Abstract 214

BAALC & ERG are aberrantly expressed in younger (<60 y) adult CN-AML patients (pts), where expression levels of these genes are also associated with clinical outcome. Whether aberrant BAALC & ERG expression also occurs in older (≥60 y) CN-AML pts is unknown. To assess the BAALC & ERG expression levels in older CN-AML & their impact on outcome we studied pts ≥60 y (median age, 68 y; range 60–83) enrolled on cytarabine/daunorubicin-based protocols [CALGB 10201, 9720, 9420, 8923, 8525], with diagnostic blood samples available for quantitative RT-PCR analysis (n=158), that were also characterized for other molecular prognosticators (FLT3-ITD, FLT3-TKD, NPM1 & WT1 mutations). BAALC & ERG expression values were normalized to an internal control (ABL1), & the median gene expression value was used to define high & low expressers for BAALC & ERG. Gene (GEP) & microRNA (MEP) expression profiling were done using, respectively, Affymetrix U133 plus 2.0 & OSU CCC v4.0 arrays. At diagnosis, lower BAALC expression was associated with mutated NPM1 (P<.001) & low ERG expression (P<.001). Lower BAALC expressers had a higher complete remission rate (CR; 86% v 54%, P<.001) & longer disease-free (DFS; P=.006; 3y rates 19% v 12%) & overall survival (OS; P<.001; 3y 29% v 10%) than higher expressers. Lower ERG expression was associated with lower WBC (P=.005), % marrow (BM; P=.001) & blood (P<.001) blasts & absent FLT3-ITD (P<.001) & low BAALC expression (P<.001). Lower ERG expression also associated with longer DFS (P=.001; 3y 18% v 14%) & OS (P<.001; 3y 24% v 15%). In multivariable models (Table 1), low BAALC expression independently associated with CR & longer DFS. BAALC & ERG expression were the only factors associated with OS. Comparison of age-groups (60-69 y v ≥70 y; Table 2) showed BAALC expression had a stronger prognostic impact in ≥70 y pts; lower expression was associated with higher CR rates & longer DFS & OS. ERG expression had instead stronger prognostic impact in 60-69 y pts (Table 2); lower expression was associated with longer DFS & OS. GEP (482 probes) & MEP (22 probes) differentiated low from high BAALC expressers. Low BAALC expressers had down-regulation of stem cell markers (CD34, CD133) & unfavorable outcome predictors (HGF, MN1, CD200), & up-regulation of HOX genes & miR-10a & miR-10b. GEP (1554 probes) & MEP (11 probes) differentiating low from high ERG expressers showed low ERG expressers had down-regulation of DNMT3B & up-regulation of topoisomerase 1 (TOP1), which is associated with enhanced chemotherapy sensitivity. Among up-regulated microRNAs in low ERG expressers was miR-208a, which is predicted to target ERG. In conclusion, lower expression of both BAALC & ERG associated with better outcome in older CN-AML pts even in the context of other established prognostic molecular markers, but have different impact on age-groups. GEP & MEP provided novel information that may elucidate how differential expression levels of these genes contribute to leukemogenesis & aid in developing novel risk-adapted therapies.

Table 1:

Multivariable analyses for CR, DFS & OS

Variables in the final modelCR
DFS
OS
OR*PHR†PHR†P
BAALC, low v high (median cut) 4.32 <.001 0.64 .03 0.49 <.001‡ 
ERG, low v high (median cut) - - - - 0.69 .03 
WBC, each 50 unit increase 0.68 .02 - - - - 
NPM1, mutated v wild-type 2.34 .04 - - - - 
FLT3-ITD, any ITD v no ITD - - 1.79 <.001‡ - - 
Age, each 10 year increase - - 0.65 .047 - - 
Variables in the final modelCR
DFS
OS
OR*PHR†PHR†P
BAALC, low v high (median cut) 4.32 <.001 0.64 .03 0.49 <.001‡ 
ERG, low v high (median cut) - - - - 0.69 .03 
WBC, each 50 unit increase 0.68 .02 - - - - 
NPM1, mutated v wild-type 2.34 .04 - - - - 
FLT3-ITD, any ITD v no ITD - - 1.79 <.001‡ - - 
Age, each 10 year increase - - 0.65 .047 - - 
*

OR, odd ratio, >1 (<1) mean higher (lower) CR rate for the 1st category listed for the dichotomous variables & for the higher values of the continuous variables.

†

HR, hazard ratio, <1 (>1) indicate lower (higher) risk for an event for the first category listed for the dichotomous variables & for the higher values of the continuous variables. Variables considered in the models were those significant at α=0.20 in univariable analyses.

‡

Variable did not meet the proportional hazards assumption, a covariate was used to account for time dependence.

View Large
Table 2:

Age-Group comparison for CR, DFS & OS

Clinical endpointBAALC expressionERG expression
60-69 y Pts (n=86)≥70 y Pts (n=72)60-69 y Pts (n=86)≥70 y Pts (n=72)
LowHighPLowHighPLowHighPLowHighP
% achieving CR 84 71 .19 88 37 <.001 80 76 .80 71 53 .15 
DFS, % in CR at 3 y 16 10 .10 23 14 .03 19 6 <.001 17 25 .48 
OS, % alive at 3 y 24 12 .01 35 8 <.001 24 12 .006 23 18 .03 
Clinical endpointBAALC expressionERG expression
60-69 y Pts (n=86)≥70 y Pts (n=72)60-69 y Pts (n=86)≥70 y Pts (n=72)
LowHighPLowHighPLowHighPLowHighP
% achieving CR 84 71 .19 88 37 <.001 80 76 .80 71 53 .15 
DFS, % in CR at 3 y 16 10 .10 23 14 .03 19 6 <.001 17 25 .48 
OS, % alive at 3 y 24 12 .01 35 8 <.001 24 12 .006 23 18 .03 

The age-group clinical endpoints significantly associated with BAALC or ERG expression levels are denoted by bold type.

View Large
Disclosures:

No relevant conflicts of interest to declare.

Topics:
biological markers, brachial plexus neuritis, cancer and leukemia group b, cd34 antigens, chemosensitivity, complete remission, cytarabine, daunorubicin, dna topoisomerases, gene expression

Author notes

*

Asterisk with author names denotes non-ASH members.

© 2009 by The American Society of Hematology
2009
Sign in via your Institution