BMC Medical Research Methodology,
Год журнала:
2023,
Номер
23(1)
Опубликована: Ноя. 4, 2023
Subject-level
real-world
data
(RWD)
collected
during
daily
healthcare
practices
are
increasingly
used
in
medical
research
to
assess
questions
that
cannot
be
addressed
the
context
of
a
randomized
controlled
trial
(RCT).
A
novel
application
RWD
arises
from
need
create
external
control
arms
(ECAs)
for
single-arm
RCTs.
In
analysis
ECAs
against
RCT
data,
there
is
an
evident
manage
and
analyze
same
technical
environment.
Nordic
countries,
legal
requirements
may
require
original
subject-level
anonymized,
i.e.,
modified
so
risk
identify
any
individual
minimal.
The
aim
this
study
was
conduct
initial
exploration
on
how
well
pseudonymized
anonymized
perform
creation
ECA
RCT.This
hybrid
observational
cohort
using
clinical
arm
completed
phase
II
(PACIFIC-AF)
Finnish
sources.
were
within
(k,
ε)-anonymity
framework
(a
model
protecting
individuals
identification).
Propensity
score
matching
weighting
methods
applied
RWD,
balance
potential
confounders
data.
Descriptive
statistics
overall
survival
analyses
conducted
prior
after
weighting,
both
sets.Anonymization
affected
baseline
characteristics
only
marginally.
greatest
difference
prevalence
chronic
obstructive
pulmonary
disease
(4.6%
vs.
5.4%
compared
respectively).
Moreover,
changed
anonymization
by
8%
(95%
CI
4-22%).
Both
able
produce
matched
Anonymization
impacted
22%
-21-87%).Anonymization
viable
technique
cases
where
flexible
transfer
sharing
required.
As
necessarily
affects
some
aspects
further
careful
consideration
strategies
needed.
Clinical Pharmacology & Therapeutics,
Год журнала:
2023,
Номер
114(2), С. 325 - 355
Опубликована: Апрель 20, 2023
Real‐world
data
(RWD)‐derived
external
controls
can
be
used
to
contextualize
efficacy
findings
for
investigational
therapies
evaluated
in
uncontrolled
trials.
As
the
number
of
submissions
regulatory
and
health
technology
assessment
(HTA)
bodies
using
rises,
light
recent
HTA
guidance
on
appropriate
use
RWD,
there
is
a
need
address
operational
methodological
challenges
impeding
quality
real‐world
evidence
(RWE)
generation
consistency
evaluation
RWE
across
agencies.
This
systematic
review
summarizes
publicly
available
information
outcomes
from
trials
all
indications
January
1,
2015,
through
August
20,
2021,
that
were
submitted
European
Medicines
Agency,
US
Food
Drug
Administration,
and/or
select
major
(National
Institute
Health
Care
Excellence
(NICE),
Haute
Autorité
de
Santé
(HAS),
Institut
für
Qualität
und
Wirtschaftlichkeit
im
Gesundheitswesen
(IQWiG),
Gemeinsamer
Bundesausschuss
(G‐BA)).
By
systematically
reviewing
context
guidance,
this
study
provides
quantitative
qualitative
insights
into
how
control
design
analytic
choices
may
viewed
by
different
agencies
practice.
The
primary
aspects
identified
discussion
include,
but
are
not
limited
to,
engagement
regulators
bodies,
approaches
handling
missing
(a
component
quality),
selection
endpoints.
Continued
collaboration
these
other
will
inform
assist
stakeholders
attempting
generate
controls.
ESMO Real World Data and Digital Oncology,
Год журнала:
2023,
Номер
1, С. 100003 - 100003
Опубликована: Ноя. 1, 2023
•Real-world
evidence
in
oncology
is
evolving
rapidly
with
many
particularities.•This
guidance
provides
key
recommendations
for
reporting
real-world
studies
oncology.•Recommendations
are
based
on
a
review
of
current
and
the
authors'
collective
expert
opinion.•Authors
multidisciplinary
group
experts
from
different
institutions
countries.•Guidance
provided
full
article
development,
including
title,
introduction,
methods,
results,
discussion,
conclusion.
The International Journal of Neuropsychopharmacology,
Год журнала:
2024,
Номер
27(2)
Опубликована: Фев. 1, 2024
Abstract
Background
The
3
paliperidone
palmitate
(PP)
long-acting
injectable
antipsychotic
formulations,
PP
1-month
(PP1M),
3-month
(PP3M),
and
6-month
(PP6M),
have
shown
to
reduce
the
risk
of
relapse
in
schizophrenia.
current
phase-4
study
constructed
external
comparator
arms
(ECAs)
using
real-world
data
for
PP3M
PP1M
compared
prevention
rates
with
PP6M
from
an
open-label
extension
(OLE)
adult
patients
Methods
were
derived
a
single-arm,
24-month,
OLE
(NCT04072575),
which
included
schizophrenia
who
completed
12-month
randomized,
double-blind,
noninferiority,
phase-3
(NCT03345342)
without
relapse.
