Newer Diabetes Drugs, Including GLP-1 Agonists, Tied to Autoimmunity Risk

Analysis of hundreds of thousands of patient records showed differing risks for autoimmune disorders for new users of the more recent types of antidiabetic medications.

Rates of new-onset autoimmune conditions were compared among patients starting on DPP-4 inhibitors, GLP-1 receptor agonists, and SGLT-2 inhibitors, according to a report by Jeffrey A. Sparks, MD, MMSc, of Harvard Medical School and Brigham and Women’s Hospital in Boston, and colleagues in ACR Open Rheumatology.

Compared with GLP-1 agonists, DPP-4 inhibitors were associated with less risk for plaque psoriasis (HR 0.79, 95% CI 0.70-0.85), psoriatic arthritis (HR 0.65, 95% CI 0.53-0.79), and autoimmune thyroiditis (HR 0.68, 95% CI 0.59-0.76), but more risk for bullous pemphigoid (HR 1.78, 95% CI 1.24-2.46) and dermatomyositis (HR 2.18, 95% CI 1.24-3.53). For other autoimmune disorders, however, such as rheumatoid arthritis (RA) and lupus, no significant differences were observed.

DPP-4 inhibitors were also associated with differential risks in comparison with SGLT-2 inhibitors (a.k.a. gliflozins). Risks with the former were doubled for bullous pemphigoid (HR 2.07, 95% CI 1.34-3.05) and dermatomyositis (HR 2.03, 95% CI 1.35-3.02), and by nearly as much for giant cell arteritis (HR 1.83, 95% CI 1.31-2.65). On the other hand, DPP-4 inhibitors were associated with less risk for psoriasis (HR 0.81, 95% CI 0.70-0.91) and autoimmune thyroiditis (HR 0.76, 95% CI 0.63-0.90).

Meanwhile, when GLP-1 drugs and SGLT-2 inhibitors were compared, no significant differences were seen for any of the 19 autoimmune conditions examined in the study. None of the between-class comparisons, in fact, showed differences for RA, lupus, inflammatory bowel disease, multiple sclerosis, celiac disease, systemic sclerosis, or other autoimmune conditions not mentioned above.

Sparks and colleagues stopped well short of suggesting that these drugs were actually causing autoimmune diseases, although in a few cases the associations they found dovetailed with previous reports. Rather, they said the study should be fodder for more focused and detailed studies to determine whether they reflect true safety signals.

“The study’s findings do not establish mechanistic certainty and should be regarded as preliminary signals that may help frame the scope and direction of more targeted mechanistic inquiry to identify potential unifying biologic principles underlying the observed associations,” was how they put it.

An almost complete lack of previous research on autoimmune risks with newer diabetes medications was the researchers’ primary motivation for conducting the study. They did point to a previous study showing that DPP-4 inhibitors can also block inflammatory processes in the thyroid triggered by lymphocytes, which might explain how these agents were associated with lower risk for autoimmune thyroiditis. But in general Sparks and colleagues were starting from scratch in searching for such associations.

Their method involved taking data from the massive TriNetX database of patient records, searching for individuals with diabetes and new prescriptions for drugs in the three classes, and then looking for subsequent diagnosis of autoimmune disorders. Using the so-called target trial emulation methodology, the researchers generated pairwise matches of patients receiving the different medication types.

In this way they developed three comparisons of patients: those taking DPP-4 inhibitors versus GLP-1 agonists, DPP-4 inhibitors versus SGLT-2 inhibitors, and GLP-1 agonists versus SGLT-2 inhibitors, with more than 100,000 pairs of patients in each case.

This meant there was no control group of patients receiving older types of medications such as metformin or sulfonylurea drugs, or no medication (which would be very unusual anyway). However, they did look at incidence of non-autoimmune diagnoses such as hand lacerations and inguinal hernia in each set of patient pairs — i.e., “negative controls” — and in these analyses there were no differences between medication classes.

“Given the lack of true untreated comparators,” Sparks and colleagues wrote, “the findings reflect relative differences between antidiabetic classes rather than absolute effects on autoimmune disease risk; future work incorporating mechanistic endpoints and external untreated comparators is essential to further elucidate the degree to which relative differences may translate to clinically relevant impacts.”

Other limitations included reliance on administrative data and the potential for confounding by unmeasured parameters, although the researchers were able to adjust for such factors as baseline HbA1c, body mass index, tobacco and alcohol use, and renal function.