| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1Centre for Child Health Research, University of Western Australia, Perth, Western Australia, Australia; and 2Unit of Developmental Vascular Biology, Institute of Child Health London, United Kingdom
Submitted 7 May 2004 ; accepted in final form 17 September 2004
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
lung function; morphometry; collagen; elastin; airway; rabbit
The majority of studies exploring the effects of exogenous glucocorticoids on alveolar development have been conducted using animal models, namely the rat, ferret, sheep, and primate. Most studies in sheep and primates have focused on glucocorticoid exposure during the antenatal period and not postnatal. Ellington and coworkers (7) found that injected steroids during the postnatal period retarded body weight and total lung capacity and that peripheral airways were less sensitive than central airways to steroid treatment. Massaro and colleagues (15, 16) reported that exogenous administration of injected glucocorticoids to rats during the period of septation (days 4–13) almost completely abolished the development of secondary septa and alveoli. The effects of exogenous glucocorticoids appear to be time specific, with excess glucocorticoids at critical times in development precociously inducing septal thinning and microvascular maturation (28), occurring via halting of proliferation and enhancement of differentiation (28). Further studies (2) have confirmed that this deleterious effect is long lasting, with alveolar surface area 25% lower, 2.4 times fewer alveoli, and alveolar volume 2.2 times greater than controls in 60-day-old rats treated with dexamethasone (Dex) between days 4 and 13 (2). In humans, the process of alveolar development starts antenatally and continues after birth (4, 6, 11, 25–27); therefore postnatal steroids are given on a lung that is more mature than that of the newborn rat. Our laboratory has previously published data confirming the rabbit as a suitable model for studying the effects of steroids on postnatal lung growth and development because alveoli develop in part before birth and continue to develop after birth (13).
Glucocorticoids are known to inhibit lysyl oxidase activity, resulting in diminished cross-linking of the collagen and elastin fiber network, thus altering the structural integrity of the lung (1). Ramchandani and colleagues (20) found that immature rabbits (4 wk old) had a greater proportion of smooth muscle and a lower proportion of cartilage compared with mature animals (6 mo). There is an absolute increase in cartilage in older animals. It is possible that an increase in the proportion of cartilage occurs via either a laying down of new cartilage or the loss or thinning of smooth muscle. Either of these processes has the potential to be affected by inhaled glucocorticoid treatment. In addition, studies have found that connective tissue is synthesized in the airways for a longer period of time than in the parenchyma (3). There is thus the potential that inhaled glucocorticoid treatment in the early postnatal period may have differential effects on airways and lung parenchyma. At present there are no data in the literature reporting the effects of inhaled glucocorticoids on postnatal lung growth and development, and, given that they are readily prescribed in clinical practice to young infants and children, it is important to establish what effects they are having on the developing lung. Therefore, the aim of the present study was to investigate the effects of treatment with inhaled glucocorticoids on early postnatal lung development of rabbit using lung function studies together with postmortem morphometric analysis of the alveoli and airways.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Lung function. After the 4-wk treatment period, animals were weighed and anesthetized. Local anesthetic (lidocaine, 1% in saline) was injected subcutaneously along the midline of the neck and chest before incision. Tracheotomy was performed with a plastic neonatal cannula (3 mm internal diameter, 3.5 cm long). Animals were mechanically ventilated (Harvard rodent ventilator model 55-3438, South Natick, MA) with a tidal volume of 10 ml/kg at a frequency of 40 breaths/min and positive end-expiratory pressure set to 2 cmH2O. Animals were allowed to stabilize on the ventilator, and volume history was standardized by inflating the lungs to double the tidal volume (a sigh) before lung function measurements began. Pulmonary function was measured by the low-frequency forced oscillation technique (10). Respiratory input impedance was measured within a frequency range of 0.25–19.625 Hz. Five measurements of input impedance were made at 1-min intervals at functional residual capacity (FRC) set by end-expiratory pressure of 2 hPa. These were followed by volume-dependence measurements in which the lungs were inflated to 0, 2, 5, 10, 15, and 20 hPa, and the signal was again applied (23). Four measurements were made 1 min apart at each pressure. The parameters obtained were averaged at each lung volume. Lung function data were also volume corrected by multiplying airway resistance or dividing tissue damping (G) and elastance (H) with the lung volume at which the measurement was made, as calculated from quasi-static pressure-volume (P-V) curves.
