The imprint of radiofrequency in the management of hepatocellular carcinoma

Abstract

Background. This article reviews the current results of radiofrequency application in the management of hepatocellular carcinoma (HCC) with reference to the comparison between the different surgical modalities. Method. An electronic search was performed for studies on the treatment of HCC. Results. Thermoablation by means of radiofrequency (RFA), microwave coagulation therapy (MCT) and laser-induced thermotherapy (LITT) provides tumor necrosis with a low complication rate. These methods are still not predictable and it is difficult to monitor the extent of necrosis in a real-time manner. Combined transarterial embolization and RF ablation is a promising strategy for large HCCs. Radiofrequency-assisted liver resection is unique and has become very popular recently because it permits parenchymal transection with minimal blood loss. Conclusion. Many alternative techniques have been applied recently for the management of HCC but their exact roles need to be defined by randomized studies. Advances in technology and refinements in technique may provide an effective and predictable way to ablate liver tumors using radiofrequency devices.

Introduction

Hepatocellular carcinoma (HCC) is one of the most common malignancies worldwide with an annual occurrence of one million new cases ¹,²,³,⁴. An etiologic association between HBV infection and the development of HCC has been established with a relative risk 200-fold greater than in non-infected individuals ⁵.

Hepatitis C virus is also proving an important predisposing factor for this malignancy, with an incidence rate of 7% at 5 years and 14% at 10 years. The prognosis depends on tumor stage and degree of liver function, which affect the tolerance to invasive treatments ⁶.

Surgical resection has generally been accepted as the treatment of choice for HCC but new treatment strategies have been developed nowadays including local ablative therapies, transarterial embolization and liver transplantation. With increasing detection of small HCCs from screening programs for cirrhotic patients it is foreseen that locoregional therapy will play an important role in the near future.

Major advances have been made with radiofrequency energy application for ablation and bloodless liver resection and the key questions with the future perspectives are presented in this article.

Radiofrequency ablation (RFA)

Radiofrequency ablation (RFA) of liver tumors was pioneered in 1993 by Rossi et al. ⁷. RFA induces deep thermal injury in hepatic tissue while sparing the normal parenchyma. Its basic principle includes generation of high-frequency alternating current (400 MHz) which causes ionic agitation and conversion to heat, with subsequent evaporation of intracellular water which leads to coagulation necrosis. The area of the injury depends on the size, position and shape of electrode used.

RFA has been performed by percutaneous, laparoscopic or open techniques ⁸,⁹. Percutaneous RFA usually is performed under sedation due to severe pain ¹⁰. The disadvantages of the method are considered the inability for vascular inflow occlusion through the percutaneous approach, and difficult access for deep tumors located near blood vessels, or neighboring the diaphragm or the bowel. On the other hand, imaging of RFA performed during laparotomy or laparoscopy is limited to ultrasound (US), while the percutaneous approach offers the ability to use US, magnetic resonance imaging (MRI) or computed tomography (CT) guidance. The main indication is recurrences after open procedures and patients with poor performance status ¹¹.

The electrode is placed through normal liver tissue close to the tumor margin and guided by US. The tissue is ablated at a temperature >90°C for 5–12 minutes or until the impedance increases rapidly, although multiple overlapping ablations are necessary to completely destroy a tumor exceeding 3 cm in diameter. At the end of RFA the ablated tissue takes on the appearance of a lacrosse racket.

Currently, three devices are approved. RITA Medical Systems produce a 50-W 460-kHz alternating current generator with a variety of needles. Three, four or seven retractable prongs are deployed once the needle is positioned. A thermocouple at the tip of each prong monitors the temperature. Radiotherapeutics manufacture a 90-W generator with a 10-prong deployable needle. Therapy is monitored by measuring the impedance with power roll of when impedance increases associated with completeness of tissue ablation. Finally Radionics has an internal circuit of cold water in the needle which may increase the size of the thermal injury.

There are limitations related to the physics of the RF process. Tissue charring causes increased impedance that results in decreased energy absorption and a smaller treated tissue volume. Although large amounts of tissue can be ablated in vitro, the charring and ‘heat sink phenomenon’ are difficult to overcome ¹².

Inflow occlusion by means of the Pringle maneuver increases the size of the zone of coagulated necrosis and enhances the likelihood of complete tumor cell kill even if the tumor abuts a major intrahepatic blood vessel. RFA treatment combined with vascular inflow occlusion can also produce complete circumferential necrosis around major vessels without damaging the integrity of their walls. This modified method of intraoperative RF takes advantage of the inflow occlusion and proved superior to the standard method in the retrospective study by Yamasaki et al. ¹³. Alternative techniques based on the same concept include angiographic balloon occlusion of the hepatic artery of the liver segment involved with tumor ¹⁴ and percutaneous transhepatic portal balloon occlusion ¹⁵,¹⁶.

