ESTRO 2024 - Abstract Book

S2278

Clinical - Upper GI

ESTRO 2024

response was assessed through the rate of tumor necrosis relative to the total tumor volume. Previously unknown HCC lesions, ypT and tumor grading (G) were also reviewed. Intrahepatic progression of disease (I-PD) before LT and adverse effects, including radiation-induced liver disease (RILD) were evaluated. Acute and late toxicities were scored using the CTCAE 5.0 grading. Disease-free survival (DFS) and Overall survival (OS) were estimated from the day of LT using the Kaplan-Meier method. Cox regression model was used to assess associations between covariates and these oncological outcomes. Radiographic and pathologic response concordance rates were analysed using the κ statistics.

Results:

Between September 2012 and December 2022, 82 patients (117 lesions) with liver- confined HCC, ECOG PS ≤2 and Child- Pugh class preferably ≤ B7 were treated with SABR, followed by LT. The median age was 58 years (IQR:54 -65). 70 patients (85.4%) received SABR as a bridge to LT, while the others were treated with downstaging purposes being potential LT candidates. SABR was mostly delivered on a single node (69.5%) and it was the first-line local treatment for 50% of patients. Approximately 66% of treated lesions was peripheral. The most frequent sites of treatment were the V (22%) and VII (22%) liver segments. Up to 3 sites affected by HCC lesions were treated simultaneously (some small lesions were included in a single PTVs), with a median PTV of 58.46 cc. The most frequent fractionation schedules were 48 Gy/3 fractions (50%) and 40 Gy/5 fractions (39%), at 80% isodose. Median time from completion of SABR to LT was 146 days. At the restaging imaging prior to LT observed response rates were: 49.2% complete response, 16.4% partial response and 31.1% stable disease. Only 2 in-field PDs (with a LC of 96.7%) and 15 extra field PDs were detected. Pathological complete response, significant and minimal partial response rates were 36.6%, 25.6% and 23.2%, respectively ( Figure 1 ). Histopathological examination of liver explants revealed tumors which had not been identified in restaging imaging tests in 19 patients (23.2%). Most patients had ypT2 (42.7%) and G2 disease (41.5%). A weighted κ statistic of 0.38 revealed a minimal level of agreement between radiographic and pathological response. Acute toxicities after SABR included ascites (G2: 3; G3: 2), nausea (G2: 2), chest wall pain (G2:1) and rib fracture (G2:1). 10% of patients developed classic RILD and 8.5% non-classic RILD. With median follow up of 23 months (IQR:12-36 months) from LT, the median OS was not reached (1, 2-year OS: 88% and 86%). Following LT, 8 patients (9.8%) suffered extrahepatic PD, 4 of whom together with intrahepatic PD. The 1 and 2-year DFS rates were 92% and 90%. In Cox multivariable analysis only ypT≥3 was predictive of worse OS (HR:10.6; IC 95%: 1.85-60.73; p=0.008), while BCLC stage at time of SABR (HR: 3.87; IC 95%: 1.40- 10.70; p=0.009) and tumor grading ≥3 (HR: 7.96; IC 95%: 1.48-42.74; p=0.016) had a negative impact on DFS ( Figure 2 ). Crude 30 and 90-day mortality rates were 2.4% and 3.6% respectively.

Figure 1. A) Overall survival (OS) of the entire cohort; B) OS stratified for ypT; C) Disease Free Survival (DFS); D) DFS stratified for grading (explant pathology).