Validation of a plasma metabolomics model that allows anticipation of the occurrence of cytomegalovirus DNAaemia in allogeneic stem cell transplant recipients

A plasma metabolomic model obtained by means of untargeted ¹H nuclear magnetic resonance, to which taurine, choline, methylamine, total glutathione, trimethylamine N-oxide, lactate, lysine, isoleucine, total fatty acids and unsaturated fatty acids contributed, was validated for the prediction of first episodes of cytomegalovirus (CMV) DNAaemia in a cohort of 79 allogeneic stem haematopoietic stem cell transplant (allo-HSCT) recipients. The predictive success rate was nearly 65 % for patients at both low and high risk of CMV-related complications according to their baseline characteristics. Plasma metabolomics profiling shortly after engraftment (day 21 after transplantation) allowed the anticipation of the occurrence of CMV DNAaemia in 71 % of patients. Plasma metabolomics analyses may be ancillary for identifying allo-HSCT patients at the highest risk of CMV DNAaemia who may benefit from early targeted antiviral prophylaxis.