Hyperthermia in Bone Generated with MR Imaging controlled Focused Ultrasound: Control Strategies and Drug Delivery 1

Transcription

1 Note: This copy is for your personal non-commercial use only. To order presentation-ready copies for distribution to your colleagues or clients, contact us at Robert Staruch, BASc Rajiv Chopra, PhD Kullervo Hynynen, PhD Hyperthermia in Bone Generated with MR Imaging controlled Focused Ultrasound: Control Strategies and Drug Delivery 1 Purpose: To evaluate the feasibility of achieving image-guided drug delivery in bone by using magnetic resonance (MR) imaging controlled focused ultrasound hyperthermia and temperature-sensitive liposomes. Original Research n Experimental Studies 1 From the Centre for Research in Image-Guided Therapeutics, Sunnybrook Health Sciences Centre, 2075 Bayview Ave, Room C713, Toronto, ON, Canada M4N 3M5 (R.S., R.C., K.H.); and Department of Medical Biophysics, University of Toronto, Toronto, Ontario, Canada (R.C., K.H.). Received June 14, 2011; revision requested August 5; revision received August 26; accepted September 29; final version accepted November 20. Supported by a Terry Fox Foundation New Frontiers Program project grant from the National Cancer Institute of Canada and by an Ontario Research Fund grant from the Government of Ontario, Canada. R.S. supported by a National Sciences and Engineering Research Council Alexander Graham Bell Canada Graduate Scholarship. Address correspondence to R.S. ( staruchr@sri.utoronto.ca). Materials and Methods: Results: Conclusion: Experiments were approved by the institutional animal care committee. Hyperthermia (43 C, 20 minutes) was generated in 10-mm-diameter regions at a muscle-bone interface in nine rabbit thighs by using focused ultrasound under closed-loop temperature control with MR thermometry. Thermosensitive liposomal doxorubicin was administered systemically during heating. Heating uniformity and drug delivery were evaluated for control strategies with the temperature control image centered 10 mm (four rabbits) or 0 mm (five rabbits) from the bone. Simulations estimated temperature elevations in bone. Drug delivery was quantified by using the fluorescence of doxorubicin extracted from bone marrow and muscle and was compared between treated and untreated thighs by using the one-sided Wilcoxon signed rank test. With ultrasound focus and MR temperature control plane 0 mm and 10 mm from the bone interface, average target region temperatures were 43.1 C and 43.3 C, respectively; numerically estimated bone temperatures were 46.8 C and 78.1 C. The 10-mm offset resulted in thermal ablation; numerically estimated muscle temperature was 66.1 C at the bone interface. Significant increases in doxorubicin concentration occurred in heated versus unheated marrow (8.2-fold, P =.002) and muscle (16.8- fold, P =.002). Enhancement occurred for 0- and 10-mm offsets, which suggests localized drug delivery in bone is possible with both hyperthermia and thermal ablation. MR imaging controlled focused ultrasound can achieve localized hyperthermia in bone for image-guided drug delivery in bone with temperature-sensitive drug carriers. q RSNA, 2012 Supplemental material: /suppl/doi: /radiol /-/dc1 q RSNA, 2012 Radiology: Volume 263: Number 1 April 2012 n radiology.rsna.org 117

2 Bone metastases occur in approximately 70% of patients with advanced breast and prostate cancer (1). Their sequelae impair a patient s quality of life with pathologic fractures, spinal cord compression, hypercalcemia, and debilitating pain. For localized pain related to skeletal metastases, palliative radiation therapy is the standard of care, but 20% 30% of patients do not experience adequate relief (2). For some patients, bisphosphonate therapy has shown promise in breaking the vicious cycle of tumor progression and osteoclast-mediated bone resorption, which decreases and delays the incidence of skeletal complications and pain (3,4). In patients with intractable pain refractory to radiation therapy and bisphosphonates or systemic chemotherapy, thermal ablation of osteolytic bone metastases by using percutaneous radiofrequency ablation with ultrasound or computed tomographic image guidance achieves pain relief by ablating periosteal nerve endings and reducing tumor burden (5,6). Thermal ablation by using magnetic resonance (MR) imaging guided focused ultrasound demonstrated effective palliation of pain arising from both osteolytic and osteoblastic bone metastases in patients who had exhausted all other treatment options (7 9). Focused ultrasound has several advantages over radiofrequency ablation, most important of which is noninvasive heating with the ability to cover larger lesions. One strategy to improve the efficacy of thermal therapy is the combined administration of heat with anticancer drugs. Several chemotherapeutic Advance in Knowledge nn Local drug deposition enhancements in image-targeted regions of bone marrow (8.2-fold) and surrounding soft tissue (16.8- fold) were achieved noninvasively by combining focused ultrasound hyperthermia and systemic administration of temperature-sensitive drug carriers. agents show enhanced cytotoxicity when combined with mild heating and can be encapsulated in long-circulating nanoparticle drug carriers such as PEGylated liposomes that accumulate preferentially in tumors, an effect that is enhanced by mild heating (10). Temperature-sensitive liposome formulations release their contents when heated to temperatures greater than 40 C for as little as seconds (11), which provides an immediate cytotoxic effect in sublethally heated treatment margins (12). The efficacy of radiofrequency ablation combined with thermosensitive liposomes in curative treatment of hepatocellular carcinoma is being evaluated in phase III trials against radiofrequency ablation alone (13), and MR imaging controlled focused ultrasound has been used to achieve localized drug delivery with thermosensitive liposomes in vivo (14). In patients with painful bone metastases, combining localized drug release with thermal pain palliation could increase cell kill at treatment margins and decrease the required energy to achieve a therapeutic effect, potentially reducing treatment time and risk of normal tissue damage. Localized drug release mediated by using hyperthermia, rather than high-temperature thermal ablation, may reduce pain related to treatment and possibly achieve local tumor control without compromising skeletal stability. This noninvasive localized therapy would depend on the ability to maintain desired temperatures in targeted regions, with minimal heating of surrounding tissue. Focused ultrasound under closedloop feedback control based on MR thermometry has been used to achieve precise noninvasive heating in soft tissue (15 17), and MR imaging has been used to target and evaluate the effects of focused ultrasound on bone (18,19). Temperature control in bone is made difficult by the inability of proton-resonance frequency shift MR thermometry to help measure temperature in cortical bone. In this study, a feedback control system that incorporates proton-resonance frequency thermometry to maintain desired temperatures in bone on the basis of temperatures measured in adjacent soft tissue was proposed. Its ability to generate localized hyperthermia in bone for localized drug delivery was evaluated. The purpose of this study was to evaluate the feasibility of achieving image-guided drug delivery in bone by using MR imaging controlled focused ultrasound hyperthermia and temperature-sensitive liposomes. Materials and Methods Lyso-thermosensitive liposomal doxorubicin (Thermodox; Celsion, Lawrenceville, NJ) was provided by the manufacturer. The authors alone had control of the data and the information submitted for publication. Animals Experiments in this study were approved by the Animal Care Committee at Sunnybrook Research Institute. Nine male New Zealand white rabbits weighing 3 4 kg (Charles River Laboratories, Sherbrooke, Québec, Canada) were anesthetized by using intramuscular injection of ketamine (50 mg per kilogram of body weight per hour) and xylazine (10 mg/kg/h). Anesthetized rabbits had one ear vein cannulated and both thighs depilated and were placed on a stage above the degassed water tank of an MR imaging compatible focused ultrasound system (Fig 1). During hyperthermia, rectal temperature was Published online before print /radiol Content codes: Radiology 2012; 263: Author contributions: Guarantors of integrity of entire study, R.S., K.H.; study concepts/study design or data acquisition or data analysis/ interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, R.S.; experimental studies, all authors; statistical analysis, R.S., R.C.; and manuscript editing, all authors Potential conflicts of interest are listed at the end of this article. 118 radiology.rsna.org n Radiology: Volume 263: Number 1 April 2012

3 Figure 1 Figure 1: Experimental setup for MR imaging controlled focused ultrasound heating in bone. A, Axial localizer MR image along the ultrasound beam path shows the location of the three coronal planes (1, 2, 3) used for MR thermometry, two of which are shown in B and C. B, Coronal T2-weighted fast spin-echo MR image (repetition time msec/echo time msec, 2000/75) in muscle near the bone interface, in which MR thermometry was used to control treatment. C, Coronal T2-weighted MR image through the bone, in which MR imaging thermometry was used to detect excessive heating adjacent to the bone. ROIs = regions of interest. monitored by using a fiberoptic temperature probe (3100; Luxtron, Santa Clara, Calif) and was maintained by manually regulating the temperatures of a hot water blanket covering the animal and the degassed water reservoir below. Focused Ultrasound System Acoustic energy was delivered by using continuous sonication with a spherically focused air-backed ultrasound transducer (fundamental frequency, MHz; curvature radius, 10 cm; aperture diameter, 5 cm) driven at its fifth harmonic (2.787 MHz) by using an arbitrary waveform generator (33250A; Agilent, Mississauga, Ontario, Canada) and radiofrequency amplifier (NP2912; NP Technologies, Newbury Park, Calif). Electrical power was measured with a power meter (438A; Hewlett Packard, Palo Alto, Calif) and a dual directional coupler (C1373; Werlatone, Brewster, NY). The transducer was impedancematched to the amplifier with a passive matching circuit. At the driving frequency, emitted acoustic power was approximately 70% of the applied electrical power (forward minus reflected) when measured with a radiation force technique (20). This efficiency was used to calculate the electric power required to achieve a desired acoustic power during sonication. The transducer was incorporated into an MR imaging compatible positioning system similar in function to that described in Chopra et al (21) and was programmed to move along a 10-mm-diameter circular trajectory at a speed of one revolution per second for simultaneous ultrasound heating and MR thermometry. Degassed water coupled the transducer through a window of 25-mm polyimide film (Kapton; Du- Pont, Wilmington, Del) into the target