AJR 2001; 176:3-16
© American Roentgen Ray Society
Radiofrequency Ablation of the Liver
Current Status
John P. McGahan1 and
Gerald D. Dodd, III2
1
Department of Radiology, University of California, Davis Medical Center, 4860
Y St., Ste. 3100, Sacramento, CA 95817.
2
Department of Radiology, University of Texas Health Science Center at San
Antonio, 7703 Floyd Curl Dr., San Antonio, TX 78229-3900.
Received February 22, 2000;
accepted after revision June 5, 2000.
Address correspondence to J. P. McGahan.
Introduction
Hepatocellular carcinoma (HCC) and metastases from colorectal carcinoma are
the two most common malignant tumors to affect the liver. When these tumors
are left untreated, the prognosis for both is dismal, with essentially 100%
mortality at 5 years. Conventional therapies such as systemic chemotherapy or
radiation have proven ineffective. Surgical resection of the tumors is
considered the only potentially curative therapy
[1,2,3,4].
Successful resection of targeted tumors with negative tumor margins is
achieved in approximately 80-90% of patients undergoing hepatic resection
[5,6,7].
Unfortunately, surgical resection has many factors limiting its overall
usefulness. Of all patients presenting with a malignant hepatic tumor, few are
surgical candidates. Contraindications to hepatic resection include too many
tumors, tumors in unresectable locations, insufficient hepatic reserve to
tolerate resection, and other medical conditions that make the patient a poor
surgical risk. It has been estimated that only 5-15% of patients with HCC or
hepatic metastases are eligible for resection
[1,2,3,4].
For those patients who undergo hepatic resection, there is considerable
postoperative morbidity—a small but real risk of death related to the
operation—significant monetary expense, and only a modest improvement in
long-term prognosis. The 5-year survival rate for patients undergoing
resection of HCC or hepatic metastases is only 20-40%
[1,2,3].
Most patients die from recurrent hepatic tumors. Although in some instances
surgery may be repeated to resect recurrent tumor, at most institutions
hepatic resection is a "one-shot" therapy. In light of these
shortcomings, an effective, minimally invasive technique is needed for
treating these tumors—one that can be repeated as necessary to treat
recurring tumor.
A number of alternative therapies have been used for the treatment of
malignant hepatic tumors. These include chemoembolization, ethanol injection
therapy, and thermal ablation techniques. Chemoembolization has been studied
extensively and is often reserved for patients with unresectable hepatic
tumors [8,
9]. Ethanol injection therapy
has gained fair international acceptance as a safe, inexpensive, and effective
therapy for small HCC tumors
[10,
11]. However, it has failed to
generate much enthusiasm in the United States. This lack of enthusiasm is
based in part on the need to perform multiple consecutive therapeutic sessions
to completely kill even the smallest HCC tumor and the fact that the technique
is typically performed under sonographic guidance
[11]. Furthermore, ethanol
injection therapy has been shown to be ineffective for the treatment of
colorectal metastases
[12].
Thermal ablation techniques for the treatment of malignant hepatic tumors
include both freezing (cryoablation) and heating (radiofrequency, microwave,
laser, and high-intensity focused sonography) techniques. Of these techniques,
cryoablation has been the most extensively investigated
[13,
14]. The two advantages of
cryoablation over surgical resection are that it can be used to treat liver
tumors that, by number or location, are not surgically resectable, and that it
is associated with diminished morbidity and mortality relative to resection.
The overall prognosis for patients undergoing cryoablation is reported to be
the same as for hepatic resection. However, the limitations of cryoablation
are similar to the limitations of hepatic resection—namely, it is
invasive, with a laparotomy being performed in most cases
[14].
During the last 10 years, considerable interest has developed in the
thermal ablation techniques that produce heat. Methods that are being
investigated include microwave, laser, high-intensity focused ultrasound, and
radiofrequency ablation. Most of the research on microwave ablation has been
performed in Japan, with minimal experience or knowledge of the technique
outside that country [15,
16]. Laser ablation has been
tested most rigorously outside the United States
[17,18,19].
One group of German researchers, Vogl et al.
[19], claim that the technique
is highly effective for the treatment of both HCC and colorectal metastases.
However, one of the primary investigators of laser ablation in England has
essentially abandoned the technique in favor of radiofrequency ablation
[20]. High-intensity focused
ultrasound has been shown to be successful in ablating hepatic tumors in
animal models but has not been used to treat liver tumors in humans
[21]. Overall, the interest
and enthusiasm for radiofrequency thermal ablation has far exceeded that for
either microwave or laser ablation. This article will review the current
status of radiofrequency ablation in the treatment of liver neoplasms
Background
The initial investigation of the use of radiofrequency waves in the body is
credited to d'Arsonval [22] in
1891, who showed that radiofrequency waves that pass through living tissue
cause an elevation in tissue temperature without causing neuromuscular
excitation. These observations eventually led to the development in the early
to mid 1900s of electrocautery and medical diathermy
[23,24,25].
The best known of these developments is the surgical Bovie knife (Liebel
Florsheim, Cincinnati, OH)
[26]. This device is used to
cauterize bleeding tissue. The device consists of an alternating electric
current generator operated in the range of radiofrequency, a small knifelike
electrode, and a large grounding pad. The physical principles of the operation
of the device are fairly simple. The alternating electric current passes back
and forth through the patient between the grounding pad and the Bovie knife.
