| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Medical Imaging, University Health Network/Mount Sinai Hospital,
Toronto, Ontario, Canada.
2 Princess Margaret Hospital 3-923, University of Toronto, 610 University Ave.,
Toronto, Ontario M5G 2M9, Canada.
3 Sunnybrook and Women's College Health Sciences Centre, Toronto, Ontario,
Canada.
Received December 17, 2004;
accepted after revision March 3, 2005.
Address correspondence to M. J. Dill-Macky
(Marcus.Dill-Macky{at}uhn.on.ca).
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Twenty-two lesions (1.5-3.7 cm) were studied in 19 patients. Enhanced sonography was performed before and within 1 hr after radiofrequency ablation. At routine 2-week follow-up CT or MRI, additional enhanced sonography was performed. The findings of preablation CT or MRI and enhanced sonography were compared with those of postprocedural and follow-up enhanced sonography by three radiologists experienced in these techniques. The reviewers were unaware of the follow-up CT or MRI results (reference standard). Technical adequacy, ablation zone targeting, and identification of residual disease were assessed by each reviewer, and the results were analyzed by consensus.
RESULTS. One postprocedural sonographic study was considered technically inadequate. Postprocedural sonography predicted the follow-up CT or MRI results in 76% (16/21) of subjects (sensitivity, 88%; specificity, 40%; positive predictive value [PPV], 82%; negative predictive value, [NPV] 50%). Follow-up CT or MRI identified accurate targeting in 17 of 22 subjects. Follow-up sonography agreed with CT or MRI in 82% (18/22) of subjects (sensitivity, 88%; specificity, 67%; PPV, 88%; NPV, 67%). Postprocedural sonography predicted the follow-up CT or MRI results in 81% (17/21) of subjects (sensitivity, 40%; specificity, 94%; PPV, 66%; NPV, 83%). Follow-up CT or MRI detected residual disease in six subjects. Follow-up sonography agreed with CT or MRI in 91% (20/22) of subjects (sensitivity, 83%; specificity, 94%; PPV, 83%; NPV, 94%).
CONCLUSION. Postprocedural enhanced sonography has the potential to guide completion of radiofrequency ablation at the time of initial therapy when residual disease is detected. The procedure is less accurate in detection of residual disease than is either delayed enhanced sonography or CT or MRI.
Keywords: ablation • contrast media • interventional radiology • liver • radiofrequency • sonography
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Local ablative therapies destroy the tumor while maximally preserving surrounding normal liver parenchyma. Patients may have multiple treatments in one session, allowing complete ablation of a tumor when a single treatment is inadequate.
Therapy usually is targeted and monitored in real time with sonography; however, CT or, rarely, MRI with or without contrast material may be used for lesions occult to sonography. Factors affecting the size and distribution of the ablated zone include probe gauge, length of the exposed tip, probe temperature, local blood flow, and the duration of treatment [4-7]. During therapy, an area of acoustic shadowing develops around the probe because of gas formation secondary to thermal effects. This shadowing severely limits sonographic visibility, and a poor correlation between lesion size seen on sonography and actual lesion size seen at histologic examination has been reported [7-9].
|
|
|
|
At follow-up imaging, CT and MRI findings on the size of the ablation zone have been shown to correlate well with histologic findings [8]. On completion of an ablation, multiphasic contrast-enhanced CT or MRI is commonly used to evaluate ablation success and detect tumor recurrence [2, 3, 12].
The ability to detect residual disease immediately after ablation, allowing retreatment in the same session, could potentially reduce the number of local treatment failures. Sonography would be an ideal technique for this purpose because it is readily available in an interventional suite, easy to perform, and inexpensive. The microbubble contrast agents now available for use with sonography allow identification of enhancement analogous to that seen on contrast-enhanced CT and MRI. Contrast-enhanced sonography can show lesional vascularity in real time with the high resolution afforded by gray-scale sonography [13].
|
|
|
|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Subjects
Between December 2002 and February 2004, 23 nonconsecutive subjects
referred for radiofrequency ablation of hypervascular lesions (clinically,
hepatocellular carcinoma) were recruited into our study. Of these, four failed
to complete the protocol. Two were lost to follow-up; in one, transient severe
back pain developed immediately after injection of contrast material
(incidence of reported adverse events, 1.7%), precluding him from further
contrast injections; and one declined to complete the study for unrelated
reasons. The remaining 19 subjects included 15 men and 4 women (age range,
43-83 years; mean, 64 years). A total of 22 lesions were assessed, ranging
from 1.5 to 3.7 cm in diameter. Three lesions were a localized recurrence of a
previously ablated tumor. Nineteen lesions were hypervascular masses showing
interval growth. Three subjects had two lesions evaluated, two simultaneously
and one at different treatment sessions. Biopsy was attempted for all lesions
before radiofrequency ablation, as is our usual practice. Fourteen of 22
lesions were proven hepatocellular carcinoma. One lesion could not undergo
biopsy because of its position. Seven of 22 lesions had no evidence of
neoplasia at biopsy. When biopsy was not attempted or had negative findings
(8/22), the diagnosis was made from typical imaging characteristics of
hepatocellular carcinoma on multiphasic CT (5/8) or dynamic MRI (3/8) of the
liver, with evidence of enlargement on serial scans (8/8) and increased
-fetoprotein levels (7/8) (range, 4-256 µg/L; mean, 88 µg/L).
Radiofrequency Ablation Technique
In our institution, referral for radiofrequency ablation is considered in
patents who are not a candidate for surgery or need a bridge to
transplantation, have no extrahepatic disease, have three tumors or fewer,
have tumors less than 4 cm in diameter, have an estimated life expectancy
greater than 6 months (and are expected to die from the tumor rather than from
liver failure), and have correctable coagulopathy.
Radiofrequency ablation was performed using the RF 3000 (Boston Scientific) 200-W radiofrequency generator and Levine needle electrodes between 2 and 4 cm in diameter. The probe was positioned under sonographic or CT guidance. The manufacturer's suggested heating algorithm was followed, attaining roll-off for two heating cycles per electrode placement.
Imaging
All recruited subjects were evaluated with enhanced sonography targeted to
the index lesion immediately before and within 1 hr (15-60 min) after the
ablation at our interventional day unit. Within 2 weeks to 1 month after
treatment, the subjects returned for routine follow-up multiphasic CT or
gadolinium-enhanced dynamic MRI of the liver, at which time additional
enhanced sonography of the ablated lesion was performed.
Contrast-Enhanced Sonography
Definity (Bristol-Myers Squibb) is a microbubble contrast agent consisting
of a perfluoropropane gas surrounded by a phospholipid shell. It is relatively
robust, allowing real-time detection without significant agent destruction
when low-mechanical-index techniques are used. Enhanced sonography was
performed immediately before and after the procedure using this contrast agent
from the same vial (1.3 mL). Contrast agent for the preprocedural sonography
was withdrawn slowly, without venting, to extend the half-life of the agent
remaining in the vial. Before the postprocedural sonography, the previously
used vial was vigorously shaken by hand to resuspend the bubbles. Contrast
agent was then withdrawn using a venting needle to minimize barotrauma. Pre-
and postprocedural imaging examinations were separated by up to 3 hours,
without significant detriment to contrast quality. Follow-up sonography was
also performed using a single vial of microbubble contrast agent.
All sonography was performed on ATL 5000 machines (Advanced Technology Laboratories) with a C5-2 probe and contrast-specific software allowing low-mechanical-index imaging using pulse inversion. We used a mechanical index of between 0.1 and 0.2. All images were stored on our PACS as cine loops acquired by performing a smooth manual sweep through the lesion and surrounding liver as the subjects held their breath in deep inspiration.
