Clinical metabonomics for advanced laboratory medicine

Abstract

Metabonomics is an unbiased and comprehensive quantitative study of metabolomes in a biological system. Using a metabonomics approach, new insight in disease pathogenesis and novel diagnostic biomarkers with improved accuracy can be identified with clinical impact.

This PhD study focused on clinical metabonomics in advanced laboratory medicine. In part I, it was focused on the biomarker discovery for urinary tract infection (UTI) using nuclear magnetic resonance (NMR)-based metabonomics. Urine acetic acid and trimethylamine (TMA) were determined to be the best biomarker for diagnosing bacterial and Escherichia coli (EC)-associated UTI respectively. The area-under-curve (AUC) for acetic acid was 0.97 (95% confidence interval (CI): 0.94 – 1.00) and that for TMA was 0.85 (95% CI: 0.76 – 0.92). The performances were superior to conventional laboratory tests, i.e. dipstick urinalysis and urine microscopy. The high urine acetic acid was secondary to bacterial anaerobic respiration whereas TMA was a microbial-mammalian co-metabolite which was converted from trimethylamine N-oxide (TMAO), a human metabolite by EC TMAO reductase.

In part II, the diagnostic application of acetic acid in bacterial UTI was extend to spontaneous bacterial peritonitis (SBP). Rapid diagnosis for SBP is vital because antibiotics treatment could dramatically reduce morbidity and mortality. Using NMR-based metabolomics approach, peritoneal fluid acetic acid was significantly elevated in SBP patients but undetected in controls. This finding echoed the results in part I and supported acetic acid as a biomarker for active bacterial infection.

In part III, it was focused on the classification of pleural effusions (PE) using NMR-based metabonomics. PE lipoprotein was determined to be the best biomarker to differentiate exudates from transudates with an AUC of 0.96 (95% CI: 0.89 – 0.99). The diagnostic performance was superior to the Light's criteria recommended by International guideline. Elevated lipoprotein in exudates was secondary to increased capillary pore-size of the inflamed pleura. Therefore, quantification of PE lipoprotein will help assessing disease severity which is a novel and only clinical test for in vivo pleural permeability measurement.

To determine the aetiology of exudates, liquid chromatography-tandem mass spectrometry (LC-MS/MS)-based metabonomics was applied for cancerous and tuberculous PE, the two major causes of exudates. High levels of PE oleic acid and ceramide (d18:1/16:0) were found in cancerous and tuberculous PE respectively. The high oleic acid was secondary to fatty acid synthase (FASN) over-expression in cancer cells while the high ceramide was secondary to host immune response. Using a ratio of the two to predict cancerous and tuberculous PE, the AUC was increased to 0.99 (95% CI: 0.91 – 1.00).

In part IV, it was focused on the rapid diagnosis of inborn errors of metabolism (IEM) using NMR spectroscopy. Patients with IEM can present with acute metabolic decompensation. Currently, most clinical laboratories rely on gas chromatography-mass spectrometry (GC-MS) for the diagnosis of IEM which usually takes a considerable time, i.e. day(s). The feasibility of NMR spectroscopy as a rapid diagnostic test was evaluated using various IEM conditions. Pathognomonic metabolites were identified using NMR-based urinalysis in <15 minutes. This supports the essential clinical role of NMR spectroscopy for rapid diagnosis of IEM.