Genotyping of Classical Hodgkin Lymphoma on the Liquid Biopsy

Abstract

Background: In classical Hodgkin lymphoma (cHL) the low representation (1-5%) of Reed-Sternberg cells (RS) challenged tumor genotyping on the diagnostic tissue biopsy. Consistently, the mutational profile of newly diagnosed cHL is poorly characterized, and the genetics of refractory disease is completely unknown. Cell free DNA (cfDNA) is shed into the blood by tumor cells undergoing apoptosis and can be used as source of tumor DNA for the identification of somatic mutations. In addition cfDNA is representative of the entire tumor heterogeneity, thus allowing the identification of mutations from tumor cells residing in non-biopsied sites. Aims: This study aims: i) at providing the evidence that the mutational profile of cHL can be tracked by using plasma cfDNA; and ii) at characterizing the genetics of newly diagnosed cHL and, for comparative purposes, of refractory cHL. Methods: The study incudes 28 newly diagnosed cHL and 9 chemorefractory cHL. All cases were provided with cfDNA from plasma collected at baseline, before treatment start, and paired DNA from granulocytes as source of germline DNA to filter out polymorphisms and sequencing noise. Paired genomic DNA from formalin fixed paraffin embedded (FFPE) tumor tissue biopsies was available for 17 patients, including 3 cases for which RS enriched areas were macrodissected. A targeted resequencing panel optimized to include the coding exons and splice sites of 77 genes (192Kb) that are recurrently mutated in B-cell lymphomas was used for genotyping. Libraries were prepared from plasma cfDNA, germline gDNA and tumor gDNA according to the CAPP-seq targeted enrichment strategy (Nimblegen) and subjected to ultra-deep-next generation sequencing (NGS) on the MiSeq platform (Illumina). The sequencing was tailored to obtain a depth of coverage >2000x in >80%of the target region in all samples, which allowed a sensitivity of 3x10-3. The somatic function of VarScan2 was used to call non-synonymous somatic mutations, and a stringent bioinformatic pipeline was applied to suppress tha background noise and to filter out sequencing errors. Results: In newly diagnosed cHL patients, genotyping of plasma cfDNA identified non-synonymous somatic mutations in STAT6 (43%), TNFAIP3 (43%), ITPKB (32%) B2M (21%), GNA13 (14%), CIITA (7%), XPO1 (7%) and CD58 (4%) among the most recurrently affected genes (Figure 1A-B). In refractory cHL patients, genotyping of plasma cfDNA identified non-synonymous somatic mutations in ITPKB (44%), TNFAIP3 (33%), KMT2D (33%), B2M (33%), GNA13 (33%), XPO1 (22%), TET2 (22%), IKBKB (22%), BIRC3 (22%) and STAT6 (22%) among the most recurrently affected genes. Mutations of KMT2D (33%) and TET2 (22%) were enriched in refractory cHL patients compared to newly diagnosed cases, suggesting that they contributed to the chemorefractory phenotype (Figure 1C-D). By using highly sensitivity techniques, most of the mutations discovered in cfDNA were also identified in pair tumor DNA from the tissue biopsy and/or macrodissected RS cells, thus confirming their tumor origin (Figure 1F). By pathway analysis, the mutational profile pointed to the involvement of PI3K/AKT signaling, cytokines signaling, NF-kB signaling and the immune escape in cHL. ITPKB (a negative regulator of the PI3K/AKT signaling pathway) was specifically mutated in cHL across aggressive B cell lymphomas.(Figure presented) Summary/Conclusions: This study provides the evidence that cHL can be genotyped using plasma cfDNA as source of tumor DNA, pointed to a non overlapping genotype between newly diagnosed and refractory cases, and identified ITPKB as a new gene specifically involved in ∼30-50%of cHL patients.