Caren van Roekel

27 Radioembolization 3.2. Coil embolization of culprit vessels Since extrahepatic deposition of radioactive microspheres may cause serious complications, it should be assessed whether arterial branches pose a risk. The 3 most common culprit vessels that cause extrahepatic deposition are the gastroduodenal artery, the cystic artery and the right gastric artery. Previously, it was advised to preventively coil-embolize these arteries. However, there are quite some disadvantages to coil-embolization: there is an increased radiation dose and an increase of procedure time and complexity, potential vessel damage and complications of coil deployment. Moreover, in the time between the preparatory angiography and treatment, new collateral vessels may develop. Nowadays, most centers try to avoid coil embolization unless extrahepatic deposition of activity is found on pretreatment simulation scout dose SPECT/CT. Sometimes, a more distal injection position or the use of an antireflux catheter can provide a safe treatment procedure without the need for coil embolization (2). 4. SCOUT DOSE IMAGING AND PRETREATMENT DOSIMETRY 4.1. 99m Tc-MAA Worldwide, pretreatment simulation of radioembolization with 90 Y is currently based on 99m Tc-MAA planar imaging and SPECT/CT for assessment of extrahepatic depositions and lung shunting. MAA are biodegradable particles. Their size is not well-calibrated, estimated at between 10 and 150 µm, with 90% of the particles falling between 10 and 40 µm, and 1 to 2% below 15 µm (46). Typically, between 150 and 250 MBq of 99m Tc-MAA are injected into the hepatic artery for therapy simulation, with around 1.5x10 6 MAA particles (47). As a comparison, normally around 3-5x10 6 spheres for 3 GBq of glass microspheres are injected, and around 40-80x10 6 spheres for 1.5 GBq of resin microspheres. The size, density and number of injected particles differs between 99m Tc-MAA particles and glass/resin microspheres. Thus far, 99m Tc-MAA particles are the only particles used for the pretreatment simulation of radioembolization with 90 Y, but there are controversies regarding the predictive value. Wondergem et al. have studied the relation between pre-treatment 99m Tc-MAA distribution and post-treatment 90 Y distribution. They found a difference of >10% in 68% of the 2