preheader BJH 1

header website

Prospective H102 study: leukemic stem cell frequency combined with MRD is an important biomarker to predict relapse in acute myeloid leukemia

Despite the use of risk algorithms, outcome in acute myeloid leukemia (AML) patients is still difficult to predict. Even in good risk patients relapses occur. Thus, further refinement of currently used risk classifications is needed. This study shows that leukemic stem cell frequency combined with MRD performs well in the identification of very poor risk patients

Minimal residual disease (MRD) is a well-known risk factor and the prognostic impact of MRD was shown for patients independent of risk groups. Nowadays, prospective studies are designed in which therapy is adapted based on MRD positivity or negativity. However, relapses occur in a substantial proportion of MRD negative patients. Previous retrospective studies have shown that the leukemic stem cell (LSC) frequency harbors important prognostic information, even for MRD negative patients. In this study, presented during the EHA, data of the HOVON/SAKK H102 trial were used to prospectively define, the leukemic CD34+CD38- stem cell frequencies and MRD frequencies to investigate impact on patient outcome.

In 242 patients who achieved morphologic complete remission, both LSC and MRD data after two cycles of chemotherapy treatment were available. MRD-positivity was defined as a percentage of MRD-positive cells above 0.1% (as compared to total amount of white blood cells) and LSC-positivity was defined as a CD34+CD38-LSC percentage above 0.0000% (LSC cut-off 0.0000%; thus no CD34+CD38-LSC events measured).

Cumulative incidence of relapse (CIR) and overall survival (OS) data were investigated for 4 different MRD/LSC groups: 1) MRD negative + LSC negative patients (n=136), 2) MRD positive + LSC negative patients (n=28), 3) MRD negative + LSC positive patients (n=58), and 4) MRD positive + LSC positive patients (n=20). 3-year CIR for the four above-defined groups was 35%, 43% , 53% , and 100%, respectively. Similar results were found for OS: 3-year OS was 66%, 68%, 53%, and 0%, respectively.

Subsequently, the impact of MRD/LSC status in the good, intermediate, poor and very poor risk group (according to HOVON) were investigated. Although patient numbers were sometimes small, results show that MRD positive + LSC positive AML patients in all different risk categories have a very poor prognosis. Moreover, multivariate analyses, containing all well-known risk factors including risk group and post remission treatment, showed that MRD positive + LSC positive patients have a significantly worse cumulative incidence of relapse (hazard ratio [HR] 5.89; 95% CI 3.32-10.47) and overall survival (HR 3.62; 95% CI 1.86-7.04) as compared to the MRD negative + LSC negative patients.

Overall, the investigators concluded that these prospective results demonstrate that CD34+CD38-LSC frequency has important additional value in MRD assessment and that it especially enables to identify very poor risk patients in all different currently used risk categories. These data urge to include both MRD and LSC in future AML risk classification to better inform post-remission treatment.


W. Zeijlemaker, R. Meijer, A. Kelder, et al. Leukemic stem cell frequency combined with NRD is an important biomarker to predict relapse in acute myeloid leukemia. Results from a prospective H102 study. EHA 2017, oral presentation, abstract S113.


Speaker Wendelien Zeijlemaker


Wendelien Zeijlemaker, MD, Noordwest Ziekenhuisgroep, Alkmaar, The Netherlands


See: Keyslides


Back to Top