Quantitative lung particulate analysis links increased anthracotic pigment burden to burn pit exposure in post-9/11 veterans with distal lung disease Authors: Jeremy T. Hua, Lauren M. Zell-Baran, Carlyne D. Cool, Silpa D. Krefft, Leonard H.T. Go, Robert A. Cohen, Susan M. Majka, Cecile S. Rose Published: 2025 ABSTRACT Military deployment to Southwest Asia and Afghanistan is linked to complex hazardous airborne exposures, but ascertaining the contributions of specific environmental toxicants in causing or contributing to deployment-related distal lung diseases (DDLD) remains challenging. We applied a quantitative microscopy for particulate matter (QM-PM) technique to measure the in situ burden of pigmented and birefringent lung particulates in 24 veterans with DDLD, 10 smokers with respiratory bronchiolitis (RB), and 10 healthy controls. Veterans were more likely to be male (88%) and were significantly younger than RB and healthy controls (mean age 41 vs. 47 vs. 53 years, respectively). Adjusting for age, the anthracotic pigment fraction in lung tissue from DDLD was similar to RB (1.69% vs. 1.37%, P = 0.72), and was significantly greater than healthy controls (0.52%, P = 0.02). Pigment fraction in veterans was significantly associated with higher reported burn pit smoke exposure (P = 0.02), but not sandstorms or diesel exhaust. Birefringent dust density, indicating retained silica/silicates, was significantly increased in RB (28.89/mm2, P < 0.01) compared to DDLD and healthy control groups (8.93 and 9.44/mm2). Findings suggest that lung deposition of anthracotic dust from burn pit smoke exposure in previously deployed post-9/11 military veterans may be important in the pathogenesis of DDLD.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-025-27793-1. Introduction Military deployment to Southwest Asia and surrounding theaters of war during the First and Second Gulf Wars (1990–1991 and post-9/11/2001) has been linked to a growing spectrum of both proximal and distal respiratory diseases, termed deployment-related respiratory diseases (DRRDs)1–6. Proximal DRRDs include chronic sinusitis and rhinitis, vocal cord dysfunction, expiratory central airways collapse, and asthma. Distal DRRDs that affect the terminal bronchioles and alveoli, also known as deployment-related distal lung diseases (DDLD), include bronchiolitis, emphysema, and granulomatous inflammation. Deployed military personnel report diverse inhalational exposures including smoke from oil well fires and burn pits, sandstorms, desert dusts, combat blasts, jet fuel vapors, and diesel exhaust. Environmental assessment tools such as deployment exposure matrices, quantitative air sampling, geospatial modeling, and historical video footage have helped characterize exposures encountered by military personnel in these regions1,7–16. However, it remains difficult to determine the relative contribution of various exposure sources in the development of DRRDs. Advanced microscopy techniques characterizing in situ lung particulates have enhanced understanding of hazardous exposures in several occupational cohorts. Elemental analysis of lung tissue from historical and contemporary coal miners using scanning electron microscopy/energy dispersive spectroscopy (SEM/EDS) helped unmask the causal role of respirable crystalline silica in the modern-day resurgence of progressive massive fibrosis among Appalachian coal miners in the United States17,18. A study of military personnel with DDLD by Lowers et al. (2024) using SEM/EDS showed a greater proportion of silica (SiO2) and aluminosilicates in retained lung dust in veterans compared to healthy controls19. Other studies using SEM/EDS have shown higher numbers of total particles in deployed compared to non-deployed service members, with geologic clays (illite, kaolinite, feldspars, quartz/silica and titanium-rich silicate mixtures) accounting for most of these differences20. While these analytical approaches help inform lung health effects from exposure, SEM/EDS has several shortcomings that constrain interpretation. Lack of standardized methods and variability in analytical techniques limit comparability among studies. Selection of particle-rich areas of lung tissue for in situ SEM/EDS analysis may lead to sampling bias. Analyses using lung tissue digestion risk contamination during sample preparation and require both larger specimen sizes and sample volume correction. For some methods, the use of carbon-coated planchets precludes characterization of carbonaceous lung particulates. All the current methods are time-intensive and require sophisticated equipment and expertise, unavailable to most practicing