Pulmonary Vascular Disease in Veterans with Post-Deployment Respiratory Syndrome

Background: Increased frequency of exertional dyspnea has been documented in U.S. military personnel after deployment to Southwest Asia and Afghanistan. We studied whether continued exertional dyspnea in this patient population is associated with pulmonary vascular disease (PVD). Methods: We recruited five Iraq and Afghanistan Veterans with post-deployment respiratory syndrome (PDRS) and continued exertional dyspnea to undergo a detailed clinical evaluation including symptom questionnaire, pulmonary function testing (PFT), surface echocardiography, and right heart catheterization (RHC) with exercise. We then performed detailed histomorphometry of blood vasculature in 52 Veterans with PDRS, 13 patients with advanced idiopathic pulmonary arterial hypertension (PAH) and 15 non-diseased (ND) control subjects. Results: All five Veterans involved in clinical follow-up study had a continued dyspnea at exertion. On transthoracic echocardiography, we identified borderline or overt RV enlargement in three out of five Veterans. Right ventricle outflow tract (RVOT) acceleration time, a well-established surrogate measure of pulmonary pressure, was mildly reduced in three out of five Veterans. Of the five Veterans with PDRS who underwent RHC at exercise, we found that three had evidence of post-capillary PH at rest and one had PH at exercise. Morphometric evaluation of lung biopsy samples showed mild/moderate increase of fractional thicknesses of intima and media, and significant fibrosis of adventitia in pulmonary arteries in Veterans with PDRS compared to ND controls and PAH patients. Veterans with PDRS did not display plexiform or dilation/angiomatoid lesions, specific for PAH. Pulmonary veins showed similar levels of intima and adventitia fractional thickening in Veterans with PDRS and PAH patients compared to ND controls. In Veterans, IA veins were characterized by marked fibrous intima and adventitia thickening, usually with increased thickening and formation of multiple layers of elastic laminae, but without features of luminal occlusion, muscular hyperplasia or dilation/angiomatoid lesions seen in pulmonary veno-occlusive disease or chronic thromboembolic PH. Conclusions: Our studies suggest that vasculopathy and PVD may explain exertional dyspnea and exercise limitation in some Veterans with PDRS. Evaluation for PVD should be considered in Iraq and Afghanistan Veterans with unexplained dyspnea.


Introduction
Increased frequency of exertional dyspnea has been documented in U.S. military personnel after deployment to Southwest Asia and Afghanistan. [1][2][3] Cohort studies have shown a wide spectrum of identifiable lung diseases in deployed Veterans (such as asthma and laryngeal disorders), yet up to 1/3 have been classified as undiagnosed exertional dyspnea. 4 In 2021, we reported comprehensive analysis of open lung biopsy samples from 50 Veterans with exertional dyspnea unexplained by standard non-invasive testing and found complex pathological changes involving all distal lung compartments. 5 We described chronic intrapulmonary lymphocytic inflammation, small airways pathology consistent with constrictive bronchiolitis (ConB), hypertensive-type pulmonary vasculopathy, diffuse interstitial fibrosis of alveolar tissue, and fibrosis involving visceral pleura. We proposed post-deployment respiratory syndrome (PDRS) as an umbrella term to define this complex of exposures, symptoms, and lung pathology. 5 The dominance of exertional dyspnea and presence of hypertension-type vascular pathology, in the absence of substantial pulmonary function testing (PFT) abnormalities, in affected Veterans raise the possibility that pulmonary vascular disease (PVD) could account for exercise limitation in these patients. To determine whether Veterans with biopsy-proven PDRS and continued exertional dyspnea might have evidence of PVD, we recruited five Veterans (11-16 years after lung biopsy) to undergo a detailed clinical evaluation including symptom questionnaire, PFT, surface echocardiography, and right heart catheterization (RHC) with exercise. We also performed detailed histomorphometry of blood vasculature on lung biopsies from a larger group of subjects with PDRS to better define the pathological basis of PVD in this cohort.
for use under a CC0 license. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted May 16, 2023. Echocardiography. Echocardiograms were performed on a Philips iE33 machine (Netherlands) according to American Society of Echocardiography guidelines. 10,11 Data were analyzed by a board certified echocardiographer (Evan Brittain) using Tomtec software (Germany).
Exercise right heart catheterization and invasive cardiopulmonary exercise testing. Right heart catheterization was performed per clinical guideline as previously described. 12 Briefly, in the supine position and under sterile conditions utilizing Seldinger technique, a 7-8Fr sheath was inserted under ultrasound guidance into the right internal jugular vein. A 7Fr Swan-Ganz pulmonary artery catheter (Edwards Life Sciences, Irvine, CA) was inserted through the sheath, balloon inflated, and sequentially advanced under fluoroscopic and hemodynamic guidance into the right atrium, right ventricle, pulmonary artery, and pulmonary wedge positions. Pressure measurements were zero balanced at the level of the right atrium while supine, and all pressure measurements at rest were obtained at end-expiration in mmHg. Cardiac output (CO) was estimated by the thermodilution method. Three Veterans underwent supine exercise RHC and two Veterans underwent invasive cardiopulmonary exercise testing (CPET), which consisted of seated upright RHC and gas exchange measurement. Research participants were subjected to a progressive load exercise protocol with a cycle ergometer either in the supine or seated upright position (Lode B.V., Groningen, Netherlands). For participants who underwent supine exercise, repeat hemodynamic measurements were made at various loads of exercise until the participant had reached symptom-limited peak exercise. Cardiac output was assessed by thermodilution at rest and peak exercise. For participants who underwent invasive CPET, a metabolic cart (MGC Diagnostics, St. Paul, MN, USA) was utilized to monitor gas exchange. Adequate test effort on invasive CPET was determined by respiratory exchange ratio (RER) via metabolic cart, and the for use under a CC0 license. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

Histopathological analysis and morphometry
Four serial paraffin tissue sections (5 μm) were cut from the same tissue blocks used in previous studies. 5  This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted May 16, 2023. ; https://doi.org/10.1101/2023.05.15.23289956 doi: medRxiv preprint adventitia were measured for all cross-sectioned BVB arteries and at least 10 randomly chosen IA arteries on digital images using ×10 or ×20 objectives depending on size. To determine intima fractional thickness, area of intima was normalized to the length of internal elastic lamina. To determine media and adventitia fractional thicknesses, areas of media or adventitia were normalized to the length of external elastic lamina. Data were presented as V:SAIntima, V:SAMedia or V:SAAdventitia, respectively. Fractional thicknesses of intima and adventitia were measured in at least 10 randomly chosen IA veins on digital images using ×20 objective. V:SAIntima and V:SAAdventitia in these veins were measured as intima or adventitia areas with following normalization to elastic lamina length. All morphometric measurements were done using a computerized image analyzer system (Image-Pro Plus, Media Cybernetics) by investigators (Sergey Gutor and Vasiliy Polosukhin) who were blinded to the study groups.

Statistical analysis
Demographic, clinical, morphological, and physiological data are reported as median (range) or mean ± standard deviation (SD) according to their distributions. For categorical variables, proportions were used. Comparisons between symptomatic Veterans, PAH patients and ND controls were conducted using t-test or Mann-Whitney U-test for continuous variables.
Follow-up data were compared with initial values by paired t-test or paired Wilcoxon test according to distribution. Hierarchical cluster analysis was used to identify and differentiate subgroups based on pathological patterns. 16 A hierarchical cluster tree (dendrogram) was constructed based on Euclidean distance. Subgroups were formed in an agglomerative manner, starting with each participant as his/her own subgroup and pairing the two closest subgroups together at every step until only one group of all participants remained. We applied the Gap for use under a CC0 license.
This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted May 16, 2023. ; https://doi.org/10.1101/2023.05.15.23289956 doi: medRxiv preprint statistic and Total Within Sum of Square statistic method for determination of the optimal number of clusters. 17 The number of subgroups was also confirmed by a visual inspection of the dendrogram and a Multi-Dimensional Scaling plot. All analyses were performed using Rsoftware version 3.5.2 (www.r-project.org).

Clinical follow-up in Veterans with PDRS
We recruited 5 Veterans who had previously undergone lung biopsy for unexplained exertional dyspnea (11-16 years prior to follow-up) and had findings consistent with PDRS after an extensive clinical evaluation and lung biopsy examination. Veteran participants were all males, ages 41-62, and non-smokers. All had a continued dyspnea at exertion and high symptom burden on the St. George Respiratory Questionnaire (SGRQ), which measures respiratory symptoms, activities limited by symptoms, and the impact of symptoms on well-being and social functioning. Four Veterans had normal PFT parameters, and one had mild obstruction. Of note, this Veteran had mild obstruction at his initial assessment (at the time of biopsy) without further decline in PFT parameters ( Table 2).
Veterans underwent echocardiography and exercise RHC or invasive CPET, which included exercise RHC with a metabolic cart to determine anaerobic threshold. Four of the five Veterans achieved submaximal workloads (cases 1,3,4, and 5), while one Veteran (Case 2) achieved 105% predicted workload. Of the two patients who underwent invasive CPET, one Veteran (Case 5) had reduced peak VO2 (67% predicted) with a reduced anaerobic threshold (37% of predicted VO2 max) 18 consistent with a primary cardiocirculatory limitation and one Veteran (Case 2) had a normal peak VO2 (93% predicted) and normal anaerobic threshold.
for use under a CC0 license.
This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.  19 Resting PH, defined as a mean pulmonary arterial pressure (mPAP) > 20 mmHg, was present in three out of five Veterans (Cases 1, 2, and 4). Two Veterans (Cases 2 and 5) showed evidence of mildly increased pulmonary artery wedge pressures at rest (PAWP) > 15 mmHg, suggestive of left heart congestion. With peak exercise, one Veteran (Case 3) demonstrated exercise PH with mPAP/CO slope > 3 mmHg/L/min and three Veterans demonstrated evidence of moderate isolated post-capillary PH (Cases 1, 4, 5) and met criteria for heart failure with preserved ejection fraction (HFpEF), which is defined by either an exercise PCWP > 25 mmHg or PCWP/CO slope of > 2 mmHg/L/min ( Table 2).
On transthoracic echocardiography, we identified borderline or overt RV enlargement in three Veterans (60%). 20 RVOT acceleration time, a well-established surrogate measure of pulmonary pressure, 21-23 was mildly reduced in three out of five Veterans (Cases 3-5).
Echocardiographic measures of left ventricular size and systolic and diastolic function were normal ( Table 2). Together, these findings suggest that some Veterans with PDRS in our study have mild/moderate PH that is predominately post-capillary, and/or HFpEF, which could explain their continuing exertional dyspnea.

Pathological evaluation
Microscopic re-evaluation of lung biopsy sections from Veterans with PDRS confirmed our previous findings of intrapulmonary inflammation and multicompartmental lung pathology. 5,24 In this study, we focused on vascular pathology. In BVB arteries, pathological changes were characterized by mild-to-moderate media thickening due to smooth muscle for use under a CC0 license.
This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. hyperplasia/hypertrophy, and fibrosis of adventitia, usually without thickening. Neointimal formation was variable and often had eccentric, cushion-like localized manifestations ( Figure   1A). Pathological changes were not limited by BVB arteries but extended to distal vessel generations located within alveolar tissue. IA arteries were characterized by moderately increased intima and media thicknesses and significant fibrosis and thickening of adventitia.
Neointimal lesions were accompanied by moderate-to-severe fibrosis (Figure 2A). Some arteries (~ 5-10%) showed laminar fibrosis of the intima due to the concentrically arranged fibrous layers leading to partial obstruction (also known as "onion-skin lesions"). In comparison to PAH patients, Veterans with PDRS had more prominent fibrosis at all generations of pulmonary arteries analyzed (Figures 1, 2). Comparison analysis of morphometric parameters showed a mild increase of fractional thickness of intima and moderate increase of media in both BVB and IA arteries in Veterans with PDRS compared to ND controls and PAH patients. Fractional thickening of adventitia was similar in BVB arteries in all study groups, but gradually increased in IA arteries in Veterans with PDRS and PAH patients compared to ND controls (Figures 1 and   2). In contrast to PAH patients, Veterans with PDRS did not display plexiform or dilation/angiomatoid lesions.
Although end-stage PAH patients showed more severe remodeling in BVB and IA arteries, IA veins showed similar levels of intima and adventitia fractional thickening in Veterans with PDRS and PAH patients compared to ND controls (Figure 3A-C). However, in Veterans, IA veins were characterized by marked fibrosis of intima and adventitia, usually with increased thickening and formation of multiple layers of elastic laminae, but without features of luminal occlusion, muscular hyperplasia or dilation/angiomatoid lesions seen in pulmonary venoocclusive disease or chronic thromboembolic PH. [25][26][27] for use under a CC0 license. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. To determine which anatomical compartment, arterial or venular, is predominately involved in vessel pathology in PDRS, we calculated the ratio of wall thicknesses between IA pulmonary arteries and veins. This index was measured for each study participant as the mean of sum of intima, media and adventitia fractional thicknesses for arteries divided by the mean of sum of fractional thicknesses of intima and adventitia for veins. Figure 3D shows a modest increase in the artery/vein index in PAH patients compared to ND controls, while lungs from PDRS patients had a reduced artery/vein index compared to both ND controls and PAH patients.
This finding indicates more prominent remodeling of pulmonary veins relative to arteries in PDRS, which is distinct from the predominant pre-capillary vessel pathology in PAH.
To further investigate whether pathological changes in the lung vasculature in Veterans with PDRS were distinct from ND controls and PAH patients, we performed unsupervised hierarchical clustering using all morphometric parameters. This approach showed that the optimal number of clusters was 3 (Supplemental Figure 1) and separated ND controls and PAH patients; however, some Veterans were incorporated into the clusters containing ND controls or PAH patients. The dendrogram in Figure 4 shows that the 1 st cluster combines all ND controls and 16 Veterans; the 2 nd cluster consists of Veterans only; and the 3 rd cluster combines 10 Veterans and all PAH patients. To confirm whether this distribution reflects mild, moderate, and severe vascular pathology, we grouped all morphometric parameters from Veterans according to clusters and performed comparison analysis between clusters. We found a significant increase of all measured parameters across the three patient clusters, consistent with increased pathological remodeling in all vessel wall components (Supplemental Figure 2). Secondary analysis of morphometric parameters from the five Veterans who participated in the clinical follow-up study showed that pathological remodeling in three (Cases 2, 4 and 5) fell into the least affected cluster for use under a CC0 license. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

DISCUSSION
This clinical follow-up study showed that all five Veterans with biopsy-proven PDRS had continued dyspnea at exertion and evidence of PVD. RHC showed that three of five Veterans had mild resting PH and a fourth had exercise PH. In all cases, PVR was < 2 Wood units, suggesting post-capillary PH. Although post-capillary PH is well-described in the setting of left heart failure, [28][29][30] no participants in our study had echocardiographic evidence of heart failure with reduced or preserved ejection fraction (HFrEF or HFpEF). However, we observed marked fibrotic remodeling of intra-acinar veins in Veterans, which was relatively more severe than in PAH patients. Thus, we speculate that fibrous remodeling of pulmonary veins and resulting narrowing of the cross-sectional luminal area could contribute to the post-capillary PH observed in these Veterans. Although we found no echocardiographic evidence for diastolic dysfunction in our study, left-heart catheterization with assessment of the diastolic pressure-volume relationship will be needed to conclusively exclude diastolic dysfunction as a contributor to post-capillary PH in these Veterans.
We previously reported multicompartmental lung pathology with hypertensive-type changes in BVB arteries and capillary rarefaction in alveolar tissue were important pathological features of PDRS. 5 In this study, we extend these findings to show that vascular pathology involves multiple vascular compartments (arteries, capillaries, and veins) and multiple anatomic levels (BVB arteries, IA arteries and veins). Generally, these findings are similar to, although less severe than, patients with idiopathic PAH who underwent lung transplantation. Together, for use under a CC0 license. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted May 16, 2023. ; https://doi.org/10.1101/2023.05.15.23289956 doi: medRxiv preprint our studies suggest that vasculopathy and subsequent PH may explain exertional dyspnea and exercise limitation in some Veterans who have an otherwise unremarkable non-invasive evaluation.
We speculate that PH in this cohort of Veterans with PDRS is associated with mixed inhalation exposures received during deployment. 5,6 Recent epidemiologic and animal studies have been demonstrated a causative relationship between inhalation of noxious substances and development of PH. Thus, PH has been reported in association with chronic exposure to tobacco and woodsmoke, air pollution, particulate matter with a diameter < 2.5 µm (PM2.5), and exposure to other chemicals, dust, gases or fumes exposures. [31][32][33][34][35][36] Experimental works demonstrated that exposure to diesel exhaust or urban ambient PM2.5 is sufficient to induce marked pulmonary arterial thickening and increased right ventricle systolic pressure in mice, supporting a causal relationship between inhalation of noxious particles and the development of PH. [37][38][39] Conceptually, vasculopathy in the setting of inhalation of noxious substances could result from "inside-out" or "outside-in" processes. In the inside-out model, absorption of toxins such as cigarette smoke or pollution into the bloodstream at the capillary/alveolar interface could result in direct damage to endothelial cells through well-established pathways involving oxidative stress and induction of endothelial cell dysfunction. [40][41][42][43][44] Alternatively, vasculopathy could result from an "outside-in" process whereby inflammation from adjacent airways induced by inhalation of noxious particles causes pathologic remodeling of blood vessels. In support of this concept, the extent of perivascular inflammation in PAH has been shown to correlate with pulmonary hemodynamics, vascular remodeling, and clinical outcome. [45][46][47][48] Persistent airway and parenchymal lymphocytic inflammation in association with pure Th1 adaptive immune response was defined as a central feature of PDRS that raises the possibility that the PVD could progress for use under a CC0 license. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted May 16, 2023. ; https://doi.org/10.1101/2023.05.15.23289956 doi: medRxiv preprint in this patient population over time. 5,24 The relative contributions of inside-out and outside-in pathology in each vascular compartment and anatomic level await further testing in relevant animal models.
Our study has several important limitations. First, because vasculopathy in these Veterans occurs in association with pathology in other lung compartments including small airways, the lung parenchyma, and the pleura, we cannot definitely link PVD with symptoms in individual Veterans. Emerging non-invasive techniques capable of interrogating multiple lung compartments simultaneously (such as hyperpolarized xenon MRI 49 ) may help to assign the relative importance of pathology in various lung compartments in individual patients. Second, pathologic analyses were performed on archived biopsies obtained more than a decade prior to exercise RHC. Thus, we cannot ascertain whether PH was present at the time of biopsy or developed or progressed over time. Third, we performed exercise RHC in three Veterans using a multistage exercise protocol prior to developing the capability to perform invasive CPET using a continuous ramp exercise protocol. In these Veterans, exercise could not be standardized against anaerobic threshold, and thus the exercise level likely varied between these participants and those who performed invasive CPET due to differences in exercise protocol. Finally, we acknowledge that our case series of five patients is insufficient to conclude that development of PH in these patients is widespread. However, it is important to note that several of the Veterans who underwent RHC had relatively minor vasculopathy present on surgical lung biopsies compared to other PDRS patients, suggesting more prominent findings may exist in other Veterans with PDRS. Larger studies involving a group of Veterans with heterogenous civilian and deployment exposure profiles will be necessary to estimate the prevalence of PH in this population.
for use under a CC0 license.
This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. In summary, here we report complex vascular pathology in surgical lung biopsies from Iraq and Afghanistan Veterans with unexplained dyspnea accompanied by PVD in a subset. Our findings suggest PVD should be considered in the evaluation of Iraq and Afghanistan Veterans with unexplained dyspnea.
for use under a CC0 license. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
Median and range (minimum and maximum) are indicated for age. Number (percent) is indicated for gender and smoking history.
for use under a CC0 license. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.