Percutaneous RFA is a minimally invasive technique that uses thermal energy to damage cellular proteins, enzymes, and nucleic acids in tissues.1,2 During RFA, an electrode inserted into the target tissue is used to induce coagulation necrosis in the central zone and sublethal hyperthermia injury in the transitional zone of the lesion.3 The necrotized tissue hardens, becomes surrounded by a thick fibrous capsule, and decreases in volume at a variable rate over a period of 6 to 12 months.4,5 In humans, RFA has been used for nonsurgical treatment of lesions in solid organs, including the liver, kidneys, lungs, and bones, with a particular focus on treatment of cancer.1,2 Radiofrequency ablation can be performed repeatedly on solitary or multifocal hepatic lesions and is associated with a lower cost, shorter procedural time, and lower morbidity and mortality rates, compared with those for surgical treatments. In particular, RFA has been applied in patients with metastases to the liver involving a maximum of 4 or 5 small (< 4 cm) lesions.6 However, in veterinary medicine, the application of RFA for cancer treatment has been limited to ocular squamous cell carcinoma in cattle and horses and thyroid gland tumors in dogs and cats because of a lack of baseline information regarding suitable types of neoplasms for RFA or appropriate procedures for assessing lesion size.7–10
Achieving an appropriate margin is essential for complete ablation and reduces the risk of local tumor recurrence. Recommendations for humans state that the targeted tumor should be ablated, including an adjacent margin of normal parenchymal tissue of at least 5 mm but ideally 10 mm.11 After RFA, the tissue margins should be evaluated for the presence of viable residual tumor to allow consideration of retreatment or additional procedures that could improve patient outcomes.12 Therefore, an assessment of the extent of targeted tumor obliteration or the amount of surrounding ablated tissue is crucial to treatment success. Unnecessary biopsies and retreatment (or additional treatments) can be prevented by the use of accurate diagnostic imaging of the ablation zone and minimizing false-positive diagnoses of residual tumor. This result can be achieved by examining specific postablation changes by use of various diagnostic modalities. The feasibility and reliability of various diagnostic imaging techniques, including CT, MRI, positron-emission tomography-CT, and ultrasonography, for evaluation of RFA efficacy have been investigated in human studies.13–16 Color flow Doppler ultrasonography and contrast-enhanced CT can be used to evaluate changes in ablation zone perfusion.17 Ultrasonography allows real-time monitoring of needle insertion into a lesion and reduction of blood flow during RFA. However, nitrogen microbubbles that arise during coagulation induce an initial hyperechoic appearance in the ablation zone (ie, echogenic cloud), and resultant acoustic shadowing within the ablation zone hinders visibility of the ablation margin.4,17 The echogenic cloud generally persists for 15 minutes to 6 hours after RFA.4
Computed tomography-based real-time monitoring of needle insertion during RFA is technically demanding. However, CT can be used to accurately determine the size and extent of the ablation lesion and is useful for assessing tumor remnants, tumor recurrence, and complications in human patients after RFA. On unenhanced CT images obtained immediately after RFA, the ablation zone appears hypoattenuated or heterogeneously hyperattenuated because of coagulation necrosis and hemorrhage, although visible changes are not evident in some patients.4,5 On contrast CT images, the ablation zone is clearly demarcated, and lesion size measured in the portal venous phase correlates closely with gross lesion size.18 Adequate treatment and the absence of viable tumor can be determined on the basis of a lack of enhancement in the ablation zone.19 A transient hyperemic area is induced by thermal injury, an inflammatory reaction, and granulation tissue at the periablational region and is manifested as a poorly defined thin rim of enhancement during the arterial phase or, less commonly, the portal venous phase of contrast CT imaging.5 If residual tumor is present or a tumor recurs after RFA, irregular peripheral enhancement of the ablation margin can be observed on contrast CT images. Therefore, the hyperemic periablational area should be carefully evaluated to avoid confusion with pathological enhancement. Repeated CT evaluation to monitor the ablation zone has certain limitations with respect to veterinary applications because of the cost, radiation exposure, and necessity for heavy sedation or anesthesia of dogs and cats.
Elastography and CEUS have been used to evaluate RFA lesions in humans in an attempt to overcome or minimize the limitations of ultrasonography and contrast CT. Contrast-enhanced ultrasonography reportedly is an efficient tool for immediate evaluation of RFA efficacy. However, CEUS is not free from limitations intrinsic to ultrasonographic imaging, such as low image quality and interference with ablation zone visibility caused by nitrogen microbubbles.4
Strain elastography can be used to quantify and display organ tissue stiffness during and after RFA. Ablated organ tissue becomes dehydrated, and protein denaturation occurs. Therefore, relative stiffness of the tissue will increase relative to that of the unablated parenchyma.20,21 In humans, elastography can be used to estimate real-time differences in tissue stiffness, which are represented on a color map that depicts the ablation zone boundary in accordance with the difference in stiffness between the ablated and nonablated zones. Unfortunately, strain elastography is limited to predictions of the ablation zone size and is a highly operator-dependent technique.22 In dogs, the liver is located within the rib arch; thus, it is difficult to obtain the optimal ultrasonographic acoustic window necessary to obtain images of the hepatic lobe located in the far field. In particular, a manual induction of strain for hepatic lesion imaging might not be effective for strain elastography in the context of RFA monitoring.