pathologists and clinicians. To address these limitations, we leveraged the microscopic properties of common fibrogenic dusts to characterize retained lung particulates. Crystalline silica and silicates have inherent birefringent properties that are apparent under cross-polarized light microscopy (PLM) in lung tissue from exposed workers21–23. Carbonaceous dust from environmental combustion products and smoking can lead to deposition of pigmented (anthracotic) particulate matter (PM) in lung tissue which is readily visible under conventional brightfield microscopy24–27. We developed and validated a more accessible quantitative microscopy for particulate matter (QM-PM) technique to characterize these in situ lung dusts28. QM-PM utilizes a brightfield microscope, polarizing filter, and semi-automated digital imaging processes to assess the burden of pigmented and birefringent PM across an entire tissue specimen. Our initial validation study demonstrated increased silica and silicates in lungs from contemporary coal miners compared to historical counterparts, and showed that QM-PM results were concordant with the labor-intensive scoring method used by a panel of pulmonary pathologists18,28. Our aim in this study using QM-PM was to explore the linkage between dust mineralogy characteristics and histologically well-characterized distal lung disease, by measuring the quantitative burden and features of in situ lung PM from deployed veterans. We hypothesized that the quantitative burden of in situ pigmented and birefringent particulates in military personnel with DDLD would be comparable to smokers with respiratory bronchiolitis (RB), but greater than healthy controls. Methods Study design and population We obtained surgical lung tissue from deployed veterans with biopsy-proven distal lung disease defined by histologic findings of smooth muscle hypertrophy, peribronchiolar metaplasia, lymphocytic and granulomatous inflammation, constrictive/obliterative bronchiolitis, emphysema, and/or pleuritis6. Details about both sample collection and study participants have been described previously6,19. Briefly, all study participants were military personnel who were evaluated in the Center for Deployment-related Lung Disease at National Jewish Health (NJH) in Denver, Colorado. We analyzed tissue samples from participants who underwent surgical lung biopsies between 2001 and 2017 as part of their clinical workup for unexplained exertional dyspnea. Study participants provided informed consent, approved by the Institutional Review Board (IRB) Biomedical Research Alliance of New York (BRANY; protocol: HS-2689). We obtained lung tissue specimens from two comparison groups that have been described previously6. The first group consisted of lung tissue from participants with a histopathologic diagnosis of respiratory bronchiolitis (RB) obtained via the University of Colorado School of Medicine Pathology Department (IRB protocol HS-3022). RB is an inflammatory lung disorder caused by tobacco smoking and characterized by the presence of pigmented macrophages and chronic peribronchiolar inflammation in distal airways29,30. For the second comparison group, we used control lung samples donated by accident victims without known lung disease (IRB protocols HS-3209 and HS-3429). We selected control samples from participants who were as similar as possible in age and smoking status to deployed participants6,19. Research in all participants in this study was performed with informed consent from participants and/or their legal guardians, and in accordance with relevant guidelines/regulations including the Declaration of Helsinki. Deployment respiratory hazard exposure indices We calculated individual weighted respiratory hazard scores for post-9/11 deployment exposures in participating veterans10,19. As previously described, we applied a modified inhalational exposure matrix for each military service member by combining their deployment duration and the reported frequency/intensity for each exposure including burn pit combustion products, sandstorms, and diesel exhaust. Quantitative Microscopy – Particulate matter (QM-PM) The technique of QM-PM measures the total area of tissue on a glass slide, the area of pigmented tissue with retained anthracotic dust, the density of birefringent silica and silicate PM, and characteristics of individual birefringent particles. We performed QM-PM using previously published methods28, with the addition of several updated modifications described herein. Total Tissue Area: We obtained brightfield microscopy images of H&E-stained lung tissue slides, scanned at 40x magnification, using Aperio XT (ImageScope