The most apparent differences between the three types of microspheres are the type of radiation they emit and their specific activity. 166Ho emits γ-radiation, which allows it to be visualized by SPECT/CT. It is also paramagnetic, so it can be visualized by MRI as well. On the contrary, imaging of 90Y is possible by using either bremsstrahlung-SPECT/CT or PET/CT.
The differences in specific activity are significant: typically low for 90Y resin and high for 90Y glass, with 166Ho having a specific activity in between. This has many consequences: first, the difference in specific activity is translated into the number of particles injected, leading to a larger embolic effect (and potential stasis) for 90Y resin and 166Ho than for 90Y glass. Also, the tolerability of the liver is different for the three types of microspheres, leading to a difference in safety thresholds. A higher number of microspheres leads to a higher number of targeted liver clusters (i.e. healthy liver parenchyma) and a more homogeneous distribution in the liver.  With 90Y glass microspheres, there generally is a more heterogeneous distribution of the microspheres, leading to a greater tolerability at high absorbed doses to the healthy liver. This explains why the thresholds for safety are different for the three types of microspheres. Likewise, a more homogeneous distribution, as obtained with 90Y resin and 166Ho, most likely also requires lower tumour absorbed doses to be equally effective. This is indeed reflected in the thresholds found in the literature. 
Dose thresholds depend on the microsphere type, but also on the tumour type. Hepatocellular carcinomas are generally hypervascular, and will receive a high tumour absorbed dose, whereas colorectal carcinoma metastases are generally more hypovascular, resulting in a lower tumour absorbed dose and higher absorbed dose in the healthy liver tissue. In any patient, TARE should therefore be performed by using a personalized treatment plan based on a sufficient tumour absorbed dose and safe healthy liver absorbed dose.
This is especially true when treating with a combination of systemic treatment and TARE. Based on multiple RCTs in colorectal carcinoma in the first line setting (SIRFLOX, FOXFIRE and FOXFIRE-Global studies), the combination of first line chemotherapy and TARE cannot be recommended at this time. However, even without using a personalized treatment plan (i.e. a one-size-fits-all approach was used, which likely lead to underdosing of most patients), a statistically significant difference in cumulative incidence of first progression in the liver was still found.  EPOCH confirmed this positive signal in the second line setting with a significant improvement not only in hepatic PFS, but also in overall PFS.  Similarly, no overall survival benefit could be established in the SORAMIC study when combining sorafenib with TARE in HCC.  But, with an individualized treatment plan in the DOSISPHERE-01 study, a very clear anti-tumour effect was found in HCC.  Most ongoing studies therefore include personalized treatment as a basis for state-of-the-art TARE.
State-of-the-art TARE, also including high-quality pre-, peri- and post-procedural imaging guidance, paves the way towards new treatment scenarios and new indications in and outside the liver.