The grounding pad is applied to the patient's thigh and acts as a large
dispersive electrode that allows the current to pass freely through the
patient without producing any significant heat except around the point of the
Bovie knife. the Bovie knife has a small tip and, when brought into contact
with the patient, acts as a focal point for the electric current. The current
arcs between the Bovie knife and the patient, desiccating and charring the
tissue at the point of contact. Work by Organ
[27] elucidated the physical
principles of the interaction of the alternating electric current with living
tissue. He showed that at low-power settings, the alternating current causes
agitation of the ions in the adjacent tissue. The ionic agitation causes
frictional heat that extends into adjacent tissues by conduction. However, at
high-power settings the ions are quickly destroyed through desiccation and
charring of the superficial tissue, and heat production is minimal
[27,
28]. Subsequent modifications
of the Bovie knife have led to its use in other superficial applications, such
as destruction of the neuronal pathways in the heart in patients with
intractable arrhythmias [29,
30].
In the early 1990s, two independent groups of investigators using modified
radiofrequency equipment studied the percutaneous creation of focal thermal
injuries deep in the liver
[31,
32]. In their studies they
used equipment similar to that of the Bovie knife; that is, an alternating
electric current generator operated in the radiofrequency range, a grounding
pad placed on the tissue, and an insulated needle as a monopolar electrode.
The original needle design was simple and used standard stock needles
insulated to the distal tip (Fig.
1A,1B).
The studies consisted of placing needle electrodes deep in the liver and
creating focal thermal injuries around the noninsulated tip of the needles.
Gross and histologic examination of the ablated livers showed that the process
produced a well-defined concentric region of coagulative necrosis around the
exposed needle tip. This was similar to coagulative necrosis as described in
prior experimentation [33].
However, the size of the thermal injuries was small; the radius of the
coagulated tissue surrounding the exposed needle tip was approximately 1 cm
(Fig.
2A,2B).
Superficial charring around the needle tips similar to that produced by the
Bovie knife limited the size of the thermal injury that could be created in
the liver. Nonetheless, both investigators proposed that radiofrequency
ablation might be an effective technique for destroying small malignant liver
tumors [31,
32].
View larger version (19K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1B. —Original needle design. Drawing shows theoretic lesion that
would be produced if noninsulated needle tip were used during monopolar
radiofrequency electrocautery. (Reprinted with permission from
[75])
|
|
View larger version (120K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2A. —In vivo sonographic and histologic correlation for
radiofrequency coagulation of swine liver. Sonogram of monopolar
radiofrequency lesion shows hyperechoic regions surrounded by hypoechoic rim
(arrow).
|
|
View larger version (118K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2B. —In vivo sonographic and histologic correlation for
radiofrequency coagulation of swine liver. Photograph of in vivo liver reveals
central area of charred tissue (1) surrounded by coagulative necrosis (2) and
hyperemic rim (arrow, 3). L = healthy liver. (Reprinted with
permission from [28])
|
|
Subsequent research has shown that effective tissue coagulation requires
local temperatures in excess of 50°C
[34]. Furthermore, several
factors, including a slow increase in generator power, prolonged
radiofrequency application, and an increase in the exposed surface area of
radiofrequency needle electrodes, have been shown to increase the volume of
coagulated tissue
[34,35,36].
Modern radiofrequency ablation equipment can create thermal lesions of
sufficient size (3-4.5 cm) to be clinically relevant. In the United States,
three primary companies (RITA Medical Systems, Mountain View, CA;
Radiotherapeutics, Mountain View, CA; and Radionics, Burlington, MA) market
radiofrequency devices for tissue ablation. Each of the companies received
approval of their devices from the Food and Drug Administration under a
510K-like device exemption. The radiofrequency tissue ablation devices were
deemed similar enough to the Bovie knife in design and application that a full
Food and Drug Administration application was unnecessary. Each company was
allowed to market its device for generic tissue ablation; however, none could
market its device specifically for the ablation of hepatic tumors. However,
just recently the Food and Drug Administration has approved the use of these
devices for the treatment of liver tumors that are surgically
unresectable.
Equipment
To overcome the problem of tissue desiccation and charring and a limited
radius of coagulated tissue around the radiofrequency needle electrode, each
of the three companies in the United States has experimented with different
radiofrequency needle designs and generator algorithms. Each of the marketed
devices uses a different radiofrequency needle design, generators of different
wattage, and generator algorithms that vary significantly from each other. The
following is a description of the different equipment and operating
parameters.
One of the manufacturers markets a system with two retractable needle
electrodes (Model 70 and Model 90 Starburst XL Needles; RITA Medical Systems).
The needle electrodes consist of a 14- or 15-gauge insulated outer needle that
houses seven or nine (respectively) retractable curved electrodes of various
lengths. When the electrodes are extended, the devices assume the approximate
configuration of a Christmas tree, with the length and diameter of the
electrode clusters measuring 3 or 5 cm, respectively
(Fig. 3A). Four of the
electrodes are hollow and contain thermocouples in their tips that are used to
measure the temperature of the adjacent tissue. The alternating electric
current generator comes in either a 50- or 150-W model; both are operated at
460 kHz. To perform a typical ablation, one or two grounding pads are placed
on the patient's back or thigh. The tip of the needle (with retracted
electrodes) is advanced to the desired location, and the electrodes are
deployed to approximately two thirds their length. The generator is turned on
and run by an automated program. The program starts the generator at
approximately 25 W and then gradually increases the wattage to peak power in
30-120 sec. The program monitors the temperature at the tips of the electrodes
and maintains peak power until the temperature exceeds the preselected target
temperature (typically between 95° and 105°C). The operator watches
the temperature display, and after the target temperature is achieved,
advances the curved electrodes slowly to full deployment while maintaining the
target temperature. When the electrodes are fully deployed, the program
maintains the target temperature by regulating the wattage. As the tissue
begins to desiccate, the amount of power needed to maintain the target
temperature decreases. The company recommends that the target temperature be
maintained for 8-12 min for the smaller electrode and 25 min for the larger
electrode. After the ablation cycle is completed, a temperature reading from
the extended electrodes in excess of 50°C at 1 min is reported to indicate
a satisfactory ablation
[34].
Another radiofrequency ablation device (LeVeen Needle Electrode;
Radiotherapeutics) has retractable curved electrodes and an insulated 14-gauge
outer needle that houses 10 solid retractable curved electrodes that, when
deployed, assume the configuration of an umbrella
[37]
(Fig. 3B). The electrodes are
manufactured in different lengths (2- to 3.5-cm umbrella diameter), with most
users choosing the 3.5-cm device. The alternating electric current generator
is 100 W operated at 480 kHz. Two ground pads are used with the device; both
are placed on the patient's thighs. In application, the tip of the
radiofrequency needle is advanced to the target tissue and the curved
electrodes are deployed to full extension. The generator is switched on and
immediately set to 30 W. Every 60 sec the wattage is increased by 10 until
peak power (90 W) is attained. The device is maintained at peak power for 15
min or until it "impedes out" (i.e., a rapid rise in impedance
stops current flow and the ablation). If the device impedes out, it is turned
off for 30 sec and then restarted at 70% of the maximum power attained at the
time the generator impeded out. The application continues until the generator
once again impedes out or for 15 min. If the device did not impede out during
the first cycle, it is switched off for 30 sec and then restarted at maximum
power and run until it impedes out or 15 min has elapsed. The
Radiotherapeutics ablation algorithm is based on tissue impedance rather than
tissue temperature. The power settings are increased slowly to minimize early
tissue desiccation and charring. The impedance of the tissue increases as the
tissue desiccates. It is believed that an ablation is successful if the device
impedes out.
A third manufacturer has taken an entirely different approach to needle
electrode design. They use an insulated hollow 17-gauge needle with a exposed
needle tip of variable length (2-3 cm) (Cool-Tip radiofrequency electrode;
Radionics). The tip of the needle is closed and contains a thermocouple for
recording the temperature of the adjacent tissue. The shaft of the needle has
two internal channels to allow the needle to be perfused with chilled water.
Two groups of investigators have shown that the cooled needle produces a
larger ablation than does a similar non-perfused needle
[38,
39]. Conceptually, the cooling
is believed to prevent desiccation and charring around the needle tip; thus,
ionic agitation, frictional heat, and the size of the thermal injury produced
are maximized. In an attempt to further increase the size of the ablation, the
company placed three of the cooled needles in a parallel triangular cluster
with a common hub (Fig. 3C). Goldberg et al. [40] tested
this device and found that it produces a significantly larger ablation than
does a single cooled needle.
The generator provided with the cooled-tip radiofrequency needles is the
most powerful of the three commercially available generators. It has a peak
power output of 200 W and is operated at 480 kHz. In clinical application,
four grounding pads are placed on the patient's thighs. The tip of the single
or cluster electrode is advanced to the desired location in the tissue to be
treated. To maintain the correct spacing and triangular configuration of the
cluster needle, a guidance thimble is used
(Fig. 4). The single or cluster
electrode needles are connected both to the generator and to a perfusion pump.
The pump is switched on and sterile chilled water is circulated through the
needle. To begin the ablation, an automated program gradually increases the
power for 1 min to a peak of 200 W and maintains the power at that level until
the impedance rises 20 
over the starting level. The power is then
reduced automatically to 10 W for 15 sec and then returned to maximal power
until the impedance increases again. If the power cannot be maintained for at
least 10 sec without a rise in the impedance, the power is reduced for
subsequent cycles to minimize elevations in impedance. Successive cycles are
continued for a total ablation time of 12 min. Successful ablations usually
increase the temperature of the ablated tissue to between 60° and
80°C.
Clinical Application
The clinical application of radiofrequency ablation of hepatic tumors
usually includes these steps: preoperative evaluation; choice of approach:
percutaneous, laparoscopy, or laparotomy; anesthesia and medications; needle
placement and treatment strategy; and followup. The following represents a
summary of the practice in our programs, but the practice may vary from
institution to institution.
Preoperative Evaluation
The preoperative evaluation begins with a review of the pertinent imaging
studies. Goodquality abdominal CT or MR imaging is the fundamental imaging
examination on which the candidacy of a patient for radiofrequency ablation is
based. These preoperative imaging studies are used to determine the number and
size of tumors and their relationship to surrounding structures such as blood
vessels, bile ducts, gallbladder, diaphragm, and bowel. Patients are
considered potential candidates if they have fewer than five tumors, each less
than 5 cm in diameter, and no evidence of extrahepatic tumor
[6,
40,41,42].
In practice, patients with more than two tumors approaching 5 cm are poor
candidates because of the sheer bulk of their tumor burden and the difficulty
in completely ablating tumors of such size. By far the most common liver
tumors that have been treated with radiofrequency ablation are HCC tumors and
colon metastases. In addition, we have treated metastatic tumors from the
pancreas, breast, stomach, and neck, as well as metastatic neuroendocrine
tumors. In fact, any type of tumor isolated to the liver that meets our
criteria can be treated by radiofrequency ablation. However, patients with
potentially resectable tumors should be told that hepatic resection is the
standard of care, and then referred for surgical evaluation.
