Published in the February 2005 issue of Today’s Hospitalist
A debilitating “and even deadly “reaction to heparin is receiving a growing amount of attention, both because the syndrome can be so confounding to physicians who have little experience with it, and because early recognition and treatment can usually prevent catastrophe.
The condition, heparin-induced thrombocytopenia (HIT), occurs in only 3 percent to 4 percent of patients exposed to unfractionated heparin, so it may not be a top concern of physicians prescribing anticoagulants. Experts say that HIT, however, deserves more attention than it has received, particularly in the inpatient setting.
While patients may face a low risk of developing the syndrome, the number of Americans who are exposed to heparin each year “12 million, by most estimates, or one-third of all inpatients “is so high that a significant number of patients are potentially at risk. When HIT isn’t diagnosed or treated, it can lead to problems ranging from the loss of a limb to death.
What makes HIT so challenging to diagnose and treat is that it defies many of physicians’ instincts.
Try to detect HIT by looking for a platelet count of less than 100,000, for example ” the classic marker of thrombocytopenia “and you’re likely to miss many cases of the syndrome. And try to treat the condition merely by stopping heparin “or by using a commonly used anticoagulant like warfarin “and you increase the odds that your patient will develop a clot.
Here’s a look at a condition that some physicians describe as paradoxical, and how you can make sure you don’t miss it in your patients.
HIT is an immune-mediated drug reaction that occurs when heparin-dependent antibodies develop, usually after patients have been taking heparin for between five and 15 days. The syndrome’s name implies that thrombocytopenia is the end result, but the real danger is a thrombotic state, not bleeding. And while the risk of clotting is highest early in the condition, the danger can persist for several weeks.
The risk of HIT is highest in patients who have undergone surgery, making it a special concern for hospitalists who co-manage surgical patients. Experts say that the condition is probably most common in patients who have undergone cardiovascular surgery or percutaneous coronary interventions, largely because heparin is typically given in such large doses during these procedures.
“Any time that patients in the hospital have a fall in platelet count or show signs and symptoms of a new clot, you need to think about HIT,” says Lawrence Rice, MD, a hematologist who is professor of medicine and program director of hematology at Baylor College of Medicine in Houston. “You have to consider the diagnosis because the implications are so important.”
After cardiac surgery, orthopedic surgery is thought to present the next highest risk factor for developing HIT. While medical patients are thought to have a lower risk of developing the syndrome “particularly when they are receiving prophylactic doses of heparin “Dr. Rice points out that any exposure to heparin causes risk, even when it is something as simple as a catheter flush.
Experts also emphasize that while patients who receive low-molecular-weight heparin instead of unfractionated heparin have a lower risk of developing HIT “less than 1 percent, by most estimates “these patients can still develop the syndrome.
What makes HIT so difficult to diagnose, particularly in the medically complex patients that are the bread and butter of hospitalists, is the fact that the symptoms can be masked by so many other conditions and factors.
“HIT can be anything from a seemingly unimpressive and unrecognized drop in platelet count to a lethal thrombotic storm,” says hospitalist Tracy Minichiello, MD, who is director of the anticoagulation service at the University of California, San Francisco. “It seems paradoxical. But once you understand the biology, it makes sense.”
So how do you detect this deadly condition? The first and most important step, experts say, is to track patients’ platelet counts. But if you start by looking for thrombocytopenia “a platelet count under 100,000 “you’ll likely miss many cases of HIT.
That’s because instead of looking for an absolute platelet count, you need to watch for a percentage drop. In more than 90 percent of HIT cases, explains Dr. Rice, platelet counts fall more than 50 percent from baseline. It’s an important distinction, he points out, because while you wouldn’t consider a patient with a platelet count of 200,000 thrombocytopenic, that individual is a candidate for HIT if his baseline platelet count was 400,000 “and he is or was recently taking heparin.
“Part of the problem with HIT is that the name of the condition is often incorrect, because thrombocytopenia is not necessarily part of the syndrome,” says William H. Matthai Jr., MD, a cardiologist and clinical associate professor of medicine at the University of Pennsylvania Medical School in Philadelphia. “Heparin-induced falling platelet syndrome would be a more correct term.”
