January 2012
Feature Story

William Check, PhD

And on this team, the winners and still champions aspirin, heparin, and warfarin. Like the ancient Titan Cronos, who rebuffed many challenges from his powerful offspring before finally being vanquished by his son Zeus, these three venerable graybeards continue to dominate their clinical realms despite the birth of many new antiplatelet agents and anticoagulants. Michael Laposata, MD, PhD, calls the efforts to displace these therapeutically effective elders “a big growth industry,” saying, “We keep getting new agents.” This is especially true of warfarin and heparin. So far there has been no “Zeus” of anticoagulants, but one or more of the new arrivals could soon fill that historic role.

Aspirin has been used for thousands of years; as a pure synthetic compound it dates from 1850–1870. Clinical trials of heparin began in 1935; it entered clinical use in 1937. Warfarin, the youngster of the bunch, took on its anticoagulant incarnation in the mid-1950s. In contrast, many of the upstart antiplatelet agents and anticoagulants are callow youths. Just in the past year three oral agents were approved for prevention of stroke in atrial fibrillation—rivaroxaban (Xarelto), dabigatran (Pradaxa), and apixaban (Eliquis). The financial stakes are huge. Broad substitution of one of these oral anticoagulants for warfarin could increase health care costs by $20 billion over the next 10 years, in the estimation of Dr. Laposata, who is the Edward and Nancy Fody professor and executive vice chair of pathology, microbiology, and immunology and professor of medicine at Vanderbilt University School of Medicine and pathologist-in-chief at Vanderbilt University Hospital. Do the therapeutic benefits of the newer agents justify this increase? How will the health care system deal with this huge new expense? At the present time, there are no clear answers to these questions.

However, several important principles in this area are very clear, Dr. Laposata says, including some that perhaps many are not fully aware of. One is reversibility. If a patient starts bleeding on Coumadin, you can reverse the anticoagulant effect with vitamin K or a transfusion of fresh frozen plasma, or both. However, with the newer agents there is much less experience with safe reversal. For example, Dr. Laposata says, “If a patient starts bleeding on dabigatran, we don’t have clear evidence about what to do.”

Another potential drawback of newer oral anticoagulants comes from their pharmacokinetics. They have a rapid onset of action, but, on the flip side, they also have rapid offset. “If a patient forgets one dose of rivaroxaban or dabigatran, they could be in big trouble,” Dr. Laposata warns. In one day much of the anticoagulant effect is lost.

Another salient fact is that there is more than one replacement for warfarin. “People will have to choose,” Dr. Laposata says. In these early days, definitive differences among them are not apparent.

A final principle concerns laboratory monitoring. Newer drugs are touted as needing no monitoring, but all of these drugs need some monitoring at some time, Dr. Laposata cautions. “A lot of people think that when you don’t need to monitor, you don’t need to test.” In fact, he points out, “We will have to be able to determine plasma concentrations for those newer drugs that don’t require routine monitoring when patients on these drugs experience bleeding or thrombosis. We really want to get it right.”

Dr. Laposata presented an overview of antiplatelet agents and anticoagulants as well as methods of monitoring for them at a course at CAP ’11 last fall. Antiplatelet agents—aspirin, clopidogrel (Plavix), prasugrel (Effient), ticagrelor (Brilinta), and glycoprotein IIbIIIa inhibitors—are indicated for many conditions in which platelet aggregation leads to arterial thrombosis. Summing all major indications—heart attack and angina, stroke, diabetes in those over age 30, and peripheral arterial disease—gives a total clinical potential of 30 to 40 million users, Dr. Laposata said. “Maybe everyone should be taking an antiplatelet agent unless they have a contraindication,” he opined. His statement is not an exaggeration. The U.S. Preventive Services Task Force recommends prophylactic low-dose aspirin for all men between the ages of 45 and 79 and all women between 55 and 79 “when the potential benefit due to a reduction in myocardial infarctions outweighs the potential harm due to an increase in gastrointestinal hemorrhage.”

Aspirin blocks thromboxane A2 synthesis irreversibly. Clopidogrel, prasugrel, and ticagrelor intake produce inhibition of platelet ADP receptors. All are oral agents. The IV drugs abciximab (Reopro), eptifibatide (Integrilin), and tirofiban (Aggrastat) block glycoprotein IIbIIIa receptors. Each of these groups of drugs has a different impact on platelet function. “Imagine that I am a platelet,” Dr. Laposata said in his course. “Your goal is to immobilize me.” Acting out his descriptions, he then said, “Aspirin ties up my hands, Plavix ties up my feet, and a glycoprotein IIbIIIa inhibitor is like a large piece of Saran wrap completely covering my body.”

Unfortunately, though there are laboratory tests to determine whether a drug is inhibiting platelet function, Dr. Laposata said, “these tests often do not agree with each other.” Moreover, “There is no gold standard test.” As a result, even though antiplatelet agents are used for life-threatening conditions, patients are typically not monitored. “Going on faith for most patients” is how Dr. Laposata puts it.

