| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Experimental Studies |
1 From the Departments of Radiology (M.S., A.M., O.W., G.A.K., S.S., D.S., C.B.H.) and Medicine (R.L.), University of California San Francisco, 505 Parnassus Ave, San Francisco, CA 94143-0628; Philips Medical Systems, Best, the Netherlands (A.M.); and Bioheart, Santa Rosa, Calif (M.L.). From the 2002 RSNA scientific assembly. Received April 28, 2003; revision requested July 10; revision received August 21; accepted October 8. Address correspondence to M.S. (e-mail: maythem.saeed@radiology.ucsf.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Experiments were performed in six dogs by using an x-ray/MR fluoroscopy system. The arterial guide wire and catheter were advanced into the heart with x-ray fluoroscopic guidance. The dogs were injected with 0.5, 1.0, and 2.0 mL of iohexol. For passive catheter tracking, a steady-state free precession MR imaging sequence was used. A steerable dual-lumen catheter was used to transendocardially inject a mixture of gadodiamide (0.05 mol/L) plus Evans blue dye (3%). An electrocardiographically gated dual-inversion-recovery MR imaging sequence was used to visualize the myocardial delivery of the gadodiamide–blue dye mixture. A high concentration of gadodiamide (0.5 mol/L) was used to demarcate the borders of the area of interest, or "hit the target." Blood pressure, heart rate, and oxygen saturation were measured before and after the intervention. Analysis of variance, Scheffé, and paired Student t tests were used for data analysis.
RESULTS: LV catheterization via arterial access was feasible with two-dimensional x-ray fluoroscopic and three-dimensional MR fluoroscopic guidance. Delivery of the gadodiamide–blue dye mixture and the consequences of the procedure were monitored with MR imaging. Gadolinium-enhanced regions were bright on T1-weighted MR images, but they varied in size as a function of injectant volume. The mean sizes of these regions were 1.5% ± 0.6 of the LV after the 0.5-mL injection of the mixture and 7.0% ± 0.5 of the LV after the 2.0-mL injection (P < .001, Scheffé test). The corresponding mean sizes of the blue dye–enhanced regions were 2.3% ± 0.6 and 8.3% ± 0.4, respectively (P < .001). A high concentration of gadodiamide caused signal intensity loss around the gadolinium-enhanced regions.
CONCLUSION: Transendocardial delivery of potential therapeutic solutions is feasible with x-ray/MR fluoroscopic guidance. The injection catheter can be navigated with MR imaging guidance to hit the target.
© RSNA, 2004
Index terms: Animals • Catheters and catheterization, 51.122, 51.1269 • Experimental study • Heart, interventional procedures, 51.122, 51.1269 • Magnetic resonance (MR), guidance, 51.121411, 51.121413, 51.121416, 51.12143, 51.12149
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Intramyocardial delivery of therapeutic agents has been suggested (1–4) to accomplish the following: (a) the delivery of a high local concentration of the agent, (b) the delivery of the agent to pathologic regions only, and (c) minimization of the potential for systemic adverse effects, such as hypotension, accelerated atherosclerosis, or oncogenesis (9). Intramyocardial delivery of therapeutic agents has been accomplished with surgery (ie, minithoracotomy) or selective coronary arterial catheterization. More recently, percutaneous nonfluoroscopic electromechanical mapping technology (NOGA; Biosense Webster, Diamond Bar, Calif) has been used to map the myocardial scar and thus demarcate the target for therapy delivery (10–12). Real-time magnetic resonance (MR) imaging also has been used to monitor the delivery of therapeutic solutions to the myocardium and vascular wall (13–15).
The purpose of our study was to demonstrate the feasibility of using a combination of x-ray fluoroscopic and MR fluoroscopic (ie, x-ray/MR fluoroscopy) guidance for left ventricular (LV) catheterization and transendocardial delivery of extracellular tissue markers. An additional goal was to determine whether this method is precise enough to enable the delivery of solution to a specified site (ie, target) in the myocardium in a simulation of delivery to a specific region of ischemic myocardial injury (ie, infarction or scar).
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
After percutaneous catheterization of the femoral artery, the dog’s heart rate and blood pressure were continuously measured by using a pressure transducer (Gould, Cleveland, Ohio). Phasic and mean arterial blood pressures were recorded on a multichannel recorder (Gould). Oxygen saturation was continuously monitored by using an oxygen sensor placed on the tongue (3150 MRI; InVivo Research, Orlando, Fla).
