Background

Nucleoside analogs depend on cellular hENT1 expression for entry into cells and cytotoxic activity. Studies suggest low cellular hENT1 levels correlate with poor response to such chemotherapies in solid tumors, data on AML and MDS is scarce.

Aim

To examine hENT1 expression by multiparameter flow cytometry (MFC) in newly diagnosed AML and MDS and correlate results to morphologic, cytogenetic (CG) and molecular genetic (MG) findings. To examine hENT1 expression with respect to clinical outcome in AML patients (pts) treated with intensive cytarabine-based chemotherapy (CHT).

Methods

We studied pts with newly diagnosed AML (n=145) and MDS (n=96), 133/108 male/female, median age 67.3 (AML) and 73.3 years (MDS). CG was done in 130 AML and 86 MDS. Pts included 107 de novo AML, 9 t-AML, 29 s-AML; FAB: 9 M0, 27 M1, 50 M2, 9 M3, 21 M4, 8 M4eo, 7 M5, 14 not classified; by CG (MRC): 21 favorable, 75 intermediate, 34 adverse. 91 were de novo MDS, 5 t-MDS; 1 RARS, 17 RCMD-RS, 37 RCMD, 3 5q- syndrome, 3 RAEB-1, 5 RAEB-2, 1 CMML, 24 not classified; 2 IPSS-R very low, 55 IPSS-R low, 8 IPSS-R intermediate, 8 IPSS-R high, 13 IPSS-R very high. hENT1 expression was quantified by a novel four color intracellular staining assay using monoclonal antibodies against hENT1, CD45, CD64 and myeloperoxidase. Median fluorescence intensities (MFI) of hENT1 were determined in myeloid progenitors (MP), granulocytes (G) and monocytic cells (Mo) and correlated to hENT1 MFI in lymphocytes to derive hENT1 index (index).

Results

No correlation of index to age, gender, hemoglobin level or counts for blasts, WBC or platelets was detected.

In AML, we generally saw higher index by trend in the more favorable prognostic subgroups. M3/t(15;17)/PML-RARA+ displayed higher index in MP than non-M3 AML (4.24 vs 2.56, p<0.001). G index was lower in M0 (3.01) vs M1, M2, M4 and M4eo (5.66, 4.34, 5.35, 4.77; p=0.01, 0.028, 0.004, 0.043, respectively) and in M2 compared to M1 and M4 (4.34. vs 5.66 and 5.35, p=0.01 and 0.033, respectively). M2 showed lower MP index than M5 (2.42 vs 2.99, p=0.016). Considering CG, index in MP was higher in favorable vs intermediate and adverse pts (3.05 vs 2.58 and 2.53, p=0.034 and 0.023, respectively), Mo index was higher ín favorable vs adverse pts (3.17 vs 2.71, p=0.044). By MG, higher index in Mo and G was observed in RUNX1-RUNX1T1+ AML (4/83, 4.32 vs 3.04, p=0.01; 8.16 vs 4.60, p=0.002, respectively). Higher index for MP was found in FLT3-ITD mutated (mut) (18/111; 3.19 vs 2.62, p=0.012), CEPBA mut (4/26, 3.15 vs 2.35, p=0.004) and for Mo in NPM1 mut AML (23/104; 3.72 vs 2.84, p=0.02), whereas lower index for MP was found in RUNX1mut pts (13/65; 2.17 vs 2.59, p=0.031). De novo AML displayed higher MP index than s-AML (2.7 vs 2.28, p=0.008).

Using lowest quartile of index for MP (2.1185) as cut-off, AML pts in the MRC intermediate group treated with CHT (n=38) had inferior OS if MP index was below vs above this cut-off (OS at 6 months 63% vs. 95%, p=0.017, median follow up 4.6 months).

MDS showed lower Mo and MP index than AML (2.68 vs 2.96, p=0.021, 1.84 vs 2.65, p<0.001, respectively). By IPSS-R, significance was reached for higher index in Mo and MP in very low risk compared to low risk pts (3.39 vs 2.54, p=0.013 and 4.07 vs 1.78, p<0.001, respectively), MP in very low compared to intermediate and high risk pts (4.07 vs 1.95, p=0.004; 4.07 vs 1.76, p=0.002), and MP and G in very low vs very high risk pts (4.07 vs 1.71, p=0.005; 5.86 vs 3.85, p=0.001, respectively). IPSS-R intermediate vs poor and very poor showed lower G index (5.47 vs 3.59, p=0.018 and vs 3.85, p=0.034 respectively).

Conclusion

AML with genetic and molecular genetic good risk profile had higher hENT1 expression in MP, G and Mo, suggesting a causal mechanism for better response to CHT and better outcome. Consequently, AML with poor risk molecular genetics (RUNX1 mut) showed lower levels of hENT1 in MP. The detection of higher levels in FLT3-ITD mut pts is in line with reportedly good response to CHT, overall worse outcome being mostly due to early relapses. Strikingly, we saw differences in outcome in pts treated with CHT according to hENT1 expression with shorter OS in pts with low index for MP. Higher index in de novo AML than s-AML and MDS may be causal for better response to nucleoside-based CHT in de novo AML. Data for MDS may be interpreted accordingly, lower risk cases showing higher index in MP, G and Mo. Further analyses are needed to explore hENT1 expression in AML and MDS more comprehensively.

Disclosures:

Bellos:MLL Munich Leukemia Laboratory: Employment. Davis:Trillium Diagnostics, LLC: Equity Ownership. Culp:Trillium Diagnostics, LLC: Equity Ownership. Schnittger:MLL Munich Leukemia Laboratory: Employment, Equity Ownership. Haferlach:MLL Munich Leukemia Laboratory: Employment, Equity Ownership. Haferlach:MLL Munich Leukemia Laboratory: Employment, Equity Ownership. Kern:MLL Munich Leukemia Laboratory: Employment, Equity Ownership.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution