Demonstration of a novel quantitative microscopy technique for automated characterization of in situ particulate matter in coal miners with progressive massive fibrosis

Abstract

Rationale: Increasing exposure to respirable crystalline silica (RCS) linked to changes in mining production processes has been implicated in the resurgence of severe lung disease in U.S. coal miners. Lung mineralogy can provide insight into particle pathogenesis. However, standard approaches to characterizing in situ particulate matter (PM) by pulmonary pathologists have poor inter-rater comparability, and consensus agreement is time consuming. Scanning electron microscopy/energy-dispersive spectroscopy (SEM/EDX) is technically complex and labor intensive. We developed a method for quantitative in situ PM characterization using conventional polarized light microscopy (PLM) and explored PM features in lung tissue of coal miners with progressive massive fibrosis (PMF). Methods: With institutional review board approval, PLM images were obtained from 30 miners with PMF, classified by pathologists consensus based on PM profusion; 10 from each profusion group (mild/moderate/severe) were selected. Automated PM counting and characterization (including dimensions and grayscale intensity of PM > 0.3 m diameter) was performed on image samples using PLM with modified cell-counting software (BZ-X800 light microscope, Keyence Corporation, Osaka, Japan) (Figure 1). Quantitative PM density loge(PM count)/mm3 tissue was calculated for each sample and compared to pathologist PM profusion groups. PMF lesion type using consensus pathologist classification (13 coal-type, 9 mixed-type, and 8 silicotic-type) was compared to automated PM birefringence level (% particles with mean grayscale intensity <65, range 0-255). RCS particles are expected to be weakly birefringent (lower intensity) relative to other common minerals (e.g., silicates) contained in coal mine dust. PM features were analyzed in R 4.0.3 using one-way ANOVA for between-group comparisons. Results: Measured PM log-density increased significantly with higher qualitative profusion group (mild=10.480.98/mm3, moderate=11.460.81/mm3, severe=12.520.86/mm3, p<0.0001). Prevalence of weakly birefringent particles was significantly higher among silicotic-type PMF samples (31.57.9%) compared to either coal-type (21.510.1%, p=0.022) or mixed-type lesions (21.510.5%, p=0.025). Conclusion: This pilot study demonstrates the feasibility of a novel quantitative microscopy technique for counting and characterizing in situ lung PM in coal miners with PMF. Quantitative PM burden was comparable to pulmonary pathologists consensus profusion classification, but this method was substantially less time consuming and labor intensive and provided additional information about relevant PM features. The higher prevalence of weakly birefringent particles seen in silicotic-type PMF lesions may help inform mineralogic pathogenesis of RCS. Future efforts will expand the number of PMF cases analyzed, further validate our mineralogic findings using data from SEM/EDX and lung tissue digestate methods, and compare findings in historical versus contemporary coal miners with PMF.