UNITED STATES OF AMERICA

FOOD AND DRUG ADMINISTRATION

CENTER FOR DEVICES AND RADIOLOGICAL HEALTH

CIRCULATORY SYSTEM DEVICES PANEL

 

MEETING

 

TUESDAY

SEPTEMBER 11, 2001

 

The panel met at 8:00 a.m. in Salon 8 of the Gaithersburg Marriott Washington Center, 9751 Washingtonian Boulevard, Gaithersburg, Maryland, Doctor Warren K. Laskey, Acting Chairman, presiding.

PRESENT:

Warren K. Laskey, M.D., Acting Chairman

Salim Aziz, M.D., Member

Michael D. Crittenden, M.D., Temporary Voting Member

Robert A. Dacey, Consumer Representative

Thomas B. Ferguson, M.D., Temporary Voting Member

Michael Morton, Industry Representative

Janet T. Wittes, Ph.D., Member

Megan Moynahan, M.S., Executive Secretary.

A-G-D-E-N-D-A

page no.

Introduction

Warren Laskey, M.D. 3

Special Presentation to

Doctor Michael Crittenden 7

Sponsor Presentation: CryoLife, Inc.

P010003, BioGlue Surgical Adhesive

Introduction

James C. Vander Wyk, Ph.D.

VP Regulatory Affairs and Quality

Assurance, CryoLife, Inc. 9

Non-Clinical Performance Results

David M. Fronk, VP Clinical Research

CryoLife, Inc. 17

IDE Clinical Protocol and Clinical Results

Joseph S. Coselli, M.D.

Baylor College of Medicine 28

Clinician's Information

Joseph E. Bavaria, M.D.

Hospital of University of Pennsylvania 37

FDA Presentation

Lisa Kennell, Lead Reviewer 47

Open Committee Discussion, Recommendations

and Voting 57

Panel Recommendation Options for

Pre-market Approval Applications and Vote 112

P-R-O-C-E-E-D-I-N-G-S

8:03 a.m.

DOCTOR LASKEY: Good morning. My name is

Warren Laskey. I am an interventional cardiologist from the University of Maryland, and I'd like to call this session to order.

This morning's topic is a discussion of a pre-market application for the Cryolife BioGlue Surgical Adhesive. I'd like to have Megan Moynahan, the Executive Secretary, read the conflict of interest statement.

MS. MOYNAHAN: Thank you. The following announcement addresses conflict of interest issues associated with this meeting and is made part of the record to preclude even the appearance of an impropriety.

To determine if any conflict existed, the agency reviewed the submitted agenda for this meeting and all financial interests reported by the committee participants. The conflict of interest statutes prohibit special government employees from participating in matters that could affect their or their employers' financial interests.

The agency has determined, however, that the participation of certain members and consultants, the need for whose services outweighs the potential conflict of interest involved, is in the best interest of the government.

Therefore, a waiver has been granted for Doctor Janet Wittes for her interest in firms that could potentially be affected by the panel's recommendations. Copies of this waiver may be obtained from the agency's Freedom of Information Office, Room 12-A-15 of the Parklawn Building.

In the event that the discussions involve any other products or firms not already on the agenda for which an FDA participant has a financial interest, the participant should excuse him or herself from such involvement, and the exclusion will be noted for the record.

With respect to all other participants, we ask, in the interest of fairness, that all persons making statements or presentations disclose any current or previous financial involvement with any firm whose products they may wish to comment upon.

DOCTOR LASKEY: Thank you. At this point, I'd like to have the panel members introduce themselves starting with Mr. Morton.

MR. MORTON: I'm Michael Morton. I'm the industry representative, and I'm an employee of W.L. Gore and Associates.

DOCTOR CRITTENDEN: Michael Crittenden, cardiac surgeon, West Roxbury V.A., Harvard Medical School.

DOCTOR FERGUSON: Tom Ferguson, cardio-thoracic surgery emeritus, Washington University, St. Louis.

DOCTOR AZIZ: Salim Aziz, cardiac surgeon, University of Colorado, Denver.

DOCTOR WITTES: Janet Wittes, statistician from Statistics Collaborative here in D.C.

MR. DACEY: Robert Dacey, consumer representative from Boulder County, Colorado.

MR. DILLARD: Jim Dillard. I'm the Director of the Division of Cardiovascular and Respiratory Devices in the Office of Device Evaluation for the Food and Drug Administration.

DOCTOR LASKEY: And, again, Warren Laskey, interventional cardiologist and to my right hand is.

MS. MOYNAHAN: Megan Moynahan. I'm the Executive Secretary of the Circulatory System Devices Panel.

DOCTOR LASKEY: Thank you. Thus far, there were no individuals requesting time to speak at the open public hearing. Has that changed? Then Megan.

MS. MOYNAHAN: I'd like to read through the voting status statement for today.

Pursuant to the authority granted under the Medical Devices Advisory Committee charter and dated October 27, 1990 and as amended August 18, 1999, I appoint the following individuals as voting members of the Circulatory System Devices Panel for this meeting on September 11, 2001: Michael Crittenden and Thomas Ferguson.

In addition, I appoint Doctor Warren Laskey to serve as panel chair for the duration of this meeting.

For the record, these people are special government employees and are consultants to this panel under the Medical Devices Advisory Committee. They've undergone the customary conflict of interest review and have reviewed the material to be considered at this meeting.

DOCTOR LASKEY: Thank you. Sorry to get out of order. If there were no individuals who request time at the open public hearing, I open and close the public hearing. Thank you.

We divert for a moment. We have a little bit of ceremonial function to pursue, so I'll turn things over to Mr. Dillard for a moment.

MR. DILLARD: Thank you, Doctor Laskey. I appreciate it. It's my honor occasionally to get to really highlight some of the service that you all really provide to us. It's with great pleasure today that we'd like to honor Doctor Michael Crittenden. Michael has served for many, many years on this panel and has been one of the real stalwarts and leader from the cardiovascular surgery perspective and has really been one of the people that we've looked to time and time again to not only serve in his role as a full voting member of this advisory panel but from time to time, like today, bringing him back for his services. Michael, I'd certainly like to extend my gratitude for all the years of service that you've dedicated to the agency as a special government employee and, for that, I've got a plaque for you and I'd like to just read a nice little note that Doctor Henney, before she retired, wrote about your service.

"Dear Doctor Crittenden, I'd like to express my deepest appreciation for your efforts and guidance during your term as a member of the Circulatory Systems Devices Panel of the Medical Devices Advisory Committee. The success of this committee's work reinforces our conviction that responsible regulation of consumer products depends greatly on the participation and advice of the non-governmental health community.

In recognition of your distinguished service to the Food and Drug Administration, I am pleased to present you with the enclosed certificate. And it's signed Doctor Jane Henney, Commissioner of Food and Drug Administration."

The plaque reads, "In recognition of distinguished service to the Circulatory System Panel of the Medical Devices Advisory Committee term from March 18, 1998 to June 30, 2001," and it's signed in addition by Doctor Jane Henney and David Feigal, the Center Director for Devices and Radiological Health.

With that, Michael, thank you very much.

DOCTOR CRITTENDEN: Thank you.

(Applause)

MR. DILLARD: And with that, I'll turn it back over to you, Doctor Laskey.

DOCTOR LASKEY: Thanks, Jim, and congratulations again, Mike.

At this point, I think we can get on with the task at hand and the sponsor's presentation.

MS. MOYNAHAN: And just a reminder for the folks who are speaking to introduce yourselves and state your conflict of interest.

DOCTOR VANDER WYK: Mr. Chairman, members of the panel, good morning. I am James Vander Wyk, Vice President, Regulatory Affairs and Quality Assurance, for CryoLife.

This morning I will be introducing the BioGlue Surgical Adhesive for vascular and cardiac repair. Further presentations will be made by David Fronk, Vice President, Clinical Research, concerning non-clinical performance, Doctor Joseph Coselli of Baylor College of Medicine, a participant in the BioGlue Surgical Adhesive IDE trial concerning clinical trial data and concluding, Doctor Joseph Bavaria of the Hospital of the University of Pennsylvania, also a trial participant, providing a clinician's viewpoint.

Additional CryoLife personnel and consultants present and available for further discussion include the individuals listed here.

In the following few slides and two brief videos, I will present the regulatory history of BioGlue Surgical Adhesive, its mechanisms of action, operation of the delivery device, and several examples of clinical applications.

BioGlue Surgical Adhesive was first approved for use in the United States in June of 1998 under an IDE for repair of acute type A aortic dissections. Two years later, a second separate investigative arm was added to expand the trial to cardiac and vascular repair. Using interim data, CryoLife sought and was granted a humanitarian device exemption in December of 1999 for acute thoracic aortic section repair.

To date, institutional review boards representing over 600 hospitals have approved the use of BioGlue for this indication. We estimate more than 5,000 patients have been treated in the U.S. under this HDE.

In January of this year, CryoLife submitted a PMA application for BioGlue Surgical Adhesive to be used as an adjunct to standard methods of cardiac and vascular repair such as sutures or staples to provide hemostasis.

BioGlue Surgical Adhesive may be used prophylactically or after a leak is detected to bond tissue layers, seal, reinforce anastomoses in cardiac and vascular surgical repairs.

In Europe, CryoLife received a CE Mark in November of 1997 for the adhesive permitting its use for sealing, reinforcing and adhering tissue in vascular repair and in March, 1999 for sealing air leaks in pulmonary tissue repair. This approval was followed by similar approvals in Canada and Australia. The approval process is ongoing in Japan. To date, Cryolife has received commercial distribution authorization in over 36 countries.

At this time, we have not received any reports of adverse events involving deaths or serious injuries attributable to BioGlue or reports of reactions or immune responses from the use of over 45,000 units commercially distributed world-wide. There has been no withdrawal of approval in any country nor any causes or actions to do so undertaken in any jurisdiction.

In our first video, we introduce the simple components of the BioGlue Adhesive system and demonstrate the assembly of the delivery device. In the basic starter kit, we have the main trigger-activated unit, a cartridge plunger, a dual chamber cartridge pre-filled with the adhesive components, and a tortuous path mixing tip of which four are supplied.

To assemble, the cartridge locks are released, the ratchet plunger is inserted into the base unit oriented by size and then pulled fully back. The cartridge, which requires only room temperature storage, is also oriented by barrel size, inserted into the slot in the base, and locked in place.

After removing the cartridge tip cap, a mixing tip is connected to the cartridge. Corresponding notches aid in aligning the tip. Final attachment is achieved by turning the locking ring at the base of the tip, either by hand or with a twist ring tool as illustrated here.

In the last assembly step, the plunger is advanced to the cartridge fill point. The device is now ready for priming which must be done only immediately prior to adhesive application.

The active components of BioGlue Surgical Adhesive are 45 percent bovine serum albumen and 10 percent glutaraldehyde mixed in a 4:1 ratio. In this ratio, there is a slight stoichiometric excess of aldehyde groups over free amine groups. The glutaraldehyde reacts with free amino groups of protein, first of the BSA, then tissue, forming permanent covalent bonds.

This process is initiated immediately in the mixing tip and continues on contact with the tissue. With synthetic material a vascular graft, BioGlue adheres by mechanical bonds as it polymerizes within the interstices of the graft material.

The unpolymerized BioGlue is stable in the body for greater than one year. Any reabsorption that does occur is a slow process. This is typical of any glutaraldehyde cross-linked protein such as enforcing heart valves. In areas of resorption, there is evidence of cellular infiltration accompanied by collagen formation in a normal healing process.

Regarding safety issues of sourcing the BSA, all serum is derived from cattle, only of U.S. or Canadian origin, which are slaughtered in USDA-inspected abattoirs. The purification process of the source BSA is capable of a 16.2 log reduction of the BSE based upon testing inoculated samples. This material holds a certificate of suitability from the European Pharmacopoeia.

Three examples of clinical use of BioGlue are presented in the following video. They are typical of the methods used and results obtained. These video segments were selected to illustrate prophylactic application of the BioGlue in commonly encountered anastomotic repairs and are used in a repair of an actively bleeding site in a vascular graft.

In this first section, we are observing the repair of the distal end of an aortic dissection. Excess tissue is removed in a normal manner to prepare for joining to a vascular graft. Thrombus and blood is removed from the false lumen. The BioGlue is primed immediately to prior to the first use by simply expelling a small amount onto a disposable sterile material. The false lumen is obliterated and the walls adhere together by filling the false lumen with BioGlue.

