UNITED STATES OF AMERICA

FOOD AND DRUG ADMINISTRATION

CENTER FOR DEVICES AND RADIOLOGICAL HEALTH

MEDICAL DEVICES ADVISORY COMMITTEE

* * *

CIRCULATORY SYSTEM DEVICES PANEL

* * *

MEETING

* * *

THURSDAY,

OCTOBER 2, 2003

                       * * *

 

            The Panel met at 9:00 a.m. in Salons A, B and C of the Gaithersburg Hilton, 620 Perry Parkway, Gaithersburg, Maryland, Dr. Warren Laskey, Acting Chairman, presiding.

 

PRESENT:

WARREN K. LASKEY, M.D.       Acting Chairman

SALIM AZIZ, M.D.             Member

THOMAS B. FERGUSON, M.D.          Consultant

PRESENT (Continued):

ALLEN A. HUGHES, Ph.D.       Consumer Representative

MITCHELL W. KRUCOFF, M.D.    Consultant

WILLIAM H. MAISEL, M.D., M.P.H., Consultant

DOUGLASS A. MORRISON, M.D.   Consultant

MICHAEL C. MORTON            Industry Representative

GARY G. NICHOLAS, M.D.       Consultant

SHARON-LISE NORMAND, Ph.D.   Consultant

JOHN C. SOMBERG, M.D.        Consultant

CYNTHIA TRACY, M.D.          Member

CHRISTOPHER J. WHITE, M.D.   Consultant

GERETTA WOOD                 Executive Secretary

 

FDA REPRESENTATIVES:

BRAM ZUCKERMAN, M.D.                      

JOHN P. HOLDEN, Ph.D.

BARBARA KRASNICKA, Ph.D.

WOLF SAPIRSTEIN, M.D.

 

 

 

 

SPONSOR REPRESENTATIVES:

CHRIS REISER, Ph.D

BRUCE GRAY, D.O.

JOHN LAIRD, M.D.

VENKATESH RAMAIAH, M.D.


                  C O N T E N T S

                                              PAGE

Conflict of Interest Statement ................. 5

Introductions .................................. 8

Open Public Session ........................... 13

Spectranetics Corporation:

Sponsor's Presentation:

      Dr. Chris Reiser ........................ 13

      Dr. John Laird .......................... 18

FDA Presentation:

      Dr. John P. Holden ...................... 63

      Dr. Wolf Sapirstein ................. 70, 85

      Dr. Barbara Krasnicka ................... 79

Presentation of Lead Reviewer, Dr. Gary Nicholas 99

Commission Discussion ........................ 107

FDA Questions ................................ 190

Open Public Hearing .......................... 236

Panel Vote ................................... 238


               P-R-O-C-E-E-D-I-N-G-S

                                       (9:04 a.m.)

            CHAIRMAN LASKEY:  If we can all take our seats, we'll get started.

            Thank you.

            Good morning.  I'd like to call us to order.

            My name is Warren Laskey.  I'd like to call this meeting of the Circulatory System Device Panel to order today.

            The topic discussed will be the pre-market application for Spectranetics CVX-300 Excimer Laser System, P910001.

            If Geretta could please read the conflict of interest statement.

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

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

            The agency has determined, however, that the participation of certain members and consultants, the need for whose services outweighs the potential conflict of interest involved, is in the best interest of the government.  Therefore, waivers have been granted for Drs. Mitchell Krucoff and Christopher White for their interest in firms that could be affected by the panel's recommendations.

            Dr. Krucoff's waiver involves consulting on a competing technology firm's unrelated product for which he receives an annual fee of less than $10,001.

            Dr. White's waiver involves a grant to his institution for the sponsor's product study in which he had no involvement in data generation or analysis and for which funding to the institution was less than $100,001 per year.

            The waivers allow these individuals to participate fully in today's deliberations.  Copies of these waivers may be obtained from the agency's Freedom of Information Office, Room 12A-15 of the Parklawn Building.

            We would like to note for the record that the agency took into consideration other matters regarding Drs. Krucoff and Cynthia Tracy.  These panelists reported past or current interests involving firms at issue, but in matters that are not related to today's agenda.  The agency has determined, therefore, that they may participate fully in all discussion.

            We would also like to note that Michael Morton, the industry representative for the panel, has reported interests in firms at issue. 

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

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

            CHAIRMAN LASKEY:  Thank you.

            If I can have the panel members introduce themselves beginning with Mr. Morton.

            MR. MORTON:  My name is Michael Morton, and I'm an employee of CarboMedics, and I'm the industry representative.

            DR. HUGHES:  My name is Allen Hughes, and I'm on the faculty at George Mason University, and I'm the consumer representative.

            DR. NICHOLAS:  Gary Nicholas.  I'm at Lehigh Valley Hospital, the faculty of Penn State University.

            DR. TRACY:  Cindy Tracy.  I'm an electrophysiologist at Georgetown University.

            DR. MAISEL:  William Maisel, a cardiologist at Brigham & Women's Hospital in Boston.

            DR. WHITE:  Chris  White.  I'm an interventional cardiologist from the Ochsner Clinic Foundation.

            DR. FERGUSON:  Tom Ferguson, cardiovascular surgeon, Washington University School of Medicine, St. Louis.

            MS. WOOD:  Geretta Wood, Executive Secretary.

            CHAIRMAN LASKEY:  Warren Laskey, interventional cardiologist, National Naval Medical Center.

            DR. MORRISON:  Doug Morrison, interventional cardiologist, University of Arizona in Tucson VA.

            DR. SOMBERG:  John Somberg, Rush University.

            DR. KRUCOFF:  Mitch Krucoff.  I'm an interventional cardiologist at Duke University Medical Center and the Durham VA and the Director of devices trials at Duke Clinical Research Institute.

            DR. AZIZ:   Salim Aziz, adult cardiac surgeon practicing in the D.C.-Maryland area, and a clinical associate professor at University of Colorado.

            DR. NORMAND:  I'm  Sharon-Lise Normand, associate professor of biostatistics at Harvard Medical School and Harvard School of Public Health.

            DR. ZUCKERMAN:  I am Bram Zuckerman, Director, Division of Cardiovascular Devices at FDA.

            CHAIRMAN LASKEY:  Thank you, all.

            If I can have Ms. Wood read the voting status statement, please.

            MS. WOOD:  Pursuant to the authority granted under the Medical Devices Advisory Committee charter, dated October 27, 1990, and as amended August the 18th, 1999, I appointed the following individuals as voting members of the Circulatory System Devices Panel for this meeting on October the 2nd, 2003:

            Thomas Ferguson, M.D.

            Sharon-Lise Normand, Ph.D.

            Mitchell W. Krucoff, M.D.

            William Maisel, M.D., M.P.H.

            Douglass A. Morrison, M.D.

            Gary G. Nicholas, M.D.

            John C. Somberg, M.D.

            Christopher J. White, M.D.

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

            In addition, I appoint Warren K. Laskey, M.D., to act as temporary chairperson for the duration of this meeting.

            And it's signed by David W. Feigal, Jr., M.D., M.P.H., Director, Center for Devices and Radiological Health, on September 30th, 2003.

            CHAIRMAN LASKEY:  Thank you.

            It is traditional at this point to have the open public hearing portion of this morning's meeting, and before I open the session up to the public, I have a statement to read which is as follows:

            Both the Food and Drug Administration and the public believe in a transparent process for information gathering and decision making.  To insure such transparency at the open public hearing session of the Advisory Committee meeting, FDA believes that it is important to understand the context of an individual's presentation.

            For this reason, FDA encourages you, the open public hearing speaker, at the beginning of your written or oral statement to advise the committee of any financial relationship that you may have with the sponsor, its product, and, if know, its direct competitors.

            For example, this financial information may include the sponsor's payment of your travel, lodging, or other expenses in connection with your attendance at the meeting.

            Likewise, FDA encourages you at the beginning of your statement to advise the committee if you do not have any such financial relationships.

            If you choose not to address this issue of financial relationships at the beginning of your statement, it will not preclude you from speaking.

            That being said, is there anyone in the audience who wishes to address the panel on today's topic or any other topic?

            (No response.)

            CHAIRMAN LASKEY:  Great.  Than I'd like to close this portion of the open public hearing and move right into the sponsor's presentation.

            DR. REISER:  Good morning.  My name is Chris Reiser, and I'm the Vice President of Technology and Clinical Research at Spectranetics.

            I'd like to introduce the presenters that will be assisting me this morning. 

            Immediately to my right is Dr. John Laird from the Washington Hospital Center.  Dr. Laird was principal investigator for LACI and is a member of the LACI Steering Committee.

            Just to his left is Dr. Bruce Gray from Greenville Memorial Hospital.  Dr. Gray was also on the LACI Steering Committee and a LACI investigator.

            To his left is Dr. Venkatesh Ramaiah from the Arizona Heart Hospital.  Dr. Ramaiah is a staff surgeon there and was the LACI PI at that site.

            We have prepared several presentations for you this morning.  The top four, the ones in yellow, will be the ones that we'll give during our hour's presentation this morning.  We prepared several others which are in white, which we'll hold back in case there is a question from the panel and the Chairman would so desire us to show that slides.

            We reviewed today an application of excimer laser atherectomy.  Our device uses a xenon-chloride excimer laser which emits pulses of ultraviolet light at 308 nanometers.  These pulses are delivered via fiberoptic catheter which is specifically designed to deliver this intravascularly.

            It was first approved, this technology, by FDA in 1993 for use in coronary arteries.  This technology is similar, but slightly different from LASIK, which you probably heard advertised on the radio. 

            A quick comparison shows that LASIK uses a slightly different ultraviolet wavelength at 193 nanometers.  One, ninety-three does not propagate through fibers, and so in a LASIK machine the beam basically propagates through free air inside a work station, which if roughly the size of a small room.  One, ninety-three is good for working on corneas.

            Three, oh, eight, by comparison, does travel down fibers, and hence we can make catheters which deliver 308 nanometers through arteries and veins. 

            Both of these technologies use cool, ultraviolet ablation, which basically shaves away small layers of tissue either on the cornea or in the arteries without burning or charring, like previously used CW lasers.

            This is a picture of the laser system that was used in LASIK.  The generation four CVX-300 Excimer Laser System was approved by FDA in 1993, and we've been building and shipping the system for about ten years.

            The same system is used for all of our applications, coronary atherectomy, and pacing lead removal.  That particular application uses a specialized catheter that still plugs into the same laser.

            And in Europe we've been shipping this unit for peripheral atherectomy, but not in the United States.

            A few facts about this unit.  It's about the size of a console television set.  It weighs 650 pounds, has four wheels, and it plugs into the wall in a cath. lab.

            Really what does the work is the excimer laser catheter.  All of our catheters look like this picture here.  At the proximal end is a black connector that plugs into the laser.  Connecting the proximal end to the distal end are three meters of fiberoptic cable.  In this black connector the fiberoptics shown here as a rectangular array of about 250 fibers are arranged so that their shape is the same shape as the beam in our laser, which happens to be rectangular.

            Those fibers go through the whole catheter.  The last 130 centimeters is the patient contact end.  Those fibers end in the distal tip.  We put an optical polish on the tip, including the fibers and the metal bands on the inside and the outside of the fiber bundle.

            These fibers deliver the ultraviolet light directly to the tissue.  That light goes into the tissue and penetrates about 50 microns.  Now, human hair is about 60 microns.  So the penetration depth is about the depth of a human hair.  That's the tissue  that's affected by the light.  That's where the light goes.  It doesn't go to the side.  It doesn't penetrate deeply like a search light.  It basically shaves its way through the lesion that's in contact with the tip.

            All of our catheters, regardless of whether they're coronary designed or peripheral designed, have these same features and work the same way.

            How did we get interested in peripheral angioplasty?  Well, we looked at our seven indications for the coronary arteries.  We noticed that the top three in yellow here are a reasonable description of the disease that we expect to see in CLI patients, that is, long, diffuse lesions, total occlusions, and moderately calcified lesions.

            We did investigate our technology in the legs under an FDA approved IDE early in the 1990s.  That data is not part of our PMA submission, and so we won't be reviewing it today, but the point here is that we had a reasonable expectation that our technology would work well in the legs based on some data obtained about ten years ago.

            We also had experience in Europe since we have commercialized this product there for the past six years.  FDA advised us that we should try a pivotal trial and bring that data to the regulatory authorities, and that was the purpose of LACI Phase 2.

            At this point I'd like to invite Dr. Laird to the podium to give us a review of the LACI protocol and the LACI results.

            DR. LAIRD:  Well, thank you very much, Chris.

            It's a real privilege for me to be able to present the LACI Phase 2 registry study to this distinguished panel.

            I'd like to disclose up front that I do not have any financial interest in Spectranetics Corporation.  I do not own any stock or stock options, although I am paid as a consultant for my time here today.

            Just as a brief introduction, the patients being treated in the LACI trial are patients with critical limb ischemia, which is basically end stage peripheral arterial occlusive disease, which results in breakdown of the skin with ulcers or gangrene or pain in the foot at rest.

            In general, in the literature patients with peripheral arterial disease are classified either according to the Fontaine classification or the Rutherford classification.  In general, patients with critical limb ischemia have either Fontaine III or IV peripheral arterial disease, and those patients with rest pain or tissue loss are categorized as Category 4 or 6 on the Rutherford classification.

            Patients with a rest pain are Fontaine Class III or Rutherford Category 4.  Patients with tissue loss are Fontaine Class IV or Rutherford Category 5 or 6.

            The Rutherford classifications scheme does also tend to try and further differentiate patients to either having minor tissue loss, which is Rutherford Category 5, or major tissue loss, which is Rutherford 6, and in general, Rutherford Category 6 patients are felt the most likely to require amputation.

            When these patients undergo study, in general they have very diffuse, multi-level disease with frequent calcification, and predominantly they have occlusive disease in either most or all of the tibial peroneal arteries.  Only a minority of the patients, in one study approximately 22 percent of patients, present with lesion morphology that would be felt to be suitable for balloon angioplasty.

            And it wasn't the goal of the LACI to try and test laser angioplasty for patients with lesions that were otherwise suitable for balloon angioplasty, nor was it a trial that was meant to compare laser angioplasty with the gold standard, which is surgical revascularization for critical limb ischemia.

            What we were looking at was basically the no option patient, the patient that had diffuse, multi-segment, occlusive disease of the lower extremities which was not favorable for balloon angioplasty, and patients who were felt to be not good surgical candidates.

            The obvious criticism of this trial is that it's not a randomized trial, and I would like to kind of address this subject right up front.  I think it's important whenever you talk about a randomized trial you have to have a reasonable or specific therapy to test in the control group, a therapy that's an accepted therapy for a particular disease state.

            When we talk about the alternative therapies for patients with critical limb ischemia, we could talk about balloon angioplasty medications, primary amputation, or bypass surgery.

            In our discussions with the FDA initially with regards to study design, up front we decided that we were only going to treat patients who were felt to be poor or non-surgical candidates.  So that sort of takes surgical bypass out of the randomization scheme.

            Also, I think the majority of us who take care of patients with peripheral arterial disease would not consider medical therapy or primary amputation as an appropriate alternative therapy for the majority of patients who present with critical limb ischemia.

            So that leaves us primarily with balloon angioplasty, and the question is whether balloon angioplasty would be an appropriate randomization strategy for these patients.

            If you look at the published literature for balloon angioplasty for critical ischemia, it's really all over the place.  There's wide variability in the published results for balloon angioplasty.  When the technique is used in very well selected patients with discrete lesions in the femoral popliteal or tibial vessels, the results can be quite good with limb salvage rates that have been published in the 70 to 85 percent range.

            However, when the technique has been more broadly applied to diffuse disease or long occlusions, the results are not nearly as good, and if you look at the studies in general, there's a lot of faults that could be found in these studies.  They are all retrospective, single center studies.  The follow-up intervals vary widely.  The adjunctive use of other treatments are often not completely reported.

            And if we look at the published literature over the last 15 years, there have been really no trials at all comparing balloon angioplasty in any randomized format with other therapies.

            In the year 2000, in January in the Journal of Vascular Surgery, the TASC document or the TransAtlantic Inter-Society consensus document for the treatment of peripheral arterial disease was published, and this is a document which provides very strong guidelines for those of us who treat patients with peripheral arterial disease.

            And in the TASC document, they basically stated that there's really only one lesion subset that was recommended for balloon angioplasty for patients with critical limb ischemia, and that was the patient who presented with very discrete lesions in the femoral popliteal segment or an infrapopliteal segment.

            For those patients who present with more complex disease, such as the Type B patient with multiple short stenoses, the Type C patient with long stenoses or short occlusions, or the Type D patient with long tibial occlusions or diffuse disease, angioplasty was either not felt to be recommended based on insufficient data or surgery was the recommended therapy.

            And we'll discuss this in greater detail if we look at patients included in the LACI trial.  Eighty-eight percent of the patients or limbs in this trial had TASC C or TASC D type disease, very complex, diffuse disease with long occlusions.  So the type of patients that were felt not to be suitable for a balloon angioplasty.

            And, again, if we try and present some of the data from the literature with regards to balloon angioplasty for critical limb ischemia, looking at the last eight years or so, we see several reasonable studies with reasonable follow-up of at least six months in patients with critical limb ischemia, and most of these modern studies include use of balloon angioplasty, stents, a modern anticoagulant and antiplatelet regimens, and in many cases arterial closure devices.

            This is just a sampling of the literature, and if we would try and compare just in general terms some of these publications with the patient population treated in the LACI trial, in general, these were less complex and less long lesions, fewer lesions treated in patient, and overall in terms of the important endpoints of death and major amputation and re-intervention, I think that the results from the LACI trial compare very favorably with higher death rates and major amputation rates seen in these three studies.

            And, again, the study from Danielsson and Dorros published in 2001, again, less complex lesion subsets, similar results with regards to death or major amputation.

            So to try and summarize the literature for balloon angioplasty, all of the reports are single center, retrospective case series, and we know how that can sort of impact on the results of the data that is presented.

            Some of these articles report only on initial successes.  So those initial technical failures were not included in the final outcomes with regards to limb salvage and complications.

            So why not randomize this patient population to balloon angioplasty?  Well, the disease subsets that we're treating in this study, diffuse disease, multi-segment disease, long occlusions, were not recommended to be treated with balloon angioplasty in the TASC document, and there's really little evidence that angioplasty can be successfully used to treat patients with diffuse disease and critical limb ischemia who were felt to be surgical candidates.

            And, clearly, there is no study that has made balloon angioplasty the gold standard for the treatment of patients with critical limb ischemia.

            So we have basically a situation where there's really no one therapy that's appropriate or standard of care for patients with critical limb ischemia who are not otherwise good surgical candidates, and although certainly not optimal, it is true that those of us who were investigators involved in the design of this trial, we all felt pretty strongly that we could not randomize patients to balloon angioplasty, and that that trial design was unworkable.

            So we're left with either doing a self-controlled study design or a historical control study design.  Well, the main goal of the control group being that we want to demonstrate that we somehow do not alter the natural history of these patients in some negative way, either end up with greater incidence of major amputation because of our interventions or a higher mortality in those patients who are intervened on with the laser system.

            The ideal best case historical control would be an exact match in patient characteristics, a large enrolling study, full statistics with excellent follow-up, and a treatment plan that basically defines the standard of care in the community, which in this case would likely be a mixed set of modalities that might include revascularization with surgery, medical therapy, or other supportive care.  And that document should conform to the TASC definitions.

            Luckily, right around the time that we were trying to formulate the appropriate study design for the LACI Phase 2 registry, a study was published that we felt met fairly well our requirements for a good historical control group, and this was the ICAI study, which was an Italian, multi-center, randomized study of Prostaglandin E-1 and critical limb ischemia patients.

            Basically, they randomized over 15 patients to Prostaglandin E-1 versus standard treatment, which in this control group consisted of bypass endarterectomy in approximately 43 percent of patients and then medical therapy, in a few cases balloon angioplasty.  And this was felt to be standard or best care possible in the community for which these patients were treated.

            This study was published in the Annals of Internal Medicine in 1999 and did conform to the TASC definitions in good clinical practice.

            There are some important differences between this control group and the LACI registry group.  The ICAI trial enrolled critical limb ischemia patients regardless of their candidacy for surgery, and ultimately 35 percent of these patients received bypass surgery as their primary treatment option.

            As previously described, patients were enrolled in the LACI trial only if they were felt to be poor or nonsurgical candidates.  So really the ICAI treatment plan was not an alternative for these LACI patients, nor was it a fall-back plan for these LACI patients.  The LACI population was really not eligible for the same treatments, and hence may not have enjoyed the same outcomes as the ICAI patients.

            The ICAI statistics do represent though the benchmark, and if we believe that the ICAI  statistics reflect safe and effective therapy, then a treatment plan with equal statistics must also be safe and effective.

            We did set, I think, a very high bar for comparison.  If you look at the data from the ICAI trial, the mortality rates in this trial were lower than or at expected levels compared to the published literature for critical limb ischemia, and it was a very low frequency of major amputation in this group, lower than one would have expected based on historical literature controls.

            So we chose, I think, a very high bar to compare against when looking at the data from the LACI registry.  The control statistics are benchmarks, not a true measure of alternatives available to the LACI population.

            With that background, I'd like to present the LACI trial and the results from this trial.  I think it is worthy of mentioning that this is the very first study of its kind, a prospective, multi-center study evaluating a device for the treatment of the endovascular treatment of patients with critical limb ischemia.

            As we previously mentioned, all of the previous studies using angioplasty are retrospective, single center experiences.  We were treating patients with critical limb ischemia, Rutherford Category 4 to 6 who were felt to be poor or nonsurgical candidates.  We were treating lesions in the SFA, popliteal, and/or infrapopliteal arteries with adjunctive balloon angioplasty and optional stenting.

            The primary endpoint of the trial was limb salvage at six months and basically freedom from major amputation at or above the level of the ankle.

            The primary safety endpoint was death at six months.

            Patients were treated with a 2.2 or 2.5 millimeter Spectranetics peripheral laser catheter or any of the available coronary laser catheters.

            They were poor surgical candidates because of any of the following reasons:  either poor or absent vessel for an outflow anastomosis, basically very diffuse and distal vascular disease, so not good candidates for a distal bypass operation, or the absence of any venous conduit for bypass, or significant cardiac or medical co-morbidity that would place them at high risk for surgical mortality or complications, basically ASA Class 4 or higher.

            The enrollment period was from April of 2001 to April of 2002, and we ultimately enrolled 155 limbs and 145 patients.

            These were the enrolling sites in the trial.  This was a multi-national, multi-center study.  There were three sites from Germany, two of which contributed significantly to the enrollment in this trial.  Thirty-six percent of the site enrolled 68 percent of the patients in the study.

            These are the patient descriptors.  It has been said that there were no demonstrable differences in terms of morbidity or preoperative morbidity in the LACI trial, and one of the FDA reviewers focused on the fact that there was a higher incidence of smoking in the control group.

            But I think if you look very carefully at the patient demographics, there were very significant differences in terms of patient morbidities in the LACI group.  Almost half of the patients in the LACI trial were women, compared to 28 percent in the control group, and I think we're all aware of the poor outcomes for women who have critical limb ischemia and undergo revascularization therapies.

            There was a higher incidents of myocardial infarction and stroke in the LACI population, higher instance of hypertension, hyperlipidemia, and obesity.  And I'll focus here on the fact that 66 percent of the patients in the LACI group had diabetes compared to only 39 percent in the control group.

            I think we're all very aware of the worst outcomes in patients with diabetes mellitus with regard to limb salvage and outcomes following endovascular or surgical revascularization.