The sonographic imaging protocol was as follows: The optimal plane for visualization of the index lesion was established and marked on the skin for reproducibility. Contrast-specific imaging parameters were optimized, including positioning of the focal zone deep in relation to the lesion. Baseline cine-loop sweeps were acquired through the index lesion. Contrast agent was prepared according to the manufacturer's instructions and was injected IV (0.3 mL over 5 sec, flushed with 10 mL of normal saline) via a 20-gauge cannula placed in a large arm vein. The liver was observed in real time, and an arterial-phase sweep was acquired as soon as bubbles became visible in the hepatic arteries (15-30 sec). A portal-phase sweep was then acquired immediately after the arterial-phase cine loop had been stored (50-70 sec). Finally, two delayed-phase sweeps were acquired at least 3 min after injection. Preparation and injection of contrast agent were repeated as required, with 10 min allowed between injections. Immediately before injection of additional boluses, residual bubbles were destroyed from the field of the index lesion using a high-mechanical-index sweep [13, 14].
All CT was multiphasic and performed on multislice scanners (GE Healthcare). Up to 200 mL (2 mL/kg) of IV contrast agent (iodixanol, Visipaque 270, Amersham Health) was administered at a rate of 5 mL/sec by power injector. Data sets were obtained through the liver at 30 and 60 sec during the hepatic arterial and portal phases, respectively. These were acquired in deep inspiration, with 5-mm collimation reconstructed at 2.5-mm intervals.
All MRI was performed on 1.5-T magnets (Signa, GE Healthcare). Gadolinium-enhanced data sets of the liver were acquired dynamically as 3D spoiled gradient-echo volumes at 30, 60, 90, and 300 sec, corresponding to the arterial, portal, equilibrium, and delayed phases, respectively.
Imaging was stored on our PACS and reviewed at dedicated soft-copy reporting stations.
Image Analysis
Data were analyzed in a masked fashion by three radiologists experienced in
the use of sonographic contrast agents and in the interpretation of CT and MRI
liver scans.
Postprocedural sonography was compared with preprocedural CT or MRI and sonography. Technical adequacy, ablation zone targeting, and identification of residual disease were assessed by each reviewer independently, and the results were analyzed by consensus. Follow-up sonography and CT or MRI (our reference standard) were then independently assessed in the same way. Targeting was assessed by directly comparing the position of the ablation zone on the sonographic portal-phase sweep with the preprocedural images. Residual disease was identified by discrete, nodular, noncircumferential arterial-phase enhancement at the ablation margin or adjacent to the ablation zone.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Assessment of Targeting
Follow-up CT or MRI found that targeting had been accurate in 16 of 22
subjects. Reviewers recorded accurate targeting for the postprocedural
sonography in 17 subjects and for follow-up sonography in 16 subjects.
Postprocedural sonography predicted the CT or MRI results in 76% of subjects
(16/21) (sensitivity, 88%; specificity, 40%; positive predictive value [PPV],
82%; negative predictive value [NPV], 50%). The results of follow-up
sonography agreed with those of CT or MRI in 82% of subjects (18/22)
(sensitivity, 88%; specificity, 67%; PPV, 88%; NPV, 67%).
Identification of Residual Disease
Follow-up CT or MRI detected residual disease in six of 22 subjects.
Residual disease ranged from 9 to 25 mm in diameter (mean, 16 mm). Sonography
detected residual disease in two subjects postprocedurally and in six at
follow-up. Postprocedural sonography predicted the CT or MRI results in 81% of
subjects (17/21) (sensitivity, 40%; specificity, 94%; PPV, 66%; NPV, 83%). The
findings of follow-up sonography agreed with those of CT or MRI in 91% of
subjects (20/22) (sensitivity, 83%; specificity, 94%; PPV, 83%; NPV, 94%). If
our masked interpretation showed residual disease on CT but not on sonography,
our reviewers retrospectively reviewed the sonography studies to decide if
residual disease was depicted but had not been recognized. In all but one
case, residual disease was detected retrospectively. Allowing for this case,
adjusted results show that postprocedural sonography predicted the CT or MRI
results in 90% (19/21) of subjects (sensitivity, 75%; specificity, 94%; PPV,
75%; NPV, 94%) and that the findings of follow-up sonography agreed with those
of CT or MRI in 95% of subjects (21/22) (sensitivity, 100%; specificity, 94%;
PPV, 86%; NPV, 94%).
