Evaluation of enzyme-linked immunosorbent assay for the determination of polycyclic aromatic hydrocarbons in house dust and residential soil.

Two commercially available enzyme-linked immunosorbent assays (ELISA) for total polycyclic aromatic hydrocarbon (PAH) and carcinogenic PAH (C-PAH) were evaluated. The testing procedures were refined for application to screening PAH and C-PAH in house dust and soil samples for human exposure studies. The overall method precision expressed as percent relative standard deviation (%RSD) of triplicate real world dust and soil samples was within +/- 29% (12-29%) for PAH ELISA and +/- 21% (5.9-21%) for C-PAH ELISA. Spike recoveries from real world dust/soil samples were 114 +/- 30% for phenanthrene from PAH ELISA and 120 +/- 8.2% for benzo[a]pyrene from C-PAH ELISA. The overall method accuracy for PAH and C-PAH assays cannot be assessed for multiple PAH components in dust/soil samples (which represent real-world samples), because of the assays' cross reactivities with other PAH components. Over 100 dust/soil samples from 13 North Carolina homes and 22 Arizona homes were analyzed by PAH and C-PAH assays, as well as by the conventional gas chromatography/mass spectrometry (GC/MS) method. Statistical analysis showed that dust/soil PAH data from ELISA and GC/MS methods are significantly different. In general PAH ELISA responses were higher than PAH GC/MS responses. The regression analysis showed that the linear relationship between ELISA and GC/MS measurements is not strong in the combined data. The relationship became stronger for the data from the same type of dust/soil samples. The screening performance of ELISA was evaluated based on the frequency distribution of ELISA and GC/MS data. The results indicated that the ELISA PAH and C-PAH assays cannot be used as a quantitative analytical tool for determining PAH in real-world dust/soil samples. However, the ELISA is an effective screening tool for ranking PAH concentrations in similar types of real world dust/soil samples.