Dec. 20, 2024
Mayo Clinic researchers are working to apply their latest discoveries about brain tumor heterogeneity to personalized patient care. These efforts use artificial intelligence and multiplatform molecular data to build predictive modeling of tumor behavior.
"AI modeling using genomic, methylation and other testing is transforming care for patients," says Gelareh Zadeh, M.D., Ph.D., chair of Neurosurgery at Mayo Clinic in Rochester, Minnesota. "We have established the clinical benefit and utility of this predictive modeling. What is needed now is widespread adoption of these approaches to personalize care and inform clinical trial design."
The research encompasses benign and malignant tumor types. The results of this laboratory work, as well as intraoperative molecular testing of tumors, can help guide decision-making about tumor aggressiveness and potential responses to treatment. It also helps guide decision-making for patients on the role of surgery, radiation therapy and surveillance.
As a major quaternary center, Mayo Clinic has the expertise and resources to expand the reach of this translational work. "The motivation behind our research is to maximize patient outcomes and allow for patients to make better informed decisions about their brain tumor care," Dr. Zadeh says.
Research breakthroughs
Dr. Zadeh's research findings have direct clinical applications. Meningioma is a particular focus of research in her laboratory. Her team also studies glioma, brain metastasis, pituitary tumors and neurofibromatosis.
As described in Nature Medicine, one study used multiplatform molecular, treatment and outcome data to identify molecular predictors of treatment response in meningiomas. The researchers characterized the benefits of differential degrees of tumor resection and dural margin treatment across different molecular classifications, in addition to identifying a group of molecularly defined radiotherapy-resistant meningiomas.
"Our research findings support the rationale for investigating radiotherapy results for meningiomas in the context of molecular classification," Dr. Zadeh says. "We also need to consider clinical trials informed by molecular pathology to investigate treatments for radiotherapy-resistant meningiomas."
Another multiplatform study uncovered a novel group of IDH-mutant gliomas that share metabolic features with IDH-wild-type tumors. As reported in Acta Neuropathologica, the researchers integrated matched epigenome-wide methylome, transcriptome and global metabolome data in patients with glioma. That study demonstrated the importance of characterizing each individual patient tumor.
"This metabolic heterogeneity among IDH-mutant gliomas has considerable implications for managing patients and for future clinical trials," Dr. Zadeh says. "Genome-based classifications are commonly used in clinical practice. But it's important to remember that a tumor's genotype doesn't always reflect the phenotype and behavior. The metabolic profile can facilitate a more comprehensive understanding of glioma biology, when coupled with genomic data."
DNA methylation profiling is an especially important tool. In a study published in Nature Medicine, Dr. Zadeh and colleagues described their use of methylation profiling to predict brain metastases from lung adenocarcinomas.
Methylation testing has revolutionized the understanding of brain tumor biology. "We now have multiple clinically relevant classification and outcome prediction tools using methylation testing," Dr. Zadeh says. "Our focus is to reach more patients and be able to take our discoveries to all patients at the first step of their journey in managing their brain tumors and cancer care. The next step is to operationalize this technology — or a more accessible surrogate — into routine clinical workflows."
"Knowing the specific genetic composition of a patient's tumor helps us design the most effective treatment strategy and management care plan," Dr. Zadeh says. "Our priority is always what is best for the patient."
For more information
Wang JZ, et al. Molecular classification to refine surgical and radiotherapeutic decision-making in meningioma. Nature Medicine. 2024;30:3173.
Nassiri F, et al. Metabalogenomic characterization uncovers a clinically aggressive IDH mutant glioma subtype. Acta Neuropathologica. 2024;147:68.
Zuccato JA, et al. Prediction of brain metastasis development with DNA methylation signatures. Nature Medicine. In press.
Refer a patient to Mayo Clinic.