This second post is a forwarding of a post I originally sent to a person
off-list.  I think it would be useful to repeat it for the benefit of
others:
 
Hi Chris.
 
>Hi, something in Anne Ryan`s post from the AFA&LIFE bothers me slightly,
>the phrase ( high incidence of false positives }.  It`s a term bandied
>about in the north of the UK mostly by people who do not ( will not ) test
>their ferrets.  The people at the UK testing lab said the test was very
>accurate.  A positive test is nearly 99.98% accurate.  If it is taken to
>mean that a ferret which has tested positive for the antibodies but has or
>did not show any clinical symptoms for a long time ( 6 to 12 months ) is a
>false positive be very wary.  I know of one hob purchased as a kit who in
>his second year of life infected his 2 play mates, this ferret is now
>nearly 4 and is very fit and healthy.  The 2 infected ferrets were both
>dead within 6 months.  Would someone please define ( false positive ) for
>me.
 
I'm not sure how much background you have in immunology, so I'll try to
explain it succinctly.
 
Most infectious diseases these days are diagnosed using a test called an
ELISA (enzyme linked immunosorbent assay).  Basically, without going into
the nuts and bolts of the assay, an ELISA tells the diagnostician how much
antibody reactive with a particular disease protein is in the sample being
tested.  To do this the test uses a sample of the disease protein (an
'antigen') to 'capture' any antibodies in the sample.  Implicit in the
interpretation of the test is that only antibodies 'specific' for the
antigen will bind to it - antibodies to protein other than the antigen
being tested for should (theoretically) not bind to it, and hence not be
detected in the assay.
 
Trouble is, sometimes antibodies that are 'intended' by the immune system
of the the patient to react with something else will coincidentally react
with the antigen being tested for.  As a hypothetical example, a ferret may
have developed antibodies to a proteinacious dietary component that may
structurally resemble the protein being used in the ELISA to detect
antibodies produced against 'flu, (or distemper, or whatever).  In this
case, the 'cross- reactivity' (or non-specificity) of the anti-dietary
component antibody gives the same result in the ELISA that a specific
anti-disease antibody would.  This is a false positive result.  A 99.98%
accuracy in the context of positive result means that out of ten thousand
random positive results in an ELISA, two results will be ones that are
caused by a cross-reactive or nonspecific antibody binding, rather than
by the intended specific reaction of an anti-disease antibody to the test
antigen.  False positive may sometimes occur when antibodies from a
(possibly insignificant) disease to which the patient has been exposed
cross-react with the antigens of a (more significant) disease in an assay.
 
In these cases the patient giving the fasle positive result is usually
entirely healthy - the test result is an 'artifact' of the serendipitous
cross-reactivity.  On the other hand the hob you spoke of would in all
likelihood give a very strong true positive result in an ELISA.  The fact
that he remains healthy is an indication that his immune system is able
to maintain the disease at a level that does not endanger his well-being.
However, at the time of exposure to the other two ferrets he may still have
been producing infective virus particles, hence their sickness and death.
Often an apparently healthy animal or person can produce infective microbes
for months or even years without showing clinical signs of the disease
themselves - the classic example in immunology was a lady named 'typhoid
Mary'.
 
For significant human diseases such as HIV there are often secondary
assays that may be used to discriminate between true positives and false
positives.  A procedure called a 'Western blot' can often help in this.
The best way to confirm a true positive is to actually isolate a sample of
the target virus.  However, these tests are usually expensive and may not
be used if it is not 'cost-effective' in the context of the disease being
screened for.  Clinical history is also an integral component in arriving
at a reliable diagnosis for a particular disease.
 
CEPs (or as they're known in Britain and Australia, counter
immunoelectrophoresis - CIE) work on the same antibody-antigen binding
principle as an ELISA, but the execution of the assay is different.  One
of the main things with CIEs is that they generally use a non-purified
source for the antigen, and hence there are many different proteins in the
mix.  This greatly increases the chance that antibodies in the patient's
sample may cross-react, leading to a false positive result.  The advantage
of ELISAs is that nearly without exception they are based on a purfied
capturing antigen.
 
Harko Werkman,
Frog Research Group,
Department of Biological Sciences,
University of Newcastle,
Callaghan,
NSW. 2308.
Australia.
Ph: (+612) 4921 6253
Fax: (+612) 4921 6923
Email: [log in to unmask]
[Posted in FML issue 2954]