00331
1
2
3
4 VOLUME III
5 (Morning Session - November 16, 1999)
6
7
8
9
10 HUMAN TUMOR ASSAY SYSTEMS
11
12 HEALTH CARE FINANCING ADMINISTRATION
13 Medicare Coverage Advisory Committee
14 Laboratory & Diagnostic Services Panel
15
16
17
18
19
20 November 15 and 16, 1999
21
22 Sheraton Inner Harbor Hotel
23 Baltimore, Maryland
24
25
00332
1 Panelists
2 Chairperson
John H. Ferguson, M.D.
3
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.
10
Temporary Voting Member
11 Kathy Helzlsouer, M.D.
12 Consumer Representative
Kathryn A. Snow, M.H.A.
13
Industry Representative
14 James (Rod) Barnes, M.B.A.
15 Carrier Medical Director
Bryan Loy, M.D., M.B.A.
16
Director of Coverage, HCFA
17 Grant Bagley, M.D.
18 Executive Secretary
Katherine Tillman, R.N., M.S.
19
20
21
22
23
24
25
00333
1 TABLE OF CONTENTS
Page
2 Welcome and Conflict of Interest Statement
Katherine Tillman, R.N., M.A. 5
3
Opening Remarks & Overview
4 Grant Bagley, M.D. 10
5 Chairman's Remarks
John H. Ferguson, M.D. 28
6
Brian E. Harvey, M.D., Ph.D. 30
7
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
15
LUNCH 196
16
VOLUME II
17
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
20
Open Committee Discussion 304
21
Day One Adjournment 330
22
23
24
25
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
6
Adjournment 487
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
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