BEGIN:VCALENDAR
VERSION:2.0
METHOD:PUBLISH
PRODID://Telerik Inc.//Sitefinity CMS 13.2//EN
BEGIN:VTIMEZONE
TZID:UTC
BEGIN:STANDARD
DTSTART;VALUE=DATE:20210101
TZNAME:UTC
TZOFFSETFROM:+0000
TZOFFSETTO:+0000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DESCRIPTION:Date: Wednesday 21st September  Thursday 22nd September 2022\n
Time: \;\n21st = 12:3017:00  \;22nd \;= 09:0017:00 BST\n21s
t \;= 13:3018:00  \;22nd \;= 10:0018:00 CEST\nLocation: Onl
ine\n\nWho is this event intended for? \;PreClinical Statisticians in
the Pharmaceutical Industry\, with interest and/or some basic knowledge o
f R/STAN/BUGS and Bayesian statistics.\nWhat is the benefit of attending?&
nbsp\;This workshop offers the chance to meet with colleagues across indus
try and learn more about Bayesian Methodology and its applications in pre
clinical.\nOverview\n\nThis is the preclinical SIG&rsquo\;s 10th workshop
and will be the first one run virtually. Our theme for this workshop is B
ayesian\; the workshop will run over a day and a half and will include a t
raining course on Bayesian methods (see more below)\, two presentations on
applications of Bayesian methodology in a preclinical setting and a brea
kout session.\n\n\nCourse\nBayesian Statistics for Preclinical Research: N
ew Opportunities\n\nThe course will start by introducing the key concepts
of Bayesian statistics\, emphasizing the context and key objectives of pre
clinical research in pharmaceutical and medical device development. \;
We then move on to show how Bayesian thinking and practices are a fitfor
purpose paradigm. \; \; Over the last decade\, preclinical resear
ch has been identified in the literature as an area of research suffering
from a lack of reproducibility. Causes for this are many\, but in this cou
rse\, we&rsquo\;ll show how to frame a Bayesian strategy to address reprod
ucibility concerns by proposing new study designs\, modelling\, and decisi
onmaking. Preclinical research is a learning process\, making Bayesian st
atistical learning a very natural partnership. \;\n\nKey topics covere
d by the course include:\n\n Define the question and the research objec
tive\n Strategies for determining\, using\, and checking robustness of
prior distributions\n Replacing experimentbased decisions in favor of
projectbased decision\n Use of informed control groups and unbalanced
designs\n Design of the overall project\, integrating the potential sou
rces of irreproducibility in advance\n Progress under uncertainty\, ado
ption of adaptive designs\n Designing experiments using Bayesian assura
nce\, rather than power\n Understand risks and predictive probability o
f success to meet the objective\n Bayesian incorporation of real world
evidence (RWE)\n Examples of Bayesian programming using R/STAN/BUGS and
SAS\n\nBruno Boulanger\, Senior Director\, PharmaLex\nBradley Carlin\, Se
nior Advisor\, PharmaLex\n\n\nTalks\nBayesian Tumor volume analysis with B
RMS R package\nMarie Miossec\nIn cancer drug development\, demonstrated ef
ficacy in tumor xenograft experiments on severe combined immunodeficient m
ice who are grafted with human tumor tissues or cells is an important step
to bring a promising compound to human. A key outcome variable is tumor v
olumes measured over a period of time\, while mice are treated with certai
n treatment regimens. The tumor growth inhibition delta T/delta C ratio is
commonly used to quantify treatment effects in such drug screening tumor
xenograft experiments In this presentation\, we propose a Bayesian approac
h to make a statistical inference of the T/C ratio\, including both hypoth
esis testing and a credibility interval estimate. Through a practical case
\, implementation\, diagnosis\, model selection and results with the BRMS
