Hyperpolarized 3He diffusion MRI and histology in pulmonary emphysema

Chronic obstructive pulmonary disease (COPD) is predicted to move from 12th to fifth place in the conditions that contribute to the global burden of disease by the year 2020 (1). Its defining feature is irreversible airflow limitation (2,3), which is estimated by the volume of air that can be forcibly expired from the lung in 1 s (FEV1), and the ratio of this value to the forced vital capacity (FVC), the maximum volume that can forcibly expired with no time limit. These measurements reflect the increased time required to empty the lung due either to an increase in the resistance of the small conducting airways (4-6) and/or an increase in lung compliance and early airway closure due to emphysematous destruction of the lung’s elastic recoil (7). Emphysema can be diagnosed using computed tomography (CT) scanning, and several studies have shown an association between CT measurements (8) and histology (9,10). However, the precise separation of fully expanded normal lung from mild emphysema using CT remains a subject of debate (25). In addition, the radiation dose required for the repeat measurements needed to monitor disease progression in clinical trials is excessive. In order to monitor regional emphysematous progression for longitudinal studies, and develop future pharmaceutical agents to slow or stop progression, a non-ionizing imaging standard of the severity of emphysema must be established. Since magnetic resonance imaging (MRI) of the restricted diffusion of hyperpolarized gases shows much promise for becoming such a standard, and since whole-lung sets of images can be acquired during a single 10-s breath-hold in vivo, we focused on this technique in the present study. For 3He at or near the dilute limit in air or N2, the free (unrestricted) diffusivity is 0.88 cm2/s. At a typical value of diffusion time Δ of 2 ms, the root mean square (RMS) free displacement (2DoΔ) is 0.59 mm. A typical mean alveolar “diameter” is near 0.3 mm, and the average major and minor diameters of acinar ducts are around 0.7 mm and 0.3 mm, respectively (30). Normal acinar spaces restrict the apparent diffusion coefficient (D, or ADC) from its unrestricted value of 0.88 cm2/s to 0.2 cm2/s, so the diffusion is substantially restricted. Expansion of the alveoli and tissue destruction associated with emphysema result in fewer gas-atom collisions with alveolar walls during typical experimental diffusion times (around 2 ms). Thus, the expanded alveolar spaces increase the ADC by up to four times (21). Emphysema is defined by a larger than normal expansion of lung tissue with associated alveolar destruction (13). Duguid et al. (14) were among the first to try to estimate the severity of this destruction by measuring the average distance between alveolar walls and calculating the total alveolar surface area of the lung. Thurlbeck (15,16) subsequently made an extensive study of the alveolar surface area of postmortem lungs and concluded that there was too much variance in the measurement of total alveolar surface area for it to be useful in comparing the amount of emphysema present in different subjects. Normalizing the surface area to lung volume ratio (SA/V) appears to be better suited for quantifying the severity of emphysema (17). Recent studies of an animal model (18) and human disease (17) showed that the SA/V decreases in mild forms of emphysema, whereas the total surface area itself only becomes measurably decreased when emphysema becomes more severe. Because 3He MRI is sensitive to changes in geometry within the acinus, it may offer a very sensitive means of separating normal from emphysematous lung in patients with COPD in vivo. The present study was designed to correlate the restricted 3He diffusivity with histologic measurements of SA/V and the mean linear intercept (Lm, the average distance between alveolar walls). Since histological sampling is only feasible with the ex vivo lung, we used lungs that had been removed from patients at lung transplantation and histologically-normal donor lungs that were not used for transplantation. Most of the previously reported measurements of 3He diffusivity were obtained in vivo in humans and large animals (12,19-22). In this study we took advantage of the opportunity to study explanted human lungs from patients with advanced COPD to allow a direct comparison with histology of human disease. In addition, studying the ex vivo lung offered a few technical advantages. First, the absence of the chest wall removed most of the radiofrequency (RF) loss associated with saline and tissue, and permitted the use of a sensitive, high-performance coil for a high signal-to-noise ratio (SNR). Second, oxygen, which depolarizes 3He nuclear spins, was replaced with nitrogen in the explanted lung and the lung was maintained at a near-constant volume, which allowed multiple imaging experiments to be performed for hundreds of seconds. Crucially, the ability to inflate the entire lung and rapidly freeze it in the position of study allowed us to make a direct regional comparison of the diffusion measurements with histology. Although in this study we used the ex vivo lung for the above practical reasons, in vivo 3He diffusion MRI has been employed for several years in healthy subjects and patients with severe disease (11-12,19-21).