Older AML patients have low complete remission (CR) rates and short disease-free (DFS) and overall (OS) survival; unfavorable karyotypes, prior myelodysplastic syndrome (MDS) and multidrug resistance (MDR) mediated by the cellular drug efflux pump Pgp have been implicated. The significance of other MDR proteins expressed in AML cells, including the cellular drug efflux pumps MRP-1 and BCRP and the major vault protein LRP, is unknown. We correlated MDR parameters with response in 170 previously untreated AML patients ≥60 years old receiving ADE (cytarabine 100 mg/m2, daunorubicin 60 mg/m2 and etoposide 100 mg/m2 induction for 7+3+3 days and consolidation for 5+2+2 days) (CALGB 9720). Pretreatment blasts were studied by flow cytometry for expression of Pgp, MRP-1, BCRP and LRP with the MRK-16, MRPm6, BXP-21 and LRP56 antibodies; for efflux of mitoxantrone, a substrate for Pgp, MRP-1 and BCRP; and for modulation of mitoxantrone efflux by the Pgp, MRP-1 and BCRP modulators PSC-833, MK-571 and fumitremorgin C (FTC) and by cyclosporine A (CsA), which modulates all three proteins (CALGB 9760). Karyotypes of 141 patients (83%) were centrally reviewed and classified as favorable, intermediate or adverse with respect to likelihood of CR, cumulative incidence of relapse (CIR) and OS (
). Ages were 60 to 89 years (median=71). 135 patients (79%) had de novo AML, 35 (21%) prior MDS and 9 (5%) therapy-related AML. Karyotypes were favorable for CR in 2 (1%), intermediate in 105 (74%), adverse in 23 (16%) and unclassified in 11; favorable for CIR in 2 (1%), intermediate in 92 (65%), adverse in 31 (22%) and unclassified in 16; and favorable for OS in 2 (1%), intermediate in 88 (62%), adverse in 40 (28%) and unclassified in 11. CR rate was 46% and median DFS and OS 6.2 (95% CI:5.0–8.8) and 8.1 (95% CI:5.3–9.5) months. Karyotype risk groups (favorable/intermediate vs. adverse) were associated with CR (n=130; p=0.0105) but not with DFS or OS, and age and prior MDS were not associated with CR, DFS or OS. Pgp, MRP-1, BCRP and LRP were expressed in 31, 39, 50 and 52% of samples. Mitoxantrone efflux occurred in 68% and was inhibited by PSC-833, MK-571, FTC and CsA in 30, 5, 7 and 35%. No MDR parameter correlated with age, prior MDS or karyotype risk groups. CR rate, DFS and OS did not differ in patients with (n=115) vs without efflux, but patients with, vs without, PSC-833 modulation of efflux, indicating Pgp function, were less likely to achieve CR (24% vs. 55%; p=0.0022) and had shorter OS (p=0.0066). In multivariate analysis adjusting for karyotype risk, patients (n=88) with PSC-833 efflux modulation remained less likely to achieve CR (odds ratio=0.21; 95% CI:0.07–0.64). Patients with efflux with, vs without, CsA modulation were also less likely to achieve CR (25% vs. 57%; p=0.0015), even adjusting for karyotype risk (odds ratio=0.29; 95% CI:0.11–0.78), likely due to effect on Pgp. MK-571 and FTC modulation did not predict CR or OS, and no MDR parameter predicted DFS. Thus Pgp function is the only MDR parameter associated with outcome, and strongly predicts CR induction failure. These data support development of Pgp modulators and modulation regimens that can be safely administered to older patients with AML with Pgp function.
This feature is available to Subscribers Only
Sign In or Create an Account Close Modal