College of American Pathologists
Printable Version

  Coag collision course—cancer and thrombosis


CAP Today




November 2008
Feature Story

Karen Titus

Everyone knows malignancy and thrombosis cross paths. So, what else you got?

A lot and a little. Physicians are growing more savvy about this relationship—turns out the two are close cousins, if not siblings. Researchers are also digging into the underlying mechanisms and drug interactions that help explain this kinship. But like certain awkward family relationships, no one quite knows what to do once everyone arrives at the gathering.

In the future, thrombotic events in patients with malignancy may become less of a concern, as patients are treated with drug regimens that carry less thrombotic risk. That could one day close the door on certain lab tests. For now, though, laboratories will have an ongoing role in monitoring low-molecular-weight heparin and measuring D-dimer levels in certain patients; indeed, it may turn out that an even wider group of patients may benefit from D-dimer testing.

Two recent papers and two relatively new guidelines are evidence of the ongoing interest in teasing out the whys and hows of malignancy and thrombosis.

A comprehensive review article published in Clinical Lymphoma & Myeloma late this summer (Adcock DM, et al. 2008;8:230–236) explores the mechanisms by which malignancy promotes thrombosis. “Pathologists need to understand this relationship,” says lead author Dorothy Adcock, MD, medical/laboratory director, Esoterix Coagulation, Englewood, Colo.

The cancer/thrombosis link represents initiation of a vicious cycle, she says, with each part feeding off one another. Tumors promote clot formation, and activation of coagulation promotes tumor growth and dissemination. “It’s symbiotic—which is fascinating,” Dr. Adcock says. When cancer and thrombosis present simultaneously, she says, up to 40 percent of patients already have evidence of distant metastases.

Delving a little deeper, Dr. Adcock points to two cornerstones of the relationship: tissue factor (which, of course, promotes clotting) and thrombin (on which the conversion of fibrinogen to fibrin is based). Not only do these entities function in clot formation, but they’re crucial signaling factors, a fact that may be somewhat underappreciated by pathologists, Dr. Adcock suggests. These factors interact with cell surface receptors and modulate cell behavior by activating cellular pathways. And it just so happens, says Dr. Adcock, that the cellular pathways activated by thrombin and tissue factor promote tumor growth and invasion. Her article, as well as another review published in the New England Journal of Medicine (Furie B, Furie BC. 2008;359:938–949) discuss tissue factor’s role in mediating intracellular signaling events. Dr. Adcock also draws attention to the role fibrinolysis plays in cell movement. “Cell movement is critical for tumor growth and angiogenesis. This is because you have to break down the matrix that either a tumor lives in or a blood vessel is trying to grow into. And that matrix is broken down by the tissue fibrinolytic system in conjunction with matrix metalloproteases.”

For the Clinical Lymphoma & Myeloma article, Louis Fink, MD, co-author on the paper, helped develop a graphic showing the intra- and extracellular factors and pathways that enhance the expression of tumor factor on tumor cells (see “Regulation of angiogenesis by the hemostatic system,” at left).

“Signaling pathways are like alphabet soup,” says Dr. Fink, director of laboratories, Nevada Cancer Institute, Las Vegas. “Sometimes I lie awake and see all these pathways.”

Like Dr. Adcock, he sees a tre­men­dous growth in understanding the mechanisms of thrombosis—including cases involving malignancy—in recent decades. “When I started to work in this area, about 30 years ago, we could only explain maybe 10 percent of thrombosis that had no underlying cause. Now we’re up to well over 80 percent.” That’s much better—but still not good enough, says Dr. Fink. “We need to be more predictive about who’s going to get thrombosis, because we can’t anticoagulate everybody.”

In the next few years, he expects physicians to close in on that goal, as genetic mechanisms and coagulation testing become even more refined.

That should dovetail nicely with efforts in the past several years to expand awareness clinically of malignancy and thrombosis, says Jawed Fareed, PhD, professor, departments of pathology and pharmacology, and director, Special Coagulation Laboratory and the Hemostasis and Thrombosis Research Program, Loyola University Medical Center, near Chicago. “There are more articles, more coverage, more professional societies. Even the ­hospital groups have alerts on cancer patients now” to warn caregivers of the risk of thromboembolism, says Dr. Fareed. Patients know more, too. “A cancer patient is a very intelligent patient because they know what drugs they’re using, what kind of cancer they have, and what kinds of complications” they might encounter. Adding to the dialogue are the two aforementioned guidelines, one from the American Society of Clinical Oncology (Lyman GH, et al. J Clin Oncol. 2007;25:5490–5505) and the other from the American College of Chest Physicians (Hirsh J, et al. Chest. 2008;133[6 suppl] 67S–968S). The latter is the eighth edition of Antithrombotic and Thrombolytic Therapy: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines, published as a monumental supplement to the June issue; for a brief over­view, see the “Medical News & Perspectives” article on page 890 of the Aug. 27, 2008 issue of JAMA (vol. 300).

