Home About us Editorial board Ahead of print Current issue Search Archives Submit article Instructions Subscribe Contacts Login 

 Table of Contents  
Year : 2016  |  Volume : 12  |  Issue : 7  |  Page : 153-158

Effect of heat sink on the recurrence of small malignant hepatic tumors after radiofrequency ablation

1 Department of Interventional Radiology, First Affiliated Hospital of Fujian Medical University, Fuzhou, China
2 Department of Radiology, Shaxian Hospital, Shaxian, China
3 Department of Pain Management, Mindong Hospital of Fujian Medical University, Fuan, China

Date of Web Publication21-Feb-2017

Correspondence Address:
Zheng-Yu Lin
Department of Interventional Radiology, First Affiliated Hospital of Fujian Medical University, 20 Chazhong Road, Fuzhou 350005
Login to access the Email id

Source of Support: None, Conflict of Interest: None

DOI: 10.4103/jcrt.JCRT_959_16

Rights and Permissions
 > Abstract 

Aims: The aim of this study was to investigate the effect of heat sink on the recurrence of hepatic malignant tumors <3 cm after percutaneous radiofrequency ablation (RFA).
Subjects and Methods: This study included 564 hepatic malignant tumors <3 cm in 381 patients. Preoperative images were used to determine whether these tumors were adjacent to vessels, and the diameter of adjacent vessels was measured. RFA was performed computed tomography (CT), magnetic resonance imaging (MRI) and ultrasound (US) guidance, and postoperative imaging follow-up was then conducted.
Statistical Analysis Used: SPSS software version 17.0 was used for data processing, and the χ2 test was used for comparative analysis. Two-sided P < 0.05 indicated statistical significance.
Results: A total of 33 recurrences were found: 15 in the MR group (15/468), 12 in the US group (12/53), and 6 in the CT group (6/43). Of the 101 lesions adjacent to blood vessels larger than 3 mm, 20 showed recurrence: 10 in the MR group (10/77), 7 in the US group (7/17), and 3 in the CT group (3/7). The recurrence rate of perivascular lesions was higher than that of nonperivascular lesions, and the rate in the MR group was lower those in the US and CT groups.
Conclusions: The curative effect of MRI-guided RFA is better than those of US- and CT-guided ablation. The heat sink effect is an important factor affecting recurrence of hepatic malignant tumors after RFA.

Keywords: Heat sink effect, hepatic malignant tumor, magnetic resonance imaging, radiofrequency ablation, recurrence

How to cite this article:
Lin ZY, Li GL, Chen J, Chen ZW, Chen YP, Lin SZ. Effect of heat sink on the recurrence of small malignant hepatic tumors after radiofrequency ablation. J Can Res Ther 2016;12, Suppl S3:153-8

How to cite this URL:
Lin ZY, Li GL, Chen J, Chen ZW, Chen YP, Lin SZ. Effect of heat sink on the recurrence of small malignant hepatic tumors after radiofrequency ablation. J Can Res Ther [serial online] 2016 [cited 2021 Mar 1];12:153-8. Available from: https://www.cancerjournal.net/text.asp?2016/12/7/153/200610

 > Introduction Top

Percutaneous radiofrequency ablation (RFA) involves insertion of a radiofrequency (RF) electrode into tumors for ablation through a percutaneous puncture under the guidance of various imaging techniques, such as computed tomography (CT), ultrasound (US), or magnetic resonance imaging (MRI).[1] It results in coagulative necrosis of tumor tissues and subsequent tumor inactivation. Among the main organs, the liver is especially susceptible to primary and secondary malignancies. Because it causes relatively less trauma and has reliable local efficacy, RFA is widely used for treating malignant hepatic tumors that lack surgical indications.[1] Although the curative effect of RFA is similar to surgery for small hepatic tumors, the tumors tend to recur.[2] In the case of tumors adjacent to large blood vessels or those abundant with blood vessels, heat loss can occur during RFA, a condition referred to as the “heat sink effect.” This is an important factor affecting RFA efficacy and leading to residual tumors or recurrence.[3],[4]

In the present study, we retrospectively analyzed malignant hepatic tumors <3 cm that were treated with RFA at our hospital, with the aim of exploring the efficacy of RFA and the effect of heat sink on postoperative local recurrence.

 > Subjects and Methods Top


This study included 381 patients with 564 malignant hepatic lesions ≤3 cm, as confirmed by pathological examination or clinical diagnosis, who underwent RFA between September 2009 and September 2014. There were 322 men and 59 women with an average age of 56.7 years (range, 26–81 years). Two hundred and ninety-one cases were of primary hepatic cancer (393 lesions) while the remaining (90) were of hepatic metastatic carcinoma (171 lesions). MRI-guided RFA (MR group) was used in 301 cases (468 lesions); US-guided RFA (US group), in 49 cases (53 lesions); and CT-guided RFA (CT group), in 31 cases (43 lesions).

Equipment and materials

For RFA, a RITA 1500X RF generator, StarBurst XL Multipolar RF probe (14-gauge, 10/15 cm), and MRI-compatible RF probe (StarBurst XL MRI) were used.

A GE Signa Infinity Twinspeed 1.5 T dual-gradient MR scanner and torso body coil were used for MRI guidance, with the following scan sequence and parameters: repetition time (TR) = 6000 ms, echo time (TE) = 87 ms, flip angle = 90°, slice thickness (ST) = 5.0 mm, gap = 1.0 mm, field of view (FOV) = 38 cm × 28 cm for fat suppression, fast relaxation, fast spin-echo T2-weighted imaging (T2WI) and TR = 4.8 ms, TE = 1.1 ms, ST = 3 mm, FOV = 38 cm × 28 cm for three-dimensional dynamic T1WI. A Toshiba Aquilion 4-slice spiral tomography was used for CT, while a Siemens Sequoia-512 color Doppler ultrasonic diagnostic apparatus and Philips IU22 color Doppler ultrasonic diagnostic apparatus were used for US.

Measurement of vessel diameter

From preoperative MRI images, two senior imaging diagnosticians determined whether the tumor was adjacent to vessels, and they jointly measured the diameters of adjacent vessels. Tumors that had adjacent vessels with diameters ≥ 3 mm were placed in the vessel group (Gv; perivascular lesions), while those adjacent to vessels with diameters <3 mm or without adjacent to vessels were in the nonvessel group (Gn; nonperivascular lesions).

Radiofrequency ablation procedure

Conventional imaging (MRI, CT, or US) was first performed to identify the puncture point and determine the puncture angle (such that the puncture path was as short as possible and avoided important structures). The puncture site was routinely disinfected, and local anesthesia was administered. During the guidance imaging, the RF probe was gradually introduced into the puncture site until it reached the tumor edge; the fine electrode was then deployed and scanning was performed again to confirm that the electrode placement was satisfactory. The RF probe and skin electrode wire were connected to the RF generator, the power was set at 150 W and the target temperature at 105°C, and the appropriate ablation procedure and time were selected according to the desired ablation size (2 cm required 5 min, 3 cm required 5.5 min, 4 cm required 8 min, and 5 cm required 15 min). After ablation, imaging was performed again to check for therapeutic as well as adverse effects. The ablation range was generally 0.5–1.0 cm beyond the lesion edge. Supplementary ablation was performed if residual lesions were detected. Finally, the electrode probe was withdrawn, needle-path ablation was performed, and the needle was withdrawn.

For follow-up analysis, imaging was conducted 1 month after RFA and then once every 2–3 months. Two senior imaging diagnosticians jointly evaluated the local efficacy of RFA and analyzed the causes of any recurrence. If the primary cancer lesion did not show any enhancement during the arterial, portal venous, or parenchymal phase in contrast-enhanced US (CEUS), CT, or MRI, complete tumor ablation was considered. If imaging examination showed an increase in the size of lesions with enhanced nodules around them, residual tumor or recurrence was considered. Local recurrence of an ablation lesion or death was considered as the end point of follow-up. If the lesion was adjacent to vessels ≥ 3 mm preoperatively and the region of recurrence was close to these vessels, the recurrence was deemed to have been caused by heat sink effect [Figure 1].
Figure 1: A 39-year-old man with postoperative hepatic metastasis of colonic carcinoma. Preoperative T2-weighted imaging (a) shows a lesion in the left lateral lobe of the liver adjacent to the portal vein. T1-weighted imaging at 1 month after operation (b) shows a gap in the high-signal ring surrounding the ablation lesion adjacent to the vessel (black arrow), suggesting recurrence. T1-weighted imaging at 5 months after operation (c) shows an obvious increase in the size of the recurrent lesion

Click here to view

Statistical analysis

SPSS software version 17.0 (IBM Corp., Armonk, NY, USA) was used for data processing, and the χ2 test was used for comparative analysis. Two-sided P < 0.05 indicated statistical significance.

