Toxicodynamic effects of ciclosporin are reflected by metabolite profiles in the urine of healthy individuals after a single dose

Abstract

WHAT IS ALREADY KNOWN ABOUT THE SUBJECT * Ciclosporin's nephrotoxicity initially targets the proximal tubule and is, at least in part, driven by increased formation of oxygen radicals. * (1)H-nuclear magnetic resonance spectroscopy (NMR)- and mass spectrometry (MS)-based biochemical profiling (metabolomics) allows for the sensitive detection of metabolite pattern changes in urine. * In systematic studies in rats we showed that ciclosporin caused urine metabolite pattern changes typical for proximal tubule damage and that these pattern changes seemed to be more sensitive than established clinical kidney function markers such as serum creatinine concentrations. WHAT THIS PAPER ADDS * This study showed that urine metabolite pattern changes as assessed by (1)H-NMR and HPLC-MS are sensitive enough to detect the effect of ciclosporin as early as 4 h after a single oral dose. * In our previous rat studies, changes in urine metabolite pattern in response to ciclosporin translated into healthy humans, indicating the involvement of the same toxicodynamic mechanisms. * The results provide proof of concept for further development of this combination molecular marker strategy into diagnostic tools for the detection and monitoring of drug nephrotoxicity. AIMS The immunosuppressant ciclosporin is an efficient prophylaxis against transplant organ rejection but its clinical use is limited by its nephrotoxicity. Our previous systematic studies in the rat indicated urine metabolite pattern changes to be sensitive indicators of the negative effects of ciclosporin on the kidney. To translate these results, we conducted an open label, placebo-controlled, crossover study assessing the time-dependent toxicodynamic effects of a single oral ciclosporin dose (5 mg kg(-1)) on the kidney in 13 healthy individuals. METHODS In plasma and urine samples, ciclosporin and 15-F(2t)-isoprostane concentrations were assessed using HPLC-MS and metabolite profiles using (1)H-NMR spectroscopy. RESULTS The maximum ciclosporin concentrations were 1489 +/- 425 ng ml(-1) (blood) and 2629 +/- 1308 ng ml(-1) (urine). The increase in urinary 15-F(2t)-isoprostane observed 4 h after administration of ciclosporin indicated an increase in oxidative stress. 15-F(2t)-isoprostane concentrations were on average 2.9-fold higher after ciclosporin than after placebo (59.8 +/- 31.2 vs. 20.9 +/- 19.9 pg mg(-1) creatinine, P < 0.02). While there were no conclusive changes in plasma 15-F(2t)-isoprostane concentrations or metabolite patterns, non-targeted metabolome analysis using principal components analysis and partial least square fit analysis revealed significant changes in urine metabolites typically associated with negative effects on proximal tubule cells. The major metabolites that differed between the 4 h urine samples after ciclosporin and placebo were citrate, hippurate, lactate, TMAO, creatinine and phenylalanine. CONCLUSION Changes in urine metabolite patterns as a molecular marker are sufficiently sensitive for the detection of the negative effects of ciclosporin on the kidney after a single oral dose.

Figure 1

Calibration curves of key metabolites as measured by ¹H-NMR. Heparinized whole blood was enriched with the metabolites and extracted using a methanol : chloroform extraction procedure [–33]. Abbreviations: GSH, glutathione; Glu, glutamate; TMAO, trimethylamine oxide. TMSP (1.1 mm) has been used as an internal standard compound for quantitation. Data are presented as means (n= 3). Standard deviations (<13.1%) are not shown to facilitate visual comparison. GSH (); Glu (); lactate (); creatinine (); citrate (); TMAO ()

Figure 2

Ciclosporin blood (A) and urine concentrations (B) after oral administration of a single ciclosporin 5 mg kg⁻¹ body weight dose in its Neoral formulation. All concentrations are presented as means ± SD (n= 13)

Figure 3

Individual comparisons of 15-F_2t-isoprostane concentrations 4 h after administration of placebo or 5 mg kg⁻¹ ciclosporin in its Neoral formulation. The results of seven subjects are shown. One subject had urine isoprostane concentrations below the lower limit of quantitation. The samples of six subjects were lost due to an analytical error

Figure 4

Principal component analysis based on ¹H-NMR spectra of blood samples after oral administration of a single dose of 5 mg kg⁻¹ body weight ciclosporin or placebo. All blood samples were drawn 4 h post dose. Fourier transformation, phase and cubic splines baseline corrections were performed on ¹H-NMR spectra prior to reduction to bucket histograms with 0.04 ppm bucket width. The spectral regions between 4.3 and 6.0 ppm and 3.33–3.38 ppm were excluded due to the water pre-saturation sequence used for data acquisition and rest methanol signal. Bucket values were scaled to the total integral. One outlier spectrum is marked by an arrow. placebo 4 h (○); Neoral 4 h (▵)

Figure 6

Comparison of the Neoral solutions with and without ciclosporin (placebo) used during the study. The polyethylene glycol (PEG) signal matched that in the subjects' blood and urine samples and the concentrations in both Neoral batches were similar. TMSP: trimethylsilyl propionic-2,2,3,3,d₄ acid (used as internal shift reference standard)

Figure 5

Principal component analysis based on ¹H-NMR spectra of urine samples after oral administration of a single dose of 5 mg kg⁻¹ body weight ciclosporin or placebo. Urine samples were taken 4 h after administration. Fourier transformation, phase and cubic splines baseline corrections were performed on ¹H-NMR spectra prior to reduction to bucket histograms with 0.04 ppm bucket width. The spectral region between 4.3 and 6.0 ppm was excluded. Bucket values were scaled to the creatinine signal region from 3.02–3.07 ppm. placebo (○); Neoral (▵)

Figure 7

Partial least square fit (PLS) analysis of ¹H-NMR spectra of urine samples after oral administration of Neoral without or with ciclosporin (5 mg kg⁻¹ body weight). Only urine samples collected 4 h after administration were included. Fourier transformation, phase and cubic splines baseline corrections were performed on ¹H-NMR spectra prior to reduction to bucket histograms with 0.04 ppm bucket width. Bucket values were scaled to the total integral. The spectral regions between 4.3 and 6.0 ppm was excluded due to the water pre-saturation sequence used for data acquisition. placebo (○); Neoral (▵)