v12.4.3, Leica Biosystems, Buffalo Grove, Illinois). We then used the built-in Aperio Positive Pixel Count Algorithm feature to identify pixels containing lung tissue in an automated fashion, by selecting pixels that met threshold criteria based on predetermined hue, saturation, and intensity along the red, green, and blue (RGB) color spectrums. The desired pixels were any that did not include the white background. We implemented these macro algorithms across all tissue samples using the automated Spectrum Batch Analysis mode in Aperio eSlide Manager, the virtual platform for Aperio ImageScope. Compared to our previously described QM-PM methods that used ImageJ (https://imagej.net/ij/)31 software for total tissue area measurements, these optimized processes in Aperio use full resolution images rather than requiring reduced image resolution because of computation limitations in ImageJ. Pigmented (Anthracotic) Particle Characterization: We performed automated identification of pigmented particulates in a similar manner as with total tissue area. Using Aperio Positive Pixel Count Algorithm and eSlide Manager batch mode, we optimized macros for measuring the total pigmented area of black and brown pixels in tissue. We calculated the pigment fraction for each sample by dividing the pigmented area by the total tissue area. Birefringent Particle Characterization: We applied the same methods as previously described to identify and characterize in situ birefringent particles, primarily crystalline silica and silicates22,23,32. Briefly, H&E-stained tissue samples were loaded on a conventional light microscope, Keyence BZ-X810 (Keyence Corporation, Osaka, Japan), with a Jigtech N11599 polarizer lens and filter cube (Daitron Incorporated, Wilsonville, Oregon). We obtained high-resolution, digital, polarized light microscopy images at 20x magnification across the entire tissue sample, totaling approximately 1000–4000 images depending on specimen size. All images were obtained using the same microscope, illuminator, and polarizing filter. We used the built-in Keyence Macro Hybrid Cell Count feature (BZ-X800 Analyzer software v1.1.1.8) to identify birefringent particles which became visually apparent under polarized light. Pixel grayscale brightness ranges from 0 for black to 255 for white, where crystalline silica particles are typically weakly birefringent (duller) and silicates are more strongly birefringent (brighter). We applied a brightness threshold under the blue RGB spectrum to capture birefringent particles while excluding background lung tissue and birefringent collagen fibers28. We excluded particles that were greater than 50 µm2, as these likely represented artifacts rather than in situ dust. We analyzed individual particle features such as long- and short-axis dimensions (µm), area (µm2), and mean grayscale brightness. We then calculated the birefringent dust density (birefringent particle count/mm2 total tissue area) for each sample. Previous studies demonstrated the utility of QM-PM in characterizing lung tissue dust from coal miners with progressive massive fibrosis (PMF), the most severe form of dust-related pneumoconiosis18,28. QM-PM showed substantially greater birefringent dust density in miners compared to healthy controls, and greater pigment fraction in those with coal-type PMF. Having further optimized the QM-PM methodology since this initial study, we applied the new protocol to lung tissue samples from all 85 coal miners (in accordance with relevant guidelines/regulations approved under IRB protocol HS-3039-528) and compared their updated QM-PM findings in a sensitivity analysis to those from deployed veterans, positive controls with RB, and healthy controls. Additional details about this secondary analysis including demographics and exposure details for the 85 coal miners are included in the Online Supplemental Information. Statistical analysis Demographics are presented as mean and standard deviations or n (%) and were compared using one-way analysis of variance (ANOVA) for continuous data and Fisher’s Exact Test for categorical information. To compare pigment fraction, birefringent dust density, and individual particle-level characteristics among deployed veterans, RB, and healthy control groups, we used a mixed model (lme4 package) with a random intercept for each subject to account for multiple lobe samples and accounted for multiple comparisons using Tukey adjustment (emmeans package). Birefringent dust density was log-transformed for statistical analyses. We adjusted all models for age at biopsy due to potentially confounding differences among diagnosis groups. We performed sensitivity