            Twenty-seven percent of the LACI patients and 30 percent of the control group had ischemic rest pain at the time of presentation.  Seventy-two percent of the LACI patients and 70 percent of the control patients had tissue loss at the time of the presentation.

            It is worthy of mention that seven percent of the patients in the LACI registry had Rutherford Category 6 limb ischemia at the time of presentation, basically major tissue loss at the time of presentation.

            We cannot from the ICAI study determine how many of those patients had major tissue loss at the time of presentation because they were classified according to the Fontaine classification.

            Of course, all of the patients had at least one reason for being a poor candidate for surgical revascularization in the LACI study by definition.  But actually one-third of the patients had at least two reasons for not being a good surgical candidate, and six percent of these patients had all three reasons for not being a good surgical candidate for revascularization.

            I'd like to present a couple of representative case examples from the study.  This is a 61 year old Hispanic woman with diabetes for greater than 20 years who had end stage renal disease and had been on hemodialysis for one year.  This is really the most difficult of all patient populations for us to treat, the longstanding diabetic with renal failure on dialysis.

            And she presented with multiple ischemic ulcers on both feet to the Arizona Heart Institute, where she underwent excimer laser assisted angioplasty on the 14th of August 2001.  This was her two feet at the time of presentation.  She had a very large ulcer on the planer aspect of her right foot, as well as a second ischemic ulcer on the heel.

            The left foot, again, had multiple ischemic ulcers, including a very large ulcer on the left heel.

            After a successful revascularization procedure at three months, the ulcer on the right heel had resolved completely.  There was significant improvement in the foot ulcer on the right and significant improvement in the foot ulcers on the left, and at six months complete healing of all of the ulcers on both feet.

            And I think this is a testimony certainly to a very good revascularization, but also to very good wound care.

            This is a patient from our institution who is a 45 year old female with longstanding diabetes and morbid obesity.  She weighed about 350 pounds, and my surgeons actually begged me to treat this patient so that they wouldn't have to operate on her.  She had distal popliteal and tibial occlusive disease with a painful ischemic second toe on the left foot.

            After a successful revascularization procedure, she had near complete healing of the toe at three months, complete healing of the toe at six months to the point where she was now able to put toenail polish back on her toes.

            As part of the LACI trial, we photographed patients at baseline and follow-up and performed digital morphometry to measure the area of the ischemic ulcers, and we have really collected an incredible library of photographs from before and after treatment, and I think these photographs really tell a very compelling story.

            This is another example from the trial of a patient with an ischemic ulcer in the lower leg.  This ulcer had been present for over six months and was not healing nor responding to standard therapy. 

            After a successful revascularization, near complete healing at three months, complete healing at six months.

            This is another example of a very challenging patient, a patient with a large ischemic ulcer on the heel.  After successful revascularization, significant improvement at three months and continued improvement at six months, although this ulcer is not completely healed at six months.  And this patient would be considered to have persistent critical limb ischemia in the trial, although clearly has had significant improvement after revascularization therapy.

            Not all patients can avoid minor amputation.  This was a patient with severe gangrene of the second toe and a deep and severe ulcer on the forefoot.  After a successful revascularization procedure, this patient underwent a transmetacarpel amputation with successful healing of that surgical incision at six months and avoidance of a higher level of amputation, either below knee or above knee amputation.

            In the LACI study, just over 40 percent of the lesions that were treated were in the superficial femoral artery.  Over half of the lesions were either in the popliteal artery or in the infrapopliteal vessels.

            We treated a mean of 2.7 lesions per patient.  The mean length of each lesion was 6.1 sonometers.  So we treated overall lesion length of around 16.4 sonometers in each of the limbs of these patients.  So these are very long, complex lesions.

            The majority, over 70 percent of the lesions treated or of the patients treated had a combination of stenoses and occlusions.

            As previously shown, 88 percent of the lesions treated in the study would have been classified as either TASC C or TASC D or subsets.

            The lesions could be successfully crossed with a guidewire in 92 percent of cases.  There were some cases where the lesion could not be crossed by the guidewire, and the laser catheter was used in  step-by-step manner to facilitate crossing of these refractory occlusions.

            Adjunctive balloon angioplasty was performed in 96 percent of cases and stent implantation was performed in 45 percent of cases.

            Procedure of success is defined as a residual percent diameter stenosis, and less than 50 percent of all of the lesions treated in a given limb was 85 percent.  And straight line flow to the foot was established in 89 percent of cases.

            The median hospital stay was only one day.  There were a few patients who had very long hospital stays.  So the overall mean hospital stay in the study was three days.

            Half of the improvement in luminal gain that was seen following intervention was a result of the excimer laser angioplasty.  Final percent diameter stenosis following all interventions ranged from 16 to 24 percent.  In general, in the infrapopliteal arteries, the final percent diameter stenosis was slightly higher.

            Stenting was performed preferentially in larger vessels.

            Again, I'll show a few representative angiograms from the study.  This is the normal infrapopliteal anatomy, anterior tibial artery, tibial peroneal trunk, peroneal, and posterior tibial artery.

            This is a typical patient from the LACI trial basically demonstrating occlusion of all of the tibial peroneal arteries without any visible distal vessels for bypass grafting, at least in this image.

            This is after the laser catheter.  A channel is reestablished in the anterior tibial artery.  Adjunctive balloon dilatations were performed in this vessel in multiple locations, with an excellent angiographic result, with wide patency now established in that anterior tibial artery.

            Another LACI type case, again, very diffuse disease below knee with occlusion of all three of the tibial arteries, some faint reconstitution of a diffusely diseased peroneal artery distally.  After laser assisted recannulization and adjunctive balloon dilatation, now straight line flow is established through this patent peroneal vessel.

            A little bit simpler anatomy, but again difficult lesions to treat with just balloon angioplasty alone.  Occlusion of the popliteal artery and sub total occlusion of the tibial peroneal trunk, post laser and nice channel is established in both of these lesions, and then following adjunctive balloon dilatation without the use of stents, an excellent angiographic result at the popliteal artery and in the tibial peroneal trunk.

            Again, another sort of representative case to give you an idea of the severity of disease treated in this trial for patients who have significant orthopedic work done.  Occlusion of the distal popliteal artery with diffuse disease and occlusion of all of the proximal tibial arteries with faint reconstitution of the small peroneal artery, again, the kind of case that would be difficult for surgical revascularization.

            After use of the laser catheter, a nice channel is established with now patency into the peroneal artery through this occluded popliteal segment.  Then after balloon angioplasty alone an excellent angiographic result with good flow and wide patency of the peroneal artery.

            And I think this is an example of a case where potentially with the use of laser recannulization, we can now demonstrate actually a very good distal target vessel which could potentially be used for surgical revascularization in the future if the patient did not heal his ischemic ulcers.

            Many patients were treated in the trial with long occlusions and the superficial femoral artery.  This is an example of a patient with total occlusion almost over the entire length of the right SFA, with reconstitution of the distal SFA adduct or canal.

            After passage of the laser catheter, we see a nice channel now through the superficial femoral artery with really a channel that is actually larger than the diameter of the laser catheter that was used, and I think this reflects the ability of the laser to vaporize some of the relatively acute or more chronic thrombus that's present in some of these SFA occlusions.

            After only adjunctive balloon dilatation, no stents, an excellent angiographic result is achieved.

            As I previously mentioned in the control group of the ICAI study, 35 percent of these patients ultimately underwent bypass surgery or endarterectomy.  Five percent received angioplasty, and three percent underwent a thrombectomy procedure.

            Fifty-seven percent of these patients underwent supportive care with analgesics, vasoactive therapies, and antithrombotic therapies.  Some of them received hypobaric oxygen.

            If we compare the groups with regards to serious adverse events, there's overall no difference between the two groups with regards to SAEs, no difference in terms of mortality; no difference in terms of major amputation or nonfatal MI or stroke.

            There is a higher re-intervention rate, 17 percent in the LACI group compared to four percent in the control group, although it is important that this four percent reflects the entire group as a whole, and if you look at those patients, the 43 percent of patients who underwent some surgical or angioplasty treatment, they had a re-intervention rate of 11 percent.

            I think if you consider the lesions that were being treated, 2.7 lesions per patient with an overall lesion length of 16 sonometers, a re-intervention rate of 17 percent I don't think is unacceptable.  In fact, I think it's actually quite good and would compare very favorably with any historical angioplasty study.

            I think if you think of some of the alternative therapies, such as surgical bypass, this also compares very favorably.  Multiple series in the literature, and some of these are included in your packet, show that within six months of the surgical bypass the re-intervention rate may be as high as ten to 20 percent, and I see these patient all the time who come in with intimal hyperplastic lesions at the proximal or distal anastomoses, retained valves in the graft, or actual acute graft thrombosis.

            Also, if you look at primary amputation as a mode of therapy, 19 percent of patients who undergo primary amputation for critical limb ischemia will require some revision of that amputation to a higher level within 30 days of that initial amputation.

            In this LACI registry, the incidence of acute limb ischemia or the need for bypass surgery endarterectomy was strikingly low during the follow-up time period.

            With regards to the primary endpoint of limb salvage, we can analyze the data in several ways.  If we analyze it on a per patient basis, we enrolled 145 patients in this trial.  Fifteen patients died during the follow-up period.

            It is important to note that only two of those patients had a major amputation before they died, and as we all know, a lot of these patients end up dying because of the cardiovascular complications of their severe and diffuse atherosclerosis.

            Eleven patients were lost to follow-up, leaving us with 119 patients who reached their six-month endpoint.  Major amputation was required in nine of the surviving patients, giving us 110 patients who survived with limb salvage.

            If we look at the limb salvage rates on an intent to treat analysis, which is basically the worst case scenario where we're considering all of our deaths and all of our loss to follow-up patients as a treatment failure and a limb loss, our limb salvage rate is 76 percent, and that is a number that you'll see presented by the FDA in their discussions today.

            But I think if we look at the more meaningful statistic, which is basically what is our limb salvage rate in our surviving patients, it is a strikingly high 92 percent, which compares very favorably with any study published in the literature.

            If we analyze the patients on the basis of or the study on the patients that had limbs treated, this would include 155 limbs, seven deaths, 11 lost to follow-up, nine major amputations.  Again, limb salvage on an intent to treat analysis, 76 percent; on a survival analysis, 93 percent.

            Then when we compare the LACI group with the control group with regards to our main endpoints at six months, there was no difference with regards to death, no difference with regards to limb salvage, no difference with regards to persistent critical limb ischemia, and  no difference with regards to serious adverse events.

            So at the very least, we did not alter the natural history of these patients in a negative way, increasing the mortality or the risk for major amputation.

            As mentioned, we performed digital morphometry from the photographs that were performed on these patients at baseline and during follow-up, and we saw a decrease in the area of the ischemic ulcers over the course of the study with 50 percent completed healed at six months.  Seventy-three percent of these ulcers were healed or improved at six months, and only 13 percent were noted to be larger or had new ulcers at the follow-up time period.

            And if we look at overall functional outcomes in this slide presented as the category at baseline versus the category at six months, where all of the patients are in green, reflect patients who are improved, and those in yellow reflect patients who are in the stable condition, and those in red reflect patients who have gotten worse during the follow-up time period, we see a great majority of patients who are improved and are in a non-critical limb ischemia category at the six-month follow-up period, and only a small percentage of patients have had progression of their critical limb ischemia and worsening of their ischemic ulceration.

            There's only one major or only one predictor of major amputation in the trial, and that's not surprising a Category 6 Rutherford classification or major tissue loss at the time of presentation, and only one predictor of death.  That was advanced age at the time of presentation.

            One of the confounding variables obviously in this study is what was the impact of stenting in the study, and we certainly would have liked to have not had any stents implanted in the study if at all possible.  But in the modern era it's just impossible to keep investigators from placing stents for full vessels.

            It did appear that those vessels that were stented had a higher procedural success compared to non-stented vessels and a greater incidence of straight line flow compared to non-stented vessels.  But there was no statistical difference, statistically significant difference with regards to limb salvage in those vessels that were stented compared to those that were non-stented.

            And the same size is small enough that it does make, you know, definitive analysis of this data difficult.

            So in conclusion, we were treating a very complex patient population, patients with multiple lesions, often long occlusions, diffuse disease, who were felt to be poor surgical candidates because of the diffuseness of their disease, their significant co-morbidities, with the lack of venous conduit for a bypass surgery.  So we were taking basically the worst of the worst, patients with critical limb ischemia who had the worst of all possible disease.

            Despite that, we had very high procedural success rates with few in-hospital serious adverse events and short hospital stays, and an excellent limb salvage rate despite treating a very high risk patient cohort.

            The outcomes in this study met all of the hypotheses in the protocol, and the statistics met the benchmarks of safety and effectiveness.

            So if we're going to try and define the clinical benefit of this strategy of laser assisted angioplasty, I think we can say that this strategy did save limbs in the study with an efficacy endpoint that equaled the controlled benchmark without affecting the patient's chances of survival or significantly increasing these patients' risk of serious adverse events.

            And as an endovascular approach to the treatment of vascular surgery, it avoids the perioperative risks and morbidity associated with surgery.  It shortens the initial hospital stay and quickens the return to normal function, and it does not jeopardize any further surgical options for these patients.

            And as we demonstrated in that one case, it may, in fact, create surgical options for these patients by opening up occluded tibial arteries and demonstrating an adequate distal vessel for possible future surgical revascularization.

            The results in the LACI Phase 2 trial were achieved really with virtually no surgery.  Less than three percent of these patients went on to have either bypass or endarterectomy during the six-month follow-up time period.

            As we've seen, the LACI strategy is applicable to a wide range of vascular disease states.  We basically were taking on all comers with the worst of all possible disease, with lesions that would generally not have been considered amenable to just balloon angioplasty alone.

            And I think this is a strategy that is useful for patients as a last ditch or last option prior to major amputation. 

            The results were predictable with very high procedural success, and I think remarkably low serious adverse events.

            And with that I'll conclude my presentation.  Thank you for your attention.

            CHAIRMAN LASKEY:  Thank you.  That was excellent.

            Although I'll remind the panel that everyone will have their opportunity to provide comments and insights later on, are there any panel members who have any questions for the sponsor and/or Dr. Laird at the moment?

            Mitch, yeah.

            DR. KRUCOFF:  John, just one quick question.  You guys have a reasonable number of patients who were sort of roll-in or learning curve.  Could you tell us a little bit about your experience or share any data you have on what the learning curve in applying this technology in these patients looked like compared to the --

            DR. LAIRD:  That's a very good question.  I think it is important to recognize that around the time of this study initiation there were not that many operators in the United States who had experience with laser angioplasty in the periphery.  In fact, there were only a few places where this was being done.

            And as part of the both this trial and another trial, there was a roll-in period where patients were treated, some with LACI type anatomy, some with less complex disease, and there was a LACI Phase 1 trial which included roughly around 20, 25 patients who were treated with critical limb ischemia with somewhat different enrollment criteria.  It did not include patients with critical limb ischemia.

            And overall -- I don't have the data, and I can't present it to you -- but overall, the results were favorable in that study.  And I think a lot of these very good results were achieved despite the fact that there wasn't a long historical experience with laser in the centers that were included in the trials, reflecting, I think, a relatively short learning curve for the use of this device.

            CHAIRMAN LASKEY:  In that regard, you had the two German centers which were preponderant in enrolling.  Did they just have a longer track record?  Is that why?  What was --

            DR. LAIRD:  Well, I think one of the sort of pioneers in the use of excimer laser angioplasty for peripheral arterial disease is Giancarlo Biamino, and his site was one of the sites that was included in this trial.

            The referral mechanisms in Germany, I think, are a little bit different.  The patients, you know, tend to get referred to the larger centers, and so they really had a remarkable, you know, referral base from around Germany where they were getting patients from all over the country, difficult patients with critical limb ischemia.  So they tended to have more patients than a lot of the sites in the United States to be included in the trial.

            CHAIRMAN LASKEY:  Yes, Dr. Somberg.

            DR. SOMBERG:  This type of therapy is highly specialized and is not, as you point out, an alternative in terms of something that could be just whipped in as an alternative to.  So what would happen either prior in your center or concomitantly in other centers to these type of patients?  You didn't really draw that out.

            You said they were not candidates for balloon angioplasty.  At this point would they have all gone to amputation?

            DR. LAIRD:  I think a significant percentage of them would have gone to amputation.  It's possible that some of these patients may have been referred for surgery and undergone a distal bypass with synthetic graft, and I think we're all, I think, aware of the generally poor results when synthetic grafts are used for long bypasses in the popliteal arteries.

            Some of these patients may have undergone balloon angioplasty, but I think our expectation from the literature trying to treat patients such as this with balloon angioplasty, the results would not have been very good, and some of them would have undergone the other sort of modalities that are tried, such hyperbaric oxygen, some of the newer drugs for ischemic ulcers, things like that.

            CHAIRMAN LASKEY:  Dr. Aziz.

            DR. AZIZ:  John, I see that in your presentation there were no perforations.  I think there were zero.  Are you aware of in any other centers, maybe even outside this trial, there was a perforation of the vessel?

            DR. LAIRD:  I thought actually there were a couple of perforations in the study.  I don't know if I have reflected on the slide, but there were no serious adverse outcomes related to perforation.

            In a related experience, we've done the PELA trial for long occlusions.  We did see perforations in roughly eight percent of cases, none of which really had any significant sequelae because we're often in the middle of a long occlusion when this occurs.  So the patients don't bleed, and it's a self-sealing perforation.

            I think it's worthy of mention that the incidence of distal embolization in this trial, I think, was also remarkably low.  It's three percent, and when you're talking about treating long occlusions with a total lesion length of 16 sonometers in these limbs, that rate of embolization is very, very good.

            CHAIRMAN LASKEY:  Yes, sir.  Dr. Nicholas.

            DR. NICHOLAS:  A question regarding the definition of "inadequate venous conduit."  Were these patients looked at with duplex for upper extremity, lesser saphenous, contralateral vein?

            DR. LAIRD:  We really don't have the full data on that as to whether upper extremities were looked at in terms of inadequate upper extremity conduit.  I think that sort of practice is very much, you know, center dependent.  I think there are centers who specialize in using veins from the upper extremity and can do good work with that, but not all centers.

            And so it was pretty much left up to the individual sites to make the determination of whether, quote, unquote, adequate venous conduit was present.

            All of these patients had no saphenous vein for bypass.

            DR. NICHOLAS:  While I'm asking, one other question.  In ASA Class 4 patients, how many had that as the sole exclusion?  Because we do a fair number of those.

            DR. LAIRD:  Forty-six percent of patients were in ASA Class 4 or greater.  Now, some of those patients may also have had another reason, either lack of vein or poor distal targets, too.  I don't know that we know for sure just the pure ASA-4.

            Chris?

            But at least 46 percent of them had that as one of their criteria for being in the trial.

            CHAIRMAN LASKEY:  Well, I'll be the bad guy here.  You glossed over the statistics, although I'm sure that will get dredged up again this afternoon.  Could you just tell us why a delta of ten was felt to make everybody happy with this?

            DR. LAIRD:  I think I'm going to punt that one to Chris.

            DR. REISER:  When we designed the trial, it was anticipated that the basic population would exhibit significantly more co-morbidities than the ICAI control population.  As Dr. Laird presented, we believe that turns out to be true.  That was the rationale that the Steering Committee gave me for the delta of ten.

            As Dr. Laird points out though, we didn't see in the LACI results that the results were inferior to the control population.  So we didn't invoke the delta of ten when we inferred that the results were equivalent.

            CHAIRMAN LASKEY:  Let's make it short because you'll have another opportunity.

            DR. NORMAND:  Okay.  I'll just ask a very straightforward question, I think.

            Had the two populations been randomized or really comparable, what would the delta have been?

            DR. REISER:  I don't think we ever had that conversation.

            DR. NORMAND:  Well, I'm just asking now.  I'm trying to understand the choice of a delta, and part of helping me understanding the choice of that delta, I guess I would wonder had I -- I don't want to say "done it right" -- but had I been able to randomize or had really truly comparable patients, I guess that's the way I would think of the delta.

            And if you haven't thought about that, that's okay, but in terms of me trying to understand that, that would be a useful number.

            CHAIRMAN LASKEY:  I'm sorry.  Are you thinking about a response or just --

            DR. REISER:  No, we haven't had that conversation.  So I really don't have a good response for that.

            CHAIRMAN LASKEY:  Okay.  Dr. Ferguson.

            DR. FERGUSON:  I was going to ask this later, but it seems appropriate now.

            It was a very nice presentation.  I'm curious as to why you were not able to get the raw data from the Italian study because there's several points that come up during the course of what I've read and what you've said that would make it at least better had you known what the raw data were, and just curiosity.

            DR. REISER:  Frankly, it's fairly rare in my experience that a study coordinator actually gives away the raw data.  It has been my experience that you can buy an analysis of data.  It depends on how much you would like to pay, but I as a sponsor have never received raw data from any other sponsor coordinator.

            DR. FERGUSON:  I thought we were all scientists.   I don't understand that.

            DR. LAIRD:  Well, that was an industry sponsored trial, and they may not have been so forthcoming for that reason.

            DR. FERGUSON:  I would just comment.  I think you're limited to have to go to a journal to gather your data.

            DR. LAIRD:  Agreed.

            CHAIRMAN LASKEY:  Great.  We are moving right along nicely.  If it's okay with the transcriptionist, can we just move to the next portion without the break?  We'll take the break after the FDA presentation.  Is that okay?

            Great.  Thank you.  Thank you, gentlemen.

            And I'd like to invite  the FDA.

            (Pause in proceedings.)

            CHAIRMAN LASKEY:  We could have taken a bathroom break at this point, eh?

            Your clock.

            (Laughter.)

            DR. HOLDEN:  I'm not sure where we got this computer, but it's a great ad for Windows XP, I suppose, that probably loads faster.

            CHAIRMAN LASKEY:  I've been suggesting McIntosh for years up here, but I don't --

            (Laughter.)

            DR. HOLDEN:  Finally.  Good morning.  This presentation will include a brief summary of FDA's review of the preclinical, clinical, and statistical data for Spectranetics' PMA supplement application.

            Well, good.  Apparently we have to reboot this and go through this procedure again.  Maybe a --

            CHAIRMAN LASKEY:  I'm reluctant.  If we do a break, we may lose some.  We're three minutes on a reboot, right?  It shouldn't be --

            DR. HOLDEN:  I suppose.  I didn't --

            CHAIRMAN LASKEY:  Well, if nature calls, nature calls.  Let's see if we can keep right on schedule or before schedule.

            DR. HOLDEN:  One suggestion.  Perhaps while this is happening we could begin the presentation if you all have handouts of our slides while this is happening.

            CHAIRMAN LASKEY:  That's fine up here.

            DR. HOLDEN:  Mine in particular are all -- let me get a copy.  So while this is happening, I'll speak to slide one, meaning top right.

            Again, the presentation will include a summary of our reviews of the data for Spectranetics' PMA supplement application for its excimer laser system to treat critical limb ischemia.

            On the next slide my  name is Dr. John Holden.  I'm an engineer and a review scientist in the Division of Cardiovascular Devices, and I'll begin with just an introduction and a very brief description of the preclinical evaluation.

            And as an overview, I'll describe the FDA review team, a brief history of the clinical trial, and the PMA supplement application, and then FDA comments on the various evaluations, in particular, the clinical and statistical reviews, and we'll conclude with just a few key points from the clinical and statistical summaries.

            The next slide in the lower left shows the FDA review team for this PMA supplement.  Dr. Sapirstein will provide FDA's clinical comments, and Dr. Krasnicka will give the FDA's statistical summary.

            And you should note that there were quite a number of other FDA reviewers who were involved in the review of the IDE application and of earlier supplements to this PMA application.

            The currently proposed indications for use are as follows, as shown on the slide in the lower right:  for facilitating limb salvage in patients with critical limb ischemia associated with Rutherford Categories 4, 5, and 6 who have angiographically evident culprit stenoses and/or occlusions in the SFA, popliteal and/or infrapopliteal arteries, who are poor surgical candidates and who are acceptable candidates for revascularization.

            The next slide at the top of the second page of our presentation.  The device, again, is the CVX-300 Excimer Laser System, plus 15 models of Spectranetics' excimer laser atherectomy catheters.  Three types of these ELA catheters have been evaluated in the LACI trial, over the wire, graft exchange, and eccentric models.

            All of these models, as the sponsor described, include a proximal end that couples exclusively with this CVX-300 Excimer Laser, and then that system is controlled by software that instructs the laser to deliver the correct energy for each particular catheter model.

            Ten of the catheter models have been previously approved for use in coronary arteries.  Five of them are new catheter models that are specific to this use in peripheral arteries to treat critical limb ischemia.

            And if I might take just a break to catch up.

            Okay.  The applicant refers to its IDE study as the LACI trial.  It was conditionally approved; Phase 1 was conditionally approved in December in 1998 as a feasibility study with 25 patients, and that study was completed using the device to treat 25 limbs from 23 patients.

            An expansion of the study was conditionally approved in January 2001.  This pivotal trial was proposed by Spectranetics who was a single arm registry to treat 167 patients.  Thirty of these were reserved for training, although not all of those were used.

            As described by the sponsor, the trial used as an historical control the control group from a publication by the ICAI group in Italy.  That publication is listed in the middle of this slide.

            The primary effectiveness endpoint was freedom from amputation at or above the ankle at six months, and the primary safety endpoint was death within six months of the laser procedure.

            FDA's conditional approval letter for the pivotal trial included several comments and recommendations to the sponsor as noted here on this slide.  The trial began that a risk-benefit analysis would be needed for the PMA application and should include an analysis of all device and procedure related adverse events.

            We suggested that this analysis also quantify the purported benefits of the device, for example, reducing stent use and/or the creation of surgical options.

            And we indicated that it would be necessary to show that stenting, which was to be discouraged in the trial, did not confound the analysis of the study endpoints.

            In February of 2002, three models of the Extreme II catheters were added to the trial.  Also, the lacing parameters for these catheters were changed from a cycle of five seconds on, ten seconds off to ten seconds on, five seconds off.

            The laser catheter usage in the trial is indicated on this slide.  There were a total of 203 laser catheters used.  Note that four of the catheter models were used only once in the trial.

            This led to a PMA supplement application that was submitted in January of this year with the clinical data from the LACI II study.  Since that time we've had a highly interactive review process, including a face-to-face meeting with the sponsor to discuss FDA's questions related to the risk analysis, the safety and effectiveness results, and certain aspects of the clinical protocol, the device software, and the instructions for use.

            And this interaction has led to today's consideration for the Advisory Panel's comments and recommendations.

            The peripheral catheters underwent a variety of preclinical testing, including tissue ablation and tip integrity testing, artery model and collateral tissue testing, some software validation, biocompatibility, and qualification of the sterilization process.

            And at this time there are no additional questions from the FDA about this preclinical testing.  So at this time I'd like to introduce Dr. Wolf Sapirstein, who will provide FDA's clinical review summary.

            DR. SAPIRSTEIN:  Good morning.  In the spirit of full disclosure or possible conflict, I must say that I am a cardiovascular surgeon by background, but I count many interventional radiologists and cardiologists amongst my friends.

            (Laughter.)

            PARTICIPANT:  Is that a waiver?

            (Laughter.)

            DR. SAPIRSTEIN:  Okay.  The sponsor has described in very great detail the laser trial that was undertaken to assess the use of laser energy for ablation of vascular obstructions in patients suffering from critical limb ischemia in end stage arteriosclerotic disease.  The patients were felt to be poor candidates for surgical revascularization.  As such, they presented a desperate situation such that primary amputation can be a management option.

            The LACI study single arm design with a historical control was thought to be ethically necessary where a randomized controlled trial was ruled out for lack of clinical equipoise.

            This has resulted in an outcome open to conflicting interpretation which does not, however, detract from the excellent conduct and monitoring of the study by the sponsor and the investigators.

            Critical limb ischemia was determined by patients enrolled in the study in Rutherford Classes 4, 5, and 6, which were comparable to the Fontaine Stages III and IV used in the ICAI study.  Raw data from the non-treatment control arm of ICAI that was selected for the historical control of LACI  is not accessible.  This is a common problem when using historical controls generally, but is particularly troublesome in denying robust comparison of many important secondary endpoints for evaluating treatment of critical limb ischemia.

            The sponsor justifies this design because LACI patients are perceived as not suitable candidates for surgical intervention.  However, an extensive literature review suggests that these patients can be very effectively managed with a variety of alternative treatments.  This is especially evident for patients available to an endovascular procedure, particularly when an infragenicular vessel is present, which is a LACI entrance criterion.

            And international consortium of experts in management of peripheral arterial disease, the TASC group, has published evidence based recommendations for a study on management of this condition.  Critical issues are listed for situations where such evidence is as yet not defined.

            This is the case, for example, for the treatment of lesions described as Type B and C.  TASC recommendations has become an accepted standard for treatment of peripheral vascular disease and study of peripheral arterial disease to which both ICA and the LACI study groups have attempted to conform.

            The LACI study enrolled patients with critical limb ischemia who were considered unsuitable for surgical revascularization defined by at least one of three entrance criteria:  operative risk for surgical survival of at least American Society of Anesthesiology physical Class 4 or higher; absence of a suitable autogenous vein conduit; or by the extent of disease.

            In fact, only 46 percent of patients fell into the ASA Class 4.  Only 32 percent had no suitable autogenous vein conduit, and many of these criteria existed in the same patient.

            Forty-one percent of lesions were in the superficial femoral artery, and 27 percent in the popliteal artery or the peroneal tibial trunk.  The mean number of lesions per limb was 2.7, and the mean length, six centimeters.

            Many of these lesions are often TASC Class B and C for which the role of endovascular treatment has not been established, as listed in TASC critical issue number 14.

            The LACI exclusion requirement of a patient infragenicular artery is a factor favorable for PTA where bypass is not an option.

            The surgical risk status is claimed indicative of co-morbidities placing LACI patients in a higher risk category for a poor outcome.  Therefore, demonstration of equivalence to the selected control, many of whom underwent a surgical procedure, is a conservative estimate of effectiveness that can be disputed.

            The ASA classification classifies the patient's general clinical status by probability for postoperative survival, which may be unrelated to the surgical intervention.  Although linked to co-morbidities, it does not necessarily impact on the risks for treatment of a regional condition, such as critical limb ischemia or under benefits derived from such treatment.

            Univariate analysis of LACI outcomes identified that the only statistically valid predictors of outcomes were age for mortality and gangrenous changes for amputations, both of which occurred as similar baseline incidences in LACI and the ICAI control.  A smoking history was more prevalent in the ICAI control, while other risk factors occurred more frequently in LACI.

            However, many of these risk factors were moot considering the patient's ischemic level of Class 4, 5, and 6.

            Delivery of the laser energy is controlled over a guidewire.  The guidewire successfully navigated the culprit lesions in 85 percent of cases, and the step-by-step laser procedure was necessary in only 13 cases.

            Procedural success defined as equal or less than 50 percent residual stenosis was achieved in 91 percent of cases, but required adjunctive PTA.  It is, therefore, conceivable that PTA alone may have been affected  and the laser angioplasty superfluous.

            Use of stents in 70 percent -- I'm sorry -- in 70 cases, 46 percent of the LACI patients, the majority placed in the superficial femoral artery further clouds the issue.  Although similar levels of salvage was achieved in stented and non-stented limbs, stents may have significantly impacted the short-term, six-month limb survival.

            Block randomization insured that cases were withdrawn equally from both arms.  The race for the outcomes have been calculated for patients not lost to follow-up.  Re-intervention was required to maintain limb perfusion, was significantly higher in LACI.  The rehospitalization necessary can only adversely impact on the already constrained quality of life.

            During the six months of follow-up, persistence of critical limb ischemia categories occurred in 30 percent of patients.

            The sponsor claims that the review of literature base for alternative treatments supports both the LACI study design and the clinical advantage of laser arthroplasty in these patients.

            These are extrapolations from heterogeneous studies and patient populations.  Disease conditions were also heterogeneous, and studies varied considerably with follow-up evaluations that ranged from 30 days to five years.

            Furthermore, these reports cover two or three decades, and many antedate the more recent significant advances in both the technology and clinical management of peripheral arterial disease.

            The benefit of limb salvage is not in controversy.  However, a documented need for amputation, end stage critical limb ischemia, to manage local and/or systemic effects of ischemia is recognized in the TASC recommendations, and it's critical issue number 45.

            Medication absent an alternative is a procedure essentially of desperation as studied in the ICA trial.  Newer drugs can provide short-term palliation and perhaps adjunctive benefit to borderline revascularization procedures.

            Bypass surgery remains the gold standard for critical limb ischemia.  Considerable success has been documented for periods extending to five years for aggressive revascularization, including a pedal bypass.  Modifications in surgical approach and advances in use of alternative conduits, together with newer radiological active drugs may have permitted improvement in tissue perfusion over the short term to allow healing and then salvage even in the absence of a suitable autogenous vein.

            Improvement in endovascular interventions also fueled by technology have generally excellent results in critical limb ischemia for even the periphery located disease.

            In summary, therefore, why the LACI study single hypothesis of survival with limb salvage in critical limb ischemia at six months has been met, any benefit claimed for this modality is diminished by the re-intervention rate and the persistence of critical limb ischemia as symptomatology.

            The value added to a management strategy employing only percutaneous angioplasty and not utilizing laser therapy remains uncertain.

            Thank you.

            I would now like to introduce Dr. -- pardon me.

            DR. KRASNICKA:  Good morning.  My name is Barbara Krasnicka.  I am an FDA statistician.

            In my presentation, I will focus mainly on the problems connected with the study design and statistical analysis.

            As Dr. Sapirstein already mentioned in his presentation, the objective of this clinical study was to determine safety and effectiveness of the use of excimer laser for ablation of vascular obstructions.

            The primary effectiveness endpoint was the percentage of live patients without major amputation at six months.  The primary safety endpoint was any death occurrence during six months of the follow-up period.

            I will discuss only some issues connected with the statistical analysis for the primary effectiveness endpoint and for one of the secondary endpoints, namely, the survival time to six months of the follow-up.

            Note, again, that this study was designed as a non-randomized clinical trial with one prospective treatment group called LACI and an historical control group based on a single published paper in 1999.

            Please note there is no data available at the individual patient level for the control group.  Only summary statistics, such as estimated proportions, means, standard deviations, and figures were accessible from the paper.

            Now let me summarize again some general information about the study.  The prospective LACI trial was carried out at 14 sites, 11 located in the U.S. and three in Germany.  Altogether, 145 patients with 155 treated limbs were enrolled in the treatment group.

            However, during the six months of the study, 11 patients dropped out of the study.

            The historical control study was carried out at 56 sites in Italy.  Six hundred seventy-three patients enrolled in this study were considered in the analysis.  Only seven patients were lost to follow-up in the control group.

            Essentially, the sponsor's objective was to show that the results in the treatment group were at least as good as for the control one.  FDA agreed to accept the equivalence design based on the  assumption that the control patients would be less sick than the LACI patients.

            The results of the statistical analysis on the primary effectiveness endpoint are as follows.  The percentage of alive patients without amputation for the LACI group was 75. 9 percent, and for the control group was 73.4 percent.  The 95 percent two-sided confidence interval of the difference of limb salvage rates for the two intent to treat populations was roughly minus 510.  This means there was no statistical difference between the two groups.

            Let me now pose a question.  Should we accept the point estimates and confidence interval without any restrictions?

            The study was not randomized.  Therefore, the real treatment effect is uncertain.  The LACI and control groups are not comparable.  Patients' smoking history and previous major amputations were statistically more prevalent in the control group than in the LACI group.

            However, more diabetes, prior stroke and obesity were noted in the LACI group of patients.  Additionally, the LACI and control studies were carried out in different countries and hospitals, with likely differences with respect to manufacturers connected with the patients.

            All of the above-mentioned factors may impose a bias on the results.

            Now the question is whether we could overcome the bias problem.  Information on the historical control study was based on a single published paper without any possibility to use the individual patient's information.  Therefore, the visible differences at baseline between these two groups of patients could not be taken into account in the statistical analysis.

            It is worth to mention that small differences in important covariates, like previous major amputations could reasonably explain the differences in outcomes in treated and control groups.

            Also, a formal sensitivity analysis for hidden bias due to the non-randomized study is impossible to perform because of the lack of the raw data of the patient level.

            Additionally, in the LACI group, the laser treatment was not the only single procedure applied.  All patients received also balloon angioplasty and 45 percent of them received stents.

            In the control group, the treatment was conventional medications for blockage of arteries.  Only sometimes the treatment included a bypass surgery and/or other procedures at the time of enrollment.

            Due to the concomitant procedures that were not included in the control group treatment, it is very difficult to evaluate the effectiveness of the use of the laser device alone.  For this reason, the statistical analysis given by the sponsor does not uniquely prove the advantage of the laser therapy.

            One of the secondary endpoints of the study was survival time during a six-month follow-up period.  The average survivor functions for the LACI and control groups were approximated using Kaplan-Meier estimated.  Results are shown in the figure prepared by the sponsor.

            The upper curve corresponds to the LACI group and the lower one to the control group.  The visual impression is that the LACI patients could survive longer than the patients from the control group, but based on the Wilcoxon test, the difference between these two groups is not significant at the .05 level, P equals .17.

            Now the question is how perfect this survivor analysis is.  It is very well known that heterogeneity between patients may have considerable impact on the estimation of the difference between the two treatment groups.  In the LACI group, the range of patients' age was 41 to 91.  Evidently even the LACI group of patients is not a really homogeneous one.

            It is worth nothing that the stratified comparison of survivor times with or without adjustment for important covariates could solve the problem of biases in survivor analysis.  However, it is impossible to apply such methodology without having access to zero data at the patient level.

            As I already  mentioned before, the laser treatment was injunctive with  PTA.  Because of all the above problems, the survivor analysis is also questionable.

            Thank you very much.

            And now Dr. Sapirstein will present the clinical and statistical summary.

            DR. SAPIRSTEIN:  Thank you, Dr. Krasnicka.

            Well, in conclusion, the sponsor has demonstrated equivalence to the selected control for the endpoints of survival with salvage of the limb.  The data does not, however, support an assumption that LACI registered patients were at greater risk than the control on which the hypothesis was based.

            The non-randomized study design dictates caution in analyzing the data.  The extent to which non-assessed covariates influence effectiveness of the laser treatment is difficult to determine.  Re-intervention rate impacts negatively on the risks and the benefits of the laser intervention.

            Thank you very much.

            CHAIRMAN LASKEY:  And thank you.

            Questions from the panel members for the FDA presenters?  Yes, Sharon.

            DR. NORMAND:  I have a question about I think it's slide 29, and this is where the FDA apparently agreed to accept an equivalence design where, in fact, the patients were different.  And perhaps you could elaborate, someone could elaborate on the reasoning for doing that, on the one hand, and, on the other hand, it seems that you're criticizing the sponsor or there are questions raised about the sponsor on the second half for the effectiveness endpoint, saying that, in fact, the patients are different.

            So I guess if someone could help me understand maybe the decision to accept a design in which the patient populations are different.

            DR. SAPIRSTEIN:  We discussed this with the sponsor.  They felt that they could not develop a control other than this historical control, except that they were accepting of the fact that these patients were permitted to enter the study with multiple interventions.  They felt that this was a more rigorous control of their study, which they felt they could justify with a, I think, ten percent delta.

            CHAIRMAN LASKEY:  This may be something at Dr. Zuckerman's level, but traditionally I think of pivotal trials as randomized.  Could you give us some insight or guidance into non-randomized pivotal trials?  What is an acceptable study design at this level that is acceptable to the agency?

            DR. ZUCKERMAN:  Okay.  Let's take a step back.  There's nothing in our regulations that says that the pivotal study design for a PMA device or PMA supplement needs to be a randomized controlled clinical trial.  Certainly from a clinical trial design, an agency perspective we would prefer this study design for the obvious reasons.  The obvious reasons are at the end of the day, it's frequently the easiest way to analyze in a fair manner any differences between control and experimental therapy.

            Now, as Dr. Sapirstein has recollected, the sponsor and their investigators believed at the time of design of this trial that a randomized controlled clinical trial for this patient population would not be a doable one.  As such, they attempted to develop a control group that they thought could provide a fair comparison with their experimental therapy.

            As pointed out by multiple FDA comments, when one goes to  non-randomized designs, frequently the number of questions regarding a fair comparison of control in experimental groups seems to increase, and that's one of the things that the panel will deal with.

            The other thing to note is, again, going back to  Dr. Holden's initial comments, although the agency can give conditional approval for a pivotal trial design, and in this situation that's what happened, the usual FDA letters always have a final paragraph that explains that while the trial can go forward, this is not a total FDA endorsement, and several caveats were mentioned by Dr. Laird in his initial presentation.

            CHAIRMAN LASKEY:  Dr. Somberg.

            DR. SOMBERG:  What I would appreciate from the FDA, whoever wants to take this, is some further clarification.  I understand that the equivalency hypothesis was accepted with a caveat, as I stated by  Dr. Zuckerman, but was that for safety or was that for efficacy? 

            Because I'm confused in that there was a selection of a control group for a pharmacologic study where essentially there was current practice indications for whatever was done.  So you were trying to be current with a trying to be current practice in this control group, whatever that might be -- and that's, of course, varied in the study, and it's current practice in Italy in the early '90s one would have to say as opposed to the United States approximately ten years later.

            So is that an efficacy acceptance of equivalence or was it a safety acceptance, the latter, I think, being more logical?

            DR. SAPIRSTEIN:  It was an acceptance of an effectiveness equivalence, feeling that that was the basis for the study design, the belief that these patients were in a desperate situation of inevitable limb loss.  So it basically was an effectiveness measured by limb salvage.

            DR. ZUCKERMAN:  Does that answer your question, Dr. Somberg?

            DR. SOMBERG:  Well, if it was a superiority to that group, the answer is no.  If that was a superiority to that group, I would understand that, to the control group, but if it was equivalent to the control group, I don't see how that could prove effectiveness.

            You know, the internal parsimony in that study is that Prostaglandin E-1 was superior to standard of care therapy.  I think what the sponsors were trying to do was show that laser angioplasty systems was superior, but if it was equal to the standard of care, how does that prove that something was done that was effective in the FDA statement of effectiveness, you know, by the Federal Register?

            DR. ZUCKERMAN:  I think what you have to accept -- and, again, this is for panel discussion -- is that there was an implicit belief that the control arm from ICAI would be a less sick control arm, and therefore, if there was shown in the sicker experimental group equivalence to that less sick control arm, these were acceptable results.

            CHAIRMAN LASKEY:  Dr. Krucoff first.

            DR. KRUCOFF:  I think what Bram just said is key to trying to figure out where we are.  The hypothesis here of equivalence literally, as I understand it, and I had to go through this pack twice because I didn't get it the first time around, is that essentially doing something in a higher risk group would be equivalent to doing nothing in a lower risk group.  And at least statistically you're going to have to help because I don't know how to do that.

            But I think that's the issue that seems to be on the table in this data set, and it really gets back to the question of how do you deal with equivalence in different patient populations,b ut I guess we'll have time to go back through some of that.

            I just wondered from a statistical point of view is there a way of doing this.  I mean, are you actually -- can you share with us a statistical approach that defines equivalence in a higher versus a lower risk patient population where you're intervening in one and not in the other?

            Oh, we're going to a higher authority.

            (Laughter.)

            DR. GERRY GRAY:  I'm Gerry Gray, the team leader for Cardiovascular and Ophthalmic Statistics.

            And the answer to that is, of course, -- well, the answer to that is really no.  I don't think that it's meaningful to show that equivalence when the patient populations are that much different.  It's kind of hard to -- I'm not sure how you would actually interpret the results of that even if you were successful in showing equivalence.

            DR. KRUCOFF:  Okay.  Then just a brief follow-up to you or Bram or whoever, Gerry.  So my sense of the process is the sponsors go to their investigators, their steering committee.  They get their wisdom, and they come back and say this is the only way you can do a trial ethically or whatever in this patient population.

            And you guys send a letter that says, "We will conditionally let this go ahead.  However, here, in fact," as John Holden showed with us and is in the pack, "here are some thins that we suggest that you also do," like the risk-benefit analysis, et cetera.

            But then the sponsor actually going ahead with the trial, I mean, is there any further dialogue after that letter or they just go ahead and then here we are today?

            DR. ZUCKERMAN:  I think there is always dialogue with the sponsor.  At least I seem to find that to be the case, but you know, in the end the sponsor was prepared to take a certain risk with this trial design.  It's a difficult situation for evaluation, and that's why, you know, the eventual trial results are here at this panel deliberation.

            The agency is looking for outside expert advice.

            CHAIRMAN LASKEY:  All right.  Dr. Morrison.

            DR. MORRISON:  Well, I'd like to take a risk and add one more level of complexity.  I've heard non-random, non-comparable, equivalence versus superiority.  My question at the end of the day is:  why is the control group medical therapy as opposed to angioplasty, particularly given that 83 percent of these patients had wire crossing and 98 percent of them received balloon angioplasty and 45 percent received stents?

            It seems to me there are three different levels of difficulty here.  One is noncomparable groups.  The other is why is the addition of laser not better, but secondly, why compare it to medical therapy as opposed to angioplasty?

            And that's addressed to anyone.

            CHAIRMAN LASKEY:  Well, we're supposed to confine our questions to the FDA at the moment.  I think Spectranetics is twisting in the wind here, but we should confine this to the FDA.

            DR. MORRISON:  Did the FDA agree to that as part of the agreement in the first place, that this particular study would be the control group, or was it just an agreement that in principle some control group?

            DR. ZUCKERMAN:  When designing a study with a sponsor, there are always multiple factors that are considered both at the FDA and sponsor, and certainly the agency did provide caveats in its interaction with the sponsor, but this is the design that we have at the end of the day.

            DR. SAPIRSTEIN:  And to add to that, PTA was an adjunctive intervention in this condition.  So we felt that it was an adequate control to have this historical control where a multitude of interventions were permitted at the initiation of the ICAI control included PTA and surgical intervention.  So we thought it was an adequately robust control.

            CHAIRMAN LASKEY:  I guess for those interventional people on the panel here, if you get a wire across something, you can follow it with anything, and I guess we're just sitting here wondering, as did everyone else, why not a balloon.

            We've heard the track record with balloons for long lesions, but these are expert investigators and best hands, best case scenarios.  So it's a very uncomfortable feeling at the moment.  Really if you get your wire down there, anything goes after that.

            DR. WHITE:  Well, I just was going to add actually the same comment, and that is that I think in some ways the FDA shares responsibility for not providing a little better guidance here for this trial.  I disagree with the interpretation of the sponsor's reading of the literature.  I've found several prospective trials, one of which they've cited themselves, a Soder paper.  I think it's number 18 in their reference list in Package 3, which were prospective PTA trials in chronic limb ischemia patients that did quite well.

            And I know that the standard practice in my laboratory, and I'm sure some of the sponsor's investigators' laboratories are that we treat every day patients with this type of disease without lasers.  The real question that we'd like to be answered today is what the laser adds to other interventional therapy, and I'm afraid we're not going to be able to come to that conclusion, and I think the FDA let the sponsor down by not insisting that that happened.

            Dr. Sapirstein says that ICA was used because of the interventions, but you know, Wolf, there was only four percent angioplasty done in that series.  Only a very small, small number of patients in the ICAI trial actually got an intervention.  It was not part of their treatment.

            DR. SAPIRSTEIN:  Well, we thought that the control was sufficiently robust.  Somewhere around 40 or 50 percent of the control arm had an active intervention, either by bypass or an angioplasty or thrombolytic agent.  So I think that a control to PTA per se might have been even -- and many of these patients were considered unsuitable for PTA, but as we pointed out in our review, we didn't feel that any -- we thought it was questionable whether any added value was provided by the laser to PTA.

            CHAIRMAN LASKEY:  Okay.  I think we beat that one.

            Well, congratulations to all for keeping us way ahead of schedule.  My suggestion is if the lunchroom is ready for us that we break for lunch.  It's a tad early, but I hesitate to start the open committee discussion at this point.  Okay?

            It is ten after 11.  Can we regroup at 12:15 and resume?  That way we pretty well will get everybody to the airport on time.  So let's adjourn for the moment and regroup at 12:15.

            Thank you very much.

            (Whereupon, at 11:11 a.m., the meeting was recessed for lunch, to reconvene at 12:15 p.m., the same day.)

 


         A-F-T-E-R-N-O-O-N  S-E-S-S-I-O-N

                                      (12:23 p.m.)

            CHAIRMAN LASKEY:  All right.  Thank you for returning on time.  We have a quorum here.

            So we will lead off with Dr. Nicholas who is the lead reviewer for this PMA, and then we'll just go around the table from there.

            Dr. Nicholas.

            DR. NICHOLAS:  Thank you.

            Some of what I say, I think, unfortunately will be a bit redundant, but I feel I've got to sort of put it all in one place.

            By way of summary, the investigators presented the study using excimer laser as a catheter based treatment for lower extremity peripheral vascular ischemia.  The device has been previously approved with smaller catheters, and the investigators have utilized several new, larger catheters previously unapproved for the peripheral vascular interventions.

            An initial feasibility study was conducted on 23 patients, 25 limbs.  The pivotal study identified laser angioplasty for critical limb ischemia, known by the acronym LACI, was then conducted, 15 sites, 12 in the United States, three in Germany.  There were 145 patients enrolled.  Fifty-eight of these patients were from centers outside of the United States, and 69 patients were enrolled from three sites, two of which were in Germany and one in the United States.

            Preclinical studies, as you heard, of biocompatibility have all been taken care of and tested.

            The results of the registry, patients who entered the LACI protocol were compared to a historic group published in the Annals of Internal Medicine in 1999.  This was a study from 56 departments within the Italian National Health Service.  In that study, patients were randomized into two groups, one receiving Prostaglandin E-1 and the control group received standard medical care at the same institutions.

            The current LACI study utilized this control group from that protocol to compare their results with the excimer laser.

            The primary endpoint of the LACI study is limb salvage without major amputation at six months.  Secondary endpoint of death, peripheral vascular endpoint which was defined as a major amputation or persistent limb ischemia, and other secondary endpoints were wound healing, surgical bypass, reduction of degree of planned extremity amputation.

            Serious adverse events were defined as those events that were fatal, life threatening, disabling or resulted in prolonged hospitalization.  The patients were monitored throughout a six-month follow-up period.

            That's my summary of the protocol as I read it and we heard today.  I'm going to express now some of my observations and concerns about the protocol and the strengths and weaknesses that I see.

            The protocol uses a historic control rather than a prospective randomization.  A stronger control population would have been a random sample of patients who received only balloon angioplasty and stenting without the use of the excimer laser.

            Given the control group as designed in this study, it is difficult to sort out the beneficial effect of the excimer laser.

            The investigators suggest that they have shown equivalency of their results to the control group because the patients in the registry of their protocol were sicker than those in the historic control group.  It is difficult to be assured that the LACI patients were, indeed, less medically fit.

            Since the exclusion characteristics of both the control and the LACI population included patients with a limited six-month prognosis and those with severe or unstable cardiac disorders, these were also in the control group.

            The primary safety endpoint of all cause mortality occurred in 11.2 percent of the registry and in 14.4 percent of the control group.  These rates are not significantly different and mitigate against a significant difference in the medical fitness of the selection process of the two groups.

            Thirdly, the multiple comparisons of the LACI registry to the surgical literature and medical literature fail to achieve Level 1 or Level 2 evidence.  The control group contained a significant number of higher men and current smokers.  We also heard about a number of other risk factors that differed between the control group and the LACI protocol group.

            The degree of ischemia is difficult to estimate.  The control study did not use the Rutherford classification for chronic lower limb ischemia, and with regard to the Rutherford Categories 5 and 6, the distribution is not available, and this would significantly impact the reality of the outcome of the LACI protocol.

            Rutherford Group 6, as defined in the original manuscript, states that these are patients who have major tissue loss above the metatarsal level such that a functional foot is no longer salvageable.  Including these patients in a protocol would seem of question since the stated protocol was limb salvage.  The benefit of entry into the protocol would only be to raise the level or -- excuse me -- to lower the level of amputation.

            It would appear based on the control group that the LACI results would have to stand on their own merit since the control population is not clearly managed.

            Another concern is that the sponsors indicated that they measured the ankle/brachial indices upon entry into the study as well as at intervals throughout the six-month follow-up.  This data has not been presented in the proposal.

            It would certainly be of interest to know if there was a significant improvement in the index of .15 or greater in the patients in the registered group from their control level to the subsequent measurements after intervention.

            The effectiveness of the procedure is of concern because at the end of six months, 39 percent of the LACI protocol patients remained in Rutherford Class 5 or 6, 39 percent, including Class 4.  Excuse me.  In the control population, 43 percent were still considered to have critical limb ischemia.

            The amputation rate in the LACI registry was 7.6 percent at the end of six months.  If we include the two patients who died after amputation, it was nine percent.  This is similar to the 13 percent from the control population in the literature of 13.3 percent.

            The occurrence of adverse events which has been talked about occurred at approximately the rate of five to six percent for each month of the follow-up.  The trend on the graph appears to be a continuum, and it would be of interest to know the longer term follow-up and necessity for further interventions after that six-month period.

            And just two more points.  One should not lose sight that the gold standard for care of people with severe limb threatening ischemia is distal bypass grafting with a venous conduit.  The investigators clearly did not evaluate their patients for alternate sites of venous conduit.

            One of the entry criteria into the study was the presence of at least one vessel runoff, which is also an entry criteria for having a distal bypass.  Several of the examples that we looked at showed what appeared -- and it's difficult to say from across the room -- appeared to be a graftable peritoneal artery, and possibly with more distal films, there might have been more ankle vessels that were available.

            Additionally, anesthesia ASA Class 4 is unfortunately a classification in which we frequently find ourselves as vascular surgeons, and at my institution, speaking with my anesthesia colleagues, about a third of our patients who come for tibial bypass are anesthesia ASA-4.

            A final comment.  With regard to the illustration of ulcer healing and measurement, it is difficult to estimate the impact of this form of intervention in ulcer care since these patients obviously received improved management.  We are unaware of the natural history of this particular lesion in the patients.

            The lesions shown were clearly neurotrophic ulcers which may have been mixed with vascular insufficiency, and we are well aware that that type of ulcer responds well to multiple modality care on a local basis.

            That concludes my introductory comment.

            CHAIRMAN LASKEY:  Gary, do you have any specific questions then for the sponsor?

            DR. NICHOLAS:  Well, I've asked two of them this morning that related to ASA class as an exclusion criteria, and then related to the use of alternate venous conduit, which is truly standard of care.

            Thirdly, one of the -- and this has been confusing to me -- one of the entry criteria is the presence of at least one named vessel for distal runoff.  This is certainly the same criteria we use in doing surgery.  We used to have a named vessel to bypass to, whether it be in the popliteal level, the calf, the ankle or even out into the foot.

            Was there any evaluation by your surgical colleagues in the investigation as to whether they thought these might be graftable vessels?  And if not, why not?

            The other questions related to vascular are renal insufficiency.  We did see illustrated at least one patient who was on renal replacement therapy.  How prominent was that in your study protocol since that clearly is the most difficult of populations that we're called upon to deal with?

            CHAIRMAN LASKEY:  So who wants to tackle the question?

            DR. RAMAIAH:  Dr. Venkatesh Ramaiah from the Arizona Heart Institute.

            I have no interest in Spectranetics and  no stock options, but I'm here as a consultant for this session.

            As a surgeon, you  know, you would say why are we doing all of this because as a surgeon, as Dr. Nicholas will say, I can graft anybody; I can bypass anybody.  And the whole basis of this study was to evaluate patients that were not surgical candidates, and that's the whole crux of the LACI study.

            The three criteria that were really important to enroll patients then were, one, very poor quality of vessel, yes.  One of the criteria was they should have a named vessel, but it does not elaborate on the named vessel, whether this is less than one millimeter, highly calcified, intermittent long lesions in that one named vessel.

            So the quality of that vessel is very important, and yes, as a surgeon enrolled in this study and as a major enroller into this study, these patients were evaluated at least in our center by a vascular surgeon.

            Vein at our center, there was no saphenous vein.  To be very frank, we didn't go all the way in duplex down veins.  So that leaves a little questionable doubt as to whether these patients should have gone one extra step, especially if they didn't have a target vessel.

            And the third one is, of course, ASA criteria, which was greater than four, and as Dr. Nicholas said, all of our vascular patients generally fall in that category.

            So to enter these patients in, they had to have at least one criteria, and just judging by the one example,  which is actually a very truly representative sample of the entire group was that patient  with renal failure who ultimately Dr. Laird showed had healed in six months after skin grafting and wound care.

            That patient really didn't have or even if she did have a named vessel, it was not an ideal, suitable target vessel for a long distal bypass.  Yes, there have been studies which clearly show that vein bypasses, pedal bypasses, results have been good.  I can quote hundreds of studies where they show, you know, patencies of 80 percent at one year, you know, Taylor, et al., DHR, Albany Group.  There are a lot of studies of what the pedal bypass is.  There's no question about it.

            But they are also associated with a certain degree of perioperative mortality, anywhere from three percent to about seven to eight percent perioperative mortality with these long bypasses, in addition to revisions and re-interventions and things like that.

            So from a surgeon's point of view, at first I shouldn't even be here because the patients that were selected for the study were nonsurgical candidates.  So basically the surgeon shouldn't really be commenting, other than the fact that these were nonsurgical candidates.

            And at our center a surgeon did interview these patients and they found that they were not the best candidates, and so LACI got together a group of patients which were really the worst candidates for any kind of re-intervention in terms of surgical options.

            DR. NICHOLAS:  Thank you.

            CHAIRMAN LASKEY:  You had another question about the renal insufficiency subgroup?  Did you want to pursue that?

            DR. NICHOLAS:  Only insofar as the example was a patient with renal replacement therapy.  Were there other patients in that situation in your group?  How many of the 145?

            DR. RAMAIAH:  The reason why renal insufficiency was not added in comparison to the control group was because in the control group there was no renal insufficiency as a marker.  We definitely had -- and for the panel, chronic renal failure is the most predictor of a failed bypass or bad outcomes, and in our group, I think they're going to look up the total number of renal insufficiency, but to the best of my knowledge, there was a highly significant number of patients with renal insufficiency.  But we will get the numbers.

            DR. NICHOLAS:  If I can expand on my question a little bit, one of the most frustrating parts of being a vascular surgeon, and I'm sure you'll agree, is a successful bypass in a patient with dialysis therapy, and you still lose the limb.  Did that occur in your protocol?

            DR. RAMAIAH:  Well, in our protocol, we had an amputation rate of nine patients in terms of the laser assisted angioplasty.

            DR. NICHOLAS:  Right.

            DR. RAMAIAH:  Of those patients, two of them actually underwent a bypass, and the identification of the target vessel was made possible because of the laser, where we feel or we felt that the channel of the distal artery that was thought to be a candidate for a bypass was opened up because of opening of proximal lesions by the laser.

            So the LACI study, one does not go beyond that extra step of creating a problem and taking the patients off from any other modality of treatment in terms of surgical bypass.

            On the other hand, it goes a step further by the possibility of evaluating or reimaging a vessel that we may not have seen rather than blind exploration of a target vessel.

            DR. NICHOLAS:  Thank you.

            CHAIRMAN LASKEY:  Dr. Tracy.

            DR. TRACY:  I'll try to keep this pretty brief.  I'm not going to touch on any of the discussion that we had earlier, the questions of the control group and the appropriateness of that.  I just have a couple of questions about the -- forgive the electrophysiologist for asking this question, but it seems to me when you stick a laser down somebody's vessel that it is extremely likely that you'll, as 90-plus percent -- will require something else to be done.  Yet it seems like stents, which are in other vessels better than sticking balloons in, were sort of viewed as a bad thing.

            And I just don't quite understand that, didn't quite pick up on why the desire was to avoid other interventions like placing stents.  If somebody could explain that to me, it might help me understand that.

            And the other little part of things, there was a higher re-intervention rate required in the laser treated patients, and if you could just explain what those re-interventions were for and what was done at the re-intervention.

            DR. BRUCE GRAY:  I thank you for the opportunity to be here.  My name is Bruce Gray.  I'm from Greenville, and I have no stock options or stock interest in Spectranetics, but I am a paid consultant to be here today.

            I do have a great passion for the use of laser in the periphery.  That's all I do.  That's the mainstay of my practice, and it's from that context that I'd like to address the issue of the avoidance of stents.

            Many of these patients have very long lesions, and median lesion length in this study was 16 centimeter, which is akin to a study that we published in 1997 that also had a mean lesion length of 16 and a half centimeters, in which we treated all of them with balloon angioplasty and wall stents.

            Our primary patency rate at the end of one year was 22 percent.  Our secondary patency rate was 43 percent in that patient cohort.  We could get a pristine look angiographically after doing the procedure.  The problem is we just couldn't keep it open.  The restenosis rate was very high, and those were honest data.

            Now you're faced with a scenario where you're walking into taking care of patients with long segment occlusive disease, and they really don't have a good surgical option.  You're saying, "I don't really want to just put a balloon in the stent down there.  I'd like to do it without a stent for the secondary animal hyperplastic response of the stent."

            And that's where my interest in the laser came to be.  There's something special about putting a laser down an occlusion in that it takes out the chronic thrombus that's there.  The lesion is typically an atherosclerotic lesion at the distal portion of the occlusion, and then thrombus propagates up the vessel.

            When you pass the laser down through, as evidenced in that superficial femoral artery example, you take the chronic thrombus out.  We have no other tool in which to take it out, but that will do it.

            It will also take some plaque out, but it's really changing the milieu in then which you're going to use a balloon and then a stent.  The problem that we're all having is that you're saying, "Well, what is the role of the laser?  What good does it really do?"

            Well, if you can change the milieu and safely treat it with a balloon with minimal embolization, then you've really accomplished something safely in that particular patient population.  I think that's what the data shows, is you can take long lesions, not a six centimeter lesion.  These are long lesions, and these are patients that have compromised out-flow.  They're only dealing with one runoff vessel.

            So if you knock off that one vessel with an embolus, your amputation risk ought to be substantial, and I think we'd all agree that in that context if we're sending a lot of stuff downstream that we're going to knock off a few limbs, but that's not what we're seeing.

            What we're seeing is we can preserve a lumen down through.  Now, your point of how many other problems are you going to run into, six, nine, 12 months down the line is a very good one, and it is an issue.  And re-intervention isn't that uncommon at the tune of 17 percent, but if you just used a balloon alone, your intervention rate is going to be in the 30 to 40 percent range.  If you add a stent to it, that increases your initial technical success rate, but you often lose it six, nine months down the line.

            So I don't think that that strategy is necessarily the best, but the complex strategy of being able to use multiple modalities to first simplify a lesion and then treat it with the other tools that you have to address the plaque burden underlying that occlusion is what's most helpful.

            So you can't just view it as just the laser alone.  It's a combination, and then when you selectively use the stents to give you your initial high technical success rate, that's the value of the laser, is you're just using them selectively then.

            If you take a 30 centimeter SFA occlusion and you put stents all the way along it, that's one thing.  But if you take a 30 centimeter occlusion and only have a stent a two centimeter segment distally, that's a totally different issue, and that's what you're able to see here.  You don't have to lay 30 centimeters of stent in.  You can use a much shorter stent, perhaps a four centimeter stent distally where the plaque burden is greatest, and that's where the great advantage is of the laser, and then the combination of the laser and the balloon in selectively stenting.

            Did that answer your question?

            DR. TRACY:  Yeah, I think it answers the question.  The higher re-intervention rate you're saying is you've changed the substrate in some way and you may or may not have in that higher re-intervention rate group have had a stent in place or had an angioplasty done in conjunction with the initial laser.

            And once again, we're running into the problem of comparison then with a control group where it was just angioplasty that was done on, I assume, a smaller length lesion.  So it's an equivalency problem that I'm having, I guess, to try to figure this out.

            You're saying that you're opening something that otherwise you wouldn't even be able to have placed a stent, and the higher re-intervention rate is as a result of --

            DR. BRUCE GRAY:  No, you could have -- excuse me for interrupting -- you could have --

            DR. TRACY:  No, please do.

            DR. BRUCE GRAY: -- been able to place a balloon in the stent.  You could have been able to place a balloon in the stent, but the potential complications of ballooning along lesion are much more substantial than what you would see in these data.

            Also, the length of stents used and the number of stents used would be substantially different, which would then have a consequence down the line.

            DR. TRACY:  So one stent instead of 40 stents in a vessel or something like that.

            DR. BRUCE GRAY:  A huge difference.

            DR. TRACY:  Okay.  I think you've answered my question.  It, too, I think, is a problem of not having a good thing to compare to, and I think the panel is just going to have to struggle with what the data are without a comparison and try to make a decision based on that.

            DR. MAISEL:  Good afternoon.  I just have a couple of related questions, I guess.  First, I was wondering if the sponsor could maybe clarify a little bit how the ICAI study was selected as the control study.  What process played out that that paper out of a vast literature was selected?

            Was there some literature search that was performed?  What other studies were looked at?  Why were those other studies excluded?  Was there any thought to pooling other studies?

            Maybe you could just play out what actually happened to pick that study.

            DR. REISER:  Let me try to address that.

            During the conduct of LACI Phase 1, we struggled with the question of what an appropriate control group might be for Phase 2.  We did an extensive literature search, both in the PRA literature and in other literature, that is, literature describing other modalities to try to define what the standard of care would be for this particular patient subset.

            It was difficult to find any paper in PTA that claimed a reasonable large -- in a population, in a patient population that was close to what we expected to enroll in LACI.  That was one of the rightful criticisms of some of the literature that Dr. Laird discussed this morning.

            It covered a very heterogeneous group.  We also looked at the TASC document because it gave us what the scientific definitions were.  The TASC document suggested that they needed more data to determine what they would recommend for this particular patient subgroup as a standard of care, as a standard therapy.

            So after reviewing PTA literature, we despaired of finding a suitable control group there or even statistics on which we might base expectations should we design a study using PTA as control.

            At about the time that we were designing the LACI Phase 2 protocol, that is, between LACI Phase 1 and LACI Phase 2, the ICAI study group paper appeared.  It occurred to us that the control group in that paper defined what the standard of care was for all CLI patients, and that standard used a variety of treatment modalities, as Dr. Sapirstein, I believe, told us this morning.

            In a sense, it set a benchmark which we thought would be the best one in the sense that it used all treatments to try to treat a very large patient population.  In that sense, we found that this particular publication was fairly authoritative.  We reasoned that if what they published represented the standard of care for all CLI patients, then that was the standard against which we should be compared, against which we should compete, so to speak.

            That was the rationale.  We did note when we wrote the protocol that we expected our patients to be more morbid because the exclusion was all good surgical candidates would be excluded.  And Dr. Laird hoped to make that point during one of the slides this morning.

            So taking all of that together was our rationale.

            DR. MAISEL:  Okay.  That helps clarify that.

            Just in follow-up a little bit, it's mentioned that 288 patients were screened, and 128 patients were screened failures.  And I thought I recalled seeing what the reasons for failures were, but I couldn't find it.

            What specifically were the reasons for screen failures, and do we have any follow-up on those patients?

            DR. REISER:  We didn't analyze the reasons for screen failures.  Screen failures were not entered into our database, and we didn't follow them.

            DR. MAISEL:  That potentially would have been a very valuable resource because presumably these patients are more equivalent to the actual study patients who might have been screen failures because they were a little more sick or a little less sick, and it would have been, I think, very valuable to have some follow-up data on what happened to those patients and what their outcomes were.

            But we don't have it.  So that's a moot point.

            DR. REISER:  Sorry.

            DR. MAISEL:  And then just one final question regarding the re-interventions.  Certainly I guess I was not surprised by the 17 percent re-intervention rate.  Perhaps you could just clarify a little about what the timing of the re-interventions was, meaning were there some acute interventions that happened within hours or during the same hospitalization, such as vessel occlusion, thrombosis, et cetera, or were these, you know, re-intervention that were happening a month or two or three later?

            DR. REISER:  It's my interpretation of the data that very few of them were in hospital re-interventions, the vast minority (phonetic) of them, as in approximately four of the total.

            I think a word from one of our investigators would be useful at this point.

            DR. LAIRD:  Actually there were only two acute re-interventions during the hospitalization pursuant to complications or closure of the vessel.  The remainder of the re-interventions occurred in the six-month time period mostly based on clinical grounds.  An ulcer that was initially healing and then stopped healing and started progressing again led to reevaluation and re-intervention.

            And as mentioned, the re-interventions were for the most part repeat angioplasties, you know, with or without laser.  Only a couple of patients went on to have a surgical procedure, either bypass or endarterectomy.

            DR. MAISEL:  Okay.  Thank you.

            CHAIRMAN LASKEY:  In my mind this is a terribly important point.  Do you have a plot, the cumulative frequency distribution of re-intervention?

            I think that six months is kind of short on a follow-up, and if things are accelerating from four to six months, you may be on the steep limb, and you may cutting some events off if you stop at six months.  Do you have any idea why this is picking up speed as time goes on or it levels off or what your rate here --

            DR. LAIRD:  Well, what we saw -- the short answer to the question is, no, I don't have that data, but what we saw over the time course of the six months is that there was continued decrease in size of the aneurism and sort of a stabilization of the process.

            So there weren't -- and there wasn't an apparent increase in late interventions or late problems.  And, in general, you know, when you're dealing with patients with critical limb ischemia, it takes less blood to keep the tissues healed once they are heeled.  So if you can get the vessels open well enough to heal the ulcers even if the vessels do re-narrow during that time period, there's a very good chance that the patients will stay healed.

            CHAIRMAN LASKEY:  Thank you.

            Dr. White.

            DR. WHITE:  Thank you.

            I'd like to as an operator in this field commend the investigators in the execution of this trial.  We may not be very happy with the way the trial was designed, but clearly, they were able to get excellent results in an extremely difficult population, and the outcomes clearly are excellent in their hands for this device.

            But I think the -- well, I just have a couple of questions maybe.  Could you tell me?  I didn't see in the panel package.  Did you report the average fluences that were used for ablation?  Can you tell me what the energy was per lesion?

            DR. REISER:  Let me find the table.

            DR. RAMAIAH:  (Speaking from an unmicked location.)

            DR. WHITE:  And what is the fluence to ablate calcified tissue?

            MS. WOOD:  Excuse me.  Someone answered behind you, and he wasn't speaking into the microphone.  Could he please step up and answer into the mic for the transcriptionist's benefit?

            Thank you.

            DR. RAMAIAH:  While Chris looks for the exact numbers, off the top of my head the average fluences are used anywhere from 25 to 40, and the pulses were mainly from 25 to 30.

            In terms of fluence you've got to use for what kind of lesions, soft, calcified, dense, thrombus, I think Chris will look up into the exact technical details of that.

            MS. WOOD:  Please re-identify yourself for the transcriptionist also when you come up to the --

            DR. RAMAIAH:  Dr. Ramaiah, Arizona Heart Institute.

            DR. WHITE:  The reason to tease that out, if you have that, is that there's always a little bit of debate about the mechanism of the laser opening, the issue and whether or not we actually are ablating tissue or whether we're mechanically getting through these occlusions, and so I think understanding the fluence and the required fluences for these diffused calcified lesions might help.

            Go ahead.

            DR. REISER:  According to my data, the mean maximum fluence was 51.  That's 51 millijoules per square millimeter.  The mean repetition rate was 32 pulses per second.  Mean laser pulse is 5,371.  That's per limb.

            DR. WHITE:  So when you say mean maximum, just so everybody understand, there may have been multiple fluences used on a single patient, but what the operator actually used at a maximum across all of the patients was 51 millijoules per millimeter squared?

            DR. REISER:  You got it.  Thanks.

            DR. WHITE:  Now, can you tell me what I need to ablate a calcified tissue specimen?

            DR. REISER:  Our technology does not do well in rocky calcium, as I'm sure you've had the experience.  That's why our coronary indications, for instance, call for moderately calcified lesions.  So --

            DR. WHITE:  But I think we've heard today that part of the reasons why these lesions were not so good for surgery was that they were diffusely calcified and heavily calcified.  So is it your impression that you were, as Dr. Gray was talking about, the debulking of these lesions, or did you really get the feeling you were taking away tissue with this device?

            DR. REISER:  I think I'll let Dr. Gray answer that.

            DR. BRUCE GRAY:  Thanks, Dr. White.

            Bruce Gray for the transcriptionist.

            Yeah, I think the energy that you need to bust through the top cap of the occlusion to begin with, the 45 fluence rate is plenty adequate.  If you have a heavily calcified lesion, atherosclerotic plaque in the intima, then typically the laser will not do much at that fluence.  You can turn it up to 60, but it still isn't going to do a whole lot with it. 

            That's contrasted to the calcium on the adventitial layer of the artery that we typically see under fluoroscopy that makes it hard to sew into that type of artery, and we're talking about two different things there.

            So the incidence of intimal calcification of the plaque is less prominent than the adventitial calcification you see in the artery.  So on the case report form when you put down was there calcification in the artery, will you see it under fluoroscopy, yes, but the type of composition of the plaque you can only discern after you've already passed it through and say, yeah, I've really done something here.

            So it's only anecdotal at best as how you manage that.  Now, technically getting through it, then you just turn the fluence up and you'll feel it.

            DR. WHITE:  Okay.  In my laboratory and, I'm sure, in many, the treatment of choice for chronic limb ischemia is actually a percutaneous procedure, not surgery, and I'm sure that varies across the country in different hospitals, but clearly we get a chance at most of these patients, and then if we fail, the surgeons will take a crack at them.

            But typically our treatment does not require the use of a laser, although we do have the laser available.  So I am a little bit concerned that we're relying a little bit too much on the TASC document, which I think is a wonderful tool and full of a fund of knowledge that we can all use, but is, I think we all agree, a bit outdated in that it doesn't address a lot of the stenting issues that we currently practice.

            So I'm just concerned that angioplasty was not chosen as the design for this trial.  I believe that the laser is an adjunctive treatment.  In fact, in your study you used some other treatment, balloon angioplasty, in virtually all of the patients, and so what I'm really wanting to know is what does the laser add, obviously, and the data doesn't allow me to determine that.

            I also would like to say I think that ICAI -- that ICA population was -- I don't believe that is the standard of care or practice for patients with critical limb ischemia, and again, only four percent of those patients got an interventional treatment, although a third of them did get surgery.

            I'd like to know, I think, in terms of the risk what the cost benefits are here.  Can you tell me on average what the laser catheter is going for these days, what it would cost me to use one of the approved devices in the leg?

            DR. REISER:  Chris Reiser.  The street price of a laser catheter is roughly $1,500 in the United States.

            DR. WHITE:  Okay.  So one of the issues that we'd want to know is are we getting a benefit that justifies the expense.  Stents also cost that much money.  Drug eluting stents are going to cost more, and I think it's important for us to have some reasonable idea of the bang for our buck as opposed to simply what equivalent therapies that work.

            DR. ZUCKERMAN:  Dr. White.

            DR. WHITE:  Yes.

            DR. ZUCKERMAN:  It's important to understand that in formulating a risk-benefit profile for this panel, we don't take into account cost.

            DR. WHITE:  That's because you don't treat anybody.

            DR. ZUCKERMAN:  Correct, but the advice that we need is one of a clinical nature, and it's really important to look at the problem that way.  I mean the other stuff is very interesting, but the advice that the agency is looking for unfortunately does not include that area.

            DR. WHITE:  I appreciate your sensitivity to the cost issues, but I think that when you're doing an equivalency trial and you have equivalent therapies, it's interesting to be able to weigh some of those issues.

            Did you want to say something, John?

            DR. LAIRD:  If that's okay, Chris.

            John Laird from Washington.

            I just wanted to sort of readdress the issue about, you know, how well we could potentially treat these patients with balloon angioplasty.  I mean, obviously, we've all been doing this, using balloons for these type patients before we had a chance to use the laser, and I think it's worthy of mention that these are not sort of the everyday patients and not sort of the everyday critical limb ischemia patients.

            We're talking about patients with a mean lesion length of 16 sonometers, you know, roughly 2.7 lesions per patient, each of these lesions around six sonometers in length, and certainly we as interventionalists and skilled interventionalists, and we have some very talented people on the panel, can get wires through these lesions and dilate them.

            But the question is:  what are the result when you do that?

            And there is a body of literature that's pretty strong on balloon angioplasty for patients who have this diffuse type disease, long occlusions, diffuse disease, where the results are not good.  I mean, you can pull a number of studies from the literature with patency rates and primary limb salvage rates of less than 50 percent at a year.  So I could cite several studies like that.

            And then to talk about the TASC document being outdated, perhaps it is, but the TASC document was released at the time that this study was formulated, and 60 percent of the patients in this trial had TASC D lesions which wa, according to the authors of the TASC document not suitable to balloon angioplasty, basically surgical lesions.

            So we, I think -- I think you have to look very carefully at the patients we are trying to treat here.  They were very sick, very badly diseased patients, and I think really the results that were achieved were with a strategy of using laser in addition to adjunctive balloon dilatation and selective stenting were excellent.

            It's clearly difficult to sort out the true benefit of the laser itself.  We debulked, you know, these lesions significantly.  Half of the lumenal gain was from the laser.  We treated very long lesions with extremely low rate of distal embolization, and a very, very low rate of recurrent or acute limb ischemia in the hospital, and an acceptable rate of stenting, I think, and low stenting below knee.

            So I think excellent results were achieved with this strategy of laser first, combined with additional modalities.  So perhaps at the end of the day we cannot absolutely call out what the exact benefit of the laser was, but using the strategy in a very difficult patient population we got excellent results.

            DR. WHITE:  Well, I agree you have excellent results, but you say yourself, John, that you compare yourself to PTA, and in fact, in the panel packet yourself, whoever wrote Section 3 did cite the Soder paper that is a prospective trial, and I found Jameson in 2002, JVIR, and of course, you talked this morning about the Dorros paper.

            I mean there are some excellent results with PTA in a population that may or may not compare.  I mean obviously there are apples and oranges.  They weren't randomized.

            So all I'm saying is that I think it's not the fact that you can find crappy reports of angioplasty.  I agree they're out there.  The question is in your hands what would have happened.

            And I think that you know, a lot of this business comes down to a strong believe and bias that we all have as clinicians, and I think that we should take on the objective benefit of showing that there is a ten percent.

            I mean, for example, what about the 15 patients or so that couldn't be crossed with a guidewire.  I mean, you know, that's a group of patients that couldn't be treated perhaps another way.  You know, that's clearly an adjunctive benefit of having the laser catheter available to you.

            So, I mean, those are the kind of measurements that I think would be meaningful to us in that kind of population.

            DR. LAIRD:  I think just since you brought up the Soder paper, I mean, it's very difficult when you review the literature on balloon angioplasty for critical limb ischemia, but that paper, for instance, was a single center, small study, quote, unquote, prospective study of 60 patients, but they sort of retrospectively dropped out about five patients from the study because they didn't quite satisfy their angiographic criteria.

            The mean lesion length in that study was 3.8 sonometers compared to 16 sonometers on our study.  Their technical success rate for total occlusions was 60 percent, and their patency rate, angiographic patency rate at a year for infrapopliteal occlusions was 40 percent.

            So we could discuss the literature, and that study is an example and find, I think, that we are doing really well with the strategy.

            DR. WHITE:  You know, it's interesting you raised the patency issue because patency and limb salvage are two separate things, as you well know.  The idea is to heal the ulcer whether the artery stays patent.

            And you guys are also paying a penalty for re-intervention compared to your control group because your control group didn't get any intervention.  So how can you have restenosis when you don't do angioplasty?  And you're doing restenosis.  I mean, your intervention rate is 17 percent.  Obviously it's low for an angioplasty trial.

            But you might have done better had you looked at balloon angioplasty with a much higher intervention rate in a group that represented intervention at a higher rate because you can't have re-intervention if you don't ever do it.

            So you kind of put the study in a hole by --

            DR. LAIRD:  Yeah, in retrospect, I guess we knew it prospectively that it was an issue, but there was a real concern that we could not complete a trial if we randomized to balloon angioplasty.  One, there just was no enthusiasm for a strategy that we at the time thought was ineffective.  So we didn't want to randomize against a treatment that we did not think worked for these patients.

            DR. WHITE:  That's all I have.

            CHAIRMAN LASKEY:  Tom.

            DR. FERGUSON:  I agree with Dr. White.  I think the presentation has been superb.

            I guess my one question relates to what I understand that the laser treatment does for the lumen of the vessel.  You enlarged on that just now when you said that you achieve a 50 percent larger lumen.

            It's a small catheter, and the direction of the beam is straight ahead, and you don't do any kind of Roto-root situation.  So I guess my question is:  does not at least in perhaps the larger peripheral vessels, which would include all the way half down, you know, the lower leg, require balloon angioplasty as an adjunct?

            And the reason I bring that up is that I could find only this one statement.  We've been talking about using the balloons all day long with your treatment modality, and on page 5 of the projected material that goes out with the catheter, number 12 says, "Following laser atherectomy, perform a follow-up angiography and balloon angioplasty if needed."

            My question is:  how often is it needed?  And it sounded to me like it was needed a great deal of the time.  And the converse of that, how many times can you use the laser treatment and not use the catheter, angioplasty?

            DR. LAIRD:  Yeah, those are very good questions and observations.  I think the real limitation is the size of the catheter.  In general when we're treatment the SFA or popliteal artery, we're treating vessels that are five to six millimeters in diameter, and we're limited by the catheter size, our largest catheter being 2.5 millimeters in diameter.

            We have seen -- I think it's an anecdotal observation -- that all of us who do a lot of these procedures have experienced that.  We get larger lumens than that though, particularly when we're treating occlusions in the superficial femoral artery where you have an atherosclerotic lesion which occludes the vessel, and then the vessel fills up with thrombus more proximal to that occlusion, up to the nearest large collateral or branch vessel.

            And it has been shown in a number of in vivo and in vitro experiments that the laser does ablate a vaporized thrombus.  So I think we get more of an effect in that thrombus than we do in plaque per se in terms of increasing the lumen diameter, but we're still limited by the diameter of the catheters.

            Below knee I think there is a greater chance to get a stand alone result, but I think the majority of us still have had a tendency to use a balloon after the laser irregardless perhaps just to treat ourselves to try and maybe smooth out the lumen a little bit or get perhaps just a little bit more lumenal improvement with the balloon.

            As a practicing clinician right now in the absence of an approved peripheral laser catheter, we use the coronary catheters in an off-label manner, and we're hampered by the diameter of the catheter.  So it would be nice to have access to the 2.2 and 2.5 millimeter catheters to allow us to treat these vessels more effectively.

            DR. FERGUSON:  Yeah, that's a nice explanation.  I appreciate it.

            And, again, my question goes for information, but it also goes for labeling, and the question still remains about that, I think.

            CHAIRMAN LASKEY:  Dr. Morrison.

            DR. MORRISON:  Well, I first of all would like to applaud both the sponsor and the investigators for taking on a group of patients that are desperately in need of an innovative approach, and having proposed what we call salvage angioplasty as an alternative for patients who were refused coronary bypass graft surgery some 15 years ago, I can relate to the difficulties in trying to get your arms around these people definitionally and in designing a randomized trial.

            I guess like  most everyone else who has spoken now, I still have difficulty with comparing your results to people who for the most part got medical therapy as opposed to an alternative, and I don't see an easy way out of that.

            I think the suggestion that there's some 15 percent that you need at the laser to get across in the first place is an important one, and the concept of how many or what proportion you really didn't need additional balloon is also important.  If I read your data, that would only be about two percent, however.

            And other than that, at this point I really have no further questions.

            CHAIRMAN LASKEY:  Dr. Somberg.

            DR. SOMBERG:  Well, I think I, too, concur with the previous panelists here on the importance of this area and the efforts that the investigators put to this, and I was also, I must say, very impressed with the anecdotal pictures.  You know, while they do not dictate a study, they certainly -- outcome -- they certainly for those who take care of patients speak for themselves.  There was a lot of benefit here.

            I'd like to clarify for my own benefit.  Is it correct to say, as I think Dr. Morrison was just saying, that in only 15 percent of the patients you were not able to cross the lesion in its entirety with a guidewire, or is that incorrect?

            DR. LAIRD:  The technique of step-by-step laser recannulization was used in 13 percent of cases where the guidewire initially did not cross, and then we used the laser to penetrate the occlusion to try and get through that fibrous cap, and then once that's penetrated, often you can advance the wire down with the assistance of the laser, or you may have to advance the laser for a short distance and try the wire again.

            So around 13 percent of the time that strategy was used to facilitate crossing the occlusions.

            DR. SOMBERG:  So were the other 77 percent of or 87 percent of cases -- sorry -- you were able to cross both the initial fibrous cap and the rest of the lesion without needing the assistance of the laser angioplasty device?

            DR. LAIRD:  That's correct.

            DR. SOMBERG:  Or laser whatever.

            DR. LAIRD:  That's correct, and I think with the modern day availability of hydrophilic guidewires and a very experienced cohort of investigators have done, you know, thousands of interventions, and a very high guidewire crossing success rate was achieved.

            DR. SOMBERG:  But is it correct to say then that in 13 percent of patients without this modality you would have not been able to proceed and the interventional procedure would have terminated at that point?

            DR. LAIRD:  It's my understanding from the data that that's correct, yes, that the laser allowed crossing in 13 percent of cases when the wire itself wouldn't go.

            DR. SOMBERG:  But I think that's an important point.  So there are at least three investigators here who have each done patients, and I don't know if that finding distributes to these three investigators, but I would like to know if there is a subset of patients where the intervention would have stopped and nothing further could have been done for these patients without this modality of therapy being available.

            And specifically, I'm assuming you'd need the small, 2.5 millimeter, leads.

            DR. RAMAIAH:  Dr. Ramaiah from Arizona Heart.

            The interesting part about my experience in this whole study is we were the surgeons involved in this study and we were also the interventionalists who did the study when it came to LACI and angioplasty stenting, and, yes, you may ask that you used the laser, you know, most importantly in only 13 percent of those cases, and in the other 87 percent or so you got your wire through, and so you can just go ahead and do a balloon angioplasty.

            But I think what the data hides or which we didn't highlight is to start crossing an occlusion, and this is an important feature that I think hasn't really been highlighted, is when you start crossing a lesion right at the onset with the wire, there's a high indication, a high chance that you may create a subintimal intersection, and it has been my experience and the experience of the investigators also that with the use of the laser even to initiate the initial traversing of an occlusion, the chances of staying intimal rather than subintimal are much more.

            And that's a group of patients of 13 percent who were not able to cross with the wire.  The laser definitely did help us gain access, at least create an initial channel to which we could then either with the laser or the wire get through the whole thing.

            DR. SOMBERG:  But you're sort of bringing up another issue.  You're saying that the laser may facilitate the crossing without having the dissection or tear, if you will.  That may be the case, but there is a finite number of patients who without the leads that are not approved or without the catheter size that are not approved, that the interventional procedure would have ended, and there would have been no therapeutic benefit whether it be with balloon angioplasty or balloon angioplasty plus stent, and I'm just clarifying that statement.

            DR. RAMAIAH:  Right, and I don't think we have the exact number.  We were not able to traverse the lesion, but we know that nine patients did eventually go on to have an amputation.

            Yes, if you do not have the larger size catheters or even a laser catheter to traverse the lesion, which we cannot get with a wire, yes, the intervention would have stopped, and you would look at other options of an amputation or even a blind exploration for revascularization.

            DR. SOMBERG:  Well, maybe someone wants to comment.  Dr. Morrison, just help me clarify.  On page 35 of the materials distributed, there's a statement of eight percent.  I'm not pushing you on the number exactly, but I just want to feel comfortable that there is maybe, if you will, a very special indication that has been demonstrated by this study, and then there may be other indications we might want to parse later on or feel they have not been approved.

            Later.  I mean I don't want to take up the time.

            DR. BRUCE GRAY:  Yeah, I just wanted to address the issue of the verbiage used in the over-the-wire technique versus the step-by-step technique.

            The over-the-wire technique, where you place a wire through the entire lesion and then place your laser catheter down, many of the patients that I do, you place a laser catheter right at the top of the lesion, start the laser light to first initiate a channel and then pass your wire.

            So there was no wire traversal of the lesion to begin with.  Just the laser light was used to initiate the channel, and then the wire was placed.  That would be labeled as an over-the-wire use, whereas a step-by-step in my vernacular would be where you start with your laser catheter, try the wire; the wire still doesn't want to go; push the laser catheter further.  The wire still doesn't want to go and you end up going through the bulk of the lesion without wire lead.  That would be the step-by-step technique, and that you see in the minority of cases.

            But in most cases you're leading with the laser catheter to begin with.

            DR. WHITE:  Bruce, was that the way the protocol was done, without attempt at passing the wire?

            DR. BRUCE GRAY:  Well, no.  You can put the wire to the top of the lesion, but if you really push the J portion of the guidewire, you're going to go subintimal, and so the --

            DR. WHITE:  I guess what I'm asking is in this trial what you just described I don't think is the protocol.  Is that the protocol, John?  Was that right?

            I'm just confused now because I understand what you said.

            DR. BRUCE GRAY:  As an operator you have your choice. 

            PARTICIPANT:  It was an option.

            DR. BRUCE GRAY:  It really wasn't -- it's a little bit operator dependent on what you feel most comfortable with or if you were just pushing a laser catheter without a wire, you have a certain level of comfort, but if you have a wire right where you want, then you're going to push a little bit more easily.

            DR. SOMBERG:  I wanted to add a comment.  Unfortunately you've confused me a bit further, and the reason I was looking from my own point of view to see if there's a benefit is I understand there is this system out there.  Certain leads are available because they're used in the coronary.  Certain catheters, there's an advantage for peripheral vascular using smaller catheters.

            The FDA is asking this panel for guidance.  It seems to me that safety has been established.  The device works.  It doesn't blow up.  It doesn't kill anyone, the operator or the patient, in terms of an immediate problem.  It is functioning.  It has rationale behind it.

            With that -- and I might add that I do not worry about the re-intervention rate because you're comparing a controlled study where whatever was done, it's a random rate of re-intervention.  It's not all people had an intervention.  So you can compare the two.

            And that's the problem I lead into the host study, is you really can't compare the two groups.  We don't have enough information.  The concept of noninferiority and, therefore, you proved efficacy because this group is sicker is a very tenuous one.  That type of statistical basis is not there, and I feel you haven't proved that.

            So I'm looking to say is there a special subset of patients that this provides an overwhelming benefit for and should be available to without a level of evidence that would give it a general indication, and I think that's why I asked those questions, and it would be very important for me.

            I do think there is a need to find out where laser angioplasty or the laser device fits in and whether one should lead with the wire first and then do angioplasties or only do stents where the lesion is most bold (phonetic), or whether you need to use a laser to get down to that level, and that would help people.  That's what you need for the indications.

            And I don't think this study, for a variety of reasons -- and I don't think anyone is guilty.  Please don't take it that I'm criticizing the sponsor or the investigators -- but I don't think this type of study lends itself to a determination of efficacy compared to an historic control where there was nothing done to the control and then, finally, where significance was not shown.

            Because even if you accepted all of your premises, it didn't reach the .05 level, and that's a very minor level of surety when we don't even know half of the confounding variables in the control group.

            So I really would hope that you can come back to me and say that there is demonstrated by just one arm of the study, the arm you had control over, that there was clearly a standard approach that the manipulators of the wire used, and that there was a finite number of patients that nothing could be done for.  They were ill, et cetera.  They needed some sort of interventional procedure, and I'll accept that they probably were very poor surgical candidates, and that these small catheters added something to what we don't have at the moment.

            Thank you.

            CHAIRMAN LASKEY:  Dr. Krucoff.

            DR. KRUCOFF:  Okay.  I'm going to also try and not go over the same ground, but I have to start by recognizing that (a) this is a patient population who really suffer, and trying to advance that therapy, I think, is probably pretty solidly placed in all of our hearts.  It's very clear the investigators' passion in this comes from largely just that fact, and being very aware of this technology and its applications to vascular over the years, I'm equally confident that the sponsor has that same intention.

            So I think the struggle that we're all wrestling with is can we, based on these data, come to important conclusions, and I guess my starting point is when we use the word "equivalence," as far as I know -- and I'm going to ask the sponsor the same question I asked the status group from the FDA -- to my knowledge, other than equivalence as a philosophical term, are you all aware or, Chris, do you all have a statistician who can help us understand how an equivalent statistic can be generated and what really is a doing something in a high risk group compared to doing nothing in a lower risk group?

            Do you all have a statistician who could help us understand how equivalence would be measured in this trial design beyond just sort of as a philosophical "we must be as good as"?

            DR. REISER:  This is Chris Reiser.

            No, I didn't bring a statistician with me.  Perhaps you could help me through this, and I think the statistician --

            DR. KRUCOFF:  Well, maybe I'll let the real expert go back to this.  At least to my understanding, you know, feeling the gestalt as a clinician that if we can achieve the outcomes that we see in lower risk patients in higher risk patients by doing something, I get that.  That's a clinical definition of equivalence.

            But ultimately in a clinical trial, equivalence is a very formal statistical concept, and the basis for equivalence calculations could not, at least to my -- I'm not a heavyweight -- but to my knowledge, that's a problem, and I just thought maybe you guys had a statistical dimension that we could add here, or maybe not.

            And I guess the FDA answer was no, and that may be where we reach.

            I do have a question while I've got you at the podium though that I'm very interested in, the dissolution of thrombus.  In my memory of the intercoronary applications where thrombotic coronary lesions were addressed with excimer, it was not a favorable thing to do, but that's a little different.

            What I'm understanding you guys are saying is that in vitro you dissolve thrombus, and possibly in vivo you think some of your observation of a -- for instance, the illustration you may see a lumen that's larger than the channel that would just be identified with the catheter, and your interpretation of that is that you're vaporizing thrombus.  Is that what's being said?

            DR. REISER:  That's a good thumbnail sketch, yes.  Based on bench evidence and also clinical evidence, we believe that our technology vaporizes or at least liquifies thrombus inside the artery.

            It's true that patients who had angiographic evidence of thrombus were excluded from our initial coronary IDE in 1989 because at that time it was thought that these patients had active lesions.

            Since that time though, we have mustered a respectable body of evidence that shows that our technology actually works fairly well in such thrombus laden lesions. 

            Recently, I don't know if this is relevant, but FDA has allowed us to restructure our instructions for use in the coronaries to move the patient who has acute thrombosis into individualization of treatment.  They did this based on a body of bench and clinical evidence that we submitted.

            Perhaps one of the investigators could comment on the usual nature -- well, they have, in fact -- on the usual nature of the high thrombus content that's usually found in legs, and they've opined that this is one of the reasons that our technology seems to work so well in occlusions, especially, in the legs.

            DR. KRUCOFF:  Great.  Thank you.

            I was intrigued by the selection in the original protocol of death as the primary safety measure since I think a lot of us think of the safety of peripheral procedures as potentially being driven by other more minor outcomes.  The one thing that I just wondered, even though the mortality rates overall in these patient populations, they're just sick people.

            Were there any deaths that were actually during the indexed hospitalization?  It looked from the one set of -- the one Kaplan-Meier that was supplied, it looked to me like the registry population had zero deaths out to about 20 days or something.  Is that all after discharge?

            DR. LAIRD:  There were no deaths during the hospitalization or within the first 30 days. There was on in-hospital major amputation in a patient who presented with Category 6 Rutherford ischemia and had major tissue loss, and as a last ditch, the LACI procedure was attempted, but no procedure related or in-hospital deaths.

            DR. KRUCOFF:  Okay, and I think to me what this set of data may be most useful for is really to understand the safety of working with this technology in the very, very sick, and then how or when we can define efficacy.  At least if nothing else, you could build on this work rather than start over, that kind of thing.

            We, I think, can discuss that a little bit later.

            My last questions for you, for the sponsors.  You got back this conditional approval letter from FDA which they sort of made a point of clarifying for us that some of the suggestions in that letter, like a risk-benefit analysis and emphasis on understanding the potential confounding role of stents were two elements that were in that letter.

            Can you fill me in at least on how you thought about those two requests?

            As I've gone through the panel pack, it actually looks to me, in fact, from the table, John, I think you showed earlier like there's a suggestion that actually stenting did have or would have a role in outcomes if you were to look at a slightly larger population.  Now, .09 is not a proof, but which does suggest, as several people have mentioned, that both PTA and stents may have therapeutic benefit, and that takes us back to the whole question of what's the role of the laser, et cetera.

            But given this letter, is there a risk-benefit analysis?  Was there any kind of quality of life data involved?

            I realize you guys were trying to concentrate on harder endpoints like amputation, but I just wonder in the whole picture and the ankle indices which were not presented and at least as far as I can tell are not in the panel pack, but are there any supportive data or did you consider in the FDA's conditional approval level these two requests for how you were going to present or deal with the confounding influence of stents and/or an overall risk-benefit analysis?

            DR. REISER:  Chris Reiser again.

            With respect to stents, it was our intention to subanalyze as we showed you by stented versus non-stented patients.  However, the study was not designed to power subanalyses such as this.

            So the statistical power in that subanalysis is rather small, as Dr. Laird pointed out this morning.

            The other question, would you remind me what that is?

            DR. KRUCOFF:  The other suggestion in that letter was an overall risk-benefit analysis.

            DR. REISER:  My staff is trained in ISO standards, and we did provide an ISO standard risk analysis.  That was part of our submission.  I think it's somewhat different than the kinds of risk-benefit analysis that have traditionally been put forth, which are mostly a discussion of the risks and the benefits.

            That is, a positive risk-benefit profile should be lower risks with the same benefit or higher benefit with the same risks.  That's more of a qualitative sort of discussion.

            When we designed the endpoints, certainly the primary efficacy endpoint of limb salvage appeared to be the logical inverse of the risk, the risk being major amputation.  To find an endpoint which was different, but still talk to safety, we had to pick a different endpoint, that being death.

            Both of those endpoints were suggested in TASC.  So if you look at risk-benefit, the risk is -- what is the risk?  The risk is either death or major amputation.

            What is the benefit?  The benefit is either lack of major amputation or lack of death.

            So we were a bit cornered by this particular patient population.  Typically the risk is in one particular variable, and the benefit is in another particular variable, but in our case in this patient population, they were both flip sides of the coin.

            When Dr. Laird showed the benefits of LACI slides this morning, at the end of the LACI results, we tried to make clear that the benefit was that this was an intravascular procedure, an intravascular strategy compared with a strategy that might contain only medication or a strategy that contained a lot of surgery, 35 percent surgery.

            In the publication that we found, it defined the closest thing to a standard.  We went through the other benefits, as well. 

            So I was hoping that those slides that he showed this morning would clearly define what we thought the benefits of LACI were.

            DR. KRUCOFF:  Okay.  Just to share in the last comment and I'll quit.

            You know, some of the ways you've characterized this patient population, the no option patients and later the all comers, obviously there's an odd mixture.  Some of this is a morphologic feature of the stenosis and the residual vasculature, whatever you can see or not see.  Some of this is characterized by co-morbidities, the obese patient or the, you know, multiple co-morbidities who are high risk for surgery.

            And I would at least challenge the group on one issue, which is the ethics of randomization.  I think in a population like this where we're adding a component of our procedure, that to randomize them against standard care where standard care is whatever else you would do and do today with these patients  is not only ethical, but is probably very reasonable, and that opens the door to a lot of other ways of gathering information on risk and benefit that are not just death or amputation.

            And perhaps later if we talk about where do we go next we could amplify on that, but I think that's a territory, randomizing against standard care in a population who, frankly, are as ill and difficult to manage and who suffer as much as these people do; I think actually you've got a pretty fat target there.

            And if, indeed, what's conveyed by the investigators today is that the laser component, whether you pull a coronary laser off the shelf off label or do it as part of a protocol with a little better designed instrument, that there's an important adjunctive role for this instrument that proving it is probably worth doing and feasible to do.

            DR. BRUCE GRAY:  Can I just make one comment to that, sir?  Bruce Gray.

            We've walked through many of the alternatives of what good clinical practice could be in that patient population, and to take a patient and say, "Ma'am, we can either try this endovascular treatment strategy or we can randomize you to primary amputation or to medical therapy," I just don't think that would be  a workable alternative for a lot of the reasons mentioned.

            Was that what you were suggesting?

            DR. KRUCOFF:  Well, no and yes, Bruce.  I mean, the reality is today without the LACI protocol running, what we all do with these patients is standard care.  Now, we do different things.  You know, we may go find the surgeon who turned them down and say, "Come back here.  You know, I need you to look again."

            We may go after it with a balloon even though we have little hope for a durable solution, but that's the real world.  That is standard care.

            And it may be that this is not a population in whom you could prescribe that anything would be prohibited or prescribed.

            DR. BRUCE GRAY:  I guess what I'd rather have is the opportunity to use the bigger catheter.  I'm, therefore, going to be using a coronary device in an off-label territory, and I think it would behoove us to use the more appropriate size device in that territory.

            DR. KRUCOFF:  Well, I would say if you're going to experiment on human beings without their informed consent by pulling a coronary device off the shelf and putting it in their leg, why don't you just do it in a randomized trial and then you could find out what you're doing and standard care, too.

            CHAIRMAN LASKEY:  This is why we try not to endorse off-label use of the devices up here.  So enough of that.

            Are you okay, Dr. Krucoff?  Great.

            Dr. Aziz.

            DR. AZIZ:  Thank you.

            I've just got a few comments and a few questions.  Is there any evidence either in animal data previously or from looking at vessels of patients who died or who were amputated after having laser done what exactly happens to the vessel?

            DR. REISER:  We did not do postmortems on patients who died in this study, no.

            DR. AZIZ:  What about animal data?  You  know, that's when the coronary excimer was initially brought about.   Does it get re-endothelialized over time or do you just have a raw channel of collagen lying down there?

            DR. REISER:  Those studies were done many years ago, as I mentioned.  I can't quote those results off the top of my head.  Is there a particular outcome that you're interested in?

            DR. AZIZ:  Well, it would be interesting to know, for example, if it gets re-endothelialized, I mean, it would be interesting to know once the tissue responds to injury.  I know that in the acute stage obviously you're zapping, and you may be vaporizing and getting rid of the clot, but in terms of long-term patency, if you have the data, I mean, it's not going to change what you're doing, but it would be useful at least from my point of view if you had that, but if you don't, I mean, it doesn't really matter.

            DR. REISER:  Sorry.  I don't have an explicit answer for that question.

            DR. AZIZ:  Okay.  In patients who do get a perforation -- and I realize obviously they're different from the coronary circulation where obviously if you get a perforation, and I've seen that, it's quite a major event -- I saw that one or two of the patients had to go to surgery.  Now, could you use a covered stent if that happened rather than taking them to surgery?

            Maybe, John, you could answer that.

            DR. LAIRD:  The great majority of the time when we have a perforation in the periphery, it's a non-event, particularly when that perforation occurs in the SFA, and when it happens, it's usually in the middle of an occlusion.  So there's very little, if any, bleeding.

            The time when it is an event is when it happens in a tibial artery where there could be bleeding in the compartment and a compartment syndrome.  I'm not aware that in the LACI trial we had anything like that occur.  Certainly if it happened in the SFA, there are off-label -- sorry, Warren -- but there are devices that can be used off label to seal a perforation, you know, wall graft or the VIABAHN stent graft from W.L. Gore, and conceivably if you had a perforation in a tibial artery you could use another coronary covered stent in an off-label manner for that patient's benefit.

            But it's generally a non-event.  It's a very low likelihood or low risk event, and in all of the studies of laser angioplasty in a perforation, it's shown to be safe in that regard.

            DR. AZIZ:  You know, looking at some of the angiograms you showed, even though they have critical limb ischemia, some of the patients had these "tweaky" collateral vessels, and some obviously were sort of black and there was really nothing distally.

            When you looked at the data, was there any correlation in terms of long-term patency, patients who had, let's say, better runoff in terms of, you  know, better collaterals, that dated (phonetic) better than if it was just like in a dead end, black territory?

            DR. LAIRD:  We have limited data, and it's, you know, small numbers where clearly when we're able to establish straight-line flow to the foot we have better limb salvage, and that's consistent with previously published literature on the treatment of patients with critical limb ischemia.

            But other than that, no, we don't really have that kind of angiographic analysis.

            DR. AZIZ:  Looking, again, at the mortality sort of figures and going case by case, in addition to the fact that a large number of these patients have cardiac problems and congestive cardiac failure, there were a number of patients who had sepsis as a sort of endpoint.

            Some of the patients that you see with critical limb ischemia may have dry, gangrene, and some obviously were quite colorful feet that you showed that were wet.  Was there a difference in terms of who got sepsis if it was dry gangrene foot versus a wet, soggy, infected area?

            DR. LAIRD:  I don't think that we have that data to provide.  Obviously, these were sick patients.  Two-thirds of them were diabetics and source of sepsis in those patients can be, you  know, multiple, pulmonary, gall bladder, peripheral, but we don't really have -- I don't have that data t provide.

            DR. AZIZ:  The other thing, in your database was there a lot of information on fibrinogen levels, hematocrit, and things like that?  I mean, I didn't see a lot of that stuff, but do you have the data?

            DR. LAIRD:  It would not be normal practice to do fibrinogen levels in patients who are undergoing any kind of peripheral intervention, other than perhaps thrombolysis.  So we don't have that data.

            DR. AZIZ:  Because, you know, years ago, I think, Dormandy -- I think you've got his name in the back here somewhere -- but he's in English and showed that some of the long-term outcomes in patients related to fibrinogen and viscosity.

            Anyway, that's just a side issue.  I thought if you had it, it would have been helpful.

            I see also that the anti-platelet and anticoagulation regimen was really quite variable, and you left it to the investigators, which maybe that's just the way trials are designed, but I know that in the coronary literature, you keep that fairly tightly controlled.

            DR. LAIRD:  I would say in general in all of the peripheral trials that are being done now, a lot of these issues are left now to the discretion of the operator.  There is clearly no standard of care with regard to anticoagulant or anti-platelet regimen after peripheral interventions.

            Certainly, we assume certain things, that aspirin and ticlopine or Plavix would be better than aspirin alone, but there is no basis for that in the literature.  There's an ongoing randomized trial looking at that.

            There's no randomized data or good scientific information about the role of Coumadin chronically in these patients either.

            DR. AZIZ:  Just something that might come up in the future, in patients who have peripheral limb ischemia, obviously some of them have had bypasses either using a saphenous vein or, let's say, a manmade material, you  know, PTFE or the like.

            You know, this technology could be used in occluded or occluding PTFE grafts.  Is there any evidence of, against --

            DR. LAIRD:  There is only anecdotal experience.  I had some of my own where we've opened up chronically occluded bypass grafts, and sometimes, you know, you're not sure as you're getting into the vessel whether it's a native vessel or the graft because there's no stump found.

            And that has been in anecdotal experience successful or short term, but no, there's no systematic study of that group.

            DR. AZIZ:  Okay.  Thank you.

            DR. NORMAND:  Hi.  I'm last, and I have a few -- well, I think I'm not last -- but I have a few technical questions.

            CHAIRMAN LASKEY:  You're last.

            DR. NORMAND:  I'm last?  All righty.

            I realize we have talked a lot today about the comparability of the two cohorts, and what I want to do, not to beat a dead horse, I just want to be able to understand sort of the directionality of some of the findings, and so I just want to clarify a little bit of that.

            And the first question I think may just be my misunderstanding.  I had thought that the LACI group were sicker.  They've been characterized as sicker, and they weren't supposed to be candidates for surgery; is that correct?

            Well, then there's a variable that you describe as high surgical risk.  Yet only 46 percent of the cohort, patients in the LACI group, are considered high surgical risk.  Should that be 100 percent?

            That might be my misunderstanding of the variable.

            DR. LAIRD:  There were three reasons why patients were felt to be not good surgical candidates.  Significant cardiac or medical co-morbidity was only one of the three.  The other two were an absence of a vein, which is a very important component when you're talking about distal revascularization, and the other component was poor distal vasculature, poor targets for bypass.

            Forty-six of the patients met that one criteria, which was significant cardiac and medical co-morbidity with high ASA classification of four or greater.

            In general though, this patient group was a higher risk patient group than the historical control, and that one slide I showed I think outlined that very well.

            DR. NORMAND:  But just so that I can understand, you're saying that empirically you would not expect that variable to be 100 percent, that is, 100 percent of the patients in the LACI group.  You wouldn't expect 100 percent of them to be characterized as a high surgical risk.

            DR. LAIRD:  Correct.

            DR. NORMAND:  Less than 50 percent doesn't seem reasonable, but anyhow, you're saying less than 50 percent are characterized as high surgical risk.

            This is important because the reason why I'm saying this is because I want to talk about the directionality of some of the findings.  So just to emphasize --

            DR. REISER:  That particular criterion properly expressed should be high risk of surgical mortality as evidenced by ASA Class 4.

            ASA Class 4 means that you expect at least two percent and perhaps higher surgical mortality just because of the patient's co-morbid conditions.

            DR. NORMAND:  Okay.  So -- I'm sorry.

            DR. REISER:  So it's mostly a marker of expected mortality under surgical conditions.

            DR. NORMAND:  Okay.  But taken as a whole, if I look at at least the measurable characteristics, it seems to me when I look at them on the page that sometimes the LACI group looked sicker and sometimes the control group looked sicker, at least if I look at the variables univariately, that is, one at a time.

            And I could pull up some numbers, but the point being that sometimes the control group had higher rates on bad conditions.  I mean, I think smoking was one of them and a bunch.

            The reason why I'm raising this issue is because it has got to do with how you collected the data and your other endpoints, but we have two cohorts that for the measurable characteristics that you do have, it certainly does seem sometimes the LACI group is sicker, but as I mentioned, sometimes the control group is sicker.

            And the problem that I have is how to weigh how some of the -- you know, what outweighs the "sickerness"?  I mean, these variables are more important than the variables on which the control group is sicker.

            And so hence, when I'm trying to interpret some of the findings in terms of the size of the difference between, you know, your primary endpoint as well as your safety endpoints, I'm trying to figure out, gee, you know, as people have said already, apples to apples.

            To go to the first question that was asked earlier today of how you actually find the delta for two groups of patients that are comparable, there really isn't any way to do that.  At least there's an ad hoc way which no statistician in their right mind would advocate.  I mean, you could say are they additively; you know, is it multiplicatively the sicker or additively sicker?

            I mean you just can't do that.  So I felt, you know, looking at it from the beginning you had a really tough statistical problem, and that is how do you define, you know, the size that makes sense for a group of people that aren't comparable.

            And I would even argue it's still even a bigger problem -- not a bigger problem.  It still is a problem even if you say overall they're sicker.  I mean, statistically you can't -- the idea of equivalence, as you mentioned, to me at least makes no sense.

            So with those comments then, I wanted to say something about your safety endpoint, and that has to do with mortality, if I recall, and with the fact in mind that you're saying that the LACI group are sicker, then I get concerned when I see more loss to follow-up in your cohort relative to the control group because presumably you're saying they're much sicker, and so I get worried about that lack of information.

            I just have no way now to sort of figure out.  I get worried with missing data and with missing data in which you say that the cohorts aren't comparable, and moreover, the LACI group is sicker.  It just raises a lot of, you know, how can we sort of take that missing information and make an inference about it.

            Because I'm assuming obviously the mortality rates are only for the observed data, the people that you do have information on.  So that's another question.

            The third question relates to the -- and this also is important in terms of how the information is collected.  I realize I actually -- you know, I think you can answer some questions without a randomized controlled trial.  So I'm not objecting to that. 

            I guess what I'm objecting to is the type of observational analysis that you actually conducted was a challenging one. 

            And so one question I did have was whether or not -- how sure are you that the variables that you're using to characterize the two cohorts, as well as the endpoints, are measured in the same way between the two groups?

            I mean, are there standard protocols?  Is it obvious?  What information can you provide us that the way the data were collected in the two studies is comparable?

            DR. REISER:  Boy, you asked a lot of questions.

            DR. NORMAND:  That's why they leave me to the end.

            DR. REISER:  Let's see if I can remember them.  The last question was how are the data collected and how were we sure that the data collected were in the same way.

            DR. NORMAND:  Well, just really it's that last one in terms of how can you be sure that they were -- you're measuring the same thing.

            DR. REISER:  The risk factors that were noted on our case report form are basically yes/no patient conditions.  I think they're pretty straightforward in the sense that they don't require a laboratory measurement or an angiographic core lab or some other core lab to measure those risk factors.

            So I think the risk factors that were specifically mentioned, say, in the control paper are fairly straightforward to note in a patient's physical and history.

            There were two blood lab results, creatinine and blood urea nitrogen that we did note, but the control publication did not reveal those for their particular population.  So there's no comparison to make there.

            Whether a person was a smoker or not, I think that's pretty straightforward.

            Was there something specifically that caught your eye that might be procedurally related?

            DR. NORMAND:  Well, all of them, all of the variables that you're using to say that.  You know, sometimes you're saying the patients are sicker and then I can refer to a page in the handouts where you say the patient population is balanced, which is contradictory to the fact that you're saying that they're sicker.

            But it's all of those characteristics that are important.  It's not just the endpoint.  It's also whether or not -- again, it's my trying to get the directionality of the findings that, indeed, if the history of hypertension and diabetes was collected in the same manner in both studies, maybe that's a no-brainer, maybe.

            But I could have in Italy maybe -- I don't know who fills it out there and who fills it out here, but it's those types of questions I have about the risk of your two populations in terms of their comparability.

            DR. REISER:  Perhaps I could ask a clinician to comment on taking a patient's history.

            DR. NORMAND:  It's really not who's taking it.  It's who's filling out the forms, and in looking at the paper I couldn't get a sense of not knowing exactly what you did.

            I mean all you need to sort of -- my point is, the question I'm asking is whether or not you feel that the data were collected in a similar manner so that, indeed, when we look at history of CABG, when it's recorded in the control group versus LACI, that you can say, you know, it's basically the same protocol in terms of how they're reporting that information.

            That's essentially what I'm asking.

            DR. REISER:  I believe it was.  I believe both studies were run according to modern standards, reading the hospital records and writing down the information on case report forms.

            DR. NORMAND:  And I just have --

            DR. LAIRD:  Excuse me, Sharon. 

            Then who categorized the ASA class?  Did you investigators do that or did you have a third party do that?  How did they get into the various categories

            DR. REISER:  I believe the investigators did that.

            DR. LAIRD:  So a non-blinded observer.

            DR. REISER:  Correct.

            DR. LAIRD:  And in Italy do you know how the --  I didn't read the small print in there, but did they have some kind of overseer, third party do this?

            DR. REISER:  For?

            DR. LAIRD:  For the Italian study, for the ASA class or the Rutherford class.  How did they put the patients in those classifications?

            DR. REISER:  I don't believe that was specified in the paper.

            DR. NORMAND:  And I just had one last question, and that has to do with the use of the overall rates of the endpoints in the Italian study.  It was mentioned earlier it was conducted a while ago.  Presumably at the time of the events it was a while ago.  Also in Italian settings, hospitals.

            And so I guess the question I have is would we really expect that rate to be lower or higher now?

            I mean, that's a concern I have.  If we had a contemporary study at the same time, I don't know if we would think that the population that is characterized by the control group in Italy, that it takes place by Italians as Italian patients and Italian hospitals in the early '90s; whether or not that overall endpoint, which I can't remember what it was, 73 percent or something, would we expect that to be 73 percent right now or do we actually expect that to be 68 percent?

            In other words, you've got -- I'm worried about the fact that it's in a different country, different patients, different hospitals, different time. It's not the randomization.  It's where it's taking place in terms of how that number should compare to the group that your particular cohort rate now.

            DR. REISER:  I don't know how to allay that concern with a solid statistical argument.  I don't have enough information to compare, for instance, Italian data to American data.  I don't have an equivalent study that was done after 2000 to show you that there are trends that may make the results of the particular control paper that we used different than the ones that are published.

            So I'm at a loss to make any sort of reasonable statistically based argument that --

            DR. NORMAND:  Actually it wouldn't be statistically based.  It would be substantively based.

            Is there any reason to believe that somehow things got much better?  The trends are such that the salvage rate would be 90 percent now in such a group.

            DR. REISER:  Not to the best of my knowledge.  There are no wonder drugs for this patient cohort.  I don't believe bypass surgery has made quantum leaps forward in terms of its technology or its implementation.

            Those two cohorts comprise the bulk of the ways that these patients were treated.

            DR. NORMAND:  Okay.  Thank you.

            CHAIRMAN LASKEY:  All right.  Thank you.

            Are you sure that's it?

            DR. NORMAND:  I'll stop.

            CHAIRMAN LASKEY:  Okay.  And we're on schedule.  Does anybody want to break?  Yes, okay.  Need some fresh air.

            All right.  I have two o'clock.  Let's regroup at 2:15, and the panel will do its thing.

            (Whereupon, the foregoing matter went off the record at 2:03 p.m. and went back on the record at 2:22 p.m.)

            CHAIRMAN LASKEY:  Okay.  Thank you again.

            Let's resume with the Executive Secretary now reading the FDA questions.

            MS. WOOD:  Okay.  I will read the questions, and then the panel members will have a chance to respond. 

            John, go ahead and bring up the first one.

            Study Design:  the sponsor of LACI predicated the sample size on demonstrating that freedom from major limb amputation at six months, the primary endpoint, was not more than ten percent worse than the control.  Entrance criteria for the LACI trial were intended to insure that LACI patients are at greater risk from co-morbidities than the control, justifying the ten percent difference.

            LACI intended to enroll a cohort of patients that were not candidates for surgical revascularization based on the inclusion criteria of:  ASA risk of Class 4; or higher or absence of suitable autogenous vein, SAV, for conduct; or the extent of vascular disease.  Patients were not excluded if they were candidates for endovascular procedures.

            Sixty-six, 46 percent, of the 145 LACI patients were classified as being in ASA 4 anesthesia risk status.  Forty-six, 32 percent, of the 145 patients were described as lacking SAV.

            Univariate analysis established that only Rutherford Class 6 was a predictor for major amputation in the LACI study and occurred with similar incidence in the treatment and control groups at baseline.

            1.  Please comment on the following aspects of the study design:

            (a)  Please comment on whether or not the characteristics of patients in the LACI trial and the control group demonstrate an increased risk for limb loss in LACI sufficient to justify the ten percent difference for the primary effectiveness endpoint.

            CHAIRMAN LASKEY:  All right.  Did you want to do (b) or we can --

            MS. WOOD:  No, go ahead.

            CHAIRMAN LASKEY:  All right.  I think it has been the consensus of the panel throughout the day that there is not enough on the table here to not even justify the ten percent delta, but the whole concept of approaching this with a non-inferiority study design due to the inability to compare the two study populations.

            Is that a fair summary of today's conversation?

            Okay.  So it transcends the delta.  It starts at the beginning of the study design.

            MS. WOOD:  Okay.  (b)  An active intervention for limb salvage in LACI is compared to a control arm of non-intervention.  Please comment on whether the outcomes for this endovascular procedure can be satisfactorily assessed without comparison to balloon percutaneous transluminal angioplasty, PTA.

            CHAIRMAN LASKEY:  Do you want to summarize the panel's -- do you have a question?

            PARTICIPANT:  I have a comment.  Do you want to summarize it?

            CHAIRMAN LASKEY:  Okay.  I'll just try and summarize then.  I think, again, this is just a corollary to Part A, that the outcomes cannot possibly be assessed or compared without more in depth knowledge of the patient population in the Italian study vis-a-vis similarities or differences in the underlying risk.

            As far as any reference or illusion to PTA, again, I think the better part of today has been devoted to the advisability perhaps of structuring the study from the get-go with the PTA arm, but it's a little late for that.

            Mitch.

            DR. KRUCOFF:  Just a brief comment.  I do think it's entirely conceivable that you could identify a patient population in whom PTA is not a reasonable alternative and potentially investigate this.  This is an open ended piece of the question as an alternative to standard care where standard care is not PTA, and I think that would be entirely conceivable.  It's not extractable from these data.

            CHAIRMAN LASKEY:  Yes.  That's the question of how can we be helpful here, but the answer to Geretta's Question B is they cannot be assessed.

            I'm sorry.  Dr. Somberg.

            DR. SOMBERG:  I concur with your summary.  I just would like to add that I can conceive -- and I think other members of the committee had mentioned this as well -- of a study where one would have this as an adjunct, not necessarily -- and I understand the investigators' concern that a lot of these patients have long lesions, and if this were primary angioplasty, and that would not be the most appropriate therapy, but one could have two groups getting primary angioplasty and stent possibly and one group having the utilization of these specific modalities to aid one, and this may be an appropriate adjunct to that, or you might want to randomize it as the Chairman said against PTCA.

            But there's a whole host of control groups.  And another group would be  just as this retrospective control, which you had no control over though, would be to standard of care, to maybe surgery, maybe a little angioplasty, maybe a little of this versus a sequential, procedurally exact approach using this modality would also provide you a lot of information.

            CHAIRMAN LASKEY:  If you can remember to say that again for Question 4, we'll be all set.

            Thank you.

            That's Question 4 actually. 

            Sir?  Oh, I'm sorry, Cindy.

            DR. TRACY:  I guess it gets back to my struggle with what the right control group is for this thing, and I think you have a set of data now, and I think it's going to be important to identify if there's a way to salvage that set of data by identifying a better group against which to compare the information that's available with laser.

            I agree that PTA may not be the right thing, but I completely think that some type of control could be identified that would be much more appropriate than the group that was chosen, and I disagree that there's an ethical issue because without laser, everybody in the country is doing something to treat these patients.  So there is a control out there that can be identified.  It's not necessarily versus PTA.  It may be versus standard therapy.

            But I think that the control that was chosen is not appropriate, and I just don't think that PTA is absolutely necessarily the control to compare against.

            CHAIRMAN LASKEY:  Okay.

            MS. WOOD:  Safety:  The primary safety endpoint was death within six months.  This occurred in 15/134, 11.2 percent, patents in the LACI study and was not significantly different from the 113/782, 14.5 percent, patient deaths in the enrolled control group.  Patient age was the sole predictor for this outcome and was similar at baseline for both study arms.

            Secondary safety endpoints were serious adverse events, SAEs, as adjudicated by an independent Clinical Events Committee.  SAEs occurred in 48/134, 36 percent, patients and 58/144, 40 percent, limbs in LACI, including patients lost to follow-up.  These SAEs included 24/134, 18 percent, re-interventions and 11/134, eight percent, major amputations.  The SAE rate in the control group was 239/666, 36 percent, and 10/666, one percent, re-interventions and 76/666, 11 percent, major amputations.

            To put these results in context, the sponsor noted that the rate of adverse events at six months is comparable to the rates reported for PTA for periods that extend to five years.

            2.  Re-interventions were significantly higher in the LACI study than the control group.  Please comment on whether the adverse event data from the LACI study provide reasonable assurance of the safety of ELA used to treat CLI.

            CHAIRMAN LASKEY:  Gary, as the primary reviewer, do you want to just talk to the safety component of the study?

            DR. NICHOLAS:  Surely.  I think that comparing the interventions to the chosen control group is obviously inappropriate, and that has been pointed out before.

            I think comparing it to the literature at large in terms of interventions in similar studies that have had catheter based interventions is more reasonable, and if you look at it in that regard, the rate of intervention then is not excessive.

            The only question I would still be concerned about is since the line of interventions continues to slope upwards at the six-month period, is that going to continue to do so?  And my guess is that it would.

            CHAIRMAN LASKEY:  My concern as well.  I guess the other variable here is that it wasn't clear from the package what the indications for re-intervention were.  Either they certainly weren't prospectively identified, and it was hard to tease that out.

            So it most likely is an underestimate of the re-intervention rate in the real population.  So it's probably higher than that, but how much higher we'll never know.

            DR. ZUCKERMAN:  Can we just have some clarification, Dr. Laskey, on Dr. Nicholas' comment?  The question asked if there's reasonable assurance of safety with the six-month data.  You know, it would be nice to have one-year data or beyond, but when you see the word "reasonable" as a modifier, Dr. Nicholas, how would you answer that?

            DR. NICHOLAS:  I think there is reasonable safety at the six-month point compared to the literature available, again, not the strongest based medicine in the world, but I think it's reasonable to state that it appears safe.

            CHAIRMAN LASKEY:  Tom.

            DR. FERGUSON:  As I read that, I had difficulty trying to decide whether the way the study was performed, whether re-interventions had anything to do with safety.

            In other words, to me a re-intervention in a patient who has had a myocardial infarction and so forth, where I can define what the problem is, the re-interventions here I don't think had -- this is just a comment -- had much correlation with the safety aspect.

            DR. ZUCKERMAN:  Okay.

            CHAIRMAN LASKEY:  So perhaps we need to reconstruct the MACE (phonetic) endpoint there, which I guess we'd be more comfortable calling re-intervention a MACE (phonetic) and then that's the safety aspect.  That wasn't done here.

            Okay. 

            MS. WOOD:  Effectiveness:  the primary effectiveness endpoint of the LACI II study was limb salvage, absence of major amputation, at six months.  In the LACI II study, documented limb salvage at six months was achieved in 110 patients, 75.9 percent of 145 patients enrolled.  Of the other 35 patients, 15 dies, 11 were lost to follow-up, and nine had major amputations.  Two other major amputations were performed on patients who subsequently died.

            By comparison, limb salvage at six months in the control group was achieved in 494 of the 673 patients, 73.4 percent.  Of the other 179 patients, 96 died, seven were lost to follow-up, and 76 had amputations.

            Rutherford Class 6 was the only significant univariate predictor for this effectiveness endpoint.  Eleven, 7.5 percent, LACI patients were in this class at baseline.  By comparison, 60, 7.6 percent, control patients were listed at enrollment as being in Fontaine Class V, which includes both gangrenous ulceration and tissue loss.

            Of the 110 LACI patients who were evaluated at six months and were free of major amputation, 43, 39 percent, continued to be classified with CLI.  This is compared to 211, 43 percent, cases of persistent CLI in the control group reported by ICAI.

            3.  The clinical objectives of the study were states as:

            (i)  Protection from acute amputation;

            (ii)  limb salvage;

            (iii)  resolution of CLI; and

            (iv)  preservation of surgical options.

            Please comment on whether the outcomes for the LACI study demonstrate that these objectives have been achieved.

            CHAIRMAN LASKEY:  Well, again, I'll let Gary speak to this.  He was the lead reviewer and, I think, articulates the panel's sentiments.

            DR. NICHOLAS:  I'll just take it A through D.

            Protection from acute amputation.  Again, we are comparing to what we all agree is a very weak control population, but there's no difference.

            Limb salvage. Exactly the same answer, I'm afraid.

            Including the people that remain in the categories that we call critical limb ischemia seems to be the same in the control and the study population.

            And I think the whole protocol here, if I can digress for just a moment, suffers from the control population.  I think there is some merit here.  I think it needs to be sorted out, and I think that this is a technology and a technique that many of our patients will be able to use who are in this desperate situation.

            To come back to then number D, preserving surgical options.  Yes, in the study group of 145 people they did demonstrate that there were two in whom they identified bypass vessels that were not previously present.  We don't know the follow-up on those two patients, and certainly it's too small to make any positive assertion about preservation of surgical options.

            They did have a low incidence of distal embolization, which obviously is a very positive finding, but again, I think the study really didn't demonstrate that we saved surgical options.

            CHAIRMAN LASKEY:  So in summary, overall we have a safe measure, but no real convincing measure of efficacy.  Is that a good way to sum --

            DR. NICHOLAS:  That's the way I look at it.

            CHAIRMAN LASKEY:  Because of the conversation we've had all day about compared to what.  Okay.

            MS. WOOD:  Are you ready for the next one?

            CHAIRMAN LASKEY:  yes.

            MS. WOOD:  Laser ablation requires crossing of the culprit lesions with a guidewire for control of energy delivery.  Where standard guidewire crossing cannot be achieved, "step-wise" use of the laser can assist in achieving guidewire crossing. 

            In LACI, the guidewire negotiated the lesion without need of laser in all but 25/155, 16.7 percent, limbs.  Following the use of laser energy, balloon angioplasty was required in all cases for the final reduction of lesion obstruction to less than 50 percent angiographically.  This procedural success was attained in 132/155 limbs, 85 percent.

            4.  Please comment on the added value provided by the laser therapy, which is used as an adjunct prior to the PTA required for final resolution of the lesion obstruction.

            CHAIRMAN LASKEY:  So if you could rephrase the first third of your comments a few moments ago, I think that was germane to the adjunctive value of the laser here.

            I'm not sure we all share your sentiment, but at least rephrase it for discussion.

            DR. SOMBERG:  Well, my feel, and I'm only going to say my feeling was that the current study did not demonstrate the adjunctive value.  It could be demonstrated in a number of potential study designs, one of which was being that it could be randomized to two groups receiving interventional therapy.  One group only has the laser added as an adjunctive therapy.  That was essentially what I said before.

            I will also interject that within the database that this company has presented, there may be a small but finite group where they could demonstrate benefit because nothing could be done for those patients until the laser was used, although we did see some discussion of what was the approach.  Did they put the guidewire in first?  Did they use the laser first?

            But that might be something that the agency and the company would discuss at a later date.

            DR. FERGUSON:  Warren, could I make a comment?  Somebody else?

            CHAIRMAN LASKEY:  Go ahead, Tom.

            DR. FERGUSON:  No, I'm getting back to the point about using the word "adjunctive," and again would stress from my point of view it is almost necessary to use both modalities, the laser and the balloon as a part of the treatment package, and I just bring that up again because, again, I don't see how the two can be separated, frankly.

            I don't do this work, but --

            CHAIRMAN LASKEY:  Well, and this study is certainly not as near universal use of PTA.  So there would have to be another design.

            Anybody else?

            DR. TRACY:  Warren, I just had a thought.  I completely do not think that we can understand the adjunct value of this thing because initially I thought that I think it was 13 percent that were crossed by laser that could not have been crossed by wire alone, but then it sounded like there was some different technique of laser-wire, laser-wire, laser-wire, laser-wire.

            So I don't think there's data in here that can help us understand.  I think it probably is an adjunct, but I don't think we can look at that.

            DR. FERGUSON:  My comment related to the other end though.

            CHAIRMAN LASKEY:  Yeah, the other 98.

            DR. FERGUSON:  The other end being if you use only the laser I don't think you've done a complete, and they can correct me if that's wrong.

            DR. KRUCOFF:  Warren.

            CHAIRMAN LASKEY:  Mitch, yeah.

            DR. KRUCOFF:  Yeah, I mean, the way I would go would be to do your best to identify people in whom all of the angiographic, morphologic criteria suggest that getting a guidewire through the lesion or a balloon over that guidewire are simply not going to be doable, and in that population even just technical success with or without the laser as an adjunct could give you, I think, some very quick information that would support the ability to answer this question.

            CHAIRMAN LASKEY:  Another study.

            Move on, yeah.

            MS. WOOD:  Risk-benefit:  co-morbidity associated with CLI has accounted for mortality greater than 50 to 60 percent in patients out to five years and as high as 40 percent at two years in some reports.  Primary amputation has been recommended as an acceptable alternative to revascularization attempts in some cases. 

            While freedom from amputation was obtained in 110 of the 155 limbs in this study, 15 patients died and 43 patients remained in Rutherford classifications for CLI.  In addition, rehospitalization for SAEs was necessary for 48, 36 percent, patients.

            5.  Please comment on whether the benefit demonstrated in this study, particularly, with respect to quality of life-years, outweighs the adverse events that occurred and the persistence of CLI documented.

            CHAIRMAN LASKEY:  Well, we gave the sponsor what I thought was a wide open opportunity to talk about quality of life, but he chose not to run with that.  I was curious why.

            I mean, we went after some of the other endpoints here in the study that are certainly important and should be in every study henceforth in this kind of patient population.  But I guess the panel did not discuss in depth the quality of life-years or the risk-benefit ratio, if you will.

            We tried to get that from the sponsor, and we didn't get very far.  That's my take-home message.  It's there.  I mean, you have it.  I don't know why you didn't present it.

            MS. WOOD:  Moving on to labeling, six, labeling for a new device should indicate which patients are appropriate for treatment, identify potential device-related adverse events, and explain how the device should be used to optimize its risk-benefit profile.

            If you recommend device approval please address the following:

            (a)  Do the indications for use, as stated below, adequately define the patient population and procedural use for which the device will be marketed?

            "The Spectranetics CVX-300 Excimer Laser System is indicated for facilitation of limb salvage in patients with critical limb ischemia (associated with Rutherford Categories 4, 5 and 6) who have angiographically evident culprit stenoses and/or occlusions in the SFA, popliteal and/or infrapopliteal arteries, who are poor surgical candidates, and who are acceptable candidates for revascularization."

            CHAIRMAN LASKEY:  Well, these are certainly the inclusion criteria, but I'm not sure we can go any further with that based on where we're hung up with respect to the efficacy of the device.  So I'm not sure where else we can go with that.

            DR. ZUCKERMAN:  That's a fair response.

            MS. WOOD:  (b)  Based on the study results, please discuss whether the proposed warnings, precautions, and contraindications are acceptable.

            CHAIRMAN LASKEY:  Again, with the same caveats, I don't think that these questions really are relevant in this particular --

            DR. ZUCKERMAN:  Well, I think there's a standard policy where we need to review the labeling regardless of what happens in the next section.  For example, I'm looking at Section 3 of your panel pack, which is the labeling, and under the contraindications section, for example, it is written "no known contraindications for laser."

            We need generic type help here.  Is that appropriate?  Are there certain things in warnings and precautions for peripheral vascular device, such as the laser, that one would want to see in an FDA label?

            DR. TRACY:  I think from the standpoint of the warnings and precautions it looks inclusive of the types of things that you'd worry about generically with the use of a laser device.  So I have no problem with what's stated here.

            I think we just need a little bit more clinical information to know whether such statements as no known contraindications, whether that's appropriate.  But I think this how to use a laser device, I think these are appropriate warnings and precautions.

            CHAIRMAN LASKEY:  Is it fair to say it should be consistent with your coronary indication, or is there some reason to broach that?

            DR. ZUCKERMAN:  We can review that labeling, but, Dr. Nicholas, for example, would you have any special concerns that you would want to see in the labeling for a peripheral vascular device?

            DR. NICHOLAS:  No, I think the labeling and the indication as it's written is very well written and I would support that, given the efficacy issue.

            MS. WOOD:  (c)  Please discuss whether the instructions for use adequately describe how the device should be used.

            CHAIRMAN LASKEY:  Well, here I don't think that it has been adequately described.  There are nuances to the approach of these lesions.  You can wire first; you can lase (phonetic) first; you can do a little of both.  I think the IFU needs to be fairly specific if the company is intent on furthering this technology.

            What we heard today were a number of non-protocol regulated approaches left to the discretion of the operator, et cetera.  So perhaps that could be cleaned up.  That's a simple matter of what to do if Wire A doesn't work and laser -- I don't think that's been well delineated.

            Are there any other parts of the IFU that people would like to elaborate on in terms of actual use, hands on?

            DR. KRUCOFF:  Warren, I guess the only question we haven't gotten to asking that was in the back of my mind is whether in general you finish a case over this wire or whether there are cases where you would withdraw or swap through the balloon or whatever.

            I agree with you.  I think there are probably some technical nuances that could go into instructions for --

            CHAIRMAN LASKEY:  Yeah, and now that I think of it, this whole issue about stents and value of adjunctive stents I think needs to be sorted out in your indications for use.  If it's part and parcel of this procedure, of this strategy for patient management, then  I think it needs to be clearly laid out where you recommend stenting and where you don't.

            We heard some interesting options today in real life.  That just needs to make its way into the package.

            DR. SOMBERG:  But if it weren't -- I just need to interject something.  Would you be satisfied if, let's say, it was effective as initial strategy to try to put a wire down?  Twenty percent of patients you could not do it.  The laser was helpful and then they went ahead and got down to a small lesion and had to put one stent in, et cetera.

            That may not test all of the possibilities, and maybe you don't need the stent.  Maybe you need, you know, an extra three centimeters to stent above and below that area, but you wouldn't be advocating they have to explore all of those possibilities?

            CHAIRMAN LASKEY:  No, no, but in the era we live in where stents rule and probably will for quite some time, I think we're being naive if we leave our head in the sand on this, that many peripheral vascular procedures will involve stents and with some chemicals attached to them.  So I think we need to just be aware of that.

            This appears to be a useful adjunct, sort of standard.  Perhaps together they're better, but you just need to clear that up in the instructions for use.

            DR. TRACY:  Warren, can I just add that if you just look at the section on directions for use without getting too hand tying to the clinician, if this was an approvable device, this is fairly generic and also fairly good at describing the technical directions for use.

            So I think some of the other concerns that we have about the specifics that you're discussing may never end up in the directions for use.  I think if you're just analyzing directions for use, you take it out of the pack.  You put this wire down such-and-such.  They look fine to me.

            I don't know what the question is.  If it's a question of does this look okay, I'd say the answer is yes.  Is there more than could be there?  Not necessarily even if we had more specific information.  So I'm not sure what exactly the FDA question is.

            DR. ZUCKERMAN:  Okay.  Dr. Tracy, in general, I would agree with your gestalt as to how we look at contraindications, warnings, and precautions.  I think what Dr. Laskey is leading you to is to Question 7, which is more directed towards what should go into the description of the clinical trial.   It's on page 3 of the label, and it has some of these subset analysis that you've talked about, and maybe if you look at Question 7 it will help you determine whether some of this information should be in the clinical trials description.

            CHAIRMAN LASKEY:  And this is a clinical trial in which there was 98 percent use of balloons and X percent use of stents.  So that's going to be hard to overlook.  So I think it's part of the package.

            So seven.

            MS. WOOD:  Please indicate if the following findings are sufficiently robust to warrant incorporation in the label:

            (a)  The 110 LACI patients in Rutherford Clinical Categories 5 and 6 experienced 15 percent mortality and an amputation rate of seven percent.  This contrasted with one percent mortality and two percent amputation rate in 45 Category 4 patients.

            (b)  Seventy limbs in the LACI study also required stent placement.  Stents were placed in 56 superficial femoral arteries (SFAs) in the 104 limbs with SFA lesions.  Forty-nine, or 87.5 percent, of the SFAs with stents remained amputation free at six months.

            DR. NICHOLAS:  My response is yes.  I think both should be included because 7(a) regarding the Rutherford classes and outcome bears upon case selection, and it might have a significant influence on choice of patients for whom the procedure would be recommended.

            Seven (b) gives support to the comments that Warren just made about virtually everybody gets balloon angioplasty.  Then the 80-some percent get a stent in their SFA if this type of procedure is expected.

            So I think the operator looking at those instructions would be well served that he or she knows they are going to be moving on to balloon and/or stenting after they've utilized the laser.

            CHAIRMAN LASKEY:  So as statements they certainly should stand.

            MS. WOOD:  Okay.  Are you ready to move to eight?

            CHAIRMAN LASKEY:  Yes.

            MS. WOOD:  The sponsor has proposed the following training requirements in the draft instructions for use:

            "The use of the CVX-300 Excimer Laser System is restricted to physicians who are trained in atherectomy, percutaneous transluminal coronary angioplasty, PTCA, and who meet the training requirements listed below.  These requirements include, but are not limited to:

            "1.  Training of laser safety and physics.

            "2.  Review of patient films of lesions that meet the indications for use.

            "3.  A review of cases demonstrating the CLiRpath catheters in lesions that meet the indications for use.

            "4.  A review of laser operation followed by a demonstration of the CVX-300 Excimer Laser System.

            "5.  Hands-on training with the CVX-300 Excimer Laser System and appropriate model.

            "6.  A fully trained Spectranetics representative will be present to assist for a minimum of the first three cases.

            "7.  Following the formal training session, Spectranetics will make available additional training if so requested by the physician, support personnel, the institution or Spectranetics."

            Please comment on whether these training requirements are adequate.

            DR. WHITE:  Warren, I don't see why we're asking for coronary angioplasty as a certification for this.  It should be peripheral angioplasty, not PTCA, but PTA.

            And I'm not quite sure what we mean by "atherectomy," since that's generally a procedure that we don't do anymore in the leg.  So I would think that the qualification for using this device would simply be someone who was angioplasty credentialed in the periphery.

            CHAIRMAN LASKEY:  Yes, Dr. Maisel.

            DR. MAISEL:  I'm not sure I see a need for Number 6, that a fully trained Spectranetics representative needs to be present.  Certainly that would make sense for a physician who's not at all trained in this, but if the device were ultimately approved and a physician were trained and it's passed on from physician to physician or physician to fellow, I'm not sure that that is a necessity.

            CHAIRMAN LASKEY:  That's a CYA kind of thing.

            DR. WHITE:  Actually I think that it's important that that be there because you don't want the company to withdraw support, and for the initial -- I mean, it's certainly up to the institution.  If you've been using this device in the coronaries for three years, it's not going to be a problem to use it in the legs, but I think if you're going to use it for the first time in the legs, it's important to have, I think, someone who understands the operation of the device.

            So I think that's fair enough to leave.

            CHAIRMAN LASKEY:  A proctor would be better, but that's an opinion.

            DR. NICHOLAS:  To start at the bottom, I think Number 6 should be rephrased also, three proctored cases, but I think also the first paragraph of italicized qualifications should not be there because it becomes very restrictive and, again, brings into the issue of which group of doctors is going to be able to take care of these patients.  And you get access to the right tool rather than the individual skills of an individual.

            DR. WHITE:  Did you just say that you think a proctor needs to be there?  The first part of that, did you --

            DR. NICHOLAS:  Well, the question of Number 6 which was raised of do you really need somebody there to watch you do the first three if you have already been doing coronary lasers.  Probably not, but if you're going to write a standard for the use of the device, supervised three times or have experience with X number of procedures at the coronary level, I think, would meet the needs.

            DR. WHITE:  I guess I just want to make sure that we're not overreaching a little bit because I really don't think a proctor -- it would not be a good use of my time to go watch somebody do this.  I don't think this is a -- I mean, it's a skill, and there's some sense, some tactile sensation, but this is not something that an expert, a company person can't easily walk you through.

            This system is, I think, actually pretty user friendly.  The hardest part to me in the system is actually setting up the laser and the software, and that's what generally the company guy does better than anything else.  Actually advancing catheters over guidewires, whether they're lasers or balloons or stents, are all kind of -- so I think a proctor is probably not a great addition to that, but I think the company support is.

            DR. NICHOLAS:  But you'd rather see the company representative there?

            DR. WHITE:  I would.

            DR. NICHOLAS:  I have no dispute with that.

            DR. KRUCOFF:  Maybe another way of approaching this is do you really need to concentrate on training a physician or could you profile how to establish whether or not a site is ready, and where certification could be to have at least one physician on site who has done at least three cases and a staff who knows how to operate the device, and from there on they can train their own, you know, if they have younger people coming in.

            But one way of approaching this might be in the same way we've done other technologies that have multiple pieces like this, is for the company to make sure that a site has the resources on site that it needs to know what it's doing and then after that let them do their thing.

            CHAIRMAN LASKEY:  Gary, to pick up on your thought, how about the use of the CVX-300 Excimer Laser System as restricted to physicians who are trained in peripheral vascular intervention?  Do you to like that?

            DR. NICHOLAS:  No, I'd be fine with that.

            CHAIRMAN LASKEY:  Okay.  I think we ought to get rid of the PTCA.  We ought to get rid of cardiologists -- I mean in this sense --

            (Laughter.)

            DR. WHITE:  Agreed.

            CHAIRMAN LASKEY:  Okay.  So wording is "trained in peripheral vascular intervention."  All right.

            DR. ZUCKERMAN:  And is there consensus on Dr. Krucoff's comments that Points 1 through 6 could be rewritten with the sponsor to certify site training as opposed to individual physician training for each physician at that site?

            CHAIRMAN LASKEY:  You could add that to one through six.  I'm not sure it supplants it.

            DR. KRUCOFF:  I would supplant it.  Just make it a condition of selling the catheters to a hospital and let the hospital carry  it forward from there.

            CHAIRMAN LASKEY:  I don't have any strong feelings on that.

            DR. NICHOLAS:  I think it needs to define what the package is going to be that gets hospital approval, and I think one through six or seven really do that.

            DR. KRUCOFF:  Yeah, basically I agree, Gary.  I think, you know, if you have one through six for at least one doc on site and you train the staff because, as Chris says, the interventionalist is just a point and shoot person, and apposition and I mean there are a lot of important elements to that, but the staff setting of operating the instrument is the other piece.

            And after that I think you can train your own.

            DR. WHITE:  The one concern I would have is that in many hospitals, in ours certainly, this device is used in multiple locations.  It can be taken down to the operating room and used by a surgeon.  It can be brought to the cath. lab.  It can be used in radiology.

            And so if you simply train a guy in the cath. lab to do this, that expertise may not travel to the operating room, and so I think that if you actually link the usability to the user, then if the surgeon wants to use it in the OR, he's going to get this education.  It may be redundant for what the cath. lab has done, but it probably is not a bad thing to do.

            DR. ZUCKERMAN:  So what is the consensus?

            DR. TRACY:  I would leave it more training is better, more is better.

            CHAIRMAN LASKEY:  We would leave it, but add niches.  I mean, that's an option.

            DR. ZUCKERMAN:  Okay.

            CHAIRMAN LASKEY:  May I take this opportunity to point out there's no patient information brochure, nothing for the patient?  I think that needs to be added to the pile.

            Good.  Dr. Zuckerman, does the FDA have any additional comments?

            DR. ZUCKERMAN:  Just one moment.

            (Pause in proceedings.)

            DR. ZUCKERMAN:  Okay.  No additional comments.  Thank you.

            CHAIRMAN LASKEY:  Thank you.

            And for the sponsor, does the company have any additional comments or questions before the vote?

            DR. LAIRD:  I would like to thank you for your time, and I would like to make a few additional comments.

            I think obviously there were some limitations to this study design that the FDA helped us device, and they have been well, you know, ground through today. 

            The challenges of trying to demonstrate efficacy against a  historical control where the majority of the patients did not receive an intervention, I think, is really sort of an insurmountable  problem, but I feel extremely confident that we have demonstrated safety for this device in this population of very sick people, and by any measure of historical data treating patients with critical limb ischemia, we achieved excellent results.

            I think if I were to ask any of you would you accept a procedure for your patient that had zero percent, 30-day mortality and a limb salvage rate at six months of 92 or 93 percent, I think in general you would be very happy with that therapy.

            And we can do randomized studies, and I can predict what that randomized study will look like.  We will randomize laser assisted angioplasty against PTA, and in that study, despite our best efforts, 50 or 60 percent of the patients in each arm of the trial will get stents.

            And three years from now we'll sit here and we'll try and tease out what the benefit of the stent was and how it impacted on the laser or the balloon results, and we will be nowhere and we will have lost a lot of time, and I think our patients will have suffered because of this.

            I don't personally, as a person who takes care of patients with peripheral disease every day, see any other alternatives in terms of a randomized trial.  We can randomize against amputation, but I would put you in my shoes.  How would you like to offer a patient the alternative of a percutaneous intervention or having a below knee amputation?

            We could certainly try and randomize against surgery with a synthetic conduit in patients who don't have any lower extremity saphenous vein, but even trying to do any kind of randomized trial where you randomize against surgery is challenging at best.

            I think we have done the best we could with a very difficult patient population and have demonstrated extremely good efficacy despite this challenging patient population with very low complication rates, and I think you have the opportunity here to approve a device for these patients that can help them, and it can be labeled in a manner; say, perhaps this can be labeled as a device that in conjunction with the usual tools, balloon angioplasty, perhaps stenting, has the opportunity to provide an excellent limb salvage rate for these patients.

            Thank you.

            CHAIRMAN LASKEY:  Thank you, Dr. Laird.

            Comments from industry?  Mr. Morton?

            MR. MORTON:  Well, I'd like to echo what we've heard all day, to just acknowledge that the presentations have been excellent.  Obviously, the investigators are passionate about the benefits of this device for a very sick patient population.

            I'd like to give special thanks to the Agency because earlier today they helped clarify what the requirements were for valid scientific evidence.  There was a question about randomized controls, prospective controls, and as a matter of fact, those are not required by law, and I appreciate that clarification.

            Today we've seen an example of the dilemma that the FDA and the sponsor often have when they're trying to design a study, and a control may not be available.  It may not be ethical in some cases, and a study is developed the best that we can with what we know at the time.

            And then as that study goes on, other things happen.  New medications, new techniques, new devices come out, and at the end of the day when the study is complete, you might not have designed the study that way, but nonetheless, you've done the study, and you must make use of the data that you have.

            So, again, I thank the panel.  I thank the Agency for that clarification, and I'd ask the panel to keep that in mind today and also in future reviews.

            CHAIRMAN LASKEY:  And Dr. Hughes on behalf of the consuming public.

            DR. HUGHES:  Thank you very much.

            I also want to commend the sponsor and the FDA for their presentations, and also my colleagues here on the panel for their in depth and insightful analysis and review, you know, of this device.

            I have just one question.  I hope it's okay to ask this at this point to the sponsor.  Those patients, you  know, who had a limb amputated, to what extent or to what degree would they be -- any of them -- be candidates for a prosthetic device?  Any hope at all?  Any chance at all?

            I didn't quite get a sense of whether in terms of quality of life that, you know, once amputated are we talking about just having a stump and there not being any chance at all of any kind of a prosthetic device?

            DR. RAMAIAH:  Well, I think it all boils down to the question of amputation versus revascularization, and a lot of studies have been done.  The Delphi Consensus Study is there which evaluated about 956 patients.  Between radiologists, cardiologists, and surgeons, there was only a small percentage, nine to ten percent, which said that amputation should be the primary treatment.

            Quality of life question is quality of life studies have been done on patients who have been amputated, and obviously it has shown that those who revascularize or those who have options for revascularization do a lot better in terms of depression, in terms of social affability, social interaction, and in terms of physical mobility.

            So revascularization is definitely the way to go in terms of amputation.

            Having said that, if there is no other option of revascularization, then obviously amputation is the only treatment, and there is -- at the current rate of prosthetic development, these patients can be rehabilitated, but if you compare them to the patients who have been revascularized, obviously quality of life is definitely better for the patient with revascularization than the patient with an amputation that is being rehabilitated.

            DR. HUGHES:  Okay.  Thank you.  I think I understand, you know, what you're saying there.

            And also, I guess, trying to get a really clear sense of alternatives, I believe it is outlined rather well in Section 2, the summary of safety and effectiveness, somewhere around, I believe, page 26, but I really want to get a clear sense.

            This particular device and procedure, this LACI, it would be pretty much considered last resort, wouldn't it, or would it not?  Last resort?

            DR. LAIRD:  Well, I think the standard of care in most places for patients with this problem is surgical revascularization, and the study design was basically looking at a group of patients who, in essence, had very, very few options.  They were not good candidates for surgery, and they had very diffuse disease, and they had critical limb ischemia.  So they were at great risk for losing their limb.

            So, in essence, yeah, it's sort of a last resort, you know, last stop before potentially going on to amputation.

            DR. HUGHES:  Okay.  Thank you.

            Okay.   Those are just a couple of things I really wanted to get clear in my mind in terms of a consumer representative.

            Once again, I think that the panel has, you know, done its job very, very well in ferreting out, you  know, these issues, very complex issues in terms of comparing the population for the LACI procedure to those in the control group and that not being really appropriate is the way that I see it.

            But I think that the panel in the end most likely will have some very good recommendations, you know, concerning that kind of issue.  I think they're coming out already.

            So anyway, I just want to leave it at that.  I think everyone has done as best a job as conceivable and reasonable under the circumstances.

            Thank you.

            CHAIRMAN LASKEY:  Thank you.

            I'd like to just briefly open the open public hearing portion again.  Is there anyone who wishes to step forward and address the panel on today's topic?

            (No response.)

            CHAIRMAN LASKEY:  If not, I'll close the open public hearing portion and ask Ms. Wood to read the voting options.

            MS. WOOD:  The Medical Device Amendments to the Federal Food, Drug and Cosmetics Act, the Act as amended by the Safe Medical Devices Act of 1990, allows the Food and Drug Administration to obtain a recommendation from an expert advisory panel on designated medical device pre-market approval applications, PMAs, that are filed with the Agency.

            The PMA must stand on its own merits, and your recommendation must be supported by safety and effectiveness data in the application or by applicable publicly available information.

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

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

            Your recommendation options for the vote are as follows:

            Approval if there are no conditions attached;

            Approvable with condition.  The panel may recommend that the PMA be found approvable subject to specified conditions, such as physician or patient education, labeling changes, or a further analysis of existing data.  Prior to voting, all of the conditions should be discussed by the panel;

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

            Following the vote, the Chair will ask each panel member to present a brief statement outlining the reason for their vote.

            CHAIRMAN LASKEY:  Thanks, Geretta.

            So the recommendation of the panel may be approval, approvable with conditions that are to be met by the applicant, or denial of approval.

            I will now ask for a motion on the PMA.

            DR. NICHOLAS:  I would move that the proposal not receive approval based on the fact it has not been shown to be effective, but certainly, I think, has shown to be safe.

            DR. SOMBERG:  I second the motion.

            CHAIRMAN LASKEY:  Is there a second?

            DR. SOMBERG:  Second.

            CHAIRMAN LASKEY:  So it has been moved and seconded that the PMA is denied approval.

            DR. ZUCKERMAN:  Now the panel needs to vote on that motion.

            CHAIRMAN LASKEY:  Okay.  So can we engender, at risk of prolonging this discussion of this motion?  If not, I suggest we vote on the motion.

            Again, the motion is to deny approval.  All in favor of denying the approval, raise hands, please.  High.

            (Show of hands.)

            CHAIRMAN LASKEY:  We're counting right?    So one, two, three, four, five, six, seven, eight, nine in favor of the denial of approval.

            All against?

            (Show of hands.)

            CHAIRMAN LASKEY:  One.

            DR. ZUCKERMAN:  For the record, Dr. Laskey, can you indicate who voted for and against?

            CHAIRMAN LASKEY:  Yes, I can.  Voting for were Drs. Nicholas, Tracy, Maisel, White, Ferguson, Morrison, Somberg, Krucoff, and Normand.

            And voting against was Dr. Aziz.

            So shall we just finish up with each person's short rendition of why?  You stated your position very well, Gary.

            DR. NICHOLAS:  Well, I think that there's been a strong argument made by the investigators that there's a role for this excimer laser possibly in that short lesion propagated with clot proximal to it.  I think Dr. Gray presented that very well.

            There's clearly a nidus here for providing information that will allow us to approve this technology, and I'd encourage the investigators to design the study that will allow us to do that.

            DR. TRACY:  I also voted for not approvable, and I just echo the opinion that effort needs to be put into finding a control group that's suitable, and that may not in my mind require additional investigation, a new clinical study, but may require identification of some better control group from the literature.

            DR. MAISEL:  I voted  for not approvable for all of the reasons we have discussed previously and agree that I'm quite comfortable with the safety data that's been presented, and it's been an issue of effectiveness and appropriate control group comparison.

            DR. WHITE:  I voted for not approvable based upon my conviction that there needs to be, I think, a contemporaneous control group so that we can tease out, I think, the adjunctive benefit gained from the laser.  I think that there may be an option to look at a group of patients who are not candidates for intervention, and that's certainly -- particularly given the 13 percent or 14 percent of patients who are not treatable with a guidewire crossability, but I don't think there's enough data in this PMA to support that as an exception.

            DR. FERGUSON:  I voted for not approvable for the reasons that have been given around the table, with considerable angst, I might say, because in my heart I feel that this is a viable option, and it's a good option, and I think I agree with Dr. Tracy.  I think that there are ways to salvage this by appropriate multi-institutional studies or some other way without going through a very large study again as you've done this time.

            DR. MORRISON:  Well, I also voted for not approvable with considerable reluctance because I think this is a very sick group, and I think demonstration that there are even a small cohort where the adjunctive use of laser would allow a successful procedure really could be adequate, but unfortunately I don't see the current control group as providing that evidence.

            So with some reluctance, I voted for not approvable.

            DR. SOMBERG:  I also voted for not approvable for essentially the reasons that have been mentioned by fellow panelists.  I am very concerned that this could discourage the development of catheter sizes that are necessary for peripheral vascular, and I do think there's benefit here, and I think it may be culled from the current data set or from additional data sets.

            And I also would like to underscore what I was trying and I think other people have mentioned as well, that there are other things between mortality and more drastic endpoints like amputation and not, such as quality of life, healing, et cetera, which could be compared if a new days is set or was necessary that would not take as long as or be as arduous as this particular study.

            But lacking evidence and proof that there is efficacy, it would be a gross violation of our mission to approve.

            DR. KRUCOFF:  I also voted for not approvable for many of the same fundamental -- I think this is a data set that clearly illustrates safety in a highly frail population.  I can echo the reluctance of saying no to the obvious impression of the individuals who have used this device in these patients that it may well have an important adjunctive role, and that this may set back the time line of our ability to reach those patients.

            I do think on the efficacy side that I would really encourage the sponsor and the investigators to think about short, doable ways.  You know, John, I feel the spirit of the randomization issue, but it's very different if you approach somebody and say, "I've got a trial that's 50-50, 50 percent chance we're going to cut off your leg, 50 percent chance we'll use a laser."

            That's a different conversation than going to a patient who's imminently going to have their leg removed and saying, "We have a trial that at least would have a 50-50 chance of trying something different."

            And I do think randomization in some of these patients is feasible, and in a randomized cohort what I would urge you to do is think about a technical endpoint that would show efficacy at a technical level in patients in whom routine intervention techniques are unlikely to work without laser adjunct, and to prose that in addition to this safety data you've already collected, I would personally find that a very favorable way to try and briskly with a modest randomized trial bring this device forward.

            Another suggestion in the really, really, really sick patients who have bad anatomy and multiple co-morbidities, to consider and even dialogue around whether a human device exemption, an HDE path, a non-randomized path might be something that could be discussed in really the ultra sick where there are truly no other options, and see if those patients could be identified.

            So I really hope that some additional data would be enough to help us all understand data supporting efficacy in addition to all of the hard work that has been done that has provided, I think, key data on safety.

            DR. AZIZ:  I voted that it should be approved.  I agree that the trial design was not perfect.  I think it did demonstrate that it was safe.  I think the efficacy, I think, in this trial under this device and in other devices where you have ongoing concurrent, other therapies like angioplasty and stents is going to confuse the patent both with this device and in the future.

            And I think I don't know quite how to answer that sort of dilemma, and even though the effectiveness was not pure, I think there are a group of patients who really are -- who have the only option is that of amputation.  So I hope that even though that this is obviously not going to pass down, that there would be an exemption or a compassionate use because I think as the data here showed in some of the cases, those legs were truly saved.

            DR. NORMAND:  I voted not approvable basically for the reasons that were mentioned earlier, but I want to emphasize I'm not necessarily advocating use of a randomized trial.  My concerns were related to what I believe to be a poor analysis of observational data, and I think there's a good way to go forward with registry data.

            You unfortunately didn't have the covariate information, but I think there's reasonable and surely sound statistical methods to help go forward without a randomized trial to adjust appropriately for differences.

            CHAIRMAN LASKEY:  Well, we are clinicians up here, and I just want to applaud the sponsor and applaud Dr. Laird for really a cogent presentation.  I think we're all sensitive to how dire these patients are.  This is almost destination therapy, if you will, and perhaps some clever configuring of the adjunctive/conjunctive aspect of this device will go a lot further than is apparent right now.

            But again, I'd like to thank the sponsor, again, Dr. Laird and my panel members. 

            This concludes the report --

            DR. ZUCKERMAN:  Dr. Laskey, before we conclude, can we just comment on several options made by panel members here?

            There have been two potential ways to move forward.  Dr. White and Dr. Somberg have developed the idea of perhaps another trial where this is looked at as more of a niche device for patients where guidewire crossing is not possible.

            Our general experience with that type of trial design has been somewhat problematic in defining when a guidewire can cross a lesion, and you should try a different modality.

            Could you give any other helpful hints, Chris?

            DR. WHITE:  Well, I think I've participated in  trials that required guidewire failure, and I think that, I mean, those trials aren't dependent upon the integrity of the investigator.  I mean, you have to trust somebody sometime, and while I understand that you could possibly subvert the intention of the trial, I still think if you make the argument that 15 percent of these patients or 13 percent of these patients were not treatable had the laser not been available, then that well could be an indication for this device in the periphery.

            And so I think that is worth pursuing for that reason.  I don't know how to make people more honest or I don't know how to quantify water failure.  I'm not trying to make a laser pass, and I don't know how to -- you know, the guidewire police can only visit so many institutions.  So I think you just have to trust the integrity of the investigator.

            DR. SOMBERG:  Just very quickly, I would inject, Dr. Zuckerman, that it's one thing to do a trial only like that and present you the data, but in this particular case there is all of this other data, and there is a trend to feel, from what I'm understanding from most of the panel, that there may be some benefit here, but the problem is there was no way to show that scientifically.

            So, therefore, if you're going to say, well, this is going to be an adjunctive device and you have a choice of no data or having a feeling that something may work, but now all of a sudden you have demonstration that it's a technical tool, it might be useful.

            Also, the way to get around the guidewire police is not to randomize everybody, but obviously to have the population get some other therapy or the procedure is stopped and then some people get the laser therapy.  So that would be able to allow for the fact that sometimes somebody might have been able to squeeze a guidewire through or something like that to see if it really opens that lesion up.

            But I mean, there are ways of getting around it, but I think the point people were trying to convey is that there's a lot of information here.  Unfortunately, it's not one that could be codified in a statistically significant package.  There may be a technical tool package might be useful to bring this rapidly to the fore.

            DR. ZUCKERMAN:  Fine, and then one final question for Dr. Krucoff, who suggested in a subsequent randomized trial a technical endpoint rather than the endpoint used here could be utilized.  Do you have some suggestions for that technical endpoint?

            DR. KRUCOFF:  I actually think they're related because it would be a reach, cross, and dilate in a cohort of patients who could be selected for a likelihood that you're going to start doing some of those or actually in this data set, and then how you handle the interventionalist bias, I think, again, one way to do it is to say in patients who you can't reach, cross, and dilate, although you tried to randomize them or to just count on the integrity of your selected investigators and randomize them ahead of time so that if you are unable to reach, cross, and dilate without adjunctive laser, could you then apply the laser and come to a different end?

            And I guess what that would beg would be necessarily the six-month follow-up and which actually could be treated more in a modular way as safety elements that could be reported later for completeness, but allow a decision about bringing the device to market to be earlier perhaps with an earlier 30-day or even indexed hospitalization primary endpoint of efficacy, given what you already have in hand about safety. 

            If that was something the Agency would consider, that would at least accelerate the time line that would be required to gather a sufficient cohort of patients to bring the question of the effectiveness of this adjunctive use back to the table.

            CHAIRMAN LASKEY:  I would like to suggest to the Agency that they go beyond six months.  I think that's an overly optimistic point at which to truncate the observation.  I think that there's enough events out there which are cumulative that I don't think we have a real picture of what the success is.

            So this concludes the report and recommendations of the panel on PMA P910001 from Spectranetics Corporation for a CVX-300 Excimer Laser System for the treatment of patients with critical limb ischemia.

            Again, thank you all.

            (Whereupon, at 3:31 p.m., the meeting was concluded.)