The key to detecting HIT before a thrombosis develops is to get a baseline platelet count, and then recheck the patient’s platelet count every day or two. Regular platelet counts are critical to catching HIT, experts say, because when the count falls, you need to compare it to the baseline, not just the previous day’s level.
Experts worry that many physicians may not be collecting enough data on platelet counts to spot HIT.
“Particularly in an era of trying to save money and control costs,” says Stan Nasraway, MD, associate professor of surgery, medicine and anesthesia at Tufts University School of Medicine in Boston, “physicians may not regularly think about following platelet counts. But it’s really important, because you’re not necessarily looking for an absolute platelet count.”
The importance of timing
While platelet counts can provide the first evidence of HIT, the timing of those changes is critical. In patients who have just undergone cardiac surgery, for example, it’s not unusual for platelet counts to plunge the first few days after the procedure.
This drop can be the result of a dilutional effect from transfusions, or from a less dangerous form of HIT that is sometimes referred to as type 1 HIT or heparin-associated thrombocytopenia. Dr. Rice says this other condition, which should not be confused with HIT, is not immune-mediated, occurs early after exposure to heparin (usually within 48 hours of exposure), and is benign.
“If the platelets are at 100,000 two days after heart surgery,” Dr. Rice says, “that’s normal. But if they’re still at 100,000 five or six days post-operatively, then you need to think about HIT.”
And because HIT can develop five to 15 days after exposure to heparin, some patients may develop signs and symptoms of HIT after you’ve discharged them.
“Patients may only be exposed to heparin for five or six or seven days before they are transitioned to something else, but the consequences of HIT can continue even though the heparin has stopped,” says Dr. Nasraway. “Most people aren’t even thinking about that once the heparin has stopped.”
Dr. Nasraway, who is also director of the surgical care ICUs at Tufts New England Medical Center, gives the classic case of a patient who undergoes successful elective cardiac surgery and is discharged from the hospital in five days, before any signs of HIT develop.
“The patient goes home, then re-presents to another hospital with a DVT two weeks later, and no one has any idea of what has happened or why,” he says. “And what’s their initial treatment for a DVT? Heparin. So HIT caused the DVT while the patient was home, and the first thing the doctor in the ED or the hospitalist admitting the patient will do is put the patient on heparin.”
When it comes to timing, there are also growing concerns about delayed onset of HIT. Recent literature indicates that patients who have been exposed to heparin in the last 100 days and still harbor the antibody responsible for the syndrome may experience a rapid onset of the syndrome after being re-exposed, possibly within hours or days.
While there are several tests that can detect HIT antibodies and help confirm a diagnosis of HIT “the enzyme-linked immunosorbent assay (ELISA) test is most popular “it typically takes days to get these test results back. If you think your patient is at risk of developing a thrombotic event, you don’t want to wait for antibody results to take action. That’s why experts say physicians need to use their best clinical judgment to choose a treatment and then re-evaluate that decision when the test results come back.
Delays aren’t the only problem with the ELISA test. Dr. Rice points out that while up to 50 percent of patients who have undergone heart surgery will test positive for HIT antibodies, less than 5 percent will actually develop the condition.
There are, however, some classic signs that a patient may have HIT. Dr. Rice says that studies suggest that in orthopedic surgery patients, for example, clots tend to be venous. And Dr. Matthai from the University of Pennsylvania says that if your patient develops an unusual thrombotic event or numerous events, HIT is a distinct possibility.
“If you have a patient who has been lying in bed with heart failure and has a DVT,” Dr. Matthai explains, “that’s not necessarily unusual. But if a patient has an upper extremity DVT for no apparent reason, that’s unusual. Or if the patient has an arterial event, that’s unusual.”
“We’ve seen some patients with subacute stent thrombosis after an angioplasty, which is uncommon,” he continues. “If you have a perfect result and no reason to suspect a thrombosis, HIT should be floated as a possible diagnosis.”
