4:03 PM Abstract No. 29 Streamlining Y90 in early stage hepatocellular carcinoma by eliminating the lung shunt study: analysis of a 403-patient cohort

Tc-99m macro-aggregated albumin (MAA) imaging to estimate hepato-pulmonary shunting (LSF) is required prior to radioembolization (Y90) to mitigate risks of radiation pneumonitis. We hypothesize that patients with hepatocellular carcinoma (HCC) within Milan criteria for liver transplant have negligible LSF. This would potentially allow eliminating MAA studies when treating this cohort of HCC patients with Y90. With IRB approval, we retrospectively studied HCC patients who underwent Y90 from 2004-2017. Inclusion criteria were early stage HCC within Milan criteria (solitary ≤5 cm, 3 nodules ≤3 cm). LSF was determined using pretreatment MAA scan in all patients. Tukey test was used to detect outlier LSF values. Associations between LSF and baseline characteristics were investigated using 2 x 2 odds ratio test. Statistical significance set at P <0.05. 403 patients, 271 (67%) males with mean age 65.6 years, met inclusion criteria. 320 (79%) had one tumor and 83 (21%) had multifocal disease. 43 (11%) had T1 disease and 360 (89%) had T2 disease. Mean tumor size was 2.7 cm (range, 1-5). 364 (90%) had unilobar disease while 39 (10%) had bilobar HCC. Median LSF was 3.9% (IQR: 2.4%-6.2%) where 377 (93.5%) patients had LSF ≤ 10%. Those with LSF ≥ 12.3% were identified as outlier values. Only the presence of TIPS was associated with high LSF >10% (OR = 10.79, CI: 4.47-26.03, P <0.0001). Median LSF was 3.9 (IQR: 2.4-5.9%) versus 6.3 (IQR: 3.7-14.2%) in no-TIPS versus TIPS patients (P = 0.0001). There were no significant differences in LSF among other variables including etiology, size, CP, tumor size, ascites and AFP. Median segmental volume was 200 mL (IQR: 139.5-340.5). Median administered activity was 0.96 GBq (range, 0.18-4.63); median lung dose was 1.9 Gy (IQR: 1.04-3.4). LSF is negligible in early stage HCC patients within Milan criteria and no-prior TIPs placement. Further, the volume of perfusion is typically low, allowing segmental treatments. These patients appear to be appropriate candidates for a No-MAA paradigm provided they receive selective Y90.