Dual-modality imaging with CEUS and MRI for early recurrence assessment in hepatocellular carcinoma following combined radiofrequency ablation and transarterial chemoembolization

Abstract

Objective This study aimed to evaluate the diagnostic performance of combined contrast-enhanced ultrasound (CEUS) and magnetic resonance imaging (MRI) for detecting early recurrence within 12 months after radiofrequency ablation combined with transarterial chemoembolization (RFA-TACE) in patients with hepatocellular carcinoma (HCC). Methods This retrospective cohort study included 173 patients with HCC who underwent RFA-TACE between July 2021 and July 2023, including 143 men and 30 women, with a mean age of 57.35 ± 7.61 years. All patients underwent CEUS and 3.0-T MRI at 1 and 3 months after RFA-TACE, with additional examinations performed when recurrence was subsequently suspected. Two senior radiologists independently reviewed the images while blinded to the clinical outcomes. CEUS assessment included a whole-liver grayscale survey followed by targeted contrast-enhanced evaluation of the treated area and any additional suspicious intrahepatic lesions. MRI assessment included whole-liver evaluation of qualitative enhancement and diffusion characteristics together with quantitative apparent diffusion coefficient values and ratios. Early recurrence was assessed using CEUS-only interpretation, MRI-only interpretation, and an exploratory combined CEUS–MRI interpretation strategy. The combined strategy represented expert integrated judgment rather than a prespecified algorithmic model. Disagreements were resolved by a third senior radiologist. Recurrence status was determined using a composite reference standard based on histopathology when available or, otherwise, longitudinal clinical and imaging follow-up adjudicated independently of the index-test interpretations. Diagnostic performance was summarized using sensitivity, specificity, positive predictive value, negative predictive value, and accuracy. Results Early recurrence occurred in 109 of 173 patients (63.01%). In unadjusted comparisons, the early recurrence group had higher proportions of BCLC stage C disease (48.62% vs 18.75%; OR = 4.10, 95% CI: 1.97–8.52; P<0.001), Child-Pugh class B liver function (32.11% vs 12.50%; OR = 3.31, 95% CI: 1.43–7.69; P = 0.004), and pretreatment AFP levels >100 ng/mL (59.63% vs 42.19%; OR = 2.02, 95% CI: 1.08–3.79; P = 0.026). Tumor diameter >5 cm was not significantly associated with early recurrence (45.87% vs 31.25%; OR = 1.86, 95% CI: 0.97–3.57; P = 0.059). On CEUS, arterial-phase nodular or irregular enhancement, early contrast washout, peritumoral enhancement, and irregular margins were more frequent in the early recurrence group (all P ≤ 0.001). On MRI, DWI hyperintensity, arterial-phase nodular or irregular enhancement, delayed-phase washout, and irregular margins were also more frequent in the early recurrence group. The mean ADC value was lower in the early recurrence group than in the no early recurrence group (1.05 ± 0.18 vs 1.32 ± 0.15 ×10-3 mm²/s, P<0.001). CEUS alone achieved a sensitivity of 80.7%, specificity of 82.8%, and accuracy of 81.5%. The corresponding values were 88.1%, 87.5%, and 87.9% for MRI and 92.7%, 93.8%, and 93.1% for combined CEUS–MRI. Conclusion Exploratory combined CEUS–MRI interpretation showed higher descriptive sensitivity, specificity, and accuracy than either CEUS-only or MRI-only interpretation for detecting early recurrence after RFA-TACE. CEUS and MRI provided complementary information on vascular perfusion, tissue structure, and water-molecule diffusion. Because the combined strategy was based on expert integrated judgment rather than a prespecified algorithm, its findings should be interpreted cautiously and require prospective validation.