Noninvasive Urinary Monitoring of Progression in IgA Nephropathy

Abstract

Standard methods for detecting and monitoring of IgA nephropathy (IgAN) have conventionally required kidney biopsies or suffer from poor sensitivity and specificity. The Kidney Injury Test (KIT) Assay of urinary biomarkers has previously been shown to distinguish between various kidney pathologies, including chronic kidney disease, nephrolithiasis, and transplant rejection. This validation study uses the KIT Assay to investigate the clinical utility of the non-invasive detection of IgAN and predicting the progression of renal damage over time. The study design benefits from longitudinally collected urine samples from an investigator-initiated, multicenter, prospective study, evaluating the efficacy of corticosteroids versus Rituximab for preventing progressive IgAN. A total of 131 urine samples were processed for this study; 64 urine samples were collected from 34 IgAN patients, and urine samples from 64 demographically matched healthy controls were also collected; multiple urinary biomarkers consisting of cell-free DNA, methylated cell-free DNA, DMAIMO, MAMIMO, total protein, clusterin, creatinine, and CXCL10 were measured by the microwell-based KIT Assay. An IgA risk score (KIT-IgA) was significantly higher in IgAN patients as compared to healthy control (87.76 vs. 14.03, p < 0.0001) and performed better than proteinuria in discriminating between the two groups. The KIT Assay biomarkers, measured on a spot random urine sample at study entry could distinguish patients likely to have progressive renal dysfunction a year later. These data support the pursuit of larger prospective studies to evaluate the predictive performance of the KIT-IgA score in both screening for non-invasive diagnosis of IgAN, and for predicting risk of progressive renal disease from IgA and utilizing the KIT score for potentially evaluating the efficacy of IgAN-targeted therapies.

Keywords: IgA nephropathy; KIT assay; KIT-IgA score; diagnostics; noninvasive; prediction.

Conflict of interest statement

The samples from the patient cohort in this study came from an investigator-initiated study sponsored by Genentech/Roche, Inc. and the Fulk Family Foundation. The sponsors had no role in study design, protocol development, data analysis, or preparation of the manuscript. M.M.S. and J.Y.C.Y. are founders of KIT Bio, Inc. (San Francisco, CA), now operating under NephroSant, IP for which is exclusively owned by the Regents, University of California San Francisco and licensed to KIT Bio. R.L. is a scientific advisor of KIT Bio. All other authors declare no conflict of interest.

Figures

Figure 1

Study design and patient disposition. (Left) In the original trial, 34 patients met inclusion criteria and were randomized into Rituximab and standard of care treatment groups. At least one urine sample was available from 28 of the 34 patients, with 14 having urine samples at all three designated time-points. (Right) Pictorial depiction of patients, treatment, and sample availability. Patients were segregated based on treatment, either with standard of care (turquoise) or Rituximab (coral), with individual patients as rows. A yellow square indicates a urine sample was available at the indicated time-point, while gray indicates that no urine sample was available for analysis due to failure to collect or insufficient sample volume.

Figure 2

Changes in eGFR (by Modification of Diet in Renal Disease (MDRD) [16]) from baseline. Data shown here for the change in eGFR values over the time-course of the study for a subset of 14 patients who had complete urine samples collected at all three study time-points. Each line represents an individual patient trajectory. Trajectories in teal represent patients who received standard of care while those in salmon represent those who received Rituximab.

Figure 3

The urinary KIT biomarkers could segregate healthy controls from those with IgA nephropathy. (A) An IgA risk score ranging from 0 to 100 segregated healthy control patients from those with IgA nephropathy. Urine samples were collected from healthy controls (n = 64) who had no evidence of kidney disease or injury as assessed by both absence of proteinuria and eGFR greater than 120. All urine samples from IgA patients (n = 67) were used, as none of these patients had remission of IgA during the treatment duration. (B) Receiver–operator characteristic (ROC) curves of the IgA risk score with AUC of 0.994 (p < 0.0001) and proteinuria. For the IgA risk score, the sensitivity and specificity were 95.5% and 98.4% respectively. **** p < 0.0001.

Figure 4

KIT Assay biomarker modeling of progression status after one year of treatment. Modeling was performed on endpoint, midpoint, and baseline biomarker data on either (A) the set of KIT Assay biomarkers or (B) proteinuria alone. The y-axis shows the probability of progression as determined by a nominal logistic regression model.