Three types of RF electrodes are currently available commercially: two brands of retractable needle electrodes (model 70 and model 90 Starburst XL needles, RITA Medical Systems, Mountain View, CA, USA; LeVeen needle electrode, Boston Scientific, Boston, MA, USA) and an internally cooled electrode (Cool-Tip RF electrode; Radionics, Burlington, MA, USA)[15].

The needle electrodes of RITA consist of a 14-gauge insulated outer needle that houses nine retractable curved electrodes of various lengths. When the electrodes are extended, the device assumes the approximate configuration of a Christmas tree. Nine of the electrodes are hollow and contain thermocouples in their tips in order to measure the temperature of adjacent tissue. The alternating electric current generator comes in a 250-W model at 460 kHz (Model 1500X RF Generator, RITA Medical Systems). The ablation algorithm is based on temperature at the tips of the electrodes. After the ablation cycle is completed, a temperature reading from the extended electrodes in excess of 50°C at 1 min is considered to indicate satisfactory ablation.

Another RFA device (LeVeen Needle Electrode; Radiotherapeutics) has retractable curved electrodes and an insulated 17-gauge outer needle that houses 10 solid retractable curved electrodes that, when deployed, assume the configuration of an umbrella. The electrodes are manufactured in different lengths (2- to 4.0-cm umbrella diameter). The alternating electric current generator is 200 W operated at 480 kHz (RF 3000; Boston Scientific). The ablation algorithm is based on tissue impedance, and ablation is considered successful if the device impedes out.

The third RFA device (Cool-Tip radiofrequency electrode; Radionics) has an insulated hollow 17-gauge needle with an exposed needle tip of variable length (2- or 3-cm). The tip of the needle contains a thermocouple to record the temperature of adjacent tissue. The shaft of the needle has two internal channels to allow the needle to be perfused with chilled water. In an attempt to further increase the size of the ablation area, the manufacturer placed three of the cooled needles in a parallel triangular cluster with a common hub. The generator has a peak power output of 200 W and is operated at 480 kHz (CC-1; Radionics). The ablation algorithm is based on tissue impedance, and ablation is considered successful if the device impedes out. As a result, successful ablations usually increase the temperature of the ablated tissue to above 60°C.

In patients with HCC, exclusion criteria should include evidence of extrahepatic metastases and/or lobar and local portal venous thrombosis or uncontrolled liver disease decompensation, patients with clotting impairment, renal failure, or Child-Pugh class C cirrhosis. In the EASL Consensus Conference criteria[20], all patients that had tumor nodules with a maximum diameter of 3 cm and not more than three in number with contraindications for surgery are included.

The technical effectiveness of ablation is commonly assessed by findings on contrast-enhanced computed tomography (CT) or magnetic resonance imaging. A tumor was considered to have been successfully ablated when there were no longer any enhanced regions within the entire tumor during the arterial phase and at least a 0.5 cm margin of apparently normal hepatic tissue surrounding the tumor during the portal phase (Figure 1)[21-23]. This safety margin for RFA therapy is necessary from the perspective of partial volume effect. Failure to establish a sufficient ablative safety margin was shown to be an independently significant risk factor for local tumor progression on multivariate analysis[24]. Part of the tumor was diagnosed as remaining viable when images of the ablated area showed nodular peripheral enhancement[25].

Open in New Tab   Full Size Figure   Download Figure

Figure 1 A 61-year-old man with 1. 5-cm recurrent hepatocellular carcinoma after ablation therapy in segment 5 of the liver. A: Early-phase dynamic computed tomography (CT) scan shows recurrent tumor (circle). Non-enhanced area (arrowheads) was previously treated by radiofrequency ablation (RFA); B: Contrast harmonic ultrasound (US) using Levovist shows enhancement of viable focus of a hepatocellular carcinoma (HCC) nodule (circle); C: Portal-phase dynamic CT scan, which was obtained 3 d after RFA shows that the tumor was not enhanced, indicating complete necrosis of the lesion (arrow); D: Contrast harmonic US, which was obtained 3 d after ablation shows non-enhanced area (circle).

A randomized control trial (RCT) has shown that RFA achieved survival rates similar to those achieved following resection[26] (Table 1). Chen et al[26] conducted a RCT on 180 patients with a solitary HCC ≤ 5 cm indicated to receive either percutaneous RFA or surgical resection. This study showed percutaneous RFA achieved the same overall and disease-free survival rates as surgical resection for patients with small solitary HCC. The 1- and 4-year overall survival rates after percutaneous RFA and surgery were 95.8%, 67.9% and 93.3%, 64.0%, respectively. The corresponding disease-free survival rates were 85.9%, 46.4% and 86.6%, 51.6%, respectively. However, in cases of primary liver cancer in which local curative therapy was achieved by securing a safety margin, the 4-year survival rate was relatively high, at 66%-82% (results in Japan)[27,28]. Percutaneous RFA has an advantage over liver resection in providing a better short-term postoperative result because local ablative therapy is a less invasive procedure. Although promising, these data need to be confirmed in larger RCTs before local ablative therapy can replace partial hepatectomy in the treatment of good surgical candidates.

