Transcript Lecture - Berkeley Law

Business of
Drug Discovery
David Ridley
January 2009
Health Sector Management
David Ridley



Assistant Professor at Duke
University
Teaching: health economics
for MBA students
Research: co-author of
priority review voucher
proposal
2
Bill Gates described priority review
voucher at Davos last year
“But some of the highest-leverage
work that government can do is to
set policy and disburse funds in ways
that create market incentives for
business activity that improves the
lives of the poor. Under a law signed
by President Bush last year, any drug
company that develops a new
treatment for a neglected disease
like malaria or TB can get priority
review from the Food and Drug
Administration for another product
they've made. If you develop a new
drug for malaria, your profitable
cholesterol-lowering drug could go
on the market a year earlier. This
priority review could be worth
hundreds of millions of dollars.”
3
Drug Discovery:
Who, What, When, Where, How
When? Drug development timeline
 How much money?
 By whom? License in R. Contract out D
 Where? Clinical trial location
 What? Orphan and neglected diseases

4
R&D Is Risky & CostlyCompound Success
Years
Discovery:
(2-10 years)
0
2
4
Phase I: 20-80 healthy
volunteers to determine
safety & dosage
Phase III: 1000-5000
volunteers to monitor
adverse reactions to
long-term use
Additional
post-market
testing
Rates by Stage
6
8
10
12
14
16
Source: DiMasi et al. 2003, Tufts
Preclinical:
laboratory &
animal tests
5,000–10,000
Screened
Phase II: 100300 volunteers
to look for
efficacy & side
effects
FDA Review
Approval
250
Enter Preclinical
Testing
5
Enter
Clinical
Testing
1
Approved by
the FDA
Net Cost: $802 million
invested over 15 years
5
Time
DiMasi et al. 2003
www.fda.gov/
Mean
Mean time
cder/rdmt/
length
to next
Discovery
Phase I
Phase II
Phase III
FDA approval
1.8
2.1
2.5
1.0
2.2
2.8
0.5-1.5
6
Vary by disorder
Source: Chris Adams & Van Brantner (FTC), Health Affairs, 2006
Probabilities and duration estimated using PharmaProjects data
Did not observe actual costs
7
Tuberculosis
Probability of anti-TB by 2010 < 5%
Source: Glickman et al., Science, 2006
But anti-infectives already used now in
phase III so might be higher
Preclinical Phase I Phase II Phase III
Expected duration (years)
1.6
1.8
2.5
5.5
Patients
-
104
264
1000
USA expected cost ($m)
5.1
0.7
3.4
26.6
Uganda expected cost ($m)
5.1
0.2
1.6
8.2
8
Patent = 20 years
Effective patent < 20 years


Patent life is 20 years
Effective patent life is considerably shorter
–
–

Patent restoration:
–
–
–

Clinical trials
FDA time
1/2 R&D and all FDA time
<=5 years extra
<= 14 years total
But time value of $ so 1 year delay for blockbuster
costs $300 million (Ridley et al. 2006)
9
Drug Discovery:
Who, What, When, Where, How
When? Drug development timeline
 How much money?
 By whom? License in R. Contract out D
 Where? Clinical trial location
 What? Orphan and neglected diseases

10
$802 million



Cost per new drug averages $802 million
(year 2000 dollars) (Source: DiMasi, Hansen,
Grabowski, 2003)
The estimate has been criticized, even used
as the title for a critical book
Is it really “just” $200 million?
–
–
Cut in ½ for opportunity cost of capital
Cut in ½ again for share of dry wells
11
Cost of R&D
$200m Clinical $130m+Pre-clinical $70m
$400m Dry wells
$800m Cost of capital
$900m Phase 4
$1300m Inflation
•We often hear $800 million but with post-approval costs it’s $900
million and with inflation adjustments it’s > $1300 million
•Accountants might say, however, $200 million (+ inflation)
12
Sarafem different?





