Withering heightshigh factor levels a new thrombotic risk?
Blood group and thrombosis
For those who find coagulation testing a daunting and bewildering
process, Wayne L. Chandler, MD, offers a catchy algorithm.
"Test for everything no matter what," says Dr. Chandler, tongue
only partly in cheek. For with the recent publication of several
studies suggesting elevated levels of certain procoagulants may
be risk factors for thrombosis, his directive may gain added credence
in the not-too-distant future. After years of focusing on the risks
of coagulation factor deficiencies, researchers are exploring the
meaning of high factor levels. Their findings are fine-tuning current
understanding of the coagulation cascade and may well transform
the makeup of coag test panels.
"I will be very surprised if these procoagulants don’t weigh in
with the same importance that we’ve found with anticoagulants,"
says John D. Olson, MD, PhD, chair of the CAP Coagulation Resource
"This is potentially going to be an extremely interesting story,"
he continues. "But we’re only at the beginning of the story, and
we’re going to have to watch it unfold."
The tale begins with a handful of papers, published in
the last five years, that looked at elevated factor levels. It has
gained momentum only this year, however, with the appearance of
two papers in The New England Journal of Medicine that concluded
high levels of factor VIII and factor XI are risk factors for recurrent
venous thromboembolism and venous thrombosis, respectively.
"We’ve had suggestions of this all along," says Michael Laposata,
MD, PhD, director of clinical laboratories at Massachusetts General
Hospital, Boston, and a member of the resource committee. What makes
these two studies noteworthy is the care with which they have been
done, as well as their size and reliance on improved testing methods.
Most compelling has been the factor VIII study, which was published
in mid-August -(Kyrle PA, et al. 2000;343:457-462). The researchers
followed 360 patients-taken from the ongoing Austrian Study on Recurrent
Venous Thromboembolism-who had had a first episode of spontaneous
venous thromboembolism. Of those, 38 developed recurrent venous
thromboembolism-and all 38 had higher plasma levels of factor VIII
than those without a recurrence.
While earlier papers have implicated high levels of factor VIII
in the occurrence of deep vein thrombosis,* this one stands out
because it takes other risk factors into account. After adjusting
for age, sex, the presence or absence of factor V Leiden, the presence/absence
of the G20210A prothrombin gene mutation, and the duration of oral
anticoagulation, the researchers found that higher levels of factor
VIII remained an independent risk factor.
"The fact that they were able to put in all these other variables
gives the study credibility," says Dr. Laposata, adding that previous
studies may not have looked at other risk factors simply because
so many of them are new. Indeed, the researchers note in their discussion
that factor V Leiden and the G20210A mutation were unknown when
their study began.
When the Austrian researchers calculated the relative risk associated
with each of several different ranges of factor VIII levels, they
found the risk of recurrence was nearly seven times as great among
patients whose levels exceeded the 90th percentile, compared to
those whose levels placed them below the 25th percentile. Those
falling into the latter category had factor VIII levels of less
than 120 IU/dL, while those in the former category had levels greater
than 234 IU/dL. For patients with factor VIII levels above the 90th
percentile, the likelihood of recurrence at two years was 37 percent,
compared with a five percent likelihood among patients with lower
What Dr. Laposata finds particularly absorbing is the sharp surge
in relative risk that occurs at high levels. "They reached the conclusion
that the risk goes from basically nothing to a huge relative risk
of 11.4, at a threshold. So it’s a giant step upward."
Also worth noting is the number of patients who may be affected
by high levels of factor VIII. Among patients with venous thrombosis,
the prevalence of such an elevated plasma level is roughly 20 percent,
the researchers observe. Says Dr. Chandler, an associate professor
of laboratory medicine at the University of Washington, Seattle:
"It looks like it’s common enough that it’s interesting. It looks
like it could have enormous impact, roughly on the same order of
population risk as factor V Leiden, perhaps even more common, if
you believe the literature."
"In fact," he continues, "factor VIII looks like it’s probably
going to go."
While factor VIII may be nearing readiness for entry onto
a coag test panel, factor XI appears to be a few steps further away.
"But it’s probably coming up from behind," says Dr. Chandler, sounding
a bit like a racetrack announcer.
In a study published in early March (Meijers JCM, et al. N
Engl J Med. 2000;342:686-701), a team of Dutch researchers evaluated
whether high levels of the factor increased the risk of venous thrombosis,
looking at 474 patients with deep vein thrombosis and 474 people
the same age who did not have such clots. As with the recent factor
VIII study, the researchers adjusted for sex and age as well as
other risk factors, including oral contraceptive use, protein C
and protein S deficiencies, antithrombin deficiency, homocysteine,
fibrinogen, factor VIII, and the factor V Leiden and G20210A mutations.
