A bs tr ac t. n engl j med 363;19 nejm.org november 4,

Transcription

1 The new england journal of medicine established in 1812 november 4, 2010 vol. 363 no. 19 Safety of Recombinant Activated Factor VII in Randomized Clinical Trials Marcel Levi, M.D., Jerrold H. Levy, M.D., Henning Friis Andersen, M.Sc., and David Truloff, D.V.M. A bs tr ac t Background The use of recombinant activated factor VII (rfviia) on an off-label basis to treat life-threatening bleeding has been associated with a perceived increased risk of thromboembolic complications. However, data from placebo-controlled trials are needed to properly assess the thromboembolic risk. To address this issue, we evaluated the rate of thromboembolic events in all published randomized, placebo-controlled trials of rfviia used on an off-label basis. Methods We analyzed data from 35 randomized clinical trials (26 studies involving patients and 9 studies involving healthy volunteers) to determine the frequency of thromboembolic events. The data were pooled with the use of random-effects models to calculate the odds ratios and 95% confidence intervals. Results Among 4468 subjects (4119 patients and 349 healthy volunteers), 401 had thromboembolic events (9.0%). Rates of arterial thromboembolic events among all 4468 subjects were higher among those who received rfviia than among those who received placebo (5.5% vs. 3.2%, P = 0.003). Rates of venous thromboembolic events were similar among subjects who received rfviia and those who received placebo (5.3% vs. 5.7%). Among subjects who received rfviia, 2.9% had coronary arterial thromboembolic events, as compared with 1.1% of those who received placebo (P = 0.002). Rates of arterial thromboembolic events were higher among subjects who received rfviia than among subjects who received placebo, particularly among those who were 65 years of age or older (9.0% vs. 3.8%, P = 0.003); the rates were especially high among subjects 75 years of age or older (10.8% vs. 4.1%, P = 0.02). From the Academic Medical Center, University of Amsterdam, Amsterdam (M.L.); the Emory University School of Medicine, Atlanta (J.H.L.); and Novo Nordisk, Bagsværd, Denmark (H.F.A., D.T.). Address reprint requests to Dr. Levi at the Department of Medicine (F-4), Academic Medical Center, University of Amsterdam, Meibergdreef 9, 1105 AZ Amsterdam, the Netherlands, or at m.m.levi@amc.uva.nl. This article ( /NEJMoa ) was updated on November 16, 2011, at NEJM.org. N Engl J Med 2010;363: Copyright 2010 Massachusetts Medical Society. Conclusions In a large and comprehensive cohort of persons in placebo-controlled trials of rfviia, treatment with high doses of rfviia on an off-label basis significantly increased the risk of arterial but not venous thromboembolic events, especially among the elderly. (Funded by Novo Nordisk.) n engl j med 363;19 nejm.org november 4,

2 T h e n e w e ngl a nd j o u r na l o f m e dic i n e Recombinant activated coagulation factor VII (rfviia) (NovoSeven, Novo Nordisk) is approved for the treatment of bleeding in patients with hemophilia A or B who have inhibiting antibodies to coagulation factor VIII or IX. Indications have been broadened to include the treatment of episodes of bleeding and the prevention of episodes of bleeding related to surgical or invasive procedures in patients with congenital and acquired hemophilia, factor VII deficiency, or Glanzmann s thrombasthenia (the last indication is approved only in Europe). The mechanism of action of rfviia may offer the potential for its use in preventing or treating severe or life-threatening bleeding in patients with other clinical conditions. 1 A growing number of case reports and small, controlled or uncontrolled studies have shown the successful use of rfviia for various clinical indications other than the treatment of hemophilia, including management of severe traumatic injury, control of bleeding during surgery and transplantation, treatment of intracerebral hemorrhage, and management of bleeding due to anticoagulation therapy. 2 Many of the patients who have received rfviia for these indications have been at high risk for death because of frank hemorrhaging, which was a prerequisite for the clinical trials. The primary concern surrounding the use of rfviia is its potential as a hemostatic agent to induce thromboembolic events. It is administered at doses that are up to 1000 times the physiologic level and has a half-life of approximately 2.5 hours. 3 Although rfviia acts by generation of thrombin on thrombin-activated platelets and is theoretically localized to the site of vessel-wall injury, systemic activation of coagulation may occur. 4 An evaluation of 483 published studies (mostly uncontrolled and retrospective) of the use of rfviia in patients with hemophilia, liver disease, trauma or a condition requiring surgery, anticoagulation reversal, and coagulation disorders showed an incidence rate of thromboembolic events of 1 to 2%. 