ALBERT

Description: Background. TP53 mutations (TP53mut) along with 17p13 deletion (del17p) are strong predictors of poor survival and refractoriness to chemo-immunotherapies (CIT) in chronic lymphocytic leukemia (CLL), and their analyses should be always performed before treatment. Studies based upon ultra-deep-NGS have shown that TP53mut can be present at very low level in CLL cell populations, although their detrimental clinical impact in this setting is still matter of debate (Rossi, Blood 2014), and ERIC recomendations discourage to report TP53 mutations if subclonal Malcikova, Leukemia 2018). Aim. To investigated the presence and clinical relevance of clonal/subclonal TP53 aberrations in a large CLL cohort. Methods. The study includes 1,058 out of 1,613 CLL patients (509 treated with standard CIT) diagnosed between 1991 and 2018, and consecutively referred to a single institution for del17p analyses by FISH (167-kb 17p13 orange probe, MetaSystems), and TP53mut by ultra-deep NGS (MiSeq Illumina; median coverage 〉2,000X with an amplicon-based strategy covering exons 2-11 using 40ng DNA/test) in CD19-purified (〉85% pure) CLL samples, collected before treatment (as per ERIC recommendations). For TP53mut analyses, FASTQ files were aligned to the Hg19 reference with Burrows-Wheeler Aligner-MEM algorithm, and allele variants called by FreeBayes (Garrison & Marth, arXiv 2102) with non-stringent parameters. To calculate random/systematic errors we generated a specific database with all the variant allele frequencies (VAF) observed in a subset of TP53 wild type (wt) subjects (n=362). TP53mut were accepted if: i) validated by Fisher exact test after Bonferroni correction (p 12.5%) and 67 subclonal (all clones 0.4% for the clinical impact of TP53mut (Fig.D), and the c-index of combined clonal/subclonal TP53mut (0.645) was significantly higher than the c-index of clonal TP53mut alone (0.602; P10% of nuclei had significantly shorter OS than cases with del17p in 0.5% as the best cutoff. In keeping with this cutoff, TP53 mutated patients experienced a significantly shorter OS than wt patients. Again cases with clonal and subclonal mutation experienced the same OS (Fig.I). Importantly, the cutoff found in the training cohort was able to reproduce the very same results also in the validation cohort both in term of mutation per se and in terms of clonal and subclonal TP53 mutations (Fig.J). Conclusion. i) By applying ERIC recommendations and a rigorous pipeline of analysis, TP53mut impacted on OS also with VAF 10% of nuclei. These cutoffs may be employed for the clinical management of CLL patients. Figure Disclosures Di Raimondo: Takeda: Consultancy; Amgen: Consultancy, Honoraria, Research Funding. Rossi:Gilead: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; Abbvie: Honoraria, Other: Scientific advisory board; Janseen: Honoraria, Other: Scientific advisory board; Roche: Honoraria, Other: Scientific advisory board; Astra Zeneca: Honoraria, Other: Scientific advisory board.

Description: Background. CD49d, the alpha-chain of the integrin heterodimer VLA-4, has been identified among the strongest predictors of overall survival (OS) in chronic lymphocytic leukemia (CLL), along with IGHV mutational status and deletion of the 17p (17p-) chromosome involving TP53 (Bulian et al, JCO, 2014). In addition to TP53, the clinical relevance of NOTCH1, SF3B1 and BIRC3 gene mutations has been recently emphasized. Aim. To test the relevance of CD49d in molecular subgroups of CLL defined by NOTCH1, SF3B1 and BIRC3 gene mutations. Methods. The study was based on a cohort of 778 unselected CLL (422 untreated and 356 treated cases) all characterized for CD49d expression (CD49dhigh, ≥30% positive cells by flow cytometry in 229 cases), IGHV mutational status (unmutated, UM, in 262 cases), karyotype abnormalities according to the hierarchical stratification (13q- in 308 cases, +12 in 103 cases, 11q- in 64 cases), tested at diagnosis, along with mutations/deletions (hereinafter, disruption) of TP53 (disrupted in 84 cases, including 58 cases with 17p-), BIRC3 (disrupted in 59 cases), and mutationsof NOTCH1 (mutated in 81 cases), SF3B1 (mutated in 54 cases)tested at diagnosis in 65% of cases, in all cases before therapy. Chromosome abnormalities by FISH were defined by using a 5% cut-off; IGHV, TP53, BIRC3, NOTCH1 and SF3B1 mutations were investigated by Sanger sequencing. Median follow-up of patients was 80 months with 173 deaths. Results. i) association with CD49d: a CD49dhigh phenotype associated with age ≥65 years (p=0.0001), Rai stage I or higher (p〉0.0001), UM IGHV status (p〉0.0001), absence of 13q- (p=0.0001), presence of +12 (p

