Performance of the Li-Rads Treatment Response Algorithm V2024: A Systematic Review and Meta-Analysis of Response Assessment Accuracy and Non-Definitive Categories Publisher Pubmed



Zare M ; Mohebbi A ; Abdi A ; Mohammadi A
Source: Abdominal Radiology Published:2026

Abstract

Background: The LI-RADS Treatment Response Algorithm (LR-TR) was comprehensively updated in 2024, but its clinical performance has not been systematically validated. We performed the first systematic review and meta-analysis to synthesize the prevalence of non-definitive categories (Equivocal/Nonprogressing) and the classification performance of LR-TR v2024. Methods: Following PRISMA guidelines, we systematically searched four databases for studies applying LR-TR v2024. Methodological quality was assessed using QUADAS-2. Data for prevalence and response assessment accuracy (sensitivity, specificity) were pooled using random-effects models. Evidence certainty was evaluated using GRADE. Results: We included 14 studies (1706 patients, 2036 lesions). For non-radiation therapies, the pooled prevalence of LR-TR Equivocal was 9% (CI: 6–11%), withsubstantial heterogeneity (I2=64.75%). For radiation-based therapies, the pooled prevalence of LR-TR Nonprogressing was 43% (CI: 27–60%), withsubstantial heterogeneity (I2=94.75%). The pooled response assessment performance yielded a sensitivity of 72% (CI: 54–84%) and specificity of 95% (CI: 78–99%), and accuracy of 83% (CI: 78–88%). Risk of bias was high, primarily in Patient Selection and Reference Standard. Conclusion: LR-TR v2024 demonstrates robust response assessment performance for detection of residual Viable tumors. It significantly improves response assessment for non-radiation therapies by reducing Equivocal assessments.Conversely, for radiation-based therapies, while Nonprogressing lesions (43%) represents non-definitive classification that appropriately directs continued surveillance rather than retreat, based on its high prevalence, further prospective research employing standardized imaging protocols and longitudinal assessment at defined intervals is essential to enhance the proportion of lesions achieving definitive categorization at early post-treatment assessments and ultimately optimize clinical outcomes. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2026.