Assessing tumor drug sensitivity by a new in vitro assay which preserves tumor heterogeneity and subpopulation interactions.

We have designed an in vitro assay to assess the influence of tumor subpopulation interactions on drug response. The assay is based upon inhibition of growth of 1 mm3-pieces of tumor embedded in a collagen gel matrix. Tumor growth is quantitated by planimetry of each colony's image, formed with a split image tracing device attached to an inverted microscope. That expansion of the colonies in collagen gel represents growth through cell replication was demonstrated by releasing and counting cell nuclei. Outgrowths from pieces of tumors produced by a series of mouse mammary tumor subpopulation lines expanded in collagen gel at a rate characteristic of each cell line: the growth rate of tumor pieces was similar to that of the corresponding tumor line embedded as a cell bolus of cultured cells, indicating that growth of pieces of tumor is due to the tumor cells rather than to stromal components. When two cell lines were grown together in collagen cultures, interactions affecting growth rate were observed. Both tumor pieces and cell boluses from cultured cells of the relatively homogeneous cell lines displayed similar, characteristic sensitivities to adriamycin (ADR) in the collagen gel assay. Advantages of the collagen assay over cloning assays are (1) preservation of potential cellular interactions which may be important in assessing tumor drug sensitivity; (2) maximization of growth of all cell populations within the tumor, as compared to growth in agar; and (3) reflection of the zonal distribution of different subpopulations within tumors; and (4) simulation of the three-dimensional growth architecture found in vivo.