2137 - High-Resolution CBCT-Guided VMAT Lattice Radiotherapy for Bulky Hepatocellular Carcinoma: Early Tumor Regression as a Trigger for Adaptive Radiotherapy to Preserve OAR Sparing

Purpose/Objective(s):

Bulky hepatocellular carcinoma (HCC) remains challenging to treat with radiotherapy. VMAT-based lattice radiotherapy (LRT) enables intratumoral high-dose vertices with a lower-dose peripheral component, potentially improving tumor control while preserving critical organs at risk (OARs). However, rapid early tumor regression may alter tumor-OAR geometry during treatment and compromise dosimetric integrity. We evaluated the feasibility, tumor regression profile, toxicity, and adaptive radiotherapy (ART) requirements of VMAT-based LRT for bulky HCC using high-resolution daily CBCT-guided on-treatment surveillance.

Materials/Methods:

We retrospectively reviewed 8 patients (9 treatment courses) with unresectable bulky HCC treated with VMAT-based LRT. All patients had received prior locoregional and/or systemic therapies (including HAIC, TAE, and atezolizumab/bevacizumab), and several continued systemic therapy during or after LRT. LRT plans incorporated high-dose vertices (18-20 Gy) delivered upfront, followed by peripheral hypofractionation (20-52.5 Gy in 5-15 fractions) while meeting institutional OAR constraints. Daily high-resolution CBCT was used for setup and on-treatment anatomic surveillance. ART was considered when CBCT suggested compromised plan integrity, including =10% target shrinkage/necrotic collapse affecting vertex geometry, =5 mm change in or loss of clearance at the tumor-OAR interface, or anticipated violation of OAR/mean liver constraints. Acute and late toxicities were graded using CTCAE, and radiographic response was assessed on follow-up imaging.

Results:

Median age was 72 years (range, 50–93), and initial target volume ranged from 324 to 2150 cm³ (median, 645 cm³). LRT was feasible across all courses, with a median mean liver dose of 13 Gy while preserving OAR constraints. Dose heterogeneity was maintained (median PVDR [D_peak/D_valley], 3.65; range, 3.08–11.90). At 3 months, most lesions showed marked regression with necrotic transformation (median absolute reduction, 333 cm³; median relative reduction, 62%). Acute toxicity was limited to Grade 0–2; late toxicity was Grade 0 in evaluable patients, and no RILD was observed at 3 months. Three patients required ART during treatment (two for rapid regression with >10% shrinkage within the first 1–3 fractions; one for loss of target–stomach clearance). Four patients continued immunotherapy after LRT. From RT completion, median OS was 6.3 months, while median DSS was not reached.

Conclusion:

VMAT-based LRT is a feasible and well-tolerated approach for bulky HCC, often resulting in substantial early tumor regression and necrotic transformation. Given the frequent and dynamic anatomic changes observed during treatment, daily high-resolution CBCT monitoring and predefined ART triggers are critical. This strategy supports safe treatment delivery by enabling timely plan adaptation, thereby maintaining target coverage while preserving liver and OAR constraints.