R package will be discussed.\n\n\nA Bayesian\, Generalized Frailty Model f
or Comet Assays\nHelena Geys\nThis paper proposes a flexible modelling app
roach for socalled comet assay data regularly encountered in preclinical
research. While such data consist of nonGaussian outcomes in a multilev
el hierarchical structure\, traditional analyses typically completely or p
artly ignore this hierarchical nature by summarizing measurements within a
cluster. NonGaussian outcomes are often modelled using exponential famil
y models. This is true not only for binary and count data\, but also for\,
e.g.\, timetoevent outcomes. Two important reasons for extending this f
amily are: (1) the possible occurrence of over dispersion\, meaning that t
he variability in the data may not be adequately described by the models w
hich often exhibit a prescribed meanvariance link\, and (2) the accommoda
tion of a hierarchical structure in the data\, owing to clustering in the
data. The first issue is dealt with through socalled over dispersion mode
ls. Clustering is often accommodated through the inclusion of random subje
ctspecific effects. Though not always\, one conventionally assumes such r
andom effects to be normally distributed. In the case of timetoevent dat
a\, one encounters\, for example\, the gamma frailty model (Duchateau and
Janssen 2007). While both of these issues may occur simultaneously\, model
s combining both are uncommon. Molen berghs et al (2010) proposed a broad
class of generalized linear models accommodating over dispersion and clust
ering through two separate sets of random effects. In Ghebretinsae et al\,
we used this method to model data from a comet assay with a threelevel h
ierarchical structure. Whereas a conjugate gamma random effect is used for
the over dispersion random effect\, both gamma and Normal random effects
are considered for the hierarchical random effect. Apart from model formul
ation\, we place emphasis on Bayesian estimation.\nWorkshop Cost\nThis Wor
kshop is open to both Members and NonMembers of PSI. Please see below for
confirmation of fees.\nPSI Members \;= £\;125+VAT\nPSI NonMembe
rs = £\;125+VAT\nRegistration\nPlease note: this event will take plac
e online via Zoom\, and has a limited number of places available. \;\n
To register for this workshop\, please click here.\nSpeaker details\n\n\n\
n \n \n \n Speaker\n \n
\n Biography\n \n \n \n
\n Bruno Boulanger\n \n \n \n
\n \n Bruno Boulanger has 25 years of ex
perience in several areas of pharmaceutical research and industry includin
g discovery\, toxicology\, CMC and early clinical phases. He holds various
positions in Europe and in USA. Bruno joined UCB Pharma in 2007 as Direct
or of Exploratory Statistics. Bruno is also since 2000 Lecturer at the Uni
versité\; of Liè\;ge\, in the School of Pharmacy\, teaching De
sign of Experiments and Statistics. He is also a USP Expert\, member of th
e Committee of Experts in Statistics since 2010. Bruno has authored or co
authored more than 100 publications in applied statistics and coedited on
e book in Bayesian statistics for pharmaceutical research.\n \n
\n \n \n \n Brad Carlin\n
Brad Carlin is a statistical researcher\, methodologist\, consu
ltant\, and instructor. \; He currently serves as Senior Advisor for D
ata Science and Statistics at PharmaLex\, an international pharmaceutical
consulting firm. \; Prior to this\, he spent 27 years on the faculty o
f the Division of Biostatistics at the University of Minnesota School of P
ublic Health\, serving as division head for 7 of those years. \; He ha
s also held visiting positions at Carnegie Mellon University\, Medical Res
earch Council Biostatistics Unit\, Cambridge University (UK)\, Medtronic C
orporation\, HealthPartners Research Foundation\, the M.D Anderson Cancer
Center\, and AbbVie Pharmaceuticals. \; \; He has published more t
han 185 papers in refereed books and journals\, and has coauthored three
popular textbooks: &ldquo\;Bayesian Methods for Data Analysis&rdquo\; with
Tom Louis\, &ldquo\;Hierarchical Modeling and Analysis for Spatial Data&r
dquo\; with Sudipto Banerjee and Alan Gelfand\, and "Bayesian Adaptive Met
hods for Clinical Trials" with Scott Berry\, J. Jack Lee\, and Peter Mulle
r. \; From 20062009 he served as editorinchief of \;Bayesian An
alysis\, the official journal of the International Society for Bayesian An
alysis (ISBA). \; During his academic career\, he served as primary di
ssertation adviser for 20 PhD students. \; Dr. Carlin has extensive ex
perience teaching short courses and tutorials\, and won both teaching and
mentoring awards from the University of Minnesota. During his spare time\,
Brad is a health musician and bandleader\, providing keyboards\, guitar\,
and vocals in a variety of venues.\n \n \n \n
\n Marie Miossec\n \n \n
Marie is a biostatistician engineer at IT&\;M STATS. She was gradua
ted from ENSAI (National School of Statistics and Information Analysis\, F
rance) in 2019 with a master's degree specializing in statistics for life
sciences. She has been working for SANOFI as a contractor for three years
in the team in charge of biostatistical support to nonclinical efficacy &
amp\; safety studies.\n \n \n \n \n
\n Helena Geys\n \n \n
Helena Geys is Global head of the Discovery and Nonclinical Safety Stati
stics group at Johnson and Johnson. Helena joined J&\;J 18 years ago du
ring which period she has made significant contributions in various areas
of nonclinical statistics: discovery\, toxicology\, manufacturing. She is
an active participant in many professional organizations\, and has shown h
erself a contributor to many successful external and crosspharma initiati
ves and academic collaborations leading to impactful successes in drug dev
elopment strategies. The results of her research have been published in &g
t\;100 methodological and applied publications on clustered nonnormal dat
a\, risk assessment\, spatial epidemiology\, translational medicine and su
rrogate marker validation. In addition to her assignment at Janssen\, Hele
na has a strong passion for teaching and mentoring. She combines her work
at Janssen Pharmaceutica with a position as tenuretrack professor in bios
tatistics at the Data Science Institute of Hasselt University (Belgium) an
d has mentored and coached >\;30 master and PhD students.\n \
n \n \n
DTEND:20220922T160000Z
DTSTAMP:20221004T155217Z
DTSTART:20220921T113000Z
LOCATION:
SEQUENCE:0
SUMMARY:PSI PreClinical SIG Workshop 2022
UID:RFCALITEM638004955372446593
XALTDESC;FMTTYPE=text/html:Date<
/strong>: Wednesday 21st September  Thursday 22nd September 2022
\n<
strong>Time: \;
\n21st = 12:3017:00 &
nbsp\;22nd \;= 09:0017:00 BST
\n21st \;= 13:3018:
00  \;22nd \;= 10:0018:00 CEST
\nLocation: Online
\n
\nWho is this event intended for?&
nbsp\;PreClinical Statisticians in the Pharmaceutical Industry\, with int
erest and/or some basic knowledge of R/STAN/BUGS and Bayesian statistics.<
br />\nWhat is the benefit of attending? \;This works
hop offers the chance to meet with colleagues across industry and learn mo
re about Bayesian Methodology and its applications in preclinical.
\n<
h4 style="textalign: justify\;">Overview\n\nThis is the preclinical SIG&rsquo\;s 10th workshop and will be t
he first one run virtually. Our theme for this workshop is Bayesia
n\; the workshop will run over a day and a half and will include
a training course on Bayesian methods (see more below)\, two presentations
on applications of Bayesian methodology in a preclinical setting and a b
reakout session.
\n
\n
\nCo
urse
\nBayesian Statistics fo
r Preclinical Research: New Opportunities
\n
\nThe cour
se will start by introducing the key concepts of Bayesian statistics\, emp
hasizing the context and key objectives of preclinical research in pharmac
eutical and medical device development. \; We then move on to show how
Bayesian thinking and practices are a fitforpurpose paradigm. \;&nb
sp\; Over the last decade\, preclinical research has been identified in th
e literature as an area of research suffering from a lack of reproducibili
ty. Causes for this are many\, but in this course\, we&rsquo\;ll show how
to frame a Bayesian strategy to address reproducibility concerns by propos
ing new study designs\, modelling\, and decisionmaking. Preclinical resea
rch is a learning process\, making Bayesian statistical learning a very na
tural partnership. \;
\n\nKey topics covered by the course include:
\n\n
 Define the question and the research
objective
\n  Strategies for dete
rmining\, using\, and checking robustness of prior distributions
\n
 Replacing experimentbased decisions in
favor of projectbased decision
\n  Use of informed control groups and unbalanced designs
\n  Design of the overall project\, integrating the
potential sources of irreproducibility in advance
\n  Progress under uncertainty\, adoption of adaptive des
igns
\n  Designing experiments usi
ng Bayesian assurance\, rather than power
\n  Understand risks and predictive probability of success to meet
the objective
\n  Bayesian incorp
oration of real world evidence (RWE)
\n  Examples of Bayesian programming using R/STAN/BUGS and SAS
\n\nBruno Boulanger\, Senior Director\, PharmaLex
\nBradley Carlin\, Senior Advisor\, Pharma
Lex
\n
\n
\nTalks
\nBayesian Tumor volume analysis with BRMS R package
\nMarie Miossec
\nIn cancer drug development\, demonstra
ted efficacy in tumor xenograft experiments on severe combined immunodefic
ient mice who are grafted with human tumor tissues or cells is an importan
t step to bring a promising compound to human. A key outcome variable is t
umor volumes measured over a period of time\, while mice are treated with
certain treatment regimens. The tumor growth inhibition delta T/delta C ra
tio is commonly used to quantify treatment effects in such drug screening
tumor xenograft experiments In this presentation\, we propose a Bayesian a
pproach to make a statistical inference of the T/C ratio\, including both
hypothesis testing and a credibility interval estimate. Through a practica
l case\, implementation\, diagnosis\, model selection and results with the
BRMS R package will be discussed.
\n
\n
\nA Bayesian\, Generalized Frailty Model for Comet As
says
\nHelena Geys
\nThis paper proposes a fle
xible modelling approach for socalled comet assay data regularly encounte
red in preclinical research. While such data consist of nonGaussian outc
omes in a multilevel hierarchical structure\, traditional analyses typica
lly completely or partly ignore this hierarchical nature by summarizing me
asurements within a cluster. NonGaussian outcomes are often modelled usin
g exponential family models. This is true not only for binary and count da
ta\, but also for\, e.g.\, timetoevent outcomes. Two important reasons f
or extending this family are: (1) the possible occurrence of over dispersi
on\, meaning that the variability in the data may not be adequately descri
bed by the models which often exhibit a prescribed meanvariance link\, an
d (2) the accommodation of a hierarchical structure in the data\, owing to
clustering in the data. The first issue is dealt with through socalled o
ver dispersion models. Clustering is often accommodated through the inclus
ion of random subjectspecific effects. Though not always\, one convention
ally assumes such random effects to be normally distributed. In the case o
f timetoevent data\, one encounters\, for example\, the gamma frailty mo
del (Duchateau and Janssen 2007). While both of these issues may occur sim
ultaneously\, models combining both are uncommon. Molen berghs et al (2010
) proposed a broad class of generalized linear models accommodating over d
ispersion and clustering through two separate sets of random effects. In G
hebretinsae et al\, we used this method to model data from a comet assay w
ith a threelevel hierarchical structure. Whereas a conjugate gamma random
effect is used for the over dispersion random effect\, both gamma and Nor
mal random effects are considered for the hierarchical random effect. Apar
t from model formulation\, we place emphasis on Bayesian estimation.
\n
Workshop Cost
\nThis Workshop is
open to both Members and NonMembers of PSI. Please see below for confirma
tion of fees.
\nPSI Members \;= £\;125+VAT<
br />\nPSI NonMembers = £\;125+VAT
\nRegistr
ation
\nPlease note: this event will
take place online via Zoom\, and has a limited number of places available.
\;
\nTo register for this workshop\, please click here.
