Histologic patterns of chronic interstitial lung disease in dogs

Abstract

Chronic interstitial lung disease (cILD) is uncommon in dogs and little is known of the pathogenesis, apart from the condition in West Highland White Terriers. This study aimed to characterize histologic lesions of canine cILD, compare the lesions and clinical features, and classify the histopathologic patterns according to criteria used in humans. The study included 24 postmortem cases of subacute or chronic ILD in >6-month-old dogs with respiratory signs. Histologic lung lesions included attenuated bronchiolar epithelium, alveolar edema, type II pneumocyte proliferation, fibrosis of alveolar septa, fibrin or fibrous tissue within alveoli or bronchioles, and hyaline membranes. Of the 24 cases, 8 were classified as organizing diffuse alveolar damage, 4 as organizing pneumonia, and 3 as acute fibrinous and organizing pneumonia; 9 were unclassifiable and considered as nonspecific interstitial lung disease. None fulfilled criteria for usual interstitial pneumonia. Potential causes included aspiration of gastric or foreign material, prior acute respiratory distress syndrome, or failed healing of pneumonia. Left-sided heart failure was identified in 12 of 24 cases but was not considered to directly cause the interstitial lung lesions. Gross lesions of cor pulmonale were associated with organizing pneumonia and longer clinical duration. The cases had diverse histologic lesions and patterns of lung fibrosis, but the results suggested that these may represent divergent responses to overlapping causes of lung injury rather than distinct diseases. These findings clarify the pathogenesis of cILD in dogs, the mechanisms of initial damage, and the future development of approaches to delay or predict disease progression.

Interstitial lung diseases in dogs include acute diffuse alveolar damage (DAD), which is often related to the clinical condition acute respiratory disease syndrome (ARDS),^6,38 and chronic interstitial lung diseases (cILDs) that have a longer clinical course or histologic lesions of type II pneumocyte proliferation or fibrosis. cILD is recognized in West Highland White Terriers ²⁰ (WHWTs) but poorly described in other dog breeds.^{7,15,25,30,31} In humans, specific histologic patterns of cILD are combined with knowledge of diagnostic imaging to assign a clinicopathologic diagnosis ( Table 1 ).^19,27,37 A single cause (such as rheumatoid arthritis or systemic sclerosis) can result in different histologic patterns. Thus, the histologic pattern and clinicopathologic diagnosis are not mainly used to identify a specific cause, but instead to predict the likely progression, disease course, or effective therapy. ¹⁹ Idiopathic interstitial lung diseases have been identified in humans for >50 y, and considerable progress has been made in understanding the nature of these conditions and developing a clinically meaningful classification scheme, even if the causes often remain obscure.^1,23,37

Table 1.

Classification of canine lung lesions according to criteria used for human chronic idiopathic interstitial lung diseases.^4,17,24

Criteria for diagnosis		Histologic patterns
	Organizing DAD	OP	AFOP	UIP	NSIP
Major histologic differential diagnoses	UIP, NSIP, OP	Organizing DAD, constrictive bronchiolitis obliterans	OP, BP, Organizing DAD	Fibrosing NSIP, Organizing DAD	UIP, Organizing DAD
Histopathology definition	Hyaline membranes; increased fibroblasts in alveolar septa with or without fibrosis; diffuse, monophasic; lung architecture is visible	Air spaces contain fibroblastic polyps; patchy, monophasic; lung architecture is visible; interstitial inflammation of affected alveoli	Fibrin “balls” in alveoli or alveolar ducts, usually with OP lesions	Patchy distribution, temporal heterogeneity, including areas with normal lung, immature fibroblastic foci, confluent scarring, and honeycombing	Diffuse monophasic interstitial fibrosis ± chronic interstitial inflammation
Distribution of fibrosis	Diffuse/uniform or patchy	Patchy	Patchy	Patchy, variegated	Diffuse
Location of fibrosis	Mainly alveolar septa, ± some polyps in alveoli ± bronchioles	Dominant feature is polyps in alveoli ± bronchioles	Alveoli, usually peribronchiolar	Alveolar septa	Alveolar septa, may be some in alveoli (OP lesion)
Temporal heterogeneity (mature fibrosis + fibroblastic foci)	Absent	Absent	Absent	Required for diagnosis	Absent/minimal; dense or loose connective tissue, but the same everywhere
Temporal heterogeneity- fibroblastic foci	Absent	Absent or present	Absent	Required for diagnosis	Absent or present
Honeycombing (cystic spaces)	Normal lung architecture	Normal lung architecture	Normal lung architecture	Yes	Absent or inconspicuous
Hyaline membranes	Present	Absent	Absent	Absent, or present in cases with accelerated decline or acute exacerbation	Absent
Inflammation	Minimal	None, or mild-moderate	Mild-moderate, lymphocytic	Absent, or mild interstitial or lymphoid aggregates	None, or mild-moderate, not marked
Vascular lesions	None	Not specified	Not specified	Medial thickening and intimal fibrosis	Not specified

AFOP = acute fibrinous and organizing pneumonia; BP = bacterial bronchopneumonia; DAD = diffuse alveolar damage; NSIP = nonspecific interstitial pneumonia; OP = organizing pneumonia; UIP = usual interstitial pneumonia.

Proposed causes of cILD in humans include aspiration of acidic gastric content, inhalation of inorganic dusts (pneumoconiosis) or other toxicants, drug toxicity, hypersensitivity to inhaled allergens (hypersensitivity pneumonitis), other immune-mediated diseases, radiation exposure, and chronic consequences of respiratory virus infection. ³⁰ Many cases in dogs are idiopathic, or have only limited evidence of the above causes.^15,31

Our first objective was to characterize the range of histologic lesions in fatal cases of cILD in dogs, and to identify associations among different clinical and pathology features. A second objective was to apply the classification scheme for human idiopathic interstitial pneumonia, and to investigate if similar categories in dogs have consistent patterns of histologic and clinical features that could advance our understanding of this condition.