thigh. The thigh was positioned such that the ultrasound beam penetrated the skin at approximately normal incidence and was focused at the musclebone interface. To prevent undesired tissue heating in the contralateral thigh, a saline bag and a polyurethane rubber acoustic absorber (AptFlex F28; Precision Acoustics, Dorchester, England) were placed between the rabbit s thighs (22). MR Imaging controlled Focused Ultrasound Hyperthermia Experiments were performed with the positioning system placed in a clinical 1.5-T MR imaging unit (Signa; GE Healthcare, Waukesha, Wis). MR imaging controlled focused ultrasound was used to heat 10-mm-diameter regions of thigh muscle near the muscle-bone interface to 43 C for 20 minutes. This temperature elevation and exposure duration were chosen to be sufficient to trigger drug release from temperature-sensitive liposomes (11) without causing substantial tissue damage (23). A single-loop receive coil with a square opening of 85 mm was placed underneath the animal around the polyimide film window to improve the signal-to-noise ratio in the heated region. Before heating, anatomic images were acquired with T1 weighting and T2 weighting (Table 1) for target definition and verification of acoustic coupling. These images were reacquired after heating to evaluate changes in tissue perfusion and/or tissue damage caused by ultrasound heating, which Radiology: Volume 263: Number 1 April 2012 n radiology.rsna.org 119

4 were identified as regions of either increased signal intensity or decreased signal intensity with a gadolinium-enhanced rim. Postsonication T1-weighted images were acquired before and 1 minute after bolus injection of MR imaging contrast agent (0.2 mmol/kg gadodiamide, Omniscan; GE Healthcare, Mississauga, Ontario, Canada) in the ear vein. During mechanical scanning of the ultrasound transducer, coronal fast spoiled gradient-echo images were continuously acquired in three planes: at the location of the bone, in the muscle directly beneath the bone, and in the muscle closer to the skin (Fig 1). When new images were available (every 5 seconds), a real-time acquisition interface (24,25) transferred k- space data to a control computer for image reconstruction and calculation of spatial temperature distribution by using software written in Matlab (MathWorks, Natick, Mass). Phasedifference maps were calculated by using complex phase subtraction between treatment images (26) and the average of five baseline images acquired during transducer motion prior to heating. Temperature maps were obtained from phase-difference maps by using a proton-resonance frequency shift coefficient of ppm per degree Celsius (26) and adding the baseline temperature measured by using a fiberoptic probe in the rectum. Temperatures averaged across a mm 2 image region in a mineral oil reference phantom were subtracted to account for magnetic field drift during the experiment (26). Temperature images were then averaged across a 30-second sliding window, which was shown previously to be effective in reducing the effect of periodic susceptibility-related phase shifts caused by circular transducer motion during imaging (14). Spatial temperature elevations in muscle just below the bone interface were controlled by using eight independent proportional-integral feedback controllers along the scanned heating trajectory, rapidly switching the applied power as the transducer scanned along the circular trajectory, and updating controller outputs after each image acquisition according to the following equation ( ) n ( []) [ ] = - [ ] P n K T T n 1,...8 P goal 1, K T - T i. I goal 1,...8 i=1 For temperature image n, the difference between the temperatures T 1,...8 in the eight predefined 1-mm 2 control regions and the target temperature T goal, as well as the integral of the difference over each previous image i, was scaled by proportional and integral gains, K P and K I, to specify the acoustic powers, P 1,...8, delivered by the transducer as the focus crossed each region. The average lag time for data transfer, reconstruction, and controller update was less than 1 second. Two control strategies were investigated. In the first, the ultrasound focus and the image plane used for temperature control were set in muscle 10 mm below the bone interface (10-mm offset) by using gain settings reported previously for controlled heating in muscle (K P = 4.5, K I = 0.03) (14). In the second, the control plane was defined immediately under the bone with the focus centered at the bone interface (0-mm offset); gains were selected by using simulations of scanned focused ultrasound heating (K P = 3.0, K I = 0.0) (Appendix E1 [online]). For each strategy, simulations were also used to estimate temperature elevations in bone and marrow. To evaluate the temporal and spatial uniformity of heating, the mean, T 90 (temperature that 90% of the target exceeds), and T 10 (temperature that 10% of the target exceeds) temperatures were measured at each time point in circular image regions matching the intended target diameter. The steadystate mean and standard deviation of each parameter was calculated across all treatment images starting once T goal reached 95% of the final target temperature elevation. Thermal dose in the target region was calculated in cumulative equivalent minutes at 43 C by using the Sapareto and Dewey timetemperature equation (27). Drug Administration and Tissue Harvesting Rabbits were administered a doxorubicin dose of 2.5 mg/kg diluted as equal parts lyso-thermosensitive liposomal doxorubicin and 5% dextrose sterile solution, infused