Prognostic factors of localized extremity rhabdomyosarcomas: The results of a pooled analysis from U.S. and European cooperative groups.

Abstract

9505 Background: We present pooled data from 643 patients (pts) treated in 14 studies conducted by European and American cooperative groups. Methods: Clinical factors, including age, histology, site of primary (hand and foot vs others), size, invasiveness (T), nodal involvement (N) and post surgical group (complete R0, microscopic residue R1 or macroscopic residue / biopsy only R2), chemotherapy (CT) type and duration, radiation therapy (RT) and period of treatment (before or after 1995) were looked at for their impact on survival (OS). Results: The 5 and 10 year OS are 67% (se 1.8) and 62% (se 2) respectively. By univariate analysis, OS is adversely influenced by age > 3 years, size > 5 cm, T2 stage, N1 stage, cooperative group (Europe vs US), incomplete initial surgery and treatment before 1995. Site and alveolar histology are less significant. RT has no impact. Many of these variables are associated. Pts treated by the US group were more often operated on and had more complete surgery at diagnosis than pts treated in Europe. In multivariate analysis OS is strongly correlated to age, T stage, N stage, initial surgery, treatment period and cooperative group. The better survival rate of tumors treated by the US group is observed for grossly resected tumors (R0 or R1) only. OS survival is 86% (se 3) for US pts as compared to 68% (se 4) for European pts (p=0.004). When considering the R0 and R1 tumors separately (n=252) by univariate analysis, OS is significantly influenced by the completeness of initial surgery (R0 vs R1, p=0.05), T stage (p=0.02), and site of the tumour (p=0.02). Longer CT duration (> 27 weeks) is associated with a better OS (p=0.002). RT had no impact. In multivariate analysis, if CT duration is not considered, cooperative group (p=0.001), site (p=0.001) and T stage (p=0.05) are significant. Completeness of initial surgery has no impact. However, adding duration of chemotherapy to the model, site of the primary remains significant (p=0.006) but T status and cooperative group are no longer significant. Conclusions: This analysis confirms the role of many prognostic factors, but in grossly resected tumors it identifies that the duration of chemotherapy has an important impact. No significant financial relationships to disclose.