Materials and methods

Postmortem cases of cILD in dogs were retrieved from the Animal Health Laboratory (University of Guelph, Guelph, Ontario, Canada) archive (2007–2024). Inclusion criteria were dogs >6-mo-old with a clinical history of respiratory signs (such as cough, tachypnea, or dyspnea) that were considered the cause of death or euthanasia, with available histologic sections and paraffin-embedded tissue, having subacute or chronic interstitial lung lesions identified histologically (with epithelial hyperplasia and/or fibrosis) that were not attributed to eosinophilic lung diseases, renal failure, bacterial or fungal pneumonia, or pulmonary hyalinosis. For the 24 cases that met the inclusion criteria, case information was obtained from clinical records and pathology reports (Tables 2, 3). The median number of available lung sections per case was 3.5 (quartiles: 3–13; range: 1–11). Histologic lung sections stained with H&E, Masson trichrome, Perls Prussian blue, and Verhoeff–van Gieson stains from all cases were initially evaluated (M. Kozawa, J. Caswell) in a blinded manner without knowledge of the clinical history or pathology report, lesions were scored and classified, and consensus was reached by discussion. Subsequent case interpretation was additionally based on information obtained from clinical and pathology records. Lesions were scored for items listed in Table 3 as follows: absent (score 0); present but a minor feature of the lesions (score 1); present as a prominent feature of the lesions (score 2). Iron in alveolar macrophages in Perls-stained sections was scored as absent (0), rare (1), widespread but faint (2), visible at lower magnification (3). For selected cases, sections were immunolabeled for smooth muscle actin (n = 4), cytokeratin (n = 2), CD31 (n = 1) or IBA1 (n = 1), or stained with periodic acid–Schiff reaction (PAS; n = 4). Test results for infectious agents were retrieved from the pathology records, but we did not do additional testing.

Table 2.

Summary of clinical and pathology findings in 24 dogs with chronic interstitial lung disease.

Continuous data	Median [quartiles]
Age, y	9 [7–11]
Time from onset of respiratory signs to death, d	11.5 [5–75]
Clinical and gross pathology data	No. of cases
Sex
Female	17
Male	7
Onset of respiratory signs
Abrupt	11
Gradual	7
Exacerbation	6
Aspiration suggested by history
Yes	6
Suggested	3
No	15
Pulmonary hypertension
Clinically diagnosed or suspected	6
No	4
Echocardiography not done	14
Cor pulmonale*
Yes	9
No or not tested	15
Left-sided heart disease score †
0 = no disease	6
1 = minimal changes	2
2 = disease without failure	4
3 = limited pathology evidence of failure	5
4 = clinical diagnosis of failure	7
Left-sided heart lesions, pathology
Yes	18
No	6

^* Right-sided heart remodeling or failure, based on pathology findings.

^† Left-sided heart disease, considered a contributor to respiratory signs based on all clinical and pathology information.

Table 3.

Categories and scoring of histopathology findings in the lungs of 24 dogs with chronic interstitial lung disease.

Histologic finding	No. of cases*
Uniformity across lung sections
Non-uniform	10
Uniform	13
Distribution in affected areas
Patchy	15
Diffuse	9
Mono- or polyphasic
Monophasic, subacute	5
Monophasic, chronic	7
Polyphasic	12
Intralobular distribution
Centrilobular	5
Panlobular	18
Histopathology score†	No. of cases
Uniformity across sections
0	2
1	5
2	16
Organizing pneumonia
0	9
1	8
2	7
Fibrosis in alveolar septa
0	1
1	7
2	16
Type II pneumocytes
0	3
1	6
2	15
Intra-alveolar fibrin
0	6
1	12
2	6
Hyaline membranes
0	12
1	4
2	8
Inflammatory cells in alveolar septa
0	6
1	11
2	7
Alveolar edema
0	2
1	6
2	16
Alveolar macrophages
0	1
1	4
2	18
Iron in alveolar macrophages
0	4
1	4
2	12
3	3
Perivascular fibrosis
0	11
1	5
2	8

^* Number of cases, based on those with available information.

^† Histopathology scores: 0 = absent; 1 = present but infrequent or a subtle feature of the lesions; 2 = present as a major feature of the lesions in the case.

The histologic pattern in each case was assigned using human criteria for organizing diffuse alveolar damage (organizing DAD, with prominent hyaline membranes), acute fibrinous and organizing pneumonia (AFOP, with epithelialized fibrin balls in alveoli), organizing pneumonia (OP, with polyps of fibrous tissue in alveolar and/or bronchiolar lumens), usual interstitial pneumonia (the histologic lesion of idiopathic pulmonary fibrosis), and others (Table 1).^11,19,27,37 Cases that did not align with these histologic patterns were considered as nonspecific ILD; these had chronic interstitial fibrosis, type II pneumocyte hyperplasia, and/or bronchiolar epithelial attenuation but no or minimal hyaline membranes, alveolar fibrin balls, or fibrous polyps. Because the human criteria have not been validated in dogs, the canine cases in our study were classified separately by both the human criteria and by the predominant histologic lesion.

The distribution of histologic lesions within the lung tissue were categorized as uniform (lesions present in each of the lung sections examined) or non-uniform (lesions in some lung sections but not others; presumably reflecting a regional distribution in the lung), and as diffuse (within an affected lung section, lesions were present in all areas, even if there was variable severity of lesions) or patchy (within an affected lung section, lesions were present in some areas but other areas had essentially normal lung). Histologic lesions were categorized as centrilobular if they predominantly affected the alveoli and alveolar ducts located near bronchioles) or panlobular if lesions affected the entire lobule; we did not observe perilobular lesions (predominantly affecting the periphery of the pulmonary lobule) in our study. Histologic lesions were categorized as monophasic subacute (interstitial or intrabronchial fibroblast proliferation with minimal or immature collagen), monophasic chronic (interstitial or intrabronchial fibroblast proliferation with mature collagen), or polyphasic (lesions with areas of mature fibrosis as well as areas with fibrin and necrosis, or interstitial fibroblasts and minimal collagen).

We use cor pulmonale here to describe lung disease that resulted in right-sided heart remodeling with or without right-sided heart failure, as was suggested by clinical or gross pathology evidence of right atrial dilation, right ventricular dilation or hypertrophy, and/or hepatomegaly, hepatic congestion, and ascites. Left-sided heart disease (with or without failure) was based on clinical diagnosis or gross or histologic evidence of left-sided heart lesions including myxomatous valve disease (endocardiosis, n = 15), infective endocarditis (n = 1, with concurrent endocardiosis), dilated cardiomyopathy (n = 2), and left ventricular systolic dysfunction of uncertain cause (n = 1). Left-sided heart changes were categorized as absent (score 0), minimal (such as mild endocardiosis; score 1); left-sided disease with no clinical or pathology evidence of left-sided heart failure (score 2); left-sided disease with limited pathology evidence of failure, such as severe cardiac lesions accompanied by hemosiderin-laden pulmonary alveolar macrophages and/or diffuse pulmonary edema in areas with minimal interstitial lesions (score 3); or left-sided heart disease with conclusive clinical evidence of left-sided heart failure based on thoracic radiographs or echocardiography (score 4). Pulmonary congestion, edema, and hemorrhage with hemosiderin-laden macrophages can occur in interstitial lung disease; hence, these lesions were not conclusive indicators of left-sided heart failure (score 4) in this study.

