Whether emphysema is centrilobular or panlobular determines pulmonary artery remodelling in COPD

Extract

Emphysema, along with small airway diseases, is a crucial component of COPD. Although the type of emphysema smokers develop is not usually considered, two emphysema types with distinct morphological and physiological characteristics might be present in COPD: centrilobular emphysema (CLE), in which the centre of the acinus-respiratory bronchiole and alveolar duct are destroyed with preservation of alveoli, and panlobular emphysema (PLE), seen with α-1 antitrypsin deficiency, but also in smokers with normal levels of the protein.

Shareable abstract

Centrilobular emphysema has significantly worse remodelling of pulmonary arteries than panlobular emphysema, suggesting distinct pathological phenotypes in COPD, a concept that might improve the understanding of the disease https://bit.ly/4oZtm09

To the Editor:

Emphysema, along with small airway diseases, is a crucial component of COPD. Although the type of emphysema smokers develop is not usually considered, two emphysema types with distinct morphological and physiological characteristics might be present in COPD: centrilobular emphysema (CLE), in which the centre of the acinus-respiratory bronchiole and alveolar duct are destroyed with preservation of alveoli, and panlobular emphysema (PLE), seen with α-1 antitrypsin deficiency, but also in smokers with normal levels of the protein. In PLE the entire acinus-respiratory bronchioles and alveoli are destroyed [1].

Pathological remodelling of the pulmonary vascular tree in COPD has been well-described [2–5], however the possible differences in the vascular abnormalities in the two very distinct types of emphysema have not been investigated. To determine if the type of emphysema could affect the vascular tree differently, we assessed the structural abnormalities of pulmonary arterioles and small airways in surgically excised lungs of 27 smokers with CLE, 21 with PLE with normal α-1 antitrypsin levels, seven control smokers and eight nonsmokers, undergoing lobectomy or pneumonectomy for lung cancer. Lungs were classified as having PLE, CLE, or no emphysema using macroscopic and microscopic criteria [6, 7].

Pulmonary function tests were performed before surgery. Excised lungs or lobes were fixed in formalin by intrabronchial infusion at a pressure of 25 cm H₂O for 24 h [8]. Two to five tissue blocks per patient were collected using a template, and 5 μm sections were cut and stained for lung, airways, and pulmonary arteries morphometry as previously described [2, 4, 9]. The mean number of arteries examined per patient was 7±5 in CLE, 5±2 in PLE, 6±2 in smokers and 8±3 in nonsmokers. Lung samples belonged to an historical cohort, recruited prior to the latest American Thoracic Society/European Respiratory Society standards [8], therefore sampling procedures do not align completely to them; this and the number of arteries may be limitations of our study. Arteries with a perimeter along the internal elastic lamina <1.5 mm (diameter 0.5 mm), and a double elastic lamina visible for at least half the circumference were selected. Total wall area (between lumen and outer border), intimal area (from lumen to internal elastic lamina), medial area (between internal and external elastic lamina) and adventitial area (from external elastic lamina and outer border) were measured and normalised by perimeter (figure 1a).

FIGURE 1.

a) Representative photomicrographs of pulmonary arteries in CLE and PLE showing the method of measurement of intimal (I), medial (M) and adventitial (A) layers using a computer linked digitiser. b) Total, intimal, medial and adventitial areas normalised by perimeter in centrilobular emphysema (CLE), panlobular emphysema (PLE), smokers (S) and non-smokers (NS). In each box-plot, the solid line represents the median; the bottom and top the 25th and 75th percentiles; and the whiskers the minimum and maximum. p-values in the figure represent the results of Mann–Whitney U-tests. Kruskal–Wallis test: p<0.001 for total, intimal and medial area; p=0.02 for adventitial area.

Interalveolar wall distances in lung parenchyma were measured in 20 fields per slide with a single cross hairline and were expressed as the mean linear intercept (Lm) for the lung. Section area was compared with template area for correction for tissue shrinkage [6]. Protocols were approved by the institutional ethical committees of the Royal Victoria Hospital in Montreal (Canada) and all subjects provided written informed consent.

Subjects with emphysema had similar age (CLE 67±9; PLE 63±10 years) and smoking history (CLE 47±35; PLE 59±33 pack-years). Forced expiratory volume in 1 s (FEV₁)% and FEV₁/forced vital capacity (FVC) were similar in CLE and PLE (78±23% pred, 67±13% versus 76±28% pred, 63±16% respectively) but lower than in smokers and non-smokers (p<0.01 for all). Global Initiative for Chronic Obstructive Lung Disease (GOLD) stages were similarly distributed between CLE and PLE. Most patients presented with GOLD 0–2, while two CLE and three PLE cases displayed GOLD 3.

