Real-Time Positron Emission Tomography Evaluation of Topotecan Brain Kinetics after Ultrasound-Mediated Blood–Brain Barrier Permeability

Glioblastoma (GBM) is the most common primary adult brain malignancy with an extremely poor prognosis and a median survival of fewer than two years. A key reason for this high mortality is that the blood–brain barrier (BBB) significantly restricts systemically delivered therapeutics to brain tumors. High-intensity focused ultrasound (HIFU) with microbubbles is a methodology being used in clinical trials to noninvasively permeabilize the BBB for systemic therapeutic delivery to GBM. Topotecan is a topoisomerase inhibitor used as a chemotherapeutic agent to treat ovarian and small cell lung cancer. Studies have suggested that topotecan can cross the BBB and can be used to treat brain metastases. However, pharmacokinetic data demonstrated that topotecan peak concentration in the brain extracellular fluid after systemic injection was ten times lower than in the blood, suggesting less than optimal BBB penetration by topotecan. We hypothesize that HIFU with microbubbles treatment can open the BBB and significantly increase topotecan concentration in the brain. We radiolabeled topotecan with 11C and acquired static and dynamic positron emission tomography (PET) scans to quantify [11C] topotecan uptake in the brains of normal mice and mice after HIFU treatment. We found that HIFU treatments significantly increased [11C] topotecan brain uptake. Moreover, kinetic analysis of the [11C] topotecan dynamic PET data demonstrated a substantial increase in [11C] topotecan volume of distribution in the brain. Furthermore, we found a decrease in [11C] topotecan brain clearance, confirming the potential of HIFU to aid in the delivery of topotecan through the BBB. This opens the potential clinical application of [11C] topotecan as a tool to predict topotecan loco-regional brain concentration in patients with GBMs undergoing experimental HIFU treatments.


Introduction
Glioblastoma (GBM) is the most common primary adult brain tumor with an extremely poor prognosis and median survival of less than 2 years [1,2]. One key reason for  18 F was produced on the Siemens Eclipse HP cyclotron RDS 111  in fluorine radioisotope nuclear reaction 18 O(p,n) 18 F. The synthesis was performed as previously reported [23] on the ORA Neptis Synthesizer Unit. Briefly, 6 mg of FSPG precursor, di-tertbutyl (4S)-4-(3-((2-naphthylsulfonyl)oxy)propyl)-N-trityl-L-glutamate, was transferred to a 10 mL vial with 1.4 mL of acetonitrile, mixed, and loaded into the reaction vessel. Siemens Eclipse cyclotron was loaded with 2.4 mL of [ 18   Carbon-11 labeled topotecan was synthesized using a modification of the previously reported scheme by Yamasaki et al. [24]. Briefly, desmethyl topotecan (0.4 mg) was dissolved in anhydrous DMF (300 µL) and 0.5 N sodium hydroxide (5 µL). A Siemens Eclipse HP cyclotron RDS 111  was bombarded for 30 min at 50 µA, and the radioactive carbon-11 carbon dioxide [ 11 C] CO 2 gas was converted to [ 11 C] iodomethane via GE F x MeI module. The [ 11 C] iodomethane gas was bubbled into the reaction mixture at room temperature. Once activity had plateaued, the reaction vessel was moved to a heating block set at 80 • C and heated for 5 min. The reaction was then removed from heat and quenched with 300 µL of semi-prep mobile phase and taken up into a high-performance liquid chromatography (HPLC) loop. The crude product was purified by way of HPLC using a mobile phase consisting of methanol/water/trifluoroacetic acid (40/60/0.06) at a flow rate of 4.0 mL/min. The desired fraction was collected (r t~8 min) and loaded onto a pre-conditioned t-C 18 plus cartridge (conditioned via 10 mL ethanol followed by 10 mL of sterile water for injection) with the use of 100 mL of sterile water for injection. The cartridge was washed with 10 mL of sterile water for injection and then eluted to a final product vial through a sterile GVM 0.22-µm filter with the use of 1.0 mL ethanol followed by 9 mL of 0.9% saline to afford [ 11 C] topotecan (42-min post start of synthesis (SOS)), a pale-yellow solution. The radiochemical purity of [ 11 C] topotecan determined by HPLC analysis was 94 ± 3% (n = 4).

Mice
C57BL/6NTac mice (B6, Taconic) were maintained on a normal mouse diet. All animal experiments were conducted according to protocols approved by the Institutional Animal Care and Use Committee of Columbia University Medical Center (AC-AAAT8456 (25 January 2021)).

