An Inter-Laboratory Comparison for the Urinary Acrolein Biomarker 3-Hydroxypropyl-Mercapturic Acid (3-HPMA)

• Main Authors: Scherer Gerhard, Heller Wolf-Dieter, McEwan Michael, Göen Thomas, Joza Peter, Liu Nan, Newland Kirk, Schettgen Thomas, Wang Sheng, Sohn Hyung-Ok, Troude Valerie, Yuki Dai, Zheng Saijing, Zhou Guojun
• Format: Article
• Language: English
• Published: Sciendo 2017-01-01
• Series: Beiträge zur Tabakforschung International
• Subjects: Inter-laboratory comparison; 3-hydroxypropyl-mercapturic acid (3-HPMA); intra-laboratory variation; inter-laboratory variation; repeatability; reproducibility
• Online Access: http://www.degruyter.com/view/j/cttr.2017.27.issue-5/cttr-2016-0006/cttr-2016-0006.xml?format=INT

An inter-laboratory comparison study on the acrolein biomarker of exposure 3-hydroxypropyl-mercapturic acid (3-HPMA) with 12 laboratories from 7 globally distributed countries was performed. The laboratories received coded triplicates of 4 spiked and lyophilized urine samples (LU, 12 samples) as well as 5 authentic urine pool samples (PU, 15 samples) covering the 3-HPMA concentration range from background (non-smoking) to heavy smoking levels for analysis by using their own (in-house) analytical method. All laboratories applied liquid chromatography with tandem mass spectrometry (LC-MS/MS), with most of them (10 of 12) using solid phase extraction (SPE) as sample work-up procedure. The intra-laboratory variation (indicating repeatability) was determined by calculating the standard deviation (sr) and the coefficient of variation (CVr) of the triplicates, whereas the inter-laboratory variation (indicating reproducibility) was determined by calculating the standard deviation between laboratories (sR) and the corresponding coefficient of variation (CVR). After removal of outlier samples or laboratories, the mean CVr values for LU and PU test samples ranged from 2.1–3.6% (mean: 2.8%) and 2.4–3.7% (mean: 3.3%), respectively, indicating good repeatability for the determination of 3-HPMA in both sample types. CVR for LU and PU test samples ranged from 9.1–31.9% (mean: 18.8%) and 13.9–27.0% (mean: 18.5%), respectively, indicating limited reproducibility in 3-HPMA analysis for both sample types. Re-calculation of the PU results by applying an embedded calibration (EC), derived from the reported peak areas for the LU test samples, somewhat improved the CVR values (range: 9.6–28.8%, mean: 16.7%).