Relationship between cytotoxicity and site-specific DNA recombination after in vitro exposure of leukemia cells to etoposide.

Etoposide, an inhibitor of the normal religation activity of the nuclear enzyme topoisomerase II, can induce a secondary acute myeloid leukemia characterized by site-specific DNA rearrangements. The schedule of drug administration appears to be a clinical risk factor for this devastating treatment complication. We tested the hypothesis that prolonged exposure of leukemia cells in vitro to low concentrations of etoposide, compared with short exposures to high concentrations, could produce equivalent or greater desired cytotoxic effects, with decreased occurrence of undesired site-specific double-stranded DNA recombinational events (i.e., recombinogenesis). We used the frequency of V(D)J (variable-diversity-joining) recombinase-mediated deletions of exons 2 and 3 of the hypoxanthine phosphoribosyltransferase (HPRT) gene as a biomarker of etoposide-induced, nonhomologous, site-specific DNA rearrangement. A polymerase chain reaction-based technique was used to measure exon 2 + 3 deletions in human lymphoid leukemia CCRF-CEM cells 6 days after either 4-hour or 24-hour treatment with etoposide at clinically relevant concentrations. Cytotoxic effects of etoposide (determined by the number of viable cells present in the treated compared with the control [i.e., untreated] cells) were measured 6 days after treatment of the cells. The frequency of the exon 2 + 3 deletion following the two treatment-duration conditions was compared by use of the Mantel-Haenszel statistic. All P values resulted from two-sided tests. Cytotoxicity increased with increasing etoposide concentration and exposure duration, as expected. By day 6, the frequency of exon 2 + 3 deletions was significantly higher (global P value = .0003) after the 4-hour treatment than after the 24-hour treatment, regardless of whether the frequency was assessed at etoposide concentrations achieving equivalent (e.g., 95%) cytotoxicity (14.2 x 10(-7) versus 4.1 x 10(-7) or at equivalent etoposide concentrations (e.g., 1 microM) (10.8 x 10(-7) versus 1.3 x 10(-7). Thus, the ratio of desired cytotoxic to undesired recombinogenic effects was higher with the 24-hour schedule. After the treated cells were subcloned at limiting dilutions, the frequency of the exon 2 + 3 deletion increased from 16.3 x 10(-7) to 4.33 x 10(-3), indicating that the recombinational event is not necessarily lethal. For all drug concentrations and levels of cytotoxicity studied in CCRF-CEM cells, there was a greater ratio of cytotoxicity to genetic recombination following prolonged exposure to etoposide than following brief exposure. These data suggest that recombinogenesis is not inextricably linked to cytotoxicity. If confirmed in the clinical setting, the use of prolonged dosage schedules may provide a means to decrease the risk of etoposide-induced acute myeloid leukemia without compromising treatment efficacy.