at 1.2 ml/min into the ear vein during hyperthermia, starting when the average temperature in the target region reached 43 C. One leg was heated in each rabbit; the other served as the control. In four rabbits, the focus and the image plane used for the control were offset 10 mm from the femur; in five rabbits, a 0-mm offset was used. Two hours after lyso-thermosensitive liposomal doxorubicin infusion, unabsorbed liposomes were flushed from the vasculature by using transcardiac saline perfusion before collection of tissue samples for evaluation of drug release. After sacrifice, 10-mm-thick axial slices of the femur with its marrow, as well as mg samples of adjacent muscle, were harvested from the targeted region of the treated leg and matching regions of the untreated leg, snap frozen in liquid nitrogen, and stored at 280 C. Analysis of Drug Concentration in Bulk Tissue Samples Tissue doxorubicin concentrations were measured by using the fluorescence intensity of doxorubicin extracted from homogenized tissue samples (28). Tissue samples were weighed and added to 20 volumes of acidified ethanol extraction solvent (0.3N hydrochloric acid in 50% ethanol) before homogenization with a tissue grinder (PYREX Ten Broeck; Corning, Corning, NY), overnight refrigeration in acidified ethanol solvent, centrifugation ( g, 30 minutes), and storage of supernatants in the dark at 220 C. A total of 1.5 ml of acidified ethanol was added to 0.5 ml of supernatant in 3-mL fluorometry cuvettes, and fluorescence intensity was measured by using a benchtop fluorometer (VersaFluor; Bio-Rad 120 radiology.rsna.org n Radiology: Volume 263: Number 1 April 2012

5 Table 1 MR Imaging Parameters Sequence Repetition Time (msec) Echo Time (msec) Flip Angle (degree) Echo Train Length No. of Signals Acquired Field of View (cm) Matrix Size Section Thickness (mm) Bandwidth (khz) Imaging Plane T2-weighted fast spin echo T1-weighted fast spin echo Fast spoiled gradient echo Coronal, axial Coronal, axial NA Coronal Note. All MR imaging examinations were performed at 1.5 T (Signa; GE Healthcare). T2- and T1-weighted fast spin-echo images were acquired for treatment planning, with no phase wrap. Fast spoiled gradient-echo images were acquired for thermometry. NA = not applicable. Laboratories, Hercules, Calif) with 480-nm excitation and 590-nm emission filters. Relative fluorescence intensities were scaled to doxorubicin concentrations by using a fluorescence calibration curve of a serial dilution of free doxorubicin (Doxorubicin HCl; Teva Novopharm, Toronto, Ontario, Canada) added to 0.5-mL blank tissue homogenates in 1.5 ml of acidified ethanol. Statistical Analysis Descriptive statistics were calculated by using means 6 standard deviations. The one-sided Wilcoxon signed rank test was used to identify drug concentration enhancements in heated versus unheated tissues taken from the same rabbit for samples of marrow and muscle at the bone interface (GraphPad Prism 5.0; GraphPad Software, La Jolla, Calif). Differences were considered statistically significant for values of P less than.05. Results MR Imaging controlled Focused Ultrasound Hyperthermia Figure 2, A and B, shows temperature distributions in the same coronal imaging planes depicted in Figure 1, B and C, with temperature greater than 37 C overlaid on the corresponding magnitude image, at a selected time of 10 minutes after heating began. In Figure 2, B, pixels in bone where signal-to-noise ratio was too low to produce accurate temperature measurements were masked out for display. In this experiment, the focus was set at the muscle-bone interface, with the top of the 5-mm-thick imaging plane used for temperature control set just below the bone. Temperature evolution within the 10-mm targeted region of the control plane is shown in Figure 3, as well as the power applied at each control point over the course of time. At each time point, the mean, T 90, and T 10 temperatures are shown, as well as the mean temperatures in unheated regions mm from the edge of the circular target, the controller input function T goal, and the start and end times of lyso-thermosensitive liposomal doxorubicin infusion. Figure 4 shows the radial mean 6 standard deviation of the steady-state temperature distribution and median thermal dose, centered on the target region, for the control plane and the bone. Hyperthermia experiments are summarized in Table 2, showing average steady-state mean, T 90, and T 10 temperatures of 43.2 C, 42.0 C, and 44.4 C, with standard deviations of 0.4 C across the experiment duration. Numerical Simulations of Temperature Elevations in Bone during MR Imaging controlled Hyperthermia Temperature elevations in bone caused by scanned focused ultrasound heating with beam offsets of 0 and 10 mm from the bone were estimated by using numerical simulations (Appendix E1 [online]) across the tissue volume depicted in Figure 5, A. Calculated temperatures were averaged across the image planes shown. Figure 5, B, shows coronal temperature distributions in the control plane for the 0- and 10-mm offset scenarios, both with average control region temperatures near 43 C, similar to in vivo results. The respective sagittal images in Figure 5, B, show higher temperatures in bone and adjacent soft tissues with the 10-mm offset. The median, T 90, and T 10 temperatures within 10-mm-diameter regions centered around the beam axis in cortical bone, marrow, and 5-mm-thick regions of muscle centered at depths of 2.5 and 12.5 mm below the bone are summarized in Table 3, showing bone temperature elevations 4.2 to 4.7 times higher than the control region for the 10- mm offset case, as opposed to 1.1 to 1.8 times higher for a 0-mm offset. These elevated temperatures in the 10-mm offset case resulted in thermal damage identified on posttreatment T2-weighted and contrast material enhanced T1-weighted images. Temperature variability in cortical bone results from high temperatures along the scan trajectory where applied power is dynamically modulated versus the central regions heated by conduction, as shown in simulated radial temperature distributions in Figure 6. Doxorubicin Concentrations in Bone Marrow and Muscle Table 4 shows the doxorubicin concentrations measured in tissue samples Radiology: Volume 263: Number 1 April 2012 n radiology.rsna.org 121

