Hyperpolarized Helium3 Encapsulated in Microbubbles: A New Class of Blood Pool MRI Contrast Agent1 V. Callot, E. Canet, J. Brochot, H. Humblot, A. Briguet, H. Tournier, Y. Cre´millieux
RATIONALE AND OBJECTIVES A novel blood pool agent, based on hyperpolarized helium3 (HP 3He), was investigated for tissue perfusion using MRI. The results are reported in this paper. Thanks to the considerable polarization value obtained by optical pumping process, HP 3He has already demonstrated a great potential for anatomical and functional lung ventilation studies (1,2). In order to benefit from this large polarization and NMR signal advantages, we have studied the feasibility of this new contrast agent for perfusion imaging. To overcome the problem of the low helium solubility in blood, methods based on helium transportation using carrier agents have been investigated (3). In our study, a phospholipid-based substrate has been developed for helium encapsulation. Injections of 3-m-diameter microbubble solutions have been used for in vivo intravascular imaging. In this work, the lung parenchyma perfusion imaging (4) as well as imaging of the coronary arteries (5) and myocardium perfusion are investigated. MATERIALS AND METHODS Helium3 Polarization Helium3 was polarized at the Lau¨e-Langevin Institute (Grenoble, France) using the metastable-exchange method Acad Radiol 2002; 9(suppl 2):S501–S503 1
From the Laboratoire de RMN, CNRS UMR5012, Universite´ Claude Bernard Lyon, Domaine Scientiﬁque de la Doua–CPE–Aile C, 3 rue Victor Grignard, 69616 Villeurbanne, France (V.C., A.B., Y.C.); Creatis, CNRS UMR5515, Hoˆpital Cardio-Vasculaire Louis Pradel, Lyon, France (E.C.); Bracco Research SA, Geneva, Switzerland (J.B., H.T.); and Institut Lau¨eLangevin, Grenoble, France (H.H.). Address correspondence to Y.C.
(6). The typical polarization level varied between 50% and 55% at the end of the optical pumping process. Helium3 Encapsulation and Microbubble Characteristics Lyophilized substrates used for helium encapsulation, provided by Bracco-Research (Geneva, Switzerland), were made of a combination of phospholipids and pharmaceutical grade polyethyleneglycol. This type of substrate and the encapsulation technique are similar to those used to generate microbubble vascular contrast agents for ultrasonography. 300 mg of lyophilized substrate, 4 ml of saline and 5 to 6 ml of helium lead to a number of microbubbles equal to 8.108/ml and to a microbubble mean diameter equal to 3.0 ⫾ 0.2 m (measurements were done using a Coulter Multisizer II instrument (Coulter Electronics Ltd, London, UK)). NMR Methods NMR studies were performed on a 2 Tesla (Oxford Instrument, Oxford, England), 17-cm-bore magnet, interfaced to a SMIS console (Guildford, England). Lung perfusion experiments were achieved using a 6-cm-diameter volume coil tunable to both 1H and 3He resonance frequencies (respectively equal to 85.13 and 64.86 MHz ). Coronary imaging on the isolated pig heart experiments were performed using a 12-cm-diameter double tune volume coil. Longitudinal relaxation times T1 of encapsulated helium were measured using a single “pulse-acquire” sequence repeated n times. Transverse relaxation times T2 were determined using a multiple spin-echo acquisition. Apparent transverse relaxation time T *2 was measured from spectrum linewidth. Details of the measurement have been described elsewhere (4).
CALLOT ET AL
Academic Radiology, Vol 9, Suppl 2, 2002
Figure 1. Coronal projection of the lung parenchyma acquired after injection of a microbubble solution in the jugular vein (a). Transversal projection of the lung parenchyma acquired after injection of the microbubble solution, prior (b) and after (c) an experimental embolism model. Images (b) and (c) are superimposed with the corresponding proton image. FOV⫽80 mm.
