MYC amplification and protein expression as prognostic markers in pediatric and young adult osteosarcoma.

Abstract

10043 Background: Risk stratification in osteosarcoma relies on metastatic status and tumor necrosis after chemotherapy. Despite a complex genomic landscape, genomic biomarkers are not yet used for predicting therapy response. MYC amplification has previously been identified as a potential biomarker for chemotherapy resistance but studies of MYC expression are limited. This study evaluated the relationship between MYC amplification and protein expression and between these biologic features and survival in pediatric and young adult osteosarcoma. Methods: In a cohort of 93 patients with high-grade osteosarcoma, MYC copy number was evaluated using a targeted sequencing panel, and MYC protein expression was quantified via IHC with an H-Score. The relationship between copy number and protein expression were evaluated via spearman correlation. Amplification (AMP) was defined as > 7 copies, and high expression (EXP) as > 175 H-score. The primary outcome, overall survival (OS), was assessed using Kaplan Meier analysis (median [IQR]) for unadjusted models and cox proportional hazard analysis (HR±SE) with models adjusted for metastatic status at diagnosis. Results: Among the 93 patients, 64% were male, with a median age of 14 years [range: 4–29]. MYC AMP was present in 16 (17%) patients, high MYC EXP in 19 (20%), and both AMP + high EXP in 8 (9%) patients. Copy number status was positively correlated with expression (r = 0.57, p < 0.0001). Patients with AMP were more likely metastatic at diagnosis (75% vs 38%, p = 0.011) and more so for AMP + high EXP (100% vs 39%, p = 0.0009), but not high EXP alone (63% vs 39%, p = 0.11). OS was reduced for AMP compared to non-AMP (median OS: 1.3 [1.0, 2.6] vs 6.2 [2.8, 12.8] years, p < 0.0001; Adj HR 3.2±0.4, p = 0.001) and high EXP compared to low EXP (median OS: 1.5 [1.0, 3.1] vs. 6.2 [2.8, 12.8] years, p < 0.0001; Adj HR 5.6±0.4, p < 0.0001). A dose-response relationship was seen with higher copy number or expression linked to reduced OS. Compared to non-AMP + low EXP (median OS: 7.23 [3.1, 12.8] years) there was reduced OS for AMP-only/high EXP-only (median OS: 3.1 [2.0, 4.0] years, p = 0.01; Adj HR 2.8±0.4, p = 0.009) and further reduced OS for AMP + high EXP (median OS: 1.0 [0.5, 1.0] years, p < 0.0001; Adj HR 15.3±0.5, p < 0.0001). Metastasis at diagnosis predicted reduced OS compared to localized disease (median OS: 2.0 [1.0, 4.5] vs 6.2 [4.0, 12.8] years, p < 0.0001), but age, sex, and tumor necrosis were not associated with OS. Conclusions: MYC amplification and protein expression are positively correlated, and both independently and synergistically predict poor OS in osteosarcoma, even after adjusting for metastatic status at diagnosis. Incorporating MYC amplification and expression status into risk stratification may help identify patients with worse prognosis. These data also underscore the need for therapeutic approaches tailored to the genetic basis of osteosarcoma tumor biology.