Proximal Tumor Location and Fluid-fluid Levels on MRI Predict Resistance to Chemotherapy in Stage IIB Osteosarcoma

• Published in: Clinical orthopaedics and related research Vol. 472; no. 6; pp. 1911 - 1920
• Main Authors: Jeon, Dae-Geun, Song, Won Seok, Cho, Wan Hyeong, Kong, Chang-Bae, Cho, Sang Hyun
• Format: Journal Article
• Language: English
• Published: Boston Springer US 01-06-2014; Lippincott Williams & Wilkins Ovid Technologies
• Subjects: Adolescent; Adult; Aged; Antineoplastic Agents - therapeutic use; Bone Neoplasms - drug therapy; Bone Neoplasms - mortality; Bone Neoplasms - pathology; Chemotherapy, Adjuvant; Child; Child, Preschool; Clinical Research; Conservative Orthopedics; Disease-Free Survival; Drug Resistance, Neoplasm; Female; Humans; Kaplan-Meier Estimate; Logistic Models; Magnetic Resonance Imaging; Male; Medicine; Medicine & Public Health; Middle Aged; Multivariate Analysis; Neoadjuvant Therapy; Neoplasm Staging; Odds Ratio; Orthopedics; Osteosarcoma - drug therapy; Osteosarcoma - mortality; Osteosarcoma - pathology; Predictive Value of Tests; Proportional Hazards Models; Retrospective Studies; Risk Factors; Sports Medicine; Surgery; Surgical Orthopedics; Time Factors; Treatment Failure; Tumor; Tumor Burden; Young Adult; Histologic Response; Proximal Location; Osteosarcoma; Tumor Enlargement; Independent Poor Prognostic Factor
• ISSN: 0009-921X; 1528-1132
• DOI: 10.1007/s11999-014-3521-1

Background
Primary tumor growth during neoadjuvant chemotherapy is believed to be a sign of resistance to chemotherapy (chemoresistance), and often is associated with poor histologic response, local recurrence, and poorer survival. Currently there are no proven indicators to predict poor response to chemotherapy at the time of diagnosis.
Questions/purposes
We asked (1) what clinicopathologic factors present at diagnosis predict primary tumor growth during neoadjuvant chemotherapy, (2) what factors at presentation predict survival, and (3) when the factors at presentation and the treatment-related factors are considered, what factors independently correlate with survival.
Methods
We studied 567 patients with Stage IIB osteosarcomas. The factors assessed included age, sex, location, pattern on plain radiographs (radiodense, radiolucent, mixed), MRI findings, pathologic subtype, initial tumor volume determined by MRI, tumor volume change after chemotherapy, surgical margin, and histologic response to preoperative chemotherapy. Logistic modeling was used to identify risk factors.
Results
Independent risk factors associated with primary tumor growth after neoadjuvant chemotherapy were proximal tumor location (p < 0.01; relative risk [RR], 2.41; 95% CI, 1.5–3.86) and fluid-fluid level on initial MRI (p < 0.01; RR, 5.56; 95% CI, 3.48–8.87). Among factors at presentation, large initial tumor volume (p < 0.01; RR, 1.58; 95% CI, 1.22–2.04), proximal tumor site (p < 0.01; RR, 1.61; 95% CI, 1.19–2.19), and presence of fluid-fluid level (p < 0.01; RR, 1.83; 95% CI, 1.37–2.5) independently predicted reduced event–free survival. When we consider the factors at presentation and treatment-related factors, large initial tumor volume (p < 0.01; RR, 1.54), tumor growth after neoadjuvant chemotherapy (p < 0.01; RR, 3.88), inadequate surgical margin (p < 0.01; RR, 2.42), and poor histologic response (p 
=
 0.03; RR, 1.43) were independent poor prognostic factors of event-free survival.
Conclusions
Proximal tumor location and the presence of the fluid-fluid level on initial MRI were predictors of tumor progression and poor survival in patients presenting with Stage IIB osteosarcomas. If confirmed in other studies, patients with these risk factors should be considered for trials of other treatment strategy.
Level of Evidence
Level III, prognostic study. See the Instructions for Authors for a complete description of levels of evidence.