Radiographic assessment of the ablated lesion should be delayed for 1 month following treatment because of the inability to distinguish between edematous tissue surrounding the lesion and residual tumor early after the ablation ¹⁰. Successful radiofrequency ablation of lesions initially causes peritumoral hyperemia at the interface of the ablated and normal tissue on postoperative CT, which is identified by characteristic hypervascular rims with intense contrast enhancement (Figure 1). Unlike cryoablation, in which the evolving ablated area can be monitored by US, the hyperechoic area in RFA does not correspond exactly to the area of tumor ablation, nor does it indicate whether tumor ablation is complete. Thus the extent of necrosis can be more accurately assessed by helical CT, MRI or color Doppler scan with bubble contrast. Spectophotometry is also a unique method of tissue necrosis detection, although it is not widely available.

Figure 1. .

Contrast-enhanced CT images. (a) A 5-cm hypervascular hepatocellula carcinoma (HCC) is identified on VIIth liver segment. (b) Percutaneous RFA CT-guided procedure. (c) Image obtained after RFA showing necrosis of the tumor and a contrast-enhancing rim.

Follow-up is done with CT scans and tumor markers. The patient should undergo clinical evaluation, serial liver tests and tumor marker determinations every 3 months for 2 years ¹⁰. Persistent non-enhancement is seen in coagulated tissues but not in viable tumor on delayed-opacification images (Figure 2). A CT scan performed during the first 3 months usually demonstrates a cystic-density area that is larger than the original tumor. Decrease in size of this ablated zone over time is a sign of successful treatment ¹⁷. A gas bubble sign can be also noted, but Mitsuzaki reported only 4 of 19 lesions being related to abscesses ¹⁸. On MRI local recurrence is expressed by contrast enhancement, increase of ablated lesion size, or evidence of areas showing low signal intensity or high signal intensity on T1- and T2-weighted images accordingly.

Figure 2. .

Follow-up contrast-enhanced CT scan of an ablated centrally located HCC with no signs of residual mass or recurrence.

Using the aforementioned diagnostic tools several studies have shown complete tumor necrosis in 80–90% of HCCs smaller than 3–5 cm after a single session. Because RFA is a relatively new treatment modality for HCC, the data in the literature are preliminary results with short follow-up (Table I) Rossi et al. reported their 7-year experience in treating HCC with RF in 39 patients with tumors smaller than 3 cm, having survival rates 94 to 40% at 1 and 5 years, respectively ¹⁹. Curley et al., in a series of 110 patients, reported that 79 patients (72%) were alive at a median of 19 months with 48% of them free of disease ²⁰.

Table I. Follow-up results after radiofrequency (RF) ablation therapy for HCC.

Study	Number of patients	Tumor size (cm)	Route of RFA	Necrosis (%)	Recurrence (%)
Rossi et al., 1996 21	39	≤3	P	95	41
Rossi et al., 1998 19	23	≤3.5	P	100	28
Francica & Marone, 1999 54	15	1–4.3	P	90	33
Curley et al., 1999 42	48	–	P and I	100	2.1
Nicoli et al., 2000 55	47	1–6	P and I	100	–
Curley et al., 2000 20	110	Mean 3.4	P, L, and I	95	49
Poggi et al., 2001 56	15	1.5–6.2	P	88	20
Buscarini et al., 2001 57	88	≤3.5	P	93	39
Elias et al., 2002 58	19	≤3.5	P
Kuvshinoff & Ota, 2002 41	11	Mean 4.0	P, L, and I		27
Adam et al., 2002 41	18	Mean 2.8	P	76	17
Yamasaki et al., 2002 35	31	≤4.0	P
Shibata et al., 2002 31	36	Mean 2.3	P	96	4

P, percutaneous; L, laparoscopic, I, intraoperative RF.

Like other local ablative therapies, intrahepatic recurrence of tumor is common, ranging from 20 to 49% ¹⁹,²¹. However, the majority of recurrences are new lesions that are probably related to multicentric carcinogenesis. Tumor recurrence often occurs at the radial margins of the ablated tissue (Figure 3). Patient-to-patient variability of tumor geometry, tissue response and location of the tumor close to vasculature can lead to an incomplete ablation margin and predispose to local tumor recurrence.

Figure 3. .

(a) Contrast-enhanced CT image of a percutaneously ablated tumor in VIth liver segment. (b) On the follow-up CT scan 9 months after RFA an enhanced mass is demonstrated at the margins of the ablated area near a dilated vessel, consistent with recurrence.

RFA versus other ablative techniques

Recently, RFA has been compared with other local ablative therapies and the data suggest that RFA may be a superior option. The clinical trials ²²,²³ comparing RFA to percutaneous ethanol injection (PEI) demonstrate a clear advantage for RFA. Necrosis induced by RFA is more predictable and treatment by a single session is sufficient in most patients with small tumors. PEI is preferable for patients with relatively advanced liver disease and tumor located close to the hilum, making RFA a difficult and less attractive option. The disadvantages of PEI compared with the other local ablative modalities and RF in particular is that the distribution of ethanol is impeded by the presence of intratumoral septa or tumor capsule and repeated sessions are usually required for a better tumoricidal effect. However, the combination of PEI and super-selective embolotherapy makes the lesions more receptive to the ethanol destroying tumor septation ²⁴. Tumor seeding has been reported to be one of the rare complications after ethanol injection and is considered an advantage compared with RF with an incidence of 0.6–1.1% ²⁵ due to the small thickness of the needle.

Livraghi et al. ²² showed that for tumors <3 cm, there is a clear benefit in tumor necrosis after RFA (90% vs 80% after PEI). A Japanese study also demonstrated complete tumor necrosis after RFA but only 94% after ethanol injection ²³. RFA demands fewer treatment sessions for the same rate of necrosis when compared with PEI ²⁰,²⁶,²⁷. The study by Lencioni et al., apart from the benefit concerning the percentage of tumor necrosis (91% vs 85%), demonstrated a clear advantage in recurrence rate (4% vs 17%) when comparing RFA with PEI ²⁶. Better survival has also been demonstrated by Olschewski et al. ²⁸ in a prospective randomized study when RFA was compared with PEI.

Apart from the more predictable rate of necrosis, the risk of tumor seeding as reported earlier seems to be higher among patients who underwent RFA compared with the PEI group. A biopsy-proven needle track tumor seeding 4–18 months after RFA was 12.5% in one study ²⁹, compared with a low incidence (0.6–1%) of seeding after PEI. Tumor cell release associated with intratumoral explosion resulting from the increase in temperature during ablation and the thickness of the RF needles (15–18G), play an important role in the pathogenesis of tumor seeding ²⁵,³⁰.

The advantage of RFA compared with microwave (MCT) is the need for fewer sessions. A randomized study from Japan demonstrated similar therapeutic results, complication rates and residual disease when comparing the two methods, but patients who underwent MCT needed more sessions for the same result (2.4 vs 1.1) ³¹. An experimental study showed that RFA may be superior to MCT regarding the capacity to produce a larger area of necrosis ³².

Although there is a lack of evidence from randomized trials, conflicting data also exist from two nonrandomized studies where cryoablation (PCS) was compared with RFA. In the study by Pearson et al. ³³, cryotherapy was identified to have higher complication (40.7%) and local recurrence rates (13.6%) than radiofrequency (3.3% and 2.2%, respectively). As a result, this group expressed its preference for RF rather than cryosurgery for managing primary or metastatic liver tumors. In contrast, Bilchik et al. ³⁴ compared PCS and RFA using a variety of approaches. Despite the similar findings in morbidity rates, they observed no significant differences in mortality, and overall local recurrence rates (3% vs 2.5%, and 15% vs 10%). Their analysis identified significant differences in local recurrence for tumors exceeding 3 cm (17% for cryotherapy vs 38% for RFA) and therefore they recommend a tailored approach for every patient according to the tumor's size ³⁴.

In a French retrospective study by Adam et al. ³⁵, a group of 36 patients with unresectable HCC underwent either PCS or PRFA. There was no statistical difference in tumor local recurrence rate between the two methods, while similar results were observed in the initial success rate for both treatments (77% and 81%).

The ease of performing percutaneous RFA has also raised the debate as regards replacement of repeat hepatectomies as treatment of choice for tumor recurrence. The non-invasiveness of the method, the shorter length of hospitalization, and the lower mortality and morbidity rates make the method quite an attractive alternative to surgical resection ³⁶,³⁷,³⁸,³⁹,⁴⁰ for small lesions.

