BAALC Expression Is a Feasible Marker for Risk Stratification and Detection of Minimal Residual Disease in Cytogenetically Normal Acute Myeloid Leukemia

During the last years it has been shown that PCR-based assessment of minimal residual disease (MRD) is of great importance for risk stratification and early detection of relapse in acute myeloid leukemia (AML). Most frequently, monitoring of MRD was restricted to patients carrying specific genetic markers such as fusion genes and gene mutations. However, many patients lack such markers amenable to sensitive detection by PCR. For this reason, it is still crucial to identify molecular targets that are applicable for MRD assessment in patients with AML. High brain and acute leukemia, cytoplasmic (BAALC) expression defines an important risk factor in cytogenetically normal AML (CN-AML) and has been proposed to represent a clinically useful biomarker to monitor response to therapy. A significant correlation of BAALC expression levels with the mutation ratio of internal tandem duplications in FLT3(FLT3-ITD), partial tandem duplications in MLL (MLL-PTD) and NPM1 mutations has been shown previously (Weber et al., BCJ 2014). The objective of this study was to further validate the applicability of BAALC expression as a marker for detection of MRD.

Patients and Methods

We investigated BAALC mRNA expression levels in 550 diagnostic and follow-up samples from 116 de novo CN-AML patients showing high BAALC expression at initial diagnosis. BAALC mRNA expression was assessed by quantitative real-time PCR and normalized against the expression of the control gene ABL1. To identify high BAALCexpressers the previously defined cutoff ratio of 33.1% BAALC/ABL1 (Weber et al., BCJ 2014) was used. The cohort was composed of 55 females (47%) and 61 males (53%). Age ranged from 18 to 85 years (median: 55). The median number of follow-up samples per patient was 3 (range: 1-21) with a median follow-up time of 32 months (range: 16-48 months). All patients included in this study were treated according to standard AML protocols in curative intent. In addition to BAALC expression, a full molecular characterization for the presence of FLT3-ITD (42/116, 36%), MLL-PTD (10/116, 9%) and mutations in NPM1 (38/116, 33%) was performed.

Results

At diagnosis, BAALC expression ranged from 33.2 to 5655.7% BAALC/ABL1 with a median of 163.9%. Of the 116 patients, 62 patients (53%) harbored at least one of the previously established MRD markers (FLT3-ITD, MLL-PTD and NPM1), while 54 patients (47%) lacked these MRD targets. To validate stability of BAALC expression we analyzed 25 patients with paired diagnostic and relapsed samples available. The respective BAALC expression levels at both time points showed good correlation (Mean % BAALC/ABL1 of diagnosis vs. relapse: 343.1 vs. 297.4, r=0.642, p=0.001). Three of the 25 patients relapsed after allogeneic stem cell transplantation. Also in these patients BAALCexpression levels were similar between diagnosis and relapse (Mean %BAALC/ABL1: 686 vs. 684), supporting the use of BAALC expression for follow-up evaluation. Moreover, in 5 cases with cytomorphologic complete remission during follow-up a molecular relapse based on elevated BAALC expression levels (32.4–315.2% BAALC/ABL1) could be detected 21-42 days before cytomorphologic relapse. To investigate the predictive value of BAALC expression levels during follow-up, BAALC expression was analyzed in 58 patients after the second course of induction chemotherapy, but before start of consolidation chemotherapy. To separate low from high BAALC expression in this follow-up analysis, the cutoff established for diagnostic samples (33.1% BAALC/ABL1) was used. According to this cutoff, 29% (17/58) of the patients had high BAALC expression levels, while 71% (41/58) exhibited low BAALCexpression levels. High BAALC expression was associated with a shorter EFS (median: 5 month vs. not reached, 1-year-EFS: 61% vs. 21%, p<0.001)

Conclusion

This data validates the applicability of BAALC expression as a target for MRD monitoring in CN-AML. A significant number of CN-AML would benefit from molecular monitoring using a quantitative BAALC expression assay aimed at following disease course and early detection of relapse.