Anticoagulants, which are mostly used for venous thrombosis and to prevent stroke in patients with atrial fibrillation, include the oral agents warfarin (Coumadin), dabigatran, and rivaroxaban; and the IV/SQ agents unfractionated heparin, low-molecular-weight heparin [enoxaparin (Lovenox), dalteparin (Fragmin)], fondaparinux (Arixtra), lepirudin (Refludan), and argatroban.

In non-valvular atrial fibrillation, a clot can be formed in the heart and travel to the brain. “You would think it might be good to use antiplatelet therapy in this condition,” Dr. Laposata told CAP TODAY. “For run-of-the-mill stroke we do give antiplatelet therapy. But people tried it for atrial fibrillation and it didn’t work as well as warfarin.”

Warfarin had a circuitous route into medical use, which Dr. Laposata calls “a fascinating story.” During the Depression a farmer in Wisconsin fed spoiled clover to his cows. After a few weeks they began to bleed and some died. The farmer submitted samples to the University of Wisconsin, which identified the responsible compound. (Warfarin is named after the acronym for the Wisconsin Alumni Research Foundation.) In its first test as a blood thinner, warfarin was used at too high a dose, with disastrous results, which made people think it was too dangerous to use for anything but rat poison. Then, in 1952, a person in the Navy attempted suicide with warfarin. He lived, leading to its reintroduction in lower doses. When president Eisenhower had a heart attack in the mid-1950s they gave him Coumadin, and its reputation was made.

Warfarin’s effect comes not from blocking coagulation factors, but from reducing the synthesis of four of them (II, VII, IX, and X). That explains why it takes several days to reach full effect, and why in some situations patients have to be treated with heparin as a bridge. Laboratory monitoring of warfarin with prothrombin time uses reagents from either homogenized rabbit brain thromboplastin or a recombinant thromboplastin, leading to different results with different lots of reagents for the same sample. Standardization was accomplished with the International Normalized Ratio, which depends on knowing the International Sensitivity Index for any lot of thromboplastin.

The recommended INR for warfarin is 2.0–3.0 for most indications or 2.5–3.5 less commonly. “For all other anticoagulants, the preventive dose is lower than the treatment dose, but not so for warfarin,” Dr. Laposata told CAP TODAY. “The implication is that when you are trying to prevent a clot you have a higher risk of bleeding with warfarin than with other anticoagulants. Warfarin is a very difficult drug to take.”

Also important is that patient self-testing for warfarin has clearly better outcomes for bleeding, thrombosis, and in-range results than management by a physician practice, and it has similar outcomes to management by an anticoagulation clinic. “My first reaction when they told me about point-of-care testing,” Dr. Laposata says, “was complete silence for 15 seconds, then I broke out laughing. I thought, That is totally ridiculous to risk a life-threatening mistake with a home INR test.” That was in the 1990s. In 2006 an article in The Lancet summarized major studies of home monitoring of INR. “The results were far superior to having a doctor manage your INR,” Dr. Laposata says. One possible explanation is that if patients go to a hospital for monitoring, they go once per month. If they do it at home, they might do it four times per month. “So they are more likely to catch a very high or very low value and adjust their Coumadin dose before an adverse event.” If Dr. Laposata himself took Coumadin, he would have a home monitor and check his value weekly. “Every few months I would compare results on my home instrument with results on a lab instrument,” he says.

The heparin story is one of decreasing length and increasing specificity. Unfractionated heparin, a polysaccharide that comes from pig intestine, is heterogeneous in length and unpredictable in effect. It was observed that shorter heparin molecules have better pharmacokinetic properties. Low-molecular-weight (LMW) heparin is made by cutting unfractionated heparin. When people set out to make a synthetic substitute for heparin, they found that only five sugars are needed for its effect. Fondaparinux was born.

Heparins bind to antithrombin, changing its configuration and allowing antithrombin to bind activated coagulation factors and inactivate them. Unfractionated heparin inactivates many factors. LMW heparin inactivates factors Xa and IIa (thrombin); fondaparinux inactivates only factor Xa.

Unfractionated heparin is monitored in the laboratory with the partial thromboplastin time; the therapeutic range is on the order of 55–85 seconds. However, different lots of partial thromboplastin have different heparin response curves, so the therapeutic range must be established locally. Keeping a patient in the therapeutic range with unfractionated heparin typically requires multiple dose adjustments. Dr. Laposata demonstrated this with data from Massachusetts General Hospital (where he was director of clinical laboratories before joining Vanderbilt): After patients were brought into the therapeutic range, only 29 percent remained there on the next two measurements; only seven percent maintained a therapeutic partial thromboplastin time over four consecutive days.

LMW heparin promised improvements. “When [LMW heparin] first came out, we were very excited about it,” Dr. Laposata told CAP TODAY. “I wore a badge on my lab coat saying ‘Low-molecular-weight heparin is better.’” Most importantly, the patient response is more predictable: A standard dose puts patients into the therapeutic range a high percentage of the time.

