CIO 2021-30 Accuracy and Safety of Scout Dose Y90 for Personalized Selective Internal Radiation Therapy Planning

Purpose: To evaluate accuracy and safety of scout dose yttrium-90 (Y90) resin microspheres vs. technetium-99m (Tc99m) macroaggregated albumin (MAA) in predicting the therapeutic dose Y90 biodistribution for personalized Selective Internal Radiation Therapy (SIRT) planning for hepatocellular carcinoma (HCC).

Material and Methods: In a prospective single-arm clinical trial (NCT04172714), patients with HCC undergoing Y90 SIRT were recruited. Each patient underwent mapping angiography and MAA SPECT/CT followed by same-day mapping using 15 mCi Y90 resin microspheres and Y90 PET/CT. During MAA mapping, cone beam CT (CBCT) was used to detect and treat potential non-target embolization pre-MAA administration. Partition model was used for treatment planning using MAA data with the goal of >200 Gy to tumor. Prescribed Y90 activity minus previously administered scout activity was administered 3 days post mappings. Same microcatheter type and administration position were used for all stages of the study (including CBCT) in each patient. Sureplan®(MIM Software, Cleveland, OH) was used for biodistribution analysis. Using paired t-test and Pearson correlation, tumor:normal ratio (TNR) and lung shunt fraction (LSF) calculated by MAA SPECT and scout dose Y90 PET were compared to that of therapeutic dose Y90 PET. Safety of scout dose Y90 was evaluated by presence of non-target radiotracer uptake seen on scout dose Y90 PET not seen on MAA SPECT.

Results: Overall, N=30 patients with 33 tumors were treated (mean tumor volume: 44.9 cc). Mean TNR was 2.67, 2.75 and 2.71 by MAA, scout dose Y90 and therapeutic dose Y90 respectively (p’s>0.05). There was a moderate linear correlation between MAA and therapeutic Y90 TNR’s (r=0.53) and a strong linear correlation between scout dose Y90 and therapeutic Y90 TNR’s (r=0.904;p’s <0.001). Mean LSF was 5.7%, 6.4% and 5.2% by MAA, scout dose and therapeutic dose Y90 respectively (p’s>0.05). There was a weak linear correlation between MAA and therapeutic Y90 LSF’s (r=0.39;p=0.031) a moderate linear correlation between scout dose and therapeutic Y90 LSF’s (r=0.562;p=0.001). Two patients required coiling of right gastric arteries before MAA administration. No non-target uptake was seen on MAA SPECT/CT and no non-target discrepancy was noted on scout dose Y90 PET.

Conclusions: Compared to MAA, Y90 scout dose is a more accurate surrogate for therapeutic Y90 biodistribution with similar safety profile when used in conjunction with CBCT during mapping to detect and treat potential non-target embolization.