Biobanking and genomic research: Some special needs

1
(No Transcript)
2
Biobanking andgenomic researchSome special
needs

• William W. Lowrance, PhD
• (lowrance_at_iprolink.ch)
• September 27, 2007

3
Need to make clearer distinctions among the
kinds of data

• Biobanks, major platforms of data
• and biospecimens curated as broad resources -
  banks - for research
• Genetic/genomic project data and biospecimen
  collections, cordoned-off
• by research protections
• Clinical genetic test data, held under
• medical confidentiality

4
Distinctions, cont.

• Genetic-related public-health data,
• protected by public-health laws
• Other sets of biological materials and derived
  data, such as forensic collections, held under
  various regimes.
• Failure to make firm distinctions is causing
  serious confusion in dialogue and regulation!

5
Genetic/ genomic/ biobank exceptualism? My
view is that

• Clinical genetic-test and related data should be
  treated like other sensitive medical data
• But until a number of issues get sorted-out and
  reliable protections are in place, most genomic
  and research biobank data deserve special
  attention.

6
Need to refine aspects of the whole suite of
research protections

• - Informed consent
• - De-identification
• - Research-ethics oversight
• - Safeguards
• - Data-release policies and practices
• - Barriers against access for
• non-research purposes
• - Sanctions against misuse.

7
Need to rethink the construal of consent

• Traditionally, consent has been meant to be
  "fully informed"
• But biobanking and genomic data,
• and their risks, are very difficult for most
  people to comprehend
• So, in consenting informed of what?

8
In my view, consent negotiations should inform
of

• - the purposes
• - the overall plan and any data- or
• specimen-collecting that may involve
• them directly
• - the disclosure risks generally
• - the auspices and protections that
• make "the deal" trust-worthy
• - anything else asked about.

9
Need to cope with identifiability

• The challenge is that genomic data
• - are extensive
• - are very fine-grained
• - influence many personal attributes
• - hold implications about family
• - are intrinsic to the body
• - don't change during the lifetime
• - are unique to the individual.

10
Identifiability, cont.

• Whether and how to de-identify depends on the
  character of the data, the intended uses,
  consent, disclosure risks, and safeguards.
• As an alternative or complement to
• de-identification, controlled data-release
  should be seriously considered.

11
Need to improve data-release

• For open release, must become clearer as to "how
  much" genome can be exposed without undue
  disclosure risk
• For controlled release, must attend to the terms
  of release agreements, stewardship, security, and
  enforcement.
• Ref Lowrance and Collins,
• "Identifiability in genomic research,"
• Science 317, 600-602 (August 3, 2007).

12
In all of this

• We must facilitate health research for the
  collective public good and at the same time
  protect the individuals with whom the data and
  biospecimens were, or are, associated!