Limited predictive value of pretreatment P-glycoprotein immunostaining for chemotherapy efficacy and survival in long-follow-up osteosarcoma patients.

The clinical utility of P-glycoprotein (P-gp) in osteosarcoma remains controversial. This retrospective cohort study aimed to investigate the association between P-gp expression and clinicopathological factors, chemotherapy (CTx) response, and prognosis in osteosarcoma. Twenty-three osteosarcoma patients treated at our institution between 2007 and 2013 were included. P-gp expression rate was measured by immunohistochemical staining of biopsy specimens obtained at the time of diagnosis. CTx response was evaluated by image evaluation (Response Evaluation Criteria in Solid Tumors criteria) and pathological evaluation (Rosen and Huvos classification). The association between P-gp expression rate and clinicopathological factors, CTx response, disease-free survival (DFS), and overall survival (OS) was statistically analyzed. The P-gp expression rates ranged from 1.7% to 67.9% (mean 46.0%). No significant association was found between P-gp expression rate and surgical stage, CTx response, or clinical outcome. The receiver operating characteristic curve analysis revealed an optimal P-gp cutoff value of 47.0% for predicting poor CTx response, with a sensitivity of 73% and a specificity of 67%. When patients were divided into P-gp positive (n = 13) and negative (n = 10) groups based on the 47.0% cutoff value, no significant differences were observed between the 2 groups in terms of surgical stage, CTx response, OS, and DFS. Although the P-gp positive group showed a trend towards worse OS and DFS, no statistically significant differences were observed (OS: P = .73, DFS: P = .32). P-gp expression in osteosarcoma was not significantly associated with surgical stage, CTx response, or prognosis. These findings suggest that P-gp may not be a useful biomarker for CTx selection or prognosis prediction in osteosarcoma.