2780

Quantitative characterization of calcified and lipid-laden blood clot in vitro at 3T

Spencer D Christiansen^1,2, Junmin Liu¹, Trevor Wade¹, Joy Dunmore-Buyze¹, Michael B Boffa³, Luciano Sposato⁴, and Maria Drangova^1,2

¹Imaging Research Laboratories, Robarts Research Institute, Western University, London, ON, Canada, ²Dept. of Medical Biophysics, Schulich School of Medicine & Dentistry, Western University, London, ON, Canada, ³Dept. of Biochemistry, Schulich School of Medicine & Dentistry, Western University, London, ON, Canada, ⁴Dept. of Clinical Neurological Sciences, Western University, London, ON, Canada

Synopsis

Thrombus composition in embolic occlusion, particularly the presence of thrombolysis-resistant components such as calcium and fat, can significantly influence treatment efficacy, yet current MR methods for inferring composition are qualitative and sensitive only to red blood cells. We examined the ability of novel post-processing algorithms applied to a tailored GRE acquisition to discriminate and quantify important components within in vitro blood clots of varied hematocrit over a nine-day ageing period. Calcium and lard were readily discernable throughout the experiment, while clots were differentiable from one another between two to six days, demonstrating this protocol's potential for thrombus characterization in vivo.

Introduction

Thrombotic occlusion is the underlying cause behind a number of common and devastating pathologies including stroke, heart attack and pulmonary embolism. Knowledge of thrombus composition may provide highly useful clinical information towards the treatment of such conditions, including predicting the efficacy of thrombolytic agents¹ and mechanical thrombectomy procedures,² and possibly determination of etiology.³ Current MR methods for inferring thrombus composition rely on a qualitative “susceptibility vessel sign” obtained from late-echo GRE, a metric sensitive only to red blood cell (RBC) concentration and capable only of global assessments of composition, rendering invisible other informative components that may be present such as calcium⁴ and fat/cholesterol.⁵ This work evaluates the ability of a tailored multi-echo GRE acquisition paired with recently developed novel post-processing algorithms to characterize relevant thrombus components in a cohort of in vitro clots of varying composition and throughout clot ageing over a biologically relevant timescale.⁶

Methods

Imaging- Scans were performed at 3T with a 32-channel transmit/receive head-coil using a custom dual echo-train 3D GRE sequence (TE₁/∆TE/TE₅ = 3.20/1.46/9.04 ms, TE₆/∆TE′/TE₁₀ = 16.75/7.15/45.35 ms, TR: 47.6 ms, resolution: 0.94x0.94x1 mm³, matrix: 192x192x42, BW: 142.86 kHz, flip angle: 10°, scan time = 6 min 28 sec). The dual echo-train design enables the acquisition of both in- and out-of-phase images for water-fat separation and T2* decay sensitivity for Quantitative Susceptibility (QS) and R2* maps.

Image post-processing- Individual channel phase data were saved and the inter-echo variance channel-combination algorithm⁷ was used to create local frequency shift (LFS) maps for QS mapping,⁸ and the non-iterative B0-NICE algorithm⁹ was used to calculate fat fraction (FF), B0 field and R2* maps.

Phantom- Arterial porcine blood anticoagulated with sodium citrate was used to create duplicate 1.5mL blood samples of 0, 20, 40 and 60% hematocrit. Samples were clotted inside 1cm diameter polystyrene tubes by the addition of calcium chloride and thromboplastin. To emulate clinically observed emboli with calcified or lipidic components,¹⁰ 2.5mm length pieces of either calcium carbonate or lard were added to clots of different hematocrit. Platelet-poor plasma filled the remainder of each tube. Tubes were inserted into an agar phantom and kept at 37°C throughout the experiment except when scanning. The phantom was scanned at 2, 6 and 17 hours and daily thereafter up to 9 days post-clotting.

