Based on experimental and preclinical studies, DPP-4 inhibition was expected to play a role in protecting vascular complications beyond glycemic control [3]. However, controversies and concerns still exist regarding the safety of DPP-4 inhibitors, especially in diabetic retinopathy. DPP-4 inhibitors not only increase glucagon-like peptide-1 concentrations but also chemokines such as stromal cell-derived factor 1α (SDF-1α), which is closely related with angiogenesis and neovascularization. In murine models, a DPP-4 inhibitor aggravated diabetic retinopathy by increasing vascular permeability through SDF-1α/C-X-C chemokine receptor type 4 (CXCR4)/Src/vascular endothelial (VE)-cadherin signaling pathway [4].

In this issue, Chung et al. [5] demonstrate the effects of DPP-4 inhibitors as add-on therapy to metformin on the progression of diabetic retinopathy in real-world setting. They compared DPP-4 inhibitors with sulfonylurea or thiazolidinedione using national database, and DPP-4 inhibitors showed greater benefit than sulfonylurea on diabetic retinopathy progression regardless of comorbidities, duration of metformin use, intravitreal injections and calendar index year (hazard ratio [HR], 0.80; 95% confidence interval [CI], 0.66 to 0.97). Although the changes were statistically insignificant after adjustment including fasting blood glucose levels in subgroup analysis (HR, 0.89; 95% CI, 0.64 to 1.24; P=0.493), DPP-4 inhibitors did not aggravate diabetic retinopathy progression. These results are consistent with a prior study [6] based on the same database (but different enroll period and design), so DPP-4 inhibitors do not seem to aggravate the progression of diabetic retinopathy. From a clinician's perspective, these evidence supports that the use of DPP-4 inhibitors and does not violate the principle of ‘Primum non nocere’. However, there are some factors to be considered before hasty conclusions are made.

First of all, the lack of critical risk factors, as the authors commented, does not strongly guarantee the safety of DPP-4 inhibitors in this study. Since higher glycosylated hemoglobin and longer duration of diabetes were undeniable risk factors for the progression of retinopathy, careful interpretation is required when comparing safety among drugs. Recent studies conducted in Korea showed that the average time from diagnosis to first panretinal photocoagulation is over 17 years, and one-third of patients with over 15 years of disease duration were free from diabetic retinopathy [78].

Second, the lesson to be learnt from the clinical trials is that the safety of each DPP-4 inhibitors could be different between drugs in the same class [9], so more detailed analysis might be needed to determine the class effect of DPP-4 inhibitors. In addition, there is still no specific antidiabetic agent class showing protective effects against the progression of diabetes. Thus, further long-term and large prospective cohort studies would clarify whether a specific class has protective effects beyond the glucose-lowering effect.