An implantable ultrasound-powered device for the treatment of brain cancer using electromagnetic fields.

Brain tumors have been proved challenging to treat. Here, we present a promising alternative by developing an implantable ultrasound-powered tumor treating device (UP-TTD) that electromagnetically disrupts the rapid division of cancer cells without any adverse effects on normal neurons, thereby safely inhibiting brain cancer recurrence. In vitro and in vivo experiments confirmed the significant therapeutic effect of the UP-TTD, with ~58% inhibition on growth rate of clinical tumor cells and ~78% reduction of cancer area in tumor-bearing rats. This UP-TTD is wireless ultrasound-powered, chip-sized, lightweight, and easy to operate on complex surfaces, with a largely boosting therapeutic efficiency and reducing energy consumption. Meanwhile, various treatment parameters could be tuned from the UP-TTD without increasing its size or adding circuits on the integrated chip. The tuning process was simulated and discussed, showing an excellent agreement with the experimental data. The encouraging results of the UP-TTD raise the possibility of a new modality for brain cancer treatment.

INTRODUCTION

Although some significant progress has been made on the fundamental questions in brain tumor research (–), successfully diagnosing and curing brain tumors is still very challenging for clinicians (–). Tumor resection through surgery is often the best hope for patients; however, most patients still cannot survive more than 5 years after surgery (, ). Recurrence and metastasis are the leading cause of patients’ death, which come from the remaining cancer cells that are difficult to remove entirely during the surgery (–). Recently, tumor treating fields (TTFs), an antimitotic therapy that uses alternating electrical fields to interrupt cancer cells’ ability to divide, appear to be a new promising strategy for treating brain tumors. The TTFs act upon specific highly charged proteins that are essential to the process of cell division; thus, they could slow down a tumor’s growth and its ability to spread. Since 2011, TTF technology has become a Food and Drug Administration–approved anticancer treatment for glioblastoma multiforme (–), proven to be safe and improve survival for many patients (, –). Using insulated external electrodes, TTFs were applied to animal cancer models (, ) and showed encouraging therapeutic outcomes. However, traditional TTF techniques suffer from the electrode routing difficulties of a suitable wiring harness, as well as rough spatial resolution and low treating efficiency.

Wireless-powered implantable medical devices provide a solution to overcome the limitations of traditional TTFs, avoiding the wiring issues and improving the treating efficiency (, ). Implantable devices deliver more precise field control and consume less energy. Meanwhile, the adoption of wireless power transmission (WPT) eliminates the need for wires (, ). Ultrasound or acoustic power transfer features significantly higher efficiency than inductive coupling approaches for WPT and is more suitable for implanted biomedical applications (–). Moreover, the low propagation loss of ultrasound leads to an extended penetration depth in the body and reduces unnecessary biological hazards (). Therefore, as a promising tumor treating strategy, TTF techniques still have a lot of room for improvement, and by combining with ultrasound WPT techniques, more effective clinical medical TTF devices are expected.

For now, we have not found any published work on the implantable ultrasonic-powered TTF device; here, we demonstrate an implantable ultrasound-powered device for the treatment of brain cancer using electromagnetic fields. The device is made entirely of biocompatible materials, and the whole system was markedly miniaturized as one single flexible membrane (with a thickness of less than 500 μm) for easy implantation. A relatively high inhibition rate of tumor cell proliferation was verified in both in vitro and in vivo experiments.

RESULTS

Working mechanism and characterization of the ultrasound-powered tumor treating device

The ultrasound-powered tumor treating device (UP-TTD) is an implantable biomedical device that is excited by an external ultrasound power and produces alternating TTFs to inhibit tumor growth in the brain. The schematic of the device is presented in Fig. 1A. After receiving external ultrasonic excitation, the UP-TTD generates a tunable alternating electric field (with low intensity, 1 to 3 V/cm; intermediate frequency, 100 to 300 kHz). The resulting electrical TTF exerts biophysical forces on charged and polarizable molecules that are essential to cell division, preventing cellular proteins from moving to the correct locations and, thus, disrupting tumor cell mitosis and triggering apoptosis of the dividing cells (Fig. 1B) (). The TTFs mainly act upon rapidly dividing cells of cancerous tumors without affecting normal neural cells, so they can be safely implanted in a patient’s brain (). The UP-TTD hybridizes multiple layers of flexible structural substrates (Fig. 1C), integrating a wireless ultrasonic energy converter and flexible metal electrodes to realize the construction of the alternating electric field.