HIFU with Microbubbles
DEFINITY ® microbubbles were prepared according to manufacturer instructions (Lantheus Medical Imaging, Inc., Billerica, MA, USA). HIFU was applied through the intact skin to the right cerebral hemisphere using the RK50 system (FUS Instruments, Toronto, ON, Canada). Ultrasound was delivered as a series of burst exposures (10 ms duration, 1 s repetition frequency, ultrasound frequency of 1.5 MHz, and 0.7 MPa peak negative focal pressure). Simultaneous with the start of HIFU application, DEFINITY ® microbubbles (4.5 × 10 7 microbubbles in 0.5 mL of 0.9% NaCl sterile solution) were infused through a catheter at a flow rate of 50 µL/min into a mouse-tail vein [25]. Immediately after HIFU treatment, 75 µL of Cy7-albumin (Cy7 ALB) (2.6 mg/mL) and 50 µL of 2% Evan's blue (EB) solution were injected through the same tail vein catheter to verify the opening of the BBB. ] FSPG injection, mice were placed into a 4-mouse bed, and 30 min static PET images were acquired using a micro PET scanner (Siemens, Munich, Germany) followed by microCT (MIlabs, Houten, The Netherlands) on the same bed for anatomical reference.
For both [ 18 F] FSPG and [ 11 C] topotecan studies, immediately after PET/CT scans, mice were euthanized, and their dissected brains were placed into 4% PFA and imaged for 10 min with static PET. Regions of interest (ROI) were manually drawn over the heart and the right brain hemisphere around the area targeted by HIFU using microCT as a reference. All PET images were reconstructed using the 3D-OSEM algorithm with 3-iterations in 256 × 256 matrix (Inveon, Siemens, Munich, Germany) and analyzed using VivoQuant ver 4 (Invicro, Boston, MA, USA).

[ 11 C] Topotecan Kinetic Analysis
The [ 11 C] topotecan radioactivity curve over time in the heart ROI was used as a proxy for the arterial input function during the kinetic modeling of the tracer time-activity curves (TACs) in the brain ROI. In both groups of control mice (mice not treated with HIFU, n = 4) and mice after HIFU with microbubbles treatment (n = 4), the raw radioactivity values measured in the heart ROI in each individual animal were first averaged across animals within the group and then fitted using a linear interpolation before the curve peak and the sum of three decreasing exponentials after the peak. In each group, the brain ROI TACs measured in each individual animal were also averaged across animals within the group and fitted with a 2-tissue compartment model (2TCM) [26] using weights equal to the square root of each acquisition frame duration in order to provide estimates of the 2TCM rate constants (K 1 , k 2 , k 3 , and k 4 ). The tracer total distribution volume (V T ), the tracer non-displaceable distribution volume (V ND ), and the tracer specific binding potential (BP P ) in the brain ROI were then calculated from the rate constant estimates as V T = K 1 /k 2 × (1 + k 3 /k 4 ), V ND = K 1 /k 2 , and BP P = K 1 /k 2 × (k 3 /k 4 ), respectively [27]. We also computed standard errors associated with the estimates of V T , V ND , and BP P . The analysis was also repeated using each individual heart curve and brain TAC instead of the group average curves. All analyses were performed using Matlab 2016b (www.mathworks.com).

Quantification and Statistical Analysis
Statistical analysis was performed using Prism 8.0 (San Diego, CA, USA). All data are represented as mean ± standard error. Statistical p-values were calculated using a two-tailed Student's t-test for unpaired samples.

Topotecan Demonstrates Variable Cytotoxicity between Human GBM Cells
Topotecan is a topoisomerase I inhibitor that demonstrates antitumor activity against a variety of human cancer cell lines [28] and potential activity in GBM after loco-regional delivery [29,30]. To evaluate the effective topotecan dose required in the brain, we exposed G48a and U251 GBM cells to various concentrations of topotecan (0.01 to 10 µg/mL). We found that topotecan is cytotoxic on both G48a and U251 cells but was noticeably more cytotoxic on U251 cells compared to G48a cells, with LD50s of 0.002 (0.001-0.003) µg/mL and 0.08 (0.06-1.0) µg/mL, respectively ( Figure 1).

[ 11 C] Topotecan Kinetic Analysis
The [ 11 C] topotecan radioactivity curve over time in the heart ROI was used as a proxy for the arterial input function during the kinetic modeling of the tracer time-activity curves (TACs) in the brain ROI. In both groups of control mice (mice not treated with HIFU, n = 4) and mice after HIFU with microbubbles treatment (n = 4), the raw radioactivity values measured in the heart ROI in each individual animal were first averaged across animals within the group and then fitted using a linear interpolation before the curve peak and the sum of three decreasing exponentials after the peak. In each group, the brain ROI TACs measured in each individual animal were also averaged across animals within the group and fitted with a 2-tissue compartment model (2TCM) [26] using weights equal to the square root of each acquisition frame duration in order to provide estimates of the 2TCM rate constants (K1, k2, k3, and k4). The tracer total distribution volume (VT), the tracer non-displaceable distribution volume (VND), and the tracer specific binding potential (BPP) in the brain ROI were then calculated from the rate constant estimates as VT = K1/k2 × (1 + k3/k4), VND = K1/k2, and BPP = K1/k2 × (k3/k4), respectively [27]. We also computed standard errors associated with the estimates of VT, VND, and BPP. The analysis was also repeated using each individual heart curve and brain TAC instead of the group average curves. All analyses were performed using Matlab 2016b (www.mathworks.com).

Quantification and Statistical Analysis
Statistical analysis was performed using Prism 8.0 (San Diego, CA, USA). All data are represented as mean ± standard error. Statistical p-values were calculated using a twotailed Student's t-test for unpaired samples.

Topotecan Demonstrates Variable Cytotoxicity Between Human GBM Cells
Topotecan is a topoisomerase I inhibitor that demonstrates antitumor activity against a variety of human cancer cell lines [28] and potential activity in GBM after loco-regional delivery [29,30]. To evaluate the effective topotecan dose required in the brain, we exposed G48a and U251 GBM cells to various concentrations of topotecan (0.01 to 10 μg/mL). We found that topotecan is cytotoxic on both G48a and U251 cells but was noticeably more cytotoxic on U251 cells compared to G48a cells, with LD50s of 0.002 (0.001-0.003) μg/mL and 0.08 (0.06-1.0) μg/mL, respectively ( Figure 1).

PET Imaging Confirms HIFU Permeabilizes the BBB
HIFU has the potential to significantly increase the concentration of topotecan in the brain and decrease systemic exposure. To confirm loco-regional BBB opening by HIFU and our ability to detect these effects in real-time using PET imaging, we treated mice with HIFU and imaged them with a surrogate PET tracer ([ 18 F] FSPG), which has been shown not to penetrate an intact BBB both in animals [31] and in healthy volunteers [32]. penetrate the BBB and did not accumulate in the brains of untreated control mice (n = 4) in quantities detectable by PET (Figure 2a top). In contrast, after treatment with HIFU with microbubbles, the BBB was significantly permeabilized, as evidenced by [ 18 F] FSPG PET (n = 4) (Figure 2a bottom). We further confirmed the HIFU-induced loco-regional opening of the BBB by obtaining PET images of dissected brains (Figure 2b). We confirmed HIFU-induced opening of the BBB by optical imaging of Cy7-albumin and Evan's blue locoregional accumulation in the brains of HIFU-treated mice, as we previously reported [33] ( Figure 3). This strongly supports that HIFU with microbubbles opens the BBB, which can be detected using PET imaging.

