William Check, PhD
Although the course on testing for coagulation disorders at CAP ’08 was called “Controversial Topics,” it might be more accurate to say that the application of the tests discussed can be confounding. Laboratory testing for inherited thrombotic conditions is a good example. “Over the past 10 to 20 years several new etiologies for inherited thrombosis have been identified,” says George M. Rodgers, MD, PhD, professor of medicine (hematology) and pathology at the University of Utah Health Sciences Center and medical director of the coagulation laboratory at ARUP Laboratories, who spoke on this topic in the CAP ’08 course. “Tests have been developed for these conditions on the assumption that being able to identify patients with risk factors would lead to changes in patient management.” What’s controversial, Dr. Rodgers says, is that “most clinical data suggest that is not the case—making a diagnosis of an inherited disorder in a patient with a blood clot probably does not change long-term management.”
For this reason, Dr. Rodgers says, genetic tests are not recommended for patients with clotting disorders. They may have some utility in screening to identify family members at high risk, particularly affected women. “But in terms of managing a patient with a clot, there are essentially no data to support the use of genetic thrombophilia testing on a wide-scale basis,” Dr. Rodgers says.
Testing for antiphospholipid syndrome, or APS, which is one of the most common causes of acquired thrombophilia, can also be confounding. Kristi J. Smock, MD, who spoke on this topic, explained to CAP TODAY that, within APS, there are two subgroups. Some APS patients have lupus anticoagulant, or LAC, antibodies that interfere with phospholipid-dependent coagulation reactions and prolong phospholipid-dependent clotting assayed in the liquid phase. Others have anticardiolipin antibodies, or ACA, which are detected by solid-phase assays. However, these categories are not mutually exclusive. “There is definite overlap, although its frequency is unknown,” says Dr. Smock, assistant medical director of the hemostasis and thrombosis laboratory at ARUP Laboratories and assistant professor of pathology at the University of Utah School of Medicine. “In general, these patients have antibodies that either prolong clotting in the liquid phase test or can be identified by ELISA, or both.”
When considering antiphospholipid syndrome, Dr. Smock says, testing for three entities is recommended—lupus anticoagulant, anticardiolipin antibody, and antibody against β2-glycoprotein I (β2-GPI), all of which correlate with increased risk of a thrombotic event. Dr. Smock focused on lupus anticoagulant testing strategies. Her reason: “It is more difficult to test for LAC than to use an ELISA. Interpretation and algorithms are a bit more complex.”
In the CAP ’08 course, Dr. Smock discussed another problematic condition, aspirin resistance, for which a wide range of prevalence figures have been published. “I think it is a problem of using different definitions of aspirin resistance and measuring it with different tests that have different sensitivities and specificities,” she says. Moreover, she adds, testing for this condition is not generally recommended because it is not known what the treatment changes would be.
One might even wonder whether aspirin resistance actually exists. This entity was postulated on the basis of people having occlusive coronary events while taking cardioprotective doses of aspirin. However, Dr. Smock points out, aspirin lessens cardiovascular risk by only 20 to 25 percent. And cardiovascular disease is multifactorial. “It is simplistic to attribute coronary events entirely to aspirin resistance,” she says. The condition may exist but is probably rare, and prospective clinical studies are needed to document it. “True biochemical aspirin resistance may reflect a variant cyclooxygenase-1 [COX-1] enzyme that is not susceptible to inhibition by aspirin,” Dr. Smock says.
Fortunately, not all coagulation disorders are confusing. Dr. Rodgers notes that thrombotic thrombocytopenic purpura, or TTP, has become clearer in the past 10 years with the discovery that it is characterized by a lack of cleavage of von Willebrand factor due to a defective metalloproteinase. The resulting ultralarge vWF molecules induce platelet aggregation, leading to thrombotic microangiopathy.
