| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 All authors: Liver Research Unit, Chang Gung Memorial Hospital, Chang Gung University, 199, Tung Hwa North Rd., Taipei, 105 Taiwan.
Received June 5, 2002;
accepted after revision July 9, 2002.
Address correspondence to S.-M. Lin.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Of 97 patients with 112 nodular hepatocellular carcinomas, 59 hepatocellular carcinomas were ablated using a standard algorithm and 53 hepatocellular carcinomas, using an interactive algorithm. For the procedure using the interactive algorithm, the electrode's array was partially retracted or fully deployed depending on the change of impedance. Complete tumor necrosis was defined as the lack of enhancement on single-detector helical CT at least 4 months after the last radiofrequency ablation.
RESULTS. Complete necrosis was achieved in 101 (90%) of 112 hepatocellular carcinomas, with complete necrosis being achieved more frequently in hepatocellular carcinomas undergoing interactive ablation (96%) than standard ablation (85%) (p = 0.034). Power rolloff (a clinical end point in which power decreases as impedance increases) occurred in all of the 53 hepatocellular carcinomas that underwent interactive ablation, whereas power rolloff occurred in 48 (81%) of the 59 hepatocellular carcinomas that underwent standard ablation (p = 0.00053). Complete necrosis occurred more frequently when rolloff was achieved (96%) than without rolloff (36%) (p < 0.0001). Multivariate analysis determined that power rolloff was an independent factor in achieving complete necrosis of hepatocellular carcinomas (p < 0.0001).
CONCLUSION. The use of interactive radiofrequency ablation increased the frequency of power rolloff and the rate of complete necrosis in the treatment of hepatocellular carcinoma. Power rolloff was a significant determinant of whether complete necrosis was achieved in hepatocellular carcinomas treated with radiofrequency ablation.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Among various local ablative modalities, radiofrequency ablation has been developed as an effective local ablation therapy for both hepatocellular carcinoma and hepatic metastases [4,5,6,7,8,9]. Clinical end points vary among commercially available radiofrequency ablation systems. Some have thermocouples incorporated within their electrodes and monitor the temperature of the tissue immediately adjacent to the electrode [4, 6]. With temperature as an end point, the physician applies varying levels of radiofrequency energy necessary to maintain the tissue adjacent to the electrode at a temperature sufficient to cause cell death for a period of time. Another radiofrequency ablation system measures the resistance (as ohms of impedance) to the flow of radiofrequency energy (alternating current at 460 kHz) between the active electrode and dispersive electrodes [9]. As the tissue is heated sufficiently to become desiccated, the system impedance increases. When the viable tissue is no longer within the sphere of the deployed tines of the electrode, the system impedance is maximized, and the application of radiofrequency energy is halted [9]. The radiofrequency generator used with such a system is a constant-voltage generator, so as the impedance increases, the power (Watts) being applied decreases. From an electrical engineering standpoint, as the impedance rises, the power "rolls off" until at maximal impedance of less than 10 W of radiofrequency energy is applied to the electrode. A clinical end point of maximal impedance has been described as applying radiofrequency energy until power rolloff occurs.
The instructions for use of the radiofrequency ablation system at our hospital directs that a two-phase application continue until either rolloff occurs or a total of 25 min of treatment time has elapsed. However, in our initial work with this system, we noted that in following the proposed algorithm in 44 patients with 57 hepatocellular carcinomas, we did not achieve power rolloff in 11 (19%) of the treated tumors and that in nine of these 11 tumors we did not achieve complete necrosis of target tumors. A smaller series with this same system noted a similar appearance of a correlation between completeness of treatment effect and achievement of total system impedance, or rolloff [10]. Considering that these patients typically have few other treatment options, incomplete ablation due to an inability to achieve the clinical end point is a concern.
On discussion of this situation, manipulation of the current density of electrodes was proposed as a means to increase the incidence of rolloff. During application of radiofrequency energy, if the tines of the electrode are partially retracted and a high power is applied, then a higher level of power is applied to a smaller volume of liver tissue and the frequency of total impedance may be increased. Thus, to determine whether manipulation of the electrode can have a clinical effect, we proposed to compare results between radiofrequency ablation using the manufacturer's directions (standard ablation) and ablation in which the tines of the electrode were partially retracted or fully deployed depending on a change of impedance (interactive ablation) in patients with hepatocellular carcinoma.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Nine patients presented with chronic hepatitis without cirrhosis, and 51 patients were classified as having Child-Pugh class A cirrhosis, 35 with Child-Pugh B class cirrhosis, and two with Child-Pugh class C cirrhosis. Results of tests for hepatitis B surface antigen and the antibody to hepatitis C were positive in 48 and 37 patients, respectively; in nine patients, the test results were positive for both antigens. One patient had positive results for both hepatitis B surface antigen and antibody for hepatitis D virus, and two patients had cirrhosis of unknown origin.
