Frequently-asked questions relating to cell culture drug resistance testing (CCDRT; also known as "chemotherapy sensitivity and resistance assays" or "CSRAs")


Part 1: General Introductory Questions

Q1: What are the differences between "drug resistance" assays and "chemosensitivity" assays and "drug response" assays? Between "in vitro" and "ex vivo" assays?

A. The differences are largely semantic. All of these tests are geared to identify drugs which have greater or lesser than expected probabilities of helping the cancer patient. Depending on the conditions selected for the assays, some may be better than others with regard to accuracy in identifying either "good" or "bad" drugs. However, it is much more useful to define assays by the sensitivities and specificities (mathematical terms which define the performance characteristics of the assay), than to arbitrarily label an assay one way or the other. "Cell culture drug resistance testing" ("CCDRT") is a generic, descriptive term which refers to testing fresh human tumors in cell culture to determine whether they are more resistant or less resistant to individual chemotherapy drugs and drug combinations. All of the laboratories mentioned below provide cell culture drug resistance testing, according to the definition I just gave. There is no difference between "in vitro" and "ex vivo." They both refer to cell culture drug resistance testing.

Q2: What is the purpose of CCDRT?

A: To identify the best forms of chemotherapy for cancer patients on an individual basis.

Q3: If CCDRT is not performed, on what basis are treatments chosen?

A: On the basis of historical experience, preferably defined by controlled clinical trials.

Q4: Have there been controlled clinical trials to prove that CCDRT improves the results of cancer chemotherapy?

A: There have been four prospective, controlled, but non-randomized clinical trials which have been published. In addition, there has been one prospective, controlled, randomized trial presented at a cancer meeting, but not yet published.  Finally, there is a large, prospective randomized trial in progress in chronic lymphocytic leukemia, which has not yet been reported even as a meeting abstract, but which has reached its patient accrual goals, so that results should be available at some point in the future.  In breast cancer, 73 patients with advanced disease received assay-directed therapy, while 73 control patients received "physician's choice" chemotherapy. The assay-directed group had a significantly higher response rate (77% vs. 44%) and a trend for improved overall survival, which was not statistically significant, owing to the relatively small number of patients entered onto the clinical trial (it has often taken between 2,000 and 50,000 patients to "prove" statistical significance in breast cancer clinical trials). In the second study, 25 ovarian cancer patients received assay-directed chemotherapy, and these were compared with 30 non-randomized but well-matched controls. There was a response rate advantage to assay-directed therapy (64% vs. 37%) and a significant improvement in progression-free survival, with a trend for improved overall survival, although this was not statistically significant, again owing to the small size of the study. Additionally, another pilot study measured the survival of 21 stomach cancer patients following surgery with curative intent. A small group of patients receiving adjuvant chemotherapy directed by assay survived significantly longer than another small group of patients who received "physician's choice" chemotherapy.  In another ovarian cancer study, 50 relapsed patients treated with assay-directed therapy were compared with 50 matched historical controls. In "platinum-sensitive" disease, there was an improvement in overall survival from 15 months (control group) t 38 months (assay-directed group).  In the "platinum-resistant" group, there were no differences between control and assay directed groups.  In the only prospective, randomized trial (reported only as a meeting abstract and not yet in the literature), there was a non-significant trend for improved response rate in platinum-resistant patients with assay-directed therapy, with no improvement in survival.  This latter trial was very instructive, however.  Three different drug combinations were first identified by means of CCDRT (gemcitabine/threosulfan, mitoxantrone/paclitaxel, and gemcitabine/cisplatin.).  In the first year of the study, many patients in the assay-directed arm received these then-novel combinations, versus no patients in the control arm. There was a strong trend during the first year for improved response and progression-free survival in the assay-directed group. Then the clinicians saw what was happening and started to treat all of the control patients with these combinations and differences between the group narrowed.  What ended up happening was that both groups had extraordinarily high response rates (about 40%, compared to only 20% reported in the literature for conventional chemotherapy in platinum-resistant ovarian cancer).  As discussed in more detail elsewhere on this website, I believe that the results of this trial, along with the rest of the literature, suggest that the only patients with diseases like recurrent ovarian and breast cancer who should receive aggressive combination chemotherapy are those with tumors which show clear sensitivity in valid and well-controlled CCDRT.  Patients with relatively resistant tumors should not be treated aggressively merely to achieve a partial response, when the downsides of aggressive treatment are major organ toxicity, immunosuppression, and drug induced mutations to more virulent disease in genetically-unstable cancers.  Such patients (with relatively drug resistant cancers) should instead receive less toxic, less mutagenic therapy with agents such as high dose tamoxifen, "targeted" drugs such as gefitinib, erlotinib, sunitinib, and anti-angiogenic therapy with agents such as thalidomide and bevacizumab.

 The above clinical trials strongly suggest, but do not conclusively prove, that assay directed therapy provides clinical benefit.

Q5: If that is the case, then isn't it disadvantageous and even dangerous to the patient not to receive the "standard" treatments identified in the controlled clinical trials?

A: Only in cases where a clearly superior form of one-size-fits-all chemotherapy has been identified in controlled clinical trials. However, there are very few situations in cancer treatment where one clearly superior form of chemotherapy has been identified. One such case might be previously-untreated testicular cancer, where the combination of cisplatin/etoposide/bleomycin on one particular schedule has been shown to be superior to all other treatments with which it has been compared. But there are virtually no other situations in cancer chemotherapy in which this is true. In virtually all situations, there are several to many different forms of chemotherapy, all of which would be expected to have similar probabilities of benefit, if chosen by random selection and given to an "average" group of patients.

Q6: Getting back to an earlier question, how do oncologists choose between different forms of chemotherapy which would all be expected to produce similar results?

A: Sometimes on the basis of toxicity differences. Often on the basis of personal experience and personal biases. Often with consideration to which treatments will provide favorable insurance payments to the treating physician. Oncologists are, in effect, running retail pharmacies, where they (or their institutions) buy drugs wholesale, negotiating the best deal, as in any well-run business. Insurance reimbursement for different chemotherapies may range from below the actual wholesale cost, to ten or more times the actual wholesale cost. Obviously, pharmacists want to sell the drugs with the highest profit margin (or at least avoid selling drugs which lose them money), and this factor certainly does play a role in many treatment decisions. Remember, in virtually all forms of cancer, there are a wide range of treatments which would produce approximately the same results if chosen by random selection. So it is not clearly unethical for an oncologist to be aware of different profit margins with different treatments, although this certainly has at least the appearance of a conflict of interest.

Q7: Getting back to CCDRT, of what benefit is this if there are many treatments which produce the same results?

A: Notice that all the "rigorous clinical trials" have identified are the "best" treatments for the "average" patient. This has been referred to as the lowest common denominator theory of cancer chemotherapy. But cancer is not an "average" disease. Cancer is far more heterogeneous in response to various individual drugs than are bacterial infections. The heterogeneity of human cancer is shown both by the fact that some patients derive great benefit from treatments which fail to help (and often harm) the majority of patients who receive the treatment. And many patients fail to benefit from 1st line chemotherapy, only to derive great benefit from 2nd or even 3rd line chemotherapy. These patients should have received the correct treatment the first time around. Everyone agrees that the earlier in the course of the disease that the most active treatment is given, the better the result for the patient.

Q8: So what is the evidence which shows that these tests help patients?

A: There is considerable evidence validating the clinical relevance of the tests (outlined below and presented in detail elsewhere on this web site). However, there have been only the above referenced clinical trials (two) performed to prove that when drugs are selected on the basis of assay results, the results of therapy is then improved. This means that trials in other diseases have never been done, not that the trials were done and the assays were shown not to be helpful.

Q9: Why haven't the trials been done?

A: The most promising assay technologies (see below) are all public domain, non-proprietary technologies. Prospective, randomized, controlled clinical trials cost millions of dollars. We are talking hundreds of potential drug combinations and scores of diseases. As correctly noted by Dr. Maurie Markman of the Cleveland Clinic, proof of "efficacy" in one clinical situation wouldn't do anything at all to prove "efficacy" in a different clinical situation. And the trial to prove "efficacy" in even one clinical situation would become instantly irrelevant with the indroduction of the first new drug which wasn't available for testing at the time of the trial which proved "efficacy." No one is going to spend his/her own money on such trials when the result is that one hundred laboratories (some with the vast resources of very large biomedical companies) will immediately go into business to compete with the pioneer labs, if and when the pioneer labs complete successful trials.

Q10: What about getting grants from the NCI or the American Cancer Society to do these studies? What about using the huge cooperative group clinical trials network?

A: Major attempts have been made to obtain grants, design studies, and recruit investigators. These attempts were unsuccessful owing to many factors quite beyond the control of the investigators and laboratories which tried to get the studies done.

Q11 (accusatory statement): You sound as if you are trying to cop out. What if everyone that performed laboratory tests had your attitude. Patients would be victimized by lots of unproven tests.

A (retort): The tests are not "unproven." There is a vast, diverse, and entirely consistent literature documenting the correlation between test results and treatment results. The tests have been shown - without any challenge whatsoever - to identify relatively "good" and relatively "bad" forms of chemotherapy. It simply hasn't been proven that using the tests improves clinical outcome (again, because the relevant trials haven't been performed; not because they were performed and the results came back negative). We must remember that we are considering laboratory tests, which provide information, and the tests themselves are only tests and not treatments. To my knowledge, there are no laboratory or radiographic imaging tests in all of cancer medicine which have been proven to improve the results of treatment (this includes tests such as estrogen receptors, panels of monoclonal antibodies to characterize tumors, bacterial culture and sensitivity tests, and even serial CT and MRI scans performed for the purpose of measuring the size of tumors to gauge their response to therapy). The standards used to judge the utility of laboratory and radiographic tests have always been (1) acceptable accuracy of clinical correlations and (2) clinical utility, in the judgement of the physician ordering the test. Demanding proof of "efficacy," as opposed to proof of accuracy, is completely unprecedented for laboratory tests in cancer (and in almost all areas of medicine, for that matter). Cancer is a disease which has always been managed on the basis of "best evidence" and not on the basis of "conclusive evidence," which is lacking in virtually all situations in clinical oncology, including those situations in which clinical trials to identify the best treatment for the average patient have been performed and published and meta-analyzed (e.g., see discussion of Taxol in ovarian cancer elsewhere on this web site).

Q12: Why haven't universities and cancer centers taken the lead in developing assays and performing clinical trials?

A: One needs to understand the history of research into cell culture drug resistance testing (CCDRT): One must begin by understanding that there is a clear divide between CCDRT based on cell proliferation as an endpoint and CCDRT based on cell death as an endpoint. Historically, the cell proliferation endpoint received great attention, as a result of studies by Salmon, Von Hoff, and others during the late 1970s and early 1980s. These studies occurred during the heyday of the oncogene discovery period in cancer research, where oncogene products were frequently found to be associated with cell growth, and where cancer was most prominently considered to be a disease of disordered cell growth. In contrast, the concept of apoptosis (programmed cell death) had yet to become widely recognized. Also unrecognized were the concepts that cancer may be a disease of disordered apoptosis/cell death and that the mechanisms of action of most if not all available anticancer drugs may be mediated through apoptosis. When problems with proliferation-based assays emerged, there was little enthusiasm for studying cell death as an alternative endpoint. These factors explain the abandonment of research into CCDRT by American universities and cancer centers by the mid-80s. However, clinical laboratories began to offer CCDRT as a service to patients in the USA by the late 1980s, and studies of CCDRT continued in Europe and Asia.

Q13: So what has happened since the American universities and cancer centers abandoned research into CCDRT?

A: Assays based on the concept of cell death (as opposed to cell proliferation) have been proven to identify treatments associated with relatively good and relatively poor results (both with respect to shrinkage of tumors and with respect to cancer patient survival) in a wide range of human cancer, including leukemia, lymphoma, ovarian cancer, breast cancer, gastrointestinal cancer, and many other forms of cancer. Patients treated with drugs active in the assays have, on average, a 7-fold greater chance of benefiting from treatment with drugs showing good results in the assays compared with treatment with drugs showing poor results in the assays.

Q14: I'm confused by the description you just gave. Are the tests accurate or aren't they...straight answer please.

A: The tests are VERY accurate. Just as a barometric pressure reading with a precision $1,000 German barometer is very accurate. When the barometer says that the atmospheric pressure is rising, you can believe what it says. But how does this help us in predicting weather? It helps in the following way: absent a barometric pressure reading, all we have is historical experience (I'm leaving out satellite imaging, etc., to simplify the point I'm trying to make). In Southern California in November, there is about a 10% chance of rain on any given day. With a rising barometer, this decreases to less than 4%. With a falling barometer, the chance of rain increases perhaps to 35%. So the instrument is: (1) "accurate" (in correctly measuring atmospheric pressure) and (2) useful (in more clearly defining the probability of rain). Note that the instrument is "accurate," but the correlation is only "useful," and not perfect. To take this analogy one step further, let's move the barometer to Boston (still in the month of November). In Boston, historical experience indicates that there is a 30% chance of rain on any given day. With a rising barometer, this falls to less than 10%. With a falling barometer, this increases to more than 70%. Note that the accuracy of the instrument in measuring barometric pressure hasn't changed, but the correlation has changed (the instrument is still "useful," as it's much better to decide to play golf on a day with a 10% chance of rain than with a 70% chance of rain, even though the correlation between barometric pressure and actual weather is not perfect). The assays are very similar to the above analogy. They are VERY accurate in measuring tumor cell death in the laboratory. Tumor cell death in the laboratory clearly correlates with tumor cell death in the patient. Drugs are more likely to kill tumors in patients in cases where they have been found to kill tumors in the laboratory (and not to kill tumors in patients in cases where they don't kill tumors in the laboratory). But the precise correlations vary according to the underlying probability of the drugs working on average, based on historical experience, just as in the case of historical experience with probabilities of rain on any given day in either Southern California or Boston in November. Just as with barometric pressure, the correlation is about 7:1, meaning a seven-fold greater probability of clinical benefit for drugs with good activity in the laboratory, compared to drugs with poor activity in the laboratory (or a seven-fold greater probability of good weather with a rising barometer than with a falling barometer).

Let's take a few examples:

Previously-untreated ovarian cancer. Historical probability of clinical benefit with carboplatin = 70%. Probability of benefit with a favorable assay = 88%. Probability of benefit with an unfavorable assay = 18%

Previously-untreated breast cancer. Historical probability of clinical benefit with Adriamycin = 50%. Probability of benefit with a favorable assay = 80%. Probability of benefit with an unfavorable assay = 11%

Previously-untreated colon cancer Historical probability of clinical benefit with 5FU = 20%. Probability of benefit with a favorable assay = 39%. Probability of benefit with an unfavorable assay = 4%

Previously-treated colon cancer Historical probability of clinical benefit with irinotecan = 15% Probability of benefit with a favorable assay = 28% Probability of benefit with an unfavorable assay = 3%

And so on. The mathematical relationship (defined by Bayes' Theorem) between the historical "expected" probability of response to chemotherapy and the more clearly-defined "assay-predicted" probability of response is shown in a graph elsewhere on this website.


Part 2: Practical Questions

Q15: Your web site lists a number of laboratories which do this testing. Can you tell me more about the pros and cons of the different labs and tests?

A: I must preface this by saying that I have an undeniable bias that my approach is the best and that I do the best job in the world with this testing. Doubtless, others feel the same way about their own work. But you asked the question (and it's a question which is ALWAYS asked), so I'm finally going to answer it. Fairness demands that I provide an opportunity for anyone who disagrees with anything said here to provide a rebuttal, explanation, clarification, correction, or whatever. Therefore, I will immediately post on this web site any rebuttals, explanations, clarifications, corrections, or whatever which anyone wants to submit. Just e-mail it to me: mail@weisenthal.org (Larry Weisenthal). Also provided below are links to the web sites of the various laboratories, which should also be visited and reviewed to get a more complete and balanced understanding of the different laboratories.

Laboratories maintaining patient contact Web Sites:

Rational Therapeutics, Inc.
Oncotech, Inc.
Anticancer Inc
Human Tumor Cloning Laboratory (link currently inoperable)
Bath Cancer Research Unit (Bath, England)
Diatech Oncology (Montreal, Canada)
L.A.N.C.E. (Bonn, Germany)
TherapySelect (Heidelberg, Germany)
Precision Therapeutics, Inc
Genoptix, Inc.
Oncovation, Inc.
Genomic Health ("Oncotype DX" Prognostic Gene Expression Assay)
Cancer Therapeutics, Inc. (A tumor cryobanking service)

(Weisenthal Cancer Group does not maintain a website for patient communications relating to its clinical services, but information concerning these clinical services is available by calling 714-596-2100 after 9 AM Pacific Time). A general information pamphlet (in pdf format) may also be downloaded for reading or printing (click here)

(Weisenthal's biased and subjective description of the labs located in the USA follows):

Weisenthal Cancer Group. This is a private laboratory founded and run by me. I'm an MD board certified internist and medical oncologist and also a pharmacology PhD who has been working in this field on a largely full time basis since finishing my oncology training in the Medicine Branch of the National Cancer Institute in 1979. I have personally worked with most of the important technologies, including the clonogenic assay, thymidine incorporation assay, BrdU assay, DISC assay, ATP assay, the MTT assay, and a more "specific" apoptosis assay, based on measuring caspase 3/7 activity (the latter assay correlates very well with the DISC, ATP, and MTT assays, but the caspase 3/7 assay is disadvantageous, in that it requires testing at multiple time points, as the peak in caspase activation varies between different specimens and different drugs). I've done only a little work with Dr. Hoffman's "HDRA" assay, but enough to understand its advantages and disadvantages. I've never personally worked with Dr. Baker's adhesive tumor culture assay, or with Dr. Kornblith's (different type of) adhesive culture assay, or with the fluorescein diacetate assays of Rotman and Larsson/Nygren. But I believe that I have a good understanding of these latter assays and I am certainly familiar with the published data. Along with my friend and former colleague Robert Nagourney, I co-founded (originally a 4/7ths ownership share; by now much diluted) Oncotech in 1985 and served as the VP for R&D and Laboratory Director of Oncotech until January, 1992, when I left to start the Weisenthal Cancer Group.

It must be understood that none of the above technologies are truly proprietary, in the sense that existing patents are not defendable, in my opinion, and that, in any event, there are workaround solutions available to anyone who wants to perform assays based on the above principles. I do what I do because it's what I have found to be the best/most useful/most advantageous. This doesn't mean that all of the above methods could not be made to provide useful information. It just means that I think that there are advantages to doing it the way we do it in our laboratory.

We test tumors using at least two of 5 available assay endpoints, and most often use 3 of the 5. In the DISC assay, the entire contents of the cell culture are cytocentrifuged onto permanent microscope slides and differentially stained to allow discrimination of normal and neoplastic cells and living and dead cells. The endpoint for cell death is delayed loss of membrane integrity, which has been found to be a surrogate for apoptosis. The MTT assay measures mitochondrial metabolism in the entire cell culture. The ATP assay measures cellular ATP content by luminometry, based on the luciferin/luciferase reaction. The redox (resazurin) assay measures total metabolic activity in the entire cell culture, using the Alamar Blue reagent.All of the above are cell death endpoints. The caspase 3/7 assay measures the activation of caspases 3 and 7 using luminometry.  We also have an assay based on DNA synthesis available for specialized purposes.

Furthermore, results are compared with a database of assays which have similar technical characteristics, as the technical characteristics of these assays influence the in vitro results, along with the intrinsic drug resistance of the tumor cells. By controlling for technical characteristics such as spontaneous cell loss during culture, metabolic signal at the conclusion of the culture, time in transport between biopsy and cell culture, the degree of tumor cell clustering, and so forth, clinical correlations are improved.

The advantages of the above comprehensive system are the following:

1. Evaluability rates are very high (>95% of all specimens submitted produce clinically useful results).

2. It allows for very comprehensive testing. On average, we test 17 drugs and combinations at two concentrations in three different assay systems. In contrast, with the soft agar thymidine endpoint used by Oncotech and Impath, evaluability rates are much lower and typically only a half dozen drugs are tested at a single concentration in a single assay system.

3. Virtually all types of tumors and specimens may be tested: leukemias, lymphomas, malignant effusions, solid tumors, and so forth.

4. Clinical validation data are superior to those available for any other assay system(s).

5. I think that the results we obtain are the best currently available in the world for the purpose of identifying the treatments most likely to work in individual patients for virtually all of the most important available drugs.

The disadvantages are the following:

1. Such comprehensive testing is a whole lot of work and is expensive (although charges compare favorably with those of other laboratories providing much less comprehensive services).

2. It requires the continuous, hands-on presence of an experienced oncologist/laboratory person like me. When I get overburdened, it can take me a long time to get the results reported out.

3. Neither we (nor anyone else) has a whole lot of data with some of the newest agents, particularly those working with very novel mechanisms of action, such as STI571 and ZD1839.

Rational Therapeutics Institute (Dr. Robert Nagourney). Dr. Nagourney uses one of the same core technologies which is used by the Weisenthal Cancer Group, namely the DISC assay (described above), which Dr. Nagourney refers to as the "Ex Vivo Apoptosis Assay" or "EVA." Nagourney's work is quite good and I recommend his laboratory as the first alternative to our own. Comparing and contrasting "us" with "them," I spend full time doing the assays, while Dr. Nagourney maintains a clinical oncology practice, and so is more part-time in the lab. We use more assay endpoints, which I think is advantageous for a number of reasons. I consider Dr. Nagourney, however, to be the best in the world with respect to taking assay results (either ours or his own) and using the laboratory results to design a custom-tailored treatment regimen for an individual patient. When a very close family member of mine needed "custom chemo," we did the testing and gave Dr. Nagourney the results to use in designing and administering an innovative (and successful) treatment regimen. I often refer to Robert as being one of the 10 smartest and 10 most idealistic people I've ever met in my life. That pretty much describes him and his work.

Oncotech Oncotech was, as noted above, co-founded by Robert Nagourney and I in 1985. Robert left in early 1991 and I left in January, 1992, in both cases because we felt that we had given our best to start and grow the company, but that we wanted to move on to new things. Oncotech has for the last 10 years been in the capable hands of Frank Kiesner and John Fruehauf, the latter of whom replaced me as R&D and Laboratory Directors. Oncotech has continued to use as its main "bread and butter" assay the soft agarose tritiated thymidine assay, which is a direct descendent of the original Salmon/Von Hoff "Human Tumor Stem Cell" or "clonogenic" assay of the late 70s/early 80s. The assay has been validated with respect to identifying ineffective drugs which are very unlikely to work. The assay is not as good at identifying drugs which are more likely to work or to identify the disease-specific activity patterns of new drugs. While at Oncotech in the late 1980s, I identified the statistical relationships which defined "extreme drug resistance" with the assay, and I coined the term "EDR" which Oncotech subsequently used as the name of the assay in its marketing. Along with many other laboratories in the USA, Oncotech is heavily involved with R&D for molecular markers of cancer, some of which may eventually prove useful in predicting for the effectiveness of drug therapy. Oncotech's greatest asset is its bank of cryopreserved fresh human tumors, which may well be the largest in the world. This has allowed the company to establish joint venture research projects with a number of prominent institutions. Oncotech has recently entered into a technology/marketing partnership with a German laboratory, TherapySelect.

My own opinion is that the "EDR" assay is much less useful (and less well validated) than are the various cell death assays for the purpose of identifying effective treatment regimens, which is why I don't use the former assay, although I certainly could use this assay in my laboratory if I thought that it was advantageous. It doesn't mean that the information provided by the "EDR" assay is not helpful, it's just that there are, I think, better ways of getting more useful information -- namely: (1) higher evaluability rates, (2) more drugs tested per specimens, (3) multiple drug concentrations tested using multiple complementary endpoints, and (4) more extensive clinical validation of assay results.

NuOncology Labs (no longer providing clinical services): Dr. Fraser Baker's core technology was the adhesive matrix tumor assay. This used proprietary plasticware, in which cell monolayers are cultured over a proprietary "cell adhesive matrix" which is supposed to preferentially support the growth of tumor cells. Whether or not it does so was a matter of controversy in the past and there are conceptual problems with monolayer cell proliferation assays.

Studies by Teicher and Kerbel in mouse tumors showed that in vitro drug activity correlated with in vivo drug activity when tumors were tested in vitro as three dimensional clusters, but not when they were tested in two dimensional monolayers. There is now an extensive literature on what has been labeled "multicellular resistance". All published clinical correlations with true fresh tumor assays with cell death endpoints have tested the tumor cells largely in the form of three dimensional clusters. The only exception to this statement is a non-small cell lung cancer study from the NCI-Navy medical oncology group, in which subcultured cells (not true fresh tumors) were tested in monolayer culture. This latter study not surprisingly showed poor correlations; all of the other cited studies, which used true fresh (non-subcultured) tumor cells tested largely as three dimensional cell clusters (and not in monolayer culture), showed good correlations.

Theoretical objections aside, positive clinical correlations were described with this system in 1987, but confirmatory and follow-up studies have not to date been reported.

Precision Therapeutics This is a new (and apparently well-funded) laboratory currently making an aggressive sales effort. The basic technology, as I understand it, grew out of work by Dr. Paul Kornblith, a neurosurgeon. In the early 80s, Dr. Kornblith described an assay system in which cells were pre-cultured in monolayers for a period of days to more than a week and then treated with drugs. The assay endpoint was the detachment of cells from the surface of the culture plates, which correlates very well with loss of viability. The assay is conceptually a cell death assay, albeit one which is performed on pre-cultured (and ostensibly amplified) cells. Theoretical objections to the assay are that (1) it is a monolayer assay (see discussion of potential problems with monolayer assays described above, under "NuOncology Labs"), (2) it is testing a subcultured cell population, and studies by the NCI Navy medical oncology branch did not find that monolayer assays performed on pre-cultured, pre-amplified tumor cells clearly gave clinically relevant results.

The proof of the pudding are clinical validation studies. Of note is a very recent publication describing clear and significant correlations between assay results and progression-free survival in ovarian cancer. Taken in context of a growing number of studies showing correlations between the results of cell death assays and patient survival in ovarian cancer (discussed elsewhere on this website), these results indicate that the Precision Therapeutics assay (trademarked name ChemoFx) is capable of providing clinically –relevant information

Obviously, both the NuOncology and Precision Therapeutics assays are only applicable to solid tumors, and not to hematologic neoplasms, which cannot be cultured as adhesive monolayers.

Anticancer, Inc. This company uses the "histoculture drug resistance assay," which is basically an MTT assay performed on small pieces of freshly minced solid tumors cultured on a collagen substratum (sterilized pigskin). This assay has produced some very fine clinical correlations with treatment outcome in gastrointestinal cancers and more preliminary correlations in ovarian cancer. I do not, however, think that this assay offers any advantages over our own microcluster MTT, ATP, DISC, and Alamar Blue assays, and it has some serious disadvantages. The chief disadvantage is non-uniformity between culture wells, as there are big differences between the cellular composition and viability of different geographic areas in a single solid tumor biopsy. Our microcluster assays produce much better replicates and allow for much greater flexibility in testing and for much more extensive testing for tumors of similar size. Finally, our microcluster assays allow for better quality control regarding to what is actually being tested (tumor cells versus normal connective tissue, epithelial cells, and inflammatory cells).

I hasten, however, to again repeat that the proof of the pudding is in the correlations, and the correlations with the HDRA (with MTT endpoint) are quite good.

Human Tumor Cloning Laboratory. This was a research laboratory at the University of Arizona Cancer Center. This laboratory utilized several variations of the original Hamburger/Salmon soft agar "clonogenic" assay, although I believe that the endpoint most often used in later years was thymidine incorporation, similar to the endpoint used by Oncotech. The description under Oncotech, above, should be reviewed. I have a major problem with the web site of "The Human Tumor Cloning Laboratory," in that it stated that "The Human Tumor Cloning Assay (HTCA) is the only test on tumor biopsies that has been validated as being capable of predicting patient responsiveness or lack of responsiveness to specific anticancer drugs." This statement in completely false, as there are now vastly more extensive (and much more contemporary, as the HTCA data are now 15 - 28 years old) published data validating the clinical relevance of the cell death endpoint assays which I advocate.

Once again, I welcome (and will publish) all rebuttals, comments, etc. Please also visit the websites of the above laboratories, using links provided above.

April 13, 2006: [nb]: I'll try to update this section shortly, with reviews of several new labs whose links are provided above.