A Closer Look
Using Imaging to Pinpoint Genetic Mutations in Brain Tumors
Recent studies suggest that a noninvasive imaging technique can identify the presence of certain genetic mutations in gliomas, the most common type of brain tumor, by detecting a substance produced as a consequence of those mutations. If further studies validate the finding, oncologists may be able to use this approach to diagnose tumors carrying the mutations, distinguish different glioma subtypes,monitor tumor progression, and detect recurrences—all without repeated surgeries or biopsies.
The technique uses magnetic resonance spectroscopy (MRS) to detect 2-hydroxyglutarate (2-HG), a chemical that is scarce in normal tissues but accumulates in gliomas that harbor mutations in two related genes, IDH1 and IDH2. Until now, the only way to detect IDH mutations or 2-HG in a tumor was a biopsy and analysis of tissue samples. Taken together, the new work provides evidence that 2-HG could serve as a noninvasive biomarker of tumors with IDH mutations.
More than 70 percent of adults with invasive, lower-grade (grade II or III) gliomas carry IDH mutations. IDH mutations are uncommon in primary forms of glioblastoma, but they are found in many secondary glioblastomas, which arise from lower-grade gliomas. Glioblastoma is the most aggressive type of glioma.
Tumors with IDH mutations make an aberrant form of the enzyme isocitrate dehydrogenase (IDH), which plays an essential part in a metabolic pathway that cells use to generate energy. The mutations not only impair the enzyme's normal function but also give it the ability to make 2-HG.
MRS can be done at the same time and with the same magnetic resonance imaging (MRI) equipment medical centers use for diagnostic imaging. While MRI scans provide anatomical and structural information, MRS provides information on cellular activity by detecting metabolites in tissue, explained Dr. Ovidiu Andronesi, an instructor in radiology at Harvard Medical School and lead author of one of the recent studies.
Currently, a definitive diagnosis of glioma requires a biopsy and examination by a pathologist, who assigns a grade of I to IV to the tumor, with grade IV (glioblastoma) being the most aggressive. However, "two grade II tumors may look the same under the microscope, but one may be very slow-growing and another may kill the patient within 2 years," noted Dr. Howard Fine, chief of NIH's Neuro-Oncology Branch, who was not involved in the MRS studies. "We need to get to the point where we're [classifying tumors] on the basis of the genetics and the molecular biology," he said.
Thus, researchers were interested to find that patients with glioblastoma who have IDH mutations generally live longer than patients who lack the mutations. Several studies have also found that IDH mutations are associated with better outcomes in lower-grade gliomas, although this link is not as firmly established.
"Across [glioma] patients within the same tumor grade, there is a prognostic advantage to having an IDH mutation," said Dr. Susan Chang, director of Neuro-Oncology at the University of California, San Francisco (UCSF), who collaborated on a study led by UCSF radiology and biomedical imaging professor Dr. Sarah Nelson. If further studies confirm this advantage, assessing a tumor's 2-HG status with MRS could be used to divide patients into groups with similar prognoses for clinical trials of new therapies, Dr. Chang said.
MRS could also be used to diagnose IDH-mutated gliomas before surgery and to distinguish them from nonmalignant brain lesions, such as those due to multiple sclerosis, said Dr. Elizabeth Maher, a neuro-oncologist who led a study at the University of Texas Southwestern Medical Center with physicist Dr. Changho Choi. Although the approach is still experimental, Dr. Maher said she has already used it to diagnose a low-grade glioma in the brain stem of a 20-year-old patient, helping the patient avoid a risky biopsy procedure.
Researchers have used several MRS techniques to detect 2-HG, which can be difficult to distinguish from some common brain metabolites with established clinical MRS methods. Dr. Nelson and colleagues noninvasively detected and measured 2-HG in small tissue samples taken from recurrent gliomas in 52 patients originally diagnosed with grade II glioma. They found a strong correlation between the presence of 2-HG in tissue samples and the presence of IDH1 mutations in tumor tissue. (IDH1 mutations are much more common than IDH2 mutations.) They also correlated varying levels of 2-HG in IDH-mutated tumors of different grades with several histopathology parameters, including the density of tumor cells in a sample.