Dr. Rodgers notes that TTP, a rare disorder, is frequently misdiagnosed or confused with a related disorder, hemolytic-uremic syndrome, or HUS. “Until very recently, there hasn’t been a reliable test to distinguish them,” he says. As a result, some patients probably received empiric therapy for TTP—plasmapheresis—when they did not benefit from it. “Relatively new tests allow the laboratory to accurately diagnose TTP and distinguish it from HUS to allow appropriate therapy,” Dr. Rodgers says.
Several pathologies can underlie inherited thrombotic disorders: factor V Leiden (FVL), elevated factor VIII, homocysteinemia, prothrombin mutation, and deficiencies of protein C, protein S, or antithrombin. Only for prothrombin mutation is the DNA test best, Dr. Rodgers says. For the others, molecular tests are more expensive (three times as costly for FVL) and functional assays are preferred. Testing for the FVL variant in asymptomatic carriers (first-degree relatives of symptomatic probands) is not justified because the absolute annual incidence of spontaneous venous thromboembolism is low (Middeldorp S, et al. Ann Intern Med. 2001; 135: 322– 327). A CAP consensus panel recommended against testing for the altered form of the MTHFR gene, one of several possible etiologies for homocysteinemia (Olson JD. Arch Pathol Lab Med. 2002; 126: 1277– 1279).
Functional assays carry their own caveats. Routine testing for elevated factor VIII levels is controversial. If done, the test should be performed at least six months after a thrombotic event and after anticoagulant therapy is completed. When assaying for antithrombin deficiency, it is critical to remember that heparin therapy can dramatically decrease antithrombin levels; warfarin therapy may increase antithrombin levels.
Several studies support the notion that routine genetic testing or screening for thrombophilia is not beneficial. In one, deep venous thrombosis patients with a prothrombin mutation had a similar risk of venous thromboembolism recurrence at two and four years as those without a mutation (De Stefano V, et al. Brit J Haematol. 2001; 113: 630– 635). The authors suggested treating the two groups for a similar length of time. Another research team concluded that “[P]atients with factor V Leiden or the G20210A prothrombin polymorphism were not at substantially increased risk of recurrent events as compared to patients without these disorders. Moreover, the relative benefit of low-intensity warfarin therapy in preventing recurrent events was not significantly affected by the patient’s genetic status” (Ridker PM, et al. New Engl J Med. 2003; 348: 1425– 1434).
Two prospective studies of VTE/ DVT patients found that testing for congenital thrombophilia—whether by genetic or laboratory risk factors—did not predict recurrent venous thrombosis (Baglin T, et al. Lancet. 2003; 362: 523– 526; Christiansen SC, et al. JAMA. 2005; 293: 2352– 2361). “Clinical factors are more important than laboratory abnormalities in determining duration of anticoagulant therapy,” the authors of the second study said. Reviewing all extant data, Baglin concluded, “Now that high quality clinical outcome studies are being reported it is becoming apparent that despite association, testing [for heritable thrombophilic defects] has limited predictive value for the majority of unselected symptomatic patients. ... [I]n most cases decisions regarding intensity and duration of anticoagulant therapy can be made purely in relation to clinical criteria” (Pathophysiol Haemost Thromb. 2003; 33: 401– 404).
Dr. Rodgers and his colleagues recently performed a retrospective analysis of laboratory orders and results from September 2005 to August 2006 to gain insight into physicians’ ordering practices, comparing them against the ordering practices recommended by a 2002 CAP consensus conference on thrombophilia testing. The results showed that as recently as 2006, many clinicians were unaware of the limitations of thrombophilia testing (Jackson BR, et al. BMC Clin Pathol. 2008; 8:3). In general, second-line (antigen) tests were ordered almost as often as first–line (functional) assays. With regard to testing for proteins C or S, 20 to 50 percent of samples with low levels also had an INR greater than 1.3. Dr. Rodgers’ interpretation: “Probably half or more of these samples were taken from patients on warfarin, so they were falsely positive. A lot of money is being wasted, and a lot of misleading test results are being recorded.” Also, patient safety is at risk. “Patients may be wrongly labeled as having a genetic disease but it is actually a laboratory mistake,” Dr. Rodgers says. He recommends that patients be encouraged not to undergo routine thrombophilia testing unless they have female siblings or children (or both) at risk and for whom a thrombophilia diagnosis will change management.
