Importance Veterans with diabetic retinopathy have peak hemoglobin A1c levels that significantly exceed those of clinical trial populations, limiting the ability to extrapolate results on the association between hemoglobin A1c levels and treatment outcomes of intravitreal anti-VEGF injections to the veteran population.
Objective To determine whether peak hemoglobin A1c levels are correlated with the efficacy of intravitreal anti-VEGF injections in veterans with diabetic retinopathy.
Design, Setting, and Participants This was an observational retrospective study of veterans who received at least one intravitreal anti-VEGF injection for diabetic retinopathy at the VA Greater Los Angeles Healthcare System from January to March 2016. All participants had hemoglobin A1c records for a minimum of five years prior to their first injection series, defined as three injections, with the second injection no later than three months after the first injection and no earlier than three months prior to the third injection. There must also be an injection-free period of at least three months prior to the first injection in the series. Subjects sorted into the high hemoglobin A1c group had at least one hemoglobin A1c measurement exceeding 10.0%.
Methods In this retrospective chart review, data collected included number of intravitreal anti-VEGF injections, laser treatments, and steroid injections prior to and after the first anti-VEGF injection series, panretinal photocoagulation during the series, and hemoglobin A1c and central macular thickness measurements prior to and after the series.
Main Outcome and Measures Change in central macular thickness on optical coherence tomography.
Results Of 97 eyes from 68 patients, 81 were in the high hemoglobin A1c group and 16 were in the low hemoglobin A1c group. Between the two groups, there were no differences in the number of intravitreal anti-VEGF or laser injections received prior to the first injection series, type of anti-VEGF drug received during the first injection series, and central macular thicknesses immediately prior to and following the first injection series (all p values > 0.05). However, the high hemoglobin A1c group received more panretinal photocoagulation during the injection series (p = 0.002). From one month to one year after the injection series, there were no differences between the two groups in central macular thicknesses or additional intravitreal anti-VEGF, laser, or steroid injections received (all p values > 0.05). From one to three years after the injection series, there were no differences between the two groups in additional anti-VEGF or steroid injections received (both p values > 0.05). However, the high hemoglobin A1c group received more laser treatments than the low hemoglobin A1c group (p = 0.02) and had more eyes with central macular thickness under the edema cutoff of 250 um (p = 0.04). Hemoglobin A1c levels between the two groups differed significantly, prior to and following the injection series (all p values < 0.05).
Limitations Limitations to this study include small sample size, unknown confounding factors, and unequal distribution of eyes between the two groups.
Conclusions and Relevance Veterans with high peak hemoglobin A1c levels received more laser treatments during their first intravitreal anti-VEGF injection series and between one and three years after that initial series. These laser treatments may have helped to decrease central macular thicknesses at the three year follow-up and be an effective treatment adjunct to anti-VEGF injections for veterans with diabetic retinopathy.