Caren van Roekel

63 Radioembolization Careful patient selection should be applied and doses should be calculated with taking into account the administered activity per target volume. 9.2. Advances in dosimetry The need for optimized dosimetry is more and more emphasized. A huge step forward from the BSA and the MIRD method was made with the invention of the partition method for activity calculation. This method takes into account the mass of the tumors and the liver. However, it does not correct for the difference in tumor and non-tumor expected absorbed dose. Another limitation is that it is difficult to delineate ill-defined tumors. As stated above, a voxel-based method for activity calculation is proposed by Chiesa et al (63). This method uses the 99m Tc-MAA quantification as a surrogate of microsphere distribution and takes non-uniformity of the absorbed dose into account. However, limitations of this voxel-based approach are physical factors such as attenuation, scatter, noise and partial-volume effects, that should be corrected for (64). Garin et al. have developed a personalized dosimetry approach, with an aimed absorbed tumordose of ≥205 Gy. They evaluated the 99m Tc-MAA volume of distribution in the injected liver and tumor and the total injected activity. The injected activities in the tumors and liver were calculated with the MIRD formula, with the end points of attaining a tumor-absorbed dose of ≥205 Gy, a healthy liver absorbed dose of <120 Gy and a lung absorbed dose of <30 Gy. Treatment intensification was performed when the estimated tumor dose would not reach 205 Gy as calculated with the MIRD method (118). In a study of 85 patients with HCC, this method was used for activity calculation. Response rate was 80.3% on lesion-based analysis and 77.5% on patient-based analysis. There was a clear dose-response relationship, with a response rate of 89.7% for tumors with an absorbed dose of ≥205 Gy versus a response rate of 9.1% in tumors with an absorbed dose <205 Gy. Tumor-absorbed dose was also significantly associated with overall survival (p=0.005 on multivariate analysis) (140). A personalized dosimetric approach is advocated for the treatment of not only HCC patients but also for patients with hepatic metastases. Optimal tumor- absorbed doses and maximum tolerable dose to the healthy liver have to be defined for multiple tumor types. 2