PET/CT-Based Dosimetry in (90)Y-Microsphere Selective Internal Radiation Therapy: Single Cohort Comparison With Pretreatment Planning on (99m)Tc-MAA Imaging and Correlation With Treatment Efficacy

(90)Y PET/CT can be acquired after (90)Y-microsphere selective radiation internal therapy (SIRT) to describe radioactivity distribution. We performed dosimetry using (90)Y-microsphere PET/CT data to evaluate treatment efficacy and appropriateness of activity planning from (99m)Tc-MAA scan and SPECT/CT. Twenty-three patients with liver malignancy were included in the study. (99m)Tc-MAA was injected during planning angiography and whole body (99m)Tc-MAA scan and liver SPECT/CT were acquired. After SIRT using (90)Y-resin microsphere, (90)Y-microsphere PET/CT was acquired. A partition model (PM) using 4 compartments (tumor, intarget normal liver, out-target normal liver, and lung) was adopted, and absorbed dose to each compartment was calculated based on measurements from (99m)Tc-MAA SPECT/CT and (90)Y-microsphere PET/CT, respectively, to be compared with each other. Progression-free survival (PFS) was evaluated in terms of tumor absorbed doses calculated by (99m)Tc-MAA SPECT/CT and (90)Y-microsphere PET/CT results. Lung shunt fraction was overestimated on (99m)Tc-MAA scan compared with (90)Y-microsphere PET/CT (0.060 ± 0.037 vs. 0.018 ± 0.026, P < 0.01). Tumor absorbed dose exhibited a close correlation between the results from (99m)Tc-MAA SPECT/CT and (90)Y-microsphere PET/CT (r = 0.64, P < 0.01), although the result from (99m)Tc-MAA SPECT/CT was significantly lower than that from (90)Y-microsphere PET/CT (135.4 ± 64.2 Gy vs. 185.0 ± 87.8 Gy, P < 0.01). Absorbed dose to in-target normal liver was overestimated on (99m)Tc-MAA SPECT/CT compared with PET/CT (62.6 ± 38.2 Gy vs. 45.2 ± 32.0 Gy, P = 0.02). Absorbed dose to out-target normal liver did not differ between (99m)Tc-MAA SPECT/CT and (90)Y-microsphere PET/CT (P = 0.49). Patients with tumor absorbed dose >200 Gy on (90)Y-microsphere PET/CT had longer PFS than those with tumor absorbed dose ≤200 Gy (286 ± 56 days vs. 92 ± 20 days, P = 0.046). Tumor absorbed dose calculated by (99m)Tc-MAA SPECT/CT was not a significant predictor for PFS. Activity planning based on (99m)Tc-MAA scan and SPECT/CT can be effectively used as a conservative method. Post-SIRT dosimetry based on (90)Y-microsphere PET/CT is an effective method to predict treatment efficacy.