A New Online Dynamic Nomogram: Construction and Validation of a Predictive Model for Distant Metastasis Risk and Prognosis in Patients with Gastrointestinal Stromal Tumors.

Background: Gastrointestinal stromal tumor (GIST) is the most common sarcoma of the digestive tract, among which patients with distant metastases tend to have a poor prognosis. This study aimed to develop a model for predicting distant metastasis in GIST patients and to develop two models for monitoring overall survival (OS) and cancer-specific survival (CSS) in GIST patients with metastasis. This would allow us to develop an optimal, individualized treatment strategy.

Methods: We reviewed demographic and clinicopathological characteristics data from 2010 to 2017 of patients diagnosed with GIST in the Surveillance, Epidemiology, and End Results (SEER) database. The data of the external validation group was reviewed from the Forth Hospital of Hebei Medical University. Univariate and multivariate logistic regression analyses were used to confirm the independent risk factors for distant metastasis in the GIST patients, and univariate and multivariate Cox regression analyses were performed to identify the independent prognostic factors for OS and CSS in the GIST patients with distant metastasis. Subsequently, three web-based novel nomograms were developed, which were evaluated by the receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA).

Results: Of the 3639 patients who met the inclusion criteria, 418 (11.4%) had distant metastases. The risk factors for distant metastasis in GIST patients included sex, primary site, grade, N stage, tumor size, and mitotic count. For OS, the independent prognosis factors for GIST patients with metastasis included age, race, marital, primary site, chemotherapy, mitotic count, and metastasis at the lung, and for CSS, age, race, marital, primary site, and metastasis at the lung were the independent prognosis factors. Three web-based nomograms were constructed based on these independent factors, respectively. The ROC curves, calibration curves, and DCA were performed in the training, testing, and validation sets which confirmed the high accuracy and strong clinical practice power for the nomograms.

Conclusion: Population-based nomograms can help clinicians predict the occurrence and prognosis of distant metastases in patients with GIST, which may be helpful for clinicians to formulate clinical management and appropriate treatment strategies.