Comment on: Impact of breast cancer characteristics on the development and time to liver metastases: population-based study

Correspondence to: Ida Skarping (email: ida.skarping@med.lu.se)

Department of Imaging

Physiology, and Nuclear Medicine

Skåne University Hospital

Region Skåne

Sweden

BJS Open, https://doi.org/10.1093/bjsopen/zrag036, published 12 May 2026

_____

We read with great interest the article by Sundén et al. addressing the time to breast cancer liver metastases1. This is an important and clinically relevant topic, particularly as advances in local treatment and down-staging through systemic therapies have increased the need to identify patients who may benefit from metastasis-directed treatment. If local treatment of oligometastatic disease is considered, early detection, accurate assessment of the metastatic burden to avoid understaging, and robust response evaluation are essential.

We would like to further emphasize the role of imaging, which was only briefly addressed.

A key challenge in metastasis-directed therapy is patient selection. Metastatic breast cancer is predominantly a systemic disease, and the concept of oligometastatic disease depends strongly on the imaging modality used. Disease classified as oligometastatic with liver involvement on CT may no longer remain so when assessed with more sensitive techniques such as MRI or PET using different radiopharmaceuticals, especially if bone metastases are present2, 3.

ctDNA, as discussed by the authors, may complement imaging, addressing an unmet need for earlier detection of distant metastases and potentially enabling metastasis-directed treatment.

Despite the central role of imaging in diagnosis and treatment selection, current Swedish and European guidelines provide limited practical guidance on modality choice in the assessment of suspected oligometastatic disease4, 5. Without such consensus, the definition of oligometastatic disease remains inherently unstable, challenging its use as a treatment-defining concept.

The study also highlights a broader challenge: the limited availability of detailed imaging and recurrence data in registries, including the Swedish National Breast Cancer Registry. Integration of imaging, molecular biomarkers, and clinical data will be essential for advancing precision oncology. Large language models may facilitate extraction and harmonization of imaging data from clinical reports, thereby complementing traditional registry-based research.

We appreciate the valuable contribution this study makes to an evolving field.