Radiofrequency ablation for liver tumors

Published on 10/04/2015 by admin

Filed under Surgery

Last modified 22/04/2025

Print this page

This article have been viewed 2321 times

Chapter 85C Radiofrequency ablation for liver tumors

Anton J. Bilchik, Robert M. Krasny, David Allegra

Overview

Hepatic resection and transplantation are the only curative options for patients with liver metastases or primary liver tumors (see Chapter 80, Chapter 81A, Chapter 81B, Chapter 81C, Chapter 97A, Chapter 97B, Chapter 97C, Chapter 97D, Chapter 97E ). Unfortunately, resection is possible in only about 20% of patients; most hepatic malignancies are surgically inaccessible or are associated with a large tumor burden or inadequate hepatic reserve. Liver transplantation for hepatocellular carcinoma (HCC) is limited by a shortage of donor organs (see Chapter 97D). Patients with unresectable disease may be candidates for systemic therapy (see Chapter 88), local ablative techniques (percutaneous ethanol injection, microwave tumor coagulation, interstitial laser photocoagulation, cryosurgical ablation, or radiofrequency ablation [RFA]; see Chapter 85A, Chapter 85B, Chapter 85D ), or hepatic-directed therapy (hepatic artery ligation, chemoembolization, hepatic artery perfusion; see Chapters 83, 86, and 89).

Over the past decade, multiple clinical trials have evaluated RFA for the treatment of primary liver tumors and liver metastases. Animal experiments in the early 1990s determined the size and shape of ablations that could be achieved in the liver (Rossi et al, 1990). Rossi and colleagues (1996) then demonstrated the effectiveness of RFA for the treatment of HCC in humans. The Food and Drug Administration (FDA) approved RFA for generic tissue ablation in 1996 and for ablation of unresectable hepatic metastases in 2000. Since then, randomized trials have been performed in early HCC that demonstrate equivalent survival to resection (Lu et al, 2006), but prospective data in colorectal liver metastases are limited.

RFA destroys tumor by generating heat within a lesion. During RFA, a high-frequency alternating current changes the direction of ions around an alternating electrode charge. This creates frictional energy and heat conduction. As tissue temperature increases above 45° C, loss of cellular structure and protein denaturing result in tumor cell death. RFA can be performed in the operating room via celiotomy or laparoscopy, or it may be done in the radiology suite by a percutaneous approach. It can be used with other modalities of liver-directed therapy, such as resection and hepatic artery perfusion, and it can be used in conjunction with systemic therapy for other sites of metastatic disease. RFA continues to evolve with the introduction of new technology that increases the field of ablation and simplifies the technique. Long-term survival data have been reported in both HCC and colorectal hepatic metastases (Table 85C.1). A recently published position statement by the Society of Interventional Radiology (Gervais et al, 2009) describes four categories of patients that are preferred candidates for RFA: those with 1) inadequate liver function, 2) comorbid conditions, 3) anatomic distribution, and 4) for local tumor control as a bridge to transplantation.

Table 85C.1 Survival Data for Radiofrequency Ablation of Various Tumor Types

Radiofrequency Ablation Technology

RFA destroys tissue by generating enough heat to denature cell protein and cause cell death (Fig. 85C.1), which can occur instantaneously at 60° C, a temperature achieved by all RFA devices. During RFA an electrical generator produces a current that alternates directions rapidly. Charged molecules in the tissue follow the direction of the changing current: this results in friction, which produces heat. The current around the ablation electrode creates a relatively uniform zone of conductive heat that radiates out from the electrode. If tissue impedance is low, an expanding spherical zone of ablated tissue is created. As the temperature rises, tissue charring and desiccation occur. Also, nitrogen gas forms that can be seen on ultrasound (US) as air bubbles in the tissue. Both of these events raise tissue impedance and inhibit heat conduction, thereby inhibiting further tissue ablation. The spherical size of the ablated tissue is proportional to the square of the RF current (RF power density). It is also dependent on the size of the electrode and the duration of the applied energy. The RF power density decreases in proportion to the square of the distance from the electrode, resulting in a rapid decrease in tissue temperature with increasing distance from the electrode (Strasberg & Linehan, 2003). Changing technology strives to improve ablation zones by limiting impedance and increasing the RF power density.

