Undifferentiated pleomorphic sarcoma (UPS) is considered one of the most common types of soft tissue sarcoma (STS). Current studies he shown that the prognosis of UPS is related to some of its clinical characteristics, but no survival prediction model for the overall survival (OS) of UPS patients has been reported. The purpose of this study is to construct and validate a nomogram for predicting OS in UPS patients at 3, 5 years after the diagnosis.
MethodsAccording to the inclusion and exclusion criteria, 1079 patients with UPS were screened from the SEER database and randomly divided into the training cohort (n = 755) and the validation cohort (n = 324). Patient demographic and clinicopathological characteristics were first described, and the correlation between the two groups was compared, using the Kaplan–Meier method and Cox regression analysis to determine independent prognostic factors. Based on the identified independent prognostic factors, a nomogram for OS in UPS patients was established using R language. The nomogram’s performance was then validated using multiple indicators, including the area under the receiver operating characteristic curve (AUC), consistency index (C-index), calibration curve, and decision curve analysis (DCA).
ResultsBoth the C-index of the OS nomogram in the training cohort and the validation cohort were greater than 0 .75, and both the values of AUC were greater than 0.78. These four values were higher than their corresponding values in the TNM staging system, respectively. The calibration curves of the Nomogram prediction model and the TNM staging system were well fitted with the 45° line. Decision curve analysis showed that both the nomogram model and the TNM staging system had clinical net benefits over a wide range of threshold probabilities, and the nomogram had higher clinical net benefits than the TNM staging system as a whole.
ConclusionWith good discrimination, accuracy, and clinical practicability, the nomogram can individualize the prediction of 3-year and 5-year OS in patients with UPS, which can provide a reference for clinicians and patients to make better clinical decisions.
Keywords: Undifferentiated pleomorphic sarcoma, Survival, SEER database, Nomogram
BackgroundSoft tissue sarcomas (STS) are heterogeneous and rare tumors showing mesenchymal differentiation that accounts for less than 1% of all malignant neoplasms, of which undifferentiated pleomorphic sarcoma (UPS) is considered to be one of the most common types of soft tissue sarcoma in adults [1, 2]. In recent years, statistical prediction models such as nomograms he been increasingly used to predict the probability of clinical events, and for clinicians and cancer patients, providing individualized prognostic information is essential for clinical decision-making. The nomogram can convert the complex regression equation into a visual graph, individualize the prediction of the prognosis of cancer patients, and make the survival prognosis results more intuitive and easy to understand, so it has a great application prospect in clinical work. In the past, some researchers collected the clinical information of UPS patients from the SEER database from 1990 to 2015 and established a CSS survival prediction model for UPS patients [3]. Since the SEER database added tumor size, tumor depth and radiotherapy of patients diagnosed after 2004, this study selected UPS patients diagnosed between 2004 and 2015 and collected the clinical data of UPS patients through retrieval to determine the independent prognostic impact. Based on the information above, the study established the OS nomogram survival prediction model for UPS patients. Moreover, the nomogram survival prediction model constructed in this study added three more variables, tumor size, tumor depth and radiotherapy, which can better provide a reference for on-the-spot treatment decisions of UPS.
Materials and methods Research object Data acquisitionThe clinical data of patient with UPS were retrieved from SEER database through SEER*Stat8.3.9 software (http://seer.cancer.gov//seerstat/), including age, sex, race, tumor site, tumor size, tumor depth, T stage, N stage, M stage, tumor grade, surgery, radiotherapy, chemotherapy, survival outcome, survival time, etc. The data in the SEER database are publicly and freely ailable and has no privacy implications. The application and warranty for use of data had been signed, and we obtained the corresponding login account and permission of data use (username: 19352-Nov2019). Therefore, this study did not need to provide the approval and informed consent of the institutional ethical review board.
Inclusion criteriaThe year of diagnosis was 2004–2015;
Histopathological diagnosis in accordance with the International Classification of Diseases Oncology Album Third Edition (ICD-O-3) classification of UPS (8830/3);
The primary site was in the subcutaneous soft tissue (Soft Tissue including Heart in ICD-O-3/WHO 2008 site code. Primary Site was limited to the subcutaneous soft tissue, and corresponding code was C49.0-C49.9);
Exclusion criteriaPatients with single and non-primary UPS;
Patients with incomplete clinical information, such as race, tumor site, tumor size, T stage, N stage, M stage, tumor grade and other unknown clinicopathological information;
The source of patient reports was only limited autopsy or death certificate;
Patients with unknown cause of death;
Patients with age