Prognostic nomogram in patients with epithelioid sarcoma: A SEER-based study

Objective The prognostic factors for patients with epithelial sarcoma remain unclear. The study aims to develop a practical clinical nomogram that predicts prognosis in patients with ES using the Surveillance, Epidemiology, and End Results (SEER) database. Methods We extracted clinical data from 2004 to 2015 from the SEER database about patients with ES. All patients were randomly divided into training cohort and validation cohort. Kaplan-Meier analyses were used to compare outcomes between different subgroups. In order to estimate the chance of survival for patients with ES, we developed a nomogram. Nomogram performance was evaluated by discrimination and calibration. Additionally, an analysis of decision curves was conducted to evaluate the clinical usefulness of this newly developed model. Results In the primary cohort,320 met the inclusion criteria to be entered into this study. The median OS was 66.000 months (range 34.704 to 94.296 months), and the 1-, 3-, and 5-year OS rates were 70.7%, 56.1%, and 50.4%, respectively. For the validation cohort, we studied 136 consecutive patients. Age, primary site, grade, AJCC (American Joint Committee on Cancer) T, AJCC M, and surgery were included in the nomogram. The C-index values for the training set and validation set were 0.817 and 0.832, respectively. The calibration plots showed good agreement between the prediction and the observation. Based on the clinical decision curve, the model has a good clinical net benefit for ES patients. Conclusions It is the first study that developed an effective survival prediction model for patients with ES. Using this nomogram can assist in clinical decision-making as it has satisfactory accuracy. Even so, additional external validation is needed.

Automated summary

This study developed an effective survival prediction model for patients with epithelial sarcoma. Using this nomogram can assist in clinical decision-making with satisfactory accuracy. However, additional external validation is needed.