Abstract: PO0981
Association Between the Urinary Proteome and Diabetic Retinopathy in the DIRECT-Protect 1 and 2 Trials
Session Information
- Diabetic Kidney Disease: Clinical - 1
October 22, 2020 | Location: On-Demand
Abstract Time: 10:00 AM - 12:00 PM
Category: Diabetic Kidney Disease
- 602 Diabetic Kidney Disease: Clinical
Authors
- Rotbain Curovic, Viktor, Steno Diabetes Center Copenhagen, Gentofte, Denmark
- Magalhães, Pedro, Mosaiques Diagnostics, Hannover, Germany
- Hansen, Tine, Steno Diabetes Center Copenhagen, Gentofte, Denmark
- Mischak, Harald, Mosaiques Diagnostics, Hannover, Germany
- Rossing, Peter, Steno Diabetes Center Copenhagen, Gentofte, Denmark
Background
Diabetic retinopathy (DR) is a complication of paramount importance in type 1 and type 2 diabetes. Given the association of DR and diabetic kidney disease (DKD), we investigated the association between the urinary proteome and the presence and deterioration of DR in individuals with type 1 and type 2 diabetes.
Methods
Baseline proteomic analysis was performed in both the DIRECT-Protect 1 and 2 studies in a random selection of 800 and 821 subjects respectively. The DIRECT-Protect studies were designed to assess the effect of candesartan in relation to development of DR endpoints. DIRECT-Protect 1 was considered the discovery cohort and DIRECT-Protect 2 the validation cohort. Endpoints were defined as a two-step (RET2) or three-step (RET3) change in DR score according to the Early Treatment Diabetic Retinopathy Study (ETDRS) protocol. Urinary peptide levels were correlated to baseline DR score in the discovery cohort. Thereafter, identified peptide fragments were investigated for association to baseline DR score in the validation cohort and for development of endpoints in both cohorts, adjusted for sex, age, diabetes duration, smoking, total cholesterol, HbA1c, systolic blood pressure, urinary albumin excretion rate, serum creatinine, and randomization group at baseline.
Results
Follow-up ranged from 4.0-4.7 years. Eleven out of 427 peptide fragments were inversely associated to baseline DR in the discovery cohort after adjustment. In multivariate Cox regression analyses lower alpha-1 type I collagen (COL1A1) (seq. ApGD~, GKNG~, and LDGA~) was significantly (p<0.05) associated to the development of RET2 and lower COL1A1 (seq. LDGA~) and COL3A1 (seq. EDGK~) to RET3. However, when attempting to validate these results, only a KER12 fragment was inversely associated to baseline DR in the validation cohort, as well as to development of RET3. Furthermore, lower levels of one COL1A1 fragment (seq. AEGS~) was associated to development of RET2 in the validation cohort.
Conclusion
Several urinary peptide fragments were associated to the presence and worsening of DR in type 1 diabetes. However, this could not be validated in type 2 diabetes, and the identified peptide fragments were not conclusively associated to deterioration of DR across both cohorts.