Renal and hepatic concentrations of platinum: relationship to cisplatin time, dose, and nephrotoxicity.

Autopsy tissues were obtained from 30 patients who had received cisplatin antemortem; the tissues were assayed for platinum by flameless atomic absorption spectrometry. Patients with antemortem evidence of renal toxicity had higher renal cortical platinum concentrations than did patients without evidence of kidney damage. In addition, patients with nephrotoxicity were more likely than patients without toxicity to have renal cortical platinum concentrations that were higher than renal medullary platinum concentrations. Overall, the two variables most closely associated with an increase in serum creatinine with treatment were renal cortical platinum concentration (P less than .02) and cumulative dose of cisplatin (P less than .05). These two variables were important independently of one another. Renal cortex platinum concentrations correlated inversely with time from last treatment until death, whereas hepatic platinum concentrations did not. In contrast, hepatic platinum concentrations correlated with dose of cisplatin while renal platinum concentrations did not. Our results suggest the following: (1) cisplatin-induced renal toxicity is tissue-platinum-concentration dependent and cisplatin-dose dependent; and (2) cisplatin may be handled differently at the molecular level in liver and kidney.