How costly are line extensions?
Manufacturer sought a line extension for Prozac
Branded Sarafem as treatment for premenstrual
dysphoric disorder (PMDD)
Quick search of clinical studies. Appears that 3
clinical trials testing effectiveness of Sarafem for
PMDD. Trials had 260, 42, and 19 patients. At
$10,000 per patient cost is $3.2 million
Caveat: haven’t seen records, this is just back of
envelope
13
Torcetrapib different?

Average: $130 million (in 2000 dollars)
–
–

Torcetrapib: $800 million (?!)
–
–
–
–

DiMasi et al. (2003)
+ other costs + cost of capital + dry wells = $800m
“Pfizer announced that it had pulled the plug on the medicine entirely,
turning the company’s nearly $1 billion investment in it into a total loss.” –
NYT 2006
“Pfizer had spent close to US$1 billion in the development of Torcetrapib
which included US$ 800 million in the clinical trials itself” - case
About $90 million for plant in Ireland
“The numbers are correct. They are based on a simple sum of outlays
related to Torcetrapib R&D over the years. They do not include cost-ofcapital or NPV.” – Pfizer friend
Why Torcetrapib more costly?
–
More people: 7500 in treatment, 7500 in control in Illuminate; 10,000
others
14
Devices & diagnostics different?

Clinical trials often aren’t required for approval
–

For devices, 510(k) submissions demonstrate
substantial equivalence to approved
But
–
–
–
For more innovative, clinical trials necessary for
Premarket Approval (PMA)
For convincing providers and payers
For the future. Critical GAO report encouraged FDA
to subject devices to more rigorous review
15
Biotech different?

Faster
–
–

More nimble and creative (different corporate culture)
Fewer safety issues (more relevant to early biotech era
development)
Slower
–
–
Less experience in clinical development and interacting with
regulators
More costly manufacturing process R&D and production of
clinical supplies
Source: DiMasi and Grabowski 2007
16
Pre-Approval Capitalized Cost per
Approved New Molecule
1,318
Millions (2005$)
1,241
879
899
626
615
376
523
439
Preclinical
Clinical
Biotech
Pharma
Total
Pharma (time-adjusted)**
** Based on a 5-year shift and prior growth rates for the preclinical and clinical periods
New pharmaceutical R&D # is not $802 or $899 but $1318
Source: DiMasi and Grabowski 2007
17
Pre-Approval Cash Outlays per
Approved New Molecule
Millions (2005$)
672
522
559
437
361
198
136
316
150
Preclinical*
Clinical
Biotech
Pharma
Total
Pharma (time-adjusted)**
* All R&D costs (basic research and preclinical development) prior to initiation of clinical testing
** Based on a 5-year shift and prior growth rates for the preclinical and clinical periods
Source: DiMasi and Grabowski 2007
18
Clinical Trial Size Increases
with

Smaller treatment difference
relative to “noise” (variation)
(Δ)
Want smaller false positive
rate (α)
Want higher power (1- )

Intervention group size:


m
2( z1 / 2  z1  )

2
2
19
Pivotal Trials



Often 2 pivotal trials, but 1 for orphan
“The Food, Drug and Cosmetic Act… calls for ‘adequate and wellcontrolled clinical trials’ to support approval of a product for a specific
indication. In traditional drug development terms, this language
translates into at least two ‘pivotal’ trials to support registration. The
purpose of these two clinical trials is to have one trial confirm the
results of the other, thereby reducing the probability of outcomes
associated with false positives” (Brenner & Ellis-Grosse 2006).
For example, all 4 of the following groups
1.
2.