The study reported here was conducted to evaluate the feasibility and reliability of various diagnostic imaging techniques, including color Doppler ultrasonography, CEUS, strain elastography, and CT, for assessing lesion margins and to determine RFA efficacy in the liver of healthy dogs. This study was performed to test the hypothesis that CT and CEUS are the most feasible and reliable methods for evaluating the ablation area because both can be used to detect changes in vascular marking and parenchymal perfusion after RFA. The ablation zone ultimately was examined histologically to allow comparison of macroscopic and imaging findings. In addition, the appearance of the ablation zone immediately after and at 3 days after RFA was characterized by use of each imaging modality.
Materials and Methods
Animals
Six purpose-bred Beagles (5 sexually intact males and 1 sexually intact female) were included in the study. Dogs were 3 to 5 years old; mean ± SD body weight was 9.1 ± 1.4 kg. All dogs were healthy as determined on the basis of results for a physical examination, measurement of systemic arterial blood pressure, a CBC, serum biochemical and electrolyte analyses, urinalysis, and abdominal radiography and ultrasonography. Activated partial thromboplastin timea and prothrombin timea were also measured. Dogs were housed individually in cages and fed a commercial dry food; water was available ad libitum. The dogs were cared for in accordance with published guidelines,23 and all experiments were approved by the Institutional Animal Care and Use Committee at Chonnam National University (CNU IACUC-YB-2015-27).
RFA
Food was withheld from all dogs for 12 hours. Dogs then were anesthetized with a combination of medetomidineb (0.05 mg/kg, IM) and zolazepam-tiletaminec (1.25 mg/kg, IM). Tramadol hydrochlorided (2 mg/kg, IV) and cefazolin sodiume (20 mg/kg, IV) were also administered. Areas over the scapulas and hip joints of each dog were shaved, and grounding pads were affixed in the shaved areas. Dogs were placed in dorsal or lateral recumbency for conventional ultrasonographyf (with a linear probeg) of the liver. The subcostal approach was used, and > 2 accessible regions of the peripheral portion of the liver were selected for RFA. Radiofrequency ablation was performed by use of an impedance-based deviceh and a 17-gauge 20-mm slim expandable needlei with an array of 8 expandable monopolar electrodes. The generator used to power the impedance-based RFA device was operated in the constant-voltage mode and automatically controlled by an impedance feedback mechanism. By use of real-time ultrasonographic guidance, the end of the needle tip was positioned at a depth of 3 cm (from the skin), and the array was deployed from the needle tip into the liver. The initial ablation power was 30 W, which was increased at a rate of 10 W/min until the maximum impedance (determined by use of manufacturer-recommended algorithms) was reached. At that point, ablation was discontinued. After a 30-second pause, a second ablation was performed without repositioning of the electrode. For the second RFA, the initial ablation power was 50% of the power reached at the maximum impedance during the first RFA and was increased at a rate of 10 W/min. The needle was then slowly retracted. Total procedural time, including the first and second ablations, was < 5 minutes.
Experimental procedures
Radiofrequency ablation was performed on day 0. After RFA was completed, all dogs received cefazolin sodiume (20 mg/kg) and tramadol hydrochloridej (2 mg/kg) orally twice daily. The ablation zone was evaluated by use of conventional B-mode ultrasonography, strain elastography, and CEUS immediately after (time 0), 30 to 60 minutes after, and 3 days after RFA (day 3). The lesion image was obtained by use of CT at 30 minutes and 3 days after RFA.
Diagnostic imaging and image analysis
The shape, margin, and size of ablation lesions were evaluated by use of conventional ultrasonography, elastography, and CEUS. Evaluations were performed in a single plane that represented the center of the lesion.
For conventional ultrasonography, lesion size was estimated by comparing echogenicity of the ablation lesion with that of the unablated liver parenchyma. Size was quantitatively measured in the lesion area by manual tracing with integrated software.f
Strain elastographyk was performed with a 13-MHz linear transducer by 1 investigator (JC). Strain was induced by manual compression with the transducer over a transverse or oblique plane of the liver while keeping the transducer oriented perpendicular to the liver and abdominal wall. Manual compression speed and cycle were verified by use of a real-time strain graph. Elastography images were displaced with a B-mode image in the dual-screen mode. Tissue stiffness was evaluated on elastography images by use of a color map on which the hardest areas appeared blue and the softest areas appeared red. The best-fit B-mode and elastography images were selected, and elastography images were evaluated by 1 reviewer (SM), who used integrated softwarek to allow manual tracing of the ablation lesion boundaries. Lesion hardness and size in the ablated area were evaluated on elastography images.
