Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online before print February 5, 2004, 10.1148/radiol.2303021713
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2303021713v1
230/3/743    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by McDannold, N.
Right arrow Articles by Hynynen, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by McDannold, N.
Right arrow Articles by Hynynen, K.
(Radiology 2004;230:743-752.)
© RSNA, 2004

Experimental Studies

Heat-activated Liposomal MR Contrast Agent: Initial in Vivo Results in Rabbit Liver and Kidney1

Nathan McDannold, PhD, Sigrid L. Fossheim, PhD, Henrik Rasmussen, DVM, Heather Martin, BS, Natalia Vykhodtseva, PhD and Kullervo Hynynen, PhD

1 From the Department of Radiology, Harvard Medical School and Brigham and Women’s Hospital, 221 Longwood Ave, LMRC 007C, Boston, MA 02115 (N.M., H.M., N.V., K.H.); and Amersham Health, Oslo, Norway (S.L.F., H.R.). From the 2002 RSNA scientific assembly. Received December 18, 2002; revision requested February 7, 2003; final revision received June 14; accepted July 15. Supported by NIH grants CA 46627 and CA 089017. Address correspondence to N.M. (e-mail: njm@bwh.harvard.edu).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
 REFERENCES
 
PURPOSE: To evaluate by using in vivo magnetic resonance (MR) imaging the functionality of a liposomal paramagnetic contrast agent with T1 relaxivity that rapidly and markedly increases at temperatures above the gel-to-liquid crystalline phase transition temperature (Tc) of the liposome membrane.

MATERIALS AND METHODS: Liposomal gadolinium diethylenetriaminepentaacetic acid bis(methylamide) was injected intravenously at a dose of 0.4 or 1.2 mL (containing 10 or 30 µmol of gadolinium, respectively) per kilogram of body weight shortly before the application of focused ultrasound in liver (seven rabbits) or kidney (three rabbits). VX2 tumors had been implanted in liver in four of the rabbits. Eighteen locations in liver (13 in normal tissue, five in tumor) and 12 locations in kidney were sonicated. MR thermometry was performed during sonications. Signal intensity enhancement was evaluated on T1-weighted images acquired after the tissue cooled, and enhanced zones were compared with isotherms at Tc of the liposome membrane (approximately 57°C) by using Bland-Altman analysis. In liver, enhanced zones also were compared with areas of histologically verified thermal damage. The threshold temperature of enhancement at T1-weighted imaging was verified by monitoring the signal intensity increase after 10 sonications at varied powers in two locations in normal liver tissue.

RESULTS: Persistent enhancement was observed on T1-weighted images at all sonicated liver locations. In liver, enhanced zones on T1-weighted images were contiguous both with 57°C isotherms (25 measurements; mean difference ± SD, 0.4 mm ± 1.2) and with histologically verified areas of necrosis (seven measurements; mean difference ± SD, 0.1 mm ± 0.9). The threshold temperature of enhancement at T1-weighted imaging in normal liver was 53°–57°C. In kidney, enhanced zones on T1-weighted images did not match the isotherms.

CONCLUSION: The liposomal contrast agent was effective at in vivo MR thermometry in liver but not in kidney.

© RSNA, 2004

Index terms: Animals • Magnetic resonance (MR), experimental studies, 76.1291, 81.1291 • Magnetic resonance (MR), temperature monitoring, 76.12149, 81.12149 • Ultrasound (US), therapeutic, 76.12989, 81.12989


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
 REFERENCES
 
Magnetic resonance (MR) imaging techniques have been used with growing frequency to guide and monitor clinical applications of thermal therapy (1–7). MR imaging has been shown useful at all stages of thermal therapy, from treatment planning, localization, and monitoring, to posttherapeutic evaluation of treatment effects. The use of quantitative temperature mapping enables a high level of control in performance of thermal therapies. MR imaging–derived temperature information has proved useful for determining both the onset and the extent of thermal damage to tissue (8–10). MR thermometry enables verification that the heating is correctly localized with regard to the targeted volume and that the temperature increase is sufficient to produce the desired therapeutic result. Current MR thermometry techniques are limited, however, by their sensitivity to motion, ability to quantify changes only in temperature, lack of sensitivity in fatty tissue, and low sensitivity at low magnetic field strengths. Because these are serious obstacles in some therapeutic situations, an alternative technique is needed for MR guidance of thermal therapy.

Reports about the use of thermosensitive liposomes as vehicles for delivery of drug therapy have generated a great deal of interest (11,12). Some authors have also suggested that thermosensitive paramagnetic liposomes could be used for MR monitoring of thermal therapy (13,14). Liposomal contrast agents that are designed for this purpose remain dormant at physiologic temperatures but are activated when they reach a predetermined critical temperature (the gel-to-liquid crystalline phase transition temperature [Tc] of the liposome membrane). At temperatures lower than Tc, the paramagnetic element is restricted from the water in tissue and therefore has a minimal effect on the MR signal intensity. At temperatures higher than Tc, the paramagnetic element is released from the liposome carrier into the tissue water; it presumably cannot migrate from the site of its release after thermal coagulation has been achieved, and it remains in the tissue after it cools. Persistent enhancement of signal intensity on standard T1-weighted images therefore provides verification that an absolute temperature of at least Tc was achieved during treatment. The use of such liposomal contrast agents may lessen the motion sensitivity and increase the accuracy of MR guidance techniques by obviating image subtraction and enabling persistent enhancement, despite motion, in tissue that has been sufficiently heated.