Certain chemotherapy drugs, including Taxol, antimetabolites, and some of the newer antibodies, can produce endothelial damage, Dr. Fareed notes, potentially resulting in hypercoagulable states. Even erythropoietin causes thrombosis, as do growth factors.

Malignancy and thrombosis are intimately related. “The interrelationship is very striking,” Dr. Adcock says. “It’s well known that when some cancer patients are treated with anticoagulants, they have improved survival—one facet of this effect is that anticoagulant agents may inhibit angiogenesis.”

Another factor playing a role in the hypercoagulability of malignancy explored in the Clinical Lymphoma & Myeloma paper is the role of microparticles shed from tumor cells, many of which are laden with tissue factor. This, says Dr. Adcock, may be part of the mechanism causing hypercoagulability.

Adds Dr. Fink: “Almost everybody since Trousseau knows that cancer promotes increased thrombosis risk. But it’s more than most people realize—there’s an important relationship between coagulation and the behavior of cancer cells.” This includes their growth properties, which are related to tumor cells’ signaling pathways and to thrombotic pathways, both of which are linked to the activation of platelets, which carry angiogenic factors as well as growth factors.

Drs. Fink and Adcock would like to begin focusing specifically on the role of platelets in cancer and thrombosis. Several experimental animal studies have shown that inactivating platelets reduces the number of metastases. “In other words, tumor cells appear to require platelets to establish metastatic lesions,” Dr. Fink says.

As researchers gather up all the strands and work toward the equivalent of a unified theory for this field, pathologists would do well to gain a firm grasp of these basic issues. “It’s going to have a direct impact on upcoming targets for cancer therapies, and there are already so many new antiangiogenic drugs,” Dr. Adcock says. More antithrombotic agents are also emerging. Early on, the only choices were heparin and warfarin. Last year dalteparin (Fragmin), a low-molecular-weight heparin, was approved for use in the extended treatment of symptomatic venous thromboembolism to reduce recurrence in cancer patients. (An evidence-based analysis of dalteparin used in this manner was published in 2006: Bick RL. J Support Oncol. 4:115–120). Physicians can also turn to direct thrombin inhibitors, and their choices of oral anticoagulants are growing, including direct factor Xa and factor IIa inhibitors.

Dr. Fareed would like to see physicians develop protocols for different cancers based on their risk categories. Colorectal carcinoma, pancreatic carcinoma, small-cell lung carcinoma, and some of the brain cancers are very aggressive in producing thrombotic events, he notes; physicians need to be made aware of these so they can control thrombotic events by using safer, nonthrombotic cancer drugs and prophylaxis with some type of anticoagulant, ideally with a newer, safer agent.

“If you tried to get warfarin approved by the FDA today, you probably couldn’t,” says Dr. Fink. “There are so many drug interactions and problems with bleeding.” If newer agents can reduce those risks to a more acceptable level, “then we’ll see a different kettle of fish.”

Until then, there is the evergreen, and ever-underestimated, issue of D-dimer testing.

Its chief value is to rule out venous thrombosis, though because patients with an underlying malignancy tend to have higher baseline D-dimer levels, its value is somewhat limited in the cancer population, Dr. Adcock says. The sensitivity and negative predictive value for ruling out thrombosis do not change with underlying malignancy, however. “So it’s still a very useful test to rule out venous thrombosis,” Dr. Adcock says. “It’s just that you’re not going to be able to rule out as many patients. It’s very similar to what we see in an older age group because in an older population, D-dimer values also tend to be higher.”

There’s a risk to increasing the cutoff in patients with malignancy, Dr. Adcock says, because if this is done, patients with thrombosis may be missed. However, she continues, the D-dimer may have another value in this population. Recent studies have suggested that if the D-dimer value is very high—the D-dimer level used in one study (Schutgens REG, et al. Haematologica. 2005;90:214–219) was greater than 4,000 ­µg/L—that patient has a much higher chance of having an underlying malignancy than if the patient has a D-dimer level of less than 1,000 µg/L. This can be useful to help determine which patients with thrombosis may benefit from more extensive studies to identify an underlying malignancy.

More studies need to be done, Dr. Adcock acknowledges. She also points out that using high D-dimer values at presentation of a thromboembolic event to predict underlying malignancy is more valuable in a younger age group (60 being the typical cutoff age), since incidence of cancer increases above age 60, and D-dimer levels increase with age.

Thus, using this as a predictor of malignancy carries less weight when the baseline population has a higher D-dimer level; its value is greatest in patients less than age 60. “But I still think it’s going to bear out to be a useful tool to help us understand which patients with thrombosis should undergo enhanced screen­ing for an occult cancer,” Dr. Adcock says.

In case there’s any confusion, she’s referring to D-dimer tests based on quantitative immunologic methods. The old-fashioned latex agglutination assays no longer have a place in clinical medicine, she says, though she’s occasionally shocked to find them in use for ruling out venous thrombosis.