Validation of Reference Standard
Follow-up information was available for 17 subjects (17/22 lesions). Two
subjects received subsequent liver transplants allowing pathologic examination
of their explanted livers. One was transplanted within 1 month of treatment
and revealed residual tumor measuring 1.5 cm in diameter, not detected by
imaging. The other, transplanted within 3 months, confirmed the presence of
residual disease. Follow-up CT or MRI was obtained for 15 subjects and was
performed 3-13 months (mean, 6 months) after the initial radiofrequency
ablation. In one subject, enlarging residual disease not previously detected
became evident. Our reference standard thus had false-negative results in 9%
(2/22) of examinations.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Sonography would be an ideal technique for real-time monitoring of radiofrequency procedures. It is the primary method of guiding radiofrequency ablation and could thus be used quickly and efficiently without the patient's having to be moved. Conventional color and power Doppler sonography, however, are too insensitive to detect tumor vascularity reliably [19]. The development of sonographic contrast agents and advances in contrast detection technology offer the promise of a convenient and reliable method for distinguishing vascularized viable tumor from ablated tissue [13, 20-22]. Early series evaluating ablation success using color and power Doppler sonography applied a first-generation contrast agent (SH U 508A, Levovist, Schering) and showed improved sensitivity over conventional sonography for detecting residual disease. In this series, sonography had sensitivities ranging from 33% to 100% [19, 22-24]. SH U 508A comprises relatively fragile bubbles and is best imaged with a high-mechanical-index technique whereby the agent is destroyed as it is imaged. The strength of SH U 508A is its unique enhancement of the liver in the postvascular phase; this ability has been shown to improve the detection of liver metastases [25]. However, the vascular-phase imaging required to detect residual hypervascular tumor after radiofrequency ablation is performed weakly with SH U 508A. It was with the introduction of the second-generation perfluorocarbon agents, which can be scanned using a low mechanical index, that the potential for microbubble contrast agents to monitor vascular changes with radiofrequency ablation was realized [26].
The introduction of more advanced contrast-specific imaging methods, such as pulse inversion, has resulted in sensitivities of 83.3-95% [27, 28].
As with CT and MRI, however, interpretation of images obtained immediately after the procedure is particularly challenging. At the margin of the ablation zone, hypervascularity produced by a localized tissue response or arteriovenous shunting may be difficult to differentiate from residual hypervascular tumor. Reactive hyperemia is usually uniform in thickness and surrounds the ablated lesion in a rindlike fashion. Both reactive hyperemia and perfusion anomalies often are occult in portal-phase imaging because their early prominent enhancement is matched by the surrounding liver parenchyma (Figs. 1A, 1B, 1C, and 1D). Residual tumor, however, shows focal or irregular peripheral enhancement in the arterial phase. On portal venous and delayed-phase imaging, residual tumor may also be low in attenuation or intensity because of washout of contrast agent [10, 11] (Figs. 2A, 2B, 2C, and 2D). Portal and delayed-phase imaging depicts the position and size of the ablation zone as a geographic unenhanced region. In two of the six subjects in whom residual disease was identified, we saw arterial-phase enhancement with washout in the portal phase, enabling a more confident diagnosis with sonography. In four subjects, washout was not detected on sonography or on CT or MRI. In these subjects, the residual disease originated from the periphery of the tumor. Washout was not seen before ablation in these regions; instead, these areas showed pseudo-capsular enhancement, a finding well described for hepatocellular carcinoma [29].
Our study confirms the results of others who have reported a high concordance between follow-up CT or MRI and follow-up enhanced sonography in the detection of residual disease. The main drawback of enhanced sonography occurs in scanning performed immediately after the procedure and arises from technical difficulties secondary to procedure-related artifacts such as gas and uncooperative patients. Immediately after the procedure, patients often are less able to cooperate fully with the sonographer, because they are often in pain or still under the effects of conscious sedation. Achieving an adequate inspiration for visualization of the lesion may be impossible. Gas-related artifacts secondary to cavitation within the heated tissues and gas introduced at the time of the procedure may also contribute to a technically inadequate study. These artifacts may persist for 15-180 min [30]. In our experience, a delay of 20-40 min was adequate for performing a diagnostic study in all but one subject. Technical problems were minimal after the first 12-24 hr, enabling the true sensitivity and specificity of the technique to be revealed (Table 1).