As radiofrequency can be performed during open or laparoscopic approach, or even via the percutaneous route, a debate about the best approach is also coming on the scene. Kuvshinoff and Ota ⁴¹ compared the three approaches in 45 patients with HCC and metastatic lesions, concluding that the open approach has far superior results as regards local recurrence. They also demonstrated that the type of tumor has a substantial effect on overall disease-free survival. In agreement with the findings of other studies ⁴²,⁴³ they suggest that tumors <4 cm, open approach RFA, and concurrent hepatic resection are associated with improved ablation site recurrence-free survival. The laparoscopic or open approach is preferable in patients with high risk of bleeding from coagulopathy, large HCC >5 cm, superficial nodules adjacent to other visceral organs or deeply located lesions not accessible by percutaneous puncture

Indications for RFA

It is important to mention that RFA has been used for the treatment of patients with inoperable HCC or metastatic colorectal cancer. It is an emerging technology that has been proposed as an alternative to conventional PEI with similar objective response rates and fewer sessions. The main indication is considered patients with four or fewer liver tumors <3–5 cm in diameter. Goldberg et al. ⁴⁴ have reported larger tumors (>5 cm) treated successfully by RFA with overlapping of the fields. Many technical innovations such as the cool-tip electrode and portal blood flow interruption can increase the volume of induced coagulative necrosis to 4–5 cm sphere. Large HCC (up to 6 cm in diameter) are treated alternatively by a combination of segmental transcatheter arterial embolization followed by RFA.

Complete necrosis is achieved in 65, 41 and 31% of tumors measuring 3–4 cm, 4–5 cm and 5–8 cm, respectively. The overall survival for patients with small HCC is 94, 68 and 40% at 1, 3 and 5 years, respectively ²¹. Rossi et al. ¹⁹ have reported on an additional 37 patients treated with RFA. In this group there were 26 patients with HCC and 14 patients with hepatic metastases. Although 93% showed evidence of complete tumor ablation, 82% developed recurrence by 12 months of follow-up. Subsequently Livraghi and associates ⁴⁵ achieved 90% complete necrosis of HCC <3 cm, 71% in medium (3–5 cm) and 45% in large size HCC (5–9 cm) with a 3-year survival rate in Child A patients of 85%.

RFA is a well tolerated procedure with a reported complication rate of 3–7% in most series. The percutaneous approach is responsible for the largest number of documented complications. Side effects include mild fever, transient increase in serum transaminase levels and the development of pleural effusion ⁴⁶. Intraperitoneal bleeding occurs in 2–7% of patients, while portal vein thrombosis has also been reported (Figure 4). Both complications are related to severe cirrhosis ²⁰. Hepatic or subcapsular hematoma formation or hemobilia can also occur (Figure 5). Biliary complications have been noted in 1% of patients ⁴⁷; therefore Dominique et al. have used intraductal cooling during RF treatment for tumors close to bile ducts ⁴⁸. RFA leaves a necrotic tissue which is an ideal nidus for infection. Hepatic abscess is therefore a possible event after RF treatment and a few deaths have occurred, mostly due to peritonitis and inadequate aseptic technique ⁴⁹. Special care should be taken in patients with diabetes or biliary-enteric communication.

Figure 4. .

Centrally located thrombus in inferior vena cava (a) and partial thrombosis of portal vein (b), as complications of RFA treatment.

Figure 5. .

Contrast-enhanced CT scan after percutaneous CT-guided RFA, demonstrates a left liver lobe subcapsular hematoma.

Risk minimization is achieved after careful evaluation of the patient's medical condition, immune function, surgical history and liver volume relative to tumor burden. Planning the most appropriate RFA approach is also crucial in order to avoid possible technique-related complications.

The major drawback of RFA is the unpredictable necrosis rate due to the lack of correlation between image changes and extent of ablation. Even histology with special histochemistry techniques (reduced nicotinamide adenine dinucleotide staining) cannot reliably predict the viability of individual cells at the advancing edge of the ablation zone. Thus RFA is an alternative option to resection for small HCC in nonsurgical candidates since resection remains the ‘gold standard’ treatment worldwide. Randomized controlled studies are needed to evaluate the potential benefit of RFA over other local ablative therapies and surgical resection.

Bloodless liver resection

Blood loss during liver resection remains one of the most important factors affecting postoperative morbidity and mortality ⁵⁰. As a result, multiple approaches have evolved to reduce bleeding during the parenchymal transection phase. Recent efforts have aimed to use vessel-sealing devices based on RF energy for parenchymal transection to accomplish bloodless liver resection combined with less intraoperative ischemia (no inflow occlusion).

In an attempt to facilitate liver resection Habbib ⁵¹ popularized his method of hepatectomy using RF-assisted coagulation. Multiple insertions of the RF needle close to the surgical margin create a bloodless field. The needle is placed in the parenchyma under US guidance. RF should last 60 seconds to obtain a zone of tissue necrosis with a radius of at least 1 cm and a depth of 3 cm. After complete coagulation is achieved the probe is placed 2 cm away from the point of previous application. Once RF thermal ablation has been completed along the resection line, the parenchymal transection can be performed with the use of a common scalpel.

