Published ahead of print on January 26, 2006, doi:10.1164/rccm.200411-1595OC
© 2006 American Thoracic Society doi: 10.1164/rccm.200411-1595OC
Acinar Structure in Symptom-free Adults by Helium-3 Magnetic ResonanceSchool of Physics and Astronomy, University of Nottingham, Nottingham; and Department of Infection, Immunity, and Inflammation (Child Health), University of Leicester, Leicester, United Kingdom Correspondence and requests for reprints should be addressed to Prof. Mike Silverman, M.D., Department of Infection, Immunity, and Inflammation (Child Health), University of Leicester, Leicester, UK LE2 7LX. E-mail: ms70{at}le.ac.uk
Rationale: The apparent diffusion coefficient of hyperpolarized 3He in the lungs has been shown to correlate directly in animal models with the peripheral airspace size and can detect changes in lung microstructure. Objectives: To study in vivo the 3He apparent diffusion coefficient and to demonstrate its sensitivity to changes in lung morphometry as a result of aging, exposure to cigarette smoke, and lung inflation. Methods: We assessed the variation in the diffusion of hyperpolarized 3He gas in the lungs by magnetic resonance techniques. Spirometric lung volumes were recorded. Measurements: We measured the dependence of 3He diffusion on age and on reported cigarette smoke exposure in 32 symptom-free adults. We also measured the dependence of the apparent diffusion coefficient on the degree of lung inflation. Results: In healthy never-smokers, the apparent diffusion coefficient increased with age from 0.115 to 0.155 cm2 · s–1 at 20 and 70 yr, respectively, increased linearly with lung inflation and was independent of individual's lung size after correcting for age. For active and passive smokers, the apparent diffusion coefficient increased by up to 40% compared with never-smokers with mean values significantly higher (p = 0.016 and p = 0.0007, respectively). Conclusions: Peripheral airspace size increases with age and after exposure to smoke in healthy adults in agreement with previous histologic studies. We have confirmed in vivo that peripheral airspace size is independent of intersubject lung size.
Key Words: age • diffusion • hyperpolarized noble gas • smoking Alveolar dimensions and density have mainly been studied using histologic techniques applied to excised or postmortem lung samples, and the only in vivo data have been collected using aerosol washout techniques such as aerosol-derived airway morphometry (ADAM) (1, 2). These studies have yielded much information. It is thought that alveolarization is complete by the age of 2 to 3 yr (3–5), after which alveoli simply enlarge with childhood growth. In healthy adults, the size of alveoli is relatively constant, so that alveolar number is proportional to lung size (2, 5–7). With senescence, however, alveolar dimensions increase (1, 2, 8–10), an effect that is exaggerated in active smokers (11). A simple technique for measuring changes in peripheral airspace size in vivo would have many applications. It could form the basis for longitudinal studies of development and senescence and of the impact of disease, adverse environmental exposures, and therapeutic agents on these processes. It could also provide the structural basis for those measurement techniques that describe peripheral lung function, improving our understanding of structure–function relationships at the alveolar level. As the era of "alveolar therapy" dawns (12, 13), techniques for measuring peripheral airspace size can be used as an outcome measure. The prospect of this form of therapy lends urgency to longitudinal studies of peripheral airspace size to identify the onset of emphysema and the risk factors for its development. Certain stable isotopes of noble gases, such as 129Xe and 3He, when hyperpolarized, can be detected in very low concentrations by magnetic resonance (MR) techniques and can therefore be used as tracers to study gas distribution in the lungs (14). Within any porous structure such as the lungs, the rate of diffusion of a gas is restricted and is known as the apparent diffusion coefficient (ADC). The ADC of 3He can be measured by MR by monitoring the decay of spin-echo signals in the presence of a sensitizing magnetic field gradient. The ADC of 3He gas within the lung is approximately four times smaller than its free diffusion coefficient. ADC has been shown in rat studies to correlate with alveolar internal area and mean linear intercept (15). Modeling work by Yablonskiy and colleagues (16) has demonstrated that ADC can be related to the peripheral airspace size. The 3He imaging technique has also been applied in human subjects to visualize ventilation (17–19) and to assess the degree of emphysema (15). We have developed a system capable of producing hyperpolarized 3He (20, 21) and of measuring 3He diffusion with MR. We used this to investigate changes in peripheral lung morphology in symptom-free adult volunteers resulting from aging and exposure to smoke. Some of the results of this study have previously been reported in the form of abstracts (22–24).