Statistical analysis

Data were analyzed to identify associations among histologic diagnoses, specific lesions, and clinical or radiographic findings. Associations between categorical variables (such as histologic diagnosis, gradual vs. abrupt onset, presence of heart disease) were evaluated by Fisher exact or chi-square tests using Graphpad Prism. Age and duration of respiratory signs were not normally distributed and are reported as median and first and third quartiles. Associations of categorical variables (such as histologic diagnosis, presence of heart disease) and continuous variables (age, duration of respiratory signs) were evaluated by Mann–Whitney or Kruskal–Wallis tests using Graphpad Prism. Associations among variables were further analyzed using factor analysis of mixed data using R Studio, considering the 24 cases and the main continuous and categorical variables in the study.^16,28 Linearity among the variables was evaluated by the pairwise comparisons described above. Lesions that were the basis for the histologic patterns (e.g., scores for hyaline membranes, OP lesions) were included in the factor analysis, but the histologic pattern itself was excluded to avoid multicollinearity. The low number of cases is a limitation of the factor analysis.

Results

Clinical findings

Our 24 cases included 7 males and 17 females. The breeds were WHWT (4), Chihuahua (3), Yorkshire Terrier (2), Bichon Frise (2), Pomeranian (2), and 1 each of Australian Shepherd, Bearded Collie, Boxer, Cavalier King Charles Spaniel, Coton du Tulear, Doberman Pinscher, Golden Retriever, Great Dane, Labrador Retriever, Pekingese, and Toy Poodle.

The median age was 9 y (first and third quartiles: 7–11 y; range: 0.6–16 y). The median duration of clinical signs was 11.5 d (quartiles: 5–75 d; range: 1 d to 2 y), and the onset of respiratory signs was classified as gradual in 7 cases (progressive worsening over ≥7 d), abrupt in 11 (sudden onset of severe signs or rapid worsening over <7 d), and exacerbation or relapse of prior disease in 6 cases (>1 mo of respiratory signs with sudden worsening).

Pathology findings

The microscopic distribution of lung lesions was classified as uniform or non-uniform across the available lung sections, diffuse or patchy within affected sections of lung, and centrilobular or panlobular. The median number of lung sections per case was 3.5 (quartiles: 3–13). Of the 24 cases, 13 had relatively uniform lung lesions (lesions affected each of the histologic sections), and 10 had non-uniform lesions (some lung sections had lesions and others did not, perhaps reflecting cranioventral, lobar, or unilateral disease); 1 of the 24 cases had only a single section of lung (Table 3). Within the affected sections of lung, lesions were diffuse in 9 cases (involving all of the lung tissue, although sometimes with differing severity) and patchy in 15 cases (some areas with lesions and others without; Fig. 1A ). Lesions were interpreted as monophasic subacute in 5 cases, monophasic chronic in 7 cases, and polyphasic in 12 cases. Those with polyphasic lesions had intraluminal fibrin, hyaline membranes, and/or active cell death in some areas and immature or mature fibrous tissue in other areas ( Fig. 1B ). Monophasic subacute lesions had fibroplasia with many fibroblasts and immature collagen, whereas monophasic chronic lesions had fibrosis with mature collagen. Lesions affected entire lobules (panlobular) in 18 cases, tended to target bronchioles and adjacent alveoli in 5 cases (centrilobular; Fig. 1C ), or affected bronchioles with interstitial fibrosis but no alveolar damage in 1 case (Table 3).

Figure 1.

Chronic interstitial lung disease in dogs. A. Patchy distribution, with affected (a) adjacent to normal (n) areas. H&E. B. Polyphasic lesion of organizing pneumonia with intra-alveolar aggregates of fibrin (arrows) and fibrous tissue covered by pneumocytes (asterisks). H&E. Inset: collagenous tissue forms an intra-alveolar polyp. Masson trichrome. C. Centrilobular distribution, with intra-alveolar fibrin and hypercellular alveolar septa centered on a bronchiole (b). H&E. D. Bronchiolar epithelium (arrow) attenuation with a normal bronchiolar wall. Intraluminal fibrin (arrowhead) covered by epithelium. H&E. E. Bronchiolar wall destruction and loss or attenuation of epithelium. Inset: higher magnification of the area in box, with partial loss of smooth muscle and deposition of collagen. H&E. F. Organizing pneumonia, with 2 intra-alveolar discrete masses (arrows) of fibrous tissue containing collagen (inset, Masson trichrome), covered by epithelium. Hypercellularity of alveolar septa and intra-alveolar hemorrhage are present. H&E. G. Organizing pneumonia, with filling of an alveolus by collagenous connective tissue. Asterisks indicate approximate borders of the alveolus. Masson trichrome. H. Diffuse hypercellularity of alveolar septa (H&E), with minimal collagen formation (inset, Masson trichrome). Hyaline membrane is present (arrow). H&E. I. Interstitial fibrosis, with thickening of alveolar septa by collagen (Masson trichrome) and only minimal hypercellularity (inset, H&E). Type 2 pneumocytes are present.

Attenuation of bronchiolar epithelium was a prominent (score = 2) histologic lesion in 17 cases, minor (score = 1) in 4 cases, and minimal or absent (score = 0) in 2 cases ( Fig. 1D ); autolysis precluded scoring of 1 case. Attenuation affected mainly small bronchioles as well as those up to ~800 μm diameter, but not bronchi. Occasionally, the full thickness of the bronchiolar wall appeared necrotic, with degeneration and loss of the smooth muscle, as well as fibrosis of the wall ( Fig. 1E ; Table 3).

