DOAJ (DOAJ: Directory of Open Access Journals),
Journal Year:
2022,
Volume and Issue:
25(10), P. 730 - 734
Published: Oct. 20, 2022
Bayesian
statistics
is
an
approach
for
learning
from
evidences
as
it
accumulates,
combining
prior
distribution
with
current
information
on
a
quantity
of
interest,
in
which
posterior
and
inferences
are
being
updated
each
time
new
data
become
available
using
Bayes'
Theorem.
Though
frequentist
has
dominated
medical
studies,
been
more
widely
recognized
by
its
flexibility
efficiency.
Research
development
(R&D)
anti-cancer
drugs
have
so
hot
globally
recent
years
spite
relatively
high
failure
rate.
It
the
common
demand
pharmaceutical
enterprises
researchers
to
identify
optimal
dose,
regime
right
population
early-phase
R&D
stage
accurately
efficiently,
especially
when
following
three
major
changes
observed.
The
anticancer
transformed
chemical
biological
products,
monotherapy
combination
therapy,
study
design
also
gradually
changed
traditional
way
innovative
adaptive
mode.
This
raises
number
subsequent
challenges
decision-making
early
R&D,
such
inability
determine
MTD,
deal
delayed
toxicity,
response
dose-response
changing
relationships.
because
above
emerging
that
getting
attention
industry.
At
least,
decision-making,
could
potentially
help
achieve
higher
efficiency,
shorter
period
lower
investment.
expounds
application
drugs,
compares
analyzes
idea
scenarios
statistics,
aiming
provide
macroscopic
systematic
reference
all
related
stakeholders.
.【中文题目:贝叶斯方法在肿瘤新药早期临床研发中
的发展与应用】
【中文摘要:贝叶斯学派是通过综合未知参数的先验信息与样本信息,依据贝叶斯定理,求出后验分布,根据后验分布推断未知参数的统计方法。相比频率派,贝叶斯学派更加灵活、高效。肿瘤新药是全球研发的热点,但同时也存在高失败率的风险。在肿瘤新药早期研发中,高效寻找最佳剂量、优势人群、估计疗效和成功率是医药企业和研究者的共同需求。近年来,肿瘤新药研发呈现化学药物生物制品转变、单药治疗向联合治疗转变、传统设计向创新设计转变等新趋势;伴随出现的各种挑战,包括无法找到最高耐受剂量、延迟毒性、延迟反应、剂量疗效关系变化、剂量组合众多等。基于贝叶斯方法,恰当借用先验信息,能有效帮助企业在肿瘤早期研发中,实现从传统研发模式(高投入、长周期、低效率)向现代研发模式(低投入、短周期、高效率)的转变。研究还进行了贝叶斯方法在肿瘤新药早期研发的进展阐述,与频率派的理念、应用场景的比较分析,可为医药研发的所有从业人员提供宏观、系统的参考。
】
【中文关键词:早期试验;贝叶斯;
统计设计;肿瘤】.
BMJ,
Journal Year:
2023,
Volume and Issue:
unknown, P. e076387 - e076387
Published: Oct. 20, 2023
The
CONSORT
(CONsolidated
Standards
Of
Reporting
Trials)
2010
statement
is
the
standard
guideline
for
reporting
completed
randomised
trials.
Dose-finding
Extension
(DEFINE)
extends
guidance
(with
21
new
items
and
19
modified
items)
to
early
phase
dose-finding
trials
with
interim
dose
escalation
or
de-escalation
strategies.
Such
generally
focus
on
safety,
tolerability,
activity,
recommending
dosing
scheduling
regimens
further
clinical
development.
These
are
often
inadequately
reported,
hampering
their
informativeness
making
evidence
informed
decisions
difficult.
CONSORT-DEFINE
aims
develop
an
international,
consensus
driven
promote
transparency,
completeness,
reproducibility,
facilitate
interpretation
of
results.
provides
recommendations
essential
that
should
be
reported
in
greater
clarity,
informativeness,
usefulness
BMJ,
Journal Year:
2023,
Volume and Issue:
unknown, P. e076386 - e076386
Published: Oct. 20, 2023
SPIRIT
(Standard
Protocol
Items:
Recommendations
for
Interventional
Trials)
2013
provides
guidance
clinical
trial
protocol
writing.