Duke: treatment & control (placebo)
Quintiles (CRO): treatment & control (placebo)
But orphan drugs are often approved based on one pivotal Phase III
clinical trial (or a combined Phase II/III trial).
20
R&D by function, PhRMA, 2006
Prehuman/Preclinical
$12 b
27%
Phase I
$3 b
7%
Phase II
$6 b
13%
Phase III
$12 b
28%
Approval
$3 b
6%
Phase IV
$6 b
13%
Uncategorized
$3 b
6%
Total R&D
$43 b
100%
Does not account for
time value of $.
Source: PhRMA
Industry Profile 2008
21
Postapproval Drug Safety


Ridley, Kramer, Tilson, Grabowski, &
Schulman (2006)
11 companies participated
–

For 2003:
–
–
–

Accounting for 71% of 2003 sales by
top 20
Mean pharma sales: $17 billion
Mean safety spend per company:
$56 million (0.3% of sales)
Estimated spending by top 20 drug
manufacturers: $800 million
Context
–
–
Total budget for FDA’s Office of Drug
Safety in 2003: $ 22.1 million
Industry spend on R&D 2003: 15.6%
of sales
22
23
Drug Discovery:
Who, What, When, Where, How
When? Drug development timeline
 How much money?
 By whom? License in R. Contract out D
 Where? Clinical trial location
 What? Orphan and neglected diseases

24
Who innovates? Big pharma or others?
2006 Domestic R&D by PhRMA Members
Licensed-in
$6 b
17%
Self-Originated
$24 b
70%
Uncategorized
$4 b
13%
Total R&D
$34 b
100%
17% of dollars, but probably > 17% of products
Source: PhRMA Industry Profile 2008
25
Who conducts trials?
Contract research is an option





For contract research, how does the price for government and
foundations compare to the industry price?
“we provide the US government and Bill and Melinda Gates grantees
(and other non for profit entities working in the neglected diseases
space) with preferential terms of business.
One way is rather simple - discounted rates off our fees/hourly rates.
US discount and non-US discounts are different, with the more
aggressive reduction being provided for services in non US-locations.
Another way is to provide rebates at the end of a calendar year based
on revenues (tranche system). I would love to chat further if you are
interested. .”
- Vaila Clements, VP Public Health and Government Services,
Quintiles Transnational, January 2009
26
Drug Discovery:
Who, What, When, Where, How
When? Drug development timeline
 How much money?
 By whom? License in R. Contract out D
 Where? Clinical trial location
 What? Orphan and neglected diseases

27
T/F
 True
or false: clinical trials are
moving to India and China
28
Regional Distribution of
Biopharmaceutical Clinical Trial Sites
Western Europe and North
America declining but large
•Emerging (green)
grew from 8% in 2002
to 20% in 2006
•Largest losses in
Western Europe
between 2002 and
2004, while decline in
North America
between 2004 and
2006
• Growth is 5-10% so
not necessarily
leaving, just not
growing as fast
29
Source: Berndt, Cockburn, & Thiers 2008
Shares of Biopharmaceutical
Clinical Trial Sites
•Top 5 (US,
Germany,
France, Canada,
and Spain) host
66%
• Emerging now
host 17% of sites
30
Source: Berndt, Cockburn, & Thiers 2008
Average Relative Annual Growth Rates in
Shares
•Decreasing
shares in North
America,
Europe,
Venezuela
•Increasing
shares in China
and other
emerging
markets
31
Source: Berndt, Cockburn, & Thiers 2008
Drug Discovery:
Who, What, When, Where, How
When? Drug development timeline
 How much money?
 By whom? License in R. Contract out D
 Where? Clinical trial location
 What? Orphan and neglected diseases

32
Problem: Little private financial
incentive to develop treatments for
Orphan
diseases
Have $, but few
Neglected
diseases
Bioterrorist
attacks
Many, but poor
Unlikely to need and if
33
need then give away
How can we induce companies to
develop treatments for orphan
diseases, neglected diseases, and
bioterrorist attacks?
34
Solution: Push & Pull Mechanisms

