Our data have demonstrated that the 244K oligonucleotide array-CGH platform is a powerful tool to detect additional copy number alterations in ETV6/RUNX1-positive patients. A total of 155 aberrations were identified, including microdeletions as small as 400 bp. Many known or potential genes related to leukemia were also identified using this method. These data supported the secondary leukemogenic model that additional aberrations are necessary for leukemogenesis. According to our array data, 5 out of 11 patients (45%) showed deletion involving ETV6 gene from as small as 0.2 Mb to 19.4 Mb. We found more deletions (76.8%) than amplifications (23.2%), which is in agreement with a previous study . Among the deletions, 32.7% were larger than 1 Mb, while 33.3% of the amplifications were larger than 1 Mb.
Patient no. 2 harbored a 0.09 Mb deletion on 9p13.2 that involved the PAX5 gene. PAX5 is important in the normal development of B cells, in which loss of a wild-type PAX5 allele would cause differentiation arrest in ALL . Deletion of the tumor suppressor CDKN2A gene located at 9p21.3 was found in 36% (4/11) of our patients. The CDKN2A deletion is suggested to occur more frequently in T-ALL than in precursor B-ALL . The deletion is thought to vary by cytogenetic subgroup and the prognostic value of the incidence is yet to be determined . One patient (no. 7) was found to have a gross deletion (1.0 Mb) on 3p14.2 region that included the FHIT gene, which is proposed as a putative tumor-suppressor gene. The deletion on this particular gene was found to be correlated with a low clinical remission rate and poor overall survival [15–17].
Several putative target genes within the commonly gained region, including cryptic Xq duplications were also found in patient no. 3 and 6, both females. The sizes of the gains on the two patients were 2.1 Mb and 91 Mb, respectively. This result is discordant with the previous report that males are more common to harbor this aberration . This discrepancy may be explained by the small sample size used in this study. It would be interesting to study the expression level of ETV6/RUNX1 proposed genes, namely the SPANX family genes, on the X chromosome in our female's dataset.
Based on our FISH study on five childhood ALL patients, all samples showed a positive ETV6/RUNX1 fusion signal. Three patients (nos. 1, 3 and 5) showed concordant result with array CGH for ETV6 gene deletion. FISH result for patient no. 3 showed three red signals, suggesting that there was a duplication of the RUNX1 signal, but was not confirmed through the array findings. It has been reported that DNA microarray may fail to detect the chromosomal abnormalities if the abnormal clones are present in fewer than 25% of the cell population .
Patient no. 6 showed a unique FISH profile where two fusion signals of the ETV6/RUNX1 were detected. Double ETV6/RUNX1 fusion signals were found in 25% of ETV6/RUNX1 positive ALL patients . Previous studies have found that the additional ETV6/RUNX1 fusion signal may have arisen from duplication of the der(21)t(12;21) chromosome [21, 22], duplication of ETV6/RUNX1 fusion gene that was later translocated onto another chromosome  or ider(21)(q10)t(12;21)(p12;q22) . In the study by Loncarevic and coworkers (1999), gain of the der(21)t(12;21) chromosome was found exclusively in the relapsed cases . We were not able to ascertain the origin of the extra ETV6/RUNX1 fusion signal in our patient due to non-availability of metaphase cytogenetics. It has however been suggested that secondary changes such as the duplication of fusion signals may contribute to the process of leukemogenesis .
Three of the patients, namely patient nos. 4, 10 and 11, had a relapse. Of the three, patient no. 4 had multiple gross deletions as large as 90.8 Mb, whereas patient no. 11 had other gross imbalances larger than 1 Mb. However, the array report for patient no. 10 showed no gross imbalances larger than 1 Mb. We could not determine whether any subsequent aberrations happened after the sample was taken which might trigger the relapse event.