New analytical methods and old BEI lead to the risk to underestimate styrene occupational exposure

Introduction
Mandelic (MA) and phenylglyoxylic (PGA) acids are metabolites of styrene whose total concentration is the dose biomarker suggested by ACGIH for occupational exposure (BEI 400 mg/g of creatinine in end shift urine). They are considered unspecific biomarkers, as a background value is always found in human urine. Most analytical methods applied to biomonitoring use HPLC-UV, a non highly specific technique.

Methods
The styrene exposure of 13 fiberglass workers was assessed by means of personal air monitoring and biomonitoring. The quantitative determination of urinary MA and PGA was performed before and after the workshift on these workers and on 13 non exposed controls by HPLC/MS/MS using the isotopic dilution method.

Results
Results of the workers biomonitoring are lower (mean value 85, range 3 - 188 mg/g of creatinine in end shift urine) than the ACGIH BEI, but significantly higher than for controls (mean value 2, range 0.1 - 17 mg/g of creatinine ). The levels before and after the workshift do not change significantly, showing that the internal dose remains high for several hours. Personal air monitoring data are linearly correlated with those from biomonitoring (r = 0.77) and show that the average occupational exposure to styrene is 36.6 µg/m3, range 4.9–72.3 µg/m3 (ACGIH TLV = 85 mg/m3).

Discussion
Highly specific methods, like HPLC with tandem mass spectrometry, can measure urinary metabolite more reliably and at lower levels than those obtained with less specific techniques, like HPLC/UV . This is the case for MA and PGA, whose BEI have been set on the basis of HPLC/UV results. As a result the comparison of HPLC/MS/MS results with the current limit of 400 mg/g of creatinine may underestimate real styrene exposure. The collection of new results, obtained using highly specific analytical methods, is a research priority in order to revise BEIs.