Kidney Week

Abstract: FR-PO991

Urinary Metabolic Profiling by NMR Spectrometry Associates with eGFR in a Cohort of Patients with ADPKD

Session Information

• PKD Cellular Pathogenesis, ARPKD, ADTKD, Ciliopathies
  October 26, 2018 | Location: Exhibit Hall, San Diego Convention Center
  Abstract Time: 10:00 AM - 12:00 PM

Category: Genetic Diseases of the Kidney

• 1001 Genetic Diseases of the Kidney: Cystic

Authors

• Dekker, Shosha E.I., Leiden University Medical Center, Leiden, Netherlands

Shosha E.I. Dekker,

• Verhoeven, Aswin, Leiden University Medical Center, Leiden, Netherlands

Aswin Verhoeven,

• Soonawala, Darius, Leiden University Medical Center, Leiden, Netherlands

Darius Soonawala,

• Mayboroda, Oleg, Leiden University Medical Center, Leiden, Netherlands

Oleg Mayboroda,

• De Fijter, Johan W., Leiden University Medical Center, Leiden, Netherlands

Johan W. De Fijter,

Background

With the advent of therapeutic options to slow the rate of disease progression in patients with autosomal dominant polycystic kidney disease (ADPKD), there is an unmet need for biomarkers to select those at high-risk of rapid progression in the early stages of their disease. This study explored changes in the urinary metabolome associated with different chronic kidney disease (CKD) stages or estimated glomerular filtration rate (eGFR) in ADPKD, which is a first step towards identifying markers for early risk stratification.

Methods

Quantitative metabolic profiling (¹H NMR-spectrometry) was performed on spot urine samples obtained from 338 ADPKD patients with various CKD-stages, 84 healthy controls and 42 chronic, non-ADPKD, renal disease patients. Uni- and multivariate analyses were used to detect differences between the groups and associations between metabolic profiles and eGFR (CKD-EPI equation). Findings were validated in an independent cohort of 163 ADPKD patients.

Results

Twenty-nine urinary metabolites were quantified from the spectra. Logistic regression reliably distinguished ADPKD patients from healthy controls (AIC(DF 30,392)=185.7, p=6×10^-46, pR²=0.70; cross-validated AUC of 0.947 for CKD stage 1 in ADPKD). In patients with ADPKD, six metabolites were significantly decreased and seven significantly increased as compared to healthy controls, while metabolic profiling did not discriminate ADPKD from other renal diseases with comparable eGFRs. Sixteen metabolites correlated with eGFR in ADPKD patients, with the metabolite with the highest negative correlation having a pearson correlation coefficient of -0.72 (p=5×10^-55) and the highest positive correlation of r=0.66 (p<2×10^-16). A multiple linear regression model for eGFR was constructed based on all metabolite levels (F(30,308)=39.95; p=5×10^-87, R²=0.79). A much simpler model, however, based on the ratio of the two metabolites with the highest correlation only also served as a good predictor for eGFR (F(2,336)=536.2; p=1×10^-71, R²=0.615). Using the later model, the results were validated in a separate cohort (F(30,133)=6.293; p=6×10^-14, R²=0.578).

Conclusion

Quantitative NMR profiling enabled identification of urinary metabolic markers that distinguished ADPKD patients from healthy controls and markers that were also significantly associated with eGFR.