Description: Background. The pivotal role of the Immunoglobulin (Ig) receptor and antigenic stimulation have been proven to be landmarks for the understanding of the ontogeny and evolution of chronic lymphocytic leukemia (CLL). In addition, the mutational status of the Immunoglobulin Heavy-Chain Variable region gene (IGHV) was confirmed to be a reliable prognostic factor, supporting an antigen-driven model of CLL development. To clarify aspects regarding an antigenic involvement in CLL evolution, studies focusing on intraclonal diversification (ID) of Ig genes have provided relevant information, although mainly conducted in a pre-Next Generation Sequencing (NGS) era. Aim. To apply a NGS approach to investigate ID in CLL. Methods. IGHV genes from 530 CLL patients with Royal Masden Hospital score 4-5 (Fig. 1A) was sequenced using NGS (Lymphotrack). The bio-informatic pipeline was based on the pRESTO/ChangeO packages. Specific pathological clones were selected based on the presence of same IGHV, junction genes and with similar HCDR3 sequence according to Hamming's distance. Through the R-Alakazam package, we generated rarefaction curves to evaluate the clonal diversity inside the pathological clone (Fig. 1B). Focusing on the Simpson index (represented by the Hill number of order q=2), which gives more weight to larger clones minimizing the smaller ones (Fig. 1B), we selected a Diversity Score (DS) of 4 for the definition of cases without ID (clonal; DS

Description: Introduction. The clinical course of patients with chronic lymphocytic leukemia (CLL) is highly heterogeneous. The deletions/mutations of 17p/TP53 are predictors of chemorefractoriness, and for this reason, in the management algorithm of CLL patients, when present, indicate treatment with with chemo-free regimens also in the context of first-line therapy. Recent studies based on ultra-deep-next generation sequencing (NGS) have shown that TP53 mutations can be present at very low clonal abundance in tumor cell populations, although whether these mutations may have a detrimental clinical impact on disease course is still to be established. Aim. To investigate the presence of clonal and subclonal mutations of TP53 in a large cohort of CLL cases using an ultra-deep NGS strategy, and determined their clinical relevance for patients outcome. Methods. The study includes 590 CLL patients characterized for the deletion at chromosome 17p13 (FISH analysis) and TP53 mutations in samples before treatment. In all cases, analyses were carried out on DNA extracted from nearly pure (〉90%) tumor cells. TP53 mutational status was investigated by NGS with an amplicon based strategy. Sequencing reads analysis was made by the Burrows-Wheeler Aligner-MEM algorithm and by SAMtools. Variant calling was performed using the entire pipeline established on the MiSeq Reporter software. Results were expressed as percentage of mutated DNA. The minimal allelic fraction for mutation calling was set at 1%. Synonymous variants and polymorphisms described in the Single Nucleotide Polymorphism Database (dbSNP138) were removed. Outcome variable was overall survival (OS). Clinical correlations were made using Kaplan-Meier plots and log-rank test. Results. FISH and mutational analyses were performed in samples within 2 years from diagnosis in 92% of the cases (Figure 1A). A total of 125 TP53 mutations (Figure 1B) were found in 96 patients (11.7%). Subclonal mutations have similar molecular characteristics as their respective high frequency allele mutations supporting a comparable pathogenic effect (Figure 1B). According to a 15% cutoff of variant allele frequency (VAF), 78 cases were considered clonal and 18 subclonal (Figure 1C) for TP53 mutations (1% 〈 VAF 〈 15%). In this context, cases with subclonal and clonal TP53 mutations experienced significant shorter OS than TP53 wild-type (wt) cases, without differences between clonal and subclonal cases (Figure 1E). Accordingly, ROC analysis on the same cohort identified a cutoff of 〉0% for the clinical impact of TP53 mutations (Figure 1E inset). Deletion of chromosome 17p was found in 180 out of 574 patients (31.3%), and using a 10% cutoff, 61 patients presented a percentage of deleted nuclei above the cutoff (Figure 1D). Using only 17p deletion data and considering the above mentioned cutoff, patients with 17p13 deletion ≥10% experienced shorter OS than wt cases, while patients with 17p13 deletion 9% of deleted nuclei as optimal cutoff for OS discrimination (Figure 1F inset). Given the frequent co-occurrence of TP53 mutations with 17p deletions, we also evaluated the impact of isolated TP53 mutations and 17p deletions. By using the ROC cutoffs for the definition of mutated/deleted cases, 466 cases (81.1%) presented no TP53 disruption (TP53 mutations and deletion), 47 cases (8.2%) were TP53 mutated only, 15 cases (2.6%) were 17p deleted only and 46 cases (8.1%) presented a concomitant TP53 mutation and 17p deletion. Kaplan-Meier curves demonstrated comparable significant shorter OS intervals for TP53 mutated and/or deleted CLL cases respect to wt cases, while no differences were observed between these three groups (Figure 1G). Conclusion. By using a highly sensitive NGS approach, we have detected small subclones of TP53 in a relative high proportion of patients. TP53 mutations conferred a significant shorter OS irrespectively of VAF percent, while deletion of chromosome 17p impacted on OS only when detectable in more than 10% of nuclei. These cutoffs, once validated by prospective studies, may be employed in daily practice for the clinical management of CLL patients. Disclosures Zaja: Novartis: Honoraria, Research Funding; Abbvie: Honoraria; Celgene: Honoraria, Research Funding; Amgen: Honoraria; Janssen: Honoraria; Takeda: Honoraria; Sandoz: Honoraria.