Contrast-enhanced ultrasonography was performed with the following settings: transmitted energy, 7%; mechanical index, 0.07; and acquisition speed in the extended pure harmonic mode, 15 frames/s. A 0.5-mL bolus of contrast agentl was administered into a cephalic vein via a 3-way stopcock and 20-gauge catheter; contrast agent was mixed by shaking before each injection. The line was then immediately flushed with a 5-mL bolus of saline (0.9% NaCl) solution. Simultaneously, dynamic sequences were acquired for up to 110 seconds. Images for CEUS were obtained within 2 hours after injection of the contrast medium. The CEUS images of the ablated region were quantitatively analyzed to determine lesion size by manual tracing with integrated software.f
All CT images were obtained by use of the following settings: 130 mA, 130 kVp, 1-mm slice thickness, and 0.7 pitch. Iohexolm (880 mg of I/kg) was injected through a 22-gauge catheter into a cephalic vein at a rate of 3 mL/s with an automated injector.n Then, 3 phasic contrast images of the liver were obtained (arterial, portal venous, and delay phases). Scan delay time for the 3 phasic CT images was determined for each dog by use of dynamic CT scanning. The lesion area on CT images was measured during the venous phase. Lesion size as determined on the basis of CT image attenuation and contrast enhancement was assessed with manual tracing by use of a picture archiving communication system.o A lesion was defined as the hypoattenuated area relative to the untreated liver parenchyma or the area encircled by the inner surface of the contrast-enhanced rim. Lesion size was measured from the inner surfaces of the contrast-enhanced rim. Contrast enhancement in the ablation zone and normal liver parenchyma were quantified (number of HU), and regions of interest that had a uniform absence of vessels and artifacts were traced.
Postmortem examination
Three arbitrarily selected dogs were euthanized shortly after RFA by IV injection of propofolp and potassium chloride.q The remaining 3 dogs were euthanized on day 3 immediately after conventional ultrasonography, strain elastography, CEUS, and CT were performed. Dogs were necropsied; gross and histologic examinations of the lesions were conducted.
Each ablation lesion was completely resected along with the adjacent parenchyma, and additional liver samples were obtained from untreated liver parenchyma. Both types of samples were fixed in neutral-buffered 10% formalin and embedded in paraffin for histologic examination. The gross shape and color of lesions were assessed, and areas were measured on a computer equipped with imaging software.r Then, 3-μm-thick sections were stained with H&E stain, and lesions were photographed through a microscope. Slides were evaluated by 3 pathologists (K-OC, J-GP, and MMA), who assessed the sections for areas of necrosis, regeneration, scar formation, and inflammatory reaction.
Statistical analysis
Data were expressed as mean ± SD, when applicable. Differences among lesions were evaluated by use of the Mann-Whitney U test. The imagingestimated ablation areas and gross lesion appearances measured on days 0 and 3 were also compared by use of the Mann-Whitney U test. Spearman correlation analysis was used to compare lesion sizes determined by use of each imaging modality with actual tissue measurements. Significance was defined as values of P < 0.05. All statistical analyses were performed with commercially available software.s
Results
Thirteen lesions were created by use of RFA in the 6 dogs. One lesion was excluded from the analyses because an intact sample for histologic examination could not be obtained. Therefore, 12 lesions from 6 dogs were included in the analyses. There were 5 and 7 lesions analyzed for days 0 and 3, respectively.
During liver RFA, the electrode and deployed arrays were clearly observed with real-time ultrasonography, and no major adverse events related to RFA were detected. However, for 2 dogs, a small peritoneal effusion developed around the ablation zone. This effusion temporarily hindered the strain induction needed for elastography, but the effusion resolved within 3 days. On RFA initiation, multiple hyperechoic foci and microbubbles developed within the ablation zone. Therefore, echogenicity of the ablation zone increased diffusely in a circular shape at the conclusion of RFA. Appearance of the ablation zone was similar in all dogs during RFA.
Immediately after RFA, an echogenic cloud with acoustic shadowing and a poorly defined margin formed rapidly and persisted for 30 to 45 minutes on conventional ultrasonographs (Figure 1). Over time, demarcation of the lesion tended to improve, which facilitated lesion measurements, and those lesions typically were larger than lesions observed on images obtained immediately after RFA. Lesions did not differ significantly with regard to size. By day 3, the ablation zone was mildly hypoechoic and had a hyperechoic center. In 3 lesions, an echogenic outer rim that encompassed the ablation zone was evident. Size of the ablation zone did not differ significantly between days 0 and 3. Lesion margins were easily traced.
Ultrasonographic images of a lesion in the liver of a dog immediately after (time 0; A), 30 to 60 minutes after (B), and 3 days after (C) RFA. A—Notice the hyperechoic appearance in the ablation zone (ie, an echogenic cloud [arrowhead]) and acoustic shadowing (arrows). B—There is a mildly hypoechoic lesion (asterisk) and a more distinct lesion (curved arrows). Residual central echogenicity is also visible (dotted cross), which has a size of 1.6 cm (blue dotted line) by 1.4 cm (yellow dotted line). C—The lesion is primarily hypoechoic and demarcated by a hyperechoic rim (arrowheads).
