In vitro rat myocyte cardiotoxicity model for antitumor antibiotics using adenosine triphosphate/protein ratios.

Cumulative cardiotoxicity consistently limits the use of antitumor anthracyclines such as doxorubicin (DOX). Several in vivo and in vitro model systems have been developed for screening cardiotoxic agents. Problems with these models include excessive time and nonquantitative toxicity end points. We describe an in vitro system for culturing cardiac myocytes which overcomes these problems. Optimal myocyte cultures were obtained using serial 0.2% crude trypsin digestions of hearts from 1-2-day-old rats. Three-day-old myocyte cultures were treated with DOX for 6 h at concentrations of 0.1 to 10 micrograms/ml (0.18 to 18 microM). Electron microscopy performed on control and DOX-treated cultures showed characteristic histopathological signs of anthracycline damage. These changes included mitochondrial swelling, nuclear pleomorphism, chromatin clumping, and a diffuse loss of membrane integrity. Intracellular ATP, quantitated by the luciferase bioluminescence method, was shown to provide a simple and consistent quantitative biochemical marker of myocyte viability over the range of DOX concentrations used. The results showed both time- and dose-dependent decrements in ATP/protein ratios 72 h following exposure to DOX at concentrations greater than 0.1 microgram/ml. Leakage of lactate dehydrogenase activity, trypan blue uptake, and myocyte beating rates were variable and not as sensitive as ATP levels for evaluation of myocyte viability. Other cytotoxic agents which are not known to be cardiotoxic (dactinomycin, 1-beta-D-arabinofuranosylcytosine, fluorouracil, melphalan, and vincristine), required extremely high concentrations to produce myocyte damage in vitro. Tests with anthracycline analogues also demonstrated the ability of the assay to rank-order cardiotoxic agents on a weight basis: idarubicin greater than DOX greater than daunomycin greater than aclarubicin. When the in vitro drug concentrations required to lower ATP/protein ratios to 50% of controls were related to clinically achievable concentration x time products, DOX and daunomycin proved to be the most cardiotoxic in this series. These results suggest that comparative cardiotoxic screening studies may be performed in vitro using ATP levels in beating neonatal myocytes.