Description: Quantitative evaluation of IgVH genes mutations is widely considered a reliable prognosticator in B-CLL. Conversely, conflicting results have been reported regarding the prognostic impact of IgVH gene mutations when evidence of Ag-driven selection is investigated. To address this issue, the mutational status of IgVH genes was analysed in peripheral blood samples of 147 B-CLL patients, all with survivals, by a strategy of RT-PCR and cloning; B-CLL specific IgVH transcripts were analysed for both % mutation and Ag-driven selection by applying a multinomial statistical model evaluating the excess/scarcity of replacement/silent mutations in FR/CDR sequences of IgVH genes. Maximally selected log-rank statistics, applied for IgVH gene % mutations, estimated the most appropriate cutoffs capable to separate B-CLL patients into two subgroups with different survival; this approach identified an absolute peak at 0.2% IgVH mutations and two relative peaks at about 2% and 4% IgVH mutations. We therefore tested the impact of Ag-driven selection on B-CLL patient survival in the context of the cutoffs of 0.2, 2 and 4% IgVH mutations. For each cutoff, B-CLL cases were identified as mutated (M) or unmutated (UM) if the % IgVH mutations were above or below the chosen cutoff, respectively; M B-CLLs with evidence of Ag selection were named significantly mutated (sM), while cases lacking such an evidence were reported as not-sM (nsM). 1) 0.2% IgVH cutoff - The 133 B-CLLs with 〉0.2% mutations had longer survivals as compared to 14 cases with 2% mutations, had longer survivals as compared to 48 UM cases with 4% mutations had similar survivals; 4) sM vs. nsM (all cases) - In the absence of any cutoff, the 78 patients with sM B-CLLs had longer survival as compared to the 69 affected by nsM B-CLLs (p=9.6x10-4). By taking together this data, it appears that the use of % IgVH mutations remains the gold-standard for the definition of prognosis in B-CLL; however, at least when the canonical 2% cutoff is applied, a certain caution should be used by predicting patient survival in the absence of information regarding the affinity maturation of B-CLL clones. The expression by B-CLL cells of IgVH segments with evidence of affinity maturation seems to lose its positive prognostic impact when a cutoff is set at 4% IgVH mutations. In addition to its obvious clinical impact, these observations provide a putative explanation of the apparent discrepancies found in the scientific literature about this matter.

