Richter's syndrome (RS) represents one of the foremost clinical challenges in CLL, due to the disappointing responses to both standard chemotherapy and novel agents. Checkpoint blockade (CPB) with anti-PD-1 antibody treatment has shown promise in RS (but minimal efficacy in CLL), suggesting that distinct immuno-genetic features underlie differential response to immunotherapy of these two entities. We previously generated CRISPR-based multiplexed gene-edited mouse models of RS in immune competent backgrounds with genetic, epigenetic, transcriptional and functional similarity with human disease, thus enabling interrogation of changes in tumor microenvironmental (TME) features underlying CLL transformation. To this end, we analyzed the T-cell and monocyte/macrophage composition of 5 CLL, 3 CLL/RS (i.e. cases with equal co-existence of CLL and RS), and 9 murine RS cases by multi-parametric flow cytometry of splenocyte preparations, and compared them with 5 age-matched wild-type controls (15-18 month old). Whereas the overall proportional abundance of CD3+ T cells was not significantly changed across groups, we detected decreased CD4+ and increased CD8+ T cell populations in RS cases (P≤0.04, ANOVA). Both CD4+ and CD8+ T cells showed a prominent skewing toward the effector memory subset (CD44+ CD62L-), when compared to naïve (CD44- CD62L+) and central memory (CD44+ CD62L+). CD4+ and CD8+ T cell populations from RS cases also showed significant upregulation of exhaustion markers including PD-1, TIGIT, TIM3 and LAG3 (P≤0.05). The monocytic compartment was similarly skewed, with RS-associated enrichment in non-classical monocytes (CD43+ LyC6med/low, known to have lower antigen presentation capacity), when compared to classical inflammatory CD43+ LyC6high cells (P=0.04). Importantly, we measured increased abundance of CD11b+ Ly6C- tumor associated macrophages in RS cases (P=0.02), which expressed significantly reduced MHC class II (P=0.003), and increased PD-1 surface levels (P=0.02). Further, in 11 cases (3 CLL/RS, 8 RS) where at least 5-10% residual CLL cells co-existed with RS within the same animal, and the two populations were discernable by size, we measured increased expression levels of both PD-1 and PD-L1 ligands on RS compared to CLL by flow cytometry (P≤0.001, paired t test), consistent with surface PD-1 observed in human clonally related RS. Interestingly, surface PD-L1 levels were particularly enriched in RS cases carrying clonal deletion in Mga, a negative regulator of MYC (P=0.01, Mann Whitney), in line with the known role of MYC in upregulating PD-L1 gene expression. In light of these changes, we asked whether our animal models could be utilized as a reliable platform to study differential response to CPB. To this end, we leveraged transplantable models in immune competent mice. In our initial study, 40 million RS cells from a Trp53-deleted primary tumor were transplanted into 20 CD45.1 recipients (2 million/recipient), which were randomized either 1 [median disease burden: 2%(1-13.5)] or 2 weeks [37%(7-54)] post-transplant through peripheral blood (PB) flow cytometric assessment of circulating disease. Treatment with 200 micrograms rat IgG2a anti-PD-1 or isotype control (n=5 mice/group) was performed for 3 weeks, every 3 days, and animals euthanized 3 days after the last dose. Whereas animals with highest disease burden at treatment start (2 weeks post-transplant) showed minimal, if any, efficacy of CPB treatment, animals with lower disease burden at time of randomization (1 week post-transplant) showed significant reduction in RS infiltration of PB, spleen and bone marrow at euthanasia (P≤0.04, Mann Whitney). This suggests that disease burden may impact differential response to CPB. In conclusion, our analyses demonstrate marked exhaustion of the T cell and monocytic compartment and increased expression of PD-1 and PD-L1 on tumor cells and TME components upon RS transformation. Initial analyses showed variable benefit from single agent anti-PD-1 therapy in Trp53-mutant disease. We are currently exploring combinatorial strategies with targeted therapy (including PI3K and BTK inhibitors) to improve in vivo response rates in the context of tumor lines with distinct genetics and characterizing the immune microenvironmental changes induced by these treatments.
Read full abstract