Analysis of dynamic hepatobiliary contrast-enhanced MRI signal intensity after Yttrium-90 radioembolization with glass microspheres for the treatment of hepatocellular carcinoma

Purpose To analyze hepatobiliary specific contrast agent (HBA) dynamic MRI signal intensity (SI) differences between treated liver (TL) and untreated liver (UL) parenchyma in patients following transarterial radioembolization (TARE) for hepatocellular carcinoma (HCC) using yttrium-90 containing glass microspheres. Materials This was a single institution retrospective study of patients with HCC treated with lobar or segmental TARE who received pre- and post-treatment HBA multiphase MRI within a 3-year period. Patients with prior locoregional therapies or multiple TAREs were excluded. SI was obtained by drawing a 2D ROI on T1-weighted non-contrast, arterial (25 s.), portal venous (60 s.), transitional (180 s.), and hepatobiliary (HB) (1200 s.) phase sequences in the (TL) angiosome and UL. HB phase signal enhancement characteristics were correlated with TARE dose thresholds ( 190 Gy) using the medical internal radiation dose (MIRD) methodology. Results 282 patients received TARE using glass microspheres during the study period and 58 patients who met inclusion criteria were analyzed. Median dose was 141.5 Gy MIRD [IQR 122.0, 161.5; range 100–540 Gy]). Statistically significant differences were present between treated and non-treated liver on non-contrast (− 28.0, p = 0.003), arterial (38.5, p = 0.013), and HB phases (− 95.8, p ≤ 0.001). Median follow-up time to furthest post-treatment MRI was 6 months (range 3-11 months). There was no significant SI difference on portal venous or transitional phases. HB phase SI changes in the TL compared to UL were significant at all TARE dose thresholds ( p < 0.05). Conclusions SI differences between treated and untreated liver after TARE are most significant on the HB phase and present at all evaluated dose levels at a median of 6 months after treatment. These findings support the parenchymal ablative potential for TARE and the necessity to consider liver function loss within targeted liver volumes.