Holmium-166 radioembolisation dosimetry in HCC.

Purpose: To evaluate dosimetry, dose-response and dose-toxicity relationships for holmium-166 (Ho) radioembolisation in patients with hepatocellular carcinoma (HCC).

Methods: Thirty-one patients with hepatocellular carcinoma were included in the HEPAR Primary study (NCT03379844, registered on December 20th, 2017) and underwent Ho-microspheres radioembolisation. Linear mixed models assessed the association between tumour absorbed doses and response based on mRECIST both on tumour and patient level. Preliminary tumour absorbed dose thresholds were estimated based on predictive value. Linear regression models assessed the association between non-tumour absorbed dose and Common Terminology Criteria for Adverse Events version 4.03.

Results: Median tumour absorbed dose (tumour level) was 95.5 Gy (range 44-332 Gy). Median non-tumour absorbed dose based on whole liver volume was 19 Gy (range 3 - 48 Gy) and based on target liver volume was 30 Gy (range 13 - 54 Gy). There was a significant association between non-tumour absorbed dose and toxicity. Tumours with partial response/complete response (PR/CR, responders) received a 41% higher absorbed dose than tumours with progressive disease/stable disease (PD/SD, non-responders) (95%CI: 2%-93%, p = 0.04). A predictive value of 90% for tumour response was observed at a tumour absorbed dose threshold of 155 Gy, 100% predictive value was achieved at 184.5 Gy.

Conclusion: This study confirms a positive relationship between tumour absorbed dose and response and between non-tumour absorbed dose and toxicity. Dose thresholds found in this study can serve as a basis for personalized dosimetry in HCC patients treated with Ho-microspheres.