Title: The immune anti-tumor effects of allogeneic bone marrow transplantation and adoptive cell therapy with donor and recipient leukocyte infusions in mice carrying neuroblastoma
Other Titles: De immuun anti-tumor effecten van allogene beenmergtransplantatie en adoptieve cel therapie met donor en gastheer leukocyten infusies in muizen met neuroblastoma
Authors: Willems, Leen; M9818009;
Issue Date: 1-Jul-2013
Abstract: Allogeneic hematopoietic stem cell transplantation(HSCT) and posttransplant adoptive cell therapy with donor leukocyte infusions(DLI) are known to exert potent graft-versus-leukemia effects and convincing evidence is emerging that this approach can also elicit responses against solid tumors. Graft-versus-leukemia effects – specifically after DLI – largely rely on lymphohematopoietic graft-versus-host T cell alloreactivity which can result in graft-versus-host disease when it extends to epithelial tissues. Not only for hematological but also for solid tumors the association of anti-tumor effects with graft-versus-host disease represents the major obstacle to the successful application of allogeneic HSCT and adoptive DLI therapy as an immunotherapeutic strategy. Recent data in mice have shown that challenge with recipient-type rather than donor-type leukocytes (recipient leukocyte infusion,RLI) can equally provide antitumor effects in hematological malignancies. RLI provokes lymphohematopoietic host-versus-graft T cell reactivity with rejection of the graft and associated anti-leukemic responses and does not carry the risk of graft-versus-host disease. This strategy is currently being explored clinically in patients with hematological malignancies. In experimental mouse models, it was demonstrated that NK cells, activated in vivo in the course of RLI-induced T cell alloreactivity, contribute to the survival advantage produced by RLI. Allogeneic hematopoietic stem cell transplantation may also hold potential for the treatment of high-risk neuroblastoma, a common extracranial solid tumor in childhood that carries a bleak prognosis. Despite aggressive treatment including high-dose chemotherapy and lymphohematopoietic rescue with autologous hematopoietic stem cells, relapse is very frequent. This spurs research into post-transplant immunotherapeutic interventions to enhance antitumor immunity. Neuroblastoma cells are known to be sensitive to NK cell cytotoxicity and strategies targeting NK cells are therefore particularly of interest. Alternatively, case reports suggest that allogeneic HSCT in high-risk neuroblastoma patients may prolong survival and thus confers a graft-versus-neuroblastoma effect. This has been experimentally demonstrated in mice. A phase I/II trial is currently investigating the anti-tumoreffect of haplo-identical HSCT in children with neuroblastoma. The overall aim of this study was to investigate the potential combined effect of allogeneic HSCT and the novel RLI strategy on the growth of experimentally induced neuroblastoma tumors. Given the efficacy of RLI in enhancing graft-versus-leukemia effects in mice, the evidence that such effects involve not only T cells but also NK cells, and the knowledge that neuroblastoma cells are generally sensitive to NK cell cytotoxicity, we hypothesized that RLI may be an effective and safe alternative to DLI to enhance the immune-mediated graft-versus-neuroblastoma effect. Immune interventions after standard therapy with autologous HSCT for neuroblastoma can –to some extent– enhance endogenous anti-neuroblastoma immunity. In mice it has been shown that T cells devoid of anti-host alloreactivity may foster anti-neuroblastoma effects of allogeneic bone marrow transplantation(BMT) in mice. This supported our hypothesis that challenge with syngeneic mature effector cells may –in addition to causing alloreactive T cell responses– foster direct anti-tumor reactions. The specific aim of this work was to study the effects of RLI adoptive therapy on the local and metastatic growth of neuroblastoma tumors in allogeneic bone marrow chimeras. We studied RLI in comparison with DLI since this is the standard approach used in hematological malignancies ánd the approach currently studied in clinical trial for high-risk neuroblastoma. We used mouse models off local and systemic posttransplant neuroblastoma relapse in a setting of MHC-mismatched allogeneicBMT: we chose to inoculate mice after the transplant to study immune anti-tumor effects independently of direct effects of recent radiation therapy on the tumor. We used three different models. First, a focal model in which tumor cells are injected subcutaneously and give rise to a local and clinically measurable tumor (subcutaneous model). Second, a systemic model where tumor cells are inoculated intravenously and thus cause direct metastatic spread (intravenous model). Third, a more physiological and preclinical model of metastatic spread where tumor cells are inoculated under the kidney capsule and where the local tumor is removed in toto after 7 days by a nephrectomy (kidney model). For the systemic and kidney models, we generated a luciferase-expressing tumor cell line and developed a bioluminescence imaging protocol to non-invasively follow metastasis-free survival in vivo. In part I of this work we used the subcutaneous model to study the effects of allogeneic BMT and adoptive cell therapy on local neuroblastoma growth. We confirmed that allogeneic BMT in mice provides a baseline graft-versus-neuroblastoma effect. We next demonstrated that bothDLI ánd RLI, when given to mixed bone marrow chimeras, further slow the local