1184

Regional Detection of Lung Injury using Hyperpolarized Xenon-129 Mapping of Blood Hematocrit in a Rat Model Involving Partial-Lung Irradiation
Brandon Zanette1,2, Elaine Stirrat1, Salomeh Jelveh3, Andrew Hope3,4, and Giles E. Santyr1,2

1Peter Gilgan Centre for Research and Learning, The Hospital for Sick Children, Toronto, ON, Canada, 2Department of Medical Biophysics, University of Toronto, Toronto, ON, Canada, 3Radiation Medicine Program, Princess Margaret Cancer Centre, Toronto, ON, Canada, 4Department of Radiation Oncology, University of Toronto, Toronto, Canada

Synopsis

Dissolved 129Xe imaging holds promise for the detection of early functional decline due to radiation-induced lung injury through the extraction of regional quantitative parameters relating to lung physiology and gas exchange. In this work a spiral-IDEAL imaging technique was used with a partial-lung irradiation rat model to investigate regional changes lung function associated with injury and compare to histology. A significant reduction in capillary hematocrit (HCT) was observed in the vicinity of irradiation. These results were in agreement with quantitative histology of red blood cells. Imaging results were more sensitive than whole-lung spectroscopy, which was performed simultaneously.

Purpose

MRI of hyperpolarized (HP) 129Xe dissolved in the parenchymal tissue and red blood cells (RBCs) of the lungs enables the investigation of gas exchange between these compartments using chemical shift saturation recovery (CSSR) at varying saturation delay times1,2. CSSR has shown promise for early detection of radiation-induced lung injury (RILI) in previous animal studies3,4. However, these studies involved whole-thorax irradiation models and spectroscopic analysis, thereby obscuring any regional differences. Regional differences in gas exchange due to hemithoracic irradiation have been demonstrated using a spiral-IDEAL imaging approach5, but these differences were observed between left and right lungs only and were not correlated with histology of RBCs. Additionally, whole or hemithoracic irradiation may introduce cardiac damage, confounding RBC measurements. In this study we performed spiral-IDEAL HP 129Xe imaging of a rat model of RILI involving partial-lung irradiation, resulting in regional differences in injury avoiding the heart. The results of this work are correlated with regional analysis of relative RBC volume determined by histology.

Methods

Two cohorts (N=6 each) of Sprague-Dawley rats were used. One underwent partial-lung irradiation using an image-guided small animal irradiator (X-Rad 225Cx, PXi, North Brantford, CT) with the second serving as controls. A single dose of 20Gy photons irradiated the right medial lung in the coronal direction, avoiding the heart (Fig. 1).

Imaging was performed 4 weeks post-irradiation on a 1.5T MRI (Signa HDxt, GEHC, Waukesha, WI) with a birdcage rat coil (Morris Instruments, Ottawa, ON). Rats were ventilated as previously described3. A 4-point spiral-IDEAL6,7 sequence was used (FOV=5×5cm2, Cartesian Matrix=16×16, TE=0.3ms, ΔTE=700μs). An additional acquisition was performed with imaging gradients nulled for spectroscopy. Coronal projections were acquired at five saturation delay times by varying the repetition time (TR) between acquisitions (34ms, 50ms, 75ms, 150ms, 300ms) as shown in Fig. 2. Multiple averages were acquired to ensure similar SNR across all timepoints. Isotopically enriched (~86%) 129Xe was hyperpolarized using a commercial polarizer (Model 9800, Polarean, Durham, NC) yielding polarizations of ~15%.

Following imaging, rats were euthanized and the lungs were removed with major vessels ligated, to retain blood. After fixation, 5μm sections from the middle of the coronal plane were H&E stained. Twenty-five images from the right medial lung were captured and segmented into three clusters (airspace, tissue, RBC)8. The percent RBC area (PRA) of the irradiated region was calculated as the number of RBC pixels relative to the total FOV9, determining relative RBC volume.

Reconstruction and analysis were performed in MATLAB (MathWorks, Natick, MA). Images and spectra were scaled and normalized by the gas-phase as described previously2,10. Tissue and RBC spiral-IDEAL images were simultaneously fit to the Model of Xenon Exchange (MOXE) and maps of hematocrit (HCT) were calculated2. Statistical analysis was performed in GraphPad Prism (San Diego, CA). Two-tailed t-tests characterized the difference between means. Two-tailed Pearson correlation with a 95% confidence interval was used to compare imaging to histology.

Results

Representative HCT maps are shown in Fig. 3. MRI results are shown in Fig. 4a. ROI analysis of the right medial lung yielded a significant decrease (p<0.01) in HCT in the irradiated rats compared to non-irradiated rats (16.6%±1.3% and 21.7%±4.0%, respectively). HCT values calculated from whole-lung CSSR were not significantly different between irradiated and non-irradiated rats (24.7%±2.0% and 24.6%±2.9%, respectively). Imaging results were consistent with histologically-observed decreases in blood (Fig. 5). PRA was significantly decreased (p<0.05) in the irradiated cohort compared to the non-irradiated cohort (1.23%±0.76% and 4.55%±2.83%, respectively). A strong positive correlation between HCT and PRA was observed (Fig. 4b).

