Radiofrequency Ablation under Computed Tomography Guidance with Simultaneous Transarterial Chemoembolization in Patients with Early-stage Hepatocellular Carcinomas.

Introduction
Of the many treatment options for HCC, RFA is used for early-stage HCC [1]. RFA is generally performed under ultrasound (US) guidance, except for lesions located in the liver dome, which is a blind area for US [2]. In such cases, RFA is performed under CT guidance. Though US can visualize blood vessels, it has blind spots due to artifacts caused by air in the lungs and internal organs, and some lesions are difficult to detect depending on their nature. In contrast, HCC after conventional transarterial chemoembolization (c-TACE) is often easily recognizable due to Lipiodol accumulation, and CT guidance has no blind spots like US. However, it is generally difficult to visualize intrahepatic blood vessels under non-contrast-enhanced CT guidance, which is commonly used. Therefore, CT-guided puncture carries a higher risk of inadvertently puncturing blood vessels compared with US-guided puncture. In contrast, c-TACE, which is also performed to treat HCC, can be used in combination with RFA to both improve local control of RFA and to stop bleeding during the RFA procedure. The aim of this study was to evaluate the efficacy and safety of RFA under CT guidance in combination with TACE.

Material and Methods

Patients
We evaluated 186 procedures in which RFA was performed under CT guidance combined with simultaneous TACE in 142 patients with HCC between September 2016 and December 2021. The eligibility criteria were as follows: 1) patients with HCC diagnosed radiologically or pathologically; 2) HCC meeting the Milan criteria; 3) lesions difficult to detect with US; 4) Child-Pugh score A or B; 5) blood platelet count ≥50,000/μL; and 6) Prothrombin time-International Normalized Ratio (PT-INR) ≤1.35.

Procedures
RFA was performed under CT guidance combined with simultaneous TACE using an Interventional Radiology (IVR)-CT/angiography system (Aquilion ONE plus Infinix-I; Canon Medical Systems Corp., Tochigi, Japan), using the following protocols.

TACE
TACE, as c-TACE, was performed first [3, 4]. After inserting a 3-Fr catheter (CX Catheter, Gadelius Medical K.K., Tokyo, Japan) into the hepatic artery via the femoral approach, a microcatheter (Carnelian Marvel 2.0 Fr. selective, Tokai Medical Products, Inc., Aichi, Japan) was advanced into the feeding artery of the HCC nodule. A chemotherapeutic agent with an oil-based contrast agent (Lipiodol; Guerbet Japan, Tokyo, Japan) was injected, followed by gelatin sponge particles (Figure 1). Epirubicin (EPIR; Nippon Kayaku Co., Ltd., Tokyo, Japan) or miriplatin (MPT; Dainippon Sumitomo Pharma Co., Ltd., Osaka, Japan) was used as the chemotherapeutic agent, mixed at a ratio of EPIR/contrast medium/Lipiodol of 10 mg/1.0 mL/2.0 mL or MPT/contrast medium/Lipiodol of 60 mg/0 mL/3.0 mL, respectively.

RFA
RFA was performed immediately after c-TACE. CT was first performed for planning (Figure 2A-D), after which, under CT guidance, electrode puncture was performed into the area of Lipiodol that was injected during TACE under local anesthesia using lidocaine and minor sedation with pentazocine and midazolam. CT-guided puncture was sometimes performed under oblique imaging using direct MPR methods (Figure 2E and F) [5], and sometimes under axial imaging using one-shot half mode [6].
The RFA protocols were as follows:
a) Cool-tip needle (Medtronic Japan, Co., Ltd., Tokyo, Japan)
Ablation was started at 30 watts with a 2-cm, non-insulated part and 40 watts at 3 cm, and power was increased manually at 10 watts/min until breakdown occurred. The power was then decreased 10 watts and maintained to the end. Ablation time was 6 minutes at 2 cm and 12 minutes at 3 cm.
b) LeVeen needle (Boston Scientific, Co., Ltd., Tokyo, Japan)
Ablation was started at 30 watts with a 2-cm, non-insulated part and 40 watts at 3 cm, and the power was then increased by 10 watts/min for 5 minutes. If roll-off occurred within 5 minutes, the procedure was restarted 30 seconds later at half power, and the power was maintained until roll-off occurred. If roll-off did not occur, the output was increased by 10 watts/minute until roll-off occurred, and then the procedure was restarted at 70% power until roll-off occurred again.
c) VIVARF needle (Century Medical, Co., Ltd., Tokyo, Japan)
Basically, the protocol was the same as the one for the cool-tip needle. Power was automatically increased at 3 watts/30 seconds.

