FDG Extravasation in PET/CT imaging: A visual grading approach based on clinical observations

Abstract

Purpose This study aims to introduce a practical, visually based grading system for FDG extravasation in PET/CT imaging and to investigate its impact on semi-quantitative SUV measurements in reference regions. The objective is not only to quantify the effect but also to provide a framework for guiding clinical interpretation and reporting. Methods A total of 235 oncologic PET/CT scans were retrospectively analysed. Patients were categorised into four groups based on extravasation severity: Grade 0 (no extravasation), Grade 1 (minimal), Grade 2 (moderate), and Grade 3 (severe). A visual grading system was developed and applied to classify image quality degradation. Mean SUVmax and SUVmean values were recorded from liver and mediastinal blood pool regions. Statistical comparisons were conducted using ANOVA and post hoc testing. Results FDG extravasation was observed in 75 patients (31.9%). Among the 75 patients with extravasation, 43 were Grade 1 and 26 were Grade 2. Statistically significant increases were found in both liver and mediastinal SUV values in patients with Grade 1 and 2 extravasation compared to Grade 0 (p < 0.001 for all). For example, Liver SUVmax increased from 3.13 +/- 0.44 (Grade 0) to 3.58 +/- 0.50 (Grade 1), and MedSUVmean increased from 1.80 +/- 0.31 (Grade 0) to 2.17 +/- 0.40 (Grade 2), representing up to a 20.6% difference. Grade 3 cases did not show consistent trends. Conclusion Low-to-moderate FDG extravasation can significantly alter reference SUV values. Incorporating extravasation grading into clinical reporting or normalising reference SUV values may enhance diagnostic consistency and minimise the need for repeat imaging.