Ps931 adult triple negative acute lymphoblastic leukemia fusion detection: challenging aim for ph- alternative therapies

Background: The adult “triple negative” [Ph‐/‐/‐; neg for t(9;22); t(1;19); t(4;11); 61% of adult B‐ALL (Roberts KG et al., J Clin Oncol. 2017)] B‐ALL group is characterized by an high genetic heterogeneity, lack of innovative target therapies and poor survival. Many chimeric genes have been recently identified in a very large international cohort of pts (Gu Z, Nat Genet. 2019 Jan 14), and led to a refined classification of B‐ALL and to, in some cases, tailored therapies; but identified them (also Ph‐like) is challenging, cause to genetic complexity, low frequency, to not unique worldwide guidelines and to expensive algorithms that require multi‐NGS approaches. Aims: To identify the novel subtypes Ph‐/‐/‐ of B‐ALL, or to identify not previously described fusions, in order to assess know or unknown fused targetable biomarkers. Methods: we performed integrative genomic analysis including 1385 RNAseq gene Panel (illumine) and/or transcriptome sequencing (RNA‐seq) of 52 Ph‐/‐/‐ cases from adult cooperative group studies. Also Copy Number Alterations by SNP Arrays and/or dMLPA‐MRC‐Holland were performed. FISH and PCR validation were performed when feasible. Results: We identify fusion genes, not otherwise detected, in 21 patients (pts) with a fusions rate of 43% (23/52), denoting that Ph‐/‐/‐ are not deeply characterized. In six cases, we found 2 fusions/pt and in three of them we capture also the reciprocal fusions. Many of these fusions were previously described in B‐ALL (i.e. ZNF384‐TCF3, ZNF384‐EP300, TAF15‐ZNF384, MEFD2‐BCL9 , KMT2A/MLLT1, ABL1/RCSD1 and PAX5/ETV6 ) or in other diseases (i.e. NONO/TFE3 in Renal cell carcinoma). In 4 cases we analyzed patients at the time of diagnosis and relapse. In 2 pts no fusions were detected at both time point, in one case we found a fusion a the diagnosis (BAGE3–2/KMT2C) and in one case we have different fusion a the two different time points. At the diagnosis, we found THADA1‐CDH1 and at relapse, we detect TET3/ETV6 fusion. Notably, the TET3/ETV6 fusion was found also at the second relapse (FISH confirmation) after the second relapse. None of these 2 fusions were previously described. Ph‐/‐/‐ fusion detection help to sub‐classify our patients in new B‐ALL subgroups. We found a) Rare MLL fusions 3,8% (2/52); b) MEF2D r: 1,9% (1/52); c) ZNF384 r: 11,5% (6/52); d) probable RUNX1‐ETV6 like: 3,8% (2/52); e) probable Ph‐like fusions: 7,7% (4/52). Some fusion detection validations were very cryptic with conventional methods cause of fusions combined with deletion or amplification of one or both of gene partners or cause of the low frequency of the clone in a post transplanted patients or cause of region of homologue sequencies. Summary/Conclusion: we identified an high rate of secondary fusions in adult Ph‐/‐/‐ B‐ALL patients (43%) that are difficult to well characterized with conventional diagnostic methods like Karyotype and FISH. The use of an NGS approach permit us to easily detect fusions useful for a better classification and in some cases to find targetable fusions (i.e. ABL1/RCSD1 ), to give these poor outcome patients an alternative therapie. Supported by: AIL, FP7 NGS‐PTL project, Harmony.