Conclusions
ABT-737 caused tumor regression by apoptosis in H146 xenografts that mapped to a drug-specific, early increase in circulating cleaved cytokeratin 18 that subsequently declined. Circulating, intact cytokeratin 18 levels correlated with tumor burden. Cleaved caspase-3 and caspase-cleaved cytokeratin 18 in tumor correlated with treatment (P < 0.05, 2 hours; P < 0.001, 6, 12, and 24 hours; cleaved caspase-3, P < 0.05, 15 days; caspase-cleaved cytokeratin 18), indicating that events in plasma were tumor derived. These circulating biomarker data will be translated to clinical trials wherein serial tumor biopsies are rarely obtained.
Purpose
This study evaluated epithelial cell death ELISAs that measure circulating cytokeratin 18 in mice bearing small-cell lung cancer xenografts treated with a proapoptotic dose of the BH-3 mimetic ABT-737. Experimental design: H146 tumor-bearing and non-H146 tumor-bearing severe combined immunodeficient (SCID)/bg mice were treated with ABT-737 or vehicle control. Plasma collected before and 2 to 360 hours after treatment was analyzed by M30 (caspase-cleaved cytokeratin 18) and M65 (intact and cleaved cytokeratin 18) ELISA. In parallel, tumors were interrogated for cleaved caspase-3 and cleaved cytokeratin 18 as biomarkers of apoptosis.
Results
ABT-737-treated tumors regressed by 48 hours (P < 0.01) compared with controls, correlating with increased cleaved cytokeratin 18 (P < 0.01; 6 and 24 hours) and increased intact cytokeratin 18 (P < 0.01; 24 hours). Cleaved cytokeratin 18 levels decreased below baseline between 72 and 360 hours for ABT-737-treated and control mice whereas intact cytokeratin 18 decreased below the level of detection at 8 and 15 days in ABT-737-treated mice only. Apoptosis in tumors reflected changes in circulating cytokeratin 18 (cleaved caspase-3, P < 0.05 at 2 hours and P < 0.001 at 6, 12, and 24 hours; caspase-cleaved cytokeratin 18, P < 0.05 at 15 days, for drug treated versus controls). Conclusions: ABT-737 caused tumor regression by apoptosis in H146 xenografts that mapped to a drug-specific, early increase in circulating cleaved cytokeratin 18 that subsequently declined. Circulating, intact cytokeratin 18 levels correlated with tumor burden. Cleaved caspase-3 and caspase-cleaved cytokeratin 18 in tumor correlated with treatment (P < 0.05, 2 hours; P < 0.001, 6, 12, and 24 hours; cleaved caspase-3, P < 0.05, 15 days; caspase-cleaved cytokeratin 18), indicating that events in plasma were tumor derived. These circulating biomarker data will be translated to clinical trials wherein serial tumor biopsies are rarely obtained.