Risk stratification based on the clinical factors at diagnosis is closely related to the survival of localized osteosarcoma

Background Survival of osteosarcoma patients has reached a plateau with the addition of chemotherapy and in part predicted based on histologic response. Risk-adapted therapy might be an alternative approach. We aimed to identify risk groups using clinical variables available at time of diagnosis in order to better predict outcomes and form the basis for risk-adapted therapy. Methods This retrospective study analyzed 288 patients with high-grade osteosarcoma of their extremities. Clinicopathologic variables were analyzed to identify factors relevant to risk stratification. Results A risk assessment system was developed using age, maximal tumor length (MTL) and tumor location. Index scores were based on the sum total of factor scores, for each of the following: Age (years); “1” for ≥40 or 12–14 (female) or 13–15 (male), “0” for other ages; MTL (cm); “2” for ≥8, “1” for 6–8, “0” for ≤6; Tumor location; “1” for humerus, “0” for elsewhere. Index score 0 or 1 was assigned as low-risk, 2 as intermediate-risk, and 3 or 4 as high-risk. Of the 288 patients, 98 (34.0%) were designated as low-risk, 128 (44.4%) as intermediate-risk and 62 (21.6%) as high-risk. Risk group was related to histologic response and survival. While 63.3% of low-risk patients were good responders, only 43.0% and 33.9% of intermediate- and high-risk patients were good responders. Ten-year event-free survival (EFS) was 81.6 ± 3.9% for low-risk group, but 31.6 ± 6.0% for high-risk group. Conclusion We defined three risk groups that could be used as basis of risk-adapted therapy for osteosarcoma. Pediatr Blood Cancer 2009;52:340–345. © 2008 Wiley-Liss, Inc.