Description: CD49d is a remarkable prognostic biomarker of chronic lymphocytic leukemia (CLL). The cutoff value for the extensively validated 30% of positive CLL cells is able to separate CLL patients into 2 subgroups with different prognoses, but it does not consider the pattern of CD49d expression. In the present study, we analyzed a cohort of 1630 CLL samples and identified the presence of ∼20% of CLL cases (n = 313) characterized by a bimodal expression of CD49d, that is, concomitant presence of a CD49d+ subpopulation and a CD49d− subpopulation. At variance with the highly stable CD49d expression observed in CLL patients with a homogeneous pattern of CD49d expression, CD49d bimodal CLL showed a higher level of variability in sequential samples, and an increase in the CD49d+ subpopulation over time after therapy. The CD49d+ subpopulation from CD49d bimodal CLL displayed higher levels of proliferation compared with the CD49d− cells; and was more highly represented in the bone marrow compared with peripheral blood (PB), and in PB CLL subsets expressing the CXCR4dim/CD5bright phenotype, known to be enriched in proliferative cells. From a clinical standpoint, CLL patients with CD49d bimodal expression, regardless of whether the CD49d+ subpopulation exceeded the 30% cutoff or not, experienced clinical behavior similar to CD49d+ CLL, both in chemoimmunotherapy (n = 1522) and in ibrutinib (n = 158) settings. Altogether, these results suggest that CD49d can drive disease progression in CLL, and that the pattern of CD49d expression should also be considered to improve the prognostic impact of this biomarker in CLL.

Description: Introduction: Although expression of β1 integrins was analyzed in B-cell chronic lymphocytic leukemia (B-CLL), little is known about the clinical behaviour and the phenotyic profile of B-CLLs with different patterns of β1 integrin expression. Methods: Expression of β1 integrins (α1-α6) was investigated by flow cytometry along with other 30 surface molecules (B-cell markers, cell-adhesion-molecules, activation inducers, complement activity regulators) in 155 B-CLLs, 106 characterized for clinical stage (Rai’s stages), 121 for IgVH mutations, 79 for ZAP-70 expression and 109 with survival data. Results: In agreement with previous studies, also in our B-CLL series CD49c, CD49d and, to a lesser extent, CD49e resulted the most expressed β1 integrins, with mean expression values, evaluated as % of positive cells, of 59%, 36% and 26% (range 1-100%), respectively. Since expression of CD49c and CD49d have been recently identified by us as part of the immunophenotypic signatures of B-CLLs with different prognosis (J Cell Physiol204, 113, 2005), the present study was designed to define: 1) the real prognostic impact and the optimal cutoffs for CD49c and CD49d splitting patients into two groups with different survivals; 2) the immunophenotypic and molecular features of B-CLL subsets expressing different levels of CD49c and CD49d. 1) By applying the maximally selected log-rank statistics, we identified cutoff values of 40% and 30% of positive cells for CD49c and CD49d, respectively; therefore, B-CLL cases were divided in CD49chigh or CD49clow and CD49dhigh or CD49dlow if the expression of the molecules was above or below their respective cutoffs. The CD49chigh B-CLL subset (n=78) displayed longer survival as compared to CD49clow cases (n=31; p=2.0x10-2, log-rank test); conversely, CD49dhigh B-CLLs (n=46) had shorter survivals, as compared to the CD49dlow subset (n=62; p=1.2x10-3). 2) Supervised analyses were performed by comparing the expression of a wide panel of surface antigens in CD49chigh vs. CD49clow and CD49dhigh vs. CD49dlow B-CLLs. In addition, B-CLL subgroups, as identified by CD49c and CD49d expression, were compared for IgVH mutations (2% cutoff), ZAP-70 expression (30% cutoff) and Rai’s stages distribution (stages 0-I vs. II-IV). According to these analyses, CD62L, CD22, CD55, CD29, CD11c, CD54 and CD40 were the molecules significantly over-expressed in good prognosis CD49chigh B-CLLs, as compared to CD49clow cases (p