THORACIC SARCOMAS IN CHILDREN

Objective This study reviews the experience of the Intergroup Rhabdomy osarcoma Studies (IRS) II and III with thoracic sarcomas. Summary Back ground Data Thoracic sarcomas in children present special problems wit h histologic diagnosis, total resection of tumor, and local control. P revious studies have shown high rates of local and distant relapse. Me thods Data from evaluable patients enrolled in IRS II and IRS III were reviewed. Progression-free and overall survival were estimated by the Kaplan-Meier method and statistical differences were calculated by th e log-rank test. The multivariate regression analysis of data was perf ormed using the Cox proportional hazards model. Results A total of 84 patients presented with thoracic sarcomas during IRS II (51 patients) and IRS III (33 patients). Of these, 76 were chest wall tumors, 3 were lung tumors, 4 were pleural tumors, and 1 arose from the heart. The a verage age at diagnosis was 8.5 years (range, 1 month-20.5 years). Thi rty-three patients had alveolar histology, 17 had embryonal histology, and 12 had extraosseous Ewing's sarcoma, with indeterminate, undiffer entiated, and pleomorphic histologies in the remaining 22 patients. Th irteen patients presented as group I (localized disease completely res ected), -18 as group II (microscopic residual or nodal disease), 31 as group III (gross residual disease or biopsy only), and 22 as group IV (distant metastatic disease). Sixty patients (71%) achieved a complet e response. Thirty-nine patients had a local relapse, and 22 had a dis tant relapse. Thirty-five patients (42%) survived with a mean follow-u p of 1.8 years (range 0.7-11.7 years); 49 (58%) have died with an aver age survival of 1.1 years (range 0.3-11.7 years). Four patients (5%) d ied of infection or treatment complications. Progression-free survival was not significantly associated with histology, site, clinical group , or IRS study. Overall survival was significantly associated with cli nical group, size, and local or distant recurrence, but not with histo logy or IRS study by univariate analysis. In a multivariate analysis, only clinical group and local and distant recurrence showed statistica l significance (p < 0.002, <0.006, and <0.001, respectively). Conclusi ons As expected, outcome was better for patients with totally resected or microscopically residual tumor after resection. A higher rate of r ecurrence in group I versus group II, although not statistically signi ficant, suggests that identification of margin status may be difficult , and these patients may require more intensive efforts for local cont rol. These patients continued to have a high rate of local and distant relapse, and new strategies for preventing these are under developmen t.