Pull Mechanisms fund outputs
(drugs, vaccines)
Advance Markets
–
–

Guaranteed price creating $3
billion market
Malaria, TB, HIV
Transferable voucher for
extended patent life
–
–
Push Mechanisms fund inputs
(pay R&D costs)
Prize for treatment for
neglected diseases or
bioterrorism
Extra patent life for a different
drug
Push + Pull
 Orphan Drug Act
–
–

Tax credits & grants (push)
Marketing exclusivity (pull)
Priority Review Voucher
–
–
Priority review for other drug
Dengue, Chagas, …
35
1983 U.S. Orphan Drug Act

Eligibility
–
–


Push
–
–

Disease affects few people in U.S. (<200,000)
Or no reasonable expectation that profitable
Low expected returns sufficient not necessary
R&D tax credits: ½ clinical, even if not approved
Clinical research grant programs
Pull
–
–
FDA counseling and priority review
Guaranteed market exclusivity of 7 years
36
U.S. Definition of Disease Prevalence



From www.fda.gov/orphan/
(i) The disease or condition for which
the drug is intended affects fewer than
200,000 people in the United States
or, if the drug is a vaccine, diagnostic
drug, or preventive drug, the persons
to whom the drug will be administered
in the United States are fewer than
200,000 per year as specified in Sec.
316.21(b), or
(ii) For a drug intended for diseases or
conditions affecting 200,000 or more
people, or for a vaccine, diagnostic
drug, or preventive drug to be
administered to 200,000 or more
persons per year in the United States,
there is no reasonable expectation that
costs of research and development of
the drug for the indication can be
recovered by sales of the drug in the
United States as specified in Sec.
316.21(c).
37
Orphan Drug Laws
Country
Passed
Disease
prevalence
Push: grants
Pull: years
exclusivity
U.S.
1983
0.07% (200K)
½ clinical
7
Japan
1993
0.04%
(4 in 10K)
Average
$100K
10
E.U.
1998
0.05%
(5 in 10K)
Apply for
grants
10
38
Orphan Drugs Laws:
More Details on US & EU
US
EU
Tax credits for costs of clinical research
Access to centralized procedure for
marketing authorization
Annual grant funding to defray costs of
qualified clinical testing expenses
EU-funded research grants from Community
and member states
Assistance in clinical-study designs
Protocol assistance to optimize
development
7 years of exclusive marketing after
approval
10 years of exclusive marketing after
approval
Waiver of Prescription Drug User Fee Act
filing fees
100% fee reduction for protocol assistance,
50% fee reduction for marketing
authorization, 100% fee reduction for
pre-authorization inspections
39
Most Important Incentive:
Market Exclusivity


Market exclusivity was most sought after
incentive in Orphan Drug Act, said FDA
Particularly important for
–
–
biologicals that had uncertain patents
older chemical entities with useful orphan drug
indications
40
Orphan Drugs:
Lower Costs
Orphan drugs have smaller and
shorter clinical trials
 The 7 orphan drugs in 1999 were
approved with an average of 588
patients
 By contrast, non-orphan approvals
in the late 1990s had over 5,000
patients on average

41
Orphan Drugs:
Lower Lifetime Sales Profile
Millions (2000$)
600
Non-orphan
Orphan
500
400
300
Orphan
200
100
0
1 2
3 4 5 6
7 8 9 10 11 12 13 14 15 16 17 18 19 20
Year
Source: Grabowski and Vernon 2003
1990-1994 Sales of New Drug Introductions
42
Orphan Drugs Can Be Lucrative
Source: WSJ 2005
Update: Cerezyme = $300K/yr, $1.1b in
43
2007 (NYT March 2008)
Approvals of Drugs for Rare Diseases
160
146
140
120
100
87
80
60
40
20
10
0
1973-82
Source: WSJ 2005
1983-92
1993-02
1983 Orphan Drug Act
Source: FDA
44
Orphan Drug Act Conclusions



