Breaking the Blood-Brain Barrier

Using high-intensity focused ultrasound to treat malignant brain tumors

Email Print

Figure 1: Various images of a glioblastoma multiforme: Preoperative (A) contrast-enhanced CT, (B) contrast-enhanced MRI, and (C) flair MRI imagery. (D) shows an axial, intra-operative B-mode diagnostic ultrasound image. The tumor ("T") is largely iso-echoic within a dark rim in the ultrasound image, and bright rim in contrast-enhanced MRI imagery. These imaging modalities also show hemorrhage ("H") and normal brain ("B"). ("E") marks edema. Note the asymmetric structure of the brain due to the presence of the tumor – for example, the "mid-line" of the brain in (B) and (C) is displaced away from the brain's center. This mass effect, if left unchecked, alters brain function.

Figure 2: Human brain before (A) and after (B) sub-total GBM resection. Brain-tumor resection margin with implanted Gliadel wafers (C), to deliver drugs past the BBB, targeting residual tumor and invaded tumor cells.

Figure 3: An intact blood-brain barrier (left-most figure) keeps unwanted chemicals, as well as most systematically delivered therapeutic agents, within the blood stream and away from the brain. The primary piece of the BBB is the specialized endothelial cells that line the blood vessel. A disrupted BBB (right figure) allows blood-borne chemicals to directly enter brain tissue, by either going between the endothelial cells – as shown – or by going through those cells via a variety of other biological mechanisms, such as transcytosis (not shown), among others.

Figure 4: Using HIFU, Mesiwala et al (2002) opened the BBB to systematically introduced dyes (as shown in (A) plan- and (B) coronal views). Electron microscopy of brain following HIFU disruption of the BBB – done here (C) without electron-dense tracers – showed HIFU-induced BBB opening at least via disruption of tight junctions.

Figure 5: A solid cone HIFU device (A) for opening the blood-brain barrier intraoperatively after resection of bulk tumor by the neurosurgeon. When applied to the exposed cortex of rats, it reliably produced BBB opening to Evans blue (B) with flux volumes that were, on average, three-times larger than the damage volume, with damage always at the surface of the brain, where resection procedures already damage the brain. Studies done with tritiated chemotherapeutic agents relevant to the treatment of brain tumors have demonstrated enhanced flux of those drugs from the blood stream and into the same brain tissue that demonstrates enhanced dye flux.

The first medical response to the presence of a brain tumor is often its resection, both to alleviate mass effect, and to obtain tissue for neuropathology-based diagnosis, itself necessary for guiding adjunctive therapy. Malignant brain tumors typically recur, and do so, at least at the tumor resection margin. Many current chemo- and radiation-based therapies target recurrence at and near the margin, with measurable success.

In this article, the authors review a new strategy for delivering chemotherapeutics to address brain tumor recurrence. It uses high-intensity focused ultrasound (HIFU) to transiently open the blood-brain barrier (BBB) over a significantly large volume of brain at and near the resection margin in order to enhance the subsequent delivery of systematically introduced chemotherapeutic agents into a primary region of brain tumor recurrence.

High-Grade Gliomas

Many primary (i.e., glioblastoma - GBM) and metastatic (i.e., lung) tumors in the brain form a central mass (Figures 1A-1D), as well as infiltrate its periphery (e.g., reference 12 and citations within). This central mass can eventually produce acute alterations of patient's behavior and quality of life - usually the first clues to the presence of the brain tumor. This "mass effect" may be reduced or removed by resection of the bulk tumor (Figures 2A and 2B), with the added benefit of allowing histopathological analysis that guides subsequent adjunctive therapy.

A significant increase in length of survival is associated with removing the central mass if that resection includes the enhancing rim, if this can be achieved without an unacceptable reduction in the quality of survival. However, successful treatment of the bulk tumor does not cure the patient, due to individual tumor cells that have invaded well beyond the imaged tumor.

Individual GBM cells invade beyond abnormalities detected with MRI, usually invading more than 2 cm beyond the bulk tumor as defined by MRI. Co-author of this article, Daniel Silbergeld, and colleagues demonstrated that an MRI normal human brain, beyond the MRI borders of bulk tumor, is histologically normal, but contains GBM cells that can be grown in culture. Co-author Pierre Mourad, PhD, and his team showed with a rat model of GBM that by three days after implantation of GBM cells in the brain, those cells had invaded throughout the brain and achieved a stable tumor cell density until near the time of death.

These invading tumor cells are problematic for a number of reasons: They cannot be surgically removed; they are protected from systematically-applied chemotherapeutic agents by an intact BBB that curtails the flux of many chemicals from the blood into the brain; and they rarely cycle, which makes these cells difficult to treat effectively with radiation therapy and with whatever chemotherapeutic agents that make it past the BBB. Moreover, these invading cells are responsible for GBM recurrence at the resection margin (even after hemispherectomy).

Finally, these cells, by unknown mechanisms, can lead to progressive neurologic dysfunction without evidence of mass effects or recurrence of bulk tumor. Therefore, treatment requires attacking both the mass and the secondary cells beyond the mass.

Current adjunctive therapeutic modalities, including radiation applied in a number of ways, implanted biodegradable polymer wafers (Gliadel™) that gradually elute a chemotherapeutic agent and systemic chemotherapy, are generally undertaken to reduce the number of viable tumor cells in or just adjacent to (i.e., within millimeters) the tumor resection margin or bulk tumor remnants. These additional therapies can enhance patient survival.