Tumors close to the vital structures require careful consideration. We have
found that ablation of tumors adjacent to the diaphragm, liver capsule,
gallbladder, or main portal vessels will cause considerably more pain during
and after the procedure than will ablation of tumors embedded in the hepatic
parenchyma. Ablation of tumors adjacent to the diaphragm will cause transient
right shoulder pain; however, if the diaphragm is directly ablated, as
occurred in one of our patients, the pain can be severe and last for several
months. The ablation of a tumor touching the wall of the gallbladder can lead
to the development of cholecystitis, with symptoms lasting 2-3 weeks. Tumors
adjacent to large vessels may be difficult to completely ablate because the
blood flow in the vessel cools the tissue and may prevent adequate heating.
Careful consideration needs to be given to the treatment of tumors lying in
the bifurcation of the right and left portal veins or adjacent to either main
portal trunk. Aside from the increased pain for the patient and diminished
ability to effectively heat tumors in these locations, a very real risk exists
of damaging the main right or left bile ducts. If the ducts are injured, the
likely sequela is biliary obstruction. Finally, the treatment of subcapsular
tumors abutting other abdominal viscera poses the risk of damage to those
organs. Of greatest concern in this regard is the possibility of inducing
thermal necrosis in an adjacent loop of bowel
[43,
44].
If a patient is a good candidate for thermal ablation on the basis of his
or her baseline abdominal CT or MR imaging, then a series of other tests is
indicated before actual treatment. All patients should have histologic
confirmation of a hepatic malignancy. CT of the chest should be performed to
rule out the possibility of pulmonary metastases. A serum alkaline phosphatase
level should be obtained; if the level is elevated without other just cause,
or if the patient has new unexplained bone pain, technetium bone scanning
should be performed to exclude the possibility of bone metastases. If the
ablation procedure is to be performed using sonographic guidance, targeted
hepatic sonography should be performed to determine if the lesions can be seen
on sonography and to determine if a safe and adequate approach exists.
Hematologic evaluation is necessary because the ablation needles are fairly
large (14- to 17- gauge), and multiple percutaneous transhepatic needle
punctures may be necessary to complete an ablation. Hematology tests should
include a complete blood count, hemoglobin, hematocrit, prothrombin time,
partial prothrombin time, and a platelet count. Coagulopathies should be
corrected before an ablation session. A routine serum chemistry panel should
be obtained to check electrolytes and liver function. A baseline serum
-fetoprotein and a carcinoembryonic antigen level should be obtained in
patients with HCC and colorectal metastases, respectively. On the day before
an ablation session, and ECG and blood type and match should be obtained in
preparation for possible emergent surgery that may be necessary to treat a
significant complication caused by the ablation. Contraindications to
performing radiofrequency thermal ablation of hepatic tumors include excessive
intrahepatic tumor burden, untreatable extrahepatic tumor, Child's class C
cirrhosis, active infection, and inability to execute a proper consent.
Choice of Approach: Percutaneous, Laparoscopy, or Laparotomy
At our institutions, whenever possible radiofrequency ablation is performed
percutaneously. Percutaneous treatment has several advantages over other
approaches. The percutaneous approach is the least invasive, produces minimal
morbidity, can be performed on an outpatient basis, requires only conscious
sedation, is relatively inexpensive, and can be repeated as necessary to treat
recurrent tumor. However, advocates of laparoscopic thermal ablation of
hepatic tumors claim that the laparoscopic approach provides some distinct
advantages over the percutaneous approach
[45]. With the laparoscopic
technique, the entire liver can be imaged with a high-frequency transducer
placed directly on the surface of the liver. This technique allows the
visualization (and treatment) of small tumors that cannot be detected using
any other imaging technique (Fig.
5A,5B).
Furthermore, with laparoscopy the extent of a tumor can be more accurately
staged. The identification of previously undetected surface lesions or
peritoneal implants may lead to the appropriate cancellation of the planned
radiofrequency ablation [46].
This is important if the goal of the procedure is potential cure. However,
Siperstein et al. [45] have
also used laparoscopic radiofrequency for symptomatic relief of hormonally
active liver metastasis. Additionally, advocates of a laparoscopic approach
argue that they can perform a Pringle maneuver (temporary occlusion of the
hepatic artery and portal vein) and thus enhance the size of a thermal
ablation. The disadvantages of a laparoscopic approach include the added
invasiveness of the procedure, with the associated complications and added
costs, and the technical difficulties of the procedure. In particular, the
placement of the needle electrode can be problematic. Unlike the percutaneous
technique in which the sonographic probe and radiofrequency needle can be
moved to any position over the right upper quadrant to achieve the best angle
of approach to treat a tumor, the laparoscopic technique is hampered by
limited access through the existing laparoscopic ports. Furthermore, no needle
guides are currently available that can be attached to the laparoscopic
sonographic transducers, and the radiofrequency needles cannot be placed
parallel to the transducers. Most commonly, the needles are placed either
perpendicular or oblique to the sonographic transducers. The manipulation of
the needle from these angles is technically challenging and is not easily
mastered.
View larger version (150K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5A. —62-year-old man with metastases to liver from colon cancer.
Both preoperative CT scan (not shown) and sonogram show only a single large
metastasis in right lobe of liver (arrow) that was scheduled for
operative resection.
|
|
View larger version (183K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5B. —62-year-old man with metastases to liver from colon cancer.
At surgery, intraoperative sonography revealed several 4-mm metastases
scattered throughout liver (arrow). Resection was not performed.
|
|
Other investigators have published their experiences with radiofrequency
ablation performed via open surgical treatment
[42]. Disadvantages of this
technique include the associated morbidity and mortality of an open procedure
and general anesthetic, the added expense of the procedure and associated
recovery time, and the fact that the technique is typically a one-shot
therapy. Some of the advantages of the technique are the same as for the
laparoscopic approach—namely, the ability to scan the whole liver with
high-frequency transducers and to accurately stage the extent of the tumor
[47] (Fig.
5A,5B).
Additionally, the open approach allows a great deal of freedom in placing the
sonographic transducer and radiofrequency needle. The radiofrequency needle
can be placed through a needle guide that is attached to the transducer, or it
can be placed using a freehand technique. Difficult lesions adjacent to the
diaphragm, bowel, or gallbladder may be treated easily with the open
technique. These organs can be removed or isolated from the mobilized liver to
prevent damage during the ablation of a tumor
(Fig. 6). Finally, the blood
supply to the liver may be temporarily stopped during radiofrequency ablation.
A Pringle maneuver, which causes temporary occlusion of the portal vein and
hepatic artery, decreases blood supply to both the targeted tumor and the
adjacent hepatic parenchyma. The result is a larger thermal injury than is
possible with normal blood flow
[42].
Anesthesia and Medications
General anesthesia is required for laparoscopy or open surgical treatment.
However, conscious sedation is usually sufficient for a percutaneous approach.
At our institutions, we use the services of the anesthesiology department for
percutaneous radiofrequency ablation. On the morning of the procedure, an
anesthesiologist or nurse-anesthetist evaluates the patient for suitability
for anesthesia. The medical history is reviewed, the laboratory test results
are checked, and the ECG is interpreted. If deep conscious sedation is not
contraindicated, the patient is prepared for the procedure. A peripheral IV
line is started, and the patient is monitored to track blood pressure, pulse,
respiratory rate, and peripheral oxygenation. The site of the planned
percutaneous puncture with the radiofrequency electrode is anesthetized with
1% lidocaine hydrochloride (Xylocaine; AstraZeneca, Wilmington, DE). The
anesthesiologist or nurse-anesthetist administers an IV sedative. This can be
a combination of the traditional drugs fentanyl (fentanyl citrate injection;
Elkins-Sinn, Cherry Hill, NJ or Baxter Healthcare, Deerfield, IL) and
midazolam hydrochloride (Versed; Roche Laboratories, Nutley, NJ) that are used
for many percutaneous intervention procedures, or, as we prefer, the more
potent and short-acting agents propofol (Diprivan; Astra-Zeneca) or
remifentanil hydrochloride (Ultiva; Glaxo Wellcome, Research Triangle Park,
NC) [48]. These newer agents
are administered by a constant drip infusion. The major advantages of these
agents are the deep level of anesthesia that can be obtained and their short
duration of action. If the cooperation of a patient is needed to achieve
satisfactory placement of the needle electrode, the patient is easily aroused
within minutes of stopping the infusion of the agent. Likewise, if a patient
has respiratory suppression due to oversedation, he or she is quickly
recovered by stopping the infusion and ventilating the patient using a bouche
bag for a few minutes.
At the end of a radiofrequency ablation session, the patient can experience
severe nausea and pain with the decrease of the sedative. Because of the
frequency of these two side effects, we established a routine protocol to
treat these problems. Immediately after the conclusion of the procedure and
before transporting the patient to the recovery area, we administer both an
antiemetic and morphine IV. After the patient is transferred to the recovery
area, combined oral hydrocodone bitartrate and acetaminophen tablets (Vicodin;
Knoll Laboratories, Mount Olive, NJ), are given to control subacute pain. This
medication regime is usually sufficient to keep the patient comfortable for
3-4 hr, at which time the side effects of the ablation have usually diminished
to a tolerable level. Fewer than 50% of patients require additional analgesics
during the immediate recovery period, and even fewer require analgesics after
discharge. If pain persists, the patient is typically given a 3-day
prescription for combined oral hydrocodone bitartrate and acetaminophen
tablets (Vicodin).
Occasionally, patients may be unable to tolerate percutaneous
radiofrequency ablation with just conscious sedation. This may be true for
patients with cancer with chronic pain who have been taking routine
analgesics, patients with a history of alcohol or drug abuse, or patients with
a low tolerance for pain. General anesthesia may be indicated for these
patients. Likewise, general anesthesia may be preferred if an ablation is
going to be extensive and the procedure is expected to last 3 hr or
longer.
Although there has been no trial to validate the use of prophylactic
antibiotics, such use has become routine at some institutions. One of us
routinely administers IV cephalosporin just before treatment and continues
with oral cephalosporin for 5 days after treatment.
Needle Placement and Treatment Strategy
Radiofrequency needles can be placed under sonographic, CT, or MR guidance
using a percutaneous approach. The needles are usually well revealed by each
technique (Fig.
7A,7B).
However, without question sonography is the most common method used to guide
percutaneous radiofrequency tumor ablation
[49,50,51,52,53,54,55,56,57,58,59].
Its advantages over CT and MR are its realtime capabilities, vascular
visualization, availability, speed, and low cost.
The primary disadvantage of sonography is a limited ability to assess the
effectiveness of an ablation. Although the ablation process produces a dense
echogenic response, the size of the echogenic response provides only a rough
estimation of the size of the ablation
[60] (Figs.
8 and
9A,9B,9C,9D).
Furthermore, the echogenic response obscures the margins of the tumor being
treated, particularly the posterior margins (Figs.
8 and
9A,9B,9C,9D).
Both CT and MR imaging have been reported to be more reliable in this regard
[60,
61].
