Association of post-treatment ⁹⁰Y PET/CT–derived mean tumor absorbed dose with treated-lesion progression-free survival after glass-microsphere radioembolization for hepatocellular carcinoma

Abstract

To assess the prognostic value of post-treatment yttrium-90 PET/CT–based dosimetry and an absorbed dose threshold for progression-free survival in large (> 8 cm) hepatocellular carcinomas treated with transarterial radioembolization. Fifty-eight patients with unresectable hepatocellular carcinoma who underwent transarterial radioembolization were retrospectively stratified by maximum tumor diameter (≤ 8 cm vs. >8 cm). Tumor absorbed doses were quantified using pre-treatment technetium-99 m macroaggregated albumin single-photon emission computed tomography/computed tomography and post-treatment yttrium-90 PET/CT. Progression-free survival was analyzed using Kaplan–Meier and Cox proportional hazards methods. Receiver operating characteristic analysis explored a post-treatment dose threshold in large tumors. Patients with tumors ≤ 8 cm had longer progression-free survival than those with tumors > 8 cm ( p = 0.030). Post-treatment dosimetry revealed higher tumor absorbed doses in tumors ≤ 8 cm than in those > 8 cm (median, 1386.1 vs. 515.9 Gy; p < 0.001). In tumors > 8 cm, a delivered dose > 468 Gy was associated with improved progression-free survival ( p = 0.014). Pre-treatment dosimetry correlated with post-treatment doses but was not associated with prognosis. Post-treatment dosimetry suggested that a tumor absorbed dose of approximately 468 Gy may be an exploratory marker of adequate dose delivery in tumors > 8 cm, though this threshold should be interpreted alongside tumor biology, particularly alpha-fetoprotein level, and requires validation. Pre-treatment dosimetry did not reliably predict outcomes, indicating undertreatment risk when relying solely on pre-treatment imaging. Post-treatment imaging may help confirm adequate dose delivery in large hepatocellular carcinomas.

The authors' abstract, as published at the source. Cancer Imaging, 2026 · DOI ↗