Figure 2C shows the placement of Gliadel wafers at the resection margin. These biodegradable drug carriers elute chemotherapeutic agents directly into the brain tissue at that margin as well as into the surrounding cerebrospinal fluid and thereby into the rest of the brain. In this fashion, they have extended the lives of patients with re-current GBM compared with systematically delivered chemotherapy. This suggests that getting such agents beyond the blood-brain barrier (BBB) may improve the otherwise dismal prospects of brain-tumor patients. These results with Gliadel wafers have motivated us to pursue research in support of a new strategy for treating brain tumors, centered around the opening of the blood-brain barrier by HIFU in conjunction with surgery.

Blood Brain Barrier

The blood-brain barrier plays a protective role in the central nervous system (CNS) by restricting the movement of substances into the brain and spinal cord. It is composed of endothelial cell tight junctions, basal lamina and glial processes.

Adding to the protective features of the BBB is the "sink effect," a gradient favoring passage of substances in extracellular fluid from the brain to the cerebrospinal fluid (CSF), with the CSF constantly circulating and carrying substances away.

As a result of the BBB, chemotherapy has been relatively ineffective in the treatment of CNS diseases, including malignancies (both primary and metastatic) and infections. In particular, although the BBB is frequently not intact in the center of malignant tumors and abscesses, as demonstrated by contrast enhanced CT and MRI (Figures 1 and 2), the presence of an intact BBB at and beyond the proliferating edge of a tumor - which can be the site of tumor resection - has been proposed as a major factor for failure of chemotherapy for CNS neoplasia.

HIFU Applied to the Brain

HIFU can generate at many centimeter's distance from its source within a rice-grained size volume (reviewed in) profound physical changes in tissue within that volume. Those physical changes can, in turn, generate therapeutic effects.

As reviewed by Mourad, early work on therapeutic ultrasound focused primarily on the brain. Dominant in the field at the middle of the 20th century were the Fry brothers, whose work culminated in successful, short-term treatment of patients with movement disorders. Their ablative therapy required a craniotomy and a "water hat" placed above the skull defect, into which transducers were lowered to apply the appropriate acoustic protocol.

Following the development of anti-Parkinsonian medicines, however, HIFU for treatment of the brain fell into disuse almost without exception until recently, when Hynynen and colleagues began research efforts to treat brain disorders with HIFU. Their special focus has been on transcranial application of ultrasound of likely significant clinical benefit for movement disorders, where the locus of the disease is well-known and well-circumscribed, where there is no need to remove tissue acutely for therapeutic purposes and no need to retrieve tissue for diagnostic purposes.

Ali Mesiwala, MD, et al, showed that HIFU applied intra-operatively in the rat can open the BBB (Figure 4B and 4C) without inflicting collateral brain damage as determined histologically in acute studies, with comparable results in extended survival studies. Mesiwala's team also showed that HIFU-induced BBB opening persisted for at least three days following HIFU application.

Finally, they showed that electron microscopy images of the BBB after HIFU application (Figure 4D) are consistent with the hypothesis that HIFU opens the BBB at least by disrupting the tight junctions between the specialized endothelial cells that constitute the major functional component of the BBB.

Given the formidable impediment that the BBB offers to the treatment of brain tumors and movement disorders, transient and largely non-destructive opening of the BBB may someday lead to meaningful clinical impact. Recent advances have moved this basic idea a little closer to clinical reality. For example, we have constructed and tested a robust, hand-held and sterilizable "solid" cone (Figure 5A) that, when applied to exposed brain, reliably opens the BBB not only to systematically introduced dyes (Figure 5B), but also to therapeutic drugs.

Best Strategy

Successful treatment of infiltrative brain tumors requires the targeting of both the bulk tumor and the microscopically invading individual tumor cells. The bulk tumor typically requires immediate treatment - usually surgery - to relieve acute symptoms, with the added benefit of obtaining tissue to diagnose brain tumor type, which, in turn, guides subsequent adjunctive therapeutic strategy. This strategy of necessity leaves behind the invaded tumor cells that are inadequately attacked by current adjunctive therapies, and which ultimately form a bulk tumor at the resection margin.

Given that neurosurgical intervention often follows the identification and preliminary diagnosis of many brain tumors (Figures 6A and B), we propose the following strategy for use of HIFU. Apply HIFU immediately after bulk tumor resection (Figure 6C), to induce BBB opening at and near the edges of the former tumor site (Figure 6D). For up to three days after surgery, apply chemotherapeutic agents systematically at doses lower than normal, thereby producing the same dose of agent at a relatively large, but targeted volume of the brain. If preliminary engineering designs become real, this volume of enhanced drug flux could measure many millimeters deep beyond that margin. Enhanced delivery of drugs to the site of re-occurrence (both acutely after surgery, as described here, and at other stages during patient treatment, currently under consideration) would in principle and in the spirit of Gliadel™, or even in conjunction with it, retard the re-growth of new tumor at the resection margin. Much work still needs to be done before the new strategy presented here has a chance to make a meaningful impact on patient care. However, given the dismal prognosis of patients with malignant brain tumors, meaningful research on any new treatment strategy is worthwhile.

Authors' note: This work was supported in part by a training grant (K-25 NS02234-01) from the National Institutes of Health, National Institute of Neurological Disorders and Stroke (PDM) as well as the McDonnell Center for Cellular and Molecular Biology Award #26275D, and training grants (KO8 NS01730 and NS69640) from the National Institutes of Health (DLS). We thank the large number of students who have helped with this project over the years, including Brett Anderson, Elizabeth Dahl, Ali Mesiwala, Lisa Nguyen, Rachel Sparks and Sara Vaezy.

References

— Pierre D. Mourad, PhD, is a senior research scientist in the department of neurological surgery, Applied Physics Laboratory at the University of Washington, Seattle. Daniel L. Silbergeld, MD, is chief of neurosurgery and an attending neurosurgeon at the University of Washington School of Medicine. Questions and comments can be directed to editorial@rt-image.com.