Influencing Factors of Lung Shunt Fraction in Transarterial Radioembolization Treatment

Purpose We aimed to evaluate the influencing factors of lung shunt fraction (LSF) in patients who underwent transarterial radioembolization treatment. Methods We enrolled 105 patients who had either primary or metastatic liver disease. Planar and SPECT/CT images were obtained after intraarterial injection of 99m Tc-MAA to measure LSF. Planar-based LSFs (LSF planar ), calculated as the arithmetic mean (LSF AM ) and geometric mean (LSF GM ), were compared to LSF SPECT/CT . Subgroup analyses were conducted to assess the impact of scattering and spillover effects in patients with target lesions located near lung. Both lungs were segmented into lower, middle, and upper sections (RL, RM, RU, LL, LM, and LU). For control group, ten patients with normal lung perfusion scintigraphy were included. Results The mean of LSF SPECT/CT was significantly lower than the mean of LSF GM (6.5 ± 4.1 vs. 13.9 ± 8.4, p = < 0.001) and the mean of LSF AM (13.5 ± 8.4%, p = < 0.001). LSF planar methods overestimated LSF compared to LSF SPECT/CT . LSF GM divided by 2.3 can provide more accurate estimated LSF similar to LSF SPECT/CT . In subgroup analyses, we found that mean values of RL/RU, RM/RU, LL/LU, and LM/LU were 9.3 ± 7.4, 2.2 ± 1.2, 1.2 ± 0.4, and 1.2 ± 0.2, respectively. In control group, mean values of RL/RU, RM/RU, LL/LU, and LM/LU were within a range of 0.9–1.1. Conclusion LSF SPECT/CT with attenuation correction can provide more accurate results compared to LSF planar before TARE treatment. In patients who had liver lesions located adjacent to lung, the possible miscalculation of LSF SPECT/CT from scattering and spillover effect could be eliminated by exclusion of 3 cm diameter from inferior border of the right lung.