Share this article on:

Treating Warfarin-Associated Intracranial Hemorrhage

Roberts, James R. MD

doi: 10.1097/01.EEM.0000269588.19241.a0

After reading this article, the physician should be able to:

Author Credentials and Financial Disclosure: James R. Roberts, MD, is the Chairman of the Department of Emergency Medicine and the Director of the Division of Toxicology at Mercy Health Systems, and a Professor of Emergency Medicine and Toxicology at the Drexel University College of Medicine, both in Philadelphia. Dr. Roberts has disclosed that he has no significant relationships with or financial interests in any commercial companies that pertain to this educational activity.



Part 2 in a Series

Release Date: April 2007

Warfarin — a life saver and a killer — is omnipresent, and it would be an unusual shift if an emergency physician did not encounter a patient taking warfarin. The clinical indications are many, but the complications are gargantuan. Most emergency physicians have a limited knowledge of warfarin, primarily understanding bleeding complications. The literature on warfarin drug-drug interactions, initiation and reversal of coagulopathy, and a plethora of related issues comprise literature so vast that no physician can possibly read or remember it all.

Although we emergency physicians rarely start the drug, we frequently must deal with its therapeutic and untoward sequelae. I recently discussed the issue of head trauma in patients anticoagulated with warfarin, and concluded that banging your head and taking warfarin is a bad combination. Seemingly minor head trauma, especially in the elderly anticoagulated patient, is a minefield across which one must tread quite carefully. Although we may not prescribe the drug in the ED, we do prescribe many drugs and offer advice on medical issues that may interfere with the proper use of this anticoagulant.

I previously reviewed the basics of the clinical use of warfarin, and some issues about the drug that may have been forgotten by emergency physicians. The FDA has recently mandated that coagulated patients receive specific written instructions from the pharmacy about the clinical issues and dangers of warfarin therapy, and this information is well known to any individual with a computer. The sagacious clinician realizes that a patient and his family can quickly check on the doctor's proper or improper use of this drug, so let the clinician beware.

Last month, I discussed how to approach the asymptomatic patient who is serendipitously found to have an elevated INR. A very conservative approach has been promulgated, using either watchful waiting or small doses of oral vitamin K for those not bleeding and with an INR less than 9. Simply withholding warfarin for three to five days is considered a reasonable clinical approach in the vast majority of such asymptomatic patients. Last month's column discussed the epidemiology of a more dramatic and often fatal warfarin-related issue — intracranial hemorrhage (ICH), either spontaneous or associated with other risk factors.

I emphasized that it is not just highly elevated INRs that cause ICH; in fact, many such unfortunate patients have the therapeutic INR considered ideal to treat their underlying problems. Watchful waiting is clearly not the norm for these individuals. This month, I review possible treatments for ICH.

The reversal of anticoagulation is a controversial and largely unstudied issue in the ED for anticoagulated patients who present with spontaneous intracranial hemorrhage. Various techniques and interventions may be helpful in rapidly obtaining a normalized INR in critical patients. Interestingly, there are surprisingly few data on this particular issue, and even less evidence that rapid intervention favorably alters the outcome to any great extent. Nonetheless, EPs should be cognizant of the issues, even if we are not able to offer the Holy Grail to these unfortunate patients.



When one attempts to normalize an elevated INR secondary to warfarin therapy, there are three general approaches that are considered. These include fresh frozen plasma (FFP), vitamin K, and a variety of coagulation factors that are deficient in patients treated with warfarin, such as prothrombin complex concentrate and Factor VII (rFVIIa).

Review: Warfarin Therapy

Hanley JP, J Clin Pathol, 2004;57:1132

This is an excellent review article highlighting the most appropriate way to reverse the anticoagulation effect of warfarin. The authors emphasize that any approach to ameliorating warfarin-induced coagulopathy depends on individual clinical circumstances. Obviously, patients with serious bleeding (e.g., ICH) require more rapid reversal, but in other cases (e.g., hematuria), a more leisurely approach can be applied. My prior discussions have analyzed issues involved in reversing asymptomatic (non-bleeding) patients with coagulopathy secondary to warfarin, and a rather nonchalant, often simply a wait-and-see approach, has been advocated. My analysis of this particular article focuses on the nuances of rapid reversal when it is truly imperative to stop the bleeding.