6 Figure 2 Figure 2: Snapshot of temperature distribution from controlled hyperthermia at muscle-bone interface in rabbit thigh by using MR imaging controlled focused ultrasound. Control regions of interest, circular unheated region of interest, and temperatures greater than 37 C are overlaid on coronal fast spoiled gradient-echo magnitude images (38.6/10, 30 flip angle) acquired, A, in muscle used for feedback control, and, B, in bone, with low signal bone pixels masked. Images were acquired 10 minutes after start of heating. FOV = field of view. from each rabbit, collected from the marrow and adjacent muscle in both the targeted region of the treated thigh and equivalent regions in the untreated contralateral thigh. Overall, the per-animal increases of doxorubicin concentration in heated versus unheated marrow and muscle were times and times, respectively. Figure 7a shows the mean 6 standard deviation doxorubicin concentrations for each tissue type across all experiments; paired differences were statistically significant for each tissue type (marrow, P =.002; muscle, P =.002; one-sided Wilcoxon signed rank test). Figure 7b shows doxorubicin concentration data divided into experiments with the control plane offset mm versus 0 mm from the bone. With the control plane at the bone interface, increases of doxorubicin concentrations in heated versus unheated marrow and muscle were times (P =.03) and times (P =.03) higher, respectively. Enhancements were seen when the control plane was set mm from the bone interface but did not reach statistical significance, with increases of times (P =.06) for marrow and times (P =.06) for muscle. Discussion In our study, image-guided localized drug delivery in bone was achieved noninvasively by using MR imaging controlled focused ultrasound heating and temperature-sensitive liposomes. Temperatures of 43 C were generated in a 10-mm-diameter circular region 122 radiology.rsna.org n Radiology: Volume 263: Number 1 April 2012

7 Figure 3 Figure 3: Temporal evolution of controlled hyperthermia at muscle-bone interface in rabbit thigh by using MR imaging controlled focused ultrasound for the same experiment shown in Figure 2. Top: Graph shows mean, T 90, and T 10 temperatures measured in the 10-mm-diameter target region in the plane used for temperature control. Liposome infusion duration is shaded. Bottom: Acoustic power applied at each of the eight control regions at each time step. LTLD = lyso-thermosensitive liposomal doxorubicin. Figure 4 Figure 4: Graphs show radial distribution of median thermal dose and temporal mean 6 standard deviation (SD) steady-state temperature in (a) control and (b) bone planes for the same experiment shown in Figures 2 and 3. Bone pixels are masked, and the unheated region of interest is shaded. CEM43 = cumulative equivalent minutes at 43 C. at a bone interface by using scanned focused ultrasound and were maintained for 20 minutes by controlling ultrasound power based on MR imaging temperature measurements in adjacent soft tissue. Administration of lyso-thermosensitive liposomal doxorubicin during heating resulted in locally enhanced drug deposition in bone marrow and muscle adjacent to the bone surface. The proposed method controls localized bone heating in an automatic feedbackcontrol loop by using proton-resonance frequency shift MR thermometry in adjacent soft tissue. The proton-resonance frequency shift technique is based on temperature-dependent changes in the local magnetic field around water protons and is of limited use in tissues with low water content such as bone, fat, and lung. Phase subtraction is also sensitive to tissue displacement in the image plane and magnetic field distortions related to tissue motion near the image plane, which limits the proposed method to targets immediately adjacent to nonmoving aqueous tissue where stable thermometry can be achieved. Sprinkhuizen et al (29) have demonstrated that temperature dependent magnetic susceptibility changes in fat ( ppm per degree Celsius) influence the proton-resonance frequency shift field changes in nearby water; we calculated that the temperaturedependence of the susceptibility of bone is much smaller ( ppm per degree Celsius for a susceptibility of ppm [30] and linear thermal expansion coefficient of per degree Celsius [31]) and will not affect the field of neighboring soft tissue during heating. With externally focused ultrasound heating, this method is limited to targets with an acoustic window through soft tissue. At high frequencies (3 MHz), ultrasound-generated temperature elevations occur close to the bone interface, and temperatures in cortical bone can be approximated by using temperatures measured in adjacent soft tissue. This was confirmed by our simulations at MHz; simulated temperatures in muscle also demonstrated good agreement with in vivo MR temperature measurements. At lower frequencies (1 MHz), peak temperatures in bone will exceed those of adjacent soft tissue and occur further from the bone interface (32,33), which may be useful when heating thick layers of cortical bone in osteoblastic lesions. MR imaging controlled bone heating with other energy sources, including invasive radiofrequency applicators, would be limited by interference between electromagnetic heating devices and MR imaging and by poor electrical conductivity (and thus heating) in bone (34). Previous preclinical studies that used thermosensitive liposomes have demonstrated 10- to 20-fold increases in drug concentration in tissue samples harvested from animal tumors and muscle Radiology: Volume 263: Number 1 April 2012 n radiology.rsna.org 123