Fig. 1.

Schematic and design of the flexible UP-TTD.

(A) Schematic of the UP-TTD. The device is designed as an implantable biomedical device that is excited by an external ultrasound power source and produced adjustable alternating electric fields to inhibit tumor growth in the brain. (B) The UP-TTD mainly focuses on rapidly dividing tumor cells without affecting normal neural cells, ensuring the safety of the equipment to the patient’s brain. (C) Structure of the UP-TTD with key components labeled.

The whole system was markedly miniaturized for the easy implant, and all the components were integrated on one single flexible membrane chip with a total thickness of less than 500 μm (Fig. 2, A and B), capable of adapting to the irregular surfaces inside the human brain. The wireless ultrasonic energy converter is the core component of the UP-TTD, which couples ultrasonic mechanical vibrations with an implantable triboelectric generator, achieving energy conversion from mechanical to electrical in vivo. The micrometer-scale displacements were generated on the perfluoroalkoxy (PFA) membrane by ultrasound, which is then converted into electrical energy through contact electrification with the adjacent metal electrodes (Fig. 2C). The microgap between the two triboelectric pairs (a thin PFA membrane and a flexible copper film) was tuned to generate specific required electrical outputs. For biosafety, the whole device was packed with a thin layer of biocompatible material, polydimethylsiloxane (PDMS; 50 μm) (), which requires careful handling during the process to ensure that the flexibility and functionality of the devices were not compromised (Fig. 2D). The package design maintains a stable and constant output despite variations under environmental conditions. UP-TTDs also exhibited remarkable flexibility and durability, preserving its structure and function even after 500 bending tests (Fig. 2E).

Fig. 2.

Mechanical characterization of UP-TTD.

(A and B) The optical images of the UP-TTD, which demonstrate the mechanical compliance and ultrathin characteristics of the device. Scale bar, 200 μm. (C) Schematic of wireless ultrasonic energy conversion. (D) Little output decrease in the UP-TTD is observed after PDMS package; devices were tested at a distance of 30 mm from the ultrasound probe (frequency, 40 kHz; power density, 0.2 W/cm2). (E) The 500 times bending test shows strong durability of the UP-TTD; devices were tested at a distance of 40 mm from the ultrasound probe (frequency, 40 kHz; power density, 0.2 W/cm2). N.S., not significant.

Wireless tuning of UP-TTD for tumor treating

To obtain acceptable beneficial effects from our UP-TTD, two key parameters need to be considered: the intensity and frequency of the TTF. For our device, an average output intensity could reach 3 to 4 V/cm (Fig. 3A), with an ultrasonic excitation (0.3 W/cm2) at 20 mm, which is a little intensive for brain tissue. However, although it is believed that increased field intensity confers higher inhibition effects on tumor cell division (, , ), a field of 1 to 3 V/cm is considered to be safer for clinical brain tumor treatment ().We adopt a simple and easy way to tune the frequency and intensity of the TTFs, avoiding increasing size or adding circuits on the integrated chip. As demonstrated, a field intensity ranging from ~0.7 to ~4.3 V/cm could be successfully tuned by changing the distance from the ultrasonic source (Fig. 3B) or the exciting power of the ultrasound (Fig. 3C). The wireless operation allows the device to operate above 50 mm subcutaneously, requiring an ultrasound power of approximately 0.2 to 0.4 W/cm2, which is a safe power level for most medical applications (–).

Fig. 3.

Wireless tuning of UP-TTD for tumor treatment.

(A) Outputs of the UP-TTD with an ultrasonic excitation (0.3 W/cm2) at 20 mm. (B) The output intensity of UP-TTD increases as the excitation distance decreases. (C) The output intensity of UP-TTD could be successfully tuned by changing the exciting distance and excited ultrasonic power; the error bars represent the SD of output from five devices (fabricated by the same method). (D) The output frequency of the UP-TTD could be successfully tuned by changing the structure gap between the flexible PFA membrane and electrodes. (E) Simulation of the generated acoustic pressure field with or without UP-TTD. (F and G) The simulated results on the displacement of PFA membrane in UP-TTD. Multimode peaks of the displacements could be observed, thus a higher frequency (100 to 300 kHz) of electrical output could be generated by our UP-TTD. Color bar, displacement of the membrane.

