NcXHF

GENOMICS, PERSONALIZED MEDICINE, AND SUPPORTIVE CANCER CARE
FIGURE 1. Differentially Expressed Genes Defined by Signaling Functions
Major roles for TNF-alpha have been suggested in pathoetiology of radiation-induced toxicities, including mucositis and fatigue. The formation of a network based on differentially
expressed genes in patients who developed mucositis confirms a central role for TNF, illustrates its effect on connected nodes, and provides a potential target for pharmacologicallybased
intervention. 40
such as Ingenuity Pathways Analysis (Ingenuity Systems).
Network analyses identify the biologic functions and canonical
pathways that are most signifıcant to the genes in
the network (Fig. 1). 40
This methodology has been used to provide an abundance
of mechanistic data to defıne regimen-related tissue injury.
Canonical pathways associated with NF-kappa B signaling,
Toll-like receptor signaling, P13K/AKT signaling, interleukin-6
signaling, and p38 MAPK signaling are among examples that
are consistent with proposed mechanisms by which toxicities,
such as mucosal injury, occur. From the viewpoint of
drug development, the key roles played by each pathway suggest
that each is a potential therapeutic target.
An analysis of canonical pathways may also provide hints
to the global symptomatic response of patients to cancer
therapy. For example, although overexpression of genes
within the glutamate signaling pathway has been associated
with central nervous system activity, recent analyses suggest
broadened activity, including a role as a signal mediator. 23
This conclusion has implications in drug development and
might favor and inform formulation and/or route of administration
strategies.
The fact that genetics can play a key role in determining a patient’s
response to a drug creates opportunities for both expediting
drug development and individualizing care. By identifying a
genetic signature that differentiates the likelihood of response or
nonresponse to a drug, developers would have the ability to enrich
clinical trials with groups of patients who were most likely
to benefıt, and, at the same time, spare genetically-defıned nonresponders
from the risks and inconvenience of receiving an experimental
agent. Once approved, compounds that effectively
mitigated regimen-related toxicities could be selectively administered
to the most appropriate patients, favorably affecting the
physiological and fınancial costs of treating nonresponders.
ECONOMIC, POLICY IMPLICATIONS, AND
CHALLENGES FOR THE APPLICATION OF GENOMICS
TO SUPPORTIVE CANCER CARE
For a clinical genomics test to achieve acceptability by patients,
clinicians, payers, and regulators, it has to successfully answer
two questions: Does it work and is it worth the cost? 41 How to
answer the fırst question has been a topic addressed in a variety
of forums. In 2004, the Centers for Disease Control and Prevention
launched the Evaluation of Genomic Applications in Practice
and Prevention (EGAPP) initiative. The group defıned four
evaluation criteria: analytic validity (how well does the test measure
what it is supposed to measure); clinical validity (how well
does the test predict its specifıed outcome); clinical utility (how
well does the test improve or harm outcomes to patients); and
ethical, legal, and social issues. 42
In general, the effıcacy of genomic tests as toxicity-risk predictors
using the validity criteria cited above requires data drawn
from large populations of patients. As noted earlier in this article,
results of studies using either candidate gene or GWAS approaches
have been notoriously diffıcult to replicate, and, in the
case of GWAS, prone to false-positive signals. Fortunately new
big data methods and the integration of multifaceted analytics
provide a platform from which genomic parameters with toxicity
risk can be determined. Based on past experiences, such studies
likely will be most effective if they are prospective and if they
provide a strict template for the capture of accurate toxicity data.
asco.org/edbook | 2015 ASCO EDUCATIONAL BOOK 13