 > Results Top

General findings

RFA was successfully completed for all 564 lesions in the 381 patients, without any instance of intra-or peri-operative death. The maximum diameters of the lesions ranged from 0.5 to 3.0 cm with an average value of 1.66 cm. One to five repetitions of ablation were necessary for each lesion, and the average number of repetitions was 1.8 (1.9 for the MR group, 1.4 for the CT group, and 1.2 for the US group). The average follow-up time was 27.4 (range, 8–53) months. Thirty-two patients died during follow-up, and the median time of death was 18 months after surgery (range, 8–41 months).

Intra-operative imaging findings

After ablation in the MR group, the lesions showed a low-signal intensity on T2WI, surrounded by a thin high-signal zone. The signals of the primary lesions were complex, and most showed equisignals or low signal, while a few showed slightly high signals. On T1WI, the ablation lesions showed concentric changes, wherein the central primary lesions continued to show low signals but were surrounded by a distinct high-signal zone with a clear boundary [Figure 2].
Figure 2: A 74-year-old man with postoperative hepatic metastasis of gastric carcinoma. Preoperative T2-weighted imaging (a) and T1-weighted imaging (b) show long-T1 long-T2 metastasis in the left medial lobe of the liver. Intraoperative T2-weighted imaging (c) shows the deployed radiofrequency electrode wrapped around the lesion. T1-weighted imaging after ablation (d) shows increased signal intensity of the peripheral normal liver tissue after ablation but low-signal intensity of the central primary lesion. T2-weighted imaging (e) shows the ablation lesion with a low-signal intensity surrounded by a thin high-signal zone

Click here to view

In the CT group, the postoperative scans showed lesions that were dominantly mixed density though some had a slightly high density, and most were surrounded by a circular low-density zone. The primary lesion was unclearly displayed.

US enabled real-time monitoring of the echo changes in the ablation region. The echo of ablation lesions strengthened gradually although the primary lesions were not clearly displayed. CEUS, color Doppler flow imaging, and color Doppler energy imaging showed the blood flow information of the ablated region disappearing.

Local curative effect

A total of 33 recurrent lesions were detected: 15 in the MR group (15/468), 12 in the US group (12/53), and 6 in the CT group (6/43). One hundred and one lesions were adjacent to vessels ≥ 3 mm (77 in the MR group, 15 in the US group, and 9 in the CT group). The recurrence rates of hepatic tumor in all imaging groups are shown in [Table 1], while those in the perivascular (Gv) and nonperivascular (Gn) groups are shown in [Table 2]. The recurrence rates for perivascular lesions in all groups are shown in [Table 3]. The median recurrence time was 6 months (1–22 months). Of the 33 recurrent lesions, 20 were of perivascular lesions, and among these 20, 1 was preoperatively adjacent to a small vessel, which obviously expanded during ablation until its diameter exceeded 3 mm. A nodular recurrent lesion was seen around this vessel after surgery, and it was considered a perivascular recurrence [Figure 3]. In addition, in eight cases (three cases in the MR group, three cases in the US group, and two cases in the CT group), review examination conducted 1 month after operation showed that the ablation lesion had not completely covered the primary lesion. This finding was considered to indicate a residual lesion.
Table 1: Recurrence rate of hepatic malignant tumors after radiofrequency ablation guided by different imaging techniques

Click here to view
Table 2: Recurrence rate of perivascular and nonperivascular hepatic malignant tumors after radiofrequency ablation

Click here to view
Table 3: Recurrence rate of perivascular hepatic malignant tumors after radiofrequency ablation guided by different imaging techniques

Click here to view
Figure 3: A 51-year-old woman with hepatic metastasis of colonic carcinoma. Preoperative T2-weighted imaging (a) shows a high-signal lesion in the right of the liver. Intraoperative T2-weighted imaging (b) shows the deployed radiofrequency electrode wrapped around the lesion. T1-weighted imaging after ablation (c) shows a gap in the high-signal ring. Two rounds of supplementary ablation (d). Postoperative T1-weighted imaging (e) shows ablation lesions. T2-weighted imaging (f) shows the adjacent vessel significant enlargement (arrow). Recurrent lesions after operation showing a long-T1 (g), long-T2 signal (h)

Click here to view

Between-group comparisons

The recurrence rate of perivascular lesions was higher than that of nonperivascular lesions [Table 2], and the rate in the MR group was lower those in the US and CT groups [Table 1] and [Table 3].

 > Discussion Top

In the present study, the complete ablation rate of the 564 lesions ≤3 cm reached about 94.1% although there were a few recurrences. Thus, RFA seems to be an effective local treatment, and in-depth examination of recurrence factors may prevent recurrence and considerably improve the curative effects of RFA.

