Abstract
Introduction
Gastrointestinal stromal tumors (GISTs) are the most common mesenchymal tumors of the gastrointestinal tract. Imatinib and later-line tyrosine kinase inhibitors have transformed management, yet outcomes still vary widely. Social determinants of health are well established drivers of cancer outcomes, influencing both access to care and survival. In GIST, these factors remain far less studied than tumor biology. This study examines how demographic, socioeconomic, and clinical factors shape treatment and survival.
Methods
We conducted a population-based retrospective cohort study of adults diagnosed with GIST in SEER from 2010 to 2024. Patients missing grade, T stage, or treatment status, and seven receiving intraoperative systemic therapy, were excluded. Outcomes were overall (OS) and disease-specific survival (DSS). Multivariable Cox proportional hazards models estimated the independent associations of demographic (age, sex, race/ethnicity, marital status), socioeconomic and health-system (income, urban residence, Commission on Cancer accreditation), tumor (grade, T/N/M stage), and treatment factors with each outcome. ] Interactions between the treatment sequence and survival time were also evaluated to account for time-varying treatment effects.
Results
Among 2,857 patients, 53.8% were female, 49.8% non-Hispanic White, 23.7% non-Hispanic Black, and 60.5% married. Overall, 96.6% received surgery, systemic therapy, or both, and 72.6% underwent surgery alone. At last follow-up 89.5% were alive, with GIST-specific and non-GIST deaths nearly equal (5.3% vs 5.2%). Non-Hispanic Black patients had higher overall (HR 1.77, 95% CI 1.31-2.38) and disease-specific mortality (HR 1.52, 95% CI 1.00-2.31) than non-Hispanic White patients. Unmarried patients had approximately twice the hazard of death (OS HR 1.95, 95% CI 1.52-2.50; DSS HR 2.12, 95% CI 1.49-3.01). Cases reported by Commission on Cancer-accredited facilities had lower disease-specific mortality (HR 0.55, 95% CI 0.35-0.88). Income, urban residence, and time from diagnosis to treatment were not associated with either outcome. Distant metastasis (OS HR=3.96, 95% CI=2.80-5.60; DSS HR=4.67, 95% CI=3.08-7.08), high grade (OS HR=1.62, 95% CI=1.17,1.96; DSS HR=1.76, 95% CI=1.23-2.53), male sex, and older age (OS HR=1.32, 95% CI=1.25-1.39; DSS HR=1.11(1.04-1.19) predicted worse survival outcomes. Excluded patients with missing values of the study variables differed significantly across demographic, clinical, and health-system factors.
Conclusion
Demographic and health-system factors were independently impact survival in GIST after adjustment for tumor burden and treatment. Being non-Hispanic Black, unmarried, and older age were the strongest demographic predictors of adverse survival outcomes, whereas income and rurality were not, indicating disparities in GIST is not primarily driven by area-level economics. These differences persisted in the surgically treated cohort, indicating access to operation alone may not close outcome gaps. Notably, care at Commission on Cancer accreditation significantly improved disease-specific survival, and adjuvant systemic therapy conferred significant early survival benefits. Findings should be interpreted considering selection bias inherent to complete-case analysis and short follow-up.