Conventional Hepatic Volumetry May Lead to Inaccurate Segmental Yttrium-90 Radiation Dosimetry

Objective To compare radioembolization treatment zone volumes from mapping cone beam CT (CBCT) versus planning CT/MRI and to model their impact on dosimetry. Methods Y90 cases were retrospectively identified in which intra-procedural CBCT angiograms were performed. Segmental and lobar treatment zone volumes were calculated with semi-automated contouring using Couinaud venous anatomy (planning CT/MRI) or tumor angiosome enhancement (CBCT). Differences were compared with a Wilcoxon signed-rank test. Treatment zone-specific differences in segmental volumes by volumetric method were also calculated and used to model differences in delivered dose using medical internal radiation dosimetry (MIRD) at 200 and 120 Gy targets. Anatomic, pathologic, and technical factors likely affecting segmental volumes by volumetric method were evaluated. Results Forty segmental and 48 lobar CBCT angiograms and corresponding planning CT/MRI scans were included . Median Couinaud- and CBCT-derived segmental volumes were 281 and 243 mL, respectively ( p = 0.005). Differences between Couinaud and CBCT lobar volumes (right, left) were not significant ( p = 0.24, p = 0.07). Couinaud overestimated segmental volumes in 28 cases by a median of 98 mL (83%) and underestimated in 12 cases by median 69 mL (20%). At a 200 Gy dose target, Couinaud estimates produced median delivered doses of 367 and 160 Gy in these 28 and 12 cases. At a 120 Gy target, Couinaud produced doses of 220 and 96 Gy. Proximal vs. distal microcatheter positioning, variant arterial anatomy, and tumor location on or near segmental watersheds were leading factors linked to volumetric differences. Conclusion Use of CBCT-based volumetry may allow more accurate, personalized dosimetry for segmental Y90 radioembolization.