The purpose of this study was to verify the functionality in vivo of a liposomal paramagnetic contrast agent with a T1 relaxivity that rapidly and markedly increases at temperatures higher than the liposome membrane phase transition temperature Tc.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
 REFERENCES
 
Experimental Protocol
The tissue was heated with focused ultrasound, and the temperature distribution was monitored with MR thermometry. Signal intensity enhancement due to thermally induced release of the paramagnetic element from liposomes in the contrast agent was detected with T1-weighted imaging. The main goal was to verify that enhancement occurred at the sonicated locations and persisted after the tissue had cooled and that MR thermometric isotherms at the Tc of the liposomal agent matched the enhanced zones on T1-weighted images. After initial negative results in the kidney (described in Results), no further evaluation of the liposomal agent was performed in the kidney. In the liver, additional analyses were performed to compare the enhanced zones on T1-weighted images with the areas of successful treatment confirmed at histologic analysis and to verify that the threshold temperature of enhancement was identical to the Tc of the agent.

In all experiments, the process charted in Figure 1 was followed. The time between the end of sonications and the acquisition of T1-weighted images (approximately 30–60 seconds) was sufficient to allow the tissue to cool almost to its baseline temperature. The T1-weighted MR imaging plane was the same as that used for MR thermometry (ie, coronal, perpendicular to the axis of the ultrasound beam, in the focal plane). In some cases after individual sonications, and at the end of every experiment, additional T1-weighted images were acquired in other planes. The comparison of T1-weighted images acquired immediately before and after intravenous injection of the liposomal contrast agent enabled measurement of any signal intensity changes caused by the presence of the contrast agent without the effects of heating. Because multiple locations were sonicated in each animal, the time delay between the injection of the agent and the applications of ultrasound varied. MR thermometric mapping was performed during and for 26–56 seconds after the sonications. Sonication depths in the exposed liver varied from approximately 5 mm to 2 cm from the surface (depending on the tumor location and orientation and the thickness of the liver lobe) and in the kidney were about 2 cm from the skin.



View larger version (38K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1. Flowchart of the experimental protocol.

 
The liposomal contrast agent was tested over a range of experimental parameters (Table). In normal liver and kidney, between two and five nonoverlapping locations were sonicated in each animal, depending on the extent of accessible tissue and the time available. In tumor-bearing liver, the aim was to thermally ablate both the tumor and a surrounding rim of normal tissue. In the first tumor, this goal was achieved with two overlapping 20-second sonications. In the next three tumors (three rabbits), the sonication time was increased to 60 seconds so that this goal could be achieved with a single ultrasound application. In these three rabbits, an additional location was sonicated in normal liver tissue distant from the tumor (in one rabbit, in a separate liver lobe). The gadolinium dose used (30 µmol per kilogram of body weight) was the same as that used in earlier ex vivo tests (14). A reduced gadolinium dose (10 µmol per kilogram of body weight) was tested in two of the four tumors after the functionality of the agent was assessed at the higher dose level.


View this table:
[in this window]
[in a new window]

  		Summary of Experiments

 
For each sonication, a maximum temperature map was reconstructed from the time series of temperature-sensitive images to show an isotherm at 57°C, the Tc of the agent, to enable verification that this was the threshold temperature of enhancement at T1-weighted imaging. In the liver, the 57°C isotherms were measured, and their dimensions were compared with those of the enhanced zones on T1-weighted images. Because the signal intensities of normal tissue and of thermally damaged tissue were obviously different on T1-weighted images, no grading scheme was needed to determine whether enhancement had occurred.

The enhanced zones on T1-weighted images were compared with histologically verified areas of thermal damage in the liver (at two locations in tumors and one location in normal tissue). Because the tissue at histologic analysis was sliced in a direction different from that used at MR imaging (see Histologic Analysis), measurements were performed on T1-weighted images acquired in two planes perpendicular to each other: Length was measured on a sagittal image, and width was measured on a transverse or coronal image. In one of the specimens, the length of the enhancing zone could not be obtained because of an error in processing. In another case, it was possible to compare the thickness of the histologic damage in front of and behind the tumor with the corresponding thickness of the enhancing rim in addition to the length and width of the entire lesion. Thus, a total of seven measurements (width of three enhancing zones, length of two enhancing zones, and two measurements of the thickness of the enhancing rim) were made on T1-weighted images and compared with subsequent measurements obtained at histologic analysis.

Normalized signal intensity on T1-weighted images was calculated for each sonicated location in liver by using signal intensity values measured either at the center of the enhanced lesions (in locations in normal liver) or in the enhanced rim that surrounded the tumors; the signal intensity in normal liver (ie, in a non–thermally treated region on the same image as the treated region) was used as the reference for normalization. In calculating normalized values, we took into account any changes that occurred from image to image (eg, changes in gain of the MR signal amplifiers or changes in signal intensity over time that were caused by the contrast agent but were unrelated to heating).

Comparison of the dimensions of the 57°C isotherms with the dimensions of the enhanced lesions (as described earlier) served to verify that the temperature that resulted in enhancement on T1-weighted images was equal to Tc of the liposomal agent. Because thermal gradients are high at the margins of thermally treated lesions and the accuracy of visual verification therefore might be insufficient, we performed separate experiments to verify the temperature at the center of the focal location, where the spatial distribution of heat produced by the transducer is relatively uniform. For this purpose, two locations in normal liver tissue were sonicated five times each at progressively higher powers (2–11 W). The incremental increase in power from sonication to sonication was calculated on the basis of the temperature increase achieved during the previous sonication. For each of these 10 sonications, the peak temperature achieved at the focal location was compared with the normalized signal intensity on T1-weighted images acquired between sonications. The threshold temperature of enhancement on T1-weighted images was defined arbitrarily as the temperature at which normalized signal intensity exceeded measured noise by 1 SD. The highest-power sonications at these two locations were delivered after the locations began to show signal intensity enhancement. These two locations also were included in the comparison of isotherms with enhanced areas on T1-weighted images.

Statistical analysis and MR image–based analysis, including the construction of temperature maps, measurement of isotherms and enhanced zones, and calculation of normalized signal intensity, were performed by one author (N.M.) using software written in-house for Matlab. Other authors implanted the tumors and prepared the animals (H.M.) and marked the histologically verified necrotic regions (N.V.). Several authors performed the experiments (N.M., K.H., H.M., S.F., H.R.).

