Molecular therapy for rhabdomyosarcoma

Rhabdomyosarcoma (RMS) is the most common sarcoma of childhood representing about 23% of all sarcomas, and approximately 7% of all pediatric malignancies (Arndt and Crist 1999). Histologically, RMS presents as three major variants, embryonal (ERMS), representing about 60%, botryoid that are usually combined with embryonal tumors, and alveolar RMS (ARMS) that comprise 30% of RMS. The pleiomorphic variant is an unrelated high-grade sarcoma with various degrees of muscle differentiation and is rarely (if ever) diagnosed in children. Embryonal RMS is not associated with any chromosomal translocation, but loss of heterozygosity at 11p15.5 is a common feature (Scrable et al. 1987; Scrable et al. 1989) associated with loss of imprinting of the IGF-2 locus (Anderson et al. 1999). Alveolar RMS is characterized by specific translocations. Approximately 60 to 70% of histologically diagnosed ARMSs involve translocations of t(2;13)(q35;q14) leading to PAX3-FKHR gene fusion (Galili et al. 1993), whereas 10% have the t(1;13)(q36;q14) translocation that encodes the PAX7-FKHR fusion (Davis et al. 1994). Both translocations generate in-frame fusion between the PAX gene DNA binding domain and the transactivation domain of FKHR. Interestingly, both ERMS and translocation-positive ARMS may have loss of imprinting at 11p15.5 (Anderson et al. 1999), suggesting that dysregulation of IGF2 may be common to both histologies. Recent studies using expression profiling suggest that histological ARMS that are translocation negative cluster with ERMS, hence the molecular classification of these tumors may differ from the histopathologic classification. Further, a small signature comprising as few as ten genes differentiates translocation-positive from translocation-negative RMS (Lae et al. 2007). Thus, the understanding of genetic events characteristic of RMS is at a point where these may be applied to differential diagnosis, and potentially to novel treatment strategies. It is well established that ARMSs are more aggressive than ERMS, and have a poorer prognosis (Qualman and Morotti 2002; Breneman et al. 2003). Further, the prognosis for patients with metastatic disease at diagnosis is significantly worse for ARMS having the t(2;13)(q35;q14) translocation compared to those having the t(1;13)(q36;q14) variant (Sorensen et al. 2002). Thus, the genetic alterations impact on chemo- or radio-sensitivity. However, it is also clear that any advanced stage RMS still presents a clinical challenge. Essentially, the cure rate for metastatic disease has not changed significantly in 30 years, despite intensification of cytotoxic therapy and introduction of novel cytotoxic agents (Pappo et al. 2007). Thus, while introduction of novel cytotoxic agents, such as the camptothecins that target topoisomerase 1 (Furman et al. 1999; Pappo et al. 2007), may ultimately improve outcome, it is probable that it will be at the expense of additional toxicity or necessitate reduced dose intensity of other agents used in the treatment of RMS. Alternative approaches that exploit the molecular characteristics of RMS conceptually appear to offer potential benefit with lower toxicity and with reduced sequellae. The obvious example is that of imatinib mesylate in the treatment of chronic myelogenous leukemia (Druker 2003; Druker et al. 2006) or gastrointestinal stromal tumors (Rubin et al. 2007; Siehl and Thiel 2007). However, the complexities of RMS biology suggest that a single genetic driver is unlikely, and that combinations of agents that target different molecular abnormalities will be necessary to eradicate these tumors using these rational approaches. Indeed, the development of therapeutic strategies that exploit synthetic lethal interactions that are consequential to the molecular aberrations in these tumors will be the major challenge in developing curative approaches to childhood RMS. Here, we review the reported molecular characteristics of RMS as they relate to the development of molecularly targeted therapy.