TNFRSF13C (BAFFR) positive blasts persist after early treatment and at relapse in childhood B‐cell precursor acute lymphoblastic leukaemia

Tumour necrosis factor (TNF) superfamily member 13b [TNFSF13B, also termed B cell-activating factor (BAFF)] belongs to the TNF family (Daridon et al, 2008). Physiologically, TNFSF13B mediates the behaviour of B cells through interactions with its family receptors (Mackay & Browning, 2002). However, only the TNFSF13B receptor (TNFRSF13C, also termed BAFF-R) interacts specifically with TNFSF13B, being mainly responsible for primary B cell survival, selection and differentiation, in physiological and disease conditions (Rolink & Melchers, 2002) (Cancro, 2008) (Tussiwand et al, 2012). Moreover, a role for TNFSF13B/TNFRSF13C has been proposed in autoimmune diseases (Vincent et al, 2013) and in graft-versus host disease (Sarantopoulos et al, 2009). Importantly, a previous study showed that recombinant TNFSF13B supported the survival of acute lymphoblastic leukaemia (ALL) cells in the absence of stroma, and significantly attenuated the apoptosis rate caused by nilotinib, a drug used to treat Philadelphia chromosome-positive ALL (Parameswaran et al, 2010). However, few data are available regarding a direct role of the TNFSF13B/TNFRSF13C axis in haematological malignancies (Tangye et al, 2006). Herein, we aimed to investigate the potential role of TNFSF13B/TNFRSF13C axis in B-Cell Precursor (BCP) ALL. We analysed primary bone marrow (BM) and peripheral blood (PB) samples from children affected by BCP-ALL, at diagnosis, follow-up and relapse, as well as haematological tumour cell lines. Details of the methods are described in the Data S1. We demonstrated that TNFRSF13C was highly expressed in BCP-ALL cell lines (MUTZ-5 and CALL-4), in healthy donor BM and PB monoclonal cells (RNA commercial library, Clontech, Takara Bio Europe, Saint-Germain-enLaye, France) (Figure S1) and in BCP-ALL diagnostic samples, with a wide range of expression (Fig 1). TNFRSF13C expression was also detectable at lower levels in mixed lymphoid/myeloid phenotype cell lines (such as THP1 and RS4;11), in myeloid K562 cells and in U937 histiocytic lymphoma cell lines (Figure S1). Supported by these data, we further investigated TNFRSF13C protein expression in diagnostic and short-term follow-up samples from paediatric BCP-ALL patients, enrolled in the Associazione Italiana di Ematologia e Oncologia Pediatrica-Berlin-Frankf€ urt-M€ unster (AIEOP-BFM) ALL2009 protocol in our centre. Twenty-six consecutive diagnostic BM and/or PB samples, in addition to at least one follow-up sample (i.e., PB at day +8 or PB/BM at day +15). TNFRSF13C expression was analysed by flow cytometry, using a biotinilated anti-TNFRSF13C antibody revealed by phycoerythrin (PE)-conjugated streptavidin antibody. In the same sample, we used hCD19 (fluorescein isothiocyanate), hCD10 (allophycocyanin) and hCD45 (peridinin chlorophyll) direct staining to recognize leukaemic cells (CD45 expression on CD10+/CD19+ cells) among the residual normal cells (CD45 expression on CD10 / CD19+ cells). Interestingly, high levels of TNFRSF13C were detected on CD19+/CD10+/CD45 leukaemic cells in diagnostic samples (mean of 18 96% 22 68 and 32 20% 25 31, in n = 25 BM and n = 26 PB, respectively). Overall data are reported in Fig 2A, while Table S1 reports detailed results of fluorescence-activated cell sorting analyses; a representative phenotype is shown in Figure S2. Importantly, TNFRSF13C protein persisted during the follow-up treatment. At day +8, the residual tumour burden was very low due to pharmacological treatment by prednisone or dexamethasone, depending on randomization in the current protocol. At day+8, we detected a statistically