Angiography after RFA
After removing the RFA needle, screening angiography was performed to check for extravasation. If extravasation was identified, hemostasis was achieved by embolization performed immediately using gelatin sponge particles and/or coils (Figure 3A and B).

Follow-up
In all patients, either triple-phase dynamic CT or ethoxybenzyl diethylenetriamine pentaacetic acid-enhanced MRI including triple-phase dynamic imaging and blood examinations including complete blood count, albumin, total bilirubin, and serum creatinine were performed the day after the procedure and every 3 months to check for recurrence.

Evaluation
Patient background, survival, local recurrence, adverse events, post-procedural bleeding, and post-procedural treatment for recurrence were evaluated. Technical success was defined as no obvious tumor remnants on CT or MRI the day after the procedure.
In addition, the patient's radiation exposure was evaluated using the total dose length product (total DLP) for CT and the incident air kerma at the patient entrance reference point displayed on the equipment (Ka,r) for angiography.

Ethics statement
This retrospective study was approved by our institutional review board, and all procedures were performed with patients' written, informed consent.

Results
Table 1 lists the patients' characteristics. Included in the study were 28 female and 114 male patients (median age, 74 years; age range, 49-90 years). The etiology of HCC was hepatitis B/hepatitis C/hepatitis B+C/hepatitis non-B non-C in 49/27/28/38 patients, respectively (per patient). The Child-Pugh score was 5/6/≥7 in 137/41/8, and modified albumin-bilirubin (mALBI) was 1/2a/2b/3 in 97/45/42/2, respectively (per procedure). Table 2 lists the characteristics of the TACE and RFA procedures. Imaging performed the day after RFA showed clear local residuals in 7 procedures, all of which underwent additional RFA. Additional RFA was performed the same day or the day after CT was performed. The 1-, 2-, and 3-year overall survival rates were 96.1%, 87.4%, and 74.0%, respectively (Figure 4A). Local recurrence developed after 33 of 186 procedures, and the 1-, 2-, and 3-year local recurrence-free survival rates (per procedure) were 86.4%, 76.6%, and 57.5%, respectively (Figure 4B). RFA needle tract ablation was performed in 77 procedures, and post-procedural bleeding occurred in 17 of 186 procedures (9.1%), for which embolization was required in 13. In the remaining 4 cases, the bleeding stopped spontaneously. In addition, there were no cases of fluctuating vital signs and no cases of pseudoaneurysm formation. Furthermore, there was no significant difference in post-procedural bleeding between patients who underwent tract ablation and those who did not (6/77 vs 11/109, p = 0.5890). No other procedure-related bleeding was observed in any patient, and no patient required blood transfusion. Other adverse events are shown in Table 3. Severe adverse events that caused a prolonged hospital stay occurred in 7 procedures, including hepatic encephalopathy in 1 patient, liver abscess (biloma) that required percutaneous transhepatic drainage in 1, hyperamylasemia in 1, pneumothorax in 1, hyperglycemia (diabetic ketoacidosis) in 1, loss of appetite in 1, and renal dysfunction that required hydration (increased serum creatinine) in 1.
After the procedure, no patients received adjuvant chemotherapy, and 97 patients received some treatment for recurrence, as follows: resection in 2 patients (hepatectomy in 1 patient, pneumonectomy for metastases in 1 patient), RFA in 71 patients, TACE in 63 patients, hepatic arterial infusion chemotherapy in 11 patients, systemic chemotherapy including a cytotoxic agent, molecular target agent, and immune checkpoint-inhibitor in 32 patients, percutaneous ethanol injection in 2 patients, and radiation therapy including for cure and palliation in 2 patients. No patients underwent transplantation.
Radiation exposure was recorded on the CT system in 185 of 186 procedures and on the angiography system in 38 of 186 procedures. The radiation exposure was 2,183 (330-9,913) mGy-cm for CT (median total DLP) and 467.716 (73.695-1,899.945) mGy for the angiography equipment (median Ka,r).