Animals
The experiments were performed in male New Zealand white rabbits (Milbrook Breeding, Amherst, Mass). The experimental protocol was approved by the institutional animal research committee. The animals were anesthetized with a mixture of 12 mg of sodium xylazine (Xyla-ject; Phoenix Pharmaceuticals, St Joseph, Mo) and 48 mg of ketamine hydrochloride (Abbott Laboratories, North Chicago, Ill) per kilogram of body weight per hour. Prior to each experiment, the fur on the rabbit’s abdomen was removed with clippers and depilatory lotion. The rabbit’s temperature was monitored with a copper-constantan thermocouple, which was positioned in the liver near the sonication target during the liver experiments. In the kidney experiments, a rectal thermocouple was used. To maintain normal body temperature, warm water was circulated through a plastic mat beneath the animal and through a plastic coil that encircled the water bag used for acoustic coupling (see Ultrasound Application).

To prevent motion due to respiration (for accurate temperature imaging) and to create a larger target area (because sonications could not be performed through the ribs with this phased-array transducer), an incision was made below the ribs and a liver lobe was exposed. The edges of the incision were then sutured together as closely as possible around the exposed lobe so that the liver tissue would not retract when the animal was placed in the prone position. In one animal, the liver experiment was performed with two exposed liver lobes, one of which contained a tumor. It was not necessary to expose the kidneys, because they were not behind the ribs and because their motion during breathing did not impair MR thermometry.

Tumor Implantation
In four animals, tumors were implanted in the liver before sonication was performed. Approximately 50 million VX2 carcinoma cells were injected into the liver with either MR imaging or ultrasonographic guidance. The delay between the tumor cell injection and the experiments varied, depending on the time at which tumors were detectable with MR imaging. Good results were obtained when the tumors were injected 10 days before the experiments. Tumors measured on T2-weighted images before sonication had diameters of 5–21 mm (mean diameter, 10.6 mm). On T1-weighted images acquired prior to sonication, the tumors appeared smaller in diameter (range, 3–19 mm; mean, 8.5 mm) because only their hypointense centers were visible.

Ultrasound Application
Focused ultrasound waves were generated by an eight-element spherically curved air-backed phased-array transducer with a sector-vortex configuration (radius of curvature measured in centimeters/diameter measured in centimeters, 8/10; frequency, 1.71 MHz) (15). The transducer was connected to a computer-controlled multiple-channel amplifier system (16). The array could be operated in five numbered modes, from mode 0 (with a focus equivalent to that obtained with a non–phased-array transducer) to mode 4 (the largest focus) (15). All modes other than 0 produce multiple foci arranged in a ring. As tissue in the foci is heated, the center of the ring also is heated by thermal conduction, which results in a spatially flat temperature profile that allows for spatial averaging. Mode 4 was used in liver sonications to enable spatial averaging and to create the largest possible thermal lesion. Because of the high rate of perfusion in the kidney, it was necessary in kidney sonications to use a more concentrated focus (mode 0) to achieve a sufficient increase in temperature. As a result, lesions in the kidney were smaller than those in the liver. The performance of the transducer was characterized as described elsewhere (17). The full width at half maximum was 0.8 and 3.8 mm for modes 0 and 4, respectively.

The transducer was mounted in an MR imaging–compatible manual positioning system. The transducer was mounted in this system in a tank filled with de-ionized degassed water. The top of the tank was sealed with a flexible plastic membrane through which the ultrasound beam propagated upward out of the tank. The rabbit was placed in a prone position on a plastic tray arranged above the tank. This tray had a hole cut in it, which served as the opening of a plastic water bag that hung down and rested on the membrane on top of the tank. The target tissue was centered in this hole and partially submerged in the free water surface of the bag. During the liver sonications, another thin plastic membrane was attached over the top of the hole to support the exposed liver. A 12.7-cm-diameter, receive-only surface coil (GE Medical Systems, Milwaukee, Wis) was attached beneath the plastic tray.

Thermosensitive Liposomal Contrast Agent
Liposomal gadolinium diethylenetriaminepentaacetic acid (DTPA) bis(methylamide) (BMA) (Amersham Health, Oslo, Norway) was used as the thermosensitive contrast agent. Details about the liposome preparation, which is briefly described here, can be found elsewhere (13,18). Liposomes with membranes consisting of 90% (wt/wt) distearoylphosphatidylcholine and 10% distearoylphosphatidylglycerol were prepared with thin film hydration. The liposomes were subsequently minimized with membrane extrusion, and untrapped Gd-DTPA-BMA was removed with dialysis. Key physicochemical liposomal properties, measured with standard methods described elsewhere (13,18), included the following: liposome size, 110 nm; Tc, 57°C; effective gadolinium concentration, 24 mmol/L; intraliposomal gadolinium concentration, 250 mmol/L. Two doses of liposomal Gd-DTPA-BMA were tested: 1.2 and 0.4 mL (with 30 and 10 µmol of gadolinium, respectively) per kilogram of body weight. The liposomal agent was given as an intravenous bolus injection.

MR Imaging and Thermometry
MR imaging was performed with a 1.5-T clinical imager (GE Medical Systems). A T2-weighted fast spin-echo sequence (repetition time msec/echo time msec, 2,000/75; echo train length, eight; matrix size, 256 x 256; field of view, 16 cm; section thickness, 3 mm; number of signals acquired, two) was used to localize the implanted liver tumors and determine the targets for sonications. A T1-weighted fast spin-echo sequence (500/15; echo train length, four; matrix size, 256 x 256; field of view, 16 cm; section thickness, 3 mm; number of signals acquired, four) was applied before and after administration of liposomal Gd-DTPA-BMA and before and after sonications to enable detection of changes due exclusively to the presence of the liposomal agent. To investigate the cause of negative results in the kidney (see Findings in Kidney), we repeated the T1-weighted imaging sequence after bolus injection of the standard MR imaging contrast agent gadopentetate dimeglumine (Magnevist; Berlex, Wayne, NJ) at a dose of 0.125 mmol per kilogram of body weight.