View larger version (111K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 8. —Sonographic artifact seen in swine liver. Radiofrequency
ablation causes formation of microbubbles in liver. Margin near ablation
becomes echogenic (solid arrow) and produces distal acoustic
shadowing (open arrow). Echogenic response prevents visualization of
deeper anatomy.
|
|
View larger version (123K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 9A. —64-year-old man with hepatitis C and hepatocellular carcinoma
(HCC). CT scan shows 1.5-cm enhancing lesion (arrow) that was
biopsy-proven HCC. Patient also had small amount of ascitic fluid surrounding
liver.
|
|
View larger version (126K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 9D. —64-year-old man with hepatitis C and hepatocellular carcinoma
(HCC). Echogenic response has diminished after approximately 10 min, and
well-demarcated lesion measuring 2.8 x 3.8 cm with hyperechoic rim is
identified in location of HCC.
|
|
Regardless of the method used to guide radiofrequency tumor ablation,
strategy must be established before placement of the needle. The objective in
treating a tumor must be to kill the entire tumor as well as a tumor-free
margin of normal liver. The surgery literature has clearly shown that an
adequate tumor-free margin is mandatory to prevent local recurrence of a tumor
after hepatic resection. In one of the latest publications on the subject, an
adequate tumorfree margin was defined as being preferably 2 cm and no less
than 1 cm of normal liver
[62]. If the goal of
radiofrequency ablation is to duplicate the success rate of hepatic resection,
then in all likelihood the same requirements for a tumor-free margin will need
to be followed.
Given that most radiofrequency ablation devices produce an approximate 3-cm
ablation, the largest tumor that should be treated by a single ablation should
be smaller than 2 cm in diameter (Fig.
10A). A 2-cm tumor treated with a 3-cm ablation yields at most a
5-mm tumor-free margin. If error in needle placement is factored in, the
margin could be much less. Certainly, 3-cm tumors should not be treated by a
single 3-cm ablation. To treat a larger tumor, multiple ablations need to be
overlapped to build a composite thermal injury of sufficient size to kill the
tumor and to provide the desired tumor-free margin. By strict geometric
analysis, six overlapping spherical ablations are necessary to create an
intact composite thermal sphere of the next magnitude
(Fig. 10B). If six 3-cm
thermal spheres are precisely overlapped, with four in the x-y plane
and two along the z-axis, the largest intact composite thermal sphere
that can be created is just 3.75 cm, or 1.25 times the diameter of a solitary
sphere. Thus, the largest tumor that should be treated with six 3-cm
overlapping ablations should be less than 3.25 cm in diameter. Larger tumors
can be treated with radiofrequency ablation, but they require many more
ablations or require adjuvant techniques to increase the size of an ablation.
A strategy for treating large tumors consists of the creation of thermal
cylinders [63]
(Fig. 10C).
View larger version (68K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 10B. —Artist's rendition of ablation schemes. Six optimally placed
overlapping spheres produce composite spherical thermal injury with diameter
equal to 1.25 times diameter of a single ablation sphere.
|
|
View larger version (68K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 10C. —Artist's rendition of ablation schemes. Overlapping thermal
cylinders is effective way to treat large tumors. Each cylinder is created by
overlapping serial ablations by 50% along a single needle path. Adjacent
cylinders are overlapped by 50%.
|
|
When performing multiple radiofrequency ablations using sonographic
guidance, it is important to plan the order of ablations so that the deepest
ablations are performed before the superficial ones (Figs.
8 and
9A,9B,9C,9D).
This strategy minimizes the possibility that microbubbles might obscure
visualization of the deeper portions of the tumor and thus prevent completion
of the ablation session.
Follow-Up
Immediate evaluation can consist of sonography performed at the time of
treatment. This may be helpful with HCC tumors that are vascular. However, it
may not be helpful in metastatic disease. Solbiati et al.
[58] found that use of
contrast-enhanced sonography may help to detect residual tumor after
radiofrequency treatment. This immediate feedback can be used for
reapplication of radiofrequency in areas that are untreated.
Most institutions perform CT of the liver within a few hours of the
ablation to gauge the completeness of the ablation and to look for any
potential complications [61].
However, the accuracy of the completeness assessment on the immediate CT scan
is often limited because of the presence of an ablation-induced hyperemic rim
around the margin of the ablated tissue (Fig.
11A,11B).
This hyperemia is difficult to distinguish from residual tumor. Fortunately,
the hyperemia is usually resolved by 1 month after the ablation, and a more
accurate assessment of the completeness of the ablation can be made at that
time (Fig.
12A,12B).
In some instances, CT is not performed immediately, but rather 1 month after
ablation. If no evidence of residual tumor is seen on the CT scan 1-month
after ablation, patients are usually followed up by repeated CT every 3
months. Each follow-up CT scan must be scrutinized for evidence of tumor
recurrence in the liver, adjacent to and remote from the ablated site, and
outside the liver. Typical sites of extrahepatic tumor recurrence in patients
with HCC are the adrenal glands, lungs, bones, and regional lymph nodes along
the porta hepatis, gastrohepatic ligament, celiac axis, and cardiophrenic
sulcus. Patients with colorectal carcinoma tend to develop extrahepatic
metastases at the primary tumor site, in the peritoneal cavity, and in the
lungs. Because of the frequency of extrahepatic metastases in patients with
colorectal carcinoma, we usually perform both abdominal and pelvic CT every 3
months and chest CT every 6 months.
View larger version (171K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 11B. —58-year-old man with hepatocellular carcinoma. Arterial phase
CT scan immediately after ablation shows normal hyperemic rim
(arrows) around ablated tumor. Hyperemic rim may prevent accurate
assessment of completeness of ablation.
|
|
View larger version (135K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 12B. —67-year-old man with two metastases from colon cancer.
