The Critical Path Institute: transforming competitors into collaborators

The Critical Path Institute: transforming
competitors into collaborators
Martha Brumfield
The Critical Path Institute brings scientists from regulatory agencies, industry and academia
together to improve drug development and regulatory processes.
Founded in 2005, the Critical Path Institute (C-Path)
originates from the US Food and Drug Administration
(FDA)’s Critical Path Initiative, which a decade ago identified the need to connect scientific advancement with
regulatory policy through public–private partnerships
(PPPs)1. Today, the institute serves as a catalyst in the
development of new approaches to advance medical
innovation and regulatory science by leading teams that
share data, knowledge and expertise to produce sound,
consensus-based science.
Martha Brumfield is
President and CEO of the
Critical Path Institute,
1730 East River Road,
Tucson, Arizona 85718, USA.
[email protected]
Expanding the precompetitive space
The rapid pace of knowledge generation, technical
breakthroughs and medical advances is now such that it
is impossible that any one entity has all of the information or expertise to find complete answers to key drug
development challenges in isolation. Sharing knowledge and data through targeted venues is the future of
developing new methods to prevent, diagnose and treat
disease, especially for more complex areas such as neuro­
logical and psychiatric disorders.
Understanding the precompetitive space and building
common ground among different stakeholders is essential to a successful collaborative project. Less than a decade ago, a precompetitive project would only address
in vitro or preclinical data. But the boundaries are
expanding with the growing realization of the substantial benefit that can result from new discoveries (such as
biomarkers, modelling tools and clinical outcome assessments (COAs)). Even clinical data, traditionally considered proprietary, are contributed to pooled databases to
help solve prespecified research questions.
The tools and methods created through sharing information precompetitively are changing the landscape of
clinical trial design, enabling trials to be more optimally
designed and executed. One such example is a simulation tool developed through a C-Path consortium — the
Coalition Against Major Diseases (CAMD) — for clinical
trials in mild to moderate Alzheimer’s disease2. The simulator minimizes the uncertainty regarding duration, dose
and disease stage of patients to enrol in trials. The tool
was created by using data from intervention trials of
Alzheimer’s treatments, findings from the longitudinal
Alzheimer’s Disease Neuroimaging Initiative (ADNI)
study and academic publications. Initial interest in the
tool, which received favourable regulatory designation
from the FDA as well as the European Medicines Agency
(EMA) in 2013, has been high.
Another example of how collaboration can advance
regulatory science by sharing risks and costs is the development of patient-reported outcomes (PRO) instruments
through multistakeholder consortia. Working on PRO
instruments in seven therapeutic areas, C-Path’s PRO
Consortium develops instruments that consortium
members can test in their trials and then provide the
data to the consortium for regulatory qualification. The
development cost of such an instrument runs into the
millions of US dollars, but through collaboration the cost
for each participant is lowered, and the risk is reduced
owing to the active engagement of the FDA along the
path to qualification.
Secondary use of data collected in clinical studies is
increasingly recognized as a valuable source of information that can yield unanticipated insights when curated
and appropriately aggregated. Aggregation of diverse data
sources is enabled by the use of global standards such as
those developed through the Clinical Data Interchange
Standards Consortium (CDISC). Universally accepted
for use in clinical research and development, CDISC data
standards also promise to streamline regulatory review.
To further the opportunities from data aggregation, the
institute and TransCelerate BioPharma have partnered
with the CDISC to more expeditiously develop new
disease area standards (see Further information).
Information sharing in the precompetitive space is
leading the way towards more knowledge and insight
about biomarkers, disease progression trajectories, COA
tools and clinical trial process improvements. For example,
the institute has achieved regulatory qualification of
a number of biomarkers and other tools, and has an
extensive pipeline of research programmes in its seven
consortia (see Supplementary information S1 (table)).
VOLUME 13 | NOVEMBER 2014 | 785
© 2014 Macmillan Publishers Limited. All rights reserved
Addressing the challenges
The challenges in building collaborative precompetitive
research programmes arise from different concerns and
perspectives of stakeholders. A strong collaborative programme addresses key issues at the outset: how to involve
the right individuals or groups; how to reach consensus
on the science; how to ensure adequate protection for
intellectual property; how to enable aggregation of data
by applying appropriate protective mechanisms; how
to fund each research project; and how to prioritize
multiple research objectives. By starting there, the independent collaborative model can factor in and resolve
technical issues, provide a venue for dialogue and reduce
the inherent risk in drug development.
