Biology Reference
In-Depth Information
mutations that are lurking within a tumor at low frequency at the initiation of treatment but
become the predominant clone over time. Translating such genomic discoveries about het-
erogeneity within tumors into tailored therapeutics will be the challenge of the future.
1.2.1.4 Case of Pegylated Interferon (IFN) in Infectious Disease
Pegylated IFN in combination with ribavirin is approved for the treatment of chronic
infection by the hepatitis C virus. Ge et al. performed a whole genome association study in
which they compared the frequencies of almost 600,000 single nucleotide polymorphisms
(SNPs) in patients infected with genotype 1 hepatitis C virus between those who had a sus-
tained virologic response to therapy and those that failed to respond
[36]
. A C
/
T SNP in the
vicinity of the interleukin 28B gene, which encodes the type III interferon IFN-l3, was highly
significantly associated with response in whites, blacks and Hispanics. Compared with ~20-
40% of patients with a T/T
/
T/T genotype who had a sustained virologic response, 60-80% of those
with the C
/
C genotype had sustained virologic response. In heterozygotes sustained viro-
logic response was intermediate (Hispanics) or similar to that observed in T/T
/
T homozygotes
(whites and blacks). Subsequent studies also found that this SNP was significantly associated
with sustained virologic response
[37-39]
. Because there were no placebo-treated patients in
the study, it could not be determined whether the association between IL28b and sustained
virologic response was truly a treatment X genotype interaction effect (as in the KRAS exam-
ple described above) or whether genotype was a predictor of virologic response even in the
absence of treatment. Natural history studies addressed this issue
[39,40]
. These authors eval-
uated patients who spontaneously cleared their hepatitis C virus infection, and found that
the SNPs associated with response to treatment with interferon-ribavirin were also associated
with spontaneous clearance of hepatitis C infection in the absence of treatment, indicating
that SNPs in IL28b contribute to variation in virologic response regardless of therapy.
It should be noted that in both cases the analyses that led to the discovery of predictive
markers were performed outside of the drug development setting. Because of the large num-
ber of tests performed in such analyses (e.g., the genome-wide association tests evaluated
hundreds of thousands of SNPs) there is increased likelihood of false-positive associations
and initial results need to be replicated in multiple independent studies, as was described
for both the KRAS and IL28b examples. Such independent datasets are typically unavailable
during the clinical development of a novel therapeutic, making it difficult to interpret such
retrospective analyses. In the absence of replication, one runs the risk of investing significant
resources in following up potentially spurious associations.
1.3 USE OF A GENOMIC PHARMACODYNAMICS (PD) MARKER TO
AID
IN DOSE SELECTION IN CLINICAL DEVELOPM
ENT
The major objectives for first time in human (FTIH) studies for large molecules are usu-
ally to evaluate the safety, tolerability, pharmacokinetics and immunogenicity profiles. Due
to the long half-life of monoclonal antibodies (mAbs), a single-dose, cohort dose escalation
study can take a year or longer to complete. To expedite clinical development and improve
the PoS, a relevant, sensitive, and robust PD biomarker that can be easily monitored in tri-
als is of great value for finding the right dose. The PD biomarker can be used to account for
Search WWH ::
Custom Search