Radiological Assessment and Outcome of Local Disease Progression after Neoadjuvant Chemotherapy in Children and Adolescents with Localized Osteosarcoma

OBJECTIVE: We examined the interobserver reliability of local progressive disease (L-PD) determination using two major radiological response evaluation criteria systems (Response evaluation Criteria in Solid Tumors (RECIST) and the European and American Osteosarcoma Study (EURAMOS)) in patients diagnosed with localized osteosarcoma (OS). Additionally, we describe the outcomes of patients determined to experience L-PD. MATERIALS AND METHODS: Forty-seven patients diagnosed with localized OS between 2000 and 2012 at our institution were identified. Paired magnetic resonance imaging of the primary tumor from diagnosis and post-neoadjuvant chemotherapy were blindly assessed by two experienced radiologists and determined L-PD as per RECIST and EURAMOS radiological criteria. Interobserver reliability was measured using the kappa statistic (κ). The Kaplan Meier method and log-rank test was used to assess differences between groups. RESULTS: Of 47 patients (median age at diagnosis 12.9 years), 16 (34%) had L-PD (by RECIST or EURAMOS radiological definition). There was less agreement between the radiologists using EURAMOS radiological criteria for L-PD (80.9%, κ = 0.48) than with RECIST criteria (97.9%, κ = 0.87). Patients with radiologically defined L-PD had a 5-year progression-free survival (PFS) of 55.6%, compared to a 5 year-PFS of 82.7% in the group of patients without L-PD (n = 31) (Log rank p = 0.0185). CONCLUSIONS: The interobserver reliability of L-PD determination is higher using RECIST than EURAMOS. RECIST can be considered for response assessment in OS clinical trials. The presence of L-PD was associated with worse outcomes.