Sensitive determination of ethylene glycol, propylene glycol and diethylene glycol in human whole blood by isotope dilution gas chromatography–mass spectrometry, and the presence of appreciable amounts of the glycols in blood of healthy subjects

Ethylene glycol (EG), propylene glycol (PG), and diethylene glycol (DEG) are widely used as components of antifreeze liquids for automobiles and in many other products. They occasionally cause severe poisonings when they are ingested in high doses. In this study, we established a detailed procedure for highly sensitive and simultaneous determination of EG, PG, and DEG in human whole blood by isotope dilution gas chromatography–mass spectrometry (GC–MS). A 0.25-ml aliquot of whole blood containing the glycols was mixed with 100 ng each of deuterated EG, PG, and DEG as internal standards. After centrifugation, the supernatant fraction was evaporated to dryness, derivatized with heptafluorobutyric anhydride, and the derivative was extracted with n-hexane; 1 μl of the n-hexane layer was subjected to GC–MS analysis. Contrary to our expectation, appreciable amounts of EG, PG, and DEG were found, even in human whole blood samples obtained from nonoccupational healthy subjects. Because such results have never been reported, in the early stages of this study, we suspected that our results were caused by carry-over of the glycols during the GC–MS procedure; thus, we cleaned the injection port, installed a new GC column, and washed the ion-source chamber. Despite these efforts, the clear peaks were detected at the same retention times as those of the authentic glycols for the extracts of healthy subjects, which led us to confirm the presence of the glycols in the blood of healthy subjects. Because the blank whole blood samples from healthy subjects already contained glycols, we constructed a standard addition calibration curve for each sample; the horizontal intercept point, where the straight calibration line intersected the horizontal concentration axis, showed the background concentration of the glycol. The calibration curves showed linearity in the range of 0.4–400 ng/ml for EG and DEG, and in the range of 4–2,000 ng/ml for PG, with correlation coefficients larger than 0.99. The limits of detection and limits of quantitation could not be exactly determined, because of the unavailability of blank human whole blood without each glycol. However, the values were estimated to be around or lower than 1 ng/ml from the signal-to-noise ratios of peaks for each glycol at the lowest concentrations obtained by selected ion monitoring. To validate the method, intraday and interday repeatability was tested; the percent relative standard deviations were 2.5–12.2 % and 0.8–8.5 %, respectively. Recoveries of the three glycols using whole blood samples of two subjects obtained by a unique method were 61.6–80.9 %. The concentration ranges (mean) of EG, PG, and DEG in whole blood obtained from ten subjects were 39.1–97.0 (64.0), 49.1–689 (181), and 8.08–22.9 (11.1) ng/ml, respectively. To test the effect of oral intake of PG, two volunteers ingested 100 ml of a commercially available energy drink containing 33.7 mg of PG, the safety of which was accredited by the Japanese Government. The PG levels in the blood of the two subjects increased by 74.0 and 158 % at 1 and 0.5 h, respectively, suggesting that the glycols present in human blood are largely derived from food. To our knowledge, the analytical method for the glycols presented in this article is the most sensitive among those so far reported. In addition, this is the first description of the presence of the glycols in healthy human subjects, which should be useful for setting cutoff levels of the glycols in poisoning cases.