The induction of partial tolerance toward pancreatic autoantigens in the treatment of type 1 diabetes mellitus (T1DM) can be attained by autologous hematopoietic stem cell transplantation (HSCT). However, most patients treated by autologous HSCT eventually relapse. Furthermore, allogeneic HSCT which could potentially provide a durable non-autoimmune T-cell receptor (TCR) repertoire is associated with a substantial risk for transplant-related mortality. We have previously demonstrated an effective approach for attaining engraftment without graft versus host disease (GVHD) of allogeneic T-cell depleted HSCT, following non-myeloablative conditioning, using donor-derived anti-3rd party central memory CD8 veto T cells (Tcm). In the present study, we investigated the ability of this relatively safe transplant modality to eliminate autoimmune T-cell clones in the NOD mouse model which spontaneously develop T1DM. Our results demonstrate that using this approach, marked durable chimerism is attained, without any transplant-related mortality, and with a very high rate of diabetes prevention. TCR sequencing of transplanted mice showed profound changes in the T-cell repertoire and decrease in the prevalence of specific autoimmune T-cell clones directed against pancreatic antigens. This approach could be considered as strategy to treat people destined to develop T1DM but with residual beta cell function, or as a platform for prevention of beta cell destruction after transplantation of allogenic beta cells.

Clear Cell Sarcoma (CCS), also referred to as malignant melanoma of soft parts, is a rare and aggressive malignant tumor. It comprises 1% of all soft tissue sarcomas and is known to be radio- and chemotherapy resistant. CCS shares morphological and immunohistochemical features with malignant melanoma, including melanin biosynthesis and melanocytic markers. However, it is distinct for the presence of EWSR1-ATF1 translocation which activates MITF transcription factor. We report here of an aggressive case of CCS in a 9-year-old patient, which demonstrates the critical role of molecular analysis in the diagnosis and treatment of uncommon cancer variants in the era of personalized medicine. The EWSR1-ATF1 translocation induces pathological c-Met activation, and so, following unsuccessful CTLA4 and PD-1 blockade immunotherapy, the child received cabozantinib, a small molecule tyrosine kinase inhibitor, with the intent to block c-Met oncogenic effect. In parallel, active immunization, using hapten di-nitrophenyl modified autologous tumor cells was administered with monotherapy PD-1 inhibitor nivolumab. Under this “triplet” therapy, the patient attained an initial partial response and was progression-free for 2 years, in good performance status and resumed schooling. Based on our observation, cabozantinib can be used as an effective and potentially life-prolonging treatment in CCS. We suggest that priming the child’s immune system using her autologous tumor and combating T cell exhaustion with PD-1 blockade may have synergized with the targeted therapy. Combining targeted and immunotherapy is a rapidly growing practice in solid tumors and provides a glimpse of hope in situations that previously lacked any treatment option.

R.S.K and B.N.L contributed equally to this study. The induction of tolerance towards pancreas autoantigens is a promising approach in the treatment of type 1 diabetes mellitus (T1DM). This goal can be partially attained by immuno-ablation followed by autologous HSCT which mitigates the immune reaction leading to diabetes, at least temporarily, if performed soon after diagnosis. However, most T1DM patients transplanted with autologous HSCT eventually relapse (E Snarski et al., Bone marrow Transplantation; 2016; 51, 398-402). Thus, developing methodologies for allogeneic HSCT, to provide a durable non-autoimmune TCR repertoire, could be a promising treatment approach provided that such protocols can safely achieve donor type chimerism. Accordingly, transplantation of T cell-depleted allogeneic HSCT (TD-HSCT) under mild conditioning, associated with minimal toxicity and reduced risk of GVHD, offers an attractive therapeutic option. However, overcoming rejection after reduced conditioning in T1DM represents a major challenge. We previously demonstrated in wild type mice that rejection of TD-HSCT can be prevented using donor-derived veto cells. Here, we show proof of concept of the safety and efficacy of veto cell mediated non-myeloablative mismatched allogeneic TD-HSCT, in the established NOD mouse T1DM model. Veto activity, first defined by Miller (Miller, R. G; Nature; 1980; 287; 544-54), is based on the ability of specific cell populations to attack host CTL-precursors (CTLp) that are directed against the antigens presented by the veto cells themselves. This response spares cells that are not targeted against the veto cells, including those recognizing pathogens. Among different veto cell populations described in the literature, central memory CD8 T cells exhibit the most robust veto activity upon transplantation, but are also endowed with marked GvH activity. We overcame GvH reactivity by expanding naïve or memory CD8 T cells against 3rd party MHC or viral antigens, under culture conditions favoring expression of the central memory phenotype. Such anti-3rd party central memory CD8 T cells (Tcm) are endowed with marked veto activity, while effectively depleted of GVH reactivity in fully mis-matched recipients (Reviewed in Reisner Y, Or-Geva N. Semin. Hematol. 2019; 56(3): 173-182). In this study, Tcm veto cells were generated from splenocytes obtained from Balb/c donors (H2d) cultured against irradiated third-party splenocytes (FVB; H2q), under cytokine deprivation. The selective expansion of CD8 mouse T cells against 3rd party stimulators under these conditions leads to selective 'death by neglect' of bystander anti-host T cell clones that could mediate GvHD; such cells are further diluted out by subsequent expansion of anti-3rd party T clones during further culture in the presence of IL15. In addition to causing selective loss of GvH reactive T cells, these culture conditions induce a central memory phenotype shown to be crucial for robust veto activity in vivo. To evaluate the safety and efficacy of such a transplant for diabetes therapy, 8 week old NOD mice (before diabetes onset) were treated with 4.5 Gy TBI conditioning at day -1, anti-3rd party veto Tcm and megadose nude bone marrow (NuBM) on day 0, and Rapamycin treatment from day -1 to day +4 (Scheme 1). Controls were untreated or received conditioning with no transplant. In four independent experiments (n= 35 NOD mice in the transplanted group), high chimerism from 83.5% to 99.6 % was found in all transplanted mice at 6 months post-transplant. Notably, with a follow-up of 200 days, 72.4% mice in the untreated group, and 96% in the conditioned group died of diabetes, while only 8.5% diabetes-associated mortality (Fig.1A) and no GvHD or other transplant-related mortality was observed in the transplanted mice (Fig.1B). Our results demonstrate a proof of concept for the safety and efficacy of non-myeloablative allogeneic TD-HSCT in type 1 diabetes. A clinical protocol testing the safety and efficacy of anti-3rd party veto cells in the context of low toxicity non-myeloablative TD-HSCT in hematological malignancies is commencing at MD Anderson Cancer Center. If successful, our results support further extension of this platform to autoimmune diabetes and other autoimmune diseases. Disclosures Lustig: Yeda Ltd.: Patents & Royalties. Reisner:Cell Source, Inc.: Consultancy, Equity Ownership, Membership on an entity's Board of Directors or advisory committees, Patents & Royalties.

