Molecular Approach to Brain Tumors Improves Treatment

Traditionally, brain tumors have been diagnosed based solely on appearance—highly skilled neuropathologists peer at small surgical biopsies under the microscope, and assign a tumor type and stage based on what they see. But such diagnoses are sometimes problematic—different tumor samples that look identical to the trained eye could have vastly different prognoses and response to treatment.

Now, rapid advances in identifying molecular markers for brain cancer are giving rise to new diagnostic schemes that use both genetic and histological information to categorize tumors.
At Mass General Hospital, neurosurgeon Daniel Cahill, MD, PhD is a leader in the treatment of a type of glioma marked by mutations in the gene for isocitrate dehydrogenase 1 (IDH1). Cahill’s work shows the power of the new gene-based biomarkers to help physicians to make better diagnoses and tailor treatments for individual patients.

A biomarker for better survival
The IDH1 story started in 2008, when researchers at Johns Hopkins and at Duke discovered that many gliomas (approximately 20% of adult diffuse gliomas, depending on the histological type) carry a mutation in the gene for IDH1, an enzyme involved in cell metabolism. The investigators quickly realized that the IDH1 mutation marked a fraction of gliomas that were histologically indistinguishable from tumors with normal IDH1, but that behaved quite differently. The IDH1 mutant tumors occurred in younger people, tended to grow much more slowly and had a much better prognosis than gliomas without the mutation. For patients with aggressive glioblastoma with an intact IDH1 gene, median survival time is measured in months; patients with IDH1 mutant tumors often live for years to decades.

Today, a sensitive and specific antibody test for the most common IDH1 mutation is widely available, and the implications of the test are profound for physicians and patients alike. In addition to identifying a much better prognosis, it has become clear that the IDH1 mutation can also affect the response to treatment.

In an influential study of surgical outcomes in patients with and without IDH1 mutant tumors, Cahill found that in patients with malignant astrocytomas, those who had the mutation and got aggressive surgery had better survival than those with less aggressive surgery. On the other hand, patients without the mutation showed no benefit despite taking the risks of more aggressive surgery to resect their tumors.

That's important because surgery to remove a tumor carries the chance of neurological damage. Previously, the two types of tumor were impossible to distinguish histologically. “Some people say we should always do aggressive surgery in everybody, but you can’t just take unnecessary risk for no benefit. Patients pay a price and it doesn't make sense if they are in the group that does not derive a benefit from the more aggressive operation,” Cahill says.

Besides altering a surgical plan, the presence of an IDH1 mutation could also mean thinking about different approaches to radiation or chemotherapy that minimize side effects to maximize quality of life. For example, radiologists at MGH are in the midst of a unique study looking at the use of proton beam therapy in lieu of traditional radiation in patients with IDH1 mutant gliomas. Proton beam therapy can be precisely aimed at a tumor, sparing surrounding healthy tissue from damage. “We know that protons are very good at minimizing long term side effects, which is an important consideration for IDH1 mutant glioma patients,” says Cahill. “We’re really trying to personalize and optimize the therapy for this group of people.”

Tests for IDH1 and other markers are poised to play a greater role in the diagnostic schema of brain tumors. “As we look to treat patients in the future, there is an emerging imperative to arrive at an integrated diagnosis,” says Cahill, one that encompasses both histological and molecular data. That is reflected in an upcoming revision of the World Health Organization criteria for the classification of brain tumors, due out in 2016 that will require molecular markers in addition to histological characterization to define some brain tumor types.
Because most standard treatments were developed before IDH1 mutations were discovered, Cahill says that patients who find they carry the mutation should be ready to pause and reassess their treatment plans. “It’s a key time for a second opinion, to make sure things make sense before you move forward with treatment.”

Going live
For patients with brain tumors, it is not unusual to have two surgeries. The first gives surgeons an initial look at the tumor, and the chance to retrieve tissue for histological and genetic analysis. Days later, the results come back and a treatment plan is formulated. Especially in cases of IDH1 mutant tumors, the plan often involves a second surgery to more completely resect the tumor.

In a perfect world, Cahill says, the genetic information would be available during the first surgery, where it could be used to guide surgical decisions in real time, for instance to pursue aggressive resection aided by intraoperative MRI, and obviate the need for a second operation.

To that end, Cahill and colleagues are now working on speeding up the detection of IDH1 mutations. He and a team from the Broad Institute in Cambridge recently developed a rapid PCR-based IDH1 test that takes just 40-60 minutes from biopsy to result.

In that time frame, surgeons might be able to learn the IDH1 status while the patient is still in the operating room. “This could allow us to course correct during the first operation,” Cahill says. “If we could get that information to the team early, during the first operation, then they could make the decision about more aggressive or less aggressive with good information. That might save the patient from needing to have a second operation.” Cahill and colleagues are now enrolling patients in a research project, to see if the test is sensitive, specific and feasibly achievable during the span of the surgery.