Hepatocellular carcinoma was diagnosed by pathology (eight tumors),
cytology (104 tumors), or from two diagnostic imagings plus an
-feto-protein level of greater than 200 ng/mL (28 tumors). Histologic
or cytologic grading using Edmondson grading system for hepatocellular
carcinoma was performed for each tumor
[11]. None of the tumors
treated with radiofrequency ablation had been treated with transcatheter
arterial chemoembolization, percutaneous ethanol injection, radiation therapy,
or chemotherapy before radiofrequency ablation.
Before radiofrequency ablation was performed, conventional liver biochemical tests, prothrombin time, and complete blood cell counts were measured. Routine power color Doppler sonography (ProSound 5500; Aloka, Tokyo, Japan) was performed to detect any vessel or arteriovenous shunt in or surrounding the target tumors. Three-phase single-detector helical CT was performed to detect any enhancement of the tumor.
Radiofrequency Ablation
Patients were hospitalized 1 day before undergoing radiofrequency ablation.
After patients had fasted for 4 hr, meperidine hydrochloride (Demoral; Sanofi
Synthelabo, Zuellig, France) and midazolam hydrochloride (Dormicum; F.
Hoffmann-La Roche, Basel, Switzerland) were IV administered 3-5 min before the
procedure. A local anesthetic was provided using 10 mL of 2% lidocaine
hydrochloride (Xylocaine; Astra, Uppsala, Sweden) Two grounding pads were
placed on each of the patient's thighs. Under sonographic guidance, either a
12- or 15-cm-long, 15-gauge needle electrode (LeVeen Needle Electrode; Radio
Therapeutics, Sunnyvale, CA) was introduced percutaneously into the tumor. The
10 tines of the needle electrode then were deployed into the tissue
encompassing the tumor, and the electrode was connected to a radiofrequency
generator (RF2000; RadioTherapeutics). When fully deployed in the tissue, the
10 tines of this needle electrode form an umbrella-shaped array that ranges in
diameter from 2 to 3.5 cm.
Radiofrequency ablation was performed in these patients using either the manufacturer's standard algorithm or a newly developed interactive algorithm. As noted in the electrode's instructions for use, the standard algorithm begins with the power set at 30-50 W, depending on the diameter of the array of the electrode, and the power increases by 10 W per minute until a maximum power of 70-90 W was applied. This power level is applied until either a rapid rise in impedance and decline in power occurs (rolloff) or 15 min of treatment time had elapsed. A second-phase application of radiofrequency energy was applied again either until a second rolloff was achieved or until 10 min of treatment time had elapsed. As a result, no more than 25 min of radiofrequency energy was applied with each deployment of the needle electrode.
The interactive radiofrequency ablation is a slight modification of the standard algorithm. For the interactive algorithm, the initial power settings were similar and power was increased by 10 W per minute up to a maximum of 90 W; however, if impedance had not changed after 8 min had elapsed during the first-phase application of radiofrequency energy or after 4 min during the second-phase application, the tines of the electrode were retracted 0.5-1 cm. After retraction, typically the power would decrease from 90 to 87 W because a smaller surface area (higher current density) of tines was available to transmit alternating current. Power was increased manually on the radiofrequency generator back up to 90-99 W, and the tines remained retracted until impedance began to rise. Once impedance began to rise, the tines were redeployed fully into the tumor, with power remaining at the highest level, and the procedure continued until rolloff occurred.
The electrode was repositioned as necessary to fully encompass large target tumors and mimic a surgical margin in these lesions. After radiofrequency ablation was complete, the needle electrode was removed, and the puncture site was covered with a sterile dressing and compressed by a sand bag. The patient was sent to the recovery unit for at least 4 hr of bed rest before being released to the floor.
