BACKGROUND: In the revised AJCC staging system a maximal tumor diameter of 8 cm was adopted as a cutoff for the subdivision of IIA and IIB osteosarcoma, but this cutoff was chosen based on limited information. METHODS: We retrospectively reviewed 347 patients with stage II osteosarcoma that were treated at our institute. We plotted a receiver operating characteristic (ROC) curve of maximal tumor diameter for the prediction of subsequent metastasis, and calculated diagnostic indices according for different cutoffs. RESULTS: A maximal tumor diameter greater than 8 cm was found to predict subsequent metastasis with a sensitivity of 76.3%, a specificity of 49.5%, and a positive predictive value of 49.0% Almost half of stage IIB patients subsequently developed metastasis, whereas only a quarter of stage IIA patients did so. Tumor size had no prognostic relevance for proximal humeral tumors. CONCLUSIONS: The present study shows that an 8 cm maximal tumor diameter cutoff is useful for subdividing AJCC stage II osteosarcoma patients in terms of predicting of a subsequent metastatic event. Our results suggest that the IIA and IIB subdivision in the AJCC staging system provides a basis for risk-adapted therapy when used in combination with other prognostic factors. (c) 2008 Wiley-Liss, Inc.
BACKGROUND: In the revised AJCC staging system a maximal tumor diameter of 8 cm was adopted as a cutoff for the subdivision of IIA and IIB osteosarcoma, but this cutoff was chosen based on limited information. METHODS: We retrospectively reviewed 347 patients with stage II osteosarcoma that were treated at our institute. We plotted a receiver operating characteristic (ROC) curve of maximal tumor diameter for the prediction of subsequent metastasis, and calculated diagnostic indices according for different cutoffs. RESULTS: A maximal tumor diameter greater than 8 cm was found to predict subsequent metastasis with a sensitivity of 76.3%, a specificity of 49.5%, and a positive predictive value of 49.0% Almost half of stage IIB patients subsequently developed metastasis, whereas only a quarter of stage IIA patients did so. Tumor size had no prognostic relevance for proximal humeral tumors. CONCLUSIONS: The present study shows that an 8 cm maximal tumor diameter cutoff is useful for subdividing AJCC stage II osteosarcomapatients in terms of predicting of a subsequent metastatic event. Our results suggest that the IIA and IIB subdivision in the AJCC staging system provides a basis for risk-adapted therapy when used in combination with other prognostic factors. (c) 2008 Wiley-Liss, Inc.