Exploiting the Gut Microbiota to Boost Immunotherapy and Switch the Trajectory of Multiple Myeloma

Abstract

Clonal proliferation of neoplastic plasma cells in the bone marrow (BM) and organ dysfunction characterizes multiple myeloma (MM), a treatable but incurable disease. Although asymptomatic smoldering MM (SMM) often precedes full blown MM, most patients affected by SMM are only offered active observation, which increases their frustration and anxiety. In transgenic Vk*MYC mice developing de novo MM, we found a direct link between gut microbiota composition and T helper 17 (Th17) lymphocytes, which migrated from the gut to the BM during asymptomatic MM (Early-MM), and favored the expansion of neoplastic plasma cells, thus fueling progression to symptomatic disease (Late-MM) [Calcinotto A. et al. Nat Commun (2018) 9(1):4832]. Similarly, in SMM patients, higher levels of BM Interleukin-17 (IL-17) predicted accelerated disease [Calcinotto A. et al. Nat Commun (2018) 9(1):4832]. Interestingly, Prevotella melaninogenica ( P.m.), a human commensal, limited MM aggressiveness in mice challenged with Vk*MYC-derived MM cells (t-Vk*MYC) by restraining Th17 cell expansion [Calcinotto et al. Nat Commun (2018) 9(1):4832]. Because the gut microbiota also contributes to the therapeutic efficacy of immune checkpoint blockade (ICB) and human neoplastic plasma cells express PD-L1, we hypothesized that modulation of the gut microbiota by P.m. could limit Th17 cell expansion in mice affected by Early-MM, thus fully exploiting the therapeutic potential of anti-PD-L1 against MM. Clinical trials combining anti-PD-1/PD-L1 antibodies, dexamethasone and immune modulatory drugs in MM were halted by US FDA because of unfavorable benefit-risk profile and life-threatening immune related adverse events (irAEs). Because IL-17 is involved in irAEs [Johnson et al. Cancer Immunol Res (2019) 7 (6): 860-865], we also hypothesized treatment with P.m. could restrain ICB-induced expansion of Th17 cells while unleashing CD8-mediated antitumor immunity. Early-MM Vk*MYC mice and t-Vk*MYC, the latter mimicking aggressive MM, were treated with P.m. and/or anti-PD-L1 and disease progression was monitored by paraprotein quantification in blood. Anemia provided clinical evidence of symptomatic disease. At sacrifice, gut, spleen and BM were analyzed by flow cytometry. In vitro assays were conducted to clarify the mechanism by which the gut microbiota influenced the induction of Th17 cells. To mimic cutaneous irAE, the skin of treated mice was exposed to imiquimod, which induces Th17-mediated inflammation [van der Fits et al. J Immunol (2009 )182(9): 5836-5845]. Body weight loss, skin inflammation, and inflammatory infiltrate into secondary lymphoid organs were taken as measure of imiquimod-induced toxicity. In Vk*MYC mice, all neoplastic plasma cells expressed PD-L1. Treatment of Early-MM Vk*MYC mice with anti-PD-L1 antibodies delayed progression to Late-MM, thus extending previous findings on the use of ICB in this model to anti-PD-L1 antibodies and Early-MM [Guillerey et al. J Clin Invest (2015 )125(5): 2077-2089]. Treatment with P.m. exerted a similar effect on disease progression, suggesting that a treatment as simple as taking a probiotic has an impact on MM evolution. When P.m. was combined with anti-PD-L1 antibodies, evolution to Late-MM was further delayed or even blocked. Single and combined treatments also reduced MM aggressiveness in t-Vk*MYC mice. Mechanistically, P.m. limited the expansion of gut-born Th17 cells and their accumulation in the BM of treated mice without dampening the antitumor cytotoxic response elicited by the ICB. Thus, only the combined treatment provided the most favorable CD8/Th17 and CD8/T regulatory cell ratios. Ex vivo, P.m.-conditioned BM DCs showed a limited propensity to produce Th17-polarizing cytokines. In vitro, stimulation of both human and mouse DCs with P.m. reduced polarization of naïve T cells toward a Th17 phenotype. Skin application of imiquimod in C57BL/6 mice induced psoriasis-like lesions and expansion of Th17 cells both in draining lymph nodes and spleen. The contextual administration of anti-PD-L1 antibodies worsened imiquimod-induced inflammation. An ongoing experiment will clarify if mice receiving P.m. and ICB experience reduced skin lesions. Taken together, our data support the development of microbiota-based strategies in combination with ICB to treat full-blown MM and to prevent progression of asymptomatic SMM to full-blown disease.