Gene Microdeletions in Adult and Pediatric Acute Lymphoblastic Leukemia,

Microdeletions of genes involved in B lymphopoiesis and cell-cycle regulation, such as CDKN2A/B, PAX5, IKZF1, ETV6, RB1, BTG1 and EBF1 have been reported as a frequent event in pediatric acute lymphoblastic leukemia (ALL). Whether these findings are found in adulthood and the possible differences with childhood ALL, as well as its prognostic implication, are still unknown.

To assess the differences between two cohorts of children and adults diagnosed with ALL on the frequency of deletions in these genes and their relationship with clinical data and prognosis.

We studied 70 children and 83 adults diagnosed with ALL with available DNA sample at diagnosis. In children, median age was 4y. (1 – 14), median leukocytes 10.3×10⁹/L (0.7 – 675) and the cytogenetic risk distribution was 42(39%), 30(27%) and 12(11%) for favourable [t(12;21) and hyperdiploidy], intermediate (normal karyotype and miscellaneous) and high risk [t(9;22), t(4;11), hypodiploid and complex karyotype], respectively. In adults, median age was 38y. (15 – 85), median leukocytes 16.8×10⁹/L (1 – 371) and 29(35%) patients belonged to the high risk cytogenetic group.

We performed Multiplex Ligation Probe Amplification (MLPA) using SALSA kit P335-A1 (MRC-Holland). PCR products were separated on an ABIPRISM 310 DNA Analyzer and analyzed using GeneMapper v3.2 (Applied Biosystems).

Frequency of deletions in the studied genes was similar in children and adults, except for IKZF1 deletions that were more frequent in adults (P<.001) (Table 1).

In children, ETV6 deletions occurred more frequently in patients with t(12;21) (67% of patients with deletion vs. 17% without, P <.001); CDKN2A/B deletions were found in patients assigned to the intermediate cytogenetic risk group (59% of patients with deletion vs. 23% without, P =.028); and the three cases with RB1 deletions were found in patients with hypodiploidy (P <.001).

In adults, ETV6 and CDKN2A/B deletions occurred more frequently in women (67% vs. 39%, P =.022 and 77% vs. 42%, P =.021, for patients with and without deletions, respectively); PAX5 and IKZF1 deletions appeared more frequently in patients with >30×10⁹/L leukocytes (60% vs. 27%, P =.032 and 52% vs. 21%, P =.007, for patients with and without deletions, respectively); besides, PAX5 deletions occurred in patients who belonged to the standard cytogenetic risk group (55% vs. 6% for patients with and without deletions, P <.001).

In the pediatric cohort, the leukocytes >30×10⁹/L and the cytogenetic risk group were the variables that reached statistical significance for both overall survival (OS) and relapse free survival (RFS) and also age >10y. for OS, but in the multivariate analyses, just the cytogenetic risk classification remained significant [HR: 4 (CI 95%: 1.6 – 10), P =. 004 for OS and HR: 3.5 (CI 95%: 1.7 – 7.2), P =. 001 for RFS].

In the adult cohort, multivariate analysis for OS including all significant variables in the univariate analysis (age >60y, karyotype, CDKN2A/B and ETV6 deletions) showed as independent variables: age >60y. [HR: 4.3 (CI 95%: 2.1 – 8.6), P<. 001] and CDKN2A/B deletions [HR: 2.6 (CI 95%: 1.4 – 5.3), P=. 004]. Similarly, taking into account karyotype, CDKN2A/B and ETV6 deletions for the RFS multivariate analyses, just ETV6 deletions arose as an independent factor [HR: 3.8 (CI 95%: 1.5 – 9.4), P=. 004]. In fact, having CDKN2A/B and/or ETV6 deletions conferred a worse prognosis to patients in both standard risk cytogenetic group (3y. RFS: 45% vs. 70% for patients with and without deletions, respectively; P =.049) and high risk cytogenetic group (3y. RFS: 14% vs. 66% for patients with and without deletions, respectively; P =.025).

This study shows the high incidence of deletions in genes of cell-cycle and B-lymphopoiesis in adult and pediatric ALL. However, the biological and prognostic implications of these deletions seem to differ between both patient groups: while cytogenetics was the strongest variable for risk assessment in children, gene microdeletions in CDKN2A/B and ETV6 added a prognostic value to karyotype in our adult cohort.

AP-194/10, R06/0020/0031, BES2008–008053, CM10/00321, CM09/00038, and CA08/00141.

No relevant conflicts of interest to declare.