1

 

                DEPARTMENT OF HEALTH AND HUMAN SERVICES

 

                      FOOD AND DRUG ADMINISTRATION

 

                CENTER FOR DEVICES AND RADIOLOGIC HEALTH

 

 

 

                    CIRCULATORY SYSTEM DEVICES PANEL

 

 

 

 

 

 

 

                        Thursday, March 18, 2004

 

                               9:00 a.m.

 

 

 

               Hilton Gaithersburg Washington D.C., North

                           Salons A, B and C

                           620 Perry Parkway

                         Gaithersburg, Maryland

 

                                                                 2

 

                              PARTICIPANTS

 

         Cynthia Tracy, M.D., Acting Chair

         Geretta Wood, Executive Secretary

 

         MEMBERS:

 

         Salim Aziz, M.D.

         Mitchell Krucoff, M.D.

         William Maisel, M.D., MPH

         Christopher J. White, M.D.

 

         CONSULTANTS:

 

         Clyde Yancy, M.D.

         Judah Z. Weinberger, M.D., Ph.D.

         John W. Hirshfeld, M.D.

         Thomas B. Ferguson, M.D.

         Norman S. Kato, M.D.

         Brent Blumenstein, Ph.D.

         Charles Bridges, M.D.

         L. Henry Edmunds, Jr., M.D.

 

         INDUSTRY REPRESENTATIVE:

 

         Michael Morton

 

         CONSUMER REPRESENTATIVE:

 

         Chrissy Wells

 

         FDA:

 

         Bram Zuckerman, M.D.

 

                                                                 3

 

                            C O N T E N T S

 

      Call to Order

                Cynthia Tracy, M.D.                              4

 

      Conflict of Interest Statement

                Geretta Wood                                     4

 

      Introductions                                              6

 

      FDA Presentation:

                Julia Marders                                    8

                Wolf Sapirstein, M.D., MPH, FACS                17

                Kachi Enyinna                                   26

      Open Public Hearing:

 

                Randall Wolfe, M.D.                             31

                Robert W. Emery, M.D.                           41

                G. Phillip Schoettle, M.D.                      50

                Robert Frater, M.D.,                            55

                Michael Mack, M.D., STS/AATS                    66

                Mark Slaughter, M.D.                            76

                Henry Frank Martin, M.D.                        85

                Bernard Hausen, M.D.                            90

                Prof. Uwe Klima                                 96

 

      Open Committee Discussion                                107

 

      Summary

                Cynthia Tracy, M.D.                            262

 

                                                                 4

 

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

 

  2                          Call to Order

 

  3             DR. TRACY:  Good morning, everybody.  I

 

  4   would like to call to order this meeting of the

 

  5   Circulatory System Devices Panel.  The topic today

 

  6   is a discussion of type of data and study required

 

  7   to effectively evaluate performance of aortic

 

  8   anastomotic devices for marketing.

 

  9                       Conflict of Interest

 

 10             MS. WOOD:  The following announcement

 

 11   addresses conflict of interest issues associated

 

 12   with this meeting and is made a part of the record

 

 13   to preclude even the appearance of an impropriety.

 

 14             To determine if any conflict existed, the

 

 15   agency reviewed the submitted agenda and all

 

 16   financial interests reported by the committee

 

 17   participants.  The conflict of interest statutes

 

 18   prohibit special government employees from

 

 19   participating in matters that could affect their or

 

 20   their employers' financial interests.  However, the

 

 21   agency has determined that participation of certain

 

 22   members and consultants, the need for whose

 

 23   services outweighs the potential conflict of

 

 24   interest involved, is in the best interest of the

 

 25   government.

 

                                                                 5

 

  1             A waiver has been granted for Dr. Clyde

 

  2   Yancy and a wavier was previously granted for Dr.

 

  3   Judah Weinberger for their financial interests in

 

  4   firms at issue that could potentially be affected

 

  5   by the panel's recommendations.  The waivers allow

 

  6   these individuals to participate fully in today's

 

  7   deliberations.  Copies of these waivers may be

 

  8   obtained from the agency's Freedom of Information

 

  9   Office, Room 12A-15 of the Parklawn Building.

 

 10             We would like to note for the record that

 

 11   the agency took into consideration other matters

 

 12   regarding Drs. Thomas Ferguson, Mitchell Krucoff,

 

 13   Cynthia Tracy, Judah Weinberger and Clyde Yancy.

 

 14   These panelists reported past or current interests

 

 15   involving firms at issue but in matters that are

 

 16   not related to today's agenda.  The agency has

 

 17   determined, therefore, that these individuals may

 

 18   participate fully in the panel's deliberations.

 

 19             In the event that the discussions involve

 

 20   any other products or firms not already on the

 

 21   agenda for which an FDA participant has a financial

 

 22   interest, the participants should excuse him or

 

 23   herself from such involvement and the exclusion

 

 24   will be noted for the record.

 

 25             With respect to all other participants, we

 

                                                                 6

 

  1   ask in the interest of fairness that all persons

 

  2   making statements or presentations disclose any

 

  3   current or previous financial involvement with any

 

  4   firm whose products they may wish to comment upon.

 

  5             DR. TRACY:  Just before we get started, I

 

  6   would just like to ask everybody to be sure that

 

  7   you are speaking directly into the microphone,

 

  8   including the speakers who will be coming up later

 

  9   in the open public hearing.  A transcript is being

 

 10   made from these presentations today.

 

 11             At this time I would like to ask the panel

 

 12   members to introduce themselves.

 

 13                          Introductions

 

 14             MR. MORTON:  I am Michael Morton.  I am

 

 15   the industry representative.  I am an employee of

 

 16   CarboMedics.

 

 17             DR. WEINBERGER:  Judah Weinberger,

 

 18   Director of Interventional Cardiology at Columbia.

 

 19             DR. YANCY:  Clyde Yancy, Director of Heart

 

 20   Failure and Transplantation at UT Southwestern, in

 

 21   Dallas.

 

 22             DR. WHITE:  Chris White.  I am the Chief

 

 23   of Cardiology at Ochsner Clinic Foundation.

 

 24             DR. HIRSHFELD:  John Hirshfeld.  I am an

 

 25   interventional cardiologist in the University of

 

                                                                 7

 

  1   Pennsylvania.

 

  2             DR. KATO:  Norman Kato, cardiovascular

 

  3   surgeon, private practice, Encino, California.

 

  4             MS. WOOD:  Geretta Wood, Executive

 

  5   Secretary.

 

  6             DR. TRACY:  I am Cindy Tracy,

 

  7   electrophysiologist, George Washington University.

 

  8             DR. EDMUNDS:  I am Hank Edmunds, Professor

 

  9   of Surgery at the University of Pennsylvania.

 

 10             DR. FERGUSON:  Tom Ferguson,

 

 11   cardiovascular surgeon, Washington University St.

 

 12   Louis.

 

 13             DR. KRUCOFF:  Mitch Krucoff,

 

 14   interventional cardiologist, Duke University

 

 15   Medical Center; Director of Devices Trials, Duke

 

 16   Clinical Research Institute.

 

 17             DR. MAISEL:  William Maisel,

 

 18   electrophysiologist, Brigham & Women's Hospital.

 

 19             DR. BLUMENSTEIN:  Brent Blumenstein,

 

 20   biostatistician, Seattle, Washington.

 

 21             DR. BRIDGES:  Charles Bridges, Chief

 

 22   Cardiovascular Surgery Pennsylvania Hospital,

 

 23   University of Pennsylvania.

 

 24             MS. WELLS:  Chrissy Wells.  I am the

 

 25   consumer representative on the panel.

 

                                                                 8

 

  1             DR. ZUCKERMAN:  Bram Zuckerman, Director,

 

  2   FDA, Division of Cardiovascular Devices.

 

  3             DR. TRACY:  Thank you.  At this point we

 

  4   will have the FDA presentation, and Julia Marders,

 

  5   from the Office of Surveillance, will be the

 

  6   opening speaker.

 

  7                         FDA Presentation

 

  8             MS. MARDERS:  Good morning.

 

  9             [Slide]

 

 10             My name is Julia Marders, and I am a nurse

 

 11   analyst in the Division of Postmarket Surveillance,

 

 12   Office of Surveillance and Biometrics.

 

 13             [Slide]

 

 14             I will present an analysis of adverse

 

 15   event reports received by the FDA on aortic

 

 16   anastomotic devices.  My presentation will begin

 

 17   with a brief description of the Medical Device

 

 18   Reporting System, MDR, which is a system for

 

 19   adverse events and product problems, and include a

 

 20   discussion of its limitations.

 

 21             Next, I will describe the database, search

 

 22   methodology used to obtain the reports of aortic

 

 23   anastomotic devices and provide a summary of

 

 24   findings and analysis of conclusions.  Then I will

 

 25   finish the presentation with conclusions,

 

                                                                 9

 

  1   considerations and questions for the panel to

 

  2   contemplate.

 

  3             [Slide]

 

  4             The Medical Device Reporting System is a

 

  5   nationwide passive surveillance system which

 

  6   includes both mandatory and voluntary reporting.

 

  7   Since 1984 manufacturers and importers have been

 

  8   required to submit reports to the FDA of

 

  9   device-related deaths and serious injuries, as well

 

 10   as events involving device malfunctions that may

 

 11   cause or contribute to death or serious injury.

 

 12             The Safe Medical Devices Act of 1990

 

 13   introduced mandatory reporting of device-related

 

 14   deaths and serious injuries by user facilities,

 

 15   most notably hospitals and nursing homes.

 

 16   Voluntary medical device adverse event and problem

 

 17   product reports, most often submitted by healthcare

 

 18   practitioners, consumers, patients or family

 

 19   members, are received through the FDA's MedWatch

 

 20   program.  In general, approximately 95 percent of

 

 21   the reports received by FDA are from manufacturers,

 

 22   one percent from importers, and the remainder is

 

 23   equally split between voluntary and user

 

 24   facilities.

 

 25             [Slide]

 

                                                                10

 

  1             The Medical Device System, which is the

 

  2   MDR system as most people know it, while providing

 

  3   signals of actual and potential device-related

 

  4   problems, has some limitations.  Under-reporting of

 

  5   adverse events to hospitals, manufacturers and the

 

  6   FDA by healthcare practitioners is a well-known and

 

  7   recognized phenomenon.  Thus, events reported

 

  8   through MDR represent a subset of the total

 

  9   occurrence of events.

 

 10             In addition, manufacturers are not

 

 11   required to submit denominator information such as

 

 12   number of devices manufactured, distributed and

 

 13   implanted.  Thus, due to under-reporting and lack

 

 14   of denominator data accurate incidence rates are

 

 15   unable to be determined based on MDR data alone.

 

 16   Furthermore, reports received may not be

 

 17   representative and reflective of a variety of

 

 18   reporting biases.  Thus, for example, reporting may

 

 19   vary by manufacturer or by the presence or absence

 

 20   of publicity.

 

 21             Although there is a regulatory requirement

 

 22   for a minimum data set, event narrative

 

 23   descriptions vary in completeness and complexity.

 

 24   For example, one aortic anastomotic device report

 

 25   indicates failure of the connector as the entire

 

                                                                11

 

  1   event description and no further details are

 

  2   provided.  In addition, many reports do not contain

 

  3   results of manufacturer failure analyses.  Often

 

  4   devices are not returned to the manufacturer for

 

  5   evaluation because they are discarded or remain

 

  6   implanted.  Thus, root causes for reported events

 

  7   are often unable to be determined.

 

  8             [Slide]

 

  9             Now I will describe the search methodology

 

 10   used to obtain the data set of aortic anastomotic

 

 11   device reports in this presentation and present the

 

 12   findings.  First I searched the database by product

 

 13   code.  All medical devices approved or cleared for

 

 14   marketing have a unique three-letter identifier

 

 15   called the product code.  Next, I narrowed the

 

 16   search by date.  This search includes events of

 

 17   aortic anastomotic devices that were reported from

 

 18   May 24, 2001, the first marketing clearance date

 

 19   for these devices, to March 1, 2004.  I also

 

 20   performed additional database queries by brand

 

 21   names to validated that I had captured all aortic

 

 22   anastomotic device reports that have been entered

 

 23   into the database.

 

 24             [Slide]

 

 25             Now the findings, a total of 213 reports

 

                                                                12

 

  1   are in the database and most reports were received

 

  2   in 2003.  The number of death reports is 23;

 

  3   injury, 185; and malfunction, 5.  The vast majority

 

  4   of these reports, that is 203, came from

 

  5   manufacturers, with 2 from user facilities and 8

 

  6   voluntary.

 

  7             [Slide]

 

  8             Patient age was provided in 129 of the 213

 

  9   reports and ranged from ages 35 to 83 years, with

 

 10   most in the 50-65 age range.  Slightly over half of

 

 11   these events are noted in males, a quarter in

 

 12   females and a quarter were gender unspecified.  One

 

 13   hundred and seventy-three events, or 81 percent,

 

 14   occurred with patients in the U.S. and 14, which is

 

 15   7 percent, with patients outside the U.S.; 26 event

 

 16   reports, 12 percent, did not specify whether the

 

 17   event is foreign or domestic.

 

 18             [Slide]

 

 19             Of the 23 death reports, 22 were from

 

 20   manufacturers and one from a user facility.  All

 

 21   patient deaths occurred within 18 days of

 

 22   implantation and 15 of the deaths occurred within 3

 

 23   days of implantation.  Interestingly, one patient

 

 24   actually had both a dissection operatively and a

 

 25   detachment postoperatively and is included in 2 of

 

                                                                13

 

  1   the problem categories listed on this slide.

 

  2   Additionally, another patient had both thrombus and

 

  3   aortic detachment that was discovered on

 

  4   postoperative day 2 when the patient coded.

 

  5             Twelve reports indicate the problem of

 

  6   occlusion or thrombus at the connector site.  One

 

  7   report describes the patient was noted to have a

 

  8   predisposing hypercoagulable state, and 2 reports

 

  9   indicate that patients had atrial fibrillation.

 

 10   Aortic dissection associated with deployment or

 

 11   after the connector is placed was noted in 7

 

 12   reports.  Device detachment, resulting in

 

 13   hemorrhagic shock, occurred in 6 reports.  None of

 

 14   the devices associated with death were returned to

 

 15   the manufacturer for evaluation and the

 

 16   manufacturer has not been able to determine the

 

 17   root cause of the events.

 

 18             [Slide]

 

 19             Now I will present an actual report to

 

 20   illustrate these findings in a more clinically

 

 21   relevant way.  A patient was implanted with an

 

 22   aortic anastomotic device during an off-pump

 

 23   procedure.  No difficulties were encountered with

 

 24   loading or deployment of the device.  Recovery was

 

 25   good for approximately 40 hours when the patient

 

                                                                14

 

  1   suddenly lost consciousness after a dramatic drop

 

  2   in blood pressure.  CPR was initiated and blood

 

  3   appeared in the drains.  At re-operation, the

 

  4   aortic connector was detached from the aorta and

 

  5   the patient died after 10 minutes.  The autopsy

 

  6   revealed the cause of death was hemorrhagic shock.

 

  7             [Slide]

 

  8             A total of 185 injuries were reported.

 

  9   Stenosis and occlusion are overwhelmingly noted to

 

 10   be the first and second most frequently reported

 

 11   problems respectively.  Although infrequently

 

 12   reported, events involving device detachment have

 

 13   also resulted in serious injury.  Clinically, the

 

 14   reported outcomes of stenosis and occlusion

 

 15   resulted in life-threatening conditions resulting

 

 16   in shortness of breath, chest pain, arrhythmias,

 

 17   subsequent myocardial infarction and/or hemodynamic

 

 18   instability requiring either surgical or

 

 19   interventional treatment including catheterization

 

 20   for PTCA and stenting.

 

 21             The time from implantation to injury, as

 

 22   noted in 37 of the 185 reports submitted, of the 30

 

 23   noting stenosis or occlusion most, or about 60

 

 24   percent, occurred within 90 days; 4 events occurred

 

 25   within 4 days.  The other 7 are associated with a

 

                                                                15

 

  1   variety of patient problems other than stenosis or

 

  2   occlusion.  Three reports of device detachment

 

  3   occurred within one day of surgery, and another

 

  4   event atypically occurred after 97 days, possibly

 

  5   due to a fragile aorta and placement of the

 

  6   connector on a pseudoaneurysm.  Of all the

 

  7   injuries, only 2 devices were returned to the

 

  8   manufacturer for evaluation, both of which resulted

 

  9   in manufacture evaluation indicating no device

 

 10   failure detected.

 

 11             [Slide]

 

 12             Five reports indicated a device

 

 13   malfunction.  One report states the device was not

 

 14   able to be used because the anchor tip was closed.

 

 15   Two reports indicate the aortic plug was not seen

 

 16   by the surgeon upon inspection of the device.  Both

 

 17   of these patients have not experienced any adverse

 

 18   consequences.  The fourth report indicates a device

 

 19   malfunction resulting in an aortic laceration

 

 20   requiring repair.  It is not clear why the user

 

 21   facility reported this event as a malfunction

 

 22   rather than an injury, and no information was

 

 23   included about the patient's outcome.  Follow-up is

 

 24   ongoing.  The fifth report indicates failure of the

 

 25   connector, with no other details, other than

 

                                                                16

 

  1   indicating no consequences to the patient.

 

  2             [Slide]

 

  3             Conclusions--the reports of serious

 

  4   adverse outcomes related to aortic anastomotic

 

  5   device use raises a signal of a potential public

 

  6   health problem.  Some of these occurrences are

 

  7   catastrophic, such as aortic dissection or device

 

  8   detachment, and not expected.  Others, for example,

 

  9   occlusion or stenosis, may be expected depending on

 

 10   the patient's underlying condition of adequacy of

 

 11   antiplatelet therapy, or may reflect device-related

 

 12   events, for example, stenosis at the connection

 

 13   site or thrombosis potentially related to

 

 14   bioincompatibility or poor hemodynamics.  Lastly,

 

 15   the reported information to date reflects

 

 16   short-term experience.  Long-term failure

 

 17   information is also important.

 

 18             [Slide]

 

 19             Considerations--additionally, there are

 

 20   two other important points to consider, first,

 

 21   failure analyses of this adverse event data are

 

 22   lacking or limited.  The underlying root cause of

 

 23   these events, particularly occlusion and stenosis,

 

 24   is unknown.  Multiple factors may be involved which

 

 25   can make the evaluation of these events difficult. 

 

                                                                17

 

  1   Second, this adverse event data needs to be

 

  2   factored into the risk/benefit profile for these

 

  3   devices.

 

  4             [Slide]

 

  5             To conclude my presentation, I have the

 

  6   following three questions for the panel to consider

 

  7   that are based on adverse event report findings:

 

  8             First is the question of collection of

 

  9   long-term failure rate data.  Should a longer

 

 10   period of time for manufacturer collection of

 

 11   device performance data post implantation be

 

 12   required to fully understand aortic anastomotic

 

 13   device failures?

 

 14             Next, should studies comparing short- and

 

 15   long-term patient outcomes between standard

 

 16   suturing versus sutureless aortic anastomotic

 

 17   devices to address risk/benefit issues be

 

 18   undertaken?

 

 19             Finally, should further study of

 

 20   device-related events be considered?

 

 21             I encourage the panel to consider these

 

 22   questions before making final recommendations.

 

 23   That concludes my part of the presentation and now

 

 24   I will turn over to Wolf Sapirstein.

 

 25             DR. SAPIRSTEIN:  Dr. Tracy, panel members,

 

                                                                18

 

  1   good morning.

 

  2             [Slide]

 

  3             The people listed up there are members of

 

  4   the Division of Cardiovascular Devices.   My name

 

  5   is Wolf Sapirstein.  We are mandated by statute to

 

  6   regulate cardiovascular devices, and are hopeful

 

  7   that this panel will generate guidance for us in

 

  8   undertaking this activity for these new and unique

 

  9   devices used in treatment of coronary-artery

 

 10   disease.

 

 11             [Slide]

 

 12             Vascular suturing was introduced by Carrel

 

 13   in 1903 and has changed little over the next 100

 

 14   years, except for the replacement of catgut with

 

 15   synthetic suture.  After about 30 years of attempts

 

 16   by various investigators internationally, an

 

 17   automatic device to effect vascular anastomoses,

 

 18   the Symmetry Aortic Connector, was cleared in 2001

 

 19   for commercial use by the agency.  The drive for

 

 20   development of these devices has undoubtedly been

 

 21   coronary arterial bypass graft procedure which also

 

 22   underscores the clinical importance of assuring

 

 23   safety and effectiveness for these devices.

 

 24             Incremental modifications to the coronary

 

 25   arterial bypass graft procedure are seminal to the

 

                                                                19

 

  1   acceleration in development of these devices.  The

 

  2   surgeons among us will have the forbearance, I

 

  3   hope, while I undertake a thumbnail sketch of the

 

  4   changes that have taken place in the performance of

 

  5   the coronary arterial bypass grafting procedure.

 

  6             [Slide]

 

  7             The CABG procedure was the earliest

 

  8   surgical therapy validated with a randomized,

 

  9   controlled trial, the Coronary Arterial Surgery

 

 10   Study.  Autogenous venous conduits remain

 

 11   extensively employed with anastomosis performed to

 

 12   the aorta and the coronary artery distal to the

 

 13   obstructive lesion.  Induced ventricular

 

 14   fibrillation and anoxic cardiac arrest with

 

 15   hypothermic protection were initially used to

 

 16   provide the quiet field demanded by the challenge

 

 17   of suturing vessels 1-2 mm in diameter.

 

 18   Cardioplegia inducing perfusion of the coronary bed

 

 19   has since produced cardiac standstill with improved

 

 20   myocardial preservation during the ischemic period

 

 21   of conduit anastomosis.

 

 22             Resistance of the internal thoracic

 

 23   artery, ITA, to the atherosclerotic degeneration

 

 24   that seemed inexorable with vein conduits has led

 

 25   to is preferential employment since the late 1980s.

 

                                                                20

 

  1   This also provides the advantage of eliminating

 

  2   need for an aortic anastomosis.  Patient survival

 

  3   has since been shown in several studies to be

 

  4   closely related to the effectiveness of

 

  5   revascularization achieved for the anterior surface

 

  6   of the heart and left ventricle.  These

 

  7   developments, patency of the ITA and anterior

 

  8   cardiac revascularization justified introduction of

 

  9   the minimal access direct CABG procedure in the

 

 10   1990s to perform an isolated LIMA-LAD bypass.  This

 

 11   was shortly followed by beating heart and finally

 

 12   off-pump CABG with elimination of extracorporeal

 

 13   circulatory support entirely.  Thus, were the ill

 

 14   effects of cardiac arrest, extracorporeal

 

 15   circulatory perfusion and aortic clamp manipulation

 

 16   of the aorta obviated.

 

 17             [Slide]

 

 18             These modifications made to the CABG

 

 19   procedure addressed its changing role in an era of

 

 20   increasing catheter-mediated coronary treatment.

 

 21   The MIDCAB is seen as reducing the morbidity of

 

 22   incisional trauma, particularly in an increasingly

 

 23   older patient cohort and patients with more

 

 24   compromised coronary circulation not amenable to

 

 25   percutaneous coronary interventions, and these

 

                                                                21

 

  1   patients become candidates for operative

 

  2   intervention.  Dispensing with cardiopulmonary

 

  3   bypass eliminated a potent activator of both the

 

  4   systemic inflammatory response and the various

 

  5   immunological cascades.

 

  6             There is also increasing recognition of

 

  7   the frequency with which neurocognitive

 

  8   deterioration, apart from the more overt cerebral

 

  9   ischemic events, occur with CABG procedures.

 

 10   Extracorporeal cardiopulmonary bypass and

 

 11   manipulation of the atherosclerotic aorta for

 

 12   cardiopulmonary bypass perfusion, as well as

 

 13   conduit anastomosis, have been indicted as

 

 14   etiologic factors for these complications.

 

 15   Anastomotic devices, by facilitating the various

 

 16   modifications to the CABG that address morbidity,

 

 17   can certainly play a major role in reducing this

 

 18   illness.

 

 19             [Slide]

 

 20             Several studies during the development

 

 21   stage of CABG evaluated the effectiveness of this

 

 22   revascularization procedure measured as durability

 

 23   of patency.  While this slide presents a generally

 

 24   accepted distillation of these study findings, it

 

 25   should be noted that patency of CABG is dependent

 

                                                                22

 

  1   on multifactorial elements that have likely been

 

  2   affected by recent changes to the operation itself

 

  3   that are still being evaluated, and by new measures

 

  4   to inhibit the progression of coronary-artery

 

  5   disease.  This has to be considered when evaluating

 

  6   anastomotic devices in a comparison to these

 

  7   conduit patency rates.

 

  8             [Slide]

 

  9             Failure of the CABG conduit has been

 

 10   attributed to several causes which are listed here.

 

 11   They are broadly stratified by the period of their

 

 12   most prominent effects: the perioperative failures;

 

 13   6 months o 1 year, failure due to neointimal

 

 14   hyperplasia; and the continuum from 6 months on are

 

 15   both coronary-artery disease in the native vessel

 

 16   and the conduit itself.

 

 17             [Slide]

 

 18             The advent of anastomotic devices carry a

 

 19   promise for significant benefits in the performance

 

 20   of the CABG procedure that go beyond simplifying

 

 21   procedural mechanics for the benefit of the

 

 22   operator.  They have the potential for eliminating

 

 23   many of the factors contributing to poor patient

 

 24   outcome.  It must be recognized that while the

 

 25   precise benefit perceived for some of these recent

 

                                                                23

 

  1   changes to the procedures, such as beating heart

 

  2   and operations performed without cardiopulmonary

 

  3   bypass, are as yet unresolved.  The use of

 

  4   non-suture constructed anastomoses will certainly

 

  5   facilitate and increase the frequency of their use.

 

  6   These are some of the benefits that seem intuitive

 

  7   with anastomotic devices.

 

  8             [Slide]

 

  9             Well, Woody Allen has said every silver

 

 10   lining has a dark cloud, and this is exactly true

 

 11   with these anastomotic devices.  Here are listed

 

 12   some of the design characteristics that may

 

 13   contribute to graft failure which do not obtain

 

 14   with conventional sutured vascular connections.

 

 15             [Slide]

 

 16             In our evaluation of these devices for

 

 17   clearance with a 510(k) notification, we have

 

 18   required extensive preclinical data to support

 

 19   limited clinical studies.  The clinical material

 

 20   was required to substantiate equivalence to

 

 21   historical data for conduit patency, which was a

 

 22   surrogate for correcting the deficiency in

 

 23   myocardial perfusion.

 

 24             We encountered some disagreement regarding

 

 25   the study design, the duration of follow-up, and

 

                                                                24

 

  1   the instruments for assessing effectiveness.  While

 

  2   general agreement exists regarding the use of

 

  3   suture anastomosis as the gold standard to control

 

  4   for patency, there is considerable advocacy to

 

  5   employ measures of coronary perfusion for

 

  6   assessment of patency.  This is a reversal of the

 

  7   original CABG use of patency as a surrogate for

 

  8   perfusion.

 

  9             With regard to duration of follow-up, the

 

 10   initial concept was to take into consideration the

 

 11   multifactorial causes of CABG failure by accepting

 

 12   a relatively short period, such as 6-9 months, that

 

 13   focuses on the adequacy of the anastomosis

 

 14   constructed rather than the other factors in graft

 

 15   failure.  The changes made to the CABG procedure

 

 16   itself and the introduction of measures aimed at

 

 17   disease progression were not addressed.  It was

 

 18   also felt that a distinction could be made for

 

 19   devices used on the proximal aortic or on the

 

 20   distal coronary artery.

 

 21             [Slide]

 

 22             The problem encountered in designing a

 

 23   study to evaluate these anastomosis devices goes

 

 24   beyond the inherent problem of the multifactorial

 

 25   causes of CABG failure.  They involve in general

 

                                                                25

 

  1   the device-specific variables listed here that may

 

  2   frustrate attempts at one-design-fits-all study

 

  3   design for the devices.

 

  4             [Slide]

 

  5             From our initial experience with cleared

 

  6   devices, we now have the belief that the rigor of a

 

  7   randomized trials may be required unless there are

 

  8   very mitigating circumstances to justify otherwise.

 

  9   To this end, we would like input on an appropriate

 

 10   template for study design that could be modified to

 

 11   accommodate some of the variables intrinsic to

 

 12   their use.  This slide lists some of the

 

 13   considerations we have encountered for designing a

 

 14   study template and it is just put up for your

 

 15   consideration as a straw man.

 

 16             [Slide]

 

 17             This slide represents a sample size

 

 18   estimation for a one-armed study with the endpoint

 

 19   for effectiveness based on the historical values

 

 20   listed here for conduits performed with hand

 

 21   suturing.  For instance, a point estimate of 95

 

 22   percent patency, with a lower confidence level

 

 23   accepted as 5 percent, would require a sample size

 

 24   of 150 patients for study.  This is just placed

 

 25   here for your consideration or evaluation for even

 

                                                                26

 

  1   a one-armed study.

 

  2             This completes my introduction to the

 

  3   FDA's request for this panel's input in formulating

 

  4   an appropriate regulatory approach for devices that

 

  5   present the potential for critically affecting the

 

  6   treatment of coronary arterial disease, which is

 

  7   the wound stripe of modern society.

 

  8             Kachi Enyinna, our lead engineer reviewer

 

  9   for these devices, will now present or crystallize

 

 10   some of the comments that I have made in the form

 

 11   of questions that we would like this panel to

 

 12   address in helping us wrestle with the regulation

 

 13   of these devices.  Thank you very much.

 

 14             MR. ENYINNA:  Good morning.  My name is

 

 15   Kachi Enyinna, biomedical engineer and lead

 

 16   reviewer, Division of Cardiovascular Devices.  I

 

 17   will be presenting the questions we have come up

 

 18   with and seeking some kind of guidance from panel

 

 19   on how to evaluate clinical studies of these

 

 20   devices.  I would like to remind the panel members

 

 21   to keep these questions in mind while I go over the

 

 22   questions and to keep the questions in mind until

 

 23   discussion time this afternoon allow members of the

 

 24   medical community, as well as sponsors and industry

 

 25   to speak before we discuss the questions.

 

                                                                27

 

  1             [Slide]

 

  2             Regarding trial design, the first

 

  3   question, please comment on the choice of control

 

  4   in the clinical trial required to evaluate vascular

 

  5   anastomosis devices for CABG.  The gold standard of

 

  6   sutured CABG anastomoses has a well-documented

 

  7   history of over thirty years.

 

  8             [Slide]

 

  9             Can historical data from sutured CABG

 

 10   anastomosis device trials be used as the control in

 

 11   the device studies?

 

 12             [Slide]

 

 13             Alternatively, are concurrently performed

 

 14   CABG controls necessary given the multifactorial

 

 15   causes of CABG failure, for example, technical

 

 16   construction, extent and progression of native

 

 17   vessel disease, condition of conduit and

 

 18   progression of intima hyperplastic and atheromatous

 

 19   degeneration, and the introduction of drugs for

 

 20   mitigation of atherosclerotic disease?

 

 21             [Slide]

 

 22             If these trial designs are inadequate,

 

 23   should randomized, controlled clinical trials be

 

 24   performed?

 

 25             [Slide]

 

                                                                28

 

  1             With regard to device placement and device

 

  2   design, please address the following:  Given the

 

  3   considerable differences between the proximal and

 

  4   distal CABG anastomoses, what, if any, differences

 

  5   in study criteria should be required?

 

  6             [Slide]

 

  7             Are there certain aspects of the clinical

 

  8   study design, for example length of follow-up and

 

  9   endpoints, that should be required for all devices

 

 10   irrespective of device form and function?  For

 

 11   example, the U-clip performance closely duplicates

 

 12   that of a suture, whereas the Symmetry has greater

 

 13   similarity to a stent.

 

 14             It is rarely possible to determine the

 

 15   cause of conduit failure.  Can you suggest criteria

 

 16   to determine whether failure is device related?

 

 17             [Slide]

 

 18             Number three, do you believe that the

 

 19   significant differences between an arterial conduit

 

 20   and a venous conduit warrant distinct study

 

 21   criteria and assessment for each?  If so, please

 

 22   identify these criteria and analyses.

 

 23             [Slide]

 

 24             Four, should the primary effectiveness

 

 25   endpoint be graft patency alone, or include both

 

                                                                29

 

  1   graft patency and myocardial perfusion?

 

  2             Five, with regard to device safety, what

 

  3   criteria, that is, acceptable adverse event rates

 

  4   as compared to that for suture should be applied to

 

  5   the evaluation of device safety as distinguished

 

  6   from device effectiveness?  For example, myocardial

 

  7   infarction, reoperations, neurologic events,

 

  8   incidence of aortic complications.

 

  9             [Slide]

 

 10             Regarding endpoint evaluation, number six,

 

 11   with regard to appropriate patient follow-up, in

 

 12   view of the possible persisting risk of failure of

 

 13   some mechanical anastomosis sites, distinct from

 

 14   the progression of native vessel disease, what

 

 15   duration of follow-up is advisable for premarket

 

 16   evaluation?

 

 17             [Slide]

 

 18             Should postmarket follow-up be required to

 

 19   assess long-term device effectiveness?  If so,

 

 20   please define the appropriate length of follow-up

 

 21   after primary patency evaluation.

 

 22             [Slide]

 

 23             The last question, can non-invasive

 

 24   measuring instruments, for example,

 

 25   echocardiography, ultrafast spiral CT, MRA, EBT,

 

                                                                30

 

  1   etc., be used for primary assessment of graft

 

  2   patency or is angiographic follow-up necessary?  At

 

  3   what time points should patency be assessed?  Thank

 

  4   you.

 

  5             DR. TRACY:  Does that conclude the FDA

 

  6   presentation?  Does anybody on the panel have a

 

  7   question for the FDA at this point?

 

  8             [No response]

 

  9             At this point, we will move on to the open

 

 10   public hearing.  Both the Food and Drug

 

 11   Administration and the public believe in a

 

 12   transparent process for information gathering and

 

 13   decision-making.  To ensure such transparency at

 

 14   the open public hearing session of the advisory

 

 15   committee meeting, FDA believes that it is

 

 16   important to understand the context of an

 

 17   individual's presentation.  For this reason, FDA

 

 18   encourages you, the open public hearing speaker, at

 

 19   the beginning of your written or oral statement to

 

 20   advise the committee of any financial relationship

 

 21   that you may have with the sponsor, its product

 

 22   and, if known, its direct competitors.  For

 

 23   example, this financial information may include the

 

 24   sponsor's payment of your travel, lodging or other

 

 25   expenses in connection with your attendance at this

 

                                                                31

 

  1   meeting.  Likewise, FDA encourages you at the

 

  2   beginning of your statement to advise the committee

 

  3   if you do not have any such financial

 

  4   relationships.  If you choose not to address this

 

  5   issue of financial relationships at the beginning

 

  6   of your statement it will not preclude you from

 

  7   speaking.

 

  8             MS. WOOD:  I have just a couple of

 

  9   announcements for the open public speakers.  We

 

 10   have asked today, due to the number of speakers

 

 11   that have requested time, that you limit your

 

 12   remarks to five minutes each.  I would also ask

 

 13   that you provide me with either an electronic copy

 

 14   or a hard copy of your presentation for the benefit

 

 15   of the summary writer and the transcriptionist.  If

 

 16   you could see me at lunchtime, that would be great.

 

 17   Thank you.

 

 18             DR. TRACY:  There are a number of speakers

 

 19   and I will call them in order.  The first is Dr.

 

 20   Randall Wolfe, from University of Cincinnati.

 

 21                       Open Public Hearing

 

 22             DR. WOLFE:  Members of the panel, ladies

 

 23   and gentlemen, good morning.

 

 24             [Slide]

 

 25             Thank you for honoring my request to speak

 

                                                                32

 

  1   before you.  My disclosure is that I was the

 

  2   principal investigator on the multicenter U-clip

 

  3   distal anastomotic trial.  Those results were

 

  4   presented at AATS two years ago.  There is no

 

  5   financial relationship.

 

  6             I was a past consultant for Ethicon in

 

  7   laboratory and clinical evaluation of proximal and

 

  8   distal anastomotic devices, and in the past was a

 

  9   consultant to Ventrica in helping set up their

 

 10   clinical distal anastomotic connector trial.

 

 11             I am currently on the steering committee

 

 12   of the Prevent IV Core Gentech E2F Decoy trial

 

 13   which uses synthetic DNA to prevent aortic coronary

 

 14   venous graft atherosclerosis.  That study is closed

 

 15   with over 3,000 patients enrolled.  I mention that

 

 16   because I think we are going to be educated on true

 

 17   graft patency of the results of that trial which

 

 18   will be opened first quarter of next year.

 

 19             [Slide]

 

 20             My primary interest is that I have been

 

 21   presenting summary of anastomotic devices at our

 

 22   national meetings, both AATS, STS and ISMICS.  In

 

 23   the next five minutes I would like to summarize

 

 24   some of the things that have been presented at

 

 25   these meetings.

 

                                                                33

 

  1             [Slide]

 

  2             Overall, there are a lot of anastomotic

 

  3   connector devices, and this shows a convenient way

 

  4   to classify these into proximal and distal and

 

  5   subsequently into automated versus manual.  There

 

  6   are 13 to 15 different devices in these different

 

  7   categories but I find this a convenient way to look

 

  8   at connectors.

 

  9             [Slide]

 

 10             There are different value propositions

 

 11   with the connectors and they range from traditional

 

 12   CABG all the way to total endoscopic CABG.  I don't

 

 13   have time to go over this in detail but only to

 

 14   point out that there is a possibility of

 

 15   eliminating the heart-lung machine by using certain

 

 16   connectors and also reducing ischemic time.  In the

 

 17   endoscopic evaluation there is a potential to

 

 18   reduce patient pain and trauma and to truly enable

 

 19   endoscopic surgery.

 

 20             [Slide]

 

 21             This is a summary that you will probably

 

 22   hear more about from other presenters, but vein

 

 23   graft failures could be a bad vein; the vein could

 

 24   be too long; it could be too short; there could be

 

 25   a poor run-off bed; or it could be a distal or

 

                                                                34

 

  1   proximal anastomotic problem.

 

  2             [Slide]

 

  3             I think this is an important slide.  This

 

  4   is some of the science and this is based on some of

 

  5   the work of the E2F Decoy trial but there is an

 

  6   initial wave of inflammation in a venous graft.

 

  7   There is injury.  There is activation of smooth

 

  8   muscle cells.  There is migration proliferation and

 

  9   intimal soil, if you will, for atherosclerotic

 

 10   plaque and ultimately accelerated atherosclerosis.

 

 11   However, this initial wave is in the first two

 

 12   weeks after the venous graft has been harvested

 

 13   from the leg and placed on the heart.

 

 14             [Slide]

 

 15             This is a summary of how I look at graft

 

 16   failure.  I divided it into three distinct

 

 17   categories.  The first is immediate, that is a

 

 18   technical graft failure.  These are all venous

 

 19   grafts, by the way; it could be arterial as well.

 

 20   Technical failure would be identified in the first

 

 21   week.  In other words, if one obtained a

 

 22   postoperative coronary angiogram in a patient in

 

 23   one week technical failures would be disclosed.

 

 24             The next is intermediate, and this is what

 

 25   I relate to devices.  This is usually in the first

 

                                                                35

 

  1   six to eight weeks.  So, a six-month angiographic

 

  2   evaluation should pick up device failures.

 

  3             The third is chronic and this relates to

 

  4   accelerated atherosclerosis and this really takes

 

  5   years.  In the E2F Decoy trial we are looking at

 

  6   one year but, in fact, it probably occurs over five

 

  7   years.  In my opinion, if the St. Jude device had

 

  8   been evaluated at six months by angiography

 

  9   stenoses and occlusions would have been discovered

 

 10   that related to the device.  In other words, the

 

 11   intermediate category.

 

 12             [Slide]

 

 13             We now have second generation anastomotic

 

 14   devices.  They have proven to be more reliable than

 

 15   hand sewn.  There is a consistent orifice size.

 

 16   They are easier to use.  I think, importantly,

 

 17   another change that has happened with the second

 

 18   generation is a lack of vein manipulation.  So,

 

 19   these should be evaluated with six-month

 

 20   angiographic equivalency and we should also look at

 

 21   performance outcomes.

 

 22             [Slide]

 

 23             In summary, I believe the science supports

 

 24   six months angiographic data for the intermediate

 

 25   or device failure area.  Proximal stainless steel

 

                                                                36

 

  1   devices have demonstrated excellent patency, which

 

  2   will probably be discussed.  And, second generation

 

  3   distal devices demonstrate excellent patency.  I

 

  4   think we have to keep in mind as we think about

 

  5   this is that unlike stents for coronary-artery

 

  6   disease, these devices do not rearrange plaque

 

  7   morphology.  Thank you.

 

  8             DR. TRACY:  Thank you.  Are there any

 

  9   brief questions for Dr. Wolfe from the panel?

 

 10             DR. EDMUNDS:  What data do you have for

 

 11   that last statement?

 

 12             DR. WOLFE:  Which part of it?

 

 13             DR. EDMUNDS:  The last statement, how do

 

 14   you know that the device doesn't rearrange plaque?

 

 15   I mean I don't know.  I would just be interested in

 

 16   your data.

 

 17             DR. WOLFE:  The last statement is

 

 18   concerning distal devices.  This is assuming that

 

 19   the device is placed to a target site that is

 

 20   relatively free of atherosclerotic debris.  The

 

 21   second bullet point is for the proximal devices.

 

 22   The third bullet point is specifically for distal.

 

 23             DR. EDMUNDS:  That is what I am talking

 

 24   about.  Are you talking about magnetic coupling?

 

 25             DR. WOLFE:  Any type.  What I am trying to

 

                                                                37

 

  1   relate is that stents and anastomotic devices are

 

  2   not equal in that a stent is supposed to rearrange

 

  3   plaque to open up a stenosis.  For devices that we

 

  4   are using that is not their purpose.  We are not

 

  5   rearranging the plaque.  We are connecting,

 

  6   hopefully, a fairly normal vein or artery to a

 

  7   fairly normal coronary distal target.

 

  8             DR. BRIDGES:  I have a question about the

 

  9   second bullet point.  Can you also inform us what

 

 10   data that is based on?

 

 11             DR. WOLFE:  I think that will be presented

 

 12   by others, but I believe that the difference is

 

 13   that stainless steel is stronger, and in a proximal

 

 14   position where there is atherosclerotic disease a

 

 15   stainless steel device can actually hold the aorta

 

 16   open, whereas a nitinol device may not; it may

 

 17   buckle and close.  So, it is really the strength of

 

 18   the material.  The proximals are different from the

 

 19   distals.  In fact, the people that may need the

 

 20   proximal devices the most are the ones who have the

 

 21   worst aortas.  They have disease in a situation

 

 22   where it is maybe not safe to clamp the aorta.

 

 23             DR. AZIZ:  With the proximal devices, if

 

 24   you do get narrowing, how do you propose that be

 

 25   handled?  Let's say in six months you find that you

 

                                                                38

 

  1   have osteal narrowing, how would you handle that?

 

  2             DR. WOLFE:  I don't know the answer to

 

  3   that.

 

  4             DR. AZIZ:  Can they be dilated in the cath

 

  5   lab?

 

  6             DR. WOLFE:  I don't know the answer to

 

  7   that.

 

  8             DR. AZIZ:  And is there intimal

 

  9   hyperplasia that you are seeing, if you do see it?

 

 10             DR. WOLFE:  I believe so with the second

 

 11   generation devices.  With the first generation

 

 12   devices I think it was a more complicated situation

 

 13   where the graft could actually embrocate over the

 

 14   device.  But in the second generation devices it

 

 15   should be more related to disease in the aorta.

 

 16   However, if a large lumen is maintained then there

 

 17   shouldn't be significant stenosis.  So, let's say

 

 18   you get neointimal hyperplasia in every graft,

 

 19   let's say you get a millimeter in every

 

 20   graft--well, if you get a 1.5 mm opening, that is

 

 21   significant.  If you get a 3 mm opening that is

 

 22   maintained, it won't be significant.

 

 23             DR. AZIZ:  Let me ask you one other thing,

 

 24   with the proximal anastomotic devices, the angle

 

 25   that the graft comes off is really at right angles

 

                                                                39

 

  1   to the aorta.  Right?

 

  2             DR. WOLFE:  In some of the products, that

 

  3   is true.

 

  4             DR. AZIZ:  You mean there are ones where

 

  5   you can have it coming off as a cobra head?

 

  6             DR. WOLFE:  That is correct.

 

  7             DR. KRUCOFF:  Have you actually retrieved

 

  8   any of these devices and looked at them under a

 

  9   microscope when they have failed?

 

 10             DR. WOLFE:  I have not--well, I have seen

 

 11   the slides, I certainly have.

 

 12             DR. KRUCOFF:  Whose slides are those?

 

 13             DR. WOLFE:  St. Jude.  I did go over those

 

 14   at one point and, again, that is a first generation

 

 15   device and I believe the mode of failure of that is

 

 16   different from anything you might see in the

 

 17   future.  It is multifactorial but the occlusions

 

 18   tend to be flush with the aorta.  There is

 

 19   neointimal hyperplasia; there is thrombus.  First

 

 20   of all, the angiogram does not look like a typical

 

 21   angiogram that you might see with an occluded vein

 

 22   graft; it is completely different.  There is also

 

 23   the possibility that the vein graft itself has

 

 24   changed its position on the connector.  In

 

 25   addition, that was a connector that had a high

 

                                                                40

 

  1   profile.  There is also the possibility that there

 

  2   could be a right angle kink right at the end of the

 

  3   connector.

 

  4             I think in summary, I give credit to the

 

  5   pioneers for being the first ones out there.  The

 

  6   first eight patients who received a mitral valve

 

  7   replacement all died.  Fortunately, we still do

 

  8   mitral valve replacements and maybe with the first

 

  9   generation connectors we are seeing some of the

 

 10   same things, some of the mistakes.  I think many of

 

 11   those have been changed by changes in device and

 

 12   changes in material.

 

 13             DR. YANCY:  As you have worked through

 

 14   your clinical trials with these devices, have there

 

 15   been concomitant improvements or changes in medical

 

 16   management because of anticipated problems with

 

 17   these connectors vis-avis antiplatelet therapy,

 

 18   anticoagulation, aspirin, etc.?

 

 19             DR. WOLFE:  We do have some data from the

 

 20   E2F Decoy trial.  The trial has not opened but we

 

 21   have some demographic data.  It has been shown that

 

 22   when patients are followed more closely the chances

 

 23   of them going home on antiplatelet agents are much

 

 24   higher.  Although most surgeons say that they send

 

 25   their patients home on aspirin or some antiplatelet

 

                                                                41

 

  1   agent, in fact, many patients do not go home on

 

  2   that but in a careful study situation they do.

 

  3   There is a study bias.

 

  4             DR. YANCY:  So, those anticipated events

 

  5   that you thought would be predicted or captured at

 

  6   six months, do you think they are product failures,

 

  7   medical management failures or both?

 

  8             DR. WOLFE:  I expect they are product

 

  9   failures and they probably would be in an extreme

 

 10   environment such as a very atherosclerotic aorta,

 

 11   but I am not sure.  I am not sure.

 

 12             DR. TRACY:  I think we are going to have

 

 13   to move on at this point.  There is a number of

 

 14   other speakers.  Thanks very much.  Dr. Robert

 

 15   Emery?

 

 16             DR. EMERY:  While we are setting up my

 

 17   disc here, I am Robert Emery.  I am in private

 

 18   practice in Minneapolis-St. Paul, Minnesota.  I am

 

 19   not being sponsored by any companies but I have had

 

 20   relationships in terms of research grants by St.

 

 21   Jude Medical, ATS Medical, AtriCure, Congestive

 

 22   Heart Failure Solutions.  I have been on research

 

 23   advisory boards for St. Jude Medical, Medtronic,

 

 24   Myocor, Percardia, CardioGenesis, Inc.; data safety

 

 25   monitoring boards for Cardioblate and for Myocor,

 

                                                                42

 

  1   and I have received speaking fees for several of

 

  2   the aforementioned companies.

 

  3             [Slide]

 

  4             I would like to address our early

 

  5   experience in the Minneapolis-St. Paul area looking

 

  6   at why vein grafts fail, the new issues with aortic

 

  7   connectors.  We have been through the etiology of

 

  8   graft failures so I won't go into that, however,

 

  9   there are several new issues that are introduced by

 

 10   the currently used generation of connectors.  There

 

 11   can be overloading of the connector, that is, too

 

 12   much vein graft placed below the prongs;

 

 13   double-loading of the connector like putting on

 

 14   your socks where you can invert the graft and load

 

 15   that which inhibits flow through the graft.  You

 

 16   can skive the aortic punch and that make take out a

 

 17   complete circle.

 

 18             There are variations in operative

 

 19   technique.  For instance, performing your proximals

 

 20   first, as most surgical trainees in the United

 

 21   States perform distals first you are radically

 

 22   changing the way we have been trained in our

 

 23   everyday use in conduct of the operation.  Grafts

 

 24   can move.  After the patient is closed the lungs

 

 25   can push the grafts to various positions and this

 

                                                                43

 

  1   can cause loss of the 90 degree angle, that has

 

  2   been mentioned here, that is necessary for the

 

  3   current generation St. Jude connector.

 

  4             [Slide]

 

  5             Let's look at some of these issues that we

 

  6   have seen.  Here is a surgical technical error at

 

  7   the distal anastomosis that would lead to graft

 

  8   failure if not completed.  I don't think that could

 

  9   be blamed on the connector but a connector was

 

 10   utilized.

 

 11             [Slide]

 

 12             This is the first case I performed in the

 

 13   United States, the second one done in the United

 

 14   States after FDA approval.  You can see two

 

 15   technical errors here that I learned over time and

 

 16   if I had not changed my operative technique one

 

 17   would have a consistent mode of failure that would

 

 18   be uncorrected.  That is, these grafts are placed

 

 19   on top of the aorta instead of further down the

 

 20   side toward the pulmonary artery, therefore,

 

 21   maintaining a 90 degree angle.  The grafts are also

 

 22   reflected superiorly with some kinking at the

 

 23   anastomotic site, not maintaining that 90 degree

 

 24   angle.  As I mentioned, these grafts can move.  All

 

 25   grafts should be tacked to keep that important 90

 

                                                                44

 

  1   degree angle.  If you lose that you can predict

 

  2   some degree of graft failure.

 

  3             [Slide]

 

  4             There can be poor run-off, as shown on

 

  5   this slide, to a patent vein graft but a poor

 

  6   distal vessel.

 

  7             [Slide]

 

  8             Another example is shown here.  The graft

 

  9   can be too short, as mentioned.  Again, it may be a

 

 10   variation in operative technique.

 

 11             [Slide]

 

 12             Here a graft is tethered across the

 

 13   pulmonary artery and you can see the narrowing

 

 14   several centimeters distant from the connector

 

 15   device.

 

 16             [Slide]

 

 17             And a similar vein here wrapped around the

 

 18   pulmonary artery more tightly than one would like

 

 19   to see.

 

 20             [Slide]

 

 21             Improper placement of the graft is also

 

 22   important.

 

 23             [Slide]

 

 24             Here is a vein graft that was placed on

 

 25   the right side of the aorta as we traditionally

 

                                                                45

 

  1   place our saphenous vein grafts when we suture

 

  2   them, rather on the anterior surface of the aorta,

 

  3   riding over the right ventricular outflow tract

 

  4   maintaining the 90 degree angle.  You can see the

 

  5   acute bend on the right side as this graft reflects

 

  6   against the patient's pleural surface.

 

  7             [Slide]

 

  8             Aortic disease was mentioned and this can

 

  9   be important.  Here is an occluded connector in a

 

 10   diffusely diseased aorta and you can see, as Dr.

 

 11   Wolfe mentioned, the flush occlusion of the aorta.

 

 12             [Slide]

 

 13             A combination of factors--here is a small

 

 14   vein graft and poor run-off.

 

 15             [Slide]

 

 16             And here is a very small vein graft that

 

 17   has become atretic over time to a small distal

 

 18   vessel, still patent through the connector but,

 

 19   nonetheless, narrowed.

 

 20             [Slide]

 

 21             Then there is the unknown.  Here is the

 

 22   occluded connector again flush at the angiographic

 

 23   site.

 

 24             [Slide]

 

 25             Here is an occluded vein graft marked by

 

                                                                46

 

  1   the stainless steel ring in the same patient.  You

 

  2   can see the connector graft slightly to the left

 

  3   and one or two centimeters down in this example.

 

  4   There are connector related issues that are key.

 

  5             [Slide]

 

  6             This is what was addressed a little bit in

 

  7   the prior question, proximal anastomotic problems

 

  8   in the face of appropriate graft and appropriate

 

  9   distal connectors that need to be investigated.

 

 10             [Slide]

 

 11             Yet, there are technical issues.  Here is

 

 12   another proximal connector with a very good vein

 

 13   graft and a large distal run-off system.

 

 14             [Slide]

 

 15             Improved and more extensive training may

 

 16   obviate several of the modes of failure that we

 

 17   have seen.  We need to develop indications and

 

 18   contraindications for the use of these devices,

 

 19   particularly as they come out not just general,

 

 20   overall approval.  There are technical

 

 21   considerations that need to be mentioned.  Many

 

 22   modes of failure are unstudied or unconfirmed.

 

 23   Thus, prospective studies are warranted including

 

 24   operative technical detail, both visual, such as

 

 25   the photograph I showed you and verbal operative

 

                                                                47

 

  1   reports, and improved mentoring may be necessary

 

  2   even for devices that seem intuitively simple.

 

  3             [Slide]

 

  4             There are tips for success that I have

 

  5   developed in my own practice based on my

 

  6   experience.

 

  7             [Slide]

 

  8             What we do not want to do is throw the

 

  9   baby out with the bath water because these

 

 10   connecting devices offer us a great opportunity to

 

 11   improve our service to our patients.  Thank you.

 

 12             DR. TRACY:  Thank you.  Are there any

 

 13   brief questions from the panel members?  I do want

 

 14   to remind you that there are a lot of people who

 

 15   want to present today.

 

 16             DR. KRUCOFF:  Just one question.  The

 

 17   angiograms you showed us, were they part of a study

 

 18   protocol that required angiography or were these

 

 19   clinical presentations of people who came back

 

 20   sick?

 

 21             DR. EMERY:  These were clinical

 

 22   presentations in approximately our first eight

 

 23   months of use, from May, 2001 through the first

 

 24   eight months, and we have seen very few since we

 

 25   have modified our surgical techniques.

 

                                                                48

 

  1             DR. KRUCOFF:  And the denominator for

 

  2   these eight months?

 

  3             DR. EMERY:  It was about 160, and these

 

  4   are not all of them.  These are representative

 

  5   samples of technical errors that are correctable

 

  6   with proper training and changing of your

 

  7   techniques as you learn the process.

 

  8             DR. AZIZ:  When you say you tack the

 

  9   grafts, are you putting many anchoring stitches or

 

 10   what do you do?

 

 11             DR. EMERY:  Three or four generally on the

 

 12   left side.  I put one on the pulmonary artery and,

 

 13   again, depending on the length of the graft,

 

 14   because you are doing proximals first with this

 

 15   device, I will connect it so that it won't move

 

 16   with respiration.  On the right side I connect it

 

 17   down the body of the right ventricle as the graft

 

 18   goes directly up from the aorta over the right

 

 19   ventricle and down to the right coronary artery,

 

 20   the posterior descending artery, just some 6-0

 

 21   prolene suture tacking.

 

 22             DR. AZIZ:  We normally do the regular

 

 23   suturing technique; usually you don't have to do

 

 24   that?

 

 25             DR. EMERY:  No, I don't.

 

                                                                49

 

  1             DR. AZIZ:  So, why do you think you need

 

  2   to do it here?

 

  3             DR. EMERY:  Because I learned doing this

 

  4   distal first and I think my measurement of the

 

  5   length of the vein graft to the aorta is better on

 

  6   a distal first process in my hands.  So, sometimes

 

  7   I would rather make my grafts too long than too

 

  8   short because the shortness of the graft may be one

 

  9   reason for disconnection of these connectors from

 

 10   the aorta.  As the pulmonary artery fills, if the

 

 11   graft is too short you can pull these off.  I have

 

 12   pulled them off myself in the operating room by

 

 13   tugging a little bit too hard and I had to put my

 

 14   finger over the hole.  So, a short graft can lead

 

 15   to connector displacement from the aorta,

 

 16   particularly as the patient moves or the heart

 

 17   fills in the postoperative period.

 

 18             DR. TRACY:  Dr. Hirshfeld?

 

 19             DR. HIRSHFELD:  I would just like to say

 

 20   as an angiographer who has probably taken pictures

 

 21   of thousands of bypass grafts, I have heard a lot

 

 22   about considerations that I was never aware of

 

 23   before from your brief presentation, and I think it

 

 24   calls for a sharing of information between

 

 25   angiographers and surgeons about many of these

 

                                                                50

 

  1   technical considerations that affect graft

 

  2   performance.  So, I would hope that out of this

 

  3   will come that kind of sharing of information.

 

  4             DR. EMERY:  I have reviewed all the

 

  5   angiographs of patients that failed in my hands.

 

  6             DR. TRACY:  I think we have to move on; we

 

  7   have a number of speakers.  I am sorry to cut this

 

  8   short; it is very interesting.  To remind you,

 

  9   there will be more time this afternoon to discuss

 

 10   things in detail.  Dr. Schoettle?

 

 11             DR. SCHOETTLE:  Good morning.  My name is

 

 12   Dr. Phillip Schoettle.  I am a thoracic and

 

 13   cardiovascular surgeon in practice at Methodist

 

 14   University Hospital in Memphis, Tennessee.

 

 15             I am here this morning to discuss my

 

 16   experience with the Symmetry proximal anastomotic

 

 17   device.  I would like to disclose at the outset

 

 18   that I have no financial interest in this matter.

 

 19   I paid my own way to Washington, and I am not

 

 20   employed by anyone, nor intend to be employed by

 

 21   anybody with a financial stake in this issue.

 

 22             In September of 2001 I was trained in the

 

 23   use of the Symmetry proximal anastomotic device,

 

 24   along with two of my scrub assistants, by St. Jude

 

 25   Medical.  I was attracted to the device because of

 

                                                                51

 

  1   the reasons mentioned previously which would allow

 

  2   you to do a proximal anastomosis off the aorta

 

  3   without the use of a partial occluding or

 

  4   side-biting clamp with its attendant risk of

 

  5   embolic debris.

 

  6             I rapidly incorporated that device into my

 

  7   practice and used it almost exclusively for the

 

  8   next eleven months.  Initially I was very pleased

 

  9   with the results.  I had minimal, if any, technical

 

 10   issues with the device and was not aware of any

 

 11   acute or subacute saphenous vein closures.

 

 12   Unfortunately, at approximately ten months we began

 

 13   to see almost a deluge of patients returning to the

 

 14   cardiac catheterization laboratory with vein graft

 

 15   occlusions or high grade stenoses invariably

 

 16   occurring in the connector site.

 

 17             This occurrence was totally incompatible

 

 18   with my previous surgical experience.  I reported

 

 19   this to St. Jude Medical and I felt like it

 

 20   warranted a distribution to the surgical community

 

 21   and I began a review of my patients, resulting in

 

 22   the paper that you see here.  This paper was

 

 23   entitled, "Use of an Anastomotic Device in Coronary

 

 24   Bypass Surgery: A Word of Caution."  It was

 

 25   published in the January edition of the Journal of

 

                                                                52

 

  1   Thoracic and Cardiovascular Surgery.

 

  2             Without going into great detail, I would

 

  3   like to summarize the results of that paper.  It

 

  4   was a review of two years of experience.  The first

 

  5   year was the year prior to my beginning to use the

 

  6   device, while proximal saphenous vein connections

 

  7   off the aorta were done in the conventional manner,

 

  8   that is, hand-sewn with a partial occlusion clamp.

 

  9   Beginning in September of 2001, for the next eleven

 

 10   months, comprises the next group of patients where

 

 11   almost exclusively all proximal anastomoses were

 

 12   done with the St. Jude Symmetry anastomotic device.

 

 13             I divided the group in group A and group

 

 14   B.  Group A was the first group, the prior year

 

 15   with hand-sewn anastomoses.  I reviewed all

 

 16   patients who required repeated cardiac

 

 17   catheterization after coronary artery bypass

 

 18   surgery.  What we found was that even though the

 

 19   patients in group A had had a year longer of

 

 20   exposure to my cardiology colleagues, less of those

 

 21   required repeated cardiac catheterization, although

 

 22   that number was not significant between the groups.

 

 23             The number of grafts studied between group

 

 24   A and group B was also similar.  However, what we

 

 25   did find was that the group A patients, those with

 

                                                                53

 

  1   hand-sewn anastomoses, had an 80 percent patency

 

  2   rate of vein grafts studied.  Remember, these were

 

  3   symptomatic, or at least theoretically symptomatic

 

  4   patients.  So, 80 percent of the grafts were widely

 

  5   patent in the hand-sewn anastomoses, with no

 

  6   significant stenoses, and 20 percent were

 

  7   occluding.

 

  8             Unfortunately, in the group B patients,

 

  9   those with the Symmetry proximal anastomotic

 

 10   device, only 20 percent of the grafts studied were

 

 11   patent.  Fully 80 percent of the grafts were either

 

 12   totally occluded or had high grade stenoses

 

 13   uniformly occurring at the connector site.  The

 

 14   significance in p value in favor or patency of the

 

 15   hand-sewn anastomoses was standardly evaluated with

 

 16   a p value of 0.0001.

 

 17             Based on my experience with the Symmetry

 

 18   proximal anastomotic device and review of my own

 

 19   patients, I have several observations and two

 

 20   conclusions I would like to make.  The use of the

 

 21   Symmetry St. Jude proximal anastomotic device in

 

 22   its current generation results in a significantly

 

 23   higher saphenous vein closure and occlusion rate

 

 24   when compared to hand-sewn anastomoses.

 

 25             I do not believe that technical issues are

 

                                                                54

 

  1   the major factor.  I can show you arteriograms of

 

  2   what appear to be perfectly laid out saphenous vein

 

  3   grafts with a 90 degree angle off the aorta, with

 

  4   no kinking, where the stenosis arises immediately

 

  5   in the connector site off the aorta.

 

  6             In two patients that I reoperated, I was

 

  7   able to harvest the segment of aorta with the

 

  8   connector and the saphenous vein.  This was looked

 

  9   at microscopically by the pathologists in my

 

 10   hospital who reported basically a foreign body

 

 11   reaction in the connector site with associated

 

 12   neointimal hyperplasia.

 

 13             I would also point out that these

 

 14   connector stenoses and occlusions are not

 

 15   clinically insignificant.  In this group at least

 

 16   six patients have required early reoperation.

 

 17   Thirty patients, over a year ago, required PCI

 

 18   stents and angioplasty.  There have been four

 

 19   sudden deaths in these patients, two of which were

 

 20   almost certainly related to myocardial infarction.

 

 21             If I can have the liberty of making a

 

 22   conclusion, I see no clinical indication for the

 

 23   current generation of the St. Jude proximal

 

 24   connector.  The use of this connector or any other

 

 25   vascular anastomotic devices must be evaluated by

 

                                                                55

 

  1   scientifically controlled, prospective clinical

 

  2   trials.

 

  3             I do not believe that uneducated surgeons

 

  4   and uninformed patients should be the testing

 

  5   ground for these devices that have not proven to be

 

  6   clinically safe or effective.  I clearly am not

 

  7   opposed to technological advances in coronary

 

  8   bypass surgery.  I have been an early proponent of

 

  9   off-pump surgery and less invasive coronary

 

 10   surgery.  I do not want to throw the baby out with

 

 11   the bath water.  I do not believe, however, that

 

 12   the cause of less invasive coronary artery bypass

 

 13   surgery is furthered by the ill-advised use of

 

 14   these unproven devices.  Thank you.  I would be

 

 15   glad to answer questions if there is time.

 

 16             DR. TRACY:  Any brief questions?  Dr.

 

 17   White?

 

 18             DR. WHITE:  Would you just clarify for me,

 

 19   in the early part of your statement you said

 

 20   something about follow-up at ten months.  Was there

 

 21   a ten-month interval that was special to you?

 

 22             DR. SCHOETTLE:  No, I believe it just

 

 23   would have become apparent to me, you know, with

 

 24   just the overwhelming evidence of patients.  All of

 

 25   a sudden I was getting call after call from these

 

                                                                56

 

  1   patients.

 

  2             DR. WHITE:  Would six months not have

 

  3   identified these patients?  Would a six-month

 

  4   follow-up, do you think, not have been adequate?

 

  5             DR. SCHOETTLE:  I was asked that question

 

  6   last night.  I don't have that answer.  My gut

 

  7   feeling is that six months would probably be

 

  8   appropriate but I don't know that answer based on

 

  9   this review.

 

 10             DR. BRIDGES:  I have a question.  In the

 

 11   brief study that you gave us I didn't see the

 

 12   mortality in the two groups.  You said that there

 

 13   were no sudden deaths in the hand-sewn group but

 

 14   what was the overall mortality in the two groups

 

 15   and are there any updates since this paper was

 

 16   submitted?

 

 17             DR. SCHOETTLE:  The operative mortality

 

 18   was less than three percent but the overall

 

 19   mortality long-term, I don't have that; there have

 

 20   been no updates at this point although I intend to

 

 21   do that.

 

 22             DR. BRIDGES:  But both groups--

 

 23             DR. SCHOETTLE:  They were very similar.

 

 24             DR. BRIDGES:  At least for the graft

 

 25   connector patients, what would be medical therapy

 

                                                                57

 

  1   for these?

 

  2             DR. SCHOETTLE:  It is in the paper, but

 

  3   all patients were discharged on aspirin and all

 

  4   patients were discharged on Plavix for two months.

 

  5             DR. AZIZ:  But when you had to reoperate

 

  6   on them did you have to redo the whole anastomosis

 

  7   or could you immobilize it and rehook it?  How did

 

  8   you do that?

 

  9             DR. SCHOETTLE:  In a couple of cases I was

 

 10   able to continue to use that vein segment.  Conduit

 

 11   length was an issue.  Several of the veins were

 

 12   totally occluded and we just had to sacrifice those

 

 13   veins.

 

 14             DR. AZIZ:  So, the orifice was like the

 

 15   whole length?

 

 16             DR. SCHOETTLE:  That is correct.

 

 17             DR. AZIZ:  So, you probably had intimal

 

 18   hyperplasia proximally and you had full flow and

 

 19   then thrombus--

 

 20             DR. SCHOETTLE:  And then thrombus

 

 21   distally, correct.

 

 22             DR. AZIZ:  Obviously most people don't,

 

 23   and they probably should use some flow techniques

 

 24   to measure flows.

 

 25             DR. SCHOETTLE:  All patients in both

 

                                                                58

 

  1   groups had mediastinal transit time flow evaluation

 

  2   at the time of surgery, and 95 percent of cases

 

  3   were done off-pump.

 

  4             DR. AZIZ:  When you did proximal

 

  5   anastomoses with the device did you do any

 

  6   sequential grafts--

 

  7             DR. SCHOETTLE:  No, they were sequential

 

  8   grafts but I don't have that number available to

 

  9   me.

 

 10             DR. TRACY:  Thank you.

 

 11             DR. SCHOETTLE:  Thanks.

 

 12             DR. TRACY:  Dr. Frater?

 

 13             DR. FRATER:  Let me state immediately I am

 

 14   the Medical Director of St. Jude and, obviously,

 

 15   have that as a conflict of interest.

 

 16             I have a few points to make.  I had

 

 17   expected ten minutes so I am going to try and make

 

 18   them quickly.  I think we can all agree that the

 

 19   MDR system is a warning light that tells us nothing

 

 20   about incidence and, unless we are very lucky,

 

 21   doesn't give us much information on causality, hard

 

 22   as we try to look into every single report that

 

 23   comes in from the field.  I shall not elaborate on

 

 24   that.  I suspect the FDA feels the same about the

 

 25   MDR's utility as we do.

 

                                                                59

 

  1             The question of comparing anastomotic

 

  2   devices to historically published data for sutures

 

  3   is an interesting one.  The data that was obtained

 

  4   in the past has been cardiac surgeons who were

 

  5   trying to find out what they were doing 30 years

 

  6   ago when they were making venous anastomoses.  The

 

  7   patients were younger.  The vessels were better.

 

  8   The extra conditions, such as diabetes, were far

 

  9   less common and it was a different group of

 

 10   patients.  Those patients have long since been

 

 11   captured by the interventionalists and the cardiac

 

 12   surgeon today faces a very different patient.

 

 13             We need to know what the patency rates are

 

 14   today with the current set of patients.  We also

 

 15   need to know what the difference may be between

 

 16   off-pump and on-pump.  There was a paper presented

 

 17   just a few weeks ago at the ACC, the so-called Prog

 

 18   IV Trial, a randomized comparison between off-pump

 

 19   and on-pump surgery with angiography at one year.

 

 20   The patency rate of the cases performed on-pump at

 

 21   one year was 59 percent; the patency rate of those

 

 22   performed off-pump was 49 percent.

 

 23             There is a paper being published in The

 

 24   New England Journal of Medicine by Kahn.  It came

 

 25   out of Britain.  Again, a randomized study of

 

                                                                60

 

  1   on-pump and off-pump anastomoses studied at three

 

  2   months by angiography, which was performed in 80

 

  3   percent of the patients in the trial.  The patency

 

  4   rate of the on-pump cases was exemplary.  At three

 

  5   months they had a 98 percent patency rate but the

 

  6   off-pump cases had an 88 percent patency rate at

 

  7   three months.

 

  8             I present this material, which is clearly

 

  9   important in trying to assess what will be the

 

 10   target of patency that we will be looking at in

 

 11   future trials, and a recognition that times have

 

 12   changed and circumstances are clearly very

 

 13   different.

 

 14             We have done a meta-analysis of some 7,000

 

 15   patients in which angiograms were done between 6

 

 16   and 12 months.  We chose that 6- and 12-month

 

 17   period for the obvious reason that you have already

 

 18   heard today, that after 12 months atherosclerosis

 

 19   dominates the failure of vein grafts.   The mean

 

 20   occlusion rate was 16 percent in this meta-analysis

 

 21   between 6 and 12 months of sewn anastomoses.  But

 

 22   the range was from 9.5 to 26.5.  There is an

 

 23   immense diversity from different institutions and

 

 24   we can speculate forever, certainly not in five

 

 25   minutes, as to what the reasons for those

 

                                                                61

 

  1   differences are.  I am sure that the surgeons in

 

  2   the Prog IV study have not suddenly become

 

  3   incompetent; there are factors that we need to look

 

  4   at.

 

  5             The question of the extent to which

 

  6   clinical utility data is considered to be

 

  7   necessary, I think we have already begun to deal

 

  8   with this.  Six months seems to be a period of time

 

  9   that people are reaching, and that is not

 

 10   unreasonable considering that stents are a Class

 

 11   III device which may or may not be identical to

 

 12   anastomotic devices--that is debatable--are being

 

 13   evaluated with MACE and target vessel interventions

 

 14   at six months.

 

 15             Certainly, it is reasonable to state that

 

 16   it should not be more than 12 months for the

 

 17   obvious reason that by then atherosclerotic disease

 

 18   dominates.  There is intimal damage and technical

 

 19   factors in the first week, neointimal hyperplasia

 

 20   for the next few months, blending finally into

 

 21   atherosclerotic disease.

 

 22             Now, it is essential that the FDA provide

 

 23   clarity on the type of clearance that we need.  If

 

 24   the clinical data requirement reaches the point

 

 25   which would normally be required for a PMA, then it

 

                                                                62

 

  1   should be a PMA.  If you are required to produce

 

  2   the data for a PMA, then the process should be done

 

  3   under a PMA process.  Thank you very much.

 

  4             DR. TRACY:  Any questions?

 

  5             DR. WEINBERGER:  I have a question for

 

  6   you.  You said that you were concerned that the

 

  7   follow-up should be at six months because you

 

  8   thought that atherosclerosis dominates the

 

  9   subsequent natural history of graft failure.  I am

 

 10   concerned because we have a pretty good idea that

 

 11   there is distinct biological heterogeneity in

 

 12   different vascular beds in terms of the kinetics of

 

 13   responses to manipulation.  For instance, we know

 

 14   that for coronary interventions basically at six to

 

 15   nine months the process is over.  But if you look

 

 16   in the periphery, like the iliacs, the usual time

 

 17   is three years.  Do you have any data to suggest

 

 18   that the process to response to injury in vein

 

 19   grafts is over at nine months?

 

 20             DR. FRATER:  Well, if you look at the data

 

 21   from peripheral vascular intervention where it is

 

 22   far easier to follow the patients, it seems fairly

 

 23   definite that while there is an acute phase, which

 

 24   is partly technical and partly because of the

 

 25   damage we do to the vein by the various things we

 

                                                                63

 

  1   do when we take it out and manipulated it, it

 

  2   starts in the first week.  The neointimal

 

  3   hyperplasia seems to blend at 12 months in these

 

  4   peripheral vascular studies with the

 

  5   atherosclerotic process.  It would seem reasonable,

 

  6   if there is an atherosclerotic process taking place

 

  7   in veins after 12 months, not to attribute that to

 

  8   how we handled the vein at the time of the initial

 

  9   anastomosis.

 

 10             DR. WEINBERGER:  Just one follow-up, if

 

 11   there is any kinking in the vein and you have a

 

 12   jet, that jet wouldn't lead to an accelerated

 

 13   atherosclerotic process later on as well?

 

 14             DR. FRATER:  It would happen far quicker

 

 15   than that.  Usually, if you leave a kink in a vein

 

 16   there is a consequence that is soon and definite.

 

 17             DR. WEINBERGER:  Data?

 

 18             DR. FRATER:  Data?  Clinical experience.

 

 19   I am a cardiac surgeon.

 

 20             DR. TRACY:  Dr. Bridges?

 

 21             DR. BRIDGES:  Yes, there are two points.

 

 22   One is to echo Dr. Weinberger's point that I don't

 

 23   think we know exactly.  There is nothing to suggest

 

 24   that there can't be an interaction between

 

 25   mechanical factors and atherosclerosis that extends

 

                                                                64

 

  1   beyond one year.  To say that you can divide these

 

  2   into two discrete processes that are technical,

 

  3   device related and then atherosclerosis I think is

 

  4   unsubstantiated and you can't really defend that.

 

  5             Furthermore, I am sure we are going to

 

  6   hear from Dr. Mack but his own data that was

 

  7   presented at the STS meeting, just in January,

 

  8   showed, at least in his series which I am sure he

 

  9   will comment on, that it was not until you got out

 

 10   beyond one year that you started to see a

 

 11   difference in MACE, that is, you know,

 

 12   cardiovascular events.  So, that, in and of itself,

 

 13   also suggests that the idea of only looking at a

 

 14   one-year or six-month time period is clearly going

 

 15   to result in us missing failures.

 

 16             DR. FRATER:  The obvious issue is how long

 

 17   would you like it to be?  Clearly, it becomes

 

 18   extraordinarily difficult if you are suggesting

 

 19   that we should wait five years, or something like

 

 20   that.  Dr. Mack can speak for himself but I believe

 

 21   that in diabetes there was a difference and there

 

 22   may well be factors like that that make a

 

 23   difference.

 

 24             DR. BRIDGES:  My point is not to suggest

 

 25   how long we need to look, I am simply objecting to

 

                                                                65

 

  1   the concept that we can definitively or

 

  2   declaratively state at this point, based on what

 

  3   evidence we have, that we know that six months or

 

  4   nine months is an acceptable time frame in order to

 

  5   exclude device-related issues.

 

  6             DR. TRACY:  Dr. Yancy, and then if there

 

  7   is time Dr. Maisel.

 

  8             DR. YANCY:  Just a very short yes/no

 

  9   question.  I have not seen the referred to NEJM

 

 10   article comparing on-pump versus off-pump surgery.

 

 11   Were connectors used in the off-pump cases?

 

 12             DR. FRATER:  This was absolutely a study

 

 13   of on-pump versus off-pump vein patency.

 

 14             DR. TRACY:  Dr. Maisel?

 

 15             DR. MAISEL:  You have eloquently stated

 

 16   that times have changed and that historical

 

 17   controls are just that, historical, and you stated

 

 18   data that the patency rates vary greatly from

 

 19   institution to institution.  In many respects that

 

 20   is a strong argument for randomized trials but you

 

 21   didn't come out and state that.  Are you a

 

 22   proponent of randomized clinical trials to assess

 

 23   these devices?

 

 24             DR. FRATER:  You know, in the best of all

 

 25   possible worlds, yes.  I am speaking as a cardiac

 

                                                                66

 

  1   surgeon now.

 

  2             DR. TRACY:  Thank you.

 

  3             DR. EDMUNDS:  Did I hear you say that the

 

  4   one-year patency rate for off-pump proximal veins,

 

  5   the occlusion rate was 9-26 percent?

 

  6             DR. FRATER:  In the meta-analysis that we

 

  7   did between 6 and 12 months, there was a range from

 

  8   9.5 to 26.5 percent in this meta-analysis of some

 

  9   7,000 cases.  In the Prog IV study--

 

 10             DR. EDMUNDS:  For hand-sewn?

 

 11             DR. FRATER:  Hand-sewn anastomoses, in the

 

 12   Prog IV study the patency rate was 59 percent

 

 13   patent at one year on pump, 49 percent patent

 

 14   off-pump.  It was just presented at the ACC.

 

 15             DR. TRACY:  We do have to move on, I am

 

 16   sorry.  Dr. Mack?

 

 17             DR. MACK:  My name is Michael Mack and I

 

 18   am a cardiac surgeon in Dallas.  By way of

 

 19   disclosure, I am not sponsored by anybody today.  I

 

 20   paid my own way here.  I have served as a

 

 21   consultant in the past at St. Jude, also to

 

 22   Cardica.  I have received research grant support

 

 23   from St. Jude regarding anastomotic devices, and I

 

 24   am also on the scientific advisory board for

 

 25   Medtronic and Guidant, both of which have equity

 

                                                                67

 

  1   interests in anastomotic device companies.

 

  2             The thrust of my presentation was to

 

  3   discuss saphenous vein graft patency and not our

 

  4   own St. Jude device paper, which has been

 

  5   presented, in view of the fact that I only have

 

  6   five minutes but I will try and get this done in

 

  7   four and just spend the last minute discussing

 

  8   that.

 

  9             [Slide]

 

 10             Specifically, what I would like to discuss

 

 11   is I thought until I looked at all this that I knew

 

 12   what the gold standard for saphenous vein graft

 

 13   patency was.  Eeverybody throws around numbers but

 

 14   until I did a meta-analysis of the literature I

 

 15   really didn't know, and this is specifically to

 

 16   address the trial design question number one of the

 

 17   FDA, the gold standard of sutured anastomoses had a

 

 18   well-documented history of over the past thirty

 

 19   years, and I would like to go over those thirty

 

 20   years right now--

 

 21             [Slide]

 

 22             --or the why of saphenous graft failure in

 

 23   five minutes.

 

 24             [Slide]

 

 25             Since between 1979 and 2001 there have

 

                                                                68

 

  1   been thirty studies published, analyzing a total of

 

  2   28,081 grafts.

 

  3             [Slide]

 

  4             Factors impacting the studies are

 

  5   angiogram survivors.  You lose early graft

 

  6   occlusions resulting in death so, therefore, you

 

  7   are automatically losing patency in any angiogram

 

  8   series because you can only angiogram survivors.

 

  9   Studies are impacted by the completeness of

 

 10   follow-up, the percent of patients actually

 

 11   undergoing angiograms, and whether the study was

 

 12   done as a surveillance study or done for cause.

 

 13             [Slide]

 

 14             If we look at a meta-analysis of all

 

 15   studies that looked at 30 days or less, there has

 

 16   been a total of 11,000 grafts looked at.  If you

 

 17   just skip to the number at the bottom right, the

 

 18   patency rate at 30 days in these 7 studies,

 

 19   comprising 11,000 grafts, is 87.8 percent.

 

 20             [Slide]

 

 21             If you now look at 3 to 6 months and look

 

 22   at the 10 studies published here, with a total of

 

 23   2,290 grafts, at 3 to 6 months 84 percent is the

 

 24   saphenous vein graft patency.

 

 25             [Slide]

 

                                                                69

 

  1             If we now go to 12 months and look at the

 

  2   13 studies comprising almost 12,000 grafts, the

 

  3   patency rate is 82.7 percent in the literature.

 

  4             [Slide]

 

  5             Lastly, if we look at 2 to 5 years, with a

 

  6   total of 3,100 grafts in these 3 studies, the

 

  7   patency rate between 2 and 5 years is 74.3 percent.

 

  8             [Slide]

 

  9             If we summarize all this, there is a

 

 10   significant attrition in the literature of about 12

 

 11   percent of vein grafts in the first 30 days.

 

 12   Between 30 days and 3-6 months another 3 percent of

 

 13   grafts are lost at that point.  If we go between

 

 14   3-6 months to 12 months another 1.5 percent of

 

 15   grafts are lost.  Then there is a slightly greater

 

 16   attrition from 2 to 5 years.  If you look at

 

 17   overall graft patency of all 28,000 grafts done at

 

 18   any time, it is 84 percent.

 

 19             [Slide]

 

 20             Variables known to impact graft patency

 

 21   include age, gender, diabetes and how well the

 

 22   diabetes is controlled, obesity, which vessel is

 

 23   bypassed, the LAD, the circ. or the right, the

 

 24   target vessel size, the presence of distal disease,

 

 25   the size of the vein graft, harvest injury, whether

 

                                                                70

 

  1   an endarterectomy was done, what the graft flow at

 

  2   time of implant was, individual versus sequential

 

  3   vein grafts, how much myocardium was supplied, what

 

  4   the ventricular function of the patient was,

 

  5   whether lipid management was tightly controlled,

 

  6   whether antiplatelets were used, surgeon

 

  7   experience.

 

  8             [Slide]

 

  9             Variables that are not known how they

 

 10   impact on graft patency has been alluded to.

 

 11   Whether it is done on a beating heart or an

 

 12   arrested heart.  That recently has been called into

 

 13   question.  And whether anastomotic connectors

 

 14   impact positively or negatively on the saphenous

 

 15   vein graft patency.

 

 16             [Slide]

 

 17             I think how you design your study you can

 

 18   get 100 percent patency at ten years if you do an

 

 19   LAD, do it as a sequential and a three millimeter

 

 20   target with no distal disease, use a small vein,

 

 21   have a large run-off in a thin male that does not

 

 22   have insulin-dependent diabetes and normal ejection

 

 23   fraction, does not have a hypercoagulable state, is

 

 24   on antiplatelet agents and is well controlled with

 

 25   statins.  On the other hand, you can do the

 

                                                                71

 

  1   converse of all those and you will end up with a 10

 

  2   percent patency rate in less than 30 days.

 

  3             [Slide]

 

  4             In conclusion, I think that many variables

 

  5   other than anastomotic connectors impact graft

 

  6   patency.  Angiography is the only reliable method

 

  7   to determine patency.  Meta-analysis reveals an

 

  8   overall saphenous vein graft patency of 80-85

 

  9   percent.  There is no significant difference from

 

 10   3-6 months versus 12-16 months or, for that matter,

 

 11   even between 30 days and the latter two endpoints.

 

 12   An angiographically normal graft at the earlier

 

 13   study times is often likely to develop occlusion on

 

 14   later follow-up and, in my opinion, a 6-month

 

 15   angiographic endpoint is adequate to evaluate graft

 

 16   patency with anastomotic devices.

 

 17             Real quickly regarding our experience, it

 

 18   has been published on the St. Jude device.  We did

 

 19   find that there were events that happened after six

 

 20   months.  These were clinical events.  The study

 

 21   that we performed was very similar to Dr.

 

 22   Schoettle's.  We took a one-year experience with

 

 23   the St. Jude device and compared it to one-year

 

 24   previously with a similar cohort of patients and

 

 25   found that there was a higher incidence of clinical

 

                                                                72

 

  1   events in the St. Jude patients.  However, these

 

  2   were all limited to diabetics.  We looked at the

 

  3   non-diabetic population and we looked at all

 

  4   possible variable by logistic regression diabetes

 

  5   was the only thing that sorted out.  The

 

  6   confounding variable in all this is that all these

 

  7   procedures were also done on a beating heart.

 

  8   Thank you.

 

  9             DR. TRACY:  Thank you.  Panel, you have 4

 

 10   minutes and 36 seconds to ask questions.  Any

 

 11   questions?  Dr. Weinberger?

 

 12             DR. WEINBERGER:  In surgical literature

 

 13   everyone seems to focus on patency.  Are you

 

 14   interested at all in morphology, like quantitative

 

 15   angiography looking at 30 percent stenosis, 40

 

 16   percent stenosis?  Is that information valid to

 

 17   surgeons?

 

 18             DR. MACK:  Absolutely.  Because I do think

 

 19   that that is a precursor of potential total

 

 20   occlusion.

 

 21             DR. WEINBERGER:  And if that is the case,

 

 22   are your angiographic colleagues who have looked at

 

 23   these connectors able to assess the morphology

 

 24   right around the metallic connector adequately?

 

 25             DR. MACK:  I think the answer is yes.

 

                                                                73

 

  1             DR. KRUCOFF:  Not being as familiar with

 

  2   the surgical literature, in this list you sort of

 

  3   ended with do you think there is sufficient data to

 

  4   create a real propensity score in planning a trial?

 

  5             DR. MACK:  Yes.

 

  6             DR. KRUCOFF:  To actually create risk

 

  7   categories that could be sufficiently evaluated in

 

  8   new populations?

 

  9             DR. MACK:  Yes, I do.  I think that

 

 10   everything I listed there--one study or another has

 

 11   listed those factors implicating graft patency and,

 

 12   yes, I think you can develop a propensity score.

 

 13             DR. TRACY:  Dr. Edmunds?

 

 14             DR. EDMUNDS:  Mike, you said that all of

 

 15   these were off-pump bypasses.

 

 16             DR. MACK:  In our St. Jude experience,

 

 17   yes.

 

 18             DR. EDMUNDS:  Were they mostly right

 

 19   grafts?

 

 20             DR. MACK:  First of all, we did not have

 

 21   any connectors placed to the LAD so they all went

 

 22   to diagonal circumflexes or right, and which vessel

 

 23   it went to, in our experience, did not sort out as

 

 24   a factor.

 

 25             DR. EDMUNDS:  But non-LAD?

 

                                                                74

 

  1             DR. MACK:  All non-LADs.

 

  2             DR. EDMUNDS:  And these were surveillance

 

  3   angiograms, not for symptoms?

 

  4             DR. MACK:  No, the surveillance was

 

  5   clinical events only.  The only angiograms--

 

  6             DR. EDMUNDS:  So, you have bias towards

 

  7   symptomatic patients.

 

  8             DR. MACK:  The endpoint was not

 

  9   angiography.  The endpoint of our study was

 

 10   clinical events, major adverse events at now two

 

 11   years of follow-up.  We did not do a specific study

 

 12   angiogramming the patients.  The only angiograms we

 

 13   had was in patients that were done for cause.

 

 14             DR. EDMUNDS:  The 28,000 patients were

 

 15   from 30 studies, weren't they?

 

 16             DR MACK:  Okay, I am mixing up your

 

 17   question then.  Ask again, Hank.

 

 18             DR. EDMUNDS:  Well, the cohort of 28,081

 

 19   angiograms was from 30 papers--

 

 20             DR MACK:  Right.

 

 21             DR. EDMUNDS:  --and were those

 

 22   surveillance angiograms or for symptoms?

 

 23             DR. MACK:  I am sorry, I thought you were

 

 24   talking about our own experience with the

 

 25   connectors.

 

                                                                75

 

  1             DR. EDMUNDS:  No, I am sorry, Mike.

 

  2             DR. EDMUNDS:  No, all of those were

 

  3   surveillance.  Any that was done for cause and I

 

  4   did not include in that.  All those were

 

  5   surveillance studies.  Similarly, there were a

 

  6   couple of other studies that looked at just

 

  7   saphenous vein graft, the LAD, I did not include

 

  8   those because those were abnormally high.

 

  9             DR. TRACY:  Dr. Bridges, did you have a

 

 10   question?

 

 11             DR. BRIDGES:  My question really is that

 

 12   given that in the results you presented recently

 

 13   there was a difference in major adverse

 

 14   cardiovascular events, I guess in the manuscript

 

 15   that I have seen a draft of it was limited to

 

 16   diabetic patients.  However, those were non-insulin

 

 17   dependent diabetics, I believe, and I was wondering

 

 18   if you had a hypothesis as to why non-insulin

 

 19   dependent diabetics would be different than insulin

 

 20   dependent diabetics.  Given that, should we be then

 

 21   separating diabetics from everyone else in terms of

 

 22   determining the applicability of these devices?

 

 23             DR. MACK:  That is an excellent question,

 

 24   and we were a little bit surprised to find that

 

 25   that was the case also because from the stent

 

                                                                76

 

  1   experience you would expect it would be more so in

 

  2   insulin dependent diabetics but such was not the

 

  3   case.  We have hypothesized that perhaps it was due

 

  4   to the fact that with non-insulin dependent

 

  5   diabetic oral agents the blood sugar is not as

 

  6   tightly controlled, but we have no proof; it is

 

  7   total hypothesis.

 

  8             I also think that the way that we look at

 

  9   diabetes now, today, is totally blurring the line

 

 10   between insulin dependence and non-insulin

 

 11   dependent diabetics.  I think we have a lot of

 

 12   metabolic syndrome patients who are actually Type 2

 

 13   diabetics but are insulin dependent and we are

 

 14   actually categorizing them as insulin dependent

 

 15   when, in fact, they really should not be.

 

 16             DR. TRACY:  Thank you.  Dr. Slaughter?

 

 17             DR. SLAUGHTER:  Thank you.  I was asked to

 

 18   speak today on behalf of Converge, and I am a U.S.

 

 19   investigator for their ongoing trial for distal

 

 20   anastomotic studies and they did pay my travel here

 

 21   but I have no other financial relationship with

 

 22   them.

 

 23             [Slide]

 

 24             To date so far we have heard predominantly

 

 25   about proximal anastomotic devices and what I would

 

                                                                77

 

  1   like to do is to tell you a little bit about a

 

  2   current and ongoing look at distal anastomotic

 

  3   devices.

 

  4             [Slide]

 

  5             Certainly, this comes up in many issues

 

  6   and I don't think we need to belabor the fact but

 

  7   perhaps at the end I will comment briefly on some

 

  8   of the other questions asked, but there is still no

 

  9   question, and it is really sort of one of the

 

 10   unspoken issues for any outcome for the patient,

 

 11   and that is, you know, the quality of anastomosis

 

 12   and the overall revascularization and long-term

 

 13   patency.  Certainly surgeon skill is very

 

 14   important.  There are also the other issues of the

 

 15   anatomy, disease state, access and visibility that

 

 16   would affect these things.  But all these things

 

 17   are very important in determining not only acute

 

 18   but long-term graft patency and the overall outcome

 

 19   for the patient.

 

 20             [Slide]

 

 21             This has been brought up now several times

 

 22   and I think is very important.  This is just

 

 23   another way of presenting it.  It is looking at

 

 24   sort of the time scale injury.  That is, as was

 

 25   brought up by Dr. Weinberger as well, there is no

 

                                                                78

 

  1   question that there is good information and good

 

  2   data as to the initiation of the injury,

 

  3   inflammation and then subsequently intimal

 

  4   hyperplasia.  As a rule of thumb, the idea is that

 

  5   within, say, six to eight weeks the injury has

 

  6   stopped.  I don't think anybody in their right mind

 

  7   would argue that there is not heterogeneity and

 

  8   certainly there are differences within patients.

 

  9   Certainly that would show up as stenosis and

 

 10   changes in morphology, as you mentioned.

 

 11             But the idea is there is reasonably good

 

 12   science and information to suggest that within

 

 13   about 60 days a vascular anastomosis has healed,

 

 14   particularly within the coronary-arterial tree.

 

 15   So, beyond that time, if there are graft failures,

 

 16   the question is what are they due to, and it is

 

 17   generally due to ongoing atherosclerosis, intrinsic

 

 18   patient factors and/or perhaps a lack of medical

 

 19   therapy such as antiplatelet agents, aspirin and/or

 

 20   Plavix.

 

 21             So, you know, if hand-sewn anastomosis is

 

 22   so perfect, why are we here today?  The issue is

 

 23   they are not perfect and there certainly is room

 

 24   for improvement.  Certainly, by hand sewing in a

 

 25   bad distal vessel it is calcified in a diabetic. 

 

                                                                79

 

  1   They have lateral calcification.  By piercing them

 

  2   with needles--we all had that experience, you end

 

  3   up with plaque rupture.  You have hemorrhage within

 

  4   the media.  The idea is this is a traumatic event.

 

  5             [Slide]

 

  6             The other is reliability.  The issue is

 

  7   how can you do it day to day, 20,000 a year.  The

 

  8   idea is  you want to make it as reliable as

 

  9   possible and it needs to be reproducible between

 

 10   different surgeons at different institutions.

 

 11             The other is it must be reversible.  The

 

 12   idea is if you don't like it you have to cut the

 

 13   suture, take it out and redo it.  You want to be

 

 14   able to do the same thing, perhaps in a less

 

 15   traumatic fashion, with a coupler device.

 

 16             The other is it must be easy to use.  The

 

 17   idea is if anybody walks up to the podium and is

 

 18   giving you a talk, they basically should all be

 

 19   able to have the same results without any

 

 20   significant extensive training.

 

 21             The other is I think we do need to realize

 

 22   there are differences between proximals and

 

 23   distals.  I don't think we need to spend a lot of

 

 24   time on this today but the main two differences are

 

 25   the flow dynamics which clearly are different at

 

                                                                80

 

  1   the proximal and distal ends, as well as tissue

 

  2   characteristics.  On the tissue characteristics, on

 

  3   the right it is either going to the aorta or vein

 

  4   aorta or artery depending on which you conduit you

 

  5   use.  Certainly for distal anastomoses what you are

 

  6   looking at is vein to coronary artery or an

 

  7   arterial conduit to an artery but it is a very

 

  8   different scenario.

 

  9             Also, with flow dynamics there is no

 

 10   question that the size or the shape of the opening

 

 11   or the angle of the take-off is very important, the

 

 12   pressure differential, as well as the vessel

 

 13   diameter throughout the length.

 

 14             [Slide]

 

 15             I think one other issue which hasn't been

 

 16   brought up today which does need to be mentioned,

 

 17   at least just to bring it up, is actually the type

 

 18   of material.  I think this sort of goes into the

 

 19   heterogeneity or perhaps ongoing injury to intimal

 

 20   hyperplasia.  These are not new materials.  They

 

 21   have all been used before.  They have all been used

 

 22   in intravascular scenarios and the idea is there is

 

 23   good evidence to suggest, whether it is nitinol,

 

 24   stainless steel, titanium, that they are

 

 25   compatible, and I don't think that we can sort of

 

                                                                81

 

  1   imply or say that they are intrinsically the source

 

  2   of perhaps later stenoses or some ongoing failures

 

  3   beyond the eight-week time period.  Certainly there

 

  4   is the heterogeneity of healing in some patients

 

  5   but it is a relatively small number.  It is like a

 

  6   cheloid.  Some patients get cheloids but not all.

 

  7   The answer is you see it as it progresses.  If you

 

  8   follow them and you look for it you can identify

 

  9   who those patients are.

 

 10             [Slide]

 

 11             I would like to just show you a histologic

 

 12   series which I think is interesting in helps people

 

 13   visualize.  Really the sort of best description I

 

 14   think for the Converge distal anastomotic device is

 

 15   that of sort of a compression clip.  The idea is it

 

 16   is two frames which are expandable.  In the upper

 

 17   right it sort of gives you the diagrammatic picture

 

 18   of a graft into the artery.  The important thing

 

 19   here is that you now are able to mechanically

 

 20   manipulate flow dynamics as well as other

 

 21   engineering aspects so you get a perfect 30 degree

 

 22   take-off; you get perfect dynamics.  You won't get

 

 23   turbulence at the site of the anastomosis.

 

 24             The left side shows the bypass graft,

 

 25   which is CABG going down to the circumflex artery. 

 

                                                                82

 

  1   I think the important thing here is this was done

 

  2   at 90 days but, once again, the idea is it is

 

  3   completely endothelialized so the idea is if you

 

  4   get an angiogram at six months and you have a

 

  5   normal lumen you have no narrowing.  The idea is

 

  6   are you going to have ongoing intimal hyperplasia

 

  7   that would be an unexpected finding?  I think the

 

  8   answer is no.

 

  9             DR. TRACY:  If you could finish up in the

 

 10   next few sentences.

 

 11             DR. SLAUGHTER:  Sure, I can finish up in

 

 12   about 30 seconds.

 

 13             [Slide]

 

 14             The idea is you see very clearly that it

 

 15   is a well healed anastomosis and you have the

 

 16   advantages.

 

 17             [Slide]

 

 18             This has already been brought up.  The

 

 19   idea is are historical controls acceptable?  I

 

 20   think the answer is yes.

 

 21             [Slide]

 

 22             There is no question there is lots of

 

 23   existing data.  We have also lots of information to

 

 24   suggest not only at seven days but at years out

 

 25   that you can evaluate intimal hyperplasia.

 

                                                                83

 

  1             [Slide]

 

  2             Certainly angiography--we know the causes

 

  3   of failure, early failure and what we need to do is

 

  4   differentiate between a device failure and ongoing

 

  5   atherosclerosis.

 

  6             [Slide]

 

  7             I will just show--

 

  8             DR. TRACY:  I am sorry, we are just going

 

  9   to have to cut this off if we are going to have

 

 10   time for questions from the panel.

 

 11             DR. SLAUGHTER:  I apologize.

 

 12             DR. TRACY:  We have three minutes left for

 

 13   questions from the panel.  Anybody?  Dr. Hirshfeld?

 

 14             DR. HIRSHFELD:  I would just point out

 

 15   that in the coronary stent experience if we used a

 

 16   two-month follow-up we never would have discovered

 

 17   restenosis.

 

 18             DR. TRACY:  Dr. Krucoff, did you have a

 

 19   comment?

 

 20             DR. KRUCOFF:  I would just also say that

 

 21   in the stent experience I think if we started with

 

 22   historical controls based on lung literature, we

 

 23   would have left a lot of important information out.

 

 24             DR. SLAUGHTER:  I think the one difference

 

 25   though, and this has come up I think in other

 

                                                                84

 

  1   discussions with the FDA panel, is that although it

 

  2   uses a similar material and it is stent-like, it is

 

  3   not a stent.  The idea is it is just the edges that

 

  4   are present along the edges of the coronary artery.

 

  5   It is not compressed plaque and the idea is it is

 

  6   very different.  It is really sort of a compression

 

  7   clip that applies the vein graft to the distal

 

  8   coronary artery.

 

  9             DR. TRACY:  Dr. White?

 

 10             DR. WHITE:  I think there is no evidence

 

 11   for that, and I think everything that we have heard

 

 12   today sounds like it is a stent, although a stent

 

 13   in a graft.  So, the question would be if you don't

 

 14   believe it is a stent, then you should show us data

 

 15   that suggested that intimal hyperplasia within the

 

 16   tube is not the primary cause of these closures.

 

 17             DR. SLAUGHTER:  Sure.

 

 18             DR. YANCY:  And because of that, I think

 

 19   it is even more important to state that historical

 

 20   controls would be really problematic I think.

 

 21             DR. TRACY:  Any other comments from the

 

 22   panel?

 

 23             [No response]

 

 24             Thank you.  Is Mr. Lotti here?

 

 25             [No response]

 

                                                                85

 

  1             We will move on then to Dr. Martin.

 

  2             DR. MARTIN:  Good morning.  As so many

 

  3   members of the panel have already suggested,

 

  4   including Dr. White, I can make my comments brief.

 

  5   My name is Dr. Frank Martin.  I am Chairman of the

 

  6   Department of Cardiology at Methodist Care in

 

  7   Memphis, one of the largest private hospitals in

 

  8   the country.  I have no financial ties with any

 

  9   anastomotic device companies or, for that matter,

 

 10   any stent companies.

 

 11             My historical experience, I trained with

 

 12   John Simpson back in 1985, '86, and have

 

 13   relationships with many of the members of this

 

 14   panel.  I trained with people who are icons today,

 

 15   like Paul Yak, Paul Tierstien, Dean Keriakus, Met

 

 16   Selman, Morris Bookbinder, Rock Califf, Eric Topal

 

 17   and did interventional cardiology until

 

 18   approximately four years ago and made a life style

 

 19   change, and now I do only diagnostic caths and do

 

 20   my chairmanship.  Also as discussed earlier, in the

 

 21   late 1980s, with Dr. Chris White, we did brachy

 

 22   therapy because some of the early DCA slides showed

 

 23   needle intimal hyperproliferation similar to

 

 24   cancer.

 

 25             That having been said, I, as many of you

 

                                                                86

 

  1   all, have honed this sixth sense of skill with

 

  2   cardiology over the last 15 years of practice.  Dr.

 

  3   Phil Schoettle, who has already presented here

 

  4   today, and I have worked collaboratively for the

 

  5   last 20 years.  We basically make a good team

 

  6   because he knows what I do and I know what he does.

 

  7             Our group opened one of the first

 

  8   outpatient cardiac cath labs and it was a labor of

 

  9   trust on his part.  Both of us have a sixth sense

 

 10   about when patient is dissected and needs to go to

 

 11   surgery urgently, and have always had that sort of

 

 12   feel.  Obviously, in the early days of intervention

 

 13   with PRCA lots of patients went to CABG and, of

 

 14   course, more and more patients went to CABG at that

 

 15   time than do now.

 

 16             So, imagine my chagrin in September of

 

 17   2002 when I cath'd an ER nurse friend of mine and

 

 18   found one occluded and two stented Symmetry aortic

 

 19   connectors, the first patient I had ever seen.

 

 20   When Dr. Schoettle referred to September, 2002 that

 

 21   was the watershed moment.  I walked out of that

 

 22   cath, called him and said, Dr. Schoettle, I don't

 

 23   know what this device is but it is a stent and it

 

 24   will act like a stent and it will always be a

 

 25   stent.  I said, what is it?  What do we know about

 

                                                                87

 

  1   it?  And, he basically told me his experience over

 

  2   the last ten months.

 

  3             At that point I found an interventional

 

  4   colleague of mine and said, what are these devices?

 

  5   He said he had been stenting them since April; he

 

  6   didn't know much about them.  I did an Internet

 

  7   search and found out they were made in nitinol, and

 

  8   realized at that point in time that no

 

  9   cardiologists were involved in either the research,

 

 10   the design, the implementation or the roll-out of

 

 11   this device basically because all the

 

 12   cardiologists, interventional cardiologists,

 

 13   especially know the problems associated with that.

 

 14             It took me about 45 days, almost two

 

 15   months, with multiple interventional colleagues of

 

 16   mine and surgeons in Memphis to have it withdrawn

 

 17   from all the shelves of all three hospitals in

 

 18   Memphis, Tennessee, and that was in the latter part

 

 19   of fall of 2002.

 

 20             As patients have returned to the clinic,

 

 21   dozens, and dozens, and dozens have been found to

 

 22   have virtually total and/or subtotal occlusions of

 

 23   these devices.  The first contact I had with St.

 

 24   Jude was in December, 2002 after I had gone to TCT,

 

 25   in September I believe, and HA in November, telling

 

                                                                88

 

  1   them about the problem with these devices and why

 

  2   they acted like stents.  Finally, they walked in on

 

  3   me while I was cath'ing a 70 year-old ob/gyn who

 

  4   had two patently occluded Symmetry aortic

 

  5   connectors.  Basically, I said this is the problem.

 

  6   You don't understand anginal syndrome because most

 

  7   of these patients won't come back with chest pain

 

  8   for multiple reasons--denervation of the heart;

 

  9   more LV dysfunction problem.  You don't understand

 

 10   the role of clopidogrel or Plavix in these patients

 

 11   because most of them go to surgery without Plavix

 

 12   on board, and you don't understand the fact that in

 

 13   stent pathology, which we obviously cath a lot, you

 

 14   can have one patent graft, for instance the LIMA

 

 15   which most of these patients get, and the other two

 

 16   can subtotally occlude slowly and their only

 

 17   symptom is LV dysfunction.

 

 18             We, as cardiologists, as members of this

 

 19   panel, diagnose ischemia.  We send these patients

 

 20   to a surgeon for treatment and continue to reattach

 

 21   and stent these folks.  They will come back for

 

 22   years with their LIMAs.  An anecdotal experience of

 

 23   one surgeon in Jonesboro, Arkansas, close by

 

 24   Memphis, asked two of the cardiologists in his

 

 25   community, "so what's up with this Symmetry aortic

 

                                                                89

 

  1   connector?"  And the cardiologists response was,

 

  2   "are you having any problems?"  And he said, "well,

 

  3   I don't know."  And they said, "well, don't worry

 

  4   about it."

 

  5             He wasn't satisfied with that, came to

 

  6   Memphis, we cath'd him and his two connectors were

 

  7   occluded and his lumen was patent.  After

 

  8   intervention he told me that as an oral surgeon he

 

  9   uses nitinol every day to induce scar tissue

 

 10   formation and keep bridge reconstruction in place.

 

 11   The fact that you auger a hole in the aorta, hold a

 

 12   finger over it beginning the platelet clotting

 

 13   cascade, implant a metallic device with hooks

 

 14   without the benefit of loading doses of Plavix or

 

 15   predictable absorption is inconceivable.

 

 16             The idea of a connector makes sense for

 

 17   improvement of stroke risk, however, I feel the

 

 18   present device should be withdrawn and should have

 

 19   been withdrawn years ago.  Basically, I think the

 

 20   cardiologists need to be involved in any future

 

 21   trials or designs and I think to do otherwise is a

 

 22   violation of our sacred oath to our patients.

 

 23   Thank you.

 

 24             DR. TRACY:  Thank you.  Any questions from

 

 25   the panel?  Comments?

 

                                                                90

 

  1             [No response]

 

  2             Thank you.  Dr. Hausen?

 

  3             DR. HAUSEN:  By way of introduction, my

 

  4   name is Bernard Hausen.  I am the present CEO of

 

  5   Cardica.  My background, I am a cardiac surgeon by

 

  6   training.  My financial conflicts are inherent with

 

  7   my position, otherwise I have none other.

 

  8             [Slide]

 

  9             I want to use this opportunity to show you

 

 10   new generation of products that we are developing

 

 11   beyond the pioneers in this field that we have been

 

 12   discussing so far.

 

 13             [Slide]

 

 14             We have two products in the pipeline.  One

 

 15   is a distal anastomosis system.

 

 16             [Slide]

 

 17             It is called C-Port and it is based on the

 

 18   principle of simulating interrupted stitch distal

 

 19   anastomosis by applying a set of eight implantable

 

 20   clips, all simultaneously, and performing

 

 21   arteriotomy with the push of one button.  This type

 

 22   of a system results in a minimal amount of metal

 

 23   exposure.  It is a applied in distal anastomosis

 

 24   and it is in clinical evaluation as we speak.

 

 25             [Slide]

 

                                                                91

 

  1             This is just a video showing how it works.

 

  2   This is a 1.5 mm LAD.  You insert the anvil; pull

 

  3   it out and you are basically done; place one stitch

 

  4   to close the anvil insertion  hole.  This is a 1 mm

 

  5   diagonal cadaver heart and shows how it works.

 

  6             This technology, we believe, will enable

 

  7   beating heart surgery as it is quick and does not

 

  8   require any temporary ischemia of the myocardium

 

  9   during placement.

 

 10             [Slide]

 

 11             We have a second device which is called

 

 12   PAS-Port.  It stands for proximal anastomosis

 

 13   system, and it is a second generation proximal

 

 14   system.  We have the advantage of being a company

 

 15   that is going to be able to take advantage of the

 

 16   knowledge from the predecessors, predicate devices.

 

 17             [Slide]

 

 18             So we were able in our design to spend a

 

 19   lot of time on key improvements from things we have

 

 20   learned from the other devices.  We have focused on

 

 21   trying to minimize or completely eliminate

 

 22   endothelial trauma of the graft during loading.  We

 

 23   wanted to minimize blood-exposed non-endothelial

 

 24   tissue, i.e., metal exposure.  We wanted to

 

 25   maximize the orifice area and reduce the incidence

 

                                                                92

 

  1   of kinking by a low profile.

 

  2             [Slide]

 

  3             We did that by basically having nothing

 

  4   touch the endothelium of the vein during loading or

 

  5   deployment.  This is a cross-section of the

 

  6   implant.

 

  7             [Slide]

 

  8             We have a minimal amount of metal exposed

 

  9   with the stainless steel device.  It is the same

 

 10   material as is being used for coronary stents.  And

 

 11   we wanted to maximize the orifice, especially for

 

 12   small vein grafts, and have a very low profile

 

 13   height.

 

 14             [Slide]

 

 15             For all this we have done a clinical

 

 16   trial.  We have had a lab cardiologist review our

 

 17   data by QCA and determine what is the amount of

 

 18   narrowing of the implant versus the graft body.

 

 19   They first looked at some hand-sewns that were done

 

 20   concurrently in those patients and, as you can see,

 

 21   the average narrowing of a hand-sewn is about 5

 

 22   percent at discharge and about 18 percent at 6

 

 23   months.  This is in agreement with all the

 

 24   published literature.

 

 25             [Slide]

 

                                                                93

 

  1             Then we asked them to look at the PAS-Port

 

  2   data.  What you find, and you can hardly see this,

 

  3   this is a minus 7 percent narrowing, i.e., the

 

  4   grafts at the anastomosis are larger than they are

 

  5   in the graft body and that is by design.  That is

 

  6   how the implant has been designed.

 

  7             Now at 6 months, the most important

 

  8   figure, the average narrowing is 3 percent compared

 

  9   to 18 percent in hand-sewns.  I propose that if a

 

 10   device had a problem at discharge or at 6 months

 

 11   you would be seeing that in this quantitative

 

 12   analysis.  If you don't see it because the

 

 13   injurious event was at the time of surgery, you are

 

 14   very unlikely to see it going forward besides the

 

 15   normal decay of a vein graft, as alluded to by the

 

 16   previous speakers.

 

 17             [Slide]

 

 18             So, Cardica's regulatory position is we

 

 19   are applying for 510(k) clearance based on

 

 20   prospective multicenter non-randomized trials, and

 

 21   our primary study endpoint for this distal device

 

 22   is vessel patency at discharge and 6 months, and

 

 23   for the proximal device performing a vessel patency

 

 24   study at 6 months with QCA.  Thank you very much

 

 25   for your attention.

 

                                                                94

 

  1             DR. TRACY:  Thank you.  Any questions?

 

  2   Dr. White?

 

  3             DR. WHITE:  I just noticed that on the

 

  4   last slide you said you were going to do MRI

 

  5   follow-up on these metal grafts.  How are you going

 

  6   to do that?

 

  7             DR. HAUSEN:  We have done that on the

 

  8   proximal anastomotic device.  Basically, with the

 

  9   gadolinium contrast injection you can see--the only

 

 10   thing CTs and MRIs allow you to do is determine is

 

 11   the graft patent or not.  You cannot evaluate the

 

 12   degree of stenosis at the implant.  So, a preferred

 

 13   method is a quantitative angiography.

 

 14             DR. WHITE:  Do you have experience with

 

 15   MRI?

 

 16             DR. HAUSEN:  We have done five MRIs in the

 

 17   patients in this study.

 

 18             DR. WHITE:  And also CT?

 

 19             DR. HAUSEN:  And CT too and MDCT.

 

 20             DR. WHITE:  And there is no difference in

 

 21   your hands?

 

 22             DR. WHITE:  I like the MDCT much better.

 

 23   I think the image is much clearer.  The 3-D

 

 24   reconstructions are very impressive.

 

 25             DR. AZIZ:  And how does that correlate

 

                                                                95

 

  1   with angiograms?

 

  2             DR. HAUSEN:  It depends on what your

 

  3   outcome variable is.  If you want to just know if

 

  4   the graft is patent or not, there is a very, very

 

  5   good correlation.  That has been shown in the

 

  6   literature.  If you need more than that, if you

 

  7   need to know is there a degree of narrowing, that

 

  8   will not suffice.

 

  9             DR. AZIZ:  If you do distal anastomosis if

 

 10   you have bleeding, how can you control that?  Can

 

 11   you put a regular stitch over that?

 

 12             DR. HAUSEN:  Yes, you can.  It is the same

 

 13   as a steel device.  It is very firm.  The pull-out

 

 14   force of this device is very high because stainless

 

 15   steel is three times stiffer than nitinol.  So,

 

 16   what you do, you just place the first string around

 

 17   the anastomosis and slowly tighten it.  That brings

 

 18   the aorta closer to the implant--

 

 19             DR. AZIZ:  If you do distal anastomosis if

 

 20   you have bleeding, can you do regular stitches?

 

 21             DR. HAUSEN:  Yes.

 

 22             DR. AZIZ:  You have obviously shown a vein

 

 23   graft.  If you had an arterial graft can you use

 

 24   your same distal anastomotic site for that?

 

 25             DR. HAUSEN:  This generation of device,

 

                                                                96

 

  1   no; the next generation, yes.

 

  2             DR. TRACY:  Dr. Bridges?

 

  3             DR. BRIDGES:  You showed differences in

 

  4   percent stenosis of the proximal anastomoses at

 

  5   discharge and at 6 months.

 

  6             DR. HAUSEN:  Yes.

 

  7             DR. BRIDGES:  What about occlusion or

 

  8   patency at the same time points?

 

  9             DR. HAUSEN:  We have 87.9 percent patency

 

 10   rate so we had 6 occlusions in 50 implants, which

 

 11   is 100 percent in agreement with the historical

 

 12   data from the meta-analysis you saw and we did too.

 

 13             DR. BRIDGES:  So, how would you interpret

 

 14   the fact that in spite of having a higher orifice

 

 15   area or diameter you have the same patency at the

 

 16   6-month time point?

 

 17             DR. HAUSEN:  That is wonderful proof that

 

 18   it has nothing to do with the connector.  It is

 

 19   probably your distal run-off or any of the other

 

 20   200 factors that Dr. Mack said.

 

 21             DR. AZIZ:  If you had a very thick

 

 22   proximal ascending aorta--

 

 23             DR. HAUSEN:  Yes?

 

 24             DR. AZIZ:  --sometimes you do a hand-sewn

 

 25   vein graft that dunks in and obviously you don't

 

                                                                97

 

  1   want that.

 

  2             DR. HAUSEN:  Yes.

 

  3             DR. AZIZ:  Does your anastomosis

 

  4   technique--where would that fit in?  Would the vein

 

  5   also dunk in?

 

  6             DR HAUSEN:  It is inverted over the

 

  7   implant so it is in the lumen but, because it is a

 

  8   stainless steel implant, it props the anastomosis

 

  9   open and you will not have lumen reduction, if that

 

 10   is where you are heading towards.  And we have

 

 11   shown that, minus 7 percent widening of the

 

 12   anastomosis is evidence that that is exactly what

 

 13   the implant does and it accommodates the varying

 

 14   thickness of the aortic wall because it is like a

 

 15   paper clip.  It can adjust to varying thicknesses.

 

 16             DR. AZIZ:  And the angle at which it comes

 

 17   off proximally, is that oblique or head-on?

 

 18             DR. HAUSEN:  It is theoretically 90

 

 19   degrees.  We asked our core lab to evaluate that

 

 20   too.  There are hardly any at 90 degrees.  They

 

 21   vary from 10-70 degree take-offs.  Because the

 

 22   hinge point is so small, only 1.5 mm, the vein can

 

 23   come off almost at any angle it wants to.

 

 24             DR. AZIZ:  So, could you take the proximal

 

 25   along through the transverse sinus and pull it

 

                                                                98

 

  1   through?

 

  2             DR. HAUSEN:  You could, yes.

 

  3             DR. AZIZ:  You could?

 

  4             DR. HAUSEN:  Yes.

 

  5             DR. EDMUNDS:  What is the size of the shoe

 

  6   inside the vessel?

 

  7             DR. HAUSEN:  The shoe inside the vessel?

 

  8             DR. EDMUNDS:  Against which you are

 

  9   putting the clamps down.  The part of the device

 

 10   that goes inside the vessel, what are the

 

 11   dimensions of that shoe of the device?

 

 12             DR. HAUSEN:  There is really nothing

 

 13   inside the vessel.  The vein is pulled through the

 

 14   implant and inverted so there is no metal inside,

 

 15   except for the prongs that penetrated the vein and

 

 16   then go outward.  I would be more than happy to sit

 

 17   down afterwards and show you maybe some work.  I am

 

 18   kind of limited by the time here.

 

 19             DR. AZIZ:  Can you do a sequential of this

 

 20   for the distal anastomosis?

 

 21             DR. HAUSEN:  No.  Well, you could if you

 

 22   did your side by side by hand, absolutely.

 

 23             DR. TRACY:  Thank you very much.  Prof.

 

 24   Klima?

 

 25             PROF. KLIMA:  Ladies and gentlemen,

 

                                                                99

 

  1   members of the panel, my name is Uwe Klima.

 

  2             [Slide]

 

  3             I am a full professor at Hanover Medical

 

  4   School for Cardiac Surgery.  The financial

 

  5   disclosure I have to make is that Ventrica paid for

 

  6   my trip here and my lodging, and Ventrica provided

 

  7   us with an unrestricted grant for preclinical

 

  8   testing of an anastomotic device three years ago.

 

  9             [Slide]

 

 10             I expected a talking time of ten minutes

 

 11   and I will try to cut that down to five minutes.

 

 12   Basically, what I want to talk about is mechanisms

 

 13   of how wound healing takes place after an

 

 14   anastomosis; give you some of our clinical

 

 15   experience with hand-sewn anastomosis, especially

 

 16   with our MIDCAB series; more update or experience

 

 17   with our anastomotic devices; and I will have a

 

 18   little discussion of appropriate methods and

 

 19   follow-up time frames for CABG surgery.

 

 20             [Slide]

 

 21             As background, we all know that hand-sewn

 

 22   anastomoses now are more or less on the market for

 

 23   more than four decades.  Everything is pretty much

 

 24   well tested and evaluated.  We have a pretty clear

 

 25   understanding of what happens to an anastomosis. 

 

                                                               100

 

  1   What happens is a healing response--at what time

 

  2   frame this will be stable.  So, I will go into

 

  3   details with my next slide.

 

  4             [Slide]

 

  5             There are several publications out now

 

  6   which tell us exactly what happens after an

 

  7   anastomosis has been performed.  We know there is a

 

  8   lot of trauma coming after surgery.  Cell

 

  9   proliferation is coming out.  And the most

 

 10   important message that comes out of this

 

 11   publication, for example, is that the repair

 

 12   process is about to be completed two months after

 

 13   surgery.

 

 14             [Slide]

 

 15             We wanted to know what is happening with

 

 16   anastomotic devices.  Is it the same response?  Can

 

 17   we expect the same thing to happen?  Filsoufi

 

 18   published, from Boston.  He tested the Ventrica

 

 19   device and what is happening after implantation two

 

 20   months, three months and six months after surgery,

 

 21   and we could see that there is a single layer of

 

 22   endothelium covering after two months, three months

 

 23   and after six months and there was no sign of any

 

 24   inflammatory response at the site of the

 

 25   anastomosis.

 

                                                               101

 

  1             [Slide]

 

  2             As a control group--I would like to speak

 

  3   a little bit about our MIDCAB experience in

 

  4   Hanover.  We have now enrolled more than 700

 

  5   patients.  Out of the first 500 patients we did

 

  6   angiographic follow-up in 6-7 percent of these

 

  7   patients.  The first group, which was the big one

 

  8   with 297 patients, had a pre-discharge angiogram.

 

  9   What was pretty interesting was that in about 6

 

 10   percent of these patients we had a highly

 

 11   significant problem at the site of the anastomosis,

 

 12   as you can see here.  As a Swedish colleague

 

 13   presented his data with the same problem four years

 

 14   ago at the ASCTS and recommended just to wait

 

 15   because this is part of the healing response, we

 

 16   changed our politics, which you will see on the

 

 17   next slide, and just let the whole situation be as

 

 18   it was; waited for 3-6 months, reevaluated these

 

 19   patients and saw that the degree of stenosis or the

 

 20   number of intimal hyperplasia went down without any

 

 21   intervention from 6 percent to 1 percent.

 

 22             [Slide]

 

 23             There is another example here and, as you

 

 24   can see again, there is a highly significant

 

 25   stenosis here at the pre-discharge angiogram;

 

                                                               102

 

  1   perfect anastomosis 3-6 months later when we

 

  2   reevaluated the patients.  What we learned here is

 

  3   that the healing response is still evolving in the

 

  4   earlier time frame.  We changed our angiographic

 

  5   follow-up from pre-discharge to a 6-month follow-up

 

  6   so the remaining 203 patients were evaluated 6

 

  7   months after surgery instead of having a

 

  8   pre-discharge angiographic follow-up.

 

  9             [Slide]

 

 10             So, what is the Hanover experience now

 

 11   with anastomotic devices?  Just to give you a quick

 

 12   overview, Hanover does approximately 2,000 open

 

 13   heart procedures per year.  It is a large teaching

 

 14   institution.  We are affiliated with several

 

 15   research centers so we are exposed to new

 

 16   technologies and clinical trials.  The studies I

 

 17   have performed were with Ventrica, St. Jude and

 

 18   Converge.  In addition, I have a little experience

 

 19   also with Cardica and Coalescent, however, I just

 

 20   want to present you the data where I have

 

 21   angiographic follow-up.

 

 22             [Slide]

 

 23             St. Jude--we had a prospective, randomized

 

 24   trial with 11 patients where every patient received

 

 25   two proximal anastomoses.  One was hand-sewn and

 

                                                               103

 

  1   the other one was an automatic anastomosis.  This

 

  2   is the strongest study design that you can create.

 

  3   The data that we saw was that in 11 patients who

 

  4   were enrolled in the study and came back after 6

 

  5   months and there were 10 postoperative angiograms

 

  6   showing that only 3 grafts were patent.  We had 6

 

  7   occlusions and 1 highly significant stenosis at the

 

  8   site of the anastomosis, with a consequent PTCA and

 

  9   stent after graft.  Even though the patients were

 

 10   asymptomatic, the study was stopped because the

 

 11   data did not look the way we wanted to have it.

 

 12             [Slide]

 

 13             Even though all patients were

 

 14   asymptomatic, due to several reasons that we can

 

 15   discuss, I think independent of the cause a

 

 16   prospective six-month angiographic evaluation was

 

 17   sufficient in our study to detect performance

 

 18   issues of the device.

 

 19             [Slide]

 

 20             Ventrica was part of a multicenter trial

 

 21   that we did with two other centers.  We enrolled

 

 22   100 patients, 48 came from Hanover--

 

 23             DR. TRACY:  Can I ask you to start

 

 24   wrapping up?

 

 25             PROF. KLIMA:  Yes.  The most important

 

                                                               104

 

  1   information I want to give you here is that in the

 

  2   first 48 patients we had a pre-discharge angiogram

 

  3   and a 6-month angiogram to study efficacy of the

 

  4   device and performance of the anastomosis after 6

 

  5   months.

 

  6             [Slide]

 

  7             The last study we did was with Converge.

 

  8   It was also a multicenter trial.  We had 8 weeks of