Citation: American Journal of Veterinary Research 78, 3; 10.2460/ajvr.78.3.295
Strain elastography performed immediately after RFA revealed a mosaic pattern within the ablation lesions that predominantly comprised hard tissue (blue on the color map) surrounded by unaffected liver parenchyma (green on the color map; Figure 2). Good lateral boundaries for the ablation zone were observed, but no signal was detected in the deep margins of most lesions because of acoustic shadowing. Although the strain images became more demarcated with distinct distal boundaries by 30 to 60 minutes after RFA, it was difficult to achieve adequate strain and obtain consistent tissue strain data for 2 lesions because of a small adjacent peritoneal effusion that formed during RFA. Both of these peritoneal effusions had resolved by day 3, at which time sufficient strain could be induced in all lesions, except for 1 lesion for which the proper angle could not be obtained to induce strain. Margin of the ablation zone could be more clearly discriminated in images obtained later than in those obtained immediately and 30 to 60 minutes after RFA.
Strain elastographic images (color map) of a lesion in the liver of a dog immediately after (time 0; A), 30 to 60 minutes after (B), and 3 days after (C) RFA. A—The ablation lesion is hard (represented in blue [arrowheads]), and shadowing (arrows) below the lesion is present without any signal. B and C—The lesion has a mosaic pattern with a dominant blue area (arrowheads) and a clear margin distinguishing it from the surrounding hepatic parenchyma (represented in green). The distal boundary of the lesion is apparent.
Citation: American Journal of Veterinary Research 78, 3; 10.2460/ajvr.78.3.295
The CEUS images obtained immediately after RFA featured an absence of blood signals and the presence of intense hyperechoic foci within the ablation zone (Figure 3). Demarcation between the ablation zone and surrounding hepatic parenchyma was observed with enhancement. However, acoustic shadowing from the center of the ablation zone interrupted the distinct deep margin of the ablation zone, which made it difficult to obtain measurements of lesion size. However, some ablation lesions with poorly defined margins during conventional ultrasonography were more easily delineated during CEUS. The CEUS images obtained 30 to 60 minutes after RFA (ie, when the echogenic cloud began to disappear) provided clearer depictions of lesions, which appeared as unenhanced defects. By day 3, the microbubbles had been resorbed, and the hypoechoic ablation zone had a sufficiently clear margin that the lesion size could be measured without acoustic shadowing.
Contrast-enhanced ultrasonographic images of a lesion in the liver of a dog immediately after (time 0; A), 30 to 60 minutes after (B), and 3 days after (C) RFA. A—An unenhanced area with a good lateral margin (arrowheads) is visible, but no distal boundary is clearly evident (arrows). B—A relatively clear margin of the ablation zone (arrowheads), including the distal boundary, is visible. C—The image provides excellent demarcation of the lesion (arrowheads).
Citation: American Journal of Veterinary Research 78, 3; 10.2460/ajvr.78.3.295
Examination of unenhanced CT images obtained immediately after RFA revealed that most lesions had no discernible changes, except for slight hypoattenuation (Figure 4). Tiny air foci were observed within the ablation zones of 6 of 12 examined lesions. Injection of contrast agent revealed homogenously nonenhancing lesions in the arterial, venous, and parenchymal phases. The lesions were surrounded by 1- to 3-mm-thick contrast-enhanced rims during both the arterial and venous phases. The enhanced rims were delineated in the venous phase but were not discernible from the enhanced surrounding parenchyma in the delayed phase. Maximum contrast enhancement of the lesion did not exceed 10 HU for any phase (Table 1). Ablation lesions were slightly more hypoattenuated on unenhanced CT images obtained 3 days after ablation, compared with CT images obtained at 30 minutes after RFA. Gas foci were not identified. Contrast-enhanced CT images of lesions, including images of the nonenhanced ablation zone and contrast-enhanced rim, were similar to those obtained on day 0.
Triphasic contrast-enhanced CT images of the liver of a dog obtained 30 minutes after (A through D) and 3 days after (E through H) RFA. A and E—The lesion has barely noticeable hypoattenuation and a poorly defined margin in precontrast transverse images. B and F—The slightly enhanced liver parenchyma surrounds the consistently unenhanced lesion. C and G—During the venous phase, lesion demarcation reaches a maximum and periablational rim enhancement is visible. D and H—The lesion margin returns to a poorly defined state during the delay phase. Contrast enhancement of the lesion did not exceed 10 HU in any phase.
Citation: American Journal of Veterinary Research 78, 3; 10.2460/ajvr.78.3.295
Mean ± SD attenuation of the ablation zone and normal hepatic parenchyma on unenhanced and contrast-enhanced CT images obtained during the venous phase after RFA.
| Ablation zone | Normal parenchyma | |||
|---|---|---|---|---|
| Time* | Unenhanced CT Enhanced CT | Unenhanced CT | Enhanced CT | |
| Day 0 (n = 6) | 62.4 ± 3.8 | 65.8 ± 3.3 | 62.7 ± 5.9 | 125.4 ± 21.3 |
| Day 3 (n = 3) | 45.1 ± 4.1 | 48.9 ± 4.2 | 62.3 ± 4.4 | 129.9 ± 14.8 |
Attenuation was measured in HU.