Because LMW heparin is more predictable in its effect, typically no monitoring is done. Notable exceptions are for obesity and low body mass, impaired renal function, pregnancy, and for patients on long-term therapy. In these cases, monitoring is done with the anti-factor Xa assay, using a target of 0.5–1 U/mL. “A really important indication for monitoring, one that sometimes people forget, is for a pregnant woman who was previously on Coumadin,” Dr. Laposata said in an interview. Since Coumadin causes birth defects, these women are switched to LMW heparin. However, as the fetus grows, it metabolizes LMW heparin faster than the mother, so the dose must often be increased as pregnancy progresses. “There have been a number of cases in which a mom was taking Lovenox during pregnancy but was not monitored with an anti-Xa assay,” Dr. Laposata told CAP TODAY. “A few women died during pregnancy because they were underdosed late in pregnancy. So monitoring the anti-factor Xa level late in pregnancy is very important.”

Another crucial consideration is the distinction between the factor X and anti-factor Xa assays. After Dr. Laposata left Massachusetts General Hospital in 2008, one of his patients there called him about pain in her abdomen from bleeding. She also had a severe bleed in her leg. Dr. Laposata asked her what her anti-factor Xa level was. Her physician hadn’t checked it. When she asked her physician to get a level, he ordered a factor X level. “That only tells you what the factor X is,” Dr. Laposata says. “Anti-factor Xa tells you how much LMW heparin you have on board.” When the woman got the desired anti-factor Xa result, it was two times the upper limit of the therapeutic range. “She was bleeding from an overdose of anticoagulant,” Dr. Laposata says.

Protamine sulfate is recommended when emergent reversal of LMW heparin is needed.

Fondaparinux is approved for a variety of indications in which LMW heparin would otherwise be used. Monitoring is not usually done, but an anti-Xa assay can be used.

Lepirudin and argatroban are direct thrombin inhibitors that are used in place of heparin or LMW heparin in heparin-induced thrombocytopenia (HIT), a hypersensitivity reaction in which platelets clump. Both can be monitored with the partial thromboplastin time. “We are just lucky that a $5 test tells us all we need to know,” Dr. Laposata says.

Rivaroxaban and dabigatran are intended as oral substitutes for warfarin. Dr. Laposata reiterates his concern about missed doses with these agents. “One disadvantage to Coumadin is that it takes a long time to take effect,” he says. “On the other hand, if a patient forgets a single Coumadin dose, it retains its effect for a while. But if a patient taking rivaroxaban forgets one day’s drug, that omission will result in blood that is not effectively anticoagulated. It’s a really challenging problem.”

Neither rivaroxaban nor dabigatran is usually monitored. A method for monitoring rivaroxaban is in progress. “We will have something to monitor plasma concentration of rivaroxaban in our own laboratory soon,” Dr. Laposata says. “We are now doing plasma concentrations of dabigatran and thrombin times to rule out residual dabigatran. So even though the label says ‘No testing required,’ there is testing.”

“A recent article in Circulation showed that a prothrombin complex concentrate, a mix of four coagulation factors, can stop bleeding in a patient on rivaroxaban,” Dr. Laposata says. The concentrate did not reverse dabigatran. General considerations about reversibility were summarized recently (Levi M, et al. J Thromb Haemost. 2011;9:1705–1712).

Of the possible $20 billion increase if the new oral anticoagulants were to be widely used, Dr. Laposata says, “Can you imagine that? That is a huge number!” And yet data show these drugs to be safer than warfarin. In the RE-LY trial, for stroke prevention in atrial fibrillation, dabigatran was more effective than warfarin and had fewer side effects (Connolly SJ, et al. N Engl J Med. 2009;361:1139–1151). “Everyone is afraid of intracranial bleeding,” Dr. Laposata says. “For dabigatran it was less than half of what it was with warfarin.”

“There are also studies that show that dabigatran is equivalent to Coumadin for a venous clot in the leg or in the lung,” he says. “However, it is not specifically approved for that indication.”

As of Dec. 7, 2011, the FDA is investigating all reports of bleeding in patients taking dabigatran [FDA Drug Safety Communication: Safety review of post-market reports of serious bleeding events with the anticoagulant Pradaxa (dabigatran etexilate mesylate). www.fda.gov/Drugs/DrugSafety/ucm282724.htm]. Worldwide post-marketing surveillance has confirmed there have been 260 fatal bleeding events between March 2008 and Oct. 31, 2011. Safety advisories have been issued in Japan and Australia. European and U.S. agencies are focusing on the need for more frequent monitoring of kidney function to mitigate the risk. Two clinical cases, one fatal, were published on July 25, 2011 in Archives of Internal Medicine: Legrand M, et al. The use of dabigatran in elderly patients. 2011;171(14):1285–1286.

Market penetration of rivaroxaban and dabigatran has been somewhat slow at first, Dr. Laposata says. “After people gain clinical experience, if [these drugs] live up to what was experienced in clinical trials, people may go in droves to rivaroxaban and dabigatran and, when it becomes FDA approved, apixaban. The economics of these drugs will become a substantial problem because the daily dose is likely to increase from about 35 cents for warfarin to $5 for each of the new anticoagulants.”

For many years the Food and Drug Administration was reluctant to allow new drugs in this category, Dr. Laposata notes. Now the FDA seems to have opened the door. Will one, or perhaps all, of the newer drugs rush through and overwhelm the defending champion?