Biology Reference
In-Depth Information
precedents set by earlier efforts. On the other hand, translating biomarker data from early
clinical development studies into validated companion diagnostic tests is much less well-
explored territory. In this section we will offer insights and opinions gained from our experi-
ences but caution that this is very much work in progress with considerable uncertainty as to
how the regulatory framework for companion diagnostics may evolve. We will provide refer-
ences to existing regulatory guidance documents but acknowledge that there are presently
few published data on the implementation of biomarker selection and companion diagnostic
assay validation in clinical studies of candidate asthma therapeutics.
A primary challenge of companion diagnostic development is having enough data and
confidence to invest the resources to develop these assays on an appropriate platform and
with sufficient quality control in place well before the pivotal Phase III drug trials are initi-
ated. This early investment in the assays for use in Phase III is because these trials will be
used to register not only the drug, but also the specific way in which the diagnostic test is
to be used in conjunction with the drug. Given the complexity of regulatory requirements
for both drugs and diagnostics, companion diagnostic development dictates that the work
begins to be at risk very early in the development process and that as many opportunities as
possible are taken to generate data about how the biomarker performs. As such, the discus-
sion here will focus on diagnostic biomarker development during early clinical development,
up to the start of Phase III pivotal trials, by which time the diagnostic biomarker should be
established if it is to be an essential component of registration.
Figure 4.4 depicts a schematic early development timeline for a candidate asthma drug and a
companion diagnostic. Assuming suitable pilot data on the relationships between the biomarker,
the therapeutic target, and the disease state are available, assay development with a partner
diagnostic company may take several years. Thus, to allow sufficient time to develop a robust
clinical biomarker assay prior to the start of pivotal Phase III studies, exploratory biomarker
discovery must take place very early, preferably prior to the start of Phase I clinical studies.
A technically validated assay capable of reliably reporting biomarker values in relevant clinical
samples should be in place prior to unblinding Phase II proof-of-concept clinical studies to ena-
ble a prospective stratified analysis. The outcome of the proof-of-concept trial determines the
strategy for diagnostic implementation in pivotal trials. Three potential outcomes are depicted:
1. The diagnostic enriches for clinical benefit in Phase II but the Phase II study is
underpowered to identify a precise cutoff based on the outcome measures. Phase III
should enroll all comers with a pre-specified analysis of patients stratified according to
baseline biomarker levels. In this case, a provisional cutoff value may be pre-specified
with the understanding that the greater statistical power afforded by larger numbers
of patients enrolled in the Phase III study could enable post-hoc optimization of the
biomarker cutoff. The regulatory pathway for this option remains to be defined, and any
post-hoc adjustments would need to be based on relatively unambiguous data.
2. In the rare event that a precise diagnostic cutoff is identified in Phase II and there
is no clear clinical benefit to 'diagnostic-negative' patients, Phase III could proceed
with enrollment of only 'diagnostic-positive' patients. As discussed in this chapter,
given the variability of approvable outcome measures and the continuous, rather than
dichotomous, nature of asthma subtypes, we do not believe this scenario is likely for
asthma therapies currently under investigation.
Search WWH ::




Custom Search