00331

4 VOLUME III

5 (Morning Session - November 16, 1999)

10 HUMAN TUMOR ASSAY SYSTEMS

12 HEALTH CARE FINANCING ADMINISTRATION

13 Medicare Coverage Advisory Committee

14 Laboratory & Diagnostic Services Panel

20 November 15 and 16, 1999

22 Sheraton Inner Harbor Hotel

23 Baltimore, Maryland

00332

1 Panelists

2 Chairperson

John H. Ferguson, M.D.

Vice-Chairperson

4 Robert L. Murray, M.D.

5 Voting Members

David N. Sundwall, M.D.

6 George G. Klee, M.D., Ph.D.

Paul D. Mintz, M.D.

7 Richard J. Hausner, M.D.

Mary E. Kass, M.D.

8 Cheryl J. Kraft, M.S.

Neysa R. Simmers, M.B.A.

9 John J.S. Brooks, M.D.

Paul M. Fischer, M.D.

Temporary Voting Member

11 Kathy Helzlsouer, M.D.

12 Consumer Representative

Kathryn A. Snow, M.H.A.

Industry Representative

14 James (Rod) Barnes, M.B.A.

15 Carrier Medical Director

Bryan Loy, M.D., M.B.A.

Director of Coverage, HCFA

17 Grant Bagley, M.D.

18 Executive Secretary

Katherine Tillman, R.N., M.S.

00333

1 TABLE OF CONTENTS

Page

2 Welcome and Conflict of Interest Statement

Katherine Tillman, R.N., M.A. 5

Opening Remarks & Overview

4 Grant Bagley, M.D. 10

5 Chairman's Remarks

John H. Ferguson, M.D. 28

Brian E. Harvey, M.D., Ph.D. 30

Open Public Comments & Scheduled Commentaries

8 Frank J. Kiesner, J.D. 48

Larry Weisenthal, M.D. 57

9 Randy Stein 92

Richard H. Nalick, M.D. 99

10 William R. Grace, M.D. 108

John P. Fruehauf, M.D., Ph.D. 110

11 James Orr, M.D. 127

Robert M. Hoffman, Ph.D. 131

12 Andrew G. Bosanquet, Ph.D. 136

David Alberts, M.D. 142

13 Robert Nagourney, M.D. 147

David Kern, M.D. 159

14 Daniel F. Hayes, M.D. 168

Bryan Loy, M.D. 178

LUNCH 196

VOLUME II

Open Public Comments & Scheduled Commentaries

18 Edward Sausville, M.D. 201

Harry Handelsman, D.O. 227

19 Harry Burke, M.D., Ph.D. 234

Mitchell I. Burken, M.D. 262

Open Committee Discussion 304

Day One Adjournment 330

00334

1 TABLE OF CONTENTS (Continued)

2 VOLUME III

3 Opening Remarks - Introduction 336

4 Open Committee Discussion 337

5 Motions, Discussions and

Recommendations 425

Adjournment 487

00335

1 PANEL PROCEEDINGS

2 (The meeting was called to order at

3 8:05 a.m., Monday, November 15, 1999.)

4 DR. FERGUSON: Miss Tillman, our right

5 hand, is here, and has some announcements and

6 pronouncements.

7 MS. TILLMAN: Good morning, and welcome

8 again. First of all, I am going to read the

9 conflict of interest statement again. Conflict

10 of interest for the Laboratory and Diagnostic

11 Services Panel meeting November 15th and 16th,

12 1999. The following announcement addresses

13 conflict of interest issues associated with this

14 meeting, and is made part of the record to

15 preclude even the appearance of an impropriety.

16 To determine if any conflict existed, the Agency

17 reviewed the submitted agenda and all financial

18 interests reported by the committee participants.

19 The conflict of interest statute prohibits

20 special government employees from participating

21 in matters that could affect their or their

22 employer's financial interests. The Agency has

23 determined that all members and consultants may

24 participate in the matters before the committee

25 today.

00336

1 With respect to all other participants,

2 we ask that in the interest of fairness, that all

3 persons making statements or presentations

4 disclose any current or previous financial

5 involvement with any firm whose products or

6 services they may wish to comment on.

7 In addition to that, we request that

8 anyone with a cell phone please turn it off, so

9 it doesn't disrupt the discussion this morning.

10 Also, as the speakers either come to

11 the microphone or the panel members begin to

12 speak, if you could identify yourself for the

13 record, since we have a court reporter here, and

14 it would make it easier for him to identify who's

15 speaking.

16 And also, anyone who would like a

17 transcript of the meeting can contact Mr. Paul

18 Gasparotti, with Salomon Reporting Services, and

19 he can make a transcript available for you.

20 Dr. Ferguson?

21 Opening Remarks - Introduction

22 DR. FERGUSON: Thank you. This

23 morning, we'll be primarily discussing among the

24 panel members, and then voting on the questions,

25 or the questions proposed as points to vote on.

00337

1 I would like to remind those in the

2 audience that we would like to keep this

3 restricted to the panel's discussion, except on

4 points of reference where we may need some points

5 clarified from members who presented yesterday,

6 until the 11 to 11:30 session, which we can open

7 up to four or five minute remarks by some who

8 presented their work yesterday.

9 Open Committee Discussion

10 DR. FERGUSON: Now I would like to

11 start this morning and kind of go around among

12 the panel members to get their ideas and their

13 comments and critiques and concerns, and

14 questions on what we've heard, and anything they

15 think that is important that we should know

16 about. Maybe I could start over there on the far

17 right.

18 DR. MINTZ: My concern here is trying

19 to hone in on -- these tests are reasonable, and

20 it's reasonable in a setting of a malignancy to

21 do this. The question I think with which I'm

22 wrestling is when are they necessary. And it's

23 hard to hone in on the data that we have seen on

24 specific situations where we find them

25 necessary. I look forward to further comments

00338

1 this morning on can we identify situations and

2 disease states where we feel collectively that

3 this test is necessary.

4 I am just beginning to read the

5 articles by Dr. Bosanquet here in CLL, but I do

6 find them interesting, and I'm ready to be

7 persuaded, but I would like a little time to look

8 through them. But as we address these questions

9 this morning, I am interested in hearing further

10 from the participants yesterday as to where we

11 can find specific situations where we might deem

12 this a necessary test. And I look forward to my

13 colleagues trying to identify that situation. I

14 at present have not identified such a situation,

15 but I am open to being persuaded.

16 DR. FERGUSON: Very good. Kathy?

17 DR. HELZLSOUER: Yeah. I agree that

18 intuitively these make sense to be used, and I am

19 wrestling with the same things. I think cancer

20 is too large of a disease entity, and it seems

21 that there probably are settings, and maybe CLL

22 is one of them, where these tests are

23 appropriately used. There is the issue of

24 metastatic versus primary settings, or adjuvant

25 setting, or trying to sort out once they've been

00339

1 previously treated.

2 And I'm not convinced, although we

3 heard a lot about quality of life, and I believe

4 that's a very good clinical indication for this,

5 that if you can avoid unnecessary chemotherapy,

6 that's extremely relevant and important, but I'm

7 not convinced yet, given the specificity of the

8 overall test results, that we have 80 percent,

9 plus the 10 to 20 percent problem with

10 acquisition of tissues appropriate processing,

11 how many will be spared. And my readings of some

12 of the graphs yesterday and some of the articles

13 here, that if you still have 20 percent that,

14 that's the specificity in the combined group,

15 would you feel comfortable eliminating that for

16 an individual, because there is still a chance 20

17 percent of the time they would still be sensitive

18 in vivo, they will still respond. And when you

19 get down to a metastatic setting when people will

20 choose something for even a 1 percent benefit,

21 it's hard for me to see that you will be

22 eliminating a lot of chemotherapy. So that's one

23 thing that I would like more clarification on.

24 I think that the problem is that there

25 isn't much information on the clinical outcome,

00340

1 although there's a correlation with survivors,

2 it's the problem we always have with reviewing

3 issues, that responders always do better than

4 non-responders, and it's probably a good marker

5 for responders. But I think we have to see how

6 we can clinically use that in either choosing

7 chemotherapy, and I think the compelling argument

8 is the issue of avoiding unnecessary

9 chemotherapy, but I'm not sure I have the

10 evidence to say that would actually be done in

11 practice.

12 DR. FERGUSON: Thank you. Miss Kraft,

13 do you have some?

14 MS. KRAFT: Cheryl Kraft responding.

15 First of all, what was pointed out yesterday was

16 two different percentages of how many people

17 don't respond to any type of chemotherapy or

18 cancer treatment, one being 70 percent and the

19 other being 76.3 percent. So it's clear that we

20 don't know how to manage cancer patients so that

21 they can survive. So the question to ask is,

22 will the tests that are available to us, these

23 human tissue assay systems, help us in prolonging

24 or help the physician in treating the patient?

25 From what I can tell in the studies

00341

1 that I've read, that the use of these tests and

2 the way the doctors use these tests in treating

3 patients, that there were no negative effects to

4 the patient or consequence to the patients, with

5 the exception of one trial that was outlined in

6 one of the studies. So that being, do these

7 tests then have, test for drug resistance at a

8 sensitivity that is great enough so that the

9 physicians can interpret the benefit to the

10 patient? Well due to the fact that drug

11 resistance is growing and is definitely

12 multifactorial, as one of the articles said, and

13 the heterogeneity in cancer tumors is great, then

14 are we not, and myself, trying to make sense out

15 of all the articles I have read, and in which

16 cancers and which drugs should be treated for

17 which specific cancers, are we not maybe trying

18 to fit a heterogenetic tumor into a box? I think

19 analytical people try to fit everything into a

20 box.

21 And so what I would like to put forth

22 to the panel is that maybe we should step out of

23 the box and we need to look at, since again, all

24 these tumors are heterogenetic, should we look at

25 just continuing to do what has been done, that

00342

1 being continuing to test all types of these

2 cancer tumors against all the drugs available to

3 us and see, and continue to treat patients

4 accordingly? Now none of the patients in none of

5 the articles were denied treatment of drugs that

6 they were considered to be resistant to, so

7 taking that into consideration, maybe the studies

8 should continue to be done.

9 However, during that time, this panel

10 needs to think of should the patients, even

11 though this may not be definitively designed for

12 a specific tumor and a specific drug, should the

13 patients really be denied a test? And this is a

14 laboratory test we are talking about. Should

15 they be denied a laboratory test that could

16 possibly benefit them?

17 I think, again, this laboratory test is

18 a tool for a physician. The physician should

19 take advantage of all the tools available to him

20 to treat a patient. And since studies show that

21 only 25 to 30 percent, again, of patients do

22 respond to the test and/or the drugs and/or the

23 correlation of the drugs and the chemotherapy

24 that we have available to them, should we not

25 consider, due consideration to looking at the

00343

1 advantage of these human tissue assay tests and

2 the resistance that has been found to

3 chemotherapy drugs?

4 DR. FERGUSON: Okay. Thank you.

5 Dr. Hausner?

6 DR. HAUSNER: Dr. Richard Hausner. For

7 me, I would like to take the approach to try and

8 put my comments in the context of my own clinical

9 experience, my own day-to-day, I'm a working

10 pathologist, although I am on the active clinical

11 faculty of Baylor College of Medicine and the

12 University of Texas Health Science Center in

13 Houston. I practice in a community hospital, but

14 I have very long reach in terms of my clinical

15 experience. I have a big practice. And I can

16 tell you that in Houston, Texas, where there is

17 quite a bit of health care going on on a daily

18 basis, not once ever in my life, with all of the

19 cancer patients that I've seen, have I once been

20 asked to harvest tissue for this procedure. Not

21 ever. And I can tell you that if any of the

22 patients in my practice had had this testing,

23 that we would have been involved in the

24 harvesting by definition, because the surgeons

25 would have surely asked. So I know that it

00344

1 hasn't happened.

2 Nevertheless -- and I came in here

3 reading the source material with that bias,

4 because I had that bias from the very beginning.

5 But nevertheless, somewhere around the middle of

6 yesterday afternoon, my thoughts began to

7 crystallize, and they crystallized during the

8 time that, in the afternoon session when the data

9 was put up to a tremendous amount of scrutiny and

10 a very sophisticated critique, and I thought that

11 it held up pretty darned well. And I have come

12 to the conclusion that while over the past 20

13 years of the research that has developed for this

14 technique, it clearly was a research tool and not

15 ready for prime time, that the decision was

16 correct not to allow this into Medicare's realm

17 and therefore, give it the validity to go

18 forward.

19 Because what is someone's exciting

20 front line technique comes very close to someone

21 else's quackery, and at some point it would have

22 been premature to allow this. But I believe now

23 that the third generation technologies clearly

24 take this beyond a research tool and that from

25 this point forward, I would hope that the

00345

1 clinical studies will be conducted to refine

2 where this could be best used.

3 Another analogy would be that a, that

4 if this technique is not permitted in its current

5 state, then the panel ought to reconvene and

6 consider removing microbiologic sensitivity

7 testing from the armamentarium of physicians, if

8 this is not approved. The truth, I believe, lies

9 somewhere in the middle, therefore, and just like

10 so many other things we do in medicine, that this

11 is a useful tool, imperfect as it is, and the

12 ground rules may have to be carefully defined,

13 but to turn the test away in its entirety, I

14 believe would be inappropriate.

15 And in closing, I would point to the

16 final paragraph of Dr. Weisenthal's paper in

17 which he talked about whether we use the civil or

18 criminal criteria of preponderance of evidence

19 versus beyond a reasonable doubt. Beyond a

20 reasonable doubt, we don't have. Preponderance

21 of evidence, I believe we do. And therefore, my

22 conclusion is, as a rough sketch, is that

23 something ought to be done towards bringing this

24 test into, as another tool for physicians to

25 use.

00346

1 DR. FERGUSON: Thank you. Dr. Kass.

2 DR. KASS: Thank you. Mary Kass.

3 First of all, I think my first question

4 was about the testing methodology, but I think

5 that there is overwhelming evidence to show that

6 these tests meet all the normal QC, all of the

7 normal standards that all other laboratory tests

8 have to meet. I think that they're valid, I

9 think that they are reproducible, so the third

10 generation of tests for me is no longer a concern

11 in that respect.

12 The question has been raised about

13 necessary versus clinical utility. I don't know

14 how to define a necessary laboratory test; I

15 think that's really in the mind of the user.

16 When I was in training, which wasn't all that

17 long ago, the emergency room of a downtown urban

18 hospital in Washington, D.C. didn't even have a

19 laboratory open from midnight until eight a.m.

20 because there were no laboratory tests that were

21 necessary to make clinical diagnoses. But we've

22 come a long way since then, and I think medicine

23 has grown and realized that there are many things

24 that can help physicians do a better job in

25 taking care of their patients. So the clinical

00347

1 utility of this test, I think has been

2 demonstrated, to certainly my satisfaction.

3 The fact that the test is difficult to

4 do because you have to acquire fresh tissue, it

5 has to be shipped in a certain way quickly to a

6 laboratory, that doesn't bother me either. That

7 doesn't change its utility. I remember when we

8 first started doing flow cytometry, the transport

9 of specimens to do flow cytometry on was a big

10 challenge to us. Now we do it routinely and we

11 don't lose specimens in the transport process.

12 It is very intriguing to me that this

13 particular methodology may be very helpful in

14 evaluating new drugs, the number of new

15 chemotherapeutic agents that are rapidly being

16 introduced to try to help us have a greater

17 impact to the treatment of cancer. I think that

18 anything that we could use to help define which

19 modalities have a greater possibility of working

20 and which don't, would be very helpful. I think

21 it also allows the earlier consideration of other

22 treatment modalities for patients, rather than

23 going through a whole course of chemotherapy and

24 waiting for the end point of no response.

25 Earlier in the course of that, a clinician may

00348

1 have an opportunity to switch a chemotherapeutic

2 drug, or remove one which has a very toxic side

3 effect from the treatment regimen.

4 I guess in summation, I think that we

5 haven't done a terrific job in treating most of

6 the solid tumors. I think everyone is very

7 disappointed in the fact that we haven't been

8 able to have greater success than we have. I

9 think that this is another tool, one of many,

10 that could be available to clinicians that might

11 help, certainly in terms of the quality of life,

12 if we could remove drugs from the treatment

13 regimen that were not effective, and perhaps in a

14 better outcome.

15 I think the patient that testified

16 yesterday, that's one case, it's anecdotal.

17 However, I've practiced pathology for 32 years; I

18 have never seen a patient with widely

19 disseminated pancreatic carcinoma that survived.

20 You have to take notice of that. I think that's

21 worth listening to.

22 So, I think that's the summation of my

23 comments.

24 DR. FERGUSON: Thank you. Miss Snow?

25 MS. SNOW: I'm Kate Snow. I'm the

00349

1 consumer voice for this panel. I listened very

2 intently to all of yesterday's testimony and I

3 agree that Mr. Stein was very compelling, and I

4 too have never seen a pancreatic cancer

5 survivor. However, I did not know how old this

6 gentleman was, or if he had other comorbidities.

7 I believe that if I were a cancer victim, I would

8 want this study available for my use. I would

9 feel it was reasonable and I would also very much

10 feel it was necessary.

11 Listening to the quality of life and

12 the cost of life that could be gained, and to

13 decrease the burdens for individuals was also

14 very compelling. If it takes the guess out of

15 the therapy that's used, I think it's a very good

16 tool to have available to us.

17 I struggle with whether or not this

18 test will be available in a way where those of us

19 in northern rural Michigan will have access to

20 this kind of tool or not, and what that might

21 look like in the future.

22 I do feel there is a possibility for a

23 cost effectiveness. It may need some more

24 research and looking into exactly how cost

25 effective this could be, both for the medical

00350

1 community as well as the beneficiary.

2 And I think that's all I have to say

3 for now.

4 DR. FERGUSON: Thank you very much.

5 Dr. Loy?

6 DR. LOY: I'm Dr. Bryan Loy, and I

7 listened also very intently yesterday to the

8 presentations being made. I have a couple of

9 comments, first of all regarding the

10 presentations. I noticed a number of cancers

11 were being elaborated on. I am still not clear

12 at what point in the clinical progression of the

13 disease, or how often the testing should take

14 place.

15 However, having said that, this does

16 sound like this is a tool that be could be very

17 useful. But having listened to the presentations

18 yesterday, again, we were focusing on specific

19 cancers, and to try to take that tool and apply

20 it to all cancers at this point in all clinical

21 scenarios, doesn't seem to be quite reasonable at

22 this point. We really didn't talk a lot about

23 the sarcomas, or trying to talk about such broad

24 fields as hematopoietic neoplasms. I think at

25 least in my mind, I would need some more

00351

1 convincing evidence to try to apply this

2 technology wide spread, and I think that this is

3 certainly germane to a policy type discussion.

4 The other piece that's still lacking in

5 my mind is where this really fits clinically.

6 Because some cancers are clearly curable with

7 chemotherapy, or they're curable with radiation

8 therapy in combination with chemotherapy, or

9 they're curable with surgical resection, or any

10 of those combinations. And trying to really fit

11 this into that niche is going to be quite

12 difficult to do from a policy perspective.

13 Having said that, I think that there

14 certainly is some promise. I think there is some

15 utility that has been potentially demonstrated

16 here, but I am not clear on where this fits yet.

17 DR. FERGUSON: Thank you. Dr. Murray?

18 DR. MURRAY: Thank you. I am Robert

19 Murray, and I've kind of grouped my comments into

20 four areas. The first point is that I believe

21 what we are supposed to be doing is looking at

22 the questions that were presented, the six

23 specific questions that we would like to come to

24 grips with and arrive at answers to. I sense

25 that we have been taking the view from 35,000

00352

1 feet and not from the detail level that we need

2 to or that we were asked to. I am concerned

3 about that, and I really think that we have

4 looked mostly at number 5, which asks, is there

5 evidence to support clinical utility? I sense

6 that from the speakers who have voiced their

7 opinions and also reflecting my own that the

8 answer is yes, there is evidence for utility in

9 certain cases.

10 The second point is, we're stumbling

11 over reasonable and necessary versus clinical

12 utility. Reasonable and necessary is in the

13 statute, and Dr. Bagley gave us a couple examples

14 of how you can assess reasonable and necessary.

15 I view it as a term of art. I don't think we

16 look first at reasonable in isolation, and then

17 we look at necessary in isolation. We'll

18 certainly get thoroughly enmeshed in what kind of

19 necessity. As was already mentioned, no

20 laboratory test may be necessary, mathematically

21 necessary. You can certainly find alternatives.

22 But nonetheless, utility is perhaps an equivalent

23 term that we have in fact focused on. But again,

24 we are looking at a very high level.

25 And Dr. Loy's comments yesterday, and

00353

1 his comments just a moment ago, I think remind us

2 that we need to come to conclusions that are

3 going to allow a very high legal of specificity.

4 We can't just say, I don't think we should say,

5 at the end of this session, yes, there are some

6 situations when some testing might be

7 appropriate. That is simply not the level of

8 guidance that we need. Some of us went through

9 negotiated rule making over the past year and we

10 realize how difficult it is to draft a national

11 coverage decision with the uniformity and

12 specificity. So I am concerned about the fact

13 that we are, we seem to agree that there are some

14 situations in which there is utility, but we're

15 far from reaching the level of specificity that

16 we ultimately will need.

17 The third point is just my own

18 reaction. Spending my life generally in the

19 laboratory, I tend to analogize all of the

20 situations, the questions, to existing laboratory

21 tests. There is no question that many laboratory

22 tests which are routinely approved currently have

23 nowhere near the evidence, nowhere near the

24 accuracy and predictive value that the tests that

25 we're considering today, that we heard about

00354

1 yesterday, have already demonstrated. Yes, we do

2 have to look at outcomes. We have to look at

3 outcomes measured in different ways. We have to

4 look at evidence. But the evidence, even if the

5 bar is raised higher, the evidence that we have

6 heard certainly exceeds the evidence that we have

7 for many, many tests currently in use.

8 My fourth and last point is actually

9 two very minor specific objective questions, and

10 perhaps Dr. Bagley can respond to one or both of

11 them. In Dr. Weisenthal's paper that he included

12 in the packet, there is a reference to a Medicare

13 hearing in April of 1998 which seemed to indicate

14 that it was a decision of what I would assume was

15 an administrative law judge, ruled that these

16 tests would be covered. And my question is,

17 which I'm not asking for an answer now, but

18 sometime before noon, does that decision affect

19 our decision here? If a judge has already ruled

20 that they are coverable, then what are we

21 debating?

22 And the last and very minor point, a

23 question that perhaps one or perhaps several of

24 yesterday's speakers could answer, are any of the

25 tests that have been suggested, the tests

00355

1 performed that are being currently offered on the

2 market, are any of them covered by patent

3 protection? Are we doing anything, are we making

4 a decision on issues that would force or

5 encourage or would support limitation in the

6 availability of the test? And again, I am not

7 asking for an answer now, but if sometimes

8 perhaps during the open discussion, I am curious

9 what level of patent protection there is

10 currently, could these tests be offered by any

11 laboratory if they were approved? That's all of

12 my comments.

13 DR. FERGUSON: Thank you.

14 I had a number of things, mostly in the

15 form of questions myself. But I guess some can

16 be considered comments. First, there are several

17 different tests done by several different groups

18 that we were exposed to. Not all seemed to be

19 equal or equivalent to each other, they were used

20 in many different kinds of cancers. This leaves

21 a large number of combinations and permutations

22 for us to grapple with. And it's hard to put

23 them, as a matter of fact, I would say it's

24 impossible to put them all in one basket and say,

25 you know, treat them all together. At least I

00356

1 would find it difficult, given the amounts of

2 data and studies that we saw, all for different

3 tests and so on, so this in my view makes it a

4 difficult job.

5 A second, that many of the studies that

6 we saw were on the small side, small numbers of

7 patients.

8 Number three, it wasn't always clear to

9 me how the patients were chosen for these studies

10 or from what populations they were chosen. In

11 other words, what the denominator was, how did

12 these patients get into the study. Sometimes it

13 was. I'm giving sort of an overall, at least

14 what my concerns were. It's clear that these

15 patients had to be self selected in a way that

16 there was an accessible tumor to be biopsied or

17 surgically removed, that some patients who

18 perhaps had recurrences weren't available, once

19 they had tumors that recurred, because they were

20 deep or in bone, or inaccessible in some other

21 way, or weren't willing to put up with biopsies

22 and so on. So that there were patients that

23 might possibly benefit but couldn't because they

24 didn't have tumor available. Whereas the tumors,

25 the easily accessible, perhaps is in leukemia

00357

1 patients and lymphomas, where tissue is

2 reasonably easily available, and maybe that's a

3 different group. I mean obviously, maybe they're

4 self selected in that way to be better and more

5 responsive. But any way, it is a bit of an

6 issue, I think. It's hard to treat them all

7 equal when you need tissue in order to do this

8 test.

9 The fifth point, it seemed like to me

10 on most of the studies we were dealing with

11 advanced tumors, mostly recurrent after stage 2.

12 I wondered how many actually stage one and stage

13 two type patients had been studied.

14 Number six, that in -- it seems to be a

15 number of papers alluded to the fact, or studied

16 the fact that even cancer cells from the same

17 patient were different, in other words, that the

18 primary site tested different than the metastatic

19 site. Which brought up the notion, and this was

20 again stated and makes it somewhat difficult,

21 that patients had been treated, their cancers now

22 become more resistant and test differently with

23 these tests. So this is just another factor

24 which makes the testing, you know, when you test,

25 after treatment, before treatment, and whether

00358

1 you test a metastatic site, the primary site, do

2 you still have that and so on. This all adds

3 other things and as Dr. Loy mentioned, when is

4 this test most useful? And so this just raises

5 to me another set of questions.

6 Then I think what was mentioned by

7 Dr. Barnes yesterday, in a number of areas there

8 are several histologic types of cancers, so we

9 weren't always given that kind of information,

10 and whether they all test the same or might test

11 the same. Ovarian cancer is a multidimensional

12 animal, as I understand it.

13 So, those were my concerns. Having

14 said that, I also felt that in some of the

15 studies that were presented, I was impressed with

16 some of the leukemic studies and some others that

17 there is some usefulness and that it needs to be

18 mined, but mined carefully and under the right

19 conditions.

20 Just another comment about randomized

21 trials. Where I sat at the NIH as chair of the

22 technology assessment committee for the American

23 Academy of Neurology for a number of years, the

24 number of randomized trials with outcome

25 measurements for diagnostic tests, I don't

00359

1 believe I could have counted on one hand. I

2 would have to look very hard to find those

3 tests. I remember seeing reference to one or

4 two, but -- and there may be more, but I think

5 there is no question that for diagnostic tests,

6 randomized trials with good clinical outcomes are

7 extremely rare and I believe that, however, they

8 should be done. We need better standards.

9 Dr. Bagley?

10 DR. BAGLEY: I would like to bring up a

11 couple of other, or reiterate a couple of other

12 notions which I just want to sort of bring to the

13 forefront for us to keep in mind as we consider

14 these.

15 Dr. Loy brought up a very important

16 point, I think, yesterday. And it's one that's

17 easy to lose sight of when, as Dr. Murray said,

18 looking from 35,000 feet. And that is that any

19 recommendations that you make, that are then

20 placed or implemented in the policy, need to be

21 done with some specificity. We normally don't

22 write policies that simply say, pay for test

23 whenever a patient's physician thinks it's

24 necessary. Now that might be a reasonable

25 policy, but Medicare isn't designed to work that

00360

1 way. And in fact, we've learned from long

2 experience that if we do things that way, that

3 while it works 99 percent of the time, the 1

4 percent of the time that it doesn't work, it is a

5 disaster, because there are, there is fraud and

6 abuse in Medicare. It is a very very small

7 proportion of what goes on, but it accounts for a

8 large portion of the dollars, and they're the

9 dollars that belong to the beneficiaries of the

10 program, and they need to be protected.

11 And perhaps that's the reason that

12 Congress gave us the admonition that we shouldn't

13 just pay for medical service, we shouldn't pay

14 for medical service that a patient or a physician

15 thought was reasonable and necessary, but

16 actually the prescription that's written in the

17 law is written in the negative. It says, Health

18 Care Financing Administration will make no

19 payment for a service unless it is reasonable and

20 necessary. That means there has to be some

21 policy determination and there has to be some

22 review, some process by which we determine the

23 things that are reasonable and necessary.

24 Now with diagnostic tests, it's perhaps

25 a little more difficult than it is for therapies.

00361

1 When we're talking about treatments, we're

2 talking about options that a patient can take,

3 and in fact, they can select from one of the

4 options, and as long as there is evidence that

5 they are reasonable choices, it then becomes a

6 little easier to come to the notion that it's

7 reasonable and necessary. But diagnostic tests

8 become a little bit more difficult, because

9 diagnostic tests, after all, give us information.

10 Patients want information, physicians want

11 information, and we're all taught that the more

12 information we have, the better off we are, more

13 information gives us better results. But that's

14 not always the case.

15 First of all, that, when Medicare views

16 a service or a test or a drug, or anything else,

17 as a covered service, therefore, it will be paid

18 for by Medicare, it's easy to lose sight of the

19 fact that that doesn't -- it's paid for by

20 Medicare. It means it's paid for by the

21 beneficiaries in the Medicare program. Medicare

22 is after all a program which is funded by the

23 beneficiaries, and the future beneficiaries,

24 which is all of us. And in fact, the payment

25 comes from that source, and in fact it doesn't

00362

1 come entirely from that source. Some of it comes

2 directly from the pockets of the patients who are

3 receiving the service.

4 And what we're talking about here is a

5 combination service. Some of what we're talking

6 about may ultimately come under the heading of

7 laboratory testing, but a great deal of what

8 we're talking about is not laboratory testing,

9 but it's physician service. It's interpretation,

10 physician interpretation, and it really comes

11 under the heading of consultation, it's a

12 physician consultation. And when it is paid for

13 by Medicare, so called, it means that 80 percent

14 of it comes from the premiums which are paid by

15 the Medicare beneficiaries, premiums that are

16 paid for the part B Medicare service, which is

17 optional, although most beneficiaries do opt for

18 that. But they pay a premium every month and

19 that premium pays the service. That premium is

20 determined in some part by the amount of payment

21 in the program. It's a health insurance premium.

22 And the remaining 20 percent comes from the

23 beneficiary. So these tests are not tree.

24 Now, it doesn't matter if they're free

25 or not. If someone has a fatal disease and they

00363

1 offer hope and they offer an improved way to

2 treat it, then we don't really put a price on

3 that, nor do the patients. But we need to

4 remember that there was that last notion that I

5 put up yesterday when we talked about what was

6 reasonable and necessary. And that was, once

7 something's safe and effective and once something

8 has demonstrated utility, and the risks outweigh

9 the benefits, then perhaps we should also look at

10 the issue of whether or not it adds value.

11 Now value is not a new concept, it's

12 not a new concept in considering medical

13 treatment. We've always done that. We've always

14 done that with diagnostic tests. As long ago as

15 when I went to medical school, the notion was

16 given to us very early that tests are not

17 something to be used indiscriminately. When you

18 order a test you should consider, is the

19 information needed, is it going to make a

20 difference and therefore, am I properly using

21 this resource? So the notion of does it add

22 value to the patient's treatment is very

23 important, and I think that needs to be kept in

24 mind.

25 And then the next point, which follows

00364

1 on what Dr. Mintz said, is it reasonable, is it

2 necessary? I think it is an important concept,

3 because for it to be reasonable and necessary

4 both, it needs to offer not only information, not

5 only information which may be correct, but

6 information which is likely to influence the

7 course of treatment. Is it information the

8 patient and the physician need? If it's going to

9 guide therapy, then it should give us a decision

10 in which we should do this. Now there is an

11 interesting, there's an interesting interplay

12 between what's reasonable and necessary.

13 We have recently been looking at the

14 same thing in terms of what's reasonable and

15 necessary in terms of a test with regard to using

16 PET scans for many diagnostic uses. And in many

17 ways it's the same kind of a process we're going

18 through here; we're saying when is it reasonable

19 and necessary, and for what conditions, because

20 it's used for many many things, and the evidence

21 is stronger for some uses than others. And as we

22 approach that we say, well, if it's reasonable

23 and necessary, then it's diagnostic information

24 which is useful. Now we had just such a

25 situation when we considered the use of a PET

00365

1 scan for evaluation of a single pulmonary

2 nodule. The argument was that a single pulmonary

3 nodule evaluated with a PET scan in which the

4 nodule turns out to be not metabolically active

5 or occult would eliminate the need for a biopsy

6 and in fact, we can do a PET scan, if we get a

7 negative result, we don't needed to a biopsy.

8 But that's certainly a powerful argument, and for

9 a patient making a decision about whether or not

10 to have an invasive biopsy, it certainly is a

11 reasonable option.

12 But we then looked at it and said, well

13 then, if we use PET scans to eliminate the need

14 for an open biopsy, how would we view an open

15 biopsy that was performed after a negative PET

16 scan for a single pulmonary nodule? Then we

17 would be left with the dilemma of saying if the

18 PET scan was reasonable and necessary because it

19 could prevent an open biopsy, then was the biopsy

20 after the negative PET scan reasonable and

21 necessary? It's hard to say they were both. And

22 we are faced with the same dilemma here. We have

23 a test, which we are told is useful to patients,

24 because it will allow us to more accurately

25 select their chemotherapeutic agents. We can

00366

1 avoid toxic agents which won't be effective, or

2 we can select new expensive agents which are only

3 used when they are effective. Now that's a

4 persuasive argument.

5 But then we had the organization which

6 represents most of the oncologists in the country

7 stand up and say they are neutral on this

8 procedure, but the one thing they're sure of is

9 if we allow this procedure, we shouldn't pay

10 attention to the results. That's what they

11 said. It should not be used to withhold

12 therapy. Which means, if a drug is shown to be

13 resistant we shouldn't withhold the drug, based

14 on the test, or if a drug is shown to be not

15 sensitive and it's an expensive drug, we should

16 use it anyway. That seems to me to be hard to

17 understand. That you can take a neutral position

18 about a test and say it looks okay, we think it's

19 reasonable to do it, as long as we aren't asked

20 to pay attention to the results.

21 And that gets back to the final point I

22 made yesterday, is that in terms of looking at

23 the evidence and one of the things is to look at

24 the evidence and say, where does it take us

25 clinically? And not just say, it's good enough

00367

1 to pay for but not good enough to pay attention

2 to. We should say, it's not only good enough to

3 pay for, but the evidence is so strong in a given

4 area and perhaps it's a given tumor, perhaps it's

5 a given kind of patient, perhaps it's for given

6 drugs, but if the evidence is strong enough that

7 we should pay, we should not only pay, we should

8 promote and at some point we should insist,

9 because after all, if it was reasonable and

10 necessary to do the test, if we then ignored the

11 test in future therapy, would in fact that

12 therapy be reasonable and necessary?

13 So, I am simply putting those problems

14 that we deal with in writing policy into context

15 for you, because I think you need to keep those

16 in mind as we answer these questions. Because as

17 Dr. Loy said yesterday, we need to have

18 specificity because reasonable and necessary as a

19 test means it might be reasonable and necessary

20 to pay attention to the results when that

21 happens.

22 And then just finally Dr. Murray's

23 question about what happened in terms of the fact

24 that an administrative law judge overturned a

25 claim denial. We are talking about policy here,

00368

1 policy which says, this is how it's going to

2 apply to the entire Medicare population. And

3 when we write policy, it applies to everyone.

4 When we write a policy that says we will pay for

5 PET scans for single pulmonary nodules, it means

6 we pay for single pulmonary nodules for

7 everyone. And we don't pay for another use that

8 we haven't dealt with. Now that means that we've

9 written a national policy and it applies to

10 everyone, every carrier, every beneficiary, and

11 every administrative law judge in the appeals

12 process. It is binding on everyone. That's why

13 we make national policy on bright line issues,

14 when we know which side of the line the coverage

15 policy ought to be.

16 The opposite is true when we don't have

17 a bright line. We leave it to the carriers to

18 make policy based on input from the carrier

19 advisory committee and also to review claims on a

20 claim by claim basis if necessary. And when

21 carriers review claims on a claim by claim basis

22 and make a denial of a specific claim for a

23 specific individual, that individual by right can

24 appeal that claim, and that appeal process if

25 carried to its conclusion has a hearing before an

00369

1 administrative law judge. That administrative

2 law judge hears the facts and can overturn the

3 carrier's decision to deny that claim, but can

4 only overturn that decision if there is no

5 national coverage decision in place which can

6 influence that.

7 So what we're talking about here is a

8 binding national process, not an administrative

9 law judge. When the administrative law judge,

10 and administrative law judges are with the Social

11 Security Administration, they are not medically

12 trained, and an administrative law judge hears an

13 appeal, overturns it, it applies to that

14 beneficiary and that claim only. It is not

15 precedent, it does not apply to other

16 beneficiaries, other claims, other carriers, or

17 Medicare as a whole.

18 DR. FERGUSON: So, you remind me, your

19 slide yesterday said that these procedures are

20 not, there is a national coverage policy; isn't

21 that correct? So the administrative law judge

22 couldn't have overturned the noncoverage.

23 DR. BAGLEY: The administrative law

24 judges are bound by national coverage decisions.

25 In areas where we have a national noncoverage

00370

1 decision and an appeal for an individual claim

2 goes to administrative law judge, the

3 administrative law judge is bound by the national

4 noncoverage decision. There have been cases

5 where administrative law judges have overturned

6 claims denials which were ultimately in conflict

7 with national coverage decisions, and your

8 question is, how can that happen? Well, it

9 happens, and the solution to that is that there

10 needs to be an overturning of the administrative

11 law judge's position, the denial of the claim.

12 That at times doesn't happen and the claim is

13 paid, and if it's not appealed by the government,

14 the claim is paid, so that process can lead to

15 claims payment. But in general, the policy we're

16 writing, the fact remains that the policy that

17 we're writing is binding, it is national, and if

18 all of the appeals don't go both ways, that can

19 happen.

20 DR. FERGUSON: Thank you. Mr. Barnes?

21 MR. BARNES: Well, as you know, I'm the

22 industry rep on the panel and as such, I have

23 tried to be a liaison with the proponents of

24 reimbursement, which I understood was my job.

25 And to some extent I sort of feel like I'm

00371

1 sitting at the wrong table at this particular

2 moment, and you'll understand. But let me offer

3 just a couple very quick thoughts.

4 The process, a couple comments on

5 process, I guess. One is that the industry

6 representatives here yesterday heard some

7 criticisms of studies, and in fact a lot of

8 attention yesterday afternoon was paid to quality

9 of studies and science, and it, while I'm not an

10 advocate for the industry, it does make sense to

11 see if they might have any particular thoughts to

12 share with this panel today, and I would

13 encourage the chair to allow that opportunity.

14 There was a lot of interaction with

15 regard to Dr. Burke's presentation, but I think

16 Dr. Burken's review of a great number of studies,

17 and in particular the statement that a couple of

18 things were reversed on some of his slides which

19 panel members had been looking at overnight might

20 prompt the industry to want to clarify a couple

21 of points. So I guess I would encourage that to

22 be allowed.

23 A couple of panelists so far have

24 encouraged the panel to pay attention to much

25 greater detail, to really define what cancers,

00372

1 what drugs, what tests, and I think that might

2 have been a general sense of the panel

3 yesterday. And so, my second comment on process

4 is that the way that we have conducted this

5 session over the last one plus days really isn't

6 very conducive to all of the details that you

7 three gentlemen have said you'd like to hear

8 about. And I don't have a solution for that, but

9 I think basically, I just don't see how that

10 could have worked very well, given all the

11 different variables that you'd like to hear more

12 details on, so it makes it kind of difficult.

13 The comments yesterday on the quality

14 of the research did not, I agree with several

15 panelists, did not seem to outweigh the general

16 notion that there clearly is benefit to patients

17 from this test, and that from time to time I

18 guess kind of made me a little bit frustrated. I

19 was sitting here thinking that there are a number

20 of small companies or even very small labs who

21 can't quite present the randomized clinical trial

22 that many of us would like to see, the conclusive

23 once and all for cause and effect, RCT. But they

24 do seem to present a good mass of evidence

25 suggesting patient benefit, as has been

00373

1 mentioned, and I certainly, if I personally or

2 someone in my family was involved, I would like

3 them to have access to this test.

4 Speaking about access to the test, I

5 think it was Mr. Kiesner in his early comments

6 yesterday who pointed to the numbers of hospitals

7 that send specimens in the volume of testing that

8 they do. I have spoken a little bit to these

9 folks, and it would appear that a number of

10 people in U.S. managed care organizations have

11 gone through an assessment of this testing

12 technology and have indeed decided that it is

13 something that is of value and something that

14 they are willing to submit information, rather

15 tissue for. So I guess I would be interested in

16 hearing more about who some of those plans are

17 and what kinds of evaluations have happened

18 previously.

19 The negotiated rule making came up, and

20 I understand there in fact was a consensus

21 document that was put together, and I'm sorry, I

22 am not familiar with what happened during

23 negotiated rule making, but my understanding is

24 that there was a movement in the direction of

25 recommending reimbursement for this testing, and

00374

1 also that the ASCO signed on to, I'm not sure

2 what the term would be, but did indeed agree,

3 unlike yesterday's discussion where we heard they

4 were neutral, and I think I would like to hear

5 some more about that. I think there were a

6 couple panelists involved, and it would be

7 interesting to know some more about that.

8 Evidently, the gentleman from ASCO was not aware

9 of that.

10 In my real life, a health economist, I

11 appreciate Dr. Bagley's information and

12 perspective on valued added. That was the last

13 step in the stair case of defining reasonable and

14 necessary. I would be interested in hearing some

15 more from Dr. Bosanquet, if that's possible. I

16 understood he had quality adjusted life year

17 information suggesting utility, and I believe a

18 study that has been published that hasn't really

19 been made available to the panel, it's in the

20 Technology Assessment Journal, which most of the

21 panelists probably don't see. But I think that

22 kind of information has in fact been

23 accumulated. So perhaps he will have a chance to

24 share that with us later. Thank you.

25 DR. BROOKS: My comments are a little

00375

1 bit along the lines of a number of the others,

2 and I just want to start out by saying --

3 DR. FERGUSON: Do you want to say your

4 name for the record?

5 DR. BROOKS: Yeah. John Brooks. As a

6 pathologist, I certainly came to this without

7 much knowledge or interest, in a sense of one who

8 would give chemotherapeutic drugs, and kind of

9 evaluated it in the same way as I would evaluate

10 any new upcoming test that we have actually to

11 evaluate, almost every week I would say, in the

12 clinical laboratories. As a pathologist,

13 generally, my thought would be that we like to

14 see more tests done and certainly, useful tests

15 are very helpful to people. So you know, in

16 evaluating the information that I got beforehand

17 and that we heard here, I was certainly impressed

18 with how much had been done. I was actually a

19 little bit surprised at how much had not been

20 done, however. I mean, in some settings, it

21 certainly seems to me like the data is there for

22 utility. I am not doubting that the test, I mean

23 it's been mentioned before by a number of people

24 that, you know, I actually believe that the test

25 does test resistance and so forth, and that we

00376

1 may have to decide which of the tests might be

2 recommended, or maybe two tests, A and B,

3 whatever, but histology specific type of data

4 wasn't necessarily there, except in certain

5 situations.

6 You know, the hematopoietic

7 malignancies certainly seemed to have really

8 pretty good hard data. For example, one question

9 that kept occurring to me is how often in an

10 individual tumor type, not a site, not just

11 ovary, but a specific type in that ovary, because

12 that's what the clinician diagnosis is. And that

13 kind of information, for example, is available in

14 CLL. And the articles provided to us this

15 morning tell me that you know, for example, 12

16 percent of patients with CLL, which I view as a

17 pretty high number for a very uniform type of

18 disease, you know, where the markers, et cetera,

19 are quite similar case to case, showed a

20 difference. Whereas, if I had had such data in

21 small cell, et cetera, I certainly would be much

22 more persuaded that the test is useful to

23 people.

24 In other words, if we're trying to

25 define a policy, and suppose we had a tumor whose

00377

1 data showed it never, it always had the same

2 resistance pattern, then you wouldn't want to

3 necessarily, although I may be foolish, to order

4 over and over again to get the same result.

5 All that said though, you know, I do

6 view that there is data there that shows some

7 clinical utility. The question in my mind then

8 becomes, how would you write a policy,

9 et cetera. What I would like to see, just

10 apropos, forget chemotherapy is something that

11 says, you know, people who have diabetes ought to

12 get a glucose test, and we'll pay for that every

13 time. Okay.

14 So then I come to the issue of denial,

15 because as it was brought up before, we had ASCO

16 there saying that, you know, sure, go ahead and

17 do the test, be we might not pay any attention to

18 it, and we sure want to give the drug that's

19 resistant. I mean, that's kind of what I heard.

20 I was thinking about that overnight and I was

21 wondering, well, okay, can I think of something

22 to explain that position? And even though I'm a

23 pathologist, I actually talk to people a lot.

24 We're actually not in the closet, and we're out

25 in the public, and I love patients, and in fact,

00378

1 patients call me all the time. Okay.

2 So let's take diabetes. So suppose we

3 have a diabetic test, suppose we don't have a

4 glucose test, and the new proposal is, I've got a

5 better glucose test to tell if somebody has

6 diabetes. And if they have a sensitivity

7 specificity of 80ish percent, or maybe even 90

8 percent. So okay, now I have a patient and I am

9 a clinician treating this patient. The patient

10 has a negative new glucose test, and it's

11 negative. Should I be denied giving insulin to

12 this patient? It sounds pretty reasonable. But

13 maybe, you know, the test is imperfect, and maybe

14 by looking in the eyes and by looking at the

15 weight and by looking at this and that, I could

16 figure out a way as a clinician in an imperfect

17 world, and certainly not with our glucose tests

18 that we have now, that you could in fact see

19 people should have insulin because of a, not gut

20 feeling, but as a group of symptoms and signs

21 that I see in the patient, that I should be able

22 to give insulin to that patient.

23 So with that said, I'm a little bit

24 torn in fact, as to the issue of if a test is

25 really good and it shows a negative result,

00379

1 should you deny the drug being given. And I

2 guess that's kind of where I am. And in a sense,

3 you could look at things that happened in other

4 arenas. I'm aware that when they were doing the

5 surgical treatment, you know, surgery versus

6 medical treatment for coronary artery disease,

7 and the surgeons said that, you know, the

8 cabbage, or bypass was better, it took us ten

9 years to figure out what that was all about,

10 mainly because there was no clinical trial.

11 So another question that occurs to me,

12 and I don't have the answer, and I want to hear

13 what happens the rest of this morning, is if this

14 is approved, do we want just America to use it

15 willy-nilly? That is, this person uses it over

16 here to treat this person, et cetera, and nobody

17 gathers any data. And I don't mean gathering

18 data by the companies, I mean publicly gathered

19 data. So another question therefore that occurs

20 to me is that if we approve it for specific

21 histologies or, you know, certain restricted

22 diseases or otherwise, should the -- should this

23 be for a period of time and should the data be

24 gathered by an independent source? Now that's

25 not to say a clinical trial. A clinical trial,

00380

1 you know, you have mentioned is very difficult to

2 start, especially if it is only directed at the

3 DiSC assay or whatever. But for example, you

4 could just have the results go into a central

5 repository, and the physician who used the test

6 would be required to say exactly how he treated

7 that patient, for how long, and was it on

8 protocol or off protocol.

9 My final question is, what does this do

10 to clinical trials? And we have large public

11 groups, cancer, breast cancer groups, prostate

12 cancer groups, et cetera, working very hard to

13 define what are the best therapies for each of

14 these cancers. Now those cancers at least are

15 more defined amongst themselves. If we bring in

16 such a test as this, do we undercut what's being

17 done in those trials, because, in other words, I

18 want to dovetail them, but I need to know how. I

19 don't want to see suddenly everything being done

20 willy-nilly based on a test that after all, only

21 seems to have an 80 percent or so sensitivity.

22 In other words, there are patients who are

23 resistant who may respond. I think that's the

24 question.

25 So with, those are the questions I have

00381

1 in my mind, and I would like to listen to more.

2 DR. FERGUSON: Paul Fischer.

3 DR. FISCHER: In preparation for this

4 meeting I called a couple of the thoughtful

5 oncologists in the Augusta area and asked them

6 about their knowledge and experience with this

7 test. They were uniform that in their opinion

8 that it was not something that was useful because

9 people did not behave the way the test would

10 predict when they were given chemotherapy. And

11 what I realized after listening to the folks

12 yesterday was that there are really two cultural

13 views here. The one cultural view is histology

14 driven. We look under the microscope, we see a

15 particular histology, and we therefore know what

16 drugs to give. The other world view is the test

17 tube driven world view. And when they were

18 showing slides yesterday, one of the speakers

19 said, well, I know what to give because I see all

20 these cells got killed. So it's a really

21 different way of believing what is going on in

22 the world.

23 And then the question for me becomes,

24 do the champions of the technology who spoke

25 yesterday, do they represent pioneers or nuts?

00382

1 And I think that's what HCFA has to decide,

2 because does HCFA advance the standard of care

3 where the average oncologist says we don't

4 believe in this, but HCFA is going to pay for

5 this so let's try it, or do they support the

6 current standard of care, which is to let the

7 histology world decide what to do.

8 The problem I have as a family doctor

9 is that in my total practice experience I believe

10 that probably my patients have been hurt as much

11 by chemotherapy as they've been helped. And

12 that's even given some of the wonderful drugs

13 that do respond beautifully to chemotherapy. But

14 I regularly protect my patients from oncologists

15 who have this histology world view. If there is

16 a tumor, it needs chemotherapy. And if you

17 didn't respond to the first one, you get the next

18 one. So eventually you're given drugs that are

19 more and more toxic, less likely to benefit the

20 patient, but because you've still got tumor, you

21 need another course of chemotherapy.

22 Now my way of dealing with this in

23 practice is to be very selective of who I send my

24 patients to, and some oncologists are

25 conservative and some are not, and I avoid the

00383

1 latter. But I, you know, on a weekly basis sit

2 down and talk with a husband and wife and talk to

3 them because they had some advice from an

4 oncologist that they should go through this

5 chemotherapy and to be quite frank about it, it's

6 not always a very balanced view of whether the

7 patient is going to benefit from it. So I really

8 believe we need to move beyond the current

9 situation and put some brakes on this histology

10 driven world view which encourages more and more

11 chemotherapy given with less and less benefits.

12 And I'm not sure we need to say that we

13 have proven that this technology is useful for

14 every tumor and every drug. You know, obviously

15 they haven't. But clearly they have in CLL. You

16 know, I'd like the people who give chemotherapy

17 to stop for a second and say well, gee, maybe

18 there are some other ways to think about who

19 needs what, and this seems as good as any

20 approach currently available, and I would

21 therefore, I will vote to support some sort of

22 funding for this when we get to the end of the

23 morning.

24 DR. FERGUSON: Dr. Klee.

25 DR. KLEE: Hello. My name is George

00384

1 Klee, and I guess as I looked at the data that we

2 had sent to us early on, it looks like you have a

3 valid laboratory test. But where the issue sort

4 of comes is where does this fit into the practice

5 of medicine and how does it improve patient care

6 in the longer run? You know, if we were to look

7 it in sort of a protocol design, a selection of

8 which drug should be used in which diseases, that

9 seems to be a legitimate application that could

10 go forward. Looking at it on individual patients

11 is where the issue seems to come into play, and

12 if so, which patients and for what decision

13 purpose are we looking at it?

14 The data that was sent with the packet

15 seemed to indicate that the best utility for the

16 test was in the negative predictive value. That

17 is, in response to determine which patients are

18 not likely to respond to chemotherapy. And I was

19 pretty well convinced of that until yesterday's

20 presentation by Dr. Burken. And going through

21 the numbers that were in those slides, and I

22 tried to tally them up last night, there are very

23 few that are up in that 99 percent negative

24 predictive value. You know, there was a few of

25 them that had numbers less than 20 that had 100

00385

1 percent negative predicted value, but there was a

2 lot of cases that were, a lot of publications

3 that were referenced there that had negative

4 predictive values in the 20 to 60 percent range,

5 which doesn't really make too much sense if you

6 have an a priori odds of 70 percent, that

7 apparently these are not prevalence adjusted.

8 And although the presentations went

9 through a lot of explanation of the Bayesian

10 theory, it would be nice to have these numbers

11 prevalence adjusted. But even so, if you start

12 with this at a 50-50 with those lower odds, it's

13 getting up to a point where it's not looking like

14 this test would really be that useful. You know,

15 if you had something that's, you know, you have a

16 priori odds of 70 percent, and you can only get

17 it up to 80 percent chance that this drug is not

18 going to respond, that's not very much of an

19 improvement over the prior. You know, if you're

20 taking it up so you've got a chance of one in a

21 hundred that this drug is not going to respond,

22 then I think we've got something we can use in a

23 clinical decision. So I guess I would like to

24 see some of that information further clarified

25 and presented in the form that, you know, several

00386

1 presenters have indicated that the Bayesian

2 theory is needed, and it's the post odds that

3 we're looking at in terms of the negative

4 predictive value on these tests.

5 I guess I'd like to expand on several

6 of the issues that were raised by other members

7 of the panel here in terms of, we've got a

8 multitude of diseases, and where does this fit

9 in? You know, we've got a multitude of drugs,

10 we've got a multitude of subclasses of these

11 diseases in terms of whether it be histology or

12 whether it be in terms of just other clinical

13 classifications as to the stage of the disease

14 and things like that. It looked like it's too

15 big of a matrix that we're trying to deal in

16 here, and it didn't look like one size fits all

17 for the answer to that.

18 And I agree with the general assessment

19 that the leukemias, CLL in particular, seemed to

20 be one where the focus looked a little more

21 channeled in an area where we can say that there

22 is a definite improvement based on at least a few

23 trials that have been out there. But at the same

24 time, when we look to see, there was a question I

25 was asking there yesterday, has it really been

00387

1 compared to using this test versus not using the

2 test. That seems like where you would find out

3 whether or not there is a benefit for the test.

4 And those studies have not been done.

5 And I would like to maybe inquire a

6 little bit more of Dr. Bagley, that at our

7 orientation session, you had mentioned that there

8 is some new activities going on with HCFA where

9 tests can be prospectively monitored in terms of

10 utility, and perhaps introducing this in a

11 disease management strategy with controlled

12 output, you know, similar to what Dr. Brooks was

13 alluding to, but is that something that through

14 the funding mechanisms of HCFA, that this could

15 be carried out, since NCI doesn't look like they

16 have carried this out recently. They looked at

17 it many years ago. And if this is not part of

18 the clinical trial approach, we need an alternate

19 way to do this. It doesn't seem like it's

20 something that's a yes no answer, perhaps a

21 controlled introduction in a very focused disease

22 with the output monitoring required as to what

23 happened to these patients, was there benefit in

24 terms of quality of life years or in terms of

25 survival, or in terms of any other parameters you

00388

1 might put up. But it looks like if we're going

2 to take the practice of medicine to this higher

3 plane that has been alluded to several times, we

4 are going to need to do it in a controlled

5 manner. Somebody's going to have to pay for it.

6 It's not reasonable, I think, to have the burden

7 of this put back into the people that may be

8 making the test, but it needs to be looked at in

9 terms of those that would be benefitting from it

10 and as it is part of health care policy, then it

11 looks like it should be, you know, integrated

12 through, whether it may be a joint venture

13 between NCI and HCFA, to say how do we carry this

14 out in a manner that we can say, give this a

15 controlled trial over X number of years, see

16 what's going to happen, only apply it in a very

17 limited focused area, for example CLL, and then

18 try to see, did it make an impact, did it change

19 the way that we're caring for these patients?

20 Did it change the benefit in terms of life

21 expectancy or quality of life years for the

22 patients.

23 DR. BAGLEY: Let me answer that real

24 quickly, and I'll try to make it a quick answer,

25 because that's a long complicated issue. And it

00389

1 has to do with when should Medicare become

2 involved in issues that are not quite settled,

3 how should we continue to look at things, and

4 what can do? I'll give you some examples and

5 tell you where we do do it then, and then tell

6 you how limited that option is.

7 In two areas of diagnostic testing, we

8 have recently written cautious coverage policies,

9 not that different from what we're dealing with

10 here today, new technologies have a lot of

11 promise, clinical community doesn't know quite

12 how to use it, they are not so sure they're going

13 to use it to replace something else, we're not so

14 sure it's going to improve care, but we

15 cautiously advance coverage, and say we'll not

16 only pay for it, but in paying for it we will

17 collect some information. Now we can't collect

18 very much information. By law, we can only

19 collect the information necessary to process the

20 claims. But we can interpret that to the point

21 of saying we can get a certain amount of clinical

22 information because that's what we process claims

23 in.

24 Those two example list are magnetic

25 resonance angiography of the head and neck, which

00390

1 is new and controversial. We added some coverage

2 for MRA for head and neck vessels, and at the

3 same time we cover it, we gather some information

4 on what the indications were and how it's used,

5 so that we can continue to evaluate it and say is

6 it making a difference.

7 We're doing the same thing with the

8 other example I mentioned, PET scans. PET scans

9 for single pulmonary nodules which, you know, the

10 promise was, it's a better less invasive way to

11 monitor these patients. We're collecting the

12 information from the claims in such a way that we

13 will have an idea of what the experience was, and

14 are these patients avoiding biopsy, or are they

15 getting some other ones. That's one place to do

16 it.

17 But remember that in a very limited way

18 we have gone forward and said that we are very

19 cautious. And we have to be very clear that

20 we're not doing research, because research is not

21 only not Medicare's job, it's prohibited. So we

22 are not doing research, but we are at least

23 monitoring early diffusion of technology.

24 The third place we have done that is in

25 another area called lung volume reduction

00391

1 surgery, which is a new surgical procedure. It

2 has the promise of helping patients with

3 emphysema. It was early touted as a miracle

4 cure, some people still believe it is, and we are

5 promoting it very heavily. And we looked at it

6 early on, and realized that this would be another

7 issue such as is turning out to be, bone marrow

8 transplant for breast cancer, something that

9 after ten years not only was never proven to be

10 beneficial, in fact it turns out it may be

11 harmful. And so, lung volume reduction surgery,

12 we're gathering information at the same time

13 we're paying for it in limited clinical trial.

14 But in general, it is very difficult

15 and may not even be possible, although it's

16 tempting at times for an issue like this, to say

17 gee, it's early, let's pay for it and keep an eye

18 on it, and then we will sort of stimulate the

19 research.

20 In the previous panel we had a multiple

21 myeloma, which is at the current time

22 investigational, and the question was, is it

23 ready to move to prime time? We made it clear to

24 that panel and I think it ought to be clear to

25 this panel and future panels, that when Medicare

00392

1 covers in an area that -- the day of the clinical

2 trial is gone. We can accumulate information

3 from case series, we can watch patients, but the

4 day of the clinical trial is over when Medicare

5 starts to cover it. There will be no

6 randomization for that procedure ever again,

7 because we have made the decision by coverage of

8 saying randomization is no longer necessary, we

9 have the answer.

10 DR. FERGUSON: Dr. Sundwall?

11 DR. SUNDWALL: Thank you. I feel

12 fortunate going at the end of this discussion, so

13 I get the benefit of all these experts, and I

14 don't mean to be flip about that. I really

15 appreciate the expertise of the panel, more

16 analytical than those of you who carefully

17 reviewed the studies. My perspective is that of

18 a family physician like Paul, who is taking care

19 of patients, diagnosed cancer, followed them

20 through chemotherapy, and seen sometimes the

21 benefits, more often the heartache and morbidity

22 that's associated with that. That doesn't mean

23 I'm not a believer, it's just that I think we

24 need to have a very healthy skepticism about

25 current cancer treatments.

00393

1 I am also coming at it from the

2 perspective of someone who's spent most of my

3 adult life in health policy here in Washington in

4 different capacities, and I'm very familiar with

5 the tension between those advocating something

6 new and wonderful, and the payers who in their

7 responsibility to monitor how we spend our funds,

8 make sure they are done appropriately.

9 However, I must admit that through the

10 course of the discussion yesterday, I was

11 reminded of a quip I once heard about economists,

12 which was intended to be funny, but the

13 economists were described as people who, when

14 something is proven to work in practice they want

15 to find out if it works in theory. And it seemed

16 to me the preponderance of evidence was that this

17 is in fact a useful tool, it's information, I

18 think in some respects it has been oversold in

19 what it promises, but I look at it more simply as

20 Grant described, it's information. And it

21 clearly would be useful to me and my patients in

22 making decisions about chemotherapy.

23 I think to put it in context of

24 reasonable and necessary is wise. It is

25 reasonable. Whether it is necessary, I think

00394

1 must be left to the discretion of those experts

2 who are going to be using it.

3 And Grant, I just have a quick question

4 for you, because I'm not clear in my mind, maybe

5 I should be at this stage of our panel's life,

6 but I'm not. Let's assume for the sake of

7 discussion that this panel has consensus that it

8 is reasonable to pay for. However, to determine

9 what's necessary is going to, because I know

10 about this having participated in the negotiated

11 rule making, that won't be done willy-nilly as

12 you were talking to Dr. Brooks, that won't

13 happen, because Medicare doesn't willy-nilly pay

14 for anything. It will either be then subject to

15 development of a national coverage policy, or

16 left for medical review. And how do you envision

17 the next step? Will it be an LLMP or will it be,

18 will you ask us, not us, but will you convene

19 another panel or will you ask us to reconvene and

20 develop what we in negotiated rule making did?

21 For those of you who don't know what that was, we

22 did wrestle over several months and developed

23 coverage policies, was it 23 tests or 24?

24 DR. BAGLEY: It seemed like 124, David.

25 DR. SUNDWALL: For those of us who

00395

1 participated, it was a 13 month process in total,

2 and it felt like going to the dentist for three

3 days at a time, but we did it, and it was a

4 useful thing. So do you think this kind of

5 testing would be subject to a national medical

6 review policy or would you leave it at the

7 discretion of the carriers?

8 DR. BAGLEY: Well, given the fact that

9 the whole negotiated rule making process was

10 driven by mandate from Congress, and the goal was

11 that there be national policies for laboratory

12 tests, and even when they weren't national

13 policies they should be uniform across all the

14 regions, I think it would be our goal to in one

15 fashion or another to try to make the policies as

16 nationally uniform as possible. That means

17 making them as specific and data driven as

18 possible, and that means being as specific as

19 possible in what we're dealing with. And most

20 likely, it's going to require something beyond

21 what this panel will do in terms of mining

22 through the literature. So should this panel

23 make a recommendation that we be, that we go down

24 a path of looking for specific kinds of

25 indications, of drug tumor combinations in which

00396

1 the evidence is somewhat better, then I think we

2 would undertake the task of trying to mine the

3 literature for that purpose. But I don't see

4 reconvening this panel. But I also think our job

5 would be to try to find a consistent policy which

6 could be applied nationally, whether it be

7 regionally uniform or national, I think we would

8 try to have some uniformity.

9 DR. SUNDWALL: Because I think that

10 would calm a lot of the concerns that I have

11 heard about today, the idea that if some of us

12 say that it's okay, then it's just universally

13 available. I think there needs to be guidance

14 and expertise on how it's appropriately applied.

15 The last comment I would just make is,

16 I think from a health policy standpoint, we keep

17 talking about quality access and costs, and

18 access I think is important, and this is a useful

19 technology and ought to be accessible to Medicare

20 beneficiaries appropriately applied. I do think

21 the cost issue is promising. I was disturbed by

22 the testimony of the American Cancer Society

23 clinical oncologist yesterday. It seemed very

24 self serving. We're neutral, but by the way,

25 cover chemotherapy, and don't you dare tell us

00397

1 what we can't use. And I really thought that was

2 not constructive, and I think it's troublesome.

3 And I mean, you must hear this all the

4 time. A typical provider tells HCFA, I don't

5 want much, I just want more. But I thought the

6 oncologists were not constructive, and I'm

7 perplexed why they were not more supportive of

8 this when in fact they were during negotiated

9 rule making, and I assume this young man was

10 speaking on behalf of the society. But I do

11 think in the event this is approved, I think one

12 of the questions you posed before us merits

13 discussion and hopefully we'll have it through

14 all of the tests we talked about that, and that

15 is, no, I don't think HCFA should pay for things

16 that don't work.

17 DR. FERGUSON: Okay. Miss Simmers?

18 MS. SIMMERS: I think this fits into

19 the category of the tools that a physician should

20 use in order to dose, or what course of treatment

21 to endeavor and to go in that dialogue between

22 the patient and the physician and the family.

23 I think the one presenter and one panel

24 member pointed out something that I think about

25 all the time, at least in one region, and I think

00398

1 it unfair to have Medicare beneficiaries pay

2 their insurance premium and not get access to

3 that test. So I'm very much the 35,000 foot view

4 of the Medicare beneficiaries' needs and how well

5 what we decide to recommend would serve those

6 beneficiaries. And I think Dr. Bagley's point

7 about they help pay for it is very valid. They

8 do help pay for it, and as insurance carriers are

9 paid, I think the same level of access should be

10 available to Medicare beneficiaries that there

11 are provided for those who have private

12 insurance.

13 I have some concerns, however.

14 Certainly the accessibility of the test, although

15 I am convinced there are ways to address that, I

16 think there would be a duty to promote this as,

17 promote its accessibility not only in very small

18 regional areas, but nationwide, should we decide

19 to cover it. I think there are some valid

20 research and scientific questions to be

21 answered. I do believe that this is a diagnostic

22 test and should be held to that criteria, and not

23 to that of therapy. I think that was clear, and

24 although I think the science could be better,

25 it's not a perfect world, and what we see is at

00399

1 least compelling evidence to continue.

2 The policy specific issues, certainly I

3 am not the expert in that area, but is does seem

4 to be available to us to look at those cancers

5 and what circumstances with what drugs are best

6 pursued under this coverage recommendation,

7 should we make it. It seems to me the processes

8 are in place to do that, so I don't fear that it

9 can't be done. And certainly, the devil will be

10 in the details of that kind of process.

11 But after everyone has said their

12 piece, I guess if any of us in this room were to

13 face the terrible news from our clinician

14 tomorrow that we had cancer, who among us would

15 say no, this assay is not for me. And I believe

16 if it's for us, it is for Medicare beneficiaries

17 as well. So that's my comment.

18 DR. FERGUSON: Thank you. Before

19 asking some more of the panel, I have one

20 clarification, Grant, or anybody else. As I

21 understand it, our job is really to answer these

22 questions, which can be posed in the form of

23 motions. For us to add some more questions like

24 should Medicare cover this, yes or no, for our

25 panel, or under what circumstances for our panel,

00400

1 I did not think was our job. Has that been

2 changed?

3 DR. BAGLEY: No. We spent a long time

4 with these questions as staff. We spent a great

5 deal of time toiling over them. And the reason

6 is that the answers to these questions we think

7 are particularly relevant in giving us some

8 guidance in developing a policy. So I think we

9 are particularly interested, and the number one

10 goal should be, you know, to go through those

11 questions and give us not only answers, and they

12 aren't really yes or no questions necessarily,

13 but to give us some discussion and to give us

14 some rationale. And that's one of the biggest

15 reasons for having this transcribed word for

16 word, so that discussion around those questions

17 can, we can use as guidance in trying to develop

18 some policy.

19 So I think once these questions have

20 been dealt with, and we can have the discussion

21 and the committee's thinking around these

22 questions, the reason for number six is that the

23 committee can then entertain some other issues,

24 and I think we can hear the other concerns from

25 the committee or from individual committee

00401

1 members. But I think at this point, it is

2 particularly important and relevant to deal with

3 the questions presented, because we felt those

4 were necessary for us to have the direction for

5 policy.

6 DR. FERGUSON: I was going to ask that

7 we handle question five last and put six before

8 it, because it seemed that the fifth question was

9 sort of the bottom line question. Is that --

10 DR. BAGLEY: That's not inappropriate,

11 as long as, before lunch time, we deal with

12 question five.

13 MR. MINTZ: I have a question. Grant's

14 comments just prompted a question. I had the

15 privilege of serving on the myeloma panel, and we

16 went beyond the questions certainly to the very

17 direct question of coverage, you know, by virtue

18 of motions that were made during the course of

19 discussion. I'm interested in your comments

20 about that, and whether you felt the panel sort

21 of exceeded its charge in that regard, or whether

22 that was just something that happened in due

23 course and was appropriate.

24 DR. BAGLEY: Our lesson learned from

25 the myeloma panel and again, the reason for

00402

1 toiling over those questions as we did, our

2 lesson learned was, number one, the committee

3 discussion was far too short. We had lots of

4 presentations from the outside, the committee

5 didn't have much discussion, so when we took it

6 all home to try to make sense of it, we thought,

7 I wonder what this committee member was thinking,

8 I wonder what they were thinking. So that was

9 number one lesson; we wanted to know why the

10 committee is making the decisions they are, not

11 just do this, do that.

12 Lesson number two is, it's very easy

13 for the committed to sit here and say, and we've

14 heard a sense of that this morning, you know,

15 this sounds pretty reasonable and I think I would

16 want it, so other people ought to have it

17 available. On the other hand, it needs to be

18 fit, we need to look at the science, the policy

19 has to be restricted and figure out where it

20 fits, and maybe gather some data, so we think

21 HCFA ought to pay for it, but only when

22 reasonable and necessary, and they ought to

23 gather some data. We knew that coming in, so our

24 sense is that that's the policy direction we need

25 to go, and we need to hear from the committee.

00403

1 Not that the committee can say, cover it for this

2 ICD-9 code and this one, but we need some policy

3 directions on what we need to look for.

4 So, the lesson learned from myeloma is

5 -- and that's why we fashioned the questions, so

6 I don't think it's totally inappropriate for the

7 committee to make some expressions, but the

8 myeloma panel left us to cover this, but I really

9 think it's necessary, and we did have good

10 people, and the right kind of people and figured

11 it out. That's not being that critical, but you

12 know what I mean.

13 DR. MINTZ: But it did direct, it did

14 have a specific vote on coverage. It went

15 through cytogenetics and it went through each of

16 the clinical conditions and voted 11 times, I

17 think, to cover under a variety of circumstances.

18 DR. BAGLEY: That's very true, and

19 we're cognizant of that. But I think our goal in

20 fashioning these questions, at least as far as

21 threshold issues, were to give us the information

22 that we need for policy.

23 DR. MINTZ: So it is a different

24 direction in that sense?

25 DR. BAGLEY: Yeah, I think it is. And

00404

1 I think it also makes it clear that HCFA, after

2 all, cannot abrogate the coverage decision making

3 to the panel. The panel is advisory and the

4 panel needs to give us advice so that we can make

5 the coverage decisions, and simply having

6 battling coverage decisions doesn't help the

7 information we need to make the decisions.

8 MR. MINTZ: Just for this panel's

9 information, the myeloma panel voted to cover

10 everything but refractory relapse, and went

11 through a whole series of votes to do that. So

12 it is very different from what we experienced.

13 DR. BROOKS: Dr. Bagley, I guess your

14 response to one of the other questions and

15 whether or not, you know, clinical trials are

16 over when Medicare decides to make a decision is

17 like very very crucial to me, because we have

18 data on certain areas that I think we all agree

19 at least has clinical utility. But what we

20 actually don't have in each individual, patient

21 by patient, whether they got the sensitive drugs

22 or not in all those studies. And so, I still do

23 look upon it as a potential willy-nilly thing

24 unless we collect data as we go, and maybe

25 reevaluate in a few years as to what -- because

00405

1 you know, we have that 20 percent problem there,

2 and that was brought up in the negative

3 predictive value.