Day of RFA was designated as day 0; 3 dogs were euthanized after diagnostic imaging on day 0, and the remaining 3 dogs were euthanized after diagnostic imaging on day 3.
Gross examination of the ablation lesions typically revealed a circular white region with a clear margin. Lesions were firm and slightly more compressed than the surrounding hepatic parenchyma. Although no other abnormalities were observed, the abdominal walls of some dogs contained focally thickened areas that likely indicated the point of electrode insertion. Gross postmortem examination revealed that size of the lesions could be easily measured and did not differ between dogs euthanized on day 0 or day 3. Mean size of ablation lesions measured with each imaging modality and via gross examination was summarized (Table 2).
Mean ± SD area (cm2) of ablation lesions measured during necropsy and by use of various imaging modalities.
| Time* | Necropsy | Ultrasonography | Elastography | CEUS | Enhanced CT |
|---|---|---|---|---|---|
| Day 0 (n = 7) | 2.40 ± 0.54 | 2.09 ± 0.34 | 1.68 ± 0.52† | 1.97 ± 0.38 | 2.14 ± 0.28 |
| Day 3 (n = 5) | 2.41 ± 0.34 | 1.92 ± 0.41 | 1.57 ± 0.21† | 1.66 ± 0.48 | 2.24 ± 0.40 |
| Total (n = 12) | 2.41 ± 0.45 | 2.02 ± 0.36† | 1.63 ± 0.40† | 1.84 ± 0.44† | 2.19 ± 0.32 |
Within a row, value differs significantly (P < 0.05) from the value for the gross lesion area measured during necropsy.
See Table 1 for remainder of key.
Histologic examination revealed that the ablation lesions for dogs euthanized shortly after RFA could be divided into 3 zones (ie, central, transitional, and outer zones; Figure 5). Examination at low magnification revealed that the central zone was severely pale and necrotic, and examination at high magnification revealed that each hepatocyte was necrotic and pyknotic. Severely necrotic and pyknotic hepatocytes with diffuse congestion and hemorrhage were present in the transitional zone. Ablation lesions for dogs euthanized on day 3 also had 3 distinct zones. Coagulation necrosis and pyknosis were observed in the central and transitional zones, and the transitional zone also had severe diffuse congestion and hemorrhage. A heavy infiltration of fibroblasts and lymphoid cells was observed in the marginal area between the transitional and outer zones, which indicated the beginning of granulation.
Photographs of tissue samples obtained from the liver of dogs euthanized on the day of RFA (day 0; A) and 3 days after RFA (E) and photomicrographs of tissue sections for the lesions on day 0 (B through D) and day 3 (F through H). A light central region with a well-defined margin surrounded by darker brown tissue is visible after RFA (A and E). The area from which the tissue sections for histologic examination were obtained is indicated (dashed square). Histologic examination reveals 3 zones in the lesion. The central zone (B and F) is extremely pale and has necrosis with pyknosis. The transition zone (C and G) has diffuse congestion with hemorrhage. The transition zone (G) is separated from viable liver tissue in the outer zone (D and H) by a focal granulocytic infiltration of lymphoid cells and fibroblasts (asterisk). H&E stain; bars = 200 μm for panels B through D and H through F.
Citation: American Journal of Veterinary Research 78, 3; 10.2460/ajvr.78.3.295
The central and transitional zones corresponded with the nonenhanced zone on contrast-enhanced CT and CEUS images of the examined lesions (Figure 6). Size of the actual and enhanced CT-measured lesions did not differ significantly; however, use of elastography and CEUS underestimated lesion size by 31% and 22%, respectively.
Photographs of a sample of liver tissue obtained 3 days after RFA (A) and images of that area of the liver obtained by use of conventional ultrasonography (B), strain elastography (C), CEUS (D), and contrast-enhanced CT during the venous phase (E). The same margin of the ablation lesion is indicated in each panel (arrow). The nonenhancing zones on CT and CEUS images correspond with the central white zone and brown transitional zone.
Citation: American Journal of Veterinary Research 78, 3; 10.2460/ajvr.78.3.295
Spearman correlation coefficients for comparisons between lesion measurements obtained with each imaging modality and the actual lesion size were determined. The correlation coefficient was highest for enhanced CT, which had a significant strong correlation (r = 0.724) with gross lesion size measured during necropsy. Conventional ultrasonography (r = 0.536) and CEUS (r = 0.524) both had significant moderate correlations with gross lesion size measured during necropsy. Elastography had a nonsignificant correlation (r = 0.262).
Discussion
Radiofrequency ablation has not been widely used in veterinary practice because of a lack of basic data on topics such as assessment of lesion size. Accurate determination of lesion size is crucial for successful RFA treatment because recommendations for RFA in humans indicate that sufficient margins (at least 5 mm or ideally 10 mm) are needed to prevent residual tumor or tumor recurrence.11 In the study reported here, acute changes in liver parenchyma of healthy dogs after RFA were characterized by use of conventional ultrasonography, strain elastography, CEUS, and CT, and the feasibility of each imaging modality for monitoring ablation lesion changes and measuring lesion size was investigated by comparing imaging measurements with actual lesion sizes (measured during gross necropsy). We hypothesized that CT and CEUS could be used to most accurately determine ablation size on the basis of changes in parenchymal perfusion after RFA; however, CT, and not CEUS, was the most reliable modality with respect to RFA efficacy.