While genetic tests are not helpful, the D-dimer assay “shows promise” in predicting recurrent events, Dr. Rodgers says. One month after finishing a course of anticoagulation, deep venous thrombosis patients with an abnormal D-dimer test result were randomized to either no further therapy or continued anticoagulation. Untreated patients had a 2.5-fold increased risk of recurrent clot (Palareti G, et al. N Engl J Med. 2006; 355: 1780– 1789).
While testing for inherited causes of thrombophilia has limited value for managing the thrombosis patient, it may be useful for family screening. “Thrombophilia testing can identify young female family members who might be at risk of using hormones or getting pregnant and for whom a diagnosis might change how they would be treated,” Dr. Rodgers says. “But this is a very unusual circumstance.”
Addressing the fact that some physicians favor genetic thrombophilia testing, Dr. Rodgers says he is in an “unusual” position. “I’m a pathologist who runs a coagulation lab as well as a hematologist who treats patients. There is a substantial literature now, primarily from Europe, supporting the idea that knowing whether a patient has inherited thrombophilia does not change intensity or duration of treatment.” The pathologists who have contrary points of view, he says, may see this from the perspective of the tests being available and thus they should be done. “They may not be as familiar with the clinical literature,” he says.
Testing is useful in thrombotic thrombocytopenic purpura. About 10 years ago the cause of this condition was identified as a deficient metalloproteinase. “It has taken since then to develop accurate, user-friendly tests to quantitate the enzyme,” Dr. Rodgers told CAP TODAY.
Though there are only about 1,000 new cases of TTP each year, it is important to be alert for it. “If it is not diagnosed and treated correctly, most patients will die,” Dr. Rodgers says. “It is really a fulminant disease.” Now that its basis is known and there is a rapid assay and good treatment, “the lethality of this disease should improve,” he says.
TTP must be considered in all patients with anemia and thrombocytopenia not due to another obvious disorder. Only 40 percent of TTP patients have all five characteristic symptoms, which also include fever, renal disease, and neurologic defects. “Most patients only have three features—anemia, thrombocytopenia, and neurologic defects—and those can fluctuate,” Dr. Rodgers says. Coagulation studies are usually normal. In the end, TTP is a clinical diagnosis requiring a high index of suspicion that is helped by newer tests.
Two groups of researchers published evidence in 1998 that TTP results from lack of cleavage of vWF by a metalloproteinase, with the result that ultralarge vWF accumulates, leading to platelet aggregation and disseminated thrombi (for a review, see: Lammle B, et al. J Thromb Haemost. 2005; 3: 1663– 1675). Most, but not all, patients diagnosed with TTP have severe metalloproteinase deficiency. Plasma levels of the protease, since named ADAMTS13, are very low during the acute episode but return to normal upon recovery. In adults with a single episode or intermittent TTP, the etiology is an IgG autoantibody against ADAMTS13. Chronic relapsing TTP in children, on the other hand, is due to a congenital ADAMTS13 deficiency. “The cornerstone of treating TTP is plasma therapy,” Dr. Rodgers says. “This empiric therapy is now understood to be effective because infusion of normal plasma provides adequate amounts of vWF metalloproteinase.” It also removes the antibody that inhibits the metalloproteinase.
Several rapid commercial kits for measuring ADAMTS13 activity are now available. None are FDA approved. “Many chronically ill patients also have low levels of activity,” Dr. Rodgers says. “But TTP is usually associated with levels less than five percent.” The absence of ADAMTS13 activity is 89 percent sensitive and 100 percent specific for TTP (Groot E, et al. J Thromb Haemost. 2006;4:698–699). Assays for ADAMTS13 help to distinguish TTP from hemolytic-uremic syndrome, which is usually due to enterohemorrhagic E. coli infection.
