PD-1 expression on richter syndrome tumor cells does not provide a growth advantage in murine or patient-derived xenograft models In Vivo

Abstract The immune checkpoint receptor Programmed cell death 1 (PD-1) is expressed by the neoplastic B cells in the majority of clonally-related Richter Syndrome (RS) tumors, and treatment with anti-PD-1 inhibitors, alone or in combination with BTK inhibitors, has demonstrated clinical responses in patients with RS. However, data on functional consequences of PD-1 expression on RS cells are lacking. In malignant melanoma, PD-1 blockade has been reported to suppress tumor growth through direct inhibition of tumor cell PD-1 signaling (Kleffel S et al, Cell. 2015). Whether PD-1 expression on RS cells could possibly contribute to PD-1 inhibitor treatment efficacy through either direct cellular effects or by facilitating escape from immune surveillance mechanisms is currently unclear. In order to investigate the functional consequences of PD-1 expression on RS cells, we used CRISPR/Cas9 gene editing to introduce PD-1 loss-of-function (LOF) mutations in the human RS patient-derived xenograft (PDX) models RS1316 and RS9737, both of which express PD-1. Mixed populations of PD-1 knockout and PD-1 wild type RS1316 and RS9737 cells were then inoculated intravenously into immunodeficient NSG mice, which lack T, B and NK cells (n=10 and n=8 per model, respectively). The malignant cells were isolated from the spleens of recipient mice between 4 and 8 weeks after transplantation, and the PD-1 mutant and wild-type allele frequency (WAF) were evaluated in both the originally implanted and the recovered tumoral cells. Neither RS-PDX model showed a significant change in WAF between the injected and recovered cells (RS1316, mean WAF of injected cells 63.5%, mean WAF of recovered cells 56.4%, p=n.s.; RS9737, mean WAF of injected cells 19.8%, mean WAF of recovered cells 17%, p=n.s.), indicating no growth advantage for PD-1-expressing RS cells in the absence of an anti-tumor immune response. The above results were further supported by in vivo experiments using the anti-PD-1 inhibitor pembrolizumab. NSG mice injected intraperitoneally with RS1316 (control, n=5; treated, n=6) or RS9737 (control, n=7; treated, n=7) received intraperitoneal injections of pembrolizumab (200 µg per injection) every other day beginning 4 days after tumor inoculation, and were euthanized after 4-5 weeks. Tumor burden was assessed by flow cytometry analysis of human CD19-positive cells in the peritoneal lavage. Pembrolizumab treatment showed no significant reduction in tumor burden compared to controls in either the RS1316 (median number of RS cells: treated mice 53x106, control mice 64x106; p=n.s.) or RS9737 model (median number of RS cells: treated mice 5.7x106, control mice 15.5x106; p=n.s.). Consistent with these findings, in vitro treatment of RS1316 cells with either pembrolizumab or recombinant PD-L1 Fc-fusion protein had no effect on viability or proliferation rate when compared to controls. The above data indicate that RS cell-intrinsic PD-1 does not promote tumor growth through direct signaling mechanisms. To investigate whether PD-1 expression on the tumor cells may facilitate escape from the anti-tumor immune response, we used CRISPR/Cas9 editing to introduce LOF mutations in PD-1 into the mouse TCL1-863 TKO RS model that we previously generated (Chakraborty et al, Blood 2021) and investigated the growth of a mixture of PD-1 knockout and PD-1 wild type cells in immunocompetent syngeneic and immunodeficient NSG mice (n=5 per group). In both the immunocompetent and immunodeficient recipients, no preferential expansion of the PD-1 wild type compared to the PD-1 knockout population was observed (WAF of injected cells 18%, mean WAF of recovered cells from immunocompetent mice 19%, mean WAF of recovered cells from immunodeficient mice 16%, p=n.s. for both comparisons with respect to injected cells). To further validate these findings, we performed additional experiments with another murine RS model, TCL1-355 TKO, which was transplanted only in immunocompetent recipients (n=14). Again, no preferential expansion of the PD-1 wild type over PD-1 knockout cells was observed (mean WAF injected cells 30%, mean WAF recovered cells 33%, p=n.s.). Collectively, these findings demonstrate that PD-1 on neoplastic RS cells provides no growth advantage either through direct cell-intrinsic signaling or by protecting the malignant cells from immune surveillance and suggest that the therapeutic benefit of anti-PD-1 inhibitors is entirely derived from their effects on immune cells.