The 32 volunteers (27 males, 5 females) were recruited without preselection from among the staff and students of the University of Nottingham. Their ages ranged from 18 to 74 yr and included 13 individuals who had never smoked, five active smokers (2.5–85 pack-years; median, 10 pack-years), and 14 passive smokers (the latter are defined as having lived with a smoker or suffered significant exposure to smoke or fumes in their work environment). Of these 32 volunteers, 30 were symptom-free and were not using medication for any respiratory diseases. The other two volunteers suffered from asthma. One, a never-smoker, was included because the asthma was very mild and controlled using only intermittent bronchodilator. The second person with asthma regularly used a corticosteroid inhaler and was excluded. Spirometric lung function tests were performed to standard criteria (25) on all volunteers using a calibrated electronic spirometer (Jaeger Flowscreen; Viasys Healthcare, Coventry, UK). At this point, two other subjects who had never smoked were excluded from analysis because they had abnormal lung function (FEV1/FVC < 70%), leaving 11 healthy never-smokers, 13 passive smokers, and five smokers. Hyperpolarized 3He gas was produced via metastable optical pumping (21). For each acquisition, 10 ml of polarized 3He gas was mixed with 300 ml of 4He and transported to the MR scanner in a 1-L Tedlar gas sample bag (SKC Ltd., Blandford Forum, UK). All measurements were made with the volunteers' lungs in the tidal volume range. Each volunteer was asked to inhale the 310-ml bolus of helium gas by mouthpiece at the end of a normal exhalation (FRC), then immediately to take an extra small intake of air (100–150 ml) to flush the helium into the lung periphery, and finally to breath-hold for 3 s. The 3He ADC was measured during the final second of the breath-hold, at a volume of about FRC + 400–450 ml. The MR scanner used was a 0.15-T permanent magnet system (Intermagnetics General Corporation, New York, NY) with an MR Research Systems console (Surrey, UK). A global ADC measurement for both lungs was obtained using a diffusion sequence based on a 64 echo Rapid Acquisition with Refocused Echoes sequence (echo time = 14 ms, acquisition time = 896 ms) with the slice select and phase gradients switched off. This gave a series of 64 echoes with incrementally decreasing amplitude from diffusion weighting. The b-value used was 0.3 s · cm–2. During the time of this measurement, the gas can sample several alveoli and therefore the measurement is sensitive to the peripheral airspace size rather than the size of individual alveoli. A global ADC value for the lungs was calculated by fitting to the exponentially decaying echo train. For all but two volunteers, the echo decay was monoexponential, indicating that the ADC value was approximately the same over the whole lung and therefore that a global ADC value was sufficient to describe helium diffusion in the lung. A weighted mean ADC was used for the two variant subjects. The ADC value given for each volunteer is the mean of at least two separate measurements. To demonstrate that ADC depends on peripheral airspace size, we performed a series of measurements on three healthy never-smokers (aged 27, 37, and 68 yr) in which the ADC was measured as a function of lung inflation. Each volunteer was asked to exhale maximally to residual volume. The 310-ml 3He and 4He gas mixture was then inhaled and followed by an accurately measured volume of air. The measurement was repeated for a range of air volumes in 1-L increments. Further details of the calculation are given in the online supplement. The significance of differences in lung function and ADC between groups of subjects were evaluated using t tests, allowing for different population variances where needed and linear regression. Ethics approval had been obtained from the Medical School Ethics Committee, University of Nottingham. Volunteers gave verbal informed consent.
The three groups of subjects covered similar age ranges (Table 1). There was no significant difference in the mean FEV1 or FVC values, expressed as a percentage of the expected value (26), among the three groups. FEV1/FVC values for never-smokers were significantly higher than for passive smokers (p = 0.031) and active smokers (p = 0.004).
The within-subject coefficient of variation of ADC from repeated measurements within all individuals was 4% by analysis of variance. Mean ADC values for men and women were similar (Table 2). The mean ADC in individuals who had never smoked increased with age (Figure 1). There was no significant correlation between FVC and age in this group (Figure 2). The mean ADC values for passive and active smokers were significantly higher than for never-smokers (p = 0.013 and p = 0.011, respectively; Figure 1). Because of the close dependence of ADC on age of the never-smokers, ADC was adjusted to produce age-normalized ADC values (NADC) for subsequent analysis (see the online supplement). The mean values of NADC for passive and active smokers remained above those for never-smokers (p = 0.0006 and p = 0.016, respectively; Table 2).
The NADC values of the never-smokers showed a nonsignificant negative dependence on FVC (p = 0.091, r2 = 0.28; Figure 2). NADC had no significant dependence on FEV1/FVC. There were two never-smokers with particularly low FEV1/FVC ratios (FEV1/FVC < 70%) who also had high values of NADC (Figure 2). There was a simple linear relationship between change in the volume of gas in the lungs and the corresponding change in ADC, relative to their values at FRC (Figure 3), which allowed us to relate changes in ADC to changes in alveolar wall area per unit volume (AWUV). The conversion procedure from ADC to AWUV is fully described in the online supplement. We then compared our data with histologic data of Gillooly and Lamb (9). We compared our age dependence of AWUV with the histologic measurements. The AWUV calculated from ADC measurements in never-smokers declined with age at a rate that matched that described by Gillooly and Lamb (9) (Figure 4). For comparison, our AWUV data for active and passive smokers are also plotted alongside and are seen to relate closely to the smoker data collected by Gillooly and Lamb (11).