Discussion
RFA is usually performed under US guidance, which has the advantage of avoiding the possibility of major vessel or organ injury. However, US has a blind area in the infraphrenic region, due to overlying ribs, patient factors such as obesity (subcutaneous fat, visceral fat), and intestinal gas. Although the creation of artificial ascites or pleural effusions can be useful in eliminating these blind areas, there are cases in which the presence of postoperative intestinal adhesions and US attenuation due to fatty tissue cannot be addressed. Contrast-enhanced US can be used to detect lesions due to poor intrahepatic contrast caused by cirrhosis and so on, but equipment that can handle contrast-enhanced US is required. However, CT has no blind areas, but since CT-guided puncture is usually performed without contrast medium, the vessels are not depicted clearly. Therefore, CT-guided liver puncture has the potential risk of severe bleeding from vessel injury.
TACE is used to treat intermediate-stage HCC and may sometimes be performed together with RFA to improve local control of RFA. The Lipiodol used in TACE enables visualization of the HCC nodule as a puncture target.
In comparison with previous reports, in which the risk of bleeding of RFA under CT guidance was 1.1%-23.8% [2, 7, 8], in the present study it was 9.1%. Takaki et al. [7] performed pre-RFA angiography to avoid vessel puncture but did not describe post-RFA angiography, and they did not provide details of the TACE procedure (e.g., whether segmental, subsegmental, ultra-selective, or widespread). In the present study, TACE was performed in a selective or ultra-selective manner [3, 4], and angiography was performed immediately after removing the electrode; however, angiography or CT-angiography just before RFA was not performed. This sequence is similar to that reported by Mitani et al. [2], and the rate of bleeding was also similar between the studies [8]. Since the wide area included in the puncture line cannot be covered by ultra-selective TACE, the risk of bleeding is higher than for segmental or lobar TACE. Performing widespread TACE and pre-RFA angiography might reduce the risk of bleeding; however, widespread TACE may cause liver dysfunction. In addition, there were no cases of fluctuating vital signs or pseudoaneurysm formation after hemorrhage was detected. This may be due to the fact that embolization was performed as soon as it was detected.
Tract ablation after removal of the electrode is another factor related to bleeding; however, the present study found no significant difference in bleeding whether or not tract ablation was performed, and, therefore, tract ablation appears unimportant with regard to avoiding bleeding after electrode removal.
Although spontaneous hemostasis occurred in 4 of the present cases, spontaneous hemostasis may have been achieved in some of these cases by waiting slightly longer. Because angiography was performed immediately after removing the electrode, and the rate of bleeding was similar to that reported by Mitani et al. [2], 9.1% could be the true number. Alternatively, embolization for hemostasis is easily performed because the catheter remains in the hepatic artery throughout the procedure. Accordingly, embolization can be performed as soon as bleeding is detected, which in combination with simultaneous TACE improves the safety of CT-guided RFA, as described in previous reports [7, 8]. It is noteworthy that no severe complications occurred with the present method, and there were no deaths due to bleeding.
Other complications were tolerable and among those commonly experienced with RFA or TACE in clinical practice. Only a limited number of patients had a prolonged hospital stay due to complications.
Tumor control was similar to that reported previously [9-12]. Therefore, we consider that simultaneous TACE combined with RFA has several advantages.
Radiation exposure of CT-guided RFA was tolerable. The Japanese Diagnostic Reference Level is 2,100 mGy (total DLP) for multi-phase CT of the liver and 1,400 mGy (Ka,r) for TACE [13], and the present technique does not deviate significantly from these levels. In addition, regarding radiation exposure, radiation exposure data were analyzed in 185 cases, excluding 1 case of CT, but due to the absence of records, only 38 of 186 cases were analyzed for exposure to angiography equipment. However, the TACE technique is the same with or without RFA, and in addition, patients undergoing RFA are at an early stage and have smaller and fewer lesions than patients undergoing TACE in general.
This study has 3 limitations. First, this combination treatment has longer procedure time and higher radiation exposure dose than conventional TACE. However, we believe that they are justified because we performed the combination treatment only for the lesions that were difficult to detect on US. Also, the radiation exposure dose of CT-guided RFA was tolerable in this study. Since we used intermittent imaging (1-shot volume scan), its radiation exposure dose was less than that of continuous fluoroscopy. Second, an IVR-CT system is necessary for this combination treatment. TACE and CT-guided-RFA need to be performed simultaneously, and oblique puncture is essential for CT-guided-RFA. Finally, this was a single-center retrospective study with a small number of patients.

Conclusion
In conclusion, CT-guided RFA with simultaneous TACE is a useful treatment for early-stage HCCs that cannot be detected on US.

Author Contributions
All authors contributed to the study conception and design. Material preparation, data collection, and analysis were performed by Takeshi Aramaki. The first draft of the manuscript was written by Takeshi Aramaki, and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Conflicts of Interest
There are no conflicts of interest.

Ethical Approval
This retrospective study was approved by the institutional review board of Shizuoka Cancer Center (No. J2023-14-2023-1-3).

Informed Consent
All procedures were performed with patients' written, informed consent.

Consent for Publication
For this type of study, consent for publication is not required.

Disclaimer
Dr. Rui Sato, an author of this paper and a member of the Editorial Board of this journal, was not involved in the peer-review or editorial decision-making process.