The Broad Spectrum Of Tp53 Variants In Cll: Ngs Analysis Of 573 Pathogenic Tp53 Variants

TP53 aberrations, including somatic mutations of TP53 gene or 17p deletion leading to the loss of the TP53 locus, are a major predictive factor of resistance to fludarabin based chemotherapy in chronic lymphocytic leukemia (CLL) and remain an adverse prognostic factor in the chemofree era. Therefore, detection of TP53 alteration before each new line of treatment is required for theranostic stratification. In order to better characterize the distribution and combination of the TP53 variants in CLL, we collected the TP53 sequencing data of 343 patients harboring TP53 mutations from centers of the French Innovative Leukemia Organization-CLL (FILO) and established a large data base of 573 TP53 mutations. Mutations were identified through NGS sequencing (covering exon 2 to 11) allowing the detection of low frequency variants down to 1% VAF. Several distinct low VAF mutations were orthogonally confirmed by digital PCR. TP53 variants were analyzed through UMD_TP53 data gathering 90 000 TP53 mutations from all type of cancers. IGHV mutational status and FISH analysis were available for 224 and 176 patients respectively. Using ACMG criteria from the UMD_TP53 database, we confirmed that 523 could be classified as pathogenic, 42 were likely pathogenic and 8 were VUS (Variants of Unknown Significance). As expected, the mutation distribution along the p53 protein exhibited a clustering of variants in the DNA binding domain of the protein. We also confirmed the presence of a specific hotspot at codon 234 (6%) which is noticeable in other CLL cohorts but absent in solid tumors. 431 TP53 variants led to the expression of a mutant protein whereas the remaining 142 led a TP53 null phenotype. For 8 patients without 17p deletion and a mutation VAF larger than 50%, SNP analysis indicate that these tumors had a copy number neutral loss of heterozygosis at 17p with a duplication of the mutant allele leading to homozygous mutations of TP53. When focusing on the allele burden of TP53 mutations, 264/573 (46%) variants had an allele frequency <10%. Even if they were predominantly found in polymutated cases, presence of only low VAF (<10%) mutations was evidenced in 74 (21%) patients (50 patients with a single TP53 mutation and 24 patients with more than one). All these cases would have been missed by conventional sequencing. Among the 343 patients, 113 (33%) were poly-mutated and harbored more than one pathogenic TP53 variants (2 to 11 variants per patient): 57 (16,7 %) had 2 variants, 32 (9,3%) had 3, 10 had 4 (3%) and 14 patients (4%) had 5 to 11 variants. Using both long range sequencing and in silico analysis, we could show that all these variants were distributed in different alleles supporting an important intratumoral heterogeneity and a strong selection for TP53 loss of function during tumor progression in these patients. Null variants were rarely found as single alteration: only 46 patients (13,4%) patients harbored a single null mutation. Null mutations were predominantly found in patients with multiclonal mutations (87% with 4 or more). Median size of variants significantly decreased with the number of mutations and most of low VAF (less than 10%) variants were found in multiclonal combinations. Multiclonal mutations were predominantly found in previously treated patients (41% treated versus 10 % untreated) but whether all these variants preceded treatment and were further selected is currently unknown. We observed that 71,5 % of patients were IGHV unmutated and multiclonal mutations were surprisingly more frequent in mutated IGHV cases than in unmutated ones. Only 50% of cases carried a 17p deletion, highlighting again the importance of testing for TP53 mutations in addition to FISH analysis. Presence or absence of 17p deletion was unrelated to the number of TP53 mutations. Taken together these observations suggest that the TP53 mutational landscape in CLL is very complex and can involve multiple mechanisms, converging to a total loss of TP53 function and tumor progression. NGS provides a powerful tool for detecting all these alterations including variants with low VAF and should become a standard for CLL screening prior to each line of treatment. Disclosures Leblond: Amgen: Honoraria, Speakers Bureau; Abbvie: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Janssen: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Gilead: Honoraria, Speakers Bureau; Astra Zeneca: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Roche: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau. Letestu:Abbvie: Membership on an entity's Board of Directors or advisory committees, Other: speaker fee, expert contracts; Janssen: Membership on an entity's Board of Directors or advisory committees, Other: speaker fee, expert contracts; Roche: Membership on an entity's Board of Directors or advisory committees, Other: speaker fee, expert contracts; Alexion: Membership on an entity's Board of Directors or advisory committees, Other: speaker fee, expert contracts. Cymbalista:Abbvie: Honoraria; Roche: Research Funding; Sunesis: Research Funding; Gilead: Honoraria; Janssen: Honoraria; AstraZeneca: Honoraria.