Can Patient-Derived Xenografts Predict Treatment Responses in Glioblastoma?

A new study from the Society for NeuroOncology has found that xenograft-based platform-independent gene signatures may help to predict patient response to chemotherapy, radiation, and combination therapy for glioblastoma. Investigators examined gene expression and treatment response data from 31 orthotopic glioblastoma patient-derived xenografts (PDXs), and developed gene signatures for radiotherapy, chemotherapy, and temozolomide response.  Afterwards, the signatures were independently validated in a clinical cohort of glioblastoma patients.

“We found that we could use PDX models of glioblastoma to determine which tumors were likely to respond to radiation, chemotherapy or combined chemotherapy and radiation based on which genes were turned on in the tumors, said study investigator Daniel Wahl, MD, PhD, an assistant professor in the Department of Radiation Oncology at the University of Michigan in Ann Arbor, Michigan. "We then took the six genes that we discovered from the mouse experiments and asked if their levels predicted for whether human patients with glioblastoma would respond to chemotherapy, radiation or combined treatment,” he told Cancer Network.

During their research, Wahl and colleagues performed an independent validation in a heterogeneously treated clinical cohort of 502 glioblastoma patients with overall survival (OS) as the primary endpoint. They used multivariate Cox analysis to adjust for confounding variables and evaluated interactions between signatures and treatment. The researchers found that the PDX models mirrored the clinical heterogeneity of glioblastoma patients. The use of radiotherapy, chemotherapy and temozolomide correlated with benefit from treatment in the PDX models.

Among the 502 patients in the clinical validation cohort, 65% received chemoradiation, 16% received radiotherapy alone, 3% received chemotherapy alone, and 16% received no treatment. The patients who were treated with chemoradiation fared best, followed by those who underwent single modality treatment with radiotherapy or chemotherapy. The patients who received no treatment fared the worst. The study validated that three platform-independent molecular signatures may be able to predict benefit from standard of care therapies for glioblastoma. “These signatures may be useful to personalize glioblastoma treatment in the clinic and this approach may be a generalizable method to identify predictive biomarkers without resource-intensive randomized trials,” the authors wrote.

The study showed that higher radiotherapy scores were associated with increased survival only in patients receiving radiotherapy in the independent clinical validation. The same was true for higher chemotherapy scores and higher chemoradiation scores. According to the authors, the significant interaction between signatures and treatments indicates that they do predict response to therapy rather than simply being prognostic. “Looking at the levels of the six genes we discovered in human patients with glioblastoma could give physicians some idea about how a given patient might respond to standard therapies in glioblastoma,” Dr. Wahl explained to Cancer Network.

He said the second main application of these new findings may be for researchers. If a team is trying to understand which genes dictate the response of a rarer cancer to therapy, one strategy may be to grow a panel of tumors in mice and then treat the mice with a placebo or the treatment of interest and determine which genes correlate to response. “This approach may be especially useful for rarer cancers when data from human patients are limited,” said Dr. Wahl.

To date, predictive biomarkers that could help with the selection of optimal treatment in patients with glioblastoma have been greatly lacking. Dr. Wahl said oncologists currently only have one piece of information that predicts the response of a glioblastoma patient to chemotherapy (MGMT promoter methylation). However, if these new findings are confirmed in additional studies, it may be possible to achieve a more precise estimate of responsiveness to chemotherapy, radiation, or chemoradiation. Richard Bucholz, MD, a professor of neurosurgery at Saint Louis University in St. Louis, Missouri, said this study is one of the first to show the promise of personalized medicine for the management of malignant glioma. These tumors are heterogenous, making it difficult to categorize them by population. Further research may make it possible to do just that, targeting specific treatments based on the genetic profiles of a patient's glioblastoma.

References:

1)  Zhao SG, Yu M, Spratt DE, et al. Xenograft-based platform-independent gene signatures to predict response to alkylating chemotherapy, radiation, and combination therapy for glioblastoma.  Neuro-Oncology, noz090, https://doi.org/10.1093/neuonc/noz090    Published: 23 May 2019