Webinar: Knowledge-based approaches to decreasing clinical attrition rates
May 2018
Speakers: Dr. Richard K. HarrisonGavin ConeyTeresa Fishburne
2
• Clarivate Analytics is the global leader in providing trusted insights and analytics to accelerate the pace of innovation
• Delivering intelligence for Life Science professionals in discovery, preclinical, clinical, commercialization and generics
• Powering life science data and trusted content with analytics
• Clarivate Professional Services offer unrivalled data science expertise, evidence-based consulting and independent advice across the pharmaceutical R&D value chain
Xtalks Partner
for this Event
3
• Over 30 years of experience in the life sciences industry
• Career has focused on all aspects of pre-clinical drug discovery
• Has held positions of increasing responsibility at Aventis, Merck, DuPont and Wyeth Pharmaceuticals
Dr. Richard K. HarrisonChief Scientific Officer
• Supports decision making by professionals within Life Science organizations who are interested in gaining intelligence relating to: • Clinical Development• Clinical Operations • Competitive Intelligence
• Has worked within informatics for 18 years and within the Life Sciences for the last 8 years
Gavin Coney Head of Clinical Products
• Over 20 years of industry experience in Clinical Operations, Regulatory Affairs and Quality management
• Manages a team of clinical and regulatory professionals that create high-value solutions to address customers’ever-evolving needs for strategic intelligence.
Teresa FishburneHead of Clinical & Regulatory
Professional Services
Clarivate Analytics speakers
Industry Overview
Dr. Richard K Harrison
5
What is the current rate of clinical trial attrition?
5% 7%
17%
65%
91%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Animal model tomarket
Phase I tomarket
Phase II tomarket
Phase III tomarket
Submission tomarket
Pro
bab
ility
of
succ
ess
to
mar
ket
• According to the Centers for Medical Research (CMR) the probability of success moving from Phase 1 to market is less than 10% over all therapy areas
Ref: CMR Factbook, 2016
6
What are the main causes of attrition?
• Efficacy is the reason for failure in more than half of Phase II and Phase III trials
• Failure is across all therapy areas with Oncology being the greatest
52%
3%
24%
15%
6%
Reason for failure 2013-2015
Efficacy Operational Safety Strategy Commercial
32%
17%
13%
5%
7%
7%
6%
13%
Percentage failure by therapeutic area
Oncology Central nervous system
Musculoskeletal Infectious disease
Cardiovascular Alimentary
Metabolic Other
Nature Reviews Drug Discovery 15, 817–818 (2016)
7
What are the main causes of attrition?
• According to a study by Astra Zeneca
40% of their projects failed in clinical
trials because no clear link was made
between the target and the disease
• An additional 29% failed
because the compound
chosen did not have the
correct physical properties or
did not reach the target tissue
Nature Reviews Drug Discovery 13419-431 (2014)
8
Clinical failure rates and reasons for failure
• According to the Centers for Medical Research (CMR) in
the year 2015 the probability of success moving from
Phase 1 to market is less than 10% over all therapy
areas
• Efficacy is the reason for failure in more than half of
Phase II and Phase III trials
• According to a study by Astra Zeneca 40% of their
projects failed in clinical trials because no clear link was
made between the target and the disease
• An additional 29% failed because the
compound chosen did not have the correct
physical properties or did not reach the target
tissue
Nature Reviews Drug Discovery 15, 817–818 (2016)
9
Knowledge based approaches to decrease attrition
• Insights into trial endpoint success
• Increased use of biomarkers to decrease attrition
• Correlation to trial success
• Optimum number of biomarkers
• The impact of protocol amendments on trial success
• Impact budget planning,
• Impact on patient enrollment
• Impact on cycle times.
• Incorporating strategies to reduce the number of amendments
• When are amendments required
• Making amendments more effective.
Source: Clarivate Analytics, Cortellis Clinical Trials Intelligence
Insights Into Trial Endpoint Success
Gavin Coney
11
Trial protocol design: rising data / intelligence challenge
Granularity of disease understanding: more datapoints and greater specificity
Increasing data collection (Datapoints collected per patient visit)
Rising volumes of published results Real World Evidence
Increased data volumes from multiple cohorts
Digital Health
0
20
40
60
80
Phase 1 Phase 2 Phase 3
2001 - 2005 2011 - 2015
Cap Gemini, Healthcare survey
Medidata: Analysis of hosted trial data
0
5000
10000
15000
20000
25000
1995 2000 2005 2010 2015 2020
12
Dataset: Granular trial design components indexed for every trial
• Condition
• Phase
• Recruitment Status
• Interventions
• Combination?
• Action
• Class
• Intervention Category
• Trial Design
• Active Control
• Start Date
• Primary Endpoint Completion Date
• Enrolment End Date
• End Date
• Patient Segment
• Biomarker
• Biomarker Type
• Biomarker Role
• Age
• Race
• Inclusion Criteria
• Exclusion Criteria
• Endpoints
• Endpoint Types
• Endpoints Met?
• Results
• Adverse Events
• Sponsor / Collaborator
• Commercially Relevant?
