LONDON – U.S. rheumatoid arthritis patients had a significantly reduced incidence of type 2 diabetes when on treatment with either hydroxychloroquine or abatacept in an analysis of more than 13,000 patients enrolled in a U.S. national registry during 2000-2015.
The same analysis also showed statistically significant increases in the risk for new-onset type 2 diabetes in rheumatoid arthritis (RA) patients treated with a glucocorticoid or a statin, Kaleb Michaud, Ph.D., reported in a poster at the European Congress of Rheumatology.
Mitchel L. Zoler/Frontline Medical News
Kaleb Michaud, Ph.D.
“Our findings can inform clinicians about determining appropriate treatment decisions in rheumatoid arthritis patients at increased risk for developing type 2 diabetes,” said Dr. Michaud, an epidemiologist at the University of Nebraska in Omaha and also codirector of the National Data Bank for Rheumatic Diseases in Wichita, Kan., the source of the data used in his study.
Hydroxychloroquine, a relatively inexpensive drug that’s often used in RA patients in combination with other drugs, might be a good agent to consider adding to the therapeutic mix of RA patients at risk for developing type 2 diabetes, he said in an interview. Although the implications of this finding for the use of abatacept (Orencia) are less clear, it is a signal of a novel benefit from the drug that merits further study, Dr. Michaud said.
The findings also suggest that, when rheumatoid arthritis patients receive statin treatment, they are candidates for more intensive monitoring of diabetes onset, he added.
His analysis focused on the 13,669 RA patients who participated in the National Data Bank for Rheumatic Diseases for at least a year during 2000-2015 and had not been diagnosed as having diabetes of any type at the time they entered the registry. During a median 4.6 years of follow-up, 1,139 patients either self-reported receiving a new diagnosis of type 2 diabetes or began treatment with a hypoglycemic medication. Patients averaged about 59 years old at the time they entered the registry, and about 80% were women.
Dr. Michaud and his associates assessed the incidence rate of type 2 diabetes according to six categories of RA treatment: methotrexate monotherapy, which they used as the reference group; any treatment with abatacept with or without methotrexate; any treatment with hydroxychloroquine; any treatment with a glucocorticoid; treatment with any other disease-modifying antirheumatic drug (DMARD) with methotrexate; and treatment with any other DMARD without methotrexate or no DMARD treatment. A seventh treatment category was treatment with a statin.
A series of time-varying Cox proportional hazard models that adjusted for sociodemographic variables, comorbidities, body mass index, and measures of RA severity showed that, when compared with methotrexate monotherapy, patients treated with abatacept had a statistically significant 48% reduced rate of developing diabetes, and those treated with hydroxychloroquine had a statistically significant 33% reduced diabetes incidence, they reported.
In contrast, compared with methotrexate monotherapy, treatment with a glucocorticoid linked with a statistically significant 31% increased rate of type 2 diabetes, and treatment with a statin linked with a statistically significant 56% increased rate of diabetes. Other forms of DMARD treatment, including tumor necrosis factor inhibitors and other biologic DMARDs, had no statistically significant link with diabetes development.
Dr. Michaud also reported additional analyses that further defined the associations between hydroxychloroquine treatment and reduced diabetes incidence: The reduction in diabetes incidence became statistically significant in patients only when they had received the drug for at least 2 years. The link with reduced diabetes incidence seemed dose dependent, with a stronger protective effect in patients who received at least 400 mg/day. Also, the strength of the protection waned in patients who had discontinued hydroxychloroquine for at least 6 months, and the protective effect completely disappeared once patients were off hydroxychloroquine for at least 1 year. Finally, the link between hydroxychloroquine use and reduced diabetes incidence also was statistically significant in patients who concurrently received a glucocorticoid, but a significant protective association disappeared in patients who received a statin as well as hydroxychloroquine.
Dr. Michaud had no disclosures. The study received no commercial support.
On Twitter @mitchelzoler