An in-house comparison at ARUP Laboratories showed that a GTI kit identified four of six abnormal samples, while a Technoclone kit identified all six. (Gold standard was the result from a reference laboratory, BloodCenter of Wisconsin.) “We are using the Technoclone kit,” Dr. Rodgers says. They did not evaluate an American Diagnostica kit because it was unavailable at the time. Dr. Rodgers notes that the kits require a fluorometer but are not difficult to do. “Hospitals that frequently see patients who could have TTP or HUS would probably find this assay to be clinically useful,” he says.
In her presentation on antiphospholipid syndrome, Dr. Smock said this diagnosis is made when a patient has a recent history of thromboembolic disorders or spontaneous abortion and a persistently positive lupus anticoagulant or anticardiolipin antibody test. “Lupus anticoagulant” is a misnomer, since the majority of patients with laboratory findings of LAC do not have lupus and the condition results in hypercoagulability rather than bleeding. About one-third of strokes in patients younger than 50 years have been attributed to antiphospholipid antibodies. “As we learn more about antiphospholipid syndrome, more overlap [between ACA and LAC] becomes apparent,” Dr. Smock says. Many antibodies don’t actually target phospholipids but rather protein cofactors that are bound to them. That’s why testing for APS includes testing for antibody to β2-GPI. Retrospective studies show the clinical utility of testing for all three of these molecules (De Laat B, et al. Blood. 2005; 105: 1540–1545; Harris EN, et al. Lancet. 1983; 2: 1211–1214).
Dr. Smock’s laboratory uses a lupus-sensitive anticoagulant reflexive panel that satisfies the four main criteria recommended by an international consensus panel for LAC testing (Brandt JT, et al. Thromb Haemost. 1995;74:1185–1190, 1597–1603). First, in the screening step, patient plasma must prolong a phospholipid-dependent clotting assay. “To call a sample negative, it must be negative in two different screening methods,” Dr. Smock says, “and both should use lupus-sensitive reagents.” Her laboratory does lupus-sensitive APTT and the dilute Russell viper venom test.
In the second step, a mixing study, the presence of an inhibitor is demonstrated if the clotting abnormality does not correct when the patient sample is mixed with normal pooled plasma. Third, a confirmatory test provides evidence of phospholipid dependence. In the platelet neutralization procedure, for example, adding phospholipids from freeze-thawed platelets results in significant shortening of the prolonged baseline clotting value. Finally, specific inhibition of any one coagulation factor or the presence of inhibitor drugs must be ruled out. For instance, the presence of heparin can be ruled out by addition of Hepzyme. Testing for a specific inhibitor is typically done if you are unable to demonstrate phospholipid dependence, according to Dr. Smock. In this step the clinical history is “crucial,” she says. Patients with lupus anticoagulant clot, while those with inhibitors tend to bleed.
Screening and confirmation are combined in a single procedure in the newer integrated assays. Reagents in two tubes contain a low and a higher phospholipid concentration. In the Staclot LA procedure, a modified APTT assay, one tube contains hexagonal-phase phospholipid, which neutralizes LAC, shortening the clotting time. The assay system also contains a heparin neutralizer, which enhances specificity.
One group of investigators evaluated an integrated APTT test, a dilute Russell viper venom time, and Staclot LA. Of the 105 patients, 26 were taking anticoagulants at the time of testing. Staclot LA was the most sensitive test, picking up 28 of 30 patients positive for LAC (Zhang L, et al. Am J Clin Pathol. 2005; 124: 894– 901). There was no difference in performance of the integrated tests between samples from patients taking and not taking anticoagulants. “Newer integrated testing kits may be among the most sensitive and specific assays,” Dr. Smock says.