FIGURE 85C.1 Schematic of radiofrequency ablation.

Early RFA electrodes were unipolar. Because they produced a 1.5- to 2.0-cm cylindrical zone of ablated tissue, only tumors less than 2 cm in diameter could be ablated. These electrodes were eventually replaced by single electrodes that deployed multiarray tines (multiarray electrodes) to create a larger, more spherical zone of ablation. The current multiarray electrodes described below can ablate fields of 4 to 5 cm in diameter. Technologic developments to reduce tissue impedance and increase the zone of ablation will allow progressively larger tumors to be treated with single ablations.

Three different RF electrodes and generators are commercially available. The Radiofrequency Interstitial Tissue Ablation (RITA; AngioDynamics, Latham, NY) and RadioTherapeutics (Boston Scientific, Natick, MA) systems use multiarray electrodes (Fig. 85C.2). The RITA system has a single electrode through which multiple tines are deployed. The ablation is performed at sequential steps of electrode deployment. Thermoprobes measure temperature in the ablation zone, and an automated adjustment is then made to maintain a steady temperature during the ablation. The electrode/generator system measures temperature and time as end points of ablation. This system was recently modified by the addition of saline perfusion to reduce impedance and increase the diameter of the ablation field to 7 cm. The RadioTherapeutics system automatically adjusts current according to tissue impedance; temperature is not measured. The multiarray umbrella-shaped electrode can ablate fields of about 4 cm. The Cool Tip Cluster Electrode, a single or three-pronged electrode, is manufactured by Covidien (Mansfield, MA); each electrode is internally perfused with cold saline, the current is sent in pulses, and the system measures impedance and temperature. Ablation zones can reach 5 cm in diameter with the three-pronged device.

FIGURE 85C.2 RITA (left) and Radionics (right) probes used for radiofrequency ablation.

Technical Aspects of Radiofrequency Ablation

Patients who come to medical attention with liver tumors should be evaluated for curative resection or transplantation for HCC. Those whose performance status or disease location or distribution prohibits resection may be candidates for RFA. Prior to RFA, all patients should undergo computed tomography (CT), magnetic resonance imaging (MRI), and/or fluorodeoxyglucose (FDG) positron-emission tomography (PET). If these whole-body imaging techniques reveal extrahepatic sites of metastatic disease, systemic therapy may be a preferred option instead of resection or RFA. Patients with physiologically active neuroendocrine tumors are an exception in that debulking may reduce debilitating symptoms. Baseline hepatic function should be assessed through laboratory indicators of synthetic function and Child-Turcotte-Pugh (CTP) classification of liver dysfunction. A thorough history and physical examination are essential to determine patient performance and prior treatment history.

If a surgical approach is undertaken, laparoscopy can reveal extrahepatic disease, and intraoperative US can identify intrahepatic disease not detected by preoperative imaging. A 30-degree angled laparoscope can be used to examine all parietal and visceral peritoneal surfaces, the lesser sac, omentum, and viscera. If extrahepatic disease is found, RFA should not be done, and systemic treatment should be considered. A flexible 7.5-MHz ultrasonic probe is used to evaluate the liver and determine the proximity of liver lesions to major vascular and biliary structures. The exact location of hepatic metastases will affect the technical approach for RFA.