High levels of factor XI are a risk factor for DVT, they concluded,
with the risk more than double for patients whose levels are in
the highest 10th percentile. With those high levels present in 10
percent of the population, they calculated that 11 percent of all
cases of thrombosis in the general population may be attributed
to high factor XI levels. The authors found also a continuous dose-response
relation between increasing factor XI levels and the risk of venous
thrombosis. They postulate that high levels of the factor cause
thrombosis through sustained generation of thrombin, leading to
the protection of fibrin from proteolysis.
The researchers defined a high level as a value above the 90th
percentile of the distributed values in the control group, which
equaled 120.8 percent; the 95th percentile was 130.2 percent. "The
numbers were distributed more or less Gaussian, and therefore much
higher levels than 120 were obtained," Joost C.M. Meijers, PhD,
lead author of the paper, told CAP TODAY. "It is hard to determine
risks at these high values, because the numbers become low, but
the analyses that we performed point to a ’higher is worse’ direction."
How convincing are the data from the factor XI study? The researchers,
from the Leiden Thrombophilia Study (LETS), "are some of the best
epidemiologists and researchers around," says Dr. Laposata. "These
guys are good. And they’ve got big numbers of patients and controls.
The way they look at the data, it’s pretty hard to argue with increased
The risk from high levels of factor XI is on a par with that from
factor VIII, says Eduardo C. Lau, PhD, senior research scientist
at Specialty Laboratories, Santa Monica, Calif. "Because the relative
risks are 2.8 for factor VIII and 2.2 for factor XI, they contribute
about equally to thrombosis risk."
At the same time, few studies looking at increased factor XI levels
precede this one, notes Dr. Chandler. "So it’s not yet on the everybody-believes-it
trail," he says. In fact, the LETS researchers report the role of
high factor XI levels in thrombosis had not been investigated prior
to their study.
A third factor, factor IX, is also inching onto the scene
as a possible risk factor for venous thrombosis. Another study by
the Leiden group, based on the LETS study, compared 426 patients
with a first episode of DVT with 473 population controls. The data
were adjusted for age, sex, oral contraceptive use, and high levels
of factors VIII and XI and the vitamin K-dependent clotting factors;
also taken into account were several thrombophilic risk factors,
including protein C and protein S deficiencies, antithrombin deficiency,
the factor V Leiden mutation, and the prothrombin 20210A allele.
Individuals with high levels of factor IX have a more than twofold
risk of developing a first DVT compared to those with lower factor
IX levels, the researchers reported in their study, which was published
this year in Blood (van Hylckama Vlieg A, et al. 2000;95:3678-3682).
These high levels are present in 10 percent of the population, the
A high level was defined as greater than 129 U/dL. In the group
of patients whose levels surpassed that mark, the factor IX levels
ranged between 130 to 209 U/dL, lead author Astrid van Hylckama
Vlieg, MSc, told CAP TODAY. "For the controls this range is 130
to 188. From the dose-response relation it can be seen that increasing
levels of factor IX are associated with an increased risk of thrombosis."
Based on this study as well as the earlier factor VIII Lancet
study (footnote, page 26), the researchers conclude the risk of
DVT is highest when both VIII and IX are elevated.
If factor XI has had only scant papers devoted to looking at elevated
levels, elevated factor IX has barely earned a nod from researchers.
"I did an extensive Medline search, and this is the first paper
I’ve seen on the subject," says Dr. Lau. The authors of the Blood
study note their interest was piqued not by other factor IX studies,
but by their earlier research on elevated factors VIII, X, and XI.
"Because factor VIII is the cofactor of factor IXa in the activation
of factor X, it seemed plausible that elevated levels of factor
IX could also be a risk factor for DVT," they wrote.
Given the lack of supporting studies, it’s not surprising that
factor IX remains a long shot for earning a berth on coag test panels,
at least for the time being. Asked whether she would measure factor
IX as part of a hypercoag panel to assess thrombotic risk, van Hylckama
Vlieg replies, "No, it is not useful before we know what causes
factor IX levels to be high in some people but low in others. Also,
confirmatory studies are needed. The therapeutic consequences of
measuring levels of factor IX are unclear."
The same can be said for factors VIII and XI as well,
although their testing potential has created far more buzz.
"The frequency of these elevations in factors VIII and XI is such
that maybe we should have been looking at them as a first consideration,"
says Dr. Laposata. (LETS study investigator and clinical epidemiology
professor F.R. Ros-en-daal, MD, PhD, of the Hemostasis and Thrombosis
Research Center, Leiden University Medical Center, told CAP TODAY,
"This is equally true for factor IX.")
