Response to Letter to the Editor Regarding ‘Considerations on <scp>MRI</scp> ‐Based Stratification for Retreatment After <scp>TACE</scp> in <scp>HCC</scp> ’

We sincerely thank the author for the insightful and constructive comments [1] on our study proposing the MRI-based VITAL score for predicting retreatment response of viable hepatocellular carcinoma (HCC) after transarterial chemoembolization (TACE) [2]. We appreciate the recognition of the clinical relevance of our work and welcome the opportunity to address the points raised. First, we adopted visually assessed imaging features primarily to enhance clinical applicability and real-world implementation. Features such as tumour heterogeneity and peritumoral hyperenhancement are routinely evaluated in daily practice and can be readily incorporated into decision-making without additional technical requirements. Although these assessments involve some subjectivity, all features were evaluated by experienced abdominal radiologists using standardised criteria, with acceptable inter-observer agreement. We acknowledge that radiomics and AI-based quantitative imaging may better capture tumour biology; however, their clinical translation is currently limited by issues of standardisation, and workflow integration across centers [3]. Therefore, our approach prioritises accessibility, while future studies may explore integrating quantitative features to further enhance the biological interpretability of the VITAL framework. Second, we appreciate the insightful discussion regarding the biological significance of peritumoral hyperenhancement. We agree that this feature is closely associated with aggressive tumour behaviour, including microvascular invasion and hypoxia-driven angiogenesis via the HIF-1α/VEGF pathway [4]. Accordingly, a high VITAL score may reflect not only a higher likelihood of TACE refractoriness but also an underlying hypervascular and angiogenic tumour phenotype. Although our primary aim was to predict retreatment response rather than to guide systemic therapy selection, we acknowledge the potential clinical implications. Patients with imaging features suggestive of angiogenic dominance may benefit from earlier transition to VEGF-targeted or immunotherapy-based regimens. This hypothesis-generating observation warrants prospective validation and may inform future studies integrating imaging biomarkers with therapeutic decision-making. Third, we agree that incorporating biological markers could further improve the predictive performance of the model. However, the present study was designed to predict retreatment response primarily based on post-treatment MRI features of viable tumours after initial TACE, rather than longitudinal biological dynamics. Although baseline alpha-fetoprotein (AFP) levels were not significant predictors in our cohort, dynamic AFP kinetics have been shown to better reflect tumour viability and treatment response [5, 6]. Integrating longitudinal biomarker changes with MRI-based morphological features represents a promising direction for developing a more comprehensive assessment of residual disease burden, and this concept will be explored in future studies. In summary, we are grateful for the author's valuable comments, which not only support the clinical utility of the VITAL score but also highlight important directions for its refinement. In conclusion, we are grateful for the author's valuable comments, which not only support the clinical utility of the VITAL score but also highlight important directions for its refinement. We believe that continued efforts to balance clinical feasibility with biological depth will further advance imaging-based prediction models in the personalised management of HCC. The authors declare no conflicts of interest. The authors have nothing to report.