Flow cytometric DNA index, G-band karyotyping, and comparative genomic hybridization in detection of high hyperdiploidy in childhood acute lymphoblastic leukemia.

High hyperdiploid acute lymphoblastic leukemia in children is related to a good outcome. Because these patients may be stratified to a low-intensity treatment, we have investigated the sensitivity of flow cytometry (FCM), G-band karyotyping (GBK), and high-resolution comparative genomic hybridization (HR-CGH) in detecting high hyperdiploid leukemic clones. Twenty-six girls and 34 boys with acute lymphoblastic leukemia diagnosed in 1998 to 1999 were analyzed by FCM, GBK, and HR-CGH. The correlations between DNA indices obtained by FCM, GBK, and HR-CGH were significant (rs=0.61 to 0.77; P<0.001 for all comparisons). However, in 4 of 18 patients, high hyperdiploidy was overlooked by GBK or HR-CGH, and even when FCM was applied, 2 of 18 patients with high hyperdiploidy by GBK and/or HR-CGH were classified as nonhigh hyperdiploid. If high hyperdiploid subclones were included, FCM could detect all high hyperdiploid patients found by either GBK or HR-CGH, but would then in addition classify 15% to 20% of the remaining patients as high hyperdiploid. Thus, both GBK and HR-CGH overlook patients with high hyperdiploidy, and FCM only detects all high hyperdiploid patients if small high hyperdiploid clones are included. In addition, FCM detects patients with high hyperdiploid subclones, not detected by either GBK or HR-CGH, and the challenge remains to determine the prognosis of patients with such high hyperdiploid subclones.