166 Ho-poly(L-lactic acid) microspheres allow for quantitative imaging with MR imaging or SPECT for microsphere biodistribution assessment after radioembolization. The purpose of this study was to evaluate SPECT-and MR imaging-based dosimetry in the first patients treated with 166 Ho radioembolization. Methods: Fifteen patients with unresectable, chemorefractory liver metastases of any origin were enrolled in this phase 1 study and were treated with 166 Ho radioembolization according to a dose escalation protocol . The contours of all liver segments and all discernible tumors were manually delineated on T2-weighted posttreatment MR images and registered to the posttreatment SPECT images (n 5 9) or SPECT/CT images (n 5 6) and MR imagingbased R 2 * maps (n 5 14). Dosimetry was based on SPECT (n 5 15) and MR imaging (n 5 9) for all volumes of interest, tumor-tonontumor (T/N) activity concentration ratios were calculated, and correlation and agreement of MR imaging-and SPECT-based measurements were evaluated. Results: The median overall T/N ratio was 1.4 based on SPECT (range, 0.9-2.8) and 1.4 based on MR imaging (range, 1.1-3.1). In 6 of 15 patients (40%), all tumors had received an activity concentration equal to or higher than the normal liver (T/N ratio $ 1). Analysis of SPECT and MR imaging measurements for dose to liver segments yielded a high correlation (R 2 5 0.91) and a moderate agreement (mean bias, 3.7 Gy; 95% limits of agreement, 211.2 to 18.7). Conclusion: With the use of 166 Ho-microspheres, in vivo dosimetry is feasible on the basis of both SPECT and MR imaging, which enables personalized treatment by selective targeting of inadequately treated tumors. Radi oembolization is an interventional oncologic treatment during which radioactive microspheres are administered in the arterial vessels supplying the liver and its tumors. The rationale behind this intraarterial liver treatment is that liver tumors are predominantly supplied by arterial blood, in contrast to the nontumorous liver, which relies mainly on the portal vein for its blood supply. Injection of a substance into the hepatic artery will therefore selectively target the tumorous tissue (1). Currently, the commercially available microspheres that are used for radioembolization are labeled with 90 Y. To be able to quantitatively evaluate the optimal and selective distribution of microspheres to the liver tumors, posttreatment imaging is indispensable. For that reason, optimization of posttreatment imaging of 90 Y-microspheres with bremsstrahlung SPECT and PET has recently gained interest (2-5).166 Ho-poly(L-lactic acid) microspheres have been developed at our institute as an alternative to 90 Y-microspheres specifically to be able to visualize the in vivo biodistribution of microspheres after radioembolization. 166 Ho-microspheres can be imaged with both SPECT and MR imaging, using the emission of g-photon radiation and the paramagnetic properties of holmium, respectively (6-10). Exploiting these qualities, multimodal dosimetry becomes feasible,...