However,
neither
the
original
nor
its
extensions
adequately
cover
features
of
early
phase
dose-finding
trials.
The
Dose-finding
Extension
(DEFINE)
statement
is
a
new
guideline
that
recommendations
essential
items
should
be
provided
in
protocols
these
It
details
to
guidance,
incorporating
17
and
modifying
15
existing
items.
purpose
this
promote
transparency,
completeness,
reproducibility
methods,
interpretation
protocols.
envisioned
resulting
improvements
design
conduct
trials
will
ultimately
reduce
research
inefficiencies
inconsistencies,
driving
transformational
advances
care.
EClinicalMedicine,
Journal Year:
2023,
Volume and Issue:
60, P. 102020 - 102020
Published: May 25, 2023
BackgroundThe
paradigm
of
early
phase
dose-finding
trials
has
evolved
in
recent
years.
Innovative
designs
and
protocols
which
combine
phases
I
II
are
becoming
more
popular
health
research.
However,
the
quality
these
trial
is
unknown
due
to
a
lack
specific
reporting
guidelines.
Here,
we
evaluated
protocols.MethodsWe
conducted
cross-sectional
study
oncology
non-oncology
posted
on
ClinicalTrials.gov
2017–2023.
A
checklist
items
comprising:
1)
original
33-items
from
SPIRIT
2013
Statement
2)
additional
unique
were
used
assess
quality.
The
primary
endpoint
was
overall
proportion
adequately
reported
items.
This
registered
with
PROSPERO
(no:
CRD42022314572).FindingA
total
106
included
rule-based
3
+
being
most
design
(39.6%).
Eleven
model-based
model-assisted
identified
only
(11/58,
19.0%).
65.1%
(95%CI:
63.9–66.3%).
each
individual
item
varied
substantially
(range
9.4%–100%).
Oncology
showed
lower
than
non-oncology.
In
multivariable
analysis,
larger
sample
sizes
industry
funding
associated
higher
proportions
(all
p-values
<0.05).InterpretationThe
suboptimal
(65.1%).
There
need
for
improved
completeness
transparency
facilitate
rigorous
conduct,
reproducibility
external
review.FundingNone.
BMC Medicine,
Journal Year:
2023,
Volume and Issue:
21(1)
Published: July 5, 2023
Abstract
Background
Early
phase
dose-finding
(EPDF)
trials
are
crucial
for
the
development
of
a
new
intervention
and
influence
whether
it
should
be
investigated
in
further
trials.
Guidance
exists
clinical
trial
protocols
completed
reports
SPIRIT
CONSORT
guidelines,
respectively.
However,
both
guidelines
their
extensions
do
not
adequately
address
characteristics
EPDF
Building
on
checklists,
DEFINE
study
aims
to
develop
international
consensus-driven
(SPIRIT-DEFINE)
(CONSORT-DEFINE).
Methods
The
initial
generation
candidate
items
was
informed
by
reviewing
published
reports.
early
draft
were
refined
through
review
grey
literature,
analysis
real-world
examples,
citation
reference
searches,
expert
recommendations,
followed
two-round
modified
Delphi
process.
Patient
public
involvement
engagement
(PPIE)
pursued
concurrently
with
quantitative
thematic
participants’
feedback.
Results
survey
included
79
or
SPIRIT-DEFINE
(
n
=
36)
CONSORT-DEFINE
43)
extension
items.
In
Round
One,
206
interdisciplinary
stakeholders
from
24
countries
voted
151
Two.
Following
One
feedback,
one
item
added
Of
80
items,
60
met
threshold
inclusion
(≥
70%
respondents
critical:
26
SPIRIT-DEFINE,
34
CONSORT-DEFINE),
remaining
20
discussed
at
consensus
meeting.
parallel
PPIE
work
resulted
an
lay
summary
toolkit
consisting
template
guidance
notes
exemplar.
Conclusions
By
detailing
journey
decisions
undertaken,
we
envision
that
this
will
enhance
understanding
help
researchers
future
guidelines.
allow
investigators
effectively
essential
present
reports,
thereby
promoting
transparency,
comprehensiveness,
reproducibility.
Trial
registration
registered
EQUATOR
Network
https://www.equator-network.org/
).
CPT Pharmacometrics & Systems Pharmacology,
Journal Year:
2023,
Volume and Issue:
13(1), P. 41 - 53
Published: Oct. 16, 2023
Recently,
the
use
of
machine-learning
(ML)
models
for
pharmacokinetic
(PK)
modeling
has
grown
significantly.
Although
most
current
approaches
ML
techniques
as
black
boxes,
there
are
only
a
few
that
have
proposed
interpretable
architectures
which
integrate
mechanistic
knowledge.
In
this
work,
we
test
case
one-compartment
PK
model
using
scientific
machine
learning
(SciML)
framework
and
consider
an
unknown
absorption
neural
networks,
while
simultaneously
estimating
other
parameters
drug
distribution
elimination.
We
generate
simulated
data
with
different
sampling
strategies
to
show
our
can
accurately
predict
concentrations
in
extrapolation
tasks,
including
new
dosing
regimens
sparsity
levels,
produce
reliable
forecasts
even
patients.
By
scenario
fitting
complex
absorption,
demonstrate
known
physiological
structure
into
SciML
allows
us
obtain
highly
accurate
predictions
preserving
interpretability
classical
compartmental
models.
European Journal of Cancer,
Journal Year:
2022,
Volume and Issue:
173, P. 167 - 177
Published: July 21, 2022
Phase
1
immuno-oncology
(IO)
trials
frequently
involve
pharmacodynamic
(PD)
biomarker
assessments
involving
tumour
biopsies
and/or
blood
collection,
with
increasing
use
of
molecular
imaging.
PD
biomarkers
are
set
to
play
a
fundamental
role
in
early
drug
development
agents.
In
the
IO
era,
impact
for
confirmation
biologic
activity
and
their
subsequent
have
not
been
investigated.Phase
studies
published
between
January
2014
December
2020
were
reviewed.
Studies
that
reported
on-treatment
[tissue-derived
(tissue-PD),
blood-based
(blood-PD)
imaging-based
(imaging-PD)]
analysed.
results
correlation
clinical
endpoints
evaluated.
Authors'
statements
on
influence
further
decisions,
citations
study
recorded.Among
386
trials,
most
frequent
agent
classes
evaluated
vaccines
(32%)
PD-(L)1
inhibitors
(25%).
No
100
(26%).
Of
remaining
286,
blood-PD,
tissue-PD,
imaging-PD
data
270
(94%),
94
(33%),
12
(4%)
respectively.
Assessments
more
than
one
type
82
(29%).
Similar
proportions
blood-PD
(9%),
tissue-PD
(7%),
(8%)
had
positive
correlated
activity.
Results
22
referenced
trials.Most
phase
performed
assessments.
Overall,
infrequently
or
cited
suggesting
limited
development.
With
emerging
health
regulatory
emphasis
optimal
dose
selection
based
activity,
informative
integrative
multiplexed
assays
capture
complexity
tumour-host
immunity
interactions
warranted
improve
trial
methodology.
Frontiers in Pharmacology,
Journal Year:
2023,
Volume and Issue:
14
Published: Nov. 23, 2023
Due
to
the
small
sample
sizes
in
early-phase
clinical
trials,
toxicity
and
efficacy
profiles
of
dose-schedule
regimens
determined
for
subsequent
trials
may
not
be
well
established.
The
recent
development
novel
anti-tumor
treatments
combination
therapies
further
complicates
problem.
Therefore,
there
is
an
increasing
recognition
essential
place
optimizing
regimens,
new
strategies
are
now
urgently
needed.
Bayesian
adaptive
designs
provide
a
potentially
effective
way
evaluate
several
doses
schedules
simultaneously
single
trial
with
higher
efficiency,
but
real-world
implementation
examples
such
still
few.
In
this
paper,
we
cover
critical
factors
associated
optimization
review
related
innovative
designs.
assumptions,
characteristics,
limitations,
application
scenarios
those
introduced.
also
summarizes
some
unresolved
issues
future
research
opportunities
optimization.