Appropriate use of recombinant factor VIIa: an expanding and unanswered question

Patients with hemophilia A often require replacement therapy with factor VIII (FVIII). Although such treatment may provide adequate management, a percentage of patients will generate FVIII antibodies. In the case of patients with hemophilia A, anti-FVIII represent an appropriate immune response to a “foreign” protein. Certain other patient populations that do not have hemophilia A, however, can also make FVIII autoantibodies. In both cases, such antibodies can inhibit the activity of FVIII and/or decrease its t1/2. Thus, patients with anti-FVIII often become unresponsive to FVIII therapy, even at high doses. Early attempts to overcome anti-FVIII consisted of utilizing FVIII from nonhuman sources, such as porcine FVIII, that may not cross-react with antibodies against human FVIII. Although often effective, this represents only a temporary solution, because additional antibodies arise over time against novel epitopes on the nonhuman FVIII product. An alternate strategy consists of achieving hemostasis through pathways that do not require FVIII activity. Activated prothrombin complex concentrates (APCCs) consist of activated coagulation factors that operate distal to FVIII and can achieve hemostasis in the presence of anti-FVIII inhibitors. Because APCCs do not require injury to promote clot formation, however, its use is associated with unwanted or inappropriate clotting that can result in thrombosis or in extreme cases can lead to disseminated intravascular coagulopathy. Recombinant FVIIa, also called rVIIa, was developed as a novel therapy for patients with anti-FVIII inhibitors. rVIIa activates the extrinsic clotting pathway when it contacts tissue factor exposed at sites of injury. This allows hemostasis to be achieved through extrinsic pathway factors, thus circumventing FVIII antibodies. Because rVIIa requires tissue factor to function, its activity is largely restricted to sites of injury, leading to an apparently lower risk of inducing thrombosis than APCCs. Furthermore, because rVIIa is seen by the recipient immune system as a self-protein (except in the case of congenital FVII deficiency), the development of rVIIa antibodies is not a common problem. Thus, rVIIa represents a significant advancement in the treatment of patients with anti-FVIII (and all other inhibitors of the intrinsic pathway proximal to FX) possessing both excellent efficacy and low rates of unwanted side effects. As is common with new therapies, ongoing attempts have been made to apply rVIIa to the treatment of disease states beyond its approved indications. These applications have ranged from the obvious treatment of FVII deficiency, to the reasonable treatment of multifactor deficiency in the setting of liver failure, to the seemingly illogical but nonetheless successful treatment of inherited disorders of platelet (PLT) function. This latter instance is of particular interest, because the observed efficacy furthers our understanding of the role that FVIIa plays on the PLT surface in cross-regulation between the coagulation cascade and PLT-based hemostasis.1 In the current issue of this journal, Huang and colleagues2 report the use of rVIIa in the treatment of three patients with central nervous system hemorrhages and low FVII. Each of these patients, who had already failed therapy with fresh frozen plasma, responded to rVIIa with partial normalization of elevated international normalized ratio (INR) values and subsequent neurologic improvement. No complications related to rVIIa use were observed. This represents a logical application of a new therapy to a patient population in need of better interventions and corroborates other studies assessing the efficacy of rVIIa in treating intracranial hemorrhage in patients with anti-FVIII inhibitors, FVII deficiencies, and liver failure coagulopathy.3-6 Together, these studies raise important issues of optimal treatment, which may play an essential role in moving the field of coagulation medicine forward. Great care must be taken, however, regarding the strength assigned to conclusions drawn from this and other limited studies on small patient populations lacking either control patient groups or juxtaposing the new therapy with conventional therapy and/or placebo. Perhaps more than many other fields, coagulation medicine suffers from overinterpretation of small studies or single case reports and the lack of proper randomized trials. This problem is partially due to the fact that certain coagulopathies are so rare that accumulating a sufficient number of patients with a given disorder is not feasible. In some cases, however, the results of small studies are viewed as such highly logical extensions of our current understanding that they are not subsequently confirmed with statistical rigor. In this setting, we are best advised to remember that in the world of coagulation medicine, even the most sensible conclusions can turn out to be flawed. In the early days of the bleeding time test it was logical to conclude that testing bleeding after a controlled incision was a reasonable way to predict the risk of intraoperative hemorrhage. And yet, more extensive investigations eventually showed that in patients without a history of a bleeding disorder the bleeding time test has no correlation with risk of operative bleeding.7-10 This experience only goes to underscore that despite a considerable base of knowledge our predictive powers are flawed, common sense is commonly incorrect and Occam's razor is often a very blunt blade. For all of these reasons, small studies that raise the possibility of a conclusion must never serve as a definitive justification for that conclusion but rather should serve as the intellectual nidus for subsequent studies designed to assess the question utilizing trials of sufficient statistical power to provide reliable answers. Such concerns are by no means merely theoretical with regard to rVIIa usage. One example is the increasing use of rVIIa as prophylaxis against hemorrhagic sequelae secondary to minor invasive procedures in patients with prolonged clotting times attributed to hepatic pathology. A bolus of rVIIa will routinely lower a prolonged INR in this patient population. Although the correlation of INR values is well studied with respect to managing coumadin thromboprophylaxis, the relevance of INR values to bleeding risk in liver failure is unclear. Clinicians performing minor invasive procedures in liver failure patients, such as inserting a central venous catheter, typically seek an INR of 1.5 or less before proceeding. The factual basis for using this value is lacking, however. To the contrary there are published data demonstrating that an INR elevated over 1.5 is not alone an independent risk factor for hemorrhagic sequelae in this setting.11, 12 Nonetheless, rVIIa is being used with increasing frequency to lower INRs for this purpose. One might very reasonably ask where the potential harm is in such a practice? At the end of day, however, we must acknowledge that regardless of what may make sense, we simply do not know the effect of our interventions until we have rigorously tested them. Besides the question of whether use of rVIIa really has better efficacy than other interventions, it would certainly not be outside the realm of previous experiences with new therapies to envision a scenario in which rVIIa promoted immediate hemostasis but ultimately led to a worse long-term clinical outcome. Although small studies or single case reports may be of use as general guides of what other practitioners have observed, we are under the obligation to formally test every scenario that is feasible to test. In summary, the field of coagulation medicine is a dynamic and rapidly evolving area of medicine. Over several decades we have progressed from having a limited repertoire of diagnostic tests, two anticoagulants (heparin and coumadin), and rare factor concentrates to a field with a broad variety of diagnostic capabilities and numerous interventions, of which rVIIa may prove to be among the most significant. The problems outlined above concerning use of new interventions in untested patient populations, however, are an inevitable by-product of a rapidly expanding arsenal of therapeutics. We must be ever vigilant in tempering our opinions and actions, keeping in mind what we do not yet know, and we must be relentless in the ongoing pursuit of controlled trials to bring us from the world of what might help to the realm of known efficacy.