At the completion of data collection, animals were overdosed with pentobarbitone sodium (5 ml/kg, 325 mg/ml) and manual quasi-static P-V curves were performed in situ before death.
Morphometry. Lungs were removed, and wet lung weight was recorded. The right lung was inflation fixed at 20 cmH2O and immersion fixed overnight. Volume displacement was used to measure fixed volumes of the whole right lung and each lobe separately. Each lobe of the right lung was cut into 3- to 5-mm segments in the sagittal plane in the medial-to-lateral direction. A number was assigned to each slice, and one sample was chosen by using a random number generator.
Tissues were embedded in paraffin, sectioned at 5 µm, and stained with hematoxylin and eosin. All morphometric measurements were performed blind (J. Kovar) to treatment group. Volume fractions of lung parenchyma (alveoli and alveolar ducts), nonparenchyma (conducting airways and blood vessels), and pleura were estimated from photographic enlargements x23 of 5-µm sections by superimposing a cycloid point-counting grid (74 lines/148 points). A SPOT RT color camera model no. 2.2.0 interfaced with a Leica DMLS microscope and an IBM-compatible computer was used to capture 10 random nonoverlapping images (x230 magnification) from each section and lobe.
A linear point-counting grid (32 lines/64 points) was superimposed onto the images by an image analysis program (Image-Pro Plus Version 4.1). The number of points falling on alveoli, ducts, or septal tissue and the number of air/tissue tissue-air intercepts were counted. Alveolar, alveolar duct, and tissue fractions were calculated. The number of alveoli per area was directly counted for each image by using an unbiased sampling frame (area = 0.7671 mm2). Morphometric calculations were performed as described previously (13).
Airway measurements. Airway wall area was measured on 5-µm sections of tissue by planimetry (12). Transverse airways shaped close to that of a circle and defined as those with an epithelial layer of even thickness were measured by tracing around the inner airway wall perimeter (Pi) and the outer airway wall perimeter (Po). Airway wall area was calculated by subtracting the inner wall area from the outer wall area. The number of alveolar attachments to each airway was counted, and the distance between each attachment was calculated by dividing the Po by the number of attachments (12). The length of the Po that was shared with an adjacent pulmonary blood vessel was also measured and used to correct the Po by subtraction.
Total airway wall area was standardized by dividing by Pi2. A total of 738 airways were counted (see Table 4), and analyses were performed on four airway size brackets defined by the Pi length. Size categories examined were Pi < 0.75, Pi = 0.75–0.85, Pi = 0.85–1, and Pi = 1–4.5 mm. A comparison between treatment groups for each of the airway Pi categories indicated that airways of a similar size were being compared.
|
Tissue sections from the right cranial, middle, and caudal lobes of six randomly selected animals per treatment group were analyzed for elastin content in lung parenchyma. Sections were stained with Miller's elastin stain, and 10 random, nonoverlapping images of lung parenchyma were captured (x1,200 magnification), and area of elastin quantified by color thresholding performed according to Pierce et al. (18). The amount of elastin was expressed as a percentage of total area of lung parenchyma for each image, and a mean value for the lobe and the lung was calculated.
Arterial measurements.
Arteries 15–100 µm in diameter from five randomly selected animals/group were measured (
300 arteries/group). The external diameter and wall thickness were measured, and percentage wall thickness was calculated as 2 x wall thickness divided by external diameter x 100. For each group, the mean diameter of arteries accompanying respiratory bronchioli and alveolar ducts was calculated.
Statistical analyses. Differences between groups were analyzed by Student's t-tests, one-way ANOVA, repeated-measures two-way ANOVA, or Kruskal-Wallis one-way ANOVA on ranks, as appropriate. Post hoc analyses were performed by using pairwise multiple-comparison Tukey's test or Dunn's method of multiple-comparison procedure as appropriate.
