Abstract No. 558 Does lung shunt fraction estimation using injection of 99mTc-MAA still vital in 90Y radioembolization dosimetry? A hypothetical study using patient specific information

To compare and estimate the impact of actual vs. arbitrarily chosen lung shunt fraction (LSF) on 90Y radioembolization dosimetry Institutional review board approved this retrospective study of patients who underwent invasive catheter angiogram with injection of macroaggregated albumin (99mTc-MAA) to measure the actual LSF that was then used in 90Y dosimetry for glass microspheres based radioembolization for hepatocellular carcinoma in 2017 and 2018. 40 patients (mean age, 68 years) underwent 44 injections of 99mTc-MAA. Details related to the targeted liver volume, 99mTc-MAA and 90Y procedures were collected from medical records. Impact of arbitrarily chosen LSF on 90Y dosimetry was compared with the 90Y dosimetry based on actual LSF. There was on average 22 days (range, 10–44) delay between the 99mTc-MAA and 90Y procedures. The mean actual LSF was 4.9% (range, 0.8%–11.3%). Average mass of the treated liver was 0.791 kg (range, 0.3–2.36). The mean prescribed radiation dose was 153.7 Gy (range, 110– 320 Gy). The actual radiation dose deposited by glass microspheres was estimated to be 153.4 Gy (range, 100.6–318.4). Alternatively, if 90Y dosimetry is calculated using an arbitrarily chosen LSF of 5%, a minimum commercially available radioactivity (i.e., 3 GBq at the time of calibration) required for the smallest treatment mass (0.3 kg) in this study group would deposit a dose of 118 Gy at 4 pm on a Friday, and the maximum commercially available radioactivity (20 GBq) for the largest treatment mass (2.36 kg) would deposit a dose of 321 Gy at 8 am on a Monday. This study indicates the feasibility of using arbitrary LSF in 90Y dosimetry. When a range of commercially available vials of radioactivity is readily available at the treatment enter, glass microspheres based radioembolization using arbitrarily chosen LSF based 90Y dosimetry would be feasible without the need for 99mTc-MAA injection in most patients. However, an alternative method to predict and identify the small percentage of patients who could have high LSF is required to prevent significant pulmonary injury from radioembolization based on arbitrarily chosen low LSF.