analyses after restricting veterans to only non-smokers, since smoking data was limited for participants in the comparison groups and could not be analyzed in the full models as a covariate. In another sensitivity analysis, we included coal miners with PMF in the mixed models to serve as a highly exposed group. In deployed veterans only, we examined the relationship between respiratory hazard indices with both pigment fraction and birefringent dust density using a linear mixed model (lme4 package), with a random intercept for each subject to account for multiple lobe samples. We first examined exposures using a cumulative exposure index, then separately based on each individual reported exposure. Additionally, we examined the relationship between these particle characteristics with deployment duration and first deployment period. We grouped veterans into those with first deployment to Iraq or Afghanistan either before or after 2009, the year when guidance for burn pit waste segregation was recommended15. Finally, we examined differences in dust characteristics between upper and lower lobe samples from deployed veterans using a linear mixed model with a random intercept for each subject. All data management and analyses were performed in R (version 4.4.2; R Core Team, 2024). Results Demographic and exposure characteristics We obtained surgical lung tissue specimens from 24 post-9/11 deployed military veterans with histologic findings of DDLD, including 24 upper lobe, three middle lobe, and 22 lower lobe samples. One veteran had undergone a lobectomy and the remaining 23 underwent video-assisted thoracoscopic surgical (VATS) lung biopsies. Ten VATS lung biopsies were available from participants with RB, and ten donor lung samples were available from healthy controls. Demographics for all participants are summarized in Table 1. Deployed veterans were significantly more likely to be male and were younger in age compared to RB and healthy controls. Specifics of cigarette smoking status and pack-years were missing for five participants with RB and four controls. Military veterans reported a median of three deployments totaling 29.1 ± 20.5 months on average, with nine (37%) having deployed to Iraq only, six (25%) to Afghanistan only, and 9 (37%) to both countries. Veterans reported frequent high-intensity exposure (more than 1–2 times weekly) to burn pits (79%) and diesel exhaust (95%), and less commonly to sandstorms (9%). Table 1Demographics of deployed military veterans with distal lung disease, smokers with respiratory bronchiolitis (RB), and healthy controls.DemographicsDeployed VeteransN = 24RBN = 10ControlsN = 10p-value1Male21 (88)4 (40)5 (50) 0.005 Age at biopsy (years) [range]41.0 ± 7.8 [25–59]47.1 ± 14.8 [31–75]52.7 ± 18.0 [17–71] 0.04 Ever cigarette smoker9/24 (38)4/52 (80)3/6 (50)--Smoking pack-year5.0 ± 4.550 ± 012.2 ± 19.8--Non-Hispanic White22 (92)**7/9 (78)0.3 Deployment Exposure Characteristics Number of deployments, median (range)3 (1–8) -- -- Total months deployed29.1 ± 20.5 -- -- Deployment locationIraq9 (37) -- -- Afghanistan6 (25) -- -- Both9 (37) -- -- Reported high-intensity3 exposure to:Burn pit smoke19/24 (79) -- -- Sandstorms2/23 (9) -- -- Diesel exhaust21/22 (95) -- -- Exposure index score: median (range)Burn pit smoke631 (0–2760) -- -- Sandstorms46 (0–551) -- -- Diesel exhaust587 (60–2434) -- -- Cumulative1582 (255–5349) -- -- Combat MOS48/13 (62) -- -- Note: Values are Mean (± SD) or n (%), unless otherwise noted. Any differences in denominators reflect missing data.1Where appropriate, continuous variables were compared via one-way ANOVA, and categorical variables were compared via Fisher’s Exact Test.2Of note, the one participant who denied current or former cigarette smoking reported recreational cannabis smoking.3High-intensity was defined as more than 1–2 times each week.4MOS=military occupational specialty code.**Race and ethnicity information was not available for participants with RB.Note: Missing data precluded statistical comparisons for smoking status and pack-years. Table 1Demographics of deployed military veterans with distal lung disease, smokers with respiratory bronchiolitis (RB), and healthy controls.DemographicsDeployed VeteransN = 24RBN = 10ControlsN = 10p-value1Male21 (88)4 (40)5 (50) 0.005 Age at biopsy (years) [range]41.0 ± 7.8 [25–59]47.1 ± 14.8 [31–75]52.7 ± 18.0 [17–71] 0.04 Ever cigarette smoker9/24 (38)4/52 (80)3/6 (50)--Smoking pack-year5.0 ± 4.550 ± 012.2 ± 19.8--Non-Hispanic White22 (92)**7/9 (78)0.3 Deployment Exposure Characteristics Number of deployments, median (range)3 (1–8) -- -- Total months deployed29.1 ± 20.5 -- -- Deployment locationIraq9 (37) -- -- Afghanistan6 (25) -- -- Both9 (37) -- -- Reported high-intensity3 exposure to:Burn pit smoke19/24 (79) -- -- Sandstorms2/23 (9) -- -- Diesel exhaust21/22 (95) -- -- Exposure index score: median (range)Burn pit smoke631 (0–2760) -- -- Sandstorms46 (0–551) -- -- Diesel exhaust587 (60–2434) -- -- Cumulative1582 (255–5349) -- -- Combat MOS48/13 (62) -- -- Note: Values are Mean (± SD) or n (%), unless otherwise noted. Any differences in denominators reflect missing data.1Where appropriate, continuous variables were compared via one-way ANOVA, and categorical variables were compared via Fisher’s Exact Test.2Of note, the one participant who denied current or former cigarette smoking reported recreational cannabis smoking.3High-intensity was defined as more than 1–2 times each week.4MOS=military occupational specialty code.**Race and ethnicity information was not available for participants with RB.Note: Missing data precluded statistical comparisons for smoking status and pack-years. Demographics of deployed military veterans with distal lung disease, smokers with respiratory bronchiolitis (RB), and healthy controls. Note: Values are Mean (± SD) or n (%), unless otherwise noted. Any differences in denominators reflect missing data. 1Where appropriate, continuous variables were compared via one-way ANOVA, and categorical variables were compared via Fisher’s Exact Test. 2Of note, the one participant who denied current or former cigarette smoking reported recreational cannabis smoking. 3High-intensity was defined as more than 1–2 times each week. 4MOS=military occupational specialty code. **Race and ethnicity information was not available for participants with RB. Note: Missing data precluded statistical comparisons for smoking status and pack-years. Particulate characteristics in deployed veterans compared to RB and healthy controls Lung tissue samples in deployed veterans had similar mean pigment fraction compared to participants with RB (1.69% vs. 1.37%, respectively, P = 0.72), and significantly greater pigment fraction than healthy controls (0.52%, P = 0.02), (Figure 1a). These and all subsequent group comparisons were adjusted for age. In a sensitivity analysis, the pigment fraction in only the non-smoking veterans (1.58%) remained significantly greater than healthy controls (P = 0.046). A representative example of lung tissue with increased pigment fraction from a 59-year-old military veteran (never tobacco smoker) reporting high exposure to multiple hazards during 19 months of cumulative deployment to Afghanistan is shown in Figure 2.Fig. 1Lung particulate burden measured using QM-PM. Quantitative: (a) Pigment fraction reflecting anthracotic dust, and (b) birefringent dust density indicating silica/silicates comparing veterans with deployment-related distal lung disease (DDLD), smokers with respiratory bronchiolitis (RB), and healthy controls. *P<0.05 and **P<0.01 for pairwise comparisons. Lung particulate burden measured using QM-PM. Quantitative: (a) Pigment fraction reflecting anthracotic dust, and (b) birefringent dust density indicating silica/silicates comparing veterans with deployment-related distal lung disease (DDLD), smokers with respiratory bronchiolitis (RB), and healthy controls. *P<0.05 and **P<0.01 for pairwise comparisons. Fig. 2Hematoxylin and eosin (H&E)-stained lung tissue from a 59-year-old male veteran (never tobacco smoker) who reported high-intensity exposure to burn pit smoke, sandstorms, and diesel exhaust during 19 cumulative months of deployment to Afghanistan. Histopathology showed small airways smooth muscle hypertrophy, emphysema, granulomatous pneumonitis, and chronic pleuritis. Panel (a) shows the presence of retained anthracotic pigment interspersed throughout the lung interstitium and (b) shows automated identification of pigmented dust particulates (blue overlay) using QM-PM.The mean birefringent dust density in deployed veterans was similar to healthy controls (8.93 vs. 9.44/mm2, P = 0.98), and both were significantly less than in participants with RB (28.89/mm2, P < 0.01 for both pairwise comparisons) (Figure 1b). There were no significant changes when restricting the model to non-smoking veterans. Figure 3 shows representative lung tissue under brightfield microscopy and PLM from a participant with RB, highlighting crystalline silica/silicate particles. Mean birefringent particle size was smallest in those with RB (Figure 4a). Deployed veterans had significantly larger in situ birefringent particles (mean area: 5.07 µm2) compared to those with RB (3.66 µm2, P = 0.0006) and were similar to controls (4.19 µm2, P = 0.06). We observed similar trends in deployed