Temperature changes in sonicated tissue were indicated by changes in the water proton resonance frequency (19), which was measured by applying a fast spoiled gradient-echo sequence (20,21) (40/19.7; flip angle, 30°; bandwidth, 3.57 kHz; matrix, 256 x 128; field of view, 20 x 15 cm; section thickness, 3 mm; imaging time, 4 seconds). A temperature sensitivity of -0.010 ppm/°C was used (19,22). A time series of images was acquired before, during, and after each sonication. Temperature maps were created from reconstructed real and imaginary data. The rabbit’s temperature, measured with the thermocouple, was used with the MR thermometry measurements to calculate absolute temperature. Temperature measurements that were corrupted by motion or boiling were excluded. Boiling was considered to have occurred when the temperature was near or above 100°C and a large susceptibility artifact was observed (eg, sudden change in temperature distribution and sudden appearance of a signal void near the focus on magnitude reconstructions from fast spoiled gradient-echo imaging).

Histologic Analysis
Shortly (<30 minutes) after the experiments, the animals were sacrificed. The resected liver samples, including the tumor and surrounding normal tissue, were fixed in 10% neutral formalin solution, embedded in paraffin, and then cut into 6-µm slices and stained with hematoxylin-eosin. The slices were cut in a plane parallel with the upper surface of the liver. During preparation, however, the lobes did not retain the orientation they had during the experiment, so the slices were not in the same orientation as the MR imaging sections. The sonicated areas were evaluated with light microscopy to identify the boundaries of necrotic tissue. These boundaries also were marked on a digital scan of a histologic slice from the center of the thermal lesion so that the dimensions of the area of necrosis could be measured.

Statistical Analysis
Least-squares linear regression was used to correlate the measurements of the enhanced lesions on T1-weighted images with those of 57°C isotherms and areas of necrosis identified and measured at histologic analysis. The Bland-Altman technique was used to evaluate the agreement between measurements, and the limits of agreement were defined as the mean difference ± 1.96 SD (23). Zero bias between the measurements was confirmed with a paired two-tailed t test. For normalized signal intensity measured on T1-weighted images and for peak temperature measured at MR thermometry in a treatment location, means ± SDs were calculated in a region of interest (ROI) at the focal coordinates. Statistical significance of the change in signal intensity induced by introduction of the agent before heating was calculated by using a Wilcoxon paired two-sample signed rank test. P < .05 was considered to indicate a statistically significant difference.

Numerous variables might affect the magnitude of the signal intensity enhancement at T1-weighted imaging after sonication, including the dose of the contrast agent, the heating duration and peak temperature, and the delay between injection of the agent and sonication. As seen in the Table, these parameters varied over the course of the experiments. Given the number of variables, the individual effect of each could not be determined. If the boundaries of signal intensity enhancement were clearly depicted on T1-weighted images, it was assumed that these variables had not affected the basic functionality of the agent. For this reason, measurements from all sonicated locations in the liver—those in normal tissue, as well as in tumor—were combined in the statistical analysis. The merging of data from locations in tumor-bearing liver with that from locations in normal liver was justified also because a rim of normal tissue around the tumor had been thermally ablated along with the tumor and because measurements used in the statistical analysis had been obtained mostly in the lesion margins.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
 REFERENCES
 
Effect of the Liposomal Agent before Heating
The average percentage of signal intensity change (± SD) that was unrelated to heating and was measured on T1-weighted images acquired immediately after the administration of the higher dose of Gd-DTPA-BMA (30 µmol of gadolinium per kilogram of body weight) was -2% ± 11 in normal liver, -2% ± 11 in liver tumor, and 8% ± 9 in kidney cortex. In fat and muscle near the kidney, signal intensity change after administration of the higher dose was 2% ± 4 and 4% ± 9, respectively. These changes were statistically significant (P < .01). At a dose of 10 µmol gadolinium per kilogram of body weight, the liver signal intensity changed by -1% ± 5, which was not a statistically significant difference (P = .06).

Findings in Liver
On T1-weighted images acquired in liver after the tissue cooled, signal intensity enhancement was observed at all (n = 18) sonicated locations (Figs 2, 3). Enhancement was not observed, however, in the hypointense centers of the implanted tumors. The diameters of the enhanced zones measured in the focal plane (on coronal images) were 2–9 mm for locations exposed to 20-second sonications and 14–20 mm for locations exposed to 60-second sonications. A temperature increase was observed on MR temperature images acquired during all sonications (n = 17), but reliable measurements could not be achieved in two cases because of boiling and in another two cases because of motion.



View larger version (55K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2. Top: Time series of temperature maps from coronal fast spoiled gradient-echo MR imaging (40/19.7) in a 16 x 16-mm ROI during and after sonication in rabbit liver (every other time point shown). Bottom: Normalized signal intensity profiles of voxels in the same ROI, measured with coronal T1-weighted fast spin-echo imaging (500/15). Curves show the extent and magnitude of enhancement on T1-weighted images (T1WI), and circles indicate actual measurements. The gray bands represent spatial regions in which the temperature reached 53°C (dark gray) and 57°C (light gray). Insets show same ROI from the T1-weighted MR image. Straight white lines delimit the sections in which enhancement was measured along the x and y axes. Peripheral contour lines indicate the isotherms surrounding areas in which maximum temperatures (Tmax) >= 53°C (white line) and >= 57°C (black line) were reached. The small anomaly in isotherms at the bottom of this lesion was caused by a neighboring blood vessel, which appeared hypointense on T1-weighted images.