With ultrasonic excitation frequencies of 25 and 40 kHz, our device realized extensive frequency tuning from 100 to 150 kHz and 220 to 350 kHz by adjusting the microgap between flexible membrane and electrodes (Fig. 3D). The induced biological effects have been reported to vary depending on the applied frequency of TTFs (, ): Low-frequency electric fields (<1 kHz) disrupt membrane polarization, thereby interfering with neural functions in the brain. On the other hand, high-frequency fields (>1000 kHz) cause detrimental heating effects because of the violent vibration of the charged/polar molecules inside the cells (). Here, the intermediate frequency alternating electric fields (100 to 300 kHz) were chosen because they do not generate thermal hazards nor trigger abnormal neural behaviors. The gap distance was controlled using a hollow half-rigid frame structure (as shown in Fig. 1C, “Microgap”) to separate the two membranes on the four sides of the device. Upon ultrasound excitation, the distance between the two membranes changes dynamically; however, the distance on all sides can be held at the initially set distance, which is decisive for the output frequency.

To further elaborate how output frequency is tuned on our UP-TTD, simulation of the functional processes was conducted with a finite element analysis (COMSOL Multiphysics). The generated acoustic pressure field (Fig. 3E and movies S1 and S2) and the resulting displacement of the PFA membrane on the UP-TTD (Fig. 3F and movie S3) were simulated. Motivated by a low-frequency ultrasound (40 kHz), the PFA membrane produces multimode strikes (Fig. 3G) on the adjacent electrodes. For each striking of the antinode, negative charges are generated on the membrane through the triboelectric charging; meanwhile, electric holes are created in the electrode and consequently output a voltage pulse. During the separation, the potential of the electrode gradually rises as it moves away from the negatively charged PFA membrane, generating a reverse pulse. Thus, a much higher frequency (100 to 300 kHz) of electrical output could be generated by our UP-TTD, and the output frequency was determined by the distance between the PFA membrane and flexible electrodes. Our simulation results showed an excellent agreement with the experimental data.

In vitro and in vivo experiments

The cancer inhibition functionality of UP-TTD was first characterized in vitro with glioblastoma cells from clinical patients. The UP-TTDs were placed on both sides of the cultured cells with a duty cycle of 1:1 s, an applied frequency of ~150 kHz, and field intensities of 1.3 to 1.5 V/cm. After a 12-hour exposure of TTF generated by the device, the growth rates in treated and control groups were evaluated and quantified with the 5-ethynyl-2’-deoxyuridine (EdU) Cell Proliferation Reagent (Fig. 4A). As shown in Fig. 4B, the tumor proliferation rate was significantly inhibited (up to 58%; P < 0.001, t test) after treatment with the UP-TTD, indicating a successful tumor growth inhibition of the device. As simulated in fig. S1, although a slight decrease in electrical field intensity was observed in the middle of the two plates, the electric field generated in the medium can still be regarded as relatively uniform. The results showed good agreement with in vitro experiments, which did not identify significant heterogeneity in cellular responses. Because the overall intensity was still attenuated in the culture medium, to maintain the initial intensity, more charges need to be generated from UP-TTD. As tested, an approximately 53.2% power increase was required compared to the no-load condition to maintain the TTF for in vitro experiments.

Fig. 4.

In vitro experiments.

(A) Cell proliferation analysis by EdU assay; all cells were stained blue by Hoechst, and newly proliferated cells were stained green by EdU after 12 hours of UP-TTD treatment. Scale bar, 100 μm. (B) Cells treated with UP-TTD devices showed significantly reduced cell proliferation compared to control groups. Each column represents a different UP-TTD device, and dots within a column represent technical replicates. The cells for the in vitro experiments were obtained from the same patient. ***P < 0.001, t test. (C) Anatomy and composition of porcine tissue with two implanted UP-TTD locations; each location is marked with a dot. (D and E) The output of UP-TTD is affected by the implanted tissue and the degree of the curvature. (D) shows the output of the device at different depths of tissue, with a control group performed in air.