Patients
ECAs
identified
IBM®
MarketScan®
Multistate
Medicaid
Database
based
on
similar
eligibility
criteria
as
cohort.
Results
A
total
178
each
cohort
following
propensity
score
matching.
Most
men
(>70%;
mean
age:
39–41
years).
Time
(primary
analysis
Kaplan-Meier
estimates)
was
significantly
delayed
(P
<
.001,
log-rank
test).
rate
lower
(3.9%)
vs
(20.2%)
(29.8%)
cohorts.
Risk
decreased
.001)
by
82%
(HR
=
0.18
[95%
CI
0.08
0.40]),
89%
0.11
[0.05
0.25]),
35%
0.65
[0.42
0.99];
P
.043).
Sensitivity
confirmed
findings
primary
analysis.
Although
matched
mimic
characteristics
cohort,
heterogeneity
between
groups
could
exist
due
factors
including
prior
participation,
unmeasured
confounders,
variations
capture
quality,
completeness
clinical
information.
Conclusions
In
trial
setting,
time
demonstrated
treatments
settings
among
Trial
registration
ClinicalTrials.gov
Identifier:
NCT04072575;
EudraCT
number:
2018-004532-30
Blood Advances,
Год журнала:
2022,
Номер
7(19), С. 5680 - 5690
Опубликована: Дек. 19, 2022
For
the
past
decade,
it
has
become
commonplace
to
provide
rapid
answers
and
early
patient
access
innovative
treatments
in
absence
of
randomized
clinical
trials
(RCT),
with
benefits
estimated
from
single-arm
trials.
This
trend
is
important
oncology,
notably
when
assessing
new
targeted
therapies.
Some
those
uncontrolled
further
include
an
external/synthetic
control
group
as
way
indirect
comparison
a
pertinent
group.
We
aimed
some
guidelines
comprehensive
tool
for
(1)
critical
appraisal
comparisons
or
(2)
performing
trial.
used
example
ciltacabtagene
autoleucel
treatment
adult
patients
relapsed
refractory
multiple
myeloma
after
3
more
lines
illustrative
example.
propose
3-step
guidance.
The
first
step
includes
definition
estimand,
which
encompasses
effect
population
(whole
restricted
trial
external
controls),
reflecting
question.
second
relies
on
adequate
selection
controls
previous
RCTs
real-world
data
cohorts,
registries,
electronic
files.
third
consists
choosing
statistical
approach
targeting
defined
above
depends
available
(individual-level
aggregated
data).
validity
derived
heavily
careful
methodological
considerations
included
proposed
procedure.
Because
level
evidence
well-conducted
RCT
cannot
be
guaranteed,
evaluation
than
standard
settings.
Therapeutic Innovation & Regulatory Science,
Год журнала:
2024,
Номер
58(3), С. 443 - 455
Опубликована: Март 25, 2024
Abstract
Conducting
clinical
trials
(CTs)
has
become
increasingly
costly
and
complex
in
terms
of
designing
operationalizing.
These
challenges
exist
running
CTs
on
novel
therapies,
particularly
oncology
rare
diseases,
where
target
narrower
patient
groups.
In
this
study,
we
describe
external
control
arms
(ECA)
other
relevant
tools,
such
as
virtualization
decentralized
(DCTs),
the
ability
to
follow
trial
subjects
real
world
using
tokenization.
ECAs
are
typically
constructed
by
identifying
appropriate
sources
data,
then
cleaning
standardizing
it
create
an
analysis-ready
data
file,
finally,
matching
with
CT
interest.
addition,
ECA
tools
also
include
subject-level
meta-analysis
simulated
subjects’
for
analyses.
By
implementing
recent
advances
digital
health
technologies
devices,
virtualization,
DCTs,
realigning
from
site-centric
designs
virtual,
decentralized,
patient-centric
can
be
done,
which
reduces
burden
participate
encourages
diversity.
Tokenization
technology
allows
linking
real-world
(RWD),
creating
more
comprehensive
longitudinal
outcome
measures.
provide
robust
ways
enrich
informed
decision-making,
reduce
costs
operations,
augment
insights
gained
data.
JNCI Journal of the National Cancer Institute,
Год журнала:
2022,
Номер
115(1), С. 14 - 20
Опубликована: Сен. 26, 2022
Abstract
As
precision
medicine
becomes
more
precise,
the
sizes
of
molecularly
targeted
subpopulations
become
increasingly
smaller.