Although one must weigh the benefit of temporary suspension or reduction of anticoagulation in asymptomatic patients or those with minor bleeding, intracranial hemorrhage is a different breed of coagulopathy. In such cases, the risk of thrombosis and loss of anticoagulation for chronic problems such as atrial fibrillation, DVT prophylaxis, or heart valve issues are temporarily inconsequential to the overall scenario.

With regard to therapeutic options, the most widely used method to replace coagulation factor deficits is the use of fresh frozen plasma. The oral or intravenous use of vitamin K, although appropriate for less serious bleeding, cannot be the sole approach for reversing anticoagulation in the presence of ICH. It is known that FFP does not immediately or totally reverse warfarin coagulopathy, so the use of prothrombin complex concentrate (PCC) or recombinant activated rFVIIa may be effective.

FFP has been the mainstay for warfarin reversal, but it is not complete, immediate, or totally ideal. It may be quite useful in patients with non-life-threatening GI or renal hemorrhage, but it is not without downsides. While frequently used, FFP may not be the best choice for patients with ICH, but that issue is far from settled. One important fact is that FFP does not have a hemostatic concentration of factor IX. Post FFP administration, factor IX levels do not rise above 20 percent, and this may not be reflected by merely measuring the INR. Other potential issues with FFP include the time, effort, and logistics of obtaining and infusing it and possible volume overload. FFP does have a minor thrombogenicity problem and a theoretical risk of transfusion-related infection. Of course, concomitant vitamin K therapy is always given with FFP to initiate hepatic synthesis of vitamin K-dependent clotting factors. FFP is usually effective in reversing a warfarin coagulopathy in four to six hours.

Vitamin K alone has little therapeutic usefulness in the clinical setting where rapid reversal of warfarin is required because it simply takes too long to become effective. Vitamin K can be given orally or intravenously, but the subcutaneous route has little clinical efficiency for emergency situations. (Arch Intern Med 2006;166:391; Ann Intern Med 2002;137:251.) IV vitamin K begins to reverse warfarin in about four to six hours. Therapeutic effect is slightly slower with the oral route. Both routes require about 12 to 24 hours to have a significant clinical effect, sometimes longer for complete effect. Although this vitamin always accompanies FFP/PCC use, the specific use of vitamin K will be covered in detail in a subsequent column because it is not a mainstay for life-threatening hemorrhage.

Prothrombin complex concentrate provides rapid and complete reversal of warfarin coagulopathy. A number of PCCs are available, but these products are largely unknown to most emergency clinicians and rarely used in the ED without consultant input. Although PCCs are the fastest way to reverse warfarin, most EPs have never administered them. PCCs begin to have their effect within 10 to 15 minutes, and are totally effective in about one hour. PCC is a direct replacement of vitamin K-dependent coagulation factors. As with other reversal agents, IV vitamin K should be used in conjunction with PCCs to switch on hepatic synthesis of clotting factors. PCCs were previously used for patients with hemophilia B before the availability of other recombinant factor concentrates. PCCs are used infrequently in this country, and appear to have more widespread use in other countries. There are some differences in various PCC products with regard to the factors they contain, the dosing is not totally known, and there are no comparative studies comparing individual products. A PCC and vitamin K are included in the current guidelines in the U.S. and U.K. for any elevated INR producing life-threatening bleeding.

Table. C

Table. C

Recombinant factor VIIa (rFVIIa) was previously used to treat hemophilia, but it has potential for use for a wide range of bleeding disorders, including warfarin-induced hemorrhage. This product completely normalizes INR in individuals anticoagulated with warfarin within minutes. This is a relatively new use for this product, but a number of patients with ICH have been treated. The role of rFVIIa for warfarin-induced ICH remains unclear (according to the literature as of December 2006), and studies are ongoing. As of this writing, its use is experimental and not standard of care in this country, certainly not in the ED. Keep tuned, however, because this may change in the future.

Back to Top | Article Outline

Reversing Coagulopathy

ICH is the most dreaded and most fatal bleeding-related scenario in anticoagulated patients. The incidence of spontaneous ICH can be as high as one percent per year in patients on warfarin, with a 10-fold increased risk in patients over age 50 compared with non-anticoagulated individuals. Warfarin is the culprit in many patients in a neurosurgical unit with ICH. Hypertension and increasing age are the greatest risk factors for a CNS bleed while taking warfarin. An axiom stated by many authors worth remembering is that ICH is not just related to a massive coagulopathy. It can occur in patients taking Coumadin when the desired INR is well within the therapeutic range. Patients on Coumadin may be at increased risk for ICH if they have white matter changes in their brains due to small vessel vasculopathy, termed leukoaraiosis. In addition, cerebral amyloid angiopathy is another associate of ICH in patients taking warfarin. Neither condition can be evaluated prior to initiating warfarin therapy. As mentioned in my previous column, otherwise asymptomatic or self-limited intracranial microbleeds are common in the elderly, and theoretically would often go unnoticed if the patient were not anticoagulated.