Alveolar edema and increased number of alveolar macrophages were present in 22 of 24 and 23 of 24 cases, respectively, and alveoli also contained aggregates of non-cellular eosinophilic material (possibly retained or degraded surfactant or other proteins; Suppl. Fig. 1A ) in 5 cases, and numerous erythrocytes in 4 cases. Immature or mature fibrous tissue within airspaces (OP) was prominent in 7 cases, minor in 8 cases, and minimal or absent in 9 cases, and affected the lumen of alveoli (6 cases), bronchioles (1 case), or both (8 cases). This material typically formed discrete intraluminal polyps partially covered by epithelial cells ( Fig. 1F ) or occasionally filled most of the lumen ( Fig. 1G ). Fibroplasia or fibrosis of alveolar septa was prominent in 16 cases, minor in 7 cases, and minimal or absent in 1 case, appearing either as fibroblast proliferation with no or minimal collagen ( Fig. 1H ) or as mature collagen that thickened the alveolar septa ( Fig. 1I ).

Type II pneumocytes were prominent in 15 cases, minor in 6 cases, and inapparent in 3 cases. Type II pneumocytes sometimes continuously lined the alveolus; but, more frequently they formed single or clustered epithelial cells featuring irregular cell shapes, abundant and sometimes foamy cytoplasm, marked anisocytosis and anisokaryosis, and large nuclei ( Fig. 2A ). Their prominent anisocytosis and tendency to cover the alveolar wall or intraluminal material distinguished them from hypertrophied alveolar macrophages and was confirmed by immunolabeling for cytokeratin but not IBA1 ( Suppl. Fig. 1B–D ). Four cases had areas with squamous metaplasia of epithelium in bronchioles and adjacent alveoli ( Fig. 2B ). Rarely, alveoli were lined by ciliated epithelial cells, interpreted as migration of bronchiolar epithelium into alveoli ( Suppl. Fig. 1E, 1F ).

Figure 2.

Chronic interstitial lung disease in dogs. A. Alveoli contain type 2 pneumocytes (arrows), which are often individualized, with anisocytosis, abundant cytoplasm, and 1 or 2 large nuclei. H&E. B. Squamous metaplasia of alveolar epithelium, forming a stratified layer of irregularly arranged epithelial cells. Inset: higher magnification of the area in box. H&E. C. Acute fibrinous and organizing pneumonia, with intra-alveolar discrete masses of fibrin partially covered by epithelium (arrow). H&E. D. Organizing diffuse alveolar damage (DAD), with thick hyaline membranes covering the alveolar septum (arrow). Alveolar septa thickened by mononuclear leukocytes. H&E. E. Organizing DAD, with a thin hyaline membrane (arrow) separated from the alveolar septum and PAS-positive (inset, arrowheads). H&E. F. Organizing DAD. Pneumocytes are cytokeratin-positive but the thin hyaline membrane (arrow) is not immunolabeled, consistent with absence of type I pneumocytes. Immunohistochemistry. G. Collagen deposits in and around blood vessel walls and extending into adjacent alveolar septa. H&E. Inset: collagen stained blue with Masson trichrome. H. Lung from a normal dog (for comparison to Fig. 1G). Masson trichrome. I. Mineral aggregate (arrow, inset) in a bronchiole with attenuated epithelium and protein-rich exudate. Foreign material is consistent with aspiration of acidic gastric content (dog was administered oral calcium suspension, vomited, and subsequently developed tachypnea and dyspnea that fluctuated over the next 12 d). H&E.

Intra-alveolar fibrin was prominent in 6 cases, minor in 12 cases, and absent in 6 cases. In 12 cases, the fibrin formed discrete aggregates within the lumen (“fibrin balls”), often covered by thin or plump epithelial cells ( Fig. 2C ). Hyaline membranes were prominent in 8 cases, minor in 4 cases, and absent in 12 cases. Hyaline membranes formed a thick layer covering the alveolar surface ( Fig. 2D ) or had thin eosinophilic membranes that separated from the alveolar wall. The latter were confirmed as hyaline membranes rather than detached type I pneumocytes based on PAS-positivity and absence of cytokeratin immunolabeling (Fig. 2E, 2F).

Inflammatory cells in alveolar septa were prominent in 7 cases, minor in 11 cases, and minimal in 6 cases, based on H&E staining. Within the lung interstitium (distant from bronchioles), perivascular fibrosis was prominent in 8 cases, minor in 5 cases, and not evident in 11 cases; in some cases, perivascular fibrosis may have been obscured by interstitial fibrosis. This fibrosis also expanded alveolar septa adjacent to the perivascular lesions (Fig. 2G, 2H). Based on the prominent media, these vessels were interpreted as intralobular branches of pulmonary arteries, although only an external elastic lamina was present in these small branches. Adjacent to bronchioles, larger pulmonary arteries had lesions in 4 cases including intimal, medial, and adventitial hyperplasia.

To consider different patterns of lung fibroplasia, we classified anatomically the process of lung injury and the resulting fibrosis as affecting alveolar lumens, alveolar septa, or confluent. More than 1 of these 3 patterns was recorded in 15 of 24 cases. The intraluminal pattern was identified in 19 cases, with alveolar or bronchiolar lumens containing discrete masses of fibrin (n = 2), fibrous tissue (7) or both (10). The alveolar septal pattern was identified in 17 cases, with prominent hyaline membranes (2), substantial alveolar septal fibrosis (10), or both (5). The confluent pattern was identified in 2 cases; in this pattern, alveolar lumens and septa were no longer visible because of fibrous tissue from architectural effacement and confluent scarring ( Suppl. Fig. 1G, 1H ). An additional 2 cases had confluent scarring, but the lung architecture remained visible in Masson trichrome-stained sections. Only one case had fibrosis with multiple dilated airspaces, suggestive of honeycombing, although the spaces were lined by low-cuboidal rather than columnar epithelium ( Suppl. Fig. 1I ).

Overall histologic diagnosis

Classification based on the predominant histologic lesion (hyaline membranes, epithelialized fibrin balls in alveoli, or polyps of fibrous tissue in airspaces) resulted in 9 cases of organizing DAD, 9 cases of nonspecific ILD, 4 cases of OP, and 2 cases of AFOP. Classification based on the human criteria for idiopathic interstitial lung diseases (Table 1) resulted in 11 cases of organizing DAD, 9 cases of nonspecific ILD, 3 cases of OP, and 1 case of AFOP. Using the human criteria, cases diagnosed as organizing DAD all had hyaline membranes (by definition); 8 of 11 also had AFOP lesions (alveolar fibrin balls), and 4 of 11 had OP lesions (alveolar fibroblastic or fibrous polyps). For cases of nonspecific ILD, 2 of 9 had OP lesions and 2 of 9 had AFOP lesions that were minimal and thus did not justify overall diagnoses of OP or AFOP, respectively. Cases diagnosed as OP did not have hyaline membranes or prominent AFOP lesions (by definition). The one case diagnosed as AFOP had intraluminal fibrin balls (by definition) as well as OP lesions, but no hyaline membranes (by definition).

