UNITED STATES OF AMERICA

P R O C E E D I N G S

(8:45 a.m.)

MS. SCUDIERO: Good morning, everyone.

We're ready to begin the 18th meeting of the Neurological Devices Penal. I'm Jan Scudiero, the Executive Secretary of this panel and a reviewer in the Division of General Neurological and restorative devices.

There are the usual housekeeping matters. If you haven't signed in at the door, please do so.

There is agenda information at the door, and also Advisory Penal Website information about how to get summary minutes and transcripts.

Before I turn the meeting over to Dr. Becker, I'm required to read into the record the deputization of temporary voting members statement and the conflict of interest statement that was prepared for this meeting.

This is the appointment to temporary voting status statement.

Pursuant to the authority granted under the Medical Devices Advisory Committee charter, dated October 27th, 1990, and amended on April 20th, 1995, I appoint the following as voting members of the Neurological Devices Panel for the duration of this meeting on November 30th, 2004:

Alexa I. Canady, M.D.

Michael R. Egnor, M.D.

Isabelle M. Germano, M.D.

David T. MacLaughlin, Ph.D.

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

Signed by Daniel G. Schultz, M.D., Director, Center for Devices and Radiological Health, on November 18th, 2004.

And this is the conflict of interest statement.

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 panel participants.

The conflict of interest statutes prohibit special government employees from participating in matters that could affect their or their employer's financial interests. However, the agency has determined 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, waivers were granted for Dr. Mary Jensen and David MacLaughlin for their interest in firms at issue that could potentially be affected by the panel's recommendations. The waivers for Drs. Jensen and MacLaughlin involve a grant to their institution for their sponsor's study.

These panelists had no knowledge of the funding and had no involvement in the data generation or analysis. The waivers allow these individuals to participate fully in today's deliberations.

Copies of these waivers may be obtained from the agency's Freedom of Information Office, Room 12A-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 himself or herself from such involvement, and 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.

I would like to mention that Ms. Crissy Wells, the consumer representative, is participating by telephone this morning, and I'd also like to announce that the scheduling information for the year 2005 will be made public in January in the Federal Register and on our Website.

Dr. Witten.

DR. WITTEN: Yes, thank you.

We have a couple of panel members who are going to be rotating off the panel after this meeting, and I'd like to take this opportunity to thank them.

And those are Dr. Becker, Ms. Wells, Mr. Balo, and Dr. Diaz, who unfortunately couldn't be here today.

FDA relies on its panel members to provide us with input and advice on our scientific matters for the devices that we regulate, and we appreciate the time and expertise that the panel members give us.

So I'd like to thank Dr. Becker, Ms. Wells, and Mr. Balo for their service here today and at the prior panel meetings during their tenure.

Thank you.

MS. SCUDIERO: Thank you.

I'd now like to turn over the meeting to our Chair, Dr. Kyra Becker.

CHAIRPERSON BECKER: Thank you.

Good morning. As Ms. Scudiero said, my name is Kyra Becker. I'm the Chairperson of the Neurological Devices Panel, and I'm a neurologist at the University of Washington in Seattle.

At this meeting, the panel will be making a recommendation to the Food and Drug Administration on the approvability of premarket approval application P040034 for the Confluent Surgical DuraSeal Sealant System, a bit of a tongue twister, intended for uses in adjunct sutured dural repair during cranial surgery to provide watertight closure.

Before we begin this meeting, I'd like to ask our distinguished panel members who are generously giving their time to help the FDA in the matter being discussed today and the other FDA staff seated around this table to introduce themselves. Please state your name, your area of expertise, your position and affiliation.

We'll start with Mr. Balo and go around the table.

MR. BALO: Andy Balo, industry representative. I'm Vice President of Regulatory and Clinical at DexCom in San Diego, California.

DR. LOFTUS: Hello. My name is Christopher Loftus. I'm a neurosurgeon. I'm Chief of Neurosurgery at Temple University in Philadelphia.

DR. EGNOR: My name is Michael Egnor. I am a pediatric neurosurgeon. I am Vice Chairman of Neurosurgery at the State University of New York at Stoneybrook.

DR. ELLENBERG: Good morning. My name is Jonas Ellenberg. I'm a biostatistician. I am currently employed at Westat in Rockville, Maryland. As of December 13th, I will be Professor of Biostatistics at the School of Medicine at the University of Pennsylvania.

DR. JENSEN: I'm Lee Jensen. I'm an interventional neuroradiologist. I'm Director of Interventional Neuroradiology at the University of Virginia in Charlottesville.

DR. CANADY: I'm Alexa Canady. I'm a pediatric neurosurgeon in Pensacola, Florida, and formerly Chief at the Children's Hospital in Michigan.

DR. HAINES: Steve Haines. I'm a neurosurgeon at the University of Minnesota.

DR. MacLAUGHLIN: Dave MacLaughlin. I'm Associate Director of Pediatric Surgical Research Labs at the Mass. General Hospital and a biochemist with a background in toxicology.

DR. JAYAM-TROUTH: Annapurrni Trouth. I'm a pediatric neurologist and the Chair of Neurology at Howard University Hospital, Washington, D.C.

DR. GERMANO: I'm Isabelle Germano, neurosurgeon. I'm Chief of the stereotactic functional and brain tumor problem at the Mt. Sinai School of Medicine, New York, New York.

DR. WITTEN: I'm Celia Witten, the Division Director of the reviewing division for these products at FDA.

CHAIRPERSON BECKER: And, Crissy, I don't know if you can hear us or not. Crissy Wells, are you there?

(No response.)

CHAIRPERSON BECKER: I guess she doesn't hear us by her telephone link-in.

So I guess at this point I'd like to note for the record that the voting members present constitute a quorum as required by 21 CFR Part 14.

Next Commander Stephen Rhodes, Chief, Plastic and Reconstructive Surgery Devices Branch, will update the panel on several matters that have occurred since the last meeting of the panel on June 15th, 2004.

Commander Rhodes.

CDR. RHODES: Thank you, Dr. Becker.

I am Commander Stephen Rhodes, the Chief of the Plastic and Reconstructive Surgery Devices Branch here at the FDA, one of the branches that regulates neurological devices in the Division of General Restorative and Neurological Devices.

Welcome, members of the panel, members of the public, and manufacturers, to this one-day meeting of the Neurological Devices Panel.

This panel --

CHAIRPERSON BECKER: Hi, Crissy. We're just starting the update. We'll get back with you.

DR. WELLS: I'm having difficulty hearing you.

CHAIRPERSON BECKER: Crissy, we're just starting the update. Before Stephen gets going, would you like to just introduce yourself as the consumer rep.?

MS. WELLS: Good morning. My name is Chris Wells, and I'm calling in from Phoenix, Arizona. I'm the consumer rep. for this panel.

CHAIRPERSON BECKER: Thanks, Crissy.

CDR. RHODES: This panel last met on June 15th of this year, at which time you made recommendations on the premarket approval application for the Cyberonics Vagus Nerve Stimulator Therapy System intended for the adjunctive long-term treatment of chronic or recurrent depression.

The panel also met on February 23rd of this year to make recommendations on Concentric Medical's premarket notification for the MERCI Retriever.

On August 11th, the agency cleared the MERCI Retriever for restoring blood flow in the neurovasculature by removing thrombus in patients experiencing ischemic stroke. The MERIC Retriever is also indicated for use in the retrieval of foreign bodies misplaced during interventional radiological procedures in the neuro, peripheral, and coronary vasculature.

More recently, on October 24th, the agency issued a guidance document entitled "Clinical Trial Considerations: Vertebral Augmentation Devices to Treat Spinal Insufficiency Fractures."

Additionally, one regulation action that this panel recommended in previous meetings is undergoing final review and clearance in the agency. The guidance document and the final rule reclassifying the neuro embolization device and the vascular embolization device from Class III to Class II should issue within the next few months.

And today you will make a recommendation on a premarket approval application from Confluent Surgical for the DuraSeal Dura Sealant System intended as an adjunct to sutured repair during cranial surgery to provide a watertight closure.

That concludes the update. We appreciate the commitment to public health of the panel members. We value the comments of the members of the public who have requested time to address the panel. And we appreciate the PMA sponsor's presentation to the panel this morning and responses to questions that the panel may have.

Thank you for your attention.

CHAIRPERSON BECKER: Thank you, Commander Rhodes.

So we will now proceed with the open public hearing portion of the meeting, and prior to the meeting there were no requests for the public to speak. I think I'd just like to ask if there's anybody in the audience now who would like to make an address to the panel.

(No response.)

CHAIRPERSON BECKER: No. Well, if that's the case, then we'll move on to the sponsor's presentation, and Confluent Surgical will be presenting their information for the DuraSeal Dura Sealant System intended for use as an adjunct to sutured dural repair during cranial surgery to provide watertight closure.

After this presentation we'll have a short break and then proceed with the FDA presentation before lunch. After lunch, the panel will deliberate on the approvability of the PMA.

Before the panel votes for the approvability of the PMA, there will be another open public hearing and a time for FDA and sponsor summations.

I'd like to remind the public observers at the meeting that while the meeting is open for public observation, public attendees may not participate except at the specific request of the panel.

We'll begin with the sponsor presentation. The first Confluent Surgical speaker is Mr. Eric P. Ankerud -- I hope I pronounced that correctly -- VP for Clinical, Regulatory and Quality. He'll introduce the other Confluent Surgical speakers as time goes on.

Mr. Ankerud.

MR. ANKERUD: Thank you, Dr. Becker, and good morning, distinguished members of the Advisory Panel, FDA, and guests of this meeting.

Confluent Surgical today will present to you the DuraSeal Dura Sealant System. Our presenters will include Patrick Campbell, Vice President of R&D at Confluent Surgical; Dr. Tew from the Mayfield Clinic; Dr. Cosgrove from Massachusetts General Hospital; and Dr. van Loveren from Tampa General Hospital.

In our presentation, we will provide to you an overview of our technology for the device, discuss the study design that was executed in the U.S. pivotal trial, and address safety questions that were provided to the panel.

Our company was founded in 1998 by Dr. Amar Sawhney, who has innovated this technology over the last decade. The mission of the company is to address unmet needs of surgical wound healing with in situ polymerized biomaterials.

Our country is based in a suburb of Boston, Massachusetts, and we are a small company.

The DuraSeal Dura Sealant System is commercialized in Europe in select markets, as well as registered in Australia.

The project that will be presented to you today began with a formal pre-IDE submission to FDA in March of 2002. Initially submitted as a study that was seeing input from FDA on the design for clinical trial to study the DuraSeal device.

During the discussions with FDA, FDA did seek input from an advisory panel member before this trial design was finalized in an IDE submission in February 2003.

Upon approval of that IDE with input from the advisory panel member, we executed a pivotal trial in the U.S. at ten clinical sites and one European site.

In May 2004, the study follow-up was completed. We treated 111 patients in this pivotal trial, following those patients out to three months. A modular PMA submission was initiated in January of this year, and in July of 2004, the final clinical results from the pivotal trial were submitted to FDA, and it is those results that we will discuss in this meeting today.

I would like to welcome to the podium Pat Campbell. Dr. Campbell is Vice President of Research and Development at Confluent Surgical.

DR. CAMPBELL: Thank you, Eric.

Before I discuss the DuraSeal hydrogel technology and preclinical tests, I'd like to review some of the properties for an ideal Dura Sealant.

First and foremost, the sealant needs to be biocompatible. DuraSeal consists over 90 percent of water. The solids that are there, that remain, are mostly polyethylene glycol, a molecule which is widely known and used in the pharmaceutical industry and recognized as nontoxic.

Polyethylene glycol is also synthetic, which means there's no potential for viral transmission in the product.

DuraSeal, when formed, contains water sensitive linkages that allow it to break down into small molecules and be fully absorbed in the body. When it's applied on tissue, it reacts very quickly., and this quick reaction allows it to adhere very well, and that good adherence with inherent cohesive strength of the material allow it to function well as a dural sealant to withstand elevated CSF pressures.

As I mentioned, the material goes onto the tissue as a liquid and polymerizes so that it's very easy to apply and spray onto tissue, and it contains a dilute blue dye that allows it to be visualized and to determine thickness and coverage of the product.

Prior to application, DuraSeal consists of two liquids. One liquid contains a small molecular weight amine, as shown in the top box. The other liquid contains an end modified polyethylene glycol molecule. That molecule contains a water sensitive linkage, as shown in the yellow, that allow it to break apart in time.

It also contains an end modification that allows it to react with the amine when it comes into contact with that. When those liquids are sprayed onto tissue, they rapidly polymerize, forming a hydrogel network that then has interspersed in that network water sensitive linkages that then allow it to break back down over a period of one to two months into small molecules which are absorbed and cleared from the body.

This is the DuraSeal kit as supplied. You see it contains two liquid syringes. The top syringe is injected into the powder vial that contains the polyethylene glycol and the blue dye. When that powder dissolves, the blue liquid is drawn back into the syringe, and the entire device is assembled as shown on the bottom frame.

This is a movie that shows the application of DuraSeal just on a hand. The surgeon advances the syringes, sprays the liquids. The rapid polymerization allows it to coat surfaces that are even tilted without significant runoff, and the blue coloration allows the surgeon to determine the extent of coverage and the thickness.

DuraSeal has undergone all of the biocompatibility tests as mandated in ISO 10993. The only test I'll mention here is the subchronic toxicity test where the material was evaluated in rats at a dose of 40 times the human dose, and there was no noted systemic toxicity.

DuraSeal has also been evaluated in canine craniectomy model where a durotomy was created, as you can see in the top left frame, that was two to three millimeters wide by two centimeters long. Animals were then randomized to either receive DuraSeal application, as in the right panel, or to remain as controls with no sealant application.

Animals then had the bone flaps removed and were recovered from surgery. They were then evaluated at four different postoperative time points, and at the time of evaluation animals were anesthetized. The bone flaps were removed, and the Valsalva maneuver was performed in an attempt to determine what pressure the Dura leaked.

You can see the left-hand picture on the lower panel. That's at day one. The blue dye rapidly diffuses out of the sealant after application and so you can see the gel is still present. It's clear, and you can see under the Valsalva maneuver the Dura is straining, but the sealant is withholding the pressure.

At day four there's a very similar picture. At 56 days the material was fully absorbed. The dura was completely healed, and I'll show you some histology in a moment.

This is the data obtained of the actual leak pressures measured in the DuraSeal and test animals at the different postoperative time points. You can see at day one all of the control animals leaked at five centimeters of water, which was the resting CSF pressure in that model.

So upon immediate bone flap removal there was a leak. Well, all of the DuraSeal animals withstood at least 50 centimeters of water. Two of those animals made it to 55 without leaking.

A similar difference in test and control persisted at four days. At seven days you can see a slight increase in the control ability to resist pressure, and interestingly, at seven and 56 days it was noted that in the test animals the bone flap was easy to remove. There were no adhesions between the dura and the bone flat, whereas the controls had significant adhesions or scar tissue, as you would expect in this model, between the bone flap and the dura.

And interestingly, at 56 days, there was still a difference in the test and control leak pressures.

The picture on the left is what I showed earlier with the gross image of the 56 time point with the dura smooth, healed in plane. The picture on the right is the histology in this model. You can see the bone flap, the thickened dura underneath. The dura appeared very similar in histology as it did to controls. There's no signs of neurotoxicity or local mass effect in this model.

DuraSeal was also evaluated for absorption using two different techniques. The top panel of pictures is a canine imaging study where DuraSeal was implanted and then imaged using MRI at T2 weighted. At three days you can see the light area on the top left right there. DuraSeal is very visible at two weeks. Four weeks it's getting a little bit thinner. Six weeks there's a trace, and by ten and eight weeks, it's rapidly absorbed.

A similar study was performed in the rat subcutaneous model where plugs of gel were implanted and that were harvested every two weeks and evaluated. At two weeks the physical properties of the gel were very similar to what they were at time zero. The material then rapidly degraded and was completely absorbed. The pockets were empty by eight weeks.

DuraSeal has also been evaluated in neurotoxicity studies where pieces of gel were implanted into rat brain parenchyma. The picture on the left is a histology slide from four days with no neurotoxicity, no reaction shown. The square in the middle is the void where the gel was. It doesn't withstand the histological processing well.

The six weeks had a very similar non-neurotoxic response with a decrease in volume associated with the absorption of the gel.

So in summary, DuraSeal has undergone a battery of tests. It has been shown to be nontoxic. It's not neurotoxic when in contact with brain tissue, and it has been shown to be safe at high doses in preclinical models.

It effectively seals the dura, allowing it to heal underneath, and then the material end life can be imaged using MRI imaging, and it is completely absorbed over eight weeks.

It is my pleasure now to introduce Dr. John Tew, a professor in the Department of Neurosurgery at the University of Cincinnati in Mayfield Clinic.

DR. TEW: Thank you, Pat.

Good morning, ladies and gentlemen, Madam Chair. My name is John Tew from the University of Cincinnati, Mayfield Clinic, and it's my pleasure to be here today as a neurosurgeon for 35 years and to disclose to you that I do own stock options in this company; that I am a member of the Scientific Advisory Board; and that I've been involved as an investigator in the process and am paid to be here today to explain to you the project rationale.

As a neurosurgeon for 35 years, I am well aware that watertight dural closure has been an illusive objective for neurosurgery for that time and much, much longer. Achieving a watertight dural closure is a basic objective of all who are in neurosurgical practice, particularly in some parts of neurosurgical closure, such as the post dura fossa and in spinal operations because controlling intraoperative leakage is very important to preventing CSF leakage and the development of serious postoperative complications.

The strata of these complication go from minute pinholes, either between suture lines or pinholes made in the performance of tacking up the dura for getting it out of the way, and these small holes may act as a one-way valve in which the fluid is allowed to get out of the dural compartment and collect as a pressurized system in the extra dural space.

In addition, extra dural and subcutaneous collections of CSF may develop into what are called pseudomeningoceles or enclosed meningoceles and other collections of CSF which may lead to acute and chronic problems of wound healage.

Overt leakage of CSF has perhaps even more potential serious postoperative complications and lead to not only compression of the neurological tissues, brain, spinal cord, but serious interference with wound healing, leading to dehiscence or breaking down of the wound which may be complicated by meningitis, infections, and a requirement for surgical intervention which leads to prolonged hospitalization and marked increase in medical cost.

So in my opinion as a surgeon for 35 years, there continues to be a major unmet need for product that creates a watertight dural closure.

Sealing sutured dural closure is a current method which is in search of an appropriate device. There are no FDA approved devices at the present time, but yet neurological surgeons use a variety of products off label. There's no standard of care for the use of these products. They fall basically into three types: hemostasis agents, such as surgicel or gelfoam, which are approved as hemostatic agents, but I suppose in our experience are used principally as space fillers to attempt to result in some type of sealant of the dura; adhesives, such as fibrin glues, cryoprecipitates, albumin gluteraldehydes, cryanoathacrylates, all of which have some potential toxic issues and have no approval for this particular objective; and finally, dural substitutes, such as DuraGen.

I'd like to show you one representative case which illustrates the intraoperative effectiveness of DuraSeal and the ease of application and the intraoperative effectiveness.

This is a 69 year old female who is undergoing a craniotomy for a tumor in the left frontal area, a so-called supratentorial craniotomy, in which the durotomy is seven centimeters in total length. Three, point, two milliliters of DuraSeal is applied.

And you see this is a movie which shows the craniotomy in this area. And there's an overt leak at the one o'clock area, and the DuraSeal is applied. The polymerization time is three to five seconds. You can see the rapid set-up, the polymerization, and then the testing with a Valsalva maneuver for up to 20 centimeters of water in the immediate polymerization.

The testing shows visibly that there's now no leakage at the site of what was previously an overt leakage in a hole two millimeters in size.

Thereafter the DuraSeal pilot study was performed in Europe at Nijmegen Medical Center by a single principal investigator, Dr. Andre Grotenhuis, who performed in a period of eight months craniotomies for operative procedures on the brain, 45, and two spinal procedures reflecting the principal cranial nature of his surgical practice.

The objective of this study was to evaluate safety and efficacy of DuraSeal as an adjuvant to standard surgical dural repair techniques in cranial and spinal procedures.

It was a single arm, non-randomized, single center trial performed by one surgeon. The intraoperative sealing endpoint was no CSF leakage during a Valsalva maneuver after application of DuraSeal.

The results were as follows. There was 100 percent intraoperative sealing success after a Valsalva is performed in the previous representative case. The results documented a 6.4 percent incidence of CSF leak, one incisional leak which was through the incision, and one was through the nose, reflecting an unsuspected or unidentified intraoperative leakage or potential opening into a nasal sinus, and one pseudomeningocele.

There was a 4.3 percent incidence of infection, one deep and one superficial. There were no device related adverse events, and the general impression of the investigator was that of excellent wound healing. Adverse events were consistent with the complexity of the operations that were included in the study. The results of this study were felt to be adequate to serve as a basis for a U.S. pivotal trial.

I would now like to introduce my colleague, Dr. Rees Cosgrove, who is Associate Professor of Surgery at Harvard Medical School and the Massachusetts General Hospital, who is the principal study investigator.

Thank you very much.

DR. COSGROVE: Thank you, John.

My name is Dr. Rees Cosgrove. I'm a neurosurgeon at Massachusetts General Hospital and the principal study investigator.

I have been compensated for my time and travel here today. I serve on the Scientific Advisory Board and, as Dr. Tew has mentioned, I am the principal study investigator.

As you've heard, the objective of this study was to see if the DuraSeal product would provide us with a watertight closure after primary dural sutured repair in craniotomies.

We used some of the information from the Nijmegen or the European trial in order to design an appropriate study to test this objective, and I have to say this was a very difficult study to design, and there are a variety of reasons.

We deliberated internally. We brought in experts, consultants to discuss expert groups of neurosurgeons to discuss an appropriate study design. We had communication with the FDA throughout this process. We had input from the FDA and input from one of the panel members here to try and develop an appropriate study design.

Part of the problem and some of the big problems is that in terms of achieving watertight dural closure is there is no standard of care. Neurosurgeons across this country use an absolute mishmash of materials. Some people prefer surgicel and gelfoam over the durotomy. Some people prefer dural substitutes over a primary dural closure. Other people prefer fibrin glue sprayed over the durotomy.

So we all agree that a watertight closure is an important objective of our standard wound closure, but there is absolutely no standard of care in this country.

And as Dr. Tew has pointed out, none of the devices we use are FDA approved for this application. So this presented us with a problem, and we deliberated on having a control arm of using fibrin glue which is one of the commonly utilized materials, but you know, this is an unapproved device, and in our communications and deliberations with the FDA, we were told that this was inappropriate to do a trial comparing it to an unapproved device, especially when the efficacy and safety profile of that device is not known in this application. So that was not appropriate.

And then the concept of having no treatment at all as the control arm was also neither medically or ethically acceptable, and because no neurosurgeon that I know of would not supplement their dural closure in some way in an attempt to achieve a watertight closure.

So with the FDA input, with the panel members' input, we arrived at a study that chose an intraoperative endpoint as its outcome.

The study was a prospective study at multiple centers and had a nonrandomized, single arm. We did use a prospective objective performance criteria for the primary endpoint and had 11 participating sites, ten in the United States and one in Europe.

The single European site was a Nijmegen, as has been previously mentioned, and then as you can see, there are ten other major academic medical centers in this country, and what these major academic medical center tend to attract is a very complex and complicated subject population, typically sicker patients.

So we actually gave ourselves quite a challenging study population. Key inclusion criteria included adults who were to undergo an elective craniotomy or craniectomy and classified as a clean procedure per the CDC guidelines, and there were a variety of exclusion criteria, including penetration into an air sinus of the mastoid air cells which would make it a clean contaminated procedure; prior surgery in the area; previous radiation or chemotherapy or even plant chemotherapy or radiation, preexisting hydrocephalus, and then a variety of serious medical exclusion criteria.

Intraoperatively, the eligibility criteria included a durotomy of at least two centimeters in length. The durotomy had to be at least three millimeters from the craniotomy margin. The gap could not be greater than two millimeters if after the neurosurgeon made his best efforts to close the dural opening. If there was a gap of greater than two millimeters, that these patients were excluded. We allowed only autologous duraplasty materials to be used, and importantly, the patients have to demonstrate either a spontaneous leak of CSF after the neurosurgeon had done his absolute best to get his closure, what he considered his optimal closure, or they have to leak spontaneously, or they have to leak upon a Valsalva maneuver.

So our primary efficacy endpoint was, indeed, interoperative sealing, and we termed it successful if there was no evidence of CSF leak after the dural repair, after up to two DuraSeal product applications, and tested during a Valsalva maneuver, taking the intracranial pressure up to 20 centimeters of water and holding it there for at least five to ten seconds.

And we used prospective objective performance criteria of 80 percent success rate at doing that.

And in order to justify and in order to demonstrate that for statistical purposes that our 95 percent confidence interval for intraoperative ceiling would be greater than 80 percent, we concluded that at least 70 patients needed to be enrolled and for safety purposes, we targeted a full 100 patients, assuming that about ten percent of these might drop out.

So for the entire study we planned on enrolling 110 patients.

The safety evaluations and the endpoints used were typical for a study of this nature, and included everything that is demonstrated up there. We defined a postoperative CS leak inclusively, and any obvious CSF leak that required some sort of surgical intervention, i.e., breaking of the skin, either suturing, over suturing of the incision, needle aspiration of a collection, placement of a lumbar drain or a ventricular drain or reoperation, that's clearly a significant CSF leak, and that was one definition.

Any time that fluid was collected outside the head that could be confirmed with tau-transferrin as being CSF, that was clearly a CSF leak.

And finally, at any time that the principal site investigator deemed either clinically or on his physical examination that there was suspicion for a CSF leak, that was also determined to be a CSF leak.

The protocol was designed to be with a detailed preoperative baseline testing and appropriate follow-up periods in seven days, six weeks, and three months to the conclusion of the study.

It is important to note that adverse events were collected at every time point in this study.

All centers were mandated to maintain a screening log, and a total of 303 patients were screened for entrance into the study. One hundred and five of those did not meet preoperative eligibility primarily because of coexistent medical illnesses, prior surgery in an area, long-term steroid use which we excluded.

Twelve patients actually were enrolled and signed consent, but then in terms of some of the metabolic work-up, their abnormalities of BUN and creatinine that excluded them from actually entering, and 54 patients refused participation primarily for social reasons. These are big academic medical centers were people are attracted from around the country and for different reasons. They wouldn't be willing to come back at different time points and complete the requirements of the study.

So a total of 132 patients were enrolled, and at surgery 111 of these were treated with DuraSeal. There were 21 intraoperative screen failures, i.e., an inadvertent entry into a sinus, recognition that the durotomy wasn't far enough away from the craniotomy margins or the fact that the neurosurgeon used nonautologous tissue for a duraplasty because he wasn't convinced that maybe the device would be effective.

But so a total of 111 patients were treated with a DuraSeal. All 111 were available at the first follow-up time point at seven days. One hundred nine were available at the six week visit, and at the final visit at three months, 107 patients were available for follow-up, giving us a 98 percent compliance rate.

One patient fell out after seven days because of a death at 30 days. An additional patient didn't make it into this group, although it was followed to completion because she didn't make it into the time constraints that we gave for the six-week visit, but she was followed up shortly thereafter, and made her three-month visit.

There was an additional death at 85 days postoperatively, and two patients, although they were followed to the six weeks, did not make their final three-month visit. So overall I think a rather exceptional compliance rate for a study of this sort.

Patient demographic we as expected for this study population. It's important to point out that over half of the patients had a smoking history, and in speaking to the complexity and severity of the cases that these kinds of medical centers attract, 86 percent had serious cardiovascular co-morbidities with an ASA score of two or greater.

The indications for surgery are, again, as you might expect at some of these major medical centers, with tumors, AVM, microvascular decompressions, Chiari, aneurysms, eplipsy.

Next.

But what's interesting to point out is that unlike what the ratio of procedures' surgical locations in the general population, we have nearly 50 percent of our cases were infratentorial, and it's the infratentorial group that all neurosurgeons worry about the most because of the high risk and propensity to CSF leaks and related complications.

In addition, the surgeries at these different centers tended to be long. The average duration of surgery was nearly four hours, and over 90 percent were longer than two hours, with a full 30 percent of greater than four hours and up to seven hours. So these were long and involved procedures.

As expected, there was a distribution between craniotomy, where the bone is replaced, and craniectomy, where the bone is removed and left out. And nearly 50 percent of the surgeons chose some sort of autologous duraplasty material to close the dura to their satisfaction.

At the surgical procedure, 60 percent of the patients after the surgeon had done his best to repair it in a watertight fashion, 60 percent of the patients' dural repairs leaked spontaneously, and the other 40 percent leaked after a Valsalva maneuver.

The DuraSeal was applied in a single application once, and only five percent of the time did it require two applications, and 95 percent of the time it was rated by the neurosurgeons as easy or very easy to use, and this is without any lead-in patients or any training prior to the first case.

None of the patients who were treated with the DuraSeal leaked after the application. Two patients, however, did not have their Valsalva maneuver elevated to an appropriate level to 20. It was only brought up to ten, and these we then considered were not evaluable by our protocol, and therefore, on an intent to treat analysis 98 percent of our patients were successfully sealed with a DuraSeal application.

In terms of adverse events, we created a very inclusive approach. Each and every untoward event was captured, and we did not cascade the events in individual patients. So every time there was an adverse event, even in the same patient, it was reported separately as an adverse event.

Importantly, there were no unanticipated adverse device effects. There were no device related adverse events. The majority of these events were not serious, but in keeping with the complexity of the cases, there were a significant percentage of patients who had a serious adverse event, but none of these were inconsistent with the type of surgeries performed or the complexity of the cases.

In this slide and the following slide, we've taken all of the adverse events, and I attempted to order them in descending order of seriousness, and as I said, these are all typical for the patient population that we were studying, but in terms of serious adverse events, I need to point out that most of the panel members are already aware there were eight surgical site infections.

And later on in this presentation Dr. van Loveren will be speaking to address this observation.

Next.

Each and every adverse event was reviewed by an independent clinical events committee, which consisted of three neurosurgeons who had no relationship with any of the participating sites. And it was their independent conclusion that the events that they reviewed were all consistent in type of severity considering the disease state and the procedures performed; and that no concerns were raised for patient safety because of use of the device, and none of the events were determined to be device related.

Pain assessments and modified Rankin scales were also acquired during the study, and these evolved and improved over time, as would be expected with this study population.

There were no metabolic abnormalities. All of the wounds uniformly were well healed at the three-month final follow-up, and there were no unexpected findings on CT scans.

Interestingly, at the three-month follow-up there was nearly a 75 percent reduction in the extra dural space, suggesting that DuraSeal was, indeed, absorbing as expected.

So, in summary, in terms of the primary endpoint and achieving success of a watertight dural closure, we did this in 98 percent of cases, well exceeding the 80 percent OPC mark, and there were no unanticipated adverse device effects and neither were there any device related adverse events.

So I'd now like to turn over the podium to Dr. van Loveren, who will speak to some of the safety review.

DR. VAN LOVEREN: Thank you, Dr. Cosgrove.

Dr. Becker, members of the panel, I appreciate your time today.

My name is Harry van Loveren. I'm the Chairman at the University of South Florida, Department of Neurosurgery. I was principal site investigator at the University of South Florida. I'm a member of the Advisory Board for Confluent, and therefore, my time and travel today are compensated.

I want to address the two key safety findings in this study, and that is the infections postoperatively and the postoperative CSF leaks, and then a focused comparison of those results to what is available in the current literature.

In terms of overview of our infection rate, there were eight patients in this study of 111 that had deep surgical site infections. One of those had a concurrent meningitis. Seven of those eight patients underwent removal of their bond flap to eradicate the infection. One patient was a craniectomy patient. So there was no bone flap to remove. All infections resolved.

There was one patient with a superficial surgical site infection which results with antibiotic treatment, and there was one interesting patient that was included with bacterial meningitis, also resolved. The reason I say that patient is interesting is because it's a patient that had a CSF leak, had a shunt put in, was asymptomatic regarding any sign of infection, had one broth culture come back positive for coag. negative staph. and, therefore, the surgeon as a precaution decided to use prophylactic antibiotics, which makes it an automatic inclusion.

So clinically we didn't think the patient was infected from an infectious disease standpoint, but the patient is included. That's ten patients total and a nine percent rate of infection.

We need to compare that to the literature comparisons. Finding suitable comparators in the literature for this type of information are extremely difficult. Comparing this prospective analysis, which was very rigorous is difficult when the literature is ripe with mostly retrospective reports, which are notorious for underestimating the capture of adverse events.

The definitions are difficult to reconcile in terms of what type of surgery was done, where was it done, what are the patient risk profiles and what is the definition of infection compared to our study.

The follow-up intervals in the literature tend to be short, or the intervals are unspecified, and there is a serious limitation in the literature available concerning patient follow-up and compliance. Some of the best compliance data in the literature is in the range of 75 percent return for follow-up, but there's a general assumption in the literature in many of these articles that if a patient is not heard from again by the treating center and is not referred back, that there has been no adverse event, and the denominator stays all patients enrolled rather than all patients returning for follow-up.

The bottom line is that the literature, therefore, is a very conservative estimate of adverse event rates, and therefore, potentially biases comparisons against our study, which I think gives us a rather robust comparison actually.

The literature was reviewed through common search engines, Medline, PUBMED, OVID, to find relevant articles. In relation to infection, we excluded articles published earlier than 1990 because that was before the era that antibiotic prophylaxis became common, and sine antibiotic prophylaxis was commonly used in our cases, those articles and those studies had to be excluded.

That left a number of retrospective and prospective studies. The prospective studies tended to be topic specific. They were focused on prophylactic antibiotics, preparation techniques for the surgical sites, or specific risk factor assessment.

Some of those articles dropped out because the definition of infection or the definition of surgical site or surgery type was not provided or, as we said, insufficient follow-up, and that left us with one very good article that was fairly comprehensive with well stratified patients and risk factors, and that's the Narotam article shown there with 2,249 well analyzed patients, and then a series of studies that were mostly retrospective looking at duraplasty materials.

If we look at some of the articles that were not considered good comparators, for instance, the Young article, although they start out with 800 patients, 400 of those patients are laminectomies, shunt insertions, stereotactic functional procedures with no durotomy or one to two millimeter durotomy, that really needed to be excluded.

And when you get down to the 200 to 250 craniotomies that could be compared to our series, the demographics are not known. The details provided in terms of assessment definition follow-up is quite poor.

If you look at the Bullock article, for instance, they start out well with about 400 cases. They have good follow-up. About 200- cases, again, are shuts and laminectomies, and then when you look at the 200 cases that could be applied, the risk profile is dramatically different from our study.

Our study was a very complex set of cases with an average time of surgery of multiple hours, high ASA scores, and this is a study where the average surgical time was about 100 minutes. So it's very difficult to generate a fair comparison. Again, if you look at all of the articles available on duraplasty materials, you can see that infection rates reported in the literature really are all over the map, high and low, and a lot of deficiencies found in the studies, those studies that only looked at deep wound infections and had no definition inclusion for superficial infections, those studies that had poor compliance in the range of 75 percent, but still maintained that enrollment denominator.

Next.

So if we look specifically at the Narotam article, which we used as the best available comparator in the literature, it's one of the largest prospective studies undertaken to evaluate operative sepsis in neurosurgery, 2,249 cases, and the infection rates were provided by surgery classification, and you see the five categories of classification: the clear case, clean-contaminated, clean with foreign body, contaminated and dirty, and the detailed definitions of infection were well provided.

The clean-contaminated case becomes important for our series because you've heard this several times, and you will hear it several more no doubt. We chose what we perceive to be some of the best neurosurgeons in the country, and in return for that, we got some of the worst cases in the country with the longest operative time and the sickest patients, and although that's an excellent challenge for a worst case scenario for this product, it also means when we compare it to the literature, we really have to account for that and stratify cases.

So if you look at the clean-contaminated category, a clean-contaminated cases which has a higher rate of infection is any surgery that lasts longer than two hours in duration, which was a significant number of our cases, I think more than 70 percent, and certainly they had a separate classification for any operation lasting longer than four hours in duration, and that was a little over 30 percent for our series. And, in fact, some of our cases went over ten hours, and that is a significant risk for infection.

They had a separate category for cases where they were an entry into the sinuses, transphenoidal, transoral procedures. We culled that out of their data because they were exclusions in our data as well.

Not any study is really perfect, and the small deficiency of this study was their limited follow-up to time of hospital discharge or four weeks, whichever came first, and compared to our very rigorous three-month follow-up.

What we did then is look at their infection rates for each of the specific categories, clean, clean with foreign body, clean-contaminated, and clean-contaminated greater than four hours.

The column in blue reports their infection rates for each category. The next column stratifies our patients according to the Narotam criterion, and you can see we have 54 percent of our cases lasting two to four hours, which is significant; 37, 38 percent of cases in this extreme clean-contaminated greater than four hour duration surgery.

And if you apply that mathematical statistic then, you generate for our patients, for our study group a predicted infection rate according to Narotam data for our patients of 8.3 percent.

To be fair, then, for use of Narotam data as a comparator, we have to alter our capture of infections as well because the Narotam study is very liberal about including patients in the infection group, and any concern about a wound in the Narotam data is included as a potential infection.

So we had to go back and we had one patient where there were express concerns about wound erythema that resolved, and in another patient where a surgeon cited concern about poor wound healing that also resolved, and to keep the data fair on both sides of the equation, we had to add those two patients to be consistent with Narotam.

So that really adds two patients, 12 of 111 with an observed infection rate according to Narotam data of 10.8 percent, which is not significantly different than the predictor of 8.3 percent.

We also compared to the best study in the literature for duraplasty material, and this is our comparison to that study in which DuraGen was tested, and you can see in the categories where the patients are able to be stratified, in clean surgery we had a comparable infection rate, and in the clean-contaminated group we had a comparable or favorable infection rate.

Now, one of the criticisms you could make of this comparison is that the DuraGen study intrinsically is looking at a subgroup of patients in whom the surgeon decided they couldn't primarily close the dura and had to use grafting materials. So to be fair about that comparison, we went on to restrict ourselves to a group of patients that are autologous, and we'll show you that slide later because we do do a heads up comparison.

If you look through the literature, risk factors for infection are well known and well described in the literature. Certainly prolonged surgery, greater than two hours, certainly greater than four hours are well known risk factors for infection.

American Society of Anesthesia scoring of greater than two is a risk factor. The presence of any foreign implant; we did not specifically capture data on foreign implants. Drains that are left in for greater than 24 hours. Titanium mesh, methacrylate cranioplasty adjunct to replacing the bone, we didn't capture that data. That group of patients does have an increased rate of infection.

The extent of the incision in many of our durotomies were quite long, up to, I think, 19 centimeters, and sinus penetration which we excluded and smoking is a significant literature risk factor for infection.

Next.

When we did univariate analysis and looked at our own studies, we found these factors that were significant in predicting infection: the volume of DuraSeal used, the duration of surgery, the length of durotomy, the use of an intraoperative shunt or drain, the smoking status. But in a multiple regression analysis, only those factors that have an asterisk remained significant predictors in our date, the duration of surgery and the patients' smoking status.

And if you look at this, these are our temptations with infection and the risk factors that were associated with each patient, and you can see that the risk factors are rather rampant for long surgery, elevated ASA scores.

So we're taking complex operations in sick patients.

Next.

In summary, the observed DuraSeal surgical site infection rate is what we would expect, given the patient population. We were addressing the risk profile of those patients and the complexity of the procedures performed and the infection rate compared favorably to duraplasty materials.

Next.

We also need to look then at the postoperative CSF leak rate as a safety finding and look at its comparison to the literature.

Post-op CSF leaks by definition in our study included any leak through the incision and any pseudomeningoceles that required any invasive intervention whatsoever, eve if it was the simple so-called tap and wrap that surgeons are familiar with where a needle is used to aspirate the fluid and that is placed in a bandage wrap. Any penetration of skin is considered an intervention and invasive and, therefore, is a significant pseudomeningoceles.

Five patients experienced then CSF leak for a rate of 4.5 percent. That's only two that actually have incisional leaks and three that were included as pseudomeningoceles.

One patient is interesting and was included at FDA insistence, even though the investigators had some doubts that it should, but so I'll tell you about that patient. That's a patient who had infection in the wound in the posterior fossa. The surgeon debrided the wound, and in debriding the wound, scraped off all of the DuraSeal, scraped the dura clean and noted that at the end of the debridement there may have been some evidence of CSF seepage through the previous suture line, and therefore as a precaution, the surgeon used a lumbar drain.

So certainly the minute you use a lumber drain to prevent CSF leakage, you become an included patient, although that's really quite iatrogenic.

Next.

If you break the patients down into two particular risk categories, the risk category well known and cited already in this discussion by those cases that are infratentorial where you're in the posterior fossa; you're at the dependent portion of the CSF volume, and most prone to leak, and then we had 58 in the supratentorial category with only one leak and 53 in the infratentorial category with only one leak.

And if you look at this, we substratified a bit to look at the high risk categories. Infratentorial craniectomy, of course, is a very high risk category because this is now you're at the dependent portion of the CSF volume at the base of the skull, and you don't have any bone to put back to buttress or support the suture line, and that was especially grueling in this study because when we use DuraSeal in these situations we were not allowed to buttress or support the DuraSeal with any other material.

So if you look at that high risk infratentorial craniectomy group, even though the numbers are low, 19 cases, we have only one leak, and that's the iatrogenic leak created by the surgeon clearing infection.

We have a small number of acoustic neuromas, six. Statistically that's not a relevant number, but again, we set it there because it's a group that has a higher risk in the literature for CSF leak and the rate in those patients happen to be zero.

Again, for CSF leak, we have to find comparators in the literature and we use the similar research engines, Medline, PUBMED, OVID, and then as mentioned, we excluded a number of articles that focused on cases that had unusually high CSF leak rates, which would be an unfairly favorable comparator, excluded series on acoustic neuroma, skull-base approaches, translaberatine approaches, series where there was no emphasis for dural closure, no mandate to try and close the dura and higher CSF leak rates.

We were still left then with a series of retrospective and prospective articles, and again, for similar reasons to the infection articles many dropped out because of poor follow-up, poor definition of what is a leak, no information on pseudomeningoceles, whether they were included, not included, and poor definition of the operations performed.

There is still a major prospective study with a similar patient population, which we'll discuss. That's the von Wild data, and some retrospective studies that looked at similar breakdowns of procedures in risk categories, supratentorial versus infratentorial.

Next.

This is the von Wild paper which was a suitable comparator, also a prospective, multi-center trial, a bit weak in compliance follow-up. Seventy-five percent of patients returned for follow-up at the six-month period. Every patient enrolled was placed in the denominator. We've mentioned this before. So that's a difficult comparator that grossly underestimates CSF leak, but the best comparator that we could find.

Next.

Again, also looking at it in terms of risk groups, 20 percent of their cases were infratentorial versus 48 percent of our cases. So we might expect actually a slightly higher leak rate for our more complicated cases.

Next.

But, in fact, we found a lower leak rate. Our leak rate, including pseudomeningoceles is 4.5 percent, and I should mention we are now excluding the iatrogenic leak.

If you include the iatrogenic leak, it does raise us to 5.4 percent and still well within 95 percent confidence interval compared to the von Wild study with 12.9 percent leak rate, and on leak rate, again, if the criticism would be that their patients all used DuraPatch, which was what was being tested, then we excluded cases from our group where no patch was used.

So now we're looking just at the 50 cases in our study where the surgeon decided dural closure could not be performed without an autologous patch, comparing to their use of DuraPatch. So they are a better matched set of cases, and then our leak rate is six percent versus theirs of 12.9 percent.

Next.

So we still have a relatively comparable or favorable outcome breaking it down that way.

If we look at supratentorial cases in the literature, again, it's quite a spectrum. There are a lot of retrospective studies under reported, under capturing of leak rates. No clarification as to whether they're going to include or not include pseudomeningoceles and what the criteria are, and still we come out with a very comparable leak rate to the other studies.

And if we look at those studies that isolate to the more complicated, more risky procedures of infratentorial procedures, again, our leak rate of 5.7 percent is very comparable, and in fact, if you look at the one study that's in our same ball park at 5.6, the Manley study, that study was a retrospective analysis of quality assurance data at a single center, which could easily underestimate or under capture events by 30, 40 percent. But in essence, we're very comparable.

Next.

This was a Gnanglingham study which looked at that very high risk group of infratentorial craniectomy where there's no bone to put back and, again, we have comparable rates for CSF leak. They did isolate out pseudomeningoceles. We have very comparable rates for pseudomeningoceles, and in fact, you might say favorable.

Next.

So in summary, for the CSF leak safety analysis, the observed DuraSeal CSF leak rate compares favorably to rates reported in the literature, given similar patient profiles.

Next.

The overall study summary and conclusions. We think we have had one of the most complicated series analyzed submitted in the literature, the worst cases and the best follow-up in a prospective manner. Fifty percent of our patients required grafting material. Forty-eight percent were infratentorial. Nineteen percent were craniectomies. Eighty-seven percent had elevated ASA scores, and significant morbidities.

Our procedures were remarkably prolonged, 92 percent greater than two hours, some as long as ten hours; long durotomies, as long as 19 centimeters, which is a lot for a craniotomy, and yet a very rigorous assessment with 96 percent of patients completing the total study.

The primary endpoint of intraoperative dural sealing was achieved 98 percent of the time. The wound infection rate is comparable to what would be expected in this risk profile group of patients in the literature.

The postoperative CSF leak rate compares favorably to what's comparable in the literature, and the adverse events seen in this study were consistent in nature, frequency, and severity for patients undergoing this complexity of cranial surgery.

And, in fact, the CDC independently found no evidence of a device related event.

In terms of our risk-benefit conclusion, of course, we begin with the assumption that dural closure/sealing promotes wound healing and avoids the cascade of complications that follow CSF leak and wound failure. There is no product approved by FDA for dural sealing as a support to suture closure and none demonstrated effective.

DuraSeal provides a standardized, effective, intraoperative, watertight dural closure without an increase in the risk of adverse events, and it's on that basis that we ask this panel to approve this product.

Okay. Safety and effectiveness of DuraSeal has been demonstrated through valid scientific evidence. The benefits associated with the use of DuraSeal outweigh the potential risks associated with the use of the device, and DuraSeal dural sealant is an effective adjunct to sutured dural repair during cranial surgery to provide watertight closure.

I think we said that.

CHAIRPERSON BECKER: Thank you, Confluent Medical.

At this point members of the panel are able to ask Confluent Medical questions, and I want to actually ask the first very naive question of the neurosurgeons.

How did you perform Valsalva on an anesthetized patient and how did you measure ICP during the Valsalva?

DR. TEW: The Valsalva maneuver is a standard maneuver for neurosurgeons to check watertight closure, and simply in an anesthetized patient the anesthetist bags the patient to a certain level of pressure which is then transmitted into the intercranial compartment.

You don't have a direct measurement of the intercranial pressure, you know, at surgery, but you can see the pressure indirectly through the bulging and leakage through the dura.

CHAIRPERSON BECKER: The pressures that are reported here, you know, your two failures of patients who didn't get ICPs high enough, that's based on intrapulmonary pressures or --

DR. TEW: Correct.

CHAIRPERSON BECKER: Okay. Other questions?

DR. CANADY: Yes. I had a question for Dr. van Loveren. The only control group we ever have here is the DuraGen control group. How was that control group constituted?

DR. VAN LOVEREN: Well, we have control groups for infection, comparators in the literature. When you look at the --

DR. CANADY: I understand, but I'm interested particularly in the DuraGen group. What was their --

DR. VAN LOVEREN: What was significant or special about that group?

DR. CANADY: Did they leave the leaks untreated or how was it constituted, that group? How was it defined?

DR. VAN LOVEREN: These were patients who had a durotomy that could not be suture closed and required a graft to be placed, and these were then compared in that study to patients in whom DuraGen was not used.

CHAIRPERSON BECKER: Dr. Germano.

DR. GERMANO: I have a question for Dr. van Loveren.

If you could please explain the discrepancy in reporting of the data on page 27. There is a report of CSF leak, six patients, 5.4 percent; pseudomeningoceles, two patients, 1.8 percent.

On page 39 of your presentation, you explain that the CSF leak is five because one was iatrogenic. What happened to the other two pseudomeningoceles patients? They're not reported here.

CHAIRPERSON BECKER: If I could just remind people to use the microphone when they ask questions.

DR. GERMANO: Sorry.

DR. VAN LOVEREN: On the adverse event, not every pseudomeningoceles met criteria for significance if there was no intervention. So the simple event of having a pseudomeningoceles if there is no treatment required, no penetration of skin required, and the pseudomeningoceles is observed and/or resolved, that is not included.

There has to be an intervention because if you look at other studies, pseudomeningoceles, if you look at radiographic studies looking for frequency of pseudomeningoceles, a small SCF collection after suture closure becomes really rather common.

CHAIRPERSON BECKER: Dr. Jayam-Trouth.

DR. JAYAM-TROUTH: I have a question. If normally you have 60 percent of surgical closures that come to neutral leak spontaneously, you know, all with the Valsalva manner as you've shown in a nicely done surgery, my question is then, I mean, is it really all that necessary, you know, when surgeons do a nicely done surgery that you need to have a sealant on top of it? Wouldn't it spontaneously heal?

And then you use antibiotics. Wouldn't you expect that without infection that the healing rate would be better?

DR. VAN LOVEREN: Well, I think it is these leaks at the time of closure that are resulting in all these complications you see in the literature and in practice, what starts as an interoperative leak and as a pseudomeningoceles, a wound breakdown, or an infection, and I don't think that surgeons are walking away from leaking wounds. They're reaching onto the shelf for a heterogeneous group of unapproved materials to buttress that wound. They are doing something about that wound almost each and every time, with a lot of heterogeneous, off-label, unproved, unstudied use.

And this is the first, I think, attempt to bring something standardized to that dural closure.

DR. JAYAM-TROUTH: No. My question is in 50 percent of these you put some, you know, heterogeneous material anyway, and then on top of it, you put the DuraSeal, and despite showing 100 percent closure with the DuraSeal, you still had pseudomeningoceles and you still had CSF leaks.

Now, how do you explain? Because if you say that the DuraSeal lasts for eight weeks and ten weeks in the system, you know, why is it that these patients did have the pseudomeningoceles?

DR. VAN LOVEREN: Well, I think when you refer to the heterogeneous material we put down and then put DuraSeal over it, the only thing we put down is an autologous, regionally harvested graft of patient tissue to close the gap because we're only allowed to accept a two millimeter gap. So there are no other materials being applied to that opening.

And the second part of the question is?

DR. JAYAM-TROUTH: When you had a 100 percent leak closure, how do you explain, and if the material lasts for ten weeks in the system, you k now, then how do you explain the pseudomeningoceles? You shouldn't have seen a single one.

DR. VAN LOVEREN: Well, you're asking why we didn't achieve perfection. I guess I'd have to acquiesce that the product is not perfect. If you look at the individuals', for instance, the leak rate, if you look at the true leak rate, it's incredibly low, two patients in the entire study, and in fact, one of them was found to have hydrocephalus, which was really an exclusion in this study, but the patient had hydrocephalus, recognized a month before surgery. The surgeon thought it had resolved. After surgery when the patient was leaking, the surgeon decided that the patient had active hydrocephalus and needed to be shunted to stop the CSF leak.

So I think there are explanations. If you have a complex case, you are not going to seal every case. You are going to have problems of wound healing unrelated to the DuraSeal. You're going to have problems of hydrocephalus after surgery that may break the seal, and still we have this incredibly low rate of CSF leak, and I think we were extremely rigorous in the inclusion of our pseudomeningoceles patients compared to literature where often, in fact, most of the time, a pseudomeningoceles is not considered a leak. So I think that was a very liberal definition.

DR. CANADY: You really don't think your pseudomeningoceles and CSF leak?

DR. VAN LOVEREN: I'm saying that --

DR. CANADY: I mean, I understand the study defined it certainly, but you really don't believe that, do you?

DR. VAN LOVEREN: Well, we defined pseudomeningoceles as a CSF leak.

DR. CANADY: Thank you. Okay.

CHAIRPERSON BECKER: Dr. MacLaughlin.

DR. MacLAUGHLIN: Yes. I have a question for Dr. Cosgrove.

You mentioned in your presentation that you excluded patients who were planning to have chemotherapy for their tumors or were on steroid use. Do you see that as standard restriction for the use of this product? Because a lot of your patients actually are cancer patients.

DR. COSGROVE: Well, in terms of the study design, we had to be very particular about what patients we were going to allow in and not, and many of our patients do have steroids. They're on steroids.

We actually used a criteria. They couldn't have chronic steroid use greater than four weeks prior to because that actually is implicated in delayed wound healing and infections and all of the systemic immunosuppression. And so for the purposes

of this study that's why we were so particular.

I can't say whether we would exclude those patients. I wouldn't think that's being an exclusion criteria for ongoing use of it until there was more data and information.

For the similar reasons, it was purely for a uniform study design, was the issues of chemotherapy and radiation, and in some instances you can't predict that because you'll go in and you'll say, "Well, I think this is going to be this kind of tumor," and then you come out and it's a malignant glioma and they're going to need to have radiation, you know, within the three-month time period, usually within about two to four weeks.

I know there was one patient who actually was excluded later on. So it could be -- I don't think it's going to be a major problem moving forward, but we'll have to study that.

DR. MacLAUGHLIN: Thank you.

CHAIRPERSON BECKER: Dr. Loftus.

DR. LOFTUS: Now, if I may be permitted three questions of three presenters, is that acceptable?

CHAIRPERSON BECKER: Sure.

DR. LOFTUS: First, for Dr. Cosgrove, if you wouldn't mind, just a question regarding your study design. You specify, if I understand you correctly, no gaps greater than two millimeters and no tears, durotomies, as it were, within three millimeters of the bone edge. You know, these are two of the most compelling reasons to use such a product, and aside from the pragmatic view that by eliminating these situations you are enabled to surmount your 80 percent criterion, can you just shed some light for me on why the study design eliminated what would be the most obvious need for a dural sealant?

DR. COSGROVE: Well, first and foremost, you know, the directive to all of the site investigators was to perform their best dural closure using autologous materials as needed. You and I both know that there are instances where to the best of your ability you're sewing things in and, you know, you're doing more damage than good by trying to patch certain things in or let's just put another stitch in here to see if I can get watertight, and it just pulls other things apart.

So, I mean, the goal of the neurosurgeons were to get as good a primary dural closure as possible. In some instances where you can't tell the neurosurgeon, you know, against his better judgment that you should do something different. He may make the determination that there's no way I'm going to get a primary closure and there may be a gap here, and we allowed that gap to be up to two millimeters primarily because we didn't want -- not greater than two millimeters -- primarily not because in some ways we weren't sure that it would seal properly, but when you spray it on, you didn't want it falling through the gap into the intradural compartment, and to get it to polymerize as a layer, two millimeters seemed to be the appropriate because of the viscosity of the product, that you could spray it on and it wouldn't drip. It polymerizes and sets up very nicely.

The second issue about close to the bone edge, the craniotomy margin, I agree with you. It was primarily though because in order for the compound to work as designed, it has to have a certain amount of dura that it can adhere to on both sides of the durotomy so that it can adhere properly and have the appropriate coverage.

So that was really a number that we sort of pulled out of the air, three millimeters from the craniotomy margin in order to be able to spray it on.

Now, that's not to say that when you spray it on it doesn't go right up to the craniotomy margin and, you know, up on the edges of the bone, but we thought that it was important to at least have, you know, flat dura to adhere to enough on both sides to not get a flat valve effect.

DR. LOFTUS: If I may just pursue, I mean, you know as well as I do that if approved, this is exactly what surgeons are going to want to use this for, and it's going to flop down on the surface of the brain. I mean, there's no way around that as I see it.

I just want to make certain that you feel that the product is, indeed, safe if it's directly applied to the surface of the brain.

DR. COSGROVE: Oh, yes. I mean it wasn't so much that there was any concern about the toxicology of it because I think there's -- and we'll talk to that afterwards if necessary about the detailed toxicology studies -- but this is essentially an inert substance, and it does not promote any reaction at all, and it was really to get it to form the seal, to work as designed rather than concerns about, you know, falling onto the brain and touching the brain.

You obviously though don't want to have a big lump of tissue, you know, a lump of foreign material, even though it's absorbable, sitting in the intracranial compartment after you've done an operation. I mean, that's like having a hematoma in there. So it just doesn't make good neurosurgical sense.

And you know, after we have done our surgeries, typically the brain isn't right up at the dural surface.

DR. VAN LOVEREN: If I could say something, wheN I looked at this preliminarily, I mean, if you look in pigs unfortunately it seals just fine to bone, and in fact, you'll herniate the pig before the seal breaks off of the bone. So it will be effective that way, and it will be used that way.

But if we were to allow that in the study, I think we'd have to stratify for it, and I think you'd have to stratify the patients to say sealing to bone rather than sealing to dura. You'd have to do a separate study. I don't think you can assume how something seals to dural material is how it seals to bone. You'd have to prove it.

DR. LOFTUS: May I proceed? The next two are very short.

Harry, if I could, Dr. van Loveren, if I could just ask you, so for the purposes of our comparison with the literature, I mean, it may seem pedestrian, but your definition of a deep surgical site or deep wound infection versus superficial.

DR. VAN LOVEREN: Pus deep to the galea that includes any form of involvement of the bone flap, bone osteitis, meningitis, anything that is on the deep surface of the galea. Superficial wound infections is really the incisional line only.

DR. LOFTUS: Okay, and my third question was for Mr. Ankerud, and that is you present no data in the trial regarding spinal use of the product, and I wonder if you propose that the product be also approved for use in repair of the spinal dura.

MR. ANKERUD: No, that is not included in our proposed indication for this device at this time.

DR. LOFTUS: Okay. Thank you.

CHAIRPERSON BECKER: Dr. Egnor.

DR. EGNOR: This is for Dr. van Loveren.

I share Dr. Loftus' concern regarding the exclusion criteria. For both the comparison for infection and the comparison for CSF leak to other published studies, the results are fairly impressive. My concern is that the patients though with the DuraSeal were the patients who had the very lowest risk in all of those groups because of the exclusion criteria.

When you have a dural closure in particular the supratentorial closures where there's a two millimeter or less residual opening, those are cases that are essentially closed. As your numbers noted, virtually everyone leaked, whether Valsalva or spontaneously anyway.

So the question would be: were the studies that you were comparing to for the infection and for the CSF leak -- did they have the same exclusion criteria for their patients as you did for yours?

DR. VAN LOVEREN: When they did not, we took out patients that were not included in our study either. I don't think we had a favorable group in terms of the DuraSeal patients.

One of our exclusion criteria was, for instance, entry into an air sinus, certainly any transphenoidal procedure, any procedure through a contaminated space. So to be fair, when we compared ourselves, for instance, to Narotam data, that's one of the reasons we could use their data, because they had those patients stratified, and we could exclude them because they contribute abnormally to their rate of infection.

So since they're not in our series, they're taken out of their series.

DR. EGNOR: Do you know that the CSF leak patients in the studies to which you were comparing DuraSeal had dural defects that were two millimeters or less? Because the large dural defects are really the at risk group.

DR. VAN LOVEREN: Well, I think actually we had a very difficult series because if you look at other series with CSF leak, a lot of them are including patients with shunts or stereotactic procedures where there's a pinhole made in the dura, very small procedures compared to 19 centimeter durotomies.

DR. COSGROVE: Dr. Egnor, could I also response?

DR. EGNOR: Sure.

DR. COSGROVE: You know, I think a two millimeter opening is not essentially closed, and as a pediatric surgeon in a posterior fossa procedure, if the resident said, "Oh, I've closed the dura and there's only a few two millimeter gaps," you'd go, "You did what? I mean, go back and do it again."

So I don't think in the exclusion criteria we selected easier cases by any means. I think that we just tried to do the standard of care, which most neurosurgeons try and get a watertight dura or complete dura closure.

And then, of course, we demonstrated that even though we tried to get them to do that, if they could, there was spontaneous leak in 60 percent, and then, you know, 40 percent of the time the neurosurgeon said, "Well, I did a good job there." Right? And looking pretty good, and then you do a Valsalva and it leaks, you know, typically along the suture line and the suture holes.

I mean, you k now, we say, "Well, that's as good as it gets," basically, and we would describe in our operative report we performed a watertight dural closure, right?

So I don't think we preselected. I mean, I understand that the tougher ones are where the tear goes out underneath the bone. I understand that, but that's just not something that then we can evaluate. We would have to stratify it in a different way and then have, you know, another cohort completely with tears that are not able to be primarily repaired, and so this was one way of trying to keep it a uniform study population.

But I don't think we, with our exclusion criteria, preselected any great cases.

DR. EGNOR: Well, you did. I mean, you excluded all kinds of things that were at very high risk for leak like a big, gaping hole in the dura that you could drive a truck through. I mean, those things were excluded, and those are the tough cases.

DR. COSGROVE: Well, no, but in fact, you know, as you and I both know, gaping holes tend to give, in fact, -- well, we don't know the data on that. There's no data to say that a big hole is worse than a little hole. In fact, in my experience, in fact, it's the smaller holes, the little flap valves where, you know, the patient does a Valsalva. It squirts out, opens up a little bit, and then the pressure of the fluid outside now closes the flap valve, and as they do various things that's how you get these expanding pseudomeningoceles.

If you have a big opening, I mean, the French never close their posterior fossas at all. They leave a big hole, you know, so fluid can go in and out, and then what happens is that there's no pressure or valve effect.

DR. EGNOR: Then why use DuraSeal at all?

(Laughter.)

DR. COSGROVE: Yes, well, that's a good point because the French -- I mean, the French never have a complication, right? Or at least that they can report.

But, no, there's lots of reasons to still use it because I don't abide by that at all because there are issues of wound healing. There are issues of meningismus, meningitis. There's issues of any infection with an open dura becomes now a deep intradural infection with meningitis and abscess formation. There are many, many, many, many good reasons to close the dura. I'm not abiding by the

French stance.

But I'm just saying that there is no data to my knowledge that characterizes any opening in the dura as being more dangerous or less dangerous, that a big hole is not necessarily more dangerous than a smaller hole, you know, or an intermediate hole. I don't think -- there's no data in the literature that has ever characterized that.

DR. EGNOR: Well, there may be no data because it seems obvious.

DR. COSGROVE: I don't think it's so obvious, but I think we get into more problems with the small pinhole and the valves than, you know, where you have a bigger opening sometimes.

CHAIRPERSON BECKER: Dr. Haines.

DR. HAINES: Actually it's for Dr. Cosgrove.

Dr. van Loveren's presentation is one of the most eloquent expositions of why concurrent controls would be helpful that I've ever heard, and could you explain in your deliberations about deciding not to have concurrent controls in the study why not use the surgeon's standard practice as the control?

DR. COSGROVE: Well, because you can say: what is that surgeon's standard practice? What is the next person's standard practice? What is the next person's standard practice?

DR. HAINES: But that's exactly the point. It would have produced a comparison for these patients treated by these surgeons, which it would have been much less burdensome for us in terms of understanding the comparison than having to try to deal with this literature problem.

DR. COSGROVE: Well, yeah. So I'll address that in a couple of ways. The first issue is that the standard of care of a specific surgeon is one thing. The standard of care of that specific surgeon may change from case to case. So he may use surgicel in gelfoam in one instance, which again I remind you are not FDA approved.

He may use DuraGen or some other dural replacement device, not approved for this.

He may use fibrin glue for specific things that he thinks, you know, but typically in my experience people don't use the same standard for each and every case.

DR. HAINES: But at least we'd have some idea of what that surgeon's chance of getting a leak or a deep wound infection with this group of patients was, and we really don't know that now.

DR. COSGROVE: Yeah. Well, the problem is, and this is why we have these communications and got input from the FDA. You know, it really was deemed by the FDA unacceptable to compare something to nonapproved FDA devices, where you don't have safety and efficacy profiled. You know, this is the problem. This was the problem in trying to get an appropriate study design.

CHAIRPERSON BECKER: Dr. Jensen.

DR. JENSEN: A question for Dr. van Loveren.

What imaging was done on the infected patients, leak patients at the time that it was recognized that they were infected or had a leak? Was it MR or CT? And what were these findings when compared to what you expected based upon the canine model?

Who read the studies? Did you have a neuroradiology group that looked at those studies of the infected patients? And were there any findings that you would not expect based upon the canine model, i.e., early reabsorption of the hydrogel, eccentric collections focused on the edge of the hydrogel, et cetera?

DR. VAN LOVEREN: Well, I appreciate your interest as a neuroradiologist. I think I would just word it a bit different, but we had a core lab that reviewed all of the radiographic studies, and there were studies taken at routine intervals, CAT scan, and we were primarily looking at its characteristics of dissolution on CAT scan and MRI. We had no mandate to specifically investigate radiographically if there was suspicion of an infection or of a pseudomeningoceles.

DR. JENSEN: But don't you think that would have strengthened your position if you had had good MRs done with a patient with a leak and it showed that the thickness of the material was what you would expect at that level or if you had an infection, that the site was not at perhaps the edge of the hydrogel?

DR. COSGROVE: Do you want me to speak to that?

DR. JENSEN: Whoever would like to, feel free.

DR. COSGROVE: So we had a lot of discussions about what was the appropriate technique to image, and the problem is that with the product being about 90 percent water, being able to differentiate it from CSF, to be able to differentiate it from blood breakdown products, air, I mean, we had Dr. Alex Norbash, who was in charge of the imaging corps, go through images from Europe and from the European study looking at both CT and MR images, and we had a lot of discussion about this point, and it was really felt that the optimal way for imaging this was with CT and looking at it over those time points that we described.

Now, in terms of the infections, is that ‑-

DR. JENSEN: Well, you know, based upon your dog data, you know, you talk about how the gel looks in contrast to CSF in terms of hyper intensity, and you showed an MR. So I think clearly MR in this situation would probably be better than CT in trying to look for edge enhancement of the hydrogel.

And they talk about a uniform enhancement of the edge of the hydrogel that dissipates over time. So if you're looking for inflammation, right, I mean, obviously what you're going to be looking for is enhancement, which is going to be more specific with MR.

And I would think that if you're worried about an infected collection and you do an MR and you find that you see just the same enhancement that you expect with the hydrogel and the collection is either remote or not positioned on the material or is positioned subgaleally, then chances are it's not your hydrogel. It's infection in another site.

I mean, it seems to me it would have provided you more substantial data in arguing that it was not the hydrogel, which is a site or source of infection, as opposed to other, you know, postoperative complications.

DR. COSGROVE: I understand many of your points, and not being a neuroradiologist and not being an MR specialist, maybe I'll get Pat to address some of those issues, and he's not a neuroradiologist either.

DR. CAMPBELL: No. Thanks.

Those are excellent observations. The study that I showed with the images of the MR, those studies were performed both CT and MRI imaging. MRI was performed using flare, T1, T2, with and without enhancement, and Dr. Norbash completed that study, evaluated every time point.

He did find that you could differentiate using the proper imaging the gel from CSF. You could also differentiate it from a potential infected bed, and that work is in press or in publication right now. We'll be publishing that in the next year or so. So that will be available to the general public.

DR. JENSEN: Okay. However, working at a busy neurosurgical site, I can guarantee you that your patients got MRs. I mean, I can't imagine that somebody who has an infection didn't get an MR. Do you have that data?

DR. COSGROVE: To be completely inconsistent, I mean, yes, you're right that a patients undergoing intracranial procedures typically do get MRs at times, but it's completely inconsistent. Sometimes it's early on. Sometimes it's at weeks afterwards. So --

DR. JENSEN: Right, but I mean in terms, again, of your infected patients and your leak patients. Okay? When you suspect the patient is infected, I mean, maybe it's just my institution, but that patient is going to get an MR before they go to the OR.

DR. COSGROVE: Yes.

DR. JENSEN: The same with a leak. They're going to get an MR. So I have to believe the data is there. The question is whether or not you chose to collect it and show it to your core group.

DR. COSGROVE: Well, anecdotal experience is there for sure. We did not collect it in a way that you could make any rigorous conclusions from it, but I can tell you, you know, the problem patient was my patient. The big patient, he was about 415 pounds, and we did an Arnold Kiari (phonetic) on him and said, "He's the one that got infected, and he's the one who when we explored him and debride, you know, to see what the depth of the infection was, went down and you take out all foreign material and you see what's going on.

So we scraped off all of the remaining DuraSeal, and after we did that, yeah, it sort of looked like the things were leaking. So I have an MR on that guy. This was about four weeks out from the surgery, three to four weeks out from the surgery when we took that image, and I'll tell you, well, first of all, it's difficult in an RL Carey (phonetic) malformation with all of the soft tissues and al of those things in the best of times, without DuraSeal in there, it's hard to interpret.

But it was difficult to interpret. It was a mixture of signals that, you know, the decision to reoperate was on a clinical basis, and I couldn't tell what was what because there was a combination of enhancement, fluid diffusion weighted abnormalities. It was impossible to tell at that point.

DR. CAMPBELL: Can I address one other issue sideways related to your comment?

There's a lot of information in the literature concerning polyethylene glycol and infection. Polyethylene glycol has been shown through many studies to be a poor food source for bacteria. Polyethylene glycol is synthetic, unlike other products that could be a food source.

Polyethylene glycol also has been evaluated in our preclinical studies extensively with no signs of infection. It's widely known and recognized as safe and nontoxic in the industry, and we have also completed a study in a similar product we were developing that uses polyethylene glycol where we implant a polyethylene glycol product, hydrogel, into the abdominal cavity of animals and intentionally created an infection at a rate that would cause healthy animals to die.

Prior to application we determine the LD-50 for an interperitoneal infection for these animals and then challenged it with the hydrogel versus non-hydrogen and found that the presence of the hydrogen did not potentiate infection, did not change the survival rate or the abscess formation in those animals.

DR. JENSEN: And so to that, in the case ‑- and I assume, Dr. Cosgrove, it was your case where the DuraSeal was scraped off. Did it come off as a sheet? Did it scrape off the middle peels? Were you able to send any of it to the lab to be evaluated?

DR. COSGROVE: Cultures were sent of necrotic and debrided material. It does not come off as a sheet. It's actually an amazing substance. It is adherent. It's pliable so that, you know, it stretches, and what you do if you want to take it off -- and in this instance it was four weeks out. So it had already undergone a fair amount of decomposition and was absorbing on its own. But you just took a cup curette and you'd have to scrape on the dura and lift the residual parts of it off, but it does not come off, you know, as a sheet.

CHAIRPERSON BECKER: Dr. Ellenberg.

DR. COSGROVE: And in answering your question about the infection, specimens were submitted, but in fact the patient had already been placed on antibiotics prior to the surgery. They tend to get placed on the antibiotics as soon as they hit the emergency room, and by the time you take them down to get the appropriate studies, no matter how many times you say we should hold off on the antibiotics before we get specimens.

This patient had been on antibiotics.

DR. JENSEN: Any microscopic evaluation of the hydrogel or staining or anything to just see if there had been any pockets of bacteria or anything?

DR. COSGROVE: No, I don't recall the path. report indicating anything of that sort.

DR. JENSEN: Thank you.

CHAIRPERSON BECKER: Dr. Ellenberg.

DR. ELLENBERG: Are we okay on timing?

CHAIRPERSON BECKER: Yeah.

DR. ELLENBERG: Okay. If I may, I'd like to ask three questions of Dr. Cosgrove in the area of efficacy and three questions on safety. I'd be happy to stop the questioning and allow another panel member to break in if the chair so determines that's a good idea.

My first question has to do with the issue that Dr. Haines had already raised, the lack of a control group in the study, and from my reading of the material in the FDA clinical review, the major argument considered there on page 25 was the issue that one would have a heterogeneous control population if you had a control group and you and the rest of the sponsor group have referred to that on several instances.

In addition, this morning you've raised the issue which I did not read in the panel book that FDA either ruled or has thought it inappropriate to use a standard of care as a control because the standard of care might include unapproved use of products on the market.

I want to pursue the issue of the use of control group, but at this point I think it's reasonable for the panel to understand the constraints fully that you had are not using a control group because what you are offering up, as Dr. Haines has pointed out, is a very selective review of the literature to find in the massive numbers of papers that you've looked at that paper that you can find a subgroup in. So it's a subgroup of papers. Then within a paper it's a subgroup there that most closely matches the group that you're presenting today.

So with that in mind, I would like to ask FDA if there are, in fact, constraints -- and, Dr. Witten, I would ask you to respond to this -- are there constraints on having a control group where standard of care might include the use of an unapproved device?

DR. WITTEN: You can certainly have a control where the standard of care includes, you know, the use of various unapproved devices, but then the question is for us how we would end up interpreting that. So it's not that you can't use it. It could be put into the study design, but then the question would be how we would interpret.

Just for example, would the sponsor then need to show superiority to this heterogeneous standard of care equivalence to it? What would that mean?

So just I think that --

DR. ELLENBERG: That answers my question, but let me rephrase that for my benefit and, thinking aloud, the panel's benefit.

My sense of that response is that the ruling did not have to do with the issue of standard of care. It went back to the issue of having a heterogeneous control group. So let me follow on with that.

You in defining the entrance drug criteria or --

DR. COSGROVE: Can I respond to the question before I lose -- I'm trying to keep track of all the questions.

DR. ELLENBERG: I haven't asked the question yet.

(Laughter.)

DR. COSGROVE: Well, I know, but I think it's very, very important to point out that we as the investigators were very perplexed, very cognizant of these issues. I mean of the design study, of a single arm study. We did not propose this initially. We proposed a control arm, and we figured that the best control arm, albeit less than perfect for a number of reasons which I'll go into, but we figured that the best control, the most suitable control would be a fibrin glue of some sort because similar in administration, really similar in terms of some of its properties, although it is a biologic device with all of the attendant risks that can occur from a biologic device.

In terms of indications for use, it would be very similar. In terms of actually adherence, we couldn't test it properly with a Valsalva because often when you do a Valsalva with fibrin glue, it just lifts off, and then you say, "Well, now I'm going to have to scrape it all off and put on a new one," and so you couldn't test it appropriately.

But we were willing to deal with some of those issues, and then we went to the FDA. We got their input, and were advised that using a control group, using a non-FDA approved device, we were not going to be allowed to do that.

So that was the binder. That was the handcuffs that we were placed into, and then we chose the next best alternative, in our opinion.

DR. ELLENBERG: I'm afraid I don't see the handcuffs that you had, but let's pass on that. I think you've given your response to the question.

My second concern has to do in our interpreting the safety data and the efficacy data with a question as to your definition of endpoint being the watertight seal. Could you talk a bit about why the endpoint was not infection, for example?

I'm not asking if that would have been the specific endpoint, but why you did not choose an adverse event which you have listed very clearly in the past several minutes can take many forms; why that was not the endpoint rather than a watertight seal at some point close after surgery and then thereon, which seems to me in reading through the materials is more a surrogate endpoint than what you're really after, which is no complications.

Why did you choose the watertight seal at the endpoint?

DR. COSGROVE: Well, no neurosurgery is done without complications. So you know, there are so many adverse events. If you're going to choose an adverse event as your endpoint, it's very difficult to make the connection that it was anything to do with your study.

So we chose the intraoperative endpoint of a watertight seal as something that could be easily defined. It's a binary observation, and as the essential aspect in wound healing if you are not getting a watertight seal at the time of surgery when you're actually closing the dura, it is the necessary achievement or objective in order to down the road reduce the complications associated with a non-watertight dural closure.

DR. VAN LOVEREN: If I might, I would also say that the application for this device is as a sealant to prevent CSF leak, not as a protection from infection. Although infection stands as a potential adverse outcome similar to other outcomes and highlighted itself, I don't think it's a legitimate endpoint. That's not what the application is for.

DR. ELLENBERG: No, I understand that.

DR. VAN LOVEREN: Okay.

DR. ELLENBERG: But if the outcome was different, the application would have been for something different, but I understand your point.

In the area of efficacy, again, on page 18 of your joint presentation -- I believe it was the top slide -- there's a list of the post surgical eligibility, such as the size of the hole left and then there's a whole other list. So that patients who are essentially excluded from the study post surgery if they did not meet these conditions.

DR. COSGROVE: Interoperably.

DR. ELLENBERG: Excuse me, yes. I misspoke.

In terms of those patients, did you make any attempt to see how those patients did? Did you catalogue the aspects of the reasons they -- which of your criteria they missed because my sense is that that could be quite informative in terms of both efficacy and safety? Is that data available?

DR. COSGROVE: Yes, it is. It is available, and I can get you a complete analysis of those cases. There were 23 cases that were excluded on the basis of this interoperative criteria. Those are the patients you are talking about, and we have that data. We even have follow-up data because they were actually enrolled in the trial and we can get you that data in a little while.

DR. ELLENBERG: There are enrolled in the trial in the sense they followed the protocol?

DR. COSGROVE: Well, they were continued to follow throughout the trial. I mean, we do know what they -- no, I'm sorry. They weren't enrolled in the trial, but they were followed, and we have some of the data.

PARTICIPANT: I'm sorry, but we documented their --

MR. ANKERUD: Go to the microphone. State your name.

DR. COSGROVE: We know the reasons why they were excluded. Oh, okay. I'm sorry. But they were not followed to outcome. I thought we had that.

DR. ELLENBERG: Fine, okay. I think, Dr. van Loveren, if you can stay up, on the issues of safety in terms of how this was presented to various IRBs at the cooperating clinical centers, do the protocols specify that the safety review, not the efficacy comparison, but the safety review for purposes of informing various TSMB components or the IRBs themselves; did the safety review in the protocol indicate to the IRBs that the safety evaluation would be a literature based review? Comparison, excuse me. A literature based comparison.

DR. VAN LOVEREN: I'm not sure if we communicated that specifically to each IRB about how the --

DR. ELLENBERG: Well, did the protocol have that as an analytic approach to evaluation of safety? Because that would have been submitted to the IRBs.

DR. VAN LOVEREN: Right. I don't believe so.

DR. ELLENBERG: Okay. In terms of your follow-up post three months, are these patients still being followed?

DR. VAN LOVEREN: Well, they're being followed clinically, but not for purposes of this study.

DR. ELLENBERG: So you don't have control of their follow-up at this point?

DR. VAN LOVEREN: No.

DR. ELLENBERG: If the panel advised FDA that it would be useful for a long-term follow-up, would you have the capability of reinitiating the follow-up or are there informed consent issues? Are there other things that might impede the re-contacting of these patients?

DR. VAN LOVEREN: No, I don't think that would be any impediment to that whatsoever.

DR. ELLENBERG: Okay. Thank you.

CHAIRPERSON BECKER: I think that everybody on the panel has had a chance to ask at least one question. I want to see if Crissy Wells is still there, if she has a question.

(No response.)

CHAIRPERSON BECKER: Mr. Balo, any questions?

MR. BALO: No questions.

CHAIRPERSON BECKER: So --

DR. VAN LOVEREN: Could I belabor one point? This is a very dangerous move on my part, to go back to a question that apparently was answered and bring it back up, but it's on the infection as an endpoint.

I mean, I think infection is so determined by risk profile. It's so sensitive to risk profile. To set an OPC ahead of time you don't really have the ability to do that without knowing what your patient risk profile is.

If your ASA scores are all high, you should pick a number, an infection rate of ten percent. If your operation times are all going to be less than 60 minutes, you should pick a number that's in the two percent range.

DR. ELLENBERG: Certainly, but if you were in a controlled clinical trial situation, then that would be doable.

DR. VAN LOVEREN: Yes.

CHAIRPERSON BECKER: We'll have a chance for one or two more questions, and there's going to be an opportunity in the afternoon for even more questions.

Dr. Germano.

DR. GERMANO: A question on safety. I don't see any data on seizures. Should the panel assume that the 111 patients did not have perioperative seizures?

Obviously when the compound touches the brain, there is a concern that seizures can be induced, whereas seizure studies done in the rats and dogs?

DR. CAMPBELL: Yes. The preclinical studies evaluated implantation into the rat. There was also hydrogel extracts that were injected into the cisterna magna and lateral ventricle. There was preclinical studies in the canine model I showed you.

DR. GERMANO: How did you monitor the seizures in those animals?

DR. CAMPBELL: They were clinically evaluated immediately after application and regularly daily by veterinarians. There were no signs of seizures or clinical abnormalities versus the control animals which were saline alone.

DR. GERMANO: Did you do any EEG studies?

DR. CAMPBELL: No.

DR. GERMANO: For the clinical component?

DR. COSGROVE: Seizures were reported in the adverse event summary sheet. I'm just looking through the adverse events.

DR. GERMANO: It's not there.

DR. COSGROVE: It's not there. There were three seizures reported in the final report.

CHAIRPERSON BECKER: Dr. Jayam-Trouth.

DR. JAYAM-TROUTH: A couple more points. One is the cognitive problems that you say that you had in page 27 of your patients and the speech difficulties in ten your patients, five of your patients with cognitive problems, 34 of your patients with premium nerve deficits. I mean these were all relevant to the DuraSeal itself?

DR. COSGROVE: That's correct. I mean, this speaks to the patient population and the procedures performed on them, and none of these were unexpected, and upon review by the CDC, none of these were deemed relevant to the DuraSeal application.

These, you're talking about aneurysmal surgery, cranial base surgery, microvascular decompressions, tumor surgery, all of these things, and you know, these are a standard array of neurologic deficits that when you're actually recording each and every adverse event, whether it's related or not to the DuraSeal, these are sick patients and you just have to be a neurosurgeon to understand that and a neurologist, I guess, you know.

(Laughter.)

DR. COSGROVE: Of course, you may find more things than we find I'm sure.

DR. JAYAM-TROUTH: In all of your QRAs that you did --

DR. COSGROVE: Yes.

DR. JAYAM-TROUTH: -- you know, did you actually show that there was a pressure change, the CSF was being held up, you know, and that the surgery would be helpful in these patients?

DR. COSGROVE: You know, that was a clinical decision for surgical intervention on the QRA patients was made by the site investigator, and there was nothing in the protocol looking for CSF flow studies or anything like that. You know, typically they have to have the appropriate clinical symptomatology in that the tonsils typically have to be down to the level of C1, you know, before we would consider doing a decompression, but as you well know, the clinical symptomatology from a QRA malformation can be quite diffuse, and so that's a clinical decision that the site investigator took care of.

DR. JAYAM-TROUTH: Okay. Dr. Cosgrove, for the record, there is no data on seizures in your presentation today. There is no data on seizures in the presentation that you submitted to the FDA; is that correct?

DR. COSGROVE: They're on the slides. I guess it was omitted in terms of the three patients who had seizures, but I believe it is in the -- yeah, I think we just have to look a little more closely.

CHAIRPERSON BECKER: Okay. Just a reminder that we will have a chance to ask questions this afternoon of the sponsor.

I think at this point we'll take about a five minute break while the FDA gets ready to give their presentation, and we'll reconvene at 11 o'clock.

(Whereupon, the foregoing matter went off the record at 10:55 a.m. and went back on the record at 11:05 a.m.)

CHAIRPERSON BECKER: Okay. It's now 11:05, and I'd like to call the meeting back to order.

I'd like to give a couple of reminders. Firstly, when you speak, make sure you speak directly into the microphone so that the transcriptionist can actually get a transcription made.

And I'd like to remind the public that while the meeting is open for public observation, public attendees may not participate except at the specific request of the panel.

We'll now have the FDA presentations on this PMA, and the first presenter is Dr. Peter Hudson. He'll be followed by Dr. Michael Schlosser. So Dr. Hudson.

DR. HUDSON: Great. Thank you.

Good morning. I'm Peter Hudson. I'm the lead FDA reviewer for Confluent Surgical's PMA application.

The FDA review team consisted of myself. I did the lead review and the preclinical review. Dr. Schlosser, who did the clinical review. Ms. Silverman, who did the statistical review, she was unable to be with us today, and Dr. Telber Irony (phonetic) is here, another FDA statistician to help us with any statistical issues that might arise. Mr. Rangel, who looked up manufacturing information, and Ms. Braxton, who was a lead BIMO reviewer and looked clinical data integrity.

My presentation, I'm going to briefly go over the device description, look at the toxicology information, biocompatibility evaluations, and then go over the preclinical animal evaluations that were done.

The DuraSeal Dural Sealant System consists of components for preparation of an absorbable polyethylene glycol hydrogel sealant and a delivery system, the applicator and spray tips, and it's packaged in a single use kit.

The sealant is composed of two solutions of polyethylene glycol ester and a trilysine amine solution referred to as the blue and clear precursor solutions.

When the solutions are mixed within the delivery system, it provides for a rapid in situ polymerization of the hydrogel that's intended to assist in the dealing of the dura mater incision line. The mixing of the components occurs right at the tip of the applicator just as the fluid exists the applicator.

The sponsor has done preclinical evaluations to characterize the product. The gel time is less than 3.5 seconds. The pot life, or the amount of time that the precursor solutions can be used after reconstitution is one hour.

They've done in vivo animal evaluations, as well as in vitro analyses, to look at the degradation rate to get an idea of how quickly the material might resorb, and they've determined how much the material will swell once polymerized. The gel will swell less than 200 percent. Two hundred percent volumetric swelling is defined as the percent weight gain over a 24-hour period in a PPS bath would result, for a two millimeter thick layer of gel, would result in less than a one millimeter increase if the gel isotropically swelled.

The DuraSeal device consists of the following chemical components. I'm going to specifically discuss the PEG ester, the trilysine solution, the FD&C blue eye, and butylated hydroxytoluene.

Polyethylene glycol, or PEG, is approved by the FDA as a food additive and is used in topical and oral drug formulations. It's used in ointments and lotions, tablet binders, coatings for pills, suppository bases, and in veterinary drugs.

In addition, PEG has been approved by the FDA as a surgical sealant. FocalSeal by Genzyme and CoSeal by Cohension Technologies are both PEG based surgical sealants. The FocalSeal product is used in lung indications and the CoSeal product is used as a vascular sealant to assist in hemostasis.

The FocalSeal product consists of a PEG polymer of 31,500 daltons average molecular weight. In comparison, the DuraSeal product is 20,000 daltons average molecular weight.

The half-life of PEG polymers increases with an increase in molecular weight. So a general inference from that would be that the DuraSeal product, its half-life could be anticipated to be shorter than the FocalSeal product.

To address PEG clearance, the sponsor did a number of blood chemistry evaluations specifically to address concerns about nephrotoxicity due to PEG clearance. They looked at BUN and creatinine levels. They looked at preoperatively discharge and at three months there were no abnormal blood chemistries noted.

Trilysine is the synthesis product of L-lysine. L-lysine is a naturally occurring amino acid. An extensive search of the toxicological databases did not reveal any associated toxicities with trilysine.

Butylated hydroxytoluene or BHT is an antioxidant and has been designated as GRAS, or generally recognized as safe, for use in food since 1959. It, too, a source of toxicology databases did not reveal any significant associated toxicities.

The WNO, World Health Organization, or WHO recommendation for an acceptable daily intake of BHT is 125 micrograms per kilogram per day. The amount of BHT that patients would be exposed to in one application of the device is 1.3 micrograms per kilogram.

The no effect level that's been observed in mice and rats was 5,000 parts per million and 1,000 parts per million respectively for the mice and rats.

D&C blue #1 is a water soluble dye that's been approved by FDA for use in food, drugs, and cosmetic products. Lifetime exposure animal studies support an acceptable daily intake of 12 milligrams per kilogram per day. The amount that patients will be exposed to with one application of the device is approximately 1,000-fold lower than that.

The FDA has also determined that FD&C blue #1 is not is not carcinogenic is rodents after a lifetime exposure. However, the sponsor needs to submit a color additive petition, or a CAP, to the center for use of the dye in a medical device. They need to submit a CAP to the Center for Food Safety and Nutrition.

The sponsor is currently involved in that process. This is a regulatory process that the panel doesn't need to consider in their deliberations over the safety and efficacy of the device for its intended use.

The sponsor has conducted standard biocompatibility evaluations of the device in accordance with guidance recommendations. The samples of the device were prepared in a way to be analogous to how patients would be exposed to the product in that the sealant plus any extractable chemicals and unpolymerized polymer would be included in the sample. The device passed all of these biocompatibility evaluations.

In addition, the sponsor looked at the immutogenicity of the product in four standard genotoxicity evaluations. The product passed all four of these.

No carcinogenicity testing was conducted in light of these findings and also in light of the absence of any inflammation, suggesting that the individual chemical components would be considered to be transforming agents.

The sponsor has conducted preclinical evaluations to investigate the device's performance characteristics with respect to safety and efficacy. They've evaluated in vivo animal studies to look at the neurotoxicity of the product in a couple of different types of assays and also done in vivo evaluations for the persistence of the product to get an idea of its degradation and resorption characteristics.

Finally, they've also done reproductive toxicity, teratology experiments to look at that issue as well. I'm going to go over each of these evaluations.

In the canine cranial sealing study, the sponsor created a two centimeter long dural matter incision. They loosely repaired that with microsutures and then applied the hydrogel sealant or for the control dogs did not apply anything over the two millimeter gap in the dura matter.

Eleven of 11 control dogs showed CSF leakage at pressures less than 20 centimeters of water, whereas only one of 12 animals showed CSF leakage.

Marked peridural adhesions were observed in three of three controlled dogs at seven days and in one of three controlled dogs at 56 days, whereas with the DuraSeal treated animals no adhesions were observed.

Valsalva maneuvers conducted at one, four, seven, and 56 days showed CSF leakage at lower pressures in the controls than in the treated animals. Histopathology of the control also showed thick dural fibroplasias and minimal injury to the underlying brain tissue, whereas in the DuraSeal treated animals no fibroplasia was observed and, gain, limited injury to the underlying brain tissue was seen.

Implant residual material was apparent at seven days, but was not detected at 56 days out. So the results of this experiment demonstrated that the product could effectively seal a dura matter incision line; that there wasn't fibroblastic or adhesion formation observed with the device in the healing process, and that the implant material was gone within a two-month period.

In the rat brain parenchymal implant study, the sponsor investigated the local irritant and neurotoxicity of the device, as well as they looked at systemic toxicity of the product as well. They implanted one by one by one millimeter sections of polymerized DuraSeal and/or used absorbable gelatin sponge and fibrin sealant as control implants. Absorbable gelatin sponge and fibrin sealant obviously are materials that are used in closure of the dura matter.

Under microscopic evaluation, there was no evidence of a local irritancy effect or neurotoxic effect detailed examinations, the clinical science of abnormal or diseased tissue, and neurologic assessments were conducted at four, 15, 28 and 42 days. The DuraSeal product was considered to be inert, space occupying mass that did not elicit an irritant effect and did not elicit a neurotoxic effect.

In this neurotoxicity evaluation, the investigators looked for neurotoxicity due to injection of the material into the brain. Extracts from the polymerized sealant were prepared and then injected either into the lateral ventricle or cisterna magna and compared to control buffer.

There was no evidence of treatment related neurotoxicity in the DuraSeal or control animals for a 14-day take-down examination, and the only alterations seen were due to trauma induced by the cannulation of the tissue, and there was no macroscopic or you could not see any encapsulation of the material that was injected.

The sponsor also conducted an in vivo model to characterize the degradation and resorption characteristics of the material. They implanted various formulations of DuraSeal into the subcutaneous sites in rats. The various formulations were -- well, they looked every two weeks out to 14 weeks. They excised the implant sites and looked to see if the material was still there microscopically, and they found that the material was degraded with an eight-week period or of shorter duration.

And these results correlate well with what was seen in the canine cranial study. The material was gone within 56 days.

For comparison, clinical CT imaging showed a reduction of approximately 75 percent of the extradural space where the material had been applied at three months.

Finally, the material was investigated for any potential developmental toxicity or any kind of teratogenic effect. The product was injected in a single subcutaneous administration in rats. The DuraSeal did not cause any developmental toxicities on any of the parameters measured in the dams or the fetuses.

So in conclusion from the preclinical information, the device's chemical components don't raise concerns toxicologically, either the individual components themselves or the amounts of those components that patients would be exposed to.

The device, the sponsor has done standard biocompatibility evaluations of the product, and it has been demonstrated to be biocompatible. The tests that they've conducted are those that are recommended for medical devices having this type of tissue contact and for this length of duration.

The animal model evaluations approximated the use in humans and showed that the device could work as intended and did not elicit any tissue toxicities, and there's no evidence to suggest that the device can cause carcinogenesis or reproductive toxicities.

This concludes my portion of the presentation of the update presentation, and Dr. Mike Schlosser will give you the clinical information.

DR. SCHLOSSER: Good morning. I'm Dr. Michael Schlosser. I'm a neurosurgeon and medical officer for Division of General Restorative and Neurologic Devices, and I'm going to go over my clinical review of the DuraSeal study.

To start, the study was done under IDE. The objective was to evaluate the safety and effectiveness of the DuraSeal Dural Sealant System as adjunct to a sutured dural repair during cranial surgery to provide a watertight closure.

As we've heard, the design was a prospective, multi-center, nonrandomized, single arm clinical study with a three-month follow-up period.

This is the proposed indication for use statement for the device. The DuraSeal Dural Sealant System is intended for use as an adjunct to sutured dural repair during cranial surgery to provide watertight closure.

I just put that up there because one of the panel questions, Question 3, surrounds the appropriateness of the indications for use, and some of the discussion we've had this morning already kind of touches on some of our concerns about the appropriateness of the patient study and supported this particular indication for use.

I'm going to talk a little bit about the clinical trial design. We heard a lot about this already this morning, but a few important points I want to touch on, particularly some of the inclusion and exclusion criteria.

As we heard, there were two sets of inclusion and exclusion criteria, those applied pre-operatively to screen patients for enrollment and then those applied interoperatively to determine which patients would be treated. So to start with I'll talk about the preoperative inclusion criteria. As we heard, these are all elective cranial surgeries that had dural incisions. So no nonelective cases were allowed.

Adults between 18 and 75.

The surgical wound classification is expected to be clean or Class I. That's why the CDC definition. And a little bit later I'm going to talk or go through exactly what that CDC definition is, as it becomes important.

And then finally, informed consent had to be signed.

Exclusion criteria, there were some important ones that I've selected. Translabyrinthine, transsphenoidal, and transoral approaches were eliminated. This also falls in line with the CDC Class I for a clean wound, and exposures to these bases would make a clean contaminated wound.

Penetration of other air sinuses or mastoid air cells. In addition to this being a potential source of infection, these are also other routes that CSF obviously can use to escape and cause a CSF leak.

Prior procedure in the same location. So these were all first time surgeries at that location.

Prior radiation or any planned radiation to the site in the exclusion criteria.

Any evidence of systemic or local infection.

And then chronic steroid use that had not been discontinued at least six weeks prior to the trial were all reasons for exclusion.

The interoperative inclusion criteria, and these were the patients who were successfully screened, were taken to the OR as part of the study. They were then examined again interoperatively to determine if they still met the criteria. So the surgical wound had to end up being clean or Class I so that if there was an inadvertent exposure to an air sinus or another reason why the wound would no longer be classified that way, the patient would be eliminated.

Durotomy had to be at least two centimeters in length, and then CSF leak had to be present, either a spontaneous leak or after Valsalva.

I'm going to come back to that part about the spontaneous and Valsalva leaks in a couple of slides.

And finally, the interoperative exclusion criteria, the use of synthetic or nonautologous duraplasty materials. So these are all new patients who could achieve or in which the surgeon could achieve an appropriate closure using either primary closure techniques or using only autologous grafts.

A gap of greater than two milliliters, as we've heard about in a little bit of detail this morning, was a reason for exclusion, and then finally any incidental finding of the preoperative exclusion criteria.

So I'll just pause here to mention that these points I've brought up describe kind of how the population was taken from just everyone presenting for a craniotomy down to the patients who were included in the trial, and it's important for the panel members as they kind of already have started talking about to take that into account when we starting thinking about who are the patients that are studied and who are the patients that the device should be used in.

And that, again, relates to our Question 3 in the panel questions.

Moving on in the clinical trial design, the primary efficacy endpoint, as we heard, was no CSF leakage after up to two dura sealant applications. So the patients were challenged with the Valsalva maneuver. If the DuraSeal was applied, they were challenged again. If they leaked after that first challenge, they could then have an additional application, and then after that second application, any patients that continue to leak would be considered a failure.

The study success criteria was set at 80 percent. This was based on experience and pilot data submitted as part of the IDE. The plan was to use descriptive statistics of the success rate of the study and then compare it to that study success criteria.

And then in terms of safety, all adverse events as we noted were reported to FDA. We had a specific interest in CSF leak and infection for obvious reasons.

The plan during the IDE phase was to do descriptive statistics on the safety events. There was not actually a plan during the IDE phase to use a literature or other control group. The comparisons to the literature were things done during the evaluation of the PMA data after it was submitted.

Specifically, CSF leak was an important concern as a safety endpoint, and so a specific definition of CSF leak was included. We went through this. The sponsor went through this already this morning, but just to reiterate, any CSF leak or pseudomeningoceles that required a surgical intervention, which was breaking of the skin, any CSF leak confirm by diagnostic testing, and then finally any leak confirmed by clinical evaluation.

So this basically breaks down to all leaks of fluid that could be determined to be CSF, and then in addition to that, all pseudomeningoceles that required some kind of intervention. So the only thing being excluded are pseudomeningoceles that didn't require an intervention that involved breaking the skin.

I'd like to speak for a moment now about the design rationale for the study. There are several points to make here.

The first is the fact that the goal of the device was to obtain a watertight closure meant that the device lent itself to a study that used an interoperative criteria. Since that could be easily evaluated and kind of visualized interoperatively, the use of a study success criteria with a specific goal and then a single treatment group to compare to that success criteria seemed like a good match.

In addition, as we've now heard a lot about this morning, there are no approved devices for this indication. Despite that, there are many devices that are very commonly used in our surgical practice as an adjunct to sutured dural closure, such as fibrin glues or other synthetic blues that are altogether being used off label for that purpose.

Since there is no approved device with known safety and effectiveness, no single device that could be used as a control, the idea of using a heterogeneous control group which is standard of care was raised during the IDE stage and the pre-IDE stage of this device.

However, a study that would randomize patients to standard of care would be allowed by the FDA regulations, would put us in the position of having to assess a device safety and effectiveness as compared to a heterogeneous group of other devices the safety and effectiveness of which are not known.

So in a sense you have to evaluate a study whereby your control group or your benchmark is devices with unknown safety and effectiveness, and so we felt that there was significant weaknesses in that study design as well.

Just a note about valid scientific evidence. A PMA application must demonstrate safety and effectiveness through valid scientific evidence. This is per the Code of Federal Regulations 860.7, the definition of valid scientific evidence, which includes well controlled investigations, partially controlled studies, studies in objective trials without matched controls, well documented case histories conducted by qualified experts, and finally, reports of significant human experience with a marketed device.

During the pre-IDE and IDE stage the sponsor and FDA work together to determine an appropriate study design that fits within this definition of valid scientific evidence, addresses the important safety and effectiveness issues for that device and also satisfies the least burdensome criteria of the 1997 Medical Device Modernization Act. And this process was also undertaken with this particular device.

Now, moving into the study results, the population, there was 303 patients screened to enroll 132. Of those 132 patients enrolled, 111 of those patients were treated with the DuraSeal sealant.

Here the patients who were excluded out of the 132 to get to 111, there are six patients due to a sinus penetration; seven due to a gap greater than two millimeters; three dues to less than three millimeter gap from dural incision to bony edge; and six due to the use of a nonautologous duraplasty material.

It is important at this point for me to notice that there are no patients who were excluded because they didn't leak. So all the patients who were considered for inclusion in the study either leaked spontaneously or leaked with a Valsalva maneuver.

So the idea behind using the presence of an interoperative leak was to select for a population that had leaking CSF and, therefore, were at higher risk for the morbidities and mortality associated with postoperative CSF leaks.

However, in this study, all of the patients leak. So there really was no selection based on any kind of predilection towards future CSF leak, which means that the study really describes more of an all comers approach for craniotomies than a specific subpopulation at risk for leaks.

The follow-up, as we've seen in the sponsor's study, there were two patients that died before the three-month follow-up period, and there were two patients who refused to participate at the three-month assessment, giving a total of 107 patients available at 90 days.

However, since we were using an intra-operative criteria for the efficacy endpoint, 100 percent of the patients were available for that endpoint.

This is just a chart showing the different types of cases that were included in the study, and as you can see, it kind of runs the gamut of typical intracranial neurosurgical procedures, including vascular procedures, nerve decompressions, epilepsy, and a variety of different tumors.

The primary efficacy endpoint. All patients leaked intra-operatively, as I've already mentioned with the Valsalva or spontaneous leak. This is the breakdown. Sixty percent had spontaneous leaks, and then the final 40 percent had a leak after Valsalva. I'll mention there that that also plays into our Question 3 to the panel when we ask whether or not the difference between someone spontaneously leaking and someone who leaks after Valsalva is important in determining how the product should be used in the future.

One hundred and five out of 111 subjects had no CSF leak after the first DuraSeal application. So they had the sealant of Valsalva was then done to 20 centimeters. One hundred and five of those patients didn't leak.

The remaining six had a second application, and no patients leaked after their second application. However, there were two patients who only had a Valsalva to ten centimeters of water rather than the required 20, and so if we take the conservative approach, assuming those two patients would have been failures had they had the 20 centimeter Valsalva, then we get 109 out of 111 for the success rate, which comes to 98.2 percent.

Looking at this statistically, this is the study success, at 98.2 percent the success criteria set out during the IDE phase of 80 percent. The brackets here represent the 95 percent confidence interval, and as you can see, the lower bound of the confidence interval which is at about 93 percent is still well above the 80 percent study success criteria set out at the beginning of the study.

So now I'll move on to talk in a little bit more detail about safety. This is a summary of kind of the important serious adverse events seen in the sponsor's data. The items selected in yellow are the items that I've chosen to look at in a little more detail.

The deep wound infection, there were nine such events in eight patients. As the sponsor mentioned, they didn't cascade events. So there was one patient who presented on two separate occasions with a wound infection that was counted as two separate events even though it appears from the clinical history that the patients simply had an ongoing infection over the course of the follow-up. But that was counted as two separate events, giving nine events in eight patients.

CSF leaks, six events, and then bacterial meningitis, two events in two patients. Again, there's overlap here in that one patient had both a deep wound infection and associated meningitis. So that patient had two events recorded even though it was probably one infection.

The other events listed here are stroke, hydrocephalus, aseptic meningitis, cognitive disturbance, cranial nerve deficits, are typical events you'd seen in a post craniotomy population and not of a significantly high magnitude to raise a concern.

I'll start by examining postoperative CSF leak in a little more detail. This was looked at as both a safety/adverse event endpoint as it was collected, but we also examined CSF leak to determine if any additional information about the benefit of the device to these patients could be gleaned from the CSF leak results.

Post-op CSF leaks, as I mentioned, occurred in six cases. There were three pseudomeningoceles which required some kind of surgical intervention, thus fitting the criteria. There were two overt CSF leaks through the incision, two out of 111, giving a rate of overt incisional CSF leaks, and there was one leak discovered intraoperatively which we've heard some detail about that case from Dr. Cosgrove. This is a patient who underwent a debridement of a wound infection, removal of the DuraSeal, at which point in time there seemed to be some pooling of CSF, and a lumbar drain was put in to prevent future leak of that CSF through the wound, but since the patient underwent a surgical procedure, being the lumbar drain, it was felt they met the criteria for CSF leak set down in the study and, therefore, were counted, giving us an overall leak rate of six out of 111, or 5.4 percent.

So as I mentioned, the plan in the study was to do descriptive statistics, which was done giving us that 5.4 percent rate. However, to understand what that rate means a little better, FDA undertook a comparison to the literature.

And so I've selected out a few studies from that large literature review that was done that I think are interesting to point out. The first is the BioGlue study, which was done by Kumar, et al. This was a study done outside the United States on the synthetic glue that was used as an adjunctive dural sealant, but who was not approved in the United States for that use.

Two hundred sixteen elective craniotomies were included. There was only a six-week follow-up period required. CSF leaks were screened for by physical exam only, and only overt CSF fistula is reported. In the literature article there was no mention of pseudomeningoceles, and so there were two cases, or 1.2 percent, of overt CSF fistulae, but as it's obvious from this slide, their definition of CSF leak was different from the one used by the sponsor. So it's difficult to compare apples and apples with this study, but if we look at the rate of overt CSF fistula in the DuraSeal study, which was 1.8 percent, it's similar to the 1.2 percent seen here. However, we don't know anything about the other types of leak in this BioGlue study.

Another study on DuraPatch, which is a dural substitute, involved -- and this was published by Von Wild in Surgical Neurology in '99. One hundred and one elected craniotomies, so again, only elective cases just like the DuraSeal study. They excluded lesions of the skull based on invasion of the frontal sinus, and all of these cases were such that an allograft to patch the dura was needed.

So the exact details of the types of procedures is not robust since this is just a literature article. You can make the assumption that these are more complicated dural problems, larger dural holes that could only be fixed with an allograft patch.

You can certainly make the assumption that none of these cases could be closed simply primarily with stitches. So a slightly different population in terms of the problems facing the surgeons in getting the dura closed.

Follow-up in this case was six months and did include CT and MRI. However, only 75 percent of the patients were available for that six-month follow-up and the only other follow-up was actually seven days or at discharge, and so most of the information they have is on that seven-day follow-up, and then they have a substantial 25 percent loss when they go out to their six months.

CSF leaks were clinically diagnosed. Again, a very specific definition like was used in the DuraSeal study is not provided. However, they had a much higher rate of 12.9 percent, a numerically higher rate.

All of those patients had some kind of CSF leak that would have been included in the DuraSeal study, but given that we don't have all of the details on how they selected the CSF leaks, it's tough to know if the number had they used the same rigorous criteria we used would have actually been higher or lower. But just for comparison's sake, we see a higher rate here of 12.9 percent.

And then the last study I'll mention is a study of aerosolized fibrin sealant. So as we've mentioned now, fibrin glue is very commonly used in these neurosurgical procedures as an adjunct. This study looked at using an aerosolized delivery system versus the standard fibrin sealant delivery through a syringe. It was a retrospective study, 295 cases with the aerosolized variety and 214 with the normal application. It was only elective supratentorial craniotomies. So that's a subset of the population that was seen in the DuraSeal study, which also included infratentorial craniotomies, and they excluded skull based approaches.

There was only a two-week follow-up minimum required, and again, a specific definition of CSF leak was not given. For the aerosolized group the leak rate reported is 3.1 percent, and 8.9 percent for the non-aerosolized group.

And, again, all nine leaks that were reported in the aerosolized group were described in the paper as either being treated with subcutaneous punctures or with lumbar drains, meaning that they did fit into the criteria for the DuraSeal study. However, since we don't have a specific definition given to us in the paper, we're not sure how exactly they were selecting for those leaks.

So this is summarized. It goes without saying that there are numerous reports in the literature of CSF leaks across a variety of different types of surgical procedures, and reporting a variety of different results, but I felt these three articles kind of gave us a span of what's available.

The rate of the DuraSeal study, 5.4 percent. The BioGlue study, which obviously had a rigid definition of only CSF fistula, had a lower number. The DuraPatch study, which didn't give us a definition, had a higher rate of 12.9 percent, but again, this is probably a different problem facing the surgeon in terms of achieving a dural repair than the one that was studied in this study. And then the aerosolized fibrin sealant which was a larger study and probably a more heterogeneous group of craniotomy patients, but was retrospective and, therefore, is subject to some of the biases associated with the retrospective design. It kind of shows some rates that kind of span the DuraSeal rate, 3.1 percent and 8.9 percent.

So we have seen that the rates seen in the DuraSeal study certainly fall within the range reported in the literature, and depending on how they selected their CSF fistulae, the numbers came out either higher or lower, but certainly the 5.4 percent fell within the range.

Moving on to infection, there were nine wound infections, as I mentioned, eight deep infections and one superficial. All of the deep wound infections required a reoperation, one being debridement and the other seven debridement and bone flap removal.

The one superficial infection was treated with antibiotics. The overall wound infection rate, therefore, is 8.1 percent. The 95 percent confidence interval on that rate is actually quite wide, going from 3.8 to 14.8 percent.

There were additionally two cases of meningitis. As I mentioned, one of those cases was in a patient who also had a wound infection, and so if we look at a number of patients who had a procedure related or neurosurgery related infection, it would be ten out of 111 or nine percent.

I mentioned I would come back to the CDC definition of wound classification, and here it is. Clean or Class I wound is an uninfected surgical wound in which no inflammation is encountered, and the uninfected respiratory alimentary, genital, and urinary tract is not entered.

In addition, clean wounds are primarily closed and, if necessary, drained with closed drainage. Surgical incision wounds that occur after nonpenetrating or blunt trauma can be included in this category, which obviously means penetrating trauma would not be.

And then clean-contaminated or Class II includes penetration of the air sinuses, the alimentary, genital, or urinary tracts, if done under a controlled situation, and also includes cases in which there's unusual contamination, meaning some kind of breach in sterile technique in the OR resulting in a contamination, but no obvious infection. So breach of the air sinuses in the presence of an infection would then bump it up into the next level which would be a contaminated case.

So this is the definition by which the patients for the study were selected. The literature, however, doesn't really make use of the CDC definition just by itself because studies have identified other factors which are important for predicting infection.

We've heard about some of those from the sponsor's presentation. They include the length of procedure being greater than two hours, implant of foreign body, particularly shunts, which neurosurgeons in the room are well aware of, and then ASA score.

There actually are other risk factors as well, but I'll focus on these.

We've heard already about the Narotam study. This was the 2,294 patients in which he sought to determine what the risk factors for infection were in neurosurgical cases. He used slightly different criteria. He defined clean as elective surgery, not containing one of the above risk factors, and those risk factors are entry into paranasal sinuses, cranial base fractures, breaches in standard surgical technique, and surgery greater than two hours.

So it becomes quickly apparent that there are some things in here that weren't included in the CDC definition. There are also things in here that were not included in the Class II clean-contaminated CDC definition. So this kind of lies somewhere in the middle and doesn't really fit into one definition or the other very well.

And then contaminated in this study were open fractures, contamination of the site known to have occurred, CSF leakage, and repeat surgeries.

So if you break this down now and look just at the clean contaminated cases in this study, he then subdivides even further. So we're looking at the subgroup of clean contaminated and then subgroups of that subgroup being just the patient in which entry into the sinuses occurred, fractures of the cranial base, surgery at two to four hours, and surgery at greater than four hours.

And then down here at the bottom I have the infection rate for the truly clean cases. So patients who had none of those, and that rate is extremely low, 0.8 percent.

So in comparison to all of these rates, the clean-contaminated class as a whole had a statistically higher rate than the clean case. However, an important point to note here is that this surgery two to four hours was 5.6 percent. Surgery greater than four hours was 13.4 percent.

However, that difference was not statistically significant in his study, and so these numbers look different, but in actuality all he could say was that surgery greater than two hours was a risk factor. Greater than four hours didn't prove to be statistically worse than the two to four-hour group.

It's also kind of an important point of the power of these studies. I mean, you have 178 patients here and 23 here, which was not enough to be able to tell the difference between these two rates from a statistical standpoint.

The DuraGen study, which was actually published by the same author, was a study looking at dural closure using the DuraGen product, which is a collagen product, or a control group in which it was not used, and we can see here these stratified by clean, clean with foreign body, clean contaminated in all of the cases.

As was mentioned by the sponsor, foreign body use was not rigorously collected on case report forms for this study, which is why I just haven't included anything under that column.

However, we do know about clean versus clean-contaminated for the DuraSeal study. There were no infections in the seven clean cases, and there were 12 infections out of the 102 clean-contaminated, giving us a rate of 11 percent, which is similar to the 12 percent seen in the treatment group of the DuraGen study.

The control group had a smaller rate of 4.3 percent, but again, this difference was not statistically significant in the DuraGen study, which, again, just kind of reminds us of the power of these studies given how many cases, 91 in '74 in clean-contaminated.

Looking at just the overall totals, we have the 10.8 percent in the DuraSeal study compared to five percent and 4.4 percent in the two groups of the DuraGen study. As the sponsor mentioned, this number here, 12, is higher than the ten that I presented on a previous slide because Narotam used a little more strict definition, including patients with red wounds as counting as wound infections. That wasn't part of the ID study design, and so they weren't counted kind of initially, but were just included for this comparison.

A couple of other studies. These are a little bit older studies that looked at the use of antibiotic prophylaxis. There are, however, larger studies and prospective randomized controlled studies.

The first one by Young, published in '97, looked at 846 clean procedures. Two hundred and fifty of them were major craniotomies, and they had one-year follow-up. And they defined clean cases as intact skin without evidence of infection. So again, a different definition, though it seems to be quite a liberal one in that they did not necessarily specify their sinus penetration. They didn't talk about including trauma, blunt trauma versus not including it. They just kind of had this more broad definition.

Their infection rate with antibiotic prophylaxis for the whole 846 was .9 percent. If you look just in the craniotomy, the infection rate was zero percent.

Length of procedure was not reported. ASA score was not reported. So there are a few risk factors to infection that we don't know about for this study.

Another very similar study by Bullock was another prospective randomized study of antibiotic prophylaxis. This included 416 clean craniotomies. This study did exclude breaches of air cells in a similar fashion to what was done in the DuraSeal study, and they did report OR time, with a mean OR time of 107 minutes and a standard deviation of 64 minutes. So the average being less than two hours, though with a wide standard deviation, meaning that there were subpatients greater than two hours.

Infection rate in this case was 2.1 percent without antibiotics versus 5.8 percent with antibiotics. So it seems like slightly higher than the previous study, but in terms of the confidence intervals and statistical differences, probably just very similar numbers.

We've heard a little bit about the DuraSeal pilot study. This was done in Europe. It was not an IDE study. There were 47 patients. In that group there were two wound infections, 4.3 percent. However, again, a wide confidence interval of .52 to 14.5 percent.

There was one stitch abscess, which doesn't meet the CDC criteria for a wound infection. So it wasn't counted appropriately.

And all but one case in the study were greater than two hours. So similar to the pivotal study, these were long, complicated cases, 38 percent greater than four hours, but the ASA scores were less. Only four cases that were greater than two, compared to 33 percent of the cases in the pivotal study that were greater than two.

So we have just as long procedures, but slightly healthier patients, and we get a similar number.

This table summarizes the studies that I've presented, and again, just like CSF leak, there are numerous studies in the literature that you can look at to try to estimate what infection rates are for a craniotomy. I've only selected a few that I think are descriptive. I'll move on to the next slide because it shows the same data in a graphical presentation.

On the left here we have the studies that involve only clean cases. This, again, as I mentioned, the definition of clean can change from studies from one site to the next, but these were these prospective randomized studies of clean cases.

In the center are the clean contaminated cases and in the end, the DuraSeal studies which have a combination, though they did have a majority of clean-contaminated cases.

And the important thing to look at here really are the error bars, and so I think what you can see is that the error bars on both the DuraSeal study and also on this DuraGen study really kind of span the results seen in the other studies, and so it's difficult to make a statistical comparison or to say that this is either significantly higher than this or the same as this. I think those statistical comparisons are challenging, not only given all of the differences in the study design, but just given the results.

If we even just forget about the fact that these are all different studies with different criteria and just look at their results, the error bars are wide. So really the results are kind of all falling within a very similar region.

I'm just going to come back to this difference between the spontaneous leakers versus the induced leakers with Valsalva. One of our panel questions refers to those two populations. So I broke the results down by those two groups.

The wound infection rate, 7.4 percent in spontaneous leakers, 9.1 percent in induced leakers. So really not very different.

And CSF leak, the same kind of result, 5.9 percent in the spontaneous leakers versus 4.5 percent in the induced leakers, but those numbers are close and not statistically different from each other.

In conclusion, the sponsor reached their primary efficacy endpoint as set out in the study design of a success criteria greater than 80 percent with their 98.2 percent, the lower bound of the 95 percent confidence interval being 93, and greater than 80 percent criteria.

Postoperative CSF leak rate was 5.4 percent. The wound infection rate, 8.2 percent, and the procedure related infection rate, nine percent.

I put those numbers up there by themselves because I think after the intellectual experience of examining the literature and trying to come up with a good comparison, we really come up with the conclusion that the results in the literature are varied. They use different definitions. They use different criteria. They're not IDE studies. There's a number of reasons why we can't come up with one good number as the comparison, and so I would say that the best we can learn from these studies is that with the use of the device, this is the CSF leak rate and this is the wound infection rate.

Thank you.

CHAIRPERSON BECKER: Thank you, Drs. Hudson and Schlosser.

Does anybody in the panel have a question for the FDA?

DR. CANADY: I just have one question. It is really the same question.

What was the control group in the DuraGen? What kind of defects were left?

DR. SCHLOSSER: It was patients in which the dural closure could not be completed with sutures alone, and so they didn't specify any specific number, like two millimeters that was used. It was simply patients in which an augment to the dural closure was required, and so it's a heterogeneous group in terms of the size the hole was.

CHAIRPERSON BECKER: Dr. Haines.

DR. HAINES: For Dr. Hudson, I just wonder if there's any toxicity data on direct application of blue dye in the spinal fluid.

DR. HUDSON: Of the blue dye?

DR. HAINES: Yes.

DR. HUDSON: No.

CHAIRPERSON BECKER: Dr. Jensen.

DR. JENSEN: Dr. Hudson, in the animal testing or in any of the tests, was there examination of the CSF fluid?

DR. HUDSON: I don't believe there was. Pat, do you k now?

DR. JENSEN: I didn't see it, and since the material was applied to the CSF, was that a consideration for the FDA in asking for CSF examination?

DR. HUDSON: We didn't ask them to do that. It's a good comment.

CHAIRPERSON BECKER: Dr. Egnor.

DR. EGNOR: This is for Dr. Schlosser.

Regarding the FDA's recommendations about control groups for this, why is it undesirable to compare the efficacy of DuraSeal to the standard way of managing these problems, even if the standard way involves using agents that haven't been approved by the FDA?

DR. SCHLOSSER: It really has to do with how we would interpret the study results at the end, and so I think that while as a neurosurgeon you may say that I'm comfortable with the standard way of managing these patients and if you tell me that this product is as good as the standard way, that's okay.

On the FDA side, we have to say that for all we know, all of those products are unsafe and ineffective, and in fact, maybe causing increased infections, causing increased CSF leaks because they haven't been studied.

And so to say that this product is equivalent to the heterogeneous standard of care might be to say that it's equally bad, which leaves you with the concept that maybe then you have to show superiority, but then that's a very challenging study to design. How much better do you need to be?

That's also making the assumption that those products don't work when, in fact, they may work but just haven't been studies, and then you're setting them up for a study that they can't complete because they have to show they're better at something that in actuality is equivalent.

And so it's just a challenging design. As Dr. Witten mentioned, it's not that we would not allow them to do such a study if they wanted to, but we simply advised them that we felt there was a weakness in that design and in our ability to interpret the results of that design.

CHAIRPERSON BECKER: Just to play devil's advocate, I can name you several studies that are currently being done with standard of care therapy that's not proved for stroke prevention, for instance, that are looking at neurological devices against unproved standard of care.

So I don't think it's completely out of the real of question to proceed in that way.

DR. SCHLOSSER: But, okay, to follow up with the devil's advocate though, I would say that studies that are currently underway fall under my first comment, which is that we would allow them to do it. I would be curious if you would tell me studies that have been approved based on the comparison to a standard of care.

Because as we said, we'd be happy to let them do it. Our concern was that it was not a study that would eventually lead to an approval, or it may have problems in leading to an approval.

CHAIRPERSON BECKER: Dr. Haines.

DR. HAINES: But to follow up on that and on your final comments, how does not having that control help us reach a conclusion?

DR. SCHLOSSER: I think that our comment would be that not having the control certainly isn't better than having the control and reaching a conclusion, but our feeling was the opposite, that having the control would not put you in any better situation than you're in right now, that you would have the same problem you have right now if you had that control, and that you may feel as though this number is as good as the control group, but we would feel the whole time that we don't know what that control group means, and that may be you may be relying on a number from a control group that seems okay when, in reality, that's not okay. It's actually a safety problem.

CHAIRPERSON BECKER: Dr. Loftus.

DR. LOFTUS: Unless, just as an argument, you know, as a somewhat pedagogical point, but unless you accepted a control group and developed use of no agent, which was an off-label agent, and an argument has been made that this is an unacceptable surgical standard. Many of us would disagree with that.

DR. SCHLOSSER: And that was something that was also thought about and, you know, the reasoning behind not taking that approach was simply that the neurosurgeons that were consulted, you know, by the sponsor felt as though that was not an acceptable standard of care to leave those patients open.

And I think that I would agree that the community is probably divided on that issue. I think you could probably find surgeons who, like Dr. Cosgrove mentioned, like the French, who think that closing the dura is just something you do and you probably don't even need to do it, and you could find surgeons who would tell you that you absolutely must have a watertight closure.

And so I think that that's a tough decision to make, given that there probably is an accepted standard of care, but the surgeons that the sponsor was working with, you know, they fell in the second category where they felt it was inappropriate to leave patients, especially with spontaneous leaks, you know, without any adjunct to dural closure.

CHAIRPERSON BECKER: Dr. Jayam-Trouth.

DR. JAYAM-TROUTH: In the Young study, you know, the dimension of the 846 clean craniotomies, you used that for infection. Was there any indication in that study, you know, as to what the CSF leak rate was, you know, what type of surgery it was and, you know, whether they used anything at all to stop those leaks?

DR. SCHLOSSER: Yeah, they did not mention specifically the CSF leak rate in that study. So we don't have CSF numbers from the Young study.

In addition, of the 846 cases, only 250 were craniotomies, and so all of the leaks from the spine, I would say, are a completely different physiologic problem and aren't really comparable, and then they didn't report what the leak rate was for the craniotomies in that study.

CHAIRPERSON BECKER: Dr. Germano.

DR. GERMANO: For Dr. Schlosser.

In this study, 111 patients that met the inclusion criteria leaked after experienced neurosurgeons closed the dura. Did you find in your review of the literature that this is the case? In other words, dural closure cannot be accomplished at all?

This is question number one. And question number two: if that is the case for those neurosurgeons that participated in this study, why didn't they select 50 percent of those patients to be enrolled and for the other 50 percent not to be enrolled?

DR. SCHLOSSER: Okay. The first question, I would say that the literature does not report on using Valsalva maneuver to test for a CSF leak. It is something that's done. I wouldn't say it's routinely done, but it is something that's done particularly in the spine, but also in craniotomies to test your dural closure, but it's certainly not something that's done in the 100 percent of cases, and it's not at all reported on in the literature.

In fact, the status of the dural closure prior to closing the galea was not really reported in almost any of the studies also, and so they never really comment as to whether or not there was CSF leaking through the suture holes or the incision in any of those cases that went on to develop leaks.

So that's information that we kind of have in this study that probably hasn't been really looked at rigorously in these other studies in the literature.

As far as, you know, the result, the fact that everyone leaked, I think I would like to get the sponsor's input, but I think that that would surprise me, that I would have not thought that to be the case. I would have thought that at least a portion of the sutured dural closures would have stood up to that Valsalva.

That wasn't the case. It turns out that all of those patients leaked. Now, you know, why not just exclude all those patients? Well, there's one very pragmatic answer, which is that the study design that was already approved included all of those patients, and so you really would have had to start over with a new study at that point, which you would have had to have done, of course, only after you completed the study because after the first 40 patients you may have thought, well, we're going to encounter 40 more that won't leak.

And so really at the end you would have had to decide that now after doing the whole study we need to start over.

Now, in hindsight, you know, what would the results of the study have been if we only include spontaneous leakers? Well, we don't know the answer to that question.

CHAIRPERSON BECKER: Dr. Ellenberg.

DR. ELLENBERG: Dr. Schlosser, let me follow up again on the issue of the control group. Given that the sponsor came in with an expectation of hitting above 80 percent success rate, where "success" was defined sa no leakage, it's not clear to me how that argument plays out.

If you were starting in an open field discussion of, well, we really had no concept of how this thing was or was not going to work, I'm sympathetic to that argument and probably to the approach.

But when we're talking about something based on the pilot studies' literature review is expected to work quite well and it's just a question of how quite well, and you're in the range of 80 percent and you're actually shooting to go as they did to well over 98 percent or 98.2 percent; I'm not sure how that argument works.

Because if the standard of care group was -- I'm sorry. If both groups were equally bad at the 85 percent level or at the 98.2 percent level, I think we would have had a lot of information to deal with.

So if you're talking about no knowledge and you're worried that comparing the control group to the sealant group and they were competing for a place in the eight percent level, so to speak, I'm sympathetic to your argument. But when the expectation is 80 percent, I really don't understand how that argument still holds.

DR. SCHLOSSER: Okay. I think I understand what you're asking. I think there's two questions there, and that is that, you know, why is it 80 percent, and then, you know, why is it that we don't need to test that product against something else rather than just against the number 80. Is that the correct --

DR. ELLENBERG: No, it's the issue of the current control group versus testing against what's in the literature, and in this case it turns out we're basically testing the safety against what's in the literature more than the efficacy.

DR. SCHLOSSER: Right, because I think the efficacy -- I'm not sure that this study design is any worse or any better than having a control group. However, I think that given what we now know about the results of the study, I think that, you know, especially if their control group was, you know, no treatment, you would have had zero percent versus 98 percent as your two groups because, I mean, no treatment, clearly all of those patients would have leaked.

And then if you allowed them to use standard of care and put another number, you know, other devices in, they would have had some other rate possibly as high as 98 percent or somewhere in between.

And so I think that comparison would have told you the same thing that our efficacy endpoint told you in that we kind of know that the goal of this tool is to prevent CSF from leaking out through the incision in the OR, and they achieved a 98 percent success rate at that.

And so I think that the question regarding the control group is really, as you mentioned, really more one of safety.

DR. ELLENBERG: Absolutely.

DR. SCHLOSSER: But I think that, you know, the safety of that control group is from our standpoint completely unknown. And so I think that you could speculate during the study design that if the numbers came out a little low but similar that maybe you would have some confidence, you know, that the control group was also safe and that the treatment group was safe, but I think that in the end you would have not had anymore assurance.

You know, you're comparing to an unknown. I mean, you have to make assumptions about that unknown that we're not willing to make because we make people do studies to prove safety. We don't make assumptions about safety.

And so I think in order to evaluate that result you have to make an assumption that we don't routinely make at FDA, and that is that something that hasn't been tested under an IDE study can be assumed to have a certain outcome.

DR. ELLENBERG: But you're asking us to advise you in what seems to me to be a less opportune situation where we're looking at a nonconcurring control group cold from the literature. That's not good in terms of assessment of the safety.

If the control group had a lower profile for safety -- excuse me -- a lower infection rate than the DuraSeal group and the DuraSeal group was as effective as it is now and presumably it would be more effective than the standard of care because there must have been that motivation in bringing this this far along, my sense is that we would have a much better feel for what the safety issues were.

If you didn't know -- not you personally ‑- if the world doesn't know the safety profile for standard of care, then after this study they would have a better handle on what the safety profile standard of care was in spite of the fact that the control group would by the nature that the standard of care is described, where basically the surgeon is there, there's a problem, there's a leak, and there's a shelf full of options, and the surgeon individually determines based on the type of surgery, the patient condition, et cetera. That couldn't be changed. I understand that, but that is an approach. It's a defined approach. It's what happens every day in the surgery theaters in the United States and apparently not in France --

(Laughter.)

DR. ELLENBERG: -- but it's fairly standard of care.

I simply don't understand why that comparison would have been helpful on the safety side and why it wouldn't be better than what we're being asked to judge.

DR. SCHLOSSER: Well, again, I think there's an assumption being made there, and that is that in the end of the study, the numbers would have come out in a certain way, meaning that the rate would have been higher or would have been lower.

I think that, you know, the opposite could have been true, and I think that the panel could have been given a false sense of security if the numbers had come out the same or if the control group had come out with a higher number. You might have been given the false sense of security that, oh, this device is safe because its number is the same or lower than the control group, whereas in reality all that may have been telling you is that the device is just as unsafe as standard of care.

And I think that the panel may have been, you know --

DR. ELLENBERG: But what's wrong with that answer for this particular application?

DR. SCHLOSSER: Because we don't approve devices based on the fact that they're as unsafe as other unapproved devices. We approve them based on the fact that they demonstrate a reasonable assurance of safety and effectiveness.

And so I think that the short answer to the question is that we didn't know that the panel would be in a better situation with that study than they're in now, and from a least burdensome approach, this was the least burdensome of the two studies, which in our estimation would give the same level of results and kind of put you in the same position that you would be in with that other study design.

But I will reiterate what Dr. Witten mentioned, which is that that design was an option and that it was not that the FDA would have disapproved the IDE if they had chosen to use a heterogeneous control.

DR. ELLENBERG: I understand that.

DR. SCHLOSSER: We simply advised them we thought there was weaknesses in the design.

CHAIRPERSON BECKER: I think we'll let Dr. Schlosser off the hot seat for the moment and break for lunch. We'll reconvene at one o'clock, and there will be a chance for more questions for the FDA and the sponsor as well.

(Whereupon, at 12:11 p.m., the meeting was recessed for lunch, to reconvene at 1:07 p.m., the same day.)

AFTERNOON SESSION

(1:07 p.m.)

CHAIRPERSON BECKER: It's now five minutes after one o'clock, and we will resume the panel discussion.

Two lead panel reviewers, Dr. MacLaughlin and Dr. Canady, will open this part of the meeting with the remarks to help focus the deliberations. The panel will then discuss and deliberate on the information in the submission and the information that the sponsor and the FDA presented.

The panel can ask the sponsor or FDA questions at any time. After a general discussion, the panel will address the FDA question. Then there will be a second open public hearing and FDA and sponsor summations. Then the panel will conclude the deliberations and vote on the recommendations concerning the PMA.

The first lead panel reviewer is Dr. MacLaughlin.

DR. MacLAUGHLIN: Thank you very much for setting up this overhead for me because my CD burner crashed and I don't get a chance to make a fancy presentation, but this brings me back to my old days, anyway, in school.

So what I did was to try to summarize what was done by the sponsor to sort of analyze all of the materials that go into this DuraSeal product, how it was tested, how it's made, and what sort of controls are built in for the ultimate safety of the patient.

And as we've all heard, this device is made to, you know, make sure that we close wounds in the dura that are up to two millimeters in width, and I think what's important to note, too, is that this hydrogel product is an absorbable, cross-linkage polymer of 20,000 molecular weight, and this cross-linking is done in a non-exothermic or endothermic way. It's an isothermic reaction. It happens immediately. So it doesn't generate any local heat, which can sometimes happen in chemical catalysis.

And I think that's a useful thing to point out because I feel that that's another measure of safety. It polymerizes right away, and it doesn't create any local heat, and it's pretty stable, as you've heard, to 37 degrees C.

And the desired performance characteristics I mention again because they're part of the testing procedure that went on at Confluent Surgical in order to evaluate how well the product that they were getting is performing.

So it needs to be easy to use and needs to be absorbable, and it had to adhere to the dura and not to other structures, the sort of lubricous characteristic that we've heard about already, and it needs to be biocompatible. And I think many of these things were tested in these products over time, and I think it's important to note also that everything you use in this product is bought off the shelf. I mean certain items are made to Confluent's specifications, but they're all available and used widely in lots of other applications, and that was important to me in this analysis, and there are three or four different vendors of the materials. I didn't mention all of the vendors for the plastic stuff, for the syringes and the caps and the containers and all of that because they have all been covered, I think, by the FDA under many other applications.

But all of the material that the company gets is delivered to the site, and they sort of package it together after testing it. It goes to another company to ship it out. So there are controls built in that I'll talk about in a minute for that.

So, anyway, I think a couple of points that I wanted to raise in this analysis was what Confluent does once they get the product and why they've arrived at certain specifications for the product in particular, some questions I really want to raise in that.

The other thing that is important is that the breakdown products of this polymer that you heard about in this morning's discussion are basically the same as the product itself. So you don't need to worry about a new, you know, actor in the game for toxicity. You're really looking at the same thing, going, dissolution, being cleared at the end of the day. So that's important to me.

So how happy am I with all of this? You know, I sort of looked at it to say what would I really want to have done and have accomplished, and the performance characteristic testing at the site -- this is not in the patient. It's either in the animal or in vitro -- has to meet certain standards that the company sets, and this is a few questions I wanted to raise here as to why they're set.

The reconstitution of the PEG, this polymer in its buffer, you know, should happen in give minutes. It's simple. You're just going to dissolve the material. It has to be completely in solution quickly. That's easy to analyze, easily understood.

The gel time is three and a half second. They tested this by taking the product and just squirting it from one of those syringes into a beaker that has a stir bar in it. Boom, in three seconds it has solidified. Simple test, not hard to confuse. That's important because it relates to the, you know, chemical composition of the products as they're mixed.

One thing I did have an issue with though is this so-called swelling characteristic. This is 200 percent, and that's I understand why it's not good to have a lot of swelling in the brain. I don't understand why 200 percent is the standard they picked -- excuse me -- the specification they picked because when they analyzed the material, their own data shows it to be way lower than that. So why pick a huge window when it really should be maybe smaller. I'd like some feedback on that

The hydrolysis in vitro is one and a half to four days at 60 degrees Centigrade, which is called an accelerated test. So you know the material is going to be put together. You know it's going to go into a patient. You know it's going to dissolve and be reabsorbed. So one of the chemical characteristics you can test periodically is to make your polymer, put it in a solution, heat it up, and decide how long it takes to fall apart.

So they have this accelerated test and then they have the 25 degrees C. test, and I don't understand why we have those two tests, why they're necessary. I think the 25 degree test makes sense to me. The 37 degree test I have to say doesn't make sense to me because I don't understand what it's telling us. It's not what's going on with the patient. The patient is on 140 degrees, you know. They're 25 degrees.

So, you know, I want some feedback on why that's a standard that they picked. What is that telling you about the safety of that product to have that measure? Real time makes more sense. Real temperature makes more sense to me.

So the application, this is the syringe integrity polymer, tips and all of that. They went through a series of trials actually using different kinds of products, spraying them, testing for the pattern of spray, how well things polymerized in place, and arrived at, I think, a reasonable set of materials, a reasonable set of syringes, a reasonable set of tips, applicator tips. All of that seems to make good sense to me. I don't have any concerns about that.

The other thing that's careful to inspect every time new products are shipped -- remember this is coming from vendors into your facility -- you have be sure that their oxygen content, especially of the sealed glass vial, is important and the buffer pHes of the mixing reagents are proper. That's something they test all the time. I think they should test all the time because it does affect how much polymerization one gets and how stable the product is.

So just doing a squirt test and seeing polymerization doesn't tell you how long it's going to last. It has to be many different levels of testing, which I think, in fact, they do.

The absorption and the sealing tests Dr. Hudson spoke about, I think they're very straightforward. I didn't have any trouble understanding the goals, understanding the data, or coming to the conclusion that I didn't think there was any toxicity, especially when you consider the historical controls which were done on a lot of these materials. Lots of studies have been done on these materials in the literature, and you look at how much of this material is available in one or two applications into a head. You've got so little of this product around. I don't see toxicity being a major player here of any of the components.

What I'm more concerned about is why certain specification standards were set and how they're tested for.

So another issue is this package integrity, which you have to consider. They're putting lots of different components into a plastic container. It's going to be stored for so many months. It's going to be shipped out to place. How hot can it be? How cold can it be? Is it going to keep bacteria out? Are you going to introduce, you know, bad things through the package itself? I think that's pretty well controlled for, too.

I don't have any difficulty either understanding their goals, the analysis that they used, or the results that they have. I think it's fairly clear. No problems there.

The shelf life issues, though, is another one of these accelerated versus nonaccelerated types of test. When they sterilize the material, it gets irradiated, and if you measure where the irradiation falls and measure how much radiation occurs from the surface through the material, you can get minima and maxima of radiation.

So you're like to be sure everything is stable. So you do a series of experiments in which everything gets the maximum dose versus the standard irradiation of the material.

That is underway, and as far as I know, those results on performance testing have not been completed, but are pending, and I'd like to know if they are completed now because you can have effects on the ultimate product based on irradiation, not of the patient, but of the material as it's sterilized.

So that's another thing I was interested in hearing some more about, and that has to do with acceleration, too. That's a shorter feedback loop to find out if your product is clear or not.

So the toxicity studies and the biocompatibility studies, I think, are also very straightforward to me. All of the non-hydrogen products have historical controls which I have no quibble with, and everything else was tested, I think, pretty much by very standard and well accepted criteria for, you know, genotoxicity, all of the things that have been mentioned actually by the FDA presentation.

Carcinogenicity, as I say, was not tested because of historical controls, which I think are reasonable, and I think the in vivo testing for biocompatibility relating to the seal test in the dogs and the imaging studies, all of the other in vivo animal studies I thought were reasonable because I think they did approximate what happens in the patient. I think it approximated how much material you put in, where you're putting it in, how long they're going to be in there.

So it sort of matched the four to eight week study period, not the three-month control stuff, but the four to eight-week stuff. I thought it matched pretty well what was going on in the patient and no untoward or no adverse effects were not, and I think that's pretty reasonably done.

The extraction was a slight variation on the theme where the hydrogel was extracted and ejected into these spaces referred to by the FDA, and there was no adverse effect there either.

So when we talk about, you know, the dye or specific components having effects, I think of the worst case scenario is right next to the implant or right next to the injection. That's where the dose is highest. That's where if you're going to have a bad effect you're going to see it there, and none was seen.

So I'm kind of back and forth in my own mind about whether that's a useful study to do in a different way.

The last point I think I want to reach is the fetal toxicity study and the proliferation inhibition study. The fetal toxicity study and the maternal fetal compartment study was begun at four days of pregnancy. So a small caveat is while it may be difficult to establish when a rat is pregnant, you know, a lot has happened in four days.

So you start injecting at four days. You know things are pregnant, and you know the animals are pregnant. So from that day on you know there's no untoward effect.

It's just a caveat. I'm not saying do anything sooner. It's just a limitation. It doesn't cover, you know, the nidation period or getting pregnant or anything like that, but again, I don't have any suspicions of any of this material causing any problems, but it's a caveat that you've already had fertilization. You've already had nidation. You're now starting to develop. In a 21-day pregnancy, you're already four days in. So that's a small point.

The proliferation and inhibition studies on the cell growth where they took extracts of the material, put it into cell culture with four or five cell lines I thought was completely uninformative actually. I didn't know exactly what they were going for. I understand you want to see if it inhibits our, you know, causes proliferation of cell growth, but to me proliferation is changing rate of growth. Awful hard to do in four days. Okay?

If you put something into culture, there's no discussion of what the doubling times of the cells were. You know that it was an empty T assay, but you don't know what its states of competency were. We don't have any other data around that, and I'm not sure what it was designed to tell us.

You know from the histology data that there isn't a lot of proliferation at the site of these things. You don't see inhibition of cell growth. You don't see inhibition of cell growth. You don't see wound failure. You don't see the things that would be characteristic of stimulation or inhibition of cell growth. So I don't know what that was done for, and maybe I could be informed about that.

So overall, I think I agree pretty much with the FDA's determination that this material does not contain anything that I think is risky. I don't think by themselves those components contribute to any of the side effects we've been talking about in sealing the dura. I don't see any smoking gun there, and I think they've been reasonably tested.

My concerns are what happens at the factory evaluating all of the things that come in from different sources and what their standard of performance is going to be every time you get a new lot, every time you ship things out.

How long are things stable? Six months it says on the label now. That's the only thing you have real time data for. Any extension of that needs more data, that sort of thing.

I'm staying right within the confines of physiology and your own data.

So that's really all I have to say.

CHAIRPERSON BECKER: Thank you.

Does anybody on the panel have any questions for Dr. MacLaughlin?

(No response.)

CHAIRPERSON BECKER: Would anybody at Confluent Surgical like to address some of the questions raised by Dr. MacLaughlin at this point or in the summation later? Your choice.

DR. CAMPBELL: Thank you, Dr. MacLaughlin. Those are some excellent observations. We'd like to address that.

I'd like to introduce Amar Sawhney. He's the president and CEO of Confluent Surgical, founder of the technology.

I tried to keep a list of your questions one by one. So I'll try to address them. If I miss anything, I trust you'll let me know.

DR. MacLAUGHLIN: I sure will.

DR. CAMPBELL: The first comment you had concerning the swelling and the 200 percent swelling specification, you're correct. That is a specification that we test for. Every lot that is released we evaluate the amount that the hydrogel expands.

The way the test is performed is we weigh it initially, a sample. Then we put it in PBS for 24 hours, weigh it after 24 hours, and the percent increase in weight is the 200 percent specification.

DR. MacLAUGHLIN: I know how you do it. I'm just wondering why you picked 200 percent.

DR. CAMPBELL: The 200 percent specification was like several ways. One, we've looked at competitive products that are currently used in neurosurgery of those gelfoam, flow seal, surgicel, others. Those products can swell in a similar test that much or more, 50 to 200 percent or more.

We've also performed as you're aware studies in canine and rat models. The canine model arguably is a worse case model where you have a durotomy which has been performed in an animal with a fairly small cranial vault compared to humans. You've applied an appreciable amount of DuraSeal there, similar thicknesses to what you would have in humans. You have not removed any kind of brain parenchyma or tissue underneath. So any swelling is felt by the brain. There's no space or void to fill, and the bone flap is replaced and the tissues are sutured over the top.

So arguably, that's a worst case scenario. we perform two different preclinical studies in canines using that model, and in both studies we found no mass effect, no residual effect from that.

DR. MacLAUGHLIN: If I could say, I have to agree with that. I agree with your data. What I'm saying is you're allowing, you know, 100 percent more space to be in this product than you have. I'm just saying make it the standard that you have because if you allow more space, you don't have that data in the dog. You have the data that you have, which is maybe 110 or whatever it is. I forget the specific number, how much percent you actually get of swelling.

DR. CAMPBELL: Well, a lot of those testings were performed with formulations where we were getting up to 200 percent swelling.

DR. MacLAUGHLIN: But none seen. I didn't see any in your data.

DR. CAMPBELL: We have, as you mentioned, refined manufacturing processes, and typically our swelling is less than that right now.

DR. MacLAUGHLIN: Sure.

DR. CAMPBELL: However, we have data that shows that it's safe at 200 percent, and to maintain manufacturability so that lot to lot variations don't affect this, we feel that 200 percent is an acceptable, safe level to select.

DR. MacLAUGHLIN: Well, I have to say I haven't seen the 200 percent data. You know, it was like looking at your volatiles, how much organic volatiles. I didn't mention that in the presentation, but there's a specification that say how much organic volatiles you can have, which are toxic if you get them in high enough concentrations.

I'm not saying we're there yet. We're definitely not there, but the window is really big compared to what you actually have in your lot after lot testing. So I'm just trying to make some determination as to why you need these big windows when your product isn't that big.

DR. SAWHNEY: Amar Sawhney. I'm the president and CEO of Confluent.

Let me attempt to respond. The window is actually sort of not that big because volumetric swelling takes place with the cube function. So while thickness doesn't expand that much --

DR. MacLAUGHLIN: Yes.

DR. SAWHNEY: -- the weight gain can be substantial. So it doesn't take much to reach that, the 200 percent, and when we had done the studies, the data that we have reported on the lot to lot variation is for the more recent lots.

The testing that was done on the canine study with the original materials did have that amount of swelling. So while it is not explicitly pointed out for that particular lot, those were studied. Then we have backed down and proved our manufacturing techniques, but we have tested the worst case scenario in those animal studies.

Also, the animal studies are predisposed because of the limitations, the limited space and the fact that no parenchyma is removed. We believe we have tested the worst case scenario both from a formulation and an animal study perspective.

DR. MacLAUGHLIN: Right. I think it's important for us to see the data. We've only seen your latest stuff, not the earlier stuff. I think that's an important consideration in deciding what the specifications of this material would be.

DR. SAWHNEY: Okay. Good point.

DR. CAMPBELL: A second point you mentioned was our disappearance testing, our in vitro disappearance testing. We initially started off by doing a test which is similar to the swelling test that I described where we get a piece of gel, put it into 37 degree PBS, and then observe it on a daily basis and determine the time at which the gel has completely gone into solution and there's no solids remaining.

If you do that in PBS doing that test, it's up to 40 days or so at which that test occurred or takes for the material to dissolve.

In order to streamline and since this is a test which is used for lot release, in other words, every lot that we manufacture needs to pass this test, we formed in-house testing where we determined the correlation of disappearance rate with temperature.

In other words, as you know, as you increase the temperature, the hydrolysis rate will increase also, and we did it with multiple lots using multiple lots of polymer. We determined the correlation of temperature and degradation rate and correlated that and determined a way to do the test, the same test, where you're determining -- you're demonstrating disappearance, but you do it at a much higher temperature, and it allows you to do it in less than a week.

DR. CAMPBELL: Right.

DR. SAWHNEY: Let me amplify on that a little bit. It's a standard chemical reaction. It's a first order kinetic that's taking place. It's an erraneous (phonetic) plot that you do. Very similar work is done if you look at pharmaceuticals. Their stabilities and standard kinetics can be accelerated.

It's also a bulk hydrolysis. So it is not relative to say sutures which may not have a penetration of the water. Here the material is entirely permeable because it is substantially water. So the bulk hydrolysis can be adequately accelerated with first order kinetics using an elevated temperature and provides a robust extrapolation and allows you to conduct a study and a test as a release criterion and an appropriate time, and we have data demonstrating that correlation.

DR. MacLAUGHLIN: But I guess my point about this is the same as the previous point. You have data in the patient or in the animals. You know how long it takes to go away at that temperature. I agree with you it's first order kinetics, but three or four major elements play: pH, oxygen concentration to get your ultimate right cross-linking.

And what you're doing is correlating one temperature with another, and that higher temperature has no correlate in the animal. So you don't know that that's telling you about the structural integrity of this material. You know that is' faster degrading than at 25 degrees, and I like the conformity of the sort of real time/real temperature data analysis of this material because it goes together really fast. Your own data show oxygen concentrations are very important, and I'm just saying I want a little more justification then.

You can release it faster because there isn't a correlation going back to the patient.

DR. SAWHNEY: Actually let's talk about oxygen. Oxygen concentration during the cross-linking is, frankly, not important. Oxygen is important as part of the manufacturing process wherein oxygen radicals in the presence of radiation sterilization can end up with G incision (phonetic) after molecules, and that's why the keep the oxygen concentration.

Once the solution is reconstituted, the presence or absence of oxygen, it really doesn't have any material effect to it.

DR. MacLAUGHLIN: I'll concede that point. What I'm saying is that when you look at your own analysis of what a product is, all I'm saying is that I guess I don't understand why faster is better. I mean, what advantage does that bring to the table?

DR. CAMPBELL: The main purpose for this disappearance test was just to demonstrate that the material went into complete solution after a certain amount of time.

DR. MacLAUGHLIN: Yes, I understand that. I'm talking about the elevated temperature analysis.

DR. CAMPBELL: Exactly. And the elevated temperature just allows us to demonstrate that in a week rather than 40 or 50 days.

DR. SAWHNEY: It's just a release study. It's a test, and once we have studied the material and we understand its behavior in vivo, now it's more a test of showing that one lot is similar to another lot, and that allows us to do the testing.

DR. MacLAUGHLIN: I understand. We can agree to disagree on this, I guess.

DR. CAMPBELL: And the test has been --since we established the correlation, we validated it by repeating the test with similar lots at 37 and at higher temperature --

DR. MacLAUGHLIN: Understood.

DR. CAMPBELL: -- to show that that correlation held true across multiple lots.

DR. MacLAUGHLIN: Get it.

DR. CAMPBELL: Now, if I get it right, the next question you raised had to do with an update on our current stability studies following the radiation.

DR. MacLAUGHLIN: Yes. I don't think that was complete, was it? It looked --

DR. CAMPBELL: No, it's ongoing. We have multiple lots of product which have been put onto room temperature and higher temperature, accelerated aging, following irradiation.

We have recently completed one year of shelf life testing at room temperature, and we will be sharing that data with the agency soon.

We also are working on accelerated aging data and establishing or determining that correlation, the erraneous (phonetic) for accelerated aging.

DR. MacLAUGHLIN: Okay.

DR. CAMPBELL: And we have that information also.

DR. SAWHNEY: And those studies, it's an ongoing process. Our hope is to demonstrate two year stability, but you know, you don't have that information until you actually complete that length of time. The company is a young company, and we don't see any obstacles to achieving that. It's just it takes time.

DR. MacLAUGHLIN: Right. I don't need it. I just wanted to see where you were.

DR. CAMPBELL: Okay. You also raised a question about the reproductive toxicology.

DR. MacLAUGHLIN: Just a comment. I don't think you should do anything differently. It's just maybe an interpretation issue because you're not talking about the earliest stages. That's all.

DR. CAMPBELL: Okay. If I recall, the last question you raised had to do with the in vitro cell line test, where we evaluated four different human cancer cell lines. We looked at galea blastoma, lung cancer, ovary cancer, and colon cancer, I believe, cell lines. We evaluated them at four days in the presence of the product. It was just an initial screening test. It was not meant to be a definitive test to replace potential in vivo studies, but yet as you mentioned, there's nothing in the raw materials to suggest there's a carcinogenicity problem. It's just a screening test to gather initial information on those four cell lines.

DR. SAWHNEY: Let me provide some extrapolation. As we start, as the product is launched internationally, and clinicians begin using it, somebody may have a concern question of what does this do to seeding of cancer cells.

The short answer is nothing, but what data do we have? So this was an attempt to try to see if we changed the growth rate of a few different cancer cell lines in the presence of the material. Were we having any kind of nutritional supplement effect where we enabling the cells to adhere and proliferate?

The answer we found it's an inert substance. it really doesn't do anything. Whether or not these studies are the most appropriately designed, I don't think these are standard studies. So they are somewhat speculative, and the conclusions one can draw from these are somewhat limited.

I would submit to you that I don't think this is the most robust way to look at it, but it was work that we've done, and we thought in the interest of completeness and the spirit of openness and sharing that we would share that information.

DR. MacLAUGHLIN: I appreciate that. With all due respect, I don't think the way you're going about it is the way to answer those questions, but that's a matter for another day. I think it just doesn't tell me anything.

DR. SAWHNEY: I t wasn't required of us to do those.

DR. MacLAUGHLIN: No, I understand. I thought I'd respond to it when I saw it.

Thank you.

PARTICIPANTS: Thank you.

CHAIRPERSON BECKER: The next lead panel reviewer is Dr. Canady.

DR. CANADY: It is clear from the conversation that's already taken place that my fellow clinicians need not much help in sorting out the issues here. I think that the elephant in the room is really the change from a non-evidence based approach to medicine to an evidence approach where we have a practice that is not really validated. So trying to assess new experimental methods on top of that practice is extremely difficult.

I mean, it's interesting to me as you go through the report how often the comment is the clinicians will be uncomfortable, not that there's data to say that you can't do this, but that the clinicians will be uncomfortable, and I think that's the crux of the issue here.

Yes, I believe that's true. The clinicians will be uncomfortable, but as Dr. Germano brought up earlier, is that a valid basis for that discomfort or just the training practices that we've gone through through the years?

Plus in the absence of historical controls and the range for infection is zero to 20, the range for CSF leak is zero to 20, you can pick any number and any study that will be comfortable.

So I think there is a daunting task faced by this PMA in the ability to compare itself to current practices and to compare itself to the literature.

The second issue, I think, is the use of a clinical endpoint rather than an interoperative endpoint. If we accept that the standard of leakage is comparable to other studies, then we have to question whether the 100 percent or essentially 100 percent -- and I'll give you the ten to 13 Valsalva patients, interoperatively is not a useful standard. If we end up with the same kind of CSF leak that other people end up with, then that standard has no validity, and I think that that's an issue, although, again, we end up at a five to six percent leak rate which is comparable in general to other studies, although two of the prospective studies showed significantly less leaking rate at two percent.

Also, I think when you evaluate the clinical risk of leaking, you have to include the excluded patients, which would include those that would be most apt to leak, which would be all of the patients with chemotherapy, radiotherapy, infection, and all trauma patients were excluded from this study.

So we have a highly selective population that ends up with substantially equivalent statistics, and I think we have to struggle today with what that means.

Similarly, on wound infection, all of the same issues apply. We don't have a comparable comparative group, and I've been trying all morning to tease out the DuraGen control, which seems to be the closest to a real control group, and I'm still not sure what that group constitutes, and also the numbers are low, but on a regular basis, that number of patients, the infection rate was lower.

When you look at all of the studies -- I won't say "all" -- the majority of the studies that have other kinds of materials implanted, their infection rates were higher than without them as well, which raises the question as to whether or not there may be some acceptable increased risk with another kind of implant, which I don't think I can tease the answer to, but I think is a real question.

The final question is the easy one, which is I think that as we saw on the MR today, this has an MR appearance. So I think that with the labeling issues and education issues, we need to make sure people don't get operated because of the appearance of the materials, and we need to make sure that in the labeling and somehow in the education materials that the radiologist in particular who may not be talking to the neurosurgeons and they may not be aware don't read those out in such a way that these people end up with operations that they don't need.

CHAIRPERSON BECKER: Thank you, Dr. Canady.

Does anybody have an questions for Dr. Canady?

(No response.)

CHAIRPERSON BECKER: I guess I should also open it up for any general questions for the sponsor or the FDA from the panel.

Dr. Jensen.

DR. JENSEN: A question to Dr. Campbell. I'll ask the same question I asked to Dr. Hudson. Since this material is applied in such a fashion that it's in contact with CSF, why did you choose not to study CSF parameters in the animal studies?

And I bring this up because, number one, it just makes sense to me that you should examine the CSF. You've gone to all of the trouble to inject the material into the ventricle and you do all of these other studies, but you don't do a CSF examination, and number two, some of your complications you had in your patients included hydrocephalus and aseptic meningitis, and there have actually been some anecdotal cases of those similar complications, quote, unquote, in patients that had hydrogel implanted into aneurysms.

So it does sort of bring up the question: is hydrogel actually as -- and it's a different type of hydrogel. I'll give you that, because it does not degrade and it's intravascular. However, yours is directly applied to the CSF or is adjacent to the CSF in some cases, and it also degrades.

So it just seems intuitive to me that you would have studied the CSF, and I'm curious as to why you did not and whether or not that should be done.

DR. CAMPBELL: As you mentioned, we performed the neurotoxicity study where an extract of the gel was injected directly into the lateral ventricle or cisterna magna of rats. As you're all familiar with, we performed the canine study where the durotomy was performed. So the material was applied. Certainly extracts of the gel that was implanted was certainly in contact with the subdural space.

And in that study we performed neurological examinations of the animals, looked at nine different neurological indices. There was no neurological deficits, but you are correct. We did not do specific analysis of the CSF of animals that had DuraSeal applied.

DR. SAWHNEY: However, we did look at the ventricle enlargement. What we were most interested in, it's a question of what are you looking for. What we were most interested in is are the extracts of the materials since it will be placed on the dura and can be in contact with CSF, is any material being extracted out which can impede the resorption or the production of CSF and deaths contribute to ventricle enlargement, hydrocephalus, things of that nature? Are we somehow clogging up the system?

And for that we actually did histology on the animals after they were sacrificed to look at ventricles and did not see any ventricle enlargement.

We looked for inflammation, which would have been a sign of aseptic meningitis. We did not see any of that.

The hydrogel that you allude to, which is on the hydrogel quartered coils, it's a very different substance, and I would immediately like to very clearly establish it's a different material. It's a nonabsorbable polyacrylamide gel that you allude to. Acrylamide is a known neurotoxin and PEG is not.

So I would really say that that's pretty different, and we did look at all of the reasonable things that we felt were important with indwelling catheters where in the extract materials are directly administered inside, direct intracortical implantation. There's not much else that we could do.

We did comparator studies with other materials that were implanted to kind of look at controls. So we studied it as best as we could, and we were picking up any signs of inflammation.

We should have seen that either in the blood values of the animals or in the histological sections of the brain that were done and microscopically examined.

DR. JENSEN: So you think that if you had elevated proteins and elevated white cells in a CSF, that that would have adequately been reflected in the peripheral system? Because, I mean, you went to the trouble of --

DR. SAWHNEY: No, you would see irritation of the meninges.

DR. JENSEN: Right, but I mean, you have examined a focal area of the meninges, right? I mean there are other things that we have --

MR. SAWHNEY: No, we did a whole brain section, and we looked at the meninges, the ventricles, the parenchyma. All of those were examined by the pathologist.

DR. JENSEN: Okay.

CHAIRPERSON BECKER: Dr. Egnor.

DR. EGNOR: This is a question for the FDA, perhaps Dr. Witten.

It's a philosophical question. If we find that the safety and the effectiveness of DuraSeal is commensurate with the safety and the effectiveness of standard practice, but we don't know if standard practice is safe or effective by FDA definition, what do you do?

DR. WITTEN: Well, I guess I'll give you a philosophical answer. Well, I'll give you two answers. One is that you're going to be read the statutory definition of safety and effectiveness. You should go by that.

But I suppose to answer your question more directly, it would probably depend on what you are -- you know, if you think this is safe and effective as some other practice that you believe is safe and effective, then that would help you make your answer. If you think there's a question, that would give you another answer.

CHAIRPERSON BECKER: Any other questions?

(No response.)

CHAIRPERSON BECKER: Perhaps we'll just get some general comments no and we'll go around the table. Mr. Balo, do you have any comments you'd like to make?

MR. BALO: Yeah, I guess we sort of talked a lot from a sponsor perspective, and I am the industry rep., and it's pretty difficult. In this study design, I agree with Dr. Cosgrove and with what Dr. Schlosser said relative to trying to compare a study to an unapproved device.

I mean, I'm in the industry, and usually when you get into a situation like that you will talk to the FDA. You'll ask for guidance from the FDA, and if the FDA gives guidance to the sponsor that says, "Well, we don't think this is really going to be a good control arm," from their perspective, usually the sponsor will listen to that. Most good sponsors will listen to that.

And they'll say, "Okay. The FDA is sort of guiding us in this direction," and I think this heterogeneous type of control will cause statistical issues. It will cause evaluation issues, and from this perspective, let's just come up with a different type of study arm.

And in addition to that, the sponsor went to outside help and consultants, to other neurosurgeons and asked them for their advice. So I think from my perspective, the sponsor went to the avenues that they had accessible to them relative to helping design the study.

Secondly, a literature search is a very difficult thing to do. Being on the other side of the fence and having to do literature searches in the past, it's always difficult to get up common definitions. I think we talked about that this morning.

But it's hard when a company is trying to find something in the literature to make a comparison when they run a single arm study, and I think the panel should take that into consideration for what the sponsor did because I think they did a good job, as Dr. Schlosser did relative to trying to come up with some comparable with the standard of care.

I'm not saying that they couldn't have used the standard of care. I'm just saying I think that the sponsor did the guidance from the FDA and tried to put it all into perspective in the design of this study.

CHAIRPERSON BECKER: Dr. Loftus.

DR. LOFTUS: Yes, thank you.

I assume that we'll have time to review these questions and give a summary later. This is not the appropriate time, but the one comment I would make that just keeps recurring in my mind, and I want to learn from this; I don't know the answer, but that is that I've been involved in a number of NIH trials, and every time the patient goes through the informed consent process in such a trial and is entered, they are followed to the same endpoints and with tracking of all the same criteria as the patients who actually undergo the intervention.

And I guess from what I heard this morning this data is not available, but you know, all of our or a lot of our questions here could be solved if we have the data on the 23 operated patients who were then excluded, if we knew what their infection rate was and what their CSF leakage rate was.

In an NIH trial, we would have that information, and I guess we don't have it here, and my understanding is for my questions that it's not requisite, but it would certainly be interesting to know.

CHAIRPERSON BECKER: Dr. Egnor.

DR. EGNOR: I agree with Dr. Loftus. i think that I can understand from the standpoint of the sponsor and the standpoint of the FDA that the use of these controls would not be probative in ultimately making this decision, but it sure would add information, but I understand that that's not a fault.

The dog studies are fairly convincing and I think are fairly well done, and I think it just comes down to the notion that is the way that we manage this problem in general safe and effective, because it seems to me that there's reasonable evidence that the safety and effectiveness of DuraSeal is commensurate with other ways of managing this.

CHAIRPERSON BECKER: Dr. Ellenberg.

DR. ELLENBERG: Let me raise a new question with regard to the issue of the lack of a concurrent randomized control situation. The standard, the gold standard perhaps, for testing efficacy and safety is to have a control population, but in addition to that, a sophisticated protocol will make sure that the definitions used in both groups are the same, which has just been pointed out, and that sparked my new point here.

The definitions are the same, but also everything about a protocol is done up from. It's prescribed in a protocol that then goes out to the sites and you do a prospective study.

On reflecting on the evidence that's being presented as a surrogate control group, it's not clear that in the beginning in the protocol the comparison group was defined. The comparison group being perhaps the 2,800-plus study and the subgroup within that study was not as I understand it put into the protocol as the group that was going to be compared to the safety and the efficacy results of DuraSeal.

And even that very previous standard that is used in every single comparative trial was not met here. I'm not saying that my view on the use of the literature as a control to test both the efficacy and safety might be dramatically different, but it would be somewhat different.

My sense is we're picking this control group after the fact, after the study has been completed, and that is simply not good science in my view.

CHAIRPERSON BECKER: Dr. Jensen.

DR. JENSEN: Well, Dr. Loftus made the point I was going to make. So I have nothing further to add.

CHAIRPERSON BECKER: Dr. Canady.

DR. CANADY: I don't have anything further to add.

CHAIRPERSON BECKER: I don't think I have anything further to add either.

Dr. Haines.

DR. HAINES: Yeah, I would just make a couple of points. The primary purpose stated for this device is a very limited technical one, which is that at the time after dural closure that the sealant is applied and you do a Valsalva maneuver, you don't see spinal fluid coming out.

That's what we've been asked to look at in terms of efficacy, but it's just important to understand that we have absolutely nothing to tell us that meeting that standard leads to a reduction in clinical CSF leaks.

Secondly, to beat the horse that apparently isn't dead --

(Laughter.)

DR. HAINES: -- having no information about the leak and infection rates of the surgeons involved in this study, given that there is no agreed standard in the literature for those rates fails in my opinion to meet any minuscule standard of valid scientific evidence.

And while that approach may be least burdensome to the sponsor, it is most burdensome to the panel and it really is creating a lot more difficulty, I think, for the panel than really needs to be.

And while I'm sympathetic to the sponsors who act on the guidance from the agency, very sympathetic, our responsibility ultimately is to the public and not to the sponsor, and we have to deal with the lack of information that we have.

CHAIRPERSON BECKER: Dr. MacLaughlin.

DR. MacLAUGHLIN: Thank you.

I think I conflict a little bit. I'm not a clinician, but I see the Catch-22 that we're up against with respect to what a good design is and what could be done.

The only thing that occurs to me is while it would be very interesting to see that other information, what are the other surgeons doing? What is the leak rate? What's the infection rate in these other nonapproved things, nonapproved approaches?

It would kind of give some special weight to those, I think, in our deliberation when it probably doesn't deserve it.

I understand the scratching for the information, but I'm not sure how much weight you could put on it at the end of the day because you don't have -- that's not controlled, you know. That work is not controlled. They're doing what they feel is right, but there isn't a protocol that they're following. You don't have the same trail of information that you would have in a well controlled study.

So that's my problem, is I see or recognize what really had to be done and with advice and so on from the FDA, but I'm not sure how much weight I would have paid. Again, I'm a little bit outside the loop. I'm more interested in the material side of things, but how much that would influence me because it isn't controlled either.

So it's out there. You know, like you say, it's this giant, you know, elephant in the room. You want to see what everybody else is doing, and what's the infection rate in any institution taking care of patients like this? You know, some sort of denominator you can compare it to.

The trouble is what in the heck is the right denominator. That's what I'm struggling with.

Thank you.

DR. JENSEN: Dr. Jayam-Trouth.

DR. JAYAM-TROUTH: I think many have voiced the problems of a clinician. Dr. Canady just put it all together there, and I'm struggling with the same thing. I have no question that this is, you know, a good product. It's safe; it's efficacious. It stops all leaks. I mean it looks like the product is good, but I'm not going to use it.

Why am I going to use it? Is this any better than anything else on the market? I mean, do I have to use it?

You know, if you have to do a Valsalva and show me a leak and even in spite of showing with the Valsalva there's no leak, I'm getting a six out of 111 leak; that means something went wrong somewhere.

You know, and is it really necessity? And we don't have that background there, and we don't have that information, and that is really what is the problem that I see.

CHAIRPERSON BECKER: Dr. Germano.

DR. GERMANO: What I find fascinating about the study is that there is an incidence of leak in the hands of reputable neurosurgeons of 100 percent, and that has never been reported before.

(Laughter.)

DR. GERMANO: So the question is: can this data be reproduced either or the other question is does an interoperative CSF leak really result in a clinical leak, and although some derogatory comments were made about an ethnicity, there are reputable neurosurgeons in this country that were born and raised in the United States and trained in the United States, and some of those are panel members of this panel that do no close the dura, and their data is available at the hospital or state Q&A showing that the incidence of CSF leak is virtually close to zero.

And so the question really is: do we really need to close the door?

Now, with that said, each of us, including myself, struggled with complications from CSF leaks. So I don't want to try to seem cavalier about the fact that CSF leaks can and do pose a very challenging management in patients.

But usually those patients are patients that are operated in the posterior fossa, that have had multiple surgeries, that had radiation and chemotherapy, and trauma. In none of those categories, except for the 19 patients of the acoustics, none of those other categories are represented in the study.

So I don't think we have enough answers to the clinically relevant questions.

CHAIRPERSON BECKER: Dr. Witten, do you have anything to say?

DR. WITTEN: No.

CHAIRPERSON BECKER: So I think at this point we can focus on the FDA questions. What I'll do is I'll go ahead and read each question, and after the question we'll allow the panel members to make a statement regarding that question, and we'll provide a summation to Dr. Witten.

So while they're setting up the questions, let's just start. The first question has to do with infection.

The safety evaluation included adverse events collected to three months after surgery. The overall rate of surgical wound infection in the DuraSeal clinical study was nine out of 111 patients, or 8.1 percent, with a 7.2 percent rate of deep surgical infection, all requiring repeat surgery.

Please discuss whether this infection rate raises concern.

And, Dr. Germano, we'll start with you and go around the table in the opposite direction.

DR. GERMANO: Yes, it does raise a concern. I think that the infection rate is high. Again, if you pull the articles from the literature, you can go from an infection rate of zero to 20 percent. So it's hard really to find a denominator.

And I think that what Dr. Haines pointed out should be stressed, and that is that if we had a denominator for each of the surgeons that participated in this study, it would be a little bit easier to understand if for their practice this is falling within the norm or out of the norm.

CHAIRPERSON BECKER: Dr. Jayam-Trouth.

DR. JAYAM-TROUTH: I think, again, there is no good comparison. So it kind of stands by itself, but if these were selected patients and these were clean patients and they were done under the best cares selectively, electively, I mean, I would have then expected that the infection rate should have been lower.

CHAIRPERSON BECKER: Dr. MacLaughlin.

DR. MacLAUGHLIN: I think as a biochemist I'll have to abstain from commenting on how bad these infections are.

CHAIRPERSON BECKER: Dr. Haines.

DR. HAINES: Well, I think this is the biggest concern on the safety issue. I agree. I think the infection rate is higher than I would expect from this group of surgeons for this group of patients.

I think that the attempt to find a comparison in the literature is completely unconvincing, and not useful. I will even quote myself. In the published review comments on the Narotam article where it says, "The absolute infection rates reported by the authors, therefore, should not be used as a standard for comparison unless their liberal definition of wound infection is also used."

So I just don't think that we have any valid comparison and, therefore, we have to fall back on our assessment that this looks like a high infection rate and that this probably is an issue for this device.

CHAIRPERSON BECKER: I would have to concur, and I think the infection rate is high, and these weren't infections that were simply treated with antibiotics. These patients all had re-surgery. So I think that is a big issue, and again, without an adequate comparison group, it's just hard to know what to do with it.

DR. CANADY: I concur.

CHAIRPERSON BECKER: Dr. Jensen?

DR. JENSEN: I concur.

CHAIRPERSON BECKER: Dr. Ellenberg?

DR. ELLENBERG: I concur.

DR. EGNOR: I just raise one issue. These are not clean cases or many of them were not clean cases. I thought that Dr. van Loveren did a nice job in showing that when one looks at clean-contaminated cases in neurosurgery that these infection rates are not high. When one considers the breakdown of the kinds of cases, most of the clean-contaminated cases came from the duration of the operation. Some of these were major procedures.

So I do agree. Looking at the rates, it seems awfully high for craniotomies, but for major procedures I do think there is some literature support for the notion that these are infection rates that are consistent with that.

CHAIRPERSON BECKER: Dr. Loftus, before you make your point, Dr. Haines would like to respond.

DR. HAINES: Well, let's remember that in the study protocol, the CDC definition of wound classification was proposed, and then when the results were called into question a different classification was sought, and that, number one, is a real violation of any kind of sensible study design and analysis.

Secondly, the classification by Narotam has not been validated by anybody else. It was first published in a nursing journal. The data supporting the inclusion of cases lasting longer than two hours as clean-contaminated is not solid, and it is not an established part of our understanding of surgical wound infection.

So it's one paper. It was found post hoc, and I don't think that it's a valid way of looking at it.

CHAIRPERSON BECKER: Dr. Loftus.

DR. LOFTUS: Yeah, it's a thorny issue. I mean, in and of itself we could certainly justify either position on this panel among the surgeons in this room that based on the co-morbidities, this is an acceptable infection rate or it's not.

But I don't think that's the fundamental question. To me the fundamental question is: is there a linkage between the use of the product and the infection rate as stated?

And unfortunately, to my mind we have not been able yet to come up with a credible answer to that question, but this remains, you know, a sort of nagging doubt.

DR. CANADY: The only question was in the initial univariate analysis the amount of material used was associated, although it fell out on the multi-variate.

CHAIRPERSON BECKER: Mr. Balo.

MR. BALO: Yeah, I can't comment on it.

CHAIRPERSON BECKER: So, Dr. Witten, I think that overall the panel has some concerns about the infection rate in this study, but probably the bigger concern is that we have no comparison to know how significant this infection concern really is.

DR. WITTEN: Thank you.

CHAIRPERSON BECKER: The second question has to do with postoperative CSF leaks. The primary efficacy endpoint of the study was the number of patients with continued CSF leak interop. after DuraSeal application. The study design specified a greater than 80 percent study success criteria. The sponsor achieved a success rate of 98.2 percent.

The purpose of establishing a watertight closure of the dura is to limit the postoperative CSF leak rate and associated morbidity. There are five cases, 4.5 percent of the population of the protocol defined postoperative CSF leaks observed in this study. Three patients had a pseudomeningocele seal and the two other had incisional CSF leaks. There's one additional case of a CSF leak during reoperation for a deep wound infection. Including this event, the rate is six out of 111 patients, or 5.4 percent.

Please discuss the observed postoperative CSF leak rate.

And we're going to start with Mr. Balo's end of the table on this question.

MR. BALO: Again, like Dr. MacLaughlin, not being a physician, I'm not going to comment on this.

CHAIRPERSON BECKER: Dr. Loftus.

DR. LOFTUS: I'll just keep this brief. I believe that this product is effective in obliterating CSF leaks at the time of surgery in the fashion that was described, considering somewhat of the artificial nature of the testing parameters.

The linkages between that and the clinical CSF leak related problems, as Dr. Germano has iterated, is a little bit more unclear.

CHAIRPERSON BECKER: Dr. Egnor.

DR. EGNOR: I think it's generally accepted by neurosurgeons that the absence of a CSF leak interoperatively by whatever technique one uses to achieve that is the goal. In that sense DuraSeal helps achieve that goal, and it makes that particular operation equivalent from that respect to a good mechanical closure of the dura.

The question then is what influence does that have ultimately on how the patient does regarding clinically significant CSF leaks. We don't know that answer at all, but at least intraoperatively the DuraSeal seems to accomplish what we all try to accomplish surgically, which is to not see CSF when you do a Valsalva.

CHAIRPERSON BECKER: Dr. Ellenberg.

DR. ELLENBERG: I think the response to this question from a non-clinician will come back to the issue of compared to what.

In addition, as was raised during the discussion in the morning, the endpoint defined could have been focused on CSF leaks. It could have been focused on long-term infection rates, but instead it was focused on whether the leak was stopped.

So this question is going on to perhaps an unfair level in the sense that you're going beyond what the sponsor was tasked with doing and what they proposed to do, which was simply to measure whether or not this device stopped the leaks.

CHAIRPERSON BECKER: Dr. Jensen.

DR. JENSEN: I think the sponsor has shown that it stops intraoperative leaks. I agree with Dr. Ellenberg. I'm not sure what to compare it to in terms of preventing further leaks. I would say that I don't know if the last case should actually be included because it sounded like that at surgery the dura was not leaking until after the material was removed. So I think that one is kind of not fair to include.

CHAIRPERSON BECKER: Dr. Canady.

DR. CANADY: I would agree that it clearly stops interoperative leaks. The significance of that, however, is unclear.

CHAIRPERSON BECKER: Yeah, I would fully agree. I mean, there's no question that it seems to stop interoperative leaks, but the question is what does that really mean for the patient.

Dr. Haines.

DR. HAINES: Yeah, as a tool, for a neurosurgeon who wants to stop a leak, at the time of surgery it seems to be incredibly effective, and that's a tool, I think, most neurosurgeons would like to have at their disposal.

CHAIRPERSON BECKER: Dr. MacLaughlin.

DR. MacLAUGHLIN: Yes. I agree completely that it's a great way of stopping leaks in the operating room, but I really can't comment on -- and I agree, too, with Dr. Ellenberg. That was one of the goals of their study, and I can't comment on the clinical ramifications.

CHAIRPERSON BECKER: Dr. Jayam-Trouth.

DR. JAYAM-TROUTH: Concur.

CHAIRPERSON BECKER: And Dr. Germano.

DR. GERMANO: No additional comments.

CHAIRPERSON BECKER: So, Dr. Witten, I think that with regards to CSF leaks, I think the panel in general agrees that this product is very effective at stopping interoperative leaks. The bigger question remains as to what that means long term for the patient and whether or not perhaps the way they look for leaks actually has been used in other studies. So is this product really better than other products that have been used?

The third question: to be included for treatment, patients were assessed for CSF leaks after sutured dural closure. If CSF was observed leaking from the standard incision, either spontaneously or during an induced Valsalva maneuver to 20 centimeters of water, the patient was included for treatment with DuraSeal. The selection process was intended to include a subset of patients at risk for postoperative CSF leak. However, all of the patients tested leaked.

The proposed instructions for use for all patients was sutured dural closure. So the first part of this question is: do you believe that the results of this study support an adequate risk-benefit ratio in spontaneous leakers?

The second part is: do you believe the results of this study support an adequate risk-benefit ratio in all patients with sutured dural closure as described in the proposed indication for this study?

So I'll allow you to make comments regarding both parts of the question, and we'll start with Dr. Germano.

DR. GERMANO: I don't think we have the answer for Question 3(a) because all patients were include -- sorry -- because all 111 patients were included. So we don't know whether or not this product is good for spontaneous CSF leak because the sponsor did not test for this hypothesis.

CHAIRPERSON BECKER: If I can point out that 60 percent of their patients had a spontaneous leak at the time of surgery; isn't that right?

PARTICIPANT: That's correct.

DR. CANADY: No, 40 percent spontaneous.

CHAIRPERSON BECKER: Forty percent and 60 percent Valsalva, right.

Sixty percent spontaneous, 40 percent Valsalva.

DR. GERMANO: So I would rather comment there was a percentage of patients that had a spontaneous CSF leak, and so then what we have to do, and I don't remember the data, is to go back and see how many of those that leaked spontaneously had a CSF leak after the surgery.

DR. CANADY: There wasn't good overlap. I think I looked at that. There wasn't an overlap between the patients.

DR. GERMANO: So on page 9 of the document it shows that spontaneous leaks intraoperatively was 5.9 percent and leak induced by Valsalva was 4.5. So that does not seem to be statistically different.

CHAIRPERSON BECKER: Correct.

DR. GERMANO: But then again the question here is risk-benefits. So I think I stated previously that there are some concerns about the product, and there are some concerns about the benefits.

I guess my answer is that there are concerned.

Question 3(b) is the same for all patients, and so then I guess the label if this product were to be approved, the label should say that it is approved for all patients after the dura has been closed because the authors show that there is leakage in 100 percent of the patients.

So I still have an issue with this.

CHAIRPERSON BECKER: Dr. Jayam-Trouth.

DR. JAYAM-TROUTH: I guess the answer to Question 3(a) is yes. I mean, it seems as if it does work, you know. It does stop the leaks and spontaneous CSF leakage. Now, what I am not really sure about is, you k now, do you really need to do the Valsalva maneuver to show that the Valsalva increasing the pressure to 20 centimeters produces a leak or induces a leak?

You know, which means then that if those people do not do the maneuver, you know, then in all cases they have to then opt to put in this as a dural sealant, and that is where I have a problem because, I mean, most surgeons don't do the Valsalva. I mean, correct me if I'm wrong. Many surgeons probably don't do the Valsalva every time there's a general closure that they're doing.

If they see a spontaneous leak, you put the sealant; it works. But for 3(b), I mean, should you put it in every case? Now, I don't think so.

CHAIRPERSON BECKER: Dr. MacLaughlin.

DR. MacLAUGHLIN: Yeah, again, I'm in a little bit of this gray area for me. I think it shows clearly that you get the closure, and there seems to be no significant difference at least for 3(a), I guess, between the group. So I agree that, you know, it has some early effect and early benefit, but again, for the same reason of not being a clinician, I want to comment on my thoughts on the other areas.

CHAIRPERSON BECKER: Dr. Haines.

DR. HAINES: I don't see a way to distinguish between (a) and (b), and I think that all of the next three questions address the risk-benefit ratio. I'm not sure when to answer that.

CHAIRPERSON BECKER: I agree. Go ahead.

DR. HAINES: Well, I think that effectiveness for the limited purpose that's stated has been well demonstrated that three are no major safety issues, that there is a significant concern about the possibility of a clinically important but relatively small increase in deep wound infection rates, and that with some -- and I'll put my card on the table -- with some adjustment in labeling, that the risk-benefit ratio is achieved.

CHAIRPERSON BECKER: I guess for my part I'm not exactly sure what the benefit is. I think it's effective for closing the dura. I don't know what benefit that has led to in this study. We certainly know what the risks are.

Dr. Canady.

DR. CANADY: I think it turns on the risk-benefit ratio. Clearly, it's effective in the short term for both groups.

CHAIRPERSON BECKER: Could you use the microphone, please?

DR. CANADY: Clearly, it's effective for both groups in the early stages, and the question becomes what is the relative, if any, change in infection rate and whether or not the patients in whom -- which of the patients and what is the criteria by which patients go on to clinical leaking, and those are questions for which we don't have answers.

Do I think it's terribly unsafe? No.

CHAIRPERSON BECKER: Dr. Jensen.

DR. JENSEN: Again, I think that you've demonstrated that the material closes interoperative leaks, and again the question comes down to the risk-benefit ratio.

In terms of benefit of stopping postoperative pseudomeningoceles and leaks, I mean, you still have them and they're still five to six percent, which is what we see in other studies using other materials. So it's hard without, again, a control group to figure out whether or not you're improving things there.

And I still have a problem with the potential infection rate.

CHAIRPERSON BECKER: Dr. Ellenberg.

DR. ELLENBERG: On a technical point, we have not seen, I believe, the confidence intervals for the CSF leak rate stratified by the spontaneous versus Valsalva maneuver. So I'm not sure we can comment on whether we feel comfortable with the simple proportions given on the slide on the lower left of page 9.

So I would be very cautious about splitting these questions into (a) and (b) at this point without further data, and with regard to the risk-benefit ratio, I agree with all of the comments that have been made to date that we have to look at this with a safety profile that we understand compared to what else.

CHAIRPERSON BECKER: Dr. Egnor.

DR. EGNOR: Regarding the risk-benefit ratio, we certainly can't say anything coherent about the risk-benefit ratio as regards the ultimate outcome of clinically significant CSF leaks. It seems that the evidence interoperative CSF leaks are preventive by DuraSeal is quite strong, and that risk-benefit ratio we can say something about in the sense that as neurosurgeons, we typically will spend whatever time is necessary, particularly in the posterior fossa, to get an anatomically watertight dural closure even if it adds quite a bit of time to the operative procedure.

I sometimes will spend an hour trying to make sure the dura is really watertight. So clearly, in the everyday surgical decision making of risk-benefit, I'm willing to risk the extra hour of anesthesia time for the benefit of a watertight dural closure. Therefore, it would seem that the risk-benefit ratio of achieving that watertight closure using DuraSeal is sensible. That does seem to make sense.

The long-term risk-benefit ratio we don't have a clue about. The science is woefully inadequate there.d The infection stuff is of concern. I wonder if we could request that that be studied in time.

CHAIRPERSON BECKER: Dr. Loftus.

DR. LOFTUS: You know, it's a puzzle, and so my answer will be cryptic, but let me give you the positive and the negative.

I mean, the positive is, as I see it, -- I mean, we shouldn't deny this -- this is a common surgical practice. Now, supratentorial patients, at least in my practice, don't customarily undergo the Valsalva maneuver.

For those of you who are not surgeons, you should understand that in the spinal dura and in the posterior fossa dura, this is a common paradigm, and I think I would say that most surgeons do that.

Therefore, that being said, it would be nice and ideal, and the public would be well served, to have an on label product that was FDA vetted, validated and approved, to subserve this function.

The negative is these are serious infections in these patients, and basically ten percent of these patients, serious, morbid infections, and if there is a linkage -- and I don't know whether there is or not -- it is troublesome.

CHAIRPERSON BECKER: Mr. Balo.

MR. BALO: Yeah, I really can't comment about the infection rate, but you know, just like Dr. Loftus says, I don't know if it's basically correlated with a DuraSeal or not, but I do think that the company has demonstrated DuraSeal does seal and provides a watertight when it is applied, and also I think a benefit of it would be to reduce surgical time and anesthesia time to provide that watertight seal.

CHAIRPERSON BECKER: Okay. So in summary for Question 3, Dr. Witten, it seems that the panel doesn't believe that you can artificially separate out the spontaneous leakers from the Valsalva induced leakers, and that while the DuraSeal is very effective for closing the dura, the risk-benefit ratio, I think, still is a question that cannot be adequately answered based on the data we have.

So Question 4, and as Dr. Haines already alluded to, the next couple of questions are really on the same theme, but Question 4 states that 21 CFR 860.7(d)(1) states that there's a reasonable assurance that a device is safe when it can be determined that the probable benefits to health from use of the device for its intended uses, when accompanied by adequate instructions for use and warnings against unsafe use outweigh any probable risks. Please discuss whether the data in the PMA provided to us today, provided reasonable assurance of safety.

And, Mr. Balo, we'll start with your end.

MR. BALO: Well, just starting where Dr. Loftus left off with the previous question, you sort of left that open in here relative to we have no way to correlate if the dura is still material, basically correlate some of the infections that were produced. So for my end of it, I still think that the device does as it's intended to do, and from my little experience, I think it would be safe.

CHAIRPERSON BECKER: Dr. Loftus.

DR. LOFTUS: Yeah, it's really hard for me to know the answer to this, but I'm going to take a stab at it, and once again, it's based on pragmatic information from my own practice. That is, I use all the time off-label material to subserve this exact function for which I don't have data whether or not infections are present or not.

And so I would say my answer to this question is since at least this product has been scrutinized in a more rigorous way, although the answers are imperfect, I have to say that the safety profile is at least commensurate with what I'm doing in my off-label uses of devices in my daily practice.

CHAIRPERSON BECKER: Dr. Egnor.

DR. EGNOR: Yeah, I agree with Dr. Loftus. This product has been studied, I believe, more carefully certainly with regard to infection and so on than the stuff I use every day, and it would seem to me that the risk-benefit ratio would be acceptable in light of that.

CHAIRPERSON BECKER: Dr. Ellenberg.

DR. ELLENBERG: The nature of Question 4 is, of course, the question that's asked of this panel at every meeting. Well, we need to look at the standards for both efficacy and risk.

In terms of the standards for efficacy for the primary endpoint as stated in the application, I think that has been adequately proven, and in terms of the measurement of risks, my sense is that we do not have an adequate comparison group even in the protocol defined endpoint of three months, let alone long-term follow-up.

So I find it difficult to make an objective statement just based on data presented to this panel that the benefits outweigh the risks.

CHAIRPERSON BECKER: Dr. Jensen.

DR. JENSEN: Well, unfortunately, not being a surgeon, or maybe fortunately -- I'm not sure ‑-

DR. CANADY: We would consider it unfortunate.

DR. JENSEN: -- I don't have a practice to base infection rate upon. All I really have is what has been presented in the package, and again, I believe that the material can be used safely. I still have an issue with whether or not the infection rate is substantially greater than a group that you see, Dr. Loftus, you know.

So the infection rate is not substantially higher than what's been reported in the literature, but again, like Dr. Ellenberg said, there's no good control.

So I personally have difficulty since I don't have the clinical experience that you have in saying definitely it's safe.

CHAIRPERSON BECKER: Dr. Canady.

DR. CANADY: I think it turns on your perception of the material, and I guess if I classify this material as an implant like my shunts and say that, given that it's an implant, there's an additional infection risk associated with the implant, which is essentially what's been done in the comparison to DuraGen, DuraSis, Bio; then you can say, yes, this material is safe, but it's an implant, and with an implant you have an additional risk of infection every time you implant something, and that has to be factored into your decision to use this product.

I think that then it sounds reasonable to me. I think that's not what's going to clinically happen. I mean, people aren't going to read it and make that kind of decision, but I think that given the information that we have today of an infection rate of 11 percent, which is comparable to other implants, that I would be comfortable with that decision tree, with the knowledge that this is an implant and not just a material that's there.

CHAIRPERSON BECKER: I have to say that I think the infection rate isn't out of line with the other infection rates reported in the literature, although they're imperfect comparisons. I guess still the issue I have is it's not clear to me what the benefits to the patient are long term compared to doing nothing or doing nothing different than has been done.

Dr. Haines.

DR. HAINES: Well, I'd have to say after Dr. Loftus and Dr. Egnor's comments, since I don't routinely use off-label stuff to reinforce my dural incisions --

(Laughter.)

DR. HAINES: -- and think that I have the same CSF leak rate that they do, I'm actually a little more concerned that if there is actually an increased serious infection rate associated with the use of this material and it is that widely used, that we might actually have a concern.

I mean if this doubles the infection rate, the deep wound infection rate, and it's used in 70 or 80 percent of the craniotomies done in this country, and we approve it, we have done a bad thing.

CHAIRPERSON BECKER: Dr. Loftus?

DR. LOFTUS: Dr. Egnor was first.

DR. EGNOR: Go ahead.

DR. LOFTUS: You know, Alexa brought up a very interesting point, but I disagree. I mean, this is no more of an implant than a dissolving suture is an implant. I mean, this is a temporary expedient meant to disappear, and I think we do need to --

DR. CANADY: But does it hold bacteria in place?

DR. LOFTUS: -- evaluate it in that term.

DR. CANADY: Does it hold bacteria in place the time that it's there? I mean, I don't think we know the answer to that, but I think, you know, that that's a very real possibility.

DR. EGNOR: To comment on Steve's point, there's, I think, quite a difference in the extent to which neurosurgeons place materials with a specific intent of preventing CSF leaks. I think that practically every craniotomy wound that's closed is closed at least with gelfoam in the upper dural space, certainly with some suture material, deep and superficial sutures, and often with plates and things.

So while there may not be routine for everyone placement of material to prevent CSF leaks, there's a lot of material that everybody puts in every craniotomy.

CHAIRPERSON BECKER: Dr. MacLaughlin.

DR. MacLAUGHLIN: These discussions actually point out the trouble of designing a study like this, you know, with the varied approaches by different institutions and different world class surgeons.

I think I'd prefer to answer this question with a time line. It seems to me hearing the surgeons talk about how -- and I know actually from my own experience it's better, you know, to be put to sleep for less time than more time -- if you have a procedure and there's some interoperative benefit, (a) no leak, (b) let's say shorter time under anesthesia, that feels like a real benefit to me.

The longer term issues, you know, infection rate, that as I say I can't comment on because I don't have any personal experience -- I'm really a noncombatant in this issue -- but I don't think we have enough data to support that. So I compartmentalize this thing saying, yeah, it stops leaks. that's a good thing in the operating room. It may be a good thing overall. It may not be any significantly different, let's say, than the other devices that are used. Maybe it's easier to use. Who knows?

But we don't have that, you know, to look at. So I guess overall I think it does have a benefit, and I just can't assess, you know, how real the risk is. It doesn't seem worse than other studies, but I know that those are flawed.

CHAIRPERSON BECKER: Dr. Jayam-Trouth.

DR. JAYAM-TROUTH: Kind of listening to my neurosurgical colleagues again, and from what I can see then, you know, the problem or the time, the area where it might be most applicable will be in those that are prolonged surgery, those that are in the posterior fossa, maybe those in the spine which they are not asking for approval, you know. So perhaps we may have to look at it and say in those situations where it's complicated and where you need excess time and anesthesia and stuff, that is where perhaps, you know, we take that additional risk and use the material.

But in those that are regular, normal craniotomies with clean wounds, you know, I don't see, you know where the potential or where the extra benefit is in using the material.

CHAIRPERSON BECKER: Dr. Germano.

DR. GERMANO: I don't have enough data presented to be able to deliberate on the safety of this product.

CHAIRPERSON BECKER: So, Dr. Witten, I think there's a lot of mixed feelings and controversy with regards to Question 4. I think that the panel doesn't really seem to have a consensus on whether this product is both safe and effective, whether the benefits outweigh the risks. I think that's unclear.

DR. WITTEN: Thank you.

CHAIRPERSON BECKER: So Question 5, 21 CFR 860.7(e)(1) states that there's reasonable assurance that a device is effective when it can be determined, based upon valid scientific evidence, that in a significant portion of the target population use of the device for its intended uses and conditions of use, when accompanied by adequate directions for use and warnings against unsafe use, will provide clinically significant results.

Please discuss whether the data in the PMA provide a reasonable assurance of effectiveness.

And we'll start with Dr. Germano and come around the other way.

DR. GERMANO: So, again, the question is does the interoperative CSF leak result in clinical leak, and the results presented on page 40 of the company presentation, for the infratentorial craniectomy, 19 patents, and acoustic neuroma, six patients, with a total of five percent leak are very, very promising. I would like to see those numbers with a zero after each and then we consider the product.

CHAIRPERSON BECKER: Dr. Jayam-Trouth.

DR. JAYAM-TROUTH: Well, as worded, you know --

DR. GERMANO: I'm sorry. Zero meaning to go through 190 and 60 for the denominator, not zero incidence of CSF leaks. So have the same study done with 190 infratentorial craniectomy and at least 60 acoustic neuroma with the CSF leak remaining five percent and then we consider the product.

CHAIRPERSON BECKER: Okay.

DR. JAYAM-TROUTH: I guess when you look at the wording of the question and it says, you know, when the device is used for its intended use and conditions of use, you know, is it safe, yes, it's safe. It does seal. It does do a job, but then is there enough valid evidence that we need to use the device? That is where I have my problems, and I guess that, you know, depending on how you answer this question, I'd say that, yes, it is safe to use and it does produce a good sealant.

But is there enough scientific evidence that it needs to be used? And I think the answer to that, I'm not convinced that it is.

CHAIRPERSON BECKER: Dr. MacLaughlin.

DR. MacLAUGHLIN: Yeah, I think I agree with a lot of what you just said. It definitely closes, and in that interoperative space, it certainly seems safe.

I think whether one uses it or not is up to the surgeon, among other choices. So I'm not tying so much weight on that aspect of things. I just think, you know, if it's out there, it's out there.

CHAIRPERSON BECKER: Dr. Haines.

DR. HAINES: For the limited purposes for which the product is evaluated, it is effective.

CHAIRPERSON BECKER: For me, I think the important part of this question has to do with whether or not the product will provide clinically significant results, and to me that's not clear.

Dr. Canady.

DR. CANADY: I concur.

CHAIRPERSON BECKER: Dr. Jensen.

DR. JENSEN: It seems to do no worse in terms of clinical outcomes. So it's clearly effective interoperatively. It's not worse than --

CHAIRPERSON BECKER: Doing nothing.

DR. JENSEN: -- than doing nothing, but I will say that one of the things that is appealing is the fact that it does appear to markedly diminish the amount of interoperative time, which I think is a benefit.

CHAIRPERSON BECKER: Dr. Ellenberg.

DR. ELLENBERG: Let's see. I think I may be mincing words here, but in terms of how Dr. Haines responded, it seems to me that the effectiveness has been shown as defined explicitly by the submission. Introducing the clinically significant phrase is an interesting turn of words at this late stage.

From what we've heard from the surgeons this afternoon, as was just said, just cutting the anesthesia short by an hour seems to me as a layperson to be a clinically significant result.

But that's not what we've been tasked to assess in our review of the materials or in the discussion today. So I think it meets its limited standard as submitted.

However, the one comment that I would like to make now is if this device were to be approved, it would be approved based on perhaps a high risk group of subjects, but that high risk group of subjects is by no means, as was mentioned throughout the day, the totality of subjects for which this device would be used.

So while I have limited belief in a restriction in labeling for certain types of cases being an effective stopping of a surgeon using this off label, if this were to be approved, I think we have to realize that it would be used based on no patients that were not covered, and if it were approved, if it were to be considered for approval, I think as a condition for approval even though it might not be efficacious one might consider limiting its use to the patients studied in this cohort or some definition like that rather than essentially saying its use for stopping leaks is the intended use.

CHAIRPERSON BECKER: Dr. Egnor.

DR. EGNOR: I certainly think that were it to be approved, that it could only be approved for the limited indications in this clinical trial. There's absolutely no evidence that it's of any value in any other indication, except these patients and for the purpose of intraoperatively stopping CSF leaks.

However, we do exert a great deal of effort during surgery to accomplish that, and by diminishing anesthesia time, one could certainly say that it would seem that the benefits might outweigh the risks.

DR. GERMANO: We have not seen any data showing that there was a decreased anesthesia time in any of the material that was submitted to us.

DR. EGNOR: Right, right.

DR. GERMANO: So we are basically now all speculating on one sentence --

DR. EGNOR: Absolutely.

DR. GERMANO: -- that Dr. Loftus interjected ten minutes ago, a beautiful sentence.

DR. EGNOR: Yes, but I get the sense that what we're doing here is our focus isn't really to evaluate the science. The science here is woefully inadequate. It's a terrible study.

What we're evaluating is the product, and if we demand utterly perfect science, I don't know that any product will come to market. So the --

DR. GERMANO: No, but now if you're saying that what is striking to you is the decreased operative time, then I would like to see that data, and if the data is not available, it has to be produced.

DR. EGNOR: Well, it's not striking, but it's sort of intuitive.

DR. HAINES: But no. I mean, the surgeons were instructed to do everything they normally could do to reconstruct the dura first and then take another five minutes to apply the stuff.

So, in fact, although it's not significant, the time has increased.

DR. EGNOR: Well, we don't know in fact. I mean, we don't know in fact that it was increased.

DR. HAINES: Well, it can't be decreased because they had to do everything they had to do before applying the sealant.

DR. EGNOR: I mean, in the real world one could certainly imagine spending more time if you don't have some adjuvant to help you.

DR. HAINES: Well, again, there's another danger that we're looking at the possibility that this will be -- the overwhelming likelihood that this will be used in place of closing the dura.

DR. EGNOR: Right, right, and that's another perfectly valid concern, a very valid concern.

CHAIRPERSON BECKER: Dr. Loftus.

DR. LOFTUS: You know, this is sufficiently muddy and sufficiently gray that I want to focus exactly on the question. So what do I know with reasonable assurance?

I know basically one thing, and that is that in the study that this stuff, this product, will close the dura very effectively within the population which was studied, which consists of the patients who are by far the easiest to close anyway, but that all of the difficult cases where dural closure is too problematic were eliminated from consideration.

And that's okay, but that's what I know and anything beyond that is an extrapolation, and that may be acceptable, but that's what it is.

CHAIRPERSON BECKER: Mr. Balo.

MR. BALO: It's kind of hard to follow all of these comments since I don't practice medicine, but I do think from the data that was presented, from what I read, the limited amount, I agree with what the panel has said, that the device does seal when it is applied, you know, but relative to infection, relative to the safety question we discussed, I think that just has to be discussed among the clinicians.

CHAIRPERSON BECKER: So with regards to efficacy, Dr. Witten, I think that everybody is in agreement that this device works interoperatively to close the dura. I think that the data is not really adequate to judge it against anything else, to know whether it's of clinical significance in the long term regarding CSF leaks.

So then the final question, Question 6, which is really kind of the meat of the matter, is that reasonable assurance of safety and effectiveness, as defined in Questions 5 and 6, must be demonstrated for device approval. If you believe this has been demonstrated but think there are specific focus questions regarding this device that still remain and can be addressed in a post approval study, please identify those questions.

So, Dr. Germano, we'll start with you.

DR. GERMANO: Well, my conclusion with the data that I have today is that reasonable assurance of safety and effectiveness is not demonstrated.

CHAIRPERSON BECKER: And do you think there's anything the sponsor could do to address it in a post approval study?

DR. GERMANO: Yes. As I said, I would like to see those cases that have a tendency for CSF leak that is of clinical significance, and that is to prolong the length of stay and/or result in additional surgery.

So I would like for that patient population to be expanded, and I used before the number 190 instead of 19 and 60 instead of six, and I would definitely defer the correct n to Dr. Ellenberg because I don't know if what I stated would be statistically meaningful. It might be that it's less or more. So I would defer that to Dr. Ellenberg.

DR. ELLENBERG: One hundred and thirty-two.

DR. GERMANO: Thank you.

(Laughter.)

DR. GERMANO: In addition to that, I would like to have, as the panel already recommended, some information on the infection rate of the surgeons that participated in the study and see whether or not the infection rate that we have here is comparable or not to what the standard of those surgeons are.

And then I think the panel already addressed other concerns with the possibility of having a controlled arm and so on and so forth.

CHAIRPERSON BECKER: Dr. Jayam-Trouth.

DR. JAYAM-TROUTH: I concur with Dr. Germano, and I want to add one more thing. I think where the device really needs to be studied is in the complicated cases. You know, in those cases that are difficult, in those cases that are posterior fossa, in those cases that are three millimeters close to the suture lines, in those cases where there's jagged, you know, kind of a dural tear. You know, I think that is where they really need to try and see.

If you're going to put an implant without strong evidence that it does any better, you know, we cannot say that just because it is just as bad as everything else it is better. You know, so, therefore, to me where it would be really effective or where it could be better is if it were shown that in the complicated case it makes a difference.

CHAIRPERSON BECKER: Dr. MacLaughlin.

DR. MacLAUGHLIN: I agree with the first comments of my previous two panel members here, and I feel like I can't really well define a post approval study that would shed the most light on the problem that you surgeons are seeing.

Thank you.

CHAIRPERSON BECKER: Dr. Haines.

DR. HAINES: I believe that for the specific proposed indication that effectiveness has been demonstrated, that the overall safety of the product has been demonstrated, and that if an effective post approval study of the actual clinical CSF rates and infection rates could be done, that that would be a very adequate solution to our dilemma.

CHAIRPERSON BECKER: I agree. I think if this device were to be approved that we'd definitely need some way of following patients to look for infection rates as well as clinically significant CSF leak rates so that we know whether or not in the long run this device is effective or whether it's actually safe.

DR. CANADY: The big question here is really whether closure of the dura watertight is useful or not, and that's really not an appropriate question for the sponsor to answer, but maybe Steve, sine you like those kind of studies.

It's clearly, just in the conversation here, something that needs to be established because the expense that's going to come with this kind of a thing being approved is significant because it's going to be used, you know. Neurosurgeons are still belt and suspender people, and even if you think you have a good closure, people are going to throw it out.

I think it is effective in sealing the dura in the short term, clearly.

The third point would be that I could be comfortable with this device labeling it as an implant with the possible risk of increased infection with any added device.

And I would strongly support additional collection of data regarding infection, although if we just collected data, we're going to have the same problem at the end. You don't have a comparison group.

CHAIRPERSON BECKER: Dr. Jensen.

DR. JENSEN: I agree that for the interoperative use it's safe and effective. I still struggle with the clinical follow-up.

I have a question for Dr. Ellenberg.

Is there a way to get some statistically significant data by retrospectively reviewing the 23 patients that were excluded based upon interoperative criteria if the company were able to do so, to help us get some sort of control group?

DR. ELLENBERG: It seems to me that the 23 excluded patients might represent an extraordinarily heterogeneous cohort of subjects that were excluded for a whole slew of reasons. So I'm not sure that this could add to our knowledge base.

But if the data were available, it could. But I would probably say that there's no selection bias operating in why they were excluded and probably not very beneficial.

I don't find that I can agree with the premise of the start of the second sentence and, therefore, will not comment on the latter part of the second sentence.

CHAIRPERSON BECKER: Dr. Egnor.

DR. EGNOR: I certainly think that the only scenario in which I could vote for approval would be if it were specifically for the cases that were, in fact, studied by the sponsor.

I also believe that post approval studies would be imperative, and the two post approval studies, I think, that would be critically important, first of all, would be to track infections with case controls. You really have to know. If a high infection rate is associated with this product, and that can be clearly shown, then the product shouldn't be used.

But I don't know that we right now have reason to think that, but that should be studied carefully with concurrent controls.

The second thing is it probably ought to be studied with patients who are difficult patients to stop from leaking instead of easy patients to stop from leaking, and that could be of great value to the patients if it, in fact, demonstrated some benefit there.

CHAIRPERSON BECKER: Dr. Loftus.

DR. LOFTUS: I turn this over minute by minute as we sit here and talk, as we all do, but let me tell you what I believe in my heart. I mean, I believe in my heart that this manufacturer has made a credible and sincere effort to work with the FDA to design a trial that would answer these questions, and in many respects, you know, it hasn't worked out.

Nonetheless -- and I would temper those comments only if it became clear, and I don't think it is clear, that the data which I continue to seek on the 23 patients had been withheld or suppressed in some duplicitous way, and I don't think that's the case, but obviously that would change my opinion.

And I think that couched in that framework then, when I consider the alternatives that are available to me, as I said before, which is off-label use of things that haven't been tested and might have an infection rate even higher, even though I don't think in my hands they do, that this is a credible application that has met my standard.

But I do think, as Dr. van Loveren was talking, I sketched out what I thought would be the ideal study, and that is, say what you like, best surgical practice to seal the dura with or without DuraSeal and study the infection rates, and I really think that ultimately that would be very, very useful.

CHAIRPERSON BECKER: And Mr. Balo.

MR. BALO: Again, I do concur from the information I heard today and from what I've read that DuraSeal did show its effectiveness in the population that they had selected and that they used it on.

Safety, again, I've heard a whole spectrum of analysis from the different panel members, and I just heed up to their expertise on that.

CHAIRPERSON BECKER: So with regard to Question 6, I think that the panel believes that if this product were to be approved there are some studies that would need to be done post approval specifically to address the infection rate, as well as to address high risk patients and clinical outcomes in those patients.

So I think at this time we'll take a break. Let's say to three o'clock or five after three. Five after three, and at that point we'll resume and have the rebuttals by the sponsor and have more questions.

Thank you.

(Whereupon, the foregoing matter went off the record at 2:49 p.m. and went back on the record at 3:06 p.m.)

CHAIRPERSON BECKER: It's now five after three, and before we move on Dr. Egnor has asked for the opportunity to clarify a point that was made in the discussions prior to the break.

DR. EGNOR: If I may ask a question of Dr. van Loveren, please.

Dr. van Loveren, Dr. Haines had mentioned earlier in the sessions that the higher infection rate in the longer operations is not necessarily widely accepted as being the normal situation, and you had quoted a study. You had shown us a study in which that assumption was made.

Do you know the basis that the people who wrote that study used to make that assumption that the Class II operations have a higher or that the longer operations have a higher infection rate?

And what is the basis for that assumption? Because the fact that the infection rate is a bit high for our sense of what clean craniotomies would normally have, I'm willing to accept that if there's clear evidence that the longer cases are intrinsically associated with higher infection rates, but Dr. Haines implies that that may not necessarily be the case.

DR. VAN LOVEREN: Well, I think there are two responses. It comes at two levels. One is that the finding that longer cases have a higher infection rate is simply a statistical monitored finding when you analyze and divide cases.

The second component is searching for mechanisms or explanations for why the longer cases have a higher infection rate, and then you get into hypotheses about wound exposure, the progressive vascularity of the wound, the progressive bacteria load on the wound, and the association of longer cases with multiple surgeons, multiple episodes of contamination, and the use of additional equipment.

Some studies have said any operation with a microscope, a plastic drape and a surgeon with his mouth against the drapes and on the handles is a contaminated case. So as the case grows longer, there are multiple reasons for infection rate to increase both to do with the physiology of the patient and the nature of the case and how it's being done.

DR. EGNOR: I certainly understand those considerations, but are there other neurosurgical studies of infection rates for otherwise clean neurosurgical cases that clearly show this increase in infection rate with operative times that are commensurate with what is seen in your study?

DR. VAN LOVEREN: Yes. I mean, the only reason I hesitate is because I don't know if they would meet Dr. Haines' need for statistical relevance at the highest degree, but there are several studies even in the packet that we present where there is stratification of cases with demonstrable increased infection rates.

But as we showed, especially, for instance, in the Narotam study, greater than two hours was a statistical increase in infection rate. The greater than four hours looked like a twofold increase, but was not statistically relevant because of low numbers.

So I don't know with which of each data points would reach statistical relevance.

CHAIRPERSON BECKER: Dr. Haines.

DR. HAINES: May I?

As usual, Dr. van Loveren says it very well, and, yes, in univariate analysis the duration of surgery is associated with increasing infection rates, but it all of the factors that Dr. van Loveren mentions that create the question, and when you do apply good clinical science to the question, you find out that you have a great deal of difficulty blaming the duration of the operation, focusing on that as the cause of the increased infection rate.

So that to take duration of operation and turn that into a reason to classify a case as clean-contaminated is a novel idea that Narotam did, but we can't compare it to the other studies because the other studies don't do that.

DR. VAN LOVEREN: But wouldn't one anticipate that the other factors would also be present in the DuraSeal cases? I mean, one agrees that --

DR. HAINES: One would like to know. One would like to know.

DR. VAN LOVEREN: And yet it's probably true. I mean longer cases have a higher infection rate, and, yes, the duration of surgery may be a surrogate for other physiologic and technical factors, but there does seem to be an association between very long operations and higher infection rates.

CHAIRPERSON BECKER: Thank you.

Okay. So now that the panel has responded to the FDA questions, we'll have the second open public hearing of this meeting. Is there anybody in the audience who would like to address the panel at this point?

(No response.)

CHAIRPERSON BECKER: No. Okay.

So does anybody on the panel have any further questions for the FDA? Would the FDA like to make any further comments or clarifications?

(No response.)

CHAIRPERSON BECKER: So at this point we'll all the sponsor to make further comments and clarifications.

MR. ANKERUD: Thank you, Dr. Becker.

Eric Ankerud from Confluent Surgical.

We do have some closing remarks, and I would like to just briefly comment on labeling as that was a discussion point in the afternoon session here. Dr. van Loveren and Dr. Cosgrove will also make some closing remarks.

As you know, we proposed an indication for use for the DuraSeal system and conducted a pivotal study against which we were measuring performance to that stated indication. Our intent as a company should this product be approved is to commercialize the product and label it in a way that matches the patient population that was studied. We do not have an intent to commercialize this product in any other way.

The study measured interoperative sealing efficacy, and we are seeking an indication that indicates the product for sealing interoperatively for watertight dural closure, and I can assure you that is the intent of our company should this product gain commercial approval.

At this time I'd like to invite Dr. van Loveren to the podium.

Thank you.

DR. VAN LOVEREN: Thank you.

It's not one of my closing remarks, but the Coranet (phonetic) study also looked at infection rates and found significant statistical increase with duration past four hours, but you'd have to look at it statistically.

As for closing remarks, I agree with everyone in the room that this is a very difficult study design and would have benefitted from a reasonable control arm. The problem, of course, is that none existed and now we are contemplating alluding to the redesign of the study to a methodology that long ago in discussions with the FDA panel we were warned against.

So there is, as people have said, an elephant in the room, but I think it goes beyond the absence of a control arm. It goes to the very heart of the absence of an FDA approved substance or device for this purpose, and I do believe that there is a time, there comes a time to move forward and be first, and there is a need for there to be a first product on market approved, first device approved for this purpose so that this burden cannot be placed again on other people coming to trial to be told that there is no suitable FDA approved control, which would thwart studies and thwart innovation.

I think we have shown that this device is reasonable first step to come onto market. It has done what it was supposed to do. It seals CSF leaks. It seals the dura interoperatively.

When you look at the overall study, in 111 patients we had two CSF incisional leaks. This is an incredibly robust number that could have been put up against any device or sealant studied in the literature and would have come out equivalent or better.

That's not to say that it is equivalent or better. That's to say that no matter what you have designed this to be studied against, it would have succeeded.

There is an incredible need for this product. I think the characterization that we looked at the easiest cases is not really correct. The cases that were enrolled in this study, the cases I enrolled, are on the ridiculous end of complicated neurosurgery. These are seven to ten hour cases with 20 centimeter durotomies, almost 50 percent infratentorial, and a third of those craniectomies without replacement of bone. There's no neurosurgeon that would consider that set easy, and there's no neurosurgeon that would think that closing the dura at six in the evening for a case that started at 7:30 isn't a hazardous high risk case.

As a skull base complex cranial neurosurgeon, you know, I am pursuing the study of dural sealants not because I'm interested in the company. I'm interested in my patients stopping leaks, closing dura. This stuff actually works, and I think a number of us will be disappointed if a month from now we find ourselves in the posterior fossa honestly trying to close a dura that won't close and because of this technical absence of control, this product is not able to come off the shelf, and it's a remarkably easy product to use when you compare it to what's available in the market because it's an off-the-shelf product. It's not fibrin glue which has dubious results to begin with, which you have to order ahead of time, which takes 20 to 30 minutes to arrive in the OR, which has some small concern of transfusion effects.

So I think this is an incredible unmet need. I think this is a good first product to set that standard so that second generation studies can begin.

DR. COSGROVE: Dr. Rees Cosgrove.

First of all, I would like to thank all of the panel members for spending so much time and effort in reviewing this information, and it's probably only slightly less than the time we put into getting this product in front of you.

I think that I agree with many thing that the panel members have said, that we clearly in terms of the design of the study, we clearly have satisfied the objective that we set out to satisfy, which was to get watertight dural sealing at the time of closure.

But as Dr. Haines pointed out, we have satisfied a limited objective, and all of the clinicians in the room are saying, "Well, that's great." And I think the neurosurgeons are saying, "That is great. I mean that is an important thing to do because in our gut we say we've got to do this."

And as we're all operating to closing up a posterior fossa, and there would be nothing better than to do it quickly, but we actually get our pericranial tissue. We spend an extra 45 minutes to an hour sewing it in. We check it. We do it again because we know the consequences of a CSF leak, and it's far better to put in an extra 45 minutes to an hour at that point, get it right, than dealing with the complication five, ten, 15 days down the road and reoperating in all of the issues that are involved there.

So as neurosurgeons we have this inherent acceptance in some ways of a watertight sealing. However, we also have this very great discomfort about, well, yeah, but does it have clinical efficacy down the road, and I think everybody, myself included, all physicians on the SAB have the same feeling. Does it really then translate into clinical efficacy?

And I think that's that big question, and this study doesn't answer that. However, in a surrogate way, you can look at some of the numbers and say it's in our comfort zone in terms of it did extremely well on the overt CSF leaks because nearly half of our patients, and these are the ones as pediatric neurosurgeons, you know, doing a lot of posterior fossa work and some of us doing a lot of posterior fossa work, these are the ones that we really take the extra efforts.

I agree with some of the supratentorials. It's less of a problem because there's no dependency there, but certainly these are the ones we take extra care of, and we have nearly half of the patients. So I probably wouldn't agree that this was an easy population to do. There's not many neurosurgeons or not many neurosurgical series that are about half posterior fossa procedures.

So we all have this sort of general discomfort, and I have it, too, because I'd like to know that it really is effective, but I don't think we can say that on the basis of this study, although I do think that it's in, again, our comfort zone for certainly safety, and there's still some issues, and I understand your issues about some of the infection and efficacy.

But we're in the zone although it's not statistically significant and it wasn't a well designed trial. It wasn't designed to answer that question. We all know that.

And I do understand the panel's responsibility to the public, but as a neurosurgeon, I have a responsibility to my patient, and as Harry said, there is nothing out there that does this, nothing. We use inferior, off label, non-FDA approved devices, and we use them because there's no alternative, and we know that -- I know personally that, you know, you can't test it in the operating room to see if you really got everything covered with fibrin glue, which is the one that we use the most, especially in the posterior fossa.

And I'm not going to do a Valsal. but to test and see it because you can do a Valsalva and then it flips off and then where are you at? You have to do it again and scrape it off and do it again. And so you won't test it.

And we continue to have complications. CSF leak related complications have not gone away with our off-label use of fibrin glue. They have not gone away, and I mean, the pediatric neurosurgeons and any posterior fossa neurosurgeon knows this.

So we have a very pressing need, and having used this product, it's a remarkable product, and, yes, I am a consultant for the company, but I like this product, and there's nothing out there, nothing out there for us at the moment.

So I would ask you to give very careful consideration to the things that we have set as clinicians and listened to the neurosurgeons on the board who we all have a discomfort with the study design. I accept that, but listen to honest practitioners and see what they have to say because it's not a perfect world.

Thank you.

CHAIRPERSON BECKER: Thank you.

So Ms. Scudiero will now read the panel recommendation options for premarket approval applications.

Ms. Scudiero.

MS. SCUDIERO: These are the three panel recommendation options for premarket 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 Food and Drug Administration to obtain a recommendation from an expert advisory panel on designated medical device premarket approval applications, or PMAs, that are filed with the agency. The PMA must stand on its own merits, and your recommendation must be supported by the 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 the conditions of 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 when labeled will provide clinically significant results.

Your recommendation options for the PMA vote are as follows:

One, approval if there are no conditions attached;

Two, 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.

Three, not approvable. The panel may recommend that the PMA is not approvable if 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 his or her vote.

CHAIRPERSON BECKER: Thank you.

Are there any questions from the panel about the voting options before we begin?

So is there a main motion for approvability, approval with conditions, or disapproval from the panel? Dr. Canady.

DR. CANADY: I move approval with conditions.

CHAIRPERSON BECKER: So is there a second for the motion?

DR. EGNOR: I second.

CHAIRPERSON BECKER: Dr. Egnor.

So everybody in favor of a vote for approval with conditions, please raise your hand.

(Show of hands.)

CHAIRPERSON BECKER: Okay. So that's Dr. Jayam-Trouth, Dr. MacLaughlin, Dr. Haines, Dr. Canady, Dr. Jensen, Dr. Egnor, and Dr. Loftus.

Okay. Well, I think that's the majority of people voting for approval with conditions.

DR. WITTEN: But you have to vote on the conditions first before you vote on the whole motion.

CHAIRPERSON BECKER: Right. I guess we need to start now with laying out what those conditions are, and since Dr. Canady made the main motion, why don't you tell us your conditions?

DR. CANADY: Post market surveillance of infection and labeling for possible infection risk increased.

DR. MacLAUGHLIN: Excuse me. Could you expand upon what you mean by the labeling change? I understand the post approval monitoring of the patients.

DR. CANADY: Who's talking to me?

DR. MacLAUGHLIN: I am. I'm over here.

DR. CANADY: Oh.

(Laughter.)

DR. CANADY: By labeling I would label that there's a possible increased risk of infection with this device.

DR. MacLAUGHLIN: All right. Thank you.

DR. ELLENBERG: Madam Chair, point of order. Can you record the abstentions?

CHAIRPERSON BECKER: Sure. Before we talk about the motions or the conditions for approval --

MS. SCUDIERO: I think we got a little bit out of order. We voted. We had a main motion and second for this, and then the next point of order is any discussion on the motion, and then we go into identifying the specific conditions.

A vote wasn't required at that point. So we would go into what the conditions are since we have a -- is there a condition? There was one seconded. Was it seconded?

CHAIRPERSON BECKER: Yes.

MS. SCUDIERO: And then we discuss that condition and vote upon it, and then we go through if there are other conditions. Then we will vote on the whole package with all of the conditions.

Should the conditions be voted down, then we will start over with a new main motion.

CHAIRPERSON BECKER: Okay. So as I see it, the first condition that has been brought forth is that there be some requirement for a post market surveillance of patients who receive the DuraSeal device. So I guess we shall vote on that condition initially.

Any discussion on that particular motion? Dr. Loftus.

DR. LOFTUS: You know, I'm not familiar with this process. What's the mechanism for that post approval? Is there a periodic reporting function to the FDA?

CHAIRPERSON BECKER: That's a great question that maybe Dr. Witten could answer.

DR. WITTEN: If we ask the sponsor to do a post approval study, we would typically agree on the outlines of the study prior to approval. The sponsor would then after approval submit the study in a supplement for us to approve, and then they are required to report on their progress during the study, and then at the end of the study we would typically add it to the label for the product.

I'm going to answer the following question because I know it comes up, which is whether or not we have any actual hammer if the sponsor doesn't perform the study, and the fact is we try very hard to work with the sponsors to get the studies done, and I would say we have a fair amount of success, but there's not some specific action that we have taken when sponsors don't do this, and so I guess that answers that question, although it wasn't asked. I'm assuming somebody will ask me that. So I thought I'd answer it first.

DR. LOFTUS: Well, my question would be somewhat different. What happens if the data comes back and it's unfavorable?

DR. WITTEN: Certainly what we've done is we will put any additional information on the label, but if you're asking whether or not it would come back to the panel and possibly a product would get pooled, the answer is no. So the expectation is that if the panel is recommending that a product get approved that the panel believes that reasonable assurance of safety and effectiveness has already been demonstrated.

DR. LOFTUS: Thank you.

CHAIRPERSON BECKER: Dr. Haines, did you have a question?

DR. HAINES: A further follow-on is simply to ask if there really are the resources to supervise such a post marketing study and be sure that it actually gets done and the results are disseminated.

DR. WITTEN: Well, I really don't know what I can add to what I have already said. We certainly have resources to work with the sponsor as we would engage our Office of Surveillance and Biometrics that does, you know, look at post approval issues, and we certainly have the resources to work with the sponsor and make sure that there's something that we agree on as a study after approval that would take place.

But in terms of our ability to insure that those studies actually occur, we have had a fair amount of success in working with sponsors and getting studies to happen, but we haven't -- you know, our options are limited if the studies don't take place.

DR. JAYAM-TROUTH: Can I ask another question?

DR. WITTEN: Sure.

DR. JAYAM-TROUTH: How expensive is this product?

DR. SAWHNEY: The same as fibrobryl.

CHAIRPERSON BECKER: So the question is how expensive is the product?

DR. JAYAM-TROUTH: How expensive is this product?

DR. SAWHNEY: Again, Amar Sawhney, president of Confluent.

It's --

CHAIRPERSON BECKER: Excuse me. Before you answer that question, can I just ask a question of the FDA?

Is this supposed to be something we consider?

DR. WITTEN: No.

CHAIRPERSON BECKER: Okay. So never mind.

DR. GERMANO: Just for the record, am I correct in saying that the panel has not voted yet because the motion was put on the floor, was seconded; there was no discussion and there were no conditions?

CHAIRPERSON BECKER: So what's happened is that there was a motion for approvability with conditions. There was a second for that motion. So now we're going to list out the conditions, vote on each separately, and then we'll vote on the approvability of conditions as an entire package at the end, after we've laid out all of the conditions.

DR. GERMANO: So we have not voted yet.

CHAIRPERSON BECKER: Correct.

DR. JAYAM-TROUTH: We're still discussing conditions.

CHAIRPERSON BECKER: We're still discussing conditions, and we'll vote on each condition after discussion.

DR. EGNOR: I'm a little bit concerned that the approval can't be pulled if danger is seen, that is, if we do a study of the infection rate and find a year from now that the infection rate is much higher than the infection rate one would typically encounter in cases like this, we couldn't do anything except add something to the label?

DR. WITTEN: Well, let's say that we haven't done anything of that nature and so the expectation is that if it's approved that there's reasonable assurance of safety and effectiveness.

I will say that I think if some adverse information became available that a product -- and became public, you know, the hope would be that the user community would adjust their expectations of the appropriate setting for the use of that product.

However, the answer to your question is no or yes. I've forgotten how you phrased it.

CHAIRPERSON BECKER: Any other questions or comments regarding the first condition of approvability?

(No response.)

CHAIRPERSON BECKER: So then I think that we should vote on the first condition of approvability, which would be that the sponsor conduct some sort of post approval surveillance for infections in the patients treated with the DuraSeal device.

So everybody in favor of this condition. So everybody in favor of this condition, please raise your hands.

(Show of hands.)

CHAIRPERSON BECKER: So in favor is Dr. Jayam-Trouth, Dr. MacLaughlin, Dr. Haines, Dr. Canady, Dr. Jensen, Dr. Egnor, and Dr. Loftus.

Everybody opposed to this condition raise your hand.

(No response.)

CHAIRPERSON BECKER: Everybody abstaining from voting.

(Show of hands.)

CHAIRPERSON BECKER: Dr. Ellenberg and Dr. Germano.

Thank you.

The second condition that was brought forth was that there be some change in the labeling of the device to reflect that there may be an increase in infection related with this device. So would people have any comments or questions regarding this condition?

DR. HAINES: Yes.

CHAIRPERSON BECKER: Dr. Haines.

DR. HAINES: I would have two specific recommendations about the labeling. The first is that in the table of adverse effects, that the infection complications be brought together and listed together under a title "infection," rather than being separated and alphabetically listed in ways that make them hard to find, and that the total infection rate we listed as well as the subsection infection rates.

That was done for neurologic symptoms, but it should be done for infection as well.

And secondly, I think that there should be an explicit warning in the warning section that says the use of the DuraSeal Sealant System may increase the risk of deep surgical site infection.

CHAIRPERSON BECKER: Any further comments? Dr. Jayam-Trouth.

DR. JAYAM-TROUTH: And I think it should also state that it should not be used in lieu of closing up the dura. You know, the dura should be closed, and this is, you know, in addition to closing up the dura.

CHAIRPERSON BECKER: Any other comments, thoughts?

DR. ELLENBERG: That sounds like a separate condition.

CHAIRPERSON BECKER: Yeah, I think that's going to be the third condition.

What you bring up is a third condition, not the second condition. So the second condition that we'll deal with is that the labeling won't be changed to reflect the fact that there may be an increase in infection with an explicit warning that the use of this device may be associated with increased risk of an infection, and that the infections not be separated out as they were in the data we saw, but as overall infection risk, not listed as bacterial meningitis, deep surgical infection, superficial infection, but this is potential infection rate.

DR. CANADY: And that has the additional hammer of that if you do sufficient postmarket surveillance to show that that's not there, then you could apply for relabeling, which would be an incentive to complete the study.

DR. MacLAUGHLIN: Pardon me. This is Dave MacLaughlin.

Could you please repeat that comment? I didn't hear it.

DR. CANADY: That the specific warning would be an additional incentive to complete the study which would allow them to apply for relabeling.

DR. MacLAUGHLIN: Thank you.

CHAIRPERSON BECKER: So at this point let's take a vote on the second condition for approval, which is a change in the labeling for the device to express concern about the risk of an infection, making it an explicit warning, and that the infections be listed as a conglomerate whole risk of infection and not separate out as individual types of infections.

So everybody in favor of this labeling change, please raise your hand.

(Show of hands.)

CHAIRPERSON BECKER: So it's Dr. Loftus, Dr. Egnor, Dr. Jensen, Dr. Canady, Dr. Haines, Dr. MacLaughlin, Dr. Jayam-Trouth.

Everybody against this labeling change?

(No response.)

CHAIRPERSON BECKER: Everybody abstaining from the vote.

(Show of hands.)

CHAIRPERSON BECKER: Dr. Ellenberg and Dr. Germano.

The third condition that was brought forward was that this product be used after every reasonable attempt has been made to close the dura. So in addition to closing the dura, not as opposed to closing the dura.

Did anybody want to second that motion?

DR. EGNOR: Second.

CHAIRPERSON BECKER: Okay. Dr. Egnor.

Any discussion regarding that?

DR. LOFTUS: Yes, I think that's an inappropriate condition to impose because I think that implies that surgical decision making is being dictated by this panel, which is inappropriate.

Now, if indeed my opposition is sustained, I would propose another labeling issue that might supplant this one.

DR. HAINES: Well, in fact, the indication says that it's intended as an adjunct to dural closure. So I think that's already in the label.

DR. MacLAUGHLIN: That was going to be my point, too. It's already on there. That was their indication for use, was an adjunct to suture.

DR. JAYAM-TROUTH: It should be emphasized though to me.

DR. EGNOR: Yeah, it should be emphasized because people are going to use this for the tough cases regardless of what the label says, but that should be stressed on the label that you've got to get watertight closure. That's the only way it's contested, is with a closure that's nearly watertight.

CHAIRPERSON BECKER: So it sounds like maybe we should take a vote on this particular condition, that the labeling be changed to emphasize the fact that every reasonable attempt needs to be made to get closure prior to application of the device.

So everybody in favor of this condition, may I see your hands?

DR. CANADY: Could I reword that?

CHAIRPERSON BECKER: Sure.

DR. CANADY: You know, just because as a surgeon already your hackles are up when you hear that one.

That this is to be used only as a adjuvant to primary dural closure.

CHAIRPERSON BECKER: Okay. So to change the labeling to state that the DuraSeal device be used only as an adjunct to primary dural closure.

Everybody in favor of that condition for approval, may I see your hands?

(Show of hands.)

CHAIRPERSON BECKER: Drs. Loftus, Egnor, Jensen, Canady, MacLaughlin, Jayam-Trouth.

Everybody opposed to that change in labeling, may I see your hands?

(Show of hands.)

CHAIRPERSON BECKER: Dr. Haines.

And everybody abstaining from the vote?

(Show of hands.)

CHAIRPERSON BECKER: Drs. Ellenberg and Germano.

Okay. So that's three conditions now. Is there a motion for a fourth condition?

DR. LOFTUS: Yes, I would propose a motion for a fourth condition, and it would state in some word crafted way, wordsmithed way the following: that this device has been demonstrated effective only in cases where overt, spontaneous CSF leak or CSF leak documented by Valsalva maneuver has been demonstrated and that it should not be considered as standard therapy for primary closed dura where no leak is evident.

CHAIRPERSON BECKER: Is there a second for that condition?

DR. EGNOR: Second.

CHAIRPERSON BECKER: Dr. Egnor. Okay.

Any discussion points surrounding this condition?

DR. CANADY: Every case leaked.

CHAIRPERSON BECKER: Dr. Canady makes the point that every case that was treated here leaks.

DR. LOFTUS: Well, every case in the trial, but that may not be someone else's clinical experience.

CHAIRPERSON BECKER: Any further discussion on this condition?

(No response.)

CHAIRPERSON BECKER: So let's take a vote then on the fourth condition, which is that the device should only be used in cases of overt CSF leakage or in leakage associated with Valsalva maneuver. Everybody in favor of that condition, may I see your hands?

(Show of hands.)

CHAIRPERSON BECKER: Dr. Jayam-Trouth, Dr. MacLaughlin, Drs. Loftus, Egnor and Jensen.

Everybody opposed to that condition?

(Show of hands.)

CHAIRPERSON BECKER: Dr. Canady and Dr. Haines.

And everybody abstaining from voting?

(Show of hands.)

CHAIRPERSON BECKER: Dr. Ellenberg and Dr. Germano.

Any motions for further conditions for approval?

DR. MacLAUGHLIN: I have one, Madam Chairman, that was brought up in relation to the fact that the device is radiopaque. I think that was mentioned. You know, it should be stated somewhere that it is so that clinicians know that if they do some studies that that's what they're looking at.

The MRI study.

DR. GERMANO: It is not necessarily radiopaque, but it does show on the MRI as an increased -- the MRI that was shown, it looked like an area of increased signal intensity on T2 weighted images. So it should be specified what to look for on those postoperative cases, and I don't know if the sponsor did any study with and without gadolinium, and that would be also very interesting to know because obviously one of the ways to check for infections is to inject gadolinium, and I don't know if this substance will or will not have an increased uptake after gadolinium.

DR. JENSEN: Well, they stated in the canine imagines that there was a homogeneous enhancement along the edge, which diminished over time.

DR. GERMANO: With gadolinium? I don't think so.

DR. JENSEN: Well, to enhance something you've got to give something. So you would have to have gadolinium. If it was an MR, and they talked about hyperintensity. So I assume it was MR. You're correct that they did not go into great detail about what the actual imaging parameters were, and they didn't talk about the different sequences that were used, but I think it would be appropriate to have a statement in there that talks about the fact that this device can be imaged. They saw both CT and MR, and what those imaging characteristics are.

Look at the dog study.

DR. GERMANO: Was that enhancement or was it increased signal on T2 weighted?

DR. CANADY: No, it had increased signal, plus it had enhancement at the margins.

DR. GERMANO: Both. Thank you.

DR. MacLAUGHLIN: I think you got the spirit of what I wanted. I don't know how you word it, just so that the others know that it will appear on imaging of whatever type.

CHAIRPERSON BECKER: And I think it would also be important to know the time or the duration of that signal abnormality because, you know, a month or two months out you're going to be faced with a patient who may have an infection. You don't know if this is really infection or we're still seeing the end of this abnormal signal from the DuraSeal. So I think that's going to be important.

DR. JAYAM-TROUTH: One other question. Is it different on the CAT scan? Is it radio-opaque? Is it hyper dense or hypo dense?

DR. JENSEN: Well, if it's primarily water, it's going to look closer to --

DR. JAYAM-TROUTH: So it would be hypo.

DR. JENSEN: -- CSF I would think on CT.

Again, we didn't get much in the way of radiographic data, and that is something that the sponsor is going to have to have specific imaging characteristics of.

DR. GERMANO: So on page 8 of the material that was provided, there is an MRI of a dog, and it looks like a T2 weighted image, and it looks like an area of increased signal on T2 weighted imagine, and I think Dr. Jayam-Trouth's comments are very important because it is possible that the patients undergo CT instead of MR, and it would be important to know what their appearance is on a standard, conventional emergency room admission CT.

CHAIRPERSON BECKER: So it sounds like the condition that was put forth and not yet seconded has to do with the change in the labeling to reflect the fact that there are abnormal radiographic characteristics associated with the use of DuraSeal, definitely on MR, perhaps on CT, that aren't well defined.

So I guess is there a second for the motion that the labeling for this device be changed to reflect the fact that there is changes in MR and perhaps CT signal characteristics associated with the use of this and that should be made very apparent to the clinicians using it? Is there a second for that?

DR. JAYAM-TROUTH: It should be not only defined, but also the time frame that it's present should be defined.

DR. EGNOR: Second.

DR. JENSEN: Yeah, and I wouldn't call it abnormal because there's just certain imaging characteristics this material has on both CT and MR which would be clearly defined.

CHAIRPERSON BECKER: So can I actually step back and say that maybe there are two conditions here. One is that the sponsor further study the imaging characteristics associated with the DuraSeal and then secondly the labeling reflect that.

So I'm going to put out a motion that --

MS. SCUDIERO: We have a motion on the floor though.

CHAIRPERSON BECKER: I think that they need to be separated, and first we need to know what the imaging characteristics are before we can change the labeling to reflect them.

So I would actually make a motion, supplant -- do we have to do them in the order they're brought up or can we kind of table the one?

MS. SCUDIERO: Well, we have the motion on the floor now.

CHAIRPERSON BECKER: Right. We have a motion on the floor which is to change the labeling characteristics or change the label to reflect the fact that there are radiographic characteristics associated with the use of DuraSeal.

Is there a second to the motion to change the labeling characteristics in that regard?

DR. JENSEN: Second.

CHAIRPERSON BECKER: The second is Dr. Jensen.

So we've already discussed it, I think, in great detail, and so we should take a vote now for this condition, that the labeling for the device be changed to reflect the fact that there are CT and MR changes associated with DuraSeal.

Everybody in favor of changing the labeling in that regard?

DR. JAYAM-TROUTH: Were you going to add the --

CHAIRPERSON BECKER: Yeah, we're going to go back and add where they need to define what those changes are. They went a little out of order, but we're following protocol.

(Show of hands.)

CHAIRPERSON BECKER: Dr. Jayam-Trouth, Dr. MacLaughlin, Dr. Haines, Dr. Canady, Dr. Jensen, Dr. Loftus.

Everybody opposed to changing the labeling to reflect this?

(No response.)

CHAIRPERSON BECKER: And everybody abstaining from the label characteristic changes or the radiographic changes?

(Show of hands.)

CHAIRPERSON BECKER: Dr. Egnor, Dr. Ellenberg, and Dr. Germano.

So now the next motion put forward, I'm not allowed to make. Does someone else want to take up my motion to define what those MR and CT --

DR. GERMANO: Could I shed some light? Because I found the piece of paper with some sentences provided by the sponsor. This came in a blue folder and has holes in it and page 21, and it does say that DuraSeal MR and CT imaging.

So following recovery both animals -- this is talking about two dogs where the craniotomy was done -- both animals underwent MR and CT images at three days, two, four, six, eight, and ten weeks. Gel appearance at each time point was characterized and compared with pathological finding obtained 14 weeks following implantation.

Results: the investigator found that the sealant could be viewed with MRI and CT and could be distinguished from CSF.

There's no note of gadolinium. There is no characterization on how the images look like. The only mention is that --

DR. JENSEN: But look further down.

DR. GERMANO: -- is that it's different from CSF.

DR. JENSEN: Right, but look further down where it says with MR/CT imaging investigators note the following resorption characteristics, and they talk about on page 22 homogeneous circumferential marginal enhancement.

So, again, the sponsor did not provide us with enough information to say exactly what type of scanning was performed with what parameters and with and without enhancement, although mention of enhancement indicates to me that they gave gadolinium.

However, part of your protocol had imaging. So you should have that data, and the sponsor has neuroradiologists on their board. So I think that information -- but I think the time points that were given in the dog are very important.

DR. GERMANO: I just point out that they are in the dog.

DR. JENSEN: That's right.

DR. ELLENBERG: Point of order.

CHAIRPERSON BECKER: Dr. Ellenberg.

DR. ELLENBERG: My sense is that we're going to be asked our opinions and --

DR. WITTEN: Mic, please.

DR. ELLENBERG: My sense is that this will be asked for in humans, and I would like to get the advice of FDA on the issue of this requiring additional work from the sponsor pre-approval. How do you want us to handle this particular recommendation for a condition?

DR. WITTEN: Information that's additional data that you think needs to be provided pre-approval would be not a condition of approval, but would be a ‑- that would be a recommendation for not approval. If there's information that you think the sponsor should provide but could be provided after approval, then that's a condition of approval.

So if there's an additional study you want performed prior to approval, then that's not really a condition of approval.

DR. ELLENBERG: So then how do we get past the issue of labeling?

DR. WITTEN: Well, that's where you have to decide how critical it is. If it's something that's an additional study you think needs to be performed prior to approval and the sponsor doesn't have that data in hand, that is, it's an additional study, then that would be a recommendation for not approval.

If it's something where you think they should get the information later and can add it to the label later, then that could be a condition of approval.

DR. ELLENBERG: Okay.

DR. WITTEN: Does that -- yeah.

DR. CANADY: You know, in the outline of the study they have CTs on everybody at six weeks and three months. So it's a matter of analysis rather than collection. We really want a characterization. We don't want just a descriptive statement.

So I don't see the need for additional studies.

CHAIRPERSON BECKER: But additional data perhaps.

DR. CANADY: We're going to ask them for the labeling and they've got the data. They'll just put it in the label.

CHAIRPERSON BECKER: Okay. So there was a motion then for better defining the radiographic characteristics of the DuraSeal. Is there a second for that motion?

Dr. Jensen. Dr. MacLaughlin.

Okay. So everybody in favor of the sponsor looking at the data that has been collected and defining the CT characteristics of DuraSeal and the MR characteristics where they have it following approval, may I see your hands?

(Show of hands.)

CHAIRPERSON BECKER: Dr. Jayam-Trouth, Dr. MacLaughlin, Dr. Haines, Dr. Jensen, Dr. Egnor, and Dr. Loftus.

Everybody opposed to this condition?

(No response.)

CHAIRPERSON BECKER: Everybody abstaining from the vote?

(Show of hands.)

CHAIRPERSON BECKER: Dr. Germano and Dr. Ellenberg.

Do I have a motion for any other conditions for approval? No further motions?

Okay. So then as I see it there has been a motion made for approval with conditions, and there are six conditions laid out. Let me see if I can remember what those six are, and then we'll vote on those as a group.

The first condition is that there be post approval surveillance studies done for the risk of infection.

These second condition is that there be explicit labeling warning about the risk of infection with this product, that the risk may be increased, and that the change in labeling reflect the total infection rate and not separate out the different kinds of infection.

The third condition for approval is that this device be used as an adjunct to dural closure. Make that very explicit.

The fourth condition is that this device be used only in patients where there is overt CSF leakage interoperatively after dural closure or where the CSF leakage is induced by Valsalva interoperatively after dural closure.

The fifth condition --

DR. LOFTUS: May I? That wasn't exactly what I said.

CHAIRPERSON BECKER: I'm sorry.

DR. LOFTUS: I want to make sure. What I said, I believe, if I stated it right, was that the device has shown to be effective only in cases where ‑-

CHAIRPERSON BECKER: Okay.

DR. LOFTUS: And the meaning is quit different.

CHAIRPERSON BECKER: Yeah, you're right.

DR. LOFTUS: And the corollary was but should not be considered standard of care where no CSF leak can be identified.

CHAIRPERSON BECKER: So to restate condition four then, that the device should be effective in patients who had overt CSF leakage interoperatively or who had Valsalva induced CSF leakage interoperatively. It's not considered the standard of care, but it is shown to be effective only in a specific patient population.

DR. LOFTUS: But I wouldn't use standard of care. I wouldn't use that terminology.

CHAIRPERSON BECKER: That it's not standard of care?

DR. LOFTUS: Because it implies that otherwise it is.

CHAIRPERSON BECKER: For other patient populations it's not standard of care.

DR. LOFTUS: Right.

DR. GERMANO: Also it raises the question that Valsalva maneuver now is standard of care for all craniotomies, supratentorial, infratentorial; is that correct? We're not here to determine standard of care today.

CHAIRPERSON BECKER: So maybe we should just leave out the whole standard of care part and just state that it's show to be effective only in patients with overt CSF leakage or Valsalva induced CSF leakage interoperatively.

DR. LOFTUS: I could accept that.

DR. EGNOR: Or say the data supports its use only.

CHAIRPERSON BECKER: Fair enough.

DR. JAYAM-TROUTH: Can you restate it, please?

CHAIRPERSON BECKER: So the fourth condition was that the device has been shown to be effective in patients with overt or Valsalva induced CSF leakage interoperatively.

DR. ELLENBERG: Only.

DR. JAYAM-TROUTH: Only.

CHAIRPERSON BECKER: Only. Okay. The fifth condition is --

DR. GERMANO: Effective in doing what?

CHAIRPERSON BECKER: I'm sorry?

DR. GERMANO: Effective in doing what?

CHAIRPERSON BECKER: Effective in closing ‑-

DR. GERMANO: Of interoperative CSF leak cessation.

CHAIRPERSON BECKER: So let me see if I can restate that correctly. So the device has been shown to be effective in stopping interoperative CSF leaks in patients who have overt CSF or Valsalva induced CSF leaks --

DR. LOFTUS: Do you know what? It's totally redundant because the other patients didn't have a leak. Maybe I should just withdraw the motion altogether because that's totally redundant.

CHAIRPERSON BECKER: So let's take another vote on this motion just to be sure that everybody is clear.

DR. LOFTUS: You see what I'm driving at. I wanted to get a motion that protects the neurosurgeon who chooses not to use this from being accused of a violation of the standard of care. That's what I'm trying to achieve, as a protective mechanism, you know, for our profession.

DR. JAYAM-TROUTH: How would you word it?

DR. LOFTUS: Well, I've made every attempt I can think of.

(Laughter.)

DR. EGNOR: I mean, could one say that the use or lack of use of this material is not reflected in the standard of care?

CHAIRPERSON BECKER: It was just made apparent to me that the FDA can actually help the sponsor work out this wording and we probably don't need to fret about that too much.

So now the fourth condition, since we've gotten rid of the previous fourth condition, is that the labeling be changed to reflect that there are imaging characteristics associated with the use of this device and that the clinician should be warned about that.

And then the fifth condition, the new fifth condition, is that the sponsor actually do a little bit of research with the data that's already been collected, a little analysis of the data that's already been collected to help us define exactly what those changes are in humans and how long they last.

DR. ELLENBERG: Point of order. I'm sorry. I'm confused. Have we deleted as a condition for approval the limitations on what has been shown in terms of the usefulness in a certain patient population or have we put it into another area of the labeling?

CHAIRPERSON BECKER: I think it's my impression that it's actually very clearly laid out in the indications for labeling that this is where the device has been shown to be effective. All right? So it's not actually in the --

DR. ELLENBERG: Why would they state that this is shown to be effective in X, Y, Z populations in the label if we don't put it in as a condition of approval? That's a major limitation that we've been talking about all day.

CHAIRPERSON BECKER: So, Dr. Witten, can we have a little help with that one?

DR. WITTEN: Yes. If we don't specify in the label the statement who it's used for, and there's no specific contraindications, which I don't think is what you're talking about anyway --

DR. ELLENBERG: No.

DR. WITTEN: -- then the only way that it would be described in the label would be under the clinical study where we describe the patients in whom it is studied.

Now, what I could say is if there's some concept you're trying to get across, then rather than try to work out the specific wording here, you could just explain what the concept is and then if you all agree on that concept, whatever it is we'll try to get it to show up in the label in some reasonable fashion, recognizable fashion.

CHAIRPERSON BECKER: Dr. Loftus.

DR. LOFTUS: Yes. The concept that I'm trying to address is that there is not enough efficacy from this particular trial design to say that this needs to be used in every case where a patient is operated on and a durotomy is performed, and it would be remiss not to use the product. That's what I'm trying to get across.

DR. GERMANO: Second.

CHAIRPERSON BECKER: Dr. Haines.

DR. HAINES: And I would say that since the indication clearly states that it's intended for use as an adjunct, it's not required. I mean somebody will try to infer things regardless of what the wording is, but it says adjunct. This is clearly not intended as a primary, as stated, as being essential to closure of the dura.

DR. GERMANO: But, Steven, those are two different concepts. What you are saying is that this is not the only way we close dura.

What I think we're trying to say is that not only this is not the only way we close dura with, but if we don't use it, it's only because now it's going to be the only FDA approved product, and so then if you have a question in your mind that the dura is leaking, then you have to use it unless we state that there is not enough evidence today that this is actually efficacious down the line.

There is evidence, 90 percent, 98 percent of the time of the surgery, but whether or not that has any clinical meaning we don't have.

DR. CANADY: What you really want to say is that we have demonstrated interoperative effectiveness, but we have not yet demonstrated the clinical prevention of CSF leaks.

DR. GERMANO: Correct. Thank you.

DR. EGNOR: So why don't you say an optional adjunct?

CHAIRPERSON BECKER: So it sounds like there's a motion for another labeling. We'll let Dr. Loftus make a comment and then we'll --

DR. LOFTUS: Yeah, I'm not changing the motion. It really is important to me emotionally because we have had tangible, serious questions about safety issues here today that we have, you know, to some extent overcome, and yet it needs to be very clear that this is not a treatment without risk, and that it's not mandatory in routine surgery.

DR. CANADY: I think if we can separate it from this product does not clearly prevent clinical CSF leaks, then that's what we're going to get sued on. So if we say there's no clear data to show you that this is effective for that, then you've separated the two and protected the surgeon.

CHAIRPERSON BECKER: So it sounds like maybe rewording the condition that you had to reflect that would be appropriate.

DR. GERMANO: I think what Dr. Canady worded is perfect, that this product is approved or has been shown to be effective to stop intraoperative CSF leak and no clear benefit for postoperative CSF leak has been demonstrated yet.

MS. SCUDIERO: What about just withdrawing Dr. Loftus or voting his motion down and starting over with the clear --

DR. GERMANO: Yes.

MS. SCUDIERO: I think that one is good.

DR. GERMANO: And also this, going back to using Dr. Canady's wording before, this is a nice incentive to do those postoperative, post labeling study because the sponsor can reapply and say now we have demonstrated that this clearly decreases the postoperative CSF leak, and so it's a good incentive for the sponsor to show that data to the agency.

CHAIRPERSON BECKER: So everybody in favor of withdrawing the previous condition set forth by Dr. Loftus, may I see your hands?

(Show of hands.)

CHAIRPERSON BECKER: So that's Dr. Germano, Dr. Jayam-Trouth, Dr. Haines, Dr. MacLaughlin, Dr. Canady, Dr. Jensen, Dr. Egnor, Dr. Loftus, yourself.

So then there's been a motion to reword or change the condition essentially to state that this device has been used to effectively stop interoperative CSF leaks, and that the clinical significance of this is unknown. Is that reasonable?

Is there a second for that motion?

DR. JAYAM-TROUTH: Second.

CHAIRPERSON BECKER: Okay. So everybody in favor of that condition, may I see your hands?

DR. MacLAUGHLIN: Wait. Could I have a little discussion?

CHAIRPERSON BECKER: Sure.

DR. MacLAUGHLIN: Just I have a technical question. Not having read a lot of these labels, can you qualify? Do you see other qualified statements like that for a product? You know, this efficacy hasn't been proven yet. It implies lots of things. Has that ever been seen before?

CHAIRPERSON BECKER: I could actually give you an example from the recent concentric device approval stating that it was used to remove clots from the CNS vasculature, but not for the treatment of stroke, right? So --

DR. MacLAUGHLIN: No, I can understand that qualification. I'm talking about, you know, it's a clear demonstration that it stops leaks interoperatively, but the other phrase after that I didn't know. Can you be that qualified?

CHAIRPERSON BECKER: Can you, Dr. Witten?

DR. WITTEN: Well, I don't think we say what things don't do. You know, we would say what it's for, not what it's not for. We would describe the study and what the study -- for any product, this one or any other product. In the study description in the label we would describe what the product had or hadn't been shown to do, but we don't put in the label it does this, but it doesn't do that.

I'm not sure exactly what you're quoting from, you know, because when we put out -- well, that example wasn't a PMA, but I think we might have had a press release, but in this we'd have a safety and effectiveness that would say something.

But we wouldn't say, for example, it doesn't -- I mean, if you're voting to approve it, you're voting that there is reasonable assurance that it will provide clinically significant results. So we're not going to say it works, but it doesn't work. I mean we could say in the indications this is what it's intended for and in the study this is exactly what was shown and what wasn't shown.

But I can't imagine us saying it works and it doesn't work.

DR. MacLAUGHLIN: I guess I was trying to be real sensitive to the clinicians who need to be, you know, protected from someone reading it and saying, "Oh, gee, you didn't do this. So you've been negligent," and you know, making other conclusions. I don't know. It just sounded --

DR. CANADY: We'll just subpoena Dr. Witten and she'll come and testify that we couldn't do what we wanted to do.

DR. WITTEN: Yes. So maybe somebody will have to read more than just the first sentence in the label of the SS&E. I mean that's also a possibility.

CHAIRPERSON BECKER: So, Dr. Jensen, did you have a point you wanted to make?

DR. JENSEN: I guess just looking at the indication all it says is that it's for use in watertight closure and that's it. It doesn't say anything else about "and to prevent this," or whatever.

Now, I know you probably want it to say that in more glaring language, but that can be said -- correct me if I'm wrong, Dr. Witten -- over in the clinical experience aspect of it?

DR. WITTEN: Well, yes. When we put a label and we describe a study, we typically describe the study, the study population and the important endpoints that were assessed in the study, both the primary endpoint and some of the other things.

Clearly, this is one that we put in that would be the clinical CSF leak experience.

DR. ELLENBERG: Let me -- oh, I'm sorry.

CHAIRPERSON BECKER: Actually, we'll let Dr. Haines make his point and then.

DR. HAINES: In the label under warnings, it starts and says the "safety and performance of DuraSeal hydrogel has not been established," colon, in Section 2, the first part of the label. Would it be acceptable --

DR. WITTEN: I need to see what you're talking about here.

DR. HAINES: Okay. The first column right down at the bottom, warnings. "Safety and performance of DuraSeal hydrogel has not been established," colon. Would it be acceptable to add another bullet that says "for the prevention of clinically significant CSF leaks"?

DR. WITTEN: A warning or these contraindications and warnings are, you know, circumstances in which, I mean, I'm not sure exactly whether we agree with precisely how they have these in here, but they are to indicate situations in which you wouldn't want to use it either because there's a safety issue for the patient -- well, that's pretty much -- a contraindication is when there's a known safety issue, and a warning is when there's a potential safety issue with how you might use it.

So we wouldn't do what you just said because that doesn't fit in with the warning.

DR. HAINES: Well, how about precautions, the last bullet, safety and performance has not been established in persons younger than 18 years of age and procedures involving petris bone drilling and add a third bullet for the prevention of clinically significant postoperative CSF leaks.

DR. WITTEN: Well, that precaution is -- I'm not sure where you're reading from, but a precaution is something that you need to take into account, you know, during the actual process of use. It's not to say something about where it doesn't work.

So, again, I have to go back to, you know, there's a definition of reasonable assurance of safety and effectiveness, and if we're approving it for a specific indication, then we're saying there's a reasonable assurance of safety and effectiveness for that indication. And if there are some limitations about what it doesn't do, those aren't described as it doesn't do that. During the study description, that's under the -- I see. It's called the clinical experience section of the label. That's where it would be described, what happened, what didn't happen, you know, what we didn't see, what we did see. That's where that would go.

DR. HAINES: Maybe that's a good solution, you know, is to have -- I'm really, like I said, I'm very concerned that people feel protected in this kind of setting, and I don't know how -- I'm not at risk. So maybe we're just trying to decide where to put that kind of information, you know, defining where it's really working.

CHAIRPERSON BECKER: To kind of summarize, the indication, I think, is pretty clearly stated that the dural sealant system is intended for use as an adjunct to sutured dural repair during cranial surgery to provide watertight closure, and then the issue really comes up about warning that this may not have the long-term effectiveness of preventing CSF leaks.

And I guess the question is: can we put that in the clinical description? Is that sufficient?

Does someone want to make a motion about that or how should we proceed?

DR. WITTEN: Well, you don't need a motion for it to go in the clinical experience section because that's something we would decide whether that was important, which in this case it clearly is and we would put it in there.

And I'm not sure whether there's anywhere else that it could logically go. So you know, like I say, unless there's an underlying concern about reasonable assurance of safety and effectiveness.

DR. JENSEN: What about in the clinical experience where they say the incidence of postoperative CSF leaks in the study was low? But then can you put after that "but no different than other" --

DR. WITTEN: I mean, we will have to work on the label, but we usually stay away -- I'm just saying in general we usually stay away from describing things as low or high, but we just put in a number saying this is what we saw and not say --

DR. JENSEN: Okay.

DR. WITTEN: -- you know, it was high, it was low, it was good, it was bad. It's just usually pure statements of material fact go in here like what it was.

CHAIRPERSON BECKER: So it sounds like if we want to include some kind of labeling that states that this was of short term benefit or it's effective in the short term, but the long-term benefits are unclear, then that would almost be a vote for nonapproval is what you're saying.

DR. WITTEN: Yes, yes. Either you think it has reasonable assurance of safety and effectiveness, and then we describe the experience under clinical experience, or you don't.

CHAIRPERSON BECKER: Dr. Jayam-Trouth, did you have something to say?

DR. JAYAM-TROUTH: It was different from the present discussion, and that was maybe in the warning there was another bullet we might try to put in and that is that we don't have studies on what it does at CSF and CSF inflammatory responses because that hasn't been studied.

CHAIRPERSON BECKER: Dr. Loftus.

DR. LOFTUS: I don't want to belabor this, but are you basically saying, Dr. Witten, that we shouldn't tinker with this wording in this document? Because I'm going to suggest one more tinkering if you don't say that.

DR. WITTEN: No, I'm not suggesting that you not tinker with the wording. In fact, quite the opposite, but I'm just pointing out that the specific tinkering that's under discussion about saying it works and then warning that it doesn't work is just not something that, you know, I could understand.

DR. ELLENBERG: It works in the indicated usage in the patient population that was studied.

DR. WITTEN: Yes.

DR. ELLENBERG: If we don't put in the patient population that was studied in here, we're giving carte blanche. So just put it into the clinical experience --

DR. WITTEN: Yes.

DR. ELLENBERG: -- it seems to me, and in the clinical experience to describe the patient --

DR. WITTEN: No, sorry. I'm sorry if I've been misunderstood. If you want to put something in about the patient population and the indication like only use it in patients with two millimeter, you know, or patients who don't have it near the bone or only use it in patients who are, you know, a certain age range, you know, you could suggest that. I mean, you certainly could suggest that for the indications.

I was merely commenting on the specific business of putting a warning in saying that, you know, it doesn't work long term and where that would go.

DR. ELLENBERG: Well, my sense would be the warning for the long term would be a statement either in clinical experience that simply says there is no evidence concerning long term. That's not denying or abrogating the concept of safe and effective for what it was studied.

DR. WITTEN: Yes, something describing what happened in the study, what was or wasn't discussed or reviewed in the clinical experience, it would go in the clinical experience section, or responding to the suggestion that it be put in as a warning about the product.

CHAIRPERSON BECKER: Dr. Loftus.

DR. LOFTUS: Yes. Thank you.

May I suggest that we consider this possibility? Under indication, the DuraSeal Dural Sealant System is intended for use as an adjunct to sutured dura repair during cranial surgery where watertight dural closure or primary watertight dural closure cannot be either achieved or assured.

CHAIRPERSON BECKER: So you're making a motion for that change in labeling?

DR. LOFTUS: Yes, ma'am.

CHAIRPERSON BECKER: Someone second that motion.

DR. EGNOR: Second.

CHAIRPERSON BECKER: Dr. Egnor.

So can we take a vote on that? And could you read it one more time so that I don't get around?

DR. LOFTUS: I was hoping somebody else would do it. The DuraSeal Dural Sealant System is intended for use as an adjunct to sutured dural repair during cranial surgery where primary watertight dural closer cannot be assured.

DR. CANADY: Why don't you just say "obtained"? Because "assured" over --

DR. LOFTUS: Obtained, achieved, assured, I'm happy to consider amendments.

DR. GERMANO: But then you have to say it has to be only two millimeters and away from the bone because that's what they studied. So you don't know if it works if it's three millimeters and next to the bone.

CHAIRPERSON BECKER: And in many ways that's reflecting the condition three that we had laid out as an adjunct, right? To primary dural closure. So it's almost a repeat of that.

DR. LOFTUS: Well, you can certainly have the opportunity to vote my motion down.

CHAIRPERSON BECKER: So let's take a vote on that motion, that the indication be changed as Dr. Loftus had just read. Everybody in favor of making that change.

(Show of hands.)

CHAIRPERSON BECKER: Dr. Jayam-Trouth, MacLaughlin, Haines, Egnor and Loftus.

Everybody opposed to that change?

(No response.)

CHAIRPERSON BECKER: And everybody abstaining from voting?

(Show of hands.)

CHAIRPERSON BECKER: Drs. Ellenberg, Jensen, Canady, and Germano.

DR. ELLENBERG: I think you have to vote.

CHAIRPERSON BECKER: Was it four to four?

So everybody in favor of that change, please, again raise their hands.

(Show of hands.)

CHAIRPERSON BECKER: Jayam-Trouth, MacLaughlin, Haines --

DR. ELLENBERG: I miscounted.

CHAIRPERSON BECKER: Yeah. -- Egnor and Loftus. Five to four. Okay.

Any other motions?

DR. JAYAM-TROUTH: As I said before, we have to somewhere indicate that the CSF inflammatory process has not been studies.

CHAIRPERSON BECKER: Is there a second for that motion?

DR. JENSEN: Second.

CHAIRPERSON BECKER: Dr. Jensen.

So that would be potentially a warning?

DR. JAYAM-TROUTH: A warning.

CHAIRPERSON BECKER: A bullet for warning. So everybody in favor of adding a warning that states that the CSF inflammatory response has not been studied in these patients?

DR. GERMANO: Could we have discussion for that?

CHAIRPERSON BECKER: Sure.

DR. GERMANO: How good or bad does it look that we approve something without knowing what it does?

DR. ELLENBERG: That's what we've done.

CHAIRPERSON BECKER: Yeah.

(Laughter.)

DR. EGNOR: Well, you never know everything. I mean, there's no way you can know everything. We're just pointing out things.

DR. GERMANO: Something like CSF is so simple to do.

DR. MacLAUGHLIN: I don't know. I think we kind of have to deal with the cards we're dealt at the moment. You know, the claim is that it stops the leak in the CSF in the operating room. I think these kinds of discussions always bring up lots of other studies to do and lots of other things to do, but I'm always trying to weigh one thing, you know, the safety issue against what other stuff we'd like to know. Do you know what I mean?

I'm just thinking of the sort of motion in front of us about whether this seals and whether we think it's safe, I guess, not what hasn't been done.

DR. GERMANO: Well, wouldn't it make it a stronger label if some of those "ifs" were removed and this application is reviewed when the CSF data is available, the radiographic data is shared?

I mean, maybe we cannot answer all of the questions, but we can answer some and make a very strong label with less questions.

CHAIRPERSON BECKER: I think that's, sure, the label would be improved, and I think that's what we're being asked to vote on, is how strongly do we want to prove this now and how strongly would we like the sponsor to come up with information first.

Since we have had a motion and a second for changing the labeling about the CSF issue, reflecting the fact that there is no information about the effect of the DuraSeal system on CSF inflammatory response, how many people are in favor of changing the labeling to indicate that? Let me see your hands.

(Show of hands.)

CHAIRPERSON BECKER: Drs. Jensen and Jayam-Trouth.

Everybody opposed to that labeling change.

(Show of hands.)

CHAIRPERSON BECKER: Dr. Canady, Dr. Haines.

Everybody abstaining.

(Show of hands.)

CHAIRPERSON BECKER: Drs. Loftus, Egnor, Ellenberg, MacLaughlin, and Germano.

So it's a vote of two to two for the labeling change, and I'll vote in favor of it. We'll put that warning in.

So any motions for other conditions at this point? Dr. Ellenberg.

DR. ELLENBERG: I would like to make a motion that the characterization of the population studied be included in the clinical experience, first paragraph, as a condition of approval.

CHAIRPERSON BECKER: We don't necessarily need to make a motion about things included in the clinical experience?

DR. WITTEN: Well, you certainly can if you want to. We typically would put it in, but if you feel strongly enough, you can suggest that, and like I say, that's the kind of thing you could also suggest for the indication. I didn't mean to say you couldn't put things somewhere else. It was just the one specific suggestion I was commenting on.

DR. ELLENBERG: I would like to accept a friendly amendment to put that in the indication with the wording that FDA will determine.

CHAIRPERSON BECKER: Is there a second to that motion about stipulating that this device is effective in the patient population studied?

DR. EGNOR: Second.

CHAIRPERSON BECKER: Second, Dr. Egnor.

Any discussion on that issue?

(No response.)

CHAIRPERSON BECKER: So everybody in favor of stipulating that the DuraSeal Sealant System is effective in the patient population studied, as laid out in the study design, may I see your hands?

(Show of hands.)

CHAIRPERSON BECKER: Drs. Loftus, Egnor, Jensen, Canady, MacLaughlin.

DR. WITTEN: Actually, can I ask for some specific clarification for that condition? Because I know in advance that one thing will come up, which is the age. You know, if we look at the population, if we look at all of the inclusion criteria and all of the exclusion criteria, then one we will end up having some discussions about later is that the age for this study was 18 to something.

So is that part of this recommendation as far as the indications?

CHAIRPERSON BECKER: Let's open it up for panel discussion. Is age going to limit the use of this product?

DR. EGNOR: There certainly is nothing about what we've seen about the product that would make one think or make me think that there would be a differential outcome regarding age. How would this be dealt with with other devices by the FDA?

DR. WITTEN: Well, it's the kind of question we'd ask the panel for their advice on, and the panel could either tell --

(Laughter.)

DR. WITTEN: No, seriously, could either tell us that they think that it should be limited in the indication also in terms of age. They could recommend that they think that there's no reason to expect that it would be different in the pediatric population, and recommend that the age not be limited in the indication or, you know, some third option not specified. I don't know what that would be.

DR. HAINES: There's already a statement in the precautions about age. I think it would be redundant to put that in the indications, and I think we're beginning to actually increase the risk of the things that Dr. Loftus is concerned about by making things way too specific and creating opportunities.

There are certain social problems we can't solve in the label in this device.

DR. ELLENBERG: Well, the final statement in the current label says the safe and effective use of this DuraSeal sealant for its intended use is supported by the findings of this study, and if we don't indicate the nature of the cohort in the study, it seems to me that's giving a general license to use for all patients.

DR. CANADY: But won't it be listed in your summary of the study?

DR. WITTEN: Yes.

DR. ELLENBERG: I think that's a much lesser --

DR. CANADY: Do you think this label is going to limit how people are going to use it at all?

DR. ELLENBERG: No. I thought you argued that you read labels.

DR. CANADY: I read them, but that doesn't mean I don't just do --

(Laughter.)

DR. CANADY: I think we're over engineering.

CHAIRPERSON BECKER: So the motion that we had just voted on, actually, Dr. Jayam-Trouth, I didn't see whether you had voted for or against stipulations about the device being effective in a patient population studied for the indications.

DR. JAYAM-TROUTH: I put against because I think it was random.

CHAIRPERSON BECKER: So who else is opposed to making that change in the indications?

DR. CANADY: I am because I think it's just as good in young people as it is in old.

CHAIRPERSON BECKER: And who is abstaining from voting there?

(Show of hands.)

DR. ELLENBERG: Has it passed?

CHAIRPERSON BECKER: Okay. Motions for other conditions.

DR. ELLENBERG: Has that passed or failed?

CHAIRPERSON BECKER: It's passed. Four to three is my count.

DR. GERMANO: Four abstained?

CHAIRPERSON BECKER: Now. Four for, two against, three abstained.

DR. GERMANO: Four abstained.

CHAIRPERSON BECKER: Can I see hands again? Everybody who abstained from voting on that change?

(Show of hands.)

CHAIRPERSON BECKER: So four. Okay. Everybody opposed to that change indication?

(Show of hands.)

CHAIRPERSON BECKER: Two.

DR. JAYAM-TROUTH: Three.

CHAIRPERSON BECKER: Ah. Everybody for that change in indication?

(Show of hands.)

CHAIRPERSON BECKER: Two. Okay. I'm sorry. That one does not get approved.

DR. ELLENBERG: I would like to make a motion that the nature of the patient population be specifically stated in the first paragraph of clinical experience.

CHAIRPERSON BECKER: So it's a change in the motion from moving the stipulations about the patient population from the indications to the clinical experience part of the labeling indication.

Is there a second for that motion?

(No response.)

CHAIRPERSON BECKER: No second?

DR. JAYAM-TROUTH: Second.

CHAIRPERSON BECKER: So there is a second. So any discussions on this point?

It's essentially just moving what we just talked about to a different part of the label.

DR. ELLENBERG: Can I speak on this?

CHAIRPERSON BECKER: Sure.

DR. ELLENBERG: It seems to me that this is the crux of what we've been discussing all day and not to highlight it in the clinical experience to me does not make much sense because it seems to me that it would allow that this could be used for any cohort of patients regardless of whether or not this particular data set given to us today justified it.

CHAIRPERSON BECKER: Okay. Dr. Loftus.

DR. LOFTUS: There is to my mind a tangible risk in over engineering this as Dr. Canady had said. I sought to protect the surgeon against an error of omission. This potentially exposes the surgeon to an error of commission if they were to use this in a trauma patient and it was defined that the study had not -- or in a pediatric patient -- who were not included in this very limited study. And I find that troublesome.

DR. JENSEN: But the FDA is not saying that it's disapproved for those uses. It would be an off label use of the device, correct?

DR. WITTEN: No, if it's not in the indication, if the indication stays as it is and the contraindications warnings and precautions stay as they are, that's a pretty general indication, and it wouldn't be off label for these various populations. We would typically put in the clinical experience section the study so that the physician reading the label, of which hopefully there would be some, would have some understanding of the basis for the use and the instructions for use and what to expect in terms of the results.

DR. JENSEN: All right. So outlining the patient population up front in the clinical experience doesn't --

DR. WITTEN: Doesn't limit the indication, and frankly, we do that. We will expect to do that in any case. I mean, that is what we do when we describe the study, is who was studied, but you don't expect -- I mean, just in general for a lot of our products the study doesn't study the entire universe of patients in whom our product might be used, and we don't go, you know, specifically -- you know, there are some cases where we might feel that the study was a model for one particular group, and then we would label it in the indications for that group and put that in the clinical experience.

But then there's other cases where the study, you know, perhaps supports a broad indication even though those weren't specifically who was studied. It would still go in the clinical experience section as statements of fact about the study but wouldn't make it into the indication.

CHAIRPERSON BECKER: So I think that we had made a motion and a second regarding making sure that information got into the clinical experience. Can we take a vote on that? Everybody in favor of making sure that gets into the clinical experience?

(Show of hands.)

CHAIRPERSON BECKER: Dr. Jayam-Trouth, Dr. Haines, Dr. Jensen, and Dr. MacLaughlin.

Everybody against putting that in the clinical experience?

(Show of hands.)

CHAIRPERSON BECKER: Drs. Canady, Loftus ‑-

DR. CANADY: That's just a routine place where you would put it? No, I'm for that.

CHAIRPERSON BECKER: Okay. Dr. Loftus, you were against or for?

DR. LOFTUS: I'm going to abstain.

CHAIRPERSON BECKER: Anybody against putting that information?

(No response.)

CHAIRPERSON BECKER: Anybody abstaining from putting that information in?

(Show of hands.)

CHAIRPERSON BECKER: So that's Dr. Loftus, Egnor, Ellenberg, and Germano. Five to four.

Any motions for other conditions?

So I think that now we have to go back through and read all of the conditions again since it has been so long since we've done that and hopefully get them right.

So this is a vote for approval of the DuraSeal Sealant System with the following conditions. Firstly, that there be post approval surveillance for infections;

Secondly, that there be explicit warnings for the risk of an infection in the labeling and that the labeling be changed to reflect all of the infections as a single entity as opposed to separated out into type of infections.

Thirdly, that we make sure that the indications are for using this device as an adjunct to primary dural closure.

Fourthly, that there be information in the labeling warning the clinicians that there are CT and MR changes associated with use of the device.

Fifthly, that the company get some information for us about the nature of these changes and the duration of the changes after device implantation.

Next, that there be warnings in the labeling that there is no information about the CSF inflammatory response with -- actually, I'm going to take that one out. I think we voted against that. No, we didn't. I'm sorry. So that one is in.

So CSF inflammatory response is unknown after device implantation.

Next, that the patient population studied in the dural sealant system pivotal study the explicitly outlined in the clinical experience section.

And then someone may be actually able to help me because I have another condition here that's very similar to one that is already mentioned, and that is that in the indications we mentioned that the DuraSeal Sealant System is intended for use as an adjunct to sutured dural repair during cranial surgery to provide watertight closure where it can't otherwise be obtained, yeah, otherwise can't be obtained.

DR. GERMANO: I think we voted against that one because it was redundant.

CHAIRPERSON BECKER: Yeah, I thought so, too.

DR. GERMANO: Yeah, in the precaution they already say do not use it if it's more than two millimeters and do not use if it's three millimeter next to the bone.

CHAIRPERSON BECKER: So that last one falls off then.

DR. LOFTUS: I'm sorry. I thought we voted in favor of the change of the indication to say that or to provide a watertight closure where it could not be obtained, where primary watertight closure could not be obtained. Did we not vote in favor of that motion?

CHAIRPERSON BECKER: I think we voted for it, yeah, although it seems very similar to the wording that this only be used as an adjunct. It seems very similar to me.

DR. CANADY: Why don't we put them together and let them work out the specifics?

CHAIRPERSON BECKER: Is that fair that we put those two conditions together and let the FDA work out the wording?

So with those conditions as outlined, can we have a final vote then on the approvability of this premarket approval for the DuraSeal Sealant System?

Everybody in favor of approval, may I see your hands?

(Show of hands.)

CHAIRPERSON BECKER: Drs. Jayam-Trouth, MacLaughlin, Haines --

DR. CANADY: Point of order. Are we voting for the approval with limitations?

CHAIRPERSON BECKER: We're voting for the approval of all the conditions as outlined.

Canady, Jensen, Egnor, and Loftus.

Everybody against approval, may I see your hands?

(Show of hands.)

CHAIRPERSON BECKER: Dr. Ellenberg and Dr. Germano.

And anybody abstaining from voting at this point?

(No response.)

CHAIRPERSON BECKER: So it looks like the panel has voted in favor of approval with conditions, seven people voting for the approval, two people abstaining from voting.

DR. CANADY: No, they didn't abstain. They opposed.

CHAIRPERSON BECKER: Opposed. I'm sorry. They abstained at all the other points. Two opposed to voting.

And I think at this point it's usually customary for everybody to go around and give some closing remarks on their thoughts about its approvability. So why don't we start with Mr. Balo?

MR. BALO: I mean, what can you say after everything has been said all day today? Let me tell you, but this is sort of my last panel meeting, and I'd sure like to take the time to really thank Dr. Witten and the panel members and really the sponsors for the hard work that they do to bring this to the panel.

It takes a lot of hard work to get here, and I think the panel does a great job to really try and sort out the facts, look for the public benefit. I think as we said through the questions that the FDA asked, I think we all agree that it was effective as the study was designed, and I think we sort of had questions about safety, but we put some conditions on that will sort of sort that out.

So I'd just like to applaud the panel for their work today.

Thank you.

CHAIRPERSON BECKER: Dr. Loftus.

DR. LOFTUS: Yeah, if I may just summarize, you know, the decision making, the facts that entered into my decision making, first, the type of agent is clearly in common use. This is a fact of life. As Dr. van Loveren finally alluded to the thing, I thought they would have said first thing this morning that the current product has a very distinct advantage, no lineage from human material, which cannot be overstressed.

The study has some artificial attributes regarding eligibility and interoperative testing. It's a pity the data is not available for the 23 patients, et cetera. I do believe the product works as intended.

And finally, if approved, when approved, now approved, right or wrong, I believe the product is going to be used off label for spinal CSF egress repair probably as much or more than cranial repair.

CHAIRPERSON BECKER: Dr. Egnor.

DR. EGNOR: Regarding the safety of the product, it's very good that it involves off-shelf components, and there's no major toxicological issue. The dog studies were well done, and while the clinical studies, I believe, need to be done further, there's no compelling evidence in my view that this is unsafe, and I believe that the benefit outweighs the risk overall.

As far as effectiveness, again, the dog studies were well done. Clinical studies should be continued. However, it seems to be as best one can see at least as effective as what is routinely used for this, and it's better studied than most things that we routinely use.

CHAIRPERSON BECKER: Dr. Ellenberg.

DR. ELLENBERG: I voted no only because of the issue of lack of control in assessing the safety of this product and, therefore, I am not able to judge the risk-benefit ratio, which is a requirement for voting for approval.

CHAIRPERSON BECKER: Dr. Jensen.

DR. JENSEN: I think the data supports the safety and efficacy for immediate closure to obtain a watertight seal. I strongly encourage the sponsors to get the postmarket data to the FDA as soon as it's reasonably available, and otherwise I would just like to thank Dr. Becker for doing a great job.

CHAIRPERSON BECKER: Thank you.

DR. JENSEN: And wish her well.

DR. CANADY: I'd just like to say I think the entire discussion today has been certainly an interesting one. I think it highlights the need to really assess from a clinical perspective the issue of CSF leaks and identify those factors that are related and in a much more scientific way than has been done in the past.

CHAIRPERSON BECKER: I wasn't asked to vote. I think the sponsor did a great job in doing what the FDA asked them to do. I think we still find ourselves in this place where we don't know what the long-term clinical benefit of this m is, and I am hoping that at some point in time we can actually get that data, and I think that we really can't say that this device does work to prevent CSF leaks long term, but we don't know that it doesn't either.

So I think we're just kind of left in a quandary.

Dr. Haines.

DR. HAINES: I believe that the reasonably thorough nonclinical evaluation of safety was a very important part of my decision. This is clearly an application for which we need a good product.

I think the clinical evaluation was more burdensome than it needed to be, and I think that with a small investment of a little bit more effort in a more appropriate clinical evaluation we could have had a much, much easier time in arriving at this conclusion.

CHAIRPERSON BECKER: Dr. MacLaughlin.

DR. MacLAUGHLIN: Yes, I voted in favor because I think if you look at the sort of mandate of what we were asked to evaluate, which was the closure, the interoperative time, I think that was well met. I think the materials taken by themselves are not, you know, toxic or dangerous. I think the future though will show us what the long-range efficacy is, and I think you have to start somewhere in a study like this.

You know, we couldn't agree -- I think all of us -- what the good control would be for these patients, and maybe as more experience is there and more data is accumulated we'll begin to learn, you know, how this begins to compare to other modes of treatment, but I think I didn't see anything to stop me from moving ahead.

CHAIRPERSON BECKER: Dr. Jayam-Trouth.

DR. JAYAM-TROUTH: Well, I guess I also go with the crowd, and I voted yes because there's definitely a need for some material to, you know, close CSF leaks, and I think the clinicians are struggling with that, and they put all types of autologous materials in there for which we don't have any data at all.

Certainly this is safe. You know, it is effective under the circumstances that are shown, and I think there's definitely need, you know, for more trials and especially in difficult circumstances. You know, I think those are the circumstances where surgeons really need something.

And to me, I mean, to show it and the regular case with just spontaneous leaks doesn't make as much sense as to show it in difficult cases where there is nothing else that would work.

CHAIRPERSON BECKER: Dr. Germano.

DR. GERMANO: I voted no, and I think that this was a study that had struggled putting together the sponsors with the FDA view, and I think this is why we had so much struggle today with the vote.

And it seems that some of the issues that were raised today perhaps should have been dealt with and raised up front when the study was being created.

In any event, my vote against was based on the lack of being able to establish safety and efficacy on the data that was presented, and I believe that some of the data is available and that the sponsor does have some of the information that this panel wanted to see, some of it fairly simple like CSF and radiographic, and probably by investing in a little extra time looking at cases where CSF leak is really known to be a problem, such as posterior fossa, and I would have hoped that by having that data the label would have been crystal clear and much friendlier than what was suggested today.

CHAIRPERSON BECKER: And Dr. Witten, do you have any comments?

DR. WITTEN: No. I'd just like to thank the panel for their work today, as well as the FDA review team and the sponsor, too.

CHAIRPERSON BECKER: So I guess this then concludes the 18th meeting of the Neurological Devices Panel.

Thank you very much.

(Whereupon, at 4:42 p.m., the panel meeting was concluded.)