AcademicIndustry Relationships in the Life Sciences: A U'S' Perspective

1
Academic-Industry Relationships in the Life
Sciences A U.S. Perspective

• David Blumenthal MD, MPP
• Cambridge University
• October, 2002

2
(No Transcript)
3
Agenda
1. The Reason 2. The Record 3. The Rub 4. The
Resolution
4
The Reason

• Public policy
• Bayh-Dole Act of 1980.
• Public support of research, and expectation of
  return.
• Intellectual property law
• Lincoln add the fuel of greed to the fire of
  genius.

5
The Reason (cont.)

• Changing role/norms of the university, its
  faculty, and medical professionals
• Obligation to apply research results.
• Erosion of norms of science and professionalism
  under the influence of market pressures?

6
The Record

• Data Sources
• Prevalence and magnitude
• Commercial benefits
• Academic risks

7
The Data

• Surveys of university and medical school faculty
• 1985, 1995, 1996-97, 1999-2000.
• Surveys of industry
• 1985, 1995.
• Surveys of students/trainees.
• 1986.

8
Prevalence

• Research support from industry
• 21-28 percent of faculty.
• Equity in related companies
• 7-8 percent of faculty
• Consulting 50 percent of faculty

9
Industry research support (PIs) Percent of PIs
with industry support
Blumenthal et al, Science 232 1361, 1986
Blumenthal et al, NEJM 325 1734, 1996
Blumenthal et al, Unpublished data from a survey
of faculty at academic health centers
10
Industry research support (PIs) Proportion of
research budgets from industry
Blumenthal et al, Science 232 1361, 1986
Blumenthal et al, NEJM 325 1734, 1996
Blumenthal et al, Unpublished data from a survey
of faculty at academic health centers
11
Personal financial relationships (PIs) Consulting
Blumenthal et al, Final Report to HHS AIRs
in Biotechnology, 1987 Blumenthal et al,
Unpublished data from a survey of faculty at
academic health centers
12
Combined relationships Geneticists and other
life scientists (2000)
(Last three years)
Source Unpublished data from a survey of
faculty in geneticists and life sciences
13
(No Transcript)
14
Magnitude

• Size
• 71 less than 100,000/year
• 6 more than 500,000/year
• Estimated annual level of support
• 10-15 of total university funding for life
  sciences RD

15
Duration
Percentage of Firms
16
CommercialBenefits
Percentage of Firms
17
Benefits Commercial
Commercial productivity


Percent of faculty

significantly different from comparison group
18
ConsequencesProductivity of university and
other research
Large Firms
Small Firms
Patents Academia 1.7 6.7 Elsewhere 1.2 3.5 Prod
ucts Academia 5.0 26.8 Elsewhere 1.4 27.9 Sales
Academia 18.3 22.0 Elsewhere 88.0 29.5
per 100 million invested
19
Commercial benefits

• Silicon Valley Route 128
• Academics participated in founding twenty-four
  F500 and over six hundred NF500 life science
  firms
• Local spillovers into NBEs associated with star
  scientists

20
Consequences Academic Average publications in
the last 3 years
All differences are statistically significant
(p-value lt0.001)
Source Unpublished data from a survey of
faculty in geneticists and life sciences
21
Consequences Academic (1995)
Data withholding
56
47
Agreements sometime require withholding beyond
patenting
Data actually withheld beyond patenting
22
Research funds and Data-withholding(faculty
survey 1996-7).
23
Commercialization and Data-withholding
24
Publication delays and data withholding


Odds ratios

significantly different from reference value
25
Data Withholding Has Caused
26
Why Have You Denied?
27
Consequences Academic
Change in direction of research

Percent faculty
significantly different from comparison group
28
Bias in science

• Accumulating evidence that studies supported by
  industry or conducted by scientists with
  financial relationships are more likely to favor
  industrial interests.

29
Data on Bias

• Stelfox et al, NEJM, 1998
• Calcium Channel Blockers 96 percent of
  supporters v 37 percent of critics had financial
  relationship with manufacturer.
• Rochon et al, Arch. Int. Med., 1994
• 99 percent of industry trials of NSAIDs found
  agents equal or superior to competing drugs.

30
Data on Bias (cont.)

• Friedberg et al, JAMA, 1999
• 5 percent of industry sponsored trials reported
  unfavorable findings for cancer drugs compared to
  38 percent of non-profit funded studies.
• Cho and Bero, Ann Int Med, 1996
• 98 percent of studies funded by industry in
  symposium proceedings/journals favored drug.

31
Threats to patients

• Gelsinger case, U. of Penn, 1999.
• No convincing evidence that COI has directly
  injured a patient who was subject of research,
  but appearance has been created.

32
The Rub Conflict of Interest
--Occurs when two or more interests
collide. --Interests are things of value to
somebody or something.
33
Conflict of interest

• Usually means that pursuit of one interest
  detracts from pursuit of the other.
• To resolve a conflict of interest, we must
  choose between them, or find a new way to
  reconcile them through creative management.

34
Hierarchy of interests

• Not all interests are equal.
• Implicitly or explicitly, we assign them a rank
  in making decisions among them.

35
Colliding interestsin biomedical research

• Interests of Science vs. interests of researcher
• financial interests of researchers cause concerns
  about bias, secrecy and diminished
  quality--threats to scientific enterprise and
  scientific progress

36
Colliding interestsin biomedical research,
(continued)

• Interests of patients vs. interests of researcher
• financial interests of researcher cause concerns
  about harm to research subjects - patient
  protection

37
Colliding interests (cont.)

• Interests of student v. interest of faculty.
• Faculty members interests in commercial outcomes
  of research may adversely affect educational
  experience of students/trainees.

38
Colliding interests in biomedical research
(continued)

• Interests of public health/economy v interests of
  patients, science, researchers, students.
• AIRs that threaten patients/scientific enterprise
  also hasten practical application of research
  results, with corresponding economic and health
  benefits.

39
The ResolutionWhat should be done?

• COI, real or apparent, that threatens patient
  well-being not acceptable.
• C0I that may cause bias, secrecy, or change
• in research direction requires understanding,
  weighing, and where possible, managing benefits
  and risks.

40
Specific actions

• Research
• Exploring benefits and risks.
• Exploring effects of alternative policy
  interventions.
• Prohibition
• Meaningful financial conflict of interest
  involving living human subjects
• Institutions/supervisors/faculty holding equity
  in same companies (Boston University)

41
Specific actions (cont.)

• Management of AIRs
• Disclosure of all relationships, including
  institutional equity holding in related
  companies.
• Limitations on gains in certain circumstances
  (sale of equity in companies supporting faculty
  research).