 


View larger version (146K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3. Coronal (A, B) and sagittal (C, D) T1-weighted fast spin-echo MR images (500/15) of tumor-bearing liver (A, C) and normal liver (B, D) show focal signal intensity enhancement due to thermal activation of the liposomal agent after 60-second sonications at an acoustic power of 88 W. A and B were acquired in a plane perpendicular to the direction of the ultrasound beam, and C and D were acquired parallel to the direction of the ultrasound beam (arrow in C). Contour lines in A and B indicate 53°C (white lines) and 57°C (black lines) isotherms. Dotted lines in C and D indicate the approximate orientation of tissue slices at histologic analysis. The signal void to the left of the lesion in B is a chemical shift artifact.

 
The isotherms estimated from MR temperature images were within a few voxels of perfect agreement with the boundaries of the enhanced zones on T1-weighted images (Figs 2, 3). The correlation between the diameters of the isotherms and those of the enhanced zones was significant (R2 = 0.94, P < .01) (Fig 4, A). The difference between these measurements was not significant, verifying zero bias (P < .01). The mean difference in the measurements, calculated with the Bland-Altman method, was 0.4 mm ± 1.2, and the limits of agreement were -2.0 and 2.7 mm (Fig 4, B). Signal voids caused by chemical shift artifacts at the boundary between the liver and the surrounding fat on temperature images and T1-weighted images were the source of the greatest measurement error at the lesion boundaries (Fig 3, B). Blood vessels near the sonicated location affected comparison in one lesion (Fig 2). In addition, three of the sonicated locations were near the surface of the exposed liver lobe and bordered the water used for acoustic coupling. The temperature maps did not agree with the enhanced zones on T1-weighted images of these locations, presumably because of artifacts caused by the motion of the water during sonication. One measurement from each of these three locations was thus excluded from the data shown (Fig 4, A and B).



View larger version (51K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4. Scatterplots show the results of correlation analysis (A, C) and Bland-Altman analysis (B, D) of the dimensions of enhanced zones observed on T1-weighted images (T1WI) acquired after tissue cooling, compared with the dimensions of 57°C isotherms (in A and B) and areas of histologically verified thermal damage (in C and D). In A and C, solid lines show linear regression and dotted lines show 95% CIs. {circ} = enhanced edge around tumor, {square} = lesion length, x = lesion width.

 
The correlation between the dimensions of areas of histologically verified damage and areas of signal intensity enhancement at T1-weighted imaging was significant (R2 = 0.99, P < .01). The mean difference in measurements, calculated with the Bland-Altman method, was -0.1 mm ± 0.9, and the limits of agreement were -1.9 and 1.8 mm (Fig 4, C and D). The difference in the measurements was not statistically significant (P < .01). Thermal damage was observed both at gross examination and with light microscopy (Fig 5). The lesions comprised a central zone of necrosis that appeared consistent with thermal coagulation, which progressed outward to a zone in which necrosis appeared consistent with ischemia. The tumor centers appeared as zones of spontaneous necrosis.



View larger version (163K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5. A, Photograph shows ultrasound-induced thermal damage in a liver tumor (large arrow) and in normal liver tissue (small arrow). B, Photomicrograph (hematoxylin-eosin stain; original magnification, x1) shows the boundaries of the damage. C-F, Photomicrographs (hematoxylin-eosin stain; original magnification, x100) show necrosis from coagulation (C) in tumor and (D-F) in surrounding liver tissue, as evidenced by coagulated erythrocytes (arrow in C and E), nuclear pyknosis and cytoplasmic shrinkage (in C), intensive staining and shrinkage of hepatocytes (in D), and faint staining of cell nuclei and distended sinusoids congested with erythrocytes (in E and F). Black bar = 1 cm in A and B, 50 µm in C-F.

 
In experiments performed to verify the threshold temperature of enhancement at T1-weighted imaging, the lowest mean temperature in an ROI in which enhancement was observed on T1-weighted images was 57°C ± 3; the highest mean temperature in an ROI in which enhancement was not observed was 53°C ± 1 (Fig 6). The estimated threshold temperature of enhancement was thus 53°–57°C. In ROIs in which the mean temperature was greater than or equal to 57°C, the normalized signal intensity on T1-weighted images was more than 1 SD above the noise level.



View larger version (19K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6. Normalized signal intensity measured at T1-weighted imaging (500/15) after tissue cooling and presented as a function of maximum temperature measured at MR thermometry (40/19.7) during heating. Data points are means ± SDs of the measured values per sonication in a 3 x 3-voxel ROI at the center of the focal location. Signal intensity at T1-weighted imaging in the sonicated location was normalized with reference to two 5 x 5-voxel ROIs in nonsonicated liver. From these data, the threshold at which enhancement was detected at T1-weighted imaging was estimated to be 53°-57°C. The relatively low signal intensity at these sonicated locations compared with those observed overall may have been the result of the low dose of liposomal agent, the length of time between the injection and the sonications (23-59 minutes), or a relationship between the peak temperature or heating duration and the magnitude of enhancement at T1-weighted imaging. Note that the two sonications that produced the two highest temperatures were applied after enhancement began and were performed at higher power than earlier sonications.

 
Normalized signal intensity enhancement on T1-weighted images at the center of the lesions in normal tissue and in the enhanced rim surrounding the tumors ranged from 12% to 42% (mean, 26% ± 9). For all (18 of 18) locations in the liver, enhancement on T1-weighted images persisted after the tissue had cooled (Fig 7). Enhancement on T1-weighted images was observed throughout MR imaging performed after sonication (range, 5–119 minutes; mean, 51 minutes ± 34). The boundaries of the enhanced zones did not change during this time.



View larger version (25K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 7. Normalized signal intensity measured at T1-weighted imaging (500/15) after tissue cooling and presented as a function of time after liposomal contrast agent injection at four sonicated locations (each represented by a different symbol) in normal liver. At each time point, the mean signal intensity was measured in the center of the enhanced zones on T1-weighted images and normalized with reference to the signal intensity in normal (nonsonicated) liver tissue on the same image. In the interest of clarity, error bars are not shown. The mean error bar ± SD was 9% ± 2.

 
Findings in Kidney
The liposomal contrast agent did not behave as expected in the kidney (Fig 8). On T1-weighted images acquired after the tissue had cooled, the sonicated areas appeared as small enhanced zones surrounded by rings with lower signal intensity. The dimensions of these enhanced zones clearly did not match the isotherms on temperature maps from MR imaging during sonications. To investigate whether this mismatch had been caused by errors at MR thermometry, T1-weighted MR imaging was repeated with gadopentetate dimeglumine to enable detection of regions in which the tissue vasculature had been compromised in a way that indicated thermal damage (24). On images acquired within 2 minutes after gadopentetate dimeglumine injection, the dimensions of the nonperfused areas coincided with those of the isotherms. On images acquired with a longer delay after injection, the nonperfused zones were substantially reduced in size—a finding that indicated either a partially restored blood supply or leakage of gadopentetate dimeglumine into the sonicated region. The diameters of lesions resulting from sonication, measured from the nonenhanced areas on T1-weighted images acquired immediately after administration of gadopentetate dimeglumine, were 4–5 mm.



View larger version (48K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 8. A, Coronal map shows isotherms from MR thermometry performed during five sonications in one kidney (outer boundary of gray voxels = isotherms with maximum temperature >= 53°C, outer boundary of white voxels = isotherms with maximum temperature >= 57°C). B, Coronal T1-weighted (500/15) MR image acquired shortly after the fifth sonication shows enhanced zones (arrows) presumably caused by thermal activation of the liposomal agent that are smaller than expected on the basis of temperature measurements and that are surrounded by hypointense rims. C and D, Coronal T1-weighted fast spin-echo MR images of the same ROI, acquired approximately 2 minutes (C) and 7 minutes (D) after injection of gadopentetate dimeglumine. The dimensions of the nonenhanced areas in C match the isotherms fairly closely, but in D they have shrunk and are closer to the enhancing zones seen in B. All images were acquired in a plane perpendicular to the direction of the ultrasound beam.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES
 
The Gd-DTPA-BMA–bearing liposomes were designed to release their gadolinium content at the threshold temperature of thermally induced necrosis (ie, approximately 57°C) and to remain in the tissue in areas in which the temperature had increased to this level. This functionality was verified in vivo both in normal liver and in tumor-bearing liver, but not in kidney.

Many variables, including the dose of the liposomal contrast agent, the peak temperature and heating duration, and the length of time between contrast agent injection and heating, might have affected the magnitude of signal intensity enhancement at T1-weighted imaging. Nonetheless, throughout the range of variables used in this study, persistent enhancement was observed on T1-weighted images in the liver. Systematic variations in the contrast agent dose, time between contrast agent injection and heating, peak temperature, and heating duration would be necessary to accurately determine the effects of these variables on the performance of the liposomal agent. Such effects may be important for the monitoring of thermal therapies (eg, laser or radiofrequency ablation) in which thermal deposition is not as uniform as it is in focused ultrasound.

The dynamics of enhancement from the extracellular contrast agent (gadopentetate dimeglumine) given after kidney sonications may indicate that blood flow in the thermally damaged areas was partially intact and that the paramagnetic agent released from the liposomes was eliminated by the kidneys soon after sonications. Because the nonpersistence of the liposomal agent may indicate its limited usefulness in the kidney, these observed effects warrant further study.

Future investigations also should include measurement of the speed of signal intensity enhancement during heating and the duration of enhancement in the thermally ablated zone at T1-weighted imaging, as well as more detailed assessment of contrast agent dose requirements, pharmacokinetics, and toxicity. In addition, future studies should investigate whether the signal intensity enhancement due to the liposomal agent interferes with the ability of standard MR contrast agents to depict residual tumor after thermal treatment. Although it seems unlikely that the threshold temperature of enhancement is substantially tissue dependent, additional studies are needed to verify that small differences in threshold temperature do not exist among tissue types. The results of other preliminary investigations (14) suggest, however, that there is no gross difference between the threshold temperature in kidney and the Tc found in liver in the present study.

A shortcoming of liposomal Gd-DTPA-BMA is that it provides no warning when low-level heating approaches the threshold temperatures of tissue damage or of boiling, both of which present risks to patient safety during thermal therapy. Also, the use of this contrast agent is predicated on the assumption that a particular temperature is predictive of thermal tissue damage; with a sufficient heating duration, however, tissue damage may occur at temperatures substantially lower than 57°C (25). Because heating with focused ultrasound may be rapid and result in sharp thermal gradients, the damage to tissue may be only slightly underestimated; with longer exposure times, Tc may have to be lowered. Another shortcoming of this contrast agent was its failure to enhance signal intensity in the centers of tumors on T1-weighted images acquired after heating. The tumors in our study contained areas of spontaneous necrosis, however, so low signal intensity and lack of contrast enhancement in the tumor centers were to be expected. Even in such cases, liposomal Gd-DTPA-BMA enabled verification that the ablated region extended a sufficient distance beyond the tumor. Further testing of this agent is needed in tumors that do not contain necrotic areas.

Several other nonstandard contrast agents have been proposed for use in MR image–based guidance of thermal therapies (26–33). An advantage of liposomal Gd-DTPA-BMA is that Tc is linked with the liposome rather than the contrast agent (13), which allows for freedom of choice with regard to the threshold temperature. Additionally, the kidneys rapidly excrete many of the other nonstandard contrast agents, and it is unknown whether they all are usable in vivo. The experience gained from the substantial work that has been and is being done with thermosensitive liposomes for drug delivery (11) is another advantage of liposomal Gd-DTPA-BMA.

Although others have suggested that the temperature dependence of magnetic susceptibility might be increased in the presence of paramagnetic agents and might result in errors in MR thermometry (34), our study results indicate that the use of liposomal Gd-DTPA-BMA did not greatly affect MR temperature measurements.