Discussion

HCT measured with spiral-IDEAL was significantly decreased in the region of irradiation in agreement with histologically determined PRA. This work illustrates that for focal injury, regional detection is not only desirable, but necessary since whole-lung spectroscopy HCT was not significantly different, suggesting the effect of damage may be obscured by partial-volume effects. The physiological source of this RBC reduction requires further investigation. A reduction of vessel density through vessel destruction and/or shunting of perfusion away from the injured region may be the cause. In future, the imaging described in this work may be used to answer radiobiological questions about damage to lung vasculature. This method holds additional promise for RILI detection through other MOXE parameters that are not discussed here (e.g. septal thickness, capillary transfer time). Clinical implementation of spiral-IDEAL may allow for early RILI detection during radiotherapy, potentially allowing for intervention (e.g. treatment plan adjustment, administration of pharmaceuticals) improving treatment outcomes.

Conclusion

Spiral-IDEAL imaging is sensitive to regional HCT reductions associated with irradiation in a rat model, in agreement with histology.

Acknowledgements

This work was supported by a NSERC, CIHR, and RESTRACOMP. Thanks to Dr. Richard Hill and Dr. Ozkan Doganay for assistance with the irradiation scheme and MRI methodology, respectively. The authors would like to recognize the Spatio-Temporal Targeting and Amplification of Radiation (STTARR) facility for assistance with irradiations. Special thanks to Nikhil Kanhere, Andras Lindenmaier, Yonni Friedlander, and Felipe Morgado for assistance with imaging experiments.

References

1. Ruppert K et al. NMR of hyperpolarized 129Xe in the canine chest: Spectral dynamics during a breath-hold. NMR in Biomedicine. 2000;13(4):220–228.

2. Chang Y V. MOXE: A model of gas exchange for hyperpolarized 129Xe magnetic resonance of the lung. Magnetic Resonance in Medicine. 2013;69(3):884–890.

3. Fox MS et al. Detection of radiation induced lung injury in rats using dynamic hyperpolarized 129Xe magnetic resonance spectroscopy. Medical Physics. 2014;41(7):72302.

4. Li H et al. Quantitative evaluation of radiation-induced lung injury with hyperpolarized xenon magnetic resonance. Magnetic Resonance in Medicine. 2016;76(2):408–416.

5. Doganay O et al. Quantification of regional early stage gas exchange changes using hyperpolarized 129Xe MRI in a rat model of radiation-induced lung injury. Medical Physics. 2016;43(5):2410–2420.

6. Wiesinger F et al. IDEAL spiral CSI for dynamic metabolic MR imaging of hyperpolarized [1- 13C]pyruvate. Magnetic Resonance in Medicine. 2012;68(1):8–16.

7. Doganay O et al. Hyperpolarized 129 Xe imaging of the rat lung using spiral IDEAL. Magnetic Resonance in Medicine. 2015;76(2):566–576.

8. Sieren JC et al. An automated segmentation approach for highlighting the histological complexity of human lung cancer. Annals of Biomedical Engineering. 2010;38(12):3581–3591.

9. Takahashi A et al. Noninvasive assessment for acute allograft rejection in a rat lung transplantation model. Physiological Reports. 2014;2(12):e12244–e12244.

10. Chang Y V. et al. Quantification of human lung structure and physiology using hyperpolarized 129Xe. Magnetic Resonance in Medicine. 2014;71(1):339–344.

Figures

Fig. 1: Representative treatment plan from on-board cone beam CT planning in the coronal view (beams-eye view). Legend shows isodose colours of percentage of total dose (20Gy). Yellow line indicates lung contours. Green line indicates diaphragm contour. Approximately 0.5cc of tissue of the right medial lung was irradiated using a 1×1cm2 collimated beam avoiding the heart. Anatomical features were used to target the same area in each rat.

Fig. 2: Representative compartmental spiral-IDEAL images from a healthy rat as a function of increasing saturation delay times.

Fig. 3: Representative HCT maps from a non-irradiated and irradiated rat. Red arrow indicates region of decreased HCT in the right medial lung, in the vicinity of irradiation. ROI analysis of the right medial lung was used to analyze HCT for all animals.

Fig. 4: a) Mean HCT calculated from the right medial lung of HCT maps acquired using Spiral-IDEAL was significantly reduced in the irradiated rats compared to non-irradiated. This was in agreement with histology. Conversely, HCT calculated from whole-lung spectroscopy was not significantly different between cohorts. b) Scatter plot of histology-derived PRA versus MRI-derived HCT from Spiral-IDEAL shows a strong positive correlation. This indicates as histologically measured blood content decreases as a result of irradiation, this causes HCT as measured by MRI to decrease accordingly.

Fig. 5: a) Coronal tissue section from right lung of a non-irradiated rat. Dashed lines indicate the medial area of the lung from which magnified images were captured. Subfigures show images captured at 20× magnification (865×600μm2) before and after segmentation (white=airspace, blue=pulmonary tissue, red=RBCs). b) Tissue section from an irradiated rat for comparison. RBC content in the medial region of the lung is significantly decreased due to irradiation.

Proc. Intl. Soc. Mag. Reson. Med. 25 (2017)
1184