Before the in vivo functional characterization of UP-TTD, the influences of the covering tissue need to be examined. Porcine tissue, which presents a similar anatomy and composition to human skin (Fig. 4C), was used to investigate the loss of UP-TTD outputs after implantation. Expectedly, UP-TTD generated slightly lower outputs under the tissue than that of control groups (Fig. 4D), which was caused by the increased impedance and attenuation in the tissue. Specifically for 20 mm under layered tissues including skin and fat, the performance of UP-TTD decreased by ~15%. The UP-TTDs still generated output signals of ~2.1 V/cm at a depth of 20 mm with the exciting power of 0.2 W/cm2 (Fig. 4D), guaranteeing adequate TTF therapeutic fields for in vivo experiments. Because of the complex surface adaptation required during implantation, the influences of device curvature also need to discuss. As shown in Fig. 4E, the UP-TTD was still able to produce an output more than 1 V/cm when curved to 90°. However, as the curvature increases, more energy is lost, which should be considered before therapy.

The in vivo tumor therapy effect of UP-TTD was evaluated in tumor-bearing rats with a mouse malignant melanoma cell line (B16) placed into the brain of female Sprague Dawley (SD) rats (Fig. 5, A and B). Rats were randomly selected for characterization after a period of tumor implantation. All selected rats showed similar tumor areas (coefficient of variation, 0.1093; n = 4), indicating a comparable tumor growth status in these rats. Further experiments were conducted on the basis of the same batch with a similar variance of tumor growth status. The UP-TTD devices were then implanted and excited with ultrasound (0.2 to 0.4 mW/cm2) (Fig. 5A), and resected brain tissue was analyzed after 1 week of treatment. The induced tumor cells proliferated rapidly in the control groups, showing a sizeable black shadow in the rat’s brain (Fig. 5B). Compared with control groups, the treated rats showed much less cancer areas (about 78% smaller than the control; n = 4; P < 0.01; Fig. 5, C and D), indicating a significant tumor-inhibiting effect of our UP-TTD in vivo.

Fig. 5.

In vivo experiments.

(A) Images of an implanted demonstration for large UP-TTD located on the top brain region of an SD rat. (B) The induced tumor cells proliferated rapidly in the control groups, showing a sizeable black shadow in the rat’s brain. Scale bar, 0.5 mm. (C) Compared with control groups, the treated rats showed much less cancer area. Scale bars, 0.3 mm (middle) and 500 μm (right). (D) Statistical results of tumor area between UP-TTD and control groups. ***P < 0.001, t test.

For effective TTF therapy, electrical attenuation was a very important factor to consider. An intracranially implanted design was preferred to eliminate the major attenuation of traditional TTF, which is from the skull. When tested in a rats’ skull, ~12% electrical attenuation was observed, and higher attenuation is expected in human skulls. Electrical attenuation in brain tissues was also tested; for 10-mm rat brain tissue, the output potential difference between parallel plates drops from 1.7 to 1.17 V, indicating about 31% attenuation from 10-mm brain tissue. To further minimize adverse effects from the heterogeneous field intensity, a parallel plate electric field is proposed, which balances the attenuation of the electric field between the two plates and could be approximated as a uniform field. Simulations of the intensity distributions in the brain tissue in parallel mode are provided. As shown in fig. S2, although a slight decrease in intensity occurs in the center of two plates, a relatively uniform electric field can still be achieved in between, and a stable therapeutic effect could thus be provided in the experiments. Note that, for the in vivo demonstration in this work, UP-TTD was not intracranially implanted because of the small size of the rat’s head and the difficulties of the surgery; however, a significant positive therapeutic effect was still induced from our UP-TTD. In the future, a better treatment could be expected for intracranial cases.

As an implantable medical device, although the results demonstrate the feasibility of UP-TTD for the treatment of tumors, the biological effects induced by ultrasound are worth discussing. High ultrasound energy exerts unexpected thermal () and mechanical effects (, ), which may cause tissue injury including necrosis, apoptosis, and abnormal cell behaviors (, , ). For the biosafety demonstration, we recorded the temperature changes in real time after implantation. As shown in fig. S3, no significant temperature rise was recorded even after hours of operation, and an average increase of 0.3°C was observed after 60 min at the ultrasound power of 0.4 mW/cm2, which could be easily eliminated by pulse operation mode. A sham control group was also tested, with tumor-bearing rats treated with ultrasound for the same period without implanting UP-TTD (fig. S4). Neither therapeutic nor side effects were produced by the pure ultrasound treatment. The long-term compatibility of the UP-TTD was further investigated in fig. S5, which proved that no inflammatory responses were caused after 21 days of implantation. In this proof-of-concept paper, consideration of ultrasound-induced biological effects was primarily tested; however, in future studies, more extensive evaluations should be explored before further clinical applications.