Further confounding a differential diagnosis of HIT, experts say, is differentiating HIT from conditions like sepsis. Dr. Rice notes that while recent studies indicate that up to one-third of all ICU patients may have thrombocytopenia, sepsis is also a common cause of the condition in those patients.
He adds that sepsis, HIT and thrombocytopenia due to other drugs, in fact, are the three most common causes of thrombocytopenia he sees in hospitalized patients. He also notes that devices like intra-aortic balloon pumps can lower platelets.
Scott Keller, MD, a hospitalist and senior associate consultant in the division of general internal medicine at the Mayo Clinic in Rochester, Minn., says that while he considers HIT a cause when a patient’s platelet count drops, it is one of many possibilities.
“There are so many things we see that cause low platelets more frequently than HIT,” Dr. Keller explains, “that I have to consider all the possibilities. You keep it in the back of your mind, and if a patient meets the clinical criteria and has a positive PF4 antibody, there’s no doubt about what to do. But as internists, we always have to be aware of the differential diagnosis.”
Treating high-risk patients
With patients who have a high risk of developing HIT “individuals with a history of HIT, patients on anticoagulation because of a thrombosis, and patients who have undergone cardiac surgery “the recommended treatment is to stop the heparin and immediately give the patient a direct thrombin inhibitor. Once the platelet count returns to normal, resume anticoagulation with an agent like warfarin.
The problem, experts say, is some physicians think that because heparin is the cause of the syndrome, simply stopping the drug is the solution.
“The single most important thing that physicians need to realize is that if they suspect HIT, stopping the heparin by itself is not good enough,” explains Dr. Nasraway from Tufts. “They have to not only send an antibody test to confirm their diagnosis, but they have to appreciate that patients who have HIT have a 50 percent likelihood of developing a thrombosis even after the heparin has been stopped in the following 30 days.”
Dr. Matthai says that in the first few days after a diagnosis of HIT is made, the incidence of thrombotic events averages between 5 percent and 10 percent per day, so immediate treatment with an alternative anticoagulant like a direct thrombin inhibitor is critical for high-risk patients.
“We often use direct thrombin inhibitors to buy time for patients,” Dr. Matthai says. “We can continue the anticoagulation without continuing heparin, which is important because these people are often taking heparin for some reason. They have unstable angina or they already have a DVT or they have a prosthetic valve, so you need to continue anticoagulation.”
While stopping heparin without administering an alternative anticoagulant is a mistake, so is giving the patient a drug like warfarin. Because the drug disables naturally occurring anticoagulant protein C before disabling procoagulants, warfarin can actually increase the chances of clotting in an HIT patient.
“You can never give warfarin to a patient with HIT in the absence of a direct thrombin inhibitor,” Dr. Matthai says. “You have to treat the patient for several days with a direct thrombin inhibitor, generally until the platelet count moves to over 100,000. Unopposed warfarin therapy could predispose a patient to thrombosis.”
When HIT is less certain
While experts say that an aggressive strategy is in order for patients who you strongly suspect of having HIT, that doesn’t always hold true for patients on the wards who may have a lesser risk of the syndrome.
For example, how do you treat the patient receiving prophylactic levels of heparin who shows a fall in platelet count, but no other signs of thrombotic activity? Should you immediately move that patient to a direct thrombin inhibitor, which, like any anticoagulant, presents its own risks of bleeding? Experts acknowledge that choosing a treatment for a patient with a low risk of HIT who has some signs and symptoms of the condition can be a tough call.
“I work in the ICU and all my patients are really sick,” says Dr. Nasraway, “so it’s easy for me to prophylax using direct thrombin inhibitors. But for patients on the hospital ward who are transitioning along, that decision is a little more murky.”
Dr. Nasraway says that giving direct thrombin inhibitors to a patient who you think has a low or medium risk of developing HIT can be a tough sell for several reasons. For one, the patient has to stay in the hospital a few extra days, which adds costs.
“You’re also prophylaxing in something that’s not well-appreciated,” he adds, “so many people don’t readily understand why we should keep these patients in the hospital for a few more days on an IV medication when they have never even seen a complication from HIT. It’s a tougher sell because it’s an uncommon phenomenon.”
