Investigating ultra-low-dose total-body [18F]-FDG PET/CT in colorectal cancer: initial experience

Abstract

This study was to evaluate the effects of an ultra-low dose of [18F]-FDG on the image quality of total-body PET/CT and its lesion detectability in colorectal cancer (CRC). Sixty-two CRC patients who underwent total-body PET/CT (uEXPLORER, United Imaging Healthcare, Shanghai, China) with an ultra-low dose (0.37 MBq/kg) of [18F]-FDG were enrolled in this retrospective study. The PET images were reconstructed with the entire 15-min dataset first and then split into 13-, 8-, 5-, 4-, 3-, 2-, and 1-min duration groups to simulate fast scanning images. For simplicity, the images reconstructed with the data from 15 to 1 min were referred to as G15, G13, and so on until G1. Subjective image quality was assessed with 5-point Likert scales. The objective image quality parameters included the SUVmax, SUVmean, and signal-to-noise ratio (SNR) of the liver and blood pool and the SUVmax and tumor-to-background ratio (TBR) of the lesions. G15 served as the control to evaluate lesion detectability. A total of 62 patients (43 men, 19 women; age 41–88, mean\u2009±\u2009SD 64.0\u2009±\u200910.9 years) with 64 CRC primary tumor lesions and 10 low-grade intraepithelial neoplasia (LGIN) lesions were enrolled in this study. The subjective scores were highest for G15 (4.5\u2009±\u20090.5) and then decreased from G13 (4.3\u2009±\u20090.4) to G8 (3.7\u2009±\u20090.5). The liver SNR increased with the extension of acquisition time from G8 (17.2\u2009±\u20092.8) to G13 (20.6\u2009±\u20093.4) and G15 (21.9\u2009±\u20093.4). The liver SNR of G8 was not significantly different from that of G13 (p\u2009=\u20090.15) and was significantly different from that of G15 (p\u2009=\u20090.001). All 64 CRC lesions could be identified in all image groups, even on G1. One of ten LGINs was missed on G1, G2, and G3, and one LGIN was missed on G1, G2, G3, and G4. G15 served as the control, and 100% (48/48) lymph nodes could be found on G13 and G8 compared to 93.8% (45/48) lymph nodes on G5 and G4, 85.4% (41/48) lymph nodes on G3, 81.3% (39/48) lymph nodes on G2, and 77.1% (37/48) lymph nodes on G1. For liver metastases, there were no missed liver lesions on G13 and G8 and 3, 4, 6, 7, and 9 missed liver lesions on G5, G4, G3, G2, and G1, respectively. For other areas of metastasis, including the lung, peritoneum, and ovaries, there were no missed lesions in any group. Total-body PET/CT with an ultra-low dose of [18F]-FDG can maintain satisfactory image quality and lesion detectability in CRC.