With current therapy regimens over 75% of patients with de novo acute promyelocytic leukemia (APL) can be cured. Approaches to further improve patient outcome by stratifying patients at the time of initial diagnosis according to their individual risk and to adjust therapy accordingly have been based on clinical features only. Molecular markers have not been established for risk stratification as yet. Recently, we have shown that high expression levels of the genes brain and acute leukemia, cytoplasmic (BAALC) and ets related gene (ERG) are associated with inferior outcome in APL patients. In addition, data indicate that aberrant expression of the gene Wilms’ tumor 1 (WT1) is a negative prognostic factor with regard to overall survival (OS) after complete remission (CR) and relapse free survival (RFS) in APL. In this study we evaluated the prognostic relevance of a combined score integrating the expression levels of the above mentioned genes to further improve risk stratification in APL patients.


Expression levels of BAALC, ERG and WT1 of 62 patients with newly diagnosed APL were retrospectively analyzed in bone marrow mononuclear cells using multiplex reverse transcriptase quantitative real-time PCR (qRT-PCR). Median age of patients was 47 years (range: 19 to 82y). All patients gave informed consent. Patients were diagnosed and treated in the German AML Cooperative Group (AMLCG) study with a treatment of simultaneous ATRA and double induction chemotherapy including high-dose ara-C, consolidation and maintenance chemotherapy. The following gene expression levels were identified as negative risk factors in preceding studies: BAALC expression ≥25th percentile (BAALChigh), ERG expression >75th percentile (ERGhigh) and WT1 expression ≤25th percentile or ≥75th percentile (WT1low/high). A risk score was developed as follows: for the presence of one of the mentioned risk factors one scoring point was assigned to a respective patient, i.e. a maximum of 3 points (one point for BAALChigh, ERGhigh and WT1low/high, respectively) and a minimum of 0 points (i.e. presenting with none of the aforementioned risk factors) could be allocated to one patient. Accordingly, patients were divided into four risk groups: 7 patients scored 0 points (= low risk), 27 patients scored 1 point (= intermediate 1 risk), 19 patients scored 2 points (= intermediate 2 risk) and 9 patients scored 3 points (= high risk). Subsequently, OS, RFS and relapse free interval (RFI) were calculated using the Kaplan-Meier method and a log-rank test was used to compare differences between the four risk groups (p<0.05).


The integrative risk score divided patients into four groups with significantly different outcome. The low risk group showed a RFS of 100% at 10 years of follow-up compared to the intermediate 1 risk group with 81%, the intermediate 2 risk group with 58% and the high risk group with a RFS of 42% only (median survival: 4.6y) (p=0.02). In accordance, the RFI differed significantly between the four groups: low risk 100%, intermediate 1 risk 100%, intermediate 2 risk 89% and high risk 71% (p=0.049). There was no statistically significant difference between the 4 groups with regard to OS in the entire patient cohort. However, there was a clear trend towards a difference in OS in patients who achieved a CR after induction therapy: low risk 100%, intermediate 1 risk 81%, intermediate 2 risk 68% and high risk 53% survival at 10 years of follow-up (p=0.09).


Integration of expression levels of the genes BAALC, ERG and WT1 into a scoring system identifies 4 risk groups with significantly different outcome with regard to RFS and RFI. It might be a promising approach to guide therapeutic decisions in patients with APL. However, multivariate analyses and validation of these data in an independent patient cohort is warranted.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.