Dr. Minichiello from UCSF agrees that deciding how to treat patients with a low risk of developing HIT is a particular challenge for hospitalists. “On the medical service,” she explains, “if your average length of stay is three or four days, your patients are probably going to get one or two days of heparin, so their exposure and their risk of getting HIT is lower. Your chances of seeing HIT increase in patients who stick around longer and receive more heparin.”
As a result, she uses a pre-test probability to guide how she treats HIT. In patients with a low pre-test probability of developing HIT, watchful waiting may be the first step of a process in which she carefully searches for alternatives to the syndrome.
Dr. Minichiello offers as an example the patient whose platelet count has dropped significantly, but the details do not necessarily indicate HIT. Perhaps the platelet count has not dropped 50 percent, for example, or the timing is not typical.
“I think at the very least you are required to stop their heparin and watch them if they have moderate pre-test probability,” she says. “I wouldn’t send them home at that point. But if they have a high pre-test probability, then I would give them a direct thrombin inhibitor.”
Ronald A. Sacher, MD, professor of the medicine and pathology division of hematology/oncology at the University of Cincinnati, agrees with the approach outlined by Dr. Minichiello. He says that if there is little likelihood that a patient has HIT, he has no problem discontinuing heparin without instituting another anticoagulant while he waits for the results of antibody testing.
“If it’s a prophylactic dose,” says Dr. Sacher, who is also interim dean of the College of Medicine at the University of Cincinnati, “I wouldn’t be quite as aggressive in using a direct thrombin inhibitor as quickly as in an individual who has an active thrombotic event.”
He explains that his decision-making hinges on an assess- assessment of how likely the patient is to have HIT. If the patient has only a moderate risk of the syndrome, he weighs each case individually. If the patient has a high risk of HIT, he doesn’t hesitate to immediately prescribe a direct thrombin inhibitor.
Dr. Minichiello adds that when choosing a therapy for patients with a relatively low risk of developing HIT, she also weighs the risks of direct thrombin inhibitors.
“A lot of these patients are complicated,” she says, “and giving them a direct thrombin inhibitor markedly increases their risk of bleeding, especially if it’s someone who was just on preventive doses of anticoagulation. It’s a little bit easier with someone who was being treated with heparin for an acute clot, because you’ve already done your risk benefit analysis for high-dose anticoagulation.”
When the decision is that complex, Dr. Minichiello says, she often calls for a hematology consult. While she wrote UCSF’s protocol for diagnosing and treating HIT, there are times when even she needs a little help.
“Deciding whether a patient has HIT is difficult,” Dr. Minichiello explains, “and the therapy is not something that everyone is comfortable with. Besides, if it’s not HIT, the specialist can help you figure out why the patient’s platelets are plummeting. It’s a nice side effect of the consult.”
Edward Doyle is Editor of Today’s Hospitalist.
Latest guidelines: Monitor patients you suspect have HIT
In September 2004, the American College of Chest Physicians issued new recommendations on heparin-induced thrombocytopenia as part of its seventh conference on antithrombotic and thrombolytic therapy. The guidelines establishes the following platelet monitoring schedule for patients who have a risk of 0.1 percent or higher of HIT:
“¢ If patients are receiving therapeutic doses of unfractionated heparin, platelets should be monitored at least every other day until day 14, or until the unfractionated heparin is stopped, whichever happens first.
“¢ If patients are receiving prophylactic heparin after surgery, platelets should be monitored every day from four to 14 days after the procedure, or until heparin is stopped.
“¢ In the following patients, platelet counts should be monitored every two or three days from day 4 to 14 or until heparin is stopped: medical/obstetric patients receiving prophylactic doses of heparin; postoperative patients receiving prophylactic doses of low-molecular-weight heparin; postoperative patients whose catheters are flushed with unfractionated heparin; and medical/obstetric patients receiving low-molecular-weight heparin after initially receiving unfractionated heparin.
“¢ No routine monitoring is needed for two groups of patients: medical/obstetric patients are receiving only low-molecular weight heparin, and medical patients whose catheters are being flushed with unfractionated heparin.
For more information, an abstract of the guidelines is online.