Purpose: Immunotherapy for cancer has been most well developed for malignant melanoma, renal cell carcinoma and chronic myelogenous leukemia where interleukin-2 and cell therapies have been clearly shown to have antitumor activity. It remains unclear whether the vast array of other cancers are equally susceptible to immune based therapies and if not, what factors limit the immunogenicity or immune responsiveness of specific tumors. This project seeks to determine whether pediatric tumors are immunogenic and to identify specific targets that can be used to develop effective immune based therapies. Materials and Methods: In order to answer this question, new materials needed to be generated that accurately model pediatric tumors in mice. We spent considerable effort to create novel models of pediatric cancer on the C57BL/6 background in mice and have made substantial progress in this regard. These models address a critical barrier to progress in this field since animal models provide the essential first step in understanding host:tumor interactions and immune studies are much more efficient when undertaken in the BL/6 model where many reagents and mice are available with genetic modifications in specific immune pathways and are critical reagents for identifying nodal points in the host:tumor interactions. Major accomplishments of this project are the development of three pediatric tumor models. First, C57BL/6 mice which were transgenic for hepatocyte growth factor and knock-out for p53 developed an embryonal rhabdomyosarcoma, which was adapted for growth in culture and induces local and metastatic rhabdomyosarcoma in C57BL/6 mice. The line is termed M3-9-M. We fully characterized this model and demonstrated that it accurately mimics embryonal rhabdomyosarcoma histologically, and with regard to gene expression. It also is remarkably immunogenic, as evidenced by full tumor protection from rechallenge after immunization with irradiated M3-9-M. We further demonstrated that the immune effects are T cell mediated, and that regulatory T cells contribute to immune escape by M3-9-M. Finally, using adoptive immunotherapy from mice immunized with M3-9-M, combined with Treg depletion, we can induce regression of established tumors. This work has been submitted for publication. We believe it is essential to molecularly identify the tumor antigens responsible for immune mediated effect in this model. We hypothesized that survivin, an anti-apoptotic protein that is highly expressed in M3-9-M was a candidate tumor antigen. While immunization with survivin induced partial protection against M3-9-M, we could find no evidence that whole tumor vaccination with M3-9-M induced survivin specific immunity. Nonetheless, we created a TCR-transgenic mouse with specificity for an immunodominant epitope derived from murine survivin. We see inconsistent results with some evidence for recognition of the murine rhabdomyosarcoma and the murine neuroblastoma with the TCR transgenic mouse and in other experiments, the reactivity is not observed. Furthermore we have observed differences with different founders, confounding our results. Ongoing work will seek to determine whether survivin is a relevant tumor antigen pediatric solid tumors. We also have developed and are undertaking studies of the immunogenicity of a murine model of neuroblastoma derived from N-Myc transgenic mice (in collaboration with Javed Khan) and a similar mobel of E2A-PBX leukemia (in collaboration with Terry Fry). These models form the basis for a series of studies to investigate the immunobiology of these three pediatric tumors. Finally, we are conducting collaborative studies with Dr. Terry Fry investigating the infiltration of pediatric tumors with myeloid derived suppressor cells. These cells have been shown in several model to mediate immune escape. We are also seeking to correlate findings in these murine tumor models with information gleaned from our clinical trials of tumor lysate based vaccine in pediatric sarcomas. Using surgical samples or fine needle aspirates, we are conducting proteomic analyses of tumor lysates to assess for the presence of defined tumor antigens and to explore the possibility that markers of myeloid derived or Tregulatory cells may be present. These studies are ongoing but utilize clinical samples gleaned from 07-c-0206.
Showing the most recent 10 out of 24 publications
