A nomogram to predict metastasis of soft tissue sarcoma of the extremities

Abstract

Soft tissue sarcoma (STS) of the extremities are a rare tumor. Metastases develop in about 40%-50% of patients, most of whom die from their disease. We sought to identify potential risk factors associated with metastatic diseases upon presentation for patients with STS and established a reliable nomogram model to predict distant metastasis of STS at presentation. The current study retrospectively analyzed 3884 STS of the extremities or trunk patients from the Surveillance, Epidemiology, and End Results (SEER) database between 2010 and 2015. Based on patient registration, all patients were randomly allocated to training sets and validation sets (2:1). Then, univariate and binary logistic regression analysis was used to determine the significantly correlated predictors of metastasis. Finally, the nomogram model was established, using these predictors and validated it. 311 (8.21%) of the cases experienced distant metastatic disease was present at the time of presentation. The nomogram was developed from age, histology subtype, primary site, tumor size, grade and depth. Encouragingly, the nomogram showed favorable calibration with C-index 0.790 in the training set and 0.801 in validation set. The DCA showed that the novel model was clinically useful. This nomogram model had a high precision to predict the metastasis of soft tissue sarcoma of the extremities. We expect this model could be used in different clinical consultation and established risk assessment.

Figure 1

The graphs show the nomograms which predict the probability of metastasis of STS patients of the extremities or trunk. Points of each variable were acquired by drawing a vertical line between each variable and the Points scale. By totaling the points of each variable, we then draw a vertical line between the Total Points scale and risk of metastasis scale to calculate the probability of metastasis. Age: 1: age<36; 2: age >36, histology: 1: Fibrosarcoma 2: Leiomyosarcoma 3: Liposarcoma 4: Malignant fibro histiocytoma 5: MPNST 6: Synovial 7: Other, tumor site: 1: Upper extremity; 2: Lower extremity; 3: Thoracic or trunk, tumor size: 1: <7.6cm; 2:≥7.6cm, grade: 1: I; 2:II; 3:III; 4: IV, depth: 1: Superficial; 2: Deep.

Figure 3

Receiver operating characteristic (ROC) curve analysis for distant metastasis with Soft tissue sarcoma of the extremities in the training set (a) and the testing set (b). The AUC was calculated, and its 95% CI was estimated by bootstrapping. The P values were two-sided. ROC = receiver operating characteristic; 95% CI = 95% confidence interval.

Figure 2

Calibration curves for predicting distant metastasis in the training set and validation set.

Figure 4

Decision curve analysis (DCA) for the new model in the training set (a) and the testing set (b). The horizontal solid black line represents the assumption that no patients will experience the event, and the solid gray line represents the assumption that all patients will relapse. On decision curve analysis, the nomogram showed more net benefit compared with that of treat-all or treat-none across a range of threshold probabilities.