18F-FDG PET/CT for Risk Stratification and Prognosis of Patients with Hypermetabolic Gastrointestinal Stromal Tumors

<b>Objectives:</b> We aimed to evaluate the value of various PET parameters derived from ¹⁸F- FDG PET/CT for risk stratification and prognosis of hypermetabolic gastrointestinal stromal tumors (GISTs). <b>Methods:</b> This study included 43 patients who underwent ¹⁸F-FDG PET/CT imaging with hypermetabolic (SUVmax > 2.5) GIST and underwent surgical treatment. Clinicopathological characteristics, risk stratification, PET parameters including standard uptake values (SUVs), metabolic tumor volume (MTV), total lesion glycolysis (TLG), and heterogeneity index (HI), and follow-up data were reviewed. The relationship between PET parameters and risk stratification based on the modified National Institutes of Health (NIH) criteria was analyzed. PET parameters were assessed to predict relapse-free survival (RFS) and overall survival (OS), based on Cox regression analysis and Kaplan-Meier analysis. <b>Results:</b> The median follow-up duration was 50 months. During follow-up, 11 patients (25.58%) experienced recurrence and 8 (18.60%) died. In risk stratification, the high-risk group exhibited more frequent extragastric location, larger tumor size, higher mitotic count, and elevated PET parameters except SUVmax. MTV (≤32.68 vs. >32.68, 95% CI: 1.358-72.048, <i>p</i> = 0.024) emerged as an independent PET parameter of risk stratification. In univariate analysis, tumor location (gastric vs. extragastric), SUVmax (≤10.25 vs. >10.25), and HI (≤2.44 vs. >2.44) were significant prognostic factors for RFS. Tumor location and SUVmax were significant to OS on univariate analysis. However, in multivariate analysis, only SUVmax (95% CI: 1.549-46.071, <i>p</i> = 0.014) was an independent prognostic factor for both RFS and OS. <b>Conclusions:</b>¹⁸F-FDG PET/CT demonstrates predictive value for hypermetabolic GIST patients. MTV derived from ¹⁸F-FDG PET/CT improves the ability of predicting risk stratification. SUVmax is an effective predictor of both RFS and OS.