Neoadjuvant therapy for localized extremity osteosarcoma.: Experience of the Cooperative Osteosarcoma Study Group COSS with 925 patients

Background: Owing to twenty years of multicentric interdisciplinary cooperation, the COSS group has been able to collect data on a large group of osteosarcoma patients treated by neoadjuvant therapy. This paper reviews results achieved in patients with localized extremity tumors. Patients and Methods:Inclusion criteria: Registration into a completed neoadjuvant COSS-Study. Histologically confirmed, primary, localized, high-grade, central osteosarcoma of an extremity; age < 40 years; no pretreatment; interval diagnosis to chemotherapy less than or equal to 3 weeks; no severe comorbidity. Chemotherapy: PID-methotrexate +/- doxorubicin +/-, cisplatin +/- ifosfamide +/- BCD. Scheduled local therapy: Surgery. Results: 925 evaluable patients from 101 institutions. Median age 15 years, m:f 1.4: 1. Primary site: femur 510, tibia 251, humerus 100, fibula 51, other 13. Tumor-size <1/3 of the involved bone 616, greater than or equal to 1/3 304. Definitive surgery in 903/925 cases, 443 limb salvage procedures. Good response (>90% necrosis) in 469/806 (58.2%) evaluated tumors. Median follow-up for surviving patients: 5.42 years. Actuarial survival after 5 and 10 years: 72.5% (95%-CI 69.3-75.7) and 66.3% (62.5-70.0), relapse-free 62.1% (58.7-65.4) and 59.4% (55.8-63.0). 683/925 alive (601 first remission), 242 deceased (212 tumor progression, 30 other causes). 66.2% (97.3%) of all relapses within 2 (5) years. Prognosis correlates with tumor-size(< vs, greater than or equal to 1/3: 69.9% vs. 58.3% at 10 years) and -site (tibia: 74.2%, humerus: 54.5%) and -response (,good vs. poor: 78.2% vs. 52.5%) (all p<0.01). Actuarial 10-year survival by response grading I-VI according to Salzer-Kuntschik 80.9%,82.8%,71.1%, 60.7%, 47.7%, 27.3%. COSS-studies with preoperative 4-drug therapy more efficacious than less aggressive protocols. No impact of doxorubicin scheduling (sequential: rapid vs. 48h-continuous infusion) or cisplatin scheduling (randomized: 5h vs. 72h-infusion) on prognosis detected. Conclusions: Intensive multiagent chemotherapy and delayed surgery for localized extremity osteosarcoma led to excellent oncologic results in the COSS-studies. Tumor-size, -site, and -response as well as the intensity of upfront chemotherapy correlated with outcome, Giving doxorubicin and cisplatin by continuous infusions did not result in discernible prognostic disadvantages.