While the UCSF researchers used MRS to detect 2-HG ex vivo (outside the body), Dr. Andronesi's team showed that noninvasive MRS could also detect 2-HG unambiguously in vivo (in the body) in two glioma patients with IDH-mutated tumors. They did not detect 2-HG in four healthy volunteers or in four primary glioblastoma patients who lacked IDH mutations.
In a perspective article accompanying the two studies, Drs. Philippe Metellus and Dominique Figarella-Branger of the Hôpital de la Timone in Marseille, France, detailed the potential clinical applications of "these major findings." However, they wrote, "the MRS methodology used in these studies is not available in all clinical radiology settings. Also, in vivo proof-of-principle data will need to be reproduced in a larger cohort."
Those concerns were addressed at least in part by the UT Southwestern study and by a study led by Dr. Linda Liau, professor and vice chair in the Department of Neurosurgery at the University of California, Los Angeles. Both used in vivo MRS methods that could be done routinely in most hospitals and medical centers with relatively minor modifications of standard MRS techniques.
Dr. Liau's team showed that, in 24 of 27 patients with gliomas of various grades, in vivo MRS could detect the higher 2-HG levels found in tumors with IDH1 mutations. They also showed that 2-HG levels measured noninvasively by MRS in patients prior to surgery correlated with 2-HG levels measured in corresponding tumor samples with a laboratory technique known as liquid chromatography-mass spectroscopy (LC-MS).
And, in their study of 30 patients with grade II, III, or IV gliomas, Dr. Maher and colleagues showed that in vivo MRS detection of 2-HG correlated with mutations in IDH1 or IDH2 and with increased 2-HG levels measured in tumor samples by LC-MS. The team also estimated the concentrations of 2-HG in patient's tumors.
Taking It to the Next Level
The MRS methods used by each group have advantages and disadvantages, and most teams are still refining and optimizing their approaches. "The jury is still out on which is the best approach," Dr. Nelson said. "One would have to do a head-to-head comparison under similar circumstances in order to establish [which is best]."
As Dr. Andronesi noted, researchers now need to "establish the sensitivity and specificity of these methods—so-called validation…using genomics as the gold standard" for detecting IDH mutations. "It's important that validation is done not only in one center and by one group, but that other people can replicate [the findings]," he said.
In anticipation of testing possible clinical applications, researchers are working to show that in vivo MRS can be used not only to detect 2-HG but also to accurately measure the concentration of 2-HG in a tumor. "Showing [that] something is there is one thing," Dr. Nelson noted. "Being able to measure it to a plus or minus 10 percent accuracy, which you need to do in order to look at subtle changes, is something else."
If 2-HG levels are shown to reflect changes in tumor growth or progression, physicians could use MRS measurement of 2-HG levels to monitor responses to therapy, or to detect early signs of tumor progression in patients with IDH-mutated gliomas. "Eventually, about 80 percent of low-grade tumors will transform into the highest-grade tumors, and we have no way of picking that up [on an MRI] before there's an outgrowth of a big tumor," Dr. Maher said.
Similarly, physicians hope to use MRS to noninvasively detect tumor recurrence after surgery. Doing so is "one of the biggest challenges that face oncologists today," Dr. Nelson noted, because changes on a standard MRI that suggest a recurrence may also be due to effects of treatment.
"In the broader scheme of things, it may be that 2-HG is not the only important metabolite," Dr. Liau said. Her group as well as Dr. Maher's and Dr. Nelson's are using in vivo MRS to study changes in a range of metabolites in gliomas. Such information could be used to establish "molecular fingerprints" of tumor grade and disease progression and to understand more about the basic biology of why gliomas form and develop.
Related reading: "Newly Discovered Mutation Found in More Brain Tumors"