Drug resistance in acute myeloid leukemia : Pharmacokinetic and in vitro studies

Chemotherapy has improved clinical outcome in adult acute myeloid leukemia (AML) but still the majority of the patients die of their disease. One important explanation is transport mediated drug resistance resulting in refractory or relapsing disease. P-glycoprotein (Pgp) is located in the membrane and can extrude cytostatic drugs. This will cause a reduction of intracellular concentrations of the drug and reduce the effect on the leukemic cells. Down regulation of the enzyme topoisomerase II alpha (topo II alpha) is another resistance mechanism. The aim of the first part of this thesis was to study effects of interaction with transport mediated resistance. In a phase I-II pharmacokinetic study 22 de novo AML patients were included and received a high single dose of mitoxantrone (30 or 40 mg/m2) in combination with ara-C. In the leukemic cells we found a high accumulation of mitoxantrone which, in contrast to plasma, remained stable during the 48 hours studied. Compared with previous results with mitoxantrone 12 mg /m2 the area under the curve (AUC) for intracellular concentrations was increased by 150% (30 mg/m2) and 260% (40 mg/m2) respectively. The toxicity was acceptable but a prolonged duration of neutropenia was noted. Several studies have shown that Pgp can be blocked and in the next study 10 AML patients were included and given a continuos infusion of daunorubicin in combination with the Pgp modulator PSC 833. The intracellular in vivo concentrations of daunorubicin in Pgp positive leukemic cells increased as well as the ratio of the AUC for daunorubicin in leukemic cells to the AUC in plasma suggesting that the mechanism was a specific inhibition of PgP by the modulator. Another aspect of drug resistance and cytostatic treatment is the potential interaction of Pgp, and drugs used in the supportive care of AML patients. 21 different drugs were tested in a Pgp expressing leukemia cell line and the change in intracellular rhodamine 123 was measured. However, in this model we found no significant interactions of the tested drugs on Pgp. In the second part of the thesis we studied two different chemosensitivity assays and markers for drug resistance. In the bioluminiscence ATP chemosensitivity assay cell death is measured as reduced ATP levels. 83 samples from 77 AML patients were included and the in vitro effect of six different cytostatic drugs were tested. We found that in vitro sensitivity to daunorubicin was associated with complete remission and prolonged disease free survival. One problem with in vitro assays is the possible contamination of non malignant cells which can interfere with the results. Therefore we established a new assay were selected myeloid cells were analysed using flow cytometry. 63 samples from 60 AML patients in different stages of their disease were tested. The method was feasible and we found correlations to clinical parameters and Pgp expression. Topoisomerase II alpha is a well known target enzyme for many cytostatic drugs. In this final investigation the expression of the enzyme was studied in different phases of the cell cycle in 25 acute leukemia patients. In contrast to normal cells we found that topo II alpha was expressed in GO/G1 phase. A low expression of topo II alpha in the GO/G1 phase seemed to be associated with resistant disease, which suggests that topo II alpha expression in the different cell cycle phases may have a predictive value. In conclusion a high single dose or the inhibition of Pgp increased the intracellular concentrations in vivo of mitoxantrone and daunorubicin respectively. Drugs used in the supportive care of AML patients did not interact with Pgp in vitro. In vitro sensitivity to daunorubicin was associated with both short and long term outcome using the bioluminiscence ATP assay. In vitro chemosensitivity testing of selected myeloid cells and low expression of topo II alpha in the GO/G1 phase, using flow cytometry techniques, correlated to clinical outcome but further studies are needed to evaluate the predictive value