The outcome of ICH in anticoagulated patients is quite dismal, and much worse than the already-discouraging morbidity and mortality in those who bleed intracranially absent this physician-induced coagulopathy. The 30-day mortality may be as high as 60 percent, and few patients escape serious morbidities. Overall, patients on warfarin have an increased risk of initial and spontaneous CNS bleeding, but also a greater chance for hematoma expansion as a result of rebleeding. Although neurosurgical intervention may be appropriate for some patients with ICH, there is a relatively short window of opportunity for warfarin reversal to decrease or limit the hematoma enlargement that initiates a progressive downward spiral. This window of opportunity is often elusive to the clinician or clinically impossible to embrace.

Comment: Although the specific topic of this column is warfarin-associated ICH, similar caveats can be promulgated for a plethora of other serious warfarin-induced hemorrhage, including retroperitoneal, intraocular (not conjunctival), muscle complicated by compartment syndrome, pericardial, or active traumatic bleeding in the presence of hypotension or a precipitous drop in hemoglobin. My comments will usually apply to all of these limb-, life-, and sight-threatening complications of warfarin.

I was quite flummoxed that there are no universally accepted guidelines or any large, rigorously designed, controlled studies that provide absolute answers for this dilemma. The best I could find were general older guidelines from Europe (Brit J Hematol 1998;101:374) and the U.S. (Chest 2001;11922 Suppl.) The executive summary for life-threatening, warfarin-induced bleeding is that PCC and IV vitamin K are the most common recommendations, but FFP has its advocates.

I have seen a number of patients with spontaneous ICH while taking warfarin, usually elderly individuals with hypertension and other co-morbidities. They can present in a very dramatic or deceptively subtle, even clandestine, fashion. The oldster can be a bit confused or unconscious, and may concentrate his complaints on a number of peripheral symptoms (off balance, headache, nausea, vision change, etc); regardless of presentation, they frequently die or are ultimately vegetative.

Maintaining the INR in the therapeutic range may be partly protective, but it's no secret that one can have a spontaneous fatal CNS bleed with an INR just perfect for atrial fibrillation or a replaced cardiac valve. This is important because it supports intervention in an aggressive manner in the presence of ICH by merely knowing the patient is taking Coumadin, and not having to wait for the INR result (about a one-hour turnaround in our lab on a good day). Many colleagues will wait until they obtain the INR before embarking on reversal strategies, and there is even a consensus among some of my colleagues that a leisurely approach is adequate because the INR is not sky high. This issue never has been resolved to my satisfaction, but it makes most sense to me to treat ICH aggressively with anti-warfarin measures. Intervening quickly, often with cowboy bravado, is one tactic. But there is no universal agreement on how timid or brave one should be in the ED with limited data, and a surfeit of omnipresent unknowns. A sooner-the-better argument is, however, only speculation and wishful thinking that has not been realized in currently available data. Of course, there is the often cited risk of thrombotic events from rapid reversal, and the lack of prophylaxis in the patients taking warfarin for heart valve replacements, so it's a dicey issue.

While definitive therapeutic machinations are elusive, the magnitude of the problem of warfarin-mediated ICH is well known. In one study, 26 percent of patients with ICH at a university hospital were taking warfarin. Interestingly, most episodes of warfarin-induced ICH occur during anticoagulation intensities that are considered therapeutic (INR between 2.0 and 3.5). The degree of INR prolongation does correlate with progressive hematoma enlargement, poor functional outcome, and mortality. No degree of anticoagulation is totally safe, however, and no INR elevation is desirable for a patient bleeding into his brain. The mortality rate of patients with ICH who are unconscious, either on admission or before the initiation of active reversal, is over 80 percent. Long-term morbidity is almost 100 percent, and such patients are rarely returned to their pre-ICH level. If the patient is still conscious when the INR is normalized, mortality rates are less but still high, about 30 percent versus the previously cited dismal one.