2 O Connell et al. 5 described 168 reports of 185 rfviia-associated thromboembolic events in the Adverse Event Reporting System of the Food and Drug Administration (FDA), involving both approved and off-label uses of rfviia. In 38% of these reported cases, however, other hemostatic agents were used concomitantly with rfviia. Most of the safety data on off-label indications are retrospective and involve subjects with a relatively high risk of adverse events, including thrombosis, making interpretation of these findings difficult. We systematically analyzed rates of thromboembolic events in 35 placebo-controlled trials (involving 4468 persons) of off-label indications for rfviia, including various coagulopathic states, coexisting conditions, and dosing algorithms. In all these trials, rfviia was administered for the treatment or prevention of bleeding. Me thods Identification of Safety Data To evaluate the profile of thromboembolic events associated with the off-label use of rfviia, we used data from two main types of clinical trials: 26 randomized, placebo-controlled trials involving 4119 patients with various clinical conditions, and 9 randomized, placebo-controlled trials involving 349 healthy volunteers. These trials included both Novo Nordisk-sponsored trials (29 trials involving 4064 patients and 263 healthy volunteers) and investigator-initiated trials (6 trials involving 55 patients and 86 healthy volunteers). The investigator-initiated trials were identified through a Medline search of randomized clinical trials (spanning the period from 1996 through 2008) with the use of the terms rfviia, factor VIIa, eptacog alfa, activated, or NovoSeven. The data from the investigator-initiated trials were obtained from the published studies, through direct contact with the authors of the studies, or both (for more details, see the Supplementary Appendix, available with the full text of this article at NEJM.org). These randomized, controlled clinical trials were broadly categorized according to the underlying diagnosis. The resultant seven major categories of bleeding were spontaneous central nervous system bleeding and bleeding due to advanced liver disease, trauma, cardiac surgery, traumatic brain injury, spinal surgery, and other conditions. The study was designed by the first author, and the data were collected by all the authors. Data analysis was performed by the first, second, and last authors, and the statistical analysis was performed by the third author. All authors vouch for the data and the analysis. The first two authors wrote the first draft of the manuscript and made the decision to submit the manuscript for publication with the approval of all the authors n engl j med 363;19 nejm.org november 4, 2010

3 Recombinant Activated Factor VII in Clinical Trials Definition of Terms All reported adverse events that were considered to be possibly or probably thrombotic or embolic in nature (on the basis of the use of standardized terms in the Medical Dictionary for Regulatory Activities [MedDRA]) were reviewed by the industry authors and categorized as an arterial thromboembolic event, a venous thromboembolic event, or an event that was not thromboembolic. Reports of adverse events that used broad terms that may have been suggestive of thromboembolic events were reviewed. Cases in which clear evidence showed that the event was not thrombotic in nature were excluded. All thrombotic events were confirmed by means of objective tests. Arterial thromboembolic events were classified as coronary, cerebrovascular, or other events. Coronary events (MedDRA terms: [acute] myocardial infarction, silent myocardial infarction, unstable angina, and increased troponin) were confirmed by means of electrocardiographic studies, laboratory tests, echocardiographic studies, or a combination of these tests. Cerebrovascular events (MedDRA terms: cerebral infarction, cerebellar infarction, ischemic stroke, cerebrovascular accident, hemiparesis, lacunar infarction, thromboembolic stroke, or ischemic cerebral infarction) were confirmed by means of computed tomography (thereby excluding cerebral hemorrhage as a cause of the symptoms). Other arterial thromboembolic events (MedDRA terms: arterial thrombosis of the legs, embolism, graft thrombosis, hepatic-artery occlusion, hepatic-artery thrombosis, iliac-artery thrombosis, intestinal infarction, intestinal ischemia, intracardiac thrombus, peripheral arterial occlusive disease, renal-artery thrombosis, retinal-artery embolism, splenic infarction, and vascular-graft occlusion) were confirmed by means of vascular imaging. Venous thromboembolic events (MedDRA terms: deep-vein thrombosis, graft thrombosis, jugular-vein thrombosis, mesenteric-vein thrombosis, pelvic venous thrombosis, phlebitis, portalvein thrombosis, pulmonary