Abstract: SA-OR43
Biomarker Panels for Discriminating Risk of CKD Progression in Children
Session Information
- Pediatric Nephrology and Development: Research Abstracts
October 24, 2020 | Location: Simulive
Abstract Time: 05:00 PM - 07:00 PM
Category: Pediatric Nephrology
- 1700 Pediatric Nephrology
Authors
- Greenberg, Jason Henry, Yale University, New Haven, Connecticut, United States
- Xu, Yunwen, Johns Hopkins University, Baltimore, Maryland, United States
- Abraham, Alison G., Johns Hopkins University, Baltimore, Maryland, United States
- Schelling, Jeffrey R., Case Western Reserve University, Cleveland, Ohio, United States
- Feldman, Harold I., University of Pennsylvania, Philadelphia, Pennsylvania, United States
- Shlipak, Michael, University of California San Francisco, San Francisco, California, United States
- Sabbisetti, Venkata, Brigham and Women's Hospital, Boston, Massachusetts, United States
- Warady, Bradley A., Children's Mercy Hospitals and Clinics, Kansas City, Missouri, United States
- Coca, Steven G., Mount Sinai Health System, New York, New York, United States
- Kimmel, Paul L., National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, Maryland, United States
- Bonventre, Joseph V., Brigham and Women's Hospital, Boston, Massachusetts, United States
- Denburg, Michelle, University of Pennsylvania, Philadelphia, Pennsylvania, United States
- Parikh, Chirag R., Johns Hopkins University, Baltimore, Maryland, United States
- Furth, Susan L., University of Pennsylvania, Philadelphia, Pennsylvania, United States
Group or Team Name
- CKD Biomarker Consortium
Background
We used multivariate survival trees to identify plasma and clinical biomarkers to predict CKD progression in children.
Methods
The CKiD study prospectively enrolled children aged 6 months to 16 years old with an eGFR of 30-90 and eGFR was assessed annually. The primary outcome of CKD progression was a composite of 50% decline in eGFR or incident ESKD. We used multivariate survival trees to determine combinations of baseline clinical predictors and plasma biomarkers as well as identify optimal thresholds for predicting the time to the composite event.
Results
Of the 651 children included, median age was 11 years [IQR,8-15], 405(62%) were male, 195(30%) had a glomerular cause of CKD, and baseline eGFR was 53 [IQR,40-67]. 223(34%) out of 651 children reached the primary outcome over a median follow-up time of 5.7 years. The Figure shows the best-sized multivariate survival tree and 4 prognosis groups selected after bootstrapping the sample. Plasma KIM1, TNFR1, TNFR2, and baseline eGFR were used to define branching patterns, while MCP1, YKL40, suPAR and the known risk factors of sex, age, glomerular diagnosis, BMI, hypertension, and proteinuria were not included as they did not reach a level of predictive importance. In the final model, KIM1 was the variable with the highest importance, with a level of 1335 mg/L(98th percentile) determining the first branching split and identifying the highest risk group of 14 children with predominantly glomerular types of kidney disease and nephrotic range proteinuria. When the tree-based prognosis classification was added to the clinical risk factors, the C-statistic increased from 0.81[95%CI:0.78-0.84] to 0.85[95%CI:0.83-0.88].
Conclusion
Using multivariate survival trees we identified a biomarker panel of plasma KIM1, TNFR1, TNFR2 and baseline eGFR which improved discrimination for CKD progression.
Funding
- NIDDK Support