A comparison of clonogenic, microtetrazolium and sulforhodamine B assays for determination of cisplatin cytotoxicity in human ovarian carcinoma cell lines.

An assay based upon quantitative staining of cellular protein by sulforhodamine B (SRB) has recently been adopted by the NCI for large-scale screening of new drugs. However, there are few data available regarding whether the SRB assay is comparable to other established methods. Cisplatin cytotoxicity was determined in 16 human ovarian carcinoma cell lines by both SRB and clonogenic assays, and by microtetrazolium (MTT) assay in seven cell lines. Cell lines were derived from untreated patients (some of which were selected for cisplatin resistance in vitro) and from patients clinically refractory to cisplatin-based chemotherapy. There was excellent linear correlation between SRB staining and cell number in all cell lines (r = 0.972-0.999). IC50 values obtained by the SRB and clonogenic assay (r = 0.824, P = 0.000022) were highly correlated, although values obtained in the SRB assay were uniformly higher. IC50 values obtained by SRB assay also correlated well with results obtained by MTT assay (r = 0.906, P = 0.0010). Overall, the SRB assay permitted rapid and reliable assessment of cisplatin sensitivity in these cell lines and compared favourably with clonogenic and MTT assays.