Vascular Signature Mapping of Brain Tumor Genotypes

A glioma is a primary brain tumor in adults that is characterized by a highly
variable, but overall poor survival. The optimal timing of treatment is in part
determined by the expected biological behavior of the tumor. At present the
expected biological behavior, determined by the tumor genotype, can only be
determined by tissue analysis, which requires brain surgery. Non-invasive and
improved diagnostic methods are sought to obtain insight into the molecular
profile of the tumor and the expected biological behavior to avoid surgery
performed solely for diagnostic purposes. Vascularization is an important
aspect of the biological behavior of a primary brain tumor. Tumor
vascularization characteristics can be assessed by Magnetic Resonance Imaging
(MRI), but with the currently available technology this can only be achieved
with unacceptably long scan times. In this proposal, we will develop and
optimize a novel MRI protocol to gather a large set of quantitative
vascularization parameters within an acceptable scan time. The hypothesis is
that from such a *vascular signature* the tumor genotype can be inferred by
means of machine learning.

The primary objective is to develop and clinically validate a fast
multi-parametric MRI acquisition technique, for non-invasive and comprehensive
characterization of the tumor*s vascularization, *vascular signature mapping*,
at 3 Tesla (3T) and 7 Tesla (7T) MRI. The secondary objective is to limit
difficult and time-consuming visual interpretation of the acquired vascular
information by developing a computer-aided diagnostic algorithm that
automatically and accurately predicts the brain tumor genotype from the
vascular signature maps.

The study *Vascular Signature Mapping for Brain Tumor Genotypes* is a
multi-center observational diagnostic study, which consists of two parts. The
first part of this study aims to develop and optimize a new MRI protocol that
will exploit the effect of contrast agent on the MRI signal to infer
information on the vascular properties of a tumor. It combines scans during the
pre-contrast injection phase, the dynamic phase during and right after contrast
agent injection, as well as the quasi static post-contrast phase. This research
will focus on studying the optimal way of encoding the vascular architecture
into the MRI signal and the decoding approach. In addition, the image
processing methodology will be optimized. The second part of this study is a
proof-of-concept clinical study. This part aims to link the vascular parameters
with molecular profiles of tumors by using the collected data for the
development of machine learning algorithms for predicting the tumor*s genotype
based on its vascular signature.

For the first cohort, the additional burden will not be substantial for the
participant. The additional scan time will not exceed 10 minutes and therefore
the impact on the patient will be limited. For the second and third cohort, the
additional burden includes a prolonged MRI examination at a clinical MRI
scanner (3T) and an optional additional examination at 7T MRI including
additional CA injection. The ultrahigh field 7T MRI system is commonly used for
research and no serious adverse events have been reported1. Patients
participating in this study will have no personal benefit; their participation
aids in the development of a novel MRI method for the non-invasive
determination of the tumor*s molecular profile. Moreover, there is a small
chance that the additional 7T MRI scan would provide more information on the
status of the disease in the participant.