Practical application: Liposomal Gd-DTPA-BMA offers a less motion-sensitive means for monitoring thermal therapies, compared with current MR thermometry techniques. There have been several clinical studies of MR image–guided thermal therapy of various targets in the abdomen, where motion can be a substantial problem (5,6,35–37). The effectiveness of treatment in abdominal locations may be improved with use of this agent. The persistence of signal intensity enhancement at T1-weighted imaging after the tissue has cooled may enable improved targeting and guidance of thermal ablation with focused ultrasound, in particular, because the tumor and the treated area can be tracked more accurately between multiple breath-holds and between multiple sonications.


   ACKNOWLEDGMENTS
 
The authors thank Randy King for his help with the experiments.


   FOOTNOTES
 
Abbreviations: BMA = bis(methylamide), DTPA = diethylenetriaminepentaacetic acid, ROI = region of interest

Author contributions: Guarantor of integrity of entire study, K.H.; study concepts and design, N.M., S.L.F., H.R., K.H.; literature research, N.M.; experimental studies, N.M., S.L.F., H.R., H.M., K.H.; data acquisition, N.M., S.L.F., H.R., H.M., K.H.; data analysis/interpretation, N.M., N.V., K.H.; statistical analysis, N.M.; manuscript preparation, N.M.; manuscript definition of intellectual content, N.M., K.H.; manuscript editing and revision/review, all authors; manuscript final version approval, N.M., K.H.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
 REFERENCES
 

