Novel therapeutic strategies are required for pediatric solid tumors with poor prognosis such as metastasizing neuroblastoma, rhabdomyosarcoma and Ewing's sarcoma. A prerequisite for the development of such new therapies is the availability of murine models. To be useful for therapeutic studies, these models should not only recapitulate the genetic alterations characteristic of the human disease but should also mimic the metastatic process and the response to current therapy, both of which ultimately determine the fate of children with these tumors. This review scrutinizes the utility of existing murine models of neuroblastoma, rhabdomyosarcoma and Ewing's sarcoma for investigating novel therapies. Much experience has been gained with both syngeneic and xenogeneic transplantable models of these tumors, while transgenic and knockout mice are just beginning to be available for therapeutic investigations. Modeling the genetic aberrations characterizing these tumors may provide faithful models for therapeutic studies in the future.Key words: mouse model; neuroblastoma; rhabdomyosarcoma; Ewing's sarcoma; therapy In contrast to adults, most children with localized solid tumors can be cured today. However, subsets of pediatric solid tumors still carry a poor prognosis. Besides brain tumors, which will not be considered in this review because of their unique features, these tumors include metastatic neuroblastoma, rhabdomyosarcoma and Ewing's sarcoma. As a paradigm, the dismal prognosis for advanced-stage neuroblastoma has not changed much for decades. While current therapy is largely based on clinical studies of combined chemo-and radiotherapy, the development of new therapies has to rely on clinically relevant murine models of these tumors. This is especially important in pediatric oncology, given the low number of children eligible for phase I trials. 1 The purpose of this review is to discuss the utility of existing murine models of poor-prognosis extracranial pediatric solid tumors for investigating novel therapies. Available mouse models for neuroblastoma, rhabdomyosarcoma and Ewing's sarcoma are reviewed. These tumors have been selected because of their wellcharacterized genetic alterations amenable to mouse modeling and molecular therapeutic strategies.
THE "IDEAL" MURINE MODEL OF A SOLID TUMOR FOR THERAPEUTIC RESEARCHThe general limitations of therapeutic research with both transplantable tumor models (syngeneic and xenogeneic) and "spontanously" arising tumors in transgenic or knockout mice have recently been reviewed. 2,3 The characteristics of an "ideal" murine model of a solid tumor have been well defined, as follows:1. Some tumors, especially tumors of childhood, are well described on the genetic level and show characteristic genetic alterations. These changes should obviously be present in the tumor model if the experimental approach targets these changes. However, if the therapy to be tested does not specifically target the etiologic mutation but the ensuing molecular pathology, mouse models that show...