PET Imaging Confirms HIFU Permeabilizes the BBB
HIFU has the potential to significantly increase the concentration of topotecan in the brain and decrease systemic exposure. To confirm loco-regional BBB opening by HIFU and our ability to detect these effects in real-time using PET imaging, we treated mice with HIFU and imaged them with a surrogate PET tracer ([ 18 F] FSPG), which has been shown not to penetrate an intact BBB both in animals [31] and in healthy volunteers [32]. To test the effect of HIFU on [ 18 F] FSPG brain uptake, mice were treated with HIFU with microbubbles and then injected i.v. with [ 18 F] FSPG. We found that [ 18 F] FSPG did not penetrate the BBB and did not accumulate in the brains of untreated control mice (n = 4) in quantities detectable by PET (Figure 2a top). In contrast, after treatment with HIFU with microbubbles, the BBB was significantly permeabilized, as evidenced by [ 18 F] FSPG PET (n = 4) (Figure 2a bottom). We further confirmed the HIFU-induced loco-regional opening of the BBB by obtaining PET images of dissected brains (Figure 2b). We confirmed HIFUinduced opening of the BBB by optical imaging of Cy7-albumin and Evan's blue locoregional accumulation in the brains of HIFU-treated mice, as we previously reported [33] ( Figure 3). This strongly supports that HIFU with microbubbles opens the BBB, which can be detected using PET imaging.   HIFU has the potential to significantly increase the concentration of topotecan in the brain and decrease systemic exposure. To confirm loco-regional BBB opening by HIFU and our ability to detect these effects in real-time using PET imaging, we treated mice with HIFU and imaged them with a surrogate PET tracer ([ 18 F] FSPG), which has been shown not to penetrate an intact BBB both in animals [31] and in healthy volunteers [32]. To test the effect of HIFU on [ 18 F] FSPG brain uptake, mice were treated with HIFU with microbubbles and then injected i.v. with [ 18 F] FSPG. We found that [ 18 F] FSPG did not penetrate the BBB and did not accumulate in the brains of untreated control mice (n = 4) in quantities detectable by PET (Figure 2a top). In contrast, after treatment with HIFU with microbubbles, the BBB was significantly permeabilized, as evidenced by [ 18 F] FSPG PET (n = 4) (Figure 2a bottom). We further confirmed the HIFU-induced loco-regional opening of the BBB by obtaining PET images of dissected brains (Figure 2b). We confirmed HIFUinduced opening of the BBB by optical imaging of Cy7-albumin and Evan's blue locoregional accumulation in the brains of HIFU-treated mice, as we previously reported [33] ( Figure 3). This strongly supports that HIFU with microbubbles opens the BBB, which can be detected using PET imaging.

HIFU Significantly Increases Loco-Regional Brain Concentration of Topotecan
To study the effects of HIFU on topotecan BBB permeability, we treated mice with HIFU with microbubbles and then immediately injected them i.v. with [ 11 C] topotecan. PET images at 30 min post tracer injection demonstrated significantly higher levels of [ 11 C] topotecan in the brains of HIFU-treated animals (n = 4) compared to control animals (n = 4) (Figure 4a), directly indicating real-time effects of HIFU on topotecan brain uptake. These data were confirmed in the postmortem PET images of dissected mouse brains (Figure 4b).
To study the effects of HIFU on topotecan BBB permeability, we treated mice with HIFU with microbubbles and then immediately injected them i.v. with [ 11 C] topotecan. PET images at 30 min post tracer injection demonstrated significantly higher levels of [ 11 C] topotecan in the brains of HIFU-treated animals (n = 4) compared to control animals (n = 4) (Figure 4a), directly indicating real-time effects of HIFU on topotecan brain uptake. These data were confirmed in the postmortem PET images of dissected mouse brains (Figure 4b). Figure 4. HIFU with microbubbles treatments permeabilize the BBB for topotecan and significantly increase the locoregional accumulation of topotecan in the brain. (a) Representative PET images of [ 11 C] topotecan uptake in brains of control (top panels) and HIFU-treated (bottom panels) mice. Images were quantified (n = 4) and demonstrated a significant increase in [ 11 C] topotecan brain uptake after HIFU. (b) Immediately after PET imaging, the brains of all mice were dissected, imaged, and quantified with PET to confirm in vivo PET signal. PET images and quantification of [ 11 C] topotecan uptake in dissected brains confirm PET results on live mice. Arrowhead, [ 11 C] topotecan signal in the brain of mice after HIFU.

Kinetics of [ 11 C] Topotecan in Mice Treated with HIFU
For kinetic analysis of topotecan, we obtained dynamic [ 11 C] topotecan PET scans of control (n = 4) and HIFU-treated (n = 4) mice. We found lower levels of [ 11 C] topotecan in Figure 4. HIFU with microbubbles treatments permeabilize the BBB for topotecan and significantly increase the loco-regional accumulation of topotecan in the brain. (a) Representative PET images of [ 11 C] topotecan uptake in brains of control (top panels) and HIFU-treated (bottom panels) mice. Images were quantified (n = 4) and demonstrated a significant increase in [ 11 C] topotecan brain uptake after HIFU. (b) Immediately after PET imaging, the brains of all mice were dissected, imaged, and quantified with PET to confirm in vivo PET signal. PET images and quantification of [ 11 C] topotecan uptake in dissected brains confirm PET results on live mice. Arrowhead, [ 11 C] topotecan signal in the brain of mice after HIFU.

Kinetics of [ 11 C] Topotecan in Mice Treated with HIFU
For kinetic analysis of topotecan, we obtained dynamic [ 11 C] topotecan PET scans of control (n = 4) and HIFU-treated (n = 4) mice. We found lower levels of [ 11 C] topotecan in the brain of control untreated mice compared to HIFU-treated mice, with a gradual decrease from a peak value of 0.88 ± 0.4% ID/g at 3 min post-injection to 0.2 ± 0.1% ID/g at the end of the scan (63 min post-injection) ( Figure 5).
HIFU induced a three-to five-fold increase in the [ 11 C] topotecan brain uptake, which peaked at 2.1 ± 0.7% ID/g at 3 min post-injection and gradually decreased to 1.1 ± 0.3% ID/g by the end of the scan. We used these data to analyze the kinetic rate parameters for [ 11 C] topotecan injected into untreated mice (n = 4) and compared them to mice treated with HIFU (n = 4) ( Figure 6). We found that [ 11 C] topotecan kinetics in the brain are best described by a 2-tissue compartment model (2TCM) in comparison to a 1-tissue compartment and a 2-tissue-irreversible compartment models (Figure 6a). Kinetic data demonstrated a 2.2-fold higher V T (volume of distribution) value of [ 11 C] topotecan in the HIFU-treated mice (0.4 ± 0.02) compared to control mice (0.18 ± 0.02). Interestingly, Pharmaceutics 2021, 13, 405 8 of 12 both V ND (the distribution volume of a non-displaceable [ 11 C] topotecan) and BP P (the distribution volume of specifically bound [ 11 C] topotecan) were significantly increased after HIFU treatment, while K 1 (rate constant for transfer from arterial plasma to the brain) was only moderately (1.3-fold) increased after HIFU (Figure 6). This suggests that HIFU significantly improves brain uptake of topotecan (both specific and non-specific) and increases the exposure time of topotecan in the brain, achieving considerably higher loco-regional topotecan concentration in the brain.
Pharmaceutics 2021, 13, x FOR PEER REVIEW 8 of 12 the brain of control untreated mice compared to HIFU-treated mice, with a gradual decrease from a peak value of 0.88 ± 0.4% ID/g at 3 min post-injection to 0.2 ± 0.1% ID/g at the end of the scan (63 min post-injection) ( Figure 5). HIFU induced a three-to five-fold increase in the [ 11 C] topotecan brain uptake, which peaked at 2.1 ± 0.7% ID/g at 3 min post-injection and gradually decreased to 1.1 ± 0.3% ID/g by the end of the scan. We used these data to analyze the kinetic rate parameters for [ 11 C] topotecan injected into untreated mice (n = 4) and compared them to mice treated with HIFU (n = 4) ( Figure 6). We found that [ 11 C] topotecan kinetics in the brain are best described by a 2-tissue compartment model (2TCM) in comparison to a 1-tissue compartment and a 2-tissue-irreversible compartment models (Figure 6a). Kinetic data demonstrated a 2.2-fold higher VT (volume of distribution) value of [ 11 C] topotecan in the HIFUtreated mice (0.4 ± 0.02) compared to control mice (0.18 ± 0.02). Interestingly, both VND (the distribution volume of a non-displaceable [ 11 C] topotecan) and BPP (the distribution volume of specifically bound [ 11 C] topotecan) were significantly increased after HIFU treatment, while K1 (rate constant for transfer from arterial plasma to the brain) was only moderately (1.3-fold) increased after HIFU ( Figure 6). This suggests that HIFU significantly improves brain uptake of topotecan (both specific and non-specific) and increases the exposure time of topotecan in the brain, achieving considerably higher loco-regional topotecan concentration in the brain.

Discussion
HIFU with microbubbles is currently in clinical trials to permeabilize the BBB and allow efficient delivery of therapeutics into the brains of patients with intracranial malignancies [12]. This is especially beneficial for agents like topotecan that have significant systemic adverse effects and little or no BBB permeability. PET imaging can play a valuable role in quantifying BBB permeability as well as providing direct visualization of a radiolabeled drug. We confirmed the ability of PET to image HIFU-mediated BBB permea-