Prototype Endovascular Catheter
The prototype endovascular catheter (which is not Food and Drug Administration approved) consists of steering and needle-adjusting components (scale, 1–7 mm) at one end and a nitinol needle (Bioheart, Santa Rosa, Calif) for puncturing the myocardial wall at the other end. The steering component controls the tip deflection, the needle depth adjustment, and the needle advancement. The handle allows rotation of the catheter tip in all directions inside the LV chamber for delivery of the injectant to any segment of the LV wall (Fig 1). The catheter has two lumina: one for injection and/or blood sampling and the other for pressure measurement. The delivery system is composed of a 9-F-long sheath, a 9-F dilator, and a 6-F catheter. Each injectant was delivered manually for 15–20 seconds. In the last two animals, we modified the same catheter slightly by coating it with gadolinium oxide powder (Sigma Chemical, St Louis, Mo). This small modification led to increased visibility of the tip and of the catheter shaft in the LV.
|
A solution of gadodiamide diluted with saline solution (10:1 ratio) (10% of a 0.5 mol/L concentration equals a 0.05 mol/L concentration) plus 3% Evans blue dye was used to compare the distribution of marker at MR imaging with the distribution of marker at postmortem tissue examination. Three volumes of the gadodiamide–blue dye mixture (0.5, 1.0, and 2.0 mL) were injected into the myocardium of each dog. The size of the gadolinium-enhanced region determined at three-dimensional MR imaging was compared with the distribution of blue dye in the myocardial tissue at postmortem tissue examination. All tissue markers were extracellular water-soluble agents with comparable molecular weights. The molecular weights of iohexol, gadodiamide, and Evans blue dye were 821.1, 573.6, and 960.8, respectively.
The Hybrid X-ray/MR Fluoroscopy System
A hybrid interventional x-ray/MR fluoroscopy system was used in the current study (Fig 2). This system consists of a short-bore MR imaging unit (Intera I/T; Philips Medical Systems, Best, the Netherlands) connected to a digital fluoroscopic catheterization laboratory (Integris V5000; Philips Medical Systems). A sliding door, which shields against radiofrequency energy or x rays, separates and permits combined or independent use of the two units. When this door is open, the two units are connected by a movable tabletop for rapid transport of the subject back and forth between the two units. Each experiment began with the subject in the catheterization laboratory; the animal was then moved to the MR imaging unit.
|
MR Fluoroscopy for Catheter Guidance
The animal was transferred to the MR imaging unit, without its position being changed, on the movable tabletop (Fig 2). The time to transport the animal between the two units was less than 3 minutes. A steady-state free precession MR imaging sequence and a dual-inversion-recovery T1-weighted contrast material–enhanced black blood MR imaging sequence were used for catheter guidance and for depiction of the myocardial enhancement caused by the injectant, respectively.
The advancement of the endovascular catheter from the descending aorta to the ascending aorta and then to the LV was monitored by using a steady-state free precession MR fluoroscopic sequence with the following parameters: 3.8/1.89 (repetition time msec/echo time msec), 60° flip angle, 5-mm section thickness, 200-mm field of view, 144 x 144 matrix, and 500-msec temporal resolution in continuous imaging. An 8-second breath hold per section was used for cine MR imaging.
A dual-inversion-recovery T1-weighted contrast-enhanced black blood MR imaging sequence was used to visualize the consequences of delivering the gadodiamide–blue dye mixture (16). The imaging parameters for this examination were 500 (every heart beat)/40, 5-mm section thickness, 200-mm field of view, 256 x 160 matrix, turbo spin-echo factor of 16, black blood double-inversion pulse, two signals acquired, and acquisition time of approximately 9 seconds per section (ie, breath hold). Cardiac horizontal long-axis and sagittal planes were used to locate the tip of the catheter and the gadolinium-enhanced regions.
The 0.5-, 1.0-, and 2.0-mL volumes of the gadodiamide–blue dye mixture were injected into three LV regions in each dog (total of 18 injections administered in a total of six dogs). The sizes of the gadolinium-enhanced regions were measured by using the dual-inversion-recovery T1-weighted contrast-enhanced black blood MR images. To avoid perforation of the LV wall, a constant depth of 4 mm was chosen after the wall thickness was measured.
In two dogs the injection catheter was navigated by using real-time MR imaging to inject contrast agent into the borders of the area of interest, or "hit the target." In these animals, after the last 2.0-mL injection of the diluted gadodiamide (10:1 ratio), two injections of the high concentration of this agent (ie, 0.2 mL of 0.5 mol/L) were administered around the highly enhanced region to demonstrate the feasibility of the precise delivery of therapeutic agent. A high concentration of gadodiamide produces a T2* effect—that is, a signal intensity loss—in the myocardium. Two investigators (A.M., O.W.) acquired the MR images together in all animals.
