A Novel Germline PAX5 Single Exon Deletion in a Paediatric Patient with B-Cell Leukaemia

Abstract

Germline predisposition underlies leukaemia development in children in 5-10% of the cases. Over the past years, several somatic drivers for leukaemia, including IKZF1, RUNX1, ETV6 and PAX5, have also been identified as leukaemia predisposing genes. For example, two germline PAX5 missense variants (p.Arg38His and p.Gly183Ser) have been identified in five families with a high incidence of B-cell precursor acute lymphoblastic leukaemia (BCP-ALL). The susceptibility to BCP-ALL in these patients is inherited in an autosomal dominant pattern with an incomplete but relatively high penetrance. The leukemic cells in the patients showed loss of the wild-type allele or a somatic mutation, suggesting a second hit is required for leukaemia development in this context. We present a boy who developed BCP-ALL at 5-years-old, with a father who was diagnosed with acute undifferentiated leukaemia at infancy. Because of his family history, germline testing of the patient and his parents was performed using SNP-based copy number analysis and whole exome sequencing (WES). This revealed a PAX5 exon 6 deletion in the child, which was absent in both parents. The PAX5 deletion causes a frameshift with premature stop in exon 7, resulting in a truncated protein with loss of the C-terminal transactivating and inhibitory domains. In contrast to what has been observed in patients with germline PAX5 missense variants, a second somatic hit in PAX5 was not observed. Genotype analysis, using heterozygous SNPs around the PAX5 deletion, revealed that the PAX5 deletion had occurred on the paternal allele. To test whether this deletion was present in (low) mosaic state in the father, we developed a sensitive breakpoint spanning PCR. Also with this method, the PAX5 deletion was not detected in DNA derived from blood and buccal cells of the father. Since we expected a causative germline variant to be inherited from the father, we re-analysed all sequence and copy number variants in the patient-parents' trio datasets for any other aberrations, but no plausible alternative candidates were identified. Since routine diagnostic analysis of PAX5 deletions is usually performed only in leukaemia samples, germline deletions could potentially be missed, particularly in isolated cases. To investigate the incidence of germline PAX5 deletions, we used a multiplex-ligated probe amplification (MLPA) dataset from 2041 ALL patients, from which we selected 303 cases with single-exon or partial (multiple exon) PAX5 deletion. Subsequently, we performed MLPA on 261 DNA samples from bone marrow or peripheral blood at time of remission. No additional case was found, suggesting that germline PAX5 deletions, if recurrent, are very rare. In conclusion, we present a boy with ALL carrying a de novo germline PAX5 exon 6 deletion leading to a truncated protein. Germline PAX5 deletions have not yet been described in ALL patients but since somatic PAX5 aberrations are common in ALL and germline PAX5 variants have been associated with ALL, we consider that this focal germline PAX5 deletion did contribute to leukaemia development in the boy.