Clinico-biological characterization of a subset of variant B-CLL, defined according to a combined cytofluorimetric/fish diagnostic approach

Sixty three cases of variant B chronic lymphocytic leukemia (v-B-CLL), characterized by a intermediate CLL/mantle cell lymphoma immunophenotype, atypical cytomorphology in absence of t(11;14)(q13;q32) in FISH analysis were compared with a series of 130 B-CLL. The v-B-CLL were significantly different from the B-CLL in terms of: age <70 yrs (p <.001), lymphocytosis <20 x 109/ (p <.001), lymphocyte doubling time £ 12 months (p = .02), high serum b2-microglobulin levels (p <.001), and splenomegaly (p = .002). CD38 and CD49d expression was significantly different between v-B-CLL and B-CLL (p <.001) whereas, no statistical difference was observed for ZAP-70 reactivity. There were more patients mutated in the v-B-CLL group (80.0%) than in the B-CLL group (50%) (p = .001). FISH analysis demonstrated that trisomy 12 was more frequent in v-B-CLL ( p<.001), while del13q14, considered as a single alteration, was more frequent in B-CLL (p=.008). Gene expression profiling of a panel of 9 v-B-CLL compared with 60 B-CLL samples indicated that the variant group is characterized by an up-regulation of different genes involved in oncogenesis (TPD52, AFF1, GMPS, PICALM, JUN, REL, RAC2), regulation of apoptosis (IL-7, HSP90B1, NOTCH2, BECN1, ANXA4, MCL1) and involved in the I-kB kinase/NF-kB cascade of the canonical NF-kB signaling pathway (TRIM38, EEF1D, CASP1, MALT1, RHOH0). Among the genes found differentially expressed, CD1c, OSBPL3 and ITGA4 were upregulated. After a median follow-up of 55 months (range 4-196) and 60 months (range 6-180 ), 25/42 (59%) v-CLL and 55/93 (59 %) CLL pts were treated. Time to treatment was significant different between 2 groups when the IgVH mutational status was considered (p= .006). Median OS of v-CLL subset was 112 months vs 171 months of CLL subset. When the IgVH mutational status was considered, mutated cases showed a worse OS even if a statistical difference was not observed (p= 0.062). In conclusion, our study identifies a form of B cell leukemia that shows peculiar biological and clinical features and should not be misdiagnosed as a B-CLL. The inclusion of this form in B-CLL study could alter the interpretation of results, especially related to biological markers.