Marked microbubble development during RFA was visible with real-time ultrasonography. These microbubbles persisted for 30 to 45 minutes and induced acoustic shadowing that limited the delineation of deep margins of the ablation zone. Clear margins of the ablation zone were observable with ultrasonography 30 to 60 minutes after ablation and on day 3. Efforts have been made in clinical settings to develop the use of several modalities, including CEUS and elastography, to overcome ultrasonographic limitations and detect pathophysiologic properties of ablation lesions. Limitations of conventional ultrasonography, such as an echogenic cloud and acoustic shadowing, also affected the ultrasound-based modalities of strain elastography and CEUS to a certain degree. However, distal interference was cleared relatively more rapidly with elastography and CEUS than with conventional ultrasonography in the present study.
The imaging modality used to determine RFA efficacy should accurately predict the ablation lesion size and clearly reveal the lesion margin. In the present study, lesion measurements with CT were best correlated with the actual lesion size, whereas conventional ultrasonography, CEUS, and elastography tended to underestimate actual lesion size. Investigators of a previous study21 found good correlation between strain elastography- and CT-based measurements of ablation lesion size. However, in other studies,22,24 investigators have found that distal acoustic shadowing degraded the elastography signal (although area estimations remained possible via extrapolation) and that elastography underestimated the ablation lesion size. In the present study, lesion areas estimated by use of elastography were significantly different from the actual lesion area, which resulted in a poor correlation. This discrepancy with elastography data is concerning with respect to determining RFA efficacy and might have been attributable to the high operator dependency of this technique and relative evaluation principle within the region of interest.25 The anatomic characteristics of dogs result in a small acoustic window, and this might have reduced the accuracy of elastography in the present study more than in previous studies of humans, pigs, and cattle. Moreover, strain could not be sufficiently induced in the presence of peritoneal effusion around the ablation lesion.
Investigators have found that CEUS (even real-time CEUS) can precisely distinguish between viable tissues and necrotic lesions. Furthermore, lesion measurements obtained with CEUS correspond with actual lesion sizes measured histologically.26,27 However, lesion area measurements on days 0 and 3 obtained by use of CEUS in the present study underestimated the actual measurements by approximately 22%. In a recent study,26 investigators found that ablation lesions undergo an initial rapid expansion within the first 2 hours after creation. This expansion is followed by a slow and static increase in size until lesions reach their maximum size between 2 hours and 2 days after RFA. In the present study, the lesion area measured on day 0 (when the lesion was undergoing initial rapid expansion) differed significantly from the gross lesion size because of the time interval between CEUS and gross examination of the ablation lesion. However, on day 3 (after the lesion underwent a static size change), there was no significant difference between lesion size determined with CEUS or during gross examination.
Ablation zone images obtained with CT had similar features at 30 minutes and 3 days after RFA. Unenhanced CT did not reveal remarkable changes in ablation lesions, although the ablation zone became more homogeneously hypoattenuating over time. Use of contrast medium revealed a demarcated ablation zone, which was visible as an unenhanced area that differed from the surrounding hepatic parenchyma. On CT images obtained 30 minutes after RFA, hyperattenuating tiny air bubbles were evident in some lesions. However, these findings did not substantially influence the image quality or estimated ablation lesion size. For triphasic contrast CT, an enhanced rim was seen in the arterial and venous phases, but it was more prominent in the venous phase of all lesions. Histologic examination determined that this enhanced rim was the outer zone, which was characterized by a heavy infiltration of lymphoid cells and fibroblasts and subsequent granulation. These findings were particularly prominent on day 3. In humans, this hyperemic area is usually dominant in the arterial phase but may also be dominant in the venous phase.5 If the periablational rim is enhanced during the arterial phase, it may be confused with residual tumor tissue.2 Because hepatic tumors (eg, hepatocellular carcinomas) are usually hyperattenuating in the arterial phase and hypoattenuating in the venous phase relative to the surrounding parenchyma, tissues with enhancement during the arterial phase should be suspected as tumor tissues.28,29 In contrast to findings in humans,5 periablational rim enhancement was more prominent in the dogs of the present study during the venous phase. This result, particularly the finding that the enhanced periablational rim was distinguishable from residual tumor after RFA, should be verified in a larger population of dogs.
Radiofrequency ablation of the liver was safely performed in dogs with no major adverse effects noted during diagnostic imaging or necropsy. In the study reported here, real-time ultrasonography was used to guide positioning of the ablation electrode. Radiofrequency ablation can also be performed with CT guidance to precisely target lesions in other tissues, including the lungs or mediastinal organs. Ablation tissues in the present study had a circular appearance because a slim, expandable needle with an expandable monopolar electrode array was used to create lesions. The shape and extent of the ablation zone are affected by several factors, such as advanced RFA generators that yield appropriate power output (eg, pulsed current depositions), electrode configuration, or local heat loss attributable to vascular geometry or tissue perfusion properties.30 The ablation zone may be spherical, oval, or oblong, depending on the number and type of electrodes used during the procedure. Adjacent blood vessels around the ablated tissue can cause an irregularly shaped ablation zone because the blood flow carries heat away from the area of ablation, thus decreasing RFA efficacy (so-called heat-sink effect).31 Thus, evaluations of RFA efficacy in normally vascularized livers might differ from those with pathological conditions, particularly if the internal vascularity is altered. The present study had a limitation because RFA was performed in normally vascularized tissue.