For coronary and tissue perfusion imaging with encapsulated helium, a short echo time (30 s), low flip angle (10°), projection-reconstruction (PR) sequence was used. Two hundred radial directions were sampled and images were reconstructed onto a 1282 array. Animal and Isolated Reperfused Heart Model Preparation For lung perfusion experiments, anesthetized rats (280 – 350 g) were used. The bubble suspension was administered (1 ml in 3 s) via a catheter inserted in the jugular vein. For coronary and myocardium perfusion experiments, isolated heart, harvested from pig after sternomy, reperfused with a Krebs solution thanks to canula inserted in both coronaries, was used. The suspension was injected (8 ml in 4 s) in the left or right coronary. RESULTS The NMR parameters T1, T2, T *2, of the encapsulated helium3 have been measured in vitro in a syringe filled with microbubbles. The values were found respectively equal to 140 s, 300 ms and 4.5 ms. Results obtained in the lung after injection of the microbubble solution in the rat jugular vein are displayed in Figure 1. In Figure 1-a, the jugular vein, as well as the shape of the lung are clearly delineated; the SNR was measured in the lung parenchyma equal to 25 and the bright signal is attributed to the heart cavities. Since microbubbles are confined into the lung capillaries, their
distribution reflects the tissue perfusion. This was confirmed with an experimental embolism model, performed after injection of an air-bubble (7) in the tail vein of the rat (placed on its side), which created a perfusion defect after obstruction of the upper lung pulmonary artery. Results before and after emboli are shown in figures 1-b and 1-c, superimposed with the corresponding proton image: the area without signal corresponds to the embolized regions of the lung. Imaging of the coronary arteries displayed in Figure 2-a was performed after injection of the microbubble solution in the left coronary: the interventricular and circumflex arteries as well as smaller vessels are clearly visible. The measured diameters range from 1 mm (small vessels) to 3.6 mm (coronary). Besides the visualization of coronary branches (down to the 3rd generation), an enhancement of the myocardium corresponding to the distribution of the microbubbles in the capillaries can also be observed. This tissue enhancement reflects the myocardium perfusion. SNR have been measured between 10 (myocardium) and 30 (coronary). The relationship between microbubble distribution and tissue perfusion was demonstrated with the obstruction of the left coronary. Images before and after coronary occlusion are shown in figures 2-b and 2-c. In the reference image, the left and right coronaries are easily visualized, as well as the corresponding perfused areas. In image 2-c, only the right coronary is visible; the absence of signal in the territories normally perfused by the left coronary characterized the perfusion defect.
HELIUM3 MRI CONTRAST AGENT
Academic Radiology, Vol 9, Suppl 2, 2002
Figure 2. Projection acquired after injection of the microbubble solution in the left coronary (a). Transverse projection acquired after microbubble injection in both coronaries, prior (b) and after (c) occlusion of the left coronary. FOV⫽140 mm.
CONCLUSION The measured diameter of the microbubbles is close to an optimum value: it is large enough to transport a sufficient amount of polarized gas and it enables a safe passage through the microcirculation. Moreover, measured relaxation values of the encapsulated helium are long enough for in vivo imaging. It appears interesting to compare the technique of helium microbubbles with standard contrast agent technique. One important point concerns the SNR which has been estimated around 3 times larger than the SNR achievable with gadolinium-based contrast agent. A very peculiar property of the microbubble technique is the total absence of background signal coming from fat or surrounding tissues: the lung parenchyma, the coronary arteries as well as the myocardium tissue can be directly visualized. Another feature is the purely intravascular behavior of the microbubbles: their distribution in the capillaries reflects the tissue microcirculation and a perfusion defect is characterized by the absence of signal in the non-perfused area. Kinetics analysis of the microbubble passage through the tissue should be easily applied for perfusion quantification with the advantage of negligible recirculation and no contrast remanence effect. However, the sensitivity of the technique depends linearly on the microbubble concentration of the solution. This point may limit the use of the technique in the case of peripheral injection (inducing strong dilution of the solution into venous blood before the microbubble arrival into the organ of interest).
We have demonstrated, with encapsulation of the gas, that it is possible to benefit from the strong potential signal source of the hyperpolarized helium3 to perform perfusion imaging. Lung parenchyma imaging with large SNR values compared to standard lung proton imaging have been obtained after intravenous injection in rats, and a perfusion defect model has demonstrated that microbubble distribution reflects the tissue perfusion. Encapsulated helium has also been employed to successfully image the coronary arteries as well as the myocardium tissue. Microbubbles may therefore offer a new alternative for imaging and tissue perfusion quantification. REFERENCES 1. Kauczor H, Ebert M, Kreitner KF, et al. Imaging of the lungs using 3He MRI: preliminary clinical experience in 18 patients with and without lung disease. J Magn Reson Imaging 1997; 7:538 –543. 2. de Lange EE, Mugler JP, Brookeman JR, et al. Lung air spaces: MR imaging evaluation with hyperpolarized 3He gas. Radiology 1999; 210: 851– 857. 3. Chawla MS, Chen XJ, Moller HE, et al. In vivo magnetic resonance vascular imaging using laser-polarized 3He microbubbles. Proc Natl Acad Sci U S A 1998; 95:10832–10835. 4. Callot V, Canet E, Viallon M, et al. MR perfusion imaging using encapsulated polarized 3He. Magn Reson Med 2001; 46:535–540. 5. Callot V, Canet E, Brochot J, et al. Coronary MR imaging using laserpolarized helium3 microbubbles: preliminary results in isolated pig heart (abstr). In: Proceedings of the Ninth Meeting of the International Society for Magnetic Resonance in Medicine. Berkeley, Calif: International Society for Magnetic Resonance in Medicine, 2001; 1:517. 6. Colegrove FD, Schearer LD, Walters GK. Polarization of 3He gas by optical pumping. Phys Rev 1963; 132:2561. 7. Bertheze`ne Y, Vexler V, Price DC, et al. Magnetic resonance imaging detection of an experimental pulmonary perfusion deﬁcit using a macromolecular contrast agent. Invest Radiol 1992; 27:346 –351.