Facing up to anti-animal antibody interference
October 2003 Anne Paxton
It was way back in 1973 that researchers, looking at one of the
original RIAs for hepatitis, first noted anti-animal antibody interference
in immunoassays. But, says Larry Kricka, DPhil, professor of pathology
at the University of Pennsylvania Medical Center, laboratories have
not made a significant dent in the problem of interference.
“Here we are, 30 years later, still discussing this problem, still not
having the total solution to it. And it’s causing major, major
issues for the clinical laboratory.”
In a presentation at the annual meeting of the American Association for Clinical Chemistry
in July, Dr. Kricka explained that interference can be due to cross-reactivity,
interferents, or signal modulation. “It can occur because
an antibody doesn’t have the specificity you hoped it would have.”
Anti-animal antibodies, rheumatoid factor, and a host of other things such as
drugs, Chinese medicine, herbal drinks, and paraproteins can create
interference. Other things in the immunoassay process can modulate
signals to produce false-positives and false-negatives, he noted.
Examples are fluorescent molecules that might be present in the
sample, quenchers, and inhibitors of enzymes.
“Despite the fact that immunoassays have been with us so long and we’ve
developed so many of them and used them so extensively, these interferences
are still there,” Dr. Kricka pointed out.
Two recent studies of interferences in immunoassays have shown a significant false-positive
rate. In the first one, 11 specimens, some of which were rheumatoid-factor
positive, were circulated to 66 laboratories in seven countries,
and laboratories reported results with an 8.7 percent false-positive
rate. A second study involving 500 patients looked at thyroid-stimulating
hormone and gonadotropins, and there the percentage of incorrect
results was 0.5 percent.
Interferences occur in the typical sandwich immunoassay because the capture antibody,
derived from an animal, encounters antibodies that have a reactivity
toward it. “You have this anti-animal antibody looking at
this molecule and saying it’s from a mouse, and I like mice,
and it will bind to it,” Dr. Kricka explained. “It looks
at the conjugate, which is also a mouse antibody, for example, and
it says I like mice as well and binds to it. And you end up with
a false-positive.”
False-negatives
are perhaps more obscure and less understood, he said. In these
cases, an antibody saturates the capture antibody and saturates
to some extent the conjugate in solution. “So there is no
real possibility of forming any real complexes at all. No sandwiches
are created in complete form, and you get very low results.”
Whereas high
values in an immunoassay are alarming and noticed, low values are
usually not, and they may cause a false sense of security. “People
notice when someone’s got a high CEA or a high CA 19-9 or
high hCG or PSA. But false-negative interferences probably go much
more undetected and are infrequently seen in the laboratory, because
the values don’t cause any concern.”
Of the antibodies
that have been found in human serum that have an animal specificity,
human anti-mouse antibody, or HAMA, is probably the best known,
he said, but there are many others, including human anti-rabbit,
anti-goat, anti-chimera for the chimeric antibodies that are being
used now in therapy, anti-pig, even anti-human, anti-bovine, rat,
and horse.
It’s not
always clear how some people get anti-animal antibodies, but clearly
antibody therapy is one source, he noted. “Monoclonal antibodies
from mice are immunogens, so if you give them to people, surprise,
surprise—you get an immune response.”
Animal husbandry
and keeping pets have also been implicated. “For example,
if you’re looking for people who have got anti-mouse antibodies
circulating, probably the best place to go to is your local mouse
colony to see if the keeper of the mouse colony will give you some
blood.”
Since animal-based
reagents are also used extensively in pharmaceuticals, these can
be another source of human anti-mouse antibody, human anti-rat antibody,
or antibodies against other species.
Although assays
have been developed to measure anti-animal antibodies in blood,
they vary widely and are not standardized, he said. “The literature
would tell you that there’s between micrograms and grams per
liter of anti-animal antibodies circulating in the blood of some
people, so when you see numbers like this in the literature, you
need to just be a little cautious about them.”
Anti-animal
antibodies can last from days to years. “You can have a patient
who’s been given a mouse monoclonal some time ago and they
could still be positive for human anti-mouse antibody. They would
probably not even think to tell you about the administration of
that antibody.”