|
Imaging immediately after the procedure is potentially the most clinically relevant application of enhanced sonography in radiofrequency ablation. To our knowledge, Solbiati et al. [31] and we have the only series evaluating the utility of enhanced sonography in the first hour after ablation. The study of Solbiati et al. included both hyper- and hypovascular tumors, and direct comparison thus was not possible. Though the sensitivity of enhanced sonography, 60%, is low, in our series the 94% specificity is clinically useful. In selected patients in whom residual disease is detected, enhanced sonography may facilitate completion of radiofrequency ablation at the time of initial therapy.
Only one patient in our study had false-positive interpretations in the postprocedural and follow-up sonography examinations. In retrospect, this false-positivity was due to misinterpretation of a large marginal vessel. We had one technically inadequate postprocedural study. Imaging was performed too soon after a prolonged procedure, and intralesional gas precluded visualization of the ablation zone. Residual disease was masked and was subsequently detected on follow-up sonography. False-negative results occurred in three postprocedural sonography studies and one follow-up study. In three cases, the residual disease could be identified in retrospect, illustrating how difficult these studies are to interpret, especially in a masked setting (Figs. 3A, 3B, 3C, 3D, 3E, 4A, 4B, 4C, and 4D). We believe that incorrect timing of the sweeps may have affected the results in the fourth case. Hypervascular masses may be only transiently visible, compared with surrounding liver, and thus may be not appreciated if the sweep is performed too early or too late.
|
|
|
|
|
|
|
|
|
Preprocedural sonography clearly depicted the hypervascular lesions to be treated in all our subjects. These masses frequently are visible only on contrast-enhanced CT or MRI. Accurate targeting for deploying the tines of the radiofrequency probe is an additional potential benefit of introducing enhanced sonography into the interventional suite. In one of our subjects, inaccurate targeting appeared to be responsible for residual disease that lay just outside the ablation zone (Figs. 2A, 2B, 2C, and 2D).
We used enhanced sonography as a targeted study for the purposes of our trial. In 23% (5/22) of our patients, new foci of disease appeared in the same lobe of the liver on subsequent imaging up to 11 months after radiofrequency ablation. The role of enhanced sonography in detecting these foci is promising, but it is not yet a generally accepted clinical tool [26].
This study had limitations. Our reference standard (CT or MRI) has been shown to be of limited accuracy. In 14% (2/14) of the patients described by Solbiati et al. [7] and 7% (3/45) of the patients described by Choi et al. [19], marginal recurrence within 7 months was reported in areas of previous ablation when initial CT had negative results. In 9% (2/22) of our subjects, residual disease was detected within 3 months after negative results had been obtained at 2 weeks. One case of residual disease was identified at pathologic examination of the explanted liver and the other on follow-up CT. In five patients, no imaging follow-up was yet available at our institution to validate the CT results.
Though large, prospective studies are needed to further evaluate the utility of enhanced sonography in assessing the success of radiofrequency ablation, our results concur with those of others supporting clinical application of this technique when procedure-related artifacts are minimal.
In conclusion, the use of enhanced sonography for immediate postprocedural evaluation of the success of radiofrequency ablation has a high positive predictive value when residual disease is detected. In such cases, enhanced sonography has the potential to guide completion of radiofrequency ablation at the time of initial therapy. These benefits, however, must be balanced against the technical difficulties and imaging artifacts that may be encountered.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  B. Lanka, H.-J. Jang, T. K. Kim, P. N. Burns, and S. R. Wilson Impact of Contrast-Enhanced Ultrasonography in a Tertiary Clinical Practice J. Ultrasound Med., December 1, 2007; 26(12): 1703 - 1714. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M.-H. Chen, W. Wu, W. Yang, Y. Dai, W. Gao, S.-S. Yin, and K. Yan The Use of Contrast-Enhanced Ultrasonography in the Selection of Patients With Hepatocellular Carcinoma for Radio Frequency Ablation Therapy J. Ultrasound Med., August 1, 2007; 26(8): 1055 - 1063. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
