Last updated: 5/25/2003

Cell Culture Drug Resistance Testing (CCDRT)

by Larry M. Weisenthal, MD, PhD

Introduction

Cell culture drug resistance testing (CCDRT) refers to testing a patient's own cancer cells in the laboratory to drugs that may be used to treat the patient's cancer. The idea is to identify which drugs are more likely to work and which drugs are less likely to work. By avoiding the latter and choosing from among the former, the patient's probability of benefiting from the chemotherapy may be improved.

CCDRT is currently a matter of some controversy. For example, the routine use of CCDRT in the treatment of cancer has been officially endorsed by the Southern California Medical Oncology Association, even while it has been officially opposed by the Northern California Medical Oncology Association.

I am a medical oncologist who has spent 16 years in the full time study and application of CCDRT technologies. I currently have a full-time laboratory-based medical practice in Southern California exclusively devoted to the application of CCDRT technologies as part of the routine treatment of cancer patients. Thus, I may be definitely considered to have a bias in favor of these technologies. My brief curriculum vitae is listed at the conclusion of this report.

The Pros and Cons, Whys and Wherefores of CCDRT:
Short Version

Published work in CCDRT dates back to the early 1950s, but little attention was paid until the publication of an influential paper in the New England Journal of Medicine in 1978, describing a particular CCDRT technology. This paper stimulated tremendous interest. Everyone from the NCI on down focused 100% of their attention on this single technology, rather than carrying out a comprehensive, head-to-head examination of the whole range of available and proposed technologies. When the single technology described in the NEJM paper fell short of expectations, everyone became disillusioned and the whole field became discredited, notwithstanding the fact that a whole range of CCDRT technologies had never been closely examined.

When work in NCI-funded programs essentially ceased to exist, small, mostly private sector laboratories, began to introduce newer CCDRT technologies into routine patient care. This was based on consistent data showing that the technologies were reproducibly capable of identifying drugs with significantly higher and lower than otherwise expected probabilities of being clinically effective in individual patients. Additionally, the private laboratories proved that it was possible to get very high (>90%) evaluability rates from specimens routinely submitted in a real world setting.

The NCI, NCI-designated comprehensive cancer centers (NCI-CCCs), and American universities have completely ignored the availability of these nationally licensed laboratories and the CCDRT technologies used by these laboratories. Instead of making an effort to determine if cancer treatment may be improved through the use of these CCDRT technologies, the NCI, NCI-CCCs, and American universities have continued to apply the (in my opinion failed) paradigm of the last 25 years. This paradigm seeks to identify the single best treatment to administer to the average patient with a given form of cancer through the use of prospective, randomized trials. I refer to this as the lowest common denominator theory of cancer chemotherapy. Whatever the theoretical virtues of this paradigm, it has been a failure, as established by the last 25 (largely-non-productive) years.

In my experience, I have never found a single NCI, NCI-CCC, or American University-based oncologist or researcher who has either the facts or command of concepts to discuss this issue of CCDRTs intelligently. Perhaps someone from one or more of these institutions will read this and be moved to join me in an on-line discussion of the issues raised herein.

CCDRT is of value in any situation in which there is a choice between two or more treatments. This includes virtually all situations in cancer chemotherapy, whether the goal is cure (as in acute leukemia or breast cancer) or palliation (as in far-advanced colon cancer or non-small cell lung cancer). There are a number of qualified laboratories that are nationally licensed to provide this service to patients and physicians. The cost of CCDRTs is often, but not always, covered under standard third party plans. The costs range from nothing (in the case of laboratories that provide abbreviated testing in order to obtain fresh cancer tissue for unrelated research) to $1,500 (in the case of laboratories that provide very comprehensive testing primarily for the benefit of the patient).

The main argument against CCDRT is that it has never been shown, in prospective randomized trials, that there is a clear advantage to chemotherapy selected with the benefit of knowledge of the CCDRT results compared to "physician's choice" chemotherapy selected without knowledge of the CCDRT results. This is not because studies were carried out and failed to show a difference. It is because the studies have never been supported and carried out by the NCI, NCI-CCCs, and university cancer centers. If and when such studies are carried out in the future, it will be important that they be carried out utilizing the private sector laboratories that have by far the greatest experience with the range of laboratory technologies in use today.

The main arguments in favor of CCDRT are that scores of studies have consistently shown a significantly greater benefit for treatment with drugs that are active in the tests compared to treatment with drugs that are inactive in the tests. Additionally, a growing number of studies have shown a superior survival for patients treated with drugs that are active in the tests. It should also be noted that there has never been a laboratory or radiographic test in the history of medicine that has been shown to confer a treatment benefit in prospective trials in which patients were randomized to treatment with and without benefit of the tests. Thus, non-recognition of CCDRTs until such a study has been completed constitutes an illogical, unreasonable, and unprecedented barrier to the more rational management of such a serious disease as cancer.

It comes down to which standard one chooses to apply:
1. Proof beyond reasonable doubt (argues against the use of CCDRT)
2. Weight of currently available evidence (argues in favor of the use of CCDRT)

The Pros and Cons, Whys and Wherefores of CCDRT:
Long Version

INTRODUCTION

In the course of a CompuServe Cancer Forum thread concerning high dose chemotherapy and marrow transplantation in breast cancer, the question of using "culture and sensitivity" tests to select the best chemotherapy was raised. A patient described her own situation and said, "I did mention chemo-sensitivity testing to my doctor this week, but he said the data wasn't there to show that it was reliable enough. He said that the cells which grow best in vitro are not necessarily characteristic of the tumor as a whole. He also dropped a name that I didn't recognize and said that this doctor had been exploring this technique 15 years ago and stopped doing it because it wasn't a very reliable indicator of which chemotherapy would work."

In response, I wrote the following:
"In bits and pieces, I have addressed aspects of the issue since I first logged onto the Cancer Forum in late February. I would like now to reproduce excerpts as a more unified whole. In a small number of message-length "chapters," I will present (1) a brief history of the current cancer chemotherapy paradigm as it has developed under the leadership of the National Cancer Institute, (2) a brief history of cancer "culture and sensitivity" testing, (3) a brief description of what the tests have to offer in cancer treatment, with a brief bibliography, (4) list of American, European, and Asian laboratories through which these tests are available, (5) current status regarding costs and third-party (insurance and HMO) payment for these services, (6) concluding remarks, (7) follow-up questions recently asked of me by patients and doctors, and (8) a brief curriculum vitae, for those who wish to know my background.

I will be very happy to answer specific questions or address specific concerns or criticisms from anyone.
e-mail: 72203,2235@compuserve.com
voice: 714-894-0011
fax: 714-893-3658

CHAPTER 1

Brief history of the current cancer chemotherapy paradigm,
as it has developed under the leadership of the NCI:

One of the greatest ironies in cancer treatment research was the success story in the treatment of childhood leukemia. Forty-five years ago, this was a universally fatal disease, with survivals measured in months. Over the ensuing forty years, this disease has become progressively more curable, owing to a series of prospective clinical trials, in which patients were randomized between the best current therapy and a putatively-improved, but empirical form of therapy. Gradually, therapies got better and better and today most (but by no means all, including my sister's late 7-year-old daughter Kristina) children with this disease are cured.

What happened next is that the NCI and American university investigators with NCI funding attempted to apply the childhood leukemia paradigm to all forms of cancer, with stunningly bad results. We have made virtually no measurable progress at all in treating advanced cancer over the past 20 years.

In my opinion one of the major reasons for this lack of progress is the lack of systems to model the behavior of real, clinical, human cancer. Most preclinical research has been carried out in animal tumors and in immortal cell lines (see Gerald Dermer's book: The Immortal Cell: Why Cancer Research Fails). Most clinical research is in the form of prospective, empirical, randomized trials designed to identify the best treatment to give to the average patient, in a disease notorious for its heterogeneity, where no one is average.

Of all the leading causes of death, cancer has arguably the most inadequate systems to model the behavior of human disease. The scandal is not so much that we do not have valid models with the same invaluable utility as bacterial "culture and sensitivity" tests, but rather that the NCI and American universities have made virtually no effort to develop such tests. The most obvious and promising models are those based on the study of drug effects on real human cancer tissue, freshly removed from the patient. Not only have the NCI and universities not made serious efforts to develop "culture and sensitivity" assays for cancer, but they have (both intentionally and inadvertently) done virtually everything possible to discourage research in this area (for reasons why, see chapter 2). As a result, the most important progress in this field during the past ten years has taken place in the private sector and, increasingly, in Europe and Japan.

There is an overwhelming amount of published, peer-reviewed literature to establish, beyond the shadow of a doubt, that existing assays are capable of identifying both poor prognosis therapies and good prognosis therapies, with good prognosis therapies being about seven-fold more likely to work than poor prognosis therapies (see chapters 3 and 4).

The literature is, however, complex, and cannot be intelligently interpreted without an understanding of Bayes' Theorem (a statistical concept) and a host of biological and pharmacological concepts. Good review papers exist and are increasingly appreciated, understood, and applied by private sector and European clinicians and scientists. This literature is not understood by NCI investigators and by NCI-funded university investigators, who have been satisfied to pursue their own failed paradigms. A cynic might claim that private sector clinicians and scientists like me are out to make money from promoting this testing; in my situation, this shoe doesn't fit, but , even if it did, an honest cynic would admit that the existing, failed paradigms support thousands of professional careers and hundreds of universities and cancer centers. In a big industry like cancer, there is no such thing as a truly unbiased opinion. Caveat emptier.

CHAPTER 2

Why ostensibly brilliant people, such as the typical cancer patient's own doctor, know so little about these tests: A brief history of cancer "culture and sensitivity" testing.

Cancer chemotherapy came into its own in the early 1950s. From the very beginning, scientists (Black and Speer were the real pioneers) tried to develop tests to find the best drugs for individual patients. The tests really did seem to work, but they weren't perfect, and there weren't all that many drugs to choose from, anyway. No one got real interested, except for the Japanese, who dabbled with variations of the original Black and Speer test over the next 3 decades. From time to time, a different technology was introduced. One of the great overlooked achievements in cancer research was the brilliant work of a scientist named R. Schrek in the 1960s. In his published work of 25-30 years ago, there are obvious clues to a practicable testing method which should have been developed, long ago, to improve clinical research and drug selection in clinical patients. Dr. Schrek worked at the Hines VA hospital in Chicago, right up to his recent death at the age of 87 just 2 years ago. I will always be sad that Dr. Schrek never got any real recognition at all for his work. The exciting things happening right now with regard to drug testing in leukemia and lymphoma could have happened 25 years ago, had anyone paid attention to Dr. Schrek. It would have quickly spilled over also into solid tumors, and we would by now have been way, way ahead.

In the late 1970s, a University of Arizona scientist named Syd Salmon, along with Ann Hamburger, developed a different type of test, the "Human Tumor Colony Assay" (HTCA). These authors had the great fortune of having their paper published in the New England Journal of Medicine. When the NEJM talks, people listen. One of the early listeners was Dan Von Hoff (now of the U of Texas San Antonio) who was one of my fellow clinical oncology trainees at the NCI. Dan and Syd are two of the most articulate and persuasive people around. With the NEJM as a launching vehicle and their considerable energies and talents, before too long everyone was convinced that the HTCA was going to be one of the great breakthroughs in cancer treatment. Almost everyone bought on, including the NCI.

My consistent criticism of Syd and Dan, which goes back 15 years now, is that they did not promote the general concept of "culture and sensitivity testing" as something worthy of intense investigation, but rather promoted the HTCA as a specific method and the only method worthy of attention. My criticism of the cancer research leadership is that it bought and swallowed the Syd and Dan sales pitch, despite loud protestations that the concept was much to important to be shackled to one specific technology. Rather than carrying out comparative studies of different types of assay technologies, based on different biological endpoints, the NCI and NCI-funded crowd put all of their eggs in a single basket and did, indeed, shackled the whole concept of culture and sensitivity testing to a single technology. What happened is that the HTCA technology sank like a stone and dragged the whole concept down to the bottom along with it. This occurred most precipitously as a result of a critical NEJM editorial, published in 1983 (the NEJM giveth and the NEJM taketh away). Everyone who had bought and swallowed the HTCA sales pitch felt that they had been suckered. Once burned, twice shy.

There were, however, those of us who persisted. I had personally always taken the "big tent" approach, meaning that we should learn from the clues of all who have worked in this area and should strive to develop and apply whatever systems it takes to get the answers we want, rather than claiming that a particular system is the most theoretically valid approach and should therefore be used exclusively. My own concepts (which were built on the work of others, whom I will always acknowledge) have been well documented in the literature of the last 15 years. Heretical in their time, these concepts have now been confirmed to be largely correct, beyond the shadow of a doubt. However, as of the mid-1980s, work in this field had ground almost to a halt in this country. The NCI leadership was doing nothing to encourage work in this field, and peer-review panels were shooting down virtually every proposal to study culture and sensitivity tests on fresh tumors.

What has happened since then is a tribute to the genius of the American private enterprise system (which I have come to appreciate very much, although I remain a liberal Democrat). Small companies sprouted up, supported by venture capital, private and institutional investors, the sweat equity of their founders, occasional Small Business Innovative Research Grants, and sometimes by personal IRAs, third mortgages, and children's college funds. These small companies are successfully achieving what the NCI and American universities could not and would not do -- the introduction of these tests into the mainstream practice of American cancer medicine (More about this in subsequent chapters.) But it's like something that has sneaked into the house and bitten the oncology opinion leaders from their behind-sides. They weren't expecting it; they didn't see it coming, and they still don't know a lot about it. But they are beginning to learn, as will one's own doctor, if he/she is willing to have an open mind and give all of this a fair chance.

CHAPTER 3

What the tests have to offer in the treatment of today's cancer patient.

The following is an excerpt of an earlier communication with a Cancer Forum participant. His question concerned brain tumors, but the topic and response were germane to the use of these assays in most cancer, including breast cancer and leukemia:

Dear Al: You asked (1) what types of cancers were being managed with the information provided by drug resistance assays, (2) whether there were data to document a patient survival benefit when these assays are used, and (3) whether the assays are useful in glioblastoma (a brain tumor).

In Southern California (where several laboratories are located), these assays are being used by a substantial percentage (probably a majority) of medical centers in the management of a wide variety of cancers. The assays are potentially of value in any situation where there is a legitimate choice between more than one type of chemotherapy or a legitimate choice between chemotherapy and an alternative form of treatment. If one reviews the NCI's "PDQ," describing "state of the art" therapy, one is struck by the fact that there is virtually never a situation in which there is one regimen which is clearly superior to all others. Characteristically, there are between several to one dozen drug regimens, which differ in important ways, but which would produce superimposable results if given to a population of unselected patients. Standard chemotherapy has evolved from the misguided notion that it is a good idea to perform trial and error randomized clinical trials to identify a single regimen which will produce optimum results when given to the average patient with a given form of cancer (the lowest common denominator theory of cancer chemotherapy). This approach is contrary to modern understanding of cancer biology and has not fostered either efficient or humane advancement in knowledge and improvement in cancer treatment. Cancer is characterized by heterogeneity in all aspects of its biology. It grows faster or slower and it is sensitive or resistant to different drugs in different patients. Hence the notion of testing a biopsy specimen of the cancer to determine which drugs are more likely to work and which are less likely to work. In my opinion, this is of benefit whether one has a realistic chance of cure or whether one has only a realistic chance of palliation. Given two potentially curative or palliative drug regimens, the better regimen will be the regimen which contains active ("good") drugs and which does not contain inactive ("bad") drugs.

Existing drug resistance assays have been extensively documented to be capable of placing drugs into the following general categories:

1. Drugs which have virtually no chance of working.
2. Drugs which have a lower than expected chance of working.
3. Drugs which have an average chance of working (in line with clinical expectations before the assays were performed).
4. Drugs which have a higher than expected chance of working.

All else being equal, there is about:
A 20:1 advantage in choosing category 4 drugs over category 1 drugs;
A 10:1 advantage in choosing category 3 drugs over category 1 drugs;
A 2:1 advantage in choosing category 2 drugs over category 1 drugs;
A 2:1 advantage in choosing category 4 drugs over category 3 drugs;
A 7:1 advantage in choosing category 4 drugs over category 2 drugs;
A 3 to 4:1 advantage in choosing category 3 drugs over category 2 drugs.

What are the data which support the above statements? More than 50 different peer-reviewed studies from multiple institutions reporting clinical correlations in more than 4,000 patients. There have been a number of reviews which summarize and discuss these studies (listed below).

There has never, to my knowledge, been a randomized trial "proving" therapeutic benefit for ANY laboratory or radiographic test, including drug resistance assays. There are many reasons for this. For a very succinct and to the point explanation of some of the barriers to the completion of such a study, look up the correspondence published in the New England Journal of Medicine by Gunter Umbach, et al, as well as that of Plasse (Vol. 308, No. 24, p. 147, 1983). I have written fairly extensively on this issue and I organized two national cooperative group studies in the mid- and late-80s, which were closed because of poor accrual of patients and other problems not related to the assays themselves. For too long, the use of these assays was held hostage to a hypothetical series of randomized trials which no one was willing to support. At long last, these assays are being judged by the same criteria traditionally applied to other laboratory and radiographic tests, namely the accuracy of the tests and the value of the information provided by the tests, as perceived by the physicians who order the tests. Now that this is finally happening, these tests will become more broadly available; technologies will evolve and improve at an accelerating rate, and the NCI and universities will be prodded into supporting innovative clinical trials based around the use of these assays. In my personal opinion, this will be a huge advance, and it is an advance for which the private sector will deserve most of the credit.

With regard to Glioblastoma Multiforme, according to the most recent PDQ, the following drugs are being tested in clinical trials:
carmustine, procarbazine, vincristine, cisplatin, lomustine, taxol, thiotepa, etoposide, chlorodeoxyadenosine, cyclophosphamide, topotecan, nitrogen mustard, AZQ, carboplatin, 5FU, interferon alpha (as a chemotherapy modulator), and mafosfamide.

Most of these drugs (or acceptable surrogates) may be readily tested against glioblastoma multiforme, and these drugs may then be sorted into the four groups (terrible, bad, average, and above average) described above. However, the tests demand fresh viable tissue, carefully collected and handled and sent to a laboratory via courier or Federal Express. I would not recommend brain surgery just to get tissue for the assay. However, were surgery to be performed for a reasonable clinical indication, then, in my opinion, a portion of the tissue should definitely be submitted for cell culture drug resistance testing.

Now with regard to high dose chemotherapy of breast cancer:
I am philosophical agreement with my former colleague Robert Nagourney, who says that giving megadose therapy is like shouting at a person who doesn't understand English. You will go a lot farther by speaking his language than you will by hollering.

High dose breast cancer chemotherapy utilizes combinations constructed from 2 or more of the following drugs: ifosfamide, carboplatin, etoposide, melphalan, thiotepa, mitoxantrone, taxol, and so on.

Side effects of these drugs range from the merely noxious to the fatal. Why on earth should anyone receive a noxious, ineffective drug when there were more promising alternatives?

You can sometimes make your point better by shouting, but be sure first that you are speaking the right language!

Bibliography

General reviews:

1. Fruehauf ,J.P. and Bosanquet, A.G. In vitro determination of drug response: a discussion of clinical applications. Principles & Practice of Oncology Updates, Volume 7, No 12, December, 1993, J.B. Lippincott Company, Philadelphia, PA (phone 215-238-4398).
2. Weisenthal, L.M. Clinical correlations for cell culture drug resistance assays based on the concept of total tumor cell kill. in Koechli, et al (eds) Chemosensitivity Testing in Gynecologic Malignancies and Breast Cancer, Contrib Gynecol Obstet, Basel, Karger, 1994, v. 19, pp 82-90.
3. Bosanquet, A.G. In vitro drug sensitivity testing for the individual patient: an ideal adjunct to current methods of treatment choice. Clinical Oncology (Published by the Royal College of Radiologists) 5:195-197, 1993.
4. Weisenthal, L.M. Cell culture drug resistance assays in hematologic neoplasms based on the concept of total tumor cell kill. in Kaspers, GJL, et al (eds). Drug Resistance in Leukemia and Lymphoma, Harwood Academic Publishers, Langhorne, PA, pp. 415-432.
5. Weisenthal, L.M. and Kern, D.H. Prediction of drug resistance in cancer chemotherapy: the Kern and DISC assays. Oncology (USA) 5:93-114, 1991.
6. Weisenthal, L.M. Fresh-tumor, cell-culture assays for breast cancer. in Dickson, R.B. and Lippman, M.E. (eds) Drug and Hormone Resistance in Breast Cancer, Horwood, New York, London, Toronto, and Tokyo. 1995, pp. 323-350.

Critical review of the alleged "gold standard" assay, the single basket into which the NCI and university researchers placed all their eggs, only to see them break, and which for a decade has jaundiced their view of cell culture drug resistance assays:


7. Weisenthal, L.M. and Lippman, M.E. Clonogenic and nonclonogenic in vitro chemosensitivity assays. Cancer Treat Rep. 69:615-32, 1985.

Comprehensive technical review:


8. Weisenthal, L.M. Predictive assays for drug and radiation resistance. In: Masters, JM (ed) Human Cancer in Primary Culture, Kluwer Academic Publishers, Dordrecht, The Netherlands, 1991, pp. 103-148.

Patient-oriented review:


9. Nagourney, R.A. A new paradigm for cancer chemotherapy. Cope (May/June, 1993 issue), pp. 44-46.

Editorial relating to flaws in the current methods for testing new anticancer drugs in humans:


10. Weisenthal, L.M. Antineoplastic drug screening belongs in the laboratory, not in the clinic. J. Natl. Cancer Inst. 84 466-469, 1992.

CHAPTER 4

Listing of "Reputable" Labs that I feel are likely to do more good than harm

USA:

Analytical Biosystems, Inc., Providence, Rhode Island. Ken Blackman, PhD. Solid Tumors Only. 1-800-262-6520

Anticancer, Inc., San Diego, CA. Robert Hoffman, PhD. Solid Tumors Only. 1-619-654-2555

Oncotech, Inc., Irvine, CA. John Fruehauf, MD. Solid Tumors and Hematologics. 1-714-474-9262 / FAX 1-714-474-8147

Sylvester Cancer Institute, Miami, FL. Bernd-Uwe Sevin, MD. Solid Tumors Only. (especially GYN). 1-305-547-6875

Human Tumor Cloning Laboratory, San Antonio, TX. Daniel D. Von Hoff, MD. Solid Tumors Only. 1-210-677-3827

Rational Therapeutics Institute, Long Beach, CA. Robert A. Nagourney, MD Solid Tumors and Hematologics. 562-989-6455 http://www.rational-t.com/

Weisenthal Cancer Group, Huntington Beach, CA. Larry M. Weisenthal, MD, PhD. Solid Tumors and Hematologics. 1-714-894-0011 / FAX 1-714-893-3659 / e-mail: 72203,2235@compuserve.com

EUROPE:

Bath Cancer Research Unit, Bath, England. Andrew G. Bosanquet, PhD. Hematologics and Solid Tumors. 44(Country Code)-225-824-124 / FAX 44(Country Code)-225-824-114

Haematology Research Unit, Pembury Hospital, Pembury, Kent, England. TN2 4QJ (UK) Jean Sargent, PhD. Well qualified in hematologics. Ask them about their solid tumor experience

Uppsala University, FMCA Laboratory, Uppsala, Sweden. Rolf Larrson, MD, PhD and Peter Nygren, MD, PhD. Hematologics and Solid Tumors. 46(Country Code)-18-663-000 FAX 46(Country Code)-18-502-916

Free University of Amsterdam, Department of Pediatrics, Amsterdam, The Netherlands. Prof. A. Veerman, MD, Rob Pieters, MD, PhD, Gert-Jan Kaspers, MD, PhD. Hematologics and Pediatric Solid Tumors Only. 31(Country Code)-20-548-2286 / FAX 31(Country Code)-20-548-2289

University of Zurich, Division of Gynecology, Frauenklinikstrasse 10 CH-8091, Zurich, Switzerland. O.R. Koechli, MD. Solid Tumors (especially GYN) only.

University of Heidelberg, Dept of Ob/GYN, Heidelberg, Germany. M. Kaufmann MD. Solid Tumors Only. (esp. GYN)

University of Koln (Cologne), Dept of Internal Medicine, Koln (Cologne), Germany. Robert Lathan MD. Ask Dr. Lathan about their experience.

JAPAN:

Keio University, Department of Surgery, 35 Shinanomachi, Shinjukuku, Tokyo, 160 Japan. Toshiharu Furukawa, MD. Solid Tumors Only.

IMPORTANT NOTE:
How May a Patient Arrange to Have His/Her Tumor or Leukemia Tested?
Both fluid and solid tumor specimens may be sent out via Federal Express or another overnight courier service for testing at one of more than a half-dozen labs around the country. Note that the choice of a lab is not a geographical consideration, but a technical consideration. All of the labs that I listed above are experienced and capable of providing very useful information. However, the labs vary considerably with regard to technologies, approach to testing, what they try to achieve with the testing, and cost. By investing a little time on the phone speaking with the lab directors, you should have enough knowledge to present the concept to the patient's own physician. At that point, the best thing is to ask the physician, as a courtesy to the patient, to speak on the phone with the director of the laboratory in which you are interested, so that everyone (patient, physician, and laboratory director) understand what is being considered, what is the rationale, and what are the data which support what is being considered.

CHAPTER 5

Costs of CCDRT and Insurance Issues

The cost of cell culture drug resistance testing is highly variable. Some labs are primarily interested in doing research on the tissues submitted for testing, perform relatively limited testing, and charge little or nothing. With other labs, there is a more comprehensive, patient-oriented focus, with considerably more complex testing, accompanied by specific treatment recommendations, and fees range up to $1,500. Some labs accept insurance reimbursement as payment in full, while others require co-payment or payment for denied claims. In our experience, insurance companies provide reimbursement for at least a portion of the charges in the vast majority of cases; ranging from full reimbursement (for the excellent private insurance companies such as Prudential, Principal Mutual, State Farm, Connecticut General, etc., depending, of course, on the individual policy) down to payment of only $187 or so. The average payment is the $700 - $1,200 range. The best thing, obviously, is for the patient to check with his or her own carrier prior to making the decision about whether CCDRT should be carried out. Some insurance companies dig in their heels and never pay, on the grounds that the testing is "investigational". This position cannot be supported medically or scientifically, and it is the strategy of the non-paying companies to dig in their heels, stone-wall, and provide no specific reasons for denials beyond vague references to the policy not paying for "investigational treatments" (ignoring the fact that these are laboratory tests which provide information and are not, in themselves, treatments per se).

It is getting harder and harder for third party carriers to avoid their contractural responsibilities to pay for reasonable and appropriate application of CCDRT in the management of their subscribers' cancers, as illustrated by the following earlier CompuServe Cancer Forum Communication:

To Cancer Forum SYSOP:
"On March 2, 1994, I attended the meeting of the Blue Shield of California Medical Policy Committee on Quality and Technology, held in San Francisco. There were 5 topics on the March 2 agenda. Two of these were of importance to oncology. The first topic was autologous bone marrow transplantation therapy of breast cancer. Proponents from major California transplant programs presented data favoring a policy of reimbursement. Much of the debate had to do with the strength of the uncontrolled trials, patient selection bias, and related issues. I felt that the most persuasive argument in favor of reimbursement was the statement that it was illogical to reimburse for marrow transplant therapy in lymphoma and testicular cancer (rare diseases which are eligible for reimbursement) and to deny reimbursement for breast cancer (a much more common disease with a far greater financial impact), when the type of supporting data was favorable for breast cancer, relative to the other, rarer diseases. However, Dr. Craig Henderson (Director of Clinical Oncology at UCSF) argued persuasively against reimbursement, pending availability of data from ongoing randomized trials, and the vote was unanimous (20 to 0) against reimbursement. A subcommittee was to be formed to monitor the flow of new data from the ongoing trials.

The next oncology topic was the reconsideration of cell culture assays to identify "bad" and "good" cancer chemotherapy drugs, prior to patient treatment. This topic had been previously considered at the October 13, 1993 meeting, but was tabled for 5 months after vigorous debate. The debate on March 2, 1994, was interesting, as the Northern California Oncology Association opposed reimbursement and stated its opinion that the assays are still investigational, while the Southern California Oncology Association (which has had a vastly greater experience with the clinical use of these assays over the past 6 years) came out in strong support of reimbursement. There was again vigorous debate, but the final decision was again unanimous -- this time a 20 to 0 vote in favor of reimbursement.

The official Blue Shield of California conclusion is stated (verbatim) as follows:

'Drug resistance testing in oncology is accurate and reliable. This information can affect clinical decision making and can lead to the avoidance of ineffective and potentially harmful chemotherapeutic agents. Although there are few prospective clinical trials comparing standard therapy with chemotherapy chosen by in vitro assay, there are sufficient published data to determine their safety, clinical utility, and impact on clinical decision making. RECOMMENDATION: It is recommended that the human tumor drug resistance assay is eligible for coverage when this information is required for the selection of chemotherapy.'

This Blue Shield technology review comes one year after Blue Cross of California also completed such a review and formally approved these assays for reimbursement."

Approximately 9 months later, I posted the following CompuServe Cancer Forum message:

News flash --
Following a detailed, formal technology review, Blue Cross of Colorado, New Mexico, and Nevada has officially determined that cell culture drug resistance testing is eligible for routine reimbursement when the tests are ordered by a physician for the purpose of providing information to assist in the selection of antineoplastic drugs in individual patients.
This follows by 11 months a similar decision by California Blue Shield and by 15 months a similar decision by California Blue Cross.
To my knowledge, all of the above agencies continue to deny reimbursement for high dose chemotherapy with bone marrow transfusion support in breast cancer.

With regard to HMOs, the policy varies. In Southern California, use of these tests has become widespread, and most HMOs will agree to pay for the testing after some negotiation, initiated at the request of the patient. In other parts of the country, it often depends on how committed the oncologist, surgeon, or pathologist is to the concept and value of the information in a specific case. For whatever it's worth, I'm convinced that the proper use of these tests will both help individual patients and save the health care system a pile of money, a true win-win situation. We could study this issue for a fraction of the cost of many more traditional clinical trials. I predict that we will, indeed, be able to study the cost-effectiveness of assay-directed therapy in the near future. Use of these tests is rapidly becoming too pervasive for insurance companies to stone-wall and for NCI-designated cancer centers and universities to ignore.

Medicare is a very complex area, indeed. The short answer is that, for inpatient surgical procedures, it is the hospital's responsibility to pay for the testing under the Medicare DRG allowance. Used properly, the tests could also save the hospitals money; but the initial reaction of most hospital administrators is to be skeptical and to agree to pay out of their DRG allowance only after persuasion by physicians and patients. Medicare outpatient assays must be paid for by the patient, institution, or laboratory, as Medicare does not provide payment for outpatient CCDRT.

We have occasionally received specimens from Canada. I am aware of a couple of cases where the provincial government agreed to pay for the testing, but this is obviously not a sure thing.

CHAPTER 6

Concluding Remarks

Well, that about covers the topic of cell culture drug resistance testing. It works. It's practical. It's available now, but is only widely used in areas where labs have made a concentrated effort to bring physicians up to speed. All the important progress has come outside of NCI-sponsored university-based research and it strikes at the heart of the standard NCI/university clinical research paradigms. That's why many doctors have not yet made the effort to learn about this. Because of the efforts of a dedicated private sector and the availability of media such as the CompuServe Cancer Forum and Internet, the NCI and NCI-oriented institutions will soon find themselves in the position of having to make the effort to learn and to be forced to provide sound reasons if they choose not to use these tests in the management of individual patients. It will change cancer treatment in this country more than anything else in the past 20 years, and it is quite likely to be one of the more visible and tangible early achievements of the "information highway".

CHAPTER 7

Follow-Up Questions Asked of Me on the CompuServe Cancer Forum

QUESTION A:

Asked by a sarcoma patient being treated at MD Anderson on a high dose ifosfamide (IFEX) protocol:

It was explained to me that "sensitivity testing" should really be named "resistance testing". That is, while a test of three drugs might be able to indicate to you which one was resisted the most by a particular tumor, it was not necessarily true that the drug with the least resistance would actually and affirmatively work. And, with a more or less "standard drug" like ifex, the results of resistance testing are not sufficiently predictive with a high enough of a degree of certainty that one would rule out using ifex on the basis of the results of a resistance test.

By all means go ahead with your decision to proceed with ifex therapy on an empiric basis. I agree that sometimes you just need to bite the bullet and do something. A watched cancer almost always grows. You get your opinions, and then you proceed. I think that you should proceed as planned, as you have received enough opinions to make up your mind.

On the other hand, there is so much misinformation about these tests, that I hate to see it perpetuated. Please, if you get a chance, show this message to your docs at MD Anderson.

1. Although it is true that the tests indicate resistance with a higher specificity than they indicate sensitivity, it is definitely a proven fact that drugs active in the tests have a significantly higher than otherwise expected probability of working in the patient.
2. Everything else being equal, an assay "positive" drug is, on average, seven-fold more likely to work than an assay negative drug.

O.K. Now let's take chemotherapy of sarcomas. What drugs might be used?

Just taking the clinical trials listed in the current PDQ, we have: ifosfamide, cyclophosphamide, melphalan, doxorubicin, taxol, carboplatin, etoposide, 5FU, and dacarbazine. However, there is a whole host of additional drugs which have been used or which it may be rational to consider, such as thiotepa, nitrogen mustard, cisplatin, dactinomycin, mitoxantrone, vinorelbine, vinblastine, mitomycin, bleomycin, and a variety of phase II drugs.

What the doctors at MD Anderson (and at the many other research institutions listed in the PDQ) are attempting to do is to define the "best" treatment to give to the AVERAGE patient with a given disease, otherwise known as the lowest common denominator theory of cancer chemotherapy. For 25 years we have been attempting to define the best drugs to give to the average patient, with no real progress with regard to identifying better chemotherapy regimens for the average patient. It is a scientifically bankrupt paradigm which has not fostered either efficient or humane progress in cancer chemotherapy. Ifosfamide is an old drug, not a new drug. Why doesn't everyone use it? Because not everyone thinks it is the best drug for all patients. Certainly, there are patients who fail to benefit from ifosfamide who are later shown to benefit from other treatments (at a time when the disease/clinical situation are more advanced).

If the tumor shows extreme resistance to ifosfamide but good sensitivity to carboplatin/etoposide/thiotepa/melphalan/doxorubicin, or whatever, why on earth should the patient be treated with ifosfamide? To fill out the numbers in someone's clinical study? To be a point on a survival curve in an investigator's publication? Or, in another setting (not at MD Anderson) to improve an institution's or practitioner's bottom line?

Please do show this to your doctors. I'd like to know how they respond. It would be terrific to get one or more of them on here to discuss these issues. (Note: Sadly, they never did come on-line).

QUESTION B:

Asked by a breast cancer patient offered treatment with high dose chemotherapy and ABMT at Johns Hopkins:

My doctors said that studies show that there is no difference in the result between using the results of drug sensitivity testing and an oncologist's informed opinion.

This is ABSOLUTELY INCORRECT, although you are getting close to the reason why it is possible not to use these tests and still not be guilty, prima facie, of medical malpractice.

I am only aware of a single study in which patients were randomized between assay-directed chemotherapy and physician's choice chemotherapy in which there was essentially no difference in outcome. This was a study published in the German literature about 15 years ago with a technology which was never used (and which never has been used) in the US. In contrast, there have been scores of studies of empiric chemotherapy regimens published during the same time which have shown no benefit for the new, empiric regimens. This doesn't stop anyone from continuing to try new regimens and even from continuing to use the regimens which failed to be proven superior in the trials.

In point of fact, it has never been shown, with ANY laboratory or radiographic test, that the use of the test improves treatment outcome compared with a physician's "informed opinion". According to the reasoning of your doctors, patients should never have any x-ray studies or laboratory tests performed, because they have never been proven to improve outcomes. This goes for estrogen receptors, DNA analysis, bacterial culture and sensitivity tests, CA-125 levels, CT scans, MRI scans, bone scans, and every other test you can think of also. It even goes for the practice of making a tissue diagnosis before definitively treating the cancer.

I will take this analogy one step farther. It hasn't been proven that most things that cancer chemotherapists do make any difference at all. For example, "second line" chemotherapy of virtually all types of tumors. For example, "first line" chemotherapy of many types of tumors. For example, high dose chemotherapy of breast cancer. For example, a recent review of 45 randomized trials in ovarian cancer involving 8,000 patients and 6,000 deaths concluded that "no conclusions could be made!" There was the suggestion (not proof) that platinum regimens were better than non-platinum regimens and the suggestion (not proof) that combinations were better than single agents (Advanced Ovarian Trialists Group: Chemotherapy in advanced ovarian cancer: An overview of randomized clinical trials, Br Med J 1991:303:884). The authors then had the audacity to state that they had, therefore, initiated a trial to compare one empiric regimen versus another in 2,000 patients. This is neither good science nor good medicine and is a terrific waste of human and monetary resources. However, an entire generation of academic oncologists has been trained in the paradigm of the empiric, randomized trial to select the best regimen for the average patient and an entire generation of practicing oncologists have been trained to use these empiric regimens in the management of their patients. Furthermore, an entire generation of oncologists has been trained to make extensive use of expensive, latest-generation radiographic tests and second-look surgeries to measure tumors before and during therapy to monitor and document "response" or "progression," despite the absence of even a shred of evidence that such radiographic and surgical documentation does anything at all to improve outcome beyond that which could be obtained by monitoring treatment results with history, physical, and simple blood tests.

There have been no large-scale randomized trials of assay-directed chemotherapy completed and only a small number even attempted. The very small number which have been attempted have definitely had positive results, with results trending in favor of the assay-directed chemotherapy arms. These were small, difficult studies and used a 15 year-old technology which is not being used today, because there are better technologies available. These studies are very difficult to do. For a very succinct and to the point explanation of some of the barriers to the completion of such a study, look up the correspondence published in the New England Journal of Medicine by Gunter Umbach, et al, as well as that of Plasse (Vol. 308, No. 24, p. 147, 1983). I have written fairly extensively on this issue (e.g. Weisenthal, L.M. Fresh-tumor, cell-culture assays for breast cancer. in Dickson, R.B. and Lippman, M.E. (eds) Drug and Hormone Resistance in Breast Cancer, Horwood, New York, London, Toronto, and Tokyo. 1995, pp. 323-350). I spent three years organizing two such studies in the mid-80s (one in multiple myeloma, involving 31 Veterans Administration Hospitals [VA CSP # 280]) and one in non-small cell lung cancer, involving Eastern Cooperative Oncology Group institutions [EST P-585)). Both of these studies were closed because of poor accrual and protocol violations totally unrelated to the assays themselves. There has just never been any support at all for funding or completing these studies. It is much easier for an academic institution to collect $2,000 - $7,000 per patient (on top of standard fees collected for providing medical services paid by third parties) by enrolling them on simple, empiric chemotherapy studies than to go to the trouble to participate in trials of assay-directed chemotherapy, which involve a lot more work, more expense, and fewer publications per year.

The traditional criteria on which all previous medical tests have been evaluated are (1) accuracy and (2) perceived benefit of the information by the physician who orders the test. These tests have been documented, beyond the shadow of doubt, to discriminate between active and inactive drugs with a 7 to 1 advantage (assay-positive drugs seven-fold more likely to work than assay-negative drugs). I could flip a coin and choose any two or three of the following drugs for a high dose regimen for breast cancer and achieve, overall, equivalent results: cyclophosphamide, ifosfamide, melphalan, thiotepa, nitrogen mustard, mitoxantrone, etoposide, carboplatin, cisplatin, carmustine, taxol. I could construct a standard dose regimen using any two or three of the following agents and achieve, overall, equivalent results: doxorubicin, mitoxantrone, cisplatin, fluorouracil, etoposide, vinorelbine, cyclophosphamide, ifosfamide, melphalan, nitrogen mustard, thiotepa, mitomycin c, vinblastine, taxol, and so forth. Why not choose from among the "7" drugs, rather than from among the "1" drugs?

What standard it the most appropriate to apply in this situation:

1. Proof beyond reasonable doubt?
2. Weight of currently-available evidence?

If you choose standard number 1, there will be precious few treatments available for cancer, beyond surgery and radiation therapy for local control. Please do report how your doctors view all of this. Better yet, encourage them to get online and discuss the reasons why they would defend their own approach to the management of breast cancer. (Sadly, the patient's doctors did not come on-line to discuss these issues).

QUESTION C:

Asked of the wife of a patient with a rare brain tumor.

We'd like to have the testing done. The neurosurgeon will have to go back in and biopsy the brain tumor all over again. Is this a good idea?

Let me answer your question in general.

I certainly do think that one of the areas in which these tests can be of help is in the situation of the rare tumor, in which little is known of the effects of chemotherapy in general and of many of the important drugs in particular. So, in general, I would always recommend serious consideration be given to cell culture drug resistance testing in these situations.

An always-important consideration, however, is the trade-off between the value of information the test can provide in a given situation and the morbidity (and sometimes expense) of the surgical procedure required to obtain tissue to study. We have a number of client physicians who perform major surgery for the sole purpose of obtaining tissue to study. These physicians (1) know their patients and the clinical situation of their patients, (2) understand the value and limitations of the tests, and (3) have relationships with their surgeons, so that the surgeons are willing to perform relatively major surgeries to obtain tissue in these situations.

I have a real problem when there is a patient in a remote location, whose physicians and surgeons have no experience with nor understanding of the tests. My nightmare scenario is talking local doctors into doing surgery that they really don't want to do; then having the patient suffer a serious complication during the surgical procedure; then having the assay fail for some reason to boot. Everyone will be very upset and someone may well get sued. Therefore, I would prefer to have decisions to proceed with such surgeries arise entirely from the minds of the physicians and surgeons caring for the patient, rather than from the patient of me talking them into doing something which they would really rather not.

On the other hand, if there is tissue which can be obtained through only a minor surgical procedure, under local anesthesia, then I would have greater enthusiasm for prodding the local physicians into action.

QUESTION D:

Asked by a medical oncologist.

Has anyone in the Intergroup considered doing drug resistance testing to key the most efficacious marrow ablation program in breast cancer treatment?

My answer:
No. But let's look at the issues.
What are the myeloablative drugs?
Carboplatin, etoposide, mitoxantrone, thiotepa, cyclophosphamide, ifosfamide, melphalan, carmustine, etc., to which we could also add other drugs such as taxol, vinorelbine, 5FU, doxorubicin, cisplatin, and on and on.

We test more breast cancers than any other single type of tumor. We get successful assays on 95% of all breast cancers that come in our door. We get high quality results from Tru-cut needle biopsies (as long as we can get a number of different "cores"). The testing can be as routine and simple (from the point of view of the submitting institution, if not for us) as obtaining an ER/PR, etc. panel. There are HUGE differences between the activity of the different drugs between different patients.

Recognizing the above, I wrote to both Karen Antman and Bill Peters several years ago, offering to share my insights and data relating to the application of these assays in the bone marrow transplant field. Neither so much as sent a postcard in return. Yet now we have Bill Peters reassuring patients that they don't need drug resistance testing (this happened on this forum within the past two weeks) because he "doesn't think it is useful."

He doesn't have a clue what we do; how we do it; the rationale; the data; the practicality; or anything else.

Knowing what I know, I cringe at the thought of just giving STAMP I, STAMP V, (or STAMP MCCXLVIII for that manner) to any poor patient who is betting her life on a roll-the-dice, potentially life-threatening, $100 grand procedure. But Antman (who is the current President of the American Society of Clinical Oncology, by the way) and Peters won't change their thinking until (1) someone in their own or the other's department (need to meet the competition) starts to do the work which we have been doing (and improving) full time for 16 years or (2) their patients demand to be told specific reasons why they should receive STAMP MCCXLVIII instead of individualized therapy chosen with the benefit of the information provided by these assays.

QUESTION E:

From a radiation oncologist, on why new treatments or procedures should not be introduced into mainstream medicine without first proving to be efficacious in prospective, randomized trials.

Without randomized studies comparing competing treatments, there can be no progress as patient selection is the nemesis of uncontrolled studies.

Bob, I really don't dispute this at all. This point was made abundantly clear in the Sinclair Lewis book -- Arrowsmith.

The biggest problem, though, with clinical cancer treatment research is this:
We are wasting our time doing randomized trials designed to test trivial hypotheses, e.g. CAF vs. CMF CAP vs CP Carboplatin vs. Cisplatin CP/Etop vs CP/VBL/MitC vs CAMP vs CP/VDS CP/CTX vs CP/Taxol ProMACE-MOPP vs m-BACOD vs CHOP vs MACOP-B and on, and on, and on.

What is worse than wasting our time is insisting on the above treatments being standard of care until someone proves that they have something better in a multiinstitutional randomized trial. None of the above is good enough to be considered standard of care, in my opinion. And insisting on this stifles discovery.

I don't think a single randomized trial in stage IV NSCLC should be performed until someone claims an 80% response rate and/or a median survival exceeding one year in non-randomized studies with decent historical controls. I could go in, disease by disease, but you get my drift.

I believe if we hadn't done a single randomized chemotherapy trial in adult cancer during the past 25 years that we would actually be ahead.

Sure, we might still be "misled" into thinking that "Big MACC" gives 50% response rates in stage IV NSCLC in the hands of MKSCC physicians, BUT SO WHAT?

There is entirely too much emphasis on "not letting the horse out of the barn" and protecting "standard" chemotherapy treatment from being usurped by "unproven" pretenders. As if the "standard" treatments were worth protecting from such usurpation. As if they were worth preserving at all, in many cases.

I tried for several years to organize randomized trials to "prove" the benefit of treatment based on knowledge of CCDRT results, as discussed previously, despite the fact that this has never been accomplished with any other type of laboratory test. It has been amply documented that neither the NCI nor anyone else associated with the NCI-sponsored cooperative group trials network has been willing to support such studies. I decided in 1987 not to permit the NCI-sponsored clinical trials system to hold these technologies hostage any longer. That is how we got to where we are today.

CHAPTER 8

My curriculum vitae, in brief:

Biographical Sketch and Bibliography

Larry M. Weisenthal, MD, PhD
I will be very happy to answer specific questions
or address specific concerns or criticisms from anyone.
Send e-mail to: Larry Weisenthal at 72203.2235@compuserve.com
voice: 714-894-0011
fax: 714-893-3658

This article was originally submitted Tuesday, 04-Apr-95 14:30:45 PDT .

Update added 5/25/2003

A Prospective Blinded Study of Predictive Value of Extreme Drug Resistance Assay in Patients Receiving CPT-11 for Recurrent Glioma

Rita S Mehta, M.D., Timothy F. Cloughsey, M.D., Ricardo J Parker, Ph.D., John P Fruehauf, M.D., Ph.D.

American Association for Cancer Research, 2000, Abstract #1645:

The Oncotech Extreme Drug Resistance (EDR) assay identifies non-responders to specific chemotherapy agents with more than 99% accuracy based on a retrospective blinded study of 450 patients (JNCI82:582,1990). A prospective blinded study was undertaken to confirm the predictive accuracy of the EDR assay. EDR assays were performed on biopsy specimens of recurrent glioma prior to initiating treatment with CPT-11 (SN38). For the first 15 evaluable patients, we found a significant association between EDR assay results and median time to progression and overall survival. In vitro drug resistance was categorized as either extreme (EDR), intermediate (IDR) or low (LDR). The EDR category has been associated with less than 3% clinical response rates. Compared to literature predicted response rates, the LDR category has been associated with 1.5-2-fold higher response rates. In the present prospective study, median time to tumor progression was significantly shorter at 59 days for the EDR group (n=4) as compared to 116 days for LDR/IDR group (n=11), using the Mantel-Haenszel log rank test (2-tailed p=0.035) with a hazard ratio of 3.70 (95% C.I.: 1.56-50). The one patient in the EDR group who survived a total of 283 days was crossed over to an antiangiogenic agent after progressing on CPT-11 treatment. This patient experienced a short period of stable disease on antiangiogenic therapy. Yet, there was a significantly adverse survival of 90 days for the EDR group compared to 265 days for LDR/IDR group (log rank test; 2-tailed p=0.028). Further, hundred day survival was highly statistically significant in favor of the LDR/IDR group (Fisher’s exact test; 2-tailed p=0.008). These prospective data support the retrospective findings on the predictive accuracy of the EDR assay and suggest that patients should avoid treatment with agents to which their tumors demonstrate in vitro extreme drug resistance.

Source: http://www.oncotech.com/innovation/PublicationDetails.asp?id=21