In a duration
study conducted in 1988, he noted, a patient was given a monoclonal
for colorectal cancer. “Nearly a year later they were still
positive to human anti-animal antibody, so in theory they could
still be causing interference in the very assays that are probably
being used to monitor their clinical course.”
Most of the
published literature is on human anti-mouse antibodies, largely
because mouse monoclonal antibodies are widely used to create “magic
bullets” that can improve the sensitivity of the imaging or
increase the efficacy of a drug.
“A mouse
monoclonal antibody can specifically target a particular cell in
the body, or a particular organ, and it’s a way of bringing
either a drug or a contrast agent, an imager, initially in low concentration
on the antibody, into high concentration at that particular focal
point.”
“Many
of these have been approved for all sorts of cancers and lymphomas
and leukemias, and despite everything that’s been done to
try to humanize these antibodies to stop them from producing a response,
that response still exists in some patients.”
He cited one
case of a positive HAMA interference that was a transplant patient
who had been given OKT3. “You’d have thought that people
would have been tuned in to the fact that this would give a response
and upset assays. The patient had a positive hepatitis test, and
it was a problem for a while, but then we blocked it out with mouse
IgG, to show the result was not real. Then we backed that up by
sending out for an anti-OKT3 assay, which the company ran at the
time to prove there was actually that interference.”
“To be able to do that is a luxury that you don’t often have now,”
he said, noting that for this patient the problem was fairly short-lived
but for other patients it can go on much longer.
One test that is in the public view at the moment is hCG, Dr. Kricka said. The
New England Journal of Medicine and Clinical Chemistry
have reported on a group of patients who had initial high hCGs that
were not recognized as false-positives.
As a result, all of the patients had laparoscopies, some had D&Cs, two had
hysterectomies, and all had unwanted therapies. “It’s
cases like this that are starting to get the public a little more
informed and interested in this problem.”
Although we normally think of immunoassay interference as the analytes getting
stuck in the sandwich and giving a false-positive, or overwhelming
the capture antibody or the conjugates and giving a false-negative,
anti-animal antibodies can also interfere with other components
of an immunoassay kit, Dr. Kricka explained.
He cited one particular assay, where in the patient’s sample there was
a human anti-goat antibody and in the diluent for the conjugate
there was goat serum that reacted with the anti-goat antibody. “What
it did was clog up this conjugate so that the excess conjugate bound
in the sandwich was not washed out properly, and so it remained
and gave a false-positive.”
“So when you’re thinking of interferences, you really need to look
at all the components of an assay and where it’s derived from,
and ask the question, How would an anti-animal antibody interfere
in this assay if it reacted with the different components of the assay?”
“Obviously with recombinant proteins, a lot of these issues have gone away,”
he said, “but there still may be patients out there who’ve
been exposed to foreign proteins in the past and they still have
circulating antibodies.”
Cross-reactivity frequently occurs with anti-animal antibodies, he added. Another
study took a group of specimens in which researchers believed there
was human anti-mouse antibody because they could block out that
interference with mouse immunoglobulin. They then looked at what
would happen with other sorts of species.
“What they found,” he reported, “is that sheep also cross-reacted
and would block out the interference of 78 percent of the patients,
cow in 78 percent, guinea pig in nearly 70 percent, pig in 70 percent,
rabbit in 25 percent. So if a patient has been given a mouse monoclonal,
you’re not necessarily safe relying on the fact that the immunoassays
you’re going to use are using some other species as the source
of the antibodies. Cross-reactivity would make that a problem. Just
because you have a goat-based antibody assay and someone has a human
anti-mouse antibody doesn’t mean you’re out of the woods yet.”
Bacteria such as E.coli can be yet another source of interference in
immunoassays. “It’s not just the animals, it’s
the bugs as well,” he said.
With higher awareness, progress has been made in addressing anti-animal
antibody interference, he said. The Food and Drug Administration
in 1996 and again in 1997 made pronouncements on this subject, and
most of the packet inserts contain something about human anti-mouse
antibodies if the kit uses mouse monoclonals.
“So that’s a start, but unfortunately although the laboratory reads this, the
wider general public doesn’t see it and they’re perhaps
the people who should be interested. Luckily, the Internet is helping,
because several manufacturers have put their package inserts on
the Web where the public can see them if they want to.”