If extrahepatic and/or extensive intrahepatic disease has been detected, ablation is unlikely to be of any benefit. US during surgical or percutaneous RFA or CT during percutaneous RFA is used to guide the probe into the lesion, and the ablation is monitored by real-time US imaging of an expanding hyperechogenic zone (Fig. 85C.3). Depending on the maximal ablation size achievable by the probe, multiple overlapping ablations may be needed to destroy a tumor and produce a surrounding rim of necrosis (Fig. 85C.4). This can be technically challenging. US cannot reliably distinguish between ablated and normal tissue, so ablation should begin at the most posterior portion of the tumor. The probe then can be withdrawn in approximately 2-cm increments to create sequential overlapping zones of ablated tissue. If target temperatures are not reached, as may be the case for lesions near major vascular structures because of a heat-sink effect, the tines are withdrawn slightly or rotated approximately 45 degrees to increase the temperature in the region of ablation. Each ablation should create a 1-cm margin of treated normal parenchyma to ensure complete tumor destruction and reduce the risk of local recurrence. After ablation, as the RFA needle is withdrawn, the probe tract is cauterized to prevent hemorrhage and tumor seeding of the needle tract.

FIGURE 85C.3 Ultrasound-guided radiofrequency ablation.

FIGURE 85C.4 Computed tomographic scan demonstrating large-volume ablation achieved with multiple overlapping ablation technique.

RFA can be performed by celiotomy (open), laparoscopy, or a percutaneous approach. The choice of technique depends on the patient’s condition, the number and location of liver tumors, and the skill of the physician performing the ablation. Each approach offers certain advantages and disadvantages that must be weighed to determine the best approach for the individual patient.

Percutaneous Radiofrequency Ablation

A skilled interventional radiologist typically performs percutaneous RFA as an inpatient or outpatient procedure under US, CT, or combined guidance. The patient is sedated and/or given a local anesthetic. CT guidance provides several advantages: it allows precise placement of the electrode despite the generation of gas within the ablation, and it is possible to assess the adequacy of the ablation during the procedure by performing a contrast-enhanced scan. If areas of inadequate ablation are identified, additional targeted ablations can be performed during the initial procedure.

The percutaneous approach is somewhat limited, as some liver lesions can only be accessed by a surgical approach. Tumors near the periphery of the liver may be problematic because of the risk of thermal injury to other visceral structures such as the small bowel, stomach, or transverse colon; with an open technique, these structures can be retracted away from the liver. Other lesions can be approached percutaneously but raise technical challenges. For example, ablation of lesions at the dome of the liver might result in thermal injury to the diaphragm. Introduction of an air pocket can be used to separate the diaphragm and liver to avoid diaphragmatic thermal injury (Raman et al, 2004); this requires skill and may result in complications such as diaphragmatic injury or pneumothorax. Injection of saline or use of a balloon catheter has also been reported to separate structures from the liver to avoid thermal injury; however, clinically relevant diaphragmatic injuries are rare in our experience (Fig. 85C.5). Because a percutaneous approach is less invasive than a surgical approach, it may also be preferable in patients with comorbid conditions or poor medical status.

FIGURE 85C.5 Percutaneous ablation performed adjacent to the diaphragm.

Open or Laparoscopic Radiofrequency Ablation

RFA via celiotomy or laparoscopy is performed in the operating room by a surgeon, while the patient is under general anesthesia. Unlike the percutaneous approach, a surgical approach allows visual inspection of the intraabdominal cavity for extrahepatic disease. Intraoperative US is used to evaluate the liver for intrahepatic lesions not seen on preoperative imaging. In a study of 308 patients undergoing laparoscopic RFA with intraoperative hepatic US, preoperative imaging failed to identify extrahepatic disease in 12% of patients and additional hepatic lesions in 33% of patients (Bilchik et al, 2000). Similarly, Siperstein and colleagues (2000) found that 30% of patients who underwent laparoscopic US of the liver prior to RFA had additional metastatic nodules not apparent on the preoperative CT scan. Larger liver lesions (>5 cm) are more easily treated with an operative approach than percutaneously. These lesions may require multiple overlapping ablations. Intraoperative occlusion of the portal triad can increase the ablation zone by decreasing the heat-sink effect created by nearby vessels. Furthermore, a surgical approach is sometimes necessary when other treatments are used in combination with RFA. For example, a surgical approach is ideal for patients who have bilobar liver metastases, who may benefit from concomitant liver resection. Larger liver lesions may be resected, and smaller lesions in the opposite lobe may be treated by RFA.