In the present study, contrast-enhanced CT was the most reliable method for estimating the size of the ablation zone and RFA efficacy. Nonenhanced areas on CT images were found to indicate necrosis and hemorrhage in the ablation lesion and could be measured as the ablation zone. The existence of a periablative enhanced rim was determined to be granulation tissue and should be differentiated from residual tumor tissue. This hyperemic rim was mainly observed in the portal phase, which could reduce confusion when distinguishing post-RFA changes from residual tumor. There is a clinical limit regarding the use of repetitive CT for determining RFA efficacy because of the cost and radiation exposure associated with this modality. However, CT can be used for quantitative evaluations of RFA efficacy for measuring the ablation zone and evaluating the presence of residual tumor. Ultrasonography can be used to facilitate RFA procedural guidance in real time, and CEUS can be used to delineate the deep margin of the ablation zone to a greater extent, compared with results for conventional ultrasonography, and can be used to qualitatively evaluate changes of the ablation lesions after RFA in dogs. Findings of this study can be used as fundamental data to support further studies of RFA lesions in canine patients, including those with hepatic neoplasia.
Acknowledgments
Supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF), funded by the Ministry of Science, ICT, and Future Planning (grant No.2015R1A2A2A01003313).
ABBREVIATIONS
| CEUS | Contrast-enhanced ultrasonography |
| HU | Hounsfield unit |
| RFA | Radiofrequency ablation |
Footnotes
Coag Dx analyzer, IDEXX Laboratory, Westbrook, Me.
Domitor, Orion Corp, Espoo, Finland.
Zoletil, Virbac, Carros, France.
Maritrol, Jeil Pharm, Daegu, Republic of Korea.
Cefozol, Hankook Korus Pharm, Seoul, Republic of Korea.
ProSound Alpha 7, Hitachi-Aloka, Tokyo, Japan.
UST-5412 (5 to 13 MHz), Hitachi-Aloka, Tokyo, Japan.
Boston Scientific ablation system, Boston Scientific, Natick, Mass.
LeVeen SuperSlim needle 20 mm, Boston Scientific, Natick, Mass.
Tridol, Yuhan, Seoul, South Korea.
Noblus, Hitach-Aloka, Tokyo, Japan.
SonoVue, Bracco, Milan, Italy.
Omnihexol 300, Korea United Pharm Co, Seoul, Republic of Korea.
Medrad Vistron C-T injector system, Medrad Inc, Warren-dale, Pa.
INFINITT, Infinitt Healthcare, Seoul, Republic of Korea.
Provive 1%, Myungmoon Pharm, Seoul, Republic of Korea.
KCL-40 injectable, Daihan Pharm, Seoul, Republic of Korea.
Image J software, National Institutes of Health, Bethesda, Md.
SPSS for Windows, release 21.0, SPSS Inc, Chicago, Ill.
References
1. Ahmed M, Solbiati L, Brace CL, et al. Image-guided tumor ablation: standardization of terminology and reporting criteria—a 10-year update. J Vasc Interv Radiol 2014; 25:1691–1705.
2. Claudon M, Dietrich CF, Choi BI, et al. Guidelines and good clinical practice recommendations for contrast enhanced ultrasound (CEUS) in the liver—update 2012: a WFUMB-EFSUMB initiative in cooperation with representatives of AFSUMB, AIUM, ASUM, FLAUS and ICUS. Ultrasound Med Biol 2013; 39:187–210.
3. Chu KF, Dupuy DE. Thermal ablation of tumours: biological mechanisms and advances in therapy. Nat Rev Cancer 2014; 14:199–208.
4. Kim YS, Rhim H, Lim HK, et al. Coagulation necrosis induced by radiofrequency ablation in the liver: histopathologic and radiologic review of usual to extremely rare changes. Radiographics 2011; 31:377–390.
5. Sainani NI, Gervais DA, Mueller PR, et al. Imaging after percutaneous radiofrequency ablation of hepatic tumors: part 1, normal findings. AJR Am J Roentgenol 2013; 200:184–193.
6. Solbiati L, Ierace T, Tonolini M, et al. Guidance and monitoring of radiofrequency liver tumor ablation with contrast-enhanced ultrasound. Eur J Radiol 2004; 51(suppl):S19–S23.
7. Rasor L, Pollard R, Feldman EC. Retrospective evaluation of three treatment methods for primary hyperparathyroidism in dogs. J Am Anim Hosp Assoc 2007; 43:70–77.
8. Pollard RE, Long CD, Nelson RW, et al. Percutaneous ultrasonographically guided radiofrequency heat ablation for treatment of primary hyperparathyroidism in dogs. J Am Vet Med Assoc 2001; 218:1106–1110.
