Immunology%20basics

1
Immunology basics Any normal person is ALWAYS
making antibodies against anything that enters
the body. The molecules that elicit antibody
responses are called antigens. Antigens can be of
any origin animal, vegetal or even mineral. Any
normal individual will have antibodies against
antigens commonly encountered in the environment,
such as pollen, food, and also against
microorganisms dwelling in the skin and digestive
tract (the symbiotic flora we ALL possess).
During an acute infection with a pathogen
antibodies are made, but they VERY RARELY exceed
1 of the TOTAL antibodies found in the blood.
The HIV (or any serological) test works by
detecting antibodies that recognize HIV antigens.
If those antibodies are present (HIV antibody
positive), then the person is assumed to be
infected with HIV. Mind that the same reasoning
applies to finding antibodies against ANY
antigen positivity is indicative of an encounter
with the source of that antigen (a current or
past infection).
2
What the "HIV test" measures So, the HIV test is
actually detecting a VERY small amount of serum
antibodies capable of recognizing the HIV
antigens used in the test, which probably amount
to MUCH LESS than 1. Tuberculosis (TB) the
result of infection with the bacterium
Mycobacterium tuberculosis (Mt) has been taken to
be one of the MOST representative of the AIDS
dieseases, basically because there is "strong
evidence" to support that immune dysfunction can
allow TB to progress to overt disease and also
because "several studies show" that TB and HIV
seropositivity are often present at the same time
in the same individual. The possibilities for
this frequent coincidence are two 1- HIV-AIDS
makes people more prone to develop TB as the
"defenses are down" and although TB patients
often have anti Mt antibodies, the HIV test truly
detects those antibodies directed against HIV. 2-
Mt, in a proportion of individuals, is
responsible for generating antibodies that
recognize HIV components (antigens) of the HIV
test.
3
The key points 1- At any given time, humans have
antibodies against an enormous amount of
different antigens from all sources (not only
pathogens). 2- VERY rarely the antibodies
directed against ONE antigen source (pathogen or
not) exceed a small fraction of the total
antibodies circulating. 3- If you want to
ascertain the existence of ANY antibody capable
of recognizing any particular antigen source (say
HIV, Mt, etc.), you must make sure you can pick
that small subset from a huge amount of
"irrelevant" antibodies. This is what determines
the specificity of the test.
4
So how do you go about detecting these
antibodies? First thing, you need a source of
the antigen the antibodies you wish to detect can
bind. Then you immobilize the antigens on a
surface (very much like a coat hook on the wall)
and you incubate some blood (or serum) on this
surface which is covered with the antigen. The
antibodies which can bind to the antigens will do
so and become, in their turn, immobilized, such
that you can remove the rest of the blood (or
serum), which contains many things, among them
that gt99 of antibodies that do not bind the
antigens. Finally, there are ways to measure the
amount on antibodies that stayed bound to the
antigen-coated surface, the more there are, the
more likely that the source of that antigen
elicited an antibody response (say HIV or Mt).
5
If the amount of antibodies that remained bound
to the immobilized HIV antigens used in the HIV
test is above a certain value (the "cut-off"
value) then the sample in question is positive,
and conversely, if the amount is below the
cut-off, the sample is negative. It is important
to note that reproducibility (precision) is
essential, if this is not so, a sample can score
positive in the test and negative upon
re-testing, or viceversa, and which one is the
"true" value remains to be determined. The
cut-off value is determined by testing a sample
known to be negative, and calculating the
standard deviation of multiple determinations
(generally three) and then adding the equivalent
of three times the SD of the mean. In practice,
the cut-off is the average of the negative
control values 0.1.
6
Blood of a true HIV antibody subject Addition of
serum to the immobilized HIV antigens
Antibodies capable of binding the HIV antigens
7
Blood of a true HIV antibody subject Incubation
to allow antibody binding to antigens
8
Blood of a true HIV antibody subject Wash and
remove irrelevant (unbound) antibodies
Value above the cut-off.
9
So, what happens if we remove antibodies which
recognize antigens other than HIV's before
incubating on the HIV antigen coated plates?
Preadsorption
10
Antibodies (if there are any) that bind the
non-HIV antigens will be immobilized and removed
from the sample, but the outcome of the HIV
antibody test should not change
11
Then we transfer the rest (unbound antibodies) to
a plate for the measurement of remaining anti-HIV
antibodies.
12
In principle, if the antibodies against HIV
antigens are SPECIFIC, they should not be
adsorbed by other antigens, so the subject should
remain HIV.
Value should remain above the cut-off.
13
Evaluation of cross-reactivity between
anti-Mycobacterium tuberculosis antibodies and
HIV antigens in South Africa. Study design -
Serum samples from clinically diagnosed
tuberculosis patients were tested. Of these, 99
were of known to be positives in the HIV antibody
test and 63 were of unknown HIV serological
status. All were diagnosed and in treatment for
TB. - The idea was to test if antibodies
responsible for positivity in the "HIV test"
could be removed from serum by incubation against
M. tuberculosis antigens before running the HIV
antibody test.
14
- The original study proposal called for a
stratified group of samples. most of which would
be clear but not off-the-scale positives.
However, due to operating constraints, the study
was done on an uncharacterized population of
samples which consisted of mostly very high
positives (values above 2 comprise about 90 of
the samples). - Additionally, an unexpected but
very significant variation between duplicate
samples was observed, which could be the result
of the preadsorption of the sera OR to inherent
variability in the HIV test. The latter turned
out to be the case, as will be seen. This
variability is particularly dangerous for those
that score low positive values or even negative
values.
15
High positives
Mid-positives
16
From the previous graph, no significant
differences can be observed between those samples
that were preadsorbed against Mycobacterial
antigens (of strains H37Rv and CSU93), those
preadsorbed by an inert protein control (Gelatin)
and those that were not treated. However, a
significant variation in those samples done in
duplicate (CSU93 and gelatin) was observed. The
results can be interpreted as follows - Most of
the samples are so high that removal of cross-
reacting antibodies (if they exist) diminishes
the signal insignificantly. - Variability in
the test gives values that are not-reliable for
the quantitative assessment of decrease in
signal.
17
Variability in duplicates preadsorbed with CSU93
or gelatin and comparison to untreated control.
Values are those of chosen samples with values
low to mid range.
Note the high standard deviations in many
samples. At this stage we did not know if
originating in the preadsorption step or in the
test itself.
18
So, although we had not anticipated that the test
could be so variable, we undertook to test if the
results of the HIV antibody test were reliable by
testing untreated (that is, unmanipulated)
samples in triplicate and see how precise the
test outcome was. The results were surprising,
in that we never expected this level of
variability of samples tested 3 times on the same
assay plate, as the following graph shows.
19
Intra-assay variability What happens if you test
the same sample 3 times?
20
For a number of samples, the HIV antibody status
was not known, upon testing, 14 of 58 were
negative (? 24, prevalence of 76???)
21
Conclusions The HIV antibody test used in this
study was that of Biomérieux, and was used
strictly according to the manufacturer's
instructions and under the same conditions used
in MEDUNSA's Department of Virology, where it
serves as a "confirmatory" test for HIV
seropositivity. 1- The test is highly variable
(imprecise) at low to mid-range values.2- Most
of the samples that turn out positive (by the
test's criteria) are far up in the upper limits
of detection. 3- In the upper limits of
detection, cross-reactivity cannot be assessed,
as we are far from the linear detection range of
the instrumentation to detect and quantify drops
in readings. 4- In the low to mid-range limits of
detection, we could not assess the effects of
preadsorption as the test is NOT acceptably
reproducible.
22
This pilot study has illuminated several aspects
of the HIV antibody test that were not known to
us, mostly because IF they are known, they are
not in the test's instruction manual. The most
relevant is the lack of reproducibility, and it
warrants further study, not only in the test we
used in this occasion but on as many as possible
of those which are in use in South Africa. The
preadsorption experiments must be carried out as
originally planned, with more antigens and more
samples, and the optimal conditions for
preadsorption must be sought and established in
the experimental procedures. A critical aspect
is to find samples which give positive reactivity
within the linear range of the assay, which will
have to be determined for each assay to be
evaluated. Complementary to this, is to do a
"normalization" of readings by diluting strong
positive samples down to values within the linear
range of the test.
23
To the best of our knowledge, and besides the
preadsorption experiments to be carried out, a
stringent evaluation of the HIV antibody test's
precision has not yet been conducted. Basically,
we rely on what the manufacturer says about its
test. Our preliminary results indicate that it
must be done, not only to enable the
interpretation of the preadsorption data but also
because it seems likely that an unacceptable
number of samples of low to mid-range reactivity
may turn positive or negative upon re-testing,
limiting the confidence with which an HIV
seropositivity result can be assigned. Considerin
g that an HIV result certainly ruins the life of
anyone receiving it, the level of uncertainty has
to be established to know whether it is
acceptable or not, independently of the
manufacturer's claims, which are usually based on
optimal testing conditions more or less removed
from the real working conditions encountered in
the diagnostic laboratories in South Africa or
anywhere else.