Since most lymphoma are initially sensitive to almost all chemotherapy drugs, what is the rationale for using drug resistance assays on previously untreated patients with lymphomas?
REPLY: Many tumors are initially sensitive to chemotherapy. Ovarian cancers have in excess of an 80% response rate, even in the advanced stage, yet only 29% of patients remain alive at 5 years( i.e. the other 71% relapse and succumb to their illness).
Similarly, 70% or more of women with metastatic breast cancer respond to chemotherapy yet few (<5%) are cured.
With regard to NHL, the CR rate is up to 90% yet only 50-70% of patients remain in durable CR, depending upon prognostic categories.
If we use our experience in childhood ALL the point becomes clearer. In our 25 year follow up (15 years at time of initial report; Leukemia, 1995) children with ALL found sensitive to chemotherapy had an 80% Event free survival (EVS) as opposed to only a 20% EVS in the resistant group. This
despite virtually all of the children achieving a CR with initial therapy.
The principle being that by administering the most "active therapy" administered at the time of "first-line" therapy, the degree of cytoreduction will be optimal.
COMMENT: The ALL study (name and full text not provided) does not appear to compare outcomes of the children based on selection of therapy guided by the DRA, which, presumably, is the reason some patients ask for this test.
Copying from the Physician information page of the sponsor's website: After fifty years of experience with cytotoxic drugs, why is there no widely accepted laboratory method to select chemotherapy for individual patients? From this text it appears that DRA is being offered as a way to "rationally" select therapy agents, and not to predict survival and chemo-sensitivity in general.
We'd all love to see reliable data (from controlled prospective clinical trials) proving the clinical utility of the assays for patients with lymphoma - how it helps to improve outcomes by informing selection or avoidance of treatment agents.
The significance of the data depends on the study population. You can easily set up a study to predict an outcome when the outcome (resistance to treatment) is very predictable.
Quoting from the CMS transcripts: So, if you just pick an extreme population, it turns out those are pretty easy predictions to make, but it turns out that most patients aren't in the extreme, so it's relatively un-useful. Okay? Thus, the sample must be representative of the real world in which the test is to be used. ~ Dr. Burke
Since signals in the microenvironment of the tumor are known to contribute to resistance to cell death, how do the predictive assays account for these factors in cells outside the body?
REPLY: The laboratory assays have been calibrated against clinical outcomes. They are not designed to, nor do they, recreate the human body in a test tube. They cannot take into account immune response, and certain PK/PD factors that do contribute to response. This is one of the reasons why the assays have performance characteristics in the range of 80-90% sensitivity and specificity and not 100%. However, in over 1600 peer-reviewed published clinical correlations, ASSAY SENSITIVE PATIENTS RESPONDED MORE OFTEN THAN ASSAY-NEGATIVE PATIENTS (P <0.00000001)
COMMENT: No specific study is cited in the reply.
We do not know what is meant by "calibrated against clinical trials." , and assume that by "clinical correlations" the provider means observed associations for a broad range of cancers. As above, according to the testimony of Dr. Burk, you can easily set up a study by patient selection to predict an outcome when the outcome (resistance to treatment) is very predictable. Randomized studies protect against this kind of error.
What clinical data do you have - specific to my type of lymphoma - that the assay will predict resistance to specific chemotherapy protocols
REPLY: There is a wealth and growing (see Ugurel, S Clin Canc Res. Sept 15th, 2006) body of evidence that supports the scientific basis of these assays. While we have accumulated excellent published prospective data in Lung, Ovarian, Breast, CLL and ALL (and now melanoma), not every disease at every stage of every subcategory has been prospectively studied.
COMMENT: The provider does not cite a prospective controlled clinical trial, which, again, is the foundation of evidence-based medicine. Referring to the "wealth and growing body of evidence" reminds of the comments made by Dr. Burke's at the CMS hearings:
"You have to talk a specific disease, a specific treatment, how does the test do? Not a conglomeration of diseases and treatments together."
How do the assays account for combined effects of different drugs in the body - the dosing of the agents, and the [interactions of the] metabolites of the parent drugs?
REPLY: Formal synergy analyses in thousands of patients account very well for drug
COMMENT: The provider has not provided a reference for the "Formal synergy analysis."; and the reply is not specific to lymphoma or consistent with the testimony of experts at the CMS hearing:
"But let me - as I recall reading these papers, that none of them [studies] actually routinely were testing two agents simultaneously on one. I mean, the majority of the papers that we read and that Dr. Burken presented single agents. Maybe serially they would test several agents, but not together in one Petri dish routinely.
DR. FRUEHAUF: Yes.
DR. HELZLSOUER: That's the same way, I interpreted them the same way, they were all single agent tests, and not combinations.
DR. FERGUSON: Yeah. I mean, all the published stuff we saw was single agent.
DR. HANDELSMAN: The bulk of it was, but not all of it.
DR. FERGUSON: Okay. Yes?"
Unfortunately, individual cases (testimonials) cannot be considered evidence. If so, the development of drugs and tests would be very easy indeed
REPLY: It may be useful to examine what the actual dictates of evidence based medicine are.
"Evidence based medicine is not restricted to randomized clinical trials and meta-analyses. It involves tracking down the best external evidence with which to answer our clinical questions" ."some questions about therapy do not require randomized trials (successful interventions for otherwise fatal
Sackett DJ (Professor NHS Research and Development Centre for Evidence Based
Medicine, Oxford, UK) et al BMJ, 1996
In conclusion: The evidence supporting the use of cell death assay is compelling and growing. Individuals must determine the threshold that satisfies their evidence standards.
COMMENT: Regarding the profound limitations of observed associations, we cite the comments of Dr. Steven Bratman, a nationally regarded expert in the field of alternative medicine:
"We now know that a host of "confounding factors" can easily create a kind of optical illusion, causing the appearance of efficacy where none in fact exists. The double-blind study is thus much more than a requirement for absolute proof of efficacy (as is commonly supposed) — it is a necessity for knowing almost anything about whether a treatment really works." altmedconsult.com/
Importantly, for the reader, we ask: What would be the impact on medical progress if the observations of the sponsor (with obvious financial conflict of interest) were sufficient to win marketing approval for drugs, or in this case: unproven tests to guide clinical decisions?
What are the dangers to patients, present and future, for waving standards of evidence for tests that guide treatment decisions?