Considering these diagnoses based on the human criteria, we noted the following trends. The median duration of respiratory signs prior to death tended to be shorter for cases with organizing DAD and longer for OP, nonspecific ILD, and AFOP (median: 7 vs. 45, 70, and 66 d, respectively; Kruskal–Wallis test with Dunn multiple comparisons test; Table 4 ). Abrupt onset of clinical signs (compared with gradual onset or exacerbation) was reported in 8 of 11 cases of organizing DAD compared with 0 of 3 OP, 3 of 9 nonspecific ILD, and 0 of 1 AFOP cases (p = 0.04; organizing DAD vs. others; Fisher exact test, odds ratio = 8.9). Gross lesions suggesting cor pulmonale were identified in 3 of 3 cases with OP compared with 3 of 11 cases with organizing DAD, 2 of 8 cases with nonspecific ILD, and the single AFOP case (p = 0.047; OP vs. others; Fisher exact test, odds ratio = 15.6). Left-sided heart failure was diagnosed in 6 of 9 cases of nonspecific ILD and 5 of 11 cases of organizing DAD compared with 1 of 3 cases of OP and 0 of 1 case of AFOP. Centrilobular (vs. panlobular) distribution was noted in 4 of 11 cases of organizing DAD compared with 1 of 3 cases of OP, 0 of 8 cases of nonspecific ILD, and 0 of 1 case of AFOP. Alveolar edema was prominent (score 2) in only 5 of 9 nonspecific ILD cases and 2 of 3 OP cases, in contrast to 9 of 11 organizing DAD cases. The 4 diagnoses did not differ consistently with respect to age, radiographic lesion distribution, history suggesting aspiration, uniformity of histologic lesions across different sections, distribution within affected sections of lung, monophasic versus polyphasic lesions, bronchiolar necrosis, or type II pneumocytes.

Table 4.

Clinical and pathology findings among different histologic diagnoses. Data are the number of cases except where otherwise indicated.

Histologic diagnosis	Organizing DAD, n = 11	Nonspecific interstitial lung disease, n = 9	OP, n = 3	AFOP, n = 1
Age, y (median [quartiles])	(11 [6.85–12.45])	(9 [7–11])	(7.8 [6.25–8.9])	(10 [10–10])
Duration of respiratory signs, d (median [quartiles])	(7 [4–10])*	(70 [2–105])	(45 [29–45])	(66 [66–66])
Abrupt onset†	8*	3	0	0
Aspiration‡	4	2	0	0
Cor pulmonale§	3	2	3*	1
Left-sided heart failure¦	5	6	1	0
Left-sided heart lesion#	8	8	1	1
Left atrial dilation¶	1	2	1	0
Iron, alveolar macrophages**	5	8	2	0
Patchy distribution††	7	6	1	1
Uniform among sections††	6 ‡‡	6	1	0
Mono- or polyphasic	4MS, 1MC, 6P	1MS, 5MC, 3P	0MS, 1MC, 2P	0MS, 0MC, 1P
Centrilobular distribution	4	0	1	0
Bronchiolar necrosis‡‡	9	5 ‡‡	3	0
Alveolar fibrin‡‡	5	1	0	0
Hyaline membranes prominent‡‡	8	0	0	0
Type II pneumocytes‡‡	6	6	3	0
Organizing pneumonia‡‡	3	0	3	1
Alveolar septal fibrosis‡‡	7	7	2	0
Perivascular fibrosis‡‡	2	5	1	0

AFOP = acute fibrinous and organizing pneumonia; DAD = diffuse alveolar damage; MC = monophasic chronic; MS = monophasic subacute; OP = organizing pneumonia; P = polyphasic.

^*Significantly different from other diagnoses, p ≤ 0.05, Kruskal–Wallis test with Dunn multiple comparisons test.

^†Abrupt onset of respiratory signs, rather than exacerbation or gradual onset.

^‡Aspiration suggested by clinical history.

^§Cor pulmonale (right-sided heart remodeling with or without failure) suggested by gross pathology findings.

^¦Left-sided heart failure: conclusive or limited evidence.

^#Left-sided heart lesion identified by gross or histopathology.

^¶Left atrial dilatation identified by echocardiography or gross pathology.

^**Iron in alveolar macrophages based on Perls stain, score ≥2.

^††Distribution: patchy rather than diffuse in affected sections; uniform = lesions in all lung sections examined.

^‡‡One case was omitted because there was only a single lung section (9a), or autolysis (9b).

Various other associations were identified. Dogs that had sections with diffuse rather than patchy lesions were more likely to have a history suggesting aspiration (5 of 9 vs. 1 of 15; p = 0.01, Fisher exact test). These dogs also were younger (6.9 vs. 10.0 y; p = 0.05, t test) and 9 of 9 had panlobular lesions (vs. 9 of 14 for those with patchy lesions). Cases with centrilobular (vs. panlobular) lesions often had shorter durations of respiratory signs (median [quartiles]: 10 [1.5–11.5] vs. 45 [6.75 vs. 108.8]) and tended to not have lesions suggestive of right-sided heart failure (1 of 4 vs. 8 of 17).

Suggested initiating or exacerbating causes of lung disease

Potential causes of the respiratory disease were suggested in a minority of cases, including interstitial lung disease of WHWTs, aspiration of gastric content or foreign material, supplemental oxygen therapy, mechanical ventilation, and an association with left-sided heart disease. The 4 WHWTs were 8–15-y-old, female, with either abrupt onset of respiratory signs or acute exacerbation of prior respiratory disease. The lesions in these cases were patchy or diffuse and included various combinations of intraluminal polyps, type II pneumocyte proliferation, fibrosis of alveolar septa, intra-alveolar fibrin, hyaline membranes, edema, inflammatory cells in alveolar septa, bronchiolar necrosis, and perivenular fibrosis. The lesions were interpreted as consistent with the cILD of this breed; they did not meet the histologic criteria for human usual interstitial pneumonia (idiopathic pulmonary fibrosis). The original diagnostic investigation of 6 cases included negative tests for viral causes. Streptococcus zooepidemicus and S. canis were isolated from lung in 2 of 5 cases with bacterial culture data and were considered of unlikely significance; histologic diagnoses in both cases were nonspecific ILD ( Suppl. Table 1 ).

