April 17, 2006
public meeting of Southern
Medicare contractor (National Heritage Insurance Company/NHIC)
reconsideration of Medicare coverage for culture drug resistance testing
report dated April 13, 2006 (immediately below) describes the ongoing
review process for reconsideration of Medicare coverage for cell
culture drug resistance testing (CCDRT). The first of two scheduled
public commentary sessions was held today at the downtown Los Angeles
headquarters of NHIC. The second will be held April 19, 2006, at
the Hilton Hotel at Los Angeles International Airport (LAX). I'll
be reporting the proceedings of both meetings following the completion
of the second meeting.
In preparation for the meeting, I did a last minute review to determine
if there were recent developments of relevance. I came across an April 4, 2006
technology review by BlueCross/BlueShield for the Northwestern
USA. As in the case of most Blue Cross/Blue Shield reviews,
as well as the ongoing NHIC/Medicare review, this April 4 review
relied heavily upon the series of reviews by the National
BlueCross/BlueShield TEC review panel, headquartered in Chicago, IL. My
general critiques of the entirely inadequate and misleading national
BC/BS reviews are contained in my April 13, 2006 response letter (PDF file) to
NHIC/Medicare and in my April, 2002 Superior Court testimony in support of a
successful lawsuit against Blue Cross of California.
So this most recent BC/BS review simply reiterates the same old
inadequate and misleading arguments, and only adds the following new
"An April 2006 updated search of the literature returned two new
studies. Parker and colleagues studied the use of an extreme drug
resistance assay in 18 patients with recurrent malignant glioma who
were scheduled to receive irinotecan chemotherapy. Of the 15 with
evaluable assay results, 4 were classified as extreme drug resistant.
The median time to progression was 6 weeks compared to 3 months for
those with indeterminate or low drug resistance. However, similar to
other studies, results of drug resistance were not used to direct
therapy, and there was no control group, both limiting interpretation
of this small case series."
COMMENT: Once again, the reviewers fail to realize that they are
evaluating a laboratory test and not a treatment. All laboratory tests
are evaluated on the basis of test accuracy and no laboratory test in
the history of cancer medicine has ever been shown in clinical trials
to improve the results of therapy, and this includes estrogen receptor,
progesterone receptor, Her2/neu, bacterial culture and sensitivity
testing, panels of immunohistochemical stains to subclassify tumors,
and on and on. So this latter small study was simply one more in
a long series of studies addressing the accuracy of the CCDRT with
respect to predicting for treatment outcomes. It was clearly a
The BC/BS review goes on to say:
"Staib and colleagues studied the use of the DISC chemosensitivity
index Ci assay to select effective chemotherapy for 226 patients with
previously untreated acute myeloid leukemia. (21). Of the 162
evaluable patients, there were 11 (69%) complete response rates, 34
(21%) blast reduction, no response in 17 (10%), relapse after complete
response in 53 (48%), and deaths in 89 (55%). When compared to a
non-randomized comparison group, none of the documented outcomes were
statistically significantly different. Neither study address the
limitations noted in the above discussion. Specifically, no studies
were identified that provided direct evidence comparing outcomes for
patients treated either by assay-guided therapy or contemporaneous
empiric therapy. Therefore, the policy statement is unchanged."
COMMENT: This "expert review" is once again a sad reflection on the
incompetence of these closed review panels. In point of fact, the
reviewers completely misunderstood the study. This study was NOT
a comparison of a "comparison group" and an "assay-directed treatment
group!" The DISC assay was NOT used to "select effective
This is how the study worked:
The authors first studied 76 patients between 1994 and 1997. They used
the assay results in these patients to establish retrospective cut-off
definitions between sensitivity and resistance for the various drugs
used in treatment. Then they prospectively applied these
definitions to a subsequent group of 86 patients treated between 1997
and 2000. No patients in either cohort received assay directed
therapy. All were treated with standard cooperative group acute
leukemia protocols. The purpose of the study was to see how well the
assay predicted for treatment outcomes. The "expert"
BlueCross/BlueShield reviewers were apparently confused by the authors'
Table II, the point of which was simply to show that the 1994-1997
patient cohort was similar to the 1997-2000 patient cohort. The
table showed that there were no significant differences between the
"training set" and "test set" groups (which is a strength and not
weakness of the study), yet the inept BlueCross/BlueShield reviewers
thought that the table represented a comparison between outcomes of
control and assay treated groups! This is an embarrassing mistake
representative of many such mistakes in the present and prior reviews,
all of which could have been avoided had the review been an open
process (inviting input and commentary from all interested parties,
including true experts in the field) rather than an entirely closed and
opaque process, as all of these so called "expert technology reviews"
The results as described by the study authors were as follows (Staib,P.
et al. Br
J Haematol 128:783-781, 2005):
"We found an overall predictive accuracy of 98.2% concerning treatment
response, which compares favorably with previously published data
ranging from 75% to 92%. Moreover the [assay results] proved to be the
strongest prognostic factor for overall survival in a multivariate Cox
regression analysis... (P<0.001)...and enabled the evaluation of
response to combination therapies and selection of possible treatment
alternatives. Our data suggest that [the assay] could serve as a
powerful tool for assay-directed therapy strategies in [acute
These authors identified a population of patients, who, if treated with
standard cooperative group protocols, had a 100% mortality within 12
months. These results are precisely analogous to those of Bosanquet,
et al, who identified a population of patients who, if treated with
the most popular form of "standard" therapy, had a median survival of 8
months and were all dead within 18 months, but, if treated with
something other than the most popular form of standard therapy,
survived substantially longer.
Now, these are precisely the sorts of studies which are used to
validate other types of laboratory tests used as an aid in the
selection of patient treatment. I'd like to ask any readers of
this to ask themselves if they (the readers) had acute leukemia or
chronic lymphocytic leukemia (two diseases where the National Cancer
Institute currently identifies a number of different drug regimens as
being equally likely to be active if chosen by coin flip), would they
not want to avail themselves of a simple laboratory test (which could
be performed on a simple blood or bone marrow specimen) which could
help them avoid receiving a treatment which was likely to be a death
sentence (when a large number of otherwise equal by coin flip regimens
could be chosen instead)?
But BlueCross/BlueShield and ASCO think that studies such as this are
irrelevant and ASCO specifically EXCLUDED all such studies from its
negative technology assessment review published in the September, 2004
edition of the Journal of Clinical Oncology (discussed in my April 13,
2004 letter to NHIC/Medicare [PDF file]).
Despite the fact that, if they applied the same criteria to evaluate
all laboratory tests as they apply to review cell culture drug
resistance testing, there would be absolutely zero laboratory or
radiographic tests available to assist physicians in the management of
More later, as the NHIC/Medicare re-review process continues.
April 13, 2006 Southern
Medicare contractor (National Heritage Insurance Company) reconsidering
previously-approved coverage for cell culture drug resistance testing
has no national coverage decisions regarding the types of cell culture
drug resistance testing which have been available in the USA since the
early 1990s. It has an earlier (circa early 1980s) coverage
decision denying payment for two specific forms of cell culture testing
which were available at the time. The first of these was the "human
tumor stem cell" assay and the second was the "fluorescent cytoprint
assay." The first was based on measuring tumor cell "clones," while the
second was based on measuring "tumor microorgans." Neither of
these tests has been available for general clinical application in more
than a decade (more than two decades for the latter). The new
Medicare contractor has confused the distinction between the
specifically non-covered technologies (the above, long-abandoned
technologies) and the current technologies, which have been available
for the past 15 years, as will be seen and discussed below and in my
response letter (link provided below).
In November, 1999, Medicare convened a national coverage advisory
meeting, where the issue was intensely reviewed, as described in detail
in my letter of April 13, 2006 (PDF file).
the conclusions of
this national coverage advisory committee review were generally
favorable, no national coverage decision was
made, and coverage decisions were left to the local/regional Medicare
contractors. Subsequently, TransAmerica, the previous contractor
for the Southern
California region (where most of the laboratories providing this
service are located), officially granted partial coverage for cell
culture drug resistance testing (also described in detail in my letter,
same PDF file as above).
As a direct result of two unfavorable reviews published in the
September, 2004 issue of the Journal of Clinical Oncology (discussed
also in the April 13, 2006 letter, same
the new Southern California Medicare contractor (National Heritage
has proposed to rescind
previously-approved (by TransAmerica) coverage for this testing, as
described on the National
Heritage website and
reproduced directly below. As described in my April 13 letter (same PDF
above), this proposed rescinding of
coverage cannot be justified by any rational consideration of the
issues and data. This proposed change will be discussed at open
meetings in Los Angeles, to be held April 17 and 19, as also described
on the National Heritage website (on their
website, above, click Southern California/April agenda).
The following is currently posted on the National Heritage website,
relating to the
proposed rescinding of coverage:
"DL22566 Oncologic in Vitro Chemoresponse Assays -
"06-02.5 Primary Geographic
Jurisdiction California - Southern, Oversight Region Region IX
"Medicare has a National Coverage
Decision (NCD) on human tumor stem cell sensitivity assays which states:
"Human tumor drug sensitivity
assays are considered experimental, and therefore, not covered under
Medicare at this time.
"The clinical application of the
assay, based on testing in tumor microorgans rather than in clones
derived from single cells, is considered experimental, and therefore,
not covered under Medicare at this time.
"A provider was granted a CMS
review for reversal of this noncoverage on 8/4/1999, but the request
was withdrawn during CMS’s review of the data (4/13/2000). CMS declined
to issue any form of coverage for chemoresponse testing in 2000. [nb. This
characterization is entirely incorrect,
as described in my April 13, 2006 response
letter PDF file
"In 2004, reports were published
the American Society of Clinical Oncology (ASCO) and the Blue Cross
“TEC” technology evaluation committee. The ASCO report concluded flatly
'The use of chemotherapy
sensitivity and resistance assays to select chemotherapeutic agents for
individual patients is not recommended outside of the clinical trial
setting. Oncologists should make chemotherapy treatment recommendations
on the basis of published reports of clinical trials and a patient’s
health status and treatment preferences.'
"NHIC was unable to identity major
national published protocols (e.g. www.nccn.org, for ovarian cancer)
which utilize in vitro chemoresponse testing as a necessary step in
"Indications and Limitations of
"CMS gives several specific
instructions to contractors regarding criteria for coverage evaluations.
"1. Criteria are similar for LCD
and NCD determinations; both follow the same language at Section
1862(a) of the Social Security Act (care must be “reasonable and
necessary; see Federal Register, 5/16/2000).
"2. Positions of national
associations and/or the Blue Cross TEC committee are explicitly called
out as appropriate criteria for medical necessity decisions (HCFA
Ruling 95-1, Section 5).
"3. Contractors are specifically
guided to weight consensus medical statements and association
statements over individual providers (Program Integrity Manual, Chapter
13, Section 13.7.1).
"Accordingly, NHIC does not cover
oncologic in vitro chemosensitivity tests per Medicare’s NCD (190.17).
The reciprocal tests (chemoresistance or extreme resistance tests), are
not covered per this LCD.
"Important note: Jurisdiction
with Part A may supercede this LCD where current specimen dating rules
place the date of service within a Part A hospital stay. See Claims
Processing Manual, Chapter 16, Section 120.1 (minor updates by CR 4156,
4/3/2006) and also Section 40.3 and guidance on Part A, Part B billing
and bundling issues."
Once again, my responses and
rejoinders to the above statements and proposed rescinding of coverage
are presented in my April 13 letter sent to the Southern California
Medicare contractor (click here
for PDF file of this letter).
I shall also be attending both meetings, where I will offer oral
commentary, and I shall report the results of these meetings
later next week on this website.