Assessment of the Effect of Radiofrequency Ablation
The immediate effects of ablation were assessed using three-phase helical
CT performed 10-14 days after the radiofrequency procedure. Helical CT was
performed on a ProSpeed plus system (General Electric Medical Systems,
Yokogawa, Japan) at 250 mA and 120 kVp with a 10-mm collimation and 1:1 pitch.
Nonionic contrast material was injected at a rate of 2.5-4 mL/sec (100 mL
total) using a power injector. The arterial phase began 30 sec after injection
with helical breath-hold acquisition, the portal phase began 75 sec after
injection with the same acquisition, and the delayed phase began 5 min after
injection. If foci of nodular enhancement in the treated tumor were noted on
CT, those portions of the hepatocellular carcinoma lesion were considered to
be viable and a second ablation procedure was performed to address those
areas. When performed a second time, the same ablation approach as the first
(standard or interactive) was used. Serum -fetoprotein level was also
tested within 2 weeks after radiofrequency ablation. If the baseline level was
abnormal, serum -fetoprotein level was rechecked every 1-2 months.
Long-term follow-up examinations included monitoring serum
-fetoprotein levels, assessing on sonography every 2 months, and
evaluating on helical CT every 2 months for the first 6 months after
radiofrequency ablation. Thereafter, these examinations were performed every 3
months. Local recurrence of hepatocellular carcinoma (persistence of the
original target tumor) was defined as the presence of an enhanced tumor on CT
corresponding to the initial target tumor. A new occurrence of hepatocellular
carcinoma was defined as the development of an enhanced tumor on CT in a
different segment than the original tumor.
Statistical Analysis
We used the Student's t test to compare averages between patients
receiving either standard or interactive ablation. The chi-square or Fisher's
exact test was used to compare proportions between these two groups. A
stepwise Cox hazard model was used to analyze the independent factors
associated with complete tumor necrosis, including the age and sex of the
patient, the cause of chronic liver disease, the size and number of tumor (or
tumors), the Edmondson grade, Child-Pugh grade, the level of
-fetoprotein, and the ablation algorithm used. A p value of
less than 0.05 was considered statistically significant.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
-fetoprotein level, or mean number of sessions of
radiofrequency ablation (Table
1).
|
|
|
Achieving rolloff with radiofrequency ablation was associated with a higher rate of clinical success. The procedure time was significantly shorter when the interactive algorithm was used, and in all of these cases the radiofrequency ablation achieved rolloff. Although achieving rolloff when applying radiofrequency energy resulted in a longer elapsed time, only 81% of the two-phase applications ended with rolloff when the standard algorithm was used (Table 1).
Attaining rolloff was associated with complete necrosis of the target tissue in 51 (96%) of 53 hepatocellular carcinomas in which the interactive algorithm was used as opposed to only 50 (85%) of 59 hepatocellular carcinomas in which the standard algorithm was used (p = 0.034) (Table 1). Complete necrosis was attained more frequently in tumors with small diameters (mean diameter, 23.4 mm), particularly in those smaller than 20 mm in diameter (p = 0.027), than in tumors with large diameters (mean diameter, 30.9 mm) (p < 0.05) (Table 2). The incidence of complete necrosis of hepatocellular carcinoma was higher for those procedures in which rolloff was attained (Table 2). Univariate analysis identified the variables associated with rolloff as tumor diameter, total elapsed time, and use of the interactive algorithm (Table 3).
|
|
The initial CT follow-up showed that the treatment was incomplete in 11 hepatocellular carcinomas (Table 4). The initial treatment was incomplete for the following reasons: the patient refused to undergo a second ablation treatment to treat persistent viable areas near large tumors (four tumors), the close proximity of the target tumor to a large vein (one tumor), vague margins and deep location of tumor in segments IV and VII (four tumors), and the inability to achieve rolloff (two tumors). Eight of these 11 patients were treated with transcatheter arterial chemoembolization (in one patient) or a second radiofrequency ablation that did result in complete necrosis (in seven patients). Before a second treatment could be performed, two patients died from liver failure 78 and 90 days after radiofrequency ablation, and one died from progression of a nontreated hepatocellular carcinoma 78 days after radiofrequency ablation.
|
Complications
One patient experienced a hemothorax, which required drainage 2 weeks after
one radiofrequency ablation session. Two patients experienced a pleural
effusion on the right side 1 week after radiofrequency ablation; the pleural
effusions resolved spontaneously 2-3 weeks after the initial ablation
treatment. High-grade fever (40.3°C) occurred in one patient; however,
results for bacteremia were negative, and the fever subsided soon after oral
administration of naproxen (Nagenton; Nang-Kuang, Taipei, Taiwan).
