Please wait while the transcript is being prepared...
0:00
My name is Steve Humphries.
I'm the Emeritus British
Heart Foundation Professor
of Cardiovascular Genetics at
University College London.
I'm going to talk today
about the genetics of CHD
or coronary heart disease,
and how we've been moving
research findings
into patient benefit.
0:23
The outline of the
talk is shown here.
What we're really
talking about is what we
call precision medicine.
Which can be defined
as medical care
designed to optimize
the efficiency of
the therapeutic benefit
for particular groups of
patients by using genetic
or molecular profiling.
0:43
Here are the criteria for
any useful CHD tests that
we're thinking about.
Any test should be
evaluated according to
these four ACCE criteria.
The first is
analytical validity.
How accurate is the result?
What's the possibility of lab
error and things like that?
The second is clinical validity.
How accurate is the risk
estimate that we're associating
with the marker that
we're measuring
and that we're
measuring in the test.
The third is clinical utility.
How useful is this
particular test over and
above classical risk factors
that we already know about?
What's the false positive rate?
What's the false negative rate?
Then the final issue
is the ethical,
legal, and social
impact of the test.
The issue here is,
do DNA tests cause more
anxiety than other tests?
Let's start by looking
at analytical validity.
1:39
How well can we measure some
of the classical risk factors?
For example, cholesterol,
as you know,
is a classical risk factor
for cardiovascular disease.
In the population as a whole,
its level creeps up
slowly as people get
older and there are
certainly changes
during the life course,
for example,
pregnancy in a woman,
other events as well,
where cholesterol may
change quite quickly.
However, we can
measure cholesterol in
the laboratory with
pretty good precision.
The accuracy of the
test is 3% or better
and cholesterol is quite
a stable biomarker so
that within individual
variation if you take
a sample repeated times from
the same person is about 5%.
Cholesterol is a pretty good
biomarker from this aspect.
The second example is something
called C-reactive protein,
and I'll be discussing
this later in the talk.
It's an inflammatory
marker and it
varies quite considerably
from day to day and
particularly in response to
environmental challenges such
as infection, injury,
or severe stress.
For example, if you've
just run a marathon.
Although we can measure
it with good precision,
say within 3%,
the within-individual
variation is much greater.
It's about 10%.
Now, how about for DNA?
Well, the genome of an
individual is fixed at
conception and doesn't vary
throughout lifetime
with a few exceptions.
Of course, the error
rate if we're doing
this in a properly
accredited laboratory,
is less than 1%.
What you can see
from this is that
from an analytical
validity point of view,
using DNA as a biomarker,
has several advantages
compared with some of
the other biomarkers that
are currently used in
cardiovascular risk testing.
We can multiplex our tests
to determine an individual's
genotype in a single run,
anywhere between a few
to even millions of
single nucleotide polymorphisms
or genetic variance.
We can easily obtain enough DNA
from a mouthwash sample
to be able to do this.
There are some really
good advantages
about using DNA as a test biomarker
compared to some of the others.
Now let's move to look
at clinical validity.
How accurate is
the risk estimate?
