The detection of hyperdiploidy (clones with > 46 chromosomes) in the b
one marrow of patients with acute lymphoblastic leukaemia (ALL) is imp
ortant because of the prognostic impact of this finding. The high hype
rdiploid (HeH) subgroup with 51-68 chromosomes is associated with the
best outcome, followed by the low hyperdiploid (HeL) subgroup with 47-
50 chromosomes and the triploid/tetraploid (TT) subgroup with > 68 chr
omosomes, which do less well. We present a strategy for the use of flu
orescence in situ hybridization (FISH) with chromosome-specific probes
to detect hyperdiploidy in interphase cells and to assign cases to a
ploidy subgroup. By using a model population of 252 cases, it was seen
that ten chromosomes (X, 4, 6, 8, 10, 14, 16, 18, 20, and 21) used in
particular combinations and applied in a step-wise manner enabled the
detection of 94% of hyperdiploid cases and gave an accurate predictio
n of ploidy subgroup in 96% of these cases. The detection and classifi
cation of each case required the use of four to six probes over two or
three steps. Confirmation that this strategy will achieve this level
of detection in other hyperdiploid populations was demonstrated by usi
ng 250 published karyotypes. This strategy has an application in detec
ting missing or hidden hyperdiploid cases among cases with failed or n
ormal cytogenetics. (C) 1996 Wiley-Liss, Inc.