Caren van Roekel

40 Chapter 2 starts to decline, 10 mg/day defibrotide intravenously or oral steroids can be considered. If medical treatment is ineffective, a transjugular intrahepatic portosystemic stent-shunt (TIPS) could be placed. Based on an expert panel and literature review, prevention of REILD can be pursued by excluding patients with poor liver functional reserve (such as total bilirubin >2 mg/dL or ascites) from radioembolization. Furthermore, it is recommended to adapt the calculated activity in patients with steatosis, steatohepatitis, hepatitis, cirrhosis, a liver volume <1.5 L and with multiple lines of prior chemotherapy. Also, sequential lobar treatment may improve liver tolerance to radioembolization (79). Since a high absorbed dose is an important cause of hepatotoxicity, a personalized dosimetric approach, as proposed by Chiesa et al., should be implemented in clinical practice, especially in patients with known risk factors. The most vital aspect in the prevention of REILD should be the healthy-liver tissue absorbed dose. Therefore, the parenchymal mean dose should be taken into account in treatment planning (63). Another relatively uncommon complication of radioembolization is gastrointestinal ulceration, caused by non-target delivery of microspheres (75). Gastrointestinal ulceration usually presents 2-6 weeks after treatment with symptoms of acute epigastric pain, nausea, vomiting, dyspepsia and anorexia. The incidence is about 2-3%. Symptoms can last up to 10 months despite adequate treatment with proton pump inhibitors. Normally, full recovery occurs. Prevention of ulceration by using proton pump inhibitors around treatment is often advocated but lacks scientific evidence (79). Lam et al. have performed a root cause analysis to identify risk factors for the development of gastrointestinal ulceration. In their cohort of 278 treatments in 247 patients, the following risk factors were identified: stasis, proximal administration site, young age and distal origin of the GDA. To prevent gastrointestinal ulceration, it is advised to administer microspheres exclusively distally (80). Although much attention is paid to its prevention by calculating lung shunt, radiation pneumonitis is rarely seen. It is characterized by exertional dyspnea, dry cough, restrictive ventilator dysfunction and bilateral lung infiltrates. The lung shunt is estimated based on the fraction of 99m Tc-MAA that is deposited in the lung vasculature after the pretreatment work-up. However, Elschot et al. have shown that 99m Tc-MAA usually overestimates the lung shunt, with an absolute error range of 9.4-12.1 Gy (53). The manufacturer of resin microspheres recommends