Comparison of hydrolysis and HPLC/MS/MS procedure with ELISA assay for the determination of S-phenylmercapturic acid as a biomarker of benzene exposure in human urine

Abstract The present study compared three methods for the determination of S-phenylmercapturic acid (S-PMA), a metabolite of benzene, in human urine: a HPLC/MS/MS technique with two different sample treatments (strong and partial hydrolysis) and a commercial assay based on anti-S-PMA monoclonal antibodies with chemiluminescence detection. Biological monitoring was done on 126 volunteers and the results were compared for the three methods and also with benzene exposure levels (range <3.0-592.5 μg/m(3)). The correlation between environmental monitoring data and S-PMA levels in non-smokers (n=73) was highly significant (p<0.0001, Student's t-test) for both HPLC/MS/MS methods (r=0.65 both for strong acidic hydrolysis of the urine and hydrolysis at pH 2) but not for the immunoassay, which overestimated the S-PMA levels by about 8 μg/g creatinine (creat.). Therefore the immunoassay is only useful as a semiquantitative screening test, but quantitative results need to be confirmed by a more accurate method like HPLC/MS/MS. The HPLC/MS/MS procedure with strong acid hydrolysis led to a recovery of S-PMA about double that using pH 2 hydrolysis, giving more accurate results. The difference between the results with the two methods makes it difficult to compare the strong acidic hydrolysis data with the ACGIH BEI value of 25 μg/g creat. since the BEI(®) documentation is based on data collected in pH conditions that were not always controlled, which may underestimate the true S-PMA concentration. Besides, as levels of benzene exposure were high, smoking was not considered a confounding factor. The BEI for S-PMA in end of shift urine samples could be reconsidered when sufficient data are available from studies where the analyses are carried out in comparable conditions of hydrolysis and monitoring only non-smoking subjects.