Introduction

Hyperpolarized (HP) gas apparent diffusion coefficients (ADC) and morphometry have been demonstrated for the microstructural characterization of various lung diseases, both preclinically and clinically.¹ Several studies^2,3,4 have used ADC imaging of HP ³He and ¹²⁹Xe in rats and humans to infer microstructural changes due to different ventilation pressure schemes. This tool could be useful for the early detection and monitoring of treatment of diseases affecting lung micromechanics. In the present study, we use MRI morphometry to quantify the regional response of the acinar microstructure, specifically the surface-to-volume ratio (SVR), to changes in peak inspiratory pressure (PIP) in ventilated healthy rats. In this way, we present the mapping of pseudo-compliance (C_Xe), which we believe reflects regional microelastic and micromechanical properties of the lung.⁵ We compare pseudo-compliance maps with whole-lung compliance obtained using conventional multislice ¹H MRI performed concurrently at the same PIPs.

Methods

Twelve week old Sprague-Dawley rats were anesthetized, tracheostomized, and continuously ventilated in the supine position with air and given four pre-breaths of HP ¹²⁹Xe followed by a 13 second breath-hold during which diffusion-weighted MRI was performed (3T Prisma, Siemens).⁵ To allow for imaging of the gravitational gradient, sagittal images were acquired in addition to coronal images. Six b-value images ranging 0- 30.2 s/cm² were acquired and averaged 2-4 times (Gmax = 7.9G/cm, diffusion time of 2.65 ms, TE = 8.25 ms, TR = 16ms, FA = ~4°, FOV = 70mm, and a matrix = 64 x 64, and a partial echo of 62.5%) to ensure adequate SNR. This was repeated at PIP of 6, 9 and 12 cmH₂O respectively, to span the linear region of the compliance curve.³ Images were zero-padded, Hanning, and moving-average filtered using MATLAB 2014a (MathWorks). Acinar SVR maps were calculated as previously described⁶ and co-registered using a deformable registration algorithm to account for lung expansion (imtransform.m), based on the corresponding ventilation images (b = 0) at each PIP. SVR maps were then pixel-wise fitted as a function of PIP to obtain C_Xe, defined as the slope of the SVR vs. PIP fits. Only fits with R²>0.5 were included in the C_Xe map.

Results

Figure 1 shows typical C_Xe maps for a representative rat in the coronal and sagittal projection of the whole lung, demonstrating that the spatial distribution of the pseudo compliance values is quite different along both views. Regions of increased and decreased SVR with pressure are visible in all maps; however, in Fig. 1a the dorsal regions show "truly" compliant lungs (ie. decreased SVR with pressure), while the ventral areas show predominantly increases in SVR. Maps similar to Fig. 1 were obtained for all animals (n=8; data not shown). Total lung compliance values scaled by average alveolar density values (~1.2 cm³/cmH₂O) compare favorably with values measured using ex-vivo microscopy (~0.35-0.90 cm³/cmH₂O)^7,8 as well as with compliance from the multi-slice ¹H data obtained in this study (0.44 cm³/cmH₂O).

Discussion

Figure 1 shows regional variability in C_Xe, which agrees with the high heterogeneity also seen in CT scans in human and animal studies of lung regional ventilation and elasticity.⁹ Furthermore, Fig. 1d also shows evidence of a gradient along the dorso-ventral axis in the direction of gravity, as previously established.¹⁰ Since C_Xe is the change in SVR as a function of PIP, C_Xe is affected by two major factors: expansion of terminal airways resulting in a decrease in SVR with increasing pressure (showing compliance), and alveolar recruitment resulting in an increase in SVR with increasing pressure. Areas with positive C_Xe have no distensibility while those with negative C_Xe have normal distensibility. Regions of null C_Xe are present where there is no change in the SVR with pressure (mostly seen in the airways). While the dorsal regions are measuring "true" lung compliance, the peripheral areas may be suffering from atelectasis (closed alveoli) and hence biasing the compliance characterization; changing its linearity and posing a limitation to the presented method. Atelectasis is commonly seen in the peripheral parts of the lung and can be recruited with increased pressure, thus potentially explaining an increased SVR and positive pseudo-compliance in these regions. This effect may also explain whole-lung compliance differences derived from C_Xe as compared to ¹H. Positive end-expiratory pressure maneuvers are often used to re-recruit alveoli³, and this process could potentially be tracked by mapping C_Xe to avoid over-distention or injury to the lungs. This new tool provides regional compliance information that can potentially be very useful in early detection and guiding treatment for patients suffering from lung injury, where compliance is known to change.¹¹

Acknowledgements

The authors are grateful to N. Kanhere, Y. Friedlander and F. Morgado for ¹²⁹Xe gas polarization and help with experiments, to Dr. M. Couch for helpful discussion on pulse-programming and software development and to the funding sources: CIHR operating grant (MOP 123431) and NSERC Discovery grant (RGPIN 217015-2013). AL was supported by an NSERC PGSD award.

References

1. Yablonskiy, et al. Diffusion lung imaging with hyperpolarized gas MRI. NMR Biomed 2015; doi: 10.1002/nbm.3448. 2. Chan HF, et al. Exploring Lung Inflation Mechanisms with 3D ³He and ¹²⁹Xe Whole Lung Morphometry Mapping. Proc. of The 25th International Society for Magnetic Resonance in Medicine Annual Meeting. Honolulu, USA: poster # 3315; 2017. 3. Cereda M, et al. Positive End-expiratory Pressure Increments during Anesthesia in Normal Lung Result in Hysteresis and Greater Numbers of Smaller Aerated Airspaces. Anesthesiology 2013; 119(6): 1402–1409. 4. Xin Y, et al. Hyperpolarized gas diffusion MRI of biphasic lung inflation in short- and long-term emphysema models. Am J Physiol Lung Cell Mol Physiol 2017; 313: L305–L312. 5. Lindenmaier AA, et al. Pseudo-Compliance Mapping of the Rat Lung Using Hyperpolarized ¹²⁹Xe Morphometry at Varying Pressures. Proc. of The 2017 International Workshop on Pulmonary Imaging. Philadelphia, USA: University of Pennsylvania; 2017. 6. Sukstanskii AL, et al. Lung morphometry with hyperpolarized ¹²⁹Xe: theoretical background. Magnetic Resonance in Medicine 2012; 67(3): 856-866. 7. Fox MS, et al. Comparison of hyperpolarized He and Xe MRI for the measurement of absolute ventilated lung volume in rats. Magnetic Resonance in Medicine 2014; 71: 1130–1136. 8. Lai YL, et al. Respiratory mechanics in the anesthetized rat. Journal of Applied Physiology 1978; 45: 255–260. 9. I. Galvin, et al. Distribution of blood flow and ventilation in the lung: gravity is not the only factor. British Journal of Anaesthesia 2007; 98(4): 420–8. 10. Kaneko K, et al. Regional distribution of ventilation and perfusion as a function of body position. Journal of Applied Physiology 1966; 21: 767–777. 11. Kuethe DO, et al. Magnetic resonance imaging provides sensitive in vivo assessment of experimental ventilator-induced lung injury. American Journal of Physiology-Lung Cellular and Molecular Physiology 2016; 311: L208–L218.