PSC-833 reverses multidrug resistance by P-glycoprotein at concentrations ≤ ≤ ≤ ≤1000 ng/ml. A phase I study of PSC-833 and doxorubicin was conducted to determine the maximum tolerated dose and to investigate pharmacokinetics. PSC-833 was intravenously infused as a 2-h loading dose (LD) and a subsequent 24-h continuous dose (CD). Doxorubicin was infused over 5 min, 1 h after the LD. The starting dose was 1 mg/kg for both LD and CD with 30 mg/m 2 doxorubicin; these dosages were increased to 2 and 10 mg/kg and 50 mg/m 2 , respectively. Thirty-one patients were treated. Nausea/ vomiting was controllable with granisetron and dexamethasone. Neutropenia and ataxia were dose limiting. Steady-state concentrations of PSC-833 > > > >1000 ng/ml were achieved at a 2 mg/kg LD and a 10 mg/kg CD. Ex-vivo bioassay revealed that activity in serum for reversing multidrug resistance was achieved in all patients; IC 50 of P-glycoprotein expressing 8226/Dox 6 in patients' serum was decreased from 5.9 to 1.3 µ µ µ µg/ml (P < < < <0.0001) by PSC-833 administration. Doxorubicin clearance was 24.3 ± ± ± ±13.7 (mean ± ± ± ±SD) liter/h/m 2 , which was lower than the 49.0 ± ± ± ±16.9 liter/h/m 2 without PSC-833 (P < < < <0.0001). The relationship between doxorubicin exposure and neutropenia did not differ between patients treated and not treated with PSC-833. The recommended phase II dose of PSC-833 was 2 and 10 mg/kg for LD and CD, respectively, which achieved a sufficient concentration in serum to reverse drug resistance, as confirmed by bioassay. The dose of doxorubicin should be reduced to 40 mg/m 2 , not because of the pharmacodynamic interaction between PSC-833 and doxorubicin affecting hematopoiesis, but because of pharmacokinetic interaction.

The multidrug resistance (MDR) phenotype, either intrinsic and/or acquired, is discussed in relation to several MDR-associated markers such as P-glycoprotein (P-gp) encoded by mdr1, multidrug-resistance-associated protein (MRP) encoded by MRP and lung-resistance-associated protein (LRP) encoded by LRP. Well-characterized in vitro models are required to elucidate the mechanisms of MDR. The aim of the present study is the establishment of a drug-resistant subline from human colorectal adenocarcinoma HCT-15 that intrinsically expresses moderate levels of P-gp, MRP and LRP. Three adriamycin-resistant sublines (HCT-15/ADM1, HCT-15/ADM2 and HCT-15/ADM2-2) were established by stepwise exposure in growth medium that was supplemented with 25-200 ng/ml adriamycin-resulting in a 2.2- to 7.8-fold increase in IC(50) values by using the XTT assay. They were cross-resistant to MDR-related drugs, epirubicin, mitoxantrone, vincristine, etoposide and taxol, but not the MDR-unrelated drug, mytomycin C. The resistance to adriamycin was confirmed in vivo by a lack of sensitivity in athymic nude mice. Gene expression data for mdr1/P-gp, MRP/MRP and LRP/LRP on both mRNA and protein levels demonstrated that the molecules contributing to MDR in resistant sublines are mainly P-gp and partially MRP. The newly established adriamycin-resistant sublines of HCT-15 will provide clinically relevant tools to investigate how to overcome drug resistance and elucidate possible mechanisms of acquired MDR in human colon cancer.

]vincristine) accumulation in these cells. Furthermore, a good correlation was detected between P-glycoprotein levels and the MDR-reversing effect of valspodar. In contrast, the effects of cyclosporin A could not be linked to P-glycoprotein levels in the MDR cells. In addition, the intracellular accumulation of valspodar was found to be 3-6 fold higher than that of cyclosporin A in four sublines and verapamil, an inhibitor of P-glycoprotein-mediated transport, enhanced the accumulation of cyclosporin A, but not valspodar. These results suggested that valspodar accumulation is not actively regulated by the P-glycoprotein-mediated efflux system.

P-glycoprotein is an efflux pump for a broad spectrum of hydrophobic agents. We found that bioactive peptides including somatostatin and substance P inhibit ATPdependent vincristine binding to P-glycoprotein-overexpressing K562/ADM membrane vesicles. Some of these bioactive peptides including somatostatin stimulate basal ATPase activity of P-glycoprotein; in contrast, other peptides including substance P inhibit it. The K562/ADM membrane vesicles showed an ATPdependent, osmotically sensitive uptake of somatostatin and substance P, which was inhibited by valspodar, an inhibitor of Pglycoprotein. These findings suggested that certain bioactive peptides such as somatostatin and substance P directly interact with human P-glycoprotein as endogenous substrates for Pglycoprotein-mediated transport.