embolism, renal-vein thrombosis, retinal-vein thrombosis, shunt occlusion, subclavian-vein thrombosis, thrombophlebitis, superficial thrombophlebitis, thrombosis, transverse sinus thrombosis, vena cava thrombosis, venous thrombosis, and venous thrombosis of the legs) were also confirmed by means of vascular imaging. Dose Categories A range of rfviia doses was administered across the 35 trials. The data were categorized into one of three dose groups: less than 80 μg per kilogram of body weight (low), 80 to 120 μg per kilogram (medium), or more than 120 μg per kilogram (high). This categorization of doses was chosen because 80 to 120 μg per kilogram approximates the 90-μg-per-kilogram-dose recommended in the prescribing information for treating episodes of bleeding in patients with hemophilia A or B who have inhibiting antibodies to coagulation factors VIII or IX. In trials with crossover or multiple-dose designs, the subject was assigned to the group according to the highest dose received. Statistical Analysis The statistical analyses focused on the proportion of subjects who had thromboembolic events, arterial thromboembolic events, or venous thromboembolic events. The statistical analyses were primarily performed on data collected from randomized, controlled clinical trials involving patients. The data from clinical trials involving healthy volunteers are presented separately, with the exception of Table 1, which shows rates of thromboembolic events among all persons in the 35 trials. For descriptive statistics, the rate of thromboembolic events was defined as the number of patients with events as a proportion of the number of patients who received a study drug. P values, odds ratios, and 95% confidence intervals were calculated with the use of logistic regression and SAS software (version 9.2), with adjustment for the type of bleeding, age, and rfviia treatment. The type of bleeding and age were included because it was known that they are both predictors of thromboembolic events If a statistically significant difference was observed between the event rate among patients who received rfviia and the rate among patients who received placebo, additional analyses were performed according to the type of adverse event, the type of bleeding, and the age group. In addition, P values, modelbased odds ratios, and confidence intervals were calculated only in subgroups in which there were enough events (>10 events in total) for the analysis to be meaningful. The association of dose with the risk of a thromboembolic event was also evaluated by n engl j med 363;19 nejm.org november 4,

4 T h e n e w e ngl a nd j o u r na l o f m e dic i n e Table 1. Population and Dose-Group Distribution in Placebo-Controlled Trials of rfviia.* Population and Cause of Bleeding No. of Studies Mean Age Placebo (N = 1653) rfviia (N = 2815) Total (N = 4468) rfviia, <80 µg/kg (N = 940) rfviia, µg/kg (N = 1086) rfviia, >120 µg/kg (N = 789) Reference yr no. of subjects no. of subjects (%) no. of subjects Patients (92.2) Spontaneous central nervous system bleeding (31.3) Mayer et al. 6-9 Liver disease (27.8) Bosch et al., 10,11 Carreno et al., 12 Lodge et al., 13,14 Planinsic et al., 15 Shao et al. 16 Trauma (18.7) Boffard et al. 17 Cardiac surgery (6.0) Diprose et al., 18 Ekert et al., 19 Gill et al. 20 Traumatic brain injury (2.2) Narayan et al. 21 Spinal surgery (1.1) Sachs et al. 22 Other causes (5.1) Chuansumrit et al., 23 Friederich et al., 24 Pihusch et al., 25 Raobaikady et al. 26 Healthy volunteers (7.8) Bijsterveld et al., 27,28 Erhardtsen et al., 29 Fridberg et al., 30 Friederich et al., 31 Wolzt et al. 32 * References are provided for the trials that have been published. The remaining data are provided in the Supplementary Appendix. One study included children younger than 1 year of age. Studies included patients with prostatectomy, pelvic or pelvic acetabular fracture reconstruction, dengue hemorrhagic fever, bleeding after hematopoietic stem-cell transplantation (two patients), and severe bleeding associated with vitamin K antagonist therapy. Four studies involved healthy volunteers who received anticoagulants (fondaparinux, 27 idraparinux, 28 acenocoumarol, 29 and warfarin 30 ). One study involved persons who had received the antiplatelet agent clopidogrel, one study involved persons who had received tissue factor inhibitor (nematode anticoagulant protein c2), 31 one study involved persons who had received a thrombin inhibitor (melagatran), 32 one study was a pharmacokinetics study, 31 and one study involved persons who had undergone punch biopsy n engl j med 363;19 nejm.org november 4, 2010

5 Recombinant Activated Factor VII in Clinical Trials means of logistic regression. Statistical analyses were performed by means of logistic regression with dose as a covariate. Since the administered dose of rfviia varies according to the type of bleeding being treated (with some types of bleeding, such as trauma-related bleeding, requiring the use of higher doses than other types, such as spontaneous central nervous system bleeding), the dose given is confounded by the type of bleeding. To minimize this problem, the analyses included only studies in which patients had received a study drug in at least two of the dose categories. Most of the studies in which patients were randomly assigned to at least two of the dose categories involved spontaneous central nervous system bleeding, which was associated with a higher event rate than other indications. Therefore, the effect of the dose was examined only among patients with this type of bleeding. R esult s Demographic Characteristics of the Subjects A total of 4468 subjects (1653 subjects who received placebo and 2815 subjects who received rfviia) were enrolled in 35 randomized clinical trials encompassing various clinical scenarios. Most of the subjects were patients with spontaneous central nervous system bleeding (31.3%), advanced liver disease (27.8%), or trauma (18.7%) (Table 1). Approximately 45% (2026) of the subjects received either low doses of rfviia (<80 µg per kilogram) or medium doses (80 to 120 µg per kilogram). Healthy volunteers and patients with bleeding from trauma or cardiac surgery were younger than patients with other causes of bleeding. However, regardless of the type of trial or the cause of bleeding, the mean (±SD) age was similar in the placebo groups (50±19 years) and the rfviia-treated groups (53±19 years). Thromboembolic Events in Patients Age and disease state are known indicators of thrombotic risk The odds ratios for thromboembolic events among patients who received rfviia as compared with patients who received placebo were calculated by adjusting for age and the type of bleeding. The rate of thromboembolic events was 10.2% among patients who received rfviia as compared with 8.7% among patients who received placebo (odds ratio, 1.17; 95% confidence interval [CI], 0.94 to 1.47; P = 0.16) (Table 2). When we evaluated the rates of thromboembolic events according to type (i.e., arterial thromboembolic events vs. venous thromboembolic events), we found a significantly higher proportion of arterial thromboembolic events in the rfviia-treated groups than in the placebo groups (5.5% vs. 3.2%; odds ratio, 1.68; 95% CI, 1.20 to 2.36; P = 0.003). No significant difference was observed between patients who received rfviia and patients who received placebo with respect to the incidence of venous thromboembolic events (5.3% and 5.7%, respectively; odds ratio with rfviia, 0.93; 95% CI, 0.70 to 1.23; P = 0.61). When the analyses were limited to data from industry-sponsored, placebocontrolled trials, the results were similar (odds ratio for arterial thromboembolic events, 1.72; 95% CI, 1.22 to 2.43; P = 0.002; odds ratio for venous thromboembolic events, 0.92; 95% CI, 0.70 to 1.22; P = 0.58). Further analysis of arterial thromboembolic events in the rfviia-treated groups indicated that 76 of 141 events (53.9%) were coronary thromboembolic events, and the frequency of coronary thromboembolic events in the rfviia-treated groups was higher than that in the placebo groups (odds ratio, 2.39; 95% CI, 1.39 to 4.09; P = 0.002). There was also a trend toward an increased rate of cerebrovascular thromboembolic events among patients who received rfviia (Table 3). Analysis of arterial thromboembolic events according to age, with adjustment for the type of bleeding, showed that patients who were 18 years of age or older had more arterial thromboembolic events than patients younger than 18 years of age (Table 4). When we analyzed the treatment effect, we found that the rates of arterial thromboembolic events were higher among patients who received rfviia than among patients who received placebo, particularly among patients who were 65 years of age or older (9.0% vs. 3.8%; odds ratio, 2.43; 95% CI, 1.34 to 4.41; P = 0.003) and were most pronounced in the subgroup of rfviia-treated patients who were 75 years of age or older (odds ratio, 3.02; 95% CI, 1.22 to 7.48; P = 0.02) (Table 4). Analysis of the rates of arterial thromboembolic events according to the type of bleeding, with adjustment for age, showed that among patients enrolled in trials of rfviia to mitigate spontaneous central nervous system bleeding, the rates of arterial thromboembolic events were significantly higher in the rfviia-treated groups than in the placebo groups (8.6% vs. 5.4%; odds ratio, 1.67; 95% CI, 1.03 to 2.69; P = 0.04) (Table 5). The odds ratios were similar and were higher than n engl j med 363;19 nejm.org november 4,

6 T h e n e w e ngl a nd j o u r na l o f m e dic i n e Table 2. Odds Ratios for Thromboembolic Events. Thromboembolic Event rfviia (N = 2583) Placebo (N = 1536) Odds Ratio (95% CI)* P Value number (percent) All events 264 (10.2) 134 (8.7) 1.17 ( ) 0.16 Arterial events 141 (5.5) 49 (3.2) 1.68 ( ) Venous events 137 (5.3) 88 (5.7) 0.93 ( ) 0.61 * Odds ratios were calculated by means of logistic regression with adjustment for age and type of bleeding. The percentage of thromboembolic events was calculated as the number of patients with events as a proportion of the number of patients who received the assigned study drug. Table 3. Arterial Thromboembolic Events with a Rate Greater Than 0.5%. Variable rfviia (N = 2583) Placebo (N = 1536) Odds Ratio (95% CI)* P Value number (percent) All arterial thromboembolic events 141 (5.5) 49 (3.2) 1.68 ( ) Coronary events 76 (2.9) 17 (1.1) 2.39 ( ) Acute coronary syndromes 57 (2.2) 11 (0.7) Increased troponin level 19 (0.7) 6 (0.4) Cerebrovascular events 45 (1.7) 20 (1.3) 1.27 ( ) 0.39 Cerebral infarction 44 (1.7) 19 (1.2) Hemiparesis 1 (<0.1) 1(<0.1) * Odds ratios were calculated by means of logistic regression with adjustment for age and type of bleeding. Computed tomographic confirmation of thrombosis was not obtained for the two cases of hemiparesis. 1 for types of bleeding other than central nervous system bleeding, but they were not significant, possibly because of the lower numbers of patients for most of the other types of bleeding, as well as a lower incidence of arterial thromboembolic events in the placebo group (Table 5). These analyses, however, do not account for the various doses of rfviia that were used, which may have had an effect on the risk of thromboembolic events, especially in studies involving patients with central nervous system bleeding. The rates of arterial thromboembolic events were 5.4% among 23 patients with spontaneous central nervous system bleeding who received placebo, 6.0% among 26 patients who received less than 80 µg of rfviia per kilogram, 10.3% among 45 patients who received 80 to 120 µg of rfviia per kilogram, and 11.9% among 13 patients who received more than 120 µg of rfviia per kilogram. When the dose was considered as a covariate (adjusted for age) of rates of arterial thromboembolic events among patients with central nervous system bleeding who received rfviia, this apparent dose-dependent effect of rfviia treatment was significant (P = 0.02). Thromboembolic Events in Healthy Volunteers Thromboembolic events were also analyzed in 349 healthy volunteers in five Novo Nordisk-sponsored and four investigator-initiated placebo-controlled trials. The rate of thromboembolic events was 0.9% among both healthy volunteers who received rfviia and healthy volunteers who received placebo. None of the thromboembolic events were arterial in nature. All three venous thromboembolic events (two in healthy volunteers who received rfviia and one in a healthy volunteer who received placebo) were cases of phlebitis. Discussion This comprehensive study of the safety profile of rfviia for off-label treatment of episodes of bleeding involved 4468 subjects enrolled in 35 placebocontrolled clinical trials. We found an increased 1796 n engl j med 363;19 nejm.org november 4, 2010

7 Recombinant Activated Factor VII in Clinical Trials Table 4. All Arterial Thromboembolic Events, According to Age. Age Group rfviia Placebo Odds Ratio (95% CI)* P Value no./total no. (%) <18 yr 1/70 (1.4) 1/51 (2.0) yr 73/1764 (4.1) 34/1107 (3.1) 1.36 ( ) yr 67/742 (9.0) 14/372 (3.8) 2.43 ( ) yr 33/427 (7.7) 8/225 (3.6) 2.12 ( ) yr 34/315 (10.8) 6/147 (4.1) 3.02 ( ) 0.02 * Odds ratios were calculated by means of logistic regression with adjustment for indication. The P value was not calculated for the first age group because there were only two events. The percentage of thromboembolic events was calculated as the number of patients with events as a proportion of the number of patients who received a study drug. risk of arterial thromboembolic events among patients who received off-label rfviia as compared with patients who received placebo for bleeding episodes. The rate of coronary arterial thromboembolic events among the patients who received rfviia was 2.6 times as high as the rate among patients who received placebo. Age was also associated with an increase in the risk of arterial thromboembolic events after rfviia treatment, with an odds ratio of 2.4 among patients 65 years of age or older and 3.0 among patients 75 years of age or older. The rates of arterial thromboembolic events were also higher among patients who received higher doses of rfviia. Abshire reviewed the efficacy and safety of rfviia in patients with hemophilia who had congenital or acquired inhibitory antibodies against factor VIII or IX; that study was based on data from clinical trials and spontaneous postmarketing surveillance reports. 37 With approximately 800,000 standard doses of rfviia administered during the period from May 2003 through December 2006, a total of 30 thromboembolic events were reported, 6 of which were fatal. Spontaneous reports of 71 adverse events included 14 thromboembolic events (20%), with 2 of 34 reported deaths due to a thromboembolic event. 37 Solicited reports of 40 adverse events included 5 thromboembolic events 12%), with 1 of 32 deaths due to a thromboembolic event. A comprehensive overview of thrombotic adverse events, based on the Med- Watch pharmacovigilance program, also showed a low incidence of thrombotic complications associated with the use of rfviia (24.6 events per 100,000 infusions), although the risk of thrombosis was higher among patients treated with rfviia than among those treated with other hemostatic agents. 38 As the authors of the overview correctly state, differences in adverse-event reporting practices among the various compounds may have contributed to the observed difference in the rate of thrombotic events. Taken together, the data show that the use of rfviia for an approved indication (i.e., the treatment of episodes of bleeding in patients with hemophilia) is associated with a rate of thromboembolic events of less than 1%. O Connell et al. 5 reviewed events from the FDA s Adverse Event Reporting System during the period from 1999 through 2005 and identified 185 thromboembolic events, the majority of which occurred in patients with off-label indications for rfviia. However, in this patient population, 38% of the patients received other concomitant therapies, and the study had the inherent limitations of passive surveillance. In a systematic review of all published and unpublished case reports, case series, and clinical studies from 1996 through 2004 that focused on the efficacy and safety of rfviia in patients with or without coagulation disorders, including patients with trauma and those who had undergone surgery, the incidence rate of thrombosis was 1 to 2%. 2 The difference between the rate of thromboembolic events reported in that review and the results of the pooled analysis presented here may be due to the fact that the previous review included mostly patients with congenital or acquired hemophilia or liver failure, whereas the present review includes clinical trials involving patients with other causes of hemorrhage. Our pooled analysis was conducted with a large safety data set obtained from placebo-controlled trials of rfviia. The inclusion of control groups allowed for the proper evaluation of thromboem- n engl j med 363;19 nejm.org november 4,

8 T h e n e w e ngl a nd j o u r na l o f m e dic i n e Table 5. All Arterial Thromboembolic Events, According to Cause of Bleeding.* Cause of Bleeding No. of Studies rfviia Placebo no./total no. (%) Odds Ratio (95% CI) P Value Reference Spontaneous central nervous system 5 84/974 (8.6) 23/423 (5.4) 1.67 ( ) 0.04 Mayer et al. 6-9 bleeding Advanced liver disease 7 23/795 (2.9) 6/449 (1.3) 2.19 ( ) 0.09 Bosch et al., 10,11 Carreno et al., 12 Lodge et al., 13,14 Planinsic et al., 15 Shao et al. 16 Trauma 3 19/409 (4.6) 15/428 (3.5) 1.39 ( ) 0.36 Boffard et al. 17 Cardiac surgery 3 9/153 (5.9) 4/114 (3.5) 1.59 ( ) 0.45 Diprose et al., 18 Ekert et al., 19 Gill et al. 20 Traumatic brain injury 1 2/61 (3.3) 1/36 (2.8) Narayan et al. 21 Spinal surgery 1 1/36 (2.8) 0/13 Sachs et al. 22 Other causes 6 3/155 (1.9) 0/73 Chuansumrit et al., 23 Friederich et al., 24 Pihusch et al., 25 Raobaikady et al. 26 * References are provided for the trials that have been published. The remaining data are provided in the Supplementary Appendix. Odds ratios were calculated by means of logistic regression with adjustment for age. Odds ratios were not calculated in instances with very few events. The percentage of thromboembolic events was calculated as the number of patients with events as a proportion of the number of patients who received a study drug. bolic events after the administration of rfviia. Furthermore, the data presented in our analysis were obtained from clinical trials involving patients with bleeding disorders other than hemophilia in order to establish a clear understanding of the safety of rfviia in patients with various types of bleeding. This design is important, because many of these patients also received multiple transfusions that may have contributed to adverse outcomes, especially in observational studies. 39 In a previous study, we evaluated safety data obtained from 13 clinical trials of rfviia in patients with coagulopathy due to anticoagulant therapy, cirrhosis, or severe traumatic injury, and we reported thromboembolic events in 23 of 430 patients who received placebo (5.3%) and in 45 of 748 patients who received active treatment (6.0%). No significant differences were noted between patients who received placebo and patients who received rfviia, on the basis of data from individual trials or pooled data (P = 0.57). 