Dottorato in INNOVATION IN IMMUNO-MEDIATED AND HEMATOLOGICAL DISORDERS

Titolo della tesi: Project 1 - Dissecting the role of acquired regions of homozygosity detected by microarray along the genome of a large cohort of adult patients with B-cell precursor acute lymphoblastic leukemia/ Project 2- Unveiling the genetic landscape of B-cell precursor acute lymphoblastic leukemia by novel technologies: optical genome mapping versus digital multiplex ligation dependent-probe amplification

Project 1: Dissecting the role of acquired regions of homozygosity detected by microarray along the genome of a large cohort of adult patients with B-cell precursor acute lymphoblastic leukemia INTRODUCTION Acquired regions of homozygosity (aROHs) detected along the genome are due to mitotic recombination between homologous chromosomes. This mechanism may contribute to the growth of a neoplastic clone if the involved region embraces tumor suppressor or onco-genes harbouring a mutation, with subsequent loss of the wild-type allele and duplication of the aberrant one. Cytogenetic diagnostic tools based on allelic load assay has allowed identification of aROHs also in B-cell acute lymphoblastic leukemia (B-ALL). However, real-life data about contribution of these aberrations to leukemogenesis are scarce and controversial. AIM We investigated aROHs detected in a large and uniform longitudinal cohort of adult B-ALL patients (pts), to define their potential role. METHODS AND RESULTS The study was carried out at Josep Carreras Leukemia Research Institute (Badalona, ES). Pts enrolled in clinical trial NCT04179929 with available single nucleotide polymorphisms (SNP)-array analysis performed at baseline for genetic risk-stratification as per protocol were considered evaluable. For each patient, DNA extracted from infiltrated bone marrow (BM) and amplified by PCR was hybridized to GeneChip Mapping 750 K array (Affymetrix, Santa Clara, US). Hybridization to each probe was assessed using a GeneChip Scanner (Affymetrix) and results scored using CHAS v4.5 software (Affymetrix). For the aim of the study, we first selected ROHs with size≥3 Mb and mapping in regions covered with≥20 consecutive probes. In order to identify only acquired non-polymorphic forms, we excluded ROHs overlapping ≥50% of their length with regions included in public databases of healthy controls and with variant of allele frequency (VAF)>BM infiltration. Finally, eligible calls were cross-referenced with pts' biological and clinical data for statistical analysis. From December 2019 to March 2024, 371 pts were included in LAL-19 trial. 277 cases (75%) were eligible for our study. Median age was 40 years (range 18-60). The male/female ratio was 1:1. Overall, 332 aROHs were detected in 156 pts (56%). The median number of aROHs per pt was one (range 1-9); the median size was 6.7 Mb (range 3-27.5). 94 % of aROHs were segmental. aROHs were distributed across the whole genome, 5q31.1 and 7q11.23 being the most frequently involved regions, without affecting any gene clearly involved in the leukemogenesis. No difference was observed between pts with and without aROHs in terms of age, sex and clinical features at the onset. However, the presence of aROHs was associated with the absence of hypodiploidy (<44 chromosomes) and absence of mutations in the survival regulatory pathways detected by NGS standard panel (p=0.029 and p=0.043, respectively). Based on follow-up (FU) data, survival analysis was possible in 192/277 pts (69%). After a median FU of one year, the presence of aROHs had no impact on either overall survival (OS) (p=0.594) or cumulative incidence of relapse (CIR) (p=0.893). Within the population with aROHs, no difference emerged in terms of OS and CIR stratifying by number (1 versus ˃1) and size of lesions (<20 Mb versus >20 Mb, being 20 Mb the most employed cut-off in literature). CONCLUSION Our study confirms recurrence of non-polymorphic aROHs along the genome of adult B-ALL pts, without affecting key-genes for leukemogenesis. These aberrations do not affect the prognosis. On the contrary, a protective role for the genome could be hypothesized, due to their association with a low rate of mutation and with preserved diploidy. Project 2: Unveiling the genetic landscape of B-cell precursor acute lymphoblastic leukemia by novel technologies: optical genome mapping versus digital multiplex ligation dependent-probe amplification INTRODUCTION Adult B-cell acute lymphoblastic leukemia (B-ALL) is a heterogeneous hematologic malignancy, characterized by progressive accumulation of several genomic lesions. About 10-30% of cases remain unclassifiable by routine cytogenetic tests (RT) (i.e. chromosomal banding analysis, fluorescence in situ hybridization, chromosomal microarrays). Therefore, novel technologies are required for a more extensive assessment of B-ALL genetic background, in order to improve risk-stratification and to personalize treatment. Optical genome mapping (OGM) is a high-resolution genome-wide tool revealing structural (SVs) and copy-number variations (CNVs) by fluorescent labelling and digital imaging of ultra-high molecular weight (UHMW) DNA molecules. Digital multiplex ligation dependent-probe amplification (dMLPA) combines standard MLPA with next generation sequencing for detection of CNVs and gene fusions due to interstitial deletions (dels). AIM The project aims to evaluate if OGM and dMLPA could integrate RT to improve the cytogenetic profiling of adult B-ALL. METHODS AND RESULTS For the purpose of the study, we selected a cohort of adult B-ALL patients (pts) with sufficient leftover material after RT at baseline. OGM was performed (in Spain) following manufacturer’s instructions (BionanoGenomics, San Diego, US): UHMW-DNA was extracted, labelled with DLE-1 enzyme, loaded onto a chip and run by Saphyr instrument. Data analysis was carried out through rare variant analysis algorithm on BionanoSolve software, using GRCh37/h19 as genome reference. SVs and CNVs were considered if overlapping with regions included in open-access BED files for B-ALL or if ≥ 100 kb. The dMLPA experiments were performed (in Italy), in accordance with the manufacturer’s instructions, using D007 ALL probemix (MRC Holland, Amsterdam, NL). The sample-specific products from several reactions were pooled and loaded on an Illumina MiSeq V3 flow-cell. The data were analyzed using bioinformatics software by MRC, Holland. Twenty cases were collected from May 2023 to July 2024. Median age was 50 years (range 18-70). The male/female ratio was 1:1. Based on RT, they were classified according to ICC 2022 as follows: NOS (not otherwise specified), 10 pts; Ph-like, 6 pts; KMT2A-rearranged (-r), 1 pt; PBX1-r, 1 pt; low-hypodiploidy, 1 pt; PAX5-mutated (-mut), 1 pt. OGM detected all previously known CNVs and SVs, except for two CRLF2-r. Furthermore, it detected driver SVs missed by RT with subsequent re-stratification of four cases: ZNF384-r in 2 NOS pts; MEF2D-r in 1 NOS pt and EPOR-r in PAX5-mut pt. dMLPA detected all CNVs and only SVs due to dels, included 1 CRLF2-r missed by OGM. Moreover, it detected with greater accuracy CNVs affecting genes with prognostic impact but no effect on classification (i.e. EBF1, IKZF1, CDKN2A/B, PAX5, BTG1, TCF3 and VPREB1). Overall, the two techniques provided useful information for more accurate cytogenetic diagnosis and risk-stratification in 11 pts (55%). CONCLUSION To the best of our knowledge, this is the first study to integrate simultaneously OGM and dMLPA in the cytogenetic diagnosis of adult B-ALL. Both techniques outperformed RT for the detection of SVs and CNVs, respectively, showing high sensitivity and specificity, in accordance with the literature data. In particular, they proved useful for a more in-depth characterization of challenging categories, such as NOS and Ph-like. Greater experience is required to clarify the correct placement of OGM and dMLPA in the cytogenetic diagnostic algorithm of adult B-ALL.