Immunomodulators have been implicated in increasing the risk of EBV-driven B-cell non-Hodgkin's lymphomas (NHL) in inflammatory bowel disease (IBD) patients. Whether tumour-necrosis alpha inhibitors (anti-TNFs) can also propagate this lympho-proliferative transformation remains unsettled, but their use was associated with increased NHL risk when used in combination with thiopurines. The mechanisms underlying this lymphoproliferation are poorly understood. a NHL in-vitro model was established by co-incubation of EBV-infected-human peripheral-blood mononuclear-cells (PBMC) with Cyclosporine. The resultant lymphoblastoid cell-line (LCL) were characterised by flow-cytometry staining of surface CD markers, telomerase activity assay and by next-generation sequencing (NGS) of V,D,J immunoglobulin heavy-chains after 4 weeks of culture. In parallel, EBV-titers were measured weekly by qPCR. PBMC-EBV cultures were in parallel performed in the presence or absence of major therapeutic agents for IBD (anti-TNFs, anti-integrin vedolizumab, 6mp, methotrexate) for the entire duration of the four-weeks experiment or only during the three-days of initial exposure to EBV after which drugs were washed out of the culture. Incubation of PBMC with EBV and cyclosporine caused the appearance of LCL cells, which were characterised as an expanded population of CD58hiCD23hiCD19+ B-cells. This population uniquely presented an oligo-mono-clonal profile on NGS, as well as high telomerase activity, which was absent in cells exposed to EBV alone. LCL formation was accompanied by increase in intracellular EBV-viral copies until week three and extracellular titer until week two of incubation. When thiopurines and biologics were tested, the frequency of transformed EBV-induced LCLs was significantly higher following co-incubation with anti-TNFs (Infliximab: median 20.88%, IQR 11.84-20.88, vs. median 11.12%, IQR 4.65–18.51 for EBV alone; p = 0.033, n = 20; adalimumab: median 21.31%, IQR 7.9–25.27, vs. median 9.83%, IQR 4.56–18.18 for EBV alone; p = 0.01, n = 19), and with washed-out 6MP (6MP-washout: median 31.39%, IQR 28.97–45.54, vs. EBV alone: median 17.23%, IQR 14.03–35.39; p = 0.018, n = 7), but not with the anti-integrin vedolizumab (vedolizumab: median 9.76%, IQR 3.5–19.43, vs. EBV alone: median 9.83%, IQR 4.56–18.62; p = 0.295, n = 19). These results suggest that anti-TNFs and immunomodulators - but not vedolizumab - propagate EBV-driven lymphoblastoid transformation in vitro. This model may prove useful for studying mechanisms underlying NHL risk in immunosuppressed IBD patients.
Read full abstract