Thermal effect of intravascular MR imaging using an MR imaging

Signature: Med Sci Monit, 2002; 8(7): MT113-117 PMID: 12118208

WWW.MEDSCI MONIT.COM

Diagnostics and Medical Technology

Received: 2002.02.20 Accepted: 2002.05.10 Published: 2002.07.15

Thermal effect of intravascular MR imaging using an MR imaging-guidewire: an in vivo laboratory and histopathological evaluation

Authors’ Contribution: A Study Design B Data Collection C Statistical Analysis D Data Interpretation E Manuscript Preparation F Literature Search G Funds Collection

Xiaoming Yang1 abcdefg, Christopher J. Yeung2 cdef, Hongxiu Ji3 cde, Jean-Michel Serfaty1 bce, Ergin Atalar1 abcde

MT

1

Department of Radiology, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA Department of Biomedical Engineering, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA 3 Department of Pathology, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA 2

Source of support: This work was supported in part by National Institutes of Health grants R01 HL66187 and R01 HL57483.

Summary Background:

Intravascular magnetic resonance (MR) imaging to guide interventional procedures is a rapidly growing field. A primary concern with these new techniques is their thermal safety. The purpose of this study was to evaluate, in vivo, the thermal effect of an MR imagingguidewire (MRIG) for intravascular MR imaging (IVMRI).

Material/Methods:

Two indications of potentially adverse local heating were investigated: blood coagulation disorders and pathologic changes in target vessels. Experiments were performed on ten rabbits with a 1.5 T MR scanner. Using a 0.64-mm MRIG as the RF receiver, we imaged the target aorta using a fast spin-echo pulse sequence with an average specific absorption rate (SAR) of 0.6 W/kg. The total MR imaging time was approximately 70 minutes.

Results:

There were no abnormal value changes of the coagulation factors between pre- and postIVMRI, no clinical manifestations of blood coagulation disorders, and, histopathologically, no thermal damage in target vessels.

Conclusions:

This study demonstrates, from a pathophysiological point of view, the potential safe use of the MR imaging-guidewire for intravascular MR imaging. Further study is required to precisely define the boundaries of these safe operating parameters.

key words: Full-text PDF: File size: Word count: Tables: Figures: References:

Author’s address:

interventional MR • intravascular MR imaging • MR safety

http://www.MedSciMonit.com/pub/vol_8/no_7/2562.pdf 463 kB 2211 2 3 16

Xiaoming Yang MD PhD, Department of Radiology, Johns Hopkins University School of Medicine, Outpatient Center, Room 4243, 601 N. Caroline St., Baltimore MD 21287-0845, USA, email: [email protected]

MT113


BACKGROUND We have developed an MR imaging-guidewire (MRIG) [1] based on the loopless antenna design [2]. The MRIG combines the functionality of a conventional vascular guidewire with an imaging antenna. It can be directly inserted into small or tortuous vessels or placed into the central channel of an interventional device. Thus, the MRIG functions not only as an intravascular MR receiver probe for generating high-resolution MR angiographic images and intravascular MR fluoroscopic images for device tracking, but also as a conventional guidewire for guiding endovascular interventional procedures [1–4]. Concern has been expressed about the thermal safety of the MRIG and other similar active imaging devices that have been recently developed for interventional MRI. Inserted metallic objects have the potential to couple with the transmitting radio-frequency (RF) coil and amplify the electric field in their vicinity, which results in RF heating of the surrounding tissue [5]. Guidewires, for instance, which are typically passive and uninsulated, have been shown to cause prohibitively high RF heating under normal imaging circumstances in vitro [6,7]. Indirect in vitro observations have also shown that active devices can heat significantly in receive mode [8]. However, the few indirect in vivo studies of inserted active imaging probes have so far shown no evidence of RF thermal damage. For example, no electrically induced damage to the vessel wall was found when using an active intravascular field inhomogeneity catheter [9]. In addition, an active imaging catheter was successfully used to monitor balloon dilation of iliac and femoral artery stenoses in six patients with no significant complications under MR guidance [10]. In this study devoted to safety, we conducted a detailed pathophysiological examination of thermal effects from the MRIG.

MATERIAL AND METHODS Potential RF heating when imaging with the MRIG in vivo was investigated in live rabbits. The MRIG was inserted via the femoral artery into the lower abdominal aorta, where high resolution images of the aorta were obtained. Potential blood coagulation disorders (especially disseminated intravascular coagulation, (DIC)) and histopathologic changes in the target vessels and their adjacent tissues, both of which could be related to thermal effects from RF heating, were evaluated before, immediately after, and at day 10 after intravascular high resolution MR imaging. The MRIG was constructed from 50Ω flexible coaxial cable (Pico-Coax, Axon Cable Inc, Norwood, MA) with an outer diameter of 0.64 mm. The inner conductor, 0.4 mm in diameter, was extended by 9 cm to form the antenna. Thus, the coaxial cable was 66 cm in length for a total MRIG length of 75 cm. The MRIG was connected to the MR scanner through a matching, tuning, and decoupling circuit as described elsewhere [2]. Nine

MT114

Med Sci Monit, 2002; 8(7): MT113-117

Table 1. Animal groups. Group Histologic base-line (n=1) Operative base-line (n=3) Immediate post-MRI (n=3) 10-day survival post-MRI (n=3)

Operation* + + +

IVMRI

Histology

+ +

+ + + +

*operation includes surgery and instrument insertion; IVMRI – Intravascular MR imaging

identical MR imaging-guidewires were used in the present study. Ten New Zealand White rabbits, approximately 4.0 kg in weight, were divided into four groups (Table 1): 1) one rabbit for histologic base-line control without operation or intravascular MR imaging; 2) three rabbits for operative base-line control, in which the rabbits had both operation (i.e, surgery plus the MRIG insertion) and histologic examination at day 10 after the operation, with no intravascular MR imaging; 3) three rabbits for evaluation immediately following intravascular MR imaging; and 4) three rabbits for 10-day survival evaluation after intravascular MR imaging. In groups 3 and 4, all six rabbits underwent all operative procedures and intravascular MR imaging as well as histologic examination. Group 2 was set up to eliminate possible bias from the operation. The animals were treated according to the ‘Principles of Laboratory Animal Care’ of the National Society for Medical Research and the ‘Guide for the Care and Use of Laboratory Animals’ (NIH Publication No. 80-23, revised 1985). The experimental protocol was approved by the Animal Care and Use Committee at our institution. All animals were sedated with an intramuscular injection of a mixture of ketamine (35 mg/kg b.w.), acepromazine (0.75 mg/kg b.w.), and atropine (0.5 mg/kg b.w.). An ear vein was cannulated, which permitted maintenance of anesthesia. Intravenous pentobarbital (25 mg/kg b.w.) was administered later to bring the animal to a surgical plane of anesthesia. Once the surgical plane of anesthesia was reached, a femoral arteriotomy was performed to insert a 4 Fr introducer sheath (Meditech, Boston Scientific Corporation, Watertown, MA). Then the 0.64-mm MRIG was delivered into the abdominal aorta through the introducer and positioned with its tip at a level just above the right renal artery (Figure 1). The introducer was flushed with 2 mL saline. No heparin was used in the 10 rabbits. Then, the animals were transferred to the MR scanner, while anesthesia was maintained with serial infusions of pentobarbital (25 mg/kg/hr, i.v.) approximately every 30 minutes, or as needed. Non-ferromagnetic electrodes were attached to the limbs for measurement of surface ECG. Both ECG and blood pressure signals were used to monitor the condition of the animal during the experiment. Anesthesia was monitored for the duration of the experiment using regular tests of the eyelid reflex

Med Sci Monit, 2002; 8(7): MT113-117

Yang X et al – Thermal effect of intravascular MR imaging using an MR…

average SAR, which was 1.74 W/kg (3.5 W/kg peak). The total scan duration was approximately 70 minutes. Current SAR estimating algorithms are based on human-sized samples. Since the rabbits used in the experiment were much smaller than humans, the actual peak SAR delivered to the rabbit might be quite different from the scanner’s estimate. We calibrated for this effect by measuring the actual SAR delivered subcutaneously at four sample locations. When the scanner estimates 4.0 W/kg peak power, it actually delivers 1.7 W/kg peak power to a 4.0 kg rabbit, or about a third of the power. Therefore, the actual average SAR delivered to the rabbits in our experiment was about 0.6 W/kg average power (1.2 W/kg peak). In group 2, a total of 6 ml of blood for a coagulation profile was separately sampled from an ear vein of each rabbit before and immediately after operations. In groups 3 and 4, the 6 ml of blood from each rabbit was separately taken before, immediately after, or at day 10

Figure 1. A coronal projection MR image of the MR imaging-guidewire (large arrow) overlaid on a roadmap MR image of the rabbit body. The MR imaging-guidewire is placed within the lower abdominal aorta. The small arrow indicates the right renal artery.

and mild paw compression. On completion of the experiments, the animals were euthanized. All experiments were performed on a GE Signa LX 1.5 Tesla (T) cardiac MR scanner. For inserting the MRIG into and withdrawing it from the aorta, we used an intravascular MR fluoroscopy technique with a fast spoiled gradient echo (FSPGR) pulse sequence, 4.7/1.3msec repetition time (TR)/echo time (TE), 62.5-kHz bandwidth (BW), 36×18 cm field of view (FOV), 256×128 matrix, 3 frames/sec, and no slice selection. This sequence applied 4-mW peak power to the MRIG, which was configured in transmit/receive mode. We tested the thermal effects by imaging the aorta using the MRIG as the RF receiver. We attempted to generate the maximum RF heating condition by prescribing a pulse sequence with close to the maximum allowable average specific absorption rate (SAR). This was a fast spin-echo (FSE) pulse sequence, with 2000/17-msec TR/TE, 31.3-kHz BW, 8-cm FOV, 256×256 matrix, a 1.5-mm slice thickness, number of excitations=16, and 20 slices. This sequence was not optimized for the maximum imaging quality but for the maximum estimated

Figure 2. An illustration of the target aorta. Dotted lines indicate the places from where the specimens are harvested. 1) The right iliac artery (RIA), 2) the left iliac artery (LIA), 3) the aorta at the level of the most sensitive region of the MR imagingguidewire (MRIG), 4) the aorta at the level of the tip of the MRIG, 5) the aorta at a level 4-cm above the MRIG tip, 6) the aorta at a level 10-cm above the MRIG tip, 7) the inferior vena cava (IVC) adjacent to the target aorta (at level 3, above), and 8) at fat tissue adjacent to the target aorta (at level 3, above). RRA=right renal artery; LRA=left renal artery.

MT115

MT


after MR scanning. The blood samples were measured and reported (with a normal range guideline) in the laboratory (Antech Diagnostics Company, NY). The coagulation factors, which were measured in the blood coagulation profile, included partial thromboplastin time (PTT), platelet, prothrombin, fibrinogen, and D-Dimer. Since any portion of the intravascularly-placed MRIG could be in direct contact with the target vessel wall, in each rabbit we obtained several specimens from different levels of the target vessel for histological examination. These specimens were harvested from 1) the right iliac artery (into which the introducer was placed), 2) the left iliac artery (for control), 3) the aorta at the level of the most sensitive region of the MRIG (i.e, the point between the extended inner conductor and the coaxial cable), 4) the aorta at the level of the tip of the MRIG, 5) the aorta at a level 4 cm above the MRIG tip, 6) the aorta at a level 10 cm above the MRIG tip, 7) the inferior vena cava adjacent to the target aorta at the level of the most sensitive region of the MRIG, and 8) the fat tissue adjacent to the same target aorta (Figure 2). The specimens were embedded in paraffin, cut into 5 µm slices on a cross-sectional view, and stained with hematoxylin-eosin (HE). The thermal damages on the heattargeted vessel tissues with HE staining was defined as carbonization and coagulation as well as vacuolization, based on the histological analysis by previous studies [11,12]. A pathologist (HJ) and a radiologist (XY) performed the histopathologic examinations.

RESULTS In the blood coagulation profile, all of these average values were within the normal limits (Table 2). For preMRI/immediate post-MRI/day-10 post-MRI, the average values were 14.2/14.5/13.4 (seconds) for PTT, 300/272/421 (thousand/mm3) for platelet, 5.5/5.5/5.5 (seconds) for prothrombin, 400/349/400 (mg/dl) for fibrinogen, and