+ + + + +




+ + + + +



JUNE 8, 2004


+ + + + +


The above entitled Meeting was conducted at 8:00 a.m., at the Hilton Washington D.C. North/Gaithersburg, Salons A, B, and C, 620 Perry Parkway, Gaithersburg, Maryland, Dr. Warren Laskey, Chairperson, presiding.





WARREN K. LASKEY, M.D., Chairperson, Uniformed Services University of the Health Sciences, Bethesda, MD

SALIM AZIZ, M.D., Voting Member, Capitol Cardiovascular & Thoracic Surgery Association, Takoma Park, MD

MITCHELL KRUCOFF, M.D., Voting Member, Duke University Medical Center, Durham, NC

CYNTHIA TRACY, M.D., Voting Member, George Washington University, Washington, DC

KENT R. BAILEY, Ph.D., Consultant, Mayo Clinic, Rochester, MN

THOMAS B. FERGUSON, M.D., Consultant, Washington University School of Medicine, St. Louis, MO

JOHN W. HIRSHFELD, M.D., Consultant, University of Pennsylvania Medical Center, Philadelphia, PA

JOHN C. SOMBERG, M.D., Consultant, Professor of Medicine and Pharmacology, rush University, Lake Forest, IL

NORMAN S. KATO, M.D., Consultant, Cardiac Care Medical Group, Encino, CA

JOANNE LINDENFELD, M.D., Consultant, University of Colorado Health Sciences Center, Denver, CO

JUDAH Z. WEINBERGER, M.D., Ph.D., Consultant, Columbia University, New York, NY

CLYDE YANCY, M.D., Consultant, University of Texas Southwestern Medical Center, Dallas, TX

MICHAEL MORTON, Industry Representative, Carbomedics, Inc., Austin, TX

CHRISTINE MOORE, Consumer Representative, Baltimore, MD

GERETTA WOOD, Executive Secretary, Food and Drug Administration, Rockville, MD

BRAN ZUCKERMAN, M.D., Food and Drug Administration, Rockville, MD



RODERICK M. BRYDEN, President and CEO, World Heart Incorporated

JAL S. JASSAWALLA, MSME, MBA, Executive Vice President & Chief Technical Officer, World Heart Incorporated

JAMES B. YOUNG, M.D., Chairman, Division of Medicine, The Cleveland Clinic Foundation Lerner College of Case Western Reserve University; Medical Director, Kaufman Center for Heart Failure



CHUL H. AHN, Ph.D., Biostatistician, CDRH/FDA

ILEANA PINA, M.D., Professor of Medicine, Case Western Reserve University, Consultant to FDA


Call to Order................................... 4

Open Public Session............................. 9

Sponsor Presentation........................... 11

Roderick Bryden

Jal Jassawalla

Dr. James Young

Dr. Phil Oyer

Dr. Brooks Edwards

Dr. Peer Portner

FDA Presentation............................... 67

Dr. Michael Berman

Dr. Chul Ahn

Dr. Ileana Pina


Dr. Mitchell Krucoff.......................... 113

Response by FDA......................... 149

Dr. John Somberg.............................. 160

Open Committee Discussion..................... 168

FDA's Comments

Dr. Ileana Pina............................... 234

Additional Material Provided by Sponsor....... 236

Review of Questions........................... 240

Public Session................................ 255

FDA Comments

Dr. Zuckerman................................. 256

Sponsor Comments

Mr. Bryden.................................... 257

Consumer Comments

Ms. Christine Moore........................... 259

Industry Comments

Michael Morton................................ 261

Recommendations and Vote...................... 264


8:07 a.m.

CHAIRPERSON LASKEY: Well, good morning. I'd like to call this morning's session to order. This morning's session is the Circulatory System Device Panel addressing the topic of a premarket application for World Heart Novacor N100PC and N100PC(q) left ventricular assist system.

If I may have Ms. Wood read the conflict of interest statement?

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

To determine if any conflict existed, the agency reviewed the submitted agenda and all financial interests reported by the Committee participants. The conflict of interest statutes prohibit special government employees from participating in matters that could effect their or their employer's financial interests However, the agency has determined that participation of certain members and consultants, the need for whose services outweighs the potential conflict of interests involved, is in the best interest of the government. Therefore, a waiver has been granted for Dr. Kent Bailey for his interest in a firm that could potentially be effected by the panel's recommendations. Dr. Bailey's waiver involves a contract to his institution for the sponsor's study in which he has no knowledge of the funding and has no involvement in data generation or analysis. The waiver allows this individual to participate fully in today's deliberations.

Copies of this waiver 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 certain matters regarding Dr. Clyde Yancy. He reported a current involvement with a firm at issue for which he is uncompensated. Because his involvement is not directly related to today's agenda, the agency has determined therefore that he may participate fully in the panel's deliberations.

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 interests, the participant should exclude him or herself from such involvement and the exclusion will be noted for the record.

With respect to all 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.

CHAIRPERSON LASKEY: If I may have the members of the panel introduce themselves before we begin, starting at my left.

DR. ZUCKERMAN: Bran Zuckerman, Director, FDA Division of Cardiovascular Devices.

DR. AZIZ: Salim Aziz, Adult Thoracic surgeon, clinical professor of surgery at GW.

DR. KRUCOFF: Mitch Krucoff, intraventional cardiologist, Duke University and the Director of Intraventional Clinical Device Trials.

DR. SOMBERG: John Somberg, Professor of Medicine and Pharmacology, Rush University.

DR. HIRSHFELD: I'm John Hirshfeld, Professor of Medicine at the University of Pennsylvania, intraventional cardiologist.

DR. WEINBERGER: Judah Weinberger, I'm Director of Intraventional Cardiology at Columbia, New York.

DR. LINDENFELD: Joanne Lindenfeld. I'm the Director of the Heart Transplant program at the University of Colorado.

CHAIRPERSON LASKEY: Warren Laskey, intraventional cardiologist from Uniformed Services University here in Bethesda.

MS. WOOD: Geretta Wood, Executive Secretary.

DR. BAILEY: Kent Bailey. I'm a biostatistician at Mayo Clinic.

DR. TRACY: Cynthia Tracy. I'm an electra-physiologist at George Washington University.

DR. FERGUSON: Thomas Ferguson, cardia thoracic surgeon, Washington University, St. Louis.

DR. YANCY: Clyde Yancy, Professor of Medicine, Director of Heart Transplantation, UT Southwestern in Dallas.

DR. KATO: Norman Kato, cardiovascular surgeon, Encino, California.

MR. MORTON: Michael Morton, and I'm employed by Carbomedics. I'm the industry representative.


Geretta, if you could 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 18, 1999 I appoint the following individuals as voting members of the Circulatory System Devices Panel for this meeting on June the 8, 2004:

Kent R. Bailey, Ph.D; John W. Hirshfeld, M.D.; Thomas Be. Ferguson, M.D; Norman S. Kato, M.D.; Clyde Yancy, M.D.; Judah Z. Weinberger, M.D., Ph.D.; Joanne Lindenfeld, M.D.; John C. Somberg, M.D.

For the record, these individuals are special government employees and are consults 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. This is signed by Linda Cohn for Daniel G. Schultz, M.D., Acting Director, Center for Devices and Radiological Health, and dated June the 3rd.


Before we begin the open public hearing portion, I just want to read the following statement.

"Both the Food and Drug Administration and the public believe in a transparent process for information gathering and decision making. To ensure such transparency at the open public hearing sessions 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 known as 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 this 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."

I'd like to now ask the audience if there's anyone who wishes to address the panel on today's topic? If not, I'll close the open public hearing portion and invite the sponsor to make his presentation.

MS. WOOD: I would like to remind the speakers for the sponsor to introduce themselves and to state their conflict of interest before speaking.

Your presentation is scheduled for an hour, and I would ask you to limit the presentation to data that has been reviewed in the PMA by FDA.

MR. BRYDEN: Good morning.

My name is Roderick Bryden, I'm the President and the Chief Executive Officer of World Heart Corporation. I am a full time employee of the corporation, and of course it pays my income and expenses for this panel.

With me today presenting is Jal Jassawalla, whose the Executive Vice President and Chief Technical Officer of the corporation.

Dr. James Young, whose the Chairman of the Division of Medicine of the Cleveland Clinic Foundation Lerner College of Medicine of Case Western Reserve University and the Medical Director of Kaufman Center for Heart Failure.

Also with us and available to respond to questions, Dr. Phil Oyer, a Professor of Cardiovascular Surgery at Stanford University School of Medicine.

Dr. Brooks Edwards will join us. He had to remain at his clinic last evening, so he will be in sometime during the course of the morning. He's Professor of Medicine and Cardiology at the Mayo College of Medicine and Medical Director for Cardiac Transplant Team.

Dr. Peer Portner is consulting professor of Cardiothoracic surgery at Stanford University School of Medicine.

Dr. William Anderson consulting biostatistician.

I will make a brief overview of our presentation and the bulk of the presentation will be presented by Dr. Young and Mr. Jassawalla.

We are proposing an expansion in our existing bridge to transplantation indication. That expansion is highlighted on this slide that you see on the screen and in the slides before the panel.

The purpose of the expansion us to remain as a part of the bridge to transplantation indication and to ensure that patients with relevant contraindications to transplant who are expected to become transplant candidates have access to this therapy. The judgment as to whether the patient suffers from a relative contraindication and whether that patient may be expected to become a candidate is supported by a mechanical circulatory device is a judgment that would be made by the transplant center.

That patient must also, as with the current indication, be at risk of imminent death. That judgment would be made by the transplant center. That is in distinction to a destination therapy indication, which this is not. The destination therapy indication also requires the transplant center to apply judgment, the first judgment being that the patient is not eligible for cardiac transplantation and the second judgment being that the life expectancy is less than two years.

The same centers that will make the judgment with respect to becoming eligible for destination therapy would make the judgment as to whether the patient would be eligible under the expanded bridge transplant indication.

World Heart is pursuing a randomized pivotal trial now with the acronym RELIANT for the purpose of making a submission to the FDA when that data is complete for a PMA for destination therapy. And this indication is not intended to in anyway become a part of that therapy.

There is a clinical need. The class of patients with relative contraindication to transplant but an expectation of listing for transplant do not now currently have an established consistent access to the mechanical therapy which for patients at risk of imminent death is often the only therapy available. Approval of this proposed expansion in our label would provide more of these patients with a routine method of accessing this therapy.

Our presentation will deal with two key questions. First, are the data from the Novacor bridge to transplant study sufficient to support approval of the expanded indication? And secondly, if it is, would these patients be eligible today to receive the Novacor LVAS under the existing labeling and therefore there's no need for the expanded indication? There have been specific questions raised by the FDA in its reference to the panel, and we have responded to these both during the course of the presentation by Dr. Young and Mr. Jassawalla, but also at the end of that summary presentation each question is dealt with and a brief summary of our response for your record.

To deal first with the adequacy of data. There is a clear subgroup of patients within the Novacor bridge to transplant study which experienced one or more relative contraindications at the time of enrollment. In fact, 39 percent of the 225 cases experienced such relative contraindications, 61 percent did not. One might ask then how did they become a part of the trial if they had a contraindication? The answer is in two areas.

One is, as you know, the criteria for eligibility for transplant are not legislated or regulated on a national basis. The process is for each center to define its standards within a broad set of reasonable norms and then to consistently apply those standards to the patients on whom it has to make judgments. The center practices, therefore, very from center-to-center. They did during the course of enrollment, they continue to vary from center-to-center.

Secondly, the enrollment in this trial was completed in September of 1998 and during the intervening period, indeed during the period of the enrollment itself, the evolution and experience has resulted in adjustments in the criteria from time-to-time. So there is both the eventuality of the patient having presented at a center, which in some cases enrolled patients with this contraindication and the fact that over time the effects of certain contraindications have varied as experience with this therapy has been gained.

This data arises from a prospective controlled pivotal trial, and that trial was found sufficient to result in the FDA's approval of the indication label for bridge to transplantation which exists for Novacor. All aspects of that trial were established and reviewed and the approval for the indication given by the FDA in light of it, including the nature of controls which was a thorny issue at the time in light of ethical and other issues and were carefully managed and negotiated with the FDA at that time. There have been no changes made in that database, either controls the recipient group.

Today there are twice as many patient years of data available from that trial as were available from the trial in September when the original label was issued. The trial was only completed in January of 2002, and the final patient was translated after more than three years of support on the Novacor.

The data supports this indication first by comparison with the control group, which is the control that was the basis for the entire trial, in the reduction of mortality and in the improvement in the survival to transplant. But secondly, within this group while 39 percent experienced relative contraindications, 61 percent did not. If one were to attempt to structure today a prospective trial, it is probably impossible, certainly hard to imagine, that that trial could be done with the control group receiving optimal medical therapy when they are facing imminent death and optimal medical therapy demonstratively does not succeed and with the evidence that devices are successful in that regard. So a prospective trial today would have to be in some fashion done with a control group which has different indications but also receives the same therapy. That is precisely what we have. We have here 61 percent of a group who were implanted within a controlled prospective trial who did not have contraindications and 39 percent who did. And the evidence is clear that the results for those who suffered from contraindications, both with respects to their transplant rate and with their post-transplant survival was substantially the same as the results of those who were implanted with the same device and who did not relative contraindications.

The evidence also indicates that the patients with these contraindications would not today have a routine access to the device within the current indication. Dr. Young will review both the literature and a survey of ten centers in which the details of all the data points collected for these patients, which I believe are some 59, were provided to these centers to determine whether or not today these would be listed for transplant. Both within the literature and with results of that survey, almost all of the patients with relative contraindications would not be listed at some of the centers, and a majority of those with the relative contraindications would not have been listed at any of the centers.

While the criteria vary and there is no absolute standard, there are a few standards which are generally and relatively consistently applied, one of those being that the patient should meet the accepted criteria for transplantation and be ready to undergo the transplant procedure on the day of listing. Not listed in advance in anticipation of some future event with which they would be eligible, but eligible on the day that they list. These with these relative contraindications for the majority of centers and the majority of patients would not meet that test. It is not appropriate in our view to suggest that we should rely on clinics bending the rules and not living by their own criteria as the method of giving access to these devices.

Finally, we have suggested that the indication include the terms "short" or "long term" in the labeling. We have since the completion of the trial and since the granting of the PMA in September, completed the trial with a total of twice the number of patient years of experience and up to 3.4 years of support.

Secondly, the waiting times for transplant organs is highly variable.

And thirdly, the relative contraindications are somewhat unpredictable as to the time that will be required.

The intention was simply to draw attention in the label to the fact that longer term support in the context of a bridge does have demonstratively increased rates of adverse events. It is not our intention to have any other meaning than the one that is intended, and we would be quite happy to adjust that language in the event that other words would be more precise.

I would now like to proceed with a review by Mr. Jassawalla of the overview of the trial data and also of the Novacor reliability data, and then proceed directly without my return to the podium to Dr. Young who will present clinical evidence with respect to these issues that we believe clearly support the approval of this requested expansion.

Thank you very much.

Mr. Jassawalla?

MR. JASSAWALLA: Good morning. I am an officer of the company, and an employee and as such I do have a financial interest in the company.

This slide shows system configuration and anatomic placement. The pump drive unit is implanted in the upper left quadrant of the abdomen. The inflow conduit canulates the apex of the left ventricle, the outflow graft is connected to the ascending aorta. There is a single percutaneous lead that carries the power and signal wires out to the external control. The controller is designed to have two power sources connected to it at all times.

The system is totally self regulating and responsive to flow from the left ventricle and as such, there are no user controls. The recipient simply needs to manage his or her power sources.

There is extensive history with this device starting with the first clinical use in 1984 at Stanford University. There have been over 1500 implants worldwide to date. We've accumulated over 500 patient years of experience.

The current configuration, which is the N100PC was CE marked in 1993 in Europe for all indications. FDA granted bridge to transplant approval in 1998. And there have 1,077 implants through April of last year, which is the cut off for the submission.

MS. WOOD: Excuse me, sir. It's come to my attention that this was not part of the original PMA, and I would like to inform the panel this will not be part of the discussion today.

MR. JASSAWALLA: That data was in the submission, but I won't debate. We're just trying to give you the reliability in the context of this indication.

DR. ZUCKERMAN: Right. Let me clarify the comment.


DR. ZUCKERMAN: I believe that in the subsequent discussion there is reference made to a 1,077 implants and some of the data that pertains to those 1,077 implants. That data and the review of that data are not in the PMA and cannot be utilized for this important discussion today. It's just a heads up that Ms. Wood was giving you.

MR. JASSAWALLA: May I ask in which PMA you were referring to? The original 1998?

DR. ZUCKERMAN: I'm sorry. For the PMA supplement that is under discussion today, June 8, 2004.


In support of the original 1998 approval for bridge to transplant there was an in vitro life test that was conducted. These are the results of that life test. We had a dozen systems that were placed on tests under simulated clinical conditions. The goal of this test was to run the systems to failure. The mean duration was 4.2 years with a range from a little over 3 years to 5.6 years.

The demonstrated reliability using the Wible model with a 80 percent low confidence limit gives you the results that are listed in the slide. For the first year it gives us 99.9 percent reliability. For the second year 98.5. And for the third year 87.4. There were no random failures uncovered.

And in addition to the demonstrated high reliability and multi-year durability there was a single progressive non-catastrophic wear-out mode that could be monitored invasively. The clinical experience has been consistent with the in vitro life test results and there be no deaths attributed to device failure.

Long term patient experience with this system shown here within the 1,077 implants, there have been 285 that have gone over six months, 121 that have been supported for over a year and 27 and 10 for more than two and three years respectively.

The likelihood of reoperation to replace or repair the LVAS from all causes is shown in the next table. The likelihood in the first six months is 1.6 percent. The subsequent six month period of 2.1. Eleven percent in the second year and 16 percent in the third year.

There was --

MS. WOOD: Excuse me again, sir, but it's also been brought to my attention that this was not a part of the PMA supplement and therefore, this data will not be discussed today.

MR. JASSAWALLA: The next slide.

The overview results of the BTT study where the study was conducted in a NYHA class four end stage heart failure patient population at risk of imminent death were 225 patients, 190 LVAS recipients, 35 controls. FDA approved the system for bridge to transplant in September of 1998 and at study close there were 67 total patient years of support, approximately twice the total at approval, the longest duration of support being 3.4 years.

Overall results of mortality and adverse event risk decreased substantially after the post-opt period. Sixty-eight percent of LVAS recipients were transplanted. The median survival on LVAS of about 11 months. Thirty-seven percent of the controls were transplanted with a median survival of less than half a month. The LVAS support provided a seven fold reduction in mortality of risk with a very significant p-value of .0001.

I'll now turn over to Dr. Young to talk about the specific study results and transplant listing practices.

DR. YOUNG: Thank you, Jal.

Mr. Chairman, ladies and gentlemen of the FDA and the panel, my name is Jim Young. And for those of you that don't know me, I am at the Cleveland Clinic Foundation. Today I am acting as a consultant to World Heart. I do not own any stock or equity in that corporation. They are paying me for my activities of advice given for clinical trial design and also data analysis.

And what I am going to be talking about specifically is the data analysis in the Novacor bridge to transplant study. And by way of introduction I would like to, again those of you that don't know me, point out that I am a heart failure/heart transplant cardiologist and have been involved in this arena, unfortunately to say, for over two decades now. And one of the fascinating things is the increase in our knowledge base with regard to how we can treat patients with advanced end stage heart failure. And, in fact, the roles of heart transplantation ventricular assist device therapy and mechanical support in general. And much of the issues driving why this retrospective analysis of a prospectively controlled trial was done lies in those efforts, efforts to clarify how we should be selecting patients for transplantation, selecting patients for ventricular assist device therapies and tail in well with the ongoing discussions, I won't say debate but ongoing discussions regarding which patients should be listed for heart transplant and when.

The bridge to transplant study, as has been alluded to, actually has patients in them which consensus would agree have no relative contraindications to cardiac transplantation. As we have gained knowledge over time and reviewed a variety of other databases, however, we note now that many relative contraindications to patients receiving heart transplant at a time when a ventricular assist device is placed actually do exist. And this has created contention and debate in some circles, much discussion in other circles.

We in this database had the ability to cone down on listed here seven specific relative contradictions which robust information was available in and which, as I will show you, fit the type of relative contraindications that have emerged from other databases and the focus of discussion when we select patients for listing for cardiac transplantation at review board meetings or, indeed as we do it in the state of Ohio, at a panel review by the Ohio Solid Organ Transplant Consortium.

Status II now makes up patients with one or more of these relative contradictions at enrollment into the Novacor BTT study. And these specific contradictions were creatinine greater than 2?, pulmonary systolic pressures over 60 mmHg, pulmonary vascular resistance higher than 6 Wood units, total bilirubin greater than 5 milligrams, obesity or body mass index that would suggest excessive ponderosity at 32 cubic grams per meter squared or cachexia at 19 kilograms per meter squared. And then age, which still is a terribly contentious issue at 66 years.

Could I have the next slide?

Now, those were picked based on several different things. One, literature review including rather extensive analysis of data in the cardiac transplant research database group, which now has very compulsive information on over 7,000 patients followed for ten years as well as the advice of experience clinicians, heart failure and transplant clinicians, both cardiologists and cardiac surgeons.

And these were the most clinically relevant of the 59 variables that were in the database which could be analyzed. Now, this doesn't include all of the variables that may create relative contraindications to cardiac transplantation. And three of the most vexing ones that we have to deal with on a day-to-day basis are presence of allosensitization in a patient with end stage failure who may or may not be a transplant candidate but who is coming to the decision about needing mechanical circulatory support.

Presence of a malignancy, for example. This creates a very vexing problem if you've had a patient who is only out two years from a successful resection of a breast malignancy and there's questions about those patients.

As well as other examples, and in Ohio perhaps one of the more frequent ones we see is with the state review whether or not a patient might have psychosocial issues that create relative contraindications; cigarette smoking, past history of drug abuse or whatnot. And a requirement has been raised about putting the patient through psychosocial counseling for a period of time.

Those are just examples of some of additional variables that would not be in the BTT study that could be looked at.

Next slide.

Now, these thresholds that we choose are consistent with many current consensus guidelines for transplant listing and the literature. And, indeed, if you look at a few examples here's one from Blue Cross/Blue Shield which excludes patients with pulmonary systolic pressures greater than 60, TPGs greater than 4 Woods units, obesity as they define 150 percent of ideal body weight, which would translate somewhere into the range of a BMI of about 32. And this is on vas therapies, as well. So we believe that these relative contradictions that have been listed are consistent with clinical practice and data.

Next slide.

And, in fact, if you look at the references that have focused on these relative contradictions going back to the conference led by Les Miller back in 1998 when we first really began looking at how we pick patients and list them for cardiac transplantation and focus on this question of if you list a patient, ipso facto an organ becomes available, you should accept that organ for the individual patient. And then vetting the contentious issues of those relative contradictions has been expanded into several other efforts, including the one I just alluded to the most recent publication by Jim Kirklin of over 7,000 patients in the CTRD database.

And so there's robust information in the literature about this relative contradictions and how they contribute to excessive risk post-transplantation and then begins to introduce the concept of an individual that has one or more of these relative contraindications the rationale of implanting a ventricular assist system to support hemodynamics with the expectation that these complications will in fact resolve and make the patient a better heart transplant patient.

Next slide, please.

And so group two was this group of individuals that came out of the BTT study. And overall in the BTT study now, 225 patients were in that study. And if you looked at group one, those are the individuals that were receiving LVAS therapy for hemodynamic indications that didn't have these tangible relative contraindications that were listed. In the control group, the historical control group for the BTT study, there were also individuals in whom we applied the same criteria, and you can see 23 of those original 35 control patients had no contraindications with 12 having one or more of the contraindications similar to the individuals that eventually received a VAD but had contraindications present at that time.

Next slide.

Now, if you look at the control group, this is an important element but to me perhaps not the real issue, as I'll get into in a minute. But the control group was the basis for the BTT approval and as was alluded to after a lot of discussion. And we did use the same criteria for picking in the control group as in the test group, the LVAS group, the relative contraindications.

And interestingly enough if you looked at multivaried analysis correction for the multiple covariates that are involved in the control group and the treatment group, there remained a highly significant difference as those of us clinicians might have expected.

Next slide.

Now, interesting to me was the number of relative contraindications that appears and what they were. And here you see the contraindications listed. For me the two particularly vexing difficulties when you're at the bedside looking at these patients and trying to make a decision about going forward with listing for transplant is: (1 renal insufficiency as marked by a serum creatinine of 2? or greater and; (2) pulmonary hypertension particularly after a lot of therapies have been given to try to optimize it and so pulmonary pressures at their best that are greater than 60 millimeters of mercury. And so two of the top three relative contraindications were in fact related to that.

We also have a troublesome time with many ponderous patients that come in to see us, and that represented the top number in the LVAS group. And this is the group that I think is most important. Interestingly enough, congestive hepatopathy from heart failure and even age wasn't as much of an issue in this data set.

Next slide.

Now, the important observation to me really is here on this slide, and it's the fact that even though by many different other analysis, database analysis, the relative contraindications lead to worse outcome after transplantation, when you look at those individuals who received a ventricular assist device in group one versus group two individuals that did not have in group one and individuals that did have one or more relative contraindication, you can see here that there was no statistically significant difference when those two groups were compared censoring the patients at transplantation, obviously sense that was the goal.

So it suggests that you can use left ventricular assist device therapy even in patients with these relative contraindications to support them to transplantation. And even we have data demonstrating that during the period of support, many of these relative contraindications abate or at least improve.

Also important is when you do compare it to the control groups in group and group two, again there was a highly statistically significant advantage in both group one and group two of receiving the left ventricular assist device. And, again, I think this is consistent with many other analyses, anecdotal experience and also longer term destination therapy controlled clinical trials with these patients that are being evaluated and being looked at with transplantation as an end point, as BTT looked at, are in fact quite ill.

Next slide.

Now to me also important is the rate of transplantation that occurs in this group. Because this an end point that gets to whether or not hemodynamically we are being able to resuscitate these group two patients to get them to a level, particularly with pulmonary hypertension and renal insufficiency, to get them to a level where the clinician is comfortable accepting a heart when it's offered.

What happens in reality right now is many patients who resemble individuals in group two would be listed for a heart transplant, yet when the VAD was put in would be immediately made status 7. Now, for those of you that may not know the UNOS allocation schemes, there are really three active schemes. Status 1, status 2 and status 7. A status 2 patient is an outpatient, sort of the walking wounded not requiring a lot of intensive therapies that is on the waiting list and now represents only about a third of the transplants done in the United States. A status 1 patient is an individual either in the hospital on intravenous medications with invasive lines present on hemodynamic support with mechanical circulatory sustenance.

And the practice in these types of patients would either be to make a patient status 7 or, if you left them status 1 or status 2, turn down parts listing the turn down reason being the patient "too ill" generally that's the category that would be tipped.

So what this observation demonstrates here is if you compare the patients with and without these relative contraindications, the rate of transplantation was similar statistically in these two groups, and actually pretty good from a clinicians assessment in those patients. And, of course, compared to the controlled group the rate of plantation was much, much higher again suggesting the hemodynamic support is pulling these patients back from the brink much more readily.

Next slide, please.

And then another issue which a lot of us have been interested in is what are these relative contraindications either with or without left ventricular assist device therapy mean to the post-transplant survival group. Again, as I alluded to, having renal insufficiency, having pulmonary hypertension is independent risk factors by themselves for higher adverse event rates, post-transplantation. What you can see here, again comparing the patients with relative contraindications to the patients with no contraindications is no significant difference in survival rates once transplant has occurred. And, in fact, numerically over the follow-up period survival was actually better in this group of patients.

Next slide.

So we believe that this is a unique dataset and, in some sense, this may be a fortuitous at the time not know perspective comparison of the group that creates problems for us today and the group that we were challenged to try to help. And I believe that we have demonstrated improved survival while awaiting transplantation even when relative contraindications exist and compared to control patients on a ventricular assist device therapy, I believe the dataset, again, clearly shows as it did before an improved rate of transplantation. And very important is that survival is similar in the patients with relative contraindications, those group two patients, to those without those group one patients.

And the link, I think, is this fact: There is resolution of specific conditions present and we can present data specifically regarding what happens with those seven parameters which created the relative contraindication. And I alluded to the fact that pulmonary artery pressure does fall, creatinines do improve, renal function does get better as two specific examples.

Next slide.

MS. WOOD: Excuse me, Dr. Young. But it's been brought to my attention that the information in the previous slide was not submitted to the FDA for review as part of this PMA supplement, and we will not be discussing that information today.

DR. YOUNG: Thank you.

Now, if you look at the analysis specific to these risk factors and relative contraindications and try to determine what the impact of ventricular assist device therapy is, the reduction in mortality risk is virtually unchanged when considering any of the differences in individual patient covariates. And importantly, the analysis suggests that there's a six fold reduction in mortality risk in these patients with relative contraindications. And to a clinician what this means is that we can make decisions in a patient that today if we listed him, would not clearly be a heart transplant candidate, we can make decisions to place a ventricular assist device with the expectation that those contraindications will in fact be treated and resolved. And that is the purpose of the request on the labeling.

Next slide.

Now, another interesting thing to me is that these patients in group two here in the yellow are, in fact, more ill going into these operations and during the observation period one might expect complications which are nemeses in taking care of ventricular assist device patients were seen more frequently during the post-VAD implantation follow-up period. And this data is set up so that group one patients are normalized here at 100 percent and then the relative increase in complications observed here in the group two patients are noted. And you can see that though numerically more frequent, these are 95 percent confidence intervals. There's extremely wide confidence intervals. And in the end there were no statistically significant differences between the two groups, though numerically some of the problems were more significant in the group two patients. Again, my interpretation suggesting that you can get these patients through, you can improve them and that ultimately they can be transplanted.

Next slide.

Now one of the things we also did was went back and looked at now ten, this was nine centers. We do have data now from Columbia which is not going to be in this presentation as it came it late. But these nine centers were picked because they were participants in the BTT trial and they representative of the wide spectrum of centers that were in the study. We had both very large centers, the Cleveland Clinic doing 75 to 80 transplants a year --

MS. WOOD: Excuse me, Dr. Young.

It's also been brought to my attention that this data as not part of the PMA supplement submitted for review and will not be part of the discussion today as well as the next slide.

DR. YOUNG: Okay.

The query here, however, that will be discussed demonstrated agreement, basically -- can I have the next slide -- with the relative contraindications that we had identified. And so though it won't be discussed, 83 of the 87 patients by the panel -- and this was done in a blinded fashion, agreed that there was one or more contraindications for transplanting on the day of listing.

Next slide.

So really what this boils down to is the challenge as clinicians that we have with respect to the end stage heart failure patient population which fortuitously was represented quite nicely in the BTT study here. You have those patients that clearly are not a transplant candidate nor will they ever be, and we can give many examples of those patients. Perhaps an older patient with a malignancy that's been resected and allosensitization identified pre-transplantation, they're just not going to be a transplant candidate. Perhaps they would be a candidate for destination therapy. Those are not the types of patients that were in the BTT trial, nor where they the types of patients that in our retrospective analysis made up those individuals with the relative contraindications. We're really focusing on this part here, which are individuals who fall outside the limits of transplant criteria but who may be expected to become transplant patients who then are in this status 7 or this repetitive turndown mode.

Next slide.

And also we know that transplant listing practice varies. And, indeed, this was seen in the BTT study. This is a commonly discussed issue amongst programs, and there may be some centers who are willing to list a patient who have relative contraindications particularly if a ventricular assist device is going to be placed, to move them into a category where if an organ became available, it would be utilized. And, of course, driving this has been primarily changes in practice over our time and our understanding of transplantation and ventricular assist device therapies and a lot of that has been done with the retrospective analysis of existing database. Transplantation and ventricular assist device utilization is a boutique science and unfortunately we don't have the ability to be able to do very large scale randomized clinical trials. So we know this from the UNOS databases, the CTRD databases which are readily available.

Next slide.

And so the issue of the clinical need for an indication of this sort I believe is present. And I believe when you look at the BTT study you can see patients that early on were placed into this study and went through and did demonstrate that you could rehabilitate them and get them to transplantation.

Next slide.

And so if you look at the BTT group, which was a perspective and it was a controlled clinical trial and subjected to the retrospective analysis addressing the questions that have risen more contemporaneously, we believe that the data is there to support the request. There is a six-fold reduction in mortality in this particular targeted population. This would be a population that would be problematic to randomized, given all of the information about high mortality rates and, indeed, as characterized by the control group in this particular trial although there were some problems with that.

I think that a prospective study where you randomized these sort of patients would be impractical. I would be very loathe to enter into that, particularly because of the ethical issue. And I don't think that other tactics can be used to clarify the issue. I believe that the analysis of this particular database has done that. And, obviously, databases are continuing, post-marketing surveillance is continuing. There's large registries that are being developed, responses to the NIH RFA has been made.

Next slide.

And so to summarize, I believe that a population does exist with patients with relative contraindications to transplant. And in those individuals there's no assured access to ventricular assist device and transplant therapy even if the clinician suspects that many of these parameters could resolve with ventricular assist device support. And, indeed, if you do that retrospective look at the bridge to transplant study, there were a significant number of patients in that trial that met this contemporaneous characterization. And I believe that we have demonstrated that there is effectiveness as a bridge to transplant with the LVAD survival being greater than control and perhaps even more important to me, the survival in group two patients, those with relative contraindications supported by the ventricular assist device being similar to group one patients, those without.

Next slide.

And so if you look at the benefits versus the risks, approving an expanded indication would provide more uniform access to that therapy in these patients with the demonstrated survival benefit. And those bad recipients who survive transplant do have the opportunity to benefit from really the gold standard therapy for really the terminally ill bad hemodynamic patient. And that in this analysis those patients with relative risk did have a six-fold reduction in mortality. And I personally believe that what this would lead to is more consensus regarding listing and ultimately utilization of scarce donor organs.

I think that was my final slide. Next slide. Oh, this is the final slide.

In response to the questions that have come out, I do think that we have addressed and answered them. Many of the other slide summaries are in the handout, but I do believe that the core data analysis that is here from the BTT study does in fact justify the expanded indication and the labeling rewording that has been requested.

Thank you.

MR. BRYDEN: Thank you very much. That concludes the presentation from the sponsor. And we would be pleased to take any questions now, and also of course there is the time following the FDA's presentation when we look forward to responding to question from the panel.

Thank you.

CHAIRPERSON LASKEY: So, panel members, are there any questions for an of the three presenters this morning?

DR. FERGUSON: Warren, I have a question. I'm sorry, I'm sure it's in the material, but I couldn't find it.

When you talk about the group two patients, they could have relative contraindication or seven, each individual. Do you have any data breakdown of that because the lumping seems to me to be a bit unfair, perhaps, unless you do that?

DR. YOUNG: Well, it turns out that the majority had one relative contraindication; 17 of 87 subjects had two of them and 2 of 87 had three.

DR. FERGUSON: Right. Exactly.

DR. YOUNG: So the numbers in those that had multiple contraindications were 17 with two and two with three.

DR. BAILEY: I wonder if you could just summarize the recruitment into the original BTT study; that is how were patients recruited into that prospective study, in particular the control patients versus the LVAS patients? I guess I'm interested in the timing, sort of the time flow of that.

MR. JASSAWALLA: Yes. The actual enrollment, the bridge to transplant study, the one that resulted in the approval, the LVAS patients were recruited between March of 1996 and September of 1998. Prior to that we had another study that was ongoing with the FDA. And because of the problem and ethical considerations in enrolling patients, the FDA permitted us to use the controls because the inclusion and exclusion criteria were essentially the same from the previous study when we finalized on the study that started enrollment in March of 1996.

DR. BAILEY: Okay. Then the control patients were not actually concurrently recruited?

MR. JASSAWALLA: Some of them were, but several of them were from an era slightly ahead of when the pumps were implanted.

DR. BAILEY: And what were the selection criteria for choosing controls retrospectively then?

MR. JASSAWALLA: Many of them came from centers that were in training for the LVAS, and we had a rule where we skipped no controls. That once a center was going to provide controls, they went back to their database and looked at the study exclusion criteria and then uniformly included them as controls.

DR. BAILEY: But what is the starting date? When they're first listed, is that the starting date for follow-up, when they're first listed on the transplant list?


DR. BAILEY: Okay. But then the controls would be defined by ones who are listed and who didn't receive a device?

MR. JASSAWALLA: Correct. They would have met inclusion criteria at that point.

DR. BAILEY: And would the fact that they were not going to receive the device be known at the time they were listed, that is they had already refused or were already at a center that didn't have a device available?

MR. JASSAWALLA: For the prospective controls, they would have refused a device or a device may not have been available. A surgeon may have been out of town. For the retrospective ones, there were those that would have met inclusion but the study hadn't started in terms of a trained group of people.

DR. BAILEY: Okay. Their status as far as not being able to receive the device would be known at the time of listing? You'd know prospectively?


DR. BAILEY: In other words, if they had been in a perspective study --


DR. BAILEY: -- they would have known at that time that they weren't going to get the device?


DR. LINDENFELD: The median survival of your controls in group two is 7 days? I guess when we're comparing the control group here to the device group, I mean the median survival of the medical group in REMATCH was 105 days. Seven days seems awfully short, and you have to wonder that is a pretty short median survival. These patients, this control group doesn't seem to me like the average control group listed for transplantation. I mean, the median survival 7 days after listing, it's hard for me to compare to the LVAS group because it doesn't seem like a reasonable comparison.

DR. YOUNG: Yes. First of all, I think we need to be careful when we look at the control of REMATCH. I hesitatingly mention it because they are different patients. They are not transplant candidates, number one. And number two I think their median survival is higher because those patients at risk of imminent death wouldn't get into REMATCH. At least I know for a fact they did not get into PREMATCH, which I was involved in.

Now, I agree completely. These are patients that are very ill and that when you list them, the median survival of a week after that characterizes them as a group that if you had a VAD available and were capable of putting it in, probably would have ended up getting a ventricular assist device. So I think the fact that they only have this short survival reflects the fact that these are very ill patients and yet matched well with the patients in the BTT that ended up getting a ventricular assist device.

DR. LINDENFELD: And how many of the control group and the device group where on amyotrophic therapy at the time at the initial time of either device placement or listing for transplant? It looks to me like in the briefing book it's around 12 percent?

DR. YOUNG: There's a backup slide that I think has got the -- 100 percent; that's the answer to that. And there's a backup --

DR. LINDENFELD: In both groups were on amyotrophic therapy, the control and the device group?

DR. YOUNG: Yes. Yes. Yes. And again because of the era that BTT was done, these were I think very ill individuals and the options, obviously, were not great.

DR. LINDENFELD: And then do we have data about the reversibility of the pulmonary hypertension? I think in most places there's some evaluation of the reversibility. Do we have any data for that?

DR. YOUNG: Yes. There is data both in the original PMA submission for the entire group about what happens to pulmonary hypertension and we also have data in individuals who have pulmonary hypertension is the reason for their relative contraindication. And both of them show reduction in systolic pulmonary pressures and also transplumonary gradients.

If you look up there --

CHAIRPERSON LASKEY: Dr. Young, is that pre-transplant or post-transplant?

DR. YOUNG: Pre-transplant on the ventricular assist device.

CHAIRPERSON LASKEY: So they're empirical studies?

DR. YOUNG: Yes. Yes.

DR. LINDENFELD: No, I'm sorry. I'm talking about reversibility at the time of listing or device placement.

DR. YOUNG: The reversibility would have been demonstrated after device placement. These are patients that would have been already on drugs and drips.

DR. LINDENFELD: I'm sorry I'm not being clear. What I wanted to know is if PA pressure is greater 60, most of the time the way that's evaluated is to see if patients respond to therapy and if those drop, and do we know what happened to those pulmonary pressures and the pulmonary resistance prior to either LVAD placement or listing? In other words, were they reversible prior to that?

DR. YOUNG: I can't tell you specifically what was done, but I can tell you that the PA pressures were listed as greater than 60 on optimal therapy. So I can only assume that the cardiologist and the surgeons have been trying to lower them with the best means that were available. And again, anecdotally from my experience, this is a problematic issue, probably more so in Denver at 6,000 feet. But I can't tell you specifically what was done.

What I can show you here is what happened over the period of time where hemodynamic assessments were available in the individuals who received LVAD therapy whether or not they were transplanted or not. And you can see that those patients with pressures above 60, LVAD therapy invariably dropped those systolic pressures. So one assumes that as the individuals were treated, pressures remained at 60. It was the addition of the ventricular assist device that effected that decrease in pulmonary hypertension.


DR. HIRSHFELD: I'd like to ask, do you have a slide comparable to the one you just showed for pulmonary vascular resistance? I didn't see that data in the briefing book.

DR. YOUNG: I believe that we do. We don't have it for pulmonary vascular resistance. No late wedge pressure measurements in the database.

DR. HIRSHFELD: So you don't know --

DR. YOUNG: I can tell you why that is.


DR. YOUNG: In the ICUs with the Swangantz in place, most of our surgical colleagues don't like us blowing the balloon up.

DR. HIRSHFELD: Because I think interpreting a drop in pulmonary pressure by itself may just reflect dropping left ventricular pressures and not reflect reversal of pulmonary arterial or constriction.

DR. YOUNG: Sure. And that's the one major challenge that we have at the bedside trying to figure out the patient. But I can tell you that if the blood pressure goes up, flows go up and the pulmonary artery systolics drop, we're at least much more comfortable that the patient will do okay.


DR. KRUCOFF: In the white paper section 5A of our panel pack and 5B, you all have several multi-variable models considering some of these features. Is anybody actually going to discuss those models?

MR. BRYDEN: We would be happy to discuss those. We could do it now or in the afternoon when there may be a bit more time. Which would be your preference?

CHAIRPERSON LASKEY: Well, we're actually doing well time wise. It may be somewhat more involved. If you want to save it for either your turn this afternoon --

DR. KRUCOFF: I actually just wondered if anybody was going to talk about them.


DR. YOUNG: We will, I'm sure, that given the opportunity we will present that fully this afternoon.


Dr. Aziz?

DR. AZIZ: What percent of your VAD implant patients have had prior heart surgery?

DR. YOUNG: Oh, it's a good question and I can't give you that off the top of my head.

DR. BAILEY: Do you have any analyses that don't don't -- just looking at overall survival including a post-transplant, just as all one endpoint?

DR. YOUNG: Not that I've personally looked at.

DR. BAILEY: I guess the concern is just that, you know, I guess if one group gets earlier transplants, than you're sort of looking at a different point on the curve than starting from time zero when they're listed. So, you could imagine if one group got a differential rate of getting transplants, and I guess to do that you'd have to look at the transplant endpoint, censoring it and see if practice patterns are the same in the two groups. But, you know, it sort of influences the interpretation of survival post-transplant.

MR. BRYDEN: Yes. For your efficiency, we will pull it out so it's efficiently delivered. But that data is available.

CHAIRPERSON LASKEY: Dr. Yancy, you had a question?

DR. YANCY: Thanks, Warren.

I'd like to go back to an issue that Dr. Lindenfeld raised because I think that the strength or weakness of the presentation really hinges on the aberrant outcomes in the control group. And I'd like to know beyond the inotrope question what other clinical characteristics you can share with us regarding that control group. Because I think a critical issue is whether or not the same mortality expectation would exist under a contemporary practice model, so I think it's important to know a little bit more about that comparison group.

DR. YOUNG: Well, what I can tell you is this was the characteristics of the overall group. And I can also tell you that the parameters were not significantly different between the two groups. I don't know that I actually have the control group listed out. But I think perhaps the most telling tale, again, is the low index, low systemic pressure. And to me the most important characteristics were that over 70 percent were supported with either intra-aortic balloon pump or some other mechanical circulatory support, whatever, and that 13 pumps had had cardiac arrest or cardiopulmonary resuscitation within 48 hours before enrollment. So the whole BTT group was in fact extraordinarily ill.

DR. LINDENFELD: Is it possible to see the groups with IABP and post-cardiac arrest? I mean, what percentage of each of those two groups were post-cardiac arrest and had ballooning?

DR. YOUNG: Yes, we can get that for you.

MR. BRYDEN: The datapoints, of course, were collected on both control and the implant group. And with those specific questions we can query the database and probably by this afternoon have pretty accurate answers on those questions.

CHAIRPERSON LASKEY: All right. Well, before we proceed with the FDA's presentation, I just one I guess overall philosophic question. Your BTT study looked at survival post-transplant as the primary endpoint and survival on LVAS as your secondary endpoint. Now you're looking at survival on LVAS as a primary endpoint in the same study, just switching your primary and secondary endpoints. Can you just tell us what sort of concerns were addressed up front when you do this kind of relook?

DR. YOUNG: It's a very fair point, and I did I hope openly and fairly point out this is a retrospective analyses with all of the inherent problems and weakness of that sort of database. That being said, that's one of the ways in transplantation and ventricular assist device therapy that has guided us down the road.

The questions shifted over time, and the questions initially were could you get a patient to transplant and then would in fact the post-transplant outcomes be reasonable. So post-plantation became the focus of attention. So that's why I think in the original perspective study that was the relevant issue. Because again, as a clinical it's more important to get the device out, the transplant done and have good outcomes.

Now the question has really shifting driven by the timing of listing a patient for transplant vis--vis these relative contraindications and what are our tools available to try to repair things like pulmonary hypertension and renal insufficiency and some of these other relative contraindications listed. So the attention has shifted to what the device can do before transplantation. And I think that's the best explanation that I can give for the reason that we're looking at a different outcome.

You guys want to --

MR. BRYDEN: I'd like to just comment briefly. The summary data that was reviewed and then the more detailed data that is available was reviewed first from the standpoint of survival post-transplant, which was the endpoint in the primary study. And, in fact, the survival post-transplant in the group two, those suffering contraindications was not statistically significantly better, but slightly better than the survival post-transplant by those who did not have the relative contraindications. And the survival to transplant, again, was not statistically significantly different but slightly less favorable for those who had relative contraindications than for those who did not.

So while the presentation may have been in a different order, the precision of the data collected and the judgments made, it was not intended to be adjusted.

CHAIRPERSON LASKEY: Well, I'll leave that to the statistical folks to hash out. I understand the clinical science of it, but when you do a study and you assign primary and secondary endpoints on the one hand, you assign different levels of confidence in your results. And then when you switch a secondary endpoint and make it a primary endpoint and raise the level of confidence required for those results, it just raises some questions particularly with respect to the nature of the control group.

All right. Well, we're doing very well for time, so let's proceed if we may -- do we need a break? It's so early. Okay. Well, I've been told we'll have a 10 minutes. I have 20 after 9:00. Let's regroup at 9:30. We have a real shot at getting done early today, so 9:30 it will be. Thank you.

(Whereupon, at 9:20 a.m. a recess until 9:37 a.m.)

CHAIRPERSON LASKEY: Okay. Before we begin, Ms. Wood wanted to read one statement?

MS. WOOD: Yes. I would like to correct one of my previous comments. The reliability of 1,077 patients on slide 18 of the sponsor's presentation was in fact submitted in the PMA supplement on August of 2003, however it was not included in the panel packs to the members of this panel.


And I'd like to invite the presenters from the FDA to do their thing.

DR. BERMAN: Good morning.

For the record, the matter before the panel today is a proposal or a request by World Heart Corporation for an expanded indication for use for their model N100PC and N100PC(q) left ventricular assist system, and the information was provided in supplement -- in amended 7 to supplement 4 of P980012.

I will present the FDA review summary for this file.

My name is Mike Berman. I am the lead reviewer for this file.

The FDA review team consisted of Dr. Ahn, who is a FDA statistician, myself as the lead reviewer, Dr. Ileana Pina who is a heart failure cardiologist. She is a consultant to the FDA. And Dr. Julie Swain, she is a cardiac surgeon and is a consultant tot he FDA.

So that we can focus, this the part of the sponsor's request, this is the proposed language for their expanded indication for use. The language in yellow is the language they currently had. "The LVAS is intended for use as a bridge to transplantation in cardiac transplant candidates at risk of imminent death from nonreversible left ventricular failure. The LVAS is indicated for use both inside and outside of the hospital." That's approved currently.

The sponsor wants to add language so that the indication for use will now read: "The LVAS is intended for use as a short or long term bridge to transplantation in cardiac transplant candidates and in patients with relative contraindications to transplantation who are expected to become transplant candidates with mechanical circulatory support at risk of imminent death" etcetera.

The term "indication for use" has a specific regulatory meaning, and this is what it is. This is a quote from the regulations within which the FDA must operate. "An indication for use has to include a general description of the disease or the condition that the device will diagnose or treat and it must include a description of the patient population for which the device is intended."

So, did the sponsor's request include a description of the disease? Yes, it did; it's end stage heart failure, nonreversible LV failure, risk of imminent death.

Did they include a description of the patient population? Well, they still have candidates for cardiac transplant, but now they want to add those with relative contraindications to transplant who are expected to become transplant candidates with mechanical circulatory support. That matter will be addressed further by Dr. Pina in her review.

So as part of our review process and our decision process we must determine whether there is a reasonable assurance of safety and effectiveness for this device used for this purpose and safety and effectiveness are determined under our law. We have to determine it with respect to the patients for whose use the device is intended. In this case it would be patients with relative contraindication who are expected to become, and that will be addressed further by Dr. Pina.

As well we have to take into account the conditions of use which are prescribed, recommended or suggested in the label, and we have to assess probable benefit versus probable injury.

As a reminder, this is a description of the device. The implanted components consist of an encapsulated blood pump which is placed sub-diaphragmically. There are two valved conduits and part of the percutaneous tube which connects the pump to the outside world is implanted. There are external components. There is a controller. There are battery packs, there are other power sources and there are various accessories. This is from the operator's manual.

Again, you can see the blood pump, the inflow to the pump comes from the apex of the left ventricle. It is conveyed to the pump by a conduit or a valved conduit. The blood is pumped out of the pump, again, through a valve conduit into the ascending aorta. A percutaneous tube connects the pump to the outside, to the controller. There's also a capability, a vent capability which allows the venting of the interior space of the pump to the outside. And there are battery packs shown in this view and it doesn't show the other components.

If this device system were being presented de novo for any purpose, that is if we had not seen it before, we would assess multiple characteristics of the device system determined from bench testing. For example, we would examine manufacturing processes for the device system. We would be concerned about the sterilization process, how the system would be packaged and shipped, whether the materials out of which the device was manufactured were biocompatible. We would examine device software, mechanical safety, electrical safety, electromagnetic compatibility and so on.

None of these matters are at issue in today's discussion. There are no aspects of the items listed here which are in question. They have all been examined in detail previously, mostly as part of the bridge to transplant indication, and the FDA has judged them to be adequate for bridge to cardiac transplantation and we see no concerns with any of these items for the expanded indication that's being proposed.

However, the FDA does have remaining concerns regarding the expanded indication for use, and they can be divided into clinical concerns and statistical concerns.

We're concerned that the patients evaluated are not the same as those patients for whom the device will be indicated.

We have some concerns about the term "long" or "short term" about the meaning of relative contraindication and about this idea that these patients are expected to become transplant candidates.

As far as the statistical analyses, we are concerned that subgroup analyses as presented can be extended to the intended patient population and whether or not the selected treatment and control groups are, in fact, comparable.

As far as the clinical concerns, we note that the patients in the analyses data subset which was drawn from the bridge to transplant trial were transplant eligible and patients for the expanded indication for use may not be.

As far as "short" or "long term" goes, out of 160 of the 190 LVAS patients in the BTT trial were on device six months or less, and that's from tab 5A figure 4-1. Of the 30 LVAS patients who were six months or more on device, 15 of them were at a year or more and only four went out for two years or more. So we're concerned about that in terms of use of the term "long term."

Relative contraindication, it's not clear to us from the data provided why those specific seven parameters were chosen and why the specific thresholds were chosen.

And the term "expected to become" is problematic because the patients who were analyzed in the subset analyses were all transplant patients.

And we do not see in this submission objective evidence of reversal or normalization of the relative contraindications or of an improvement of the opportunity for transplant, nor do we see any objective evidence that one can determine a priori whether a particular patient with relevant contraindication is likely to become a transplant candidate on device.

The statistical concerns we think, and Dr. Ahn will address this in more detail, that the subgroup analyses may not be extendable to the intended patient population. And there is concern about the comparability of the selected patient subgroups because covariates are not matched.

The FDA presentation will now continue. Dr. Ahn will discuss the statistical aspects of the submission.

DR. AHN: Good morning. I'm Chul Ahn, the FDA statistician for this file.

The BTT study was submitted in World Heart original PMA application for use of the LVAS device system for bridge to cardiac transplantation. This PMA was approved in 1998.

BTT trial was a two arm nonrandomized study based on 225 patients. Either the patient received the Novacor LVAS or was treated with medical management.

Among the 225 patients, 190 patients were in the treatment group and 35 patients in the control group.

The primary endpoint of this study was survival to 30 days post-transplant. The PMA supplement under consideration today use as a subset of data drawn from the sponsor's original PMA application. The sponsor proposes to expand the current indication for use to include patients with so called relative contraindication. The sponsor use seven criteria to choose a patient with relative contraindication.

They retrospectively identify 87 such patients out of 225. Among 87 patients there were 75 LVAS patents and 12 control patients. The sponsor argues that information derived from these selected 87 patients supports their proposal for an expanded indication for use.

The sponsors proposed expanded indication for use implies that patients with relative contraindication who are not eligible for transplant will become transplant eligible with LVAS support.

So the device's intended patient population is those with relative contraindication who are expected to become transplant candidates. Notice that this intended patient population is different from BTT population.

Panel members, I want to look at screen since there will be an animation.

This red circle shows the 225 patients from the BTT study. They were a sample from BTT population. The cloud shows the intended expanded patient population. It is different from BTT population. There is no overlapping between two populations.

The sponsor found the subgroup with relative contraindication from BTT sample. There were 87 patients; 75 from the treatment and 12 from the control.

Did the patient with relative contraindication improve their opportunity for transplantation with LVAS support? We know that these 87 patients were transplant eligible when they were entered into the BTT study. However, even if we accept that these 87 patients were not transplant eligible, there is no evidence to demonstrate that these 75 LVAS patients became transplant eligible while on support. The sponsor provided baseline patient characteristics but they did not provide the outcome data for the seven relative contraindication criteria that could potentially result.

Therefore, the data does not directly address intended patient population. We don't know how effective the device will be for the intended patient population. Let's revisit the previous graph once more.

The cloud population is intended patient population for the proposed expanded indication for use. They are the patients with relative contraindications who are expected to become transplant candidates with mechanical circulatory support. But the data we have is for those 87 patients from the red circle with relative contraindications who were transplant candidates when they received the LVAS and we don't know whether they result or not. Therefore, it is problematic whether the result from these 87 patients can be extended to the intended patient population.

After the sponsor identify 87 patients as those with relative contraindication, they compared the survivor curves between 75 LVAS and 12 controls. However, the question is whether these two treatment groups are comparable. We will look at three items to examine this.

They are the year implants, an example of especially covariate and baseline covariate in general and propensity scores. This graph shows the distribution of patients over the years when the device was implanted. For the control group the year of implant refers to the year when they were enrolled into the study.

In the graph blue is for the treatment group and red for the control group. Note that most of the control patients, nine out of 12, were enrolled in the first half of the 1990s while all of the LVAS patients were involved in the last half of the 1990s. There has been a drastic change in medical management during the last ten years, therefore the year of implant is a very important covariate. However, as you can see from this graph there is very little overlap in the time of enrollment between the two treatment groups.

Now I would like to show you that there are large differences in several other baseline covariates between the two treatment groups.

This is the list covariates which shows statistically significantly difference between the two treatment groups with p-value less than 10 percent. They are sorted by the magnitude of p-value. They are Milrinone, pre and antihypertension, age, history of transient TIA, dosages of Dobutamine, creatinine level and bleeding.

So far we have seen that these two treatment groups are not comparable due to imbalance of the year of implants and imbalance in multiple baseline covariates. And any direct treatment comparisons on effectiveness endpoint are problematic. Also, all p-values from direct treatment comparisons are not interpretable.

What about treatment comparisons adjusting for imbalanced covariates? We may consider two analyses methods. They are traditional covariate analysis and propensity score analysis.

For the traditional covariate analysis let's consider an example of adjustment for one covariate; health condition. We say that health condition is an important covariate when the event rate depends on health condition. Suppose the event rate is higher in the control group where there are sicker patients. Then the low event rate in the treatment group may not be due to the treatment, but simply because there are healthier patients in the treatment group. In this case we need to compare patients with similar health condition, and we want to see some overlap in the health condition between the two treatment groups. Otherwise, the two treatment groups won't comparable.

What about if there are many covariates? One solution is to replace the collection of covariates with one single number called propensity scores. Propensity score is a condition of probability of receiving the LVAS given a patient's observed baseline covariate values, such as age, gender, prior cardiac surgery and so on. Like health condition in the previous slide we compare patients with similar propensity scores. When the propensity scores are balanced across the treatment and control groups, the distribution of all the covariates are balanced in expectation across the two treatment groups. So we can use the propensity scores as a diagnostic tool to measure treatment group comparability.

And the two treatment groups will be comparable if there is enough overlap between them, however, the two treatment groups in this study did not overlap enough to allow a sensible treatment comparison. We performed propensity score analysis. We adjusted for all imbalance and clinical important baseline covariates.

This is a box plot graph. A box plot provides an excellent visual summary of many important aspects of a distribution. The left box plot shows a distribution of propensity scores for the control group. The box stretches from the lower hinge defined as the 25th percentile to the upper hinge, the 75th percentile and therefore contains the middle half of the scores in the distribution.

The median is shown as a line across the box. The right, 57 observations corresponding to 76 percent of the data are above the lower hinge. Those 76 percent of the data from the treatment group do not overlap with any observations from the control group. As we can see, there isn't enough overlap to compare to treatment groups.

This is another way of expressing two distributions of propensity scores. Here the propensity scores are categorized into five groups so that the number of patient in each group are evenly distributed. In the first bin there are eight controls and ten treatment. In the second bin there are four controls and 13 treatment. And the rest of the bins are all LVAS patients. We can see clearly the two distributions did not overlap enough to allow a sense of comparison between LVAS patients and control patients.

The propensity score analysis, therefore, tells us that any treatment comparison adjusting for imbalanced covariates are problematic.

The sponsor performed the survivor analysis, and these are the survivor curves from the original BTT study. The red line indicates the survivor curve for 190 patients in the treatment group and the black line is for the 35 patients in the control group. As you can see, there is a large difference between the two treatment groups. Now, the sponsor picked 75 patients from the treatment group and 12 patients from the control group. The sponsor was interested in whether there will be any significant difference between the two treatment groups.

Let's find out, and this is what they got. And as you can see, there is a significant difference between the two treatment groups. However, there are some concerns with sponsor survivor curves. It was mentioned before that the two treatment groups may not be comparable and there is also another concern for the censoring, because event is not independent at censoring in this case.

It implies that any difference in survivor curves between the two treatment groups may be problematic. Even if we assume that two treatment groups were comparable, we may also find other subgroups with significant difference in survivor curves between the two treatment groups.

We choose a subgroup of patients with certain characteristics. We found that 76 patients with such characteristics from BTT sample, there are 64 patients from the treatment group and 12 patients from the control group. As we can see in this graph, there is a significant difference between these two treatment groups. You might be interested in what the characteristic was. They are the patients with age evenly divisible by three.

In the BTT sample the sponsor found 87 patients with relative contraindications and it show that there is a significant difference between the two treatment groups. And we also found a subgroup X of 76 patients with age divisible by three where there is a significant difference between the two treatment groups. Our findings from these two subgroups had been expected because there was already a large difference between the two treatment groups from which they were picked.

We may also find such subgroup Y. In fact, any sample from the BTT study will likely show a significant difference between the two treatment groups.

Now let me make a conclusion. The result from subgroup with relative contraindications may not be extended to the intended patient population. The two treatment groups are not comparable so that any direct or covariate adjusted treatment comparison is problematic.

Thank you.

Let me turn the podium to Dr. Pina.

DR. PINA: Thank you, Dr. Ahn.

Panel members, ladies and gentlemen, I cannot resist but to take a dig at my friend Jim Young. I've also been doing this maybe for three years less than you, since you're a little bit older than I am, but I've known you for all those 17 years. And I am also part of the Ohio Transplant Consortium. Did not have the opportunity to see the sponsor's slides except for a few minutes prior to the panel. And I agree with Dr. Young that there are patients that still continue to make us scratch our head and wonder what we're going to do next.

You've seen this slide before, the intended population. So you've seen this before. Dr. Berman shown this in white, the current indications and in blue the intended indication for expanded use for patients who have a relative contraindication to transplantation but are expected to become transplant candidates.

Just to review one more time, the dataset that has been presented as part of the dataset for this PMA includes 190 LVAS patients with 35 controls divided up into two groups. And I'll show you a graphic of that. Group one patients with no relative contraindications. Group two the subgroup retrospectively drawn from the original dataset that have the relative contraindications chosen by the sponsor which constitutes 75 patients with LVAS and 12 controls.

This is sort of my diagram of what you have seen now several times just to remind everyone that all these patients were considered transplant eligible by their individual institutions and, in fact, had been listed for transplantation.

Here we see the 75 patients chosen with these relative contraindications of the LVAS group and 12 of the control from the 35 original controls. It should also be noted that out of these 75, 65 percent of these patients were ultimately successfully transplanted.

Here is the list of relative contraindications that now you have seen several times, and it's also included in your panel pack.

I just want to address briefly the short or long term. We really have no accepted current definition for short or long term in the context of using left sides mechanical circulatory support. In fact, of this dataset, 160 of the patients were on the device for less than six months; of 30 patients who had been on the device for more than six months, 15 patients had had device for greater than one year and four had had it for greater than two years.

I want to address the relative contraindications. I think through the years those of us who have been doing this for a long time have seen the traditional contraindications have become relative contraindications. And I think the best example of that is when I started doing this we were really not transplanting diabetes. That was something that was sort of voodoo and we stayed away from it. And as the years have gone through and we know to manage each patient better, we have better immun suppressive agents, we have better antibiotics, we are now transplanting diabetes.

One paper that's included in your pack there from Cimato and Mariell Jessup, who has extensive experience in this area, in fact talk about the old traditional contraindications now becoming relative contraindications. And I've highlighted here in blue those that have been chosen by the sponsor, but you can see that there are a lot of others that every center considers individually.

I should also note that during these years, and one of the reasons that we have become more into the relative contraindication field as opposed to traditional, is our medical therapy has changed. In 1987 consensus was published, so we were introduced ACE inhibitors. In 1992 SALT published and we were introduced to ACE inhibitors in a group of patients. Jim Young always reminds me that coronary surgery has gotten better and better and that we're doing more procedures that are bringing patients, perhaps, to us later but that are keeping patients alive for a longer period of time.

-blocker use didn't really come into its forefront until the 1998/99. And now we have nitric oxide available to us for dilating the pulmonary vascular tree.

Milrinone IV was not available also until the early '90s and didn't get really increased in use until like 1994/95. So we have really seen some dramatic changes in this area.

So of these relative contraindications it's unclear why those seven were chosen and not others. For example, high plasma reactive antibodies, which we know do exist and can exist in patients with devices and make it very difficult to transplant, and we always try to reverse these. History of cancer, psychosocial issues and, as I said before, diabetes.

And it's also unclear how the relative thresholds were chosen for contraindications. If I remember correctly, back in the SALT trial our creatinine level maximum was three, the patients were allowed into the trial and very often the creatinine does in fact get better.

Let me touch on just a few of these parameters that have been listed as relative contraindications. We realize that renal dysfunction can be a marker of poor outcome, that keeps coming out in every single study having to do with heart failure patients. We also recognize that renal function can improve with profusion, but if the renal function is abnormal due to intrinsic renal disease, it may not improve with profusion. However, I have a tough time telling ahead of time, and many of my colleagues do, who is going to reverse and who isn't. And that definition of renal dysfunction really varies quite a bit throughout the literature.

And this is taken from the Cleveland Clinic and Dr. McCarthy, a very prominent surgeon at the Cleveland Clinic. These are 25 LVAD recipients who were actually listed for transplant, critically ill patients. Six of them died of progressive organ failure. And this paper is included in the panel pack for you to review.

The LVAD survivors did tend to have lower serum creatinine and BUN. But Dr. McCarthy in this paper states that there was no predictor of failure to improve renal function. In fact, three of the survivors have the highest BUNs of greater than 50.

Another example from another transplant center, a group of 11 patients all listed for transplant. And the authors here state that no patients were excluded from listing due to renal dysfunction. Here you can see that the BUN dropped, as did the serum creatinine.

The authors go on and make a statement that they really don't see a renal function that they will not transplant, and they can address things like dialysis and many patients then get renal transplant after they get their hearts. And some centers actually do both at the same time. I know we do.

Let me address the pulmonary hemodynamics. This can be the vain of the existence of many of us who take care of these patients.

An elevated pulmonary vascular resistance that does not reverse is a risk factor for RV failure and poor outcome. However, if you get a group of us together we will tell you there is no consensus on how to do this. There is nothing in the guidelines that tell you how to do this. And these are all the agents that we use. We use nasal cannula oxygen, we use intravenous nitrates. We like nitric oxide. We use ACE inhibitors. The A2 receptor blocker is now on the forefront. Direct acting basodilators like hydralazine and IV Milrinone, which has shown very nicely to lower PA pressures. And then we know that LVADS also do this. However, you cannot predict with the current published literature which patient will reverse and if they do reverse, to what degree will they reverse.

So here are some examples from the Texas Heart Institute of patients pre and post LVAD insertion in the sort of tan bars here and then the purple bars are PA mean pre and PA mean post LVAD insertion.

This is from St. Louis, another very large transplant center showing significant decreases in PA mean pressure.

And this is a collection from the Cleveland Clinic and from Stanford published by Dr. McCarthy showing a very nice decrease in the PA diastolic pre and post implantation of an LVAD. So we know that this can happen.

Again, from McCarthy's paper in '95, cardiac index improvement, a drop in pulmonary vascular resistance, which is kind of a drop that I like to see, and all these parameters dropping in that group of patients. However, he makes a very important point in that paper. It says they could not identify a priori who would need an LVAD support. Now why is that important? Because if the PA pressures don't come down, the right ventricle can fail and you may need to put in a right ventricular assist device to support the right ventricle even when you have a left ventricular assist device. This is not something that we enjoy or like doing, but they could not identify a priori who need one and who would not.

Finally, let me just touch on hepatic dysfunction from the same paper by Burnett that's also in your pack. Eleven patients that were listed for transplant, I showed you previously their renal function, total bilirubin at listing was 4.4 and at transplant 1.6. One of the patients did not recover and in fact had cirrhoses.

There's also a statement in there stating that sometimes when the LVAD is initially put in, that the right ventricle may suddenly feel it and hepatic function actually worsens temporarily. And I think we've all seen this, and this slowly resolves.

So once again the proposed expanded indication are for patients with relative contraindications who are expected to become transplant candidates with mechanical circulatory support. So in my summary we have no standard definition for long or short term LVAS use. The number of patients with more than one year is limited so that the conclusion re long term cannot be made. That relative contraindications in fact are relative.

And including these seven relative contraindications and excluding others is not justified in the dataset. The thresholds and definitions of relative contraindications chosen by sponsor do vary in the literature quite a bit. And that the patients that are selected for the dataset were, in fact, listed for transplant and most of them, as I have show n you, were transplanted. And that even those patients with relative contraindications are currently being listed for transplantation.

From the dataset presented, panel, there is no way to predict which patients with the relative contraindications as identified by the sponsor will in fact become transplant candidates if they are not prior to device placement. Therefore, writing an FDA approved label would be difficult.

And I thank you.

CHAIRPERSON LASKEY: Thank you, FDA personnel.

Any questions from the panel here for the presenters?

DR. WEINBERGER: Yes, Warren. I'd like to ask Dr. Berman, in the original PMA when this device was approved under the BTT PMA was there an intent to send patients home with the device or is that a new request for the current PMA supplement? And I'm asking vis--vis the bioengineering examinations.

DR. BERMAN: Both devices, the N100PC and the N100PC(q) are currently approved for hospital discharge for patients who are bridged to transplant.

DR. WEINBERGER: So that the original examination of bioengineering parameters included the possibility that a patient would use the device at home?

DR. BERMAN: Yes. For example, electrical safety, EMC alarms, all of that was examined and we have no questions regarding those issues.

DR. TRACY: Did the propensity score that you did include the seven relative contraindications that the sponsor included in their list?

DR. AHN: Not all of them. I showed six baseline covariates with p-value less than 10 percent. I included those six.

DR. BAILEY: A couple of those were drug utilization I guess at baseline. Those were amazing differences, and I'm wondering what accounts for that difference in the two groups?

DR. AHN: It is more of a medical question, so --

DR. PINA: Yes. I can answer that. I think I said in my presentation that Milrinone intravenously we had an early trial called -- I think it was called a PROMISE trial with oral Milrinone. And IV Milrinone did not really start to get used extensively until later than 1991/'92. I remember in my practice really '93/'94. So dobutamine was the main drug being used at the time and then I think we've slowly turned around.

DR. BAILEY: So in a sense then the propensity score is largely a proxy for time, which as you showed very nicely, there was very little overlap in the time?

DR. AHN: Yes. And notice that when I did a propensity score analysis I did not include the year of implant.

DR. BAILEY: But I suspect if you had, you would have gotten similar results?

DR. AHN: Maybe more. More drastic.

DR. BAILEY: I don't know if this is the right forum, but you made a very nice argument about noncomparability. The same argument would have applied to the overall comparison that was the original submission. I'm wondering what was different then versus now? I mean, why was it deemed acceptable before?

DR. BERMAN: Dr. Bailey, I'm afraid you were right that this isn't the forum for that. It's not up for discussion right now.

CHAIRPERSON LASKEY: The rules of the game have changed. I'll say it.

DR. YANCY: Two questions, please.

First ask Dr. Berman has to do with whether or not in the original application the issue of durability was evaluated and if so, what was your assessment of the durability of the device, particularly since there's a question of long term application here?

DR. BERMAN: The sponsor submitted information from bench testing, which is in your panel pack, and which was submitted in support of the original bridge application to demonstrate a multi-year expectation of life with this device.

DR. YANCY: I saw those data. But in my judgment there are none physiological because it basically was a water bath.

In the clinical data submitted --

DR. BERMAN: The clinical datasets --

DR. YANCY: -- there were two patients that --

DR. BERMAN: The clinical dataset that I'm aware of is very consistent with the bench data, that is device longevity is about what the bench says it will be. And that is based on an admittedly limited dataset because there are not many patients out at multi-year. In the United States the device is currently approved for bridge to cardiac transplant. And most patients will only be on the device a matter of months. Yes, there are some at years, but it's very few.

In Europe, the device is approved for other things. It's not part of our discussion. But the number of patients multi-year is -- there's some data but it's limited, I think would be a fair way to say. But long term durability: (a) is not an issue today, and; (b) is not an obvious problem right now.

DR. YANCY: Well, the only thing that I would retort with is that the language that's requested does include the phrase long term.

DR. BERMAN: Well, we pointed out that based on the dataset we've been given and based on the dataset that you folks are deliberating about today, there were 30 patients six months or more, 15 one year or more, four two years or more. And so we consider that insufficient to justify the use of long term, especially coupled with Dr. Pina's concern that there really is no accepted definition within the community of what the term long term means.

So we don't think the data supports it and we don't really know what it means to begin with.

DR. YANCY: And this second question is unrelated, but it's for either of the panel members. It has to do with how the question of relative contraindications was addressed with the original application. Was there any comment about that, was there a statement of concern, did it come up for --

DR. BERMAN: Are you asking questions about the PMA application from which the device was approved for bridge?


DR. BERMAN: To my knowledge, and I was not the lead reviewer, the matter of relative contraindications was not brought forward by the sponsor. But if that's wrong, i would allow them to correct me.


DR. KRUCOFF: A question for Dr. Ahn. I'm actually going to resist asking yo why you showed us a survival curve of age divided by three when in the panel pack you used the example of last digit ID 01 or 2 and try and stay on the serious side of just a lay person understanding where statistics are or are not potentially useful in the application for an extended label.

So I think you did a pretty clear job helping me understand the inability to compare these groups. I guess from my limited statistical educational background, when I see numbers like three patients with a total bilirubin greater than five in one group and zero in the other group, there comes a point where populations in a dataset simply are too small to support any statistical conclusion of any kind, not just as a comparability issue between two groups but as an understanding of what role that particular feature in a treatment have with one another.

So my question is where is the lower limit in a dataset of the ability to support any kind of statistical conclusion at all about the impact of a treatment?

DR. AHN: Notice that the 87 patients sponsor selected, they are a very heterogenous group, as you indicated. They use seven relative contraindications criteria and for some criteria, there was three patients in the treatment group and none in the control group.

And there isn't -- in the frequency table we -- to compare any sensible -- to have any sense of comparison we like to see more than five observations per cell. In this case, three observations taken from treatment and none from control, that might be an also issue, too.

And also total bilirubin, there one patient from LVAS and none from control. And pulmonary resistance, one from LVAS and none from control and so on.

So it is hard to define what the population might be when we have a very heterogenous characteristic sample.

DR. KRUCOFF: Those numbers are from table 4-1 which is on page 9 of tab 5A of your panel pack?

DR. AHN: And the reason why I showed the subgroup with age divisible by three or patient ID ending in 01 or 2 is to show that the retrospective psychoanalysis is what we try to avoid as a statistician.

DR. AZIZ: I think the question about long term durability I think is an important question. And I know that, obviously, we've got to focus on the data that was presented here today. But I think it would be fair to say that of all the devices that have been implanted both here and in Europe on a long term basis, I think the data in the sort of 1,077 cases that you said we can look at, I don't think that I'm aware of any device that's malfunctioned. And even though we can't look at the data, I think we do have a general idea that -- I mean,, this device in patients in whom it's been in for more than a year or so has been very durable without, I think, stopping or having a malfunction. So I think there is a sense out there that it is a good device.

The second thing actually, this is for Dr. Pina, looking at bilirubins per se is just one aspect of liver dysfunction. You know, was any attempt made to look at the enzymes, you know, albumin, OT, PT and you know the other parameters rather than just focusing on bilirubins you could have many reasons for being --

DR. PINA: I think you make an excellent point that some of these patients have a lot of other issues with their liver function and that bilirubin is just one of the many. And, in fact, the paper that addresses the hepatic dysfunction actually addresses cytokine and inflammatory factors as being more predictive. However, we have been given as a relative contraindication the total bilirubin, and that's what we have to focus on.

But I agree with you that it is multifactorial.

DR. BAILEY: Can I just ask for my ignorance, is it reasonable to assume that if you have an improvement, let's say, in pulmonary pressure or in renal function with the device that that has the same implications as someone who hasn't been on a device and has good function as far as post-transplant survival?

DR. PINA: You know, there's never been a randomized controlled trial that looks at that specifically, but I can tell you clinically if the pressures come down with whatever format and stay down, that the patient will do much better. Early and later, because you have a problem early in the operating room and then you have a problem later. so, yes the answer yes.

And we usually wait, even if we do it with medications or we do it with the device, wait and make sure that they are down and stay down and we do repeated hemodynamic monitoring.

DR. YANCY: Let me just raise one other issue with Dr. Pina. The number of patients on LVAS who went on to transplantation, as you pointed out, was 65 percent. Do you have access to information or maybe I overlooked it in the program material, as to whether or not that group that went on in transplantation was populated towards one or another contraindication more so than the others, that is amongst the seven, the group that actually went for the transplantation did they reflect a certain profile?

DR. PINA: I have not seen that data, unless Dr. Berman has seen it. No, we have not. It would be an interesting point to see.

DR. LINDENFELD: Not only interesting, I think it's critical. And I think that when we come back to the sponsor, what I would like to see is the table of the relative contraindications. And though I recognize some of the numbers are small, I would like to see how many of each contraindication went onto transplant and what the one year survival was for those.

We're asked to say these are relative contraindications and what we'd like to see is 49 were transplanted, were those all the ones with the high BMI? Was there a much lower percentage of the high PVRs or the high creatinines? And I think that's just a very critical question as we look to say, okay, if you're not sure, we think it's okay to do this. I think we need to see how many were transplanted and subsequently what the one and two year survival in each of those contraindications. And I recognize some of the groups are one in three, but some are 20/22.

CHAIRPERSON LASKEY: A critical and continued source of confusion this relative contraindication business.

Dr. Ahn, do you find it puzzling that when they did the multivariable analyses, their proportional hazards, that three of the variables that were felt to be relative contraindications, systolic, serum creatinine, total bili were not found to be statistically significant predictors of mortality? What is that telling us, besides confusing us?

DR. AHN: When you have multiple variable and we question, for example, if even though one -- if you have one variable in the model, that variable might be significant. But if you include many variables because of interaction between the variables, some of the variables may not be significant. So I have not seen -- it might be--

CHAIRPERSON LASKEY: But the three things that failed to survive the test are those on which we're relying a great deal of credence in terms of being relative contraindications. These were important physiological that are meant to provide a definition for this patient population and yet they failed to stand up to the statistical rigor.

DR. BAILEY: I think Dr. Ahn is saying that you have to look at the joint effect of those three variables before you could rule out that they had some impact. Not just look at each individual variable as partial -- have you looked at the joint effect of those three variables?

DR. AHN: No, I did not.

DR. BAILEY: But I mean, another possibility obviously is that, you know, how abnormal or how deficient were those parameters? And if we're just at the margin maybe that's part of the story.

DR. KRUCOFF: Another feature, I don't know if this is really fair to ask Dr. Ahn or maybe we can come back in these multivariable models after lunch, but at least my understanding was that the variable entered in that model was probably the initial creatinine and how many of those patients with elevated creatinine had reversible dysfunction versus not may also impact on whether they survived well or poorly. And again, I don't know if it's fair to ask Dr. Ahn, but my understanding of the parameter entered for that model is just a single creatinine value when the patient was enrolled. But maybe we can come back to it.

DR. AHN: Yes. Right.


DR. SOMBERG: Well, just a comment and maybe the FDA reviewers would like to expand upon that. But I'm very concerned with what I hear of a number of questions from our panel suggests that in making a decision we're asking for qualifiers when in actuality we're asked to make an evidentiary determination and it's all based on comparison to something which has to be a control. If the control was inadequate, how can one ever reach a decision regarding whether parameters may go one way, another, they may change. They have to be compared something and if the control is the 12 patients or the 35 patients, if the control is inadequate and not matched, then almost anything you choose will give you a significant difference and there's been no attempt to try to validate that control with any historic other data, why should we determine anything else?

DR. PINA: I want to respond briefly one more time to Dr. Lindenfeld's concern about the lowering.

We do have data. If you go into page 11 of the sponsor's under tab 5A, they tell us that 12 of 22 of the patients who had the definition of renal dysfunction did in fact go to transplant. But what they don't say and what we've never seen is what was the creatinine at the time of transplant in those patients who have the relative contraindications. And table 4-2 shows that the serum creatinine level in that group was 3.23 and it gives other parameters, but we don't know individually what happened to those patients. And in a similar fashion with the PA pressure and the PDR if you go into the next tables. But in all fairness, we do know how many went to transplant.

CHAIRPERSON LASKEY: If there are no other questions from the panel, that means we are proceeding at an amazing efficient pace here.

Are you prepared to do your view now? Yes, well we can wait.

So what I'd like to do is to have Drs. Krucoff and Somberg give their reviews and ask questions of the sponsor.

Thank you very much FDA folks.

And after they're through, then we'll break for lunch and we'll come back for the panel queries.

DR. KRUCOFF: You want me to start?

CHAIRPERSON LASKEY: Please. Thank you, sir.

If you have a question for them, you should invite them to the table, yes.

DR. KRUCOFF: Okay. I do have a few questions.

MS. WOOD: I'm sorry, I need to correct that. You come to the podium to answer the questions, either the FDA or the sponsor.

DR. KRUCOFF: Okay. Sorry. I'll direct my questions.

And I guess I'll leave you guys to decide -- I just want to make sure we're starting on the same page.

Certainly my understanding is that for a requested expansion of an indication that the data presented to support that expanded indication should stand alone. And I realize there are certain reference points including the preclinical testing etcetera that we're not revisiting, but at least the clinical data should stand alone.

And I think it as pretty clearly -- I think Dr. Young specifically said, but I think we all appreciate that this is not a dataset that was built on a prospective hypothesis. That this is retrospective look driven pretty clearly by the dilemma that we face with patients who are sort of on that edge of are they going to be transplant candidates or not and, obviously, the dilemma of whether to employ a technology at this level and to try and better understand how to employ that technology. So that's my take, and please feel free to correct me if any of this incorrect, but that's sort of the spirit, I guess, of what I heard this morning and took from the packet.

But I do think there's an important thing, and again, Dr. Young, you mentioned that retrospective analyses have guided us in transplantation, in fact in many areas of medicine. But I also have to say that from a trials data, from an evidentiary perspective generally what retrospective analyses have guided us towards are a clearer hypothesis to be prospectively tested. And I think one of my main dilemmas with the dataset today is whether the utility of the data presented is helpful for anything except the eventuation of a useful hypothesis to actually be tested by a meaningful dataset.

And I think another element here that I've been wrestling with are the simply small numbers in many of these categories. So, obviously, if you have zero patients with a particular feature in both categories, there's no way to analyze that. If we have one patient in one group and zero in other and the one patient dies, that's 100 percent mortality. Again, obviously, statistics don't make any sense. As we get two or three or Dr. Ahn was willing to volunteer five in a cell, for certain kinds of safety analysis I think we obviously go down to those numbers and levels. But I have to say that the numbers of patients who have any evidence in some of the categories that are proposed for this expanded indication worry me greatly and make me very concerned, not only that the groups are comparable the control group, but that any sort of real statistical conclusion on the certainty of an outcome in a group with three people in it in patients this sick just defy understanding.

In section 3A in your marketing history and then later in section 5A you mentioned your experience outside of the U.S., 644 patients. And I realize nobody's had a chance to review this, but boy I have to say when I see 644 patients from 17 other countries in your experience base, my first thought is what's the data? I mean, you know, where are the patients, how many of those patients also have these relative contraindications and with a little more work would it be possible to collect enough information, perhaps, to actually have some data-based evidence with regard to some of these areas of management dilemma that might provide relative contraindications?

But my presumption from the fact that there's really no detailed data on these 644 non-U.S. patients from 17 other countries other than the survival table in table 4-2 that you present on page 8 of section 5A, that we have no other detail in the panel pack and obviously FDA wouldn't have had a chance to reveal any detail. But do you have any information available to us on the relative contraindications list that you're interested in and its behavior in any of these 644 non-U.S. patients from 17 other countries?

MR. BRYDEN: Is that a question?

DR. KRUCOFF: Yes, that's a question. And I'm sorry, I don't know who. I think they probably want you to come up so it can be recorded.

MR. BRYDEN: The data from the market implants of the device in many countries, we have data in which we can be confident in survival and in device performance because they are reported and we can audit that. But these were not done as part of the trial and we do not have access to the individual conditions of the patient in any reliable manner.

So the answer would be that aside from device failure or not and survival in the market group, we do not have reliable data.

DR. KRUCOFF: Well, because obviously that represents information that would be nice to have and perhaps in a post-market environment something that could be considered would be collecting such data with your implants, if that was feasible or logistically possible.

Dr. Ahn, you mentioned an ethical issue with regard to randomizing patients who have these relative contraindications. I just wanted to ask you a little bit.

The way I see this right now our implication is that there are a lot of patients who because of their creatinine or their bilirubin or their age or whatever, may not in fact be considered candidates a VAD or transplantation. And a randomized trial, bagging the logistics for a second, just ethically, that a randomized trial from my perspective would be an opportunity not only to afford those patients support and potential conversion to becoming transplant candidates, but in fact that would be a perfect and highly ethical perspective for a randomized trial. Can you help me understand why that would be unethical?

DR. YOUNG: Yes. That's a critically important point that we've actually grappled with. Ileana spoke about our case series which we published in 25 patients who absolutely clearly had no business being transplanted on the day that they were listed and received a VAD with the intention of rehabilitating their renal function. Now, in those patients very similar to the control group of patients here, which really were quite ill patients as we looked at, I think the invariability of death was present. And the only hope would be to VAD the patient, improve flows to the kidneys, try to attenuate all the multiple pathophysiologic reasons for the renal insufficiency. And the only way, even with all the progress has been made -- and I noticed Dr. Pina didn't include Natrecor on her list of drugs to use. But even with that, really the only thing we have in our bag that we can pull out is a VAD.

And I believe that with the data that exist today it would be unethical to do a randomized trial on this patient population. And I think instead you have to bite the bullet and make the commitment that you're going to try. But this is huge in heroic sort of therapy.

Now, in that case series where we did that and, this was looked at by appropriate regulatory purview at our institution, we were able to demonstrate that a significant number of patients did improve to a point where we felt comfortable transporting them. And I think that's pretty solid evidence that you can "get away with it" in many cases. But in sense what I'm bothered by is that we haven't clearly defined this when we're going to actually transplant the patient vis--vis when we list the patient and put the ventricular assist device in.

But I would have trouble with a randomized trial of VAD versus no VAD in this kind of patient population.

DR. KRUCOFF: So am I missing something. This kind of population patients who have relative contraindications who presumably under standard care would not get listed or transplanted, i.e., would not be candidates for VAD as currently defined?

DR. YOUNG: Correct. And --

DR. KRUCOFF: And would not get VAD or transplanted? So in a randomized trial you'd really be affording at least half of them or whatever the percentage randomized, something that they're currently not getting access to?

DR. YOUNG: Well, at many centers.


DR. YOUNG: And this represents the diverse opinion that is out there at many different centers. But for me and at my center I would personally have a great deal of difficulty participating in that kind of trial.

DR. KRUCOFF: Because these patients have an opportunity for a VAD based on the judgment of the doc?

DR. YOUNG: Right. That's correct.


DR. YOUNG: Fair enough?

DR. KRUCOFF: Thanks.

You know, I think personally I have to take a step back and visualize clearly that there are really two decisions here. One is the decision to put in the VAD or not. And the other is ultimately the decision as to whether the patient is a candidate for transplant and that the temporal sequence of these is -- in BTT you had to start with the patient is a transplant candidate and then they could be afforded a VAD. Now we're asking the question sort of the reverse way; if the patient might be a transplant candidate, should they be afforded a VAD.

So, Mr. Bryden, you've put a slide up that said it was inappropriate to rely on clinicians bending the rules. And to me what we're really talking about here is maybe less the regulatory side of indications supported by data defining populations in safety and efficacy. We're really talking about the practice of medicine is the judgment in the fuzzy zones that we all deal with in devices. So is the implication of your slide that the practice of medicine is a bad thing?

MR. BRYDEN: I think the implication is that where the regulator or those who advise the regulator are of the view that a VAD would be appropriate in the circumstance, that it should be practical to find the words by which that is approved rather than apparently prohibiting it but expecting medical profession to avoid the prohibition by making judgments which are outside the rules. That was what was intended by t hat comment and that slide.

With respect to the potential for a randomized trial, we are right now engaged in just the very early stages of a randomized trial. And the randomization is that an approved VAD is the control and the Novacor will be the trial arm and the equivalence of the two is what will be tested by the trial.

What we're saying here is that in this case the overall approval that has already been given for this entire population which includes patients who had these contraindications and were listed, that the result of that does demonstrate that these patients benefitted from that listing and as a result the device was approved for that purpose. But on the advice on a number of commission, including those who are with us today, and reviewing the data and surveying centers that do a significant share of the transplants in the United States, it was quite clear to us that a significant share of those patients who have the relative contraindications which we tested which were included in our group two, would today if presented at many of those centers not be given a VAD.

At the same time, it is clear that to be within this group at all they are at risk of imminent death. How imminent is imminent, seven days was the average within the control group. That is not, we believe, an indication of something wrong with the control group. It demonstrates an imminent means -- imminent, it doesn't mean sometime in the next two years. It means imminent.

So the fact that these patients today would not be provided within the rules that are available access to the VAD and yet within the trial that was conducted, whatever the inadequacies of the controls as they wee a decade ago, the results were clear that a very substantial share of these, in order of 65 percent, survived 30 days post-transplant. And it is very clear that that group would not and did not survive long without the device, and yet we do have a structure which unless the publications are wrong, the advice we received are wrong, and the survey of these ten centers are wrong, have a hit or miss opportunity dependent largely on the clinicians at the center deciding to implant because their choice is let this person die soon or give him a VAD even though it's not really it's approved for. We're suggesting that is not appropriate.

And the control group in this case is just as it will be in our prospective randomized trial for destination therapy. It is patients receiving the same therapy but with a different medical characteristic. We have that already. It was developed in a controlled trial under the direction of the FDA and was adequate to allow the approval, which has proven to be in the market substantially borne out in the results post-approval with what was expected in the trial results.

So what we have in front of you is not an exercise in standing of a head of a pin in statistical theory. It is that we have patients who had these characteristics, who had substantially the same results as other patients who did not have those characteristics. And our question to you is, is it not appropriate to regularize the process by which at all centers if they come to the conclusion that this patient is likely to survive to transplant if given a circulatory assist device, that they be permitted to do so within the rules rather than relying on them to bend them?

DR. KRUCOFF: Presuming you have defined the rules, which is what we're here to talk about?

MR. BRYDEN: Yes, exactly. Absolutely. That we are more than happy to be guided by both the panel's advice and the discussions with the FDA about the specifics of the wording. But it is already an established and intentional process by both the FDA and by CMS that they not practice medicine by telling each clinic exactly what will be the criteria for transplant. So what we're doing, as has been done with the approved destination therapy indication, recognizing that is not up to us to adopt it or not, it's a fact that it exists and to say the process now demands that recognized transplant centers make these judgments. They're not easy judgment. They're very difficult judgments. But the process that you use today throughout this therapy is to require them to make that judgment. All we're saying is apply that judgment in this case as well.

DR. KRUCOFF: Part of the paradox of the BTT dataset to me is that actually the cohort of patients who you have to analyze with these relative contraindications are the results of doctors making judgments --


DR. KRUCOFF: -- that these are patients who would be good candidates and, in fact, based on the evidence the practice of medicine in that case probably is not a bad way to go. But I don't want to get too stuck in this. t's just that the starting point of the BTT group as a listed group of patients creates a paradox ultimately relative to trying to deal with all the patients who might have these relative contraindications who doctors might not consider to be potential transplant candidates.

MR. BRYDEN: May I make a very --

DR. KRUCOFF: How you would actually define one group from the other, which is the rub:

MR. BRYDEN: May I make a very brief additional comment? I promise it will be very brief.

The use of a list of any kind as a shortcut to defining a population is a useful means of conducting business because it means having done something once and named it. You can just use that name and it always means the same to everyone, so you don't have to go through the whole process again. But when the name of a list does not connote consistent characteristics over time and from center-to-center, the mere fact that a name is or isn't on the list is not evidence on which a regulatory decision should be based in our view. It is the underlying characteristics that can be demonstrated and checked and tested, and judgment made. But whether the name appeared on a list is not in itself a medical characteristic.

And I think a very considerable amount of the argumentation that we have heard has been whether people were on a list or they're not on a list. The question is, is there a consistent definition of what put you on the list and if so, you know what these people are. The whole point of this exercise is it is not consistent from center-to-center or over time.

So the fact that you are or aren't on the list is neither a good thing for us nor a bad thing for us. It doesn't really tell you anything. We believe you need to examine the underlying characteristics. And in those, we believe, there is reasonable understanding and ability in the clinics to make those judgments.

DR. KRUCOFF: Well, thanks. Actually that's a very good seque into the characteristics issues and my next point. And I don't have too many more. But the one characteristic that was not only a discussion of today but dialogue in the pack between you and FDA previously is reversibility and nonreversibility of some of these features. And while you made it clear in one of your responses that you're not asking for an indication for reversal of renal dysfunction or for reversal hepatic insufficiency, I think it's pretty clear again, Dr. Young mentioned today, that whether or not these features abate or improve I believe were his words that some of these features and some of the judgment and some of the practice of medicine element here, and one of the biggest missing pieces to me of a characteristic that might be objectified would be reason or evidence that would support the potential reversibility of features like the creatinine or hepatic dysfunction.

So actually I was going to ask Dr. Young first if it's okay, how important is reversibility? As Warren mentioned, and again I don't want to dig beyond my statistical capabilities, but in the multivariable model when elevated creatinine is not predictive of death, one of the things I start to wonder about is well, maybe that's because in a good number of these patients that creatinine reversed and when they were actually transplanted, their kidneys worked fine when they're given circulation. And, boy, isn't that a great population to put a VAD in? But where is the parameter, the characteristic of reversibility on at least the reversible -- I'm going to ask you about age and body mass in a second. But on the reversible side, on the bilirubin and the creatinine?

DR. YOUNG: Those are very fair, very appropriate questions and drilled down to some of the challenges that we have when we're trying to gain insight from these kinds of databases, retrospective analyses or not. And I liked the presentation about the age or the digit numbers in my slide set about designing and implementing clinical trials. I use that great and important and distinguishing characteristic of your birthday and what sign you happen to be under. And everybody knows the rather famous analyses that have been done in multiple clinical trials that show that.

And so when you cone down into this very important question, and you're right, I don't think anybody's suggesting that we want to say that these devices are going to be put in ipso facto to cure these difficulties. It turns out that in fact there were significant changes, and I think I showed a few before.

Do we have the slides? You wanted to PA pressure, creatinine improvement, body mass, etcetera, etcetera were the seven relative exclusion factors. And we do have that I believe for everything but, was it age? Age didn't improve. I don't think we have pulmonary vascular resistance.

DR. KRUCOFF: Did it improve body mass?

DR. YOUNG: Yes. Well, actually, I'll show you. Body mass is interesting. In short term there were some rapid changes that probably were fluid and diureses, but long term there were some changes in both the cachectic and the overweight patients, if we could that up.

There we go.

So here is the resolution of these relative contraindications that were picked. And, again, you know I do respect some of the points Dr. Pina made. Choosing these relative contraindications is not entirely an exact science, and where to put the cut points is inexact, but we do have some guidance. But here you look at the VAS patients that were transplanted and those that were not transplanted in red, and you see an interesting thing; is for one reason or another many of those that weren't transplanted actually got worse. However, the preponderance of the patients that ended up getting transplanted over time, the creatinines got better. So in the individuals that wee hemodynamically supported and, obviously, this isn't necessarily done in a vacuum, but I believe that you can point towards the VAD improving this. With creatinine there was improvement.

What's the next slide?

Pulmonary systolic pressure. Again, tends to fall in everyone that the VAD goes into, whether or not they ultimately get transplanted. But rarely does the PA pressures go up in these patients. And, again, remember the cut point was systolic PA pressure of 60, as I alluded to here before. So you can see pretty rapidly you'll effect hemodynamics from a decongestion standpoint. And again as was alluded to earlier, whether this is a change in filling pressures in the left ventricle or a change primarily in pulmonary vascular dynamics, I don't know. Often times we can't sort through that. Many patients will get transplanted who have a fixed element of pulmonary hypertension. But you can see here it does what we hope it to do.

Next slide. The next one. What's the next one? Body mass. Okay.

Here's what I was referring to about body mass index. Now, these are what I would call cachectic patients. And here you can see that in patients who are transplanted there's a couple of various responses here.

Now, body mass going up like this in ten to 20 days I don't think is do to resolution of the cachexy necessarily, but maybe changes in volume status that are relative to the transplant.

These patients here, and again we're getting down to small numbers here and I don't want to make too much of this, but these patients however out 50 and 80 days probably are becoming rehabilitated. And we also know, not from this dataset but we know from other dataset, that that does happen in a cachectic patient when you can feed them.

Next slide. Oh, this is the body mass index for the ponderace patient. And, again, there's a bit of a scatter here, but you see many individuals that actually drop their weight. Why was that? Was that relief of fluid dynamics and ability to diures the patient as you're improving renal function? I assume much of that was. But there's some substantial reductions in body mass index to the area where you have problems with transplant to the area where patients do much better with transplantation. And long term support has been associated, again, in this database as well as in other databases, with improvement.

Did we go backwards or something? We need to body mass index the large patient. The next one, do we have any others? Bilirubin. Go ahead another one.

See, this refers to what you were pointing out about few patients with a bilirubin greater than 5 in the entire analysis here. But for what it's worth, the one patient that didn't get transplanted continued to get worse. The two patients that did, did in fact improve that parameter.

And, again, like I said we don't have age data for obvious reasons and we don't have pulmonary vascular resistance because of not getting the follow-up wedge pressures on these patients, also for obvious reasons.

So I think when you look at this dataset, yes, it's flawed. And, yes, it's not what we perhaps would like to have with a big randomized clinical trial answering all these questions. Because there's consistency of data in it and it goes along with other impressions that the clinicians that are dealing with these patients have. And though when you do a multivariable analysis these individuals may might not fall out because of covariate interactions, it certainly is consistent with the clinical picture that we see in small numbers of patients.

Does that answer the questions.

DR. KRUCOFF: Yes. I guess, Jim, one of the things that since clinically we would frequently use to triage patients who we think might be likely to be reversible in some of these features versus not is their history leading up to the point where you're deciding about a VAD. So if somebody had a normal creatinine, came in finally on a flare of heart failure and was rapidly going downhill and their creatinine went to two or three, I would be much more -- I mean, to me that might be a feature that could be characterized in a patient population as opposed to somebody diabetic hypertension who has a creatinine of three for two years.

DR. YOUNG: Right.

DR. KRUCOFF: Where I'd be much less enthusiastic. And I just feel like we're missing of the common sense that might in fact give us characteristics rather than just judgments for separating out who in these patients might actually benefit --

DR. YOUNG: No, I completely agree. And that's a whole another issue. Actually where I get most challenged about these decisions are the acute myocardial infarction patient who comes in with cariogenic shock, has arrested. These characteristics of that 25 patient case series that we had. Bomb, you resuscitate and the guy wakes up, you know, and they got creatinine of eight and are on hemodialysis sometimes. And you're standing at the bedside and they got a shot ventricle and they're in shock. And you're saying, you know, what are we going to do? Are we going to say this patient is a heart transplant candidate and list him for transplant and then put a VAD in and make him status 7, blah, blah. Well, that's kind of what that case series did.

But, you know, many of these patients certainly fit that criteria. And if you look at the baseline, the number of arrests prior to getting into the BTT and some of the variables, you know patients were like that.

You could quibble about where to put the creatinine cutoffs and whatnot --

DR. KRUCOFF: Okay. Let's quibble, because that is on my list. So where did you guys get these cutoffs and they're --

DR. YOUNG: Yes. I tell you, that specific data comes from the curves that were generated out of the cardiac transplant research database which shows that there is a biphasic

curve for adverse outcome at the time of transplant is the listing creatinine was above 2.5. That's just where the curve break happened to occur, and that was the most recent and the largest data analysis that we had.

And then also when you query transplant physicians and surgeons, you know, where do they put the mark where they raise the eyebrows, generally it's a creatinine clearance of below 50 and really get concerned at a creatinine clearance less than 30. And most of the creatinine clearances are calculated from the Crockroft-Gault equation. And when you get down into the less than 50 range, is at that 2.5 above, 3, 3.5 and above generally gets you down into that 30 cc less.

So even though I understand the panel's a little queasy about how we sat down and actually picked these, these are criteria that people talk about. There is evidence supporting the number. And interestingly enough, with things like pulmonary artery pressures, obesity and whatnot, there are some insurance carriers that have specifically chosen these same numbers as well as we outlined.

DR. KRUCOFF: So how do you reconcile that with the fact that in your own multivariable model and these data it is not a meaningful cut point?

DR. YOUNG: Well, the multivariable data of this particular, the BTT effort with the stratification, I think this is a numbers and an interaction problem where from a mathematical standpoint we have difficulty account for all of these interactions with the small number of patients that we have. And I'm bothered to some extent, but I think less bothered than by some others.

DR. KRUCOFF: Except that that's what you're asking for for an indication based on this dataset.

DR. YOUNG: Well, what we're asking for is an indication that if a clinician believes that a patient or expects that a patient's parameters will improve, and we've given some specific parameters if those are the ones that people want to focus on, to a point where they would be willing to accept an organ the day it was offered. The issue again is practice and what happens. and again like those patients that a VAD was put in with renal insufficiency and were listed for transplant, if we got an offer for an organ that day or shortly thereafter, it would be declined. And that is the practice that occurs. It would be declined until parameters were met such as the creatinine drops, the creatinine clearance goes up, pulmonary artery pressures come down until we believe that satisfactory marks have been made. And that, in fact, is the essence of the bridge to a bridge, if you will. Bridging to the bridge to transplant. Terminology is a little problematic here.

DR. KRUCOFF: Actually, if I can keep you here for a second, Jim, tell me about the nonreversible. Tell me about age, where is the rationale for elevated age at a relative contraindicated level.


DR. KRUCOFF: And the decision to implant or an indication for a VAD?

DR. YOUNG: This is perhaps the toughest issue and the most contentious issue, and drive perhaps by the question of age being the primary determine of whether a patient should go the transplant route or a destination therapy route.

If in fact you delve down into all of the databases, age is a consistent marker of less good, if you will to use a nonstatistical term, less good outcomes after transplantation. Now, I'm a strong believer in the relevancy of age. I mean, the oldest patient that we've transplanted was 74 at the time of transplant and was doing quite well.

And so picking a specific age is harder for me to do than many others in the community. And I have to admit I'm in the minority on the age question.

Some people in some programs will say ipso facto, age greater than 60 or age greater than 65, or age greater than 70 makes that patient not a transplant candidate, makes that patient perhaps somebody that destination therapy might be considered in.

Nonetheless, the age mark that was picked, again, was based on several different analyses which show at the elbows at the curve where these changes are occurring. Not all of the databases show the same age. I am, you know, the first to admit that.

And, again, when we look at our own personal experience at the clinic we have very good outcomes with older patients. So in fact if I were to review the contraindications, age would rarely be on that list for any given patient. And my concern would be focused on pulmonary hypertension and renal insufficiency.

DR. KRUCOFF: Okay. So really we're back to the fact that these are all relative contraindications that in certain medical centers and the discretion of certain physicians you're going to say I think this person is going to do well, and you probably would go to whatever measures would best support the person, including putting a VAD in, if you have the conviction that despite the presence of relative contraindication the overall sense is this patient will probably be a good transplant candidate? Is that where something like age would come in your --

DR. YOUNG: Yes, I think that's a very fair characterization. And, what we have with this analyses is a pretty doggone good evidence base, though flawed. Certainly one of the largest ventricular assist device databases to do that and with varied in it this inherent comparison of those with versus those without these relative contraindications, and then juxtaposed again I'm the first to admit that the control base has flaws with it. But it's right now I think the best that we can get with this type of questioning and this type of patient population.

DR. KRUCOFF: Well, up until that last phrase, "the best that we can get," I'm actually going to go beyond. I think we have spent a lot of time talking about the comparative issues. And let me just shift to one question about safety. Your slide 34, which had all the various adverse outcomes and the wide confidence intervals, some apparently higher values than others.

DR. YOUNG: Right.

DR. KRUCOFF: Is there any plot that you have available or perhaps by this afternoon could make available on the safety side relative to the timing of some of these events? How many of them cluster very early versus how many of them become issues only in later time periods after three months or six months, or a year, and then again realizing that there are a very few number of patients who have gone longer than that?

DR. YOUNG: Yes, yes. No, we do have that information. And you're right, some of the events cluster up front and then they taper down with time. As a matter of fact, they were ahead of us, adverse events right there based on the time period, two to six, seven to 12 and that has to be taken in the context that the numbers are decreasing. And so the AEs are definitely front loaded here. And this does compare it to the control patient population. But, you know, the control population, let's see --

DR. LINDENFELD: Aren't all the controls dead by two to six months?

DR. YOUNG: This is the --

DR. LINDENFELD: They can't have adverse events if they're dead.


DR. LINDENFELD: All the controls are dead after a month, right. So you can't really compare adverse events --

DR. KRUCOFF: Well, but if we're comparing, you know we can say it's obviously unsafe after seven months because there are no adverse events in the control group.

DR. YOUNG: So this was all the patients in the BTT.

DR. KRUCOFF: Okay. So one of the issues, again, as you think of extending the indication to short and long term support is not only the effectiveness issue but the safety.

DR. YOUNG: Correct.

DR. KRUCOFF: And again, having some sort of data with a comparator it would help us understand whether the seven to 12 months, 13 to 24, whether these outer bars -- basically it doesn't look like there are comparators --

DR. YOUNG: Right.

DR. KRUCOFF: -- because they have all expired by then.

DR. YOUNG: Right. And, again, looking at the types of patients that came into the study, these are not the walking wounded kinds of patients. And I might add that it was question about the patients being on inotropes at trial entry. And for BTT to get into the study, if you weren't on an a balloon pump or some other assist device, you had to be on two inotropes to actually get into the study. And then if you had a balloon pump in place, one inotrope. So that's why there was a 100 percent of both control patients and patients that went to the VAD group that wee on inotropes. There is a difference between Milrinone and dobutamine and whatnot because of the time period. This was not a pretty patient population.

DR. KRUCOFF: Thank you. I'm all done.

CHAIRPERSON LASKEY: Because of the critical importance of the whole issue of relative contraindications that we've been dwelling on here, before we ask Dr. Somberg for his comments, does anybody in the agency review team care to comment/respond/emphasize?

DR. PINA: I'd be happy to. I'd just like to go over some of the points that Dr. Young has been making.

I had personally not seen the data of the individual patients and reversibility, but I would like to point out that some of the patients that didn't reverse, still got transplanted anyway whether it was creatinine or pulmonary artery. And once more I agree that pulmonary artery systolic is not the best measure of reversibility, rather PDR would be. And it's sure that the surgeons don't like us to inflate the catheters, but we can use PAD to sort of estimate the wedge.

Another point on the conundrum of the patient that comes in with acute myocardial infarction and cariogenic shock, nowadays we use short term bridges for those patients that are available commercially. That's not the patient that you now immediately list for transplant. So times have changed for that acute very ill patient where you don't know what's going to happen to them in the near future. And I think that cariogenic shock is an excellent example of it.

The next point to be made is that insurance companies don't pick who they cover by your criteria. They pick who they cover by outcomes, and that's how they look at who they choose to pay and recommend their patients. They may want to see the list of your indications and contraindications to transplant, but it's really outcomes that cuts the mustard.

DR. KRUCOFF: Okay. I'm sorry. Then I have one last question. And, Ileana -- well, maybe, and I'll take whoever can answer.

Based on the current labeling for the LVAD is it actually contraindicated based on the current labeling to use the VAD in the setting of a patient who may have relative contraindications such as are listed in these requested extension of labeling?

DR. YANCY: And before you answer, Ileana, I would say I think that is a critical question because we need to understand what it is about the current labeling indication that really necessitates extending it with this additional language?

DR. PINA: My opinion is that the current labeling does allow the discretion of the transplant center to choose to list someone whom they believe will reverse. and I think that's what we do all the time. You always give that patient the benefit of the doubt and you go ahead and list them. And we don't make them status 7, Jim, we make them status 2s because they become status 2 nowadays.

Again, explanation for the panel. The patients are LVADs used to stay at status 1s for a while. Now they become status 2 after a month so they're no longer considered critical except for that first month. And we may keep them status for a while. They are gaining time on the list. As status 7 they gain one month total for all the time that they're status 7. So we like to keep them status 2. And, in fact, I think you've shown in your date of reversible, that some of the patients with the higher ponderosity index actually get better because they're probably moving around and exercising and we get them on diets and weight loss programs.

DR. TRACY: Mitch, was your question whether they can be listed or whether the device -- I thought your question was specific to the regulation on the device. That's my question anyway.

DR. KRUCOFF: My question is based on the current labeling. The current approved labeling for the device. Was it specifically contraindicated to put the device into patients with the creatinine greater than 2.5.

DR. TRACY: Right. And I don't think I heard the answer to that question.

DR. BERMAN: No. No. No, it is not specifically contraindicated that if a patient has any of these relative contraindications that the device may not be used. The label doesn't say that. The label says you may use it, you have it in writing. I don't remember it in my head.

Patients who are candidates for transplant, it's in your panel pack and it's in -- yes it is because --

CHAIRPERSON LASKEY: Well, that is, but can you shed light on whether there are warnings or precautions?

DR. BERMAN: Yes, the whole thing is not there.

Currently the indication for use is that the LVAS is intended for use as a bridge to transplantation in cardiac transplant candidates at risk of imminent death from nonreversible left ventricular failure, the LVAS is indicated for use both inside and outside the hospital. It doesn't say anything about not using it if the patient has PDR over 6 Wood units or creatinine over 2.5. It doesn't say you can't do it.

DR. TRACY: I'm sorry, I'm going to ask it again. What about the section on warnings, precautions and contraindications, which I didn't find in the panel pack?

DR. BERMAN: It's in the panel pack. It should be in the panel pack in the sponsor's SSED from the original indication for use. I'll go find it.

CHAIRPERSON LASKEY: Okay. Cindy, are you happy?

DR. YOUNG: Could I respond to the question that Mitch asked about the contraindications, because I do think that this is critical and I would agree with Dr. Berman that I think the language as I read it doesn't say its contraindicated. But I would use the term disingenuous. And if you look at the consensus panels, particularly that 1998 consensus panel that Les Miller led, there was a great deal of commentary in there about what listing for heart transplant meant.

To me, unless you're otherwise specifying in a clinical trial of one sort or another, that you are listing a patient for transplantation and are willing to accept an organ when that patient is listed, it becomes a disingenuous act. And whether or not you make the patient status 7 or keep the patient status 1 or 2, and then turn down organs are offered, I think it misses the spirit of what we're trying to do.

The scenario may vary from place-to-place, but even as Ileana explained with a program that might put a patient at status 2 -- there's that commentary. Even if you place a patient status 2 there are some patients that are going to get an offer pretty quick. An AA patient, an AB patient, a small female, for example.

So, yes, that explains why some might do status 7 as opposed to leaving them status 2. So I would characterize the term for better or for worse as disingenuous as contraindicated more than anything. And this labeling does take some evidence that we have, and I stress the word "some" to support the fact that we can rehabilitate the patient to get him at a point where the day an organ became available, then we would accept that organ.

And I think in the packet the indications with the contraindications are listed there, the section was section 3.

DR. KRUCOFF: Under 4 it says warnings and precautions, see warnings and precautions in the final draft labeling information for use.


DR. KRUCOFF: I see contraindications, primary --

DR. YOUNG: Right. Check section 4. Section 4. It's in that section, isn't it? It's attachment 4A that has that expanded list.

CHAIRPERSON LASKEY: We don't have IFU. So could the agency provide that for us, please?

DR. KRUCOFF: Okay. Obviously under contraindications other than the body surface area issue, none of the other relative contraindications that are being requested today --

CHAIRPERSON LASKEY: Yes. The key issue on the table is the warnings and precautions.


CHAIRPERSON LASKEY: So we just need to see that.

DR. KRUCOFF: Thank you.

CHAIRPERSON LASKEY: All right. I'm going to move ahead while we find this information.

DR. BERMAN: Sorry. Could I just have a brief comment from Dr. Oyer on that same question about included or contraindicated? Phil?

DR. OYER: I'm Phil Oyer from Stanford. I'm a cardiovascular surgeon. Conflict statement would say they paid for my trip today. I'm not a consultant and own no stock in Novacor.

MS. WOOD: Bring the microphone closer. I don't think they can hear you.

DR. OYER: Okay. As far as the conflict statement goes, they did pay for my trip today. I'm not a consultant and own no stock in Novacor, World Heart, although I did at one time serve as a consultant in past years.

With respect then to the business about contraindications in the labeling, it does say you have to be a transplant candidate. Presumably that means at the time. And the whole point we're talking about today, as many of these patients who are in the group two in fact would not be considered transplant candidates at the time. So, you know, I think in actuality it would be a violation of labeling to put patients like this into an LVAD.

And the other point to be made I think as far as listing them status 2, that's probably a fair clear violation of UNOS guidelines if in fact you don't intend to transplant them during, you know, if a donor does become available.

DR. KRUCOFF: All right. But let's stay in focus. Because the fact that those patients were in the BTT trial meant that somebody considered them a transplant candidate. And the fact that somebody else might not consider them a transplant candidate is actually a different issue. In fact, the reality is that the patients who are actually the source of the data being discussed to support this labeling change, were all patients who were enrolled in the BTT trial listed as transplant candidates.

DR. OYER: That's true, they were. But I think we're trying to give the opportunity to patients who might be appear on the doorstep of another center who would not today list those. And even in today's world, many of the patients that we, and Jim certainly talked about at his center, would not really list today for a transplant on the day that we saw them or evaluated them. So that's the whole population I think we're trying to address with this effort today.

CHAIRPERSON LASKEY: We understand that, and we've spent the better part of two hours trying to articulate exactly who these folks are. So I'd like to just finish up before the lunch hour with Dr. Somberg's review, if you would please, and then we'll break for lunch.

DR. SOMBERG: Thank you, Warren. You put me in a difficult spot being what's between everybody and lunch, but I will try to deal with being in a difficult spot.

I have a detailed review which I will give you a copy of for the record.

I think that it's very important for the panel to realize what it is being asked to give advice on to the division to change the labeling, and what we are being asked to give advice in my estimation are two considerations. One consideration is whether we should change the current wording from an indication to use a device to an indication to use the device both for short and for long; so essentially the addition it for a long term indication.

What do we have to base this on? Well, unfortunately, in the packet given to the reviewers we have statements to the effect that there is 20 years experience with the device and over 15 implants. But we've heard today, and it was my inclination from a detailed review, that we really do not have information on 15. What we have is a completed BTT trial with the device. And this completed trial compares actually the 35 control patients with the 190 device recipients. And that what we know is today is that while the trial was completed, that it was completed with a control group that preceded from another trial the intervention with the LVAS device group. And that there are a number of severe problems with the control group, as pointed out by the statistical reviewer for the FDA in the package and the presentation today.

The groups are nonconcomitant when there is a lot to give us consideration that being concomitant would be important. Sometimes there isn't in certain studies, sometimes there is. And here there's a lot because there's a constant change in what we do for these patients. So what was done in '91, '92, '93, '94 is potentially -- not is, but is potentially significant from what was done in '97 and '98 and in the latter part of '96 as well. So that worries me considerably.

There is a number of suggestions in the data, very hard to determine but there is suggestions that the control population was considerably sicker. And that's why there is a 7 day average survival as opposed to the prolong survival in the other population. One can argue well it's the device that makes the difference. But that's to the crux matter, we really don't have anything to support the validity of the control data.

As a reviewer, I would have most appreciated further assistance in this by the sponsor by looking to other publications of control groups in the area. While you only have 35 patients in the BTT trial, there are lots of patients awaiting transplant who never get a transplant who don't have many different interventions of this nature who have received and could be looked at to substantiate that.

Now, people will say well that's not randomized, it's not appropriate, etcetera. But if I saw that there were five, six, seven groups of 35 patients all with similar outcomes, maybe a little smaller groups, maybe a little larger, that would have swayed me in one way or the other. I saw nothing to support that, and no attempt to do that which I believe is devastating in terms of being able to make a decision on whether long term is adequate.

so what we have here essentially are at the completion of the BTT study with 30 patients for six months and 15 patients for one year, and without a control group to compare them to. We, obviously, know that this device causes a significant number of problems. They're more frequent in the up front than later on in the course of treatment, but we unfortunately don't have anything to base knowing if there is a risk benefit ratio because the control group is so inadequate for comparison.

So, I do not believe, and my detailed review I believe supports this, is that we have data to extend the indication from what was originally given in terms of approval. And that long term, the data is too small and inadequate control group.

In terms of the bridge to transplant, this is a very interesting concept. It certainly of consider that people with relative contraindications could with some sort of further assistance then become transplant eligible and be appropriate. And I believe the selection of the parameters, as we've already heard, has been done on an arbitrary basis and the numbers of people with each given relative contraindication very small and making the cells very hard to compare. But I do think there are constraining transplant lists and there is tremendous difficulty in deciding who gets a heart and who doesn't for transplant, and thus the concept does sound appealing to me. And I would, unlike some panelists possibly, accept this idea even though it was arbitrary and even though there was little justification presented in the handout for why these considerations were made. However, once again, there are substantial difficulties with reaching any positive conclusion here.

One is that the control population is now 12. It once against is none nonconcomitant. It once again is inadequate for a comparison and once again we have no further historic controls from any other database to try to validate why we should base this major consideration and recommendation on the data here. So we have no further substantiation.

And finally, the most disturbing conclusion I have to make is there is no evidence from the data that I was presented with to review that implantation of device changes these relative contraindications such that it would be more likely to be able to receive a device. We have survival on transplant outcome compare and that really doesn't tell me very much. It just tells me in 1996 the latter half '97 and '98 that it was more likely if you got the device, to get a heart than it was if you didn't the device in '91, '92, '93 and '94 in this very, very small group.

So, yes, it is of concern to me that the current labeling does not advise physicians what to do in patients who have in their mind a relative contraindications to transplant in terms of implanting this device. But it also very much concerns me that we do not have the information to recommend to anybody that if you do put in this device, A, B or C will happen and therefore you will have a better, a worse or the same outcome as if you did other things.

So, yes, Doctor it's bad to do harm or it's bad not to recommend something and do harm, but it's also bad to recommend something and do harm as well. So really I think what we have here, unfortunately, is a very reliable device with very little information on how to use it for these two questions we are asked for: Long term therapy and therapy when there's a relative contraindications, arbitrary maybe, but still a relative contraindication to transplant. And, thus, we really can't recommend what we should do.

And that really is a summary of my more details review that I will enter into the record.

Thank you.

CHAIRPERSON LASKEY: Did you have any queries for the sponsor?

DR. SOMBERG: No, I didn't. They have really been addressed.


Well then, thank you both. Thank you sponsor and FDA.

I suggest we break for lunch. And I'd like to resume at 1:00, it being a quarter to 12:00.

Thank you all.

(Whereupon, at 11:40 a.m. the panel was adjourned, to reconvene this same day at 1:00 p.m.)













1:05 p.m.

CHAIRPERSON LASKEY: Okay. I'd like to reconvene, if we may. And we'll proceed with the open committee discussion. We've already heard comments from Drs. Krucoff and Somberg. And I'd like to just go around the table and give the other panel members opportunities to query the sponsor for things which have not had enough clarification.

In addition, I know the sponsor was asked to provide some material this morning, and they've informed me that they have. So we'll allow a little bit of time for the presentation of that information.

Having said that, if we can being with Dr. Aziz. And I'd like to in the interest of efficiency and keeping us all on schedule, just confine each speaker to ten minutes, and I'll be watching.

Dr. Aziz, thank you.

DR. AZIZ: I'll try to be brief. I'll try to address most of my questions in the surgical arena and leave the statisticians to quibble over the statistical aspects.

I do realize we have Dr. Oyer over here who I think most of you may not realize I think did the first sort of successful transplant using the Novacor device in '83 or '84. So I think that sort of I think set the stage for these sort of rather terminally sick patients.

Let me ask you a couple of questions, maybe I could ask Dr. Oyer if he doesn't mind coming up to the podium there.

You know, for surgeons who have these patients with elevated pulmonary hypertension, clearly that's one of the risk factors for heart transplantation, in patients in whom you have to put the LVAD either as a bridge or to try to get the pressures down, what is the incidence both in your experience in those patients intra-operatably, for example, having right heart dysfunction or failure?

DR. OYER: Well, over the years we have had very few of those, in fact, with the Novacor device, at least because it unloads the left side so well, reduces the left pressure so well presumably.

I have I think only in one circumstance put an RVAD in that was a short term biomedic, so I'm not sure how many, probably 75 or 80 over the years. So I think with -- you know, in the earlier days we had other drugs -- you know, nitric oxide wasn't here. We used Prostaglandine E for a while, and it came out in the late '80s or so. But I think we've gotten away in general with managing those patients pretty well. In most cases with drugs over the years we've been able to bring those pressures down so that we've not had to use RVADs except I think one patient.

I think we may have had one other patient die of right heart failure bona fide post-Novacor that we didn't, for one reason or another, put a device on the right side. So it's been fairly limited in our experience, the need for RVADs that is.

DR. AZIZ: I mean, not only in your experience but maybe in the literature, in patients who have so called fixed pulmonary hypertension.

DR. OYER: Yes.

DR. AZIZ: You know, from what I remember of patients who have elevated PA pressure, I mean I can recall having patients who have been in the ICU for a year using various type agents and eventually the pressures came down. I can also remember putting RVADs in patients who had pulmonary -- with very high PA pressures and putting an RVAD in those patients wasn't very helpful. In fact, you know, you would get bleeding out of the AT tube. So to me it seems that even putting an LVAD in patients who have elevated pulmonary pressures, you're taking a risk. I mean, they're not like you're putting an LVAD who somebody who is just having hematein and they compromise. So you are taking a higher risk group of patients in doing so, and I think it's remarkable that the problems that one sees is not as high as it probably could or should be.

What I see from a surgical point of view, I mean we have really clear indications. You have patients for LVADs who are having hemodynamic problems and who would be transplant patients without the relatively increased risks. And then you have patients who clearly are contraindications for LVAD; infection or malignancies. But the group in the middle to me that seems to be a moving target. One is the drugs improve; nitric oxidic maybe the receptor blockers, you know, that's really that have been shown to be -- the like help patients with pulmonary hypertension. So this group of patients, I think, where we are not may not be where we will be in two or three years time.

What I'd like -- obviously, you have a great experience in dealing with the high risk patients. And what would you advise centers that don't do a lot of these sort of cases if, let's say, the indication was given that this device should be used in patients who have pulmonary hypertension of variability that hopefully that would be reversible. Do you see centers that don't do many transplants using this for these high risk patients, and do you envision let's say more problems in centers that do that?

DR. OYER: You know, I think at the end of the day it's going to be the judgment of the surgeons at those local centers and cardiologists. I think I agree with you entirely, though, that as time has gone by we've had ore and more drugs that will allow us to separate out which ones are going to have a reactive pulmonary vascular -- from those that don't. Be that as it may, there are still some patients that it's a dilemma and that we can't, you know, get those pulmonary artery pressures down enough to make us comfortable. But I think, you know, certainly if they've got grade 4 pulmonary vascular changes, some of those will not come down. And as we heard from a couple of people today, we can't always predict that. But I think that's not a reason probably to not go ahead. I mean, that problem is no different than the problem that we face with putting an LVAD in in the first place. You can't guarantee that they're not eventually going to fall off the transplant list because of a complication of the LVAD or whatever.

So I don't think that's a unique problem. And I think if -- you know, and a direct answer to your question, a center with a small number of patients per year and they put in LVADs, assume they've got enough experience to do those. I don't think, you know, a center doing three or four a year and one LVAD every five years is probably appropriate even to be doing LVADs at all, for example. But I think, you know, if assuming they have enough experience putting in LVADs, then I think it's not unreasonable to suggest that in a patient if they encounter that has pulmonary pressures and they can't be comfortable in how well they can get them down if they were to transplant them at that time, then it's not unreasonable to suggest that an LVAD would be appropriate to see. And you saw the data that we have showing that in the majority of cases those pressures do come down one way or another.

DR. AZIZ: I guess there must be some patients where you put the VAD in and the pressures don't come down. I mean, what happens to those patients? Do they become part of so called destination therapy because they can't be transplanted?

DR. OYER: Well, I hesitate to destination therapy, because that tends to confuse with bone fide destination therapy, patients that need a separate set of criteria. But I think at the end of the day, yes. If we encounter a patient whose resistance has stayed up, then at the end of the day we would not be able to transplant them and they would end up being a long term -- longer term -- whatever the term you want to use would be.

DR. AZIZ: Okay. Thank you.


Dr. Hirshfeld?

DR. HIRSHFELD: I'd like to ask the World Heart representatives to comment on just the context in which this requested indication exists. And in particular, I would like to hear comments about the relationship of this indication requested to the issue of destination therapy.

You indicated that you're embarking on a destination therapy trial now. But it's not clear to me, and this is what I would like clarified, as to what the role of seeking this indication is in terms of the actual impact on clinical practice if it's not in fact to open the door to people who would ultimately become destination therapy patients?

The reason I ask this, and this is what I would like to comment on, is that it seems that the strict request that you've put in and the strict language is actually well within purview of current accepted clinical practice that patients who are covered under the strict definition of your request are patients who current transplant cardiologists could legitimately decide could have this device implanted in a bridge to transplant mode. And so what I'd like would be for you to clarify the relationship of this request, why this request is important, what it offers the transplant cardiologist and how it relates to the ultimate goal of seeking a destination therapy indication for this device.

MR. BRYDEN: I'd like to respond to part of that question and then ask Dr. Edwards if he would mind giving you more of the clinical response.

The context of this request is that we filed nearly two years ago a request to the FDA for a PMA for destination therapy based on a Basian based statistical analysis model which included that data from around the world to the extent that there was auditable and reliable data there that included North American data which was outside the trial and it included the bridge to transplant data as well. And after very considerable work with the FDA and work by the FDA, they concluded that they were not satisfied that without a prospective randomized trial that they were prepared to approve an indication for destination therapy.

During the course of that process, however, we concluded that should we be successful in a destination therapy label of exactly the same as the one that is currently -- was then and still is currently in place, that there was a group of patients that following the rules would be neither and analyzed ourselves where will those patients fit. And it was that that caused us to come back to the FDA rather than simply withdrawing our submission and replacing it with the destination therapy submission that we proposed to the FDA that we would submit a request for and ultimately have receive their conditional approval to proceed with a randomized trial randomizing our product against heart mate for destination therapy. And that we would proceed with a very much more specific request for expanded bridge to transplant indication, which is how this particular request arose.

It is our view and I believe the view of the doctors who are speaking with us today that a significant number of patients who would be or could be assisted as a bridge to transplantation while not yet a candidate are either not receiving a therapy or being listed as a candidate at a time when if a heart were available, it would then not be implanted because the patient is not then in condition to receive it. And while we recognize that the process of deciding who is a candidate is left intentionally by the regulators, both CMS and FDA, to the individual clinics, once those clinics have established their procedures then the standard that they are held to is that they administer in a consistent manner their own procedures. And in many cases the administration in a consistent manner of the criteria that are established in the transplant clinics would not list a patient who was not at the time of listing ready to receive a transplant.

So our focus is a relatively modest number of patients, but a group of patients who in our view, if and when we are approved for destination therapy, will still require that if they're going to be served it will be by, as Dr. Young observed, being ingenuous; that is either listing them as a candidate when they're not yet truly listable under their own criteria or treating them as destination therapy when the real intention is that after being supported for six months or a year or whatever, they're going to get transplant which is not the intention of destination therapy. And a candidate means someone listed for transplantation.

So that is the basis, the background of this indication. We believe it is material to the patients who it will effect. It is material to us at the margins. It will increase the theoretical population available by some modest number, and of that some share of those may in the next three or four years actually find their way into a device use that would have otherwise not have done so. But it is a gap in our view in the approval process at this moment, and one that we have the opportunity and the data to support.

So that is why we are doing what we're doing, and we by all means intend to pursue as aggressively as we can the reliant trial and expect ultimately to be approved for destination therapy. But it will not capture those patients if they are accurately represented within the rules that most clinics apply to their own selection process.

Dr. Edwards would like to comment further, if you don't mind.

DR. EDWARDS: Thank you very much.

I'm Brooks Edwards. I'm a cardiologist, Medical Director of the transplant program at Mayo Clinic.

And it's been an interesting morning hearing the discussion. I appreciate the thoughtful consideration that the panel is obviously taking.

As a clinician I'd like to present a view that is really patient centered and not based in statistics. And coming from Mayo, we have a lot of history and adages. And there is one adage from Will Mayo himself that is quoted frequently, and that's "The needs of the patient come first." It may sound trite, but at the end of the day statistics aside and everything aside, that's really how we make decisions; the needs of the patient come first.

The dilemma here is that sometimes the needs of the patient and the labeling indications are at conflict. And what do you do when there's a conflict between the approved indications and the needs of the patient? And we really get back to the old adage: The needs of the patient come first. Ethically we have no other option. But it does put the physician in a compromised position to propose off-label use of the drug or device for a patient when you really believe that that's the best therapy for your patient.

And what I want to do is quickly tell you about one patient that I've been caring for the last several years, a fellow --

CHAIRPERSON LASKEY: Very briefly, please.

DR. EDWARDS: Okay. Forty-eight years old, my age. He's got a son in middle school, as I do. Severe dilated cardiomyopathy despite aggressive, best practices, all available therapy, he had 16 hospitalizations in the 6 month period. And it was clear to all of us that this fellow was not going anywhere but down and he was going to die. He was an ideal transplant candidate except for one problem, he weighed 315 pounds. And in our center we won't list somebody who weighs 315 pounds.

We could propose destination therapy for him, but that really wasn't what we wanted. What we wanted is bridge to candidacy. We wanted to put a device in this man to bridge him long enough so that he could lose weight, either with surgery or with conventional weight loss mechanisms. But that's the patient who falls between the cracks with the current indication. He's not a bridge to transplant because he's not a candidate right now. He's not a destination therapy patient, because he's 48 years old and I told him a destination device is not going to let him see his kid graduate from high school. What we really want to do is bridge to candidacy.

And so I think this is the kind of discussion that at the end of the day if you go back to the needs of the patient that's the indication we're looking for.

CHAIRPERSON LASKEY: This body entertains both the clinical needs as well as the scientific needs of the process. So we appreciate your input, but we are clinicians at heart as well. We wear other hats up here.

Dr. Weinberger?

DR. WEINBERGER: I don't have much substantially to add to what's been said, other than the feeling that what this labeling change will do is basically open the back door to use of the device as destination therapy. And it's very hard for me not to feel that way.

If a patient with a creatinine of 5 who you know and I know is not going to recover would, according to the new labeling indications be eligible for the device. And if three three or six, nine months from now has not turned around, what's that patient supposed to do? Anyone from the sponsor can reply to that?

What is the game plan here for patients who fail to respond to device?

DR. YOUNG: I'm very sensitive to that issue for several reasons. Number one, I'm a big believer in bridge to transport therapies. I think that's where the data gives us the greatest information about success. And I think I'm a qualified believer in destination therapy, given the information that we have and I think things are getting better. But I want to reiterate the comments that this is absolutely in no way to be construed as trying to open a back door to destination therapy. This is trying to help us clinicians do the best job that we can do for our patient.

And I can tell you that an individual that I knew wasn't ever going to be a candidate for transplantation is not the individual that I would recommend this device put in. So somebody who is a diabetic with chronic renal failure and we know has creatinine clearances that are 30 or less that is heading towards end stage renal disease and all of the implications therein would be looked at quite differently with respect to these devices than would somebody that we feel has flow induced difficulties.

You also alluded to one point that you sometimes can't predict who is going to get better and not. And I share that and I an frustrated by that fact. But this is not an attempt in any fashion to try to get a back door into destination therapy.

We'll have our trial that shows the worthiness of this particular device with destination therapy, and that trial many of us clinicians are very committed to doing and to completing. And it is not the same thing as this request.

DR. WEINBERGER: But the data that you have, it's hard for me not to get my mind around this, came from heart failure specialists who felt the patients entering were candidates for heart transplantation.

DR. YOUNG: Well, I have to go back again to comments that we had made earlier about the historic time period that these clinical trial, this particular clinical trial was ongoing and the evolution of data that has occurred since that time and the refinement of the process. Not to mention the changes in organ allocation that have occurred and also the divergence from consensus that has developed about many of these patients.

DR. WEINBERGER: So you're saying that back ten years ago patients who were acceptable as heart transplantation candidates are no longer heart transplantation candidates today?

DR. YOUNG: Many times that is the case, yes. And in other situations many cases that we didn't feel were acceptable for transplantation, we would feel would be acceptable today. And the age criteria I've already addressed a little bit.

But the other thing is, is that I do know as I termed it some disingenuous listing occur. And we know that from data that comes back from the number of refusals that UNOS tallied in individuals where organs are offered but turned down because the patient is in fact too ill.

DR. WEINBERGER: Okay. The last point is a technical point about the trial. When a patient gets a device, that patient is immediately UNOS 1-A or do you want to make him UNOS 1-A until they recover from the operative procedure?

DR. YOUNG: Right now we have three decisions that can be made. We can take 30 days of UNOS 1-A allocation, which we can activate at any time course in the patient's post-VAD placement. That's one choice.

In the patient that has no relative contraindications who the device is going in for hemodynamic stabilization, we might start the clock ticking the day or two after surgery if everything is okay.

The second choice, as Dr. Pina pointed out, listing the patient as status 2 hoping that you won't get phone calls with organ offers. Or the third choice is making the patient a status 7 where you will not get any offers until you activate the patient as either 1 or 2.

DR. WEINBERGER: What was mandated in the your BTT trial?

DR. YOUNG: Well, none of these because at the time that the BTT trial was ongoing these allocations schemes were different at that time and the standard operating procedure would be to turn down an organ if it were offered if you as the clinician didn't feel the patient was an acceptable candidate at the time.

Unfortunately and one thing that I would have loved to have done, I can't get you information about the patients that are in this clinical trial and the number of organs that were offered and turned down. What I can say from the UNOS data and also from the transplant advisory committee that has looked at this, that this is a big issue that concern has been raised about.


DR. LINDENFELD: I have just two questions.

The first is in these relative contraindications I'm having a hard time understanding what the LVAD will change about age to make the patients a candidate for transplant and why that should be on the list?

DR. YOUNG: Well, it won't per se, and as we addressed, unfortunately the device is pretty good but it's not going to last that long to make him regress in his age. But, as you know, age is a multivariable sort of factor. If you have somebody with a creatinine clearance in the 30 to 50 range and the patient is 65, you might look at that patient a heck of a lot different than the patient with a creatinine clearance that's the same, but the age is 40 or 45. So for that reason I think age is something that should be looked at and should be included, although it is certainly a relative factor, relative of the other things. God forbid that you throw in pulmonary hypertension, diabetes and a few other issues in a 60 or a 65 or 70 year old patient. That 65 year old patient rapidly looks older with all those other things. So that, to me, is why age is still a relative issue.

DR. LINDENFELD: And then a second question which comes to the issue, it's nice to know that the patients with one or more relative contraindications had nearly as many transplants as the total BTT group, 70 versus 65 percent, I think. But I think what's more important to me is that I understand that patients who had a relative contraindication didn't just get the transplant but had a similar survival to the whole group.

DR. YOUNG: Right.

DR. LINDENFELD: So I think what I would like to see is some sort of one and two year data on those two groups about total survival.

DR. YOUNG: Right.

DR. LINDENFELD: And to make sure that the survival that these creatinines don't sort of cause down the line one year problems. And we know pulmonary hypertension leads to -- so do they get transplanted but do they have a substantially worse outcome? Do we have that data? And it's hard for me to evaluate this without seeing that?

DR. YOUNG: We do have that data, and some of it is in the packets. Because the BTT trial, the primary end point as we discussed was 30 day post-transplant survival, we have that data out to 30 days. And we do have one year data out that shows that the survival rates were similar in both the groups.

DR. LINDENFELD: With and without a relative contraindication?

DR. YOUNG: With and without relative contraindications.

DR. LINDENFELD: I think it would be important to see that. If we're going to encourage people to take these marginal patients, we ought to have data that that's a wise thing to do.

DR. YOUNG: Do you have one year data?

DR. LINDENFELD: I mean for the 115 with no relative contraindications and the 75 with one or more?

DR. YOUNG: Right. Yes. We don't have that. We have it for the total, for all of the patients that were followed out --

DR. LINDENFELD: See, I find it -- again, and this is my problem with this problem with this data, but I think it's a critical problem in that we're taking patients who we've said people might be worried about with relative contraindications, and the only data we have that that's okay is that they get to transplant. But we don't know that the one and two year survivals in those groups with and without contraindications are similar. And I would like to be reassured about that.

DR. YOUNG: Well, we can get you one-twelfth of the way.


DR. YOUNG: At least a one month data.

DR. LINDENFELD: Well, but one month is not -- you know, not all of them have been transplanted even.

And then I guess the other part of a similar question that I have is we saw adverse events for the whole group and we saw that they tailed off early on. But I guess what I'd like to be reassured is that the patients without a relative contraindications and the patients with one or more relative contraindications had approximately the same number of adverse events. In other words, you're taking a high risk group.

DR. YOUNG: Right.

DR. LINDENFELD: And what kind of adverse events, what kind of hospital days do we see in this group with one or more relative contraindications? And I think those two sets of data are critical for me wanting to encourage people to expand the indications.

DR. YOUNG: Right. I think the best answer to that was that one slide that we showed where we took the no relative contraindications and set that mark at the 100 percent level and then the relative event rates that were occurring up and down with the confidence intervals. The only thing that I can say is that the wide confidence intervals created a nonstatistically significant interaction between those two groups, though numerically as you might suspect, the patients with relative contraindications did have more events. But getting to transplant was equal or seemingly equal. Similar, I guess, would be a better statistical term in the two groups.

DR. LINDENFELD: And I think that's good data, but I still have trouble. If we're going to encourage these potentially marginal candidates, we want to be sure that the ultimate thing we're aiming for, which is transplant --

DR. YOUNG: Long term survival.

DR. LINDENFELD: Post-transplant long term survival is equally as good. And I would think that that data is available. And with that, just how many hospital days, how much bleeding, those kinds of things in the two groups is critical for me for evaluating this data.

And that's all I have.

MR. BRYDEN: I'm sorry. The data to one year is available. It hadn't been broken out, so we don't have it in a file that we can actually access at this moment. Should there be a view that would result in further discussion with the FDA, we certainly can provide that and can provide it to the panel. But we do not have interactive access to our database back in Oakland that we can do it at this moment. We do have it for 12 months.

With respect to adverse events as well, those were summarized and the details of that summary can also be provided. But the one slide with the summary of adverse events did show that there was no statistical difference, although slightly higher levels of adverse events in the--

DR. LINDENFELD: Well, let me just apologize if I don't recollect that slide properly. That was the differences between the two groups in a whole bunch of different events. It wasn't the sum of all the events together. And I would say that what you would want, first of all, the serious adverse events and you wouldn't want to just say -- because the numbers are small. One adverse event comparing the two groups is not likely to be statistically significant, but you'd want to pile all those up and say, okay, were serious adverse events substantially more common than the group without a relative contraindications versus those with.

And I think with the numbers you have there's no way that each individual one is going to be different between the two groups. So we need to see a summary of that data.


DR. LINDENFELD: And hospital days would also be, of course, very valuable.

MR. BRYDEN: That is not difficult to do, and we'll be happy to provide that.

CHAIRPERSON LASKEY: I mean, at risk of exaggeration, that is efficacy and safety right there, which is something we desperately need to see.

Dr. Bailey?

DR. BAILEY: I'll try to keep brief here because a lot of points have been discussed.

Obviously, as a statistician it's always nice to see randomized data, and I've heard the arguments that this can't ethically be done. I guess I'd just like to keep that idea alive. If there's any way of changing the end point or some way of thinking harder about that, because I think that's the best way we get good data.

I mean, a lot has been said about the comparability between the two groups here. And it's laudable, you try to do everything you can to adjust for differences, but ultimately one group, the LVAS group one tends to think there may be a healthy volunteer affect if you were doing a clinical trial. With the LVAS group that may well be operating. With the other group, at least there's a subgroup who refuse to get an LVAD. So it isn't that they agreed to accept medical therapy, they refused participation or at least refused the LVAD. Obviously, they had to agree to participate in the study, I presume. But one is worried that there may be a healthy volunteer effect that's operating in one group and not the other.

I'm sort of beating a dead horse here, obviously. But just to get back to the idea that we need -- and I responded very well to Dr. Somberg's point this is one control group, but once you get away from randomized data, you know, is it enough to just look at one control group. It really behooves us to get every possible other source of data that might be more, perhaps, current in terms of being able to compare these outcomes.

Obviously, again, I'm sort of going over old territory here. But I think it gets to the point that once you've accepted the comparison between the two groups, as Dr. Ahn pointed out, it's going to be very difficult to find a subgroup that does as badly as the control group. So it's sort of a foregone conclusion that you're going to get a significant difference.

So we're left with I think then, okay, how good are the data to allow us to extrapolate the good outcome in the overall group which were patients who were listed for transplant, but now let's see if we can find a subgroup that really shouldn't have been listed for a transplant and maybe that will allow us to see if this device would work and have similar results in patients that are contraindicated. And I think there then we're being asked to believe that people who slip through the cracks and were actually listed for transplant but happened to have one or more criteria that technically should have kept them from being on the transplant list are equivalent to a group that nobody ever listed for a transplant. And the problem is that the criteria that are violated may be violated more seriously in the people we're trying to extrapolate to than the people who slipped through the cracks, so to speak, into the study.

You can't prove that's true, but what are the different possible explanations for the lack of a gradient in a survival outcome when you start adding these contraindications? Well, one possibility is that the LVAS is so good that it keeps people alive even that have these contraindications. But another possibility is simply that we don't have enough power because we're so near the fringe here that we don't have really have enough variability of these characteristics to be able to safety extrapolate the results.

So I think we're nervous about extrapolating from people that are sort of on the borderline, on the fringe, to the vast group of people who also have contraindications but maybe they have three or four of them. I mean, just because the words one or more contraindications applies to the group that was studied and the group we're extrapolating to doesn't mean they have the same number of contraindications or that they're violated as severely.

So that's where I get nervous is trying to extrapolate the data. So I guess I'd like to hear more about why it's unethical to do some form of a randomized trial here. Perhaps one could even look at functional status as an end point and have, perhaps, the ability to receive an LVAD as a backup option in those who are randomized away from it, for example.

DR. YOUNG: Well, those are extremely important points and I think queasy about control is what we're talking about here a little bit. And I share your queasiness in many senses. All of the warts of a database inquiry such as this are present and the statisticians have done a great job of pointing those out, and I certainly agree with many of their points. This is, however, I think as I indicated in my talk, an unusual and unique database. It's a very large database containing patients with a variety and spectrum of difficulties prior to transplantation who were extraordinarily ill, first of all.

And if you look at the unfortunately small control group that is present, it is just that: It's a control group. It's not a randomized control group. It's not necessarily a concurrent control group and it suffers from all of those limitations. However, despite the improvements in therapies that have occurred over the decade and the time period, this represents a group of patients that I submit really don't have many more choices. If you look at the trial entry criteria, having to be on two inotropes or balloon plus an inotrope, the proportion of patients that have had a cardiac arrest being so high prior to that; there was reasonable proportionality between the control group and the LVAS treatment group. And I think the death as early as it did, bespeaks the severity of illness that these patients have.

Now, I'm not one to abandon clinical trials easily. I look at myself as a clinical trialist. But I am one to support the concept of equipoise driving a randomized clinical trial. And I don't have equipoise in this type of patients.

I want to do something, and I feel that mandates. And so that is why I believe it's not ethical to do a randomized prospective trial in this type of patient population, and I believe we have enough support to say that an expanded indication, which is a hugely expanded indication to destination therapy for sure and isn't intended to be a back door, I think is appropriate.

DR. BAILEY: Well, what is the average number of contraindications in the patients you're trying to extrapolate to?

DR. YOUNG: The average number? Well, we had as I indicated, 18 that had two or more or three that had more than that. So the average number would be a little bit over one.

DR. BAILEY: In the current data. But what I'm wondering is it equivalent in the people that you're trying to label it for?

DR. YOUNG: I actually think there's more contraindications that we're dealing with today from my clinical perspective --

DR. BAILEY: More in the current dataset or more in the patients you're trying to get the label for?

DR. YOUNG: Yes. More than in the current dataset is my impression.

DR. BAILEY: Well, see, that's the problem. I mean, these patients are the fringe; that was my argument. You're trying to extrapolate the people that have much worse problems.

DR. YOUNG: Well, I'm not sure I would characterize it as being much worse problems, but I'm seeing more combinations of renal insufficiency and pulmonary hypertension, for example, that was in this dataset. And I don't agree that that diminishes the impact that ventricular assist device therapy can have in this group. I actually think that dropping pulmonary artery pressures and increasing renal profusion work together in many senses to improve the patient.

So I'm comfortable with the concept that this will in certain patients turn around difficulties that make them a queasy candidate, if you will. We talk about queasy controls, we can talk about queasy candidates for transplantation.

DR. BAILEY: Just a small point here. I guess in terms of adjusting, it's always very difficult to find the right -- well, there probably isn't any right adjustment model. But I guess I would argue for including all of the covariates rather than trimming them down just because you're trying to remove bias rather than variance. So in other words, I'd rather see some of those variables in there such as -- I guess prior cardiac arrest may have been in the model, but prior stroke or TIA, valve surgery, etcetera. These were things that were not nominally significant as predictors or as differences between the groups, but put it together and I think there's bias there that's not in your adjustment model.

BMI should be looked at not as a linear covariate, and I didn't see anything to the affect that you tried it as the two -- looking at it as a U shaped curve. In other words, cachexy as one end of it and then obesity as the other end. I just saw it in there as a linear term.



DR. TRACY: Just a couple of points. This whole thing strikes me as more about clinical practice than a regulatory issue. Maybe I'm being naive here, but it seems to me that although I don't have the full package labeling here, there is nothing here that tells me that you cannot put in this device in a patient who has a creatinine of 2.7. You cannot put this device in a patient has a bilirubin of whatever. I don't see anything that precludes the use of the device in these patients at this time.

And, admittedly, different centers that were involved in the baseline study, some patients would have been considered noncandidates at one center but were candidates at another, so that's the variability in clinical practice that existed even within the study.

So I'm kind of lost trying to figure out why are you trying to regulate my clinical practice? Not that I do transplants, but my point is, isn't this about clinical practice and doing what's right for the patient?

MR. BRYDEN: Well, it is exactly that. And our intention, and if we can capture it properly, I think what would be achieved by the proposal is to facilitate that clinical practice not to limit it.

The discussion just before lunch is somewhat related to your point, that is is it contraindicated. It is our understanding, and it seems to be shared by the commissions we've been working with, that the first decision is whether in this particular case whether the potential recipient of the device is a transplant candidate. It is approved for use by transplant candidates. It's not approved for use by any person suffering from the reverse left ventricular failure at risk of imminent death, they have to also be a transplant candidate. And the interpretation of what is meant by a transplant candidate I think has been quite clear, and that is that they are listed for transplant. That's how you determine are they transplant candidate or not.

So it is not a medical condition, it is a listing on a list. And our point is the whole basis of this analysis is that the listing on the list is variable and is intentionally left variable from center-to-center. And we're not making any comment about the wisdom of that, I think it's probably a fine thing, it just is a fact that from center-to-center who gets listed is not always the same. And that the standards to which the centers are held is, first, that they must establish criteria which must not be patently unreasonable. And as long as they're not unreasonable, within quite broad limits, then they must consistently apply their own standards to make the recipient a candidate by putting them on the list.

What we're saying is that it is the need to make that judgment in a situation where those clinics also and the publications which were referred to by Dr. Young and also referred to, I believe, in some of the FDA staff's papers, make it quite clear that the generally accepted practice is that to be a candidate, the patient should be one in whom the clinic would implant a donor heart if it were available at that time. Not someone who if certain things happen, which may or may not happen at sometime in the future, they would be candidates or appropriate recipients.

In order to allow those clinics to now deal with patients who present themselves with contraindications, which the clinic decides they expect will be recovered, recognizing they can never be sure, but if the clinic decides they may be recovered today they must put them on a list and then give then a devise in order to be within the label. It's not so much contraindicated for this, that or the other reason; they're either a candidate or they're not.

And by requiring that they be put on the list, it requires a judgment to be made which is contrary to the other judgment that that same center would normally make, and that is I put them on the list if I had a donor heart I'd give it to them today, but not so for a significant share of these patients.

All we're saying is that process has the effect of putting sand in the process, of making it more difficult, a different decision for a clinician than simply deciding I think that if I gave this person a VAD, they would probably recover to be a candidate and therefore I'm going to give them a VAD as a bridge to transplant. They can't do that. They can put them on the list, which they otherwise wouldn't have done in order to give them a VAD. Some do, but many do not. And it is that discrepancy that we're trying to address.

The other factor that I hope might be a satisfactory monitor on protection against opening a wide door is, first, it is the judgment of the clinic that this person, this patient, is likely to become candidate, is expected to become a candidate. It's not that they have a contraindications. It's that they're expected to become a candidate. They have to make that judgment.

If were to implement appropriate post-PMA monitoring, one could easily determine over a fairly short period of months and years whether the share of those patients who moved on to transplant was significantly different than the share of patients who were listed before having had that, which would suggest that there's, perhaps, something wrong with the judgment or it's being used too broadly. But I do not believe that what we are suggesting is interfering with medical practice, rather it's allowing the medical practice not to have to be done in the face of violating what is quite clearly their own standards or the regulatory requirement.

DR. TRACY: I appreciate that point. But I'm wondering what the practical impact would be? How often, for example, is a patient listed and transplanted the same day? And alternatively, what percentage of patients who are on a list at some point are deactivated because of something that happens, some reversible thing that happens?

MR. BRYDEN: Perhaps Dr. Young or Dr. Edwards will speak to that.

DR. EDWARDS: Well, as a practical matter, the waiting time varies from region to region. There's some regions where patients if they're listed status 1-A may get transplanted within days or some regions where that could be months.

Your other question was?

DR. TRACY: How often are patients temporarily deactivated from the list? If a heart came along today and I happened to have a fever this day, you would turn that heart down, obviously, and give it to the next person. So I'm off my priority.

DR. EDWARDS: I don't know that that data exists, because I think that often times centers aren't always deactivating patients, but they may pass on the offer for a variety of reasons whether the patient has a fever, or it's a size mismatch, or the donor hospital is so far away that the ischemic time will be long; I don't think that particular piece of data can be obtained very easy.

DR. TRACY: I don't think there's anything else that I need to ask.


DR. FERGUSON: Well, first, I want to thank the sponsors for a very lucid explanation of the problems we have here.

I'm appalled a little bit at the dilemma that we're in, "we," meaning medicine is in as you've described it because we've had transplant around for many, many years. It's a very sophisticated technical procedure done by a variety of very experienced people. And I've watched the UNOS and the way in which the transplant community has worked together, and I think there's probably give or take a few fall of the wagon, but there's never a group of people that have worked more closely together to work on the problem of donor procurement. I mean recipient procurement.

Some of my question really sort of goes with what Cynthia just has said because if I've heard what I think is the truth, or what actually goes on today, a patient who is fairly sick but for one reason or another has a condition that will not permit the group, the honest group, to put him down for a transplant listing and then at the same time he is not permitted to have at least a bridge to transplant device put in, can have other devices put in and so forth, but that truly to me is a -- that's not a crack, that's a big gap in the way in which we practice medicine. And I'm for doing something about that.

I think I was taken a bit by the comment, and I'd forgotten that that was in Les Miller material that the group came up with, and that is if you list a patient, you must a priori if a donor heart comes up, you must put that heart in that patient that day. I'd forgotten that that was in the rules.

This long preamble gets to this question: How often do we follow the rules, which leaves this fairly large group of patients falling through the cracks, as it were, and how many transplant teams either circumvent the problem by listing when they "shouldn't be listing," and how many go ahead and put an LVAD in, one of your LVADs in, when they are not certain that that patient is going to be truly on a bridge to transplant course? Because if everybody followed the rules, there is a large group of patients that are not being served; there's no question about that.

Dr. Edwards, you can answer my question.

DR. EDWARDS: Well, I think you've identified the crux of the argument. And I don't know that we know the number or really have a sense.

I know that there are certainly many patients, often time young patients, who don't fit the criteria. And as I said earlier, one has to from an ethical standpoint say what's the best interest of the patient. And if the treating clinician believes that an LVAD is in the best interest, I think that's where most centers would go. But it can be an off-label indication and we do off-label indications of drug therapy and devices. But it's particularly difficult in devices where a third party player looks at the labeled indications and say this was an off-label indication.

But at the end of the day there are patients when you put the device in they are not device in they are not transplant candidates strictly speaking and your goal is not destination therapy. Your goal is bridge to transplant or bridge to candidacy.

DR. YANCY: I wanted to focus my comments just a bit with some observations and then just raise one area of questioning.

I am very sensitive to the statement that Dr. Edwards especially made that we should be patient centered, because I too am at the bedside. And when people are within days to weeks of life, it's very difficult to reconcile the need for statistically purity with the need for clinical immediacy. So I completely appreciate that.

In my judgment, the current indication does in fact allow for this patient population to have its needs met. I think that the current statements that have given a certain degree primacy and respect to our listing criteria as controlling the behavior of transplantation in this country would be by the admission of all involved in this business much less clean than we've made it out to be. That's not a statement that any of us are proud to make, but it's a reality that we have a very dynamic environment with regards to listing criteria. It's very flexible. It's open to wide interpretation on a regional and individual basis. So I don't know that that is sufficient leverage that we specifically need this in order to move forward.

So here is where I'm going with my questioning. As has been pointed out, 65 percent of the 75 patients who received VAS with relative contraindications did go on to get transplanted. And I have anxiety about that group, because I would hate to think that that group's needs are not met if they are denied transplantation. So to that extent I think that there is something to be said about embracing it. But I'm focusing right now on the 35 percent who did not make it to transplant but have the VAS in.

We can quibble with the phraseology, but these patients fall into a chronic LVAS support. I don't want to use the word destination therapy, but a chronic LVAS support mode. If that is the case, then we need to raise the question along the lines of efficacy and safety what are the outcomes for that group?

Now, it could be that the relative contraindications enrich this patient population to the extent that they don't do as well as we'd like, but looking at the numbers that are in our packet, four patients are alive the two years. And that would suggest that of the 26 that did not get transplanted, that's about a 15 percent survival, which would appear to be below threshold. Please correct me if those quick calculations are incorrect.

So when I look at 15 percent survival to two years, and then I reflect back on table 8-1 that we haven't addressed today, which is a summary of adverse events that are related to VAS implantation, and I'm especially struck by the 41 percent neurologic deficit rate and especially struck by the infection rate, bleeding rate, etcetera.

so I guess what I need to have some clarity on and the one area where I would like to focus on is in those patients in whom the hypothesis is not realized that they are not in fact stabilized to the point that their relative contraindications are improved and are reversed and they do in fact end up with chronic LVAS support from the sponsor's perspective, give me a sense of your interpretation of their event rate, their survival, their quality of life. Help us to understand if this is in fact a reasonable option. Because as we wrestle with the question of wanting to do for the patient and we all feel that yearning, we also are guided by a desire to do no harm and to provide safe and reasonable therapy even for people who are seriously ill. So I need the sponsor to address the adverse issue, the survivability and quality of life in those people that do end up with chronic VAS support.

MR. BRYDEN: Thank you.

I'll just make a brief comment simply from the stats and then one or more of the doctors may wish to comment further.

With respect to adverse events effecting the patients who remained on the VADs for longer periods of time. The share of the total VAD recipients that were within this contraindications group was about 40 percent, 39 percent I believe. We will give you the specific details, but they can't be available in time for you to make up your mind; so there you go.

When we look at the total of all adverse events, which you had asked be a focus, the rate of all adverse events declines to very low, that is less than 5 percent, after about six months. It drops very quickly after the first month, then again within six and then it stays low. And that low level, given that there was such a significant percentage of that where the group with the contraindications would be inconsistent with a significant rate for such a large share of the total. We will break that out for you. Our belief is from recollection that the data shows not material difference after that first short period of time.

So for those who are chronic users, which was your question, the data will show that there is a relatively low in the sense of a fraction of the first 90 day period continuing event rate.

DR. YANCY: So if that is the case, then what is the reason for the detriment of 15 supported on VAS at one year versus four at two years. If we have not had a logarithmic decline in adverse events, obviously we have lost people unless they are being transplanted late, but that would not be consistent with what's been told. So what's happened to the 15 that have decreased to four?

MR. BRYDEN: The answer is they were primarily transplanted, although --

DR. YANCY: Well, that's what I need clarity on. Because it looks as if, and I can only go by what's here, it looks as if of the 75 patients that had the VAS --

MR. BRYDEN: We do have that specifically on a --

DR. YANCY: -- 65 percent transplant and 35 percent were not. I'm assuming were never transplanted. If you're telling me now that that group that was -- my presumption's never transplanted was transplanted late, that's new information. But I'd just like to understand that.

And just so we can be clear, I'm looking at slide 43 that was in Dr. Pina's presentation and it says 15 LVAS patient greater than one year, four patients greater than two years. I'm assuming that means that we lost patients from year one to two, from 15 to four and trying to understand that model

DR. PORTNER: Can I try and make some comments?

Let me first introduce myself, since I haven't been so far. My name is Peter Portner. I'm a consulting professor of cardiothoracic surgery at Stanford. I should also note that I was the founder of Novacor and involved with Novacor in one capacity or another for more than 30 years. I'm currently an advisor to World Heart.

The specific data you're asking for is available, but I mean we'd have to go back and analyze it. But what I wanted to point out just looking at some information that you have in your panel pack, that the so called group one who are the patients without contraindications form one group at risk and the group two form another group at risk. And if you look at page 11 of section 4, the patients at risk out at 18 months to 36 months are pretty comparable between those two groups.

Now, that doesn't answer the question you asked, but I don't believe that there's any substantive difference in their outcomes, either in terms of survival or in terms of adverse events out at a distant period of time. And I think the time where you would see the greatest impact, at least from my recollection, is in the first month post-transplant where we have the data and have presented it to you.

CHAIRPERSON LASKEY: What page are we on?

DR. PORTNER: This is attachment 5B of the panel pack. I'm looking at page 11 in section 4.

DR. YOUNG: Clyde, if you look at that and trace out the actuarial survivor, the Kaplan Meyer, there were three late deaths that you can see there in the LVAS group. But if you turn to page 31 and 32, that also addresses some of the complication issues, particularly the late occurring complication. And if you look at the linearized adverse event rates on this table and then the following table, you can see that the majority of them occur early on. And it doesn't completely address, I think, the concern that you have raised which I'm sensitive to, but I actually don't believe that this indicates that these patients with relative contraindications actually were set up to have worse post-transplant events.

Now, I agree with you that the database is, as I have characterized it, robust in some arenas, it's paultry in other arenas. And the late follow-up because patients were transplanted, that's why they were censored out of this, is compromised by the fact that this isn't a long term set necessarily. I prefer to use the word "longer." If we have trouble defining long, maybe longer is easier to define in many senses.

The other issue is is I'm also sensitive to your comments about perhaps the labeling as it stands right now allows us to practice what we're doing. And, you know, another solution would be for the FDA to explicitly come out and say that and make that point.

I think there's another issue why I'm sensitive to it, and that comes from my inclinations to study data and databases and whatnot. And when you go back and you look at status ETRD data, whatever, we can't get at this information because we don't know what the true intent was because, in fact, there is not an indication for this. I'll grant people that there's not a contraindications necessary, but there's surely not an indication.

And then finally the concern about branching into other patients who are even more ill with a lot of other complications that then could compromise the outcomes, I think that sometimes hinges on common sense more than anything. And, again, those of us that do this everyday, I don't think that that's our intent at all.

So there's a little bit of data I think that gets to the point and maybe sways you a little bit about this particular issue.

DR. YANCY: Well, the only reason for this topic of discussion is that if the request is to add a longer term use to the indication and if the request is to embrace patients who have a potential for not going forward to transplantation, I think cord to the decision making is to have at least a feel for what are the longer term experiences and in those patients who don't becoming a transplant candidate irrespective of time, what are their expectations. Obviously, we're dealing with a population that is prone to die, we understand that. But we have to understand based on what we've accepted before as thresholds for outcomes in that group, and we have a few limited databases to give us that.

I want to understand are we consistent with or in variance from. And if we're in variance from, is it because of adverse events or what are the circumstances.

MR. BRYDEN: If I can make a very short summary statement on those points?

The data, and in fact there is I believe sufficient summary data in the materials that are in the panel pack and certainly in the material that was supplied as a part of that long PMA process that we're going through to the FDA, to show that as patients live a longer period -- if they have first survived six months, the percentage of those who are successfully transplanted is higher than those who are in the first six months. If they have survived 12 months, the percentage of those that have lived for 12 months that are successfully transplanted is higher. And so on until the last patient has been followed to death or to transplant.

The last patient in this trial was an American who lived 3.4 years on this device and he then got a transplant just after New Year's in 2002 after 3.4 years. So while he's no longer in the trial because he was transplanted.

And the percentage of success to transplant increased at every significant advancing period.

The percentage, whether it is a linearized rate or on a percentage of people at risk of adverse events has declined throughout that period, and remained low although in one case when it was down to only three -- I believe at three years, in one case there was an infection which had the infection rate go up and you see a little blip in the three year adverse event rate.

So I believe that the data shows very consistently an improving success to transplant and a declining adverse event rate as the patient is on it for a longer period.

DR. YANCY: It's just my last statement and then I'll yield to the next person.

There may be one of you who knows this answer very quickly. Of that original cohort of 75 that had the VAS that had relative contraindications, as we stand today how many of those have been transplanted?

MR. BRYDEN: Just a moment, we'll give you that data.

DR. YANCY: Because the number that I thought was in one of the slides was 65 percent.

CHAIRPERSON LASKEY: And that was effective '03. What you want is current. Okay. If we can't produce the answer, let's --

DR. PORTNER: The answer is 49 of 75.

CHAIRPERSON LASKEY: Which is 65 -- yes.

DR. PORTNER: Sixty-five percent.

CHAIRPERSON LASKEY: So nothing changed?

DR. PORTNER: And the rest are dead, they will not be transplanted.


DR. KATO: My comments, many of which have been echoed by the panel, just wanted to make a couple of additional statements.

I would like to congratulate the sponsor for bringing to light several issues which the sponsor feels are important in order to expand their current indications in labeling. However, after reviewing the FDA presentation, I haven't really heard much from the sponsor in terms of rebuttal to the comments and criticisms posed by the FDA.

The issue, I believe, of a bridge to candidacy I think is a good one that has been brought up by the sponsor. However, I would then challenge the sponsor to do the right studies and submit the right data.

I believe that World Heart is an experienced company that's been doing device studies for a long time. And so therefore, instead of having a lot of anecdotal reports and anecdotal data that's been talked about, I believe that much of the data that in order to convince the panel of expanded indications or at least convincing the FDA for expanded indications should come from data that has been publicly presented as opposed to slides that we could not review due to the fact that this data was not submitted for this presentation.

As a comment to the issue of off-label indications, I actually do a fair amount of consulting work for insurance companies and reinsurers and I can assure that at least most insurance companies and reinsurance companies are very familiar with off-label indications. In fact, there's a very nice summary of off-label indications on the FDA website. And from everything that I've heard so far, and particularly from comments from the FDA, there's been nothing to suggest that the current labeling would preclude the current set of patients -- would preclude the specific set of patients that the sponsor is bringing to light today from receiving an LVAD.

So, those are some of my comments about it. I don't know whether you'd like to comment back.

MR. BRYDEN: Some of my colleagues may have more specific comments.

First, I'd like to observe that we followed one of the two alternative methods of presenting our data that was stated to us in the letter informing us of this panel hearing, and we're not aware that one of those two was less desirable to the panel than the others.

There is nothing in our slides except for the two which were commented on by the panel and excluded from discussion. That is not simply presenting in slide format the data which is included in the panel pack.

The data on which we rely is not in any sense anecdotal. It may not be satisfactory to you, but it is the data arising out of a prospectively done trial approved by the FDA resulting in the bridge to transplantation label which until the single approval of HeartMate for destination therapy, that and a comparable label for HeartMate done also in a similarly structured trial, are the only two significant approvals for this type of therapy in the United States until the REMATCH trial.

So we believe that the data that we've provided is substantive. It was prospectively done. It's data that was done under the guidance and reviewed by the FDA. It was found sufficient and satisfactory as a basis to provide a very meaningful indication.

We do not in any sense dispute, that's why there's no rebuttal in that sense of the view that the controls are inadequate controls. But this indication is not seeking to have approval. The principle discussion today seems to be you don't need the approval, we can get it anyway. That it is already approved. We would be very pleased if the FDA could confirm that a clinic is at liberty to implant within our label that candidate for transplantation means a person who is listed or is expected to be listed; that is just fine. But it is not our understand that that is what it means.

And I think we should be clear that we're not making this up. And the fact that many of these patients are in fact treated today, and many are, is because clinicians in spite of the rules put the patients first. That is, however, a factor that in our view does significantly limit the number of patients who are both brought forward and are given the device as compared to those that commissions may be prepared to deal with if they were not having to act contrary to the rules of their own organization and the structure of the regulatory process.

But if in fact bridge to candidacy is what's intended, then if that could be clarified even at this meeting, then we would agree with you we should all go for a drink.


Well then, lest we labor any longer in confusion, may we have the FDA's position on this issue? This should be simple enough to address. I would hate to think we're wasting our day here.

DR. PINA: This will take just two minutes. On behalf of myself and on behalf of FDA, we'd just like to reassure Dr. Edwards that we are equally worried about the patients, and that that's our very reason for being here. So I'm a little sensitive about that comment.

And to address Dr. Tracy's comment about regulating practice, I would love to know who reverses and who doesn't. The dataset presented here does not help me tell who will and who won't. And it would be terrific to have prospective data, I agree.

DR. ZUCKERMAN: Well, first thing is, I've heard multiple panel members correctly interrupt the position of the FDA, and that is that the FDA does not regulate the practice of medicine. If an individual physician believes there's a candidate who needs this device, then it's within the purview of that physician to utilize this particular device.

On the other hand, the FDA does regulate the approval of devices and the FDA does look very closely at the indications for use of these devices. that is why Dr. Berman at the beginning of the FDA presentation quite clearly tried to express the underlying conditions that FDA would need for changing an indications for use statement.

I think if one were to summarize those conditions, it begins with the need for appropriate data. And what we've asked the panel today is whether the data are there to make an expanded indications for use of this particular device. We are not asking the panel to regulate the practice of medicine.

DR. YOUNG: Could I ask one point of clarification: Does a candidate as you referred to for transplantation, mean the individual is listed for heart transplant?

DR. ZUCKERMAN: I would look at the particular indications for use of this particular device. There is a sort of gray area that you are trying to discuss today, but the bottom line is that the agency, if it works within a legal construct to expand its indication for use, needs to see appropriate data. And it's your obligation today to argue that that appropriate data is here in this panel pack.

CHAIRPERSON LASKEY: Thank you very much.

I think we're done with the panel questions. But before we move on, I understand that you had some additional information in response to a few of the queries this morning. So I'd like to take five additional minutes, if I might, and afford you the opportunity to respond to those. I forget which questions we were looking --

MR. BRYDEN: Yes. Our notes suggested that you had asked for the number of patients who had died in each of the subgroups within the contraindication. So we'll just pull that slide up.

And also, I believe you had asked whether we had done a multivariate analysis, and if so, whether that was available.

So those are the two items that --

DR. KRUCOFF: Unless you have a particular point, the latter request which I think was mine, I could step back from.


DR. KRUCOFF: I think we've addressed most of the issues that would be intrinsic to that conversation.

MR. BRYDEN: Thank you very much.

And the slides are just coming up on the screen with respect to the other one, which will take only a minute.

Jim, would you mind just dealing with these?

DR. YOUNG: Yes. I think, again, hampered by the numbers that are present, if we look at the causes of death pre-transplant that are in our identified groups, four were over age 66. And you can see the listing there, multi-organ failure and neurologic dysfunction.

The next one is total bilirubin. And, again, remember that there were only three patients. And again, multi-organ failure.

Next one BMI less than 19, four of 12. Bleeding, neurologic dysfunction, embolism, infection, spesis, arrythmia. And, again, remembering that the controls, one of which would have fallen into this group, were dying primarily of difficulties related to cardiovascular dysfunction and heart failure.

The next one. You have BMI over 32, two of three in the controls. Cardiac dysfunction. Heart failure here. And then, again, multi-organ failure, neurologic dysfunction, intracranial bleed.

Next one. Creatinine greater than 2?, multi-organ failure being the most predominate. Here, again, I'm not surprised about. I think perhaps this also refocuses attention on the touchy subject of renal insufficiency pre-transplant. Kind of 22 patients dying who were felt not candidates because of this.

And I think that would be the last one. Oh, PA pressure, we might -- that's the controls. Okay. So the problems being equally grave in the control patients. Go ahead to the control patient.

Five of six, a vast majority there with renal dysfunction.

Next one. PAS, again, heart failure in the one control death and cardiovascular dysfunction, embolism, neurologic dysfunction, intracranial bleeds, infection. And that should be the last one.

PVR we do have data. Two of five with high PVRs. Obviously, we don't have that in the control group.

So those are the causes of death in the pre-transplant group of patients that were identified as having relative contraindications compared to the controls.

MR. BRYDEN: Thank you.

And the last question that we had not fully answered, the percentage of patients who survived to transplant who were transplanted after six months was 73 percent and those after 12 months was 75 percent. So the overall was 65, a smaller percentage of those who were done in the first six months. Slightly more in each of the second two periods.


Colleagues, we're at that critical point. Should we forge ahead with the possibility of getting people to their various destinations, forgive the expression, on time. The schedule calls for a break, but can we move ahead and do the questions? Yes.

In that case, Geretta, if you would, please.

MS. WOOD: Yes. We need the computer to project the questions. Okay.

CHAIRPERSON LASKEY: As these scroll through, I will do my best to summarize the consensus of the panel discussion this morning, but as always please feel free to add or detract.

MS. WOOD: Okay. If you can move it to the next slide for me.

The sponsor makes outcome comparisons between the selected subgroups of LVAS and control patients, yet significant covariates of the two groups are not matched. Are such comparisons between groups with unmatched covariates valid?

CHAIRPERSON LASKEY: I think what we heard this morning and the consensus of the panel is that it can't be done.


MS. WOOD: Go to the next slide.

Seven variables with particular thresholds were chosen as relative contraindications to heart transplant. Comparisons were made between transplant eligible LVAS and control patients meeting the chosen criteria. All of the selected patients were transplant eligible and were listed for transplant. The majority were transplanted. Are these patients comparable to patients with these relative contraindications who are not transplant eligible and would not be listed or transplanted?

CHAIRPERSON LASKEY: That, of course, has been the crux of the debate all day long. I think it's the feeling of the panel that it's very difficult to establish comparability, certainly going in and on the tail end coming out as well.


DR. ZUCKERMAN: Would Dr. Lindenfeld or Dr. Yancy with their personal experience expand upon Dr. Laskey's comments?

DR. YANCY: I'll yield to Joanne.

DR. LINDENFELD: No, I agree with Dr. Laskey. I think that right now it appears that within these contraindications patients can still be transplanted.

DR. YANCY: The only to add to that is that it's obviously a very dynamic bucket of patients that can change very easily, and I would be very hard pressed to try to give any definitions or make a proviso of any sort.

MS. WOOD: Okay. If we can have the next slide.

Is there a sound scientific or clinical rationale for choosing the particular seven relative contraindications selected and not included the others, for example, high PRA, history of cancer, etcetera?

CHAIRPERSON LASKEY: Well, as has been alluded to, it comes down to issues of clinical judgment rather than even consensus or better yet, what we would like to see are specific guidelines. We realize that that's not likely to be the case. It is a dynamic area and we can't even establish consensus within the profession when there's so much site-to-site variability for what is a relative contraindication. So we'd have to say, unfortunately and it is unfortunate, there is no sound scientific evidence. There may be a consensus, but it's probably violated everyday with the inter-site variability.

DR. YANCY: The only amendment to that statement, Dr. Laskey, would be that to the extent that the device improves the humandynamic profile, with the exception of age, these are several relative contraindications that could be modified by a change in humandynamics.

DR. LINDENFELD: And the one other comment that might fit in there is that the data has been well presented and there is a small amount of data, I understand, in small numbers. But within this group of relative contraindications, I don't feel as if I have enough data to know if perhaps we should expand this with one, but perhaps not with another. In other words, we saw that 45 percent of patients with renal insufficiency didn't make it to transplant. That concerns me. And if the mortality is higher after transplant, we got down to a number of 10 out of 22 didn't make it transplant, it concerns me that even within this group that's been picked out we don't know that this is the right group, that they should all be encouraged. So I'm concerned about that.

MS. WOOD: Next slide, please.

Is there a sound scientific or clinical rationale for choosing the threshold values of the seven selected variables such that these variables singly or in combination are relative contraindications to transplant?

CHAIRPERSON LASKEY: Well, again, I think there's probably even less discussion here. The cut points were, admittedly, arbitrary from both sides of the room today. It is the best that people can come up with, but it is arbitrary. And, in fact, we haven't seen any information to suggest the incremental utility of combination of variables to add to the predictive validity of this approach. So, unfortunately, not enough information that's defended.

DR. YANCY: I hate to consistently be the voice, but correct that we didn't see the linear function of these risk factors nor did we see any of them in aggregate. But I will say that the discipline of transplant medicine, either by convention or by -- does in fact respect the majority of these as reasonable thresholds.

DR. BAILEY: But I guess it concerns me that we pick a cut point and we don't have data that show that there are significant variation between the two sides of the cut point. But we're not given any other criteria for saying you have to have at least this good a creatinine or this good a pulmonary pressure. So, I mean, we're just extrapolating to everybody on the other side of the cut point. And I guess I don't see any data to support that.

DR. YANCY: Well, the only way I can modify that, Dr. Bailey, is that you don't see those data in this data set, but in the broader context and particularly with clinical experience, I think the numbers are reasonable.

CHAIRPERSON LASKEY: Does that mean there's sound scientific evidence to support this? To answer their question.

DR. YANCY: My vote would be yes based on what we currently use as a threshold of practice this medicine.

DR. FERGUSON: I'd like to add a comment.

I think that -- I guess it's incumbent upon the sponsor to validate these, but these cut points come from clinical practice of medicine for years. And I don't think that we can really impugn the sponsor for selecting this particular group.

DR. YANCY: I couldn't agree more.

DR. KRUCOFF: I'd at least say that the context and the practice that to me are two key features that are missing from extracting these cut points for this type of indication are issues of reversibility and issues of multiple comorbidities, which you know if you have two or three of these in practice, that while I agree with Clyde entirely, I think empirically are these sorts of numbers, cut points in our minds, yes. But how we use them relative to their individual selection as criteria for an indication, that's where I think the actual practice with context and multiple comorbidities is a departure, actually, from their clinical meaning.

DR. SOMBERG: I just would add that we have to say they're arbitrary because they were cited post-hoc and that there was clear knowledge of the results so they were decided arbitrarily. If they were set a priori in the BTT trial, they would have considerably more weight.

DR. YANCY: The only final statement is the sentence before it says is there a sound scientific and clinical rationale, and whether it was post-hoc or not, there was a sound clinical rationale for these thresholds.

CHAIRPERSON LASKEY: I think that's fair enough. But admittedly, each of these is an arbitrary as to what to include and what to exclude.

MS. WOOD: Next slide, please.

Is the data sufficient to demonstrate the effect of the LVAS to normalize these seven variables to justify expanding the label to include patients who are expected to become transplant candidates with mechanical circulatory support?

CHAIRPERSON LASKEY: Well, we wish it were but it's not, and that's obviously a major part of the challenge today. There is no data to suggest normalization or even a change that we can evaluate.


MS. WOOD: Next slide.

Does the retrospect subgroup analysis of transplant eligible patients provide sufficient evidence of safety and effectiveness to expand the labeling to include patients not eligible for transplantation?

CHAIRPERSON LASKEY: Well again, the high road is that retrospect subgroup analysis doesn't really help us in too many respects, and in particular this approach in the material at hand does not help to expand the labeling.

DR. ZUCKERMAN: Are there --

CHAIRPERSON LASKEY: We've been over this. There is no safety data that we can fall back on. I believe Dr. Lindenfeld has articulated that. We're looking for meaningful data out to periods of time that it is likely these patients will endure the device for, and we've not seen that.

And in terms of effectiveness, we've not seen that side of the coin either. With long term data with so few patients it's very difficult to make anything of those Kaplan Meyer curves.

DR. ZUCKERMAN: Are there any differing opinions from other panel members on this critical question?

DR. AZIZ: You know, Warren, I think for the time periods that it has shown, I mean I don't think -- I think I would be -- to say that that device is not a safe device. Okay. It hasn't gone for more than two years in a lot of patients. But the time period that they have shown, and even the bench stuff, that the device itself is quite safe.

In terms of being effective and it functions as an LVAD in a number of the slides that they've shown, the data, the creatinines do decrease, the PA pressures do come down. I mean, for the time periods that they're projecting, I think it seems that they have shown that it does do that.

CHAIRPERSON LASKEY: Let's not forget the critical part of this question is that can we extrapolate the data from this subgroup of patients not eligible for transplant or from the group of patients that were eligible. And the answer is no, we can't. We've been struggling, and everyone's been struggling to find comparability. But we can't make any extrapolations from patients that are originally eligible for transplant and then to take that information and apply it to these patients that are deemed not eligible.

DR. KRUCOFF: I think I read two somewhat separate issues embedded in this question. One is the issue of long term use or whatever that is intended to mean, where the durability of the device and the track record, etcetera, may give us some insight.

To me the much murkier one is the patients in the BTT study who were analyzed who had a relative contraindication who were analyzed for one reason or another were listed for transplant. And as we get into patients who for whatever reason, be it systematic or be it good medical judgment, have renal insufficiency or one of these criteria who were not listed for transplanted in the BTT type of environment, we really have absolutely no idea how or what role, what the safety profile or effectiveness of this device would be. And that to me is the other part of this question.

So I think the long term durability is a little bit separate question, and I guess I think both of those may be wrapped up in this question a little bit.

DR. YANCY: I agree with Mitch. I mean, that was the context of my questioning. In this patient population trying to understand the longer term issues of safety, adverse events, durability, that's the struggle.

DR. AZIZ: But you know in the patients that where the creatinine came down, where the PA pressures came down, they came down fairly quickly within a matter of months. I mean, you're not talking about years. Isn't that what the data showed? You're not looking to have this device in for four years before the patient becomes transplant eligible.

DR. YANCY: But you are dealing with by their own admission 26 of 75 patients who got the device and didn't go on to transplant. And we are at that point talking about longer term questions of safety and efficacy. And that's not an insignificant percentage if you extrapolate out to larger numbers of patients. And I think we have a responsibility to know what it is we are subjecting those patients to experience if they fall into that category.

DR. KRUCOFF: That might even be an optimistic percentage if we extrapolate out into a little wilder field.

MS. WOOD: Next slide, please.

The sponsor wishes to add language indicating "short" or "long term" use of the LVAS. Of 190 LVAS patients from the BTT trial, 30 were on the device greater than or equal to six months. Of those, 15 were on the device greater than or equal to one year. Of those four were on the device greater than or equal to two years. Is that sufficient support to expand the labeling to include long term use?

CHAIRPERSON LASKEY: Well, I think there's two issues here. One, which was overt and the other which was finally stated this afternoon.

The overt issue is that there aren't enough patients that go out long enough, and long term is in the eye of the beholder, but whatever the definition is there's not enough patients out there beyond the year to provide meaningful information.

And the subliminal concern here has been it may not be an intention on your part, but many of these patients may wind up with this device as "destination therapy." We realize that no one is back dooring any of this, but it's conceivable that a significant fraction, and I would be loath to guess that number but it's going to be in double digits, would wind up with the device being ineligible for transplant even after the device, and that troubles us.


MS. WOOD: Next slide.

Does the retrospective subgroup analysis of transplant eligible LVAS patients provide sufficient data to judge whether expanding the label to include patients not eligible for transplant is safe?

CHAIRPERSON LASKEY: Well, again, it's just a corollary of what we've been saying. Unfortunately, the curves don't go out far enough and there's not enough patients out there to make that claim.


MS. WOOD: Okay. Let's move to the next slide, please.

Premarket review of an expanded indication for use for an approved product includes review of the modified labeling. The labeling must include a description of the patients for whom the expanded use is intended and an explanation of how the product is to be used for those patients to maximize clinical benefit while minimizing adverse events. Does the proposed expanded indication for use meet the requirement?

CHAIRPERSON LASKEY: The consensus of the panel seems that, no, it does not.

Would you care to elaborate primary or secondary reviewers?

DR. KRUCOFF: I think that's accurate.

CHAIRPERSON LASKEY: Yes. We don't mean to be this blunt, but I think this is fairly straightforward. These questions are specific and so are the answers.


MS. MOORE: Mr. Chair, there's one point that I as a consumer would like to make reference to, is that possible?

CHAIRPERSON LASKEY: I will be coming back to you in just a bit to ask for some input. Thank you very much.

I am obligated by protocol to open the public hearing portion of the afternoon.

Is there anyone in the audience who wishes to address the panel on today's topic? If not, it's my pleasure to close this portion of the open public hearing.

And I will be asking Dr. Zuckerman if the agency has additional comments. I'll be coming to the sponsor and then I'll be asking industry and consumer for their comments.

So, Dr. Zuckerman?

DR. ZUCKERMAN: Yes. Today we've had a very important panel discussion on an extremely important topic. One of the issues facing this panel for what is a nitch device is whether there is enough data in the present application to justify an expanded indications for use.

The panel will shortly need to decide this question. However, I think there is a more general problem that has been hinted at for the LVAD industry today, and I do make my comments to the whole LVAD industry since they're all seated here.

We have a situation where potentially in the post-market arena or what the agency has tried to underline as the total product life cycle, a lot of patients have gotten a particular device, and with study could have provided additional supportive data. Unfortunately, we don't have those data here, but the agency is committed to continuing support of these devices throughout the approval process, both in the pre and post approval domains, especially when it is difficult to recruit patients.

Thank you.

CHAIRPERSON LASKEY: I'd like to invite the sponsor to provide any final or near final comments before the vote.

MR. BRYDEN: Yes. I just want to thank you for your time and attention. We're, obviously, disappointed with your conclusions. We've learned something, I at least and perhaps others in our group perhaps have a broader understanding about the scope that the agency intends clinicians to be able to apply in determining who is and who is not a candidate. That whole discussion may prove to be helpful to the centers as they struggle with the issue of matching patient requirements.

I do think it's important to observe that in the current indication the implant of our device or any device as a bridge to transplantation cannot be assured that the patient will receive a heart or that they will remain eligible throughout the period of time it takes to wait to obtain one. So the issue with respect to longer term use is one which is not unique to the group with relative contraindications. Nonetheless, that decision is past and as the Chairman observed, I think it's clear to the panel that our intention was exactly what the proposal suggested and that our view is this would have opened the door to destination therapy. If it were, we wouldn't be carrying on the relatively large trial that we're now engaged in to achieve that indication.

So, again, thank you very much for your attention.


And, again, we acknowledge the integrity of the sponsor and in no way meant to impugn the purpose of this. It's clear, though, that what we've learned -- what I've learned is that in order to change a label, some very specific requirements need to be met. And I don't think they were met today.

Ms. Moore?

MS. MOORE: The point that I wanted to speak to was the one having to do with the values of the seven selected variables. That's 2C. And if I understand the discussion that I've heard today, then this is my comment regarding that particular point.

As a consumer, I think that one of the questions that one of us would ask would be how do you determine patient A as opposed to patient B. So I'm thinking that as something that's definitive to present to patients would help to answer that question.

I'm inclined to believe that despite all of the problems regarding statistical significance and the like, we do need consistency in defining the characteristics which will qualify one for being on that list for transplantation, even though I recognize that that list is not exhaustive. And I think there should be the caveat that the physician has the leeway then to make some judgment or a decision that may be contrary to that list if the need indicates such. And I think that -- I know you've made your decision on that item, but somehow or the other it seems to me that from the consumer's point of view more attention needs to be given to that particular item.

The other items, I think -- oh, I can say this. Can you not do whatever procedures you want to do without specifying whether it's short term or long term and just keep the statement as it is, that intended for use as a bridge to transplantation in cardia transplant candidates? Can't you do still do whatever you want to do without specifying whether it's long term or short term since we don't know what the definition of long term is or what short term is? That's a question that just went through my mind as you were discussing.

But I think those are the points that I had an interest in.

There was another minor point. I noticed your population consisted primarily of white males and a few females. I was wondering if there were any ethnic groups in some of these studies that you have done? And if there were, then what your findings?

Thank you.


MR. MORTON: Just a couple of notes to also acknowledge what's been stated several times today; that the sponsor did a good job of presentation of data. And also to recognize the commitment that the sponsor came here with a commitment to do what was best for the patient, and that was not exclusively stated by the sponsor, but also echoed by the FDA and by the panel. And that was made very clear today and made me proud to be part of this.

All parties also recognize the requirement that we have to limit discussion to the data at hand. And I think the sponsor and the FDA and the panel showed good restraint and in balance in that today, and I'd like to recognize that.


CHAIRPERSON LASKEY: All right. Geretta, you can read the voting options, please.

MS. WOOD: The medical device amendments to the Federal Food, Drug and Cosmetic Act as amended by the Safe Medical Devices Act of 1990 allows the Food and Drug Administration to obtain a recommendation from an expert advisory panel on designated medical device 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 reasonable assurance based on valid scientific evidence that the probable benefits to health under conditions on intended use outweigh any probably risk.

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

Your recommendation options for the vote are as follows:

Approval, if there are no conditions attached;

Approvable with conditions. The panel may recommend that the PMA be found approvable subject to specified conditions such as physician or patient education, labeling changes or a further analysis of existing data. Prior to voting all of the conditions should be discussed by the panel; and

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

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

CHAIRPERSON LASKEY: I'd now like to ask for a motion on the PMA. Dr. Krucoff?

DR. KRUCOFF: Move to recommend that this application is not approvable.

CHAIRPERSON LASKEY: And is there a second? Dr. Somberg seconds.

Can we have some discussion? Are we all of a like mind? Okay. We have to have some discussion.

DR. BAILEY: Do we have to have discussion on it?

CHAIRPERSON LASKEY: We must. Well, you looked like you were ready to --

DR. BAILEY: Oh, no, no. I was trying to look expectant.

CHAIRPERSON LASKEY: I see. Well, I'll alleviate that expectancy.

If there is no discussion, then I'd like to proceed with the vote.

All in favor of the motion put forth by Dr. Krucoff that the PMA is not approvable, please raise your hands. Ten in favor.

Opposed? And one opposed, Dr. Aziz.

The motion that it is not approvable passes ten to one.

And we need to go around the table and just defend the voting situation.


DR. AZIZ: I think there are a group of patients, as we've discussed, who fall in between absolute contraindications and actual indications for being transplanted. I think the boundaries are obviously are quite blurred and it's a moving target.

Having seen and looked after patients who could be transplanted but there are certain contraindications, I feel that a device such as this which clearly has not shown any evidence of device malfunction, and although our comments are related to the information presented today, I'm aware obviously of data from other trials and other centers where the device has been in for a long time where it's really been shown to be effective over a long period of time.

So I think that devices like this are needed, and even though this was, obviously, not an ideal trial, I feel that there are patients who would benefit and that the term bridge to candidacy will, I think, take an increasingly important role.


DR. KRUCOFF: I think we have to recognize the reality of our health care system. That pre-market evaluation, study and approval is certainly not identical to what physicians face in real work post-market use. And that in fact relative to their specific indications, probably more than the majority of devices in real world care are used on the basis of the practice of medicine off-label than specifically on label. And I think Dr. Zuckerman's suggestion to sponsors that they could be of important help in bridging that gap by gathering systematic post-market information would be one way for the sponsor in this case to think about a way of gathering data that might in fact come to a different conclusion on review.

I think the major social and health care issue that everyone has spoken to today, that the management and extended support decisions for patients who are dying who might be transplant candidates if they could be stabilized or improved in some way is a huge current and very difficult issue. In fact, a proportion that one might again suggest to the sponsor is to think about whether professional societies or the National Institutes of Health might be a more appropriate approach to understanding really where these devices might play an important role beyond what they already do in current practice.

I think the phrase "bridge to candidacy" is the critical phrase. It's obviously a component of the real world of bridging to transplant. But largely the way I see it and what I heard in the discussion today is that bridge to candidacy is defined by reversibility and multi-factorial complexities that simply were not a part of the data set that we had at hand today.

So while I think there are some critical issues here, and I would hope some ways that the sponsor and the clinicians associated with the presentation today might think about clarifying the extension of where and when we decide to use an LVAD in patients who are at imminent death, I think that ultimately in terms of supporting a change in indication, they simply don't based on the data available. And that was the basis of my vote.

DR. SOMBERG: Well, I think it would have been very helpful if we had the information that could answer the critical questions of whether this was an appropriate device for long term therapy and whether it's appropriate for people who had a relative contraindication to receive this device because they had a higher likelihood of then being appropriately transplanted. We didn't have sufficient data on either of these points. But I would like the sponsor and the other companies that are present to hear that my objections were not that we needed a randomized perspective controlled study to prove this point, but there could have been a substantiation of the control groups, there could have been data presented to show for instance group one versus group two fared similarly or not similarly. And these data sets were just not presented, and that prevented us from when we wanted to -- or at least where I wanted to see some progress made in identifying patients who benefit from long term device or patients who would be able to be better decided as candidates. We just don't have that data. So to approve a PMA without data is to probably do more harm than good, and that's why I voted yes on this negative proposition.


DR. HIRSHFELD: I think most of the issues have already been articulated stated.

I came away from this discussion convinced that patients in whom the efficacy of this device has been demonstrated by the data that are available to date are currently eligible for this device under its existing labeling. And so I saw no reason to expand the labeling at this point.

I think we'll eagerly await the results of the destination trial, because I think that will -- if that trial shows efficacy, then that will constitute a valid rationale for all of the indications that the sponsor seeks.


DR. WEINBERGER: I think that the sponsor probably didn't intend to create a new concept, which is bridge to candidacy, but that's in fact what you ended up doing. And that's a concept which is a worthwhile concept to flush out with clinical trials with evidence gathering at scientific meetings. But in a forum to try to use that new concept to define the change in labeling is a very, very high threshold which frankly was not sustainable by the data. And that's the reason why it could not support the proposition.


DR. LINDENFELD: I, too, as Dr. Somberg am not so concerned about the lack of a comparable control group, although we spent time discussing that. My decision was based on two things.

First, I believe right now the appropriate patients are not limited from getting this device and being transplanted. But more than that, if we're going to expand the indications, I would like to see enough data to know that those expansions are appropriate and that the outcomes, even though the numbers are small, give us some reassurance that when we expand those indications that down the line these patients are actually surviving a year or two post-transplant somewhere near with the same results as patients who don't have the contraindications. And we just didn't see any data at all like that today. The only data we saw is that they get transplant. And within that as well, we don't know that these specific relative contraindications would have the same outcome, that is that a creatinine of 2.5 is just as safe to do as somebody with a pulmonary pressure of greater than 60.

So I'm hesitant when we're not limiting any of these patients, I'm hesitant to expand those indications without some more data that it's the right thing to do.


DR. BAILEY: Not very much to add.

I think I sort of agree that the comparison control, though it might differ, is really not the issue. And it's an issue of belief and fact, and I think I believe that there is some efficacy clearly in transplant eligible patients. I think it's probably overestimated by the current control treatment comparison, but it's there.

I also believe that you can extrapolate it to some extent beyond the confines of the BTT study. The question is where to draw the line. And I think these data provide further information for physicians as they make their decisions, but I don't see how we can say that it's good evidence when it just isn't.

It's terrible, because it's such an important device and I'm sure it's very important to be used, and for the indications to be as broad as possible. But we have to make our decision based on the data.

So that basically, for me it comes down to how good are the data within the BTT study for extrapolating the performance of the device to patients who were not eligible for the study. And, you know, fortuitously there was a subset that had some of these contraindications, but it could be that these aren't a very unrepresentative subset of all patients who would have been not eligible for the study. So on scientific basis from a conservative point of view, you just can't use the data to extrapolate.

And that's the basis for my vote.

DR. TRACY: I think the sponsor's raised a very important issue of the bridge to candidacy, and I hope that the data that are here serve for additional studies that can be done.

I think within the current indications that the device is approved for there is room for clinical studies that may not be solely sponsored by industry. But I think that the data as presented and trying to extract retrospectively information from relatively small group without really overlap with what might be clinical relevant, just doesn't quite meet the standards that I need to see, and that's why I voted the way I did.

DR. FERGUSON: But I think it's important for the record to indicate that the World Heart device is not on trial here today. It's the labeling change that's on trial. And I, like the others at the table here, found it difficult to justify changing the labeling either in anyway for one of the two reasons, although my gut reaction tells me that that's the thing to do, the data simply is not there.

DR. YANCY: I supported the recommendation, which was a vote against not the paradigm, but rather the pre-market application because I not only support the paradigm in clinical practice, but I have respect for the sponsor for making this and other platforms available for a very difficult patient population when there is an incredible need for mechanical support. And I have tremendous respect for the consultants today, who I think did an admirable job with the database that was compromised for the way in which it was used.

So I respect both the presenter and the sponsor for coming forward and making this effort and raising new questions. And I hope that today's vote does not work against any momentum to answer these important questions and to believe that a registry or some other clinical experiential format will be the kind of way that we can go about doing this. As a clinician I understand it is incredibly difficult to do an evidence based scientific/rigorous trial in this patient population But as we all know, there are other ways to get the same sort of data.

So, I hope that this has been a valuable experience for all involved and we can move forward.

DR. KATO: I would also like to congratulate the sponsor for bringing to light this issue of bridges to candidacy. I have to agree with Dr. Ferguson that my vote of not approvable was for the labeling and not for the device itself.

I would like to echo Dr. Yancy's comments though that it is vitally important that on a going forward basis that the sponsor does try to establish a database for all of their patients and try to glean as much data that is potentially possible from these very, very critically ill patients so that we can learn something for the future. Because it very well may be that a strict randomized prospective study may or may not be possible, but if you can at least give us a very granular presentation of the data that you do have from the entire experience of your company, I believe that that would be a valuable asset for future presentations.

CHAIRPERSON LASKEY: And for my part, it just was begging for some evidence to support what the LVAS does in these patients, the delta. And something as simple as a multivariable analysis with the delta improvement in variable X, Y, Z would have been nice to see. But that is what you were trying to sell, it's just that you didn't have any data to support changes with the device.

With that, I'd like to thank my colleagues on the panel, thank the sponsor, thank the agency for a superb job.

And I'd like to close today's session, concluding the report and recommendations of the panel on PMA P980012 supplement 4 from World Heart for the Novacor N100PC and N100PC(q). Thank you all.

(Whereupon, at 3:16 p.m. the panel was adjourned.)