Measurement of total homocysteine in plasma and blood spots using liquid chromatography-tandem mass spectrometry: comparison with the plasma Abbott IMx method

Background: Current sampling for total homocysteine (tHcy) is problematic, requiring plasma separation within 15 min. The aim of this study was to develop a liquid chromatographic-tandem mass spectrometric (LC-MS/MS) method for the measurement of tHcy in plasma and dried blood spots and to determine whether the dried blood spot concentration could be used to predict plasma concentrations of tHcy. Methods: LC-MS/MS methodology was optimized to measure tHcy in plasma and dried blood spots. Fifty blood samples collected from heart transplant patients were used to form dried blood spots and for plasma analysis. Plasma tHcy was also measured using the Abbott IMx1 method and values were compared to the tHcy concentrations determined in plasma and dried blood spots using LC-MS/MS methodology. Results: The plasma tHcy LC-MS/MS results compared well with the IMx values: LC-MS/MS=1·18(IMx)-0·44 ( r2=0·915). The within-batch precision ( n =10) of the plasma LC-MS/MS method was < 2·0% at 14·6 and 37·7 µmol/L, respectively; the between-batch precision ( n=10) was 5·0 and 8·0%, respectively, at these concentrations. The method was found to be sensitive down to 1 µmol/L and linear up to at least 100 µmol/L. Dried blood spot LC-MS/MS results were considerably lower than the plasma IMx values ( P < 0·0001): dried blood spot LC-MS/MS=0·33IMx+1·77 ( r2=0·682). The within-batch precision ( n=20) of the dried blood spot LC-MS/MS method was 7·3% and 4·7% at concentrations of 4·0 and 7·9 µmol/L, respectively; the between-batch precision was 12·6% and 7·9% at concentrations of 5·1 and 8·0 µmol/L, respectively. To assess whether dried blood spots are suitable as a screening test to predict plasma tHcy concentrations, arbitary cut-off levels were compared. If it is assumed that a plasma tHcy concentration of >15 µmol/L is raised, a dried blood spot result of >6·8 µmol/L has a sensitivity and specificity in detecting a raised plasma tHcy of 83·3% and 96·2%, respectively, and a positive and negative predictive value of 95% and 86%, respectively, with an efficiency of 90%. Use of a dried blood spot cut-off concentration of 6·2 µmol/L for predicting high plasma tHcy concentrations (above 15 µmol/L) has a sensitivity and specificity of 95·8% and 73·1%, respectively, positive and negative predictive values of 76% and 95%, respectively, and an efficiency of 84%. Conclusions: We have developed a precise and accurate LC-MS/MS method for measuring plasma tHcy concentrations, which uses a small volume of plasma and is suitable for routine use. A satisfactory LC-MS/MS method for the measurement of tHcy in dried blood spots was also developed; this method might be useful in routine screening for raised plasma concentrations of tHcy.