US Senator Edward Kennedy of Massachusetts died in August 2009, only 15 months after he was diagnosed with brain cancer. To researchers who study gliomas, Kennedy's case is all too common. Most of these tumors are aggressive, but the real frustration is that only a fraction of patients will respond favorably to treatment, as many of these malignancies are also resistant to most therapies. Even if they could tell whether a patient would respond, most clinicians would not be able to suggest an alternate therapy because so few treatments exist.
According to the University of North Carolina School of Medicine's Ryan Miller, the key to determining disease subtypes is a more accurate tumor classification system based on molecular data gathered from DNA microarray-based gene expression profiling. In a recent British Journal of Cancer review article, Miller and his colleagues discussed the last 10 years of human glioma tumor classification research. "The problem is that we know that patients with the same tumor type under the microscope respond differently to therapy, and so the idea was to utilize molecular techniques to molecularly stratify tumors into smaller and smaller cohorts that may respond more uniformly to different therapies," Miller says.
Over the course of the decade, the field has progressed rapidly, going from proof-of-principle studies early on to the classification of smaller and smaller glioma subtypes more recently. But progress in glioma classification has not been accompanied by progress in the development of targeted treatments. "If you get diagnosed, you're going to get the same therapy that everybody else gets, which is concurrent radiation and a drug called temozolomide, with follow-up temozolomide," Miller says. For other tumor types, like breast cancers, a patient can be tested for any number of biomarkers to inform a clinician's choice for treatment. With gliomas, Miller says, the standard treatment only works in about 10 percent or 20 percent of patients.
Because of improvements in tumor classification thanks to microarray data, researchers are starting to develop assays to stratify patients based on the molecular subtypes of their tumors, which they hope will lead to more specialized clinical trials for treatments and, eventually, more personalized therapies for patients. And the advent of whole-genome sequencing is also going to allow researchers to be more precise in their molecular stratification efforts.
Although drug companies have, so far, been reluctant to spend the money, time, and effort required to develop treatments for small populations, like patients with gliomas, Miller is hopeful that understanding the different subtypes could lead to more specialized — and more successful — clinical trials. "Personalized therapy means the individual tumor and patient, but at this stage it means taking histologically or pathologically homogeneous cohorts of patients and using molecular diagnostics to divide those further into smaller, much more molecularly homogeneous, groups. Not only molecularly homogeneous, but homogeneous with respects to response to therapy," he says.