Several years ago, Rakesh Jain and his lab members began peering at tumors in living mice, using implanted glass windows that let them monitor tumors and the blood vessels that fed them. What they found was a disorganized maze of vessels. In some places, blood rushed briskly; in others, it stagnated. Some vessel walls were tight, others were filled with pores that leaked fluid. High pressure within the tumors created swelling in the surrounding tissues, further reducing blood flow in the tumor.
These observations led Jain to suggest in 2001 that angiogenesis inhibitors could be used to repair the disorganized vasculature of tumors. Although it seems counterintuitive to fix a tumor’s blood supply, Jain argued that this pruning or “normalization” could offer a clinical benefit. Repairing the disrupted roads surrounding tumors, he argued, would pave a clearer route for cancer drugs. And better circulation in the oxygen-deprived environment of tumors would make radiation and chemotherapy more successful.
A study published in the January Cancer Cell offers supporting evidence for Jain’s predictions. He teamed up with Tracy Batchelor and Gregory Sorensen, all HMS faculty members at Massachusetts General Hospital, to carry out a phase II clinical trial of an angiogenesis inhibitor in patients with recurrent brain tumors. The first results from that trial offer evidence that angiogenesis inhibitors can be used to normalize the environment of a tumor. Ideally, these drugs should be used in combination with radiation and chemotherapy to make the treatments more effective.
Seeing the Therapeutic WindowJain, the A. Werk Cook professor of radiation oncology (tumor biology) and a faculty member in the Harvard–MIT Division of Health Sciences and Technology (HST), collaborated with Batchelor, associate professor of neurology and director of the Stephen E. and Catherine Pappas Center for Neuro-Oncology at MGH. They studied the angiogenesis inhibitor AZD2171, developed by AstraZeneca, which works by inhibiting receptors for the blood vessel growth factor VEGF. Batchelor tested the drug in patients with recurrent glioblastoma, the most aggressive type of brain tumor, who had not responded to conventional treatment. Though Jain is a scientific adviser to AstraZeneca, the trial was funded through the National Cancer Institute. This study reports on the first 16 patients to enter the trial. Patients received daily doses of AZD2171 in pill form.
“My hope was that, in addition to benefiting patients, we could also learn something about normalization,” said Jain. To do that, however, the researchers needed to find a way to see inside the brains of their patients. Without the ability to do serial biopsies, they had to rely on imaging techniques to watch their progress.
They worked with Gregory Sorensen, HMS associate professor of radiology, who develops new ways to image events in the brain with MRI. Sorensen said that imaging is usually performed only every couple of months to check tumor size. In this case, the team was not just interested in the tumor’s anatomy but its physiology; they needed to visualize what was happening in the small, complex network of blood vessels around the tumor from the very beginning of treatment.
The researchers used several different MRI techniques, including novel ones developed at the MGH–HST Martinos Center for Biomedical Imaging, where Sorensen is co-director. “We investigated as many different aspects of vessel normalization as we could,” Sorensen said. One problem was how to determine whether the tiny capillaries surrounding the tumor were shrinking. Sorensen explained that MRI can be used to visualize small-scale events beyond its resolution, as long as the events occur in abundance. The team could estimate the relative size of blood vessels, as well as measure blood volume and flow in the tumors and permeability of the blood–brain barrier.
Almost immediately after beginning treatment, patients experienced a period in which small blood vessels decreased in size and became less leaky. The treatment also caused tumors to shrink, but the overall benefit varied among patients. In three quarters of the participants, tumors shrank by at least 25 percent, and they shrank by 50 percent in about half of participants. Jain and his colleagues suspected that the drug would provide a “window” of normalization before the tumors developed resistance to the therapy. With the imaging analyses, they were able to define this window as a period of about a month, lasting several months in some patients. “We saw evidence in many different kinds of imaging, and that helped reassure us that this wasn’t an artifact,” Sorensen said.
Added Clinical BenefitIn addition to the expected effects on tumor size and blood vessels, the researchers also found an unexpected but important benefit: the patients’ brains became less swollen. Batchelor explained that pressure created by a brain tumor drives fluid into the surrounding white matter. The resulting swelling can be a significant problem for patients and often requires treatment with toxic steroids. Treatment for this condition has not changed for decades, Batchelor said, but “it looks like this new class of drugs might be an effective way of treating brain edema.” When Sorensen’s team used a relatively new method for imaging the connecting fibers of white matter in the brain, “we saw white matter tracts becoming more visible” with treatment, he said, which suggests that as the swelling receded, white matter regained some of its integrity.
The current results are too preliminary to show an overall effect on patient survival. Final data from the completed trial, which included 30 participants, is forthcoming. “In order to definitively answer the question of a possible survival benefit,” Batchelor added, “we do feel that the next step is a randomized trial,” and planning is currently under way for one that will use the methods developed in this trial.
The researchers uncovered markers in the blood of patients that correlated with the tumors’ escape from treatment. Jain said that such markers would help clinicians, since they can indicate when the window of opportunity for treatment opens and closes without the need for sophisticated imaging. “They’re not only biomarkers for escape from this therapy,” Jain said, “but also potential targets.” For instance, one marker is an angiogenic molecule called basic FGF, which Jain said has not received much attention in the past decade. As anti-VEGF treatment fails, this molecule may become a secondary drug target.
The window of normalization that AZD2171 creates would be a good opportunity for delivering other cancer treatments, and the team plans to test this approach in newly diagnosed glioblastoma patients. They would like to see whether AZD2171 and other angiogenesis inhibitors can be used in concert with standard cancer treatments from the beginning of a patient’s therapy. Many had hoped that angiogenesis inhibitors themselves would kill tumors, but so far their effectiveness has proved to be limited in clinical trials. Jain, however, believes they will be an important component of a multilevel attack on cancers.