No abstract

We recently demonstrated that mouse anti-3rd party central memory CD8 veto cells (Tcm) enable engraftment of fully mis-matched T cell depleted bone marrow (TDBM) in sublethally irradiated (5.5 Gy TBI) recipients (Eran et al. Blood 2013). Here we found that Tcm generated from (host x donor)F1 or fully allogeneic donors were able to engraft and sustain their presence in such sub-lethally irradiated mice in the absence of TDBM transplants. Limiting dilution analysis revealed that the reactivity of host T cells (HTC) against donor spleen cells was significantly diminished in the chimeric mice treated with Tcm, whereas reactivity against 3rd party was comparable betweenTcm treated and non-treated mice (Fig. 1). This specific immune tolerance indicated that vetoTcm could potentially enable the use of 'off-the-shelf' allogeneic CAR T cells or any other genetically modified T cells from the same donor. To further investigate this possibility, we tested the ability of Tcmveto cells to enable engraftment of CD8 T cells from OT1 mice, expressing a transgenic T cell receptor (TCR) directed againstovalbumin (OVA) residues 257-264 in the context of H2Kb MHC-I. To that end, 1x106 OT1 CD8 T cells (H2b) were infused into Balb/c (H2Dd) recipients sublethally irradiated with 5.25 Gy TBI in the presence or absence of different doses of veto Tcm. As shown in Fig. 2, when tested at 2 months post-transplant, OT1 cells could not be detected in the peripheral blood of untreated recipient mice, while they were readily detectable in mice receiving 2x106 Tcm (0.78±0.36), reaching a higher level (1.18± 0.61) upon infusion of 5x106Tcm. To evaluate the functionality of the engrafted OT1 T cells in the chimeric mice, we developed and calibrated a new murine model, using a melanoma B16-cell line of C57BL background that expresses the ovalbumin peptide and the tdTomatomarker (B16-OVA-tdTomato) (Fig 3A). In this model, we tried to mimic a state of minimal residual disease by injecting a small number of B16-OVAtdTomato melanoma cells (0.25x106) into syngeneic C57BL mice; we then treated the mice with sublethal irradiation (6Gy TBI) to simulate treatment of the tumor, but also to clear out space for the T-cell adoptive transfer. Finally, we injected 1x106 naïve OT-1 cells generated on a (Balb x-OT1-CD45.1+RAG2-) F1 background in the presence or absence of Balb/c veto Tcm (5x106), and followed the development of the tumor. We found thatBalb/c vetoTcm prolonged the engraftment of the transgenic donor-derived F1 OT1 CD8 T cells inBalb/c recipients, thereby enabling significantly improved control of tumor growth when tested at day 20 post-transplant ( p<0.05). In contrast non OVA-expressing tumor cells grew at the same rate in all treatment groups, thereby demonstrating the specificity of the anti-tumor effect (Fig.3B). Previous studies demonstrated that Tcmare depleted of GVH reactivity by virtue of their culture with 3rd party stimulators under cytokine starvation. Our present results demonstrate that suchTcm veto cells can pave the way for functional engraftment of CD8 T cells with genetically modified specificity. Taken together, these results provide a proof of concept for the potential clinical use of such veto cells to enable therapy with 'off-the-shelf ' allogeneic CAR T cells. The relative ease of growing humanTcm in large numbers over a short period of time (9-12 days) suggests that it could be possible to harvest byleukapheresis from a single donor, a sufficient number of cells, to support CAR therapy for more than 50 allogeneic recipients. * N.O.G and R.G equally contributed Disclosures Or Geva: Yeda LTD: Patents & Royalties. Gidron:Yeda LTD: Patents & Royalties. Reisner:Cell Source LTD: Consultancy, Equity Ownership, Patents & Royalties, Research Funding.

No abstract