Glucagon-Like Peptide 1-Receptor Agonists and A1c: Good for the Heart But Less So for the Eyes?

Objective : Glucagon-like peptide 1- receptor agonists (GLP1-RA) are preferred treatments for people with DM type 2 and risk for cardiovascular disease. The FDA has recommended caution in the use of semaglutide and dulaglutide because of the risk of exacerbating diabetic retinopathy (DR). Others have suggested that the worsening of DR, is independent of the class of drug, but related to the rate of fall of A1C. We have performed meta-analyses of the major trials of GLP1-RA regarding alterations in major adverse cardiovascular events (MACE) and DR and the association of these differences with changes in A1C.

Methods: Meta-analyses and meta-regressions were performed on the 7 major RCTs (n=56004 patients) of GLP1-RA which have been the basis for determining the effectiveness of GLP1-RA regarding prevention of MACE.

Results: Six of the studies evaluated DR, and 4 defined changes as new onset photocoagulation, intravitreal injection, or hemorrhages. There was a significant increase in DR only for parenteral semaglutide (relative rate (rr)=1.73, 95% CI; 1.10:2.71, p= 0.02, rate difference (rd)= 0.013; 0.002:0.023, p=0.02). For all GLP1-RA there was no increase in DR (rr=1.05; 0.94:1.17, p=0.36). There were significant benefits for GLP1-RA in decreasing MACE (rr=0.89;0.85:0.94, p< 0.001, rd= -0.013; -0.018: -0.007, p < 0.001). Meta-regression analysis showed that the rate of improvement of MACE directly correlated with decrease in A1c (log rr=0.364 *(change in A1c), p=0.014). There was a trend for worsening of DR with decrease in A1c (log rr= (-0.67 *(change in A1c), p=0.076). Taking all studies of the class of GLP1-RA, for DR rd =0.001 (number needed to harm [NNH], 1000) compared with that for MACE rd= -0.013 (number needed to treat [NNT], 77) whereas the comparable computation for semaglutide as a single agent are NNH=77 and NNT= 43.

Discussion/Conclusion: GLP1-RA are associated with improvements in A1c, which correlates with improvement in the rates of MACE.  There is a trend showing decreases in A1c may be associated with worsening of DR. In individuals with diabetes who are high risk for cardiovascular disease, GLP1-RA should be recommended with close monitoring by ophthalmology for worsening DR. Further prospective studies of GLP1-RA on DR need to evaluate whether changes are unique to semaglutide, and whether a grading system for DR will assist in assigning patients safely to GLP1-RA therapy.