Direct analysis of mitochondrial toxicity of antiretroviral drugs

Objectives: Mitochondrial toxicity is a serious side-effect of antiretrovir al drugs, especially nucleoside reverse transcriptase inhibitors (NRTI). An in vitro assay to predict mithocondrial toxicity of in-use and development al NRTI would be invaluable. To test the ability of a cytofluorimetric tech nique to predict the mitochondrial-dependent pancreatic and hepatic toxicit y we used didanosine (ddl) alone or in combination with hydroxyurea (HU). Methods: The technique is based on the ability of the lipophilic cation JC- 1 to enter selectivrly into mitochondria and change its colour as the membr ane potential changes due to toxicity. Mitochondrial toxicity by HU and ddl was evaluated in pancreatic and hepatic human cell lines. The results were expressed as mitochondrial toxicity index (MTI), ranging from 0 to 100: th e negative control was 0 and 100 indicating maximal toxicity. Results: Dose-dependent pancreatic toxicity of ddl was evident after 14 day s of culture (MTI 34 +/- 4 at 100 muM, 10 +/- 4 at 10 muM, 2 +/- 3 at 1 muM ddl). HU alone was not toxic (MTI 7 +/- 10 at 100 muM, 2 +/- 2 at 50 muM a nd 2 +/- 4 at 10 muM HU); however, HU increased the toxicity of high, but n ot low, concentrations of ddl. For example, the MTI of 10 muM ddl plus 50 m uM HU was 54 +/- 9. Negligible mitochondrial toxicity was observed in the h epatic cell line exposed to ddl alone or in combination with HU. Conclusions: This in vitro assay might have in vivo relevance. First, ddl-r elated pancreatitis is dose dependent, and is reported more frequently than hepatic failure, consistent with our in vitro results. Second, patients wh o developed pancreatitis during randomized, controlled trials were treated with HU in combination with 400 mg ddl once daily (high peak concentration of ddl in the blood). In contrast, no pancreatitis was observed when HU was combined with 200 mg ddl twice daily (low peak concentration of ddl). Thes e in vivo results are consistent with our in vitro observation that HU incr eases pancreatic cell toxicity in the presence of high concentrations of dd l. The in vitro assay described here might be used to predict the mitochond rial toxicity of other NRTI, alone or in combination. (C) 2001 Lippincott W illiams & Wilkins.