8 Table 2 Summary of In Vivo MR Imaging controlled Focused Ultrasound Hyperthermia Experiments with the Focus Set at Two Different Offsets from a Muscle-Bone Interface in Rabbit Thigh Parameter Target Radius (mm) Temperature ( C) T 90 ( C) T 10 ( C) Radius at 42 C for Muscle (mm) Radius at 41 C for Bone (mm) Control Plane mm Away from Bone Interface (n = 4) Thermal Dose for Muscle (CEM43) Thermal Dose for Bone (CEM43) Rabbit Rabbit Rabbit Rabbit Mean Control Plane at Bone Interface (n = 5) Rabbit Rabbit Rabbit Rabbit Rabbit Mean Note. Unless otherwise indicated, data are means 6 standard deviations. Temperature (mean across 10-mm-diameter target region), T 90 (temperature that 90% of target region exceeds), and T 10 (temperature that 10% of target region exceeds) are reported as the average across all treatment temperature images starting once the target temperature reached 95% of the desired temperature elevation. CEM43 = cumulative equivalent minutes at 43 C. Figure 5 Figure 5: Numerical simulations of MR imaging controlled focused ultrasound hyperthermia at a muscle-bone interface. A, Simulation volume, including layers of water, muscle, cortical bone (CB), and trabecular bone (TB), by using tissue properties listed in Table E1 (online). B, Left: Average steady-state temperatures for experiment where focal plane is offset 0 mm from bone interface, calculated in (top) control plane 0 mm from interface and (bottom) sagittal plane. Right: Average steady-state temperatures for experiment where focal plane is offset 10 mm from bone interface, calculated in (top) control plane 10 mm from interface and (bottom) sagittal plane. 124 radiology.rsna.org n Radiology: Volume 263: Number 1 April 2012

9 Table 3 Figure 6 Numerical Simulations of MR Imaging controlled Focused Ultrasound Hyperthermia in Bone, with the Focus Set at Two Different Offsets from the Muscle-Bone Interface Simulation Plane T 50 ( C) T 90 ( C) T 10 ( C) Radius at 42 C (mm) Radius at 41 C (mm) Thermal Dose (CEM43) Control Plane Offset 10 mm from Bone Interface Muscle, 10-mm offset Muscle, 0-mm offset Cortical bone Marrow Control Plane Offset 0 mm from Bone Interface Muscle, 10-mm offset Muscle, 0-mm offset Cortical bone Marrow Note. Unless otherwise indicated, data are means 6 standard deviations. CEM43 = cumulative equivalent minutes at 43 C. Table 4 Doxorubicin Concentrations Measured by Using Fluorescence Intensity in Tissue Samples Harvested from Heated and Unheated Regions of Rabbit Thigh Muscle and Bone Marrow Doxorubicin in Muscle (ng/mg) Doxorubicin in Marrow (ng/mg) Parameter Unheated Heated P Value* Unheated Heated P Value* Control Plane mm Away from Bone Interface (n = 4) Rabbit Rabbit Rabbit Rabbit Mean 6 standard deviation Control Plane at Bone Interface (n = 5) Rabbit Rabbit Rabbit Rabbit Rabbit Mean 6 standard deviation Overall (n = 9) Mean 6 standard deviation * P <.05 was considered to indicate a statistically significant difference (one-sided Wilcoxon signed rank test). heated by using water (35,36), microwave applicators (12), and ultrasound (14,37). With nonthermosensitive liposomes, similar increases have also been seen in the periphery of lesions created in tumors, kidney, and liver by using radiofrequency ablation (38). The results of our study demonstrated the feasibility of achieving image-guided drug deposition near bone, achieving 6.1- and fold increases in heated versus unheated bone marrow and muscle adjacent to Figure 6: Graph shows numerically simulated average steady-state temperature versus radius for two control scenarios: control plane set 0 mm from bone interface (open keys) and control plane offset 10 mm from the bone interface (solid keys). For each scenario, average steady-state temperatures were calculated in cortical bone, marrow, muscle adjacent to the bone, and muscle 10 mm from the bone interface. For clarity, every fourth marker is shown. bone, respectively. In our study, liposomes were administered during heating, and drug deposition was quantified by using the fluorescence intensity of released doxorubicin in homogenized bone marrow and muscle samples. Clinical pharmacokinetic data suggest that greater drug deposition may be possible if heating is administered minutes after infusion (13). In cases of thermal damage, vascular shutdown presumably acted both to prevent liposomes from reaching the target region and to prevent released drug in tissue from returning to the vasculature, again highlighting the importance of infusion timing and target temperature. Future studies will focus on optimization of infusion and heating protocols and on performing histologic examinations to determine the spatial distribution of doxorubicin with respect to patterns of thermal damage in heated tissue. Radiology: Volume 263: Number 1 April 2012 n radiology.rsna.org 125