Laparoscopy is less invasive than celiotomy, and patients often require slightly shorter hospital stays. In some cases, a celiotomy is needed because of decreased access to the liver and limited position of the US and RFA probe under laparoscopy; However, in the presence of multiple adhesions in the abdomen from prior operations, or if the tumor cannot be approached safely, laparoscopy may not be possible. In this case, a celiotomy is warranted.

Monitoring and Follow-up

After laparoscopic or percutaneous RFA, patients are usually admitted to the hospital for 23 hours or are discharged home the same day; after open RFA (celiotomy), patients are usually hospitalized for 3 to 5 days. In all cases, red blood cell and leukocyte counts, liver function tests, and temperature are closely monitored. RFA typically causes a mild transient elevation of liver function tests, core temperature, and leukocyte count. These should normalize within about 3 days. Continued elevation, especially 5 or more days after RFA, may be an indicator of a potential complication and should prompt further clinical investigation. Tumor markers such as carcinoembryonic antigen (CEA) for colorectal cancer metastasis and α-fetoprotein (AFP) for HCC can also be used to monitor patients as appropriate.

Contrast-enhanced CT scans obtained within a month and then at 3-month intervals after RFA should be compared with preprocedure CT scans. Liver necrosis from RFA can be seen as a nonenhancing hypodense region. When complete necrosis of the tumor is achieved, the resulting necrotic zone exceeds the original tumor size. Postprocedure CT scans are important to determine the success of RFA and to monitor the liver for evidence of recurrence. PET scans may also be particularly sensitive in the assessment of the adequacy of ablation and presence of marginal recurrence (Fig. 85C.6) or to distinguish viable tumor from areas of previous surgical or ablative therapy (Fig. 85C.7).

FIGURE 85C.6 A, Marginal recurrence not clearly visualized by computed tomography. B, Recurrence is clearly identified by positron emission tomographic scan.

FIGURE 85C.7 Positron emission tomography scan demonstrating areas of viable tumor distinguished from areas of successful ablation.

Current Practice

RFA has mainly been used for unresectable HCC or metastatic disease confined to the liver (see Chapters 80, 81A, and 81B). Tumors may be deemed unresectable based on size, number, location, or doubling time. Therefore, selection of patients is based initially on a thorough preoperative workup that includes imaging to evaluate for other sites of disease, prior response to systemic therapy, and the biology of the disease. If systemic therapy eliminates extrahepatic metastases, and if its effects prove durable over several months, RFA of hepatic metastases may be considered; however, this applies to a small subset of unique patients. Some patients will have resectable disease but limited hepatic reserve. For example, a patient with cirrhosis and HCC might not have enough hepatic reserve for tumor resection but would be able to tolerate RFA. In other patients, resection of larger lesions and RFA of smaller lesions can completely eradicate tumor while maintaining hepatic reserve. Finally, patients who have multiple comorbid factors may not be good candidates for an operation and may be best treated by a less invasive approach such as percutaneous RFA.

Radiofrequency Ablation of Colorectal Hepatic Metastases (See Chapter 81A)

Resection offers the only curative option in patients with colorectal hepatic metastases, with 5-year survival approaching 60% (Ahmad et al, 2007; Pawlik et al, 2008; Tomlinson et al, 2007). The improvement in survival is attributed to more effective systemic therapy and expanded criteria for defining resectability. Despite these, many patients are poor candidates for resection, and RFA provides a locoregional alternative. In a recent review of 21 studies (Stang et al, 2009), the median progression-free survival after RFA ranged between 6 and 13 months; the median and 5-year survival ranged between 24 and 59 months and 18% to 40%, respectively.

No randomized trials have compared RFA to resection in patients with colorectal hepatic metastases. One retrospective study of 358 patients compared resection, RFA plus resection, RFA alone, and laparotomy with biopsy (Abdalla et al, 2004); RFA was used for cure when complete resection was not possible. All patient-related and tumor-related factors known to influence outcome were similar among the groups. The rate of recurrence was 84% after RFA alone, 63% after RFA plus resection, and 52% after resection alone. Local recurrence in the area treated was more common after RFA plus resection (9%) than after RFA alone (5%) or resection alone (2%). The 3-year rate of overall survival was 73% after resection, 43% after RFA plus resection, and 37% after RFA alone. Patients who underwent RFA had a survival advantage over patients who underwent biopsy with or without chemotherapy. Four recent nonrandomized retrospective studies (Aloia et al, 2006; Hur et al, 2009; Park et al, 2008; White et al, 2007) demonstrated a survival benefit in patients undergoing resection versus RFA; the 5-year survival was between 48% and 71% for resection and between 19% and 27% for RFA (Table 85C.2). The survival analyses in these studies are limited by differences in patient selection, follow-up time, tumor type, and RFA approach, which make them difficult to compare.

Table 85C.2 Radiofrequency Ablation vs. Resection for Hepatic Colorectal Metastases (Nonrandomized)

Radiofrequency Ablation of Hepatocellular Carcinoma (See Chapter 80)

Locoregional methods for ablation of HCC include percutaneous ethanol ablation (PEI), chemoembolization, cryosurgery, and RFA, although RFA offers many advantages over PEI and cryoablation. In general, PEI requires multiple treatments and needle insertions to treat a single lesion, whereas RFA can treat the entire tumor with one or two probe insertions. Two randomized trials (Lencioni et al, 2003; Lin et al, 2005) have demonstrated superiority over PEI. The 2-year disease-free survival (DFS) was 96% in patients treated with RFA compared with 62% in patients treated with PEI (Lencioni et al, 2003). Similarly, Lin and colleagues (2005) demonstrated that in patients with HCC less than 3 cm in size, the 3-year DFS was superior in those treated with RFA. When compared with cryoablation, RFA is associated with less blood loss, less thrombocytopenia, and a shorter hospital stay (Bilchik et al, 2000). Cryoablation also carries the risk of a systemic inflammatory response, which can result in renal insufficiency, coagulopathy, hypotension, and death. This type of systemic response has not been reported with RFA.

Clinical trials of RFA for HCC have shown promising results (see Table 85C.1). Livraghi and colleagues (2008) demonstrated a local recurrence rate of only 2.8% at a median follow-up of 31 months for tumors smaller than 2 cm. Poon and colleagues (2004a) achieved a complete response rate of 91% for RFA of large (>3 cm) HCC and rates of local recurrence, distant intrahepatic recurrence, and extrahepatic metastasis were independent of tumor size.

Radiofrequency Ablation Versus Resection for Hepatocellular Carcinoma

Only two prospective randomized controlled trials have been done of RFA versus resection for the treatment of primary liver carcinoma (see Chapter 80). Chen and colleagues (2006) randomized 180 patients with a solitary HCC smaller than 5 cm to either resection or percutaneous RFA. The 4-year overall survival was 68% in the RFA group and 64% in the resection group. Local recurrence rates were also similar, with less morbidity in the RFA group. Lu and colleagues (2006) randomized 105 patients with either a solitary HCC tumor smaller than 5 cm or fewer than three tumors less than 3 cm in size to either a partial hepatectomy or percutaneous RFA and found that 3-year survival rates were similar in both groups (Table 85C.3), but hospital stay and blood transfusions were significantly reduced in the RFA group.

Table 85C.3 Radiofrequency Ablation vs. Resection for Hepatocellular Carcinoma

Conflicting result have been reported in several nonrandomized studies (Hong et al, 2005; Lupo et al, 2007; Montorsi et al, 2005) that demonstrate therapeutic equivalence; others report a higher local recurrence rate and reduced survival in the RFA group (Vivarelli et al, 2004). This study was performed at two different centers; RFA was performed at one center and resection at the other. Overall and disease-free survivals were significantly better after resection; however, patients treated with RFA had more liver lesions and a higher CTP classification, both of which may have influenced outcome.

Guglielmi and colleagues (2008) also demonstrated that in HCC larger than 3 cm, significant improvement in survival was seen in patients undergoing surgical resection, but this did not translate to tumors smaller than 3 cm, nor did it apply to patients with multiple tumors.

Radiofrequency Ablation for Recurrent Hepatocellular Carcinoma

After resection or ablative therapies, most patients with HCC will succumb to progressive or recurrent liver disease. Hepatic resection is rarely feasible because of inadequate hepatic reserve, making RFA an effective therapeutic alternative. Taura and colleagues (2006) compared 610 patients who underwent a liver resection before and after 1990. The improved 5-year survival in the latter group (22% vs. 12%) was attributed to percutaneous RFA for intrahepatic recurrence. Similarly, Yang and colleagues (2006) demonstrated a survival benefit in 41 patients with recurrent HCC treated with percutaneous RFA. These studies suggest that RFA is effective in the management of recurrent HCC after resection, and RFA may be the treatment of choice for carefully selected patients with recurrent hepatic disease (Fig. 85C.8).

FIGURE 85C.8 A, Computed tomography scan of recurrent hepatocellular carcinoma after caudate lobe resection adjacent to the inferior vena cava. B, Radiofrequency ablation probe placement. Postablation follow-up computed tomography scan demonstrates hypodensity at the ablation site consistent with successful ablation.

Radiofrequency Ablation as a Bridge to Transplantation (See Chapter 97D)

Many patients with HCC will succumb to liver failure because of a shortage of donors or will be removed from the transplant waiting list because of tumor progression; indeed, the dropout rate within a year has been reported as high as 38% (Yao et al, 2004). RFA has effectively been used as a bridge to transplantation. At a mean follow-up of 12.7 months, Lu and colleagues (2005) reported a dropout rate of only 5.8 months and 14% at a follow-up of 11.9 months (Brillet et al, 2006).

These encouraging data suggest that RFA is one modality that may be considered for prolonged waiting times. The impact on overall survival is less clear; however, one prospective study of patients with HCC awaiting transplantation undergoing RFA showed no survival benefit compared with the observation group (Porrett et al, 2006).

Tumor Response and Recurrence

CT or MRI is typically used to evaluate response rates (see Chapters 16 and 17); US does not accurately predict the extent of necrosis following RFA and should not be used to determine response. CT with contrast enhancement reveals a hypovascular ablated field with a rim of hypervascular inflammatory tissue; the rim enhancement area should disappear over several months time as inflammation resolves, and the treatment area should encompass the tumor and at least a 1-cm rim of normal tissue. Comparing preoperative imaging to postoperative imaging is important to determine that the tumor was adequately treated. Patients should receive serial CT scans after RFA to identify recurrences, which appear as irregular, nonenhancing areas on contrast CT.

Complete and Incomplete Response to Radiofrequency Ablation

Response rates vary from 48% to 98% across studies, in part because of differences in the size of treated lesions. Livraghi and colleagues (1999, 2000) evaluated percutaneous RFA in patients with HCC. The rate of complete response was 48% for tumors larger than 3 cm (2000), but it jumped to 90% for smaller tumors (1999), and the success rate was lower for infiltrating versus noninfiltrating HCCs. Similarly, Lencioni and colleagues (1998) found a statistically significant difference in response rates to percutaneous RFA for metastatic liver lesions larger than 3 cm (53%) versus those smaller than 3 cm (87%). Larger lesions typically require overlapping ablations, and areas of treated tumor may obscure US imaging of untreated tumor and result in higher treatment failures and recurrences.

The approach to RFA will also influence response rates (see Table 85C.1). Curley and colleagues (2000) reported that 6 of 76 patients who underwent percutaneous RFA had an incomplete response, whereas all 34 patients who underwent laparoscopic or open RFA had a complete response. This difference might be explained by better hepatic imaging with intraoperative US during laparoscopy or celiotomy than with transabdominal US during percutaneous RFA. Another possibility is that intraoperative probe positioning may be more accurate than percutaneous placement of the probe.

Tumor location adjacent to large vessels also influences response rates. Blood flow from large vessels will create a heat-sink effect that cools surrounding tissue and increases the temperature necessary for complete ablation. Large vessels are resistant to high temperatures that can damage surrounding tissue. Lu and colleagues (2002) found that porcine vessels smaller than 3 mm thrombosed or necrosed during RFA, but those larger than 5 mm were not affected. In another porcine model, vascular occlusion during RFA increased the volume of necrosis and prolonged the exposure of tissue to high temperatures (Chinn et al, 2001). De Baere and colleagues (2002) evaluated vascular inflow occlusion in patients undergoing RFA of tumors that were large (>35 mm) or near large blood vessels and found that balloon inflow occlusion significantly increased the area of ablation. These studies support the use of a Pringle maneuver to more effectively ablate tumors near major blood vessels and improve response rates. The use of RFA combined with hepatic artery embolization or chemoembolization is discussed in Chapter 83.

Rates of Recurrence at the Ablation Site

Recurrence rates are difficult to interpret across studies (Table 85C.4). Some authors report recurrences after complete ablation as determined by early postoperative CT or MRI, whereas others do not confirm complete response by postoperative imaging and report recurrences based on follow-up imaging. Investigators may report recurrence rates within the whole liver, rather than recurrence rates at the site of ablation, and the length of follow-up also impacts the recurrence rate. Bowles and colleagues (2001) evaluated 76 patients undergoing RFA of 328 tumors. Sixteen patients underwent repeated ablation for recurrences or new lesions, and 30 recurrences were reported at the site of a prior ablation. Patients with large tumors, tumor vascular invasion, and hepatic dysfunction had a statistically higher recurrence rate. Solbiati et al (2001) evaluated 117 patients undergoing percutaneous RFA of colorectal metastases; time to local recurrence and frequency of recurrences were influenced by lesion size. Similarly, we found that tumor size significantly influenced local recurrence of metastatic disease, independent of RFA technique (Wood et al, 2000; Bleicher et al, 2003). No study has shown that differences in technique influence the rate of recurrence after a complete response.

Table 85C.4 Local Recurrence After Radiofrequency Ablation of Hepatic Tumors

Complications of Radiofrequency Ablation

Reported morbidity and mortality rates associated with RFA (Table 85C.5) can be difficult to interpret, in part because technical approaches vary. Some investigators combine RFA with other treatments, such as liver resection, but addition of a second procedure may inflate the complication rate. Ablation of multiple tumors also increases the risk of complications such as bleeding or bile leak. Early studies often used multiple sequential RFAs for treatment of a single hepatic tumor, because the monopolar electrode gave a smaller thermal ablation field than the current cluster electrodes. The multiple ablations required to destroy larger tumors increased the potential for complications.

Table 85C.5 Morbidity and Mortality Rates of Radiofrequency Ablation

Unfortunately, the reporting of morbidities is not standardized. Some authors regard low-grade fevers, transient liver function test elevations, small pleural effusions, and right upper quadrant pain as minor complications, whereas others believe these are expected events that should not be reported. As expected, studies undertaken at institutions with skilled interventionalists or surgeons report fewer complications. Variations in patient selection and disease type also confound interpretation of results: for example, patients with HCC have different comorbid factors than those with colorectal metastases, and a patient’s preprocedural state may influence the outcome of RFA.

Direct complications of RFA include biloma, biliary fistula, ascites, hepatic insufficiency, arteriovenous fistula, symptomatic pleural effusion, abscess, pain, hemorrhage, hydropneumothorax, pneumothorax, and thermal injury to surrounding structures. Burns related to grounding pads have also been reported, but these can be avoided by proper positioning of the pads, by using more pads for longer ablations, and by carefully following the manufacturer’s directions. Other potential complications are those related to an operative procedure, such as myocardial infarction, cardiac arrhythmias, and pneumonia.

As illustrated by the following two cases, RFA should be undertaken only by skilled physicians able to identify and manage its complications and only at centers equipped with appropriate staff and equipment for acute care. Figure 85C.9 shows a bile duct injury caused by RFA of a colorectal cancer metastasis near the porta hepatis. The injury was treated with endoscopic retrograde cholangiopancreatography (ERCP) and biliary stenting. Figure 85C.10 shows a hepatic artery pseudoaneurysm attributed to RFA of liver metastases. This was successfully treated with embolization by a skilled interventionalist.

FIGURE 85C.9 Biliary injury following radiofrequency ablation treated with stent placement. RFA, radiofrequency ablation.

FIGURE 85C.10 Hepatic artery pseudoaneurysm following radiofrequency ablation.

The best way to avoid potential complications of percutaneous RFA is to understand its technical limitations. A large multicenter study reported morbidity and mortality rates of 0.3% and 2.2%, respectively (Livraghi, 2003b). In this study, about 33% of all deaths and 10% of morbidities were associated with gastrointestinal thermal injury and perforation. Patients with prior abdominal surgeries resulting in adhesions and those with peripheral liver tumors had an increased risk for gastrointestinal injury during percutaneous ablation. The authors recommended consideration of open or laparoscopic RFA, instead of percutaneous RFA, for tumors within 1 cm of the liver edge adjacent to bowel.

Injury to the diaphragm is also a potential complication during percutaneous RFA, especially when tumors at the dome of the liver are treated. In an animal model, carbon monoxide was introduced into the peritoneal cavity to separate the diaphragm and the liver; although this reduced severe diaphragmatic injury during superficial hepatic RFA (Raman et al, 2004), the technique needs further clinical testing.

Abscess, one of the most frequent complications of RFA, typically occurs 1 week after ablation and requires percutaneous or surgical drainage (de Baere et al, 2003; Wood et al, 2000). Patients with bilioenteric anastomosis or biliary stenting appear to have a higher rate of abscess formation. All patients undergoing RFA should receive periprocedural antibiotics that cover coliforms as well as skin flora, and temperature and leukocyte counts should be followed after the procedure. Normally, low-grade fever and fatigue may occur immediately after RFA for up to 7 to 10 days, a so-called postablation syndrome; however, any persistent significant fever or elevation of white blood cell count should prompt CT imaging for possible hepatic abscess.

Reports of tumor seeding vary from 0.5% to 12% (de Baere et al, 2003; de Sio et al, 2001; Llovet et al, 2001), possibly reflecting differences in follow-up. Llovet and colleagues (2001) found that tumor seeding during RFA was related to subcapsular tumor location and poor tumor differentiation. Recent tumor biopsy, multiple needle insertions, and tumor hemorrhage during treatment may also increase the risk of needle-tract seeding. This complication can be avoided by limiting the number of needle insertions, angling the needle to traverse normal hepatic parenchyma prior to entering the tumor, and cauterizing the tract upon withdrawal of the needle. Nicoli and colleagues (2004) reported rapid diffusion of neoplastic cells after the creation of an arteriovenous fistula following RFA. The authors attributed seeding of tumor cells to the pressure gradient between the high-pressure tumor arteries and the low-pressure portal system; however, this is the only report of this complication in several thousand RFAs to date.

Injury to bile ducts during RFA can result in stenosis and proximal biliary dilation (see Chapter 42A, Chapter 42B ). Most physicians agree that tumors within 15 to 20 mm of a major bile duct should not be treated by RFA (Mulier et al, 2002); however, Elias and colleagues (2001) introduced intraductal cooling to prevent RFA-associated biliary stenosis. They infused cooled (4° C) Ringer’s lactate through a catheter after choledochotomy. Intraductal cooling was undertaken in 13 patients undergoing RFA of tumors within 6 mm of a central bile duct; one patient had a local recurrence, and one developed biliary stenosis (Elias et al, 2004). Biliary stenting also can prevent biliary injury during ablation of tumors near the bile ducts (Wood et al, 2000).