Don’t Blame DMARDs for Adverse Drug Effects in Rheumatology Patients

Adverse drug effects among patients under treatment for autoimmune rheumatologic disorders stemmed more often from ancillary medications — and not necessarily prescription products — than from disease-modifying anti-rheumatic drugs (DMARDs), a large case-control study found.

Some of the strongest sources of adverse drug effects were antibiotics, over-the-counter painkillers, and antacids, with odds ratios reaching into double digits, according to Titilola Falasinnu, PhD, of Stanford University in California, and colleagues.

But the adverse-effects burden wasn’t the same across rheumatologic disease categories, the group reported in a medRxiv preprint manuscript.

Patients with systemic sclerosis and systemic lupus erythematosus (SLE) appeared to suffer most from such effects, while those with rheumatoid arthritis (RA) and Sjögren’s disease had the fewest problems.

Another important finding was that rheumatologic patients across nearly all disease categories typically took a staggering number of medications. Even among controls, i.e., patients without adverse drug effects appearing in their records, the median number was around six for outpatients and 10 for inpatients.

Still, the more medications that patients were taking, the more likely they were to experience adverse effects: for most disease types, the median number of drugs taken was 50% to 100% higher among those with adverse effects compared with those without. Inpatients with systemic sclerosis who had adverse drug effects had a median of 20 different medications on board, Falasinnu and colleagues reported.

“While disease activity contributes to observed associations,” the researchers commented, “our findings suggest that medication burden, particularly from supportive therapies, represents a key and potentially modifiable contributor to harm. Strategies focused on optimizing prescribing, minimizing corticosteroid exposure, and improving pain management can reduce ADE [adverse drug effect] burden in this population.”

It’s no surprise, of course, that polypharmacy would be linked with adverse effects. As Falasinnu’s group noted, it “reflects a self-reinforcing cycle: disease activity and comorbidities necessitate multiple therapies, while treatment-related complications introduce additional medications.”

But which medications are most at fault, and whether they are primarily DMARDs given for patients’ primary rheumatologic disorders versus drugs for related symptoms or unrelated comorbidities, “remain incompletely characterized,” the investigators explained.

For their study, rheumatology patient records from Stanford’s medical system spanning a 15-year period beginning in 2010 were extracted. Autoimmune diagnoses including SLE, RA, systemic sclerosis, psoriatic arthritis, ankylosing spondylitis, and Sjögren’s disease were included.

This effort yielded 10,578 patients overall, of whom 3,154 were diagnosed with at least one adverse drug event, according to ICD-9/10 coding. Controls were selected from the remainder, matching cases by primary diagnosis and “encounter type” (outpatient, inpatient, or emergency).

Odds ratios for 85 individual medications were calculated, of which seven were conventional DMARDs and 10 were biologic or targeted DMARDs. Among the rest were corticosteroids, opioid and non-opioid painkillers, anti-infective agents, and medications for psychiatric, respiratory, gastrointestinal, endocrine, and cardiovascular symptoms or conditions. These odds ratios expressed the likelihood that a given drug was taken by a patient with adverse drug effects versus controls, and were calculated separately for each rheumatologic disease category.

Some of the worst offenders were as follows:

• Acetaminophen: ORs of 3.56 to 9.68
• Lorazepam: ORs of 4.15 to 6.52
• Famotidine: ORs of 4.18 to 5.78

Some drugs appeared problematic only for certain patients. For example, the anti-emetic agent metoclopramide had an odds ratio of 12.32 among systemic sclerosis patients but was not an issue in any other disease type. Systemic sclerosis patients also seemed to react poorly to rifaximin (Xifaxan), an antidiarrheal (OR 14.41). Similarly, ciprofloxacin was troublesome in ankylosing spondylitis (OR 13.71) but nothing else.

In general, DMARDs (both conventional and targeted) appeared far less troublesome than other medication types. Among the 17 drugs examined across the six disease categories, only two odds ratios greater than 3.0 came up: tacrolimus for patients with systemic sclerosis (OR 3.29) and certolizumab pegol (Cimzia) in those with ankylosing spondylitis. Notably, 10 of the 102 total calculations for DMARDs yielded odds ratios less than 1, as compared with three out of 408 calculations covering non-DMARD drugs.

Looking at disease types, relatively many drugs turned up with high odds ratios in SLE, systemic sclerosis, and ankylosing spondylitis, while relatively few were associated with adverse events in RA, psoriatic arthritis, and Sjögren’s disease.

Across all diseases, patients with adverse drug effects were slightly older than those without (mean 58 vs 55) and more likely to be white (66.8% vs 59.7%). Charlson comorbidity scores were also higher for those experiencing adverse effects (mean 3.47 vs 1.86), which would also help account for the increased polypharmacy burden. Nearly 40% of all patients in the study were taking more than 10 medications.

Limitations to the study included its reliance on administrative data and the potential for errors in classifying a new untoward symptom as a drug effect. As well, the researchers had no data on potential drivers of polypharmacy.