10 Figure 7 Figure 7: Mean doxorubicin concentrations measured by using fluorescence intensity in tissue samples harvested from heated and unheated regions of rabbit thigh muscle and bone marrow. Graphs show data (a) averaged across all experiments and (b) separated into experiments where focal offsets of 10 and 0 mm were used. Error bars = standard deviation. Practical applications: In clinical studies of MR imaging guided focused ultrasound thermal ablation in bone, MR thermometry has been used to monitor soft-tissue temperature elevations during treatment to enable the clinician to halt or adjust treatment parameters and avoid unintended thermal damage to surrounding tissue (7 9,19). Commercial systems are capable of closed-loop MR thermometry control in soft tissue (39) and could implement the proposed method directly for controlled heating in bone. Our results suggested that with the focus set at the bone interface, bone temperatures in clinical treatments could be safely controlled by using temperatures measured in adjacent soft tissue. They also demonstrated the importance of multiplane thermometry to observe heating outside the ultrasound focus, as already implemented commercially (40). Clinical trials are currently underway combining thermosensitive liposomal doxorubicin with thermal ablation by using percutaneous radiofrequency ablation for hepatocellular carcinoma (41) and with mild heating by using microwave hyperthermia for recurrent breast cancer at the chest wall (42). In our study, tissue drug concentrations were increased in heated muscle and marrow both for experiments where the focus was set 10 mm away from the bone and large regions of thermal damage were observed and where the focus was set 0 mm from the bone and thermal damage was limited to a narrow region at the bone interface. This suggests that clinical treatments could use mild heating to avoid thermal damage near critical structures, while achieving a similar degree of drug deposition as would be achieved with ablative heating. Further studies are required to optimize infusion protocols and heating durations for drug deposition in large targets that cannot be covered in a single sonication and to determine whether local drug delivery to bone without thermal coagulation of the soft tissue could achieve local pain relief or tumor control. Acknowledgments: The authors thank Alexandra Garces, RVT, and Shawna Rideout, RLAT, RVT, for providing their expertise during rabbit experiments, as well as Amy Qu and Samuel Pichardo, PhD, for their assistance with the simulations, Anthony Chau and Adam Waspe, PhD, for assistance with the positioning system, Melissa Togtema, BSc, for tissue sample homogenization and fluorometry, and Alex Kiss, PhD, for providing assistance with the statistical analysis. Disclosures of Potential Conflicts of Interest: R.S. Financial activities related to the present article: Celsion provided thermosensitive liposomes used in this study. Financial activities not related to the present article: none to disclose. Other relationships: none to disclose. R.C. Financial activities related to the present article: Celsion provided thermosensitive liposomes used in this study. Financial activities not related to the present article: none to disclose. Other relationships: The MR imaging guided ultrasound technology used to produce controlled heating in this study has been licensed to a company this author cofounded, FUS Instruments, by Sunnybrook Health Sciences Centre. Author has shares in the company. K.H. Financial activities related to the present article: Celsion provided thermosensitive liposomes used in this study. Financial activities not related to the present article: institution has grants from Philips for studies for MR imaging guided HIFU; Philips provided travel to 2011 ISMR meeting (500 euros). Other relationships: The MR imaging guided ultrasound technology used to produce controlled heating in this study has been licensed to a company this author cofounded, FUS Instruments, by Sunnybrook Health Sciences Centre. Author has shares in the company. References 1. Coleman RE. Metastatic bone disease: clinical features, pathophysiology and treatment strategies. Cancer Treat Rev 2001;27(3): radiology.rsna.org n Radiology: Volume 263: Number 1 April 2012

11 2. Selvaggi G, Scagliotti GV. Management of bone metastases in cancer: a review. Crit Rev Oncol Hematol 2005;56(3): Mundy GR. Metastasis to bone: causes, consequences and therapeutic opportunities. Nat Rev Cancer 2002;2(8): Coleman RE. Bisphosphonates: clinical experience. Oncologist 2004;9(suppl 4): Callstrom MR, Charboneau JW, Goetz MP, et al. Painful metastases involving bone: feasibility of percutaneous CT- and USguided radio-frequency ablation. Radiology 2002;224(1): Dupuy DE, Liu DW, Hartfeil D, et al. Percutaneous radiofrequency ablation of painful osseous metastases: a multicenter American College of Radiology Imaging Network trial. Cancer 2010;116(4): Catane R, Beck A, Inbar Y, et al. MR-guided focused ultrasound surgery (MRgFUS) for the palliation of pain in patients with bone metastases: preliminary clinical experience. Ann Oncol 2007;18(1): Gianfelice D, Gupta C, Kucharczyk W, Bret P, Havill D, Clemons M. Palliative treatment of painful bone metastases with MR imaging guided focused ultrasound. Radiology 2008;249(1): Liberman B, Gianfelice D, Inbar Y, et al. Pain palliation in patients with bone metastases using MR-guided focused ultrasound surgery: a multicenter study. Ann Surg Oncol 2009;16(1): Kong G, Braun RD, Dewhirst MW. Characterization of the