Orphan Drug Act has been a success in
encouraging many new drug approvals for
rare diseases
But few U.S. orphan drug approvals for
neglected diseases. Malaria vaccine would
qualify as orphan, but with so few highpaying customers, exclusivity does little good
Need additional incentive mechanisms for
neglected diseases
45
100
Perinatal conditions
Lower respiratory
HIV/AIDS
30.0
Diarrhoeal diseases
Malaria
Tuberculosis
Maternal conditions
Nutritional
Measles
deficiencies
Pertussis
10
20.0
Tetanus
Lymphatic filariasis
Meningitis
Syphilis
1
10.0
0.1
Annual Global Burden (Millions of DALYs)
↑Harm
Trachoma
Ascariasis
Schistosomiasis
Trypanosomiasis
Japanese
encephalitis
Dengue
Onchocerciasis
Leprosy
Hookworm disease
0.0
0.0
Leishmaniasis
Trichuriasis
Chagas disease
Diphtheria
Poliomyelitise
10.0
0
Chlamydia
Gonorrhea
Hepatitis BUpper respiratory
Otitis media Hepatitis C
0.1%
20.0
1%
30.0
10%
Developed Countries’ Share of DALY Burden
DALY=Disabilityadjusted life years
Disparity
Source:46Ridley,
Grabowski, Moe
Priority Review Voucher


Senator Brownback wrote
“After reading their proposal
in Health Affairs, I met with
Ridley and colleagues to
discuss the idea further, and
I subsequently drafted an
amendment…Indeed, their
idea is the heart of my
Elimination of Neglected
Diseases (END) amendment.”
Became law in 2007
47
Priority Review Voucher
Government/Society
2. Voucher
Developer of
treatment for
neglected disease
$
Manufacturer of
potential48
blockbuster
Priority Review Voucher
Creates Win-Win




Benefits people suffering from neglected
diseases
Benefits patients in the U.S. who receive earlier
access to the priority drug
U.S. patients also get earlier access to the
generic because drugs with faster FDA review
typically have earlier patent expirations
Concept could also be applied to treatments for
bioterrorism
49
SEC. 1102. PRIORITY REVIEW TO ENCOURAGE TREATMENTS FOR
TROPICAL DISEASES.
Subchapter A of chapter V of the Federal Food, Drug, and Cosmetic Act
(21 U.S.C. 351 et seq.) is amended by adding at the end the following:
`SEC. 524. PRIORITY REVIEW TO ENCOURAGE TREATMENTS FOR
TROPICAL DISEASES.
`(a) Definitions- In this section:
`(1) PRIORITY REVIEW- The term `priority review', with respect to a
human drug application as defined in section 735(1), means review and
action by the Secretary on such application not later than 6 months after
receipt by the Secretary of such application, as described in the Manual of
Policies and Procedures of the Food and Drug Administration and goals
identified in the letters described in section 101(c) of the Food and Drug
Administration Amendments Act of 2007.
`(2) PRIORITY REVIEW VOUCHER- The term `priority review voucher'
means a voucher issued by the Secretary to the sponsor of a tropical
disease product application that entitles the holder of such voucher to
priority review of a single human drug application submitted under section
505(b)(1) or section 351 of the Public Health Service Act after the date of
approval of the tropical disease product application.
`(3) TROPICAL DISEASE- The term `tropical disease' means any of the
following:
…
On September 27, 2007, President Bush
signed H.R. 3580, "Food and Drug
Administration Amendments Act of 2007"
into Law
50
Conclusion


When? Each of 3 phases of trials take couple of years
How much? $800 million
–
–
–

Big trials costly but need bigger if
–

–
License in: 17% of $ for in-license
Contract out: Quintiles: lower rate for Gates
Where?
–

Smaller treatment difference, Higher “noise” (variation),
Smaller false positive rate (alpha), Higher power (1-beta)
By whom?
–

½ is cost of capital, ½ is dry wells
Update: $1300
So if $325 million, you’re average
Emerging markets now host 17% of sites
What?
–
Incentives for orphan, neglected, biodefense
51