Kelly Brown was 57 when she started complaining of frequent headaches. "It must be a sinus infection," her son, Connor Ferris, recalled her saying. But a family vacation in Italy prompted him to suspect a more sinister cause. "She wasn't her usual go-go-go self," he said of his mother, a flight attendant, former nurse and single mom to three young adults. "My mother was an intrepid traveler, but on this trip, she was constantly tired and preferred to stay at the hotel. At a restaurant, she fell asleep – that wasn't my mom."

Back home in Lake Tahoe, Nevada, Brown sought medical advice. An MRI revealed glioblastoma, the most common and deadly brain tumor, with an average survival of just 15 months. Treatment involved removing as much of the tumor as safely possible, followed by radiation and chemotherapy, with the tumor expected to recur months later.

There were closer hospitals, but Brown opted for treatment at UCSF Medical Center, ranked No. 2 in the nation for neurology and neurosurgery by U.S. News and World Report's Best Hospitals survey and the site of more than 25 state-of-the-art research laboratories dedicated to developing new treatments for brain tumors.

A new diagnosis and a clinical trial

After her initial treatment, Brown did surprisingly well. It wasn't until four years later that her tumor recurred. By then, the UCSF Brain Tumor Center had developed two new tools to provide a more precise diagnosis than the standard scans and examination of cells under a microscope. Both of these molecular tools are now available to all UCSF patients who may have a malignant brain tumor.

The first, UCSF500, is a test assessing 500 cancer-associated genes that might be mutated in the patient's tumor tissue. These include mutations that can be attacked with targeted therapies that have already been approved by the Food and Drug Administration (FDA) or are being tested in clinical trials, said Dr. David A. Solomon, a molecular neuropathologist and principal investigator at the UCSF Brain Tumor Center.

The second tool, known as DNA methylation profiling, evaluates so-called "methyl marks" on the DNA within cancer cells that regulate activity of the genes involved in cancer growth. The pattern of these methyl marks across the genome provides Solomon with critical clues about where the cancer originated, the genes that fuel growth and its potential to spread. The test also offers additional information that allows doctors to reach the most precise diagnosis and gain additional insights into personalized treatment. Research shows that this tool is the most accurate method for brain tumor classification and can change the pathologist's diagnosis in as many as 1 in 8 cases.

For Brown, these molecular tests – which used tissue extracted during surgery after her cancer recurred – revealed that the tumor was not glioblastoma after all but a different type with a better prognosis. Better yet, Brown's tumor matched with a targeted therapy being tested in a clinical trial, in which she would eventually enroll.