Optimization of treatment conditions for studying the anticancer effects of retinoids using pancreatic adenocarcinoma as a model.

Retinoids are natural differentiation-inducing compounds that are promising as anticancer agents. Cancer cell lines are valuable in the investigation of the potential of retinoids for the treatment of specific cancers. However, using different treatment conditions but the same cell lines, investigators have produced markedly contradictory results for the effectiveness of retinoids. The present study examined different factors in the treatment conditions that may have masked or interfered with the effects of retinoids and, thereby, resulted in this conflict. Our studies revealed that the effects of retinoids on cancer cell proliferation were influenced by serum, the choice of vehicle (DMSO vs ethanol) and its concentration, phenol red, the degree of cellular confluence, and the method of assessing proliferation (cell number or [3H]thymidine uptake vs the MTT assay). Optimized conditions were the use of serum-free, ethanol-free, and phenol red-free media, investigating cells in the log phase of growth, using </=0.01% DMSO as the vehicle, and monitoring proliferation by cell number or [3H]thymidine incorporation into DNA measured after TCA precipitation. Using these conditions, retinoids were found to exhibit potent antiproliferative effects in pancreatic cancer cells with a variety of degrees of differentiation, even in cell lines previously documented as being retinoid resistant. Retinoids also induced morphological changes and cellular death that may indicate terminal differentiation and apoptosis. Copyright 1999 Academic Press.