Abstract
Purpose :
Dipeptidyl-peptidase 4 inhibitors (DPP-4i) are newer medications used to treat type 2 diabetes mellitus (T2DM), but there are limited studies on the association between DPP-4i use and proliferative diabetic retinopathy (PDR) development. The objective of our study was to compare the PDR development among patients who had used DPP-4i versus those who had not used DPP-4i prior to PDR diagnosis.
Methods :
Medical records of patients with diabetes evaluated at the University of Wisconsin between 2010-2022 were reviewed. Patients were excluded from the analysis if they were not diagnosed with T2DM, had missing HbA1c values, or if records contained less than 6 months of observation time before a diagnosis of PDR, and were censored when they received treatments other than DPP-4i. The hazard ratio associated with DPP-4i treatment, diabetic macular edema (DME) diagnosis, and HbA1c were modeled with separate hazard ratios as a variable of time since diagnosis of T2DM, categorized as very early (≤ 3 years), early (>3 to <6 years), mid (>6 to <12 years), and late (>12 to 25 years). A p-value ≤0.05 was considered statistically significant.
Results :
16,111 participants were included in the study: 1,405 were using DPP-4i prior to the first PDR diagnosis and 14,706 were never on DPP-4i or were given DPP-4i after a diagnosis of PDR. Participants were 50.0% male, 81.9% white non-Hispanic, average T2DM duration of 8.3 ± 5.95 years, with mean HbA1c of 7.0 ± 1.26%.
For patients with a duration of diagnosis of T2DM (≤ 3 years), DPP-4i use was associated with a hazard ratio of 4.43 (CI= 1.89 to 10.3; p= 0.0006) for PDR development after controlling for Hba1c level, diabetic retinopathy (DR) and DME status, age of T2DM diagnosis, sex, race and ethnicity. For all other time periods (>3 years duration of diagnosis of T2DM), DPP-4i use was not significantly associated with a higher hazard of PDR development. In all time periods, DME development was associated with a lower hazard of PDR development, while non-proliferative DR and HbA1c were associated with a higher hazard of PDR development.
Conclusions :
With a shorter duration of diagnosis of T2DM, the use of DPP-4i was associated with an increased hazard of PDR development, but for the longer duration of diagnosis of T2DM, the association is not significant. In periods greater than 3 years, DME diagnosis was associated with a lower hazard of PDR development.
This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.