She also notes that when reading the literature, pathologists need to make certain they understand the appropriate cutoff value, including whether it’s reported in FEUs (fibrinogen equivalent units) or D-dimer units; the difference is a two-to-one relationship (two FEUs equals one D-dimer unit). “For decades, authors did not comment as to which unit was used, as many just didn’t understand this variation even existed,” says Dr. Adcock.

Dr. Fink has been exploring several other areas where D-dimer might be useful. He and research colleagues have shown that with prostate cancer, the higher the D-dimer level, the higher the stage of the tumor. Moreover, he adds, if a patient presents with an elevated D-dimer, and other underlying conditions are ruled out, cancer might be the underlying cause. In these two situations, D-dimer “is almost like a tumor marker,” Dr. Fink says.

D-dimer could be used more frequently, says Dr. Fink. “In fact, Dr. Adcock and I are doing some studies to relate D-dimer to other cancer markers.” Toward that end, he established a biorepository at the Nevada Cancer Institute some three years ago; it now has 2,500 patients.

The other laboratory angle concerns low-molecular-weight heparin.

It’s well known that one advantage of LMWH over unfractionated heparin is that monitoring is generally not needed. In certain populations, however, monitoring might be appropriate. This would include patients with renal failure, those with severe deviation in body mass index, or those on long-term therapy (more than a month)—which includes the cancer population—in whom the drug might accumulate over time.

There’s nothing new about this, yet some physicians are still at sea, says Dr. Adcock. She recently took a call regarding a patient who had been administered LMWH but exhibited a low level. Was this patient possibly resistant to LMWH?

“My first question was, When did the physician administer the heparin, and when did you monitor it?” says Dr. Adcock. As it turns out, the patient was given the heparin the previous afternoon, then drawn the following morning.

“It’s so important that we understand the best time to draw these levels,” says Dr. Adcock. “If we’re going to monitor, then we’ve got to monitor appropriately.”

The recommended assay for monitoring LMWH is the anti-Xa assay. When should that test be drawn, and what is the target level? “Here’s where a lot of the confusion comes in,” says Dr. Adcock.

Patients on LMWH can be treated with once-a-day or twice-daily dosing. “When we monitor the anti-Xa levels, we should be measuring the peak level, generally,” she says. For once-a-day dosing, the peak level should be drawn at about four hours to six hours after subcutaneous administration. For patients receiving twice-daily dosing, levels should be drawn at three to four hours. When low-molecular-weight heparin is administered subcutaneously, which is generally how it’s done, the heparin anti-Xa levels vary throughout the day. “I don’t know how many laboratories provide clinicians recommendations as to when peak levels should be obtained,” Dr. Adcock says.

She says physicians may not appreciate this subtlety of LMWH because when patients are on unfractionated heparin, the continuous-infusion delivery is less susceptible to peaks and troughs.

Another problem: The published therapeutic ranges for LMWH are not as rigorously defined as therapeutic ranges for other drugs, including unfractionated heparin. “They’re controversial, but they’re the best we have,” Dr. Adcock says.

Basically, the peak level for patients on twice-daily LMWH dosing ranges between 0.6 to 1.1 IU/mL; the peak level target for once-daily dosing ranges between 1.0 to 2.0 IU/mL.

The problem arises because the same dose of two different LMWHs can have significantly different peak levels owing to the difference in the ratio of anti-Xa to anti-IIa activity between different drugs. “What we’re doing is monitoring these patients by looking at their anti-Xa,” Dr. Adcock says. Though a better measure might be anti-IIa activity (that is, antithrombotic activity), anti-Xa assays are more readily available and have established therapeutic ranges. Perhaps the best way to measure LMWHs would be with a thrombin generation assay, she says, though current assays are not useful, according to Dr. Fareed.

The anti-Xa assay is a window to drug concentration. When patients are administered LMWH, however, studies have shown that anti-Xa levels do not correlate with bleeding or thrombotic risk, so it’s not a good measure of what’s happening with the drug in vivo. Add to that the aforementioned problems of generalized therapeutic ranges using anti-Xa for all LMWHs, the possible failure of laboratories to tell their clinicians whether their ranges are based on once- or twice-daily dosing, and the frequent failure to draw patients at peak level, and the monitoring becomes worthless, Dr. Adcock says.

It is appropriate to monitor patients on long-term therapy to ensure the drug is being cleared appropriately. “If that’s what your clinician is after, then I’m going to propose that we should probably look at trough levels,” Dr. Adcock says. “So clinicians should consider measuring a level before the drug is administered again.”

Another well-established fact that should be well known—but isn’t, says Dr. Adcock—is when patients have underlying malignancy and thrombosis, they can be refractory to warfarin, suffering continued thrombosis despite the warfarin therapy. The culprit in these cases is an underlying DIC state. Such patients typically respond better to heparin, which is more effective with chronic DIC state in preventing ongoing or recurring thrombosis. “That’s been known for a long time, and it still amazes me that it’s not well appreciated.”

The bottom line is simple, says Dr. Adcock. “We can do better for our patients.”

Karen Titus is CAP TODAY contributing editor and co-managing editor.

Related Links Related Links