Evolution and clinical application of a rapid chemosensitivity assay.

The clinical usefulness of the soft agar colony-formation assay of in vitro chemosensitivity developed by Hamburger and Salmon is limited by long turnaround time (2-3 weeks), low success rate for small specimens, and clumping artifacts that can lead to erroneous predictions of resistance (false-negative errors). An improved technique was developed for measuring in vitro growth by incorporation of tritiated thymidine that can be performed in 5 days. With this rapid assay, 819 tumors were processed, with an overall success rate of 59.3%. This result compared favorably to the overall success rate of 58.2% for 1591 colony-formation assays because more small specimens could be submitted for the rapid assay. Melanoma and ovarian cancer specimens grew particularly well (76% and 75% successful, respectfully). Sixty-five correlations of in vitro and in vivo responses are available to date. None of 30 tumors, predicted to be resistant in vitro responded to chemotherapy clinically. Patients whose tumors were predicted to be sensitive in vitro had a 43% clinical response rate. The assay appears to be particularly accurate for predicting clinical resistance to chemotherapy, possibly because clumping artifacts do not occur in this system and peak achievable plasma concentrations of chemotherapeutic agents can be used. Optimal in vitro drug concentrations and culture conditions are still being defined, and improved success rates are being seen with more recent specimens. The introduction of this technique underscores the fact that in vitro chemosensitivity tests must continuously evolve to maximize their clinical application.