Endogenous high-performance liquid chromatography/tandem mass spectrometry interferences and the case of perfluorohexane sulfonate (PFHxS) in human serum; are we overestimating exposure?

Perfluorinated acids have received increasing scientific attention due to their widespread global distribution, environmental persistence and bioaccumulation in wildlife and humans. For perfluorohexane sulfonate (PFHxS, C(6)F(13)SO(3) (-), m/z 399), all existing human data have been generated using high-performance liquid chromatography (HPLC) and its most sensitive tandem mass spectrometric (MS/MS) transitions (m/z 399/80 [SO(3)](-) or m/z 399/99 [SO(3)F](-)), but this may be problematic because of co-eluting endogenous steroid sulfates that share common fragmentation pathways. We examined the magnitude of over-reporting for PFHxS in pregnant women (n = 29), and in pooled serum of males, non-pregnant and pregnant females (n = 3, 100 samples per pool), by comparing m/z 399/80 and 399/99 data with an interference-free transition, m/z 399/119. PFHxS concentrations in pregnant women determined using m/z 399/80 and 399/99 (p < 0.05), but not m/z 399/119, were positively correlated to the response of the steroid sulfates. This led to an average overestimation of PFHxS by 1.5- and 4.7-fold, using m/z 399/80 and 399/99, respectively, and validated the use of m/z 399/119 for the first time. The interferences were a problem in all human serum samples, and analysis of pooled serum revealed statistically significant over-reporting by m/z 399/80 and 399/99 for pregnant women > non-pregnant women > men. The magnitude of over-reporting here represents a worst-case scenario, but the extent to which the published literature values are biased is unknown due to limited details of methods in existing reports. Instead of using the less sensitive m/z 399/119 transition, we showed that an alternative selection of column and mobile phase can allow for sufficient chromatographic separation of the interferences. In conclusion, it was shown that routine analytical methods are prone to systematically overestimating PFHxS concentrations in serum of men or women, but that this can be avoided by alternative chromatographic steps or MS/MS transitions. Copyright (c) 2009 John Wiley & Sons, Ltd.