40 The limitations of the current data set include the relatively small individual study samples, differences in indications (e.g., central nervous system bleeding and bleeding from liver disease, trauma, and other causes), and the fact that the studies were conducted over a 12-year span. However, the wide scope of indications may be considered important, since subjects with or without coagulopathies were evaluated. The variation in dosing was taken into consideration by categorizing subjects into three dose groups. Confounding factors such as age and sex were also taken into consideration in the statistical analysis. The data presented provide a systematic evaluation of rates of thromboembolic events in placebo-controlled trials of rfviia. It is important to note that central nervous system bleeding occurs in an older population with an inherently increased risk of thromboembolic events. Therefore, risk-benefit considerations should be evaluated before administering any hemostatic agent. Supported by Novo Nordisk. Dr. Levy reports serving on a steering committee for Novo Nordisk, and Drs. Andersen and Truloff report being employees of and having equity interest in Novo Nordisk. No other potential conflict of interest relevant to this article was reported. Disclosure forms provided by the authors are available with the full text of this article at NEJM.org. We thank the following Novo Nordisk employees: Brett Skolnick, Ph.D., Sheba Mathew, Ph.D., and Abha Chandra, Ph.D., for valuable contributions made during the development of an earlier version of the manuscript, and Ming Ying Ching, M.S., Naum Khutoryansky, Ph.D., and Anders Rosholm, Ph.D., for providing statistical support n engl j med 363;19 nejm.org november 4, 2010

9 Recombinant Activated Factor VII in Clinical Trials References 1. Hedner U. Factor VIIa and its potential therapeutic use in bleeding-associated pathologies. Thromb Haemost 2008;100: Levi M, Peters M, Büller HR. Efficacy and safety of recombinant factor VIIa for treatment of severe bleeding: a systematic review. Crit Care Med 2005;33: Villar A, Aronis S, Morfini M, et al. Pharmacokinetics of activated recombinant coagulation factor VII (NovoSeven) in children vs. adults with haemophilia A. Haemophilia 2004;10: Hedner U. Mechanism of action of factor VIIa in the treatment of coagulopathies. Semin Thromb Hemost 2006;32: Suppl 1: O Connell KA, Wood JJ, Wise RP, Lozier JN, Braun MM. Thromboembolic adverse events after use of recombinant human coagulation factor VIIa. JAMA 2006; 295: Mayer SA, Brun NC, Begtrup K, et al. Recombinant activated factor VII for acute intracerebral hemorrhage. N Engl J Med 2005;352: Mayer SA, Brun NC, Broderick J, et al. Safety and feasibility of recombinant factor VIIa for acute intracerebral hemorrhage. Stroke 2005;36: Mayer SA, Brun NC, Broderick J, et al. Recombinant activated factor VII for acute intracerebral hemorrhage: US phase IIA trial. Neurocrit Care 2006;4: Mayer SA, Brun NC, Begtrup K, et al. Efficacy and safety of recombinant activated factor VII for acute intracerebral hemorrhage. N Engl J Med 2008;358: Bosch J, Thabut D, Bendtsen F, et al. Recombinant factor VIIa for upper gastrointestinal bleeding in patients with cirrhosis: a randomized, double-blind trial. Gastroenterology 2004;127: Bosch J, Thabut D, Albillos A, et al. Recombinant factor VIIa for variceal bleeding in patients with advanced cirrhosis: a randomized, controlled trial. Hepatology 2008;47: Carreno V, Messnern M, Arrieta J, Berthier A, Schelde P, DeMello G. The effect of recombinant factor VIIa (Novo- Seven) on haemorrhage following dental surgery in patients with liver cirrhosis a randomised placebo controlled study. Presented at the XVIII Congress of the International Society on Thrombosis and Haemostasis, July 6 12, 2001:P Lodge JP, Jonas S, Jones RM, et al. Efficacy and safety of repeated perioperative doses of recombinant factor VIIa in liver transplantation. Liver Transpl 2005; 11: Lodge JP, Jonas S, Oussoultzoglou E, et al. Recombinant coagulation factor VIIa in major liver resection: a randomized, placebo-controlled, double-blind clinical trial. Anesthesiology 2005;102: Planinsic RM, van der Meer J, Testa G, et al. Safety and efficacy of a single bolus administration of recombinant factor VIIa in liver transplantation due to chronic liver disease. Liver Transpl 2005;11: Shao YF, Yang JM, Chau GY, et al. Safety and hemostatic effect of recombinant activated factor VII in cirrhotic patients undergoing partial hepatectomy: a multicenter, randomized, double-blind, placebo-controlled trial. Am J Surg 2006;191: Boffard KD, Riou B, Warren B, et al. Recombinant factor VIIa as adjunctive therapy for bleeding control in severely injured trauma patients: two parallel randomized, placebo-controlled, double-blind clinical trials. J Trauma 2005;59: Diprose P, Herbertson MJ, O Shaughnessy D, Gill RS. Activated recombinant factor VII after cardiopulmonary bypass reduces allogeneic transfusion in complex non-coronary cardiac surgery: randomized double-blind placebo-controlled pilot study. Br J Anaesth 2005;95: Ekert H, Brizard C, Eyers R, Cochrane A, Henning R. Elective administration in infants of low-dose recombinant activated factor VII (rfviia) in cardiopulmonary bypass surgery for congenital heart disease does not shorten time to chest closure or reduce blood loss and need for transfusions: a randomized, double-blind, parallel group, placebo-controlled study of rfviia and standard haemostatic replacement therapy versus standard haemostatic replacement therapy. Blood Coagul Fibrinolysis 2006;17: Gill R, Herbertson M, Vuylsteke A, et al. Safety and efficacy of recombinant activated factor VII: a randomized placebo-controlled trial in the setting of bleeding after cardiac surgery. Circulation 2009;120: Narayan RK, Maas AI, Marshall LF, Servadei F, Skolnick BE, Tillinger MN. Recombinant factor VIIA in traumatic intracerebral hemorrhage: results of a doseescalation clinical trial. Neurosurgery 2008;62: Sachs B, Delacy D, Green J, et al. Recombinant activated factor VII in spinal surgery: a multicenter, randomized, double-blind, placebo-controlled, dose-escalation trial. Spine 2007;32: Chuansumrit A, Wangruangsatid S, Lektrakul Y, Chua MN, Zeta Capeding MR, Bech OM. Control of bleeding in children with Dengue hemorrhagic fever using recombinant activated factor VII: a randomized, double-blind, placebo-controlled study. Blood Coagul Fibrinolysis 2005;16: Friederich PW, Henny CP, Messelink EJ, et al. Effect of recombinant activated factor VII on perioperative blood loss in patients undergoing retropubic prostatectomy: a double-blind placebo-controlled randomised trial. Lancet 2003;361: [Erratum, Lancet 2003;361:1138.] 25. Pihusch M, Bacigalupo A, Szer J, et al. Recombinant activated factor VII in treatment of bleeding complications following hematopoietic stem cell transplantation. J Thromb Haemost 2005;3: Raobaikady R, Redman J, Ball JA, Maloney G, Grounds RM. Use of activated recombinant coagulation factor VII in patients undergoing reconstruction surgery for traumatic fracture of pelvis or pelvis and acetabulum: a double-blind, randomized, placebo-controlled trial. Br J Anaesth 2005;94: Bijsterveld NR, Moons AH, Boekholdt SM, et al. Ability of recombinant factor VIIa to reverse the anticoagulant effect of the pentasaccharide fondaparinux in healthy volunteers. Circulation 2002;106: Bijsterveld NR, Vink R, van Aken BE, et al. Recombinant factor VIIa reverses the anticoagulant effect of the long-acting pentasaccharide idraparinux in healthy volunteers. Br J Haematol 2004;124: Erhardtsen E, Nony P, Dechavanne M, French P, Biossel JP, Hedner U. The effect of recombinant factor VIIa (NovoSeven) in healthy volunteers receiving acenocoumarol to an International Normalized Ratio above 2.0. Blood Coagul Fibrinolysis 1998; 9: Fridberg MJ, Hedner U, Roberts HR, Erhardtsen E. A study of the pharmacokinetics and safety of recombinant activated factor VII in healthy Caucasian and Japanese subjects. Blood Coagul Fibrinolysis 2005;16: Friederich PW, Levi M, Bauer KA, et al. Ability of recombinant factor VIIa to generate thrombin during inhibition of tissue factor in human subjects. Circulation 2001;103: Wolzt M, Levi M, Sarich TC, et al. Effect of recombinant factor VIIa on melagatraninduced inhibition of thrombin generation and platelet activation in healthy volunteers. Thromb Haemost 2004;91: Abbott RD, Curb JD, Rodriguez BL, et al. Age-related changes in risk factor effects on the incidence of thromboembolic and hemorrhagic stroke. J Clin Epidemiol 2003;56: Heit JA. Venous thromboembolism: disease burden, outcomes and risk factors. J Thromb Haemost 2005;3: Rosendaal FR. Risk factors for venous thrombotic disease. Thromb Haemost 1999;82: Samama MM. An epidemiologic study of risk factors for deep vein thrombosis in medical outpatients: the Sirius study. Arch Intern Med 2000;160: Abshire T. Safety update on recombinant factor VIIa in the treatment of congenital and acquired hemophilia. Semin Hematol 2008;45:Suppl:S3-S Aledort LM. Comparative thrombotic event incidence after infusion of recombinant factor VIIa versus factor VIII inhibi- n engl j med 363;19 nejm.org november 4,

10 Recombinant Activated Factor VII in Clinical Trials tor bypass activity. J Thromb Haemost 2004;2: Murphy GJ, Reeves BC, Rogers CA, Rizvi SI, Culliford L, Angelini GD. Increased mortality, postoperative morbidity, and cost after red blood cell transfusion in patients having cardiac surgery. Circulation 2007;116: Levy JH, Fingerhut A, Brott T, Langbakke IH, Erhardtsen E, Porte RJ. Recombinant factor VIIa in patients with coagulopathy secondary to anticoagulant therapy, cirrhosis, or severe traumatic injury: review of safety profile. Transfusion 2006; 46: Copyright 2010 Massachusetts Medical Society n engl j med 363;19 nejm.org november 4, 2010