DMARDs

What these drugs are

DMARD stands for disease-modifying antirheumatic drug, and the name says what the job is. These are drugs meant to slow the disease down rather than make today more comfortable, which is a distinction that counts for a good deal more than it sounds like it would. It's the whole of why they get judged differently.

A painkiller is judged on how you feel this afternoon, while one of these is judged on what your joints look like in five years from now. Measure one by the other's test and you'll conclude it isn't working. That's the most common way somebody ends up stopping a drug that was doing its job all along without making itself felt.

Methotrexate is usually the first one tried, and others include leflunomide, sulfasalazine, and hydroxychloroquine, which has a page of its own here. Your team decides which one you get. How long you stay on it turns on which disease you have, what else you're being treated for, and what you can tolerate.

Why the wait is the hard part

These drugs take weeks, and you can see that in how the trials are built. The studies on this page report their results at 24 weeks, at one year, and at two years, and a DMARD doesn't get measured at a fortnight because there would be nothing to measure. That timescale is the thing to hold on to.

So a few weeks in with no change is expected rather than a failure of the treatment. Months in with no change is a different conversation, and that's the point at which to raise it rather than to keep waiting. Waiting longer is the more common mistake by a distance.

What covers that wait is usually something else, and often enough it's a short steroid course to bridge the wait. That's a separate decision with risks of its own attached to it. It belongs with your team rather than with a page like this one.

What the trials measure

Two things get measured here, and they don't always move together. One is how well the disease is controlled. The other is how long people stay on the drug, and those two come apart more often than anybody expects them to.

One trial makes the split clear. It took 59 people who had reached low disease activity on methotrexate and leflunomide together, then put each of them on one drug alone. Remission a year later was 90.6 percent on methotrexate against 74.1 percent on leflunomide. That difference didn't reach significance, at P equals 0.091, while the chance of still being on the assigned drug at a year was 81 percent against 55 percent, at P equals 0.025.

So on the question the trial was designed to answer, the two drugs looked similar to each other. On the question of whether people stayed on them, they didn't. Both of those findings are worth knowing, and they aren't one fact told twice.

The plan around the drug gets measured too, and it counts as much as the choice of drug does. One trial followed 43 people starting a biologic, where some had their other drugs adjusted toward a target and some were left on a fixed set, and response rates were 76.2 percent against 36.4 percent, at P equals 0.020. That is 43 people, so hold the size of it loosely and take the principle away instead.

The blood tests

You'll have blood counts checked regularly, and that's looking for something uncommon rather than something expected. It's worth knowing how uncommon it is, because the figures are on file and they're reassuring once you see them. A meta-analysis pooled 30 trials and 3858 people taking methotrexate with folic acid.

Anemia of any degree appeared in 2.55 percent, with the true figure somewhere from 0.60 to 5.47 percent, and a low white cell count appeared in 1.17 percent, from 0.16 to 2.80. A low neutrophil count appeared in 1.77 percent and low platelets in 0.19 percent. Across all 3858 of those people there were four cases of severe anemia and three cases of severe low neutrophils.

There were no cases of severe low white cells, no severe low platelets, and no pancytopenia anywhere in that pooled group at all. The authors conclude that blood count problems are an uncommon side effect at this dose with folic acid. So the tests exist because the problem is silent when it starts.

Coming off them

This is the question people most want answered, and it's the one the evidence answers least well. Nothing here is a randomized trial, so read the next part with that firmly in mind. Everything below describes what happened rather than what was tested.

One study followed 94 people whose rheumatoid arthritis had been quiet for at least six months, and over about 20.8 months 40 of them flared, which is 42.6 percent overall. Split them by what they did and the picture separates. Of the 59 who cut their drugs back, 31 flared, which is 52.5 percent, while of the 35 who didn't, 9 flared, which is 25.7 percent.

Read that carefully, because patients were not assigned to reduce and chose along with their doctors. So the people who cut back were probably already doing best of all, which should have made them flare less than the others. It went the other way.

That study found two other things that predicted a flare. One was the starting disease activity score, and the other was inflammation visible on ultrasound in somebody with no symptoms at all. That second one is the useful part, because it says feeling well and being quiet are two different measurements.

Common misconceptions.

Related terms.

• Biologics, the injected drugs used when these aren't enough.
• Biologics compared with DMARDs, for how the two sets measure up head to head.
• Hydroxychloroquine, one of these drugs covered on its own page.
• Rheumatoid arthritis, where most of this evidence comes from.