effect of hyperthermia on nanoparticle extravasation from tumor vasculature. Cancer Res 2001;61(7): Needham D, Dewhirst MW. The development and testing of a new temperaturesensitive drug delivery system for the treatment of solid tumors. Adv Drug Deliv Rev 2001;53(3): Hauck ML, LaRue SM, Petros WP, et al. Phase I trial of doxorubicin-containing low temperature sensitive liposomes in spontaneous canine tumors. Clin Cancer Res 2006;12(13): Poon RT, Borys N. Lyso-thermosensitive liposomal doxorubicin: a novel approach to enhance efficacy of thermal ablation of liver cancer. Expert Opin Pharmacother 2009;10(2): Staruch R, Chopra R, Hynynen K. Localised drug release using MRI-controlled focused ultrasound hyperthermia. Int J Hyperthermia 2011;27(2): Salomir R, Vimeux FC, de Zwart JA, Grenier N, Moonen CT. Hyperthermia by MRguided focused ultrasound: accurate temperature control based on fast MRI and a physical model of local energy deposition and heat conduction. Magn Reson Med 2000;43(3): Smith NB, Merrilees NK, Dahleh M, Hynynen K. Control system for an MRI compatible intracavitary ultrasound array for thermal treatment of prostate disease. Int J Hyperthermia 2001;17(3): Chopra R, Baker N, Choy V, et al. MRIcompatible transurethral ultrasound system for the treatment of localized prostate cancer using rotational control. Med Phys 2008;35(4): Smith NB, Temkin JM, Shapiro F, Hynynen K. Thermal effects of focused ultrasound energy on bone tissue. Ultrasound Med Biol 2001;27(10): Kopelman D, Inbar Y, Hanannel A, et al. Magnetic resonance guided focused ultrasound surgery: ablation of soft tissue at bone-muscle interface in a porcine model. Eur J Clin Invest 2008;38(4): Hynynen K. Acoustic power calibrations of cylindrical intracavitary ultrasound hyperthermia applicators. Med Phys 1993;20(1): Chopra R, Curiel L, Staruch R, Morrison L, Hynynen K. An MRI-compatible system for focused ultrasound experiments in small animal models. Med Phys 2009;36(5): Hynynen K. Hot spots created at skin-air interfaces during ultrasound hyperthermia. Int J Hyperthermia 1990;6(6): McDannold NJ, King RL, Jolesz FA, Hynynen KH. Usefulness of MR imaging derived thermometry and dosimetry in determining the threshold for tissue damage induced by thermal surgery in rabbits. Radiology 2000;216(2): Tang K, Choy V, Chopra R, Bronskill MJ. Conformal thermal therapy using planar ultrasound transducers and adaptive closedloop MR temperature control: demonstration in gel phantoms and ex vivo tissues. Phys Med Biol 2007;52(10): Santos JM, Wright GA, Pauly JM. Flexible real-time magnetic resonance imaging framework. Conf Proc IEEE Eng Med Biol Soc 2004;2: Peters RD, Hinks RS, Henkelman RM. Ex vivo tissue-type independence in protonresonance frequency shift MR thermometry. Magn Reson Med 1998;40(3): Sapareto SA, Dewey WC. Thermal dose determination in cancer therapy. Int J Radiat Oncol Biol Phys 1984;10(6): Bachur NR, Moore AL, Bernstein JG, Liu A. Tissue distribution and disposition of daunomycin (NCS-82151) in mice: fluorometric and isotopic methods. Cancer Chemother Rep 1970;54(2): Sprinkhuizen SM, Konings MK, van der Bom MJ, Viergever MA, Bakker CJ, Bartels LW. Temperature-induced tissue susceptibility changes lead to significant temperature errors in PRFS-based MR thermometry during thermal interventions. Magn Reson Med 2010;64(5): Schenck JF. The role of magnetic susceptibility in magnetic resonance imaging: MRI magnetic compatibility of the first and second kinds. Med Phys 1996;23(6): Duck FA. Physical properties of tissue: a comprehensive reference book. London, England: Academic Press, Hynynen K, DeYoung D. Temperature elevation at muscle-bone interface during scanned, focused ultrasound hyperthermia. Int J Hyperthermia 1988;4(3): Lin WL, Liauh CT, Chen YY, Liu HC, Shieh MJ. Theoretical study of temperature elevation at muscle/bone interface during ultrasound hyperthermia. Med Phys 2000;27(5): Ahmed M, Brace CL, Lee FT Jr, Goldberg SN. Principles of and advances in percutaneous ablation. Radiology 2011;258(2): Kong G, Anyarambhatla G, Petros WP, et al. Efficacy of liposomes and hyperthermia in a human tumor xenograft model: importance of triggered drug release. Cancer Res 2000;60(24): Yarmolenko PS, Zhao Y, Landon C, et al. Comparative effects of thermosensitive doxorubicin-containing liposomes and hyperthermia in human and murine tumours. Int J Hyperthermia 2010;26(5): Patel PR, Luk A, Durrani A, et al. In vitro and in vivo evaluations of increased effective beam width for heat deposition using a split focus high intensity ultrasound (HIFU) transducer. Int J Hyperthermia 2008;24(7): Ahmed M, Goldberg SN. Combination radiofrequency thermal ablation and adjuvant IV liposomal doxorubicin increases tissue coagulation and intratumoural drug accumulation. Int J Hyperthermia 2004;20(7): Enholm JK, Köhler MO, Quesson B, Mougenot C, Moonen CT, Sokka SD. Improved volumetric MR-HIFU ablation by robust binary feedback control. IEEE Trans Biomed Eng 2010;57(1): Köhler MO, Mougenot C, Quesson B, et al. Volumetric HIFU ablation under 3D guidance of rapid MRI thermometry. Med Phys 2009;36(8): Celsion. Phase 3 study of ThermoDox with radiofrequency ablation (RFA) in treatment of hepatocellular carcinoma (HCC). ClinicalTrials.gov. ct2/show/nct Updated December 20, Accessed March 19, Celsion. Phase 1/2 study of ThermoDox with approved hyperthermia in treatment of breast cancer recurrence at the chest wall (DIGNITY). ClinicalTrials.gov. clinicaltrials.gov/ct2/show/nct Updated May 24, Accessed March 19, Radiology: Volume 263: Number 1 April 2012 n radiology.rsna.org 127