The ADC in healthy never-smokers was determined in this study with a high degree of repeatability and found to increase with age (Figure 1). This finding correlates well with the previous histologic study by Gillooly and Lamb (9) (Figure 4). The histologic data demonstrated that the mean alveolar size increased with aging. We were able to simulate this process by measuring ADC as the lungs were inflated. This is discussed further in BIOLOGICAL ISSUES. The similarity between our data and histologic data is supported by the only other in vivo method, ADAM (1, 2). NADC had only a trivial dependence on lung size (FVC) in never-smokers, implying that for a given age, the peripheral airspace size is the same for all healthy adult never-smokers. Hence, assuming that ADC is a true measure of alveolar size, alveolar number is proportional to adult lung volume as has been previously described in excised lungs (5, 6) and in vivo by ADAM (1, 2). In healthy subjects, FVC is highly correlated with the other "static" lung volumes: total lung capacity, FRC, and residual volume (27). The mean NADC of active and passive smokers was significantly greater than that of never-smokers, although some individual values fell within the normal range. This difference was ascribed to an increase in mean peripheral airspace size and thus a decrease in AWUV. Published histologic data demonstrate a decreased AWUV in smokers (11), in agreement with our findings as shown on Figure 4.
Methodologic Issues No effect of cardiac motion was observed in ADC profiles. A recently published study based on an animal model confirmed a trivial effect of cardiac motion on ADC (29). A monoexponential echo decay in all but two cases indicated that the ADC was isotropic throughout the lungs, justifying our reliance on a single value to describe the mean alveolar dimensions. This was confirmed by examining the variation of ADC across the lung for all our volunteers. Images of 3He ventilation confirmed that the gas was distributed to all areas within the lung.
Biological Issues NADC values were significantly higher in healthy smokers than in never-smokers because of smoking-related damage to the alveolar walls and peripheral airways. This effect has been reported in histologic data from excised lungs of smokers (11). Damage to the intraacinar airways, known as fenestration, can create connections between alveoli and their adjacent terminal airways (respiratory bronchioles and alveolar ducts). This fenestration has also been described in lungs of smokers (30). The fenestrae are much larger both individually and in proportion to alveolar surface area than the interalveolar pores of Kohn. Even without a change in size of airspaces, the presence of such fenestrae will allow more gas diffusion and hence give the impression of larger peripheral airspaces and fewer alveoli per unit lung volume. The similar findings in passive smokers were surprising. Although it may be reasonable to assume that active smoking affects centriacinar lung structure, passive smoking would not be expected to do so to the same extent. However, this study is both small and cross-sectional. It is not possible to estimate how much of the effect on apparent alveolar size was due to a toxic effect of passive smoke inhalation during adult life, from the impact of passive smoking in early childhood on the process of alveolarization itself, or simply an effect of confounding factors that may be associated with passive smoking.
Implications and Future Directions The technique has important potential as an outcome measure for the study of both environmental agents (e.g., tobacco smoke) and pharmaceuticals, including potential therapies for degenerative conditions such as emphysema (12, 13), and drugs that could have a potential adverse impact on early alveolarization, such as corticosteroids. In conclusion, we have used the apparent diffusing capacity of hyperpolarized 3He determined by MR to detect changes in peripheral airspace size with a high degree of repeatability. The results imply that the mean peripheral airspace size increases with age throughout adult life and is larger in active and passive smokers than in healthy never-smokers. This is likely to be due to accumulated alveolar wall damage and the partial merging of alveoli. The age-normalized ADC is almost independent of FVC in never-smokers, confirming in vivo that peripheral airspace size is the same for adult healthy never-smokers of a given age. Hence, assuming that, in healthy subjects, ADC is a true measure of alveolar size, alveolar number is almost directly proportional to lung size. The technique opens up the possibility of studying alveolar development and senescence and the effects of disease and therapy in living people.
The authors thank all the staff and students from the School of Physics and Astronomy who agreed to take part in this study and Dawn Jotham for lung function measurements.
Supported by the following organizations: the EPSRC provided funding for a Ph.D. studentship, the Wellcome Trust provided funds for the design and construction of the 3He gas polarizer, and the U.K. Medical Research Council provided funds to purchase the magnetic resonance scanner. This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org Conflict of Interest Statement: None of the authors have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Received in original form November 29, 2004; accepted in final form January 26, 2006
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
METHODS