Discussion
HIFU with microbubbles is currently in clinical trials to permeabilize the BBB and allow efficient delivery of therapeutics into the brains of patients with intracranial malignancies [12]. This is especially beneficial for agents like topotecan that have significant systemic adverse effects and little or no BBB permeability. PET imaging can play a valuable role in quantifying BBB permeability as well as providing direct visualization of a radiolabeled drug. We confirmed the ability of PET to image HIFU-mediated BBB permeability using [ 18 F] FSPG, as it does not cross the BBB in animals [31] or healthy volunteers [32] and has a relatively long half-life of 110 min. PET scanning after [ 18 F] FSPG tracer administration was able to visualize and quantify BBB permeability in the area of the HIFU treatment ( Figure 2). We confirmed these PET results by direct imaging of dissected brains and by EB dye localization. PET imaging, therefore, has potential use in the clinic to validate BBB opening and personalize therapeutic dose calculations for individual patients as HIFU technology continues to develop.
Topotecan is a topoisomerase inhibitor chemotherapeutic agent used for the treatment of small-cell lung carcinoma [16] and ovarian cancers [28], and sometimes for brain metastasis [18,19]. Earlier data suggested that topotecan could cross the BBB in normal rats [17], making it an attractive therapeutic agent to use for the treatment of brain malignancies. Our study demonstrated that topotecan is cytotoxic in G48a and U251 human glioblastoma cell lines ( Figure 1). Moreover, PET imaging with [ 11 C] topotecan tracer showed that it only crosses the BBB in healthy mice at a very low level (Figures 4-6). [ 11 C] topotecan brain concentration peaked at 0.9 ± 0.4% ID/g 3-5 min after injection followed by fast decline and plateaued at 0.2 ± 0.08% ID/g. This suggests that even though topotecan can minimally cross the BBB, one would need higher-than-optimal doses to achieve effective drug concentration where the BBB is intact.
Kinetic analysis is used to model tracer behavior and estimate tracer tissue uptake and clearance rates. We demonstrated that kinetic analysis can be used to estimate [ 11 C] topotecan brain uptake and clearance in normal mice as well in mice treated with HIFU with microbubbles. In our study, we used HIFU with microbubbles to open the BBB to achieve higher topotecan concentrations in the brain. We demonstrated that injection of topotecan after HIFU at least doubles topotecan loco-regional peak concentration in the brain and significantly increases topotecan retention in the brain. Even after 1 h postinjection, we detected significant [ 11 C] topotecan remaining in the brain ( Figure 5), equal to topotecan peak concentration in the control mice. Based on our imaging data, 1-1.5% ID/gm of the topotecan dose was delivered to the mouse brain (Figure 4), primarily in the area of the HIFU application ( Figure 4). The reported MTD of topotecan in GBM patients (1.5 mg/m 2 /d per day × five days [34]) implies that it would be feasible to administer 2.14 mg per day to a 70 kg person. Our imaging data suggest that the brain uptake, which predominantly takes place at the site of HIFU application, would receive a sustained dose of at least 20 µg/gm (1% of 2.14 mg). Our cell data demonstrate that cells in vitro had an LD 50 significantly below this concentration (0.08-0.002 µg/mL). Furthermore, Kaiser et al., demonstrated that direct infusion of 2 µg/g/day for five days was effective at treating orthotopic GBMs [20]. Studying topotecan in combination with HIFU is justified because HIFU allows us to reach concentrations equal to or above these reported data. Our work strongly supports using HIFU to open the BBB before topotecan treatment in order to increase its therapeutic window and demonstrates the potential clinical application of [ 11 C] topotecan PET as a tool to predict loco-regional brain concentration in patients with GBMs undergoing experimental HIFU treatments.
Promising preclinical data that indicate the safety and efficacy of multiple doses of ultrasound plus HIFU are emerging. For example, Wei et al., recently demonstrated that two doses of HIFU (7 days apart) plus microbubbles effectively made a subtherapeutic dose of etoposide efficacious against an orthotopic model of GBM [35]. Similar to topotecan, etoposide is a topoisomerase inhibitor that does not effectively cross the blood-brain barrier. While this is exciting, it indicates that for the application of HIFU with microbubbles to become a widespread means of opening the BBB for drug delivery, more work will likely have to be conducted looking at the effects of multiple applications over a short period of time, especially when treating GBM. Alternatively, it might be necessary to change the dosing regimen from daily in certain instances to less dense dosing (weekly or monthly). Importantly, as HIFU devices are becoming more advanced, they offer sufficient convenience and comfort to permit multiple applications. The most advanced imageguided ultrasound units set up in MRI scanners allow for the repeatability and comfort needed to perform these procedures on a routine basis.

Conclusions
We demonstrate that PET imaging is an effective tool to confirm BBB permeability and can potentially be used in patients with brain malignancies treated with HIFU. We also demonstrate the potential of modeling [ 11 C] topotecan PET kinetics as a tool to predict regional topotecan brain concentration in the patients. This supports using HIFU plus microbubbles in conjunction with topotecan to potentially increase its therapeutic window in intracranial malignancies.