Aspiration of stomach content or foreign material was indicated by the clinical history in 6 cases, suggested in 3 cases, and absent in 15 cases. Clinical evidence of aspiration in the 6 cases included acute respiratory distress and localized radiographic lung lesions following an episode of vomiting, a known risk factor (pancreatitis) with acute respiratory distress and localized radiographic lesions, observed episodes of gastroesophageal reflux, or acute respiratory distress after administration and regurgitation or vomiting of a new medication ( Fig. 2I ). Potential for aspiration was suggested in 3 additional cases based on known prior episodes of vomiting, seizures, or suspected pancreatitis ( Suppl. material ).

Supplemental oxygen therapy was reported in 15 of 24 cases, for >3 h before death in 9 cases, shortly before death (<3 h) in 4 cases, and for an unspecified duration in 2 cases. Of these 15 cases, 3 cases were mechanically ventilated for <1, 2, and 6 d before death. In all cases, the onset of respiratory signs preceded the administration of supplemental oxygen, and many had a long duration of respiratory disease (median: 30 d [quartiles: 7–87 d]). Histologic lesions were similar in cases receiving supplemental oxygen for >3 h versus those not receiving this treatment (Suppl. material).

Left-sided heart disease was present in 18 of 24 cases, including 12 cases with left-sided heart failure. Dogs with left-sided heart failure (vs. those without left-sided heart disease or failure) tended to have longer durations of respiratory clinical signs, and these signs often were gradually progressive or were an acute exacerbation of chronic disease rather than abrupt onset. Histologic lesions and other case details were generally similar in dogs with left-sided heart failure, left-sided heart disease without failure, and those with no heart disease (all cases had interstitial lung disease; Suppl. material).

Clinical or gross findings suggesting cor pulmonale were present in 9 of 24 cases, namely, right atrial dilation, right ventricular dilation or hypertrophy, and/or hepatomegaly, congestion, and ascites. These cases had longer duration from onset of respiratory signs to death than those without right-sided heart failure (p < 0.05). Respiratory signs tended to have gradual onset or were an exacerbation of chronic disease, rather than an abrupt onset. All 6 cases with data had diffuse or generalized (rather than localized) radiographic lung lesions. OP was more frequent in cases with right-sided heart failure than in those without (p = 0.047; Suppl. material).

Factor analysis

There was only weak clustering of cases based on direct examination of the data and pairwise statistical tests. Factor analysis was used to objectively confirm this finding, although the number of cases in our study is a limitation for using this method. The top 5 dimensions explained only 14%, 12%, 10%, 7%, and 7% of the variance, which further supported the absence of defined clusters or subgroups among the 24 cases in our study.

Discussion

We aimed to characterize the range of histologic lesions of cILD in dogs, identify relationships between different clinical and pathology features, and apply the classification scheme for human idiopathic interstitial pneumonia to clarify the nature of this condition in dogs. Overall, we identified considerable diversity in the clinical and pathology findings, and a lack of obvious clustering of the data, suggesting that the different histologic patterns do not reflect different causes or different diseases, but instead are variations in the morphology of alveolar and bronchiolar damage and of the ensuing repair attempts. Although many cases in our study had concurrent left-sided heart disease and interstitial lung disease, evidence did not suggest that the interstitial lung lesions were directly caused by left-sided heart failure. Instead, it is presumed that both conditions contributed to respiratory failure and death or euthanasia.

Dogs in our study had combinations of 3 processes that led to lung fibrosis, comparable to those occurring in humans: intraluminal masses of fibrin and/or fibrous tissue within airspaces, hyaline membranes and/or fibrous thickening of alveolar septa, and architectural effacement from confluent scarring.^18,32,36

The first pattern—fibrin or fibrous tissue within airspaces—occurred in 19 of 24 dogs, indicating that this is a common mechanism of interstitial lung disease in dogs. Some dogs had discrete aggregates of fibrin in alveolar and bronchiolar lumens (so-called “fibrin balls”) that typically were partly covered by epithelial cells. This morphology of epithelialized fibrin balls differs from the diffuse intra-alveolar exudate of fibrin seen in bacterial bronchopneumonia; and from eosinophilic hyaline membranes that line alveolar surfaces. Some cases had OP lesions, which formed polyps (typically covered by epithelial cells) of immature or mature fibrous tissue within alveolar or bronchiolar lumens. As for humans, ¹⁹ OP lesions often had a patchy rather than diffuse distribution, and were most often within peribronchiolar (centrilobular) alveoli or alveolar ducts, although small bronchioles and peripheral alveoli were also affected. As in humans with OP,^19,27 some formed clusters of fibroblasts (or myofibroblasts) amid pale-staining matrix (known as fibromyxoid or fibroblast plugs, or Masson bodies), but others had mature fibrosis with eosinophilic strands of collagen. These AFOP and OP lesions of epithelialized fibrin and/or fibrous tissue within airspaces appeared to represent a continuum, with initial exudation of fibrin into alveolar or bronchiolar lumens, migration of alveolar epithelial cells across the surface of the fibrin, subsequent invasion of fibrin by fibroblasts and new blood vessels with loose pale-staining matrix, and eventual organization into fibrous tissue with collagen fibers. This process appears different from “accretion,” described below, because the polyp remains a discrete structure within the alveolar lumen, mostly separated from the alveolar wall. Re-epithelialized masses of intra-alveolar fibrin are recognized in cattle with chronic bronchopneumonia ⁸ ; and for some of our cases, unresolved pneumonia remains a possible basis. Although infectious agents such as respiratory viruses and Mycoplasmopsis bovis can cause obliterative bronchiolitis (bronchiolitis obliterans),^7,39 the clinical progression and polyphasic nature of the lesions in most of these dogs seemed inconsistent with an acute transient infection. In addition, viral infection was not detected in those tested.

