Figure 3. Metabolome changes as a consequence of the surgical procedure.

(a) Principal component analysis (PCA) performed on serum samples collected intraoperatively (‘pre-CPB’: after sternotomy, but before CPB; ‘post-CPB’: immediately after CPB, at 0 hour). (b) The corresponding loadings plot shows metabolite composition on PC2. (c) Partial least-square (PLS) regression analysis performed on serum NMR spectra from samples collected immediately after weaning from CPB showed cross-validated (CV) associations with the actual duration of CPB. For validation purpose a ten-fold Venetian-Blinds CV was employed. (d) Metabolites found to correlate with the duration of CPB, cross-clamp, and CABG, and their time-line trends until 2 hours post-CPB (black: positive correlation; white: no significant correlation; red: negative correlation). Abbreviations: 3-HBA, 3-hyroxybutyric acid; FA, fatty acids; PUFA, polyunsaturated fatty acids; TMAO, trimethylamine-N-oxide; ppm, parts per million; 3.73 ppm*, metabolite putatively identified as poly ethylene glycol (a widely used excipient for drug formulation).