veterans, those with RB, and healthy controls for both particle short-axis dimension (1.42 vs. 1.12 vs. 1.24 μm, respectively) and long-axis dimension (2.80 vs. 2.35 vs. 2.55 μm, respectively). Mean particle brightness was significantly different between all three groups (P < 0.001), with deployed veterans having the highest mean particle brightness (186), followed by controls (160) and those with RB (98) (Figure 4b), suggesting higher silicates and lower silica content in veterans compared to those with RB.Fig. 3Hematoxylin and eosin (H&E)-stained lung tissue from a 31-year-old male tobacco smoker. Panel (a) shows histopathologic abnormalities of respiratory bronchiolitis highlighted by small airway smooth muscle hypertrophy surrounded by pigmented macrophages using brightfield microscopy; Panel (b) shows the presence of birefringent silica and silicate particles under polarized light microscopy; and Panel (c) shows automated identification of birefringent particles (turquoise overlay) under polarized light microscopyusing QM-PM. Fig. 4Particle-level characteristics for birefringent particles using QM-PM. Cumulative frequency of particles based on (a) particle area (µm2) and (b) particle grayscale brightness under polarized light (range, 0 [black]–255 [white]; higher value indicates stronger birefringence), comparing veterans with distal lung disease (blue), smokers with respiratory bronchiolitis (green), and healthy controls (red). Hematoxylin and eosin (H&E)-stained lung tissue from a 59-year-old male veteran (never tobacco smoker) who reported high-intensity exposure to burn pit smoke, sandstorms, and diesel exhaust during 19 cumulative months of deployment to Afghanistan. Histopathology showed small airways smooth muscle hypertrophy, emphysema, granulomatous pneumonitis, and chronic pleuritis. Panel (a) shows the presence of retained anthracotic pigment interspersed throughout the lung interstitium and (b) shows automated identification of pigmented dust particulates (blue overlay) using QM-PM. Hematoxylin and eosin (H&E)-stained lung tissue from a 31-year-old male tobacco smoker. Panel (a) shows histopathologic abnormalities of respiratory bronchiolitis highlighted by small airway smooth muscle hypertrophy surrounded by pigmented macrophages using brightfield microscopy; Panel (b) shows the presence of birefringent silica and silicate particles under polarized light microscopy; and Panel (c) shows automated identification of birefringent particles (turquoise overlay) under polarized light microscopyusing QM-PM. Particle-level characteristics for birefringent particles using QM-PM. Cumulative frequency of particles based on (a) particle area (µm2) and (b) particle grayscale brightness under polarized light (range, 0 [black]–255 [white]; higher value indicates stronger birefringence), comparing veterans with distal lung disease (blue), smokers with respiratory bronchiolitis (green), and healthy controls (red). Association of particulate characteristics with deployment exposures We compared pigmented and birefringent particulate characteristics with weighted respiratory hazards scores in deployed veterans. Pigment fraction was associated with the cumulative exposure index score (0.48 per log-transformed index, P = 0.14) that includes burn pit, sandstorm, and diesel exhaust exposure, but this finding was not statistically significant. Pigment fraction significantly increased with exposure index scores for burn pits (0.45 per log-transformed index, P = 0.02) but was not statistically significantly associated with either sandstorms or diesel exhaust. Pigment fraction increased with each additional month of deployment but again was not statistically significant (0.016 per month, P = 0.19). Parameter estimates and 95% confidence intervals are presented in Supplemental Table SI for each model. There were no significant relationships between birefringent dust density and either weighted respiratory hazards scores (modeled as a cumulative score or as individual hazard components) or deployment duration (P > 0.05) (Supplemental Table SII). Nineteen (79%) military veterans had deployed prior to 2009, with all but one reporting high intensity burn pit exposure, compared to five veterans who first deployed during or after 2009 when U.S. military guidance for burn pit waste segregation was recommended15. We found no significant differences in either pigment fraction (P = 0.31) or birefringent dust density (P = 0.42) based on timing of deployment before or after 2009. Particle characteristics by lung lobe In deployed participants with both upper and lower lobe specimens available (20/24, 83%), there were no significant differences in pigment fraction, birefringent dust density, birefringent particle size, or relative birefringence between lobes (Table 2). Table 2Quantitative pigmented and birefringent dust characteristics in deployed veterans’ upper and lower lobe lung samples (N = 20)*.Upper lobeLower lobep-valuePigment fraction (%)1.42 ± 1.311.34 ± 0.910.76Birefringent dust density (particles/mm2)14 ± 1112 ± 70.61LN Birefringent dust density (particles/mm2)2.37 ± 0.742.26 ± 0.73Mean particle size (µm2)4.82 ± 0.835.25 ± 1.130.15Mean particle birefringence (0–255)189 ± 8188 ± 120.79Note: Two participants only had a lower lobe sample available. Two participants had only two upper lobe samples available. Table 2Quantitative pigmented and birefringent dust characteristics in deployed veterans’ upper and lower lobe lung samples (N = 20)*.Upper lobeLower lobep-valuePigment fraction (%)1.42 ± 1.311.34 ± 0.910.76Birefringent dust density (particles/mm2)14 ± 1112 ± 70.61LN Birefringent dust density (particles/mm2)2.37 ± 0.742.26 ± 0.73Mean particle size (µm2)4.82 ± 0.835.25 ± 1.130.15Mean particle birefringence (0–255)189 ± 8188 ± 120.79Note: Two participants only had a lower lobe sample available. Two participants had only two upper lobe samples available. Quantitative pigmented and birefringent dust characteristics in deployed veterans’ upper and lower lobe lung samples (N = 20)*. Note: Two participants only had a lower lobe sample available. Two participants had only two upper lobe samples available. Particulate characteristics in deployed veterans compared to coal miners with PMF In a sensitivity analysis of 85 coal miners with PMF, there was markedly greater mean pigment fraction (12.72%) and mean birefringent dust density (279 particles/mm2) compared to deployed veterans, those with RB, and healthy controls (P < 0.0008 for all coal miner pairwise comparisons), likely reflecting substantial differences in cumulative particulate exposures among these groups (Supplemental Figure S1). Discussion In this tissue-based investigation of post-9/11 deployed military veterans with biopsy-proven distal lung disease, we found a significantly greater burden of pigmented lung particulates in exposed veterans compared to healthy controls despite the fact that veterans were significantly younger and reported lower smoking rates and pack-years. Notably, the pigment burden in veterans was comparable to that in lung tissue from patients with respiratory bronchiolitis, a lung disease linked to heavy tobacco smoking. Further, the quantitative pigment fraction was significantly positively correlated with high exposure to burn pit fumes/vapors, suggesting that burn pit combustion products may be an important source of retained pigmented lung dust in previously deployed veterans with symptomatic distal lung disease. Major contributors to these airborne PM originate from burning waste such as electronics, vehicles, plastics, foams, and human and medical waste, often with the use of fuel accelerants like JP-811. Specific burn pit smoke hazards identified in a 2020 National Academy of Sciences report included various silicates and heavy metals, carbonaceous particulates, polycyclic aromatic hydrocarbons, and other volatile organic compounds1. Recent toxicologic studies analyzed simulation burn pit smoke emissions from burning cardboard, plywood, and plastics with diesel accelerant, under both flaming and smoldering combustion conditions33,34. Concentrations of PM emissions from smoldering burn pit combustion were substantially greater than from flaming combustion (162 vs. 8 g/kg), and burning plastics resulted in 20-fold greater PM emissions than cardboard or plywood wastes. In contrast, more polyaromatic hydrocarbons and nitric oxide gases are formed during flaming conditions. Burn pits appear to be the primary source of toxic acrolein, dioxins, and furans formed during combustion35,36. Records on waste disposal methods in military theaters of conflict were infrequent before 2005 in Afghanistan and 2007 in Iraq15. Segregated or only partially segregated waste (that is, medical or hazardous waste managed separately from more conventional waste) was unusual before 2009, when additional guidance on waste segregation for burn pits was issued15. By 2014, the proportion of military bases with burn pits had decreased to 40% as the use of incinerators became more common, and over half of all U.S. bases transported waste to landfills or municipal waste sites. We found no significant difference in the lung tissue pigment fraction between pre-2009 and post-2009 deployed military veterans in this study, though we lacked information on base- and