Three-month follow-up CT scan after radiofrequency ablation shows complete
ablation of tumors. Percutaneous radiofrequency approach was used for
treatment of the more lateral lesion. However, because of overlying bowel and
proximity to gallbladder, open radiofrequency ablation was used for the more
medial lesion.
|
|
The technique used to perform follow-up CT has a significant impact on the
early detection of intrahepatic tumor recurrence. Most early recurrences in
patients with HCC are detected only in the arterial phase of good-quality
three-phase CT. Likewise, the subtle changes of local intrahepatic tumor
recurrence in patients with colorectal metastases may be visualized only
during a strong portal venous phase contrast-enhanced CT scan
[56,
59]. We routinely perform
three-phase CT in all patients with HCC and two-phase CT in patients with
colorectal metastases. Additionally, we obtain repeated -fetoprotein
and carcinoembryonic antigen levels every 3 months in patients with HCC or
colorectal hepatic metastases, respectively. We find the levels helpful in
assessing subtle changes on the CT scans and in determining if additional
evaluation is warranted. Changes in -fetoprotein or carcinoembryonic
antigen values must be assessed on an individual basis for each patient on the
basis of their values before treatment and the normal levels used at each
medical institution.
Several investigators prefer to use MR imaging to follow up their patients
after radiofrequency ablation. One study claims that MR imaging is more
sensitive than CT for the detection of early intrahepatic tumor recurrence
[61]. The interpretation of MR
imaging for tumor recurrence is based primarily on the premise that ablated
tissue produces minimal signal on T-2 weighted sequences, whereas tumor
produces high signal. Dynamic contrast enhancement of suspected tissue is
further evidence of tumor recurrence.
Regardless of the technique used to follow up patients after ablation, the
primary objective is to detect tumor recurrence as soon as possible so the
appropriate therapy can be instituted. If recurrence is limited to the liver,
reablation may be effective (Fig.
13A,13B,13C,13D).
One of the potential benefits of percutaneous radiofrequency ablation over
other more invasive forms of treatment, such as surgery, intraoperative
cryotherapy, or intraoperative radiofrequency treatment, is that it can be
repeated as often as necessary to treat intrahepatic tumor recurrence. In our
practices, we have treated individual patients with local tumor recurrences as
many as seven times during a 3-year period. Each time the tumor appeared
completely destroyed, and the patient had 3-6 months of asymptomatic health
between treatments. If a patient develops extrahepatic tumor or extensive
intrahepatic tumor, alternative therapies such as systemic chemotherapy or
hepatic chemoembolization should be considered.
View larger version (139K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 13D. —46-year-old man with metastatic colon cancer. CT scan
immediately after reablation shows enlarged thermal injury (arrow) at
site of treated tumor recurrence and no definite evidence of residual
tumor.
|
|
Results
Long-term results of radiofrequency ablation of either primary or secondary
liver neoplasms are scant because this is a fairly new technique. The goal of
most of the early reported ablations is complete tumor necrosis and possibly a
cure. However, some published results are available on the short-term
follow-up of patients treated with radiofrequency ablation. These results are
presented by tumor type and method of treatment. Many results are with
prototype devices and do not reflect current refined technology. Furthermore,
many early results were with conventional needles similar to the original
single noninsulated needles described by McGahan et al.
[31] and Rossi et al.
[32] (Figs.
1A,1B
and
2A,2B).
Percutaneous Treatment of HCC
At least four series document the results of percutaneous radiofrequency
treatment of HCC. The original study of Rossi et al.
[56] used a conventional
needle that was insulated to the distal tip. Rossi et al.
[55] conducted a later series
using an umbrella-type needle (Fig.
3A,3B,3C).
Livraghi et al. [53] and
Francica and Marone [49]
published one study each in which they used a cooled needle device. In all
studies, near-complete necrosis of 90% of the treated tumors had occurred at 6
months. These results were for tumors smaller than 3 cm. However, for the
study by Livraghi et al. [53]
with tumors larger than 3 cm (mean, 5.4 cm,) the complete necrosis rate was
only 47.6%. Complete necrosis was 71% for noninfiltrating tumors that were
3.1-5.0 cm, and only 25% for noninfiltrating tumors greater than 5 cm in
diameter. Also, the rate of disease-free survival at follow-up was lower; the
first study of Rossi et al.
[56] found 64% at 23 months
and 71% at 12 months in a subsequent study
[55]. Francica and Marone
reported 67% of patients to be disease-free at 15 months. Thus, with HCC,
although there is a high rate of complete necrosis of ablated tumors, about a
third of the patients develop recurrent tumor. Kainuma et al.
[50] found it useful to
monitor and perform reablation in these patients. These results are summarized
in Table 1.
Percutaneous Treatment of Metastatic Liver Tumors
Radiofrequency ablation has also been used to treat tumors that have
metastasized to the liver. The results of five series are summarized in
Table 2. The original study by
Rossi et al. [56] used a
conventional needle that was not insulated to the distal tip. Solbiati et al.
[57] and Livraghi et al.
[54] conducted series with a
conventional needle not insulated at the tip combined with saline infusion.
Ten to fifteen milliliters of saline was introduced into the tumor before
radiofrequency was applied in hopes of increasing the area of tumor necrosis.
Rossi et al. [55] performed a
series using an umbrella-type needle. Solbiati et al.
[59] and Lencioni et al.
[51] reported results using a
cooled-tip needle in 1997. The rate of complete necrosis in these studies
ranged from 52% to 93%. However, few patients remained disease-free; the rates
ranged from as low as 11% at 11 months in the series of Rossi et al.
[56] to as high as 33% at 18
months in the series of Solbiati et al.
[59]. The marked discrepancy
between the complete tumor necrosis rate and the percentage of patients who
remain disease-free is in part due to the biology of metastatic liver disease
and our limited ability to detect the full hepatic tumor burden. Most patients
with a few 2- to 3-cm liver metastases also have numerous subcentimeter or
microscopic tumors that are not detectable with current imaging
technology.
Laparoscopic Results
Siperstein et al. [45]
described results of laparoscopic radiofrequency ablation of 13 neuroendocrine
tumors in the livers of six patients. The tumors ranged in size from 1 to 7
cm. Follow-up CT showed complete necrosis in 11 of 11 lesions in four patients
at 3 months. All patients showed improvement in their symptoms after
radiofrequency ablation. Cuschieri et al.
[46] used laparoscopic
sonography to guide treatment of 10 patients, two with hepatoma and eight with
metastases. No complications occurred. Patients were discharged within 2 days
of intervention. This compares with the usual same-day discharge after
percutaneous techniques, but it is a shorter hospitalization than with open
techniques. At follow-up ranging from 6 to 20 months, one patient had died of
progressive disease, one patient had further metastases, and eight were
disease-free.
Intraoperative Results
Elias et al. [64] described
the use of combined liver resection and radiofrequency ablation to treat seven
patients with liver metastasis. This is an interesting approach in that these
were liver metastases that were thought to be unresectable by previous
approaches. Radiofrequency was used to ablate those lesions that could not be
surgically resected. The Pringle maneuver was performed in all patients, and
all seven patients were disease-free up to 10-months of follow-up. Jiao et al.
[65] reported a series of
eight patients with HCC and 27 patients with metastases in which 30 of the 35
patients underwent intraoperative radiofrequency ablation using the Pringle
maneuver. Thirteen of the 30 patients had combined radiofrequency ablation and
a surgical resection. At approximately 10 months of follow-up, 24 of the 35
patients were found to have stable disease.
To our knowledge, the largest series to date of intraoperative
radiofrequency ablation of hepatic tumors is by Curley et al.
[42]. In that study,
radiofrequency ablation was used to treat 169 tumors in 123 patients. Of the
123 patients, 92 (75%) were treated with laparotomy and 31 (25%) were treated
percutaneously. Forty-eight (39%) patients had HCC and 75 patients (61%) had
hepatic metastases. All tumors were treated with an umbrella-type electrode,
and a Pringle maneuver was used on all intraoperative cases. Overall, complete
necrosis had occurred in 98% of the ablated tumors at a median follow-up of 15
months, and 72% of the patients remained tumor-free during the same time. In
the series of Curely et al., the size of the tumors treated by percutaneous
radiofrequency (2.4 cm diameter) was less than size of tumors treated
intraoperatively (3.8 cm). Those authors did not analyze the outcome
difference in the percutaneous and the intraoperative groups. However, only
three (1.8%) of 169 treated lesions had local recurrence.
Complications
As stated in the section on anesthesia, almost all patients experience pain
and nausea during and immediately after ablation. These side effects are
usually controllable and transient. Nausea rarely persists longer than 2-3 hr
after the procedure. Approximately 25% of patients will have pain that
requires continued medication at the time of discharge
[66]. In 98% of patients, pain
will be gone within 1 week after the procedure
[66].
As reported with other types of tumor ablation procedures
(chemoembolization and cryoablation), a percentage of patients (in our
experience, about 25%) will develop a delayed syndrome after ablation. The
frequency, severity, and time to onset appear to be directly related to the
amount of tissue ablated. The typical presentation consists of flulike
symptoms (low-grade fever [up to 38.8°C] accompanied by general malaise)
that begin 3-5 days after the ablation and persist for approximately 5 days.
With large-volume ablatioons we have seen the syndrome begin almost
immediately, cause fever as high as 39.4°C, produce severe lethargy, and
last for as long as 2-3 weeks. Several patients have reported night sweats.
Appropriate treatment of the syndrome is primarily supportive. We inform
patients that they may develop the syndrome after their ablation. Their fever
is treated with oral acetaminophen. They are instructed to call us if their
fever exceeds 38.8°C or lasts longer than 5 days. If their fever exceeds
38.8°C, a blood sample is drawn and cultured to determine if the patient
is septic [66].
Radiofrequency ablation of the liver is considered safe, with an extremely
low major complication rate observed by multiple groups. More serious
complications have been reported in the literature. Rossi et al.
[56] described capsular
necrosis, and Solbiati et al.
[57,
59] reported intraperitoneal
hemorrhage that did not require transfusion. Solbiati et al.
[57] also reported one case of
fairly severe hypotension that persisted for 3-4 hr after the procedure.
Livraghi et al. [53,
54] reported complications in
their patients that consisted of two pleural effusions, a perihepatic
hematoma, a hemothorax that required surgical repair, self-limited
intraperitoneal bleeding, and hemobilia and cholecystitis in one patient each.
We have experienced several complications, including three cases of
intrahepatic arterial bleeding, with one requiring selective arterial
embolization; one burn of the diaphragm that caused pain for 3 monts; one 3-cm
hepatic abscess that was treated with oral antibiotics; five 1- to 3-cm first-
or second-degree burns along the edge of the grounding pads; and three
instances of tumor seeding along the needle tract. Additionally, Lees and
Gilliams [67] reported hepatic
abscesses and a thermal burn of the transverse colon. Livraghi et al.
[52] reported a single death
(0.8% rate) in their series. This was the result of a Staphylococcus
aureus infection that developed 3 days after the procedure. Because of
this complication, those authors administer antibiotic prophylaxis of 1000 mg
of ceftriaxone sodium (Rocephin; Roche Laboratories) to all patients.
Few reports mention complications occuring after radiofrequency ablation
performed via a laparotomy. Of those that have been reported, most consist of
fever and pain. Other reported complications include a perihepatic abscess and
an intrahepatic hemorrhage that required arterial embolization
[42].
Discussion and Future Strategies
Top
Introduction
Background
