Short course infusional idarubicin plus intermittent cytarabine and etoposide for refractory hematologic malignancies: clinical and preliminary pharmacological results.

BACKGROUND AND OBJECTIVE Idarubicin (IDA) is relatively immune to the multidrug resistance P-gp mechanism that is frequently expressed in recurrent and refractory hematologic malignancies. Owing to rapid metabolism in vivo, a continuous infusion (CI) of IDA might prolong exposure time to the parent drug rather than its more P-gp susceptible alcohol metabolite. For this reason we developed a brief retreatment schedule incorporating CI IDA in order to obtain clinical as well as preliminary pharmacological data in patients with refractory leukemias and lymphomas. DESIGN AND METHODS Eligible patients had either advanced-stage acute myeloid or lymphoid leukemias (AML, ALL) or high-grade non-Hodgkin's lymphomas (NHL) which failed curative-intent frontline or salvage regimens in use at our institution during the study period (July-October 1992). CI IDA 5 mg/m2/d was employed together with intermittent (every 8 hours) intermediate-dose cytarabine (500 mg/m2) and etoposide (200 mg/m2); all drugs were given for 2-4 days. A preliminary pharmacokinetic evaluation of CI IDA was carried out in three patients, including a comparison with bolus delivery in one. The in vitro effects of CI-type vs bolus-type IDA delivery in terms of intracellular IDA accumulation and related pro-apoptotic activity were assessed in P-gp- and P-gp+ human leukemic CEM cells by means of cytofluorimetry (IDA fluorescence intensity = FI, annexin V expression), with and without the addition of P-gp inhibitor cyclosporin A (CsA). RESULTS Complete (2) or partial (4) responses were achieved in a total of 12 patients (17% and 33%, respectively), despite prior treatments with anthracyclines (100% of cases) and cytarabine-etoposide (33% of cases). Hematological toxicity caused the duration of treatment to be reduced from 4 days to 2 days after the first 4 patients. The procedural death rate was 42% (5/12), which was probably related in part to the sum of adverse prognostic characteristics: median patient age 55 years, two-thirds of cases having previously failed second/third-line regimens. The pharmacokinetic study showed an increased plasma AUC value with CI IDA in one patient (2.9-fold increase vs bolus delivery) due to the prolonged presence of low IDA plasma levels (10-20 ng/mL vs 50 ng/mL), as seen in two other cases as well. On the other hand, the in vitro study did not prove to be in favor of CI IDA because the FI threshold (> 1500 units) associated with increased apoptosis of P-gp+ cells (> 10%) was achieved only with bolus-type IDA exposure (50 ng/mL for 30') plus CsA. INTERPRETATION AND CONCLUSIONS This short regimen demonstrated activity against end-stage leukemias and lymphomas and might prove to be more effective and less toxic in younger patients and in those with less advanced disease. In view of the results from plasma pharmacokinetics and in vitro intracellular IDA accumulation and apoptosis assays in lymphoblastic CEM cells, CI IDA 5 mg/m2/day may not represent a better therapeutic option than a rapid bolus injection, particularly in P-gp+ neoplasms. If obtaining an adequate intracellular drug concentration is the primary treatment goal, a higher CI IDA dosage, the addition of a P-gp down-regulator such as CsA and others, and in vivo study focusing on tumor samples from patients could all be helpful.