CORRESPONDENCE Further characterization of complex chromosomal rearrangements in myeloid malignancies: spectral karyotyping adds precision in defining abnormalities associated with poor prognosis

In the last 30 years, information provided by cytogenetic analysis has become indispensable for the clinical management of patients with hematological malignancies. In acute myeloid leukemia (AML), the favorable prognostic subgroup is defined by the presence of leukemic blasts with t(15;17), t(8;21), or inv(16). The unfavorable cases are those with abnormalities involving more than two chromosomes, monosomy 5/5q− or 7/7q−, or rearrangements of the long arm of chromosome 3. The survival rate of this group is less than 20% at 5 years. These patients represent a considerable therapeutic challenge for whom no current treatment approach is satisfactory.1 The patients with a normal karyotype or cytogenetic abnormalities that are not included in these other categories are characterized as having an intermediate risk of relapse.1,2 In myelodysplastic syndromes (MDS), abnormalities of 7q or a complex karyotype are also unfavorable prognostic factors. Improvements in treatment of AML, t-AML and MDS have resulted in high complete remission rates even in patients with unfavorable cytogenetics; however, the majority of patients relapse.1,3 Therefore, further genetic studies redefining the patients included in the poor prognostic group are necessary. In order to characterize the karyotype more precisely, we analyzed 18 samples from patients with myeloid malignancies and a complex karyotype using three different techniques: G-banding, fluorescence in situ hybridization (FISH), and spectral karyotyping (SKY). The SKY probe mixture and hybridization reagents were obtained from Applied Spectral Imaging (Carlsbad, CA, USA). Slides for spectral karyotyping were hybridized as previously described. For analysis of cases that had complex markers or non-obvious chromosome rearrangements, FISH experiments were performed using the appropriate painting or centromere-specific probes (Vysis, Downers Grove, IL, USA). All 18 cases were successfully analyzed using SKY. In the three cases with complete G-band analysis (cases 6, 8, 14), SKY confirmed the G-banding results. In the other 15 cases, the use of SKY substantially improved the precision of karyotype analysis of malignant cells, detecting several unexpected aberrations. The approach of combining three different cytogenetic techniques allowed the identification of hidden translocations and the reconstruction of complex rearrangements. The complete karyotypes after the combined analysis are given in Table 1. Deletions and unbalanced translocations in samples with complex aberrations were particularly prone to misinterpretation based on G-banding alone, especially when chromosomal regions that have a similar G-banding pattern were involved. In 13 cases (1, 2, 4, 5, 7, 9, 11, 12, 13, 15, 16, 17, 18), chromosome material from total or partial monosomies detected by G-banding was found in derivative or marker chromosomes. On the other hand, in seven cases (7, 9, 11, 13, 15, 16, 18) deletions were actually found to be translocations after SKY. This is especially important in cases with aberrations with prognostic significance. Thirteen cases (72%) had monosomy 5 or 7, or abnormalities of 3q, 5q or 7q. SKY confirmed the G-banding results in six of these cases and detected new abnormalities in seven others. Three of our cases were −5 (cases 7, 9 and 13); based on SKY, all three were changed to a deletion 5q, eg loss of chromosome 5, bands 5q31 to 5q33 (case 7); translocation of parts of 5 to a der(7) and a der(13), that resulted in loss of chromosome 5, bands 5q13 to 5q33, in case 9. In case 13, 5q was thought to be on chromosome 18, however, SKY demonstrated that a der(5), chromosome break in