Oral Abstracts
13th Annual Global Embolization Symposium & Technologies
Julie Birch, MD
Fellow Physician
Massachusetts General Hospital
n/a
Purpose : Injection location of Tc99m macroaggregated albumin (Tc99m-MAA) to calculate lung shunt fraction (LSF) prior to radioembolization is variable in recommendation and practice, including whether repeat LSF is necessary for a subsequent contralateral lobe treatment. We compare LSFs from different distributions in the same patient, to evaluate whether repeat mapping and LSF is necessary prior to repeat radioembolization.
Material and Methods : Retrospective chart review of all patients who underwent radioembolization between February 2008 and December 2017 with LSF calculated from Tc99m-MAA injection was performed. LSF was calculated according to standard institutional method via either planar or SPECT imaging. Statistical analysis was performed using Wilcoxon Signed-Rank Test, comparing repeat LSF values from different distributions within the same patient.
Results : Of 289 patients who underwent radioembolization with LSF calculated from Tc99m-MAA injection within the study period, 71 patients underwent repeat mapping and LSF calculation, accounting for 142 mapping procedures. Mean patient age was 60.6 (range 30-89) with 53% male. Tumor types were metastatic disease (65, 87%; most commonly colorectal 29, 40%; melanoma 13, 18%; neuroendocrine 10, 14%) and hepatocellular carcinoma (HCC) (7 patients, 10%). Tc99m-MAA injection locations were: 25 proper hepatic artery, 68 right hepatic artery, 49 left hepatic artery. There was an average LSF of 7.6% (range 0 – 43, +/- 4.9%). When comparing LSF from different distributions in the same patient across all tumor types, Wilcoxon Signed-Rank test with a two- tailed hypothesis demonstrated no statistical significance with a Z-value of -1.6607 and a p-value of 0.0969. Analyzing HCC and metastatic disease separately, there was no statistically significant difference for either group.
Conclusions : No statistical significance difference in LSF was identified by injecting different arterial distributions in the same patient. This supports avoiding repeat LSF calculations and possibly mappings prior to repeat radioembolization in the same patient.