The Cool-Tip radiofrequency device (Radionics, Tyco Healthcare) uses RF energy transmitted through the electrode to the adjacent liver parenchyma causing ion vibration with heat production. In fact the device is able to achieve both necrosis of the liver tissue and sealing of blood vessels up to 3 mm in diameter by collagen fusion ⁵². The saline circuit at the tip of the RF needle offers better ion agitation and avoids charring with eschar formation. The device takes advantage of its unique function to enhance heat conduction with a better tumoricidal effect.

These findings lend important functional support to the concept of bloodless liver resection in cirrhotic patients and demanding segmentectomies (segment VIII, central hepatectomy, etc.). Of note, the demarcation line produced by the Cool-Tip device is further extended in the liver remnant, increasing the surgical margin. This effect is useful in cases with poor liver function where limited resections are indicated.

The Tissuelink Monopolar Floating Ball device (TMFB) ⁵³ uses the same principles to seal vascular structures, creating a bloodless field, and seems promising due to the low cost, since is compatible with most electrosurgical generators. Latteri et al. ⁵³ reported their experience with TMFB in nine consecutive patients. They demonstrated minimal blood loss (range 50–300 ml) with an average resection time of 20 minutes (range 10–55 minutes). While the data are encouraging, only two major liver resections were performed. Although TMFB is effective, its use is limited as a transection instrument only, because the floating ball cannot insert in the parenchyma to ablate other lesions further to the segment resected. In this situation of a combined resection–ablation procedure Radionics proved to be more flexible and effective.

The Agia Olga experience

Our group performs liver resection using Radionics as a cutting device for parenchymal transection. The protocol was initiated 1 year ago and has been actively pursued for malignant and benign tumors of the liver. Once the type of hepatectomy is decided upon based on the visual exploration and US examination of the liver, the line of resection is marked by diathermy on the liver surface. Inflow occlusion was never used in our series.

Next the Cool-tip RF (Radionics, Burlington, Tyco Healthcare, MA, USA) with a single 3-cm long needle is applied in the parenchyma. The needle is cooled by circulating cold saline and is coupled to a 480 kHz generator. RFA should last 60 seconds to obtain a zone of tissue necrosis with a radius of 3 cm. Once RF ablation has been completed a common scalpel is used to divide the tissue. In case of bleeding additional application of the needle can be used to achieve hemostasis. The probe then moves following the transection line and a new application is initiated to separate liver tissue.

Although this was not a comparative study, there was an apparent effect on the resection time (median 45 minutes). We noted that the technique is faster than the traditional methods, primarily because it facilitates oozing control. Most patients underwent minor resections (15) and 5 had major resections with minimal blood loss (median 150 ml). No blood transfusion was required nor have we noted any difference in the estimated blood loss and the underlying liver disease, albeit eight patients were cirrhotic. There was also a sense that RF-assisted resection caused less liver enzyme elevation compared with the traditional techniques applied with inflow occlusion. The 30-day mortality rate was zero. Procedure-related complications included a bile leak (which was managed conservatively and regressed after 5 days) in a patient with HCC and underlying cirrhosis who underwent segment V-VI resection at the beginning of our study. It is possible that sealing of biliary radicles requires more time than vessels.

The future

Since the clinical application of RF continues to be a possible alternative to liver resection in the management of HCC, it is useful to consider specific questions regarding how it might be effective to achieve similar results.

RF equipment

The development of advanced RF devices to create homogeneous tissue ablation in a predictable manner with few sessions requires the combined efforts of technology and science. New equipment comprising multiple RF needles to create a bloodless field during resection retains the benefit of minimal blood loss along with a shorter operating time. This concept, if brought to clinical reality, could have the potential to resect liver tumors with low morbidity.

Laparoscopic liver resection using RF

While it has been difficult to translate experimental success (from studies in pigs using RF bloodless liver resection) to the clinical arena, there have been anecdotal reports where this approach has been implemented in humans. These reports lend important functional support to the concept of minimally invasive techniques in liver surgery with low morbidity in the setting of cirrhosis and poor performance status.

The role of real-time imaging

Expected advances in hardware and software, including 3D imaging, will permit improvements in accurate lesion targeting, better evaluation of residual disease and earlier detection of tumor recurrence. PET will also play a crucial role in the future, in differentiating functioning from nonfunctioning residual ‘mass’ after ablation treatment.

The ‘Holy Grail’ of liver surgery in the management of HCC

As such, the potential to prevent tumor recurrence using new adjuvant protocols after the resection-ablation process.