To explore the issue of false-positive results, physicians at an Italian reference laboratory retested samples from 302 patients who had tested LAC-positive at one of 29 specialized clinics. Retesting excluded the diagnosis in one-fourth of the patients (Pengo V, et al. J Thromb Haemost. 2007;5:925–930). False-positive results are a serious issue, Dr. Smock emphasizes. “People should be aware that the recommendations state that a test should be persistently positive,” she says. In the case of an initial positive test for LAC or ACA, that patient should be retested 10 to 12 weeks later. “The clinical consequences of a false-positive test could be fairly drastic,” she cautions. A patient who had a thrombotic event and tests positive for LAC may be treated with long-term anticoagulation. “Because of the risks of anticoagulation, we need to document that these patients are truly, persistently positive,” she says. “Antibodies can wax and wane and may be associated with infections or medication.” Children, in particular, can have transient LAC antibodies after tonsillitis, for example.
Dr. Rodgers considers testing for APS, a major cause of acquired thrombophilia, to present “a clear contrast” to testing for inherited thrombophilia. “Unlike genetic thrombophilia tests, which do not change patient management, antiphospholipid antibody tests are very useful and change how patients with thrombotic events are managed,” he says.
Testing for APS also presents a major contrast to testing for aspirin resistance, in which results are difficult to translate into clinical management. Dr. Smock notes that, according to a strict definition, aspirin resistance would exist when aspirin does not inhibit its target enzyme, COX-1, or block COX-1-catalyzed production of thromboxane A2 (TxA2). Since TxA2 promotes platelet aggregation and a higher risk of atherothrombotic events, aspirin resistance may also be defined as “clinical failure” while on aspirin or by failure of aspirin to inhibit platelet function in laboratory assays.
Since both clinical and laboratory definitions have been used in prevalence studies, it is not surprising that a wide range of prevalence figures have been published for aspirin resistance. Even prevalence studies using laboratory methods show a wide range, since many different laboratory tests are available and testing is not standardized. In one recent summary of 11 studies using different functional assays, the prevalence of aspirin resistance ranged from 5.5 to 61 percent (Campbell CL, Steinhubl SR. J Thromb Haemost. 2005; 3: 665– 669). Dr. Smock says this finding implies that different assays vary greatly in their ability to detect the pharmacological effect of aspirin.
A Canadian group directly compared the results from six major platelet function tests in assessing the prevalence of aspirin resistance in 201 patients with stable coronary artery disease taking daily aspirin (greater than 80 mg). Prevalence ranged from 2.8 to 59.5 percent (Lordkipanidzé M, et al. Eur Heart J. 2007;28:1702–1708). Correlation among various tests was poor, as was intra-test correlation, the authors reported. Tests with a greater specificity showed a low prevalence of aspirin resistance [arachidonic acid-induced platelet aggregation, four percent; VerifyNow Aspirin, 6.7 percent; urinary dehydrothromboxane B2 (AspirinWorks), 22.9 percent]. Nonspecific tests showed greater than 50 percent prevalence. “The clinical usefulness of the different assays to classify correctly patients as aspirin resistant remains undetermined,” these investigators concluded. Dr. Smock’s conclusion: “True aspirin resistance is probably rare.”
Poor correlation has also been found between two point-of-care tests and between the POC tests and a laboratory test (Harrison P, et al. Stroke. 2005;36:1001–1005).
Dr. Smock lists a number of considerations for a laboratory thinking about offering a test for aspirin resistance. “You need to communicate with the people who will be ordering the tests so they are clear what the tests are picking up and their implications,” she says. An abnormal result may not indicate resistance but may indicate increased risk. Also, before the patient begins aspirin therapy, a baseline value should be obtained.
“It is important to think about noncompliance,” Dr. Smock continues. “Some folks identified as having aspirin resistance were just not taking their medication. That can’t be found by a lab test. Clinicians should ask about this possibility.” In addition, the clinician should ask whether the patient was taking an NSAID at the time of the test. NSAIDs also target COX-1 and have a longer half-life than aspirin.
Dr. Smock has considered offering a test for resistance, but her laboratory is not doing one now. “We do not know how to manage patients who are deemed at high risk by laboratory testing,” she says. Some patients may respond to higher doses of aspirin, or they might be switched to other antiplatelet drugs. But there is no good evidence of the utility of either of these actions.
For all of these reasons, resistance testing is not recommended now as part of the routine management of patients receiving aspirin therapy. “Many more studies using specific tests to evaluate clinical utility and cost-effectiveness are needed,” Dr. Smock says.
One therapy for which monitoring is required is anticoagulation with intravenous heparin. The APTT has been the standard assay for this purpose. Now this situation is changing, says Christopher M. Lehman, MD, associate professor of pathology and co-director of clinical laboratories, University of Utah Hospitals and Clinics. One change has been the introduction and widespread adoption of new anticoagulants with predictable, weight-based dosing regimens that do not require routine monitoring but are not readily reversible. Low-molecular-weight heparins such as enoxaparin, lovenox, and fondaparinux fall into this category.
Ultimately, heparin anticoagulation “may be limited to clinical environments in which rapid reversal of anticoagulant effect is required, e.g., the medical intensive care unit (MICU),” Dr. Lehman and colleagues wrote (Lehman CM, et al. Am J Clin Pathol. 2006; 126: 416– 421). Readily reversible anticoagulation is desirable in MICU patients, Dr. Lehman explains, because they are more susceptible to pathologic bleeding secondary to heparinization that could destabilize their critical condition and cause morbidity, death, or both. “The need to rapidly reverse the anticoagulant effect of heparin in other patient populations receiving heparin is generally less acute,” he told CAP TODAY.
A second change is that, in settings in which heparin monitoring is needed, the APTT may ultimately be replaced by automated assays for anti-activated factor X (anti-Xa), which measure heparin levels more directly. Unfractionated heparin dose appears to be more important than the APTT in predicting clinical efficacy (Eikelboom JW, Hirsch J. Thromb Haemost. 2006;96: 547– 552). Dr. Lehman notes several practical advantages of switching to anti-Xa assays. For one, he says, “Laboratories would no longer need to conduct studies comparing anti-Xa levels with PTT results in patients receiving IV unfractionated heparin therapy, provided the laboratory had notified clinicians that the anti-Xa test and not the PTT must be used to monitor these patients. These studies are increasingly difficult to conduct since unfractionated heparin is not used as commonly as it was in the past and finding patients receiving IV unfractionated heparin therapy is more difficult.”
The technical advantages are that anti-Xa assays are less susceptible to preanalytical interferences (for example, partially filled tubes) and to interference from acute-phase reactants and that they are not susceptible to factor deficiencies, with the exception of severe antithrombin deficiency.
Dr. Lehman notes that a few published studies have suggested that monitoring unfractionated heparin therapy using anti-Xa levels may be more efficient in terms of bringing a patient into the therapeutic range with fewer adjustments of heparin dose (Rosborough TK. Pharmacotherapy. 1999;19:760–766; Rosborough TK, Shepherd MF. Pharmacotherapy. 2004;24:713–719). In the case of heparin-resistant patients, monitoring with an anti-Xa assay allowed use of lower doses of unfractionated heparin. “However,” Dr. Lehman notes, “none of these studies were designed to detect a clinically significant difference in rates of complications between PTT and anti-Xa management.”
To further investigate the feasibility of routinely using anti-Xa assays, Dr. Lehman and his colleagues compared the performance of three anti-Xa assays in MICU patients receiving intravenous unfractionated heparin. Because critically ill patients can acquire an antithrombin deficiency, one of the assays they tested was supplemented with antithrombin. All three assays worked equally satisfactorily. They agreed 80 to 85 percent of the time in classifying patients according to the American College of Chest Physicians recommendations for anti-Xa heparin levels. “We did not find that antithrombin supplementation provided a clinically significant benefit,” Dr. Lehman told CAP TODAY. His laboratory hasn’t yet adopted the anti-Xa assay.
In summary, testing is not helpful in aspirin resistance. In inherited thrombophilia, genetic tests are not useful but functional assays can help. Testing has value in APS and in TTP, offering a way to identify a specific etiology in the latter. And a newer type of test, anti-Xa assays, may surpass the current gold standard, APTT, for monitoring patients receiving intravenous unfractionated heparin.
William Check is a medical writer in Wilmette, Ill.