Genetic Characterization of B-Cell Prolymphocytic Leukemia (B-PLL): A Hierarchical Prognostic Model Involving MYC and TP53 Abnormalities. on Behalf of the Groupe Francophone De Cytogenetique Hematologique (GFCH) and the French Innovative Leukemia Organization (FILO) Group

Abstract

Abstract B-PLL is defined by the presence of prolymphocytes in peripheral blood exceeding 55% of lymphoid cells. The diagnosis, mainly based on clinical and morphological data, can be difficult because of overlap with other B-cell malignancies. Because of the rarity of the disease, only case reports and small series describe its cytogenetic features. Few prognostic markers have been identified in this aggressive leukemia usually resistant to standard chemo-immuno therapy. We report here the cytogenetic and molecular findings in a large series of B-PLL. We also studied the in vitro response to novel targeted drugs on primary B-PLL cells. The study included 34 cases with a diagnosis of B-PLL validated by morphological review performed by three independent expert cytologists. The diagnosis of mantle cell lymphoma was excluded by karyotype (K) and FISH using CCND1, CCND2 and CCND3 probes. Median age at diagnosis was 72 years [46-88]. K was complex (≥3 abnormalities) in 73%, and highly complex (HCK≥5) in 45%. Combining K and FISH data, the most frequent chromosomal aberrations were: translocation targeting the MYC gene [t(MYC)] (21/34, 62%), 17p deletion including TP53 gene (13/34, 38%), trisomy 18/18q (10/33, 30%), 13q14 deletion (10/34, 29%), trisomy 3 (8/33, 24%), trisomy 12 (8/34, 24%) and 8p deletion (7/31, 23%). Whole-Exome Sequencing analysis of paired tumor-control DNA was performed in 16 patients. The most frequently mutated genes were TP53(6/16, 38%), associated with del17p in all, MYD88 (n=4), BCOR (n=4), MYC (n=3), SF3B1 (n=3), FAT1 (n=3), SETD2 (n=2), CHD2 (n=2), CXCR4 (n=2), BCLAF1 (n=2) and NFASC (n=2). Distribution of the chromosomal aberrations is shown in Fig 1. The main group of patients (21/34, 62%) had a t(MYC) that was associated with a higher % of prolymphocytes (86 vs 76, p=0.03), CD38 expression (90% vs 15%,p<0.001), and a lower K complexity (HCK≥5 : 20% vs 85%, p=0.0004). Mutations in MYC and in genes involved in RNA metabolism and chromatin remodeling were almost exclusively observed with t(MYC). Principal component analysis of gene expression data in 12 cases analyzed by RNA-Seq showed that the 7 patients with t(MYC) clustered together. These results suggest that t(MYC) form a homogeneous subgroup of B-PLL. A second group with MYC gain (5/34, 15%), was associated with HCK≥5 (100% vs 36%, p=0,01) and trisomy 3 (80% vs 14%, p=0,008). Altogether, 26/34 patients (76%) had a MYC activation, translocation or gain, that were mutually exclusive. The median overall survival (OS) for the entire cohort was 126 months with a median follow-up time of 47 months [ 0.2-141]. We found MYC activation (translocation or gain) to be associated with a shorter OS (p=0.03). Regarding MYC and del17p, we identified 3 distinct cytogenetic prognostic groups, with significant differences in OS (p=0.0006) (Fig 2). The patients without MYC activation had the lower risk (n=8, median not reached). Patients with a MYC activation without del17p had an intermediate risk (n=18, 125 months). The highest risk group corresponded to patients with both MYC and TP53 aberrations (n=7, 11 months). We performed drug response profiling on primary B-PLL cells using the ATP-based CellTiter Glo kit (Promega) (n=5). We observed that after 48h of exposure to increased doses, response was heterogeneous, with a majority of samples resistant to fludarabine (n=3), ibrutinib (n=3), idelalisib (n=4), venetoclax (n=3) and OTX015 (n=4). Annexin/PI assays using flow cytometry showed that the induced cell death could be increased by combination of ibrutinib or venetoclax with OTX015 or JQ1, two BET protein inhibitors that target MYC signaling (n=1/2). In summary, B-PLL have complex and highly complex K, a high frequency of MYC activation by translocation or gain, frequent 17p deletion, and frequent mutations in MYC, TP53, BCOR, and MYD88 genes. We identified 3 prognostic subgroups according to MYC and 17p status. Patients with MYC activation + 17p deletion have the shorter OS, and should be considered as a high-risk "double-hit" subgroup. Our results show that cytogenetic analysis is a useful diagnostic tool in B-PLL that improves prognostic stratification. We recommend to perform K and FISH (MYC and TP53) analyses systematically when a B-PLL is suspected. Our in vitro data suggest that drugs targeting the BCR and BCL2 in combination with MYC inhibition may be a therapeutic option in some patients. Disclosures Baseggio: Takeda Oncology: Honoraria.