Statistical analysis- Repeated-measures ANOVA followed by the Tukey post-hoc test was used for statistical analysis. Significance was considered at p ≤ 0.05.

Results

Figures 1-3 show multi-parametric composite images of clot of each hematocrit at an early, intermediate and late time point. Figure 4 shows multi-parametric images of 40% hematocrit clots containing lard and calcium at every time point in the experiment. Lard had a significantly higher mean FF value, while calcium carbonate had a significantly higher QS value, than the opposite component and all clots at every time point. Clot R2* values showed a parabolic relationship with time, peaking between 48-72 hours (Fig. 5A). The 40 and 60% clots did not significantly differ at any time in experiment, but between days 2 to 6 the R2* values of the 0, 20 and ≥40% hematocrit clots were all significantly different from each other.

Discussion

Both calcium and lard remained readily differentiable from clot of every hematocrit (Figs 1-3) and at all time points (Fig. 4) in the experiment. Increasing clot R2* rates over the first days of clot formation is known to be caused by the increasing deoxygenation of RBCs as their metabolic processes fail.¹¹ The 40 and 60% clots have highly similar R2* values because they have equivalent magnetic susceptibility heterogeneity.¹² The decrease in clot R2* rates observed near the end of the experiment is likely due to diffusion of eryptotic iron components into plasma following lysis of RBCs,¹³ which is supported by the clot QS values remaining relatively stable throughout this period (Fig. 5B).

Conclusion

With the proposed protocol, clinically relevant sized pieces of lard and calcium carbonate are readily differentiated inside blood clot of up to 60% hematocrit aged up to 9 days. Blood clots of negligible (0%), low (20%) and medium to high (40-60%) hematocrit can be differentiated on the basis of R2* values, but only once the RBCs have become sufficiently aged/deoxygenated.

This method shows high clinical promise for discriminating calcified and lipidic components within in vivo thrombi, and inferring an approximate clot hematocrit in thrombi that are not extremely fresh; neither task is possible with current clinical methods. As well, the acquisition is readily translatable to the clinic as the work presented here was completed entirely from a single, unaccelerated 6-minute scan.

Acknowledgements

The authors would like to thank Ralph Bos Meats for providing the porcine blood. M.D. is a Career Scientist of the Heart and Stroke Foundation of Ontario and S.C. acknowledges the Canadian Institutes of Health Research Canada (CIHR) Graduate Scholarships-Master’s Program for funding of his research.

References

1. Niessen, F., et al., Differences in clot preparation determine outcome of recombinant tissue plasminogen activator treatment in experimental thromboembolic stroke. Stroke, 2003. 34(8): p. 2019-2024.

2. Yuki, I., et al., The impact of thromboemboli histology on the performance of a mechanical thrombectomy device. American Journal of Neuroradiology, 2012. 33(4): p. 643-648.

3. Boeckh-Behrens, T., et al., The Impact of Histological Clot Composition in Embolic Stroke. Clinical Neuroradiology, 2014.

4. Almekhlafi, M.A., et al., Calcification and endothelialization of thrombi in acute stroke. Annals of Neurology, 2008. 64(3): p. 344-347.

5. Ezzeddine, M.A., et al., Clinical characteristics of pathologically proved cholesterol emboli to the brain. Neurology, 2000. 54(8): p. 1681-1683.

6. Nosaka, M., et al., Time-dependent organic changes of intravenous thrombi in stasis-induced deep vein thrombosis model and its application to thrombus age determination. Forensic Science International, 2010. 195(1–3): p. 143-147.

7. Liu, J., et al., Inter-echo variance as a weighting factor for multi-channel combination in multi-echo acquisition for local frequency shift mapping: Local Frequency Shift Mapping. Magnetic Resonance in Medicine, 2015. 73(4): p. 1654-1661.

8. Bilgic, B., et al., Fast image reconstruction with L2-regularization. Journal of Magnetic Resonance Imaging, 2014. 40(1): p. 181-191.

9. Liu, J. and M. Drangova, Method for B0 off-resonance mapping by non-iterative correction of phase-errors (B0-NICE): B0 Mapping with Multiecho Data. Magnetic Resonance in Medicine, 2015. 74(4): p. 1177-1188.

10. Walker, B.S., L.M. Shah, and A.G. Osborn, Calcified Cerebral Emboli, A “Do Not Miss” Imaging Diagnosis: 22 New Cases and Review of the Literature. American Journal of Neuroradiology, 2014. 35(8): p. 1515-1519.

11. Bradley, W.G. and P.G. Schmidt, Effect of methemoglobin formation on the MR appearance of subarachnoid hemorrhage. Radiology, 1985. 156(1): p. 99-103.

12. Janick, P.A., et al., MR imaging of various oxidation states of intracellular and extracellular hemoglobin. American Journal of Neuroradiology, 1991. 12(5): p. 891-897.

13. Lang, E. and F. Lang, Triggers, Inhibitors, Mechanisms, and Significance of Eryptosis: The Suicidal Erythrocyte Death. BioMed Research International, 2015. 2015.

Figures

Figure 1: Magnitude images and quantitative maps of in vitro clot 2 hours post-clotting. Shown are early (in-phase) and late (representing current clinical standard) magnitude images, calculated LFS, R2*, quantitative susceptibility and fat fraction maps. Hematocrit (Ht) has little effect on clot appearance as the red blood cells are still oxygenated and diamagnetic. Lard and calcium are easily discernible in the qualitative magnitude images due to their relatively low signal, but are differentiable from one another only in the quantitative QS and FF maps.

Figure 2: Magnitude images and quantitative maps of in vitro clot 4 days post-clotting. Shown are early (in-phase) and late (representing current clinical standard) magnitude images, calculated LFS, R2*, quantitative susceptibility and fat fraction maps. Here hematocrit (Ht) strongly influences clot R2* and QS values, as the red blood cells are now deoxygenated and have shifted to a paramagnetic susceptibility. Lard and calcium are undetectable, and the >0% clots indistinguishable, in the late echo magnitude images used currently to characterize clot, while their ready detectability and differentiability in FF, QS and R2* maps, respectively, remains unchanged.

Figure 3: Magnitude images and quantitative maps of in vitro clot 9 days post-clotting. Shown are early (in-phase) and late (representing current clinical standard) magnitude images, calculated LFS, R2*, quantitative susceptibility and fat fraction maps. Clot R2* values have decreased since day 4, likely due to diffusion of ruptured red blood cell contents into the above plasma (note the vertical intensity gradients present within the 40 and 60% clot’s late magnitude and R2* maps). Lard and calcium are still undetectable in the 40 and 60% hematocrit clot’s late magnitude images, yet remain detectable and differentiable in QS and FF maps.

Figure 4: Magnitude images and quantitative maps of two 40% hematocrit clots throughout the 9-day experiment. Shown are A. late magnitude images (representing current clinical standard), B. quantitative susceptibility and C. fat fraction maps. In the magnitude images, calcium and lard components are visible only in the first 6 hours and 9th day of the experiment, and are indistinguishable from each other. In contrast, QS and FF maps clearly detect and differentiate the two components within clot at all time points. A small FF is detected in the calcium due to its susceptibility-related chemical shift (mean FF calcium/lard = 0.27/0.76).

Figure 5: A. Mean clot R2* rate as a function of time since clotting. (*indicates time points in which the 0, 20 and ≥40% hematocrit clots were each significantly different) B. Mean component QS value as a function of time since clotting. Clot R2* and QS values increase together over the first 48 hours (2 days), but R2* values decrease notably after 144 hours (6 days) without a proportionate decrease in QS, implying the magnetic components of the red blood cells may be diffusing into plasma. Changes in calcium carbonate QS value over time are likely due to moisture absorption.

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

2780