The second pattern—hyaline membranes and/or substantial fibrous thickening of alveolar septa—occurred in 17 of 24 cases. Although most hyaline membranes had the typical appearance of thick eosinophilic bands covering the alveolar septa, others formed thin membranes that were separated from the alveolar wall. These membranes could be mistaken for detached pneumocytes but were PAS-positive and cytokeratin-negative. Most dogs in the study had some thickening of alveolar septa by fibrous tissue. Based on ultrastructural studies of diseased human lungs, this interstitial pattern of fibrosis can develop by “accretion” when hyaline membranes or intra-alveolar fibrin are covered by alveolar epithelium and incorporated into the alveolar wall. Alternatively, this pattern can form when alveoli collapse and permanently fuse to adjacent septa. ¹⁸ In addition, fibrosis was noted around blood vessels in alveolar septa and this fibrous tissue appeared to extend into adjacent alveolar septa, implying that collagen synthesis by interstitial or perivascular fibroblasts also contributes to interstitial fibrosis in these cases.

The third pattern—architectural effacement from confluent scarring—was infrequent. This pattern is a feature of usual interstitial pneumonia in humans, in which the affected areas of lung are replaced by scarring so that alveolar lumens and alveolar septa are no longer visible.^12,27 Only 1 case fulfilled these criteria of architectural effacement. In 2 other cases, areas of fibrosis appeared confluent on H&E-stained sections, but lung architecture including alveolar septa remained visible on Masson trichrome-stained sections. Thus, true architectural effacement by confluent scarring must be distinguished from atelectasis; fibrous thickening of alveolar septa; or partial filling of lumens by fibrous tissue, pneumocytes, and leukocytes.

Lung lesions were categorized according to criteria for human cILD. None of the dogs in our study met the criteria for usual interstitial pneumonia in humans (the histologic diagnosis leading to the clinical diagnosis of idiopathic pulmonary fibrosis). Specifically, usual interstitial pneumonia includes not only patchy lesions separated by intervening areas of normal lung (present in 15 of 24 dogs in our study), but also confluent scarring leading to architectural effacement (1 dog in our study); honeycombing defined as clusters of dilated airspaces lined by bronchiolar-like epithelium (1 dog, a WHWT); and fibroblastic foci that indicate temporal heterogeneity and appear as focal areas of immature fibroblasts amid a loose lightly basophilic matrix, typically located at the margins between fibrotic and normal areas of lung (not identified in any of our cases).^12,19,27 Other human interstitial lung diseases that were not identified in these dogs are listed in the Supplemental material.

Instead, most of our cases were classified according to the human criteria as organizing DAD, AFOP, or OP; other cases did not meet these criteria and were considered nonspecific interstitial lung disease. Consistent with the human disease, most cases of organizing DAD had an abrupt onset of respiratory signs. These cases likely represent a subacute or chronic stage of DAD, or a chronic subclinical interstitial lung disease followed by acute exacerbation and formation of hyaline membranes, as is described in human interstitial pneumonia with superimposed acute DAD. ²⁷ The cases diagnosed as nonspecific interstitial lung disease tended to have a longer duration of respiratory disease before death. This suggests that this pattern could represent organizing DAD (in which hyaline membranes have been removed ²⁷ ) or AFOP or OP (in which the intra-alveolar fibrin balls or fibrous polyps have been eventually incorporated into the alveolar wall). ¹⁸

Dogs with organizing DAD, AFOP, OP, or nonspecific interstitial lung disease had few consistent differences with respect to their age, duration of respiratory signs, timing of disease onset, radiographic patterns, or histologic distribution of lesions. In humans, these distinctive histologic patterns are associated with clinical features, such as diagnostic imaging findings, age of onset, expected clinical course, recommended therapy, and expected response to therapy. In this way, clusters of cases within the overall population of human patients with cILD have similar histologic, imaging, and clinical findings, even though these histologic patterns are not used to identify a specific cause. However, such consistent associations were limited in our study population of dogs based on the available clinical information. Our analysis was constrained by the non-standardized clinical treatments and the fact that only postmortem cases were evaluated (animals that recovered after treatment would not be not included). Nonetheless, evidence in our study suggests that the histologic patterns of organizing DAD, AFOP, OP, and nonspecific interstitial lung disease may not be distinct “diseases” of dogs (because they do not identify groups of cases that have similar clinical, imaging, and pathology findings). Instead, independent of the inciting cause, damage to alveolar and bronchiolar epithelium can result in various combinations of hyaline membranes, intra-luminal fibrin balls, and repair responses that ultimately organize into either intraluminal fibrous polyps or fibrous thickening of alveolar septa, with de novo fibroplasia in alveolar septa and around interstitial blood vessels. This interpretation does not diminish the value of identifying these different histologic patterns (i.e., organizing DAD, AFOP, OP, or nonspecific interstitial lung disease), because they allow pathologists to recognize the diagnosis of cILD. Further, they reveal different mechanisms by which fibrosis can develop in the canine lung. It is also plausible that intraluminal vs interstitial lesions could differ in their impact on lung function, clinical progression or course of disease, or response to therapy. As an example, compared with other histologic patterns, cases of OP seemed more likely to develop right-sided heart failure because of their lung disease. Thus, future studies might use these morphologic patterns or novel classification schemes to better understand pathogenesis, inform prognosis, or guide anti-fibrotic treatments.

Clinical or gross pathology findings suggesting cor pulmonale (lung disease resulting in right-sided heart remodeling, with or without right-sided heart failure) were identified in 9 cases and attributed to increased pulmonary vascular resistance. These cases typically had gradual onset of respiratory signs, longer clinical duration, and histologic lesions of attenuated bronchiolar epithelium, panlobular polyphasic lesions, and OP (fibrous polyps in bronchioles and/or alveoli). These findings suggest that OP and/or bronchiolar damage may contribute to the clinically important consequence of pulmonary hypertension and cor pulmonale, ³¹ as is recognized for interstitial fibrosis. ²⁹

Most dogs in our study had thinning of bronchiolar epithelium. This lesion typically results from necrosis of bronchiolar epithelial cells, with migration of surviving cells to cover the denuded bronchiolar wall. Lesions of diffuse alveolar damage in dogs may affect terminal and/or respiratory bronchioles, ⁸ and can be diffuse or regional within the lung.^6,36 In our cases, necrotic bronchiolar epithelial cells rarely were evident. Thus, thinning of bronchiolar epithelium could be an inactive sequela of prior bronchiolar epithelial necrosis. Alternatively, this thinning could represent improper differentiation of bronchiolar epithelial cells, similar to the way fibrosis of alveolar septa interferes with differentiation of type II into type I pneumocytes.^10,17,26,33 Thus, additional studies are needed to determine the mechanism of bronchiolar epithelial attenuation in dogs with cILD.