The most commonly adopted imaging techniques for RFA guidance are US and CT, and only a handful of hospitals use MRI-guided RFA.[5],[6] However, MRI does have some distinct advantages for use in RFA guidance, most important of which is that it provides accurate information regarding the efficacy of ablation soon after RFA.[7] Further, it does not use ionizing radiation, has a good resolution against soft tissue, can be used in arbitrary directions, and allows multisequencing and multiparameter imaging with better resolution than CT and US. MRI is also sensitive to temperature changes; the circular high-signal zone around the lesions on T1WI is due to tissue dehydration and hemorrhage after RFA, and it accurately depicts the relationship between the ablation lesion and the primary lesion, whereby minute residual lesions can be clearly identified.[7] In the present study, most cases were of MRI-guided RFA, and although the ablation time was greater than that during CT- and US-guided ablation, the ablation rate was high at 96.8%. The main reason for this was that MRI guidance allowed the timely identification of minute residual lesions, especially perivascular ones, during operation, and supplementary ablation could be immediately performed, thereby reducing the recurrence rate. Only 10 of the 77 perivascular lesions (12.99%) in the MR group showed recurrence, and this rate was significantly lower than those in the US and CT groups. Review of preoperative MR images for three cases of intraoperative residual lesions showed a similar signal on the plain scans. The scope of the enhanced lesion was obviously greater on the enhanced scan than on the plain scan, and the postoperative relationship between the primary lesion and ablation lesion was not well displayed. Thus, the advantages of MRI guidance were not realized. In addition, some patients were not eligible for MR-guided RFA, including those with coronary artery stents and metal implants and those who had claustrophobia.

US is the most commonly used guidance technique for RFA. It is superior to CT for imaging of hepatic lesions, does not use ionizing radiation, provides real-time guidance for needle insertion, and clearly displays blood vessels.[8] However, it does not clearly display metal electrodes, is easily affected by gas and bone, and cannot clearly depict lesions at certain locations (e.g., in the diaphragmatic dome). Further, when US is performed immediately after ablation, bubbles caused by the high temperature can produced a high echo, leading to overestimation of the size of the ablation lesion. Even with CEUS, because of thermal injury to residual minute lesions, most of them show mild enhancement unlike the wash-in/wash-out enhancement characteristic of primary lesions, and it is therefore easy to confuse these lesions with hepatic tissue with thermal injury or edema.[7],[8] In the present study, the recurrence rate of hepatic lesions in the US group was 22.6%; this finding may be related to case selection since 32.1% lesions (n = 17) in this group were adjacent to large vessels, a proportion significantly higher than those in the CT (16.3%) and MR (16.5%) groups. Further, three residual lesions were found at locations not suitable for US (in the diaphragmatic dome or adjacent to the intestinal canal), where gas affected the display of the lesion and RF electrode.

CT can clearly display the metal RF electrode needle, but it is inferior to MR and US for displaying lesions. CT conducted immediately after ablation can only roughly estimate ablation-related range. Contrast-enhanced CT can show the range of ablation, but like US, residual minute lesions are easily confused with hepatic tissue with thermal injury and edema.[9] In the present study, all lesions in the CT group were clearly displayed, and residual lesions were noted in only two cases. This finding indicates that the low tissue resolution defect of CT was not seen in our study.

Lu et al. performed hepatic RFA in ten alive pigs to evaluate the heat sink effect and found that 50% lesions adjacent to 3–5 mm blood vessels had living tissue residual, 100% lesions adjacent to >5 mm blood vessels had living residual tissue, and only 11.7% lesions adjacent to <3 mm blood vessels showed the heat sink effect.[3] In another study, Lu et al. ablated 15 hepatocellular carcinoma lesions adjacent to large blood vessels (≥ 3 mm) and performed liver transplantation an average of 7.5 months after ablation. Postoperative pathological examination confirmed that eight lesions had recurrence or residual carcinoma, while only 4 of 32 nonperivascular lesions had recurrence.[4] Therefore, in the present study, we selected lesions adjacent to ≥ 3 mm blood vessels to examine the heat sink effect.