The extent of emphysema, determined by Lm, was similarly increased in CLE (median (range): -302 (201–553) μm) and PLE (286 (230–433) μm) compared to smokers (213: (176–350 μm) p<0.01 for both) and non-smokers (222:, (181–254) μm, p<0.001 for both) which minimises the possible effect of the extent of emphysema on the different remodelling of pulmonary arteries. The coefficient of variation of interalveolar wall distances was larger in CLE (0.36 (0.25–0.59)) than in PLE (0.31 (0.21–0.64) p=0.048) reflecting the more uneven pattern of destruction in CLE than in PLE [6].

When the analysis was performed by pooling all single arterial values in each emphysema and control group, the total wall area normalised by perimeter was significantly larger in CLE when compared to the other three groups (PLE, smokers and non-smokers). In CLE, all arterial layers (intima, media and adventitia) were increased compared to PLE (figure 1b). Of interest the intimal layer in smokers was significantly thicker than in PLE and non-smokers. The increased vascular remodelling in CLE compared to PLE was confirmed even when the analysis was performed using the average value in each patient (p<0.05 for total, intima, media and adventitia). In six cases in which the whole lung was available (four PLE, two CLE) no significant differences in arterial remodelling between upper and lower lobes were found.

The degree of emphysema determined by Lm did not correlate to artery remodelling in either emphysema; however, we found a relation between arterial remodelling and COPD severity, as defined by GOLD stage. In CLE, but not in PLE, total wall area increased as the severity of the disease worsened (GOLD 0–1; 62 (29–136) μm versus GOLD 2–3; 71(20–175) μm; p=0.026). Moreover, in CLE GOLD 2–3 total wall area was larger than in PLE GOLD 2–3 (48 (21–144) μm; p<0.001). Importantly, the degree of emphysema determined by Lm was correlated to the FEV₁/FVC%% in PLE (r= −0.56 p=0.029) but not in CLE (r= −0.32 p=0.1).

The investigation of small airways (<2 mm diameter) revealed that, similarly to vascular remodelling, the small airways total wall area normalised by perimeter was larger in CLE compared to PLE (71.1 (26.8–546.15) μm versus 61.1 (16.7–121.2) μm; p=0.0060) and compared to non-smokers (60.8 (29.2–145.6)  μm; p=0.03), but not compared to control smokers (70.5 (34.8–169.5) μm).

Remodelling of small pulmonary arteries and their relation to abnormalities in small airways and emphysema in smokers have been well described [3, 9–12]. We now demonstrate that the magnitude of vascular remodelling is influenced by the type of emphysema (CLE or PLE) smokers may develop. In our study, the smoking subjects with emphysema had similar age, degree of airflow obstruction and extent of emphysema, yet vascular abnormalities comprising total wall, intimal, medial, and adventitial layer areas were significantly more prominent in CLE than in PLE. Of interest, these findings parallel the remodelling of small airways in our population which was significantly more prominent in CLE than in PLE, a finding that emphasises the importance of the emphysema type in the development of vascular and airway pathology in COPD.

The different behaviour of the vasculature and airways in CLE and PLE is most likely based on the pattern and site of the initial lung destruction. In CLE, lung destruction starts in the centre of the acini where the smallest bronchi and arteries arrive in tandem. An intense inflammatory reaction in these small airways [7], inflammation that has also been described in the pulmonary arteries [5], would remodel and eventually destroy airways, and likely arteries in CLE as shown by Hogg′s group [13, 14]. In PLE, where emphysema is homogeneously distributed throughout the acinus and affects the size of all alveoli and alveolar ducts, the loss of airways is smaller than in CLE [13] and small airways tend to be less inflamed [15]. While tissue destruction due to emphysema is one mechanism of vascular impairment in smokers, multiple other pathways including those modulated by inflammatory mediators, vessel remodelling, hypoxic vasoconstriction and compression due to air trapping have been described as leading to pulmonary vascular disease. Intimal layer abnormalities, already seen in smokers without COPD, are believed to be secondary to epithelial injury provoked by tobacco smoke and not due to hypoxia [3]. Contrarily, hypoxia due to inhomogeneous small airway abnormalities, that creates acinar gas exchange abnormalities, would promote proliferation of smooth muscle in the medial layer [3, 5], as seen in CLE but not in PLE.

Arterial remodelling was related to disease severity, as defined by the GOLD stage in CLE, in which small airways play an essential role to the loss of function, but not to the extent of emphysema. Conversely, in PLE, disease severity did not affect vessel remodelling. The degree of airflow obstruction was influenced by the extent of emphysema in PLE but not in CLE, underlying the different pathological mechanisms in lung destruction in the two emphysemas. All these findings emphasise the important differences between CLE and PLE and set the base for distinct pathological phenotypes in COPD to be considered.