High-Resolution SNP-Array Profiling of Chronic Lymphocytic Leukemia

Abstract 50

Genomic aberrations are important prognostic factors in chronic lymphocytic leukemia (CLL) [Döhner et al., 2000]. However, known genomic aberrations fail to fully explain the biologic and clinical heterogeneity of the disease. We sought to precisely map copy number alterations (CNA) and copy number neutral losses of heterozygocity (LOH) to better characterize known recurrent aberrations and to identify new genetic lesions.

We used Affymetrix 6.0 single nucleotide polymorphism (SNP) array analyses on CD19 sorted CLL cells. Data were analyzed using dChipSNP, a modified array normalization algorithm guided by cytogenetic abnormalities and a circular binary segmentation. We studied samples from 346 patients enrolled on the CLL8 trial of the German CLL Study Group. Data of 145 samples were analyzed against intraindividual reference DNA (paired), data of 201 samples against a pool of reference DNA (unpaired). FISH data were available for all samples, the distribution of genomic aberrations was as follows: del(13q14) in 59.8%, del(11q23) in 26.3%, trisomy 12 in 11.6%, and del(17p13) in 8.4%. IGHV was mutated in 32.9%, unmutated in 63.3%, and unknown in 3.8%.

In total, 261 tumor-specific CNA could be discovered among the 145 paired samples. Genomic aberrations were found in 85.5% of these cases. The average number of aberrations per case was 1.8; according to the hierarchical model of genomic aberrations, it was 3.5 in del(17p), 2.4 in del(11q23), 1.7 in del(13q14) single, and 0.5 in normal karyotype CLL. The minimally deleted region (MDR) on 13q14 was 277.25 kb in size and contained mir15a and mir16, DLEU1 and DLEU2; RFP2 was not part of the MDR. Deletions on 13q were highly heterogeneous in size, ranging from 294 kb to 68 Mb. On 11q23 the MDR only contained ATM, the smallest lesion of 78.5 kb being intragenic; in two of theses cases, the deletion size was too small to be detected by FISH analysis. TP53 was affected in all del(17p13) cases except two; one tumor-specific deletion of 635.7 kb was detected in cytoband 17p13.2 harboring 30 genes and a second deletion of 780 kb in 17p13.3 containing – among 15 other genes – MNT, a tumor suppressor acting as an antagonist of MYC. A partial trisomy on chromosome 12 was not detected. Of the 261 CNA, 95 were located in genomic regions that are not evaluated by our routine FISH probe panel; 17 regions were affected recurrently: del(1p35.3) [2/145], del(1q23.3) [2/145], del(1q42.12) [2/145], +2p [5/145], del(3p21.31) [2/145], del(6p25.3) [3/145], +(6p25.3) [2/145], del(6q) [11/145], del(7q23.1) [2/145], +(8q24.21) [3/145], del(9q13-q21.13) [2/145], del(10q24) [2/145], del(14q24.3) [2/145], del(14q12.3) [2/145], del(15q15.1) [2/145], +18 [3/145] and +19 [7/145]. The frequency of these CNA was subsequently evaluated within the cohort of 201 (unpaired) samples. Five of 17 regions were affected in more than 2% in the whole cohort: +2p, del(6q), +8q24.21, del(15q15.1), and +19. Gain of 2p was found in 6.9% of cases, the minimally amplified region was 1.9 Mb in size and contained e.g. BCL11A and REL. Del(6q) was detected in 6,4%, the deletions were heterogeneous, an MDR could not be identified. 16 cases had 8q24.21 gains, the minimally amplified region was delineated by three intragenic gains in MYC. 14 cases had loss in 15q15.1 focussing on MGA, a potential suppressor of transcriptional activation by MYC. 8 cases had total or partial gains of chromosome 19, among those two overlapping partial gains with a minimally amplified region of 2.17 Mb in 19p13.2. Tumor-specific LOH were identified in 6.0% (9/145) located on 13q in three cases and in one case each on 17p, 12q, 11p, 1p, 3 and 22q. The LOH on chromosomes one and three overlapped with recurrent losses in 1p35.3 and 3p21.31. Essential members of the ATR-pathway were located in these regions: ATRIP and RPA2. However, mutational analyses of the two candidate genes in 48 cases revealed no mutations.

SNP array analysis is a reliable tool to identify and further characterize genomic aberrations in CLL. MDR on 13q14 was delineated to a 277.25 kb segment affecting mir15a, mir16, DLEU1 and DLEU2 but not RFP2; the MDR on 11q23 to a segment only containing ATM. Cases with del(11q23) and del(17p) showed a higher genomic complexity than those with normal karyotype or del(13q14) as single abnormality. Relatively few novel genetic lesions were identified. Although occurring at low frequency, they may lead to the discovery of new genes involved in CLL pathogenesis.