Pre-Transplant Monocytic Myeloid-Derived Suppressor Cell Frequency Has No Prognostic Role for Outcome after Allogeneic Hematopoietic Cell Transplantation for Acute Myeloid Leukemia in Remission

INTRODUCTION: Emerging data indicate myeloid-derived suppressor cells (MDSCs) adversely impact outcomes of patients with a variety of malignancies. However, it is less clear how MDSCs affect acute myeloid leukemia (AML) disease biology or the response of AML patients to treatment. Here, we studied a cohort of people with AML undergoing hematopoietic cell transplantation (HCT) in first complete remission (CR1) to examine the relationship between the monocytic subtype of MDSCs (M-MDSCs), pre-transplant measurable residual disease (MRD) status, and outcome after HCT.

PATIENTS AND METHODS: Adults ≥18 years of age were included if they underwent first allogeneic HCT with myeloablative or nonmyeloablative conditioning for AML in CR1 from April 2006 until October 2014. Prospective MRD testing in bone marrow aspirates was performed routinely via 10-color multiparameter flow cytometry (MFC) as part of the pre-transplant work-up. Pre-transplant MFC data were used retrospectively to quantify M-MDSCs as the proportion of monocytes (identified by side scatter properties and CD14 positivity) with low or negative expression of HLA-DR. M-MDSC frequency was then expressed as a percentage of the total number of viable cells in the sample. Available combinations of antigens measured by MFC did not allow for the identification of other MDSC subsets (e.g. granulocytic MDSCs).

RESULTS: We identified 349 adults undergoing allogeneic HCT for AML in CR1 for whom pre-transplant MRD testing was performed and for whom follow-up data were available. Of these, 8 had to be excluded because of insufficient MFC events available for identification of M-MDSCs. In the remaining 341 patients, M-MDSC frequency ranged from <0.001% to 6.53%, with a median of 0.18% (25^th percentile: 0.06%; 75^th percentile: 0.40%). Patients with lower (i.e. <median) frequency of M-MDSCs were less likely male (p=0.01) and more likely had recovered neutrophil and platelet counts (p=0.001) than patients with higher (i.e. ≥median) frequency of M-MDSCs. There was no difference with regard to all other examined characteristics (age, WBC and cytogenetic risk at diagnosis, type of AML, median CR duration before HCT, donor type, and conditioning intensity). The proportion of MRD positive ("MRD^pos") patients was similar between those with lower vs. higher M-MDSC frequency (18.8 vs. 24.6%, p=0.20). Moreover, while MRD negative ("MRD^neg") patients (n=267) differed statistically significantly from MRD^pos patients (n=74) with regard to several characteristics such as more favorable disease risk and lower proportion of secondary AML, there was only a statistically non-significant trend toward lower M-MDSC frequency in MRD^neg patients (0.16% [range <0.001-5.61%] vs. 0.22% [range <0.001-6.53%], p=0.09).

The median follow-up time after HCT among all survivors was 5.8 (range, 2.0-11.4) years. The time was slightly longer for survivors with lower compared to those with higher M-MDSC frequency (6.1 [range 2.6-11.4] vs. 5.1 [range 2.0-11.3] years; p=0.0076). Stratified by the median M-MDSC frequency, the 3-year estimates of overall survival were similar for patients with lower and higher pre-HCT M-MDSCs (58.2% [95% confidence interval 50.4-65.2%] vs. 57.6% [49.8-64.7%]), as were 3-year estimates for relapse free survival (53.5% [45.7-60.7%] vs. 48.3% [40.6-55.6%]), 3-year estimates of the cumulative incidence of relapse (31.8% [24.9-38.8%] vs. 34.6% [27.6-41.8%]), and 3-year estimates of non-relapse mortality (14.7% [9.9-20.5%] vs. 17.0% [11.8-23.0%]). Analyses restricted to the subset of patients undergoing myeloablative (MA) conditioning or those undergoing MA conditioning in MRD^neg remission showed qualitatively similar results.

CONCLUSIONS: In the largest study to date examining the role of M-MDSCs in AML, we found no convincing evidence for a prognostic role of these cells for adults undergoing allografting in CR1.