In fact, Dr. Laposata adds, if elevated levels of factors VIII
and XI do account for the high number of thrombosis cases indicated
by the researchers, "We may end up doing a whole lot less of our
other tests. Because between Leiden and prothrombin mutation and
elevated VIII and XI, it could be that we’ve pretty much done it."
If, if, if-the scourge of any provocative piece of medical research.
Even Dr. Laposata, for all his enthusiasm, is quick to concede that
persuasive studies do not an accepted test make.
What hurdles will these factors need to clear before they find
themselves on a test panel alongside factor V Leiden, prothrombin
20210, and other risk factors? For starters, consider these:
Many find factor VIII vexing because it is an acute phase reactant.
"That clouds everything," says Alice Runge, MT(ASCP), technical
director of the coagulation department at Specialty Laboratories.
In their study, the Austrian researchers found no correlation
between factor VIII levels and levels of C-reactive protein, a sensitive
indicator of inflammatory response. Dr. Laposata cautions, "It’s
probably dangerous to conclude at this point that if you have a
normal C-reactive protein, that a single elevated factor VIII does
indeed reflect a true baseline. I don’t know that the kinetics of
the rise and fall in CRP precisely matches that for factor VIII."
Given that the two may fluctuate at different rates, he says repeat
testing for high levels of factor VIII are in order.
Factor XI is not an acute phase reactant, though Dr. Meijers reports
he and his colleagues did correct for CRP levels, which "did not
reduce the risk of venous thrombosis by high levels of factor XI."
Likewise, factor IX is not known to be an acute phase reactant.
Van Hylckama Vlieg and her colleagues also report that adjustment
for CRP did not affect the results of their study. "It does not
make sense to measure [factor IX] in conjunction with CRP," she
says, adding, "The absence of post-hoc phenomena is a key assumption
in the interpretation of case-control studies. Therefore, we examined
the time between the deep vein thrombosis and the venipuncture,
which ranged from six to 56 months, and we found no influence on
the factor IX levels."
Inflammation does seem to be less of an issue with factor XI,
Dr. Laposata agrees. "But still, if I saw an elevated value for
XI, I would probably repeat it to make sure it was truly up there."
"You have to be very careful not to rely on a single measurement
for any of these factors," agrees Dr. Olson, who will chair a CAP-sponsored
conference on diagnostic issues in thrombophilia-including increased
procoagulant activity-in November 2001.
Because standard coagulation assays were designed to look at deficiencies,
current testing methods may not precisely measure elevated levels.
"It’s a real challenge to take an interesting and important epidemiological
finding and convert it into something you can do in the clinical
laboratory and provide a single person with a useful, reproducible
result," Dr. Chandler says.
Even factor VIII, which boasts a fairly robust assay and clearly
has a level at which risk increases significantly, may not be a
straightforward measurement. Most research studies conducted at
a single institution use a single instrument-reagent combination
for their measurements, he notes. "Before a test is recommended
for widespread use in a new capacity, like factor VIII activity
for predicting risk of thrombosis, care needs to be taken to ensure
that the many different instrument-reagent combinations available
give the same results," Dr. Chandler says. This is particularly
true when a threshold value is used for diagnosis, as suggested
for factor VIII levels. "CAP Surveys show that a relatively wide
range of factor VIII activity levels are currently reported even
on normal samples," he says.
Will anyone pay for these tests?
"If we ordered an VIII and put an ICD-9 code down for thrombosis,
we’d get a call from someone saying, ’Factor VIII levels are associated
with bleeding; we don’t pay you for doing VIIIs on people with high
levels,’" Dr. Chandler says. "So this not only has to penetrate
to the docs, but to Medicare and insurance companies."
That’s no small task, given that most thrombosis test panels are
large to begin with and can easily cost in the high hundreds of
dollars. Tacking on another couple hundred dollars by testing for
high levels of factors VIII and XI is sure to give payers and hospitals
In his role as director of clinical laboratories for a county
health system in Texas, Dr. Olson works with a patient population
largely supported by tax dollars-which "are not sufficient to pay
for all of the bills," he says. Adding on new tests without a clear
indication of their benefit doesn’t make sense, he says, at least
not in his setting. "For routine practice, it’s probably early to
be making these measurements," he says, though he hasn’t ruled it
It’s far from clear how useful these measurements may be. With
few exceptions-such as guiding perioperative anticoagulation- current
testing for thrombophilia provides interesting diagnostic information
but little therapeutic direction, Dr. Olson says. "When we start
talking about these procoagulants, we’re in our infancy in terms
of knowing how this information is helpful to us."
Another unknown: whether certain populations may be more at risk
than others. Factor V Leiden has emerged as a risk primarily in
European populations; in Asian populations, on the other hand, the
incidence is extremely low. Likewise, "Every ethnic group might
be different for the procoagulants," says Dr. Lau.
As the list of apparent thrombotic risk factors gets longer, deciding
what to test for, and in what order, becomes a mind-boggling burden.
Dr. Laposata is in serious discussion with his colleagues about
whether to add factors VIII and XI to his lab’s hypercoag panel,
a move that would expand what he calls "an already gargantuan menu."
That panel already includes a clot-based screening test for activated
protein C resistance, followed by a PCR test if it suggests the
presence of a factor V Leiden mutation; a PCR test for the prothrombin
20210 mutation; tests for protein C, protein S, and antithrombin
deficiencies; homocysteine; lupus anticoagulants; and anticardiolipin
Dr. Chandler recalls that several hospitals once asked him to
work up patients according to a rough probability scale, starting
with factor V Leiden. The approach seemed sound until it became
apparent that patients with two risk factors were at much higher
risk. "So the algorithm became, We’ll test for factor V Leiden,
and if that’s positive, then we’ll check everything else. And if
it’s negative, we’ll check everything else.
"Of course, everyone kind of looked at me like I was crazy, like
I was Bugs Bunny and had worked out this nutty scheme," he says.
If, as Dr. Laposata suggests, factors VIII and XI, in conjunction
with prothrombin and factor V Leiden, do account for the majority
of thrombotic events, will there still be a role for the other tests?
"Who knows?" says Dr. Chandler.
Nor is the end anywhere in sight. Given the tilt in interest
toward high levels of factors VIII, IX, and XI, other coag factors
will surely be evaluated for possible links to venous as well as
"I’ll be surprised if we don’t find others," says Dr. Olson.
"Now it looks like almost all the coag factors are going to have
upper limits of normal," Dr. Chandler predicts.
Factor VII has already been the subject of numerous papers linking
high levels of the factor to risk of arterial thrombosis. Yet this
factor, which found itself in the limelight this fall thanks to
a study hinting that certain factor VII genotypes may protect against
myocardial infarction, * remains somewhat suspect. "Factor VII is
clustered with elevated triglycerides and insulin resistance and
elevated PAI-1," says Dr. Chandler. "So VII may not be the best
predictor; it’s not clear whether it’s the driving factor or if
it’s just along for the ride." The same may be true for fibrinogen,
which also appears to cluster with other risk factors.
Also of interest are factor V (high levels of which have been
linked to arterial thrombosis) and factor XII, though the number
of papers devoted to looking at elevated levels of either factor
can be counted on one hand-with fingers to spare. Reports indicate
high levels of factor II may also be detrimental in terms of arterial
as well as venous thrombosis, Dr. Chandler says.
Why a factor may be linked to either or both types of thrombosis
remains a mystery for now. "We’re still at the epidemiological association
level," says Dr. Chandler. "The pathophysiology level is pretty
simplistic: Too much is bad. But to move beyond that, we’ll have
to get to the gene level."
Researchers are already heading in that direction. Dr. Meijers
says he and his colleagues hypothesize one particular genetic mutation
may be responsible for elevated levels of factor XI. As for factor
IX, van Hylckama Vlieg says it’s unclear whether a genetic mutation
is associated with high levels of the factor, but adds, "We are
currently trying to find a possible genetic basis."
Ultimately, says Dr. Laposata, laboratories will likely turn to
genetic testing to identify mutations associated with high factor
levels. In the interim, laboratorians and their clinical colleagues
will have to pick their way through the emerging researcha
painstaking task he insists cannot be overlooked.
"Thrombosis is such a prominent disease, and thrombosis testing
is one area that you don’t want to be behind the curve on," he says.
Karen Titus is CAP TODAY contributing editor and co-managing
* Koster T, et al. Role of clotting factor VIII in effect of von Willebrand
factor on occurrence of deep-vein thrombosis. Lancet. 1995;345:152-155.
O’Donnell J, et al. High prevalence of elevated factor VIII levels
in patients referred for thrombophilia screening: role of increased
synthesis and relationship to the acute phase reaction. Thromb
Kraaijenhagen RA, et al. High plasma concentration of factor VIIIc
is a major risk factor for venous thromboembolism. Thromb Haemost.
* Girelli D, et al. Polymorphisms in the factor VII gene and the
risk of myocardial infarction in patients with coronary artery disease.
N Engl J Med. 2000;343:774-780.