NIH grant: Quantitative imaging and computational modeling to predict gene therapy coverage in glioma following resection (1R21CA307416-01A1)

retained text

NIH grant 1R21CA307416-01A1: Quantitative imaging and computational modeling to predict gene therapy coverage in glioma following resection
Institute: {'code': 'CA', 'abbreviation': 'NCI', 'name': 'National Cancer Institute', 'admin_org_id': '12500008', 'admin_funding_url': ''}. PI: Stephen L Brown. Award: $416401.
Start: 2026-09-18T00:00:00.
Glioblastoma (GBM) is a highly aggressive brain tumor with a median survival of only about
15 months despite the current standard therapy. When a glioma tumor recurs, no national
standard treatment exists. We have developed an oncolytic adenovirus (AdV) gene therapy
that enhances radiation sensitivity in tumor cells. The three-pronged therapy combines
virus, prodrugs, and radiotherapy (RT) to achieve cytotoxic eGects. We hypothesize that
inadequate distribution of AdV limits therapeutic eGicacy in locally delivered treatments.
Computational modeling can characterize AdV biodistribution and predict regions of
insuGicient coverage. Such predictions could guide dose optimization and improve tumor
response in a patient-specific manner. Our long-term goal is to personalize GBM therapy
via computational modeling supported by dynamic contrast-enhanced MRI (DCE-MRI).
Using biphasic tissue frameworks to estimate interstitial flow post-treatment, these datadriven
models incorporate tissue swelling and flow diversion around resection cavities.
Currently, the role of local flow and tissue response on the biodistribution of AdV and its
therapeutic byproducts remains underexplored. We will measure the distribution of AdV
and viral-activated factor in the presence and absence of RT. Initial validation will occur in
GBM patient-derived orthotopic xenograft (PDOX) rodent models. Animal models capture
the most relevant DCE-MRI and endpoints necessitate euthanasia to collect
tissue. Alternatives to the use of animal models is not possible; computational modeling is
used but needs input of physiological responses which are not currently known and
measures solely in humans are not possible. DCE-MRI-derived computational models from
them will be adapted for human applications. We propose two aims. The first in preclinical
models is to validate the predicted distribution at resection margins of AdV and
its byproduct using DCE-MRI, in-situ hybridization, and autoradiography. This establishes a
framework for precise treatment planning and translational scaling to humans. The second
aim is to use computational predictions in patients to optimize delivery of radiation
therapy.

document fetched 2026-09-27 02:45