Combination Therapy with BTK Inhibitor Plus Anti-PD-1 Antibody Results in a Hyperprogressor Phenotype in a Mouse Model of CLL

Chronic lymphocytic leukemia (CLL) is often driven by aberrant activation of the B-cell receptor signaling pathway. We and others have shown in preclinical animal models as well as human clinical trials that the BTK inhibitor, ibrutinib, is effective in reducing CLL tumor burden and improving survival rates in both humans and mice. Preclinical studies also suggest that BTK inhibitors likely act not only through their direct effect on CLL tumor cells but also by alleviating immunosuppression in the tumor microenvironment through BTK inhibition in myeloid-derived suppressor cells. Additionally, we have previously demonstrated that CLL cells are susceptible to cytotoxic killing by T cells targeting the aberrantly expressed TCL1 oncoprotein (Weng et al. Blood 2012). Therefore, we hypothesized that the combination of BTK-pathway inhibition in conjunction with activation of antigen-specific T-cells by immune checkpoint blockade will be a synergistic therapeutic strategy. Here, we examined the effect of acalabrutinib (previously known as ACP-196), a selective BTK inhibitor with limited effect on other kinases, alone or in combination with immune checkpoint blockade in a mouse model of CLL.

Eµ-TCL1 mice, which overexpress TCL1 in B-cells, were used as a model of CLL. We generated cohorts of Eµ-TCL1 mice and treated them with anti-PD-1 antibody (n = 18), acalabrutinib (n = 20), and acalabrutinib+anti-PD-1 antibody (n= 19). Treatment cohorts had a significant reduction in CD5+CD19+ tumor cells in peripheral blood and spleens as compared to vehicle-treated mice (n =16) (p <0.05). However, the most pronounced anti-tumor effects were observed in mice treated with acalabrutinib alone. In fact, MRI imaging of splenic volume and flow cytometry of CD5+CD19+ cells during the treatment phase revealed that acalabrutinib monotherapy was superior to the combinatorial therapy. Even more surprising was the fact that while anti-PD-1 and acalabrutinib monotherapies provided an improvement in survival compared to vehicle treatment (p = 0.05 and p < 0.0001, respectively), the combination of these two agents actually diminished overall survival (p = 0.77). Shockingly, histopathological analyses of tumors from these acalabrutinib+anti-PD-1 treated mice revealed a significant increase in lymph node involvement and tumors with a high mitotic index. Consistent with this, we observed that treatment of CD5+CD19+ tumor cells from Eµ-TCL1 transgenic mice in vitro with anti-PD-1 antibody plus acalabrutinib markedly increased the proliferative index as measured by EdU incorporation assay compared to acalabrutinib treated group.

To understand the molecular events responsible for the observed acalabrutinib+anti-PD-1 hyperproliferative phenotypes, we are examining complementary pathways that may be aberrantly activated/repressed by this combination. Currently, we have identified expression changes in both the NFkb and PKC signaling pathways when these agents are used in combination, and are working to understand how this combination may impact tumor progression. We are also assessing the possibility that anti-PD-1 interferes with the anti-tumorigenic effect of acalabrutinib by enhancing the BCR signaling pathway in CLL tumor cells.

In summary, our results demonstrate that the selective BTK inhibitor, acalabrutinib greatly enhances the survival of Eµ-TCL1 mice compared to either vehicle or ibrutinib alone. Surprisingly, we found that the combination of anti-PD-1 antibody + acalabrutinib therapy is actually detrimental in this CLL model and results in a hyperprogressor phenotype. Therefore, results from ongoing clinical trials evaluating combination strategies of anti-PD-1 antibody therapy plus BTK inhibitor need to be analyzed carefully to ensure that this combination is not leading to a similar hyperprogressor phenotype in patients. The results provided here offer some initial insights into the potential mechanisms of the "hyperprogressive" phenotypes following anti-PD-1 treatment and may highlight pathways that could be useful in blocking these deleterious effects.