Long-Term Response to Radiofrequency Ablation
After a median of 352 days (range, 78-458 days) after radiofrequency
ablation, the rate of local recurrence at the site of the initial
radiofrequency ablation was 12%. New hepatocellular carcinomas were present in
22 (23%) of 97 patients within 90-421 days after ablation. In 4% of the
patients, a new hepatocellular carcinoma developed in a different segment and
the original tumor continued to enhance after incomplete treatment with
radiofrequency ablation. At 4 months 2 weeks after the last radiofrequency
ablation, the 80 patients whose -fetoprotein level was not within
clinical norms before the procedure were found to have a normal
-fetoprotein level: the level had normalized (<20 ng/mL) in 26
patients, decreased in 41 patients, and increased in 13 patients.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Achieving complete necrosis of target lesions is the goal in patients with unresectable liver tumors, and this goal is best gained when complete system impedance, as evidenced by power rolloff, occurs [10]. However, the cooling of tissue within close proximity to blood vessels (the heat-sink phenomenon) can influence the time required for and the degree to which complete thermal ablation of hepatic tissue is attained [13]. Ablation systems that rely on point-sampling of tissue temperature can stop the application of radiofrequency energy early, and the resultant thermal lesions can be less than clinically useful [14, 15]. When we performed the radiofrequency ablation until total system impedance had been achieved, the incidence of complete necrosis was 96% of the treated tumors. However, because of the effects of the heat-sink phenomenon, the prescribed time period of 25 min of radiofrequency ablation (15 min for the first-phase application and 10 min for the second-phase application) may be insufficient to cause complete necrosis, especially of cells immediately adjacent to blood vessels.
Manipulation of the circumstances to assist in power rolloff has been reported in other studies. For the open approaches, the surgeon can perform a Pringle maneuver to reduce the blood flow into the liver and the blood supply of the liver [9, 13]. This inflow occlusion procedure during ablation limits heat dissipation and can result in shorter treatment time and more complete clinical outcomes. However, when the radiofrequency electrode is advanced using a percutaneous approach, the physician has few tools available to limit blood flow near the ablation site. Although recent research has shown that inflow occlusion, produced by balloon catheters or organic materials such as gelatin sponge particles, can be useful [16], this approach requires additional equipment. The approach we took did not require additional equipment because we simply manipulated the surface area of the tines of the electrode. Producing the temperatures necessary within the treatment area sufficient to cause coagulation is related to the output of radiofrequency power and tissue impedance, which is expressed as P = V / R where P represents the heating power, V equals the radiofrequency voltage, and R equals the impedance [17]. The achievement of maximal system impedance is related to the heat produced and the subsequent coagulation necrosis. Once the maximal impedance has been reached, the power declines and the ablation is stopped manually, so power rolloff almost always represents complete necrosis over the thermal area. In addition, the thermal lesion size depends on heat generated and heat lost [18]. The heat generated in the tissue distal to the electrode varies as 1/r4, where r represents the radius; therefore, the heat decreases as the distance from the tip increases, which explains why lesions can be treated using the radiofrequency technique [18].
Therefore, any maneuver that alters the diameter, impedance, and power of the encompassed thermal area will affect the temperature in the treated area and the effects of radiofrequency ablation. According to the relations among the amount of radiofrequency power (Watts) being applied, the resultant heating potential, and the radius of the electrode array, heating power will be concentrated, and thus increased, once the radius of thermal area is reduced after retraction of the electrode. When the impedance begins to rise and the temperature increases, radiofrequency power will be redistributed into the outer nonthermal low-temperature and low-impedance zones of the tissue as the electrode is gradually redeployed to its full diameter. These changes can explain the higher rate of complete necrosis and the shorter thermal time realized with the use of the interactive algorithm. According to the standard (noninteractive) two-phase radiofrequency ablation, further ablation might be needed if rolloff is not attained after a total elapsed time of 25 min of radiofrequency ablation treatment. Although rolloff may be achieved in most of the tumors undergoing standard radiofrequency ablation (81% in our series), 19% of these tumors required a longer treatment time than did those tumors ablated using interactive radiofrequency ablation.