The expected mortality and morbidity are critical issues in litigation. Most patients and families cannot accept the fact that modern medicine is unable to snatch their loved ones from the ravages of hemorrhagic stroke and return them to normal. A common layperson belief (and a common plaintiff expert's less-than-honest pontification) for horrific outcomes is that something must have gone astray (malpractice, negligence, or increased risk of harm). This is especially extant in this era where the television spots imply that early stroke treatment equates to a miraculous cure or magic bullet. Early recognition and intervention for stroke, the “brain attack” campaign, is basically an excellent concept, albeit often a much too optimistic prognostication. The nuances of such complex issues are often misinterpreted by the grieving family with a vegetative parent.

The typical scenario for ICH is an elderly patient with hypertension on warfarin who has sudden onset of headache, nausea, vomiting, confusion, or obtundation. It's usually a true stroke-like event, although about half the time the initial bleed is minor and the hematoma slowly enlarges during a 12- to 24-hour period. The nefarious minor bleed and elusive initial symptoms are quite subtle for the neophyte to appreciate. (I told you warfarin was a killer drug in the past four columns, so let the clinician beware.) All areas of the brain are fair game for bleeding, but the cerebellum is frequently involved. Some nascent cerebellar and intracerebral hemorrhages are amenable to truncation by reversal of anticoagulation and surgical removal, but most of the time it's a nasty event, associated with an equally nasty outcome. Note to self: Don't bleed into your head while taking warfarin.

Some authors are more zealous than others about the urgency of reversing anticoagulation, but overall all consider ICH a serious medical emergency. Small hemorrhages tend to become big hemorrhages, often over only a few hours' time. A large clot is intuitively worse than a smaller one. Although the exact role of individual reversal agents is unclear, it is universally accepted that merely administering vitamin K is not the ideal approach. Most emergency physicians have little experience with rFVIIa or other prothrombin complex concentrates. My hospital does not even stock a single PCC, but we do have rFVIIa, which is carefully guarded by the pharmacy.

The use of PCCs and now rFVIIa has generally been relegated to the neurosurgeon or hematologist, especially in treating hemophilia. Early consultation is the best hedging maneuver. Most emergency physicians are comfortable with the use of FFP. It may require six to eight units of FFP to reverse the warfarin coagulation deficit, but these approximately two liters of volume can be prohibitive. FFP must be compatible, thawed, and slowly infused. Getting FFP into a patient in under a few hours is technically impossible in the ED. Although the INR may seem to have been reversed, FFP does not totally reverse all the havoc done to the coagulation system by warfarin. I would again emphasize that regardless of the intervention chosen, vitamin K should be in the mix but not the only tactic.

Because current guidelines champion PCCs, it would behoove the prescient and sagacious clinician to be familiar with various products. Most are not familiar with the products, and cannot even name one. My hospital does not keep any PCC in the pharmacy, let alone in the ED pyxis.

The use of rFVIIa, an ungodly expensive preparation, to reverse major intracerebral hemorrhage, has gained a great deal of attention recently. It is still experimental, however, and not standard of care. At this time, rFVIIa appears to have the potential to become a remarkable product. Neurosurgeons tend to use it in all cases of intracranial hemorrhage, even for conditions that put blood into the brain and are not associated with warfarin. Most neurosurgeons will not even attempt an operation until the INR has been completely reversed, and they will often give FFP at the drop of a hat if they are going to enter the brain with their scalpel. Many neurosurgeons have turned to the use of rFVIIa to reduce INR to the normal range almost immediately. Again, full and lasting reversal with rFVIIa can be achieved only with a concomitant use of vitamin K. Changes in mortality or less morbidity are still unproven, but this is an exciting breakthrough.

Back to Top | Article Outline

Comparative Studies

There are very few studies comparing various treatments to help the clinician determine which intervention is best for ICH. In addition, there has been no proven benefit or significant difference in the neurological outcome between various treatment groups, even though the hematoma growth has been staunched and INR corrected more rapidly. Interestingly, no intervention has been associated with an improved neurological outcome when warfarin is the perpetrator. There are no studies directly comparing rFVIIa with PCC or FFP. In one small study that compared FFP with the faster acting PCCs, there was no better outcome despite a more rapid correction of the INR. (Stroke 2001,32;2567.)

So what is the emergency physician to do? There are no specific ED guidelines for reversing anticoagulation in warfarin-associated ICH. The American College of Chest Physicians advocates PCC (or rFVIIa as a possible alternative) in addition to vitamin K for this condition, with the option to use FFP. Although all authors indignantly say it's paramount to minimize delays in administering reversal agents, so far this intervention provides only hope and a modicum of wishful thinking in reducing the high mortality rates associated with warfarin-associated ICH. In other words, it seems like a good idea, but show me the data.

