Jonathan C. Strefford

We report 2 novel, cryptic chromosomal abnormalities in precursor B-cell acute lymphoblastic leukemia (BCP-ALL): a translocation, either t(X;14)(p22;q32) or t(Y;14)(p11;q32), in 33 patients and an interstitial deletion, either del(X)(p22.33p22.33) or del(Y)(p11.32p11.32), in 64 patients, involving the pseudoautosomal region (PAR1) of the sex chromosomes. The incidence of these abnormalities was 5% in childhood ALL (0.8% with the translocation, 4.2% with the deletion). Patients with the translocation were older (median age, 16 years), whereas the patients with the deletion were younger (median age, 4 years). The 2 abnormalities result in deregulated expression of the cytokine receptor, cytokine receptor-like factor 2, CRLF2 (also known as thymic stromal-derived lymphopoietin receptor, TSLPR). Overexpression of CRLF2 was associated with activation of the JAK-STAT pathway in cell lines and transduced primary B-cell progenitors, sustaining their proliferation and indicating a causal role of CRLF2 overexpression in lymphoid transformation. In Down syndrome (DS) ALL and 2 non-DS BCP-ALL cell lines, CRLF2 deregulation was associated with mutations of the JAK2 pseudokinase domain, suggesting oncogenic cooperation as well as highlighting a link between non-DS ALL and JAK2 mutations.

Richter syndrome (RS) occurs in up to 15% of patients with chronic lymphocytic leukemia (CLL). Although RS, usually represented by the histologic transformation to a diffuse large B-cell lymphoma (DLBCL), is associated with a very poor outcome, especially when clonally related to the preexisting CLL, the mechanisms leading to RS have not been clarified. To better understand the pathogenesis of RS, we analyzed a series of cases including 59 RS, 28 CLL phase of RS, 315 CLL, and 127 de novo DLBCL. RS demonstrated a genomic complexity intermediate between CLL and DLBCL. Cell-cycle deregulation via inactivation of TP53 and of CDKN2A was a main mechanism in the histologic transformation from CLL phase, being present in approximately one half of the cases, and affected the outcome of the RS patients. A second major subgroup was characterized by the presence of trisomy 12 and comprised one third of the cases. Although RS shared some of the lesions seen in de novo DLBCL, its genomic profile was clearly separate. The CLL phase preceding RS had not a generalized increase in genomic complexity compared with untransformed CLL, but it presented clear differences in the frequency of specific genetic lesions.

Inactivation of the tumor suppressor gene, CDKN2A, can occur by deletion, methylation, or mutation. We assessed the principal mode of inactivation in childhood acute lymphoblastic leukemia (ALL) and frequency in biologically relevant subgroups. Mutation or methylation was rare, whereas genomic deletion occurred in 21% of B-cell precursor ALL and 50% of T-ALL patients. Single nucleotide polymorphism arrays revealed copy number neutral (CNN) loss of heterozygosity (LOH) in 8% of patients. Array-based comparative genomic hybridization demonstrated that the mean size of deletions was 14.8 Mb and biallelic deletions composed a large and small deletion (mean sizes, 23.3 Mb and 1.4 Mb). Among 86 patients, only 2 small deletions were below the resolution of detection by fluorescence in situ hybridization. Patients with high hyperdiploidy, ETV6-RUNX1, or 11q23/MLL rearrangements had low rates of deletion (11%, 15%, 13%), whereas patients with t(9;22), t(1;19), TLX3, or TLX1 rearrangements had higher frequencies (61%, 42%, 78%, and 89%). In conclusion, CDKN2A deletion is a significant secondary abnormality in childhood ALL strongly correlated with phenotype and genotype. The variation in the incidence of CDKN2A deletions by cytogenetic subgroup may explain its inconsistent association with outcome. CNN LOH without apparent CDKN2A inactivation suggests the presence of other relevant genes in this region.

Patients with acute lymphoblastic leukemia (ALL) and an intrachromosomal amplification of chromosome 21 (iAMP21) comprise a novel and distinct biological subgroup. We prospectively screened 1630 (84%) patients treated on the UK MRC ALL97 protocol for iAMP21 and herein present demographic, clinical, and survival data on the 28 (2%) children found to harbor this abnormality. They had a common or pre-B ALL immunophenotype, were significantly older (median 9 years vs 5 years), and had a lower white cell count (median 3.9 vs 12.4) compared with children without this abnormality. Notably, patients with iAMP21 had a significantly inferior event-free and overall survival at 5 years compared with other patients: 29% (95% confidence interval [CI], 13%-48%) versus 78% (95% CI, 76%-80%) and 71% (95% CI, 51%-84%) versus 87% (95% CI, 85%-88%), respectively. As a result of this 3-fold increase in relapse risk, newly diagnosed patients with iAMP21 recruited to the current UK MRC ALL2003 trial are being treated on the high-risk arm and are considered for bone marrow transplantation in first remission.