Selective assessment of in vitro radiosensitivity of tumour cells and fibroblasts from single tumour biopsies using immunocytochemical identification of colonies in the soft agar clonogenic assay

Abstract

The assumed selective growth of tumour cells has formed the basis for the use of the soft agar clonogenic assay to test in vitro radio- and chemosensitivity of tumours. However, recent studies have demonstrated that fibroblasts proliferate in soft agar in addition to tumour cells. The present study was initiated to quantify the contaminating growth of non-malignant cells in the modified form of the Courtenay-Mills soft agar assay, in order to establish a reliable assay for estimating tumour cell radiosensitivity in squamous cell carcinomas of the head and neck. DNA flow cytometry analysis confirmed that ‘tumour fibroblasts’ (fibroblasts obtained from tumour biopsies) grow in soft agar. In contrast, white blood cells did not form colonies. Different media were tested with soft agar, but a selective medium for tumour cells was not found. Therefore, a colony filter-technique combined with an immunocytochemical analysis was developed to quantify the number of tumour cell and fibroblast colonies. In 12 tumour biopsies, 2–33% of the colonies were Cytokeratin AE1–3 positive, whereas 83–100% of the colonies were 5B5 fibroblast antibody positive. The parameter normally reported, the overall SF 2 (surviving cell fraction at 2 Gy) based on colonies in agar, was found to be statistically significantly correlated to the fibroblast SF 2, but not to the tumour cell SF 2. The overall SF 2 was significantly different from the tumour cell SF 2 in half of the tumours. Furthermore, the tumour cell SF 2 was not correlated to fibroblast SF 2. In consequence of our findings, correcting for fibroblast contamination is a necessity, when studying in vitro sensitivity of tumour cells. Combining the soft agar clonogenic assay with the new colony filter-technique and the immunocytochemical analysis appear to be useful for making this routine correction and for measuring the in vitro radiosensitivity of both tumour cells and fibroblasts from single tumour biopsies, which is of interest in future clinical studies on predictive assays.