ABSTRACT
Surgical resection is a fixture in the treatment of intracranial tumors, and there is mounting data indicating
that overall and progression-free survival improve for gross total resection compared to subtotal
resection. The current standard of care for managing intracranial tumors relies heavily on MRI of
gadolinium-based contrast agents (Gd-MRI), which plays a central role in diagnosis, surgical planning,
intra-surgical guidance, and follow-up monitoring. During surgery, patients are spatially registered to the
pre-operative MRI and MRI-derived tumor contours projected over the visual field within the surgical
microscope to guide resection. Despite the widespread deployment of these sophisticated tools in
surgery, subtotal resection rates remain stubbornly high. The primary culprits include difficulty in
identifying tumor visually and the diminishing accuracy of the pre-op registration due to brain deformation
as the surgery progresses. In this context, expansive efforts have sought to alleviate these shortcomings,
including the use of intra-operative stereovision and/or ultrasound with brain deformation models to
update the pre-op MRI and the use of fluorescent agents to label tumor in the visual field. Although
promising, both of these approaches have known shortcomings. Specifically, the data sources used for
updating pre-op MRI are only surrogate correlates with MRI, and most current fluorescence guided
surgery (FGS) efforts focus on targeted agents designed to mark molecular features of tumor cells, which
have shown high intra-patient/tumoral heterogeneity. This project aims to solve both of these
shortcomings directly by leveraging the existing clinical understanding of Gd-MRI in managing intracranial
tumors. Specifically, we will identify and evaluate fluorescent agents that mimic the kinetic behavior of
conventional MRI-based contrast agents to guide intracranial tumor surgery. This approach will transfer
the well-understood behavior of Gd-MRI directly into the visual field, enable rapid, intra-surgical
administration of the agent, and provide an ideal data input for updating of pre-op MRI during surgery.
Our approach is premised on compelling preliminary data in small animal glioma models showing highly
correlative uptake between Gd-MRI and several untargeted optical agents. To advance this strategy we
will, (1) rigorously validate these results and examine additional optical agent candidates using MRI and
our custom hyperspectral whole-body imaging cryomacrotome, (2) establish concordance between
candidate optical agents and Gd-MRI in a new porcine glioma model using our intra-operative MRI facility
and FGS instruments, and (3) assess the capacity to use the optical agent data to update the pre-op
MRI. We will also quantitatively compare uptake of the candidate agents, Gd-MRI and ALA-PPIX, the
current standard for FGS of glioma. Completing the aims of this project will establish the optical analog
strategy as a compelling approach for surgical guidance and lay the groundwork for clinical translation.