Purpose: Human solid tumor histocultures represent a clinically relevant experimental system for pharmacodynamic study. The evaluation of the drug-induced antiproliferative effect in histocultures is usually performed by manual microscopic scoring of individual cells. This procedure, because of its labor intensive nature, is performed on a single microscopic field, i.e. the field with the highest proliferative activity. Because regional heterogeneity in a solid tumor may result in different drug sensitivities in different parts of a tumor, there is the question as to whether the pharmacodynamic data determined in the most proliferative field is representative of those in the whole tumor. This question was addressed in the present study. Methods: A recently developed automated image analysis method was used to measure the labeling index of tumor cells. The drug-induced inhibition of DNA precursor incorporation into nuclei of cells in the region with the highest proliferative activity was compared to the inhibition in cells in the entire histoculture. This study was performed in human bladder tumor histocultures treated with several drugs (doxorubicin, mitomycin C, paclitaxel and 5-fluorouridine). A total of 724 pairs of data obtained from untreated and drug-treated histocultures (each data point representing the average of 1 to 6 tumor histocultures) were analyzed. Results: The absolute value of the labeling index in the most proliferative region (LIone) was significantly higher than the absolute value of the labeling index in the whole tumor (LIall), in both untreated and drug treated samples (mean difference of 18%, range 1-27%). However, when the absolute LI values in drug-treated samples were normalized to the values in untreated controls and expressed as a percentage of control, and used to construct the concentration-response curves, the two curves obtained using LIone and LIall yielded comparable pharmacodynamic parameters, i.e. curve shape parameters and drug concentrations that produce 30, 50, and 70% inhibition. Conclusion. These results indicate comparable pharmacodynamics in the most proliferative region and the whole tumor, and confirm the validity of using the most proliferative field for evaluating chemosensitivity in solid tumor histocultures.