RFA has also been investigated for treating patients with large or multifocal tumors. However, the size and number of tumors are important factors determining the local recurrence rate after RFA[29]. Apart from the larger tumor volume, large liver cancers more frequently have irregular borders and satellite lesions. Therefore, precise tailoring of the size and shape of the thermal lesion is important in RFA for large liver cancers. A number of precisely calculated overlapping coagulation zones are necessary to treat large liver cancers. To increase the size of the coagulation zone in RFA, investigators tried using vascular occlusion during RFA[30,31]. Temporary interruption of hepatic blood flow using vascular occlusion techniques (e.g. balloon catheter occlusion of the hepatic artery, transcatheter arterial embolization (TAE), or transcatheter arterial chemoembolization (TACE) has been shown to increase the efficacy of interstitial thermotherapy due to a significant increase in lesion volume. Vascular occlusion causes a reduction of heat dispersion, thus increasing the range of therapeutic thermal coagulation. Peng et al[32] reported a series of 120 patients with HCC, and the 1-, 2-, 3-, and 5-year overall survival rates for the TACE-RFA and RFA groups were 93%, 83%, 75%, 50%, and 89%, 76%, 64%, 42%, respectively (P = 0.045).

Ultrasound (US)-guided procedures are necessary but have limited use when the tumor is located under the diaphragm. However, saline solution injection into the pleural cavity can separate the lung and liver on B-mode US, i.e. artificial pleural effusion acts as an acoustic media. There are reports on the feasibility and safety of RFA with artificially induced pleural effusion for HCC located in the right subphrenic region[33-36]. In a series of 24 patients with HCC located in the hepatic dome, 200-1100 mL of 5% glucose solution was infused intrathoracically to separate the lung and liver, thus, complete tumor necrosis in a single session was achieved in 96.4% of patients[36].

Multiple RFA sessions for locally progressive HCCs were previously required because it is frequently difficult to distinguish viable tumors from necrotic tissue on B-mode US[37]. However, contrast-enhanced harmonic US imaging is able to evaluate small hypervascular HCCs even when B-mode US cannot adequately characterize the tumors[38-43]. In particular, contrast harmonic US has been improved by the development of second-generation contrast agents such as sulfur hexafluoride microbubbles (SonoVue; Bracco SpA, Milan, Italy), perflutren lipid microbubbles (Definity; Bristol- Myers Squibb, North Billerica, MA, USA), perflutren protein microbubbles (Optison; GE Healthcare, Buckinghamshire, UK), and perfluorocarbon microbubbles (Sonazoid; Daiichi-Sankyo, Tokyo, Japan). These microbubbles provide stable nonlinear oscillation in a low power acoustic field due to the hard shell of these bubbles, producing great detail in the harmonic signals in real-time[44-49]. It has been reported that contrast harmonic sonography-guided RFA is an efficient approach for guiding further ablation of hepatic malignancies that are not clearly demarcated by B-mode US (Figure 2)[50-54].

Open in New Tab   Full Size Figure   Download Figure

Figure 2 A 71-year-old man with 2. 0 cm local tumor progression of hepatocellular carcinoma after radiofrequency ablation therapy in segment 8 of the liver. A: Early-phase dynamic computed tomography (CT) scan shows outgrowth pattern of locally progressive hepatocellular carcinoma (HCC) (arrow). The lesion borders an unenhanced area, which was previously treated; B: Left: Contrast harmonic Doppler ultrasound (US) using Levovist shows enhancement of local tumor progression of HCC (arrow). Therefore, an enhanced lesion can be identified as a target for the insertion of a single RF electrode; Right: B-mode US shows a HCC nodule demonstrated as a low echoic lesion with an unclear border (arrowhead).

The use of a laparoscopic or open approach allows repeated placements of RFA electrodes at multiple sites to ablate larger tumors. The laparoscopic approach appears to be the safest and most effective method for small tumors on the liver surface, and offers the advantages of laparoscopic US, which provides better resolution of the number and location of liver tumors[55,56]. Moreover, a hand-assisted technique can be applied safely and effectively to laparoscopic liver surgery[57-59]. An intraoperative US probe is inserted into the peritoneal cavity together with the surgeon’s hand through a hand-access device. An RF electrode can be subcostally or intercostally advanced into a liver tumor under direct guidance by intraoperative US. Therefore, a hand-assisted laparoscopic US-guided method has advantages for both laparoscopic and open surgical approaches. The postoperative recovery of patients was shorter compared with that after an open surgical approach. Ishiko et al[57] reported that the surgical procedures consisted of 5 RFA to tumors in the caudate lobe with hand-assisted laparoscopic surgery (HALS), and a postoperative CT scan demonstrated sufficient ablation in all patients and there was no surgical mortality. The HALS approach has several advantages; it facilitates and expedites the procedure, reduces the stress factor on the surgeon, greatly improves exposure, and facilitates immediate and efficient control of bleeding vessels with the internal hand. The hand-access device, which essentially consists of a cuff with a spiral inflatable valve, enables withdrawal and reinsertion of the hand without loss of pneumoperitoneum during the procedure. However, the local treatment failure rate of the laparoscopic approach was higher in patients with HCC nodules situated deep within the liver and measuring 4 cm or more in diameter[60]. Great difficulty can be encountered during treatment of lesions located close to the gallbladder or in contact with the diaphragm.

Although more invasive, open RFA can be performed more easily and the puncture course of the RF needle can be more widely selected than that during the laparoscopic approach. It has been reported that patients undergoing radical open RFA demonstrated few ablation site recurrences even though the nodules measured more than 4 cm in diameter and/or there were more than three nodules[61,62]. Open RFA can be indicated for patients who are considered suitable for open surgery with large, numerous, or deeply located tumors that cannot be accurately accessed by a laparoscopic approach. Furthermore, when patients have synchronous liver metastases, open surgical RFA can be performed in conjunction with resection of the primary cancer.

The local recurrence rate after RFA for HCC ranged from 1.7% to 41%[63-70] (Table 2). As reported by Kudo[28], in a series of 141 HCC patients who underwent curative RFA therapy, local tumor progression was observed in 9 cases (local tumor progression rate, 6.3%), whereas the cumulative local tumor progression rate, calculated by the Kaplan-Meier method, was 12% at 4 years. The rate may have depended on the size of nodules treated and the skill of the surgeons. There has not been any definitive report of local recurrence of nodules measuring 2-cm or smaller, and we ourselves have not encountered any case showing such recurrence, suggesting that recurrence in such cases is exceptional. The risk of local tumor progression increases with size, but the local tumor progression rate differs markedly depending on whether or not a circumferential 5-mm safety margin is secured. In a meta-analysis of RFA vs PEI in HCC, the survival rate showed a significant benefit for RFA over PEI at 1, 2, 3 and 4 years[71]. The survival advantage increased over time with Relative Risk values of: 1.28 (95% CI: 1.12-1.45) and 1.24 (95% CI: 1.05-1.48) for RFA vs PEI at 3- and 4-years, respectively. Likewise, RFA achieved significantly lower rates of local recurrence (RR: 0.37, 95% CI: 0.23-0.59)[71].

Complications reported following percutaneous RFA of malignant liver tumors in 2320 patients treated at 41 different hospitals in Italy indicate that the mortality rate was 0.3% with an overall complication rate of 7.1%[72,73]. The authors described major complications (2.4% incidence) including death, hemorrhage, RFA needle-track seeding, RFA lesion abscess, perforation of gastrointestinal viscus, liver failure, biloma, biliary stricture, portal vein thrombosis, and hemothorax or pneumothorax requiring drainage, and minor complications (4.7% incidence) including pain, fever, and asymptomatic pleural effusion. Another recent review indicated that complication rates for percutaneous, laparoscopic, and open RFA of hepatic tumors in 3670 patients were 7.2%, 9.5%, and 9.9%, respectively[74]. Complications directly related to the liver included bleeding (1.6%), intrahepatic abscess (1.1%), biliary or hepatic vascular injury (1.7%), and liver failure (0.8%). Complications that arose in less than 1% of hepatic tumor RFA patients included pulmonary problems (pneumothorax, hydrothorax, pleural effusion), grounding pad skin burn, myoglobinemia or myoglobinuria, renal failure, coagulopathy, tumor seeding of the needle track, excessive hormone release from treated neuroendocrine tumors, cardiac problems (myocardial infarction, arrhythmia), and injury to the diaphragm or adjacent viscera. Although Llovet et al[75] reported that dissemination along the puncture route was observed in 12.5% of their patients, only a few such cases have been reported in Japan, and dissemination may not occur at such a high frequency. This complication was almost absent in many reports from Japan[28]. Overall, the frequency of major complications of percutaneous RFA was 0.6%-8.9%, which was higher than that of PEI, but generally lower than that of MCT[28].

Some investigators have suggested that tumor location is closely related to the risk of major complications. Central tumors close to the hepatic hilum were reported to be unsuitable for percutaneous RFA because of the risk of injuring adjacent bile ducts[15]. It was also suggested that RFA for nodules adjacent to large vessels might often result in incomplete necrosis because of a heat sink effect. In addition, peripheral tumors adjacent to extrahepatic organs were also suggested to be unsuitable because of the risk of heat injuries, such as intestinal perforation and pleural effusion[72,76]. Thus, there may be difficulty with RFA of nodules in such high-risk locations, possibly resulting in complications or preventing adequate treatment. However, Teratani et al[77] reported that there was no difference in early complication rates according to tumor location. The effort to achieve thorough ablation increased the total number of electrode insertions, and this may have led to an increase in complications.

To minimize complications of RFA for malignant liver tumors, knowledge of risk factors and prevention methods is required. In addition, because early and accurate diagnosis is necessary for the proper management of complications, not only radiologists but also hepatologists and surgeons should be familiar with the imaging features of each type of complication. Appropriate management of complications is essential for successful treatment with RFA.