Poster — Thur Eve — 65: Comparison of Two Fixed‐Size ROI Methods for Tumour Response Assessment Using FDG‐PET

Positron emission tomography (PET) with fluorodeoxyglucose (FDG) has shown promise for quantitative tumour response assessment. Commonly, a fixed‐sized region of interest (ROI) is centered at the maximum uptake point in the pre‐treatment (Pre‐Tx) study. However, for intra‐treatment (Intra‐Tx) measurement, the ROI is centered at the maximum uptake point or at the same location as the Pre‐Tx ROI. Therapy‐induced geometric changes of tumour and normal structures may affect the latter ROI method. We have quantified these geometric changes and compared the two ROI methods. Methods: In 15 patients, the two ROI methods were compared by placing the Intra‐Tx ROI at either the maximum uptake (ROIpeak) or the same location as Pre‐Tx (ROIsame). Normal structures and tumours were contoured both in Pre‐Tx and Intra‐Tx in 10 independent patients to quantify geometric changes measured by percentage volume losses and center of mass shift. Results: The response measure using ROIsame was on average 8.2% higher than ROIpeak, but they were highly correlated (Pearson r =0.94). The median volume loss was 76%, 58%, and 30% for the tumour, nodal disease, and salivary glands respectively. Other soft tissues and bones did not show significant volume losses. The median shift for all tumours was 5.9 mm (range, 1.8–16.7 mm). Conclusion: Tumour shift during the therapy will affect quantitative response assessment if one chooses the ROIsame method. The two ROI methods will probably have similar accuracies in differentiating responders from non‐responders but different response threshold values. ROIpeak has advantages over ROIsame due to its simplicity and observer‐independence.