Liver metastases beyond classical imaging findings.
Journal:
Abdominal radiology (New York)
Published Date:
Aug 10, 2026
Abstract
Liver metastases represent the most common form of secondary hepatic malignancy and a major determinant of outcomes in oncologic patients. Imaging plays a fundamental role in detection, characterization, therapy planning, and treatment response assessment of liver metastases. Classical imaging techniques can determine the presence, number, location, and resectability of liver metastases. Newer approaches can also assess biological characteristics of metastases important for guiding treatment strategies and predicting outcomes. Advanced imaging techniques, radiomics, and artificial intelligence (AI) can enhance the clinical capabilities of imaging by assessing tumor biological heterogeneity, thus improving predictions of therapeutic responses and patient outcomes. Nevertheless, implementing these advanced imaging approaches in clinical practice remains challenging. This review aims to present a comprehensive overview of the state of the art in imaging evaluation of liver metastases, with a particular focus on recent innovations and emerging developments. It reviews imaging features suggestive of specific primary malignancies, as well as atypical presentations of liver metastases. Additionally, it discusses imaging's potential to assess key histopathologic, molecular, and genetic characteristics of liver metastases and highlights its role in lesion characterization, prognostic stratification, therapeutic decision-making, and treatment planning. Finally, it addresses the challenges facing the use of advanced imaging techniques in the assessment of liver metastases and outlines future directions for research and clinical practice. Readers should take away an understanding of the evolving role of radiology in the management of liver metastases resulting from the paradigm shift from predominantly morphologic to biologically informed imaging.
Authors
Keywords
No keywords available for this article.