Although publications are the ‘currency’ of the acade­
mic world, data — especially patient-level, intervention
trial data — are essential for qualifying new drug development tools (DDTs). Publications typically only present
summary tables of data; consequently, the conclusions
cannot be directly verified through re-analysis. Moreover,
a single publication usually summarizes a single clinical
trial, which would probably not satisfy the data demands
for tools intended to be broadly applied across multiple
development programmes. Individual patient-level data
from multiple trials is the currency for DDT qualification,
which requires not only sharing data but also standardizing that data in order to pool individual patient data from
multiple trials. Pooling data avoids the inevitable complications of meta-analysis, and the rigour enforced through
standardization adds to the integrity of the process.
The evaluation of the performance of any DDT
depends on aggregated data, but sharing data collected
during clinical trials runs counter to the established system. Difficulties arise when: the study’s informed consent
document precludes secondary use of data; data were not
collected in a standardized manner; and stakeholders
have concerns about proprietary information. Solutions
to some of these challenges can be addressed within
a consortium. Legacy data can be remapped to a common standard, which may necessitate creating data and
measurement standards to facilitate data aggregation and
analysis. Data contributors can also specify who is allowed
access to the data. Concerns about proprietary information may be set aside when access to large-scale, integrated
data sources is agreed upon and clearly defined upfront
with input from all data contributors. The one issue that
cannot be addressed retrospectively is informed consent;
however, most companies now use consent documents
that allow more flexibility in the secondary use of data.
The inclusion of patient groups within the consortium
is a key success factor as they represent the position of
multiple patients, many of whom wish for their individual
data to be used for the greater good of science and society.
Ensuring collaboration among multiple PPPs and
alliances can prevent or at least minimize duplication of
efforts. C-Path’s Predictive Safety Testing Consortium
(PSTC) directly engages with the Innovative Medicines
Initiative’s Safer and Faster Evidence-Based Translation
(SAFE-T) consortium through shared work plans, as both
are evaluating biomarkers to assess organ-specific toxicity.
Another strategy for coordinating consortia projects is to
enlist participants of related consortia, as has been done by
the CAMD, where members coordinate with the ADNI3.
Funding is another challenge faced by consortia, partly
because regulatory science is still in its infancy. Also,
government funding traditionally has not favoured PPPs.
Some forward-thinking private foundations have recognized that it is not enough to fund bench science or even
clinical trials without addressing the critical requirement
to validate the measurement science that enables utility in
a clinical setting, especially in a development programme
intended for regulatory assessment. For example, the
National Multiple Sclerosis Society is providing research
support to develop a new COA for use as a primary or
secondary end point in clinical trials of patients with
relapsing–remitting or progressive forms of multiple sclerosis4, the Polycystic Kidney Disease Foundation is funding the development of an imaging biomarker, and the Bill
and Melinda Gates Foundation funds the development of
tools for tuberculosis therapies through the Critical Path
to TB Drug Regimens Initiative. The recently formed
Accelerating Medicines Partnership, a PPP established to
validate disease targets, indicates that the US government
could adopt the European strategy and support regulatory
science consortia through equivalent contributions from
government and industry.
The next generation of partnerships
To move into the era of precision medicine, where treatment is tailored both to the genotype and phenotype of
each individual patient as well as to the specific character­
istics of the disease, each piece of collected data needs to
be utilized effectively. Building on the lessons learned, new
collaborative efforts are tapping into the burgeoning clinical data in a fashion that protects patient confidentiality,
recognizes stakeholder contributions and maximizes
the ability of researchers to address the complexities of
disease5. As third-party conveners increase collaboration between competitors so that it becomes the norm
instead of the exception, the ability to develop innovative
treatments more efficiently and effectively will improve.
Woodcock, J. & Woosley, R. The FDA Critical Path Initiative and its
influence on new drug development. Annu. Rev. Med. 59,1–12 (2008).
Rogers, J. A. et al. Combining patient-level and summary-level data for
Alzheimer’s disease modeling and simulation: a beta regression metaanalysis. J. Pharmacokinet. Pharmacodyn. 39, 479–498 (2012).
Yu, P. et al. Operationalizing hippocampal volume as an enrichment
biomarker for amnestic mild cognitive impairment trials: effect of
algorithm, test-retest variability, and cut point on trial cost,
duration, and sample size. Neurobiol. Aging 35, 808–818 (2014).
Rudick, R. A. et al. Multiple Sclerosis Outcome Assessments
Consortium: genesis and initial project plan. Mult. Scler. 20,
12–17 (2014).
Walker, E. G. et al. Evolving global regulatory science through
the voluntary submission of data: a 2013 assessment.
Ther. Innovation Regul. Sci. 48, 236–245 (2014).
M.B. would like to thank L. Hudson for her contributions to this article.
Competing interests statement
The author declares no competing interests.
Coalition For Accelerating Standards and Therapies (CFAST):
See online article: S1 (table)
786 | NOVEMBER 2014 | VOLUME 13
© 2014 Macmillan Publishers Limited. All rights reserved