Gastrointestinal stromal tumor (GIST) is the most common subtype of sarcoma, with an incidence rate between 1.1 and 1.5 per 100,000 persons (1, 2). GIST predominantly originates in the digestive tract, with the stomach accounting for 55% of cases, the small intestine for 31%, and the colorectum for 6%. Fewer than 5% of gastrointestinal stromal tumors occur outside the digestive tract and are classified as extra-gastrointestinal stromal tumors (EGIST) (2). Although EGIST shares similar pathological characteristics with GIST, it presents a distinct spectrum of clinical features. EGIST cases exhibit a higher proportion of poorly differentiated tumors, and a higher proportion of patients with lymph node or distant metastasis (3, 4). Additionally, the majority of GISTs are caused by activating mutations in KIT (75%–80%) or PDGFRA (5%–10%). However, the incidence of KIT or PDGFRA mutation is only 40–50% in EGIST patients, the lower mutation rate of KIT or PDGFRA in EGIST patients may influence the TKI treatment effort for EGISTs (5). As the incidence of EGIST is comparatively low, with most cases reported individually, the genomic research on EGIST is limited compared with GIST, there is a lack of whole-genome sequencing, whole-exome sequencing, whole-transcriptome sequencing, and circulating tumor DNA sequencing analyses, which have been proved to predict the prognosis of patients and guiding precision treatment for GIST patients (6, 7). Despite the existing knowledge gap concerning EGIST, there is currently no specific guideline tailored for EGIST. The current understanding of EGIST is scarce, and potential prognostic factors are poorly elucidated, leading to poor prognostic outcomes for EGIST patients. Despite limited knowledge regarding EGIST’s clinical aspects, these tumors remain a significant subset of GISTs, representing a unique entity with distinct characteristics and outcomes that merit further investigation (8–10). Therefore, identifying the risk factors associated with EGISTs and developing predictive models is essential for assessing patient prognosis and informing treatment decisions.

Nomograms are widely used prognostic tools, especially for predicting tumor prognosis, as they provide personalized risk assessments and support treatment decisions, enhancing clinical management’s precision of clinical management (11). Presently, EGIST’s prognostic implications are not well understood. While some nomograms predict the prognosis of GIST patients, they generally include only cases of GIST occurring within the gastrointestinal system or a small proportion of EGIST cases. Given the evidence highlighting the heterogeneity in clinical characteristics and prognosis between GIST and EGIST, these existing nomograms are inadequate for accurately predicting outcomes in EGIST patients with EGIST (12–14). Therefore, it is crucial to develop models that address EGIST’s unique characteristics to improve patient management and guide personalized treatment strategies.

This study aims to identify the risk factors influencing survival in EGIST patients and to develop and validate a novel nomogram specifically tailored for EGIST. The Surveillance, Epidemiology, and End Results (SEER) database, a valuable resource for cancer research, includes approximately half of all cancer cases in the United States (15). Data for this study were selected from the SEER database. Our nomogram integrates key demographic, clinicopathological, and therapeutic variables to construct a comprehensive tool for estimating patient survival probabilities. This nomogram aids in guiding treatment decisions and optimizing clinical management, aiming to deepen our understanding of EGIST and enhance clinical decision-making for its effective treatment.

In this study, we identified patient age, tumor grade, tumor size, tumor SEER stage, surgical treatment, and household income as risk factors significantly associated with TSS in patients diagnosed with EGIST. Using data from the SEER database, we developed a nomogram to predict TSS for EGIST patients. This prognostic tool demonstrated robust performance in predicting patient prognosis and effectively-identified high-risk groups with poorer outcomes who may require closer monitoring and intervention. Our nomogram represents a step forward in understanding EGIST and supports the advancement of personalized medicine for patients with this rare tumor type.

EGIST, a rare form of GIST occurring outside the gastrointestinal tract, comprises approximately 5% of all GIST cases. While EGIST shares common c-kit and PDGFR-α mutations with GIST, it displays distinct clinical behaviors and prognostic characteristics (16). EGIST has received attention due to its aggressive nature and unique clinical outcomes compared to GIST within the gastrointestinal tract. Most existing studies on EGIST consist of case reports, with limited systematic analyses of clinical characteristics and prognostic factors. Using the SEER database, our study gathered a large cohort of EGIST patients, enabling us to systematically summarize clinical features and evaluate risk factors.

Our study identified several key sociodemographic characteristics among patients with EGIST. Socioeconomic factors, such as household income, have been shown to influence survival outcomes across various cancers (17). Our study further identified that low-household income is an independent risk factor for the survival of patients with EGIST, a finding that aligns with previous research on GIST (13). However, socioeconomic distributions varied significantly across different regions, given these disparities, it is essential to conduct cross-national comparative studies to explore the experiences and lessons learned by different countries and regions in addressing socioeconomic inequalities. Despite these variations, our study underscores the importance of public health strategies that target EGIST patients, particularly those with lower socioeconomic status. While gender has been identified as a prognostic factor for GIST survival, with evidence indicating that female GIST patients tend to have better survival outcomes than male patients, our research findings are consistent with prior EGIST studies, showing no significant difference in survival by gender (18–20). Further studies are needed to elucidate the detailed reasons. Furthermore, racial and marital status did not significantly affect EGIST survival, unlike GIST, where such factors have been influential (20, 21). The differences of sociodemographic characteristics between EGIST and GIST patients remain notable. The management of EGIST differs from that of GIST patients, which requires specific strategies. Developing personalized treatment plans and psychological support programs specifically designed for EGIST is crucial for improving the prognosis of these patients.

The oncologic characteristics of EGIST also differed from GIST. A clinical study comparing 62 EGIST cases with 570 GIST cases demonstrated that EGIST patients had a higher incidence of larger tumor sizes (>10 cm), more frequent signs of tumor necrosis, and higher mitotic rates, all of which correlate with poor prognosis (19). Our study also shows that 51.2% EGIST have tumor larger than 10cm, 44.4% were grade III-IV, 65.8% have a regional or distant tumor at diagnosis, which indicated EGIST have a higher degree of malignancy. EGIST’s origin remains controversial, with some researchers viewing it as a distinct subtype of GIST and others regarding it as a form of metastasis from the gastrointestinal tract (3, 5, 22). However, EGIST’s unique clinical features, prognostic factors, and outcomes support its classification as a separate subtype rather than a metastatic variant of GIST. Furthermore, our study excluded individuals with multiple tumors. The evidence of non-gastrointestinal origin further reinforces the notion that EGIST represents a distinct subtype of GIST. The higher degree of malignancy and poorer prognosis observed in EGIST underscore the limitations of current understanding of this tumor. To bridge the knowledge gap concerning EGIST, it is essential to explore its genomic characteristics and identify novel therapeutic biomarkers. Such efforts will enhance our comprehension of the pathogenesis underlying EGIST and ultimately contribute to improving its prognosis.

The treatment modalities for EGIST are typically based on GIST protocols, mainly include surgical resection, targeted therapy, and comprehensive treatment strategies. Surgery remains the primary treatment option for both GIST and EGIST (10, 23–25). However, high-risk EGIST patients often have poorer outcomes than GIST patients after surgery, indicating that adopting GIST-based strategies for EGIST may lead to suboptimal treatment (19). Radiotherapy has shown benefits for metastatic or progressive GIST (26). However, in two retrospective studies on EGIST, no patients received radiotherapy, and there was only one EGIST patient received radiotherapy in our study, making it difficult to evaluate its potential as a treatment option for EGIST (19, 27). More studies are needed to determine whether, and which subgroups of EGIST patients, will benefit from radiotherapy. For unresectable or metastatic EGIST, targeted therapies remain the preferred treatment option. However, when used as primary treatment, targeted therapies achieve a median overall survival of only 3.1 years for EGIST patients, compared to 6.9 years for GIST patients (27, 28). Given the absence of EGIST-specific treatment guidelines and large randomized controlled trials, treatment strategies for EGIST based on GIST protocols lead to poor outcome, further study should focus on the personalized and precision treatments for EGIST.

A nomogram designed for EGIST could address this need by predicting patient survival and categorizing risk levels, aiding treatment decisions, and potentially improving outcomes.

In GIST research, several nomograms have been developed to predict survival outcomes, incorporating age, race, tumor location, mitotic rate, chemotherapy, tumor size, stage, grade, and more (29–31). Sun et al. (32) created a nomogram for GIST patients post-surgery, while Bertsimas et al. (33, 34) developed prognostic models to predict the recurrence of GIST and assist in selecting patient groups suitable for adjuvant imatinib therapy. However, since patients with EGIST exhibit distinct prognostic risk factors compared to those with GIST, the existing nomograms are not applicable to EGIST patients. Our study used prognostic factors including age, tumor grade, size, tumor SEER stage, surgery, and household income to develop a nomogram predicting TSS in EGIST patients. The ROC and C-index values demonstrated the nomogram’s accuracy, and calibration curves indicated high concordance with observed outcomes. DCA further supported the nomogram’s clinical applicability with a positive net benefit. To our knowledge, our study involved the largest cohort of EGIST patients and introduced the first nomogram explicitly designed for this tumor type. Additionally, we stratified patients into high- and low-risk groups using the nomogram, with the high-risk group showing significantly poorer TSS outcomes. While the SEER staging system was limited in identifying patients with poor prognoses, our risk stratification approach effectively differentiated these patients, potentially guiding personalized and targeted treatment.

Our study has several strengths. It included the largest known cohort of EGIST patients, allowing for a comprehensive evaluation of clinical characteristics and risk factors. Additionally, we restricted the study to patients with a single primary tumor located outside the gastrointestinal tract, reducing the potential for misclassification of EGIST as metastatic GIST. There are also some limitations in our study, the SEER database spans multiple decades, and data gaps exist for some variables. Additionally, some patients were missing key characteristics such as tumor grade and tumor size. Moreover, given the extremely low incidence rate of EGIST, it is hard to collect sufficient external validation data. Therefore, we used internal validation data to assess the reliability and reproducibility of the nomogram. Despite these limitations, the high AUC and C-index values indicate that our nomogram is accurate and practical, supporting its use in predicting prognosis and guiding treatment.

We identified age, household income, surgical treatment, tumor grade, tumor size, and tumor SEER stage as significant risk factors for TSS in EGIST patients. Our prognostic nomogram offers a valuable tool for predicting TSS, enhancing patient management, and guiding personalized medical treatment for EGIST patients.