A random-effects regression analysis on log10-transformed data for alveolar number, volume, surface area, and wall thickness adjusted for (log10) lobe volume, sex, and lobe was run to test the power and sensitivity of the data set. The numbers of animals used in the present study provided sufficient power (>0.8) to detect a 12% change in alveolar surface area, a 30% change in alveolar number, a 35% change in alveolar volume, and a 16% change in alveolar wall thickness between treated and control animals.
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
|
|
|
|
22, 48, and 30%, respectively). No significant differences were seen between the three lobes making up the right lung for any parenchymal fraction (proportion of tissue, alveolar duct, or alveolar space). Therefore the values for the total right lung are given (Table 3). In addition, values standardized by the total lung volume have been calculated. The total number of alveoli was reduced and thickness of the alveolar wall increased in the Dex-treated group when compared with control (P < 0.05); however, after correction for volume, there was no significant difference in the total number of alveoli. Alveolar number was not decreased in the Bud group. The total surface area of alveoli was not significantly reduced in Bud and Dex compared with control animals when area was standardized to lung volume.
|
Effects of glucocorticoid treatment on the developing airways. There were no statistical differences in Pi between the three treatment groups, confirming that airways of similar size were compared for both airway size and collagen measurements. Steroid treatment (either inhaled or injected) reduced both airway wall area and the number of alveolar attachments and increased the distance between them in the smallest airways examined (Pi < 0.75 mm) (Table 4). Bud also reduced airway wall area and increased the distance between attachment points to airways with a Pi = 0.75–0.85 mm.
Collagen was only visible in 40% of transverse peripheral airways, and this was consistent across all three groups. Although the percentage of collagen in the Dex group (4% Pi < 1 mm; 5.85% Pi = 1–4 mm) was less than that measured in Bud (5.6% Pi < 1 mm; 9.4% Pi = 1–4 mm) and control (5.9% Pi < 1 mm; 8% Pi = 1–4 mm) animals, this difference was not significant.
Arterial measurements. In all groups, as arteries increased in size there was a decrease in percentage wall thickness (Table 5). There was no difference in the percentage wall thickness between controls and Dex-injected animals in any size range (Table 5). But in the Bud group in two size ranges (35–48 and 51–64 µm), the arteries had a reduced percent wall thickness compared with controls in all size ranges, and there was a significant increase in size of the arteries accompanying both respiratory bronchioli and alveolar ducts.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
In the present study, nebulized Bud resulted in a large amount of steroid deposited in the central airways in a patchy fashion with a sparse but even deposition throughout the small airways, ducts, and alveoli (data not shown). A small percentage of Bud deposited in the lung was absorbed into the systemic circulation; however, the dose was much lower than that of injected Dex (as evidenced by a slight reduction in endogenous cortisol observed in endogenous cortisol observed in nebulized Bud-treated animals and greater reduction in the Dex group; data not shown). The average wet lung weight of Bud-treated animals did not differ significantly from controls. However, when lung weight was standardized for body weight, there was a significantly lower ratio (12%) in the nebulized Bud-treated group compared with controls.
Systemic administration of potent glucocorticoids, including Dex, can cause body growth retardation (7, 9, 19, 21 24, 30), as well as growth retardation of the lungs (7, 19, 21, 24). It was therefore not surprising that wet lung weight in the present study was significantly smaller (30%) than in controls; however, because the mean body weight was also reduced in these animals, lung weight-to-body weight ratio in this treatment group was not different from controls, indicating a generalized but not lung-specific growth retardation.
Lung function measurements were made to determine whether steroid treatment altered the mechanical properties of the lung. There were no significant differences observed between the control group and Bud group in any of the lung mechanical properties examined, suggesting that treatment with inhaled glucocorticoids in this model did not have an impact on the mechanical properties of the lung at 2 hPa. Systemic increases in G and H, corrected for lung volume at which the measurements were made, were seen in the Dex group compared with both controls and Bud-treated animals, confirming the potential for steroids to influence lung function. These results suggest that the amount of inhaled steroids reaching the lungs was not sufficient to affect lung function.
No differences in the volume dependence of lung mechanics were observed between treatment groups. Quasi-static P-V curves showed that Bud-treated animals had more compliant lungs at volumes approaching TLC (25 hPa) than both controls and Dex-treated animals. This difference in lung compliance, however, was eliminated when data were standardized for TLC.
There was a decrease in the number of alveoli and a trend toward a decrease in the total alveoli surface area in the right lung (P = 0.056) in Dex-treated animals compared with controls; however, this could be attributed to the fact that these animals have smaller lungs, because this difference was no longer present when data were standardized for lung volume. Our data suggest that treatment with nebulized Bud retards lung growth and results in a reduced lung size per body weight; however, the parenchymal components, i.e., the number of alveoli, surface area per volume of lung, alveolar volume, and wall thickness, remain not substantially altered. Thus, during the postnatal period, the lung parenchyma is developing normally by structure and maturity; however, the lung itself is smaller than expected. This finding highlights that inhaled steroids have a discrete and localized effect on the growth of the lungs, which is different from that of systemically delivered steroids.
It is important to note that peripheral rabbit airways are thin walled for their internal perimeter and do not contain all the classic airway layers found in human airways. Thus it was only possible to measure the internal and external perimeters and to calculate the total airway wall area. A significant trend was noted across all groups that, with increasing peripheral airway size, the number of alveolar attachments increased, as did the distance between them, and within the airway size range examined, Pi = 0–4.5 mm, relative wall area decreased with increasing airway size. These trends appear intrinsic to peripheral rabbit airways.
Treatment with nebulized Bud resulted in reduced airway wall area for small peripheral airways with a Pi < 0.85 mm, and Dex reduced airway wall area in airways with Pi < 0.75 mm. This finding of airway wall thinning may be explained by apoptotic loss of airway epithelium associated with glucocorticoid treatment. Treatment of normal cultured primary central airway epithelial cells and the human airway epithelial cell line 1HAEo– with Dex or Bud has been shown to induce time- and dose-dependent apoptosis (5), via a disruption of mitochondrial polarity, with subsequent downstream involvement of caspase proteases (5). This study also found that Dex was more potent than Bud in inducing apoptosis. Equating the concentration of glucocorticoids administered to cells to that which would be delivered to the central airways in vivo is difficult. The range of concentrations tested was based on previous drug-distribution studies of Bud in the lung, which suggest that airway mucosal and tissue concentrations are at least in the nanomolar range.
When examining the effects of Dex treatment on levels of apoptosis in the postnatal rat lung, Luyet and colleagues (14) noted two peaks of apoptotic activity in normal lungs, the first directly after birth and the second toward the end of the third week. Dexamethasone was administered in decreasing doses from postnatal days 1–4 and appeared to suppress the second peak of apoptosis to levels just above baseline, and on days 3–4 there was a significant decrease in apoptosis. Schittny and colleagues (22) found that Dex treatment affected the relative number of dying cells but could not detect any significant difference in the type of cells dying (i.e., fibroblasts or type II epithelial cells). In agreement with the above study, they also found a peak in apoptosis on day 4 and a Dex-induced decrease in cell proliferation on days 2–4. They found no significant differences between treated animals and controls on days 10–36.
In the present study, the number of alveolar attachments was significantly reduced in both the Bud- and Dex-treated animals; however, this was only the case for airways with a Pi < 0.75 mm. Similarly, a recent study by Fayon and colleagues (9) reported that the number of alveolar attachments to bronchioles per millimeter of the peribronchial sheath (diameter < 0.4 mm) was reduced by one-third in Dex-treated rat pups (0.4 µg subcutaneous daily between days 4 and 14) compared with controls and hydrocortisone-treated animals. These findings also illustrate the varied effects of different types of glucocorticoids, and it is likely that this variability is dependent on the drug delivery method, dose, and potency.
Measuring the number of alveolar attachments, and the distance between them, has been used in human autopsy studies as a surrogate for alveolar number because tissue samples are often uninflated and whole lung volume may not be measured, thus making it impossible to calculate total alveolar number by the methods of Weibel (29). It is interesting to note that in the present study, although there was no difference in specific total alveolar number with treatment, the number of alveolar attachments was decreased. This discrepancy is likely to be due to the methodological differences used to calculate the variable. Alveolar attachments are counted in a given area of lung (around the outer perimeter of an airway), whereas total alveolar number is measured from a random sample and calculated by using lung volume. The two measurements thus yield different information, with alveolar attachment data giving information about the elastic load imposed on an airway rather than being a simple proxy for total alveolar number.
