Biomarker Insights (Mar 2017)

A Novel Biomarker Panel to Identify Steroid Resistance in Childhood Idiopathic Nephrotic Syndrome

  • Michael R Bennett,
  • LaTawnya Pleasant,
  • Christopher Haffner,
  • Qing Ma,
  • Wendy D Haffey,
  • Jun Ying,
  • Michael Wagner,
  • Kenneth D Greis,
  • Prasad Devarajan

DOI
https://doi.org/10.1177/1177271917695832
Journal volume & issue
Vol. 12

Abstract

Read online

Idiopathic nephrotic syndrome (NS) is the most common glomerular disorder of childhood. Response to initial treatment with corticosteroids is an indicator of prognosis, as resistant patients often present more progressive disease. In this cross-sectional pilot study, we set out to discover a panel of noninvasive biomarkers that could distinguish steroid-resistant nephrotic syndrome (SRNS) from steroid-sensitive nephrotic syndrome (SSNS). Information gleaned from such a panel could yield more individualized treatment plans and prevent unnecessary steroid exposure in patients unlikely to respond. Urine was collected from 50 pediatric patients diagnosed with idiopathic NS at Cincinnati Children’s Hospital Medical Center. Isobaric tags for relative and absolute quantitation (iTRAQ) was used to discover 13 proteins that were differentially expressed in SSNS vs SRNS in a small 5 × 5 discovery cohort. Suitable assays were found for 9 of the 13 markers identified by iTRAQ and were used in a 25 SRNS × 25 SSNS validation cohort. Vitamin D–binding protein (VDBP), alpha-1 acid glycoprotein 1 (AGP1), alpha-1 acid glycoprotein 2 (AGP2), alpha-1-B glycoprotein (A1BG), fetuin-A, prealbumin, thyroxine-binding globulin and hemopexin, and alpha-2 macroglobulin were measured and combined with urine neutrophil gelatinase–associated lipocalin (NGAL), which had been previously shown to distinguish patients with SRNS. Urinary VDBP, prealbumin, NGAL, fetuin-A, and AGP2 were found to be significantly elevated in SRNS using univariate analysis, with area under the receiver operating characteristic curves (AUCs) ranging from 0.65 to 0.81. Multivariate analysis revealed a panel of all 10 markers that yielded an AUC of 0.92 for identification of SRNS. A subset of 5 markers (including VDBP, NGAL, fetuin-A, prealbumin, and AGP2) showed significant associations with SRNS and yielded an AUC of 0.85.