Image Analysis
The sizes of the gadolinium-enhanced regions on MR images and of the blue dye–stained regions at postmortem examination were measured by using computer-assisted planimetry with the public domain NIH Image program (National Institutes of Health, Bethesda, Md) (17). The borders separating the nonenhanced and gadolinium-enhanced myocardial regions and the borders separating the nonstained and blue dye–stained myocardial regions were manually outlined on the MR images and in the heart tissue at postmortem examination, respectively. The sizes of the gadolinium-enhanced regions on the MR images were compared with the sizes of the blue dye–stained regions on slices of the heart at postmortem examination. The gadolinium-enhanced and blue dye–stained regions were measured by consensus (by M.S. and G.A.K.).
Postmortem Examination
The animals were sacrificed by means of an intravenous injection of a lethal dose of pentobarbital sodium (10 mL of 200 mg/mL). The time interval between the last injection of the gadodiamide–blue dye mixture and the excision of the heart was 5–10 minutes. To ensure the absence of perforation in the LV wall, the pericardial fluid was examined for the presence of blood or blue dye. After removal of the atria and right ventricle, the LV was cut into 5-mm-thick slices that corresponded to the MR image sections (ie, long-axis coronal views of the heart). The number of examined slices depended on the volume of injectant administered—that is, when 0.5 mL of contrast agent was injected, a total of eight heart slices from the six dogs were examined; when 1.0 mL was injected, a total 14 slices were examined; and when 2.0 mL was injected, a total of 21 slices were examined. The slices were imaged by using a flat-bed MR imaging unit. The surface areas of all nonstained and blue dye–stained heart slices were summed to determine the percentage of stained area in each heart (n = 6). The gadolinium-enhanced and blue dye–stained regions were measured (by M.S. and G.A.K.) by using the previously described public domain image analysis software (NIH Image) (17).
Statistical Analyses
Data are presented as means ± standard errors of the means. The sizes of the marked areas (ie, gadolinium-enhanced and blue dye–stained regions) after the injection of 0.5, 1.0, and 2.0 mL of the gadodiamide–blue dye mixture in each of the six dogs were measured and compared. The significance of mean differences in the sizes of the marked areas after injection of the three volumes was determined by using repeated-measures analysis of variance. We also performed post hoc pairwise comparisons of the mean differences by using the Scheffé test. Differences between the gadolinium-enhanced and blue dye–stained regions were determined by using the paired Student t test for continuous data. The same statistical test was used to compare the heart rates, blood pressures, and oxygen tensions before and after administration of the injectants.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
X-ray fluoroscopy depicted the catheter in the LV and the three enhanced injection sites after the administration of 0.5, 1.0, and 2.0 mL of iohexol in the myocardium. All 18 sites of iohexol injection showed increased radiographic density (Fig 3), but the volume of the regions with increased radiographic density could not be measured on the two-dimensional fluoroscopic images. Because of the two-dimensional nature of x-ray fluoroscopic images, the volumes of the enhanced regions on these images could not be quantified or compared with the volumes of the gadolinium-enhanced regions on the MR images or the volumes of the blue dye–stained regions on the heart tissue at postmortem examination.
|
|
|
Figure 5 shows a differentially enhanced myocardium after the administration of 2.0 mL of the gadodiamide–blue dye mixture. At postmortem examination, all 18 sites of injection of the mixture were stained blue. Analysis of variance results showed that the mean sizes of the blue dye–stained regions that resulted from the different injected volumes were significantly different (F = 46.514, df = 2, 12; P < .001). The mean sizes of the blue dye–stained regions in the six dogs were 2.3% ± 0.6 of the LV after the 0.5-mL injection, 4.5% ± 0.3 of the LV after the 1.0-mL injection (P = .015 for comparison with mean size after the 0.5-mL dose, Scheffé test), and 8.3% ± 0.4 of the LV after the 2.0-mL injection (P < .001 for comparisons with the mean sizes after both the 0.5-mL dose and the 1.0-mL dose). All of the blue dye–stained regions were significantly (P < .001 for 0.5 mL, P = .001 for 1.0 mL, and P < .001 for 2.0 mL; paired Student t test) larger than the gadolinium-enhanced regions on the MR images.
