BC Blount1 and Scott J Campbell2
1Centers for Disease Control and Prevention, Atlanta; 2SpectralWorks Ltd, The Heath Business & Technical Park, Runcorn, UK
First Published: PittCon 2004
Exposure to volatile organic compounds (VOCs) from a variety of natural and anthropogenic sources is ubiquitous in an area of public health concern. Estimating exposure to environmental toxicants is complicated by variables in human absorption, distribution, metabolism, and excretion. Therefore measuring internal dose is often the best method for assessing human exposure to environmental toxicants, including VOCs. To support studies exploring the relationship between VOCs in human blood and exposure sources we applied a software approach to tentatively identify patterns of exposure from routine exposures such as cigarette smoke and gasoline.
Materials and Methods
Blood samples from various adult volunteers were analyzed by solid phase micro extraction (SPME) coupled with GCMS. A gravimetrically confirmed amount of blood (approximately 3 g) was spiked with a cocktail of stable isotope labelled analogues of the 33 VOC analytes and capped in a pre-cleaned headspace vial. The vial was subsequently heated (30°C) and agitated (300 rpm) while the headspace was sampled (6 min) using a 75-µm Carboxen® /polydimethylsiloxane (PDMS) SPME fiber assembly and a CombiPAL autosampler (CTC Analytics AG). Following sample headspace extraction, volatiles on the SPME fiber were desorbed into a heated GC inlet (200°C, Agilent 5973 GC-MSD). Volatile analytes were trapped at the head of the GC column (VRX (30 m x 0.18 mm i.d. x 1.0 µm film, J & W Scientific) using a liquid nitrogen cooled cryo-trap (model 961, Scientific Instrument Services). After completion of splitless sampling, the cryotrap was ballistically heated and a thermal gradient used to chromatographically resolve volatile sample compounds. Data was collected using selected ion monitoring at unit mass resolution. Data processing was performed using the qualitative deconvolution and quantitative software application AnalyzerPro (SpectralWorks Ltd).
The adult volunteers were divided into 4 groups based on their likely exposure to VOCs:
Recent gas exposure
Recent gas exposure and who smoke
Neither smoker nor recent gas exposure
Levels of exposure were determined based on the blood level of each compound. The probability of each group having a specific compound present was calculated as the percentage of each group in which the compound was detected.
Figure 1. Typical section of TIC with found compounds labelled
Results and Discussion
Reliable and reproducible chromatograms were obtained from the analysis of these blood samples. Once the data was collected, it was analyzed using AnalyzerPro.
An average of 90 compounds were found in each of the unknown data files, 33 of these compounds were known and were being used as quantitation standards. Figure 1 shows a section of a typical total ion chromatogram (TIC). It shows that 19 compounds were detected in a 2 minute period and an example of the summary report for each of those detected compounds is given in Table 1.
The results from each of the files were quantified against the 33 quantitation standards and a matrix was created to calculate the average exposure levels for each compound in each of the groups, smokers, gas, smoking / gas and clean samples. Table 2 shows the average levels of exposure for some of the compounds for each of the groups. The probability of a compound being present was also calculated for each groups and results given in Table 3.
Table 1. Example found compounds
Table 2. Average levels of exposure (ng/ml)
|Gas / Smoke||Gas||Smoke||Clean|
Table 3. Compound probability per group
|Gas / Smoke||Gas||Smoke||Clean|
We were able to quantify 33 different VOCs in 86 whole blood samples by SPME-GCMS. These included control samples where there had been no known exposure to VOCs above the ambient conditions and test samples where there had been an elevated exposure from an identified source. Through characterizing single exposure events, multiple exposure events can be recognized and potentially previously unidentified exposures characterized.
Monitoring the levels of certain compounds within the blood of people allows the assessment of environmental challenge that a person has received. This poster has illustrated how software can help to calculate the different levels of exposure in the blood and exposure levels compared between the sources. Additionally, previously undetected compounds can be found using a unique qualitative deconvolution and quantitative application.
Further work is required to categorize the patterns between sample sources and probabilities so that an automated assessment of likely exposure can be established.