growth of subcutaneous implanted neuroblastoma tumors. DLI provoked a rapid conversion to full donor chimerism but mice developed severe GvHD, whereas RLI produced complete graft-rejection but left mice healthy. The chimerism of intratumoral leukocytes evolved in parallel with peripheral blood chimerism and was -both after DLI and RLI- associated with increased intra-tumor CD8/CD4ratios, CD8+ T-cell IFN-g-expression and NK-cell Granzyme B-expression. The effect of RLI was weaker than that of DLI but co-transfer of naïve recipient-type NK cells could elicit a growth-limiting effect on neuroblastoma tumors in RLI chimeras. In part II we investigated the effects of posttransplant adoptive cell therapy on the systemic spread of neuroblastoma. The final prognosis for patients with solid tumors, and especially for neuroblastoma, is determined by the presence of distant therapy-non-responsive metastases. Novel forms of immunotherapy therefore only hold clinical potential if they are efficacious on both local and systemic spread of the tumor. Using the intravenous tumor model we documented that allogeneic BMT provides a significant overall survival benefit relative to syngeneic BMT. Moreover, using luciferase-expressing tumor cells and bioluminescence imaging we showed that both DLI and RLI produce a significant delay in metastatic spread. The bioluminescence imaging studies in the kidney model -although preliminary- were consistent with those of the intravenous model and support the notion that both DLI and RLI have the potential to limit metastatic spread of neuroblastoma. In part III we addressed the question whether syngeneic BMT by itself can provide an immune anti-neuroblastoma effect. The current standard treatment of high-risk neuroblastoma includes high-dose chemotherapy and autologous HSCT which serves merely as a lymphohematopoietic rescue. However, the potential intrinsic immune effects of this procedure have so far not been investigated. Our data are preliminary but show that, relative to naïve mice, syngeneic BMT facilitates tumorigenesis of subcutaneously inoculated tumor cells while on the other hand it delays the growth of established tumors. CD8+ T and NK cells synergistically regulated tumor growth. Direct evidence of CD8+ T cells could not be found. However, the role of NK cells is supported by the evidence of increased Fas-L expression and by the fact that adoptively transferred syngeneic NK cells could reinforce the reduction in tumor growth. In conclusion, the data support the potential value of DLI in mediating graft-versus-solid tumor effects. Importantly, they also deliver the first experimental evidence that RLI is able to target a solid tumor and that it thus may represent a safe alternative to DLI for this purpose. We document that lymphohematopoietic alloreactivity extends to the environment of the tumor itself, indicating that also for solid tumors it is closely related with the antitumor effect. Studies on the leukocytes contained within the focal neuroblastoma tumors provided evidence strongly suggesting that the antitumor effect of both DLI and RLI involves not only CD8+ T cells but also cytotoxic NK cells. Whereas in the DLI setting, the antitumor effect may rely predominantly on alloreactive CD8+ T and NK cells, we provided evidence that the T cell alloreactivity provoked by an RLI challenge suffices to provide syngeneic naïve NK cells with the potential to attack syngeneic neuroblastoma cells in vivo. We conclude therefore that RLI and DLI in allogeneic chimeras can provoke an in vivo lymphokine-activated killer phenomenon leading to cytotoxically active NK cells, even in the syngeneic setting. The alloreactive T cell response produced by RLI and DLI may also facilitate the activation and expansion of tumor-specific donor or host CD8+T cells. Consistent with prior studies in leukemia models the RLI effect was weaker than the DLI effect but this could be compensated by the additional infusion of syngeneic NK cells, a procedure that does not compromise the safety profile of RLI. Lastly, although to some extent preliminary, the data indicate that the effects of both DLI and RLI are not limited to the local growth of atissue-localized tumor but can also limit the metastatic spread of a solid tumor. This is a critical requirement for a novel form of immunotherapy to beclinically relevant. The data may be particularly helpful to understand the value of allogeneic HSCT and adoptivecell therapy for neuroblastoma. We demonstrate that this approach may provide asynergistic T cell and NK cell interaction resulting in an effective antitumorresponse against neuroblastoma, even in the setting where RLI is used. This is particularly interesting since neuroblastoma is wellknown to be sensitive to NK cell cytotoxicity. With this respect, the data also suggest that NK cells equally contribute to the baseline antitumor effect of allogeneic HSCT inneuroblastoma. Last, our preliminary observations in the syngeneic BMT model show that the effects of conditioning and autologous lymphohematopoietic reconstitution on neuroblastoma growth are complex and may include paradoxical effects on tumorigenesis and tumor progression.
Publication status: published
KU Leuven publication type: TH
Appears in Collections:Laboratory of Experimental Transplantation

Files in This Item:
File Status SizeFormat
Leen Willems doctoraatsthesis 2013.pdf Published 3180KbAdobe PDFView/Open Request a copy

These files are only available to some KU Leuven Association staff members


All items in Lirias are protected by copyright, with all rights reserved.