9. Mallery KF, Pollard RE, Nelson RW, et al. Percutaneous ultrasound-guided radiofrequency heat ablation for treatment of hyperthyroidism in cats. J Am Vet Med Assoc 2003; 223:1602–1607.
10. Grier RL, Brewer WG Jr, Paul SR, et al. Treatment of bovine and equine ocular squamous cell carcinoma by radiofrequency hyperthermia. J Am Vet Med Assoc 1980; 177:55–61.
11. Wang X, Sofocleous CT, Erinjeri JP, et al. Margin size is an independent predictor of local tumor progression after ablation of colon cancer liver metastases. Cardiovasc Intervent Radiol 2013; 36:166–175.
12. Sala M, Llovet JM, Vilana R, et al. Initial response to percutaneous ablation predicts survival in patients with hepatocellular carcinoma. Hepatology 2004; 40:1352–1360.
13. Numata K, Morimoto M, Ogura T, et al. Ablation therapy guided by contrast-enhanced sonography with Sonazoid for hepatocellular carcinoma lesions not detected by conventional sonography. J Ultrasound Med 2008; 27:395–406.
14. Wen YL, Kudo M, Zheng RQ, et al. Radiofrequency ablation of hepatocellular carcinoma: therapeutic response using contrast-enhanced coded phase-inversion harmonic sonography. AJR Am J Roentgenol 2003; 181:57–63.
15. Meloni MF, Goldberg SN, Livraghi T, et al. Hepatocellular carcinoma treated with radiofrequency ablation. AJR Am J Roentgenol 2001; 177:375–380.
16. Vilana R, Llovet JM, Bianchi L, et al. Contrast-enhanced power Doppler sonography and helical computed tomography for assessment of vascularity of small hepatocellular carcinomas before and after percutaneous ablation. J Clin Ultrasound 2003; 31:119–128.
17. Goldberg SN, Gazelle GS, Compton CC, et al. Treatment of intrahepatic malignancy with radiofrequency ablation: radiologic-pathologic correlation. Cancer 2000; 88:2452–2463.
18. Raman SS, Lu DS, Vodopich DJ, et al. Creation of radiofrequency lesions in a porcine model: correlation with sonography, CT, and histopathology. AJR Am J Roentgenol 2000; 175:1253–1258.
19. Dromain C, de Baere T, Elias D, et al. Hepatic tumors treated with percutaneous radio-frequency ablation: CT and MR imaging follow-up. Radiology 2002; 223:255–262.
20. Varghese T, Zagzebski JA, Lee FT, Jr. Elastographic imaging of thermal lesions in the liver in vivo following radiofrequency ablation: preliminary results. Ultrasound Med Biol 2002; 28:1467–1473.
21. van Vledder MG, Boctor EM, Assumpcao LR, et al. Intra-operative ultrasound elasticity imaging for monitoring of hepatic tumour thermal ablation. HPB (Oxford) 2010; 12:717–723.
22. Correa-Gallego C, Karkar AM, Monette S, et al. Intraoperative ultrasound and tissue elastography measurements do not predict the size of hepatic microwave ablations. Acad Radiol 2014; 21:72–78.
23. National Research Council. Guide for the care and use of laboratory animals. 8th ed. Washington, DC: National Academies Press, 2011.
24. Varghese T, Techavipoo U, Zagzebski JA, et al. Impact of gas bubbles generated during interstitial ablation on elastographic depiction of in vitro thermal lesions. J Ultrasound Med 2004; 23:535–544.
25. Kolokythas O, Gauthier T, Fernandez AT, et al. Ultrasound-based elastography: a novel approach to assess radio frequency ablation of liver masses performed with expandable ablation probes: a feasibility study. J Ultrasound Med 2008; 27:935–946.
26. Wu H, Wilkins LR, Ziats NP, et al. Real-time monitoring of radiofrequency ablation and postablation assessment: accuracy of contrast-enhanced US in experimental rat liver model. Radiology 2014; 270:107–116.
27. Wu H, Patel RB, Zheng Y, et al. Differentiation of benign periablational enhancement from residual tumor following radio-frequency ablation using contrast-enhanced ultrasonography in a rat subcutaneous colon cancer model. Ultrasound Med Biol 2012; 38:443–453.
28. Dill-Macky MJ, Asch M, Burns P, et al. Radiofrequency ablation of hepatocellular carcinoma: predicting success using contrast-enhanced sonography. AJR Am J Roentgenol 2006; 186:S287–S295.
29. Kim SK, Lim HK, Kim YH, et al. Hepatocellular carcinoma treated with radio-frequency ablation: spectrum of imaging findings. Radiographics 2003; 23:107–121.
30. Ni Y, Mulier S, Miao Y, et al. A review of the general aspects of radiofrequency ablation. Abdom Imaging 2005; 30:381–400.
31. Lim HK, Choi D, Lee WJ, et al. Hepatocellular carcinoma treated with percutaneous radio-frequency ablation: evaluation with follow-up multiphase helical CT. Radiology 2001; 221:447–454.