Further laboratory testing to investigate specific causes was not an objective of our study. However, based on the available data, we considered aspiration, various causes of ARDS, sequela of resolved bacterial lung infection, oxygen toxicity, ventilator-induced lung injury, and left-sided heart failure as potential causes of lung disease in these dogs.

Aspiration was considered based on compatible postmortem lesions and known clinical evidence of aspiration. Aspiration of sterile acidic gastric content results in patchy lesions within the lung that begin in bronchioles and spread into alveoli. With experimentally induced aspiration of sterile acidic gastric content, the acute lesions include hemorrhage, thrombosis, edema, neutrophils, hyaline membranes, and confluent areas of necrosis.^13,14,34 Although lesions often resolve by 4–7 d after a single experimental aspiration event, ³⁴ more severe lung damage might result in fibrotic lesions, as seen in our study, and repeated micro-aspiration events are a proposed cause of cILD (including OP) in humans. ²¹ Thus, severe or ongoing aspiration of gastric content may be one cause of cILD in dogs.^2,5,38

Various causes of ARDS were considered because many cases had an abrupt onset of respiratory signs and histologic lesions of interstitial lung disease. Of the 11 cases with abrupt onset, aspiration was suggested in 5, and 2 also had pancreatitis. Bacterial pneumonia is another cause of ARDS in dogs. Although the postmortem findings did not resemble bronchopneumonia, some lesions in these dogs might represent partial healing following elimination of bacterial infection. Other known causes or risk factors for ARDS in dogs (Suppl. material) were not identified.^2,5,6,30,38

Left-sided heart failure was frequent in our study dogs. Pulmonary interstitial fibrosis and type II pneumocyte hyperplasia are described in dogs with left-sided heart failure^22,24; and, it is plausible that the resulting chronic post-capillary pulmonary hypertension, edema, and congestion could potentially lead to such interstitial changes. However, in one study, these lung lesions did not vary among dogs with different stages of heart failure, ²² suggesting that they were not causally related. In our study, 11 of 12 dogs with left-sided heart failure and interstitial lung disease had attenuation of bronchiolar epithelium, which is not anticipated in left-sided heart failure. Furthermore, cases with and without left-sided heart failure had a similar range of interstitial lung lesions, with various combinations of intraluminal fibrin, hyaline membranes, fibrous polyps in alveoli and bronchioles, type II pneumocyte proliferation, and fibrosis of alveolar septa. These observations suggest that the interstitial lung lesions were not directly caused by left-sided heart failure. Although left-sided heart failure in dogs is often treated and well-controlled for prolonged periods of time, having concurrent interstitial lung disease might impede control of left-sided heart failure and contribute to a fatal outcome. Alternatively, it is possible that acute alveolar and bronchiolar damage might be more likely to persist and result in cILD in those patients with intermittent or poorly controlled left-sided heart failure.

Experimental oxygen toxicity (hyperoxia) results in interstitial, perivascular, and alveolar edema. Less consistently, alveoli contain hemorrhage, hyaline membranes, fibrin, and neutrophils. Later lesions can include type II pneumocyte proliferation and fibrosis of alveolar septa.^3,9,35 Attenuation of bronchiolar epithelium was frequent in our study but is not described in experimental oxygen toxicity in dogs because the alveolar epithelium is considered most sensitive.^3,9,35 Four of our cases received oxygen therapy at least 4 d before death (3 also had mechanical ventilation) and 3 of these had hyaline membranes leading to a diagnosis of organizing DAD. Thus, oxygen therapy or mechanical ventilation might contribute to hyaline membrane formation but did not cause the respiratory disease that was clinically evident before these supportive interventions.

Several recommendations arise from our study. First, in these complex medical cases, the animal’s history and clinical and radiographic findings are essential for interpretation of the lesions. Second, pathologists should consider that the various histologic lung lesions could represent a primary lung disease as well as secondary lesions developing during hospitalization, such as from vomiting/regurgitation and aspiration, systemic inflammatory response syndrome, microvascular thrombosis, primary or secondary bacterial pneumonia, oxygen therapy, or ventilator-induced lung injury. Third, lesions often vary in different regions of the lung, and 1 or 2 histologic lung sections may yield incomplete information in complex cases. Specifically, separately sampling and histologically evaluating the cranial vs caudal lung could be helpful in cases of interstitial lung disease with suspected aspiration or bacterial pneumonia. Finally, special stains for collagen may be necessary to differentiate atelectasis, alveolar septal fibrosis, OP, and confluent fibrosis; and, in select cases, immunohistochemistry can be useful to distinguish hypertrophied macrophages from pneumocytes (e.g., IBA1 and cytokeratin) or to identify bronchioles and vessels within severely diseased lung (e.g. smooth muscle actin).

Our study had several limitations. As a retrospective postmortem study, these cases represent end-stage disease and it is unknown if lesions were present at the onset of disease. Identifying the cause of disease could be limited by inconsistent testing for infectious agents and omissions from the recorded clinical history. The frequency of pulmonary hypertension may be underestimated because of inconsistent clinical testing. Reliance on the non-standardized reporting of gross findings by different pathologists constrains the description of the gross appearance of the lung and the analysis of heart disease and heart failure. Diagnostic imaging data were sometimes absent or incomplete, and, when present, were limited to radiographs and/or thoracic ultrasound but not high-resolution computed tomography. Additional limitations were the non-standardized clinical treatments and the fact that the retrospective postmortem study design would exclude dogs that recovered from lung injury. These, along with the relatively low number of cases, could limit identification of clustering using factor analysis. Future studies with more cases and standardized diagnostic imaging, therapy, and sampling of tissues may be necessary to identify clinically relevant subgroups of dogs with cILD.

We found a diversity of histologic lung lesions in dogs with cILD and classified the lesions using the scheme for human idiopathic interstitial pneumonia. The postmortem lesion patterns and associated clinical findings that we identified suggest that differences in histologic patterns may not reflect different diseases or different causes. Instead, these patterns may represent morphologic variants that result from alveolar and bronchiolar damage and the ensuing repair attempts. Knowledge of cILD in humans has developed over many years, but few studies of the disease are reported in dogs. Additional investigations are needed to verify relationships between pathology patterns and clinical outcomes, identify causes based on additional testing, and report cases of cILD with specific known causes. Along with the current results, such findings will offer insight on the pathogenesis of these chronic lung diseases, mechanisms of the initial damage, and potential methods to delay or predict disease progression.