Abstract: FR-PO442
High-Throughput Proteomic Search Reveals Novel Plasma Biomarkers of Time to ESRD in Type 1 Diabetes Patients with Persistent Proteinuria and CKD Stage 1-2
Session Information
- Diabetic Kidney Disease: Clinical - I
October 26, 2018 | Location: Exhibit Hall, San Diego Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: Diabetic Kidney Disease
- 602 Diabetic Kidney Disease: Clinical
Authors
- Skupien, Jan, Jagiellonian University Medical College, Krakow, Poland
- Smiles, Adam, Joslin Diabetes Center, Boston, Massachusetts, United States
- Krolewski, Andrzej S., Joslin Diabetes Center, Boston, Massachusetts, United States
Background
The risk of end-stage renal disease (ESRD) reaches 30-50% in patients with type 1 diabetes and persistent proteinuria. An intensive research effort is needed to identify and evaluate interventions that improve prognosis in these patients. To better understand the disease process, identify novel drug targets and improve end-point definitions in clinical trials new biomarkers of renal function decline and ESRD risk are desperately needed.
Methods
We assayed baseline plasma specimens using an Olink proteomic platform (Uppsala, Sweden) in a group of 82 patients with type 1 diabetes and CKD stage 1-2 from Joslin Proteinuria cohort and replicated the findings in an additionally ascertained sample of 76 patients with similar baseline characteristics. Using an accelerated failure time survival regression model we sought association of 454 proteins with time to ESRD. Association was considered significant if false discovery rate-adjusted p-value <0.01 and a p-value <0.01 in the replication set.
Results
We identified 31 plasma proteins significantly associated with time to ESRD. The largest number of them represented tumor necrosis factor signaling pathway. All markers were associated with increased risk of ESRD, i.e. their higher concentration was associated with shorter time to ESRD. Most proteins were highly correlated and redundant as biomarkers. Three of them were independently associated with time to ESRD and are shown in the Table. Using these three proteins we estimated patients' time to ESRD. The median predicted time in patients with ESRD was 9.7 years (observed 8.2 years) and in those without ESRD the median predicted time was 15.8 years.
Conclusion
There are multiple novel biomarkers of fast progression to ESRD in patients with type 1 diabetes and proteinuria. These proteins should be further investigated for their biologic role and therapeutic potential. They may serve in the future for predicting time to ESRD in diabetic kidney disease.
Multivariate predictive model of time to ESRD
| Protein | Effect of doubling concentration on ESRD time | P-value |
| HAVCR1 | -23.3% | <0.001 |
| WFDC2 | -29.5% | 0.001 |
| PGLYRP1 | -13.3% | 0.049 |
Funding
- NIDDK Support