Contrasting in vitro effects for the combination of fludarabine, cytosine arabinoside (Ara-C) and granulocyte colony-stimulating factor (FLAG) compared with daunorubicin and Ara-C in P-glycoprotein-positive and P-glycoprotein-negative acute myeloblastic leukaemia

It has been suggested that the FLAG remission induction regimen comprising fludarabine (F-ara), cytosine arabinoside (Ara-C) and granulocyte colony-st imulating factor (G-CSF) may be capable of overcoming P-glycoprotein (P-gp) -related multidrug resistance (MDR) in patients with acute myeloblastic leu kaemia (AML). We have investigated the in vitro response of P-gp-positive a nd -negative AML clones to FLAG and compared this with their response to tr eatment with Ara-C and daunorubicin (DNR). Twenty-four cryopreserved sample s from patients with AML were studied using a flow cytometric technique for the enumeration of viable (7-amino actinomycin D negative) cells. Samples consisted of 12 P-gp-positive and 12 P-gp-negative cases, as measured by th e MRK16 antibody. The results were analysed by calculating the comparative drug resistance (CDR), i.e. the percentage cell death caused by Ara-C + DNR subtracted from the percentage cell death, caused by FLAG after 48 h incub ation in suspension culture. P-gp-positive clones were shown to have a sign ificantly higher CDR than P-gp-negative clones (P = 0.001). Furthermore, a significant positive correlation (r(2) = 0.40, P < 0.01) was found between P-gp protein expression and CDR. However, P-gp function, measured using cyc losporin modulation of rhodamine 123 (R123) uptake, was not associated with the CDR, demonstrating that there are other properties of P-gp, besides it s role in drug efflux, that modulate the responsiveness of AML blasts to ch emotherapy. These results are consistent with a potential benefit for FLAG in P-gp-positive AML, but not P-gp-negative AML, compared with standard ant hracycline and Ara-C therapy.