284
genetic landscape approach may be applied to other aggressive types of cancer, such
as ovarian and lung cancer. Eventually, networks may be created that account for
both genetic and epigenetic changes in cancer cells.
Molecular Diagnostics for Personalized Management of Brain Cancer
Several molecular biomarkers have been identifi ed in diffuse gliomas that carry
diagnostic and prognostic information. In addition, some of these and other bio-
markers predict the response of these gliomas to particular chemotherapeutic
approaches. Molecular diagnostics is an important contribution to personalized
management of glioma patients.
Diffusion MRI as a Biomarker The response to treatment of brain cancer is usu-
ally assessed by measurements obtained from brain imaging several months after
the start of treatment. A biomarker of tumor response would be useful for making
early treatment decisions and for determining prognosis. To obtain this information,
patients with GBM are examined by diffusion MRI before treatment and 3 weeks
after treatment; the images are coregistered, and differences in tumor-water diffu-
sion values are calculated as functional diffusion maps (fDM), which are correlated
with the radiographic response, time-to-progression, and overall survival. Changes
in fDM at 3 weeks are closely associated with the radiographic response at 10
weeks. The percentage of the tumor undergoing a signifi cant change in the diffusion
of water is different in patients with progressive disease as compared to those with
stable disease. fDM provides an early biomarker for response, time-to-progression,
and overall survival in patients with GBM. This method has the potential to evaluate
differences in effi cacy between patients, as well as to assess the heterogeneity of
response within an individual tumor. This technique needs to be further evaluated to
determine its usefulness in the individualization of treatment or evaluation of the
response to treatment in clinical trials.
Combined Neuroimaging and DNA Microarray Analysis This method has
been used to create a multidimensional map of gene-expression patterns in GBM
that provides clinically relevant insights into tumor biology (Diehn 2008 ). Tumor
contrast enhancement and mass effect can predict activation of specifi c hypoxia
and proliferation gene-expression programs, respectively. Overexpression of
EGFR, a receptor tyrosine kinase and potential therapeutic target, has also been
directly inferred by neuroimaging and validated in an independent set of tumors by
immunohistochemistry. Furthermore, imaging provides insights into the intratu-
moral distribution of gene-expression patterns within GBM. An “infi ltrative”
imaging phenotype can identify and predict patient outcome. Patients with this
imaging phenotype have a greater tendency toward having multiple tumor foci and
demonstrate signifi cantly shorter survival than their counterparts. These fi ndings
provide an in vivo portrait of genome-wide gene expression in GBM and offer a
potential strategy for noninvasively selecting patients who may be candidates for
individualized therapies.
10 Personalized Therapy of Cancer