National PBM Drug Monograph

Topical Thrombin Products

National PBM Drug Monograph Topical Thrombin Products (Bovine-Thrombin JMI®, Pooled Human Plasma-Evithrom®, Recombinant Human-Recothrom®) December 2008 VHA Pharmacy Benefits Management Services and the Medical Advisory Panel The purpose of VACO Pharmacy Benefits Management Services drug monographs is to provide a comprehensive drug review for making formulary decisions. These documents will be updated when new clinical data warrant additional formulary discussion. Documents will be placed in the Archive section of the PBM website when the information is deemed to be no longer current.

Executive Summary Topical thrombin, regardless of source, is FDA approved as an aid to hemostasis whenever oozing blood and minor bleeding from capillaries and small venules is accessible. In various types of surgery, solutions of thrombin may be used in conjunction with an absorbable gelatin sponge, USP for hemostasis. The off- label use of thrombin to assist in the thrombosis of vascular pseudoaneurysm has become more common over the past several years.

Efficacy  In prospective clinical trials, human thrombin and recombinant human thrombin were similarly effective to bovine thrombin in achieving hemostasis within 10 minutes in a variety of surgical settings/surgical types.  Secondary efficacy endpoints such as intraoperative blood loss, units of blood transfused, procedure duration, time in specialty unit and length of stay did not differ between groups.  Available evidence supports the use of ultrasound-guided thrombin injection for postcatheterization pseudoaneurysm (PSA). Some investigators recommend its use as first-line while others recommend reserving it for postcatheterization PSAs failing ultrasound guided compression therapy. Safety  In the majority of publications summarized in table 3 (pages 8-10), bovine thrombin was observed to produce a greater immunogenic response in comparison to human or recombinant human thrombin. However, the antibodies that developed after exposure to bovine thrombin were either not tested to determine the ability of these antibodies to neutralize native human thrombin or that information was not reported.  In the publications reporting adverse clinical events associated with bovine thrombin in comparison to human or recombinant human thrombin sources, the safety profiles were reportedly comparable.  Several authors have recommended avoidance of bovine thrombin in patients known to possess antibodies against coagulation factors (although no commercial test is available) and in those patients with a history of surgery and/or a known or suspected exposure to bovine thrombin. However, the available evidence does not support a significantly increased risk for post-operative complications with bovine thrombin when compared to human or recombinant thrombin. Although comparative data in patients with multiple exposures to any thrombin product is lacking.  It is important to note that most patients participating in the reports/studies have not had multiple exposures to thrombin from any source and the trials are relatively small. Since patients having surgery for peripheral artery bypass (PAB) or arteriovenous graft placement for dialysis are expected to undergo multiple surgical interventions, Weaver, et al.12, analyzed a subgroup of the Chapman study11 to determine the clinical safety of bovine vs. recombinant human thrombin in these patients. In the subgroup analysis, no differences in peri-operative transfusions, death, graft complications or thrombosis, serious adverse events, peri or post-operative bleeding or thromboembolic complications or differences in laboratory values (PT, aPTT, INR) were observed. A post-hoc analysis of bleeding, thromboembolic events, hypersensitivity or abnormal aPTT showed a numeric increase in events in the bovine thrombin antibody positive group that was not statistically significant when compared to bovine thrombin antibody negative group.  Because the FDA had received reports of serious adverse bleeding events associated with bovine thrombin, in 1996, all products containing bovine thrombin were required by the FDA to have a boxed warning of the risk for possible coagulation abnormalities in some patients (see boxed warning and discussion on page 5-6). There is no boxed warning for human or recombinant topical thrombin.

Updated versions may be found at www.pbm.va.gov or vaww.pbm.va.gov 1 December 2008 Topical Thrombin Products

 There is one ongoing observational study assessing the impact of exposure to bovine thrombin (Thrombin JMI) on coagulation parameters (e.g. aPTT) 48 hours after surgery in patients with a likelihood of prior exposure to bovine thrombin within the previous four years.36  The FDA reviewer had questions regarding the safety and immunogenicity of multiple exposures to recombinant human thrombin.  Because there is limited efficacy and safety data in patients having more than one exposure to recombinant human thrombin, the manufacturers of the recombinant product have committed to conducting a phase 4 study of the safety and immunogenicity of re-exposure to their product.

Dosage and Administration  The dose and administration of topical thrombin is dependent upon the amount of bleeding, the site of bleeding and preference of the surgeon. Doses for ultrasound-guided thrombin injection for postcatheterization PSA range from 50-1000+ IU.

Precautions/Contraindications Precautions Bovine (Thrombin JMI®) Human (Evithrom®) Recombinant (Recothrom®)  Boxed warning (see page 6)  May carry a risk of transmitting  Potential for allergic reactions in  Pregnancy category C infectious agents such as viruses persons allergic to snake  Allergic reactions may occur in and theoretically, the Creutzfeldt proteins patients known to be sensitive to –Jakob disease (CJD) agent,  Potential risk of thrombosis is bovine materials despite manufacturing steps to absorbed systemically  Inhibitory antibodies may reduce these risks.  Limited data are available on develop in a small percentage of  Anaphylactic reactions may repeat exposure (n=6) people occur  Safety and effectiveness not  Potential risk of thrombosis if established in children absorbed systemically Contraindications Bovine (Thrombin JMI®) Human (Evithrom®) Recombinant (Recothrom®)  Do not use in persons known to  Do not inject into the circulatory  Do not inject directly into the be sensitive to any of its system. circulatory system. components and/or to material  Do not use in individuals known  Do not use for the treatment of of bovine origin to have anaphylactic or severe massive or brisk arterial  Must not be injected or system reaction to human blood bleeding. otherwise allowed to enter large products.  Do not administer to patients blood vessels. Extensive intra-  Do not use for the treatment of with known hypersensitivity to vascular clotting and even death severe or brisk arterial bleeding. Recothrom, any components of may result. Recothrom or hamster proteins.

Conclusions  Topical thrombin (powder) is a listed on the VANF; however the source of thrombin is not identified. Both bovine and recombinant human thrombins are available in powder form. Human thrombin is available as a solution.  The efficacy of all three thrombin products in achieving hemostasis within 10 minutes is similar.  Although bovine thrombin (Thrombin JMI®) has been shown to be more immunogenic than human (Evithrom®) or recombinant human (Recothrom®) thrombin, the evidence does not support a significantly increased risk for adverse post-operative complications in prospective clinical trials. It is important to note that most of the patients in the trials have not received multiple exposures to thrombin from any source and the trials are relatively small (n=300-453).

Introduction Thrombin was first described in 1892 as “fibrin ferment” a substance that reacted with fibrinogen to form a blood clot.1 Since the 1940’s, topical thrombin has been used to stop bleeding in surgical procedures. In the 1960’s, interest in using topical thrombin as a surgical hemostat increased and in the late 1970’s, following FDA approval, demand for the use of the bovine derived topical thrombin products increased even further. Until mid 2007, bovine thrombin was the only “stand alone” topical thrombin product available. In August

2007, a pooled human plasma thrombin product was FDA approved and in January 2008, a recombinant human thrombin product received FDA approval.

The purposes of this monograph are three-fold (1) to consider the available evidence of efficacy, safety, cost and other pharmaceutical considerations that are relevant to evaluating these three topical thrombin products and their place on the VA National Formulary; (2) to determine if the clinical evidence supports substantive differences in the safety and/or efficacy between the three products; and (3) to determine if the clinical evidence supports the use of one agent over the other in individual patients and/or surgeries.

Pharmacology/Pharmacokinetics Thrombin (activated factor II, IIa) is a coagulation protein with multiple actions or effects within the clotting cascade. Among those effects, thrombin converts fibrinogen to fibrin and activates factor XI, factor V and factor VIII. Thrombin also promotes platelet aggregation and adherence to the site of injury. Thrombin’s ability to convert fibrinogen to fibrin and activate platelets results in bleeding cessation.

FDA Approved Indications and Off-Label Uses Table 1 Thrombin Product Manufacturer FDA Approved Indication FDA Approval Date (Source) As an aid to hemostasis whenever oozing blood and minor bleeding from capillaries Thrombin JMI (Bovine) King Pharmaceuticals and small venules is accessible. 1995 In various types of surgery, solutions of thrombin may be used in conjunction with an absorbable gelatin sponge, USP for hemostasis. As an aid to hemostasis whenever oozing blood and minor bleeding from capillaries Evithrom (Pooled Human Johnson and Johnson and small venules is accessible August 2007 Plasma) and control of bleeding by standard surgical techniques is ineffective or impractical. May be used in conjunction with an absorbable gelatin sponge, USP. As an aid to hemostasis whenever oozing blood and minor bleeding from capillaries Recothrom (Recombinant Zymogenetics and small venules is accessible January 2008 Human) and control of bleeding by standard surgical techniques is ineffective or impractical. May be used in conjunction with an absorbable gelatin sponge, USP.

Topical thrombin is used for bleeding cessation in an extensive variety of surgical settings including but not limited to neurologic, orthopedic, general, cardiothoracic, vascular, gynecologic, head and neck and dental surgeries. Reportedly, thrombin is commonly used to stop residual bleeding after suturing in cardiovascular and cardiothoracic surgery.2 One author has stated that “it is difficult to ascertain the extent of use of topical thrombin, however, its wide use in surgery today may be as common as saline”. 2-3 The off-label use of thrombin to assist in the thrombosis of vascular pseudoaneurysm has become more common over the past several years.1, 39

Current VA National Formulary Alternatives The generic listing of “topical thrombin, powder” is on the VA National Formulary. There is no designation specifying a particular source of the topical thrombin.

Dosage and Administration4-6 The dose and administration of topical thrombin is dependent upon the amount of bleeding, the site of bleeding and preference of the surgeon. It can be diluted to a concentration of 1000-2000 US U/ml and then applied using a variety of techniques. One commonly used technique is to immerse an absorbable sponge product (e.g. Gelfoam®) in the thrombin solution and apply to the site of bleeding using the thrombin saturated sponge. Thombin is also available as a spray and can be sprayed onto the bleeding site. In some situations, dry powder products may be applied “dry” directly onto oozing surfaces.

All three products are labeled for TOPICAL USE ONLY. Because of the appearance of the packaging, along with several other factors, there have been a number of cases of accidental intravascular administration of topical thrombin. As a result, labeling/packaging changes have been made in an attempt to reduce these errors.7

The use of thrombin to assist in the thrombosis of vascular pseudo aneurysm is becoming increasingly more common. Doses can range from 50-1000+ IU.

Table 2 How Products are Supplied, Preparation and Storage Requirements Thrombin Product How Supplied Storage/Preparation Requirements Thrombin JMI (Powder) 5,000 IU vial (5 mL NS diluent) Room temperature/Reconstitute vial 20,000 IU vial (20 mL NS diluent) contents using transfer device. Pump Spray Kit 20,000 IU (NS diluent, spray pump and actuator) Reconstituted solutions can be Syringe Spray Kit 20,000 IU (NS diluent, refrigerated for 24 hours or kept at syringe tip and syringe) room temperature for 8 hours. Epistaxis Kit 5,000 IU (NS diluent, nasal delivery device and syringe Evithrom (Solution) 2 mL, 5 mL and 20 mL vials containing 800- Store frozen. If frozen, stable for 2 1200 IU/mL of thrombin years. If refrigerated-stable for 30 days. Room temp-stable for 24 hours.

Thaw prior to use: Thaw time: Refrigerator-within 1 day; Room temp.-within 1 hour; 37º-10 minutes (2 mL and 5mL vials only). Do not heat >37º and do not keep at 37º for more than 10 minutes. Once thawed, cannot be refrozen. Once at room temp, cannot be refrigerated. Recothrom (Powder) 5,000 IU vial, 5-mL prefilled NS diluent Room temperature/Reconstitute using syringe, needle transfer device, 5-mL empty needle transfer device. syringe 20,000 IU vial Once reconstituted, may be stored at room temperature for up to 24 hours.

Efficacy Despite thrombin’s widespread use as a surgical hemostat, there is a lack of randomized trials comparing the efficacy and safety of thrombin to other hemostatic interventions. It has been reported that the FDA approved the use of topical thrombin in the late 1970’s based upon a small number of surgical cases in which intraoperative blood loss was reduced. At that time, no immunologic data were collected. One author has stated that the acceptance of thrombin in surgery is due to the fact that when thrombin is applied to a wound with blood in it, the blood clots.8

Efficacy measures: In clinical trials comparing the application of topical thrombin to either placebo or thrombin derived from other sources, success in achieving complete hemostasis at the target bleeding site within 10 minutes was the most common efficacy endpoint examined. Other efficacy endpoints include success in achieving

Updated versions may be found at www.pbm.va.gov or vaww.pbm.va.gov 4 December 2008 Topical Thrombin Products complete hemostasis at the target site within 3 or 6 minutes; as well as estimated intraoperative blood loss, units of blood transfused, length of stay in specialized care units or hospital stay.

Summary of efficacy findings: (refer to appendix A for additional study details) To date, there have been three published clinical trials9-11 and one subgroup analysis12 of the three available topical thrombin products. The first trial was a phase II study in which recombinant human thrombin was compared to placebo. Although the emphasis of this trial was on safety, efficacy, although not powered to do so, was also monitored.9 In the second trial (n=305), pooled human thrombin was compared to bovine thrombin in a variety of surgical settings.10 The percentage of target bleeding sites achieving hemostasis in 10 minutes was the primary endpoint. Secondary endpoints included success in achieving hemostasis at 3 and 6 minutes; adverse events; and blood samples were tested for seroconversion of a variety of anti- antibodies (b-thrombin, b-factor V/Va, h-thrombin, h-FV/Va). Intraoperative blood loss, units of blood transfused, procedure duration, time in specialty unit and length of stay were also recorded. No differences were noted in the primary endpoint between groups. Investigators noted that there were no clinically meaningful differences in the secondary endpoints (efficacy or safety measures) or other variable recorded. A higher percentage of patients in the bovine thrombin group developed antibodies against one or more antibodies tested (human thrombin: 3/126 vs. 0/122 bovine vs. human (NS), respectively; bovine thrombin: 10/126 vs. 3/122 bovine vs. human (NS), respectively; bovine factor V/Va: 12/126 vs. 2/122 bovine vs. human (p=0.01), respectively. No patient developed antibodies against human factor V/Va. In those three patients who developed antibodies against human thrombin, no details were provided on bleeding or other safety outcomes in comparison to the overall group. Overall, safety was reported to be similar between the groups. In this paper, the authors note that antibody development has been reported even with highly purified bovine thrombin products (<5% non-thrombin proteins); however there may be fewer clinical consequences associated with these products. The authors suggest avoidance of bovine-thrombin products in patients with known seropositivity (known antibodies against bovine or native thrombin or factor V/Va) and in those with a surgical history or known exposure to topical bovine thrombin since these individuals may be more likely to have antibodies against thrombin and/or factor V.

In the third study11, 463 patients undergoing spinal surgery, hepatic resection or vascular surgery (peripheral arterial bypass or arteriovenous graft placement) were randomized to receive recombinant or bovine thrombin. The primary endpoint endpoint was the percent of target bleeding sites that achieved hemostasis at 10 minutes. Secondary endpoints included safety assessments (patients were followed for 1 month postoperatively) and the development of anti-product antibodies. There were no differences in the primary endpoint or in safety endpoints between groups. With regard to safety, there were several categories that were prospectively identified and compared between groups. Those safety outcomes included bleeding (11% b-throm vs. 13% rh-thrombin), thromboembolism (5% b-throm vs. 6% rh-throm), cardiac events (18% b-throm vs. 20% rh-throm), hypersensitivity (17% b-throm vs. 14% rh-throm), postoperative wound infection (10% b-throm vs. 9% rh-throm) and other infections (15% b-throm vs. 12% rh-throm). The differences were not significant. With regarding to development of antibodies to the product received, more patients in the bovine thrombin group seroconverted (18.4% vs. 1.6%) or had an increase in their antibody titer to bovine thrombin versus the recombinant human thrombin group. Investigators tested the samples of patients who developed antibodies to recombinant thrombin (n=3/195) to determine if the antibodies neutralized human thrombin; they did not. The authors did not test blood samples from patients developing antibodies to bovine thrombin to determine if they neutralized human thrombin. A post-hoc analysis was performed in those patients developing anti-bovine thrombin antibodies to see if there was an association with posttreatment antibody development and adverse events compared to those that did not develop anti-bovine thrombin antibodies. They noted a numeric disadvantage in those patients developing antibodies to bovine thrombin in terms of bleeding, thromboembolism, hypersensitivity or high aPTT. However, since this was a post-hoc analysis, the findings should be considered as hypothesis generating only.

The fourth study12 consisted of a subgroup population (reference 11) of 183 patients undergoing vascular surgery (PAB and AV graft). The authors do caution that this subgroup was not identified “a priori”; however this group was analyzed since they have the potential to have multiple exposures to topical thrombin and as a result may be at greater risk for antibody formation and potentially adverse bleeding events. Aside from percent success in achieving hemostasis in 10 minutes, other observations included: Updated versions may be found at www.pbm.va.gov or vaww.pbm.va.gov 5 December 2008 Topical Thrombin Products percent achieving hemostasis at 1.5, 3, and 6 minutes, adverse events, transfusions, death, graft complications, and anti-product antibody development, etc. There were no differences in efficacy measurements (percent success in achieving hemostasis or peri-operative transfusions) or safety measurements overall. As in reference 11, more patients receiving bovine thrombin developed antibodies against bovine thrombin. Samples were not tested to determine if these antibodies were capable of neutralizing native human thrombin.

Off-label evidence for topical thrombin: Pseudoaneurysm39-40 Pseudoaneurysm (PSA) is described as a contained rupture of the arterial wall occurring as a complication of cardiac and peripheral angiographic interventions. Trauma and infection can also be causes of pseudoaneurysm. After diagnostic catheterization, the incidence of PSA may range from 0.05-2%; however, after coronary or peripheral procedures PSA can be as high as 2-6%.39 Pseudoaneurysm is often associated with pain and swelling in the groin or affected area. Although the incidence is unknown, rupture of the PSA is the most catastrophic complication and seems to be dependent upon size of the PSA, presence of symptoms, hematoma size and/or continued growth of the sac.

There are a variety of treatments used to resolve PSA including watchful waiting for spontaneous thrombosis. However, many patients are receiving anticoagulant therapies and if patients are symptomatic, treatment, other than waiting, is recommended. The most common treatments are surgery, ultrasound guided compression therapy and ultrasound-guided thrombin injection. Although surgical repair of PSA is an effective treatment, disadvantages include the need for anesthesia, incision in groin area increasing the chance for infection, cost, prolonged hospital stay and complication rate in the range of 20%. Ultrasound guided compression therapy has a success rate of about 75-98%. In patients receiving anticoagulation, success rates are lower (~30-73%). Disadvantages include the time it takes to achieve sustained thrombosis (average compression time 33 minutes, range 10-120 minutes), can be painful for patients, high recurrence rates in anticoagulated patients (25-35%), and difficulty maintaining pressure in the appropriate area for long periods of time. In 1986, Cope and Zeit41 first described the percutaneous use of bovine thrombin to treat pseudoaneurysm in four patients. Today, percutaneous treatment of pseudoaneurysm with ultrasound guided thrombin injection has become the treatment of choice for postcatheterization pseudoaneurysms and/or is used as a secondary option in patients not achieving closure of their pseudoaneurysm with ultrasound guided compression therapy. When thrombin is injected into the appropriate area, a fibrin clot forms immediately whether or not a person is receiving anticoagulation therapy; but it may take up to a few hours using compression therapy. Treatment of PSA with thrombin is successful in 91-100% of cases. Doses used are in the range of 1000 IU but more recently lower doses have been advocated (e.g. 100-200 IU) because they seem to have similar efficacy and a lower rate of complications.39-40 There are a number of authors who have reported their success in using ultrasound-guided thrombin injection and/or have recommended it to be used as the primary treatment for postcatheterization PSA.42-49 In a Cochrane Systematic Review (January 2006), authors conclude that there is evidence to support the use of thrombin injection as an effective treatment for femoral PSA. They further suggest that ultrasound-guided compression be considered as a first-line treatment and thrombin injection reserved for those who fail that approach.50

Complications of ultrasound-guided thrombin injection of postcatherization PSA include deep vein thrombosis (if injected accidentally into vein), pulmonary embolism or thrombosis of the artery if thrombin is injected into the tract or neck of the PSA or into the artery. The complication rate is reported to be around 1.3% with an embolic complication rate of 0.5%.39-40 If an embolic event does occur, spontaneous lysis has been reported in about 50% of cases while other cases may require thrombolysis or embolectomy.39-40,51 Several investigators have identified risk factors for developing ischemic complications after treatment with thrombin which can include short neck of the PSA, wide communication between main vessel and PSA, small main vessel or small PSA and thrombin injection close to or in the neck of the PSA.40,52

Adverse Events Prescribing information:4-6 Bovine:  Potential for allergic reactions in patients known to be sensitive to bovine materials. Updated versions may be found at www.pbm.va.gov or vaww.pbm.va.gov 6 December 2008 Topical Thrombin Products

 Inhibitory antibodies, interfering with hemostasis, may develop in a small percentage of people (Including a black box warning). Human:  Anaphylactic reactions may occur. Recombinant:  Limited data on repeat exposure are available.

