In vitro predictive testing: the sulfonamide era.

Since initial reports of use of a human tumor cloning system to predict response or lack of response of a patient's tumor to chemotherapy, there have been approximately 2,166 clinical correlations attempted. Overall, the percent true positives has been 69% while the percent true negatives has been 92%. Despite the high reliability of this system to predict patient response or lack of response, the cloning assay has not been put into general clinical use. Reasons for this include the past inability to grow a majority of the patients' malignancies and lack of controlled trials demonstrating an advantage of a cloning assay choice over a clinician's choice. Both of these problems are being addressed with greatly improved abilities to grow patient malignancies (70%-80% of patients' tumors can now be grown in vitro) and the reporting of results of prospectively controlled randomized trials. It is likely that the most significant limitation for the human tumor cloning assay will be the lack of agents with in vitro activity. It is proposed that we are in the "sulfonamide era" of in vitro drug sensitivity testing of human tumors. It is unlikely any predictive assay will be of major utility until a more active spectrum of agents becomes available for in vitro testing and in vivo treatment.