• Site Name
• Contact Name
• City
• State
• Country
• Region
• Enrolment Count
• Enrolment Rate
Trial / Intervention Trial milestones Trial Protocol Insights Patient Enrollment
Source: Cortellis Clinical Trials Intelligence, Clarivate Analytics, April 2018
• Over 300,000 Trials• Focus on drugs, biomarkers and devices• All therapy areas; Global coverage
13
Rising use of biomarkers%
of
Tri
als
Ap
ply
ing
Bio
mar
kers
By
Ro
le
% G
row
th In
Ap
plic
atio
n O
f B
iom
arke
r R
ole
s O
ver
5 y
ear
s
Biomarker Role
Therapeutic Effect Toxic Effect Disease
Source: Biomarker Roles Within Clinical Trials. An Analysis Of Clinical Trials from 2007-2011 and 2012-2016. Clarivate Analytics. Coney, Gavin. Clarivate Analytics, Cortellis Clinical Trials Intelligence. www.clarivate.com
Disease marker: The biomarker indicates if a disease already exists (diagnostic biomarker), or how such a disease may develop in an individual case regardless of the type of treatment (prognostic biomarker).Therapeutic effect marker: The biomarker gives an indication of the probable effect of treatment on the patientToxic effect marker: The biomarker indicates a treatment-related adverse reaction
14
Growth in trial volume by phase
Source: Cortellis Clinical Trials Intelligence, March 2018
All trials within the time period, irrespective of whether the trial endpoints were met
Nu
mb
er o
f T
rial
s
End Year
Phase 2
Phase 3
15
Trial endpoint success: Reporting bias
Nu
mb
er o
f T
rial
s
Endpoints Achieved
Endpoints Met
Yes No
• Phase 1 – 4• 4,550 trials with known outcomes with start date 2005 or later• Trial endpoint success determined by explicit statement made by sponsor referencing
the endpoint
Source: Cortellis Clinical Trials Intelligence, March 2018
16
Are phase II trials failing quietly? Focus on phase 3 successes
Nu
mb
er o
f T
rial
s
Phase 2 Phase 3 Endpoints Met
Commercially Relevant: Is the primary intervention being studied owned by the trial sponsor?
Source: Cortellis Clinical Trials Intelligence, March 2018
Phase 2 Phase 3Phase / End Year
17
Biomarkers and trial endpoint success
Pro
po
rtio
n O
f Tr
ials
Are Biomarkers Used?
Yes No
Endpoints Met
Source: Cortellis Clinical Trials Intelligence, March 2018
Is at least one biomarker applied within the trial design. Could be biomarker for efficacy, toxicity or disease
18
Number of biomarkers vs endpoint success by phaseP
has
e
Average number of biomarkers
Endpoints Met
• All trials with known outcome; Segmented By Phase and then trial endpoint status• Mean of Number of Biomarkers within each of the 8 groups
Source: Cortellis Clinical Trials Intelligence, March 2018
Phase 1
Phase 4
Phase 3
Phase 2
1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 4.2 4.4
20
Segmenting biomarker role and endpoint successN
um
ber
Of
Tria
ls –
End
po
int
No
t M
et
Biomarker Role Combinations
• All trials categorized according to the combination of biomarkers• Ranked by descending number of trials that were reported to have reached their endpoint
(green = yes; red = no)• Yellow bar displays the % of trials in each category that were reported as successful
Source: Clarivate Analytics, Cortellis Clinical Trials Intelligence
Therapeutic Effect
Toxic EffectDisease
Therapeutic Effect
Toxic EffectDisease
Therapeutic Effect
Toxic EffectDisease
Therapeutic Effect
Toxic EffectDisease
Therapeutic Effect
Toxic EffectDisease
Therapeutic Effect
Toxic EffectDisease
Therapeutic Effect
Toxic EffectDisease
Therapeutic Effect
Toxic EffectDisease
Therapeutic Effect
Toxic Effect
Disease
% o
f ca
tego
ry t
hat
did
re
ach
en
dp
oin
t
Nu
mb
er o
f T
rial
s –
End
po
int
Me
t
21
Endpoint success and application of biomarker rolesN
um
ber
of
Tri
als –
End
po
int
No
t M
et
Biomarker Role Combinations
% of Trials = Yes
Source: Clarivate Analytics, Cortellis Clinical Trials Intelligence
Disease=YesNoneAllToxic=Yes
Therapeutic Effect
Toxic EffectDisease
Therapeutic Effect
Toxic EffectDisease
Therapeutic Effect
Toxic EffectDisease
Therapeutic Effect
Toxic EffectDisease
Therapeutic Effect
Toxic EffectDisease
Therapeutic Effect
Toxic EffectDisease
Therapeutic Effect
Toxic EffectDisease
Therapeutic Effect
Toxic EffectDisease
% of Trials using this combination that were
successful
Endpoints Not MetEndpoints Met% of endpoints met in category
Endpoints Met
Therapeutic Effect
Toxic Effect
Disease
Rank order now selected as % of trials in category that were reported as reaching their endpoint
Nu
mb
er o
f T
rial
s –
End
po
int
Me
t
22
Summary
• A biomarker strategy has been seen to positively impact likelihood of meeting trial endpoint
• That effect is not uniform:-
• Positive Phase 2-4; Negative Phase 1
