Caren van Roekel

215 The efficacy of coil-embolization in radioembolization Time interval To investigate the effects of the time interval between coil-embolization and administration of the microspheres on redistribution, the patients were dichotomized using a threshold of 24 hours. This threshold was chosen as almost half of the patients included in this analysis received treatment within the same day of coil-embolization. Segment activity ratios and tumor activity ratios were then compared. Patients receiving sequential lobar treatment were excluded from this subgroup analysis. Statistical analysis Descriptive statistics were used as proportions and medians with ranges. Ratios of activity concentrations were calculated between dependent and non-dependent segments (the segment ratio) and dependent and non- dependent tumors (the tumor ratio). Since there is no definition of successful redistribution, the success rates for a 10%, 20% and 30% difference in activity concentration between the dependent and non-dependent volumes were calculated, corresponding to dose ratios of 0.9, 0.8, and 0.7, respectively. Ratios in the time interval analysis were compared with an independent samples t-test. Inter-rater reliability was evaluated by means of a weighted Cohen’s kappa. All statistical analyses were performed with IBM SPSS Statistics for Windows, version 22 (IBM Corp., Armonk, N.Y., USA). RESULTS Within the studied timeframe, a total of 517 radioembolization procedures were performed at our institute, of which 37 patients were selected for this study (Figure 3). In most cases (n=36) microcoils were used as embolization agent, and in one case cyanoacryl glue (Histoacryl®, B. Braun Surgical S.A.) was used. The baseline characteristics are shown in Table 1. Cone-beam CT images were available in 27 of 37 cases. In 10 cases, cone-beam CT series were acquired after selective injection of contrast-agent in the artery that was to be coiled and were helpful in delineating the dependent volume. 8