The significance of deletion architecture, ATM mutational status and genomic complexity in 11q deleted CLL

Rose-Zerilli, M. J. J., Parker, H., Parker, A., Rodriguez, A., Chaplin, T., Gardiner, A., Collins, A., Young, B. D., Stankovic, T., Oscier, D. G. and Strefford, J. C. (2011) The significance of deletion architecture, ATM mutational status and genomic complexity in 11q deleted CLL. Clinical Lymphoma, Myeloma & Leukemia, (supl), [S146]. (doi:10.1016/j.clml.2011.09.010).

Abstract

Genes targeted by chromosomal deletions have been identified within ‘minimally deleted regions' (MDR). Although powerful, this approach is a simplification of the architecture and gene content of deletions, which in patients are almost always larger and result in disruption of multiple regulatory sequences. One such example is deletion of 11q23. When this includes ATM, it is associated with a poor response to chemotherapy but an improved outcome after chemo-immunotherapy. In contrast to Mantle Cell Lymphoma (MCL), where 11q deletions invariably accompany an inactivating ATM mutation, no mutation can be identified in 60-70% of 11q-deleted CLL. Although there is evidence supporting a role for other genes on 11q in CLL pathogenesis, few candidates have been proposed. To fully realize the biological and clinical consequences of 11q deletions in CLL we have employed genomic profiling (SNP6.0) and ATM mutational screening (all ATM expressed and regulatory regions by DHPLC and sequencing) to determine 11q architecture in 267 patients. These include 93 cases sampled at diagnosis and 154 at disease progression, of which 134 were enrolled on the UK CLL4 trial. We identified 112 chromosome 11 deletions in 76 cases, with 69 deletions including ATM, with a mean size of 27.8Mb (521.7kb–58.2Mb). A 416Kb MDR was defined between 107.498-107.914Mb (hg18) containing 7 genes including ATM; however, the deletion was confined to the MDR in only 4% of 11q-deleted cases. In fact, 11q deletions encompassing ATM resulted in the loss of an average of 266 genes, including frequent deletion of H2AFX (26%) and MRE11 (58%). Furthermore, gene pathway analysis of 11q-deleted genes revealed that 29 genes known to interact with ATM or causally implicated in response to DNA damage were commonly deleted, suggesting that ATM is rarely deregulated in isolation. The presence of widespread 11q gene deregulation is further supported by five other chromosome 11 MDRs identified in both ATM-deleted and non-deleted cases, which contain eight genes (CNTF, EED, FERMT3, FZD4, RAG1, RAG2, PICALM, TRAF6) that are known to interact with ATM at the protein level or are implicated in response to DNA damage. Our data showed mutations of the residual ATM allele in 50% of ATM-deleted cases. Interestingly, when patients were stratified based on the presence or absence of an ATM mutation, 11q deletions were significantly larger (28.9 Mb vs 35.2 Mb; P = 0.02) when ATM remained un-mutated, resulting in the loss of additional genes in these cases. Given that regions of acquired UPD (aUPD) invariably duplicate cancer genes, it was exciting to identify four patients with 11q-aUPD in the absence of a mutation of ATM. Although ATM deletion status was associated with increased risk of genetic complexity (≥ 3 copy number changes; OR = 2.8, 95% CI: 1.6-5.0; P = 4x10-4), a residual ATM mutation did not infer a significantly greater risk of genetic complexity in ATM deleted cases. In addition, the 22% (n = 18) of non-11q deleted cases harbouring a mutated ATM gene did not have a greater risk of complexity compared to the cases with a wild-type ATM allele; suggesting genetic complexity may not solely be driven by the ATM locus. Notably, the relationship between 11q deletion anatomy and complexity revealed that ATM deletions extending as far as H2AFX (n = 18) were significantly associated with greater levels of complexity (OR = 5.2, 95% CI 1.1-25; P = 0.04). This association potentially reflects the role of a single deletion event, targeting multiple genes (e.g. ATM and H2AFX) known to interact in the DNA-damage response pathway, in driving genomic complexity in CLL. As predicted, ATM deletion was associated with reduced overall (OS) and treatment-free survival (median 78 vs 113 months, P = 0.001. and 6 vs 23 months, P = 0.003). Biallelic ATM inactivation was associated with an additional adverse impact on OS (50 vs 113 months, P = 0.003), supporting the hypothesis that 11q deletions are clinically heterogeneous. Unfortunately, when we attempted to identify a specific deletion anatomy associated with additional impact on clinical outcome, we were unsuccessful due to insufficient statistical power. In conclusion, our data show that 11q deletions potentially deregulate a large number of genes, a selection of which, other than ATM, are involved in the response to DNA damage and the maintenance of genome stability. These data also support the notion that while 11q-deleted cases can be sub-grouped based on the deletion architecture and ATM mutational status, larger studies are required to determine any impact on clinical outcome. We are currently employing target-enrichment and next generation sequencing to identify somatically acquired mutations on 11q, in particularly in those cases with 11q-aUPD that may reflect novel biomarkers or will expand our understanding of CLL biology.

This record has no associated files available for download.

Contributors

Author: J. C. Strefford

Download statistics