Laparoscopic treatment of liver tumours using a two-needle probe bipolar radiofrequency ablation device

Abstract

Background

Many hepatobiliary centres are increasingly utilizing thermocoagulative devices such as bipolar-radiofrequency ablation (B-RFA). Compared with monopolar-radiofrequency ablation (M-RFA), B-RFA does not require grounding pads, thereby avoiding dermal burn injuries, and does not position probes directly into the tumour but rather on the perimeter. Additionally, B-RFA can precoagulate parenchyma to assist in hepatic resection. Herein, we report our early experience using B-RFA.

Methods

A retrospective review identified 68 patients who underwent M-RFA or B-RFA between June 2004 and September 2010 in an academic centre. Peri-operative metrics were analysed.

Results

M-RFA was used to treat 30 patients, whereas B-RFA was used for 17 patients. There were no differences in peri-operative metrics, survival or disease recurrence between M-RFA and B-RFA. Seventeen additional patients underwent B-RFA precoagulation during laparoscopic resection (segmentectomy in eleven patients and multi-segmental resection in six patients). Four patients with multifocal disease underwent procedures that combined B-RFA with resection.

Conclusions

The early experience utilizing B-RFA demonstrates equivalency to M-RFA with respect to peri-operative metrics and survival. Moreover, B-RFA can be utilized to precoagulate tissue during a planned resection, making it not only a useful tool for tumour therapy but also a useful adjunct during surgical resections.

Introduction

Radiofrequency ablation (RFA) has undergone a refinement of its indications and techniques for the treatment of malignant liver tumours since its description in 1990.¹ The efficacy of ablation is dependent upon multiple factors, including tissue characteristics such as cirrhosis, hepatic steatosis, vascular perfusion and tumour type.² These factors cause variability in the extent of ablation, leading some to question the ability to achieve complete ablation in tumours greater than 3 cm given recurrence rates as high as 34%.^3–5

In the United States, most centres continue to use conventional monopolar-RFA (M-RFA) devices, which employ a single-needle electrode that delivers an electrical current at 400–500 kHz to create a core zone of ablation at the tumour site which extends radially into the adjacent tissue. Importantly, M-RFA requires the use of grounding pads which are placed on the patient's thighs or back to complete the electrical circuit. The single-needle tip electrode can reach target temperatures of over 100°C, but is prone to rapid charring thereby increasing tissue impedance which limits thermal energy distribution. While M-RFA is sufficient for small tumours (<2.0 cm), larger tumours require expandable electrodes (i.e. longer tines) and longer coagulation times to increase the radius of ablation.^3–5 M-RFA is also highly susceptible to ‘heat sinks,’ or convective heat loss from adjacent hepatic vessels. Experimental studies have determined that the presence and number of adjacent vessels directly impacts the efficacy and size of ablation.⁶ Other limitations of this modality include myoglobinaemia, cardiac arrhythmias, increased total operative time, risk of incomplete tumour ablation and thermal skin injury underneath the grounding pads.^7–9 The incidence of serious dermal complications ranges from 5–33% for first degree burns and 0.3–3.2% for second- and third- degree burns.¹⁰

In an effort to overcome the shortfalls of M-RFA, many centres are increasingly employing other thermocoagulative devices to treat liver tumours, including microwave ablation and bipolar-RFA (B-RFA).^11,12 B-RFA utilizes dual probes which are placed ideally in healthy parenchyma surrounding the tumour to deliver a 440–480 kHz energy field in a controlled fashion between the probes. Depending on the type of B-RFA probes used, either a rounded rectangular or a rounded cube ablation distribution is achieved to a reliable tissue ablation area up to 7 cm.² Referred to as a ‘line of sight’ delivery system,¹² B-RFA is completely different than M-RFA which requires direct penetration of the targeted tumour (Fig. 1). The advantage of ‘line-of-sight’ energy delivery is that it avoids the rapid charring effect around the electrode tip thereby effectively ablating larger tumour volumes.

Figure 1.

Monopolar (M-) vs. bipolar (B-) radiofrequency ablation (RFA): schematic showing differences between monopolar (a) and bipolar (b) ablation. Monopolar probes require single probe insertion directly into the tumour, allowing for thermocoagulation of tissue adjacent to deployable tines. Bipolar ablation requires placement of dual probes in the perimeter of the tumour, and the current stream travels between the probes

In animal studies, B-RFA has been shown to be safe and effective and can also be used with the ‘overlapping’ technique to increase the efficacy and size of the burn zone.¹³ Moreover, B-RFA is less prone to conductive ‘heat sinks,’ and in fact traps heat between probes resulting in higher temperatures, more effective cellular destruction and a larger zone of ablation.¹⁴ Theoretically, this form of RFA energy is less likely to result in necrotic liver parenchyma thereby decreasing the incidence of septic complications when compared with monopolar ablations that are prone to inaccurate energy release.^15,16

In addition to its role as a stand-alone treatment for liver tumours, RFA can also be used as an adjunctive tool during liver resections. RF energy is used to pre-coagulate liver tissue before parenchymal transection as a means of reducing intra-operative blood loss and post-operative morbidity.¹⁷ While both M-RFA and B-RFA energy sources have been expanded to be used in hepatic resections, B-RFA is more widely accepted for this indication.^17,18 Currently, various commercial B-RFA electrode systems are available which are designed specifically to perform liver resections.^15,19

The purpose of the present study was to compare the results of M-RFA with B-RFA in the treatment of liver tumours, and also to evaluate the feasibility and safety of dual-probe B-RFA technology for assisting in laparoscopic hepatic resections.

Methods

A retrospective review of a prospectively-maintained database identified 68 patients at a single centre who underwent laparoscopic M-RFA, laparoscopic B-RFA or B-RFA pre-coagulation before laparoscopic liver resection by a single surgeon between June 2004 and September 2010. Peri-operative and follow-up analysis was stratified by ablation type and procedure. Statistical comparison of demographic and peri-operative metrics was performed using the Mann–Whitney U-test for continuous data and Fisher's exact test for categorized data. A P-value of <0.05 was considered significant. Follow-up survival comparison was performed using Kaplan–Meier analysis. Statistical analysis was performed using GraphPad Prism® 5.0 (GraphPad Software, Inc., La Jolla, CA, USA).

Surgical technique for laparoscopic M-RFA, B-RFA and/or resection

All operations were performed under general anaesthesia with patients in either the supine or left lateral decubitus position. Pneumoperitoneum was induced to 15 mmHg and a 30 degree laparoscope was utilized in all procedures. The liver was fully mobilized laparoscopically off of the retroperitoneum and diaphragm to access the tumour-bearing segments. Tumours were identified and margins characterized using an intra-operative laparoscopic ultrasound probe (Aloka America, Wallingford, CT, USA). The Starburst™ (RITA Medical Systems Inc., Fremont, CA, USA) multiple needle array (2–7 cm) device was used for all tumours treated by M-RFA, with device-energy settings applied as directed by the manufacturer. Multiple overlying ablations were done to optimize tumour destruction.

All tumours treated by B-RFA were performed using the InCircle™ (RFA Medical, Inc., Fremont, CA, USA) parallel electrode system, which does not require grounding pads. Although this device is pre-packaged with the probes held in parallel by a 3-cm bracket, the majority of operations were done with the bracket completely removed allowing greater freedom of adjustment to the desired ablation zone. For the purpose of the present study, the 25-cm-long probes were used with each needle deploying six circular, planar-oriented electrodes spanning up to 3 cm (Fig. 2a). This device is designed to be used percutaneously and guided by either laparoscopy or axial imaging (Fig. 2b).

Figure 2.

The InCircle™ (RFA Medical, Inc., Fremont, CA, USA) parallel electrode system for bipolar radiofrequency ablation (B-RFA). Twenty-five-cm needle probes deploy six circular, planar-oriented electrodes spanning up to 3 cm (a). This device is designed to be used percutaneously, is guided by either laparoscopy or axial imaging (b) and does not require the use of grounding pads

An important consideration when using this device is adequate mobilization of the liver – especially the right side. Full mobilization of the liver allows precise and controlled protection of the diaphragm and viscera from thermal injuries by stray needle probes. The ability to manipulate the liver in various directions and planes is also important to provide broad access for needle probe placement. The technique adopted to maximize the zone of ablation consisted of at least three, 90-degree overlapping ablations with the needle probes deployed vertically (i.e. along the sides of tumour). The power setting for each ablation was 150 watts, with the target temperature set at 105°C. After the tumour ablations were completed along a horizontal plane moving laterally to medially across the liver, the needle probes were turned 90 degrees so that the tines were oriented horizontally (i.e. along the top and bottom of the tumour). Three more overlapping ablations were then completed in the same plane. The initial ablation generally took approximately 5 min to reach the target temperature (for 3-cm tumours). As the tissue/tumour desiccates, the current impedance rises quickly so that subsequent overlapping ablations require less time. Importantly, no RFA ablations were done in areas near the hilum of the liver to avoid vascular or biliary injures. The overall tumour ablation times necessary depended on tumour size, liver tissue type and mode of RF energy used.

All laparoscopic resections using a pre-coagulation technique were done with the B-RFA InCircle™ device (Fig. 3a). In selected operations, inflow control (Pringle manoeuver) was used depending on the magnitude of resection. In general, the techniques used for laparoscopic hepatic resections were similar to those used in open resections. After the tumour(s) had been fully evaluated by ultrasonography, the line of transection was marked on the liver surface using electrocautery. The InCircle™ needles were introduced in parallel approximately 1 to 1.5 cm apart and the circular tines deployed. A 2–3 cm zone of ablation was generated accounting for sufficient tumour margins while at the same time avoiding injury to critical intra-hepatic structures, viscera and/or the diaphragm. All parenchymal division was done using the Gyrus (Olympus Corp., Tokyo, Japan) bipolar cutting forceps (Fig. 3b).

Figure 3.

Parenchymal thermocoagulation prior to resection. The InCircle™ (RFA Medical, Inc., Fremont, CA, USA) bipolar radiofrequency ablation (RFA) probes were used to pre-coagulate liver tissue along the plane of resection (a). Parenchymal division was performed using the Gyrus cutting bipolar cutting forceps (Olympus Corp., Tokyo, Japan) (b)

Radiological follow-up protocol

For hepatocellular carcinoma (HCC), solid benign lesions and non-colorectal adenocarcinoma, patients were followed with triple-phase, 5-mm cut computed tomography (CT) scans at 3-month intervals post-operatively for the first 2 years, 6-month intervals for the following 2 years and then on a yearly basis. An incomplete ablation was recorded if the first post-RFA CT showed a persistent hypervascular area suspicious for residual disease. Local recurrence was defined by a hypervascular region located within the RFA zone not previously seen on surveillance CT scans. In several patients, perfusion artefacts were noted in the radiology reports that were indeterminate for tumour recurrence. For the purpose of the present study, all of these patients were considered to have local recurrence. The calculated local recurrence rate included patients with incomplete ablations, those with documented recurrence and those with indeterminate lesions.

Results

Comparison of M-RFA to B-RFA

M-RFA alone was used to definitively treat 30 patients, whereas B-RFA alone was used for 17 patients with contraindications to resection or as bridge therapy towards liver transplantation. Demographics and peri-operative factors are shown in Table 1. Major complications using B-RFA included two instances of a pneumothorax and one patient with hepatic decompensation, whereas M-RFA included RFA skin burns in two patients and hepatic decompensation in one; there were no deaths in either group. There were no differences in local or overall recurrence, time to recurrence, and overall and disease-free survival between the two groups (Fig. 4).

Table 1.

Comparison of demographic and peri-operative outcomes between monopolar- (M-RFA) and bipolar-radiofrequency ablation (B-RFA)

	B-RFA	M-RFA	P-value
Patients	17	30
Demographics
Age, years median (range)	63 (55–81)	57 (47–82)	0.00192
Gender (M : F)	15 : 2	30 : 0
Disease	Malignant (HCC) – 15	Malignant (HCC) – 30
	Benign (adenoma) – 2
MELD median (range)	8 (5–13)	8 (4–13)	0.402
Tumour size, cm median (range)	3 (1.2–5.0)	3 (1.8–6.8)	0.541
Peri-operative outcomes
Laparoscopic	16/17	25/30	0.390
Cholecystectomy	8/17	5/30	0.0409
Time, min median (range)	275 (152–436)	175 (85–550)	<0.001
Blood loss, mL median (range)	100 (10–500)	75 (5–300)	0.152
Hospitalization, days median (range)	7 (3–18)	5 (2–9)	0.0150
Complications32	Grade 2 = 1 (PNA)	Grade 2 = 2 (arrhythmia)
	Grade 3 = 2 (PTX)	Grade 3 = 2 (RFA burns)
	Grade 4 = 1 (hepatic insufficiency)	Grade 4 = 1 (GI bleed)
Major complications	3/17	3/30	0.702

MELD, model for end-stage liver disease; HCC, hepatocellular carcinoma; PNA, pneumonia; PTX, pneumothorax; GI, gastrointestinal.

Figure 4.

Overall- and disease-free survival over a median follow-up of 20.6 months for bipolar-radiofrequency ablation (B-RFA) (range, 3.0–39.6 months) and 31.5 months for monopolar-radiofrequency ablation (M-RFA) (range, 1.4–75.9 months). There was no difference between B-RFA and M-RFA with respect to rates of local recurrence (3/17 vs. 8/30, P = 0.733) or overall recurrence (7/17 vs. 14/30, P = 0.982). Median time to recurrence was also similar between B-RFA and M-RFA (12.6 months, range 3.4–25.1 vs. 12.4 months, range 3.0–44.7; P = 0.933)

Laparoscopic liver resection with B-RFA pre-coagulation

Seventeen patients underwent laparoscopic liver resections using B-RFA energy for coagulation before parenchymal transection (Table 2), including anatomic segmental resections in 11 patients and multi-segmental resection or hemihepatectomy in six patients. In four patients with multiple tumours, B-RFA energy was used in combination to ablate some tumours and resect others. Major complications occurred in three patients: two suffered a clinically significant pulmonary embolism and one death occurred from liver failure as a result of portal vein thrombosis 1 week after resection.

Table 2.

Demographic and peri-operative outcomes of laparoscopic liver resection utilizing bipolar-radiofrequency ablation (B-RFA) to pre-coagulate liver parenchyma before transection

Demographics	No. of patients	17
	Age, years median (range)	60 (42–91)
	Gender (M : F)	14 : 3
	Disease	Malignant = 11
		Benign = 2
		Metastatic = 4
	MELD median (range)	8 (6–15)
	Tumour size, cm median (range)	5 (0.5–10.5)
Peri-operative Outcomes	Operation	Segmentectomy – 11
		Multi-segmental – 6
	Time, min median (range)	434 (131–830)
	EBL, mL median (range)	300 (20–3000)
	Hospitalization, days median (range)	7 (3–35)
	Complications32	Grade 1 = 1 (ileus)
		Grade 2 = 1 (ascites)
		Grade 3 = 2 (hepatic insufficiency, biloma)
		Grade 4 = 2 (PE)
		Grade 5 = 1 (death)

MELD, model for end-stage liver disease; EBL, estimated blood loss.

Discussion

Given that a majority of patients with HCC are not candidates for or unable to tolerate a surgical resection, non-surgical and minimally-invasive approaches are attractive options for treatment or bridge to transplantation.^20,21 In the United States, RFA is among the most frequently used modality, with thousands of operations done yearly. Although M-RFA technology has significant limitations, it has been extensively described and is historically the most widely used thermal ablative method. In an effort to overcome the limitations of M-RFA, other thermal energy systems have been introduced for destruction of liver tumours. B-RFA energy systems have the advantage of delivering a stronger current density that reduces the need for thermal conduction resulting in more rapid ablation. Animal studies comparing M-RFA to B-RFA using the dual needle probe system have shown superiority for the B-RFA method with respect to larger ablation zone, tissue control, haemostasis and overall efficiency.²² Clinical studies are also beginning to emerge confirming improved tumour ablation times and efficacy using bipolar radiofrequency energy in comparison with M-RFA.¹²

In the present study, the initial experience with the bipolar probes demonstrates the efficacy of this device and equivalency to M-RFA with regard to peri-operative and survival outcomes. Moreover, B-RFA avoids the complication of skin burns as it does not require grounding pads. In particular, the InCircle™ device is unique in that it can also be used to assist in hepatic resections for parenchymal tissue pre-coagulation. This series demonstrates certain benefits with the laparoscopic, bipolar approach for tumour ablation. For all patients, it was possible to perform the entirety of the ablation laparoscopically without the need for conversion to laparotomy. While this point may seem trivial, it is exceedingly important in patients with compromised hepatic reserve (as in the current patient population described) prone to wound complications.

As is the case with any new technology, there is an important learning curve associated with using the InCircle™ B-RFA device. The surgeon is required to conceptually change the traditional M-RFA approach to target the tumour. While mastering laparoscopic ultrasonography is a prerequisite for successful tumour ablation regardless of the energy source, the B-RFA method demands specific thee-dimensional (3-D) spatial planning before energy delivery. As mentioned, this concept is called ‘line-of-sight’ deployment of the dual electrodes. Planning and developing a 3-D mental image of the desired ablation zone is necessary because once the energy is delivered, the targeted tumour is poorly visualized under ultrasonography.

The lack of real-time tumour destruction assessment is one important limitation of using RFA energy. However, in comparison with monopolar techniques, B-RFA tends to be more controlled and faster for similarly-sized tumours. Although reported M-RFA ablation times vary significantly, early reports document ablation times using B-RFA which are magnitudes faster than M-RFA. Yi et al. recently reported 26 consecutive B-RFA operations with an average ablation time of approximately 6 min, compared with greater than 20 min using M-RFA.^12,23 It was also the authors' experience that B-RFA was significantly faster than M-RFA and required shorter ablation times. Ablation speed was decreased because the systematic overlapping ablations lead to rapid tumour/ parenchymal desiccation and contraction thereby increasing current impedance.

Nevertheless, it must be mentioned that in the current series, B-RFA had significantly longer operative times than M-RFA. This is partly as a result of the learning curve associated with implementation of a novel device both for the surgeon as well as operating room ancillary staff. More importantly, B-RFA as a treatment modality was generally reserved for tumours which were more difficult to access and therefore required extensive laparoscopic liver mobilization for proper probe positioning. Additionally, as a result of tumour location, a significant number of B-RFA patients underwent a concomitant cholecystectomy, thereby also increasing operative time.

Laparoscopic treatment of liver tumours is becoming increasingly utilized in most major medical centres. Laparoscopy provides improved visibility with a 2.5× magnification along with several other well-established benefits such as decreased post-operative pain, shortened hospital stay and earlier return to work.²⁴ Within the current series, adequate laparoscopic liver mobilization was achieved in all patients. In addition, all tumours were successfully treated using the 25 –cm-long dual B-RFA probes regardless of location or patient body habitus. The less invasive nature of laparoscopy allowed the treatment of older patients with significant medical problems, patients with marginal hepatic reserve and those medically unfit for hepatic resection.

Within the current series of patients who underwent ablation alone, only one patient experienced hepatic insufficiency post-operatively, even although half had double-digit model for end-stage liver disease (MELD) scores. Complications in this group resulted from the need to place transdiaphragmatic, intercostal ports for improved visualization of the hepatic veins. Two of these patients underwent intra-operative thoracostomy tube placement as prophylaxis against a clinically significant post-operative pneumothorax. These tubes were removed post-operatively without complication, and these patients as well as all others recovered uneventfully from laparoscopic B-RFA. The presence of thoracostomy tubes extended length of stay, which was significantly longer than M-RFA. In part, the longer hospitalization documented in this study was also related to the more recent referrals of out-of-state veterans which occurred as the authors' medical centre transitioned into a tertiary care facility for patients with liver tumours. It is the institutional philosophy to ensure patients are completely recovered before discharge, and often they are observed as an inpatient for an extra day or two. In fact, many patients undergo ‘pre-emptive’ CT scanning to make sure they will not need further interventions. In a veteran population, it has been shown previously that this type of referral practice pattern can prolong and complicate discharge.²⁵