• Toxic Effect markers strongly associated
• Disease markers not strongly correlated
• Need for the best intelligence to inform protocol design
Impact and Strategies Relating To Trial Amendments
Teresa Fishburne
24
Amendments
Source: CMR Global Clinical Performance Metrics Programme. CMR International, a Clarivate Analytics businessSource: Getz, et al Tufts CSDD
0
100
200
300
400
500
600
2 or moreamendments
2 or moreamendments
no or 1 amendment no or 1 amendment
Internal Protocol Review -‘Protocol synopsis developed’ to ‘Protocol approved internally’
520 days
43 days62 days
383 days
Protocol Approved to End of Enrolment‘Protocol approved internally’ to “last subject enrolled’
Med
ian
Du
rati
on
(in
Day
s)
25
Budget impact and key facts
❖ $20 Billion annual spend in direct and indirect costs
❖ 45% of all amendments are avoidable
❖ 2/3 Phase III trials > 1 global amendment
❖ 74% of all Phase II trials > 1 global amendment
❖ Rare diseases amendments > non-rare
Source: https://aspe.hhs.gov/report/examination-clinical-trial-costs-and-barriers-drug-developmentSource: CMR Global Clinical Performance Metrics Programme. CMR International, a Clarivate Analytics businessSource: Getz, et al Tufts CSDD
$0
$10,000,000
$20,000,000
$30,000,000
$40,000,000
$50,000,000
$60,000,000
Average Cost of Trial Average Cost of Trial After 1Amendment
Average Cost of Trial After 2Amendments
Phase II
Phase III
$12M $13.2M
$35M
$14.4M
$50M
$20M
26
Enrollment and cycle times
Source: CMR Global Clinical Performance Metrics Programme. CMR International, a Clarivate Analytics businessSource: Getz, et al Tufts CSDD
30%
60%
10%
0%
10%
20%
30%
40%
50%
60%
70%
Start-up Enrolment Treatment Duration
Proportion of protocol amendments by timing interval
Data are shown for Phase Ip, Phase II and Phase III studies that completed enrolment (Last Patient Enrolled milestone) between 2011 and 2015 and had one or more protocol amendment. This analysis includes ongoing and terminated studies.
Amendments
Enro
llmen
t
0 1 2 3
Cycle time
Trial Budget
Strategies❖ Biomarkers❖ Inclusion / Exclusion Criteria
27
Strategy to reduce amendments – use of biomarkers
❖ Rare disease programs and programs that utilized selection biomarkers had higher success rates at each phase of development vs. the overall dataset.
❖ A three-fold higher likelihood of approval from Phase 1 was calculated for programs that utilized selection biomarkers (25.9%, n=512) vs. programs that did not (8.4%, n=9,012).
❖ Bio article concluded that the enrichment of patient enrollment at the molecular level is a more successful strategy than heterogeneous enrollment.
❖ ~150 biomarkers are included on FDA approved drug labels. Such a strategy is associated with improved approval success and reduced approval timelines
Source: Clarivate Analytics, Cortellis Clinical Trials IntelligenceSource: Hay et al; Clinical development success rates for investigational drugs. Nature Biotechnology (2014)Source: BIO, Biomedtracker (2016)
0%
10%
20%
30%
40%
50%
60%
70%
80%
Ph 1 - Ph 2 Ph 2 - Ph 3 Ph 3 -NDA/BLA
Ph 1-Approval
WithoutBiomarkers
With SelectionBiomarkers
Probability of Success With and Without Biomarkers
28
Strategy to reduce amendments – Inclusion/exclusion criteria
• Build off approved/launched competing drugs
with like MOA/indications
• Build off internal Phase 1/2
• Literature search
• New or novel criteria
• Incorporating Biomarkers:
• By roles (Therapeutic effect, disease marker, toxic
effect)
• By type (genomic, proteomic, physiological,
biochemical, cellular, structural,
anthropormorphic)
• Biomarkers to address responders vs non-
responders
28Source: Clarivate Analytics, Cortellis Clinical Trials Intelligence
• Newly emerging medical knowledge
• Regular violations of entry criteria
• Low recruitment rates
29
Overall strategy to reduce amendments – Key points
Protocol Review Biomarkers Inclusion / Exclusion Criteria
• 19 additional days of internal review= 0 or 1 amendments
• Longer internal review time = shorter trials
• Incorporate more internal review approaches
• Bundle non-urgent amendments, where applicable
• programs that utilized selection biomarkers had higher success rates at each phase of development vs. the overall dataset.
• enrollment at the molecular level is a more successful strategy than heterogeneous enrollment.
• ~150 biomarkers are included on FDA approved drug labels.
• Using internal data to build criteria
• Build off
approved/launched
competing drugs
with like
MOA/indications
• Assess feasibility / appropriateness
• Use of predictive biomarkers to identify responders vs non-responders
30
Questions
Cortellis images, CTI, CCI
To learn more go to: Clarivate.com/Cortellis