The spleen in primary immune thrombocytopenia (ITP) is not only the major site of platelet destruction but is also the main site of auto-antibody production. Accordingly, splenectomy as a second-line treatment has resulted in a durable platelet response in two-thirds of the patients. New strategy, using anti-CD20-induced B-cell depletion (Rituximab, RTX), leads to 40–50% of short term and 20% of long-lasting response. Less is known about B-cell depletion in spleen and the nature of the resistant pathogenic cells. In this study, we analyzed at the population and single cell level the splenic antibody secreting cells, in healthy donors and active chronic ITP patients, treated or not with RTX.
We took advantage of these different therapeutic outcomes to analyze the splenic B cell compartment of ITP patients, either not treated with Rituximab (n=5), or treated (n=10) and presenting with a primary failure of the B-cell depletion therapy, and compared them with healthy donors (n=5). Plasma cells (that lack CD20 expression) were the major residual population in spleens from Rituximab-treated patients (0.5% of CD45+ spleen mononuclear cells; range 0.06–1%). A fraction of these plasma cells were autoreactive (0.5–1.2% of IgG secreting cells being anti-GpIIbIIIa specific), thus accounting for the persistence of the disease and the failure of RTX treatment. Using gene profiling and single cell multiplex quantitative RT-PCR (Fluidigm dynamic arrays), we showed that such Rituximab-resistant plasma cells display a unique long-lived transcriptional program, similar to the one of bone marrow plasma cells. In contrast, splenic PC from healthy donors and from untreated ITP patients displayed an intermediate gene expression profile between the ones of long-lived PC and proliferating plasmablasts, with the simultaneous expression of cell cycle genes and survival/anti-apoptotic factors. BAFF secretion was increased two-fold in RTX-spleen supernatants compared to ITP (P<0.05), and might selectively contribute to the survival of the resident RTX-PC population.
Our results demonstrate that residual long-lived plasma cells, with a few percents of them secreting anti-GpIIbIIIa antibodies, are found in the spleen 3–6 months following RTX treatment, thus providing a possible explanation for the failure of therapy, and the success of subsequent splenectomy. As opposed to accepted schemes, our data suggest that the normal human spleen is not a niche for long-lived PC, and that auto-immune/inflammatory conditions do not create either a favorable environment for their persistence. We propose in contrast that it is the B-cell depletion itself and the specific cytokinic condition it provides that generates suitable environment for the maturation of long-lived plasma cells in the spleen. This observation suggests that interfering with the differentiation and/or survival of this RTX-induced long-lived plasma cell population could obviate the need for splenectomy in the case of RTX-treatment failure.
Off Label Use: rituximab.
Asterisk with author names denotes non-ASH members.