DISCUSSION

In this work, a promising implantable UP-TTD was presented for the effective treatment of brain cancer. Ultrasound was used to safely transfer energy through the skin and tissues, realizing the wireless power supply of the medical device. In vitro demonstration with glioblastoma cells from clinical patients indicated a significant positive tumor inhibition effect from our UP-TTD. The UP-TTD device is designed as an implantable medical device and can be placed at the tumor site during surgery, where they are close enough to induce apoptosis in cancer cells. Effective working distance with therapeutic effect is vital for the UP-TTD and is related to the operating mode of the device. The location and the number of UP-TTD devices can be adjusted accordingly based on patients’ need and surgical conditions. If circumstance permits, then the parallel plate mode is recommended, which provides a relatively stable electric field between the two plates, and all areas between the two plates are regarded as the effective range. However, in some cases, only one plate is allowed to be implanted because of limited space or surgery difficulties. For the single-plate mode, with an initial intensity of 3 V/m, the effective working distance in brain tissue is about 30 mm (estimated at 31% attenuation per 10 mm). Notably, more UP-TTD devices could also be implanted, and the therapeutic effect increases with the number of electrodes. As an implant, the UP-TTD is designed to be flexible, biocompatible, extremely thin, and easy to attach to complex surfaces, which only requires a minimally invasive technique to implant and markedly reduces risks to the patients. Tunable TTF parameters and therapeutic properties are provided by the implanted UP-TTDs through simple and tricky design changes without increasing the size or adding circuits on the integrated chip. This device not only provides a safe, effective, and reliable implantable solution for the treatment of brain tumors but also provides new hope for the rescue of patients with brain tumors.

Compared with traditional TTFs, which treat the head as a whole and, thus, only provide a rough spatial resolution of ~1450 cm3 for a human patient (), our implanted UP-TTD device enables smaller controllable volume as low as 1 cm3, offering a more than 1000 times improvement on spatial resolution compared to traditional techniques. Note that the therapeutic effect of the TTF therapy is mainly dependent on how many directions of electric fields are applied to the tumor; it has been reported that a 10% increase in tumor growth inhibition can be achieved by increasing the TTF directions from two to three (). Compared with traditional TTF technology, the implanted UP-TTD can provide more comprehensive TTFs from all directions, thereby improving the treatment effect. In addition, the TTFs provided by UP-TTD are more concentrated and efficient, requiring much less energy supply for the same therapeutic effect. WPT overcomes the wiring problem of traditional TTFs, and tunable treatment parameters of the UP-TTD guarantees the therapy effect. Meanwhile, the UP-TTD equipment is compatible with most commercial ultrasound machines, which vastly expands the feasibility for general applications.

The impact of device size on brain functions is a critical factor to consider before further clinical applications. To avoid new injuries from the device, damaged brain regions (possibly caused by tumors or surgery) are preferred as implant locations. In addition, smaller devices could be used depending on the patient’s condition to minimize the impact on brain functions. As an initial proof-of-concept manuscript, more intensive studies need to be evaluated in the future. Because of the antimitotic mechanisms, UP-TTD is believed to have a disruptive effect on most rapidly growing cancer cells. Although TTF is most helpful for gliomas on the basis of current clinical practice, positive therapeutic results have also been found on many other types of tumors (, ), such as metastatic melanoma (, ); thus, UP-TTD is believed to have great application potentials in different tumor treatments and could be further explored in the future.

MATERIALS AND METHODS

Preparation and characterization of UP-TTD

The UP-TTD is composed of several layers of ultrathin films. A 50-μm flexible copper layer was used as the bottom electrode, a 100-μm-thick PFA film from JUYOU Company was placed as top layer, and an adjustable gap was controlled by a hollow half-rigid adhesive spacer from DEYI Company on the four sides. Electric connections and components were soldered to the bottom electrode. Then, the whole device was packaged and isolated with PDMS. After integration, all components act as one flexible membrane with a thickness of less than 0.5 mm.

Characterization of electrical outputs and ultrasound

The electrical outputs generated by the UP-TTDs were measured and recorded using an oscilloscope (Rigol, DS2202) with a voltage probe (Rigol, RP3300) with 1-megohm input impedance. In most cases, two parallel plates were used to generate an electric field. The distance separating the two plates will be carefully controlled to establish a nearly uniform electric field in the middle of the plates. The field intensity (E) is calculated by dividing the potential difference under working load (U) by the separating distance (d) between plates. Ultrasound was generated and adjusted with a commercial ultrasonic transducer (THD-M1) and generator (DK40). The calorimetric method was used as a faster and cheaper means to verify excitation ultrasound energy for UP-TTD.

Simulations

COMSOL Multiphysics was used for finite element method simulations. The physical parameters of water, Cu, air, and PFA were taken from the library of the software. A large cube water area was modeled first, and a 5-cm circular ultrasound probe load with calculated pressure was then added. The UP-TTD is composed of a 2 cm–by–2 cm–by–100 μm top PFA film, a 2 cm–by–2 cm–by–50 μm fixed-bottom copper electrode, and 100- to 200-μm air gap. The entire device was placed under the probe at 5 to 50 mm. Pressure acoustics, solid mechanics, and acoustic structure physics were used to simulate the interaction and movement of the membrane. Here, attenuation inside the brain is considered to originate primarily from the white matter (σ = 0.15 S/m; ε = 3200).

Cell culture

All the research procedures were approved by the ethics committee of Sun Yat-sen University (SYSU-IACUC-2021-B0989), and informed consent was obtained from all patients. Human clinical glioma tumors were resected through surgery and collected for use. Tumor and surrounding cells were processed and separated by mechanical trituration, trypsinization, filtration, and centrifugation. Cells were cultured in Dulbecco’s modified Eagle’s medium/Nutrient Mixture F-12 (Hyclone) containing 10% fetal bovine serum (FBS) (Hyclone), 2 mM l-glutamine, and penicillin-streptomycin solutions (100 U/ml) (Gibco). Cells were incubated at 37°C in a humidified atmosphere with 5% CO2 and medium change every other day. The mouse melanoma cell line (B16) cells were purchased from American Type Culture Collection and maintained in RPMI 1640 (Hyclone) supplemented with 10% FBS, 2 mM l-glutamine, and penicillin-streptomycin (100 U/ml) (Gibco).

In vitro experiment

The effect of 12-hour UP-TTD exposure on tumor proliferation was tested on clinical tumor cells. Cells are from clinical patients with gliomas, and more specific details were provided in the Supplementary Materials. Cells were grown on 8-mm coverslips and placed between parallel UP-TTD plates separating with a distance of 10 mm. The electric field generated by the device is measured in the medium as 1.3 to 1.5 V/cm, 150 kHz. EdU (from the BeyoClick EdU-594) was added at a final concentration of 10 μM 12 hours before harvesting the cells. For the Click reaction, cells were washed, fixed with 4% paraformaldehyde for 15 min, permeabilized in the buffer for 15 min, incubated with Click-iT reaction solution for 30 min in the dark. The proliferation results were observed under a fluorescence microscope (Leica, DiM8).

In vivo experiment

All procedures involving animals were approved by the Animal Ethical Committee of Sun Yat-sen University. SD rats were purchased and housed in Sun Yat-sen University Experiment Animal Center. Female SD rats, weighing 160 ± 5 g, were randomly divided into two groups and were anesthetized with 5% chloral hydrate (intraperitoneally, 0.7 ml/100 g) to relieve pain during the surgical procedure. The rat’s hair was shaved, the skin above the head was incised, and a 1-mm-diameter hole in the skull was drilled. B16 cells (2 × 105) were implanted in the left hemisphere of rats. The UP-TTDs were implanted under the shaved head skin and above the skull to generate an electrical field. The distance between the device and the tumor injection zone is about 5 mm. After suturing the skin over the rat’s head, we placed them individually in a standard rearing cage with ultrasound transducers 3 to 5 cm above. The ultrasound transducers were customized as larger as the cage, which could provide consistent ultrasound patterns in all area. The UP-TTD groups were treated with 150-kHz TTFs at an intensity of 1.3 to 1.5 V/cm for 6 days. After 6 days of treatment, rats were anesthetized with chloral hydrate, perfused with 50 ml of normal saline, and followed by 4% paraformaldehyde. At the end of perfusion, brain tissues were removed by craniotomy and fixed in 4% paraformaldehyde for further study. Tumor areas were quantified on the basis of histology through the initial black color of the tumor, followed by confirmation of hematoxylin and eosin staining. Ten coronal brain slices were taken around the injection sites, as fig. S6, and the largest tumor area was selected for analysis.