This single-center prospective study was conducted with approval from the Seoul National University Hospital Institutional Review Board (IRB No. 1611-124-811). Written informed consent was obtained from all participants. Although funding was provided by RF Medical Co. (Seoul, Korea), the investigators retained full control over the study data and consent procedures.

Between September 2017 and January 2019, patients with recurrent HCC following TACE were prospectively enrolled, meeting the following inclusion criteria: 1) age 20-85 years, 2) ChildPugh class A liver function, and 3) radiologically confirmed local recurrence following conventional TACE. Recurrent HCC was defined as the appearance of an arterially enhancing tumor at the edge of the treated lesion, as assessed using contrast-enhanced follow-up study confirming adequate lipiodol uptake and the absence of viable tissue in the target tumor and surrounding ablation margin, according to the modified Response Evaluation Criteria in Solid Tumors criteria.5 Recurrence was primarily diagnosed using contrast-enhanced multiphasic computed tomography (CT) (n=73) or magnetic resonance imaging (MRI) (n=27). Exclusion criteria included: 1) more than three HCC nodules, 2) tumors with an abutment length of ≥5 mm to the main hepatic veins or first branches of the main portal vein, 3) HCC nodules measuring ≥3 cm, 4) platelet count ≤50,000/mm³ or international normalized ratio prolongation ≥50%, 5) extrahepatic metastasis or vascular invasion, 6) Child-Pugh class B or C liver function, 7) poor performance status defined as Eastern Cooperative Oncology Group-performance status scale 3 or higher, and 8) poor acoustic windows or inability to achieve percutaneous access (Fig. 1).

All procedures were performed under real-time ultrasound-CT/magnetic resonance (MR) fusion guidance (RS85 Prestige; Samsung Medison, Seoul, Korea; SIGNA^TM; GE Healthcare, Chicago, IL, USA; SOMATOM^TM Force, MAGNETOM^TM Vida; Siemens Healthcare, Forchheim, Germany; Brilliance^TM iCT Gantry, Ingenia^TM 7700 MR; Philips Healthcare, Amsterdam, Netherlands; Aquilion^TM ONE; Canon Medical Systems, Otawara, Japan) with conscious sedation. The procedures were conducted by a radiologist (JML) with 26 years of clinical experience in RFA. The procedure utilized two 17-gauge internally cooled-perfusion electrodes connected to a multichannel RF Generator (RF Medical Co.) operating in bipolar mode. In this mode, the electrical current flows between the electrode pair, maintaining a higher current density.13 Each TICP electrode featured two 0.02-mm side holes on the active tips, allowing the perfusion of 0.9% isotonic saline at a rate of approximately 1 mL/min for tissue perfusion and electrode cooling (Fig. 2). A peristaltic pump (VIVA Pump; STARmed, Goyang, Korea) maintained the active tip temperature between 20℃ and 25℃ by circulating chilled normal saline. The ablation protocol targeted complete ablation of the entire tumor, including the post-TACE area and locally recurrent lesion, while ensuring a 5-10 mm ablation margin.14 Technical parameters such as current, impedance, power output, and total delivered energy were continuously monitored throughout the procedure. For subcapsular tumors, artificial ascites was introduced in 80 patients (80.0%) using 5% dextrose solution to prevent organ damage. The solution was aspirated following the procedure.15

The primary endpoint was the 2-year cumulative incidence of LTP of the ablation index tumor, measured from the date of RFA to the first occurrence of LTP. Secondary endpoints included technical success rate, technique efficacy, RFA procedural characteristics, and progression-free survival (PFS) post-RFA. Data collection continued through July 31, 2020.

Technical success was defined as complete tumor treatment with full coverage by ablative margin ≥5 mm on immediate follow-up CT.16 One month after RFA, participants underwent contrast-enhanced CT (n=92) or MRI (n=8), serum alpha-fetoprotein measurements, and liver function tests. At this time point, technique efficacy was assessed, defined as complete macroscopic tumor ablation (Fig. 3).16 The ablation volume was quantified using 1-month follow-up imaging by measuring three orthogonal diameters (x, y, z) and calculating the volume using the formula: $V=\frac{4}{3}\pi \times x\times y\times z$.17 Follow-up imaging included contrast-enhanced multiphasic CT or MRI at 3, 6, and 12 months during the first year, followed by imaging every 3 months until study completion on July 31, 2020. LTP was defined as the appearance of enhancement at the ablation margin on contrast-enhanced imaging.10,18 PFS included all forms of disease progression, including LTP, intrahepatic remote recurrence (defined as the presence of HCC in a liver site not contiguous with the ablation zone), and extrahepatic metastasis (defined as the spread of HCC to locations outside the liver).10

Statistical analyses were conducted using SPSS version 27 (IBM, Armonk, NY, USA) and MedCalc version 20.0.23 (MedCalc Software, Ostend, Belgium). Continuous variables were analyzed using the Mann-Whitney U test, while categorical variables were analyzed using chi-square or Fisher’s exact tests. Technical success, technique efficacy, and LTP rates were analyzed using per-patient data. Time-to-event analyses were performed using the Kaplan-Meier method, and statistical significance was set at P<0.05.

The procedural characteristics of bipolar RFA using TICP electrodes are detailed in Table 2. Technical success was achieved in all treated tumors. Technique efficacy, assessed at one-month follow-up imaging, was achieved in 97/100 (97.0%) of patients. Three patients (3.0%) did not achieve technique efficacy at 1-month follow-up imaging due to viable tumor portions around the ablation sites; these patients were subsequently treated with repeat RFA. The mean ablation time was 8.5±3.7 minutes, with a total RF energy delivery of 8.44±5.00 kcal. Energy delivery per unit time averaged 0.99±0.34 kcal/min. The procedure resulted in a mean ablation volume of 42.1±27.3 cm³, with an ablation volume per unit time of 5.45±3.40 cm³/min.

Post-procedural complications occurred in six patients (6.0%). Five patients experienced minor complications, including mild fever and myalgia, while one patient developed a major complication (pneumothorax), which resolved completely with continuous oxygen supplementation. No procedure-related mortality was observed throughout the study period.

During a median follow-up period of 34.0 months (range, 3-41), LTP was documented in 18 patients (18.0%). All LTP cases manifested as single nodular recurrences abutting the ablation margin and were subsequently managed with either RFA (n=4) or TACE (n=14). Intrahepatic remote recurrence occurred in 60 patients (60.0%), while extrahepatic metastases developed in six patients (6.0%). The estimated LTP rates at 1 year and 2 years were 13.3% and 17.7%, respectively. The calculated PFS rates were 37.8% at 1 year and 27.7% at 2 years (Fig. 4).