This
can
make
it
challenging
to
conduct
randomized
clinical
trials
therapies
in
a
timely
manner.
To
help
with
this
problem
small
patient
subpopulation,
study
design
that
is
frequently
proposed
trial
(RCT)
intent
augmenting
RCT
control
arm
data
historical
from
set
patients
who
have
received
treatment
outside
(historical
data).
In
particular,
strategies
been
developed
compare
outcomes
across
cohorts
treated
standard
(control)
guide
use
analysis;
lessen
potential
well-known
biases
using
controls
without
any
randomization.
Using
some
simple
examples
and
completed
studies,
we
demonstrate
commentary
these
are
unlikely
be
useful
applications.
JCO Precision Oncology,
Год журнала:
2024,
Номер
8
Опубликована: Янв. 8, 2024
Advances
in
genomics
have
enabled
anticancer
therapies
to
be
tailored
target
specific
genomic
alterations.
Single-arm
trials
(SATs),
including
those
incorporated
within
umbrella,
basket,
and
platform
trials,
are
widely
adopted
when
it
is
not
feasible
conduct
randomized
controlled
rare
biomarker-defined
subpopulations.
External
controls
(ECs),
defined
as
control
arm
data
derived
outside
the
clinical
trial,
gained
renewed
interest
a
strategy
supplement
evidence
generated
from
SATs
allow
comparative
analysis.
There
increasing
examples
demonstrating
application
of
EC
precision
oncology
trials.
The
prospective
conducting
studies
associated
with
distinct
methodological
challenges,
considerations
for
use
subpopulations
been
adequately
discussed,
formal
framework
yet
established.
In
this
review,
we
present
analysis
using
EC.
Key
steps
(1)
defining
purpose
address
study
question,
(2)
determining
if
external
fit
purpose,
(3)
developing
transparent
protocol
statistical
plan,
(iv)
interpreting
results
drawing
conclusions
on
basis
prespecified
hypothesis.
We
specify
required
subpopulations,
which
include
specifying
comparator
biomarker
status
group,
lines
treatment,
assessment
testing
panels
used,
(4)
cohort
stratification
tumor-agnostic
studies.
further
discuss
novel
trial
designs
techniques
leveraging
propose
future
directions
advance
generation
facilitate
drug
development
oncology.
EClinicalMedicine,
Год журнала:
2024,
Номер
70, С. 102526 - 102526
Опубликована: Март 11, 2024
BackgroundDespite
more
than
50
years
of
research
and
parallel
improvements
in
hepatology
oncology,
there
is
still
today
neither
a
treatment
to
prevent
disease
progression
primary
sclerosing
cholangitis
(PSC),
nor
reliable
early
diagnostic
tools
for
the
associated
hepatobiliary
cancers.
Importantly,
limited
understanding
underlying
biological
mechanisms
PSC
its
natural
history
not
only
affects
identification
new
drug
targets
but
implies
lack
surrogate
markers
that
hampers
design
clinical
trials
evaluation
efficacy.
The
easy
access
large
representative
well-characterised
prospective
resources
an
important
contributing
factor
current
situation.MethodsWe
here
present
SUPRIM
cohort,
national
multicentre
longitudinal
study
unselected
patients
capturing
diversity
phenotypes.
We
describe
10-year
effort
inclusion
follow-up,
intermediate
analysis
report
including
original
results,
resource.
All
included
gave
written
informed
consent
(recruitment:
November
2011–April
2016).FindingsOut
512
patients,
452
completed
five-year
follow-up
without
endpoint
outcomes.
Liver
transplantation
was
performed
54
(10%)
malignancy
diagnosed
15
(3%).
draw
comprehensive
landscape
multidimensional
heterogeneity
illustrating
Performances
available
predictive
scores
are
compared
perspectives
on
continuation
cohort
provided.InterpretationWe
envision
as
open-access
collaborative
resource
accelerate
generation
knowledge
independent
validations
promising
ones
with
aim
uncover
diagnostics,
prognostic
tools,
markers,
by
2040.FundingThis
work
supported
Swedish
Cancer
Society,
Stockholm
County
Council,
Research
Funds
Radiumhemmet.
JMIR Medical Informatics,
Год журнала:
2024,
Номер
12, С. e55118 - e55118
Опубликована: Май 8, 2024
Synthetic
patient
data
(SPD)
generation
for
survival
analysis
in
oncology
trials
holds
significant
potential
accelerating
clinical
development.
Various
machine
learning
methods,
including
classification
and
regression
trees
(CART),
random
forest
(RF),
Bayesian
network
(BN),
conditional
tabular
generative
adversarial
(CTGAN),
have
been
used
this
purpose,
but
their
performance
reflecting
actual
remains
under
investigation.