Fully automated spleen segmentation predicts progression-free survival in HCC patients following transarterial radioembolization
- PMID: 41699276
- DOI: 10.1007/s00259-026-07792-8
Fully automated spleen segmentation predicts progression-free survival in HCC patients following transarterial radioembolization
Abstract
Purpose: Transarterial radioembolization (TARE) is a well-established treatment for unresectable hepatocellular carcinoma (HCC), though its effects on non-tumorous tissue remain a concern. In particular, the prognostic relevance of splenic volume changes after TARE is not fully understood. This study aimed to assess imaging-derived markers-specifically splenic volume dynamics-as predictors of disease progression.
Methods: We retrospectively analyzed laboratory and imaging data from 73 patients with histologically or imaging-confirmed HCC who underwent TARE with Yttrium-90 (90Y) at our institution between January 2012 and September 2022. Inclusion criteria were age ≥ 18 years, availability of baseline and 3-month follow-up imaging, and complete clinical documentation. Patients undergoing liver resection, transplantation, or additional therapies during follow-up were excluded.
Results: A relative increase in splenic volume at 3 months was the only independent predictor of progression-free survival (PFS), yielding a ROC-AUC of 0.86 (95%-CI: 0.76-0.95). An increase of 18% or more most accurately identified patients with early disease progression (< 12 months) with a sensitivity 0.74 and specificity 0.97, outperforming conventional clinical and laboratory parameters, including two-dimensional craniocaudal spleen measurements.
Conclusions: Automated splenic volumetry showed superior prognostic value over traditional markers in HCC patients treated with TARE. A post-treatment increase in spleen volume represents an additional, robust, and readily accessible imaging biomarker for early risk stratification and individualized treatment planning.
Keywords: AI-based volumetry; HCC; Imaging biomarker; Spleen volume; TARE.
© 2026. The Author(s).
Conflict of interest statement
Declarations. Ethical approval: For this retrospective analysis was obtained from the local institutional review board (reference 87/18S). Financial support and disclosures: ME reports fees from Blue Earth Diagnostics Ltd. (consultant, research funding), Novartis/AAA (consultant, speaker), Telix (consultant), Bayer (consultant, research funding), RayzeBio (consultant), Point Biopharma (consultant), Eckert-Ziegler (speaker) and Janssen Pharmaceuticals (consultant, speakers bureau), Parexel (image review) and Bioclinica (image review) outside the submitted work and a patent application for rhPSMA. Informed consent: Not applicable. Competing interests: ME reports fees from Blue Earth Diagnostics Ltd. (consultant, research funding), Novartis/AAA (consultant, speaker), Telix (consultant), Bayer (consultant, research funding), RayzeBio (consultant), Point Biopharma (consultant), Eckert-Ziegler (speaker) and Janssen Pharmaceuticals (consultant, speakers bureau), Parexel (image review) and Bioclinica (image review) outside the submitted work and a patent application for rhPSMA.
References
-
- Bray F, Laversanne M, Sung H, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2024;74:229–63. - PubMed
LinkOut - more resources
Full Text Sources