Testing assays for drug interferences is relatively straightforward, he noted.
“If this was a drug, then the source of the drug is available.
You can probably buy the metabolites, and you can find patients
taking the drug, so you immediately have three sources of materials
that would be very useful in assessing an interference.”
“When it comes to human anti-animal antibodies,” he said, “you
don’t have that luxury. What you really need is a nice set
of samples that represent the range of anti-animal antibody interferences.
Then perhaps another set to help you evaluate assays that you’re
either using or were developing to show that your optimization studies
to remove interferences have worked.”
What should these specimens be? Should you just look for human anti-mouse antibodies,
or should you have a full Noah’s ark of anti-animal specificities
available to test? From where are you going to get them, and how
can you guarantee a supply? All of these are difficult issues for
anyone wanting to assess the validity of tests or develop a test,
Dr. Kricka cautioned.
In many cases the interference can’t be investigated properly because the
specimen is no longer available. ”I think many manufacturers
encounter interferences that are reported by their customers, and
are faced with the problem that there are only 150 microliters left
of the interfering sample, and the patient won’t give any
more blood for this purpose.”
“Other patients who’ve been shown to have interferences are realizing
they’re sitting on a rather valuable resource,” he added.
“I know of at least one patient who’s charging over
$1,000 for units of blood because they had an interesting interferent
circulating in their blood. So getting some of these specimens that
would be very useful in assessing interferences may prove either
difficult, financially nonviable, or just impossible.”
As people are given new preparations, there are going to be new antibodies with
different specificities that will interfere with immunoassays, he
pointed out. “So this is a situation that never stays still,
as population changes and the protein-based drugs or imaging agents
they’re given will change the nature of the interferences
you might encounter in the sample.”
The case of Jennifer Rufer, which aired on “Prime Time” in 2001,
has brought greater public attention to the problem of interference.
Rufer was misdiagnosed with cancer because of incorrect laboratory
results. “She was one of those patients who unfortunately
went through to have treatment when she shouldn’t,”
Dr. Kricka said.
“But I really think ‘Prime Time’ did everyone a great favor,
because this put before the general public the nature of this problem.
So hopefully somewhere in this country there is some woman sitting
in a doctor’s office with an unexplained high hCG who will
say, ‘Well, when I watched Prime Time and saw Jennifer Rufer,
she said that, etc.’”
Interference has also increasingly spawned legal action. In a Washington State
case, the court ruled against the laboratory and manufacturer for
a patient who was misdiagnosed based on hCG results. “The
important message here is, when you look to see who’s paying
this money, it’s not just the manufacturer who’s found
to be liable for this. It’s also the laboratory.”
“As clinical chemists, as clinicians, we all now know about this problem,”
he noted. “No one could stand up in court and say I’ve
never heard of an interference in an immunoassay, I didn’t
know there were human anti-mouse antibodies. That is absolutely
unsupportable now.”
He is hopeful that some of the professional organizations will step up to the
plate and take stands on the issues. “The hCG case was of
sufficient notoriety that at least the American College of Obstetricians
and Gynecologists did make some recommendations, but this is only
in the special case of patients with hCG elevated for unknown reasons.”
ACOG’s answer was to perform a urine hCG. “Most of the interference
literature on anti-animal antibodies, if not all of it, is serum-
and blood-based. Antibodies are big. So unless you’ve got
renal disease, you’re not going to see them in the urine.
Therefore, there should be no interference from intact anti-animal
antibodies in urine, and I can’t think of any examples of
urine-based assays where there have been anti-animal antibodies
causing interference.”
With the variety of human anti-animal antibodies found, interference will persist
as a problem for the laboratory, Dr. Kricka predicts. “I think
we’re going to live with this for a long time yet, and probably
the only way out of this ultimately is either give up on antibodies,
or do what some people have suggested—that is move to other
species entirely that we’re not working with now.”
“All of those are major changes for the clinical laboratory,” he said,
“so in the meantime I think we have to be aware of these problems
and remain vigilant.”
Anne Paxton is a writer in Seattle.
|