Evaluation of 68Ga-DOTATATE and 18F-FDG PET in grading non-functioning pancreatic neuroendocrine neoplasms <20 mm without metastasis.

BACKGROUND

The clinical management of small (<20 mm), nonfunctional, nonmetastatic pancreatic neuroendocrine neoplasms (panNENs) relies heavily on accurate pathological grading, as Grade 1 (G1) tumors are often managed with active surveillance, whereas Grade 2 (G2) tumors warrant surgical resection. Currently, limited biopsy sampling may misrepresent the true tumor grade due to intratumoral heterogeneity, and conventional imaging lacks grading accuracy. Although 68Ga-DOTA0-Tyr3-octreotate (68Ga-DOTATATE) and 18F-fluorodeoxyglucose (18F-FDG) positron emission tomography (PET) can reflect tumor differentiation and aggressiveness, their grading efficacy-especially for small, localized panNENs-remains controversial. Therefore, this study evaluates the ability of dual-tracer PET imaging to accurately grade these specific lesions to guide optimal management decisions.

METHODS

A retrospective study was conducted between January 2019 and January 2025 on patients with histologically confirmed panNENs. All 34 patients (with 41 lesions) underwent 68Ga-DOTATATE PET imaging, and a subset of 19 patients (with 24 lesions) also underwent 18F-FDG PET imaging. The maximum standardized uptake values (denoted as G-SUV for 68Ga-DOTATATE and F-SUV for 18F-FDG) were measured in lesions with surgically confirmed histological grades. Statistical analyses were performed using the Mann-Whitney U test for continuous variables and the Chi-squared test for categorical variables. The Spearman correlation coefficient was used to evaluate both the strength and direction of the bivariate relationships. The optimal cutoff for distinguishing G1 from G2 was determined by receiver operating characteristic (ROC) analysis. Additionally, the clinical utility of visual assessment was evaluated.

RESULTS

No significant differences in G-SUV or the G-SUV/F-SUV ratio were found between G1 and G2 (P=0.355 and P=0.861, respectively). In contrast, F-SUV was significantly lower in G1 than in G2 (P=0.003), with an optimal cutoff value of 2.5, yielding a sensitivity of 73% and a specificity of 89%. Although visual assessment of neither 68Ga-DOTATATE PET nor 18F-FDG PET could distinguish G1 from G2 (P=0.548 and P=0.102, respectively), 18F-FDG PET demonstrated a high positive predictive value (89%) when using pancreatic background as the reference.

CONCLUSIONS

68Ga-DOTATATE PET could not be used to predict small panNENs grade. In contrast, 18F-FDG PET showed potential, and if validated by prospective data, it might help in selecting appropriate treatment options for patients with panNENs.