time-specific waste disposal methods. Future in situ PM assessment using larger sample sizes and more granular burn pit locations and waste disposal characteristics may further inform the link between retained lung dust and burn pit exposure. One recent study of in situ elemental and particulate lung dust characteristics showed increased proportions of crystalline silica and silicates in previously deployed veterans compared to healthy controls, though volume-corrected silica/silicate particle counts were similar between groups19. Our QM-PM technique did not show differences in the density of birefringent dust in deployed veterans compared to controls without lung disease. This is unsurprising in view of the relatively uncommon reports of high exposure to sandstorms during deployment among those included in this study. Our finding of greater measured pigment fraction using QM-PM in primarily never smoking deployed veterans, comparable to smoking-related RB, may provide insight into the pathogenesis of DDLD. DDLD is a multi-compartment syndrome with predominant findings of small airways injury including peribronchiolar metaplasia and smooth muscle hypertrophy along with emphysema and hyperinflation6. Inhalation of tobacco smoke that contains a complex mixture of carbonaceous and silicate PM along with heavy metals such as cadmium and chromium can cause respiratory bronchiolitis37. Our study supports an important role of carbonaceous particulates and other products of combustion in contributing to DDLD in post-9/11 deployed military veterans. We found that the burden of carbonaceous and birefringent particulates was similar between lower and upper lobe lung tissue samples. Though coal mine dust and silica-induced lung diseases are described as typically upper lung zone predominant, our findings are consistent with recent studies showing a relatively homogeneous spatial distribution of these exposure-related lung abnormalities6,38–40. Our findings in deployed veterans underscore the promising role of QM-PM as a tool for assessing in situ lung particulate matter in exposed workers, as shown previously in coal miners with PMF and silica-exposed workers with “silicosarcoidosis”41. Unlike QM-PM, which measures birefringent and pigmented particulate burden across an entire tissue specimen, some of the more commonly used SEM/EDS methods that rely on visually selected fields for particle identification and characterization are unable to accurately extrapolate tissue particulate concentrations. In contrast to QM-PM, SEM-EDS can provide detailed elemental composition for a subset of retained particulates. However, access to SEM-EDS is hampered by the need for highly sophisticated and specialized equipment and environmental scientists with expertise in analytical methods and data analysis. Since QM-PM relies on widely available brightfield microscopy coupled with imaging software, it is substantially less time and labor intensive. This study has both strengths and limitations. First, we showed a spectrum of QM-PM findings among several relevant groups including highly exposed coal miners with PMF, smokers with respiratory bronchiolitis, and healthy controls. However, QM-PM is unable to differentiate particles by elemental composition such as anthracotic carbonaceous dusts from pigmented metal dusts or endogenous hemosiderin. QM-PM may not detect amorphous silica particles that lack birefringence42,43. Additionally, QM-PM is unable to measure individual pigmented particle features since carbonaceous combustion smoke products may exist on the nanometer scale below the limits of detection or agglomerate into larger clusters44. Second, we were able to analyze lung tissue PM findings in relation to well-characterized demographic variables, deployment histories, and respiratory hazard scores in veterans with deployment-related lung disease to identify specific exposures of importance, mainly self-reported high intensity burn pit smoke. Respiratory hazard indices rely on historical recall and may suffer from bias. However, it is unlikely that there would be differential recall bias for the multiple deployment respiratory hazards included in the exposure index. Finally, detailed environmental and occupational histories were not available for participants with RB and controls. Conclusion Using quantitative microscopy, military veterans with biopsy-proven distal lung disease had an increased burden of retained anthracotic lung dust associated with high intensity exposure to burn pit combustion products, findings similar to tobacco smokers with respiratory bronchiolitis. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1