Background: Ewing's sarcoma (ES) is a prevalent bone malignancy. It is critical to explore new diagnostic and prognostic indicators because of the rapid progression of ES and the low survival rate of metastatic ES patients. However, few parameters of clinical significance have been found. The aim of this study was to establish a new classifier with clinical laboratory data to help ES detection and prognosis prediction.
Methods: A total of 135 ES patients, 150 healthy individuals, and 228 patients with primary benign bone lesions were included. Logistic regression on clinical laboratory indicators was conducted to establish the classifier, and then the classifier was assessed by drawing the receiver operating characteristic (ROC) curves. Patient survival was evaluated using the Kaplan-Meier method.
Results: We established the diagnostic classifier, called Ces, with clinical laboratory indicators to distinguish ES from healthy individuals. Ces showed great diagnostic performance in the test cohort (area under the receiver operating characteristic curve (AUC) 0.95) and could identify early-stage (AUC 0.93) and small-size (AUC 0.95) ES effectively. In addition, the classifier had good ability to differentiate ES from primary benign bone lesions (AUC 0.77 for Ces, AUC 0.83 for Ces + age). Furthermore, Ces was associated with tumor metastasis and event-free survival (EFS) of ES patients and showed better performance than lactate dehydrogenase (LDH) in prognosis prediction.
Conclusions: Our study indicates that Ces has the potential to be a non-invasive biomarker for ES diagnosis and prognosis.