In our study, we found that by retracting the electrodes' tines 5-10 mm while still applying the radiofrequency energy, we produced a nidus of necrosis and reduced the vascularity of the tissue in the center of the evolving thermal lesion. Once the central core of the tissue was no longer viable, redeploying the tines to their full diameter ensured that a lower volume of unaffected tissue was included in the zone of the expanded tines. When we actively manipulated the current density of the radiofrequency energy in this manner, all the ablations proceeded to rolloff, resulting in a higher incidence of complete necrosis of the target tumor in an average of 5 min less procedure time. Interactive radiofrequency ablation invariably achieves rolloff and can be used if the impedance does not rise during standard radiofrequency ablation.
The rate of complete necrosis of target tumors was 96% in our series, which is comparable to the rate of 98% of complete necrosis rate in a surgical series of comparable size [9]. The initial local recurrence rate of 16% in our series was not higher than that reported in other studies, in which rates of 30-50% have been reported [5, 6]. Incomplete ablation has been associated with tumors of medium to large sizes (>3 cm in diameter) [8]. In our series, the percentage of tumors determined to have complete necrosis decreased from 98% for the tumors that were 10-20 mm in diameter to 75% for the tumors that were 41-67 mm in diameter (Table 4). Total system impedance (power rolloff) was achieved more frequently in tumors 20 mm in diameter or smaller, irrespective of using the standard or interactive algorithm. The high rate of new tumor occurrence in these patients most likely reflects the advanced stage of chronic liver disease and high proportion (38%) of patients with a hepatitis C virus infection. One group of researchers reported that hepatocellular carcinoma tumors that are associated with hepatitis C virus infection may present with high rate of multicentric occurrence of hepatocellular carcinoma during the preclinical stage [19].
Power rolloff was a significant determinant of complete necrosis in hepatocellular carcinomas undergoing either the standard or the interactive radiofrequency ablation algorithms. Of the 59 tumors ablated using the standard radiofrequency ablation, complete necrosis was achieved without achieving power rolloff in two tumors, which means that sufficient radiofrequency energy was delivered within that tissue after 25 min to cause complete coagulation. Tumor kill can be attained even when power rolloff does not occur [10, 18]. However, when the clinical end point of this radiofrequency ablation system is shown to be a rise of impedance (power rolloff) rather than temperature, confidence in the treatment is best attained when the ablation continues until power rolloff occurs. Although in our series, four tumors (36%) were determined to be non-viable without attaining power rolloff, seven of the tumors (64%) contained viable zones of tumor cells when rolloff was not achieved (Table 2). In contrast, with the interactive algorithm 100% of the tumors attained power rolloff, and only two tumors (4%) were found to contain a rim of viable cells after ablation (Table 1). These two tumors were large (37 and 46 mm in diameter) and required repeated deployments of the 35-mm-diameter electrode array not only to encompass the tumor but also to mimic a surgical margin. Repeated insertion and positioning of an electrode may result in inaccurate positioning because treated and nontreated tissue are difficult to differentiate from one another under sonography or because an echogenic gas formation of the initial radiofrequency ablation limits visibility [20].
Although the patients in this study were not randomized to the algorithm used and the follow-up period was short, the results derived from this retrospective comparison study suggest that the use of an interactive radiofrequency ablation algorithm has benefit in terms of a higher rate of complete necrosis due to a higher incidence of achieving power rolloff. According to the significant difference in achieving complete necrosis in hepatocellular carcinoma tumors undergoing radiofrequency ablation with or without power rolloff, we think that applying radiofrequency energy until power rolloff is essential in every session to ensure complete target ablation.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  S. N. Wong, C.-J. Lin, C.-C. Lin, W.-T. Chen, I. H. Y. Cua, and S.-M. Lin Combined Percutaneous Radiofrequency Ablation and Ethanol Injection for Hepatocellular Carcinoma in High-Risk Locations Am. J. Roentgenol., March 1, 2008; 190(3): W187 - W195. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S-M Lin, C-J Lin, C-C Lin, C-W Hsu, and Y-C Chen Randomised controlled trial comparing percutaneous radiofrequency thermal ablation, percutaneous ethanol injection, and percutaneous acetic acid injection to treat hepatocellular carcinoma of 3 cm or less Gut, August 1, 2005; 54(8): 1151 - 1156. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