  1. Hynynen K, Pomeroy O, Smith DN, et al. MR imaging-guided focused ultrasound surgery of fibroadenomas in the breast: a feasibility study. Radiology 2001; 219:176-185.[Abstract/Free Full Text]
  2. Huber PE, Jenne JW, Rastert R, et al. A new noninvasive approach in breast cancer therapy using magnetic resonance imaging-guided focused ultrasound surgery. Cancer Res 2001; 61:8441-8447.[Abstract/Free Full Text]
  3. Kahn T, Harth T, Kiwit JC, Schwarzmaier HJ, Wald C, Modder U. In vivo MRI thermometry using a phase-sensitive sequence: preliminary experience during MRI-guided laser-induced interstitial thermotherapy of brain tumors. J Magn Reson Imaging 1998; 8:160-164.[Medline]
  4. Mueller-Lisse UG, Heuck AF, Thoma M, et al. Predictability of the size of laser-induced lesions in T1-weighted MR images obtained during interstitial laser-induced thermotherapy of benign prostatic hyperplasia. J Magn Reson Imaging 1998; 8:31-39.[Medline]
  5. Vogl TJ, Mack MG, Muller PK, Straub R, Engelmann K, Eichler K. Interventional MR: interstitial therapy. Eur Radiol 1999; 9:1479-1487.[CrossRef][Medline]
  6. Lewin JS, Connell CF, Duerk JL, et al. Interactive MRI-guided radiofrequency interstitial thermal ablation of abdominal tumors: clinical trial for evaluation of safety and feasibility. J Magn Reson Imaging 1998; 8:40-47.[Medline]
  7. Chen JC, Moriarty JA, Derbyshire JA, et al. Prostate cancer: MR imaging and thermometry during microwave thermal ablation—initial experience. Radiology 2000; 214:290-297.[Abstract/Free Full Text]
  8. Graham SJ, Chen L, Leitch M, et al. Quantifying tissue damage due to focused ultrasound heating observed by MRI. Magn Reson Med 1999; 41:321-328.[CrossRef][Medline]
  9. McDannold NJ, King RL, Jolesz FA, Hynynen K. Usefulness of MR imaging-derived thermometry and dosimetry in determining the threshold for tissue damage induced by thermal surgery in rabbits. Radiology 2000; 216:517-523.[Abstract/Free Full Text]
  10. Hazle JD, Stafford RJ, Price RE. Magnetic resonance imaging-guided focused ultrasound thermal therapy in experimental animal models: correlation of ablation volumes with pathology in rabbit muscle and VX2 tumors. J Magn Reson Imaging 2002; 15:185-194.[CrossRef][Medline]
  11. Needham D, Dewhirst MW. The development and testing of a new temperature-sensitive drug delivery system for the treatment of solid tumors. Adv Drug Deliv Rev 2001; 53:285-305.[CrossRef][Medline]
  12. Kono K. Thermosensitive polymer-modified liposomes. Adv Drug Deliv Rev 2001; 53:307-319.[CrossRef][Medline]
  13. Fossheim SL, Il’yasov KA, Hennig J, Bjornerud A. Thermosensitive paramagnetic liposomes for temperature control during MR imaging-guided hyperthermia: in vitro feasibility studies. Acad Radiol 2000; 7:1107-1115.[CrossRef][Medline]
  14. Il’yasov KA, Bjornerud A, Rogstad A, et al. Paramagnetic liposomes as thermosensitive probes for MRI-guided thermal ablation: feasibility study on the perfused porcine kidney (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; 324.
  15. Cain CA, Umemura SI. Concentric-ring and sector-vortex phased-array applicators for ultrasound hyperthermia. IEEE Trans Microw Theory Tech 1986; 34:542-551.[CrossRef]
  16. Daum DR, Buchanan MT, Fjield T, Hynynen K. Design and evaluation of a feedback based phased array system for ultrasound surgery. IEEE Trans Ultrason Ferroelectr Freq Contr 1998; 45:431-438.[Medline]
  17. Hynynen K, Freund WR, Cline HE, et al. A clinical, noninvasive, MR imaging-monitored ultrasound surgery method. RadioGraphics 1996; 16:185-195.[Abstract/Free Full Text]
  18. Fossheim SL, Fahlvik AK, Klaveness J, Muller RN. Paramagnetic liposomes as MRI contrast agents: influence of liposomal physicochemical properties on the in vitro relaxivity. Magn Reson Imaging 1999; 17:83-89.[CrossRef][Medline]
  19. Hindman JC. Proton resonance shift of water in the gas and liquid states. J Chem Phys 1966; 44:4582-4592.[CrossRef]
  20. Ishihara Y, Calderon A, Watanabe H, Okamoto K, Suzuki Y, Kuroda K. A precise and fast temperature mapping using water proton chemical shift. Magn Reson Med 1995; 34:814-823.[Medline]
  21. Chung AH, Hynynen K, Colucci V, Oshio K, Cline HE, Jolesz FA. Optimization of spoiled gradient-echo phase imaging for in vivo localization of a focused ultrasound beam. Magn Reson Med 1996; 36:745-752.[Medline]
  22. Peters RD, Hinks RS, Henkelman RM. Ex vivo tissue-type independence in proton-resonance frequency shift MR thermometry. Magn Reson Med 1998; 40:454-459.[Medline]
  23. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1:307-310.[CrossRef][Medline]
  24. Hynynen K, Darkazanli A, Damianou CA, Unger E, Schenck JF. The usefulness of a contrast agent and gradient-recalled acquisition in a steady-state imaging sequence for magnetic resonance imaging-guided noninvasive ultrasound surgery. Invest Radiol 1994; 29:897-903.[CrossRef][Medline]
  25. Meshorer A, Prionas SD, Fajardo LF, Meyer JL, Hahn GM, Martinez AA. The effects of hyperthermia on normal mesenchymal tissues: application of a histologic grading system. Arch Pathol Lab Med 1983; 107:328-334.[Medline]
  26. Webb AG, Wong M, Niesman M, et al. In-vivo NMR thermometry with liposomes containing 59Co complexes. Int J Hyperthermia 1995; 11:821-827.[Medline]
  27. Aime S, Botta M, Fasano M, et al. A new ytterbium chelate as contrast agent in chemical shift imaging and temperature sensitive probe for MR spectroscopy. Magn Reson Med 1996; 35:648-651.[Medline]
  28. Frenzel T, Roth K, Kossler S, et al. Noninvasive temperature measurement in vivo using a temperature-sensitive lanthanide complex and 1H magnetic resonance spectroscopy. Magn Reson Med 1996; 35:364-369.[Medline]
  29. Hentschel M, Wust P, Wlodarczyk W, et al. Non-invasive MR thermometry by 2D spectroscopic imaging of the Pr[MOE-DO3A] complex. Int J Hyperthermia 1998; 14:479-493.[Medline]
  30. Zuo CS, Mahmood A, Sherry AD. TmDOTA-: a sensitive probe for MR thermometry in vivo. J Magn Reson 2001; 151:101-106.[CrossRef][Medline]
  31. Franklin KJ, Buist RJ, den Hartog J, McRae GA, Spencer DP. Encapsulated liquid crystals as probes for remote thermometry. Int J Hyperthermia 1992; 8:253-262.[Medline]
  32. Lam MM, Webb AG. Proton-based phase-transition compounds for MRI thermometry (abstr) In: Proceedings of the Fourth Meeting of the International Society for Magnetic Resonance in Medicine. Berkeley, Calif: International Society for Magnetic Resonance in Medicine, 1996; 1709.
  33. Webb AG, Wong M, Kolbeck KJ, Magin R, Suslick KS. Sonochemically produced fluorocarbon microspheres: a new class of magnetic resonance imaging agent. J Magn Reson Imaging 1996; 6:675-683.[Medline]
  34. De Poorter J. Noninvasive MRI thermometry with the proton resonance frequency method: study of susceptibility effects. Magn Reson Med 1995; 34:359-367.[Medline]
  35. Bremer C, Allkemper T, Menzel J, Sulkowski U, Rummeny E, Reimer P. Preliminary clinical experience with laser-induced interstitial thermotherapy in patients with hepatocellular carcinoma. J Magn Reson Imaging 1998; 8:235-239.[Medline]
  36. de Jode MG, Lamb GM, Thomas HC, Taylor-Robinson SD, Gedroyc WM. MRI guidance of infra-red laser liver tumour ablations, utilising an open MRI configuration system: technique and early progress. J Hepatol 1999; 31:347-353.[CrossRef][Medline]
  37. Wacker FK, Reither K, Ritz JP, Roggan A, Germer CT, Wolf KJ. MR-guided interstitial laser-induced thermotherapy of hepatic metastasis combined with arterial blood flow reduction: technique and first clinical results in an open MR system. J Magn Reson Imaging 2001; 13:31-36.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 ScienceHome page
G. Galiana, R. T. Branca, E. R. Jenista, and W. S. Warren
Accurate Temperature Imaging Based on Intermolecular Coherences in Magnetic Resonance
Science, October 17, 2008; 322(5900): 421 - 424.
[Abstract] [Full Text] [PDF]

Home page
 Clin. Cancer Res.Home page
S. Dromi, V. Frenkel, A. Luk, B. Traughber, M. Angstadt, M. Bur, J. Poff, J. Xie, S. K. Libutti, K. C.P. Li, et al.
Pulsed-High Intensity Focused Ultrasound and Low Temperature Sensitive Liposomes for Enhanced Targeted Drug Delivery and Antitumor Effect
Clin. Cancer Res., May 1, 2007; 13(9): 2722 - 2727.
[Abstract] [Full Text] [PDF]

Home page
 JNCI J Natl Cancer InstHome page
A. M. Ponce, B. L. Viglianti, D. Yu, P. S. Yarmolenko, C. R. Michelich, J. Woo, M. B. Bally, and M. W. Dewhirst
Magnetic Resonance Imaging of Temperature-Sensitive Liposome Release: Drug Dose Painting and Antitumor Effects
J Natl Cancer Inst, January 3, 2007; 99(1): 53 - 63.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2303021713v1
230/3/743    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by McDannold, N.
Right arrow Articles by Hynynen, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by McDannold, N.
Right arrow Articles by Hynynen, K.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE