Objective: Respiratory gated PET/CT (positron emission tomography/computed tomography) of the lung is expected to increase the accuracy of quantitative determinations in lesional activities, regardless of the gating method used; reasonably, respiratory gating should increase standard uptake value (SUV; and possibly decrease lesional size), on the basis of the reduction of the "smearing effect." However, literature data are very limited, particularly for in vivo studies. The objective of this article is to test the SUV variations in a large group of lung lesion studies.
Methods: A group of 26 consecutive positive studies (21 men, 5 women, age 36-84, mean 68), performed on patients referred to our institution for known or suspected lung cancer, are examined. All studies were performed both with conventional PET/CT total body scan and with Real-Time Position Management (RPM) triggered selective gated study of the thorax. Four studies were considered technically unsatisfactory and were discarded; the remaining 22 studies are the object of this work. Max lesional SUVs were evaluated in both settings by semi-automated algorithms; for the gated studies, both values of the bin that showed more relevant variations from the clinical routinary study ("best bin") and an average value that was calculated over all bins were determined. Results were compared on a one-to-one basis.
Results: In conventional, SUV showed a mean +/- standard deviation (SD) value of 9.2 +/- 6.9 (range 0.9-26). In the averaged gated studies, the mean +/- SD value was 13.4 +/- 11.7 (range 1.4-47); in the "best bin" dataset the mean +/- SD was 14.9 +/- 12.9, ranging from 1.6 to 53.1. In general, the use of respiratory trigger induced rather variable but overall consistent increases in SUV. If the percentage variations in the average trigger dataset are considered, there is an average increase of +60%, SD +/- 97 (P < 0.05). Similar results are found in the "best bin" dataset, the average percentage increase in SUV values being +77.2% (SD +/- 04.6).
Conclusions: In lung cancer, triggering procedures increase the signal to noise ratio. The increase in SUV determined by gating is very variable, but generally relevant. This could lead to an SUV values cut-off revision, and may have an impact on smaller lesions detection.