| Abstrakt: |
Scanning electron microscopy (SEM) coupled with energy dispersive X-ray analysis (EDX) has been used in several laboratories to determine lung particulate burden. However, there has been no attempt to investigate interlaboratory variations in these analyses which may arise because of differences in tissue sample sizes, tissue preparation techniques and analytical protocols used. Consequently, we have conducted a small interlaboratory study for this purpose. Three different SEM-EDX quantification methods were used including a manual in situ procedure using 5 μm thick tissue sections (SUNY-in situ), manual analysis of particles extracted by tissue digestion (SUNY-digestion) and automated analysis of particles from low-temperature ashed tissue (NIOSH). Lung samples from 10 subjects were analysed using both the SUNY-in situ and NIOSH analysis procedures, while the SUNY-digestion procedure was used for five of these samples. The same major particle-types were seen with all three analytical procedures; i.e. various aluminium silicates, silica and metal-containing species. Spearman rank order correlations observed between the three methods for total exogenous particulate levels were: rs = 0.97 for SUNY-in situ/SUNY digestion, rs = 0.90 for SUNY-digestion/NIOSH and rs = 0.58 for SUNY-in situ/NIOSH. Weaker relationships between the SUNY-in situ and NIOSH methods were seen for percentages of aluminium silicates (rs = 0.53), silica (rs = 0.56) and metal percentages (rs = 0.49). Relationships between the SUNY-digestion method and the other two procedures for aluminium silicate, silica and metal percentages were not well defined, most likely due to the small number of samples available for comparison. In terms of absolute particle concentrations, agreements between the methods were not consistent. When the SUNY-in situ data were converted from particles per ml tissue to particles per g dry lung using the average density of the freeze-dried NIOSH lung preparations (0.1 g ml−1), the ratio of total particle concentrations between the SUNY-in situ and NIOSH analyses was close to unity, mean = 1.6 [standard deviation (SD)= 1.3]. However, total particle levels determined by the SUNY-digestion procedure were much higher than found with either of the other two methods; e.g. the mean ratio of total particle concentrations between the SUNY-digestion and SUNY-in situ procedures was 20.9 (SD = 6.6). These discrepancies are, in part, possibly due to differences in sample preparation and variations in minimum size threshold between the two SUNY and the NIOSH techniques. [ABSTRACT FROM PUBLISHER] |
