A novel approach to radioembolization treatment planning using a hybrid [^99mTc]Tc-MAA contrast enhanced SPECT/CT.

Abstract

Dosimetry-guided treatment planning is central to radioembolization. Guidelines recommend segmenting volumes on baseline contrast-enhanced CT (ceCT) and registering to a SPECT/low-dose CT. However, this conventional approach is vulnerable to spatial misregistration and potential disease progression. This study evaluates a hybrid SPECT/ceCT-based planning methodology, hypothesising that it yields significantly different dosimetry metrics compared to the conventional workflow. Hybrid SPECT/ceCT was performed in 24 patients with hepatocellular carcinoma. Treatment planning followed both the conventional and hybrid workflow, where delineation was performed directly on SPECT/ceCT. Quantification of dosimetric and positional differences between volumes from the two workflows was performed using Simplicit⁹⁰Y (Mirada Medical) and SurePlan Liver Y90 (MIM) software, respectively. Outcomes were evaluated by Bland-Altman analysis and contour intercomparison indices. Direct delineation on SPECT/ceCT yielded higher index lesion (largest single tumour) doses (6% increased) (mean: 19.5 Gy, 95% limits of agreement [LoA]: -43.9 Gy, 83.0 Gy) and lower normal tissue doses (11% decrease) (mean: -17.6 Gy 95% LoA: -104.4 Gy, 69.3 Gy). Index lesion volume increase between baseline CT and SPECT/ceCT was 14% (95% LoA: -55.9%, 82.9%). Index lesion dose (p = 0.01) and volume (p = 0.01) differed significantly between workflows. The median time interval between baseline imaging and SPECT/ceCT was 6 weeks (IQR: 3.7 to 9.3 weeks). The median deviation of the index lesion between the baseline diagnostic image and hybrid SPECT/ceCT image was 4.9 ± 2.2 mm and there was an average DICE score of 0.6 ± 0.2. Hybrid [99mTc]Tc-MAA SPECT/ceCT treatment planning is technically feasible and leads to significant differences in dosimetry metrics compared with the conventional method.