Prognostic Significance of Cryptic Genomic Aberrations In AML with Normal Karyotype

Abstract 2729

Acute myeloid leukaemia with a normal karyotype (AML-NK) is a heterogeneous clonal disorder of haematopoietic progenitor cells with an intermediate prognosis. Acquired mutations of FLT3, NPM1 and CEBPA as well as gene expression profiles have recently been used as biomarkers to further stratify prognostic subgroups, especially in AML-NK.

We have previously shown that cryptic genomic aberrations in MDS cases are prognostically significant and regions of uniparental disomy (UPD), detectable by SNP array karyotyping, can harbour mutations in specific gene(s). We analysed 83 AML patients (pts), median age 52 (range 8–78) with normal cytogenetics, excluding pts who died within 30 days of entry, who had been treated on one of the UK MRC/NCRI AML clinical trial protocols, using high resolution 250K SNP microarrays to seek genomic aberrations. In addition, we sequenced the ‘mutation hotspot’ for IDH1/2 as well as screened for the presence of FLT3 ITD and NPM1 mutations. Genomic aberrations and mutational lesions were correlated with outcome data.

To circumvent the lack of constitutional DNA, all aberrations overlapping by >50% with copy number variants from the Database of Genomic Variants as well variations observed in a cohort of 91 healthy subjects were excluded from further analysis. We demonstrate the presence of 69 cryptic genomic aberrations in 43 pts consisting of 12 deletions (del) (8 pts), 10 gains (10 pts) and 47 regions of UPD (35 pts). The median size of del was 1.6Mb (0.18Mb-3.4Mb) with chromosome 4 (q24) and 17 (q11.2 and q22.1) being the most affected. One pt had micro del on chromosome 12p13.2-p12.3 and 21q22.11-q22.12 affecting the ETV6 and AML1 genes respectively, suggestive of a t(12:21) aberration. Gains (median 0.37Mb (0.26–0.71Mb)) were identified predominantly on chromosome 8 (4 pts) with a common amplified region, 8q21.11. The median size of UPD's was 47Mb (2.3Mb-158Mb). Chromosome 13 was the most frequently affected with UPD of the entire chromosome observed in 17 pts. Interestingly, one pt had an interstitial region of UPD on q21.32-q21.33 (4.8Mb). Additional regions affected by UPD included chromosomes 6, 2, 5 in 5 pts and chromosome 11 in 4 pts.

NPM1 mutations occurred in 55 pts (65%) and FLT3 ITD's occurred in 44 pts (53%). Of 39 pts with wild type FLT3, 22 had NPM1 mutations; of 28 pts with wild type NPM1, 11 had FLT3 ITD mutations. Conversely of the 44 pts with mutant FLT3, 33 had NPM1 mutations. Six pts had mutations of FLT3, NPM1 and IDH. Of note, only 2 of 17 pts with UPD of chromosome 13 did not have FLT3 mutations, suggesting the presence of additional pathological genes in this region. IDH1 (R130H) and IDH2 (R140Q and R172K) mutations occurred in 3 and 16 pts, respectively and were mutually exclusive. Interestingly, the R172K mutation occurred in only 3 pts. Of the 19 pts with IDH1/2 mutations, 9 and 12 pts respectively, had co-existing FLT3 ITD and NPM1 mutations.

The median duration of follow-up from time of diagnosis of the study group was 29 months (range 2–160). There were no significant correlations between the presence of genomic aberrations and age, sex, WBC, secondary disease or IDH1/2 mutations; however, a significant difference was observed in the frequency of UPD in pts with mutant or wild type FLT3 (p=0.01) and NPM1 (p=0.008). A difference in overall survival (OS) was seen on comparison of pts with or without genomic aberrations, (5 year OS 10% vs 46%, hazard ratio (HR) 1.86 (1.01–3.44), p=0.05): significant differences were seen on univariate analyses when analysing gains (11% vs 27%, HR 3.07 (1.09–8.68) p=0.03), or del (0% vs 28%) HR 3.46 (1.02–11.8) p=0.05) separately, with some evidence that UPD also produced worse results (0% v 34% HR 1.82 (0.97–3.45) p=0.06). In multivariate analyses, adjusted for age, WBC, sex, secondary disease, performance status, IDH1, IDH2, FLT3 ITD and NPM1 all SNP array detected aberrations were independently prognostic for overall survival. Analysis combining all aberrations gave a HR of 2.99(1.33–0.75) p=0.006. When a model-building approach was used, after adjustment for baseline covariates, del p=0.005, gains p=0.002 and UPD p=0.02 all entered the model. We show, for the first time, that SNP microarray analysis in AML-NK pts reveals cryptic UPD and CN changes that may allow risk stratification and alter therapeutic strategies for this poor prognostic subgroup of patients.