Is CDKN2A +/− CDKN2B and MTAP Inactivation of Prognostic Significance in B-Precursor Childhood Acute Lymphoblastic Leukemia? Results of EORTC Studies 58881 and 58951.

Background: Deletions of the 9p21 chromosomal region which encompasses CDKN2A, a gene encoding both p16^INK4a and p14^ARF, are frequent in childhood acute lymphoblastic leukemia (ALL). Their prognostic relevance is controversial. Indeed, small retrospective series of patients were analysed so far, which usually mixed B- and T-lineage ALL. Moreover, the prognosis could vary according to the extent of the deletion since the inactivation of contiguous genes such as CDKN2B encoding p15^INK4b or MTAP encoding methylthioadenosine phosphorylase, may influence chemosensitivity.

Methods: 9p21 deletions were retrospectivelly studied in 227 children (aged 2 mo to 16 yrs) with B-lineage ALL, using a real-time PCR assay for CDKN2B, CDKN2A (e1b and e3), and MTAP gene dosage, and loss of heterozygosity analysis (Bertin et al., 2003). CDKN2A and CDKN2B inactivation by promoter methylation was also investigated. TEL-AML1 fusion was screened in 182 (80%) of these patients. All patients were enrolled in EORTC trials 58 881 or 58 951. Median follow-up was 6 years; total number of events was 55 and the overall 6-year EFS rate was 74%.

Results:CDKN2A bi- and mono-allelic inactivation was found in 38 (17%), and 31 (14%) B-lineage ALL respectively. Patients with a bi-allelic inactivation did not differ from undeleted ones regarding age, sex, immunophenotype and response to prephase treatment, but tended to have a higher WBC count at diagnosis and were more often allocated to the NCI high risk group (42% vs 27%). Genetic defects, including TEL-AML1, were independent of 9p21 alteration with exception of hyperdiploidy (>50 chromosomes) which was less frequent in patients with CDKN2A inactivation (5% vs 34%). The 6-year EFS rates of patients with bi-allelic, mono allelic and no inactivation did not differ significantly (68%, 80%, and 75% respectively). Bi-allelic CDKN2B inactivation by either deletion or promoter methylation was found in 36 (16%) patients and the 6-year EFS rate of these patients was similar to that of undeleted patients (71% vs 74%). MTAP bi-allelic inactivation was found in 24 (11%) of patients. Although MTAP deficient cells are in vitro very sensitive to methotrexate, the 6-year EFS rate (70% vs 75%) did not differ from that of MTAP positive patients even when the analysis was restricted to CDKN2A inactivated ALL.

Conclusion: Although bi-alleic CDKN2A inactivation was more often associated with bad prognostic parameters in B-lineage ALL, it failed to significantly influence the outcome of the patients. We did not observe either any influence of the co-inactivation of CDKN2B and MTAP on the outcome. Considering that our study is the largest so far using molecular analysis of the 9p21 region, we assume that any bad prognosis associated with 9p21 alteration, if present, should be weak in B-lineage ALL, and would probably require larger cohorts of patients to be ascertained.