S1095
Clinical – Upper GI
ESTRO 2026
Results: A total of 207 patients were included. The cohort predominantly consisted of males (80.2%) with a median age of 78 years (range 51–90) and an ECOG-PS 1 (101, 48.8%). Most patients had preserved liver function (Child–Pugh A, 88.1%), while 22 patients (10.6%) were classified as Child–Pugh B8–9. According to 2022 BCLC staging, 140 patients (67.6%) presented with advanced disease (stage C). SBRT was delivered to one lesion in 65.2% of cases, with a median dose of 50 Gy (range 30–75) and a median biologically effective dose (BED, α / β = 10) of 102.6 Gy (range 45–262). Overall, 305 lesions were treated.Median follow-up was 18.1 months (range 2.5–142). Median LC was not reached; 1- and 3-year LC rates were 90.2% and 77.5%, respectively. Median PFS was 20.6 months (95% CI 11.7–49.5) and median OS was 30.7 months (95% CI 23.9–36.5). The number of lesions and Child–Pugh class were independent predictors of PFS and OS.Acute toxicity was mild (grade 1–2), while grade 3 late toxicity occurred in a few patients, with isolated cases of ascites, chest wall pain, and hepatic failure. Conclusion: This real-world study confirms the safety and efficacy of SBRT as a treatment option for HCC across different disease stages. SBRT achieved durable local control and encouraging survival outcomes. Liver function and tumor burden emerged as key prognostic factors, supporting careful patient selection to minimize toxicity and optimize long-term outcomes. Keywords: Hepatocellular, SBRT, real-world isotolerable SBRT of pancreatic cancer in modern radiotherapy using conventional and adaptive radiotherapy Luca Nicosia, Michele Rigo, Chiara De-Colle, Andrea Gaetano Allegra, Niccolò Giaj-Levra, Carolina Orsatti, Edoardo Pastorello, Francesco Ricchetti, Andrea Romei, Ruggero Ruggieri, Filippo Alongi Department of Advanced Radiation Oncology, IRCCS Sacro Cuore Don Calabria Hospital Cancer Care Center, Negrar di Valpolicella, Italy Purpose/Objective: This study aimed to identify clinical and dosimetric predictors of tumor response and local progression- free survival (LPFS) in patients undergoing radiotherapy, comparing conventional (cRT) and adaptive (aRT) techniques. Material/Methods: Univariate and multivariable logistic regression analyses were performed to assess associations between response and factors including EQD2, age, Digital Poster 3184 A dose-response relationship model for
tumor volume, and treatment technique. LPFS was estimated using Kaplan–Meier analysis, and model performance was evaluated through Nagelkerke R ² and ROC curve analysis. Results: 95 patients were treated with pancreatic SBRT using cRT (67; 70.5%) or aRT (28; 29.5%). The median age was 66 years (range 36-89). Median EQD2 was 60Gy (range 37.5-83.3 Gy). Median LPFS was not reached. The 1- and 2-year LPFS was 69% and 56.7%. In the univariate analysis, EQD2 (OR 1.07, 95% CI 1.03-1.11, p=0.00), age (OR 0.95, 95% CI 0.91-0.99, p=0.01) and treatment technique (OR 0.14, 95% CI 0.05-0.4, p=0.00) significantly correlated with tumor response. In the multivariable logistic regression analysis, higher EQD2 was significantly associated with improved treatment response (OR 1.06, 95% CI 1.01–1.11, p = 0.02); every 1Gy increment in EQD2 significantly increase the response probability of 5.6%. The 1- and 2- year LPFS for patients treated with EQD2 > or ≤ 60 Gy was 55.3% versus 86.9 and 40.8% versus 78.2% (figure 2).Patient age was inversely correlated with response, with each additional year reducing the odds of response by approximately 5% (OR 0.95, 95% CI 0.91–0.99, p = 0.04). The treatment technique showed a strong effect, with cRT being associated with markedly lower response probability compared with aRT (OR 0.20, 95% CI 0.06–0.65, p = 0.00). In particular, aRT significantly improved LPFS over cRT, with 1- and 2-year LPFS of 64.1% versus 82% and 48% versus 82%.Tumor volume demonstrated a non-significant trend toward reduced response (OR 0.99, 95% CI 0.97–1.00, p = 0.14).The global performance of the multivariable model was satisfactory. In particular, the likelihood ratio test confirmed that the model significantly improbe prediction compared with the null model ( χ² = 31.6, df = 4, p < 0.001).The ROC curve analysis showed that the multivariable model hadgood discriminative performance, with an AUC of 0.829 (95% CI: 0.746– 0.912), confirming the model’sability to distinguish responders from non-responders. Conclusion: Higher radiation dose and aRT (over cRT) significantly improved tumor response and LPFS, while increasing age negatively impacted response. These results support dose intensification and adaptive strategies as key factors in enhancing local tumor control. Keywords: Pancreas carcinoma, adaptive RT, MR-linac
Made with FlippingBook - Share PDF online