Metabonomic profiling of renal cell carcinoma: high-resolution proton nuclear magnetic resonance spectroscopy of human serum with multivariate data analysis.

Metabonomic profiling using proton nuclear magnetic resonance ((1)H NMR) spectroscopy and multivariate data analysis of human serum samples was used to characterize metabolic profiles in renal cell carcinoma (RCC). We found distinct, easily detectable differences between (a) RCC patients and healthy humans, (b) RCC patients with metastases and without metastases, and (c) RCC patients before and after nephrectomy. Compared to healthy human serum, RCC serum had higher levels of lipid (mainly very low-density lipoproteins), isoleucine, leucine, lactate, alanine, N-acetylglycoproteins, pyruvate, glycerol, and unsaturated lipid, together with lower levels of acetoacetate, glutamine, phosphatidylcholine/choline, trimethylamine-N-oxide, and glucose. This pattern was somewhat reversed after nephrectomy. Altered metabolite concentrations are most likely the result of the cells switching to glycolysis to maintain energy homeostasis following the loss of ATP caused by impaired TCA cycle in RCC. Serum NMR spectra combined with principal component analysis techniques offer an efficient, convenient way of depicting tumour biochemistry and stratifying tumours under different pathophysiological conditions. It may be able to assist early diagnosis and postoperative surveillance of human malignant diseases using single blood samples.