Abstract： Objectives To investigate patients with acute lymphoblastic leukemia (ALL) for TEL/AML1 fusion, BCR/ABL fusion, MLL gene rearrangements, and numerical changes of chromosomes 4, 10, 17 and 21 by fluorescence in situ hybridization (FISH) and to determine the relationship and the significance of those findings.Methods Fifty-one American patients (34 men and 17 women) were included in this study. Of them there were 41 patients with pro-B cell type ALL, 9 with B cell type ALL and 1 with T cell type ALL. Chromosome metaphases of each sample were prepared according to standard protocols. Fluorescence in situ hybridization was performed using commercially available DNA probes, including whole chromosome painting probes, locus specific probes, specific chromosome centromere probes and dual color/multiple color translocation fusion probes. The digital image analysis was carried out using Cytovision and Quips FISH programs.Results An overall incidence of chromosomal anomalies, including t (9;22), MLL gene rearrangements, t (12;21), and numerical chromosomal anomalies of chromosomes 4, 10, 17 and 21 was found in 33 patients (65%). Thirty-one of them were pediatric patients and two adults. The t (12;21) was the commonest chromosomal anomaly detected in this population; 14 out of the 45 pediatric patients (31%) were positive for TEL/AML1 fusion, among which three had an additional derivative 21 [t (12;21)], four had a deletion of 12p and two had an extra copy of chromosome 21. All 14 patients with positive TEL/AML1 fusion had ALL pre-B cell or B-cell lineage according to standard immunotyping. The percentage of cells with fusion signals ranged from 20% to 80%. All fourteen patients positive for TEL/AML1 gene fusion were mosaic. Three out of the 14 patients positive for the TEL/AML1 gene fusion were originally reported to be culture failures and none of the remaining eleven samples had been found to have chromosome 12 abnormalities by conventional cytogenetic techniques. All pediatric patients with pre-T or T cell lineage and the six adults were negative for TEL/AML1 fusion. One patient had double Philadelphia chromosomes, three had a rearrangement or a deletion of the MLL gene, one had t (4;11) and two had a deletion of the MLL. One of the patients with an MLL deletion also had a large ring of chromosome 21, and r (21) was caused by AML1 gene tandemly duplicated at least five times. The second case with the MLL deletion was also unique, the patient had a t (12;21) as well. A total of 20 patients had numerical changes (gain or loss) of chromosomes 4, 10, 17 and 21. Eight patients were found to have trisomies of three or four different chromosomes. Interestingly, seven of these patients did not have TEL/AML1, BCR/ABL or the MLL gene rearrangement; one did have the TEL/AML1 gene fusion. Eleven patients with pro-B cell or B cell type ALL (9 children with ALL, 2 adults with ALL) had numerical changes of chromosome 21 (gain 1 or 2 chromosome 21), among them, 10 patients had no structural alteration of chromosome 21, and one was combined by t (12; 21). Four patients had a monosomy of chromosome 17 and three out of these patients with monosomy 17 also had a fusion signal of TEL/AML1.Conclusions FISH plays an important role in detecting chromosome changes, especially in some cryptic chromosome translocations and patients with culture failures. This study found a trend towards a division between patients who had structural changes such as t (12;21) or a ring chromosome 21 and those who had numerical changes of chromosome 21 as well as the patients with TEL/AML1 fusion and patients with the coexistence of numerical chromosomal changes of chromosomes 4, 10 and 17. In our opinion there are two separate mechanisms which lead to the development or progression of leukemia.