Phase I Clinical-Laboratory Trial of Clofarabine (CLO) Followed by Cyclophosphamide (CY) for Adults with Refractory Acute Leukemias: Evidence for Enhanced DNA Damage.

CLO is a nucleoside analog with ribonucleotide reductase inhibitory activity. As such, CLO has the ability to impede repair of damaged DNA. We designed a Phase I clinical trial of CLO followed by CY for adults with refractory acute leukemias, with escalating doses of CLO (beginning10 mg/m²/day) and CY (beginning 400 mg/m²/day) days 1–3 and 8–10. The first CY dose was split between Day 0 and Day 1 at 200 mg/m²/dose to test the hypothesis that CLO could increase CY-induced DNA damage in leukemic blasts. Peripheral blood blasts were obtained prior to CY (Pre), at end Day 0 CY 2 hr infusion (Day 0 EOI), 2 hrs after Day 0 CY EOI (Day 0 Post), and Day 1 2 hrs after CLO-CY EOI (Day 1 Post). DNA damage was measured serially by flow cytometry of cells stained with antibody to phosphorylated histone H2AX (γH2AX). Apoptosis was measured at each time point by propidium iodide (PI) staining as sub2N DNA. To date, we have studied peripheral blood blasts from 9 patients at 2 CLO doses (10 and 20 mg/m²/day) and CY total daily dose 400 mg/m². γH2AX was detected in a median 1.32% of Pre cells (range 0–6.1). Relative to Pre, there was median 2-fold increase in γH2AX (range 0.3–6) in Day 0 EOI cells with less damage detected in Day 0 Post cells (median 1.1, range 0–3.3). Day 1 Post cells evinced median 6.3-fold increase in γH2AX staining (range 1.5–38) relative to Pre, with 6/9 patients have a >2-fold increase, and median 9.2-fold increase (range 0.9–38) relative to Day 0 Post with >2-fold increase in 7/9 patients. Relative to Pre cells, there was no increase in the fraction of cells exhibiting sub2N DNA (median 0.9, range 0.34–22, >1.2-fold increase in 4/9) in Day 0 Post samples. In contrast, Day 1 Post samples had median 1.4-fold (range 0.3–61) increase in cells with sub2N DNA, with 7/9 patients having >1.2 fold increases. In this small number of patients, there was no clear relationship between CLO dose and magnitude of DNA damage or apoptosis induced in Day 1 Post cells. In 5 patients with complete tumor clearance in Day 14 marrow, there was median 11-fold increase (range 0.3–62) in γH2AX staining in Day 1 Post cells relative to Pre cells and median 5.7-fold increase (range 0.9–33) over Day 0 Post cells. In 3 patients who had residual leukemia present on Day 14, the median increase in γH2AX over Pre was 1.7-fold (1.5,1.7,38). WBC nadir was 0 (range 0–50) occurred on median day 12 (range 10–20). WBC recovery to >500/mm³ occurred in 7 patients on median day 32 (range 27–75), 1 patient with refractory AML died Day 14 of multi-organ failure, and 1 patient with refractory ALL post autologous transplant died on day 100 in marrow aplasia. Two patients with refractory AML achieved responses: 1 CR, 1 PR. Interestingly, the patient who achieved CR had the highest Pre γH2AX and Day 1 Post levels (38.6%). In summary, our in vivo correlative data suggest that DNA damage that occurs following CY, as measured by γH2AX, is augmented by prior CLO administration and correlates with tumor clearance by Day 14. With accrual and study of additional patients, we should be able to determine any relationship between magnitude of increase in DNA damage and clinical outcomes.