For each of the three topical thrombin preparations, there are several issues that require consideration with regard to safety. First, the source of thrombin; bovine, human and recombinant human should be considered.

Bovine thrombin has been available from a variety of manufacturers since 1940. The earlier bovine thrombin products (Thrombostat® and Thrombogen®), although no longer available, were less pure and contained a greater amount of nonthrombin proteins (e.g. fragments of bovine factor V, etc.) compared to Thrombin JMI®. Since bovine thrombin is antigenically different from human thrombin (differences in sequence and glycosylation patterns) and may contain nonthrombin proteins, it may be more immunogenic compared to thrombin derived from other sources. Some theorize that the higher purity of the currently available bovine thrombin (Thrombin JMI®) may reduce its immunogenicity. 1,4 (An association of antibody development after exposure to bovine thrombin with adverse events will be discussed below). Thrombin JMI® has been shown to produce a lesser immunogenic response in comparison to a less pure bovine thrombin as a result of an enhanced purification process using membrane filtration (to remove most of the antigens found in crude thrombin) in an in vitro study.34 And in an animal study, purified bovine thrombin exposure in rabbits did not result in generation of any detectable antibodies against bovine Factor V related antigens.35 Finally, bovine-derived products carry a very low risk of transmitting prion diseases such as bovine spongiform encephalopathy (mad cow disease). Thrombin from human sources can carry a risk of transmitting plasma-borne pathogens such as hepatitis B and C, and human immunodeficiency virus; in addition to a theoretical risk of prion diseases like Creutzfeldt-Jakob disease. 14 Over the past 20 years, there have been significant advances in minimizing the risk of viral transmission through viral inactivation techniques.15-16 Recombinant human thrombin has been shown in a variety of tests to be similar in structure and identical in amino acid sequence to human thrombin.6, 14 As with other recombinant products, there is no risk for viral transmission with recombinant thrombin.

Immunogenicity of Bovine thrombin: Prior to 1994, there were only a limited number of adverse events reported for topical bovine derived thrombin. However, as early as 1945, there was one published report of the development of antibodies to clotting factors in an animal model after exposure to bovine thrombin. 13 Just over a decade ago, clinicians began reporting serious adverse bleeding events resulting from the development of clotting factor inhibitors (antibodies against clotting factors) after exposure to bovine thrombin.17-27 In many of these reports, the authors conclude that the development of factor V inhibitors after exposure to bovine thrombin is responsible for the serious bleeding complications. Prompted by these reports, in 1996, the FDA required manufacturers of any product containing bovine thrombin to include a “black box” warning in their product labeling. The warning reads as follows:

Warning Topical thrombin (bovine) must not be injected! Apply on the surface of bleeding tissue.

The use of topical bovine thrombin preparations has occasionally been associated with abnormalities in hemostasis ranging from asymptomatic alterations in laboratory determinations, such as prothrombin time (PT) and partial thromboplastin time (PTT), to severe bleeding or thrombosis which rarely have been fatal. These hemostatic effects appear to be related to the formation of antibodies against bovine thrombin or factor V which in some cases may cross react with human factor V, potentially resulting in factor V deficiency. Repeated clinical applications of topical bovine thrombin increase the likelihood that antibodies against thrombin or factor V may be formed. Consultation with an expert in coagulation disorders is recommended if a patient exhibits abnormal

coagulation laboratory values, abnormal bleeding, or abnormal thrombosis following the use of topical thrombin. Any interventions should consider the immunologic basis of this condition. Patients with antibodies to bovine thrombin preparations should not be re- exposed to these products.*

(*Although the boxed warning indicates that “Patients with antibodies to bovine thrombin preparations should not be reexposed to these products”, there is no commercially available test to determine whether or not an individual has antibodies to one or more bovine clotting proteins.)

Formation of antibodies, following exposure to bovine thrombin, may manifest clinically as asymptomatic laboratory abnormalities or as hemorrhagic or thromboembolic complications. The actual risk of experiencing a serious adverse bleeding event associated with administration of bovine derived thrombin is not known. In this section, the immunogenicity of bovine thrombin and the clinical significance of antibody development in association with its use will be addressed.

Several investigators have tested blood samples from patients exposed to bovine thrombin to determine the incidence of antibody formation against bovine or human coagulation factors (most commonly thrombin and factor V) and others have examined the associated adverse events.27-31 In most cases, investigators did not determine if the formation of antibodies (against human or bovine coagulation proteins) resulted in neutralization of native human thrombin activity. There are two prospective clinical trials10-11 comparing safety outcomes of bovine thrombin with either human or recombinant human thrombin in patients undergoing various types of surgery; and two literature reviews32-33 examining adverse events and/or bleeding in patients exposed to bovine thrombin during surgery. Both of the literature reviews were supported by the manufacturer of Thrombin JMI. Since the majority of investigations in Table 3 are not controlled, prospective clinical trials, the clinical significance of their findings is unclear.

Table 3: Safety and Immunogenicity of Topical Bovine Thrombin (see Appendix B for additional details) Clinical Trial Findings Comments/Authors Conclusions Carroll28 1996  None of the 9 patients with high antibody  Antibodies were not tested to N=21 cardiothoracic surgery pts titers against bovine and human thrombin determine their ability to neutralized Bovine fibrinogen clotted with bovine and Factor V (FV) had post-operative native human thrombin or FV. thrombin. bleeding complications (Fibrastat bovine glue) Dorian29 1998  *Overall, 12/120 (10%) patients  *Antibodies not tested to see if native N=120 stored blood specimens from patients developed antibodies to b-throm human thrombin was neutralized in exposed to bovine thrombin identified via  Multiple exposures to bovine thrombin their presence hospital records. Control included 114 blood resulted in a higher risk for antibody  Authors note: patients having samples from patients not exposed. development vs. single exposure (7/18 multiple exposures to b-throm are [39%] vs. 5/102 [5%], respectively) more likely to develop antibodies to  2 patients with multiple exposure to b- coagulation factors, leading to severe throm did have severe bleeding bleeding problems in some patients, complications (both received and should be avoided if there has Thrombostat on 2nd exposure) been a previous exposure.  2/120 (1.7%) developed severe bleeding complications. Ortel30 2001  80.7% and 90.7% developed antibodies  *Antibodies against human N=151 patients undergoing cardiac surgical against b-FV and Fa, respectively. coagulation proteins were not tested procedures; immunoassay used to detect  Nearly 50% developed antibodies to b- to determine if their presence antibodies against bovine and human prothrombin, 20.5% to b-throm. neutralized native human clotting coagulation proteins (bovine and human  *51% developed elevated antibodies to proteins. prothrombin, thrombin, factor V and bovine human coagulation factors, primarily FV,  Authors suggest that the safety of b- factor Va) before and after (24-48 hrs, 4-7 thrombin or both. throm needs to be reassessed and days and 4-8 weeks) bovine thrombin  No increased risk for an adverse post-op reexposure to these agents should exposure. Hematologic and clinical outcomes outcome in patients with prior surgical likely be avoided. were also collected. history, use of higher thrombin doses, (Thrombogen) (Samples were compared to presence of elevated antibody levels to samples from 50 normal donors) bovine or human coagulation factors after surgery or abnl coagulation studies Updated versions may be found at www.pbm.va.gov or vaww.pbm.va.gov 8 December 2008 Topical Thrombin Products

 Significantly increased risk for adverse post-op outcome in patients with elevated baseline antibody levels to 2 or more bovine coagulation proteins (adjusted odds ratio 5.4, 95% CI 1.54- 18.8, p<0.01) Winterbottom31 2002  No significant correlation between  No correlation was observed when N=309 having cardiac, vascular or spinal presence or absence of detectable comparing the incidence of post-op surgery exposed to b-thrombin (Floseal® or antibodies at baseline or 6-8 weeks post- complications and presence or Gelfoam®+b-throm) (Thrombin JMI) op. There was a trend between blood- absence of antibodies to bovine (Gelfoam served as the control group) related complications (hemorrhagic, proteins (b-throm and b-factor Va). Blood samples collected pre-op and 6-8 thromboembolic, wound or other) and  No patient developed antibodies weeks post-op and assayed for antibodies to prior surgery (p=0.08) against human factor Va. (Blood bovine thrombin and factor Va. *Antibodies samples were not tested for against human thrombin and factor Va were antibodies against human factor V.) tested at 6-8 weeks post-op only in those  The authors concluded that the patients with detectable antibodies to both presence of antibodies to bovine bovine proteins. PT was measured to thrombin and/or bovine factor Va determine any clinically significant pre-op or post-op did not result in an coagulopathy. increased incidence of adverse clinical outcomes or abnl PT and “therefore the use of b-thrombin during surgery does not present an undue risk of bleeding resulting from an antibody-induced coagulopathy.” Doria10 2008 Number who seroconverted:  The authors concluded “no clinically R, DB, MC, n=305 Antibody h- b- Diff meaningful differences in other Phase III study (Ethicon sponsor) throm throm variables for blood loss, length of (Add’l details, see appendix A) Any 4/122 16/126 . stay, etc.” Statistics were not Human vs. Bovine thrombin 09-. provided 71  Authors conclude bovine and human b-throm 3/122 10/126 .09- thrombin are equal in achieving 1.01 hemostasis in 3, 6 and 10 min. and b-FV/Va 2/122 12/126 . have comparable safety profiles. 04-.  Despite no difference in clinical 67 safety outcomes, authors comment in h-throm 0/122 3/126 0- discussion section: “Antibody 1.3 development has been observed even h-FV/Va 0/122 0/126 ND with highly purified products (<5% ND=not defined, several patients are non-thrombin proteins); however, represented twice since they developed such products may be associated antibodies to 2 or more but not counted twice with less clinical sequelae. To avoid in “any” category. *No patient seroconverted potential adverse clinical outcomes for antibodies against h-FV/Va related to anti-bovine thrombin/ factor V antibodies, it is recommended that bovine thrombin not be administered to patients with known seropositivity. Because patients with a history of surgery and/or topical b-thrombin exposure are more likely to have antibodies against thrombin and/or factor V, it is also advisable that b-thrombin be avoided in such patients.” Chapman11 2007 Anti-Product Antibodies:  3 patients who seroconverted in the R, DB, MC, n=463 b- rh- Diff rh-thrombin group were tested and Phase III study (Zymogenetics sponsor) throm throm the antiproduct antibodies did not (Add’l details, see appendix A) Baseline 10/200 3/198 -- neutralize native human plasma Recombinant vs. Bovine thrombin Antibody thrombin activity. The patients who Positive 43/200 3/198 p<.0001 received b-thrombin and either Seroconvert 35/190* 3/195 -- seroconverted or had an increase in Updated versions may be found at www.pbm.va.gov or vaww.pbm.va.gov 9 December 2008 Topical Thrombin Products

>titer unit their antiproduct antibody titers; change 8/10** 0/3 -- these samples were NOT tested to determine whether or not the *Denom. includes only patients with a post- antiproduct antibodies were capable baseline measurement and no detectable of neutralizing native human plasma antiproduct antibodies at baseline. thrombin. **Denom. includes only patients with a post-  *Post-hoc analysis of bleeding or baseline measurement and detectable anti- thrombo-embolic events, product antibodies at baseline. hypersensitivity or high aPTT was conducted for association of development of post treatment antiproduct antibodies and ADEs. There were numeric differences with higher rates of these events occurring in patients who developed antibodies in the b-thrombin group. However, this is a post-hoc analysis and the denom. is lower due to missing aPTT values. Reynolds32 2008  No standardized definitions of bleeding  Risk of bleeding events not clearly Literature review-thrombin sealants events. defined but bleeding rates did not N=21 clinical trials addressing risk of  Rates of bleeding in general surgery appear to be different in the studies bleeding in surgery, 5 used thrombin sealant studies: 0.1-20.2%; most reporting with thrombin exposed patients (Supported by King Pharm) ranges from 2.6-4% versus those studies in which patients  Rates of bleeding in bovine thrombin were not exposed to b-thrombin. studies: 0.0-13%; most reported ranges  Authors recommend well-designed from 2-3%. and controlled trials to better examine the bleeding risk in surgical patients exposed and not exposed to b-throm. Clark33 2008  Immunogenicity studies did show an  Antibody elevation is increased in Literature review-topical bovine thrombin increased risk for development of patients exposed to b-throm. N=25 case reports/case series with suspect antibodies against coagulation factors  However, available evidence does not ADEs/Lab abnl attributed to b-throm. after exposure to b-throm. support this elevation in antibodies N=10 longitudinal studies b-throm used and  However, there was not consistent increases the risk for clinically ADEs reported. evidence that the development of these important ADEs. antibodies resulted in an increased risk for ADEs.  Available data mostly from case reports, case series.  Controlled trial comparing rh-throm to b- thromb did show greater antibody development in those exposed to b-throm but no important efficacy or ADE differences were observed. Abnl=abnormal, ADEs=adverse events, b-factor Va=bovine factor Va, b-throm=bovine thrombin, DB=double-blind, MC=multicenter, pre-op=preoperative, post-op=postoperative, PT=prothrombin time, R=randomized, rh- throm=recombinant thrombin

Summary of safety information In the majority of publications summarized in table 3, bovine thrombin was observed to produce a greater immunogenic response in comparison to human or recombinant human thrombin. However, the antibodies that developed after exposure to bovine thrombin were either not tested to determine the ability of these antibodies to neutralize native human thrombin or that information was not reported. In the publications reporting adverse clinical events associated with bovine thrombin in comparison to human or recombinant thrombin sources, the safety profiles were reportedly comparable. Several authors have recommended avoidance of bovine thrombin in patients known to possess antibodies against coagulation factors (although no commercial test is available) and in those patients with a history of surgery and/or a known or suspected exposure to bovine thrombin. However, the available evidence does not support a significantly increased risk for post-operative complications with bovine thrombin when compared to human or recombinant human thrombin. It is important to note that most patients participating in these reports/studies have not had Updated versions may be found at www.pbm.va.gov or vaww.pbm.va.gov 10 December 2008 Topical Thrombin Products multiple exposures to thrombin from any source and the trials are relatively small. Since patients having surgery for peripheral artery bypass (PAB) or arteriovenous graft placement for dialysis are expected to undergo multiple surgical interventions and as a result may be at a greater risk for antibody development and potentially adverse bleeding events, Weaver, et al.12, analyzed a subgroup of the Chapman study11 to determine the clinical safety of bovine vs. recombinant human thrombin in these patients. In this subgroup analysis, no differences in peri-operative transfusions, death, graft complications or thrombosis, serious adverse events, peri or post-operative bleeding or thromboembolic complications or differences in laboratory values (PT, aPTT, INR) were observed. A post-hoc analysis of bleeding, thromboembolic events, hypersensitivity or abnormal aPTT in those with antibodies to bovine thrombin compared to those without showed a numeric increase in events in the bovine thrombin antibody positive group that was not statistically significant when compared to the antibody negative group. There is one ongoing observational study assessing the impact of exposure to bovine thrombin (Thrombin JMI) on coagulation parameters (e.g. aPTT) 48 hours after surgery in patients with a likelihood of prior exposure to bovine thrombin within the previous four years.36 This study is currently recruiting subjects.

Additional information from the FDA website: (FDA Review)38 Recombinant Thrombin  In the phase 3 study by Chapman, et al.,11 antibodies to bovine thrombin developed in 43 patients, antibodies to recombinant human thrombin developed in 3 patients and one patient developed Chinese hamster ovary (CHO) antibodies after exposure to recombinant thrombin.  Events of heightened surveillance (EHS) were compared between bovine and recombinant thrombin in the phase 3 study. These events were chosen due to their potential association with the mechanism of action of thrombin, known sequelae of cross-reacting antibodies, etc. The events included: bleeding events based upon coagulopathy, thromboembolic events, cardiac events, hypersensitivity, postoperative wound infections, nausea and vomiting. Overall, there were no significant differences in these events between bovine and recombinant thrombin groups.  In the study, the manufacturer associated bleeding events (n=6) with anti-product antibody development in the bovine thrombin group. The FDA reviewer felt that since the bleeding events occurred on days 1, 1,2,3,5 and 8, it was difficult to associate the events with anti-product antibody development. Furthermore, the higher frequency of anti-product antibodies in the bovine thrombin group did not manifest clinically and the bleeding events in this group could have occurred by chance alone. The FDA reviewer advised the manufacturer of recombinant thrombin “that the bleeding adverse events in antibody positive patients do not appear to have occurred at a higher frequency than in antibody negative subjects and that the manufacturer may not advertise or promote any association with bleeding or antibody status.”  The FDA reviewer had questions regarding the safety and immunogenicity of multiple exposures to recombinant thrombin. As a result, the manufacturers of recombinant human thrombin have committed to conducting a phase 4 study of the safety and immunogenicity of re-exposure to their product.

Human Thrombin No additional information.

Precautions/Contraindications The Institute for Safe Medication Practices (ISMP) recently warned of the danger of accidental intravascular administration of topical thrombin. In several reported cases, patients accidentally injected with topical thrombin experienced adverse events ranging from an extended hospital stay to death. As a result of this warning, the Pharmacy Benefits Management Services, the Medical Advisory Panel and the VA Center for Medication Safety (VA MedSAFE) notified VA clinicians in August 2008 of the potential risk for accidental intravascular administration as well as included recommendations for limiting this type of error.37

Table 4: Warnings/Precautions4-6 Bovine (Thrombin JMI®) Human (Evithrom®) Recombinant (Recothrom®)  Boxed warning (see page 5)  May carry a risk of transmitting  Potential for allergic reactions in  Pregnancy category C infectious agents such as viruses persons allergic to snake  Allergic reactions may occur in and theoretically, the Creutzfeldt proteins patients known to be sensitive to –Jakob disease (CJD) agent,  Potential risk of thrombosis is bovine materials despite manufacturing steps to absorbed systemically  Inhibitory antibodies may reduce these risks.  Limited data are available on develop in a small percentage of  Anaphylactic reactions may repeat exposure (n=6) people occur  Safety and effectiveness not  Potential risk of thrombosis if established in children absorbed systemically

Table 5: Contraindications4-6 Bovine (Thrombin JMI®) Human (Evithrom®) Recombinant (Recothrom®)  Do not use in persons known to  Do not inject into the circulatory  Do not inject directly into the be sensitive to any of its system. circulatory system. components and/or to material  Do not use in individuals known  Do not use for the treatment of of bovine origin to have anaphylactic or severe massive or brisk arterial  Must not be injected or system reaction to human blood bleeding. otherwise allowed to enter large products.  Do not administer to patients blood vessels. Extensive intra-  Do not use for the treatment of with known hypersensitivity to vascular clotting and even death severe or brisk arterial bleeding. Recothrom, any components of may result. Recothrom or hamster proteins.

Look-Alike/Sound-Alike (LA/SA) Error Risk Potential The VA PBM and Center for Medication Safety is conducting a pilot program which queries a multi- attribute drug product search engine for similar sounding and appearing drug names based on orthographic and phonologic similarities, as well as similarities in dosage form, strength and route of administration. Based on similarity scores as well as clinical judgment, the following drug names may be potential sources of drug name confusion:

Table 6 Drug Name LA/SA Drug Name Bovine thrombin Other sources of topical thrombin (e.g. human or recombinant) Thrombin JMI® Tromethamine inj, trophamine 10% inj, Human Thrombin Other sources of topical thrombin (e.g. bovine or recombinant), human albumin Evithrom® None that would be confused in the operating room due to name confusion. Recombinant Thrombin Other sources of topical thrombin (e.g. bovine or human), other recombinant human products (e.g. hyaluronidase) Recothrom® Restasis, Rocephin, Remodulin, Repronex

Drug Interactions There are no known drug-drug interactions with the use of the agents.

VA Purchases/Acquisition Costs Table 7: VA Cost/Purchases April-September 2008 (Total $1,589,774) Cost per Unit Total Units Product/Manufacturer Packaging ($) Purchased Total Dollars ($) b-Thrombin/King 20,000 unit vial 181.82 2,239 407,117 b-Thrombin/King 20,000 unit vial 181.12 824 149,243 b-Thrombin/King 20,000 unit vial 172.70 258 44,567 b-Thrombin/King 5,000 unit vial 43.34 3,745 162,308 b-Thrombin/King 5,000 unit (nasal kit) 50.39 38 1,915 b-Thrombin/Monarch 20,000 unit vial 181.81 2,474 449,818 Updated versions may be found at www.pbm.va.gov or vaww.pbm.va.gov 12 December 2008 Topical Thrombin Products

b-Thrombin/Monarch 20,000 unit vial 172.73 419 72,374 b-Thrombin/Monarch 20,000 unit vial 181.10 809 146,509 b-Thrombin/Monarch 5,000 unit vial 43.34 3,512 152,197 rh-Thrombin/Zymogenetics 20,000 unit vial 214.33 6 1,286 rh-Thrombin/Zymogenetics 5,000 unit vial 59.51 41 2,440 h-thrombin/Johnson & 5,000 unit vial 76.80 -- -- Johnson 20,000 unit vial 300.00 No purchases for human thrombin during this time period. *Prices are effective as of 10-31-08 Monarch is a subsidiary of King Pharmaceuticals

Pharmacoeconomic Analysis There have been no published pharmacoeconomic analyses to date.

Conclusions Both topical human and recombinant human thrombin have been demonstrated in clinical trials to be similarly effective to bovine thrombin in achieving hemostasis within 10 minutes. As for safety, bovine thrombin has been shown to be more immunogenic than human or recombinant human thrombin but the clinical significance of this is not known. In clinical trials, a greater percentage of patients receiving bovine thrombin versus the comparator thrombin did develop antibodies against coagulation factors or the product received but the rate of adverse events did not differ between groups. In addition, the ability of these antibodies to neutralize native human thrombin was not determined in the trials. A post-hoc analysis of the “events of heightened surveillance (e.g. bleeding, thromboembolic events, etc.)”, in patients developing anti-bovine thrombin antibodies vs. those who did not, did not demonstrate significant differences between groups and the numeric differences in bleeding events could have occurred by chance alone.

Because the FDA had received reports of serious adverse bleeding events associated with bovine thrombin, in 1996, all products containing bovine thrombin were required by the FDA to have a boxed warning of the risk for possible coagulation abnormalities in some patients. The warning states that the hemostatic effects appear to be related to the formation of antibodies against bovine thrombin, factor V or both. In some cases, the antibodies against bovine factor V may cross-react with human factor V potentially leading to factor V deficiency. Many of the published cases of serious bleeding events associated with bovine thrombin were as a result of the development of factor V deficiency. Most of these reports involved older, less pure bovine thrombin products that are no longer available. Although there are several factors (e.g. amino acid sequence and glycosylation patterns) that can result in an immunogenic response, some theorize that the greater purity of bovine thrombin (less nonthrombin proteins) may results in a lessened immunogenic response. In one in vitro study, Thrombin JMI® was shown to produce a lesser immunogenic response in comparison to a less pure bovine thrombin as a result of an enhanced purification process using membrane filtration (to remove most of the antigens found in crude thrombin). And in an animal study, purified bovine thrombin exposure in rabbits did not result in generation of any detectable antibodies against bovine Factor V related antigens. Nevertheless, bovine thrombin does have a boxed warning (page 7) stating that “repeated clinical applications of topical bovine thrombin increase the likelihood that antibodies against thrombin or factor V may be formed and that patients with antibodies to bovine thrombin preparations should not be re-exposed to these products.” The challenge is that there is no commercially available test to determine whether or not an individual has antibodies to one or more bovine clotting proteins.

At this time, there is a lack of evidence for safety/efficacy in individuals re-exposed to recombinant thrombin. There is a trial planned to investigate multiple exposures to recombinant human thrombin.

Human and bovine thrombins carry a very small/theoretical risk of transmitting prion diseases such as Creutzfeldt-Jakob disease (CJD) agent and mad cow disease, respectively. Human thrombin may also carry of small risk of transmitting infectious agents such as viruses despite the manufacturing steps to reduce this risk. There is no risk for transmission of pathogens associated with recombinant thrombin.

Place in Therapy/Recommendations Topical thrombin (powder) is a listed on the VANF; however the source is not identified. Both bovine and recombinant human thrombins are available in powder form and human thrombin is available as a solution. Updated versions may be found at www.pbm.va.gov or vaww.pbm.va.gov 13 December 2008 Topical Thrombin Products

Although bovine thrombin (Thrombin JMI®) has been shown to be more immunogenic than human (Evithrom®) or recombinant human (Recothrom®) thrombin, the evidence does not support a significantly increased risk for adverse post-operative complications in prospective clinical trials. It is important to note that most of the patients in the trials have not received multiple exposures to thrombin from any source and the trials are relatively small (n=300-453). The efficacy of all three thrombin products is similar. The available evidence supports the use of ultrasound-guided thrombin injection for postcatheterization pseudoaneurysm.

Contact person: Cathy Kelley, Pharm.D. [email protected]

References 1. Lawson, JH. The Clinical Use and Immunologic Impact of Thrombin in Surgery. Semin Thromb Hemostasis 2006;32 (Suppl 1):98-110. 2. Lundblad RL, Bradshaw RA, Gabriel D, et al. A Review of the Therapeutic Uses of Thrombin. Thromb Haemost 2004;91:851-860. 3. Price BD. General Patient Care and Safety in Surgery Technology for the Surgical Technologist: A positive care approach. Ed BL Caruthers, P Price, TL Junge, and BD Price. Delmer Thompson Learning, Albany, New York, 2001:118-149. 4. Thrombin-JMI Product Information. Thrombin, Topical USP (bovine origin). King Pharmaceuticals, Inc. Bristol, TN 37620. November 2007. 5. Evithrom Product Information. Thrombin, Topical (Human). Johnson and Johnson Wound Management, Ethicon, Inc. Somerville, NJ 08876. September 2007 6. Recothrom Product Information. Thrombin, topical (Recombinant). Zymogenetics, Inc. Seattle, WA 98102. January 2008. 7. PBM Bulletin-Topical Thrombin (link) 8. Lawson JH. The Clinical use and Immunologic Impact of Thrombin in Surgery. Semin Thromb and Hemost 2006;32(supplement 1):98-110. 9. Chapman WC, Lockstadt H, Singla N, et al. Phase 2, Randomized, Double-Blind, Placebo- Controlled, Multicenter Clinical Evaluation of Recombinant Human Thrombin in Multiple Surgical Indications. J Thromb Haemost 2006;4:2083-2085. 10. Doria C, Fischer CP, Wood CG, et al. Phase 3, Randomized, Double-Blind Study of Plasma- Derived Human Thrombin versus Bovine Thrombin in Achieving Hemostasis in Patients Undergoing Surgery. Curr Med Res Opin 2008;24:785-794. 11. Chapman WD, Singla N, Genyk Y, et al. A Phase 3, Randomized, Double-Blind Comparative Study of the Efficacy and Safety of Topical Recombinant Human Thrombin and Bovine Thrombin in Surgical hemostasis. J Am Coll Surg 2007;205:256-265. 12. Weaver FA, Lew W, Granke K, et al. A Comparison of Recombinant Thrombin to Bovine Thrombin as a Hemostatic Ancillary in Patients Undergoing Peripheral Arterial Bypass and Arteriovenous Graft Procedures. J Vasc Surg 2008;47:1266-1273. 13. Light RU. The Antigenicity of Boving Thrombin: Clinical Evaluation. J Neurosurg 1945;2:516- 523. 14. Bishop PD, Lewis KB, Schultz J, Walker KM. Comparison of Recombinant Human Thrombin and Plasma-Derived Human α-Thrombin. Semin Thromb Hemostasis 2006;32(suppl 1):86-97. 15. Horowitz B, Ben-Hur E. Efforts in Minimizing Risk of Viral Transmisssion Through Viral Inactivation. Ann Med 2000;32:475-484. 16. Remington KM, Trejo SR, Buczynski G, et al. Inactivation of West Nile Virus, Vaccinia Virus and Viral Surrogates for Relevant and Emergent Viral Pathogens in Plasma-Derived Products. Vox Sanguinis 2004;87:10-18. 17. Christie RJ, Carrington L, Alving B. Postoperative Bleeding Induced by Topical Bovine Thrombin: Report of Two Cases. Surgery 1997;121:708-710. 18. Chouhan VD, De La Cadena RA, Nagaswami C, et al. Simultaneous Occurrence of Human Antibodies Directed Against Fibrinogen, Thrombin, and Factor V Following Exposure to Bovine Thrombin: Effects on Blood Coagulation, Protein C Activation and Platelet Function. Thromb and Hemostasis 1997;77:343-349.

19. Neschis DG, Heyman MR, Cheanvechai V, et al. Coagulopathy as a Result of Factor V Inhibitor After Exposure to Bovine Topical Thrombin. J Vasc Surg 2002;35:400-402. 20. Zehnder JL, Leung LLK. Development of Antibodies to Thrombin and Factor V With Recurrent Bleeding in a Patient Exposed to Topical Bovine Thrombin. Blood 1990;76:2011-2016. 21. Ortel TL, Charles LA, Keller FG, et al. Topical Thrombin and Acquired Coagulation Factor Inhibitors: Clinical Spectrum and Laboratory Diagnosis. Am J Hematol 1994;45:128-135. 22. Spero JA. Bovine Thrombin-Inducted Inhibitor of Factor V and Bleeding Risk in Postoperative Neurosurgical Patients. J Neurosurg 1993;78:817-820. 23. Cmolik BL, Spero JA, Magovern GJ, et al. Redo Cardiac Surgery: Late Bleeding Complications From Topical Thrombin-Induced Factor V Deficiency. J Thorac Cardiovasc Surg 1993;105:222- 228. 24. Streiff MB, Ness PM. Acquired FV Inhibitors: A Needless Iatrogenic Complication of Bovine Thrombin Exposure. Transfusion 2002;42:18-26. 25. Crow SS, Sullivan VV, Aysola AE, et al. Postoperative Coagulopathy in a Pediatric Patient After Exposure to Bovine Topical Thrombin. Ann Thorac Surg 2007;83:1547-1549. 26. Savage WJ, Kickler TS, Takemoto CM. Acquired Coagulation Factor Inhibitors in Children after Topical Bovine Thrombin Exposure. Pediatr Blood Cancer 2007;49:1025-1029. 27. Clark J, Crean S, Reynolds MW. Topical Bovine Thrombin and Adverse Events: A Review of the Literature. Curr Med Res Opin 2008;24:2071-2087. 28. Carroll JF, Moskowitz KA, Edwards NM, et al. Immunologic Assessment of Patients Treated with Bovine Fibrin as a Hemostatic Agent. Thromb hemostasis 1996;76:925-931. 29. Dorion RP, Hamati HF, Landis B, et al. Risk and Clinical Significance of Developing Antibodies Induced by Topical Thrombin Preparations. Arch Pathol Lab Med 1998;122:887-894. 30. Ortel TL, Mercer MC, Thames EH, et al. Immunologic Impact and Clinical Outcomes After Surgical Exposure to Bovine Thrombin. Ann Surg 2001;233:88-96. 31. Winterbottom N, Kuo JM, Nguyen K, et al Antigenic Responses to Bovine Thrombin Exposure During Surger: A Prospective Study of 309 Patients. J Appl Res 2002;2:1-11. 32. Reynolds MW, Clark J, Crean S, Samudrala S. Risk of Bleeding in Surgical Patients Treated with Topical Bovine Thrombin Sealants: A Review of the Literature. Patient Saf Surg 2008;2:5. 33. Clark J, Crean S, Reynolds. Topical Bovine Thrombin and Adverse Events: A Review of the Literature. Curr Med Res Opin 2008;24:2071-2087. 34. Terrab A, Pawlak D, Spaay P, et al. Further Removal of Factor V Related Antigen From Bovine Thrombin by Utilizing a Membran-Filtration Step. Clinical and Applied Thrombosis/Hemostasis 2008;4:135-140. 35. Zhu H, Hoppensteadt D, Wahi R, et al. Reduced Immunogenic Potential of Membrane Filtered Bovine Thrombin Preparations for Hemostatic Application. Clin Appl Thromb Hemost Oct 2008, EPUB ahead of print. http://www.ncbi.nlm.nih.gov/pubmed/18977777? ordinalpos=1&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsPanel.Pubmed_DefaultRep ortPanel.Pubmed_RVDocSum (Accessed 12-2-08) 36. http://www.clinicaltrials.gov/ct2/show/NCT00775398?term=thrombin&rank=4 (Accessed 12-2- 08) 37. http://www.pbm.va.gov/vamedsafe/National%20PBM%20Safety%20Alert%20_Topical %20Thrombin_%2008-07-08.pdf (Accessed 12-8-08) 38. http://www.fda.gov/cber/products/recothrom/recothrom111107mem.htm (Accessed 12-4-08) 39. Webber GW, Jang J, GustavsonS, Olin JW. Contemporary Management of Postcatheterization Pseudoaneurysms. Circulation 2007;115:2666-2674. 40. Ohlow MA, Secknus MA, von Korn H, et al. Percutaneous Thrombin Injection for Treatment of Iatrogenic Femoral Artery Pseudoaneurysms: A Case for Caution. Angiology 2008;59:372-375. 41. Cope C, Zeit R. Coagulation of Aneurysms by Direct Percutaneous Thrombin Injection. Am J Roentgenol 1986;147:383-387. 42. Olsen DM, Rodriquez JA, Vranic M, et al. A Prospective Study of Ultrasound Scan-Guided Thrombin Injection of Femoral Pseudoaneurysm: A Trend Toward Minimal Medication. J Vasc Surg 2002;36:779-782. 43. Lonn L, Olmarker A Geterud K, Risberg B. Prospective Randomized Study Comparing Ultrasound-Guided Thrombin Injection to Compression in the Treatment of Femoral Pseudoaneurysms. J Endovasc Ther 2004;11:570-576. Updated versions may be found at www.pbm.va.gov or vaww.pbm.va.gov 15 December 2008 Topical Thrombin Products

44. Maleux G, Hendrickx S, Vaninbroukx, et al. Percutaneous Injection of Human Thrombin to Treat Iatrogenic Femoral Pseudoaneurysms: Short and Midtern Ultrasound Follow-up. Eur Radiol 2003;13:209-212. 45. Mohler ER, Mitchell ME, Carpenter JP, et al. Therapeutic Thrombin Injection of Pseudoaneurysms: A Multicenter Experience. Vasc Med 2001;6:241-244. 46. Greene K. Duplex-Guided Thrombin Injection for the Treatment of Iatrogenic Pseudoaneurysms. J Vasc Nurs 2002;20:117-122. 47. Elford J, Burrell C, Freeman S, Roobottom C. Human Thrombin Injection for the Percutaneous Treatment of Iatrogenic Pseudoaneurysms. Cardiovasc Intervent Radiol 2002;25:115-118. 48. Freund G, Durschmied D, Zhou Q, et al. Occlusion of Iatrogenic Pseudoaneurysms with Percutaneous Ultrasound Guided Thrombin Injection. VASA 2007;36:96-99. 49. Krueger K, Zaehringer M, Strohe D, et al. Postcatheterization Pseudoaneurysm: Results of US- Guided Percutaneous Thrombin Injection in 240 Patients. Radiology 2005;236:1104-1110. 50. Tisi PV, Callam MJ. Surgery Versus Non-Surgical Treatment for Femoral Pseudoaneurysms Cochrane Database of Systematic Reviews 2006;issue 1, Art. No. CD0004981. DOI: 10.1002/14651858.CD004981.pub2. 51. Dasyam AK, Middleton WD, Teefey SA. Development of Nonobstructive Intraarterial Thrombi After Injection of Thrombin into Pseudoaneurysms. Am J Roentgenol 2006;186:401-405. 52. D’Ayala M, Smith R, Zanieski G, et al. Acute Arterial Occlusion After Ultrasound-Guided Thrombin Injection of a Common Femoral Artery Pseudoaneurysm with a Wide, Short Neck. Ann Vasc Surg 2008;22:473-475.

Appendix 1: Clinical Trials of Thrombin Clinical Trial Inclusion/Exclusion Intervention Endpoints/Results Adverse Events/Comments Chapman et al9 Inclusion/Exclusion: Adults, Designed to examine safety. Endpoint: Safety  Authors reported overall R, DB, PC, MC normal platelets and Efficacy examined but not  130 patients randomized. rates of ADEs, serious N=130 coagulation parameters, no powered to do so.  Because open-label rh-thrombin was ADEs and treatment- Phase II study known antibodies or Randomized to rh-thrombin allowed, only 45 patients received placebo related ADEs as similar Zymogenetics (sponsor) hypersensitivity to thrombin or placebo (1:1). TTH was alone and 88 patients were exposed to rh- between groups. ADEs (published as a letter to the or other coagulation factors, measured after intervention thrombin. were mostly consistent editor) *Certain authors with no known bleeding or was applied using a gelatin  95% completed the protocol (98% placebo with expected post- consultant agreements with hematologic disorders. sponge. If blinded treatment only, 94% rh-thrombin) operative complications. Zymogenetics failed to achieve hemostasis  AV grafts 25%; hepatic resection 22%;  Treatment related ADEs Four surgical settings: AV at 10 minutes, open-label PAB surgery 21%, spinal surgery 32% occurred in less than 5% graft for hemodialysis, major rhthrombin was used as  4.8% pla, 5.7% rh-thrombin had baseline of patients reporting an hepatic resection, PAB rescue therapy. antibodies to rh-thrombin; after exposure, ADE. surgery and spinal surgery. one patient in each group (2.4% pla/ 1.2%  Interpretation of safety rh-thrombin) developed antibodies to rh- results was difficult due thrombin that did not cross-react with to use of open-label rh- native human thrombin. thrombin.  No evidence of coagulation or laboratory  Study not powered to abnormalities. assess efficacy. Doria et al10 Inclusion: Adults (18 years or Designed to compare the Endpoint: Success in achieving hemostasis One or more ADEs were R, DB, MC >), scheduled for CV, efficacy and safety of within 10 minutes Secondary endpoint: 1) reported in nearly all patients. N=305 neurologic (spine), general or plasma-derived human Success in achieving hemostasis within 3 and 6 However, only about 12% of Phase III study post-traumatic elective thrombin to bovine thrombin minutes. 2) Intraoperative blood loss, units of the ADEs were considered to Ethicon (sponsor) surgical procedures having at for achieving hemostasis in blood transfused, procedure duration, time in be related to study drug. No *Dr Doria is an ad hoc least one bleeding site (mild to patients undergoing CV, specialty unit, and length of stay were differences noted. 17% of h- consultant to Johnson and moderate) for which the neurologic, general or post- recorded. 3) ADEs were recorded up to 5 thrombin vs. 11% b-thrombin Johnson and another author surgeon expressed a need for a traumatic surgical weeks after surgery. 4) Seroconversion was reported serious ADEs, only is an employee of Ethicon. topical hemostat. procedures. assessed for antibodies against h-thrombin, b- 1.3% (n=2) in each group Exclusion: disorders Eligible patients were thrombin, bovine factor V/Va, and human were related to study interfering with hemostasis; randomized to human or factor V/Va. intervention. Changes in lab one or more preoperative bovine thrombin ONLY  365 patients were screened, 305 values from baseline at 12 abnormal CBC, PT or aPTT when all discrete bleeding randomized (h-thrombin=153, b- hour post-op and within 24 labs that are clinically had been controlled by thrombin=152) hours of discharge were significant; antiplatelets/ surgical cauterization with  282 (92%) patients completed study (h- similar between groups. anticoagulants within 5 days diathermy and/or suture thrombin=141, b-thrombin=141) of surgery; known antibodies ligation (following reversal  248 patients (81%) had both screening and Although overall median to bovine thrombin; organ of heparin in CV procedures) f/u blood samples for antibody differences were provided and transplant; morbidly obese; and surgeons determined assessment. stated to be similar in: length liver failure; severe bleeding need for a topical hemostat Primary Endpoint: No differences overall or of stay, care in specialty units, or ongoing infection at due to failure of etc. no individual assessment

bleeding site; alcohol or drug conventional surgical when examined by surgical type. In both of those patients who user; uncontrolled DM; techniques to stop bleeding. groups, more than 90% of target bleeding sites developed antibodies to religious prohibitions to Randomization was stratified achieved hemostasis within 10 minutes. human thrombin (n=3) in the bovine or porcine products, by surgical specialty. Secondary Endpoint: 1) No differences overall b-thrombin group was done or etc. Treatments were applied to in achieving hemostasis in 3 and 6 minutes. 2) presented. the target bleeding spot with The authors concluded “no clinically a gelatin sponge and pressure meaningful differences in other variables for Authors concluded that: was applied for 3 minutes. blood loss, length of stay, etc.” Statistics were “Antibody development has The sponge was cut to the not provided been observed even with proper size and the dose of 4) Number who seroconverted: highly purified products thrombin was dependent Antibody h- b- Diff (<5% non-thrombin upon the size of the sponge throm throm proteins); however, such used. Any 4/122 16/126 . products may be associated Blood samples were 09-. with less clinical sequelae. To collected 2 weeks prior to 71 avoid potential adverse surgery and at the 5-week b-throm 3/122 10/126 .09- clinical outcomes related to postsurgical visit for 1.01 anti-bovine thrombin/factor V detection of antibody b-FV/Va 2/122 12/126 . antibodies, it is recommended formation. 04-. that bovine thrombin not be 67 administered to patients with h-throm 0/122 3/126 0- known seropositivity. Because 1.3 patients with a history of h-FV/Va 0/122 0/126 ND surgery and/or topical b- ND=not defined, several patients are thrombin exposure are more represented twice since they developed likely to have antibodies antibodies to 2 or more but not counted twice against thrombin and/or in “any” category. *No patient seroconverted factor V, it is also advisable for antibodies against h-FV/Va that b-thrombin be avoided in such patients.”

Chapman WC, et al11 Inclusion: Aged 18 years and Designed to evaluate the Primary Endpoint: Incidence of hemostasis ADEs were similar between R, DB, MC >, no history of heparin- efficacy of rh-thrombin and within 10 minutes. groups. N=463 induce thrombocytopenia and b-thrombin and secondarily Secondary Endpoint: 1) Patients were A post-hoc analysis, Phase III Study no known sensitivity to to assess the safety and anti- monitored for 1 month post-op to assess safety. examining the association of Zymogenetics (Sponsor) Thrombin-JMI components, genicity of both products. 2) The presence of antibodies was assessed to posttreatment seroconversion *Two authors were bovine materials or porcine determine the antigenicity of each product. or increase in antiproduct employees of Zymogenetics collagen. Eligible patients were  463 patients randomized, 52 were not antibody titers and during the trial. Exclusion: included in clinical randomized in a 1:1 ratio to treated, 401 completed (b-thrombin=206, development of clinical study of rh-thrombin, receive rh-thrombin or b- rh-thrombin=205) ADEs, (bleeding or undergone a surgical thrombin (1,000 units/mL). Primary Endpoint: b-thrombin 95.1% vs. thromboembolic events, procedure or been treated with Dynamic allocation assured 95.4% rh-thrombin (NS) hypersensitivity or high any experimental agent within an equal number of patients Secondary Endpoint: aPTT) was done and showed a 30 days or had received blood assigned to each treatment,  1) Nearly all patients reported at least one numeric disadvantage in those products within 24 hrs of to each surgeon and to each ADE. seroconverting and who operation. surgical type. Enrollment  Most were not considered related to study received b-thrombin. (Post- was monitored to ensure that treatment and most were moderate in hoc analyses should be Three Surgical Settings: 40% of patients underwent severity. considered only “hypothesis Spinal surgery, hepatic vascular surgery, 30% spinal  Certain categories were prospectively generating”) resection or vascular surgery and 30% hepatic resection. followed: bleeding (11% b-throm vs. 13% *Investigators did not test (PAB surgery or AV graft Treatment was applied rh-throm); thromboembolism (b-throm whether or not antiproduct placement for dialysis) topically combined with 5% vs. rh-throm 6%); cardiac events (b- antibodies (that developed either Gelfoam or throm 18% vs. rh-throm 20%); or titers increased in those Surgifoam. Gauze pads were hypersensitivity (b-throm 17% vs. rh- receiving b-thrombin) placed on top with gentle throm 14%); post-op wound infections (b- neutralized native human pressure applied. throm 10% vs. rh-throm 9%); other plasma thrombin. Why not? infections (b-throm 15% vs. 12% They did for those rh=throm). seroconverting on rh-  Common ADEs were similar thrombin.  Abnormal lab tests were similar  2) 398 evaluable for anti-product antibodies b- rh- Diff throm throm Baseline 10/200 3/198 -- Antibody Positive 43/200 3/198 p<.0001 Seroconvert 35/190* 3/195+ -- >titer unit change 8/10** 0/3 --

*Denom. includes only patients with a post- baseline measurement and no detectable antiproduct antibodies at baseline. **Denom. includes only patients with a post- baseline measurement and detectable antiproduct antibodies at baseline. + 3 patients who seroconverted in the rh- thrombin group were tested and the antiproduct antibodies did not neutralize native human plasma thrombin activity. The patients who received b-thrombin and either seroconverted or had an increase in their antiproduct antibody titers; these samples were NOT tested to determine whether or not the antiproduct antibodies were capable of neutralizing native human plasma thrombin. *Post-hoc analysis of bleeding or thromboembolic events, hypersensitivity or high aPTT was conducted for association of development of posttreatment antiproduct antibodies and ADEs. There were numeric differences (NS) with higher rates of these events occurring in patients who developed antibodies in the b- thrombin group. However, this is a post-hoc analysis and the denom. is lower due to missing aPTT values. Weaver, et al12 See reference 11 for inclusion See reference 11 for details See reference 11 for details. Additional No significant differences R, DB, MC and exclusion criteria. efficacy measure was % achieving hemostasis between b-throm and rh- N=183 Patients having either PAB or at 1.5, 3, 6 and 10 minutes. throm were observed for the Subgroup of Phase III AV grafts procedures were following: ADEs, % success Study (reference 11) included in the analysis. Since No differences were observed success in in achieving hemostasis, peri- Zymogenetics (sponsor) vascular patients may undergo achieving hemostasis at 1.5, 6 or 10 minutes. operative transfusions, peri- *Dr. Weaver is a consultant more frequent surgical The only difference was at 3 minutes was and post-operative bleeding to Zymogenetics and another procedures, they have the statistically higher (p=0.046) in the rh-throm and thromboembolic events. author is the medical director potential to receive topical vs. b-throm group undergoing PAB surgery. for Zymogenetics thrombin more than once. No other differences were noted in % achieving *Post-hoc subgroup analysis. hemostasis for PAB, AV graft or PAB+AV All findings should be *Authors acknowledge that all graft. considered exploratory. statistical analyses in this subgroup were not a priori and No difference in peri-operative transfusions. should be considered

exploratory. ADEs: no differences were noted in the following: death, graft complications or thrombosis, serious ADEs, peri and postoperative bleeding complications, thromboembolic complications, or differences in laboratory values (PT, aPTT or INR) As in reference 11, a higher number of patients in the b-thrombin group seroconverted to having antiproduct antibodies. The potential for cross-reactivity to native human plasma thrombin was not investigated. Because of the reported associations between anti-bovine antibody development and increased bleeding complications, selected ADEs were examined for this group developing anti bovine product antibodies. There was a numeric disadvantage of ADEs (bleeding, thromboembolic events, hypersensitivity, and abnl aPTT), there were no statistical differences.

ADEs=adverse effects, AV=arteriovenous, b-thrombin=bovine-derived thrombin, CV=cardiovascular, DB=double-blind, h-thrombin=plasma-derived human thrombin, MC=multicenter, PAB=peripheral arterial bypass, PC=placebo-controlled, Pla=placebo, R=randomized, rh-thrombin=recombinant thrombin, TTH=time to hemostasis

Appendix B: Immunogenicity and Safety of Topical Bovine Thrombin Study Findings Comments/Authors Conclusions Carroll28 1996 Primary focus was on development of antibodies against fibrinogen. No hemorrhagic/thrombotic complications or allergic reactions N=21 cardiothoracic surgery pts Antibodies against bovine and human thrombin and factor V were also occurred in any patient despite the presence of antibodies against Bovine fibrinogen clotted with bovine measured (n=9). Antibody titers for bovine and human thrombin and bovine and human coagulation factors. thrombin. factor V were high and titers did not differ. The antibodies were not tested (Fibrastat bovine glue) to see if the activity of native human thrombin or factor V was neutralized in the presence of these antibodies. However, none of the patients had post-operative bleeding complications. Dorian29 1998  Overall, 12/120 patients developed antibodies to bovine thrombin *Antibodies not tested to see if human thrombin was neutralized by N=120 stored blood specimens from (10%)* antibodies against bovine thrombin. patients exposed to bovine thrombin  A single exposure to b-throm: 5% (5/102) developed antibodies identified via hospital records. Control against bovine thrombin. Authors note that patients having multiple exposures to bovine included 114 blood samples from patients  Multiple exposures to b-throm: 39% (7/18) were found to have thrombin are more likely to develop antibodies to coagulation not exposed. developed antibodies. factors, leading to severe bleeding problems in some patients, and  Two patients with multiple exposures did have severe bleeding should be avoided if there has been a previous exposure. complications.  2/120 (1.7%) patients developed severe bleeding complications (both received thrombostat on second exposure) Ortel30 2001  Baseline antibodies to the bovine thrombin preparation was higher in *Antibodies against human coagulation proteins were not tested to N=151 patients undergoing cardiac the group scheduled for surgery vs. the control group (23.8% vs. 6%, determine if they neutralized the activity of the native human clotting surgical procedures; immunoassay used p=0.018). proteins. to detect antibodies against bovine and  After exposure to bovine thrombin 4-7 days after surgery, there was human coagulation proteins (bovine and a small increase in antibody levels detected. In their conclusions, the authors suggest that the safety of bovine human prothrombin, thrombin, factor V  4-8 weeks postoperative, more than 90% of all patients had elevated thrombin needs to be reassessed and reexposure to these agents and bovine factor Va) before and after IgG levels to topical thrombin (independent of type of surgery, valve should likely be avoided. (24-48 hrs, 4-7 days and 4-8 weeks) vs. coronary artery bypass grafting {CABG} or thrombin dose). bovine thrombin exposure. Hematologic  Only about 50% of patients were found to have elevated IgM and clinical outcomes were also antibody levels by 4-8 weeks. collected.  Most patients developed IgG antibodies to bovine factor V and Va (Thrombogen) (Samples were compared (80.7% and 90.7%, respectively). to samples from 50 normal donors)  Nearly half of the patients developed antibodies to bovine prothrombin and about 20.5% developed antibodies to bovine thrombin.  Fifty-one percent of patients had elevated antibody levels to human coagulation factors primarily factor V, thrombin or both.*  As far as clinical outcomes, there was no increased risk for an adverse postoperative outcome in patients with a prior surgical history, use of higher thrombin doses during the procedure, presence of elevated antibody levels to bovine or human coagulation factors

after surgery or abnormal coagulation studies.  However, there was a significantly increased risk for an adverse postoperative outcome in patients who had elevated baseline (before surgery) antibody levels to two or more bovine coagulation proteins (adjusted odds ratio 5.4, 95% CI 1.54-18.8, p<0.01). Winterbottom31 2002  Antibody data was available for 270 patients  Antibody data for 270 patients. However, numbers of those with N=309 having cardiac, vascular or spinal  Baseline: antibodies against b-throm, b-factor Va or both were detectable antibodies and those without do not seem to add up surgery exposed to b-thrombin (Floseal® detected 8.6% (12/139: Floseal) and 9.9% (13/131: (100 with, 172 without). or Gelfoam®+b-throm) (Thrombin JMI) GelFoam+thrombin) (not significantly different)  No correlation was observed when comparing the incidence of (Gelfoam served as the control group)  6-8 weeks post-op: 100/270 (37%) antibodies against bovine post-op complications and presence or absence of antibodies to Blood samples collected pre-op and 6-8 thrombin and/or bovine factor Va were detected. bovine proteins (b-throm and b-factor Va). weeks post-op and assayed for antibodies  Of these 100 patients with detectable antibodies, 17 (17%) had PT  No patient developed antibodies against human factor Va. to bovine thrombin and factor Va. >15 sec (range 16.2-61.2). All but one were on anticoagulants. The (Blood samples were not tested for antibodies against human *Antibodies against human thrombin and remaining 85 patients had a normal PT. factor V.) factor Va were tested at 6-8 weeks post-  Of the 172 patients with no detectable antibodies, 12% (21/172) had op only in those patients with detectable a PT >15 sec (range 15.1-45.2). All except 4 were on anticoagulants. The authors concluded that the presence of antibodies to bovine antibodies to both bovine proteins. PT  14.8% of patients (8/54) with detectable antibodies to both bovine thrombin and/or bovine factor Va pre-op or post-op did not result in was measured to determine any clinically proteins tested positive for human thrombin antibodies. None of the an increased incidence of adverse clinical outcomes or abnl PT and significant coagulopathy. patients tested positive for human factor Va antibodies. “therefore the use of b-thrombin during surgery does not present an  In the 8 patients with detectable human thrombin antibodies, all had undue risk of bleeding resulting from an antibody-induced normal PT. coagulopathy.”  Five patients with known prior exposure to b-throm, PT normal range.  Of the 25 patients with baseline antibodies against b-throm and/or b- factor Va (suggesting prior exposure), one patient had an abnl baseline PT (70.8) and 3 others at the 6-8 week follow-up (46, 22, 26.7 sec). All were taking warfarn at the time.  384 complications in 144 patients. 75 blood related post-op complications (hemorrhagic, thrombo-embolic, wound, and other).  No significant correlation was found between presence or absence of detectable antibodies to bovine proteins at baseline or 6-8 weeks post-op. There was a trend between blood-related complications and prior surgery (p=0.08).

Doria10 2008 The authors concluded “no clinically meaningful differences in other One or more ADEs were reported in nearly all patients. However, R, DB, MC, n=305 variables for blood loss, length of stay, etc. Statistics were not provided only about 12% of the ADEs were considered to be related to study Phase III study (Ethicon sponsor) 4) Number who seroconverted: drug. No differences noted. 17% of h-thrombin vs. 11% b-thrombin (Add’l details, see appendix A) Antibody h- b- Diff reported serious ADEs, only 1.3% (n=2) in each group were related Human vs. Bovine thrombin throm throm to study intervention. Changes in lab values from baseline at 12 hour Any 4/122 16/126 . post-op and within 24 hours of discharge were similar between

3.3% 12.7% 09-. groups. 71 b-throm 3/122 10/126 .09- Although overall median differences were provided and stated to be 2.45% 7.9% 1.01 similar in: length of stay, care in specialty units, etc. no individual b-FV/Va 2/122 12/126 . assessment of those patients who developed antibodies to human 1.6% 9.5% 04-. thrombin (n=3) in the b-thrombin group was done or presented. 67 h-throm 0/122 3/126 0- Authors concluded that the safety profiles for human and bovine 0% 2.4% 1.3 thrombin were comparable. However, more patients developed h-FV/Va 0/122 0/126 ND seroconversion for 1 or more antibodies assayed in the bovine vs. 0% 0% human thrombin group. ND=not defined, several patients are represented twice since they developed antibodies to 2 or more but not counted twice in “any” category. *No patient seroconverted for antibodies against h-FV/Va Chapman11 2007  1) Nearly all patients reported at least one ADE. ADEs were similar between groups. R, DB, MC, n=463  Most were not considered related to study treatment and most were A post-hoc analysis, examining the association of posttreatment Phase III study (Zymogenetics sponsor) moderate in severity. seroconversion or increase in antiproduct antibody titers and (Add’l details, see appendix A)  Certain categories were prospectively followed: bleeding (11% b- development of clinical ADEs, (bleeding or thromboembolic events, Recombinant vs. Bovine thrombin throm vs. 13% rh-throm); thromboembolism (b-throm 5% vs. rh- hypersensitivity or high aPTT) was done and showed a numeric throm 6%); cardiac events (b-throm 18% vs. rh-throm 20%); disadvantage in those seroconverting and who received b-thrombin. hypersensitivity (b-throm 17% vs. rh-throm 14%); post-op wound (Post-hoc analyses should be considered only “hypothesis infections (b-throm 10% vs. rh-throm 9%); other infections (b-throm generating”) 15% vs. 12% rh=throm). Not reported to be significantly different. *Investigators did not test whether or not antiproduct antibodies  Common ADEs were similar (seroconverters or increased titers in those receiving b-thrombin)  Abnormal lab tests were similar neutralized native human plasma thrombin. Why not? They did  2) 398 evaluable for anti-product antibodies for those seroconverting on rh-thrombin. b- rh- Diff The authors conclude that recombinant and bovine thrombin are throm throm comparable in terms of safety profile, however, recombinant Baseline 10/200 3/198 -- thrombin is less immunogenic than bovine thrombin. Antibody Positive 43/200 3/198 p<.0001 Seroconvert 35/190* 3/195+ -- >titer unit change 8/10** 0/3 --

*Denom. includes only patients with a post-baseline measurement and no detectable antiproduct antibodies at baseline. **Denom. includes only patients with a post-baseline measurement and detectable anti-product antibodies at baseline. + 3 patients who seroconverted in the rh-thrombin group were tested and

the antiproduct antibodies did not neutralize native human plasma thrombin activity. The patients who received b-thrombin and either seroconverted or had an increase in their antiproduct antibody titers; these samples were NOT tested to determine whether or not the antiproduct antibodies were capable of neutralizing native human plasma thrombin. *Post-hoc analysis of bleeding or thromboembolic events, hypersensitivity or high aPTT was conducted for association of development of posttreatment antiproduct antibodies and ADEs. There were numeric differences with higher rates of these events occurring in patients who developed antibodies in the b-thrombin group. However, this is a post-hoc analysis and the denom. is lower due to missing aPTT values. Reynolds32 2008  No standardized definitions of bleeding events.  Risk of bleeding events not clearly defined but bleeding rates Literature review-thrombin sealants  Rates of bleeding in general surgery studies: 0.1-20.2%; most did not appear to be different in the studies with thrombin N=21 clinical trials addressing risk of reporting ranges from 2.6-4% exposed patients versus those studies in which patients were not bleeding in surgery, 5 used thrombin  Rates of bleeding in bovine thrombin studies: 0.0-13%; most exposed to b-thrombin. sealant reported ranges from 2-3%.  Authors recommend well-designed and controlled trials to better (Supported by King Pharm) examine the bleeding risk in surgical patients exposed and not exposed to b-throm. Clark33 2008  Immunogenicity studies did show an increased risk for development  Antibody elevation is increased in patients exposed to b-throm. Literature review-topical bovine of antibodies against coagulation factors after exposure to b-throm.  However, available evidence does not support this elevation in thrombin  However, there was not consistent evidence that the development of antibodies increases the risk for clinically important ADEs. N=25 case reports/case series with these antibodies resulted in an increased risk for ADEs. suspect ADEs/Lab abnl attributed to b-  Available data mostly from case reports, case series. throm.  Controlled trial comparing rh-throm to b-thromb did show greater N=10 longitudinal studies b-throm used antibody development in those exposed to b-throm but no important and ADEs reported. efficacy or ADE differences were observed. Abnl=abnormal, ADEs=adverse events, b-factor Va=bovine factor Va, b-throm=bovine thrombin, DB=double-blind, MC=multicenter, NS-not significant, pre-op=preoperative, post-op=postoperative, PT=prothrombin time, R=randomized

Updated versions may be found at www.pbm.va.gov or vaww.pbm.va.gov 25 December 2008