Incomplete ablation of lesions is the main cause of recurrence. In our study, twenty of the thirty recurrent lesions were adjacent to blood vessels ≥ 3 mm in diameter, and all were the heat sink effect. Eight lesions were incompletely ablated because of limitations of guidance device (unclear intraoperative display of lesions or unclear postoperative display of residual lesions), and five were recurrent for unknown reasons. In the short-term review, no residual lesions were observed on CT, US, MRI, or contrast-enhanced MRI scans in any of these cases. Recurrence was detected after several months. In fact, in one case, recurrence of the primary lesion was detected in as late as 22 months after RFA. It is possible that a subclinical residual lesion was present at the center of the ablation lesion in this case, and this residual lesion slowly grew because of the thermal damage to tumor cells and lack of normal blood supply, thereby prolonging the recurrence time. It is worth noting that expansion of some small blood vessels because of RFA could also lead to recurrence [Figure 3]. In the present study, the median recurrence time was 6 months, indicating that long-term postoperative radiographic follow-up is essential after RFA.

Some methods enable reduction of the heat sink effect of ablation. For instance, transcatheter arterial chemoembolization can embolize the tumor-feeding artery to reduce blood perfusion.[10] However, in most cases, vascular embolization remains incomplete, and merely lowering hepatic arterial blood supply is not adequate for preventing the heat sink effect for tumors mainly supplied by the portal vein and those adjacent to the portal and hepatic veins. Curley et al. adopted a method in which the portal vein and hepatic artery were temporarily blocked during operation, but this method needs open operation, which has high levels of large trauma.[11] On the other hand, Goldberg et al. locally injected halothane, vasopressin, epinephrine, and other vasoactive agents that could adjust blood flow and increase the ablation area of RFA, but the curative effect of this method was limited because it was unable to completely block blood flow.[12] Through their simple technique, Hou et al. blocked blood flow through percutaneous ablation blocking the tumor-feeding artery and percutaneous tumor arterial embolism, but the recurrence rate was still 17.3%.[13] Implanting [125 I] radioactive particles near blood vessels can also prevent recurrence as it synergistically uses thermo- and radio-therapy.[14]

This study had some limitations. It was a single-center retrospective study, and the number of cases in the US and CT groups was small. Further, only the recurrence of local lesions was investigation, and survival time and other factors were not evaluated. Finally, the follow-up time was short for some patients, which may have affected the results.

RFA is an effective treatment for hepatic malignant tumors <3 cm. However, the heat sink effect is an important factor that underlies tumor recurrence. MRI seems to be the best guidance method for RFA as it clearly displays lesions and enables accurate evaluation of the curative effect of RFA.

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.

 > References Top

Jacobs A. Radiofrequency ablation for liver cancer. Radiol Technol 2015;86:645-64.  Back to cited text no. 1
Cabibbo G, Maida M, Genco C, Alessi N, Peralta M, Butera G, et al. Survival of patients with hepatocellular carcinoma (HCC) treated by percutaneous radio-frequency ablation (RFA) is affected by complete radiological response. PLoS One 2013;8:e70016.  Back to cited text no. 2
Lu DS, Raman SS, Vodopich DJ, Wang M, Sayre J, Lassman C. Effect of vessel size on creation of hepatic radiofrequency lesions in pigs: Assessment of the “heat sink” effect. AJR Am J Roentgenol 2002;178:47-51.  Back to cited text no. 3
Lu DS, Yu NC, Raman SS, Limanond P, Lassman C, Murray K, et al. Radiofrequency ablation of hepatocellular carcinoma: Treatment success as defined by histologic examination of the explanted liver. Radiology 2005;234:954-60.  Back to cited text no. 4
Hoffmann R, Rempp H, Syha R, Ketelsen D, Pereira PL, Claussen CD, et al. Transarterial chemoembolization using drug eluting beads and subsequent percutaneous MR-guided radiofrequency ablation in the therapy of intermediate sized hepatocellular carcinoma. Eur J Radiol 2014;83:1793-8.  Back to cited text no. 5
Hoffmann R, Rempp H, Schraml C, Schwenzer N, Grözinger G, Blumenstock G, et al. Diffusion-weighted imaging during MR-guided radiofrequency ablation of hepatic malignancies: Analysis of immediate pre- and post-ablative diffusion characteristics. Acta Radiol 2015;56:908-16.  Back to cited text no. 6
Lin ZY, Song QQ, Chen J, Wan RJ, Zheng H, Chen ZW, et al. Local curative effect of MRI-guided radiofrequency ablation on small hepatocellular carcinoma. Tumour Biol 2015;36:2105-10.  Back to cited text no. 7
Rajesh S, Mukund A, Arora A, Jain D, Sarin SK. Contrast-enhanced US-guided radiofrequency ablation of hepatocellular carcinoma. J Vasc Interv Radiol 2013;24:1235-40.  Back to cited text no. 8
Laspas F, Sotiropoulou E, Mylona S, Manataki A, Tsagouli P, Tsangaridