Cell Culture Drug Resistance Testing (CCDRT) Cell Death Assays:
Misconceptions Versus Objective Data
Some Technical Considerations: Frequently Asked Questions
How do you do the assays and what do they measure?
Tumor cells are isolated from solid tumors as 3-dimensional spheroids or clusters. This is often a very time-consuming, labor-intensive process. These cell clusters are cultured for 4 days in the presence and absence of drugs. In the case of hematologic neoplasms, comprised of discohesive cells, the neoplastic cells are isolated and enriched by a variety of different methods and also cultured for 4 days in the presence and absence of drugs. At the end of the culture period, the viability of the drugs is tested by three different methods. In the DISC assay, the entire contents of each individual cell culture microwell are cytocentrifuged onto permanent microscope slides and differentially stained to allow discrimination of normal and neoplastic cells and living and dead cells. The MTT assay measures mitochondrial metabolism in the entire cell culture. The redox assay measures total metabolic activity in the entire cell culture, using the Alamar Blue reagent to index the oxygen reduction potential of the culture medium.
Most disparities between the DISC and MTT assays are explained by the following: The DISC assay, though subjective, is specific for tumor cells while the MTT assay, though objective, measures net effects on cell metabolism of the entire (tumor + non-tumor) population of cells. The DISC assay measures the membrane and structural integrity of tumor cells (the former has been shown to be a surrogate for apoptosis and the latter is, in fact, analogous to assessing response by determining if the tumor mass has shrunk), while the MTT assay measures the mitochondrial metabolic integrity of all cells present in culture. DISC and MTT assays are usually in good agreement in cases where the majority of the total metabolic activity present in the cell cultures emanates from tumor cells. Assays may disagree when an appreciable total of the cell culture metabolism emanates from normal cells or when metabolic damage is disproportionate to structural damage. In the case of several specific drugs, either one endpoint or the other is considered to be more reliable, for specific biological reasons. The redox (Alamar Blue) assay is mainly used as a backup and as an independent check. Raw data from this latter assay are usually not reported, unless they are deemed to be of special importance in a given situation.
Raw data for the DISC and MTT assays are provided. We do this for the benefit of individual physicians and/or institutions who may later wish to do their own data analysis, make clinical and/or laboratory correlations, or whatever. We also frankly disclose the technical quality of the assays (criteria used to make the distinctions between technically excellent, good, fair, and poor assays are available upon request). We also provide other technical information about the assays, such as tumor cell survival in control cultures, strength of the formazan signal in the MTT assay, percentage of cells present in the cultures at the end of the culture period which are tumor cells, percentage of tumor cells in three dimensional clusters, and more.
How do you make the "Sensitive/Intermediate/Resistant/EDR (Extreme Drug Resistance)" determinations?
We compare the individual assay results to a universe of comparison, database assays. We have a large number of these comparison databases. For example, we have a master database of all results. This database contains results of high and low drug concentrations for both the DISC and MTT assay. As of August, 1999, this database contained the results of more than 300,000 individual standard (non-research) drug tests performed in fresh human tumor specimens, performed during the past 7 years alone.
Furthermore, we have individual databases broken down by type of assay and individual drug concentration tested. These can be further broken down according to tumor diagnosis, type of specimen (solid tumor versus effusion, etc.), clumpiness of the tumor specimen (which is a variable affecting in vitro drug resistance), spontaneous cell attrition of the control cultures (which is another variable affecting results), and other factors.
Basically, the situation is as follows: There are a number of independent variables which affect in vitro drug resistance. What we want to determine is the intrinsic drug resistance of an individual tumor specimen to the various drugs. But, in analyzing the data, there are a number of variables which correlate with in vitro drug resistance.
Tumor diagnosis is one variable. Prior treatment status is another variable. These variables are expected. But the clumpiness of the tumor specimen is also a variable. Work by Kerbel and Teicher suggests why this may be so, as these authors have described the phenomenon of "multicellular drug resistance." The tumor site is another variable (e.g. cells from effusions tend to be more drug resistant in vitro, which is probably reflective of a generally "healthier" state, owing to the fact that the cells survive collection and transport better than cells from solid tumor masses, and drug effects may be additive with physical damage from specimen collection and transport). The degree of spontaneous cell attrition in the control cultures is also, therefore, a variable, along with the total metabolic activity of the control cell cultures.
What it finally gets down to is comparing apples with apples and not with oranges. We start with a database of comparable specimens. By doing so, we reduce as many of the variables as possible which do not reflect the intrinsic drug resistance of the tumor cells within the patient. Once we have our proper comparison database, we then stratify the database, based on deviations from the median, for each assay system and at each drug concentration. One half standard deviation more "sensitive" than the median is the cut-off for a "sensitive" result. One half standard deviation more "resistant" than the median is the cut-off for a "resistant" result. One full standard deviation more "resistant" than the median is the cut-off for "Extreme drug resistance (EDR)." Note that we typically have 4 different determinations for each drug on which to base "Sensitivity" versus "Resistance" (high concentration DISC, low concentration DISC, high concentration MTT, low concentration MTT). In addition, we also have the back-up results from the REDOX (Alamar Blue) assay as a confirmatory endpoint. In situations where all results are in the "sensitive" range, I report out the results as "sensitive." When all results are in the "resistant" range, I report out results as "resistant." When all results are in the "intermediate" range (+/- 0.5 S.D. from the median), I report out results as "intermediate." When there is a disparity, I try to resolve the disparity by again reviewing the results and the original slides and try to determine if there is a reasonable explanation for the disparity which would indicate which of the results is most accurate. If I cannot resolve the disparity, the results would either be categorized as "intermediate" or "inevaluable," depending on the nature and degree of the disparity. Note that an "intermediate" result indicates that we can't say whether the specimen is clearly "sensitive" or clearly "resistant," meaning that the "assay-predicted response probability" remains the same as the "clinically-expected response probability."