Figure 6 shows the catheter coated with three gadolinium oxide markers in the LV before and after administration of the gadodiamide–blue dye mixture. The catheter was navigated by using real-time MR imaging to deliver the contrast agent to the borders of an area of interest (ie, target) in the myocardium. Targets were created in the myocardium by injecting a diluted concentration of gadodiamide (ie, 2 mL of 0.05 mol/L). Targets were defined as bright regions 1.5–2.5 cm in diameter. The MR images in Figures 4 and 6 depict different injection sites with low (0.05 mol/L) (Figs 4, 6) and high (0.5 mol/L) (Fig 6) gadodiamide concentrations. After injection of the low concentration of gadodiamide, the injection site appeared bright owing to the dominant T1 effect of this contrast agent. Injecting a nondiluted concentration of gadodiamide (ie, 0.2 mL of 0.5 mol/L) caused signal intensity loss at the borders of the enhanced region owing to the susceptibility, or T2*, effect of the agent. The images in Figure 6 demonstrate the signal intensity loss around each side of the highly enhanced target and thus the ability to deliver agents to the borders of an area of interest (ie, the margins of a previously highly enhanced myocardial region).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
The real-time steady-state free precession MR imaging sequence generated adequate spatial and temporal (10 frames per second) resolution for tracking the advancement of the catheter into the LV. Results of the electrocardiographically gated dual-inversion-recovery T1-weighted MR imaging examination that was used to map the distribution of gadodiamide showed that the size of the gadolinium-enhanced region varied with the volume of gadodiamide–blue dye mixture injected. The procedures used to advance the catheter and to deposit the tissue marker in the myocardium caused neither acute deleterious cardiac effects, such as arrhythmia or increased or decreased heart rate or blood pressure, nor myocardial or vascular trauma or perforation.
X-ray/MR Fluoroscopy
Cardiovascular interventions performed by using MR imaging guidance are challenging in clinical applications. Various MR imaging unit configurations have been used to perform interventional procedures; these include open- and closed-bore systems, high-magnetic-field and low-magnetic-field systems, and the currently described hybrid x-ray/MR fluoroscopy system. The described x-ray/MR fluoroscopy system, by providing the possibility to continue or extend the intervention if it is restricted by MR imaging factors, has the potential to facilitate improved flexibility in interventional procedures. It also allows the possibility of reverting to the successful vascular intervention by using fluoroscopic control if MR imaging guidance proves to be limited.
The x-ray/MR fluoroscopy system used in this study has the advantages of both two-dimensional fluoroscopy and three-dimensional MR imaging for guiding and monitoring interventional procedures (16). Since contrast-enhanced MR imaging can depict ischemia (18) and myocardial infarction (19), the possibility of using real-time MR imaging to deliver local therapeutic agents such as angiogenic growth factors and stem cells is attractive. Catheter navigation is routinely performed with fluoroscopic guidance; however, the size of the enhanced myocardial region cannot be readily quantified or localized in the LV wall with two-dimensional fluoroscopic images.
The main advantages of using a hybrid x-ray/MR fluoroscopy system are as follows: (a) The fluoroscopy system can be used to guide the catheters into the ascending aorta, or even into the heart, and the injection of solution into the myocardium can be performed by using MR imaging guidance and localization. MR imaging also can be used to visualize the distribution of the therapeutic agent in the myocardium. (b) Fluoroscopic guidance may be critical when the ileofemoral vessels are tortuous or narrowed.
The results of the current study are consistent with those described in an earlier report (10), which also showed that a delivery catheter could be guided to the myocardium by using real-time MR imaging. In addition, the current study results show that the catheter can be navigated to deliver solution to the border of the area of interest, or hit the target, and the effect of the volume of solution injected can be measured in the myocardium. Moreover, x-ray/MR fluoroscopy can enable one to continue or extend a vascular intervention when MR imaging factors restrict the procedure. The combination x-ray/MR fluoroscopy system may be useful for local drug delivery (ie, through the myocardium, pericardium, or coronary artery) and for assessment of myocardial function after the intervention.
Differential Enhancement
The 18 selected LV regions in the six dogs after injections of the gadodiamide–blue dye mixture appeared bright and had well-defined borders on the dual-inversion-recovery contrast-enhanced MR images. Lederman et al (13) observed that the transendocardial injection of gadopentetate dimeglumine causes differential enhancement for more than 10 minutes. The short-lived enhancement in the normal myocardium after focal delivery of the extracellular markers can be attributed to (a) the fast washout of the contrast material due to good myocardial perfusion, (b) a high gradient of gadolinium concentration between the blood and the interstitium, and (c) the fast renal clearance of extracellular MR imaging contrast material (19).
The described agent-delivery procedure is a potential method for the direct delivery of gene therapy and multipotential cells to infarcted myocardium. Delayed contrast enhancement of infarcted myocardium after intravenous injection of extracellular MR imaging contrast material—simulated with gadodiamide in this study—is an MR imaging approach to targeting the infarcted region. This delayed enhancement persists for 20–30 minutes, providing a window of time for an interventionist to hit the target. The prolonged enhancement of the infarcted myocardium is most likely related to poor myocardial perfusion (ie, slow washout) and a large (ie, >80%) fractional distribution volume of the contrast material (19–21).
