Abstract

Abstract 2553

Background:

High expression of the brain and acute leukemia, cytoplasmic (BAALC) gene is associated with inferior outcomes in adult AML patients with normal cytogenetics. Studies regarding the significance of BAALC expression in pediatric AML are limited. We studied BAALC expression in a cohort of children with de novo AML to evaluate its clinical significance in pediatric AML, particularly in patients with standard-risk (SR) disease.

Methods:

BAALC mRNA expression was measured via qRT-PCR in diagnostic specimens obtained from 206 patients enrolled on COG AAML03P1. Expression levels were normalized against pooled normal blood controls (PNB). To evaluate the impact of BAALC expression values on clinical outcome, samples were dichotomized at the median expression level (14.9) into a BAALClow (N= 103, median expression 2.9; range 0.0089–14.897) and BAALChigh (N=103; median expression 65; range 15.064–369.53) subset. Baseline clinical features were compared for BAALClow vs. BAALChigh patients and categorical variables compared using the chi-square test and Fisher exact test when data were sparse. The Kaplan-Meier method was used to estimate OS, EFS and DFS for the 2 groups. Estimates of RR were obtained using methods that accounted for competing events. Similar methods were used for sub-analysis of n=99 SR patients.

Results:

Compared to expression in PNB, BAALC expression varied over 5 log fold (range 0.0089–369.53); the median BAALC expression was 14.98. There were no differences between BAALClow and BAALChigh patients with respect to disease presentation (WBC, BM blast%, Hb, platelet count) though patients with BAALChigh expression were older (median age 12.2 vs. 7.8; years; p=0.001). NPM1 mutations were more frequent in the BAALClow subset (12% vs 0%, p=0.002) however CEBPA and FLT3/ITD mutations were not associated with BAALC expression. Favorable cytogenetics were associated with BAALChigh expression [t(8;21): 1% vs. 23%, p=<0.001; inv (16) 5% vs. 26%;p=<0.001] whereas normal cytogenetics and 11q23abnormalities were more frequent in the BAALClow cohort [normal cytogenetics 32% vs 13%; p=0.002; 11q23: 33% vs. 6%; p=<0.001]. High-risk cytogenetic features were not associated with BAALC expression in limited analysis (n=6 patients). BAALC expression was also correlated with disease-risk, a classification derived from prognostic cytogenetic and molecular features. BAALChigh was more common in patients with low-risk disease (19% vs. 54%, p=<0.001) whereas BAALClow was more frequently seen in SR patients (69% vs 32% p=<0.001). There was no association between high-risk patients and BAALC expression in a limited number of samples analyzed (n=26). Induction response was independent of BAALC expression (77% vs. 82%, p=0.442) although minimal residual disease (MRD) at end of induction was marginally associated with high expression (20% vs. 35%, p=0.058). There were no differences in OS and EFS from study entry for BAALClow vs. BAALChigh patients (5 year OS: 64% vs. 63%, p=0.832; EFS 49% vs. 48%, p=0.868) Corresponding DFS and RR from CR were also comparable (DFS: 58% vs. 56%, p=0.608; RR 34% vs. 31%, p=0.735).

We further evaluated the association of BAALC expression with outcome in patients with SR disease. SR BAALClow and SR BAALChigh patients had similar disease presentation (WBC, BM blast %, platelet count, Hb) and were similar in age. Abnormalities of 11q23 were associated with SR BAALClow expression (51% vs. 24%; p=0.006) but other SR cytogenetic abnormalities including normal cytogenetics were not. Rates of induction CR were similar for SR BAALClow versus SR BAALChigh patients (77% vs. 66%, p=0.225) as were rates of end induction MRD (21% vs. 37%, p=0.178). Importantly, there were no differences in OS and EFS from study entry for SR BAALClow vs. SR BAALChigh patients (5 year OS: 56% vs. 59%, p=0.464; EFS 42% vs. 37%, p=0.899). Corresponding DFS and RR from CR were also comparable (DFS: 56% vs. 47%, p=0.756; RR 36% vs. 41%, p=0.990).

Conclusions:

In this study, elevated BAALC expression was not associated with clinical outcome, within our total study cohort and in those patients with SR disease. This suggests that BAALC expression may lack independent prognostic significance within pediatric AML, despite its association with specific disease characteristics. Validation of our findings in a larger pediatric cohort, and in all disease-risk groups, is ongoing within COG protocol AAML0531.

Disclosures:

Franklin:Amgen: Employment.

This icon denotes a clinically relevant abstract

Author notes

*

Asterisk with author names denotes non-ASH members.