Long-term follow-up of patients who underwent M-RFA and B-RFA shows equivalency with respect to outcomes. Although populations of each cohort are relatively small and follow-up less than 3 years, there is still no difference between overall- and disease-free survival (Fig. 4). Moreover, local and overall recurrence rates were likewise similar between the two modalities, as was overall time to recurrence. Although this series had high local recurrence rates, they are on par with local recurrence rates of 15%–28% reported in other small series with patients who underwent M-RFA for unresectable HCC.^26–29

The observed benefit of pre-coagulation – the ability to achieve near-complete haemostasis at the planned resection plane – and the larger zone of coagulation with B-RFA adds important benefits towards increasing the tumour-free margin. The synergistic combination of pre-coagulation and laparoscopy gives the surgeon the ability to complete a hepatic resection completely laparoscopically with a decreased transfusion requirement and conversion rate to laparotomy. Only two patients of the current series required intra-operative blood transfusions. For patients with impaired hepatic and functional reserve, the ability to minimize the transfusion requirement and operative time has significant benefits with respect to post-operative morbidity and long-term survival.³⁰

Although still controversial, hepatectomy in combination with RFA in patients with multiple tumours is a viable option in highly-selected patients.³¹ Given that the majority of our patients were treated for HCC, many presented with multifocal disease. Four of these patients underwent concomitant B-RFA for tumour ablation as well as precoagulation of liver tissue before laparoscopic resection. Although small in number, these operations were no more technically difficult than either procedure alone. Moreover, patients tolerated the concomitant procedures with no greater morbidity or mortality, indicating another application of B-RFA.

There are some important cautions that need to be stated regarding probe deployment using the InCircle™ B-RFA device for both tumour ablation and pre-coagulation. The design of the parallel circular electrodes (Fig. 2) uses very thin ‘wiry’ tines that can break and remain within the liver tissue. Moreover, if the needles are positioned too closely and the tines make contact, energy conduction will stop resulting in poor tissue coagulation and bleeding. Rarely, the electrodes can arc and weld to the needle exit side holes causing significant tissue damage and bleeding when removed.

It should be emphasized that utilization of the InCircle™ B-RFA device is best suited for those with proper knowledge of hepatic anatomy. After the initial ultrasound-guided needle placement, subsequent probe placement is done blindly within liver parenchyma, increasing the risk of inadvertent injury to vascular or biliary structures. Moreover, the lack of multiple terminal-probe temperature sensors (as present in M-RFA) prevents real-time assessment of improper probe placement. It is important not only to consider the complete destruction of tissue between the bipolar probes, but also the degree of lateral spread of the energy (heat) that can cause delayed or unrecognized complications (Fig. 5). For these reasons, B-RFA is ideally suited for experienced hepatobiliary surgeons with thorough knowledge of internal liver anatomy.

Figure 5.

Zone of ablation utilizing bipolar-radiofrequency ablation (B-RFA). Contrast-enhanced computed tomography (CT) scan of a 61-year-old patient with a 4.1-cm hepatocellular carcinoma (HCC) shown before ablation (a, arrow) and 3 months after the procedure (b). A 71-year-old patient with multifocal HCC (c, arrows) with dual B-RFA ablation zones (d), shown 3 months after the procedure. The zone of ablation with B-RFA encompasses large portions of liver parenchyma, necessitating expert knowledge of liver anatomy to avoid vital structures

Nevertheless, once the surgeon gains experience using this device, it can be safely used for both tumour ablation and/or for precoagulation during hepatic resection. The ability to achieve greater efficacy with the InCircle™ B-RFA energy delivery system by laparoscopy, especially in the context of laparoscopic hepatic resection, preserves the option of liver transplant and is an important bridge to future therapies. Therefore, the promise of treating patients with more advanced local disease and those with compromised hepatic reserve affords B- RFA an important adjunctive role to hepatic surgery.

Conflicts of interest

None declared.