The data presented here suggest that both inhaled and systemic glucocorticoids cause abnormal wall thinning in small peripheral airways, decrease the number of alveolar attachment points, and increase the distance between them. The finding that steroids, both inhaled and systemic, had greater effects on small airways suggests that they are likely to be more sensitive to steroid effects. However, in the absence of any direct data on relative concentrations of steroids in the different airways and with the different delivery methods, this suggestion remains speculative.
The fact that no changes in lung mechanics were observed is not surprising, given the lack of effect of inhaled glucocorticoids on lung structure when expressed per unit volume. It would not be expected that airway wall thinning in peripheral airways would change airway caliber and thus lung function. In support of the lack of inhaled glucocorticoid effect on lung function were the findings that airway collagen content and parenchymal elastin content did not vary with treatment.
Despite some increase in muscle of the peripheral arteries of the Bud group, our results suggest that there is little direct steroid effect on pulmonary arteries. Interpretation of the importance of these changes in the rabbit model is difficult because the arterial wall is not the same as in humans. The hilar arteries are relatively thin walled, with several layers of muscle cells separated by elastic laminae. In the midlung region, the wall becomes very thick and the muscle cells are not oriented circularly around the lumen. The lumen at this time is relatively small. An area like this has been described in rat pulmonary arteries as the thin-walled oblique muscular artery (17). Beyond this area are thin-walled arteries with one, rarely two layers of muscle between elastic laminae and many only have part of the wall with muscle cells. These arteries are found with the most distal airways and in the alveolar region. They probably add little to pulmonary artery resistance because they are beyond the thick muscular area. These species differences thus caution against concluding that steroids have no effects on pulmonary arteries in the developing human lung.
The results of the present study do have potential implications for the use of inhaled steroids in young children. Although we have not shown any major effects of inhaled Bud on alveolarization or lung function, we did show adverse effects on lung growth, small airways, and alveolar attachments. In addition, systemic steroids resulted in a reduction in alveolar number, confirming the potential for high doses of steroids to adversely effect the developing lung. These findings show that the concerns held by pediatricians and parents regarding excessive use of inhaled steroids, especially in high doses in young children, are rational and stress the importance of using the lowest possible effective dose.
In summary, the main findings of the present study are that the systemic glucocorticoids lead to general growth retardation including the lung, where the total alveolar number and surface area are reduced. By contrast, inhaled glucocorticoids cause growth retardation of the lung. Structurally, the development of alveoli during the 1- to 5-wk age period was not affected by inhaled glucocorticoids but was reduced by systemic steroids. Lung volume and peripheral airway wall thickness were reduced by inhaled steroids, and the number of alveolar attachments to the airways was reduced with the distance between attachments points increased.
From a clinical perspective, the findings of this study are on the one hand reassuring: inhaled glucocorticoids do not affect alveolarization during the postnatal period in rabbits; however, on the other hand, they are disturbing given that the lung size of these animals is decreased, as was peripheral airway wall area and the number of alveolar attachments to airways. Clearly, inhaled glucocorticoids are an essential part of asthma management, with undertreated asthma having adverse effects on health and growth in children (8). The use of these potent drugs in young infants and children should therefore be monitored with caution, and only the lowest doses possible that yield significant clinical improvement should be used.
|
GRANTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. Bush and I. Annesi-Maesano "Beam Me Up, Scotty!" Am. J. Respir. Crit. Care Med., January 1, 2007; 175(1): 1 - 2. [Full Text] [PDF]
|
|
|
|
 |
|
 K. Huang, R. Rabold, B. Schofield, W. Mitzner, and C. G. Tankersley Age-dependent changes of airway and lung parenchyma in C57BL/6J mice J Appl Physiol, January 1, 2007; 102(1): 200 - 206. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. J. F. M. Merkus and J. C. de Jongste Inhaled Corticosteroids in Wheezy Infants Am. J. Respir. Crit. Care Med., October 15, 2005; 172(8): 1058 - 1059. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
) and resistance (Raw). Volume dependence of hysteresivity (G/H) (A) and resistance (B) was measured and calculated. Data points represent means ± SD; C, n = 15; Bud, n = 5; Dex, n = 5.