Introduction: Recent studies have highlighted that about 5% of children with Acute Lymphoblastic Leukemia (ALL) carry a germline mutation that predispose to leukemogenesis. PAX5 somatic mutations are frequently identified on sporadic B-ALL, but recently pathogenic PAX5 germline variants have been described in families with recurrence of B-ALL (Shah et al. 2013; Auer et al. 2014; Yazdanparast et al. 2015, Duployez et al. 2021). We describe the case of two siblings both affected by B-ALL with evidence of a new familial PAX5 variant. Case A was a male, 13 years old, diagnosed with B-II ALL. He was enrolled into the AIEOP-BFM ALL 2009 treatment protocol and according to minimal residual disease (MRD) levels he was stratified as medium risk. Eight years later, his sister (14yrs, named as case B) was diagnosed with B-II ALL, and stratified in the medium risk group of AIEOP-BFM ALL 2017 protocol. The two patients were otherwise healthy and borne from healthy unrelated parents; no history of hematological malignancies was reported across the pedigree. Methods: We performed a custom Next Generation Sequencing ALL predisposition panel on DNA extracted from bone marrow samples of both disease onset and remission phase (defined as negativity by PCR-MRD, 10-4 sensitivity). The libraries were performed following Nextera Flex for Enrichment and sequenced on Illumina Nextseq550 platform; the analyses were carried out by Congenica software. The variants of interest were validated and investigated in the family pedigree through PCR and Sanger sequencing. Digital-MLPA testing 55 ALL hot spot genes (including all PAX5 exons) was performed. Results: Through the NGS analysis, we showed that both patients shared a novel germline heterozygous PAX5 variant (c.548delG, p.Gly183AlafsTer84); moreover, the disease samples of both patients carried an additional somatic PAX5 variant (c.239C>G, p.Pro80Arg; variant fraction: 29% in Case A and 38% in Case B). The Digital-MLPA using the ALL panel, excluded the presence of shared copy number alterations (CNV) and, specifically, did not identified CNV alterations involving the PAX5 locus. The familiar segregation showed a paternal origin of the germline PAX5 variant (p.Gly183AlafsTer84); the mother was found to be wild-type. The PAX5 variant was also found in the DNA sample of the patients' grandmother of the paternal branch; a healthy brother of probands (24 years old) was not tested. To investigate a possible impairment of PAX5 function in the development of the B-cell lineage, we evaluated in both parents the B-cell repertoire on peripheral blood fresh mononuclear cells by FACS analysis. Interestingly, the father has a significant reduction on mature IgD+/CD27+ B-cells with a concomitant accumulation of IgD+/CD27- naïve B-cells. Conclusions: We are reporting a novel germline PAX5 p.Gly183AlafsTer84 in a family with recurrence of B-ALL. In contrast to the previous PAX5 described cases that carry an inherited missense variant resulting in a modest attenuation of PAX5 activity, our patients present a germline frameshift variant that brings to premature stop of the protein, suggesting a more destructive impact on PAX5 activity. As in the other reported cases, our patients presented an additional somatic event on PAX5 that further compromise PAX5 function, leading to the establishment of the leukemic clone. Interestingly and different from the literature, the two siblings share the same PAX5 p.Pro80Arg somatic variant, that has been characterized by Gu as a pathogenic variant impairing B lymphoid development and promoting leukemogenesis (Gu et al. 2019). The detailed molecular characterization, including Whole Transcriptome NGS analysis, is ongoing to investigate whether a PAX5-specific signature can be identified. Moreover, a single-cell NGS approach on remission samples of the two siblings could help to identify specific important pathways and altered gene suppression profiles, preceding full leukemia. Consent-dependent characterization of additional family members are ongoing to assess the penetrance of leukemia associated to this new variant and possible additional features. The identification and characterization of germline PAX5 variants is of high interest in the context of leukemia predisposition because of its clinical implications in terms of familial genetic counselling and because of its contribution in the understanding ALL pathogenesis.