Funded Grants

Mechanisms of Sensitivity to PDGFR Inhibitors in Human Gliomas

As a physician-scientist who studies brain cancer and cares for patients, there is an urgent need to improve our standard of care for this incurable disease. Glioblastomas are the most common primary brain tumors and recent reports have shown that glioblastoma is a heterogeneous group of diseases that can be subclassified by shared genetic aberrations. The implication is that, in part, the underlying genetics may determine the responsiveness to treatments and thus allow us to personalize therapy. Several years ago, we noted a glioblastoma patient with dramatic responsiveness to imatinib mesylate, a platelet derived growth factor receptor (PDGFR) kinase inhibitor. We investigated this case and found a molecular biomarker. We subsequently investigated clinical trial patients on a negative phase I/II trial of imatinib in recurrent malignant gliomas and found this same biomarker in some of the responders. Since then we have been evaluating these cases for mutations and other genetic abnormalities. These results illustrate the potential responsiveness of a molecularly defined subset of glioblastomas and the poor results of prior trials might in fact be due to not enriching for responsive patients. Linking such individual treatment responsiveness to the detailed genetics of tumor will be the crucial next step to truly personalizing cancer therapies. The above data has provided the rationale to initiate a prospective trial of PDGFR inhibitors in biomarker-enriched high-grade glioma patients.

With the JSMF grant, we hope to validate that PDGFR aberrations predict response to PDGFR kinase inhibitor and identify in further detail mechanisms of response and resistance. This will allow us to then plan future studies of combination trials to overcome resistance mechanisms and improve efficacy of targeted therapies.

In Aim 1, we will identify and characterize genomic aberrations of PDGFR and related genes in human glioma tissues from PDGFR kinase inhibitor based clinical trials. We will use fluorescence in situ hybridization (FISH) for qualitative assessment (high focal amplification vs. polysomy vs. non-amplified) and QPCR for quantitative assessment of copy number. Using FISH we analyzed our dataset and found a significant association (trend test; p < 0.001). We are currently obtaining tissue from several clinical trials from European consortium (EORTC Brain Tumor Group) and other sites to independently validate our findings. We have also developed a QPCR based assay to detect copy number of relevant genes (e.g. MET, EGFR, etc.) that has improved our sensitivity and screening efficiency from paraffin tissues. The samples will also be analyzed for mutations using the WAVE HS system (Transgenomic, Inc) followed by sequencing to confirm mutations. We will also use genome wide approaches (Agilent array CGH, Affymetrix gene expression profiling, high throughput sequencing, etc) where feasible to characterize the tissue samples further. We will also analyze tissues from recently initiated biomarker-enriched prospective clinical trials of PDGFR inhibitors in glioblastomas.

In Aim 2, we will correlate genetics of PDRFR and other related genes to patient response (imaging response and progression free survival) to PDGFR kinase inhibitors in retrospective and prospective clinical trials. Patients’ imaging response was defined as complete response (CR), partial response (PR), stable disease (SD) or progressive disease (PD) by MacDonald criteria and overall and progression free survival by standard neuro-oncology criteria. We will correlate the genetic findings in Aim 1 with response to PDGFR kinase inhibitors from several retrospective studies and prospective studies to validate our hypothesis that aberrations in PDGFR a predict response to PDGFR kinase inhibitors. These will be analyzed individually and pooled to increase power. Kaplan-Meier product-limit estimates and logrank tests of imaging response, overall and progression-free survival curves stratified by PDGFR and other relevant genes, Cox proportional hazards models of times to death and progression testing for association with molecular status with adjustment for age, KPS, and other relevant baseline predictors.

In Aim 3, we will functionally validate the genomic aberrations found in Aims 1 and 2 using human and mouse glioma stem cell models. We will validate the molecular basis for responsiveness to PDGFR kinase inhibitors in these systems and as means to discover mechanisms of resistance. We have been collecting and growing a library of fresh human glioma stem cell lines (“living tissue bank”) from our patients grown as neurospheres and propagated in mice. These lines are highly annotated and monitored over time for maintenance of key glioblastoma characteristics using genome-wide technologies. Several of these lines have PDGFR aberrations and have constitutively activated PDGFRA. We also have normal and tumorigenic murine stem cells (Ink4a/ARF-/- with expression of either EGFRvIII, PDGFRA, PIK3CA, BRAF, etc) using retroviral expression constructs. We will use both the human and mouse lines to correlate genetics with response to PDGFR kinase inhibitors and identify molecular mechanisms of resistance. Such data will generate new hypotheses regarding the mechanisms of sensitivity and resistance to PDGFR kinase inhibitors that can then be validated prospectively in ongoing and future human trials. This will also allow us to then plan future studies of combination trials to overcome resistance mechanisms or improve efficacy of PDGFR kinase directed therapies.

This proposal seeks to fulfill the intent of the JSMF Foundation on many levels. The work proposed is novel and unsupported by any outside funding mechanism at this point. The proposal involves validating a genetic biomarker of response to PDGFR kinase inhibitors in retrospective and prospective clinical trials of glioblastoma and elucidating the mechanisms of sensitivity and resistance. Linking such individual patient tumor genotypes to treatment responsiveness will be the crucial next step to truly personalizing brain tumor therapies and is the focus of my future translational research efforts. This proposal represents a highly collaborative effort between clinicians and scientists in neuro-oncology, neurosurgery, cancer biology, medical oncology, neuroradiology and pathology.