John Stamos - Public Service Announcement

Rational Therapeutics
750 East 29th St.
Long Beach, CA 90806
Phone: (562) 989-6455
Fax: (562) 989-8160
E-mail: Client.Services 800-542-HELP (4357)

Response Rates

Relationship Between Assay Results And Probability Of Response As A Function Of "Expected" Response Rate

Expected Response Rate of Neoplasm to Tested Drug



* "EDR": Extreme Drug Resistance Assay results are at least one standard deviation more resistant than median of results of reference database assays.

Response Rates

Relationship Between Assay Results And Probability Of Response As A Function Of "Expected" Response Rate

Expected Response Rate of Neoplasm to Tested Drug

* "EDR": Extreme Drug Resistance Assay results are at least one standard deviation more resistant than median of results of reference database assays.

This Bayesian curve was generated from the cell death assay performance characteristics established in over 500 published clinical correlations. The two curves describe the likelihood of clinical response for a "sensitive" [upper curve] or a "resistant" [lower curve] result. These assays discriminate patients into those with higher vs. lower than average likelihoods of response. The actual response probability, i.e. "predictive accuracy," derives from the assay's sensitivity* and specificity* when applied in the context of drug responsiveness by diagnosis.

Let's look at results obtained in a colon cancer patient. 5-FU based therapy provides an approximate 20% response rate, e.g. 20 of 100 treated patients respond and 80 of 100 fail. The assay's role is to distinguish this population of patients into two groups. Those found "sensitive" depicted by the upper curve have a response likelihood of 50% (16/32), while those found "resistant" depicted by the lower curve have a response likelihood of 6% (4/68).

Stringency of the definition leads to EDR

Extreme Drug Resistance is defined statistically as an EVA-PCD™ assay result falling more than one standard deviation from the median cell survival value. Patients whose ex vivo results reveal EDR have an extremely low likelihood of response to the drug in question. The stringency of the definition leads to EDR being identified in only a small fraction of all patients. Yet when present, it can be used to eliminate truly inactive agents from consideration (Nagourney, R.A. Proc. ASCO #1073, 1996; Weisenthal, INCI, 1991).

Likelihood of Clinical Response

 
Pre-Test Response
Probability %
Sensitive %
Resistant %
Clinical Decision
GROUP
I 		10 27 1
TO TREAT
OR NOT
TREAT PATIENT
20 45 2
30 59 3
GROUP
II		 40 68 5
MOST EFFECTIVE
LEAST TOXIC
REGIMEN
50 76 8
60 83 11
GROUP
III		 70 89 16
SELECT
A POTENTIALLY
CURATIVE REGIMEN
80 93 24
90 97 43


This table illustrates the clinical utility of the EVA-PCD™ assay broken down into 3 principal clinical groups.

Group I
Group I are those patients with low likelihoods of response. These would include those with melanomas, NSCLC and GI neoplasms. The principal question in this population of patients may not be which chemotherapy but instead whether or not chemotherapy should be used.
Group II
Group II are patients generally responsive but infrequently curable. These include many breast cancers, ovarian cancers and most small cell lung cancers. The principal issue is the selection of the most effective and least toxic regimen from amongst comparable choices.
Group III
Group III are the curable malignancies which are principally the hematologic neoplasms. Intelligent application of the EVA-PCD™ assay can indicate the most effective (potentially curative) regimen from amongst comparable choices.


Thus, three distinct applications, depending upon disease type, can be envisaged:

  • To treat or not to treat,
  • Selecting the most effective/least toxic combination, or
  • Selecting a potentially curative regimen.