What do the designations “sensitive,” “intermediate,” “resistant,” and “EDR” (extreme drug resistance) signify?


These are determined by the degree of tumor cell death, relative to an appropriate (“apples to apples,” not “apples to oranges”) subset of comparison, database assays.


The way this works is as follows: For each specimen, we determine relevant parameters (tumor histology, specimen transit time before testing, degree of spontaneous tumor cell loss (or gain) in control cultures, degree of tumor cell three dimensionality, strength of metabolic signal, etc.). Our computer then searches through our database of more than 500,000 assay results and retrieves the most closely-matched dataset of assays. Then for each drug tested, at each concentration, and with each assay endpoint, we determine the mean and standard deviation for degree of cell death. Plus or minus 0.5 standard deviations from the mean denotes an “intermediate” assay result. Greater than 0.5 standard deviations denotes a “sensitive” result (i.e. a clearly greater than average degree of cell death).  Less than 0.5 standard deviations denotes a “resistant” result (i.e. a clearly below average degree of cell death.


Less than a full standard deviation denotes "EDR" (extreme

drug resistance).   Applying Bayesian statistics to the

“S,I,R,E” results (lower right figure) allows us to

estimate the assay predicted probability of response

for each drug.  This (lower right) graph shows the

relationship between expected, pre-test probability

of response (i.e. the chance of the drugs working

before we did the assays) versus the new, revised

chances of the drug working, given the reality of a

“sensitive,” “intermediate,” “resistant,” or “extreme-

ly resistant” assay result. The lower right graph was

generated by applying Bayes’ theorem  to the published

clinical correlation data shown in the lower left graph.

References at: http://weisenthal.org/oncol_t.htm