The sizes of the gadolinium-enhanced and blue dye–stained regions depended on the volumes of these contrast agents injected. The sizes of the blue dye–stained regions at postmortem examination were larger than the gadolinium-enhanced regions on MR images. The slight but significant overestimation of the size of the blue dye–stained regions at postmortem examination was most likely due to the fast diffusion of the extracellular marker during the time between MR imaging and animal sacrifice.
Endovascular Catheters and MR Imaging
The prototype endovascular catheter with an adjustable 1–7-mm nitinol needle for puncturing the endocardial wall allows injection of solution to different depths. The needle-adjusting device on the handle of the catheter system is needed in cases of thin walls (eg, vascular wall or LV wall with transmural infarction) and thick walls (eg, wall of hypertrophied heart). In the current study, contrast agent was injected to only one depth, 4 mm, to simulate agent delivery to a thin transmural infarction (ie, scar tissue). The steering device on the handle controlled the tip deflection up to 180°. The handle allowed rotation of the catheter tip in the LV to various locations in the myocardium.
Visualization of the catheter and its tip was substantially improved after the shaft of the catheter was coated with gadolinium oxide markers. The catheter was coated with gadolinium oxide during the progression of this feasibility study. Thus, the coated catheter was tested in two animals, and more studies to address the utility and safety of this catheter are in progress. Visualization of the catheter tip (at imaging the last dark spot at the lower end of the catheter) improved our ability to hit the target. It should be noted that the use of these prototype catheters is not Food and Drug Administration approved.
Percutaneous nonfluoroscopic electromechanical mapping (10–12,22) is an alternative method of delivering and assessing the effects of local therapies. This technology involves the use of electromagnetic field sensors to combine and integrate real-time information from percutaneous intracardiac electrograms acquired at multiple endocardial locations. The resulting signal can be used to distinguish between normal and infarcted myocardial regions. Although percutaneous nonfluoroscopic electromechanical mapping has been used successfully to evaluate the effect of angiogenic factors in patients (21), it enables one to only approximate areas of myocardial scarring.
The main limiting factors of this study were the potential for myocardial injury caused by the nitinol needle during the delivery of the injectant and the undetermined safety of the prototype catheter in the MR imaging setting. More in vivo studies are needed to address the safety issue.
It can be concluded from the results of this feasibility study that transendocardial delivery of potential therapeutic solutions is feasible with x-ray/MR fluoroscopic guidance. The injection catheter can be navigated with MR imaging guidance to hit the target. The distribution of the extracellular markers in the myocardium was volume dependent in this study. As has been recently reported (15,23), the labeling of cells with MR imaging contrast material will enable one to monitor the delivery, distribution, and differentiation of mesenchymal stem cells.
Practical application: The described x-ray/MR fluoroscopic technique is a potential method for the direct delivery of gene therapy and multipotential cells to ischemic myocardial regions.
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, M.S., R.L., C.B.H.; study concepts and design, M.S., R.L., C.B.H.; literature research, M.S., G.A.K.; experimental studies, M.S., R.L., A.M., O.W., G.A.K., S.S., M.L.; data acquisition, M.S., A.M., O.W.; data analysis/interpretation, M.S., A.M., G.A.K.; statistical analysis, M.S., R.L., G.A.K.; manuscript preparation, M.S.; manuscript definition of intellectual content, M.S., R.L., A.M., O.W., G.A.K., M.L., D.S., C.B.H.; manuscript editing, M.S., C.B.H., A.M.; manuscript revision/review, M.S., R.L., A.M., C.B.H.; manuscript final version approval, all authors
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. Saeed, A. Martin, A. Jacquier, M. Bucknor, D. Saloner, L. Do, P. Ursell, H. Su, Y. W. Kan, and C. B. Higgins Permanent Coronary Artery Occlusion: Cardiovascular MR Imaging Is Platform for Percutaneous Transendocardial Delivery and Assessment of Gene Therapy in Canine Model Radiology, November 1, 2008; 249(2): 560 - 571. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Saeed, A. J. Martin, R. J. Lee, O. Weber, D. Revel, D. Saloner, and C. B. Higgins MR Guidance of Targeted Injections into Border and Core of Scarred Myocardium in Pigs Radiology, August 1, 2006; 240(2): 419 - 426. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. J. Lederman Cardiovascular Interventional Magnetic Resonance Imaging Circulation, November 8, 2005; 112(19): 3009 - 3017. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