4 So, I would really like to know whether

5 or not in a coverage decision, you can mandate a

6 trial that is, you know, it may not be the usual

7 clinical trial that everyone thinks about, but at

8 least, that you can say, I'm going to have the

9 NCI or the CDC or somebody, collect data as we go

10 here, that if people are not using the test to

11 give the sensitive drugs after a few years, then

12 what's the sense of having the test? Or if

13 everybody's just ignoring the results and so

14 forth, or in fact, if when they get a negative

15 result and people went ahead to treat with the

16 negative drugs and they responded 25 or 30

17 percent of the time, which is about the 30

18 percent of the time they respond to almost

19 anything, so I would like to really know that.

20 And so, what's your answer to that question as to

21 whether you can require in a coverage decision

22 that at least ongoing, much more than your

23 claim? I mean, I don't believe you can evaluate

24 this by your claim data, you know, who got what.

25 DR. BAGLEY: Well, it's -- it would be

00406

1 long and time consuming for us to try to tie a

2 coverage decision to what NCI was doing, and it

3 would probably, we would be here years from now

4 talking about the same issue. That's one

5 possibility. But the other one is that normally

6 if it requires a clinical trial, we don't cover

7 it, and if it's gone beyond that and then

8 diffuses, we do cover it. Now as I mentioned,

9 there are ways in which we can cover things and

10 we can gather information, and by law we can only

11 gather what's necessary to process the claims.

12 But it's surprising, the amount of clinical

13 information that that makes available to us. It

14 is a fairly rich data resource.

15 And I mentioned, for example, PET scans

16 for single pulmonary nodules. Now what that

17 policy says is that, it says that the purpose of

18 this is as an alternative to open biopsy, and

19 that we would not expect to see open biopsies

20 after single pulmonary nodules. That's what it

21 says in the narrative of the policy. Does that

22 mean that we would never pay for an open biopsy

23 after a negative PET scan for a single pulmonary

24 nodule? No, of course not, because there are

25 going to be clinical circumstances in which it's

00407

1 necessary. But it does say, and it's an

2 expression, and what we did, the way we described

3 it is that we said, you know, when you approach a

4 PET scan for a single pulmonary nodule, you have

5 reached a fork in the road.

6 And we would look at sensitive testing,

7 or resistance testing, as a fork in the road. We

8 have a test. We're going to do a test, and that

9 test is a fork in the road for us, and depending

10 on the result, we're going to go down path A or

11 path B. Now, I think that a physician and a

12 patient should have a full discussion of that

13 fork in the road before they go down a path. And

14 to do the test and say, well, this test was a

15 fork in the road but don't like the answer, let's

16 go the other direction, makes you wonder whether

17 the test was necessary or whether the direction

18 is necessary.

19 And so what we would do, hypothetically

20 for example, is that the policy would say that we

21 would ordinarily consider it not medically

22 necessary, or not reasonable and necessary to use

23 chemotherapeutic agents shown to be resistant or

24 not sensitive by this testing system. That means

25 that an oncologist ordering the test, and then

00408

1 embarking on a course of therapy which was not

2 shown to be beneficial by the testing would be

3 very likely to having a carrier medical director

4 say, I think these claims are not reasonable and

5 necessary, I'm going to deny them. And then they

6 could go practice their medicine in front of an

7 administrative law judge.

8 And I think that would not be an

9 unreasonable kind of policy. That would allow us

10 to gather some kinds of information, and we would

11 see where it goes. But to say we are going to

12 put a clinical trial together as part of our

13 coverage, I mean when we start paying for it, the

14 ability to collect information, and the impetus

15 out there, and the stimulus for people to do

16 clinical trials goes away. And we've seen it

17 time and time again. And bone marrow transplant

18 for bone cancer, I think is a perfect example.

19 When payments started, the clinical trials

20 stopped. And it was only after many many years

21 of accumulated data that we then found out that

22 this was not a wise course.

23 DR. FERGUSON: Dr. Hausner?

24 DR. HAUSNER: This is a question for

25 Dr. Bagley, so that I can understand the

00409

1 process. The test itself would be relatively

2 expensive, say compared to a CBC, in other words,

3 whatever the price is, and I know we're not

4 talking about price, but let's establish that. A

5 concern that I have, and this is, I practice

6 pathology in a world probably quite similar to

7 the world that Dr. Fischer practices his family

8 practice, and the question of proper utilization

9 is very much on my mind. Would the decision to

10 pay for the test eventually reflect itself in a

11 hospital DRG? In other words, would the test be

12 so expensive that the DRG for say CLL have to be

13 modified in order to incorporate the test?

14 The second question that I have, and

15 it's a concern particularly for CLL, which is a

16 malignancy that sometimes isn't even really

17 treated in an older patient necessarily, would be

18 how to avoid this becoming a standard beginning

19 of disease test that might or might not be used?

20 Is there any way to monitor, and I know you can't

21 say, well, by God, if you do this test, you'd

22 better get some chemotherapy. I am not quite

23 advocating that. But you know, do you understand

24 the flavor of what I'm asking? Because from my

25 point of view, I want to do the right thing for

00410

1 the patients, but at the same time, I don't want

2 to inadvertently give a group the key to the

3 treasury.

4 DR. BAGLEY: Well, nothing in Medicare

5 is simple. Let me start out by saying that.

6 It's always complicated. And the one, well, one

7 of my adages in Medicare is that however

8 complicated it seems to be, it will be more

9 complicated by the time you get to the end. So

10 how this would be paid for and in what fashion

11 is, I think is something for the future to

12 decide. Obviously a patient in the hospital who

13 undergoes surgery or biopsy and the tissue is

14 harvested in the hospital, then the test itself

15 is going to become part of the DRG. Now if that

16 means that the cost of many malignancies is going

17 to go up, there is a way, and the mechanism is

18 that the DRG over time can reflect increased

19 costs. But it's not rapid, it's a slow process.

20 Would all of the costs of this be part of the

21 DRG? No, it wouldn't be, because some things are

22 not included in the DRG. You know, physician

23 services are not included in the DRG. So to the

24 extent that some of these are physician services,

25 pathology services, consultation service, you

00411

1 know, and a certain portion of these tests are

2 evaluation and billed as consultation services,

3 not as laboratory services, then they wouldn't be

4 in the DRG, but they would be in the part B

5 physician fee schedule payable out of part B, and

6 they would become simply payments from that

7 mechanism. So you know, it's going to be fairly

8 complicated, but I think it's -- you bring up the

9 point that the inpatient part of this is going to

10 be in the DRG, and it's going to impact on how

11 hospitals choose to have oncology patients taken

12 care of and their admission status.

13 DR. FERGUSON: Kathy, and then Dr.

14 Fischer.

15 DR. HELZLSOUER: This is Kathy

16 Helzlsouer. I just want to raise for discussion

17 among the panel how we are going to define

18 clinical utility, because I think that's the crux

19 of it and really the first question brings this

20 out. If utility is going to define as, is this a

21 marker of response rates, I think that's what the

22 literature has been designed to show. I think

23 the appeal that Dr. Fischer did, that gee, if I

24 could get oncologists to stop treating when it's

25 appropriate and not have all the side effects,

00412

1 that's great, but we don't have any evidence that

2 this would change clinical practice, that those

3 oncologists that you don't refer to because they

4 say never never never quit, they are probably

5 still going to say never never never quit, they

6 may not even do the tests and if they do it, they

7 probably won't always use the results. And we

8 don't have any evidence that it's -- would it

9 change practice or should it, because of this

10 issue of false positives, false negatives. And

11 negative predictive values, we learned yesterday,

12 is highly dependent on the prevalence of the

13 population under study. So we have to be careful

14 of the literature that we're looking at in using

15 negative predictive value or positive predictive

16 value to guide us.

17 So I think, like I said, it seems

18 reasonable to me, it gives them added

19 information. If that's the criteria, then that's

20 fine. But if we really want more than that, and

21 we expect this to change practice, I guess I'm

22 not convinced that it would or should. And we

23 don't have -- the issue of the added value, I

24 think comes up, because in the literature that we

25 have here, with the exception of the CLL study,

00413

1 which I am just looking through this morning, is

2 what is the added benefit beyond what we already

3 know, which is what Dr. Brooks is raising. Do we

4 have all the other markers that are now being

5 used and accepted, not just histology, but other

6 prognostic markers really define who's going to

7 ultimately respond and not respond. Does this

8 add something to that information or doesn't it?

9 We don't have from the studies that have been

10 done to date anything to say what the added

11 benefit of this test truly is, with the exception

12 of perhaps CLL.

13 So I agree with you. I think that as I

14 said, a huge benefit would be if you could

15 eliminate toxicity from agents that would not be

16 used, and that may be the case, at least the

17 first line. But you're talking about a specific

18 situation too, which is the other overriding

19 issue. You're talking about metastatic end stage

20 disease when you were jumping to chemotherapy.

21 So that's what I'm struggling with. I think if

22 the diagnostic tests say it's not in the patient,

23 there's obviously no harm. The only harm would

24 be, though, it it's used to guide therapy and

25 there's a significant rate of people that would

00414

1 not get therapy that would be potentially

2 beneficial, with 20 percent response rates in the

3 face of a test that says resistant. So if the

4 policy is tied to eliminating most chemotherapy,

5 I think there is the potential for harm here in

6 how the tests would be used.

7 DR. FERGUSON: Dr. Fischer?

8 DR. FISCHER: Yeah. We talked

9 yesterday about where the bar should be set for

10 the level of evidence. I would hope that the bar

11 wouldn't be set at the level where the people who

12 were coming up with this test have to demonstrate

13 that somehow oncologist's behavior has changed

14 for the better. That's much higher than we need.

15 DR. HELZLSOUER: That's your argument

16 for using the test, though.

17 DR. FISCHER: But to borrow Greenspan's

18 term, I'm not worried about irrational exuberance

19 with this test. I mean, the fact is, most people

20 don't believe it, and I think that it's a much

21 smaller job for the champions of this technology

22 to convince us than it is to convince all the

23 oncologists of the world. And I think that, you

24 know, that will happen. The evidence that we saw

25 from '99 that has been published is a lot better

00415

1 than was published in 97, and it looks like we

2 are on to some good things here.

3 This is very different than

4 orthopedists and MRIs, you know, where you have

5 an orthopedist and an MRI, you've got somebody

6 getting a test. This is not something that's

7 going to overwhelm the oncology community

8 overnight because most people don't believe it.

9 And I think the champions are going to be

10 expected to publish data that will be disease and

11 drug specific, and to the extent that evidence is

12 persuasive, people will change their behaviors.

13 So I don't want Medicare doing any studies, to be

14 quite honest with you, and I wouldn't trust them

15 to do it. I think that the scientific community

16 is going to answer these questions over time.

17 The question is whether Medicare should pay for