Familial Predisposition to B-Cell Precursor Acute Lymphoblastic Leukemia Mediated By PAX5 Germline Variants

Abstract

Introduction: Somatic PAX5 alterations are one of the most common features present in B-cell precursor acute lymphoblastic leukemia (BCP-ALL). Recently, PAX5alt and PAX5 p.P80R have been identified as two novel PAX5-driven BCP-ALL subtypes (Gu et al., Nat Genet., 2019). PAX5 is a well-known master regulator for B-cell development, hence, deregulated PAX5 expression poses high risks for leukemia development. In this regard, families with germline PAX5 variants are a unique source to study the effect of PAX5 on leukemia development in humans. Methods: We corroborated clinical data with immunophenotyping and genomic analyses of families carrying PAX5 germline variants. Results: Here, we describe two new families with inherited PAX5 germline variants affecting amino acid 183 who developed BCP-ALL, including one family with a so far undescribed amino acid substitution p.Gly183Arg. To identify additional synergizing germline variants, whole exome sequencing of both families, together with one family previously reported by Auer et al., (Leukemia, 2014) was carried out. This analysis yielded no common germline variants apart from PAX5, which could explain the incomplete penetrance in the observed leukemia development, but confirmed the families’ diverse backgrounds in different ancestry. Interestingly, an in-depths B-cell differentiation study comparing PBMCs from PAX5 variant carriers (n=8) in the respective families and unrelated non-carriers (n=44) revealed a significant decrease of intermittent B-cells (Student's T-test, 5.796% vs 9.820%, p= 0.017) and memory B-cells (Mann-Whitney U, 10.760% vs 17.350%, p= 0.003) in carriers versus non-carriers. Moreover, bone marrow analysis from one of the patients carrying PAX5 p.G183S showed a reduction in the number of mature B cells, suggesting deregulation of the B-cell differentiation at the pre-B stage. Altogether, a total of 21 patients from seven families carrying three different heterozygous germline PAX5 variants have been reported. Statistically significant differences are observed between the age of onset of patients harboring the PAX5 p.Arg38His variant compared with patients carrying p.Gly183Ser/Arg (Student's T-test, 17.75 vs. 2.70 years old, respectively; p=0.03). Furthermore, all patients who developed BCP-ALL showed somatic chromosome 9p alterations with loss of heterozygosity of the Wildtype PAX5 locus. Concordantly, all families display incomplete penetrance, with PAX5 p.Gly183Arg and p.Arg38His being by far more penetrant than p.Gly183Ser. Regarding the outcome, a high rate of overall survival (72%) and a high rate of relapse (75%) were observed. 54% of the patients received hematopoietic stem cell transplantation, and of these, 71% survived. Conclusion: Taken together, our data provide evidence for deregulated B-cell development mediated by reduced PAX5 levels in the germline of families carrying PAX5 germline variants. Our results emphasize the benefit of PAX5 germline variant screening in patients diagnosed with BCP-ALL and somatic alterations on chromosome 9p, regardless of a family history of leukemia.