CIO 2021-34 Glass vs. Resin-based Y90 Radiation Segmentectomy of HCC: A Comparative Effectiveness, Safety and Dosimetry Study

Purpose: To evaluate the efficacy, safety, and dosimetry differences in patients with hepatocellular carcinoma (HCC) undergoing Yttrium-90 (Y90) radiation segmentectomy (RS) with glass-based (GB) and resin-based (RB) microspheres.

Material and Methods: In a single institution retrospective study, consecutive patients with HCC treated with GB and RB Y90-RS between 2019 and 2020 were included. Tumor response was assessed using modified Response Evaluation Criteria in Solid Tumors and treatment related toxicity was assessed using Common Terminology Criteria for Adverse Events (AE) v5.0; both at 3 months post-Y90-RS. Using Y90 Bremsstrahlung SPECT/CT, voxel-based dosimetry analysis was performed with MIM Sureplan® (MIM Software Inc, Cleveland, OH). Dose volume histograms of the tumor and non-tumoral tissue were determined. Receiver operator characteristic curves of the tumor dose (TD) and non-tumor dose (NTD) were used to evaluate the dose threshold able to predict, with 70% specificity, an objective (partial or complete) tumor response, complete response, incidence of any AE, and incidence of any AE with a grade ≥3. Differences in clinical and laboratory variables were assessed with student’s t-test and Chi-Square.

Results: Overall, 81 patients underwent Y90-RS (20 RB; 61 GB). Total cohort was 78% male with mean age 66.8 years. There were no statistically significant differences in baseline characteristics (eg. Child-Pugh A, ECOG 0, MELD >10, ALBI Grade 1, lung-shunt fraction, tumor volume), or incidence of post-treatment any grade AE (41% GB vs. 26% RB; p=0.3) or grade ≥3 AE (13% GB vs. 6% RB; p=0.4) between the two treatment cohorts. Overall objective response was 83% (84% GB vs 80% RB; p=0.7) and complete response rate was 58% (57% GB vs 60% RB; p=0.8). Mean TD significantly differed between the two devices (809 Gy in GB vs 241 Gy in RB; p<0.0001). Mean TD threshold to predict objective response and complete response was 176 Gy and 247 Gy for RB and 230 Gy and 481 Gy for GB radioembolization. A maximum NTD of 999 Gy predicted any grade AE in GB Y90-RS. No NTD threshold for RB Y90-RS was identified.

Conclusions: In patients with HCC undergoing Y90-RS, both resin and glass microsphere radioembolization offer similar efficacy and safety profiles. Nevertheless, there exists significant differences in Y90 dose thresholds predicting tumor response and treatment related toxicity between the two devices.