Radioembolization of colorectal liver metastases: comparative dose–response of resin and glass microspheres

PURPOSE: This study aimed to investigate the association between the mean absorbed tumor dose (ATD) and treatment response in patients who underwent transarterial radioembolization for colorectal cancer liver metastases (CRCLM). Furthermore, we sought to determine the prognostic value of mean ATD in predicting overall survival. METHODS: This retrospective, single-center observational study was conducted in a tertiary referral hospital. A total of 56 patients who received transarterial radioembolization for CRCLM were included. Posttherapy Y-90 PET/computed tomography images were retrospectively analyzed for dosimetric evaluation. Mean absorbed doses were calculated for the whole normal liver, perfused liver, and all tumors. Treatment response was assessed at 3 months after therapy using liver MRI according to the modified RECIST 1.0 criteria. RESULTS: Fifty-six patients (36 males, 20 females; median age, 64 years) were included in the analysis. Radioembolization was performed with glass microspheres in 30 patients (54%) and with resin microspheres in 26 patients (46%). The two groups were comparable in terms of age, sex, and prior therapies. In total, 234 lesions (83 treated with resin and 151 with glass microspheres) were included in the lesion-based dose-response analysis. Overall, complete response (CR), partial response (PR), stable disease, and progressive disease (PD) were observed in 73 (31%), 119 (51%), 22 (9%), and 20 (9%) lesions, respectively. In the glass microsphere group (151 lesions), CR, PR, stable disease, and PD were observed in 63 (42%), 62 (41%), 12 (8%), and 14 (9%) lesions, respectively. Receiver operating characteristic curve analysis demonstrated a significant correlation between mean ATD and treatment response (PR + CR vs. stable disease + PD), with an AUC of 0.805 [95% confidence interval (CI): 0.714-0.896, P = 0.001]. In the resin microsphere group, CR, PR, stable disease, and PD were observed in 10 (12%), 57 (69%), 10 (12%), and 6 (7%) lesions, respectively. A similar correlation between mean ATD and response was also observed in this group (AUC = 0.767, 95% CI: 0.611-0.923, P = 0.001). During the follow-up period, 43 patients (78%) died. The mean overall survival for the entire cohort was 15.7 ± 1.8 months (95% CI: 12.0-19.5 months). No significant difference in overall survival was found between patients treated with glass and resin microspheres (16.3 ± 2.4 vs. 14.2 ± 2.5 months, P = 0.7). CONCLUSION: There is a strong correlation between ATD and tumor response in CRCLM patients treated with both resin and glass microspheres. Escalation of ATD through multicompartment dosimetry may enhance treatment efficacy without increasing the risk of hepatotoxicity.