It still appears that the most commonly used intervention in the ED is FFP and vitamin K, interventions that should be available to all EPs. I personally feel comfortable starting FFP/vitamin K on even empiric evidence of ICH in a symptomatic anticoagulated patient, but if I were going to use PCC or rFVIIa, I would certainly involve a consultant early on. Over-reversal, or aggressive use of PCC and rFVIIa, carry the downside of producing thrombosis, including MI, pulmonary embolism, and ischemic strokes.

Comatose or moribund patients presenting with a large hematoma, possibly with signs of herniation, essentially have no chance of recovery. Most neurosurgeons and hematologists will not advocate a specific intervention in this scenario because of the downsides of the products and the expense in the face of a negligible chance of any functional recovery. Bad news all around.

Most neurosurgeons are reluctant to operate to remove a hematoma or dead brain in patients with cerebellar or lobar ICH in the presence of anticoagulation. My experience is that none will intervene until the coagulopathy is reversed. Perhaps the ED is the best place to initiate an all-out aggressive approach, but so far, this has not been the case in my experience.

ICH in a patient with warfarin-induced anticoagulation is a horrific condition with an extremely dismal outcome, regardless of intervention. EPs likely have been sued for the failure to reverse warfarin rapidly in the presence of ICH. (Email me if this happened to you.) If so, the accusing plaintiff expert is on very thin scientific, evidence-based ice. In the long run, no intervention may matter, but one can hope for a better outcome if you can recognize and halt a slowly bleeding intracranial vessel. Obtunded or severely symptomatic patients are likely unsalvageable. Rapid CT scanning, rapid CT interpretation, and rapid INR reversal (FFP, vitamin K, PCCs, rFVIIa) may offer hope, but this is largely just hope. Overall, the institution of FFP/vitamin K is a reasonable emergency physician decision, but other interventions are currently unsettled. Bottom line: Get the CT quickly and involve a consultant early on, and talk to the family so they know you care and are on top of this likely gargantuan insult to the brain.

Prothrombin Complex Concentrates (PCC): PCCs are prepared from pooled human plasma and are used to deactivate agents of viral diseases. They may contain prions or other unknown infective agents. The products available in the U.S. are Autoplex T, Proplex, and Feiba VH. All products contain factor VII but no fibrinogen. Thrombotic complications have been associated with high doses. Although hemophilia results in abnormalities in the aPTT test, this test does not correlate with the clinical improvement offered by these products. Reversal of the INR to normal is the goal of therapy for ICH. Concomitant vitamin K is used to initiate production of clotting factors that have been blocked by warfarin. It is not standard of care for EPs to initiate these products.

Recombinant Coagulation Factor VIIa (rFVIIa): This product contains factor VII produced by transfection of the human factor VII gene into cultured hamster cells. The brand name is Novo-Seven. It is only FDA approved for promoting hemostasis in hemophiliacs with factor VII or factor IX deficiency. It has been the object of off-label use for a variety of hemorrhagic conditions, including ICH (warfarin/other causes). rFVIIa has been poorly studied in conditions other than hemophilia. This product has been associated with thrombogenicity. The final common pathway is increased thrombin generation promoting increased activation and local accumulation of platelets, providing hemostasis. In hemophilia, rFVIIa bypasses inhibitors to factor VII and IX. Administration of rFVIIa has reduced the expansion of intracerebral hematomas, and has minimally improved functional outcome. (New Engl J Med 2005;352:777.) In December 2005, the FDA distributed a letter alerting clinicians of the increased risk of arterial thromboembolic events (MI, stroke) in nonhemophilic patients receiving Novo-Seven. It is not currently standard of care for EPs to initiate the use of this product.

Fresh Frozen Plasma (FFP): FFP is separated from single-donor whole blood and frozen within eight hours of collection. It can be stored for up to one year. FFP contains all the requisite plasma coagulation factors (but no platelets), similar to fresh plasma. FFP must be ABO compatible, but it is not cross-matched. Because it is not in concentrated form, FFP cannot be used in sole source for specific clotting factor deficiency. It's generally safe and effective, but the volume required (4–6 units or 10–15 ml/kg) may limit its use in patients with intracranial hemorrhage. When given to reverse warfarin, concomitant vitamin K should be administered.

© 2007 Lippincott Williams & Wilkins, Inc.