RA patients younger than age 65, the team found that among 4 scores for disease status in RA (DAS28, SDAI and RAPID3 as well as CDAI), only the latter did not correlate significantly with employability (as judged by patients themselves). The RAPID3 index was best at predicting ability to work, the researchers say, because unlike the others it includes a functional component as well as symptoms and patient assessments of health.

Yet another call for patient-derived outcome measures: Doctors are about three times more likely to judge patients as being in remission than the patients themselves. (Abstract 942) Concerned that a patient global score (PATGL) used in remission criteria might not adequately reflect pain, a Danish team assessed the impact of substituting a score that purely reflected pain to assess remission among more than 16,000 patients in the nationwide DANBIO registry used by all rheumatologists in public hospitals in Denmark and by many in private practice. The two patient-derived scores were “highly correlated,” the researchers report. However, doctor’s global scores “almost tripled” the fraction of patients in remission.

The above abstracts are all available online at


Related Content:

ACR | Rheumatoid Arthritis