Caren van Roekel

191 Dose-effect relationships of holmium-166 radioembolization in colorectal cancer Toxicity The median parenchymal-absorbed dose was 37 Gy (range 12-55 Gy). Toxicity incidence during three months post-treatment and CTCAE grades are summarized in Table 2. New grade ≥3 clinical toxicity was present in eight patients (20%) and new grade ≥3 laboratory toxicity was present in seven patients (17.5%). There was one patient (2.5%) who developed REILD, evidenced by hyperbilirubinemia, hypoalbuminemia and ascites, without evidence of progression or biliary obstruction. The mean parenchymal-absorbed dose of this patient was 34 Gy. The results of the linear clinical toxicity regression analyses suggested a positive association between higher parenchymal dose and increase in CTCAE grade clinical toxicity (Table S1). The mean difference in parenchymal dose for patients with CTCAE grade 0/1/2 any clinical toxicity versus CTCAE grade 3/4/5 was 11.6 Gy (95%CI 3.4-19.7, p=0.0070). The odds ratio for CTCAE grade 3/4/5 any clinical toxicity versus CTCAE grade 0/1/2 per 10 Gy increase in parenchymal dose was 7.62 (95%CI 1.95-249.03, p=0.0063) (Table S2). For laboratory toxicity, the results of the linear regression analyses for both the CTCAE grades and the relative change in laboratory parameters showed that a higher parenchymal-absorbed dose is related with an increase in laboratory toxicity (Table S3a-b and Figures 2a-f). 7