Survival after recurrence of ewing tumors: The St. Jude children's research hospital experience, 1979-1999

BACKGROUND. Despite improved therapies, 30-40% of patients with Ewing tumors (ET) experience recurrence and have a poor prognosis. The authors analyzed factors prognostic of survival in patients with recurrent ET. METHODS. The authors assessed the relation between postrecurrence survival (PRS) and demographic, disease, and treatment factors in 71 patients who experienced recurrent ET after treatment on one of three consecutive institutional protocols. RESULTS. Thirty-four patients (47.9%) had distant recurrence, 25 patients (35.2%) had local recurrence, and 12 patients (16.9%) had both distant and local recurrence at a median of 1.7 years after diagnosis. The probability of 5-year PRS (± 1 standard error) was 17.7% ± 4.5%. Recurrence ≥ 2 years after diagnosis predicted a significantly better outcome (5-year PRS, 34.9% ± 8.5%) compared with earlier recurrence (5.0% ± 2.8%; P < 0.001). Patients who had both local and distant recurrence fared worse (5-year PRS, 12.5% ± 8.3%) compared with patients who had local recurrence alone (21.7% ± 7.8%) or distant recurrence alone (17.6 ± 6.1%). Among patients with local recurrence alone, those who underwent salvage with radical surgery had significantly higher 5-year PRS estimates (31.4% ± 11.6%) compared with the other patients (9.1% ± 6.1%; P = 0.023). Pulmonary irradiation significantly improved the outcomes of patients with isolated pulmonary recurrence (5-year PRS estimate, 30.3% ± 12.5% vs. 16.7% ± 10.8%, respectively; P = 0.018). CONCLUSIONS. Although outcomes are generally poor after patients experience recurrence of ET, certain patient groups differ appreciably in their likelihood of survival. Patients who experience recurrence ≥ 2 years after diagnosis and patients who have local recurrence that can be treated with radical surgery and intensive chemotherapy have the most favorable outcomes.