Abstract
In the OS2006 study, patients younger than 18 years were treated with a methotrexate-based regimen (MTX), patients older than 25 years with a doxorubicin-cisplatin-ifosfamide-based regimen (API-AI), whereas patients aged 18-25 years received either API-AI or MTX. We herein report the prespecified subgroup analysis of the outcome of 106 patients treated with API-AI. Preoperative chemotherapy combined three doxorubicin-ifosfamide-cisplatin (API) and two doxorubicin-ifosfamide (AI) courses. Postoperative chemotherapy was assigned by risk group: localised patients with a good histological response (<10% viable cells) received two AI and two cisplatin-ifosfamide (PI) courses; patients with synchronous metastases, poor histological response or unresectable primary received five cycles of etoposide-ifosfamide (EI). Of the 106 patients, 61 were randomised to receive or not zoledronate. Median age was 30 years (range 18-67), 66 (62%) patients were >25 years. The primary tumours were axial in 28 patients (26%), and 28 (26%) presented with metastases. Ninety-six patients (91%) had surgery, conservative in 82 (85%); 36 patients (38%, 95% CI 28-48%) were good responders. Toxicity was manageable, with no significant difference in severe acute toxicity between patients aged >25 years and those younger. With a median follow-up of 4.8 years, the 5-year event-free survival and overall survival rates were 46% (95% CI 36-56) and 57% (95% CI 47-67), respectively. The primary tumour size and initial metastases correlated with a higher risk of event. In these 106 osteosarcoma adult patients, API-AI proved feasible with no excess of toxicity, and favourable activity despite poor-prognosis factors.