\nSpeaker details
\n\n\n\n \n \n
\n \n \n \n \n \n
\n \n \n
\n \n \n \n \n
\n \n \n \n \n \n\n Spea
ker \n  \n Biography \n

\n Bruno Boulanger \n \n \n \n
\n  Bruno Boulanger has 25 years of experience in several areas of pharm
aceutical research and industry including discovery\, toxicology\, CMC and
early clinical phases. He holds various positions in Europe and in USA. B
runo joined UCB Pharma in 2007 as Director of Exploratory Statistics. Brun
o is also since 2000 Lecturer at the Université\; of Liè\;ge\,
in the School of Pharmacy\, teaching Design of Experiments and Statistics
. He is also a USP Expert\, member of the Committee of Experts in Statisti
cs since 2010. Bruno has authored or coauthored more than 100 publication
s in applied statistics and coedited one book in Bayesian statistics for
pharmaceutical research. \n \n 
\n Brad Carlin  Brad Carlin is a statistical r
esearcher\, methodologist\, consultant\, and instructor. \; He current
ly serves as Senior Advisor for Data Science and Statistics at PharmaLex\,
an international pharmaceutical consulting firm. \; Prior to this\, h
e spent 27 years on the faculty of the Division of Biostatistics at the Un
iversity of Minnesota School of Public Health\, serving as division head f
or 7 of those years. \; He has also held visiting positions at Carnegi
e Mellon University\, Medical Research Council Biostatistics Unit\, Cambri
dge University (UK)\, Medtronic Corporation\, HealthPartners Research Foun
dation\, the M.D Anderson Cancer Center\, and AbbVie Pharmaceuticals.
\; \; He has published more than 185 papers in refereed books and jour
nals\, and has coauthored three popular textbooks: &ldquo\;Bayesian Metho
ds for Data Analysis&rdquo\; with Tom Louis\, &ldquo\;Hierarchical Modelin
g and Analysis for Spatial Data&rdquo\; with Sudipto Banerjee and Alan Gel
fand\, and "Bayesian Adaptive Methods for Clinical Trials" with Scott Berr
y\, J. Jack Lee\, and Peter Muller. \; From 20062009 he served as edi
torinchief of \;Bayesian Analysis\, the official journal of
the International Society for Bayesian Analysis (ISBA). \; During his
academic career\, he served as primary dissertation adviser for 20 PhD st
udents. \; Dr. Carlin has extensive experience teaching short courses
and tutorials\, and won both teaching and mentoring awards from the Univer
sity of Minnesota. During his spare time\, Brad is a health musician and b
andleader\, providing keyboards\, guitar\, and vocals in a variety of venu
es. 
\n \n Ma
rie Miossec \n  \n Marie is a biostatistician engineer at IT&a
mp\;M STATS. She was graduated from ENSAI (National School of Statistics a
nd Information Analysis\, France) in 2019 with a master's degree specializ
ing in statistics for life sciences. She has been working for SANOFI as a
contractor for three years in the team in charge of biostatistical support
to nonclinical efficacy &\; safety studies. \n 
\n \n Helena Gey
s \n  \n Helena Geys is Global head of the Discovery and Noncl
inical Safety Statistics group at Johnson and Johnson. Helena joined J&
\;J 18 years ago during which period she has made significant contribution
s in various areas of nonclinical statistics: discovery\, toxicology\, man
ufacturing. She is an active participant in many professional organization
s\, and has shown herself a contributor to many successful external and cr
osspharma initiatives and academic collaborations leading to impactful su
ccesses in drug development strategies. The results of her research have b
een published in >\;100 methodological and applied publications on clust
ered nonnormal data\, risk assessment\, spatial epidemiology\, translatio
nal medicine and surrogate marker validation. In addition to her assignmen
t at Janssen\, Helena has a strong passion for teaching and mentoring. She
combines her work at Janssen Pharmaceutica with a position as tenuretrac
k professor in biostatistics at the Data Science Institute of Hasselt Univ
ersity (Belgium) and has mentored and coached >\;30 master and PhD stude
nts. \n 
END:VEVENT
END:VCALENDAR