Initial application of respiratory-gated 201Tl SPECT in pulmonary malignant tumours

Abstract

Respiratory-gated thallium-201 chloride (201Tl) single photon emission computed tomography (SPECT) was used in preliminary investigations to reduce the adverse respiratory motion effects observed on standard ungated SPECT images and to obtain reliable fusion images with computed tomography (CT) in patients with malignant lung tumours. Fifteen patients with primary lung cancer (n=10) or metastatic lung tumours (n=5) underwent gated SPECT 20 min after intravenous injection of 148 MBq 201Tl, using triple-headed SPECT and laser light respiratory tracking units. Projection data were acquired by a step and shoot mode, with 20 stops over 120 degrees for each detector and a preset time of 30 s for each 6 degrees stop. Gated end-inspiratory and ungated images were obtained from 1/8 data centred at peak inspiration for each regular respiratory cycle and for the full respiratory cycle data, respectively. The degree and size of tumour 201Tl uptake were compared between these images by regions of interest (ROI) analysis. Gated SPECT images were registered with rest inspiratory CT images using an automated three-dimensional (3D) image registration tool. Registration mismatch was assessed by measuring the 3D distance of the centroid of 14 201Tl-avid peripheral tumours. Attenuation correction of gated SPECT images was performed using CT attenuation values of these fusion images. Gated SPECT images improved image clarity and contrast of tumour 201Tl uptakes compared with ungated images, regardless of the decreased count density due to the use of gated images. The lesion-to-normal (L/N) lung count ratios and ROI size in 18 well-circumscribed 201Tl-avid tumours were significantly higher and smaller on gated images (both P<0.0001). Gated images showed positive 201Tl uptakes in two small peripheral tumours, although negative on ungated images, and demarcated 201Tl-avid tumours from adjacent 201Tl-avid lymph node or surrounding focal 201Tl uptakes caused by other pathology, although these were not clearly demarcated on ungated images. On fusion images, gated images yielded a significantly better SPECT-CT matching compared with ungated images (P<0.0001). Fusion images accurately localized 201Tl uptakes of tumour/lymph node and other focal pathological/physiological conditions. Attenuation-corrected gated SPECT images further facilitated the detection of 201Tl uptake in small or deeply located lesions, with significantly increased L/N ratios. Gated SPECT images facilitate the detection of tumour 201Tl uptake and provide reliable SPECT-CT fusion images, which contribute to accurate interpretation and attenuation correction of Tl SPECT images.