After the graft is -- in this case, a Dacron graft, BioGlue is applied to the anastomotic site to prevent bleeding. Since BioGlue reinforces the often friable tissue, pledgets may not be required in the repair.

We are now observing the proximal repair of the same dissection. As before, BioGlue is used to obliterate the false lumen after removing thrombus and blood. Notice that no clamping or other special manipulations are required. In these examples, no pledgets were required nor is there any accessory equipment necessary to activate the adhesive. After sealing an adhesive set-up, the anastomoses is checked for leakage.

This next video example illustrates the use of BioGlue to seal a Dacron to Dacron vascular graft anastomoses in a thoraco-abdominal aneurism repair. The adhesive begins setting immediately and achieves functional strength within two minutes. The flow is restored and the anastomoses is checked for leakage.

In this final segment, we are observing a suture link that has developed in a PTFE vascular patch used in a carotid endarterectomy. The blood flow is interrupted, the field is dried, and BioGlue is applied to the bleeding site. While it is necessary that the site is dry to enhance the binding of the BioGlue to the site, it can be observed that there is a small amount of blood present showing that there is some tolerance for residual fluid. When flow is restored, the site is checked for leakage.

Thank you. This concludes my opening remarks.

At this time, I would like to introduce David Fronk, Vice President of Clinical Research, to present our non-clinical data.

DOCTOR FRONK: Good morning. My name is David Fronk. I'm the Vice President of Clinical Research for CryoLife. I'd like to take the next several minutes and share with you the non-clinical performance results of our BioGlue Surgical Adhesive.

Specifically, I will summarize the following: in vitro shear strength, biocompatability, in vivo animal studies, histopathology, and immunotoxicity testing results.

BioGlue was designed to be a high strength adhesive capable of withstanding the forces and pressures of full systemic arterial pressure. Using a mechanical shear test methodology which simulates clinically relevant loading, the shear strength of BioGlue was compared to both Fibrin-thrombin and GRF and glues. GRF or gelatin resorcinol formaldehyde glue, is a comparable adhesive available in Europe. Results show that the BioGlue is at least fourfold stronger than either of these other adhesives.

Biocompatibility testing was performed using ISO 10993 standards. These tests were conducted using derivatives of polymerized BioGlue. ISO standards were used to define passing results. The mild to moderate results for the cytotoxicity tests were expected.

As described by Doctor Vander Wyk, the mechanism of action of BioGlue utilizes the glutaraldehyde present in the polymerized BioGlue to bind to the surrounding tissue. Therefore, any cytotoxicity is acute, localized, and isolated to the first few cell layers of the tissue. Additional biocompatibility tests and their conclusions are listed in the table shown.

I will now report on a series of four different animal efficacy experiments conducted with BioGlue. The goal of these models was to evaluate the safety and effectiveness of BioGlue in a variety of different clinical indications. The aim of the first experiment was to study the effectiveness of BioGlue as a surgical adjunct in the repair of acute aortic dissection. An aortic dissection model was developed that allowed for BioGlue to be used to obliterate the false lumen by adhering and reinforcing the dissected layers of the aorta and sealing the repair. Control animals received conventional repair techniques, namely suture closure of the false lumen.

Results from the study clearly show the potential clinical benefits of BioGlue. When compared to controls, the animals receiving BioGlue had decreased rates of acute post-repair rupture of the aorta, redissection of the distal surgical repair site, dissection progression, and chronic dissection formation.

To evaluate the ability of BioGlue to seal large diameter synthetic anastomoses, a Hemashield interpositional aortic bypass graft was sewn into the thoracic aorta of a coagulopathic sheep. Hemostasis of the anastomoses was controlled, either by BioGlue or Surgical, a commercially available hemostatic agent. Animals were monitored for rate of blood loss and total post-operative drainage.

Again, the benefit of BioGlue treatment as compared to control was demonstrated. Both the rate of blood loss and total blood loss was significantly reduced in the animals receiving BioGlue as an anastomotic sealant.

The two previous studies demonstrated BioGlue's effectiveness on large diameter native tissue and synthetic grafts. To further assess its utility in small diameter repair, a glued anastomosis study was conducted on coronary artery bypass grafts. This evaluation included both an in vitro assessment of anastomotic burst strength and an in vivo goat

study.

Though not recommended by this PMA application, the anastomoses in this study were completed using BioGlue as the primary means of joining and sealing the vessels. This was done to assess the feasibility of a glued anastomoses to facilitate attaching coronary artery bypass grafts during beating heart bypass procedures.

The in vitro results clearly demonstrated BioGlue's ability to adhere the joined anastomoses. All 12 glued anastomoses easily withstood two times normal systemic arterial pressure, 10 of which withstood pressures in excess of 500 millimeters of mercury without leakage or dehiscence. In the goat model, all animals survived surgery and all grafts remained patent, non-stenotic, and free of interluminal BioGlue.

The final animal study evaluated BioGlue on small diameter synthetic anastomoses. In a growing pig model, a PTFE graft was placed as an interpositional graft. The proximal anastomoses was performed using BioGlue while the remaining distal anastomosis which was sutured served as the control. All animals survived the procedure and completed their course of follow-up. All but one graft remained patent which the surgeon investigator attributed to technical failure at the time of surgery. Even though the animals gained up to 60 kilograms during the study, the BioGlue anastomosis remained firmly adhered and sealed to the growing proximal aorta and did not dehys or leak.

I would now like to turn your attention to the histopathologic assessment of BioGlue. The following three slides show a time course of tissue response to BioGlue from the aortic dissection animal model study. BioGlue, which can be seen on the left side of each photomicrograph, appears amorphous pink in HNE staining. At seven days, we observe a tight position and homogeneous layer of BioGlue between the layers of the dissection. A slight infiltration of mononuclear cells is present but, in general, there is little evidence of an inflammatory response.

At 31 days, BioGlue is again shown to be adhered to the host tissue. Entrapped red blood cells remained at the tissue glue interface and no evidence of inflammatory response is seen. At 91 days, we continue to observe the firm position between BioGlue and the vessel wall. Also at this time, the cellular response shows signs of an inflammatory response characterized by fibrosis.

A one year explant from the coronary artery bypass graft goat study showed the longer term histologic effects of BioGlue and its response to small vessels. There is a lack of inflammatory cells present and the BioGlue remains intact as seen in the lower left hand corner. To compare the tissue response in humans, we have included a 90 day explant from a patient who underwent a repair of a Type A aortic dissection using BioGlue. As seen in this montage of photographs of different magnifications, there's a paucity of inflammatory cells present. This appears to be less than previuosly noted in animals. BioGlue is present and firmly opposed between the layers of the dissection.

To contrast this response in humans, the following is a histologic section through suture material from this same patient. There is a significantly greater inflammatory response associated with the suture material as compared to the BioGlue.

To summarize the histopathology findings, we believe that the histopathologic observations of BioGlue are consistent with a typical foreign body reaction. This is analogous to what is seen with other long-term implants.

Finally, I would like to review the immunotoxicity testing results of BioGlue. Per the FDA's testing guidance, I will address the following: hypersensitivity, inflammation, immunosuppression, immunostimulation, and autoimmunity. Data from the various biocompatibility in animal studies were used as the basis for these analyses.

Several evaluations of hypersensitivity-based immunotoxicity were performed. The Buehler test, which doesn't used adjuvant, and the Kligman test, which uses adjuvant, were conducted using topical exposure of BioGlue extract. The passing results from these two classic hypersensitivity tests showed a severity of zero which, by definition, is characterized as a grade one weak allergenic potential response.

I would like to elaborate on the previously listed antigenicity test in guinea pigs. This study had two parts: an active systemic anaphylaxis test and an antigen antibody test. For the anaphylaxis test, guinea pigs received a subcutaneous application of either BioGlue or bovine serum albumen as a sensitizing dose. At 14 days, the animals were challenged with an additional dose of intravenously administered adjuvant added extracts of BioGlue or BSA. A saline control was also administered.

As can be seen, the animals showed a significant anaphylactic reaction to bovine serum albumen but, on average, less than a weak reaction to BioGlue. Again, for ISO standards, a weak reaction is defined as a grade one passing response.

The second part of this experiment look at the antigen/antibody potential after exposure to BioGlue or bovine serum albumen. Eleven days after sensitization, diluted sera from the guinea pigs were tested against BioGlue, bovine serum albumen, and a vehicle control antigen using ELISA techniques. The bar chart summarizes these data. The three paired bars represent the original sensitizing agent. Within each paired bar are two lanes showing the titer level to re-exposure of either BioGlue or bovine serum albumen antigen. As you can see, animals that received either BioGlue or BSA had low levels of antibody against BioGlue. Animals sensitized with BioGlue had higher titers of antibody against the bovine serum albumin antigen.

Based on these hypersensitivity tests, we believe that there is a low risk of hypersensitivity reaction by the repeated use or long-term exposure of BioGlue. Once sensitized, other medical devices or medicines containing bovine serum albumen theoretically may induce an anaphylactic reaction. Please note: in recognition of this potential, clinical trials and product labeling has always contraindicated patients with a known hypersensitivity to bovine products. In addition, as previuosly stated, we have received no reports of reactions to BioGlue in over three and a half years of clinical use.

Looking at inflammation as a sign of immunotoxicity, you will recall from the previously shown photomicrographs, we conclude that there is no evidence of an abnormal inflammatory response to BioGlue.

With respect to immunosuppression, our tests revealed no evidence of post-resistance-based immunotoxicity. Similarly, our tests revealed no overt evidence of immunostimulation.

The Handbook of Toxicology recommends the histopathologic assessment of organs, particularly kidney and thyroid, for identifying signs of autoimmunity. Review of these results from our 90 day subcutaneous study in rats revealed no evidence of autoimmunity. This is clearly demonstrated in the 90 day rat kidney histopathology. Note the lack of observable inflammation in either the glomerulus or renal arterioles.

In summary, we believe that the results from the non-clinical testing demonstrated that BioGlue Surgical Adhesive is biocompatible and effective for its intended use.

I would now like to introduce Doctor Joseph Coselli who will address the IDE clinical protocol and clinical results.

DOCTOR COSELLI: Doctor Laskey and panel, I'm Joseph Coselli. I'm a cardiovascular surgeon in Houston, Texas with Baylor College of Medicine. I have had in the past nor do I have today no financial interest in the company involved. They are, however, compensating me for my expenses of being here this morning.

I'd like to present to you the information with regards to the clinical trial. Hemostasis is critical to the success of cardiovascular surgery. To all of us involved in conducting these procedures, this is intuitive as well as it to most of our patients. Also well known is that re-operation for bleeding is associated with significant morbidity, mortality and expense. It's certainly a drain on our clinical resources. Immediate hemostasis has the potential to decrease the need for blood products and associated risks thereof.

The current hemostatic devices are difficult to use. These would include devices which are of similar ilk but are not able to be stored at room temperature, difficult to apply, etcetera.

The objective that the trials seek was to demonstrate a decrease in the frequency of intra-operative anastomotic site bleeding in patients receiving BioGlue as an anastomotic prophylactic sealant as compared to patients receiving standard surgical anastomotic repair. This was done as a multi-center trial allowing each of the individual surgeons conducting the procedures to proceed, both with their usual techniques and in the environments in which they were comfortable in treating their patients.

The primary end point focused upon anastomotic hemostasis at each of the repaired anastomoses. Hemostasis equated to that the anastomosis did not require any additional agents, additional suture placement, pledgets or other hemostatic agents or even more BioGlue to control bleeding at any point during the course of the operation following completion of the initial reconstruction.

Success was evaluated twofold. 1) by anastomosis. There were multiple anastomoses carried out in most of the patients and by patient on an individual basis. A single failed anastomosis constituted a patient failure. Data was accumulated throughout the trial on a number of secondary end points to include the volume and type of blood product replacement, the use of additional hemostatic agents, returning the patient to the operating room for bleeding or re-opening the patient for bleeding, procedural complications, and mortality.

The study design included a prospective randomized control multi-center trial. The study group was comprised of standard repair for whatever that particular surgeon at that institution generally used plus the additional application of BioGlue prophylactically. The control group was the use of the standard repair without the glue.

Uncontrollable bleeding in the control group allowed the surgeon to opt out to cross over to BioGlue. This actually did occur in one patient undergoing a transverse aortic arch replacement in one institution where the distal anastomosis, the tissue was so friable that the surgeon felt that bleeding was an immediate risk to the patient and, consequently, in the trial there was one single cross over. The patient did survive.

With regards to sample size, assumptions were made -- and I might add that these were based upon information derived from the various institutions and were, I think, exceedingly conservative. The rate of anastomotic bleeding was estimated at 15 percent, BioGlue reducing bleeding to achieve five percent, and that most of the patients, the overwhelming majority, would have at least two anastomoses and actually quite a few of the patients had multiple anastomoses. Five, six, seven, eight or more.

The power calculation really estimated a sample size of 142 patients to include a potentially inflated group for possibly a 10 percent dropout and 10 percent cross-over for uncontrollable bleeding. The actual dropout for primary end point was zero and the cross-over was one single patient, as I've already mentioned. The study enrollment stopped at 151 patients.

Inclusion criteria involved patients undergoing a cardiac or vascular surgical repair and patients who, or their legal authorized representative, was willing and able to give prior informed written consent.

Exclusion criteria included proteins with a known hypersensitivity to albumen, bovine products or glutaraldehyde in patients treated with an investigational product who have not completed an ongoing study. Also, patients excluded included those undergoing repair of intra-cerebral circulation and patients undergoing repair of acute thoracic aortic dissections which was covered under a separate arm of the study and the BioGlue HDE.

The centers are listed here, and they included Houston, Philadelphia, Indianapolis, Orlando, Florida, Atlanta and Loma Linda, and all of the PIs were considered to be accomplished cardiovascular surgeons.

The study enrollment of 151 patients was randomized to 76 BioGlue for the treatment group and 75 randomized to the control group. One control patient, was I mentioned before, was crossed over to the BioGlue side. The safety data for that patient is included with the BioGlue group, but the efficacy data is excluded from all of the analysis. The two groups, the treatment group and the control group, based, as you can see here, on gender and race, were virtually identical and quite homogeneous.

The variety of procedures carried out were indeed diverse and they included quite a number of cardiac procedures, aortic procedures, and peripheral vascular operations. The distribution of these categories for the treatment group and the control group, however, was statistically similar. With regards to achieving the primary end point, both by patient and by anastomosis, it was statistically significant in favor of the treatment group using the BioGlue.

With regards to secondary end points -- and here is shown the use of intra-operative blood products, both by volume and by category of components -- the treatment group using BioGlue and the control group were essentially identical with no difference. Similarly, the use of post-operative blood products, again shown here listed by component, was identical, both for the BioGlue group and the control group.

Interestingly, again focusing on secondary end points, additional hemostatic measures -- three columns on the right -- were identical with the treatment group and the control group. However, the use of pledgets for the primary anastomosis was less in patients in the BioGlue group than in the non-BioGlue group. What this indicated to us was that surgeons as the trial progressed were evolving to use less pledgets during the construction of their initial anastomosis than they might have been otherwise as they developed comfort with the use of the material.

With regards to the use of hemostatic measures as shown on the right, the use of BioGlue did not preclude or interfere with the use of additional efforts to secure hemostasis when it was felt to be necessary. With regards to mortality, there was no statistically different significance between the control group and the treatment group.

Procedural complications. There was no difference with regards to hemorrhage, infection, inflammatory response, irreversible morbidity, ischemia, myocardial infarction, organ system and multiple organ failure between the control group and the treatment group.

There was a difficult to explain reduction in neurological deficits, primarily in the form of significant stroke, in the treatment group with the use of BioGlue. In discussing this with the other investigators, clearly in the treatment of many patients with regards to the replacement of the transverse aortic arch and aortic root, etcetera, bleeding frequently translates into episodes of hypotension and hypotension in such patients is related to the development of stroke in such patients. This possibly is an explanation for this particular data.

With regards to other complications -- and they are diverse because of the diverse procedures, types of procedures used in the trial. Everything from, as I've mentioned, root replacement to peripheral vascular procedures -- include such things as paraplegia, pleural effusion, kidney failure, pulmonary failure, stroke, thromboembolism, thrombosis and other, and there's no statistically significant difference here.

Product-related complications were few. There was one implication where the glue was incidentally applied to non-target tissue which was treated simply by removal with sharp dissection of the glue separating it from the surrounding tissue, and this was resolved without any additional sequelae whatsoever.

There was one failure of BioGlue to adhere and the surgeon at the time felt that it was applied to a field which was too wet at the time, re-cross clamping, achieving a dry field and reapplication of BioGlue resulted in a successful therapy. The risk of these problems and complications are specifically, however, provided in the labeling that we had available to us.

In summary, anastomotic hemostasis as the primary end point was statistically superior in the BioGlue group, both by patient and by anastomosis. Procedural complications were similar between the two groups and second end points, post-operative blood product administration, the need for re-operation for bleeding and mortality, were not statistically different. BioGlue significantly, however, decreased the need for pledgeted sutures to reinforce the vascular tissue at the time of the primary anastomosis.

To conclude, BioGlue is more effective in achieving immediate anastomotic hemostasis when compared to standard repair. BioGlue is safe as a standard repair, and it decreases the need for pledgeted sutures to reinforce vascular tissue.

Next, Doctor Joe Bavaria will give us the clinician's information. Thank you.

DOCTOR BAVARIA: I'm Joseph Bavaria, a cardiac surgeon and professor of cardiac surgery at the University of Pennsylvania in Philadelphia.

Regarding disclosure, I have no equity interest in Cryolife or any sort of position within the company. However, I am compensated for today's presentations and the time today, and I was a primary investigator in the study obviously.

At the University of Pennsylvania, our clinical trial had 43 test and control subjects which were equally distributed between cardiac surgery patients and vascular surgery patients. The investigators were myself as primary investigator and Doctor Jeffrey Carpenter, who's a peripheral vascular surgeon who performs some of the peripheral vascular operations.

Today I wanted to go through five cases that demonstrate various applications and various issues regarding BioGlue in the clinical arena.

The first patient is a 77 year old female who has a past medical history of previous AAA repair, abdominal aortic aneurysm repair, a right carotid endarterectomy, hypertension, hypercholesterolemia, insulin-dependent diabetes, peripheral vascular disease and was smoking up to the day of surgery. She had a distal aortic anastomotic problem from her previous AAA repair.

The arteriographic data reveals an eight centimeter Crawford Type III thorical abdominal aortic aneurism. A Crawford Type III aneurism is an aneurism that includes the abdominal aorta as well as half of the thoracic aorta. You can see that she has a previous III graft right in here. This is where the anastomosis was. She has left renal artery involvement of this aneurism. She has a left iliac occlusion. As you can see, this only has one limb and obviously a previous AAA graft.

This operation included five separate anastomoses. The proximal anastomosis was tissue to polyester and the mid-thoracic aorta are tissue to graft. The distal aorta was polyester to polyester secondary to a previous AAA graft. She had a mesenteric segment. The mesenteric segment includes a patch graft to the celiac artery, SMA artery and right renal artery. That's a side to side anastomosis tissue to polyester. And she had a separate left renal artery bypass as a side branch which actually includes two anastomoses, the graft to graft proximally and the distal anastomosis at the tissue at the renal artery. So five anastomoses. She had 100 percent hemostasis at all sites in this operation.

The second operation is abdominal aortic aneurism repair. I chose this particular one. This patient actually did not do well. He died four or five days later, but I wanted to present this to show exactly how we can get out of a pretty serious situation. This is a 79 year old very elderly frail gentleman who has a history of TIA, COPD, alcohol abuse, and again, current tobacco use. He had a common femoral artery aneurism as well, and his medications included aspirin and Plavix.

He had a six centimeter abdominal aortic aneurism and the proximal anastomosis involved accessory renal artery vessels. The distal anastomosis was to the aortic bifurcation. This is the proximal anastomotic site right here which shows that this aorta is very friable. Under a little light in the area here, you can see that there's actually a lot of calcium intra-luminal here which basically is an indicator of an aortic wall which is basically falling apart.

This patient had two anastomoses, that proximal aorta which was tissue to polyester, and the distal aorta which was tissue to graft as well. The proximal aorta was extremely friable and tore during the end to end proximal anastomotic completion. This was prior to BioGlue application. This was just during the actual reconstruction. We had to take the clamp off, reclamp at a higher level, excise that portion of the aorta, and redo this anastomosis at a higher level. Again, the aorta was extremely friable. Proximal anastomosis was redone and reinforced with BioGlue at this time. The distal anastomosis achieved primary hemostasis as well as a proximal anastomosis.

There was significant operative bleeding due to vessel friability, and my partner, Doctor Carpenter, credits BioGlue with operative survival.

The third patient is another thoroco-abdominal aortic aneurism. It's a different type. It's a Type I thoroco-abdominal aortic aneurism which includes the entire thoracic aorta and the upper portion of the abdominal aorta. This is a 48 year old female who has anxiety, hypertension, again current tobacco use, and a very symptomatic aneurism.

The radiographic data shows an eight centimeter or maybe even larger Type I thoroco-abdominal aortic aneurism. This aneurism had significant compressive symptoms, as you can see here, with the cardiac structures completely squished up to the level of the sternum here. This is the left atrium which has almost no cavity as well as left ventricular compressive symptoms. There was anterior displacement of the cardiac structures.

She had a two anastomosis Type I thoroco-abdominal aortic aneurism repair. The proximal aortic anastomosis was done at the level of the distal aortic arch at the left subclavian artery. The distal aortic anastomosis was also performed at tissue to polyester and was just proximal to the celiac access but infra-diaphragmatic. There was 100 percent primary hemostasis achieved at both anastomoses and no blood transfusions required.

A little point here is when Doctor Stanley Crawford, who performed 1,509 thoroco-abdominal aneurysms and reported this in 1991 or 1992, this was a 30 year experience that ended in January of 1991. Every single one, 100 percent of the patients who had thoroco-abdominal aneurysms over that time period had blood transfusions.

The fourth patients is an ascending aortic repair. This is a 47 year old female with rheumatic heart disease, hypertension, New York Heart Association Class II heart failure symptoms, anxiety, and a localized dissection of the ascending aorta, what might be considered DeBakey Type II dissection.

She had aortic valve insufficiency which moderately severe aortic valve regurgitation.

The ECHO anatomic data reveals a six centimeter ascending aortic dissecting aneurism, a localized dissection and cardiac function which was depressed revealing a left ventricular end diastolic diameter of six centimeters and an end systolic diameter of 4.3 centimeters. That's a typographic error. That's systolic diameter of 4.3 centimeters. She had a severely dilated aortic root.

This operation was a compositive graft. There was a four anastomosis mechanical composite graft with coronary artery transfer. What we see is we have a proximal anastomosis at the level of the heart which is the aortic root tissue to polyester anastomosis. We have a distal ascending aortic anastomosis way up at the level of the first part of the aortic arch, a right main coronary transfer, and a left main coronary transfer which may be actually the most important of the four anastomoses which is back here. You can't see it on this picture.

This is quite different than your standard aortic valve replacement which basically only has one single aortotomy suture line. The hemostases achieved was a primary hemostases achieved at all locations and there was no blood utilization required in this patient. BioGlue was placed on all four anastomoses.

The last patient is a very significant patient, in my opinion. This is a patient who I believe would not have survived in an earlier era. This is a 50 year old gentleman with hypertension, coronary disease, renal failure, and is hemodialysis depending. He actually had hemodialysis the day before hospitalization. He also has current tobacco use. He presented with an acute Type A aortic dissection. This patient was part of the Phase I HDE study.

Intra-operative TEE revealed a dissection flap noted at the aortic root, a complex dissection lap with moderately severe aortic valve insufficiency which is typical in patients presenting with acute Type A aortic dissection. What was most impressive about this patient was that his baseline hemoglobic is only nine to begin with. We got that from the hemodialysis records. He presented with a hemoglobin of approximately seven and basically not a single functioning platelet in his whole body, which is typical.

We took the patient and actually prior to surgery, actually while he was on the table prior to incision, primed his cardiopulmonary bypass machine with FFP and packed red blood cells as a CPB prime because we didn't think we could get out of this any other way. And we proceeded with repair. The repair included a BioGlue placement into the false lumen right here of the distal anastomosis as a hemi-arch basically obliterating the false lumen and sealing the adventitia to the intima and creating a quote/unquote neomedia or a new media basically of BioGlue.

That was the first application of BioGlue and then at the root after re-suspension of the aortic

valve, placed BioGlue in the false lumen at the aortic root sealing the adventitia to the intima and obliterating the false lumen at the aortic sinuses. And this is how we apply BioGlue in these patients and then after the graft is put in, we put an additional application on the grafting anastomosis.

Surgical outcome of -- catastrophic presentation. He had an intimal flap extending into his left subclavian artery. He had innominate and left carotid artery flaps repaired utilizing a Dacron graft, as I just said, and he was discharged on day 30 in good condition. The key here was he received an incredible amount of blood products, as is typical of an early era type aortic dissection, and the only spot that he was not bleeding from to my eye was the BioGlued areas. In over 200 -- aortic dissections, he's the only patient I've seen who actually bled through the interstices of the graft after completion of the operation.

This concludes my presentation and the presentation for CryoLife. Thank you, Doctor Laskey, for chairing this session.

DOCTOR LASKEY: Thank you, gentlemen, for that illuminating presentation.

We'd like to turn now to the FDA presentation and our lead reviewer for this PMA is Lisa Kennell.

MS. KENNELL: Good morning, panel members and audience. My name is Lisa Kennell and I was the lead reviewer of the CryoLife BioGlue PMA. My presentation today will focus on the regulatory history of the BioGlue, a summary of the clinical study, and an overview of the non-clinical testing issues about which we would like some panel discussion and the questions for the panel.

This next slide lists the main team members who helped in the review of the PMA, some of whom are here today. There were others who are not noted on the slide for the sake of brevity.

There has been a regulatory history with the BioGlue and some of the information provided in previous submissions led us to the PMA we are reviewing today. I would like to take a few minutes to go over this history with you as it may provide important background for today's discussion.

The first involvement that the Division of Cardiovascular Devices had with this biological adhesive was the Investigational Device Exemption submitted in 1998. The study was first proposed for the use of Bioglue as an adjunct to Type A ascending aortic dissection repair.

Shortly after the IDE study was approved for this patient group in June of '99, the sponsor opted to take advantage of a relatively novel marketing submission called an HDE or Humanitarian Device Exemption. These exemptions are granted for patient populations or diseases that are rare, that is, less than 4,000 cases in the U.S. per year, and for whom there are no options or the options are inadequate.

Another major distinction between the HDE route to approval and the PMA route is that FDA considers only the safety and probable benefit of the proposed device under an HDE and not the effectiveness when granting approval.

In their HDE submission, CryoLife expanded the patient population to include both Type A and Type B descending dissections.

After granting approval of their HDE, the sponsor began to have protocol deviations in the areas of randomization, consent and off label use. FDA encouraged the sponsor to modify their protocol to capture some of the off label use so that they could assure that the ultimate label and indication covered the gamut of patient populations and diseases desired. Thus began the cardiac and vascular study arm, which is the subject of today's discussion.

I have placed a slide containing the indication for use of the BioGlue verbatim from the proposed instructions for use to keep in mind during our discussions today. The sponsor proposes to indicate the BioGlue for adjunctive use with standard methods of cardiac and vascular repair such as sutures or staples to provide hemostasis.

The next set of slides gives an overview of the clinical study. The use of the BioGlue was randomized to standard surgical hemostases methods. The sponsor wanted to show a 10 percent improvement in hemostasis using the BioGlue. To accomplish this, a sample size of 86 patients per treatment group was estimated, factoring in cross-overs and loss to follow-up. Since there was only one cross-over, the study was completed after entering only approximately 75 patients per group.

Patients needing cardiac or vascular repair who gave consent were entered. Those who had sensitivity to bovine material, glutaraldehyde or albumin who had another investigational device who needed intra-cerebral circulation repair or acute thoracic aortic dissection repair were excluded.

This slide shows that the majority of the patients had the BioGlue used adjunctively to repair aneurysms. A few needed repair of peripheral or carotid vessels and the rest of the cases involved root dilation or valve surgery.

The primary end point was anastomotic hemostasis which was defined as no need for additional agents to control bleeding at any time. Secondary end points included exposure to donor blood productions, additional hemostasis agents needed, re-operation for bleeding, major and minor adverse events, mortality and collection of cost benefit data such as bypass, cross-clamp, operative time and time in the ICU and hospital.

FDA noted the following outcomes from the study. The superiority hypothesis of 10 percent improvement in hemostasis was met when considered using either a patient denominator or an anastomosis denominator. With the exception of the number of pledgets used for the repair, no improvement in the secondary end points was noted when using the BioGlue.

With respect to safety results shown on the next several slides, only neurological deficit was observed to be significantly different between the two groups in favor of the BioGlue. The non-clinical testing methods were acceptable and covered the all important issues. However, there is information provided in the submission relating to immunogenicity testing about which the FDA would like to have some discussion and comment from the panel today.

With respect to the issue of immunogenicity, the firm conducted a gamut of studies to address this issue which are summarized in the next several slides. They conducted several sensitization tests in which adjuvant and ELISA testing for antibody production was assessed. They also assessed compliment activation and the amount of unbound bovine serum albumen protein after various polymerization times ranging from one and a half minutes to 24 hours. Both the BCA and Lowry methods were used for this assessment. Biodegradation was also assessed in vivo.

Results of these studies suggested low titers of antibody to BioGlue and albumen were observed in the ELISA assays in the sensitized animals. Unbound protein was found in the Lowry assay but not in the BCA assay and varied biodegradation reactions were noted from encapsulation or local inflammatory response to degradation.

We solicited review and comments from Doctor Henry Homburger who is a consultant on the Division of Clinical Laboratories Clinical and Biochemical Immunology Devices Advisory Panel and the Director of Clinical Immunology at the Mayo Clinic. The next slides summarize his comments. He noted that the data are inconclusive but that there are antibodies produced that are specific to BioGlue and albumen and that T lymphocytes probably persist that could respond to BioGlue or related proteins upon repeated exposure.

He further noted that these may not be indicative of a clinically significant immune reaction. He further indicated that a transitory immune response is not likely to be clinically significant but may prime the immune system for subsequent exposures which could be clinically significant. If antigen persists at the implant site, there is a theoretical risk of immune complex mediated diseases.

He recommended that it would be difficult to design additional animal studies to evaluate the human risk. Furthermore, he recommended that product labeling should include warnings in the labeling regarding use in patients who are sensitive to bovine products -- which is already in the labeling -- and in those with a prior history of immune-mediated diseases. In addition, he recommended a warning about repeated use and a post-market clinical and in vitro study to assess antibody production.

We have the following questions for the panel, separated into those relating to safety, effectiveness and labeling issues. Questions relating to effectiveness include, #1, the sponsor proved their primary hypothesis of a 10 percent improvement in hemostasis which was defined as no need for additional agents during the procedure but did not show an improvement in the secondary end points. Please discuss the clinical implications of the primary and secondary end point data.

#2, the sponsor states in the submission that, quote, "Our clinical investigators believe that the routine use of BioGlue in these patients will allow them to modify their blood management protocol and should minimize the potentially life-threatening complication of post-operative hemorrhage." Please comment on whether there is adequate information to support the statement.

Questions relating to safety and effectiveness, #3, based on the information provided in the pre-market approval application, please discuss whether the information supports reasonable assurance of safety and effectiveness of the BioGlue.

Questions relating to safety and labeling. One aspect of the pre-market evaluation of a new product is the review of its labeling. The labeling must indicate which patients are appropriate for the treatment, identify potential adverse events with the use of the device, and explain how the product should be used to maximize benefits and minimize adverse effects. Please address the following questions regarding the product labeling and safety.

#4a, please discuss the findings of the immunogenicity testing, especially as the relate to both physician and patient labeling issues. Should patients be advised of specific adverse events to be aware of that may suggest they are experiencing a sensitization reaction from the BioGlue?

#4B, the sponsor conducted several animal studies to assess the potential for BioGlue to elicit an immune reaction. The information from these studies suggests that there may be a potential for sensitization to the bovine serum albumen and related proteins in the formulation. Information from the clinical studies is limited to assessing the product with short-term follow-up. Please discuss whether sensitization has been adequately addressed with the clinical data as supplied or are additional post-approval studies needed to assess the immune potential of BioGlue?

#5. Please comment on the indications for use section as to whether it identifies the appropriate patient population for treatment with this device. The indications verbatim from the labeling again are, quote, "BioGlue Surgical Adhesive is indicated for use as an adjunct to standard methods of cardiac and vascular repair such as sutures or staples to provide hemostasis." End quote.

#6. Please comment on the directions for use as to whether they adequately describe how the device should be used to maximize benefits and minimize adverse events and, finally, #17, do you have any other recommendations regarding the labeling of the device?

Thank you for your attention.

DOCTOR LASKEY: Thank you.

I'm going to open the panel portion now by having our lead reviewer, give his review and ask questions of the sponsor. After his portion, we'll all the other members of the panel to also question the sponsor. In the interest of just staying on schedule, I'd like to confine each panel member's delivery to 10 minutes or less. Doctor Aziz, thank you.

DOCTOR AZIZ: First, let me say I very much enjoyed listening to both the company's presentation and also the cardiovascular surgeons who were both expects in their field, Doctor Coselli and Doctor Bavaria.

I was just going over the in vivo and in vitro testing. I think a lot of the times what we find in animals really doesn't actually translate into the clinical setting. This may be one of those sort of situations. I must say I was looking at the antibody ELISA testing. There did seem to be an elevation in antibody generation, but it seems that for the large number of clinical cases it doesn't translate into a clinical problem. But I think that clearly should be watched over a period of time to see if that does propose to be a problem.

I'm going to focus most of my questions to its use in the clinical situation and a couple of these questions are targeted towards Doctor Coselli. Patients with aortic dissection are particularly troublesome and difficult, and I think there have been improvements, particularly in the grafts that we have now compared to what we had years ago. The collagen-impregnated grafts. But still, I think technically they are quite a challenge, as all of us know.

A lot of the times when you have the Type A dissection or Type I DeBakey dissection, when you apply the glue, you clearly showed that you are obliterating the false lumen at the point at which you are applying it. But the false lumen distally clearly still stays in the patient. Am I right in assuming that?

DOCTOR COSELLI: I believe you're right in assuming that. I think that the patency of the distal false lumens in a complex issue. Patency of the false lumen requires in flow and out flow. One of the things that was done to reduce that was this open distal anastomosis under hypothermic circulatory arrest eliminating the potential for cross-clamping the aorta proximal to the -- artery creating a fracturing of the inter-lining which would allow either in flow or out flow.

Other sources include the suture line and then finally, distal openings between the true and the false lumen generally at sites of branch vessels. The first situation, using hypothermic arrest eliminating the cross-clamp, takes that off the table. The BioGlue addresses the second issue. In other words, that the actual passage of suture through the tissue as a source of entry or exit from the false lumen. The remaining one, the openings due to a tearing away at the level of the branch vessels, is really not addressed in this particular approach.

DOCTOR AZIZ: When you applied the BioGlue and you obliterate the false and true channels, at least in a localized fashion, then you don't have to apply any felt at the point you do your vascular graft anastomosis. Am I right in assuming that from what I saw in the video?

DOCTOR COSELLI: Absolutely right.

DOCTOR AZIZ: When you have a dissection, obviously you look for the tear. Now, can the glue be used to obliterate the tear or should the tear be removed and your graft should actually be interposed?

DOCTOR COSELLI: The way we've managed these, if the tears in the ascending aorta are in the arch, we try to eliminate the tear as part of our replacement effort. If the tear is well down into the descending or below and we're treating a Type I dissection, then we've in effect taken an acute I dissection and converted it to an acute III for continued medical management.

I think where this particular product has altered what we've done is that before, I've been concerned about spiraling tears in and around the base of the brachial cephalic vessels. We've always wondered whether or not should we replace the base of that vessel with a separate inter-position II graft or not and we've placed pledgeted sutures inside the lumen in order to try to secure the situation and to allow for a hemi-arch or a bevel approach without being overly concerned with extension into the head vessels. I actually encountered this particular situation last Friday night on a case.

What the BioGlue allows us to do is is to seal the false lumen at the level of the brachial cephalic vessels and take that particular concern again off the table.

DOCTOR AZIZ: If you were doing a valve conduit, is there any danger of the BioGlue, for example, getting onto the valve itself? Do you see that as a potential danger?

DOCTOR COSELLI: I don't believe that it would be appropriate to allow the glue to get down onto the valve tissue. Consequently, what most of us have done, it's very simple to avoid it. It is a matter of just taking some moist gauze and placing that down inside the root covering the valve leaflets while the application is carried out into the false lumen.

Also, the applicator tip for this particular device is a fine enough instrument that you can pretty well easily control the amount of material that you're extruding from the device and control where it's going and it isn't. It's not something that is just running all over the field by and large.

DOCTOR AZIZ: There's another subset of patients who I think might be benefitted from its use which I think was included in your cases. Patients with traumatic aortic tears where the tear is really localized and you may get obviously a little dissection very, very localized. A number of patients have addressed that issue in terms of doing primary anastomosis, though most people don't do that. But it seems to me that this might provide the impetus for really using primary anastomosis techniques. What would your thoughts be on that?

DOCTOR COSELLI: I haven't used it for that particular purpose, but I certainly would consider it.

DOCTOR AZIZ: And in the cases of patients, could this also be used as a patch repair for venous tears and atrial tears? For example, if you had a right atrial tear and you were using circ arrest to sort of try to control that, both the investigators -- could you see this being applied on a pericardial patch and then placing that on the atrial tear to make that easy or IVC tears during redos?

DOCTOR COSELLI: I haven't done that, but I think the utility here would be clear. One of the things that happens when this is applied to the tissue is it takes very friable tissue and makes it far more substantial for the placement and the suture purchase and, although I haven't used it for the inferior vena cava or for the atrium, I think if you had a very thin dilated atrium where that might be a concern, this is certainly something that could be considered.

One of the things that we see in acute dissections in the way we used to carry out the anastomosis, I've personally gotten away from felt strips a long time ago. But because of access into the false lumen at the distal suture line and occasionally at the proximal suture line, you could tell that the false lumen was still patent because the adventitia and the very outer media would bulge and it would be extremely friable and you would encounter oozing through the suture line, even in the face of reasonable hemostasis. This particular material binds the false lumen fairly thoroughly and it enhances the constituency of the strength of the tissue. That particular event just really doesn't occur any more.

DOCTOR BAVARIA: I'd like to make a comment on that. We had a case during the dissection THE study. Doctor Michael Acker, one of my partners who Doctor Laskey knows, excellent cardiac surgeon, was doing a mitral valve procedure on a very elderly woman and had a AV groove tear which is almost a universally fatal complication, intra-operative complication. He basically called me up -- he knew that I was using the glue for dissections -- and said, I need the glue, I need it right now. And I called David Fronk right at that point and said, Dave, I'm going to be using the glue. We have to use this glue here. She's dead if we don't.

So I gave it to Doctor Acker, and that patient survived an AV groove tear with the BioGlue placed in the groove right after the operation.

DOCTOR AZIZ: In that particular case, you clearly had to have a bloodless field.

DOCTOR BAVARIA: Yes. The cross-clamp was on. It was a bloodless field, and the AV groove was coated, so to speak. I think it kind of goes to the finding by David Fronk's suggestion that some of these anastomoses, sutureless anastomoses, can withstand 300 to 560 millimeters of mercury pressure.

DOCTOR AZIZ: So this could theoretically be used for VSD repairs intra-operatively.

DOCTOR BAVARIA: I'm not so sure the glue should be used intra-cardiac.

DOCTOR AZIZ: That's what I was going to come to. I think if it did get into the intra-vascular space, you see that would be a problem per chance?

DOCTOR COSELLI: I would be concerned about it. I think it just should be avoided, and I think it can be avoided.

DOCTOR FRONK: This is Dave Fronk. We clearly caution against its use intra-vascularly. We do not have any data to support its use at the present time, so we caution surgeons from using it intra-vascularly.

DOCTOR AZIZ: Just a few other questions. Do you see that this could be used in an infected field? For example, if you had a patient who had SBE and you were replacing his aortic valve? Is there any evidence or would you state that there might be a contraindication to using it in patients in an infected field?

DOCTOR BAVARIA: First of all, that's a very significant operation. I would think that it could be used in an infected field after you've gone in, you're resected all the endocarditis, you've prepared your graft for anastomosis and basically everything is kind of complete and the last phase of that procedure would be to utilize it at the proximal anastomosis especially. I can see where that would be utilized, could be utilized to get the patient off the table and have a successful operation, although I would also have to say that, as any foreign body, it may be an issue regarding infections similar to a graft.

DOCTOR LASKEY: May I just interject for a second. As I was getting uncomfortable with where this was going, I got a little note to remind us we really need to stay within the purview of the indicate of the PMA and the indicated uses. So thank you.

DOCTOR AZIZ: Well, I think these are all my questions.

DOCTOR LASKEY: Doctor Crittenden.

DOCTOR CRITTENDEN: I, too, like Doctor Aziz, want to congratulate the sponsors and the presenters today for an excellent presentation. I just have a couple of questions, but let me just go back to one point about the intravascular use or problems with intravascular use.

What happens if it inadvertently gets into the blood stream? Have you done any in vitro testing just to give a bolus of this intra-vascular to see what happens? I'm just curious that if you lost some inadvertently while you're doing an open aortic anastomosis, is that in and of itself a problem or is it just a bolus that may lodge in the end organ somewhere that's the issue?

DOCTOR FRONK: Obviously it will embolize if it releases. The data we have shows that blood, because it is a protein, will stick to the BioGlue if it's administered intravascularly. So you would in essence have a thrombus attached to the BioGlue that could embolize. That's probably the biggest reason for the contra ;use.

DOCTOR CRITTENDEN: In the warning section, there was a statement about exposure to nerves. Can you talk about that a little bit? What is the problem with getting the glue on nerves? Is it the glutaraldehyde that affects it?

DOCTOR FRONK: We believe that is the case. We also believe that it's very similar to any other product out there. Electrocautery. You wouldn't want to apply electrocautery to the nerve. It's more of a safety precaution? We do have some data that shows that it does have nerve dysfunction if it's applied to it. We have conflicting data with that. We feel it's the safest thing to advise any surgeon to steer clear of nerve tissue with BioGlue.

DOCTOR CRITTENDEN: Doctor Coselli, when you guys resuscitate patients after aneurism repair, dissections, etcetera, do you use mainly crystalloid or choloid? I just want to get to the point of using human albumen in these patients who may have been exposed to this. Is that an issue or is that a non-problem?

DOCTOR COSELLI: No, it's really not an issue. We use very little human albumen in our resuscitation efforts.

DOCTOR CRITTENDEN: And then Doctor Coselli, going to your presentation, let me just find it here. There's a couple of questions about that. Slide #18, secondary end point additional hemostatic measures. I didn't quite understand the failed anastomosis and the make-up stitches. It said there were 82 percent of these anastomoses that needed make-up stitches. These are the ones that failed. These are not all of them. Is that correct? Did I understand that correctly?

DOCTOR COSELLI: That's correct.

DOCTOR CRITTENDEN: Okay. But it was basically the same between BioGlue and the control group. And then the neurologic deficits. Do you have any more insight about that? Just counting them, there was no difference in paraplegia and no difference in stroke. So are you alluding to neuro-psychiatric issues? Is that what was better that made it significantly different?

DOCTOR COSELLI: Temporary stroke and neuro-psychiatric, although not statistically significant. I think there was one less paraplegia in the BioGlue group.

DOCTOR CRITTENDEN: The sponsor is not going to make any claim in that regard, I presume. Is that correct?

DOCTOR FRONK: No, we're not.

DOCTOR CRITTENDEN: That's all I have.

DOCTOR LASKEY: Doctor Ferguson.

DOCTOR FERGUSON: I'd like to echo the comments that were made previously. I think the presentation was one of the most lucid I've been privileged to hear.

I have questions which are based on the technical aspects. A couple of them have already been asked by my colleagues. One is that in other systems, particularly the European system, they recommend that when glue is applied, say, through a dissection that for a period of time the two leaves of the dissection be compressed together. You don't agree with that. You don't need that, I gather from your discussions and so on.

The question relates, of course, to the fact that if you have two areas of dissection, say, on the proximal part of the dissection which raises the flap near the aortic valve, how do you make certain that the glue that you put in is going to obliterate the space evenly and leave you a definable suturable wall?

DOCTOR COSELLI: Actually, it's fairly easy to manage. With regards to the approach of the French with regards to compressing the false lumen by applying some sort of pressure on the inside and the outside, with this particular material, the amount of pressure that's needed is really quite minimal. We have in some cases gently inserted a Foley catheter or other balloon in order to simulate the inter lumen of the aorta in order to have a smooth configuration. It's probably not necessary. The pressure applied on the outside requires no clamps or any special devices. It can be simply accomplished by placing some wet gauze, some most gauge, on the outside of the aortic wall. It is true that you can get a slight distortion of the inner lining but it's extremely minor and probably insignificant compared to the other distortions which we're creating by our reconstruction.

DOCTOR FERGUSON: My question, of course, goes to proper use of the device and whether the instructions for use should include something where you say that the two walls should be held together or not. It's been used enough, I understand, to obviate that and you certainly have not had any problems in that regard.

DOCTOR COSELLI: In our experience, it's been optional and probably not necessary in most cases.

DOCTOR FERGUSON: I had a question about the embolic. I think you've answered that for me. And one last question relates to a dry field. I come from an earlier era of cardiac and aortic surgery and dry fields were rare and a luxury, and so I'd like to know how dry you have to have the field. I read in there somewhere about a four or five seconds period of time where you require the field to be dry for the application to work.

Have you done any studies or what is the situation if the field is moist and how moist can it get and so forth?

DOCTOR BAVARIA: I think the main way to utilize this product is to place it prophylactically prior to release of the cross-clamps. That's the way this product works. It works very well if you utilize that system. Once you have the pressurized aorta, if that aorta is bleeding, I don't think or any other product actually is going to work very well.

DOCTOR FERGUSON: We're talking about applications other than aortic dissection here. That's what we're here today about are others other than dissecting aortas, as I understand.

DOCTOR COSELLI: Just from our experience in using this material, the drier the better, but it doesn't have to be absolutely dry.

DOCTOR FERGUSON: That's what I want to hear. That's all I have, Warren.

DOCTOR LASKEY: Thank you. I, too, would like to compliment you on both a crystal clear presentation and some stellar results. I mean this is typical of controlled clinical trials and I wonder whether this is representative of general use.

For example, with respect to the secondary end point failure to find differences there, I think your outcomes were so superb here. Is 1.4 percent re-op standard for the industry or is this just representative of a bunch of great surgeons who were doing this kind of work? What is the number in the literature? I know re-op is a broad category for all the kind of surgery here, but what would be a good comparitor?

DOCTOR COSELLI: I think the institutions selected for this trial are representative of, I think, superior clinical work. The numbers are probably lower than what's generally stated in the literature. My personal experience in over 1,700 thorac abdominal aortic aneurysms, our bring back rating for bleeding is under two percent. So this particular experience reflects our general work in our institution.

DOCTOR LASKEY: I wonder. It's impossible to do a really rigorous blinded clinical trial. You did a randomized trial, yes, and there's no way to blind the operator to the use of the glue because what comes out binds immediately. But it sounds like, as the study went on, you all became converts to this and you tended to want to believe in it, and I think that may have influenced the primary end point, too. It's not truly independent of the arm in which the patient was because you believed in this and it was clear the stuff worked, so you're more likely to stand back and give it time to work as you pressurized the site and allowed the BioGlue to do its thing. Is that true?

DOCTOR BAVARIA: I think that's true. The BioGlue works exceedingly fast, especially compared to the European quote/unquote "French glue." It binds with almost 100 percent tissue strength in about two minutes, so you don't have to wait that long, which is important for surgeons. So I think we did become very comfortable with it.

DOCTOR LASKEY: I think that's understandable. An analogy here is somebody replacing the hole in their sailboat with 5200 sealant that works immediately and you feel awfully good when that water starts coming in. So I can appreciate that.

Not that I want to portray these data in any terms other than positive, but let's look at the flip side. You have a 61 percent per patient success rate versus a 40 percent in the control arm, but you have a 40 percent non-success rate with the BioGlue hemostatic. Do you want to just comment on that. Again, this is a best case scenario in superb surgical centers in highly motivated hands.

DOCTOR COSELLI: I think I can try to address that. That, I think, is probably a manifestation of the way the trial was organized. A single stitch on a single anastomosis, even if there's eight or 10 anastomoses, would qualify as a patient failure. That's a very, very low tolerance.

DOCTOR BAVARIA: For example, that one case I presented had five anastomoses. So if even one of those anastomoses failed, that would be a patient failure.

DOCTOR LASKEY: With respect to the use of polyester versus PTFE, most of the grafts in this trial were dacron. Is that true?

DOCTOR COSELLI: Absolutely.

DOCTOR LASKEY: And how much PTFE use out there is there for this kind of surgery and what would you expect to see if there were more PTFE in this trial?

DOCTOR FRONK: Obviously in the surgeons that we chose, there was a preponderance of cardiothoracic surgeons, and that's a traditional dacron use. We do have the handful of cases where the peripheral vascular surgeons did use PTFE. We augmented that in the packet that we supplied to the FDA with some data that we had internationally, looking at its use sealing PTFE and, in all cases, it was successful in the surgeon's minds in terms of sealing PTFE material.

MR. CURD: David Curd. I'm a project director in clinical research at CryoLife. About eight or nine percent of the anastomoses in this trial represented PTFE.

DOCTOR LASKEY: Right. And I think that reflects the preponderance of this surgery, but really what you're going after is abdominal and lower extremity surgery here and this, I would think, for PTFE might be more. So I just wonder how the results might differ with a material that's going to be less compliant, if you will, to behave as you like.

My only other question with respect to the immunogenicity. I was looking for fever data, not that you could ever make sense of that in a post-op patient, but can you just quickly remind me of the febrile status of these folks and was there a difference?

DOCTOR BAVARIA: I don't think we have any data on that. My personal experience is is that there was no difference between the control group and the BioGlue group regarding post-operative fevers or long-term fevers. But I don't have any real data on that. It's just a personal experience.

MR. CURD: We did not specifically collect post-operative fever data.

DOCTOR LASKEY: Right. Because you have this intriguing entry, inflammatory immune systemic allergic reactions, and they're few in number, but what were you looking for there?

DOCTOR FRONK: Obviously, we were addressing any potential concerns there might be with the administration of BioGlue. There was only two patients that experienced an inflammatory and immune reaction. It's important to point out both of those were in BioGlue patients and neither of those were related to the glue. One was after intra-operative administration of an antibiotic. The second immune reaction took place after the intra-operative administration of Protamine to reverse the heparin. So none related to BioGlue.

DOCTOR LASKEY: Nevertheless, it would have been nice to have some fever data. Thank you again. Very good.

DOCTOR WITTES: I want to echo comments from everybody else. It's very nice to be able to read a randomized study. It's much easier to interpret things and I congratulate you for doing that. I actually like the heterogeneity of the population. I think that actually adds to the strength and the follow-up, of course, is excellent which is great. And the end point is understandable. We sometimes in this panel have end points that we don't understand. It's easy to yes and no.

I have actually four questions. Two are statistical comments and then there are two larger questions. One, a question on immunogenicity that is really a clinical question. I have no comment on it. The other is related to is trying to pull together what seems to me an inconsistency in your results in the primary and secondary end points, especially the transfusion-related end points. It actually has to do in part with what Warren brought up in terms of the unblinded nature of the study.

So with that as my preamble, the two statistical issues to me are, 1) I don't believe you can analyze by anastomosis the way you have done because they're correlated within a person so that those p-values at 001 I think is an over-statement of the statistical significance. I may have misinterpreted the analysis. My interpretation of the analysis, it was just binomial. Is that true and, if not, have you done an analysis that deals with the correlation, the within person correlation?

MR. CURD: Our initial thesis for doing the study was that each anastomosis was an independent observation, that each individual site had an equal chance of bleeding/not bleeding based on anatomy, tissue friability, that type of thing. That's where that came from.

DOCTOR WITTES: Did you do the analysis that allows the correlation within a site to be estimated from the data?

MR. CURD: No, we did not do that.

DOCTOR WITTES: I think you should and I think that should -- my feeling is that that's what should be reported because otherwise, as I say, you're likely over-stating the degree of significance.

The other issue has to do with the cross-over and I struggled with this. My initial feeling was why don't you just call the cross-over a failure? Why did you hurt yourself by removing it? Then I realized of course it's an unblinded study and you have to do that, so I think that was the correct thing to do. But my first thought, as I said, was that you had actually penalized yourself.

Let me ask you my imunogenicity question and then go to the big issue. As I understand it, people are going to have bovine protein in their bodies for a very long time. Why is it not a worry in the long term that you would develop some sort of an immune complex or some kind of allergic reaction and how would a person know? What would you know five years down the line? What would you be feeling and why should we not be worried about this?

DOCTOR FRONK: I'm going to have Bill Hall, an immunotoxicologist and one of our consultants, to address that question.

MR. HALL: My name is William Hall. I'm from Pathology Associates from Frederick, Maryland. I have no financial interest in CryoLife, but they are paying me for my trip down here today.

This is a very good question, and we need to look and see exactly what kind of evidence we have so far in the studies. #1, we have some acute toxicity studies that would potentially address anaphylaxis type reactions. #1 is the Buehler test and then the Kligman test, one of which uses Froin's complete adjuvant with it.

The third study that was done is the guinea pig immunization study, and the antibody and antibody response was elicited with the guinea pig immunization study. But if one looks at that study critically, the only way -- well, not the only way because we don't know whether or not injection of this material without Froin's complete adjuvant would have produced a response. But this antigen was mixed with Froin's complete adjuvant oil, mycobacterium species and immune response was elicited. The titers were low, however, but they were IGG titers. The corresponding anaphylaxis assays were negative in this assay according ISO standard criteria.

So basically we have evidence that there can be under extreme circumstances, especially with the adjuvant administration of this material, an immune response. To get an IGG, one needs T-cells also and so, therefore, memory T-cells can theoretically be produced. But if we look at the other part of this assay, the inflammation associated with administration of this material, we have evidence in human, sheep. I've not seen the goats but I reviewed also the rat study where subcutaneous implants were administered.

The inflammatory response around each of these was nil or extremely minimal compared to the surrounding reaction that one saw, for instance, on the human graft to the collagen that was administered on that human graft. This was a resected graft that I had an opportunity to look at, but compared to the BioGlue. And I have photo micrographs of this if one wishes to see the comparison. But compared to the BioGlue, the BioGlue had virtually no host response to the BioGlue and compared with the suture material that was present and compared with the collagen reinforcement that was put onto that graft segment.

In the rat, which was a study that was done for 90 days, this was subcutaneous inoculation. Basically with the rat study, the inflammatory response around the BioGlue was minimal and it consisted of very few lymphocytes that were present. T here was some macrophages present, but there were not the large macrophages, the activated macrophages, the ones that produced cytokines and kemokines, the ones that attract lymphocytes into the area. These were not present and, if one looked at the macrophages, you could see some material that had the same tinctorial characteristics as the BioGlue itself and the graft.

So the evidence that we have for immunogenicity is based primarily on the reaction against the BioGlue itself in a number of species and the reaction quite frankly is quite minimal and benign. It generally consisted of fibrosis. The worst and most severe reactions were seen in the rat, and these reactions were quite small.

DOCTOR WITTES: So you're saying that the long-term should not be any different from the short term. Is that right?

MR. HALL: We can not estimate what might happen in the long term from the data that we have except for the minimal reactivity that's present that would suggest an immune response against the material.

DOCTOR WITTES: Thank you.

Okay. Now let me ask my questions about primary and secondary end point because naively I would have assumed that if you could cut bleeding early during the surgery that you ought to be able to see some reflection in at least the intra-operative transfusions and you don't and, in fact, what there is is a very small, not statistically significant increase in red blood cells and cryoprecipitate and I think also in the fresh frozen plasma. No, not in the plasma.

But the data -- I mean I tried to get a handle on the data and I couldn't because the data that you present has only a mean and a standard deviation, a minimum and a maximum, and the minimum is zero. So I assume there's a lot of people with zero and that there's a long tail for the red blood cells. The BioGlue group, the range is zero to 22 with a mean of 2.3. The surgical repair is zero to 12 with a mean of 1.9. So we know we got a long tail. We probably have a spike at zero and you cite only one side of the confidence interval. So it's hard to know what is the sort of worse case. So let me put my worse hat on and then tell me why this is not true.

You have the patient there. You're now convinced, as Doctor Laskey says, you're convinced that stuff works. So even though there's a little bit of bleeding, you don't suture and besides which, you want to get more successes, so you actually are anti-conservative in the BioGlue group in terms of you're holding back so there will be more success. Not that you're counting, but there's a psychological tendency to do that.

So then in fact what happens is you've held back too much and so you need a little bit extra transfusion. You don't see it in these data because you only have 75 in a group and that's not nearly enough to see more transfusion. So I need to see, first of all, the distributions so that we're not just means and standard deviations, and then tell me why I don't have to worry about this.

MR. CURD: I think the appropriate place to begin to answer this question, I think there are some clinical aspects as well. Our statistical plan prior to the study beginning was a non-inferiority analysis which is why you see the statistics presented as you do. I have looked at the medians for these and the medians are almost all zero or one. I mean it's very far towards the left, as you have stated. So that's why the data is presented as it is.

Now, as far as the clinical --

DOCTOR WITTES: But wait a minute. I mean I agree with you. You don't medians in a case like this because medians are all clumped. You don't want means because they don't reflect it. You need to look at the highest quartile. It's those people that are getting too many extra transfusions. You have to analyze these data in a very different way from data that don't have clumps at zeros.

Yes, I understand that you pre-specify something but when you're talking about worrying a little bit. When something is going in the wrong direction, it seems to me you then say, well, yes, I pre-specified that I'm going to do such and such because I thought it was going to go in the right direction but, given it's going in the wrong direction, I've got to explore the data more.

DOCTOR FRONK: I appreciate your comments. Before I let Doctor Coselli discuss this, I think it is important to know that the administration of blood products is multi-factorial and, as Doctor Bavaria presented in one of his cases, he had bleeding through the graft and not at the anastomoses. So sealing the anastomoses is one point which we're interested in, but bleeding can take place for a variety of different reasons. The administration of blood can take place for a variety of different reasons. The anesthesiologist could do it. It could be used to prime the pump.

So again, I appreciate your comments. My concern is the fact that we didn't design this trial per se to adequately control blood utilization per se and to try to train or change institution's criteria for doing blood. We let them do it the way they felt that it was beneficial.

With that, I'd rather let Doctor Coselli speak about the clinical implications of that and then maybe we can find out if we're truly answered your question or not yet.

DOCTOR COSELLI: The trial was heavily weighted because of the institutions and the people involved towards complex thoracic arterial work. As a consequence, the issues of bleeding are not all anastomotic. There's a smorgasbord of other causes of bleeding. The previous case mentioned is just an example of one of them.

One of the things that clinically we saw during the trial and I brought it out in one of my slides is surgeons became more comfortable with trusting the fact that the glue was going to work and altered, in effect, their even control anastomoses by using less pledgets as the trial went along.

We work in an institution with -- in all of this, we didn't as part of the trial alter the way our anesthesiologists functioned. For instance, an institution may have multiple anesthesiologists that deal with these cases and so, even though there was a single surgeon, there was multiple anesthesiologists and they were just going to manage the blood product administration the way they felt necessary for the case at the particular time.

Frequently in these types of cases, a lot of the transfusion work is preemptive and they've been doing things a certain way anticipating events that we didn't try to alter as part of the trial. As a consequence, early in the trial, the anesthesiologist would preemptively manage expected bleeding, and appropriately so, and we didn't try to affect that as part of what we were doing, and I think that influenced us.

I think in the latter stages of the trial, just as the surgeons became more comfortable with the fact that they were going to be faced with less bleeding at the anastomoses, the anesthesiologists slowly came onboard but that wasn't specifically targeted as what we were trying to do.

DOCTOR WITTES: Thank you. That's very helpful. That's all I have.

DOCTOR LASKEY: Great. Just a point of clarification. As I read this, when you did your sample size calculations, you did it on a per anastomotic site. That was your original unit, and then you sort of backed into per patient later on. But your point is well taken. You need to do the intra-class correlation thing. But your unit of analysis was right from the get go precise.

DOCTOR WITTES: I'm not uncomfortable with using that as unit of analysis. I'm comfortable with that method of analysis. That's all.

DOCTOR FRONK: Excuse me. Doctor Laskey, can I interrupt just one second. You asked a question earlier about PTFE. I did want to throw up one back-up slide to address that. Here's some data that we collected from international usage and looking at a variety of different types of material that BioGlue sealed and at various types of procedures that the glue was used on.

If you take a look at the far right columns, it's the success criteria rated by the investigator. Yes or no, was it effective in sealing or reinforcing or adhering and whatever they chose to rate the material. As you can see, 35 instances of tissue to PTFE grafts were sealed using BioGlue in all cases. In the 31 cases that they addressed the answer, the answer was yes, it was effective.

DOCTOR LASKEY: Great. Thank you.

We are pretty much right on schedule. Congratulations to all. So at this point, we're ready for a break but before we break, Jim Dillard had an announcement to make.

MR. DILLARD: Thank you, Doctor Laskey. This is a little bit difficult and a little bit odd for me, but I wanted to make a real quick announcement because I think it's going to be important right before break. There's been some activity that's been reported by the news media that there are various activities throughout the United States that are happening, both at the World Trade Center, the Pentagon, the White House has been evacuated, and there's a fire on the Mall downtown. All airline traffic in the United States has been suspended today. I think we have some terrorist activity potentially that is going on and so I wanted to make everybody aware that that's occurring before they step outside.

Just by way of impact on this particular meeting today, currently I'd like to just consider that we move forward with the meeting, try to move towards completion. If anything changes, we will be monitoring it today. If anything changes and we need to make other arrangements, we will certainly do so. I think under the current situation, it's probably the appropriate thing with all the right individuals here to try to move forward today. I will also be in consultation with both the companies on the break so if anything changes, I will certainly let you know. But why don't we at this point try to break for 15 minutes, come back at about 10:20 and we will resume the activities. Thank you.

(Off the record at 10:02 a.m. for a 17 minute break.)

DOCTOR LASKEY: I really appreciate everybody hanging in here while we move along. First of all, are there any further questions from the members of the panel?

DOCTOR CRITTENDEN: Yes. I have two quick questions and then Doctor Bavaria will be probably the best person to answer them. We talked a little bit about the false lumen, Doctor Coselli did, in his initial presentation and in response to some questions. The question I have. We're not saying we don't need to surveil the false lumen any more. That still ought to be followed up over time despite the fact that this may lead to less flow in the long term in the false lumen. Is that not correct?

DOCTOR BAVARIA: Let me answer that a little bit historically. The Europeans have been placing glue in the false lumen for years. When you repair a Type A aortic dissection at the distal anastomosis, what we precisely do is we try to create a neomedia of about three to four millimeters thick and about a centimeter in length or depth into the aorta. So we don't do anything targeted in any way to the distal lumen.

That strategy may or may not decrease long-term false lumen patency. If it did, that would be a good result. That's what we want. But we don't know whether long-term distal false lumen patency is affected by these surgical strategies.

DOCTOR CRITTENDEN: So the follow-up with the false lumen ought to be the same?

DOCTOR BAVARIA: The follow-up for the false lumen ought to be exactly the same with BioGlue application as it would be for any previous dissection repair. Absolutely. In fact, false lumen follow-up is probably the most important anatomic follow-up in a dissection that there is and should be life-long.

DOCTOR CRITTENDEN: And then the next question is let's say we've resuspended an aortic valve during a Type A dissection and we do an echo post-operatively and, unfortunately, there's more leakage than we think the pace can tolerate and we need to replace the valve. What is it like going through a glued anastomosis if you have to go back through this again? Can you describe that.

DOCTOR BAVARIA: Well, that one case that was presented earlier is a case that I did, and we did exactly that or a very similar operation to that. It's not that difficult. It's a very thick aorta, easily entered, kind of similar to entering a re-do aorta after a graft has been placed with all the fibrous reaction, but it's not that difficult.

DOCTOR CRITTENDEN: How about a coronary distal? It seems that some of these were applied to the distal anastomosis for a coronary. What would that be like to take down similar, not much --

DOCTOR BAVARIA: Well, first of all, no glue was applied to a distal anastomosis of a coronary bypass graft. The only application of the BioGlue in coronary work was to the buttons of a composite graft proximate to right and left coronary buttons.

DOCTOR LASKEY: At this point, I'm going to ask Mr. Dacey --

MR. DACEY: In my three years of representing the consumer on this panel and some other panels, I always try to seek out what I call the consumer/patient comfort zone regarding new technical biological treatments, measures, and this is based on the fact that the consumer when they become patients places an enormous amount of trust and faith in their physicians and in the science and in the process including the process we're engaged in today.

This is one of the few times in reviewing all the material, and I went over it very thoroughly, that I really don't have any problems from the consumer/patient perspective because consumers and patients really have to trust you and, in this case, I do.

DOCTOR LASKEY: Mr. Morton.

MR. MORTON: No questions.

DOCTOR LASKEY: In that case, we'd like to move to the open public hearing and ask is there anyone in the audience -- do you want to do the questions first? Sorry. Before we proceed with the voting, etcetera, I would like to remind the panel of the questions that we've been asked.

Question #1 relating to effectiveness. The sponsor proved their primary hypothesis of a 10 percent improvement in hemostasis and that there was no need for additional agents during the procedure but did not show an improvement in secondary end points. We're asked to discuss the clinical implications of the primary and secondary end point data.

We really should tackle each of these individually so that at least it was my impression from the discussion this morning that certainly the clinical implications of the primary end point which is adequate hemostasis is fairly straightforward and cogent and the bottom line in terms of surgical procedure. Is that a fair consensus?

DOCTOR CRITTENDEN: Yes. I think so. I didn't hear all of Janet's remarks, but I think that's pretty fair, particularly for the primary end point.

DOCTOR LASKEY: The clinical implications of the secondary end point data are hard to address because #1, the results were stellar, #2, it's possible there was some investigator bias, particularly as the study went on and the agent proved its efficacy to stand back, but I think the results speak for themselves. We would have liked to have seen some further testimony to adequacy of hemostasis but those are the data.

Is anybody uncomfortable with the absence of differences in secondary end point data?

DOCTOR WITTES: My only concern, absence of differences in secondary end point doesn't surprise me because of the sample size and the variability. My only concern is if people are withholding other ways of preventing bleeding during the procedure and, therefore, potentially increasing the probability of transfusion. That's a concern, and it sounds like you don't see that at all.

DOCTOR COSELLI: No. The short answer.

DOCTOR LASKEY: I think it's hard to argue. Your results really are outstanding.

Question #2. Sponsor states in the submission that, quote, "Our clinical investigators believe that the routine use of BioGlue in these patients will allow them to modify their blood management protocol and should minimize the potential of life-threatening complications of post-op hemorrhage."

We need to comment on whether there was adequate information to support the statement.

DOCTOR CRITTENDEN: I don't think it's there. I think there's anecdotal evidence but for that statement, I'm not sure I saw data that would support that. That's a bit strong. I think it's efficacious but that particular thing, I'd have a hard time approving that.

DOCTOR WITTES: I agree. I don't see the data.

DOCTOR LASKEY: We've heard anecdotes which are very persuasive but it's certainly not in the panel pack. So I think that that language certainly is an overstatement.

Questions relating to safety and effectiveness, #3. Based on the information provided in the PMA, please discuss whether the information supports reasonable assurance of safety and effectiveness of the BioGlue.

Well, let's take that one before we go to safety and labeling. Are we all comfortable with the effectiveness of BioGlue as outlined in the application?

DOCTOR AZIZ: I think the information presented in the clinical experience I think does validate that.

DOCTOR LASKEY: And the safety of the product as well.

DOCTOR CRITTENDEN: I guess it depends on what you mean. It's like Bill Clinton said. It depends on what reasonable is. I think in the short term we know, but in the long term, we all have experience with glutaraldehyde preserved pericardium, etcetera, valves and, over time, what they do and we don't know what this is going to do over the long term, over years and years and years and years. But I suppose it's reasonable from what we have in the application anyway.

DOCTOR LASKEY: I think reasonable assurance is the appropriate language here.

Questions related to safety and labeling. One aspect of the PMA of a new product is the review of its labeling. The labeling must indicate which patients are appropriate for treatment. Identify potential adverse events with the use of the device and explain how the product should be used to maximize benefits and minimize adverse effects.

We're asked to address the following questions regarding product labeling. A) Please discuss the findings of the immunogenicity testing, especially as they relate to both physician and patient labeling issues.

Should patients be advised of specific adverse events to be aware of that may suggest they are experiencing a sensitization reaction from the BioGlue?

DOCTOR AZIZ: I think they should.

DOCTOR LASKEY: There are warnings and precautions in the IFU for patients with a known history of sensitivity to bovine products and BSA. How much further are we recommending the language go?

DOCTOR FERGUSON: My question is what would that language include because I don't know enough about sensitization, I guess, but what would be useful to put in the labeling that would indicate what they should watch for down the line?

MR. DILLARD: I might just maybe clarify the question a little bit. I think, as Doctor Laskey mentioned, currently in the labeling we're talking about known histories to bovine product, BSA, etcetera, and I think that what we're concerned about maybe goes a little bit along the lines of what Doctor Crittenden said. Just we're at that point in time where we have both safety and effectiveness information more in the short term and could this be a product since it is around for quite some number of years potentially?

I mean we certainly saw with some of the animal results that BioGlue was present, even out to a year. If there is a long-term or delayed sensitization reaction potential associated with the device, not that we've had a lot of experience with putting that into labeling, but I think the question was back to you as clinicians. Is there something that you think would be important to at least discuss and/or suggest to patients and the clinicians about what the long-term potential may be in an area where we don't have the long-term data to support that? I don't know that we were going any farther than that, I mean in terms of what our question was. And the company may have some comment on that, too.

DOCTOR LASKEY: My own feeling -- and I'm not an immunologist -- is that it has to be a very low frequency event, that from everything we've heard in the absence of IGE stimulation, this is unlikely to be an anaphylactic response but if it is, it's likely to be extremely infrequent and there's a possibility that there may be delayed hypersensitivity to exposure to bovine products subsequently and we have no way to quantitate that other than to again advise and caution the user in this sense in just that kind of language without putting numbers on it. What is the in-house feeling with respect to the immunogenicity?

DOCTOR COSELLI: From a clinical standpoint, we use to a great extent bovine pericardial patches intracardiac and for peripheral vascular patches routinely. We use bovine pericardial aortic valves as a permanent prosthesis. We implant bovine collagen as a hemostatic agent and a protanen is administered, so just from a clinical standpoint, there's a great deal of use of bovine material in our current practices, many of which are also associated with exposure to glutaraldehyde.

DOCTOR FRONK: To address it from the company perspective, I would echo not only Doctor Coselli's comments but also yours, Doctor Laskey. I think we believe that the frequency is very, very low. We haven't seen in any of the patients in the U.S. that have received the product and, as Doctor Vander Wyk mentioned, we estimate over 5,000 patients in the U.S. have been treated with BioGlue over the last almost two years.

From a long-term perspective, Doctor Crittenden, the aortic dissection trial, we have patients out past two years on receiving that. Maybe Doctor Coselli and Bavaria, who have the longer term experience with it, can comment on that. But to our knowledge, we have not seen anything ill towards the BioGlue or the patients with longer term exposure.

DOCTOR BAVARIA: And I'm presently following I think at least 40 patients treated for acute type aortic dissection long-term in my clinic, and we've had no long-term issues out to three years now with BioGlue for aortic dissections.

DOCTOR LASKEY: Well, I mean we're stumped. We're not experts and we don't have the data, but what data is available would suggest that it's an extremely low frequency event. I think heightened awareness is advised.

Doctor Bavaria, one question. If you can't give someone porcine-derived heparin, what is the alternative? I mean there's pork mucosal, heparin. Is there another commercial source for heparin? I'm ashamed to admit this since I use heparin every day.

DOCTOR COSELLI: I'm ashamed to admit the same thing. I don't know.

DOCTOR LASKEY: Does anybody in this room know of in patients with a pork or pork-related product allergy and you can't give them heparin.

DOCTOR AZIZ: Heparin can be bovine or pork.

DOCTOR LASKEY: That's what I'm thinking so perhaps you might advise not to administer heparin derived from a bovine source in those patients. Is that fair?

DOCTOR AZIZ: I think if they don't have an allergic reaction, I think, as Doctor Coselli mentioned, a lot of the patients do get bovine products.

DOCTOR LASKEY: It's just that the heparin is given systemically and the other stuff is not put into the blood stream.

DOCTOR AZIZ: I think it'll be difficult to control that. Let's say they go for cardio-catherization. They get a bit of heparin. How are you going to control?

DOCTOR COSELLI: I've never really given it a thought that if somebody has a porcine aortic valve in and I'm doing a re-operation for aorta or coronaries or something not to go ahead and give heparin systemically in the usual way and never really had any clinical problems with it.

DOCTOR LASKEY: Then we'll let the issue rest. Are you happy?

MR. DILLARD: Yes. Thank you.

DOCTOR LASKEY: b) the sponsor conducted several animal studies to assess the potential for BioGlue to elicit an immune reaction. The information from these studies suggests that there may be a potential for sensitization to BSA and related proteins in the formulation. Information from the clinical studies is limited to assessing the product with short-term follow-up, as we just discussed. Sensitization reactions may occur longer-term, as we discussed, theoretically they may. We have no handle on incidents.

Please discuss whether sensitization has been adequately addressed with the clinical data as supplied. I think we just did that. Is that correct? Are additional post-approval studies needed to assess the immune potential of BioGlue? My own read is that this is such a low frequency event that it's probably not feasible to either organize or conduct such a study.

MR. DILLARD: Thank you.

DOCTOR LASKEY: #5) Please comment on the indications for use section as to whether it identifies the appropriate patient population for treatment with this device. The indications from the current labeling read BioGlue Surgical Adhesive is indicated for use as an adjunct to standard methods of cardiac and vascular repair such as sutures or staples to provide hemostasis. I thought that was very clear. I didn't think there was any ambiguity.

#6) Please comment on the directions for use as to whether they adequately describe how the device should be used to maximize benefits and minimize adverse events. Again, with the exception of Doctor Ferguson's question about how to get a dry field, and I'm not a surgeon, they seem perfectly clear to me. The videos were even clearer. Is there anything that we feel needs to be added to the directions for use?

DOCTOR CRITTENDEN: It just seemed to me-- I've not used it so I don't have any experience with it -- but that it's important to do it with a aorta that's still cross-clamped or perhaps even a heart that's still cross-clamped. That's an issue. I'm looking through this trying to look for that in there and I don't remember reading it before. So if it's not there, some statement about that I think is appropriate.

DOCTOR AZIZ: I think the thing is if you're going to use circ arrest, you may not have to cross-clamp the aorta. So I think I wouldn't specifically insert that statement. I think the dry field should take care of that.

DOCTOR COSELLI: Just to speak to that, that is concomitant with an identical to a dry field.

DOCTOR CRITTENDEN: Circ arrest?

DOCTOR COSELLI: Circ arrest or cross-clamping. To achieve a dry field, you have to either have systemic circulatory arrest for an arch replacement or localized circulatory arrest if you're dealing with a peripheral vascular procedure.

DOCTOR BAVARIA: I've been using the word unpressurized.

DOCTOR CRITTENDEN: I like that better. how's that? Unpressurized.

DOCTOR LASKEY: So noted.

#7) Do we have any other recommendations regarding the labeling of this device?

DOCTOR WITTES: I would like changes in table 5 and table 6 consistent with what I said before.

MR. DILLARD: Did you say tables 5 and 6?

DOCTOR WITTES: Tables 5 and 6, 5 because of the correlation and 6 because I don't think that the means and the minimum and maximum tell the story of the number of transfusions. They've very simple changes.

DOCTOR LASKEY: Thank you. Well then, I would move to the open public hearing. Is there anyone in the audience who wishes to address the panel before the vote? In that case, I will close the public hearing. Thank you.

Do we have an executive secretary to read the voting options or shall I do that?

MR. DILLARD: No. I have them. These are the panel recommendation options for pre-market approval applications. The medical device amendments to the federal Food, Drug and Cosmetic Act as amended by the Safe Medical Devices Act of 1990 allows the FDA to obtain a recommendation from an expert advisory panel and designated medical device pre-market approval applications that are filed with the agency. The PMA must stand on its own merits and your recommendation must be supported by safety and effectiveness data in the application or by applicable publicly available information.

Safety is defined in the Act as reasonable assurance based on valid scientific evidence that the probable benefits to health under conditions on intended use outweigh any probable risks.

Effectiveness is defined as reasonable assurance that in a significant portion of the population the use of the device for its intended uses and conditions of use will provide clinically significant results.

Your recommendation options for the vote are as follows. Approval. If there are no conditions attached. Approvable with conditions. The panel may recommend that the PMA be found approvable subject to specified conditions such as physician or patient education, labeling changes, or a further analysis of existing data. Prior to voting, all of the conditions should be discussed by the panel.

Not approvable. Third option. The panel may recommend that the PMA is not approvable if, #1 the data do not provide a reasonable assurance that the device is safe or if a reasonable assurance has not been given that the device is effective under the conditions of use prescribed, recommended or suggested in the proposed labeling.

Following the voting, the chair will ask each panel member to present a brief statement outlining the reasons for their vote.

Doctor Laskey.

DOCTOR LASKEY: I'd like to ask for a motion now from the panel members, please.

DOCTOR AZIZ: I would recommend that the product be approved with conditions.

DOCTOR CRITTENDEN: Second.

DOCTOR LASKEY: Okay. And before we vote, may we hear the condition. Is it one condition or several?

DOCTOR AZIZ: I think we should encompass some of the things to be talked about, particularly the bovine allergy and things of that nature. Dry field.

DOCTOR LASKEY: So for the record as well as for me, what are we placing as a condition with respect to the bovine allergy issue?

DOCTOR AZIZ: I think if a patient has a history of bovine allergy, that should be a contraindication. I think a lot of them have been covered in the data that was submitted to us.

DOCTOR CRITTENDEN: A contraindication, you're saying?

DOCTOR AZIZ: No, no. There should be suspicion. There should be a caution in patients with a history of bovine allergy.

DOCTOR LASKEY: Wasn't that in the IFU?

DOCTOR FRONK: Yes, it is. It is clearly a contraindication in our labeling. Patients with known sensitivities to bovine products or albumen.

DOCTOR LASKEY: Do you then want to revise the placement of a condition on this motion?

DOCTOR CRITTENDEN: Can I ask one of the investigators. If you knew this in a patient who was having a problem with an anastomosis that you thought was best treated with BioGlue versus a pledget or a suture, would you withhold using this device? Would you not use this in someone --

DOCTOR BAVARIA: Who had a known allergy? Yes. I would not use this. I would use something else.

DOCTOR LASKEY: I am hearing the absence of conditions on Doctor Aziz's motion.

DOCTOR CRITTENDEN: I think there were some labeling issues that we talked about, so that's one condition. I guess I'm a little miffed now because if there's a strong contraindication but we're not putting anything about surveillance of this, I thought it was minuscule in importance but it seems maybe it's not.

DOCTOR LASKEY: So should we then ask for an element of post-marketing surveillance?

DOCTOR CRITTENDEN: I don't see where we couldn't. They're going to be followed any way. Just a simple yes/no that a clinician can say does a patient exhibit any untoward reaction and that's a yes/no. I think that would be simple if they're going to be followed anyway.

DOCTOR WITTES: I don't actually understand how this would work. Is that typical whenever there's a contraindication to ask for data collection?

MR. DILLARD: Generally, and this is a confusing point so I'll spend 30 seconds maybe to talk about it. We at the agency generally look at a contraindication as either a clinical situation where we have data where there are going to be adverse patient consequences associated with the use of the product and that may come from either clinical experience or sort of general usage where a labeling may be changed, and then the other would be if there is logically from a clinical standpoint good reasons without testing it in patients why it would not be a wise clinical situation to utilize the product. Then those generally get included also.

I think in this case because of our understanding of sensitivity associated with bovine products and the outcomes and the reactions that we have seen before with other products, that we don't need to test this in a bovine-sensitive patient in order to know that perhaps clinically there could be an adverse outcome associated with it. So that also then would be a reasonable situation to put as a contraindication. It would not be general principle for us under that guise to try to force the company to get data to reconfirm that it's not very good in certain circumstances to use the product.

DOCTOR LASKEY: Ergo, are there any other--

DOCTOR CRITTENDEN: No. I withdraw my plea for that surveillance and we'll just stick with that one condition about the labeling that we discussed earlier.

DOCTOR FERGUSON: Pardon my confusion. Isn't it in there?

MR. DILLARD: Let me just clarify something again real quick. In terms of conditions for your motion, you can place a condition on the manufacturer otherwise agreeing to the fact that the way their labeling is is appropriate. It's not absolutely necessary that that be a condition if you're comfortable with the way in which it's already worded. So I think either way could certainly be appropriate under this situation. So I don't think we need to get fixated on that.

DOCTOR CRITTENDEN: But specifically what I was talking about was the claim about the life-threatening hemorrhage and that was one thing that we didn't want. Maybe I'm being picayunish about it.

DOCTOR LASKEY: No. We agreed on that and I had forgotten about that so that actually is something now that we can attach to the motion. With that exception, anything else? Great. So we have a motion on the table.

MR. DILLARD: Can I ask for one clarification before you call for a vote on the motion just because I did hear Doctor Wittes' comment on a reanalysis and I was just curious whether or not you thought it was substantial enough to add as a condition or was that sort of a recommendation to the FDA to take a look at?

DOCTOR WITTES: Yes. That's why I was being quiet. I don't see it as a condition. I would be very disappointed if you didn't do it, but I don't see it as a condition.

MR. DILLARD: Thank you.

DOCTOR LASKEY: Do we have enough members to vote?

MR. DILLARD: Yes.

DOCTOR LASKEY: In that case, can we vote on the motion that's before us to approve with the single condition that the language referring to life threatening hemorrhage be eliminated. All in favor.

(Ayes)

DOCTOR LASKEY: Opposed.

(None)

DOCTOR LASKEY: Mike, just to wrap up, just to iterate why you voted as you did.

DOCTOR CRITTENDEN: I voted for BioGlue to be approved because I thought the product was demonstrated to be safe and efficacious and, short of this one condition that we talked about, I thought the presentation was good and it deserves to be used by everyone.

DOCTOR FERGUSON: I echo that.

DOCTOR AZIZ: I echo that. I think it clearly has been shown to be safe and effective and I think it'll impact in a positive way the way we treat some of these difficult patients.

DOCTOR LASKEY: Doctor Wittes.

DOCTOR WITTES: I found the data convincing, as well.

DOCTOR LASKEY: Mr. Morton? Mr. Dacey? No further comments?

I believe we are finished with the business at hand this morning, and I'd like to adjourn this meeting and I'd like to thank everyone for a heroic devotion to the cause of duty in view of this morning's tragic events. Thank you all.

MR. DILLARD: One quick announcement. Thank you, Doctor Laskey. If everybody could just stay put for a couple of minutes, I'm going to run out and see if I can't learn anything. I will be right back. Thank you.

DOCTOR LASKEY: Thank you, CryoLife.

(Off the record briefly at 10:55 a.m.)

MR. DILLARD: If I could have everyone's attention. The update to this point is that all federal facilities have been closed and we are currently discussing with some of our senior management what the impact that might have on this advisory panel meeting. So I think we have concluded the morning's activities and I think we may need, until we reconvene, to understand the impact that it might have on the afternoon's activity.

What I'd like to suggest is that we currently do as I suggested this morning which is unless we hear otherwise, I'd like to go ahead and proceed and try to handle the second PMA this afternoon. If you can come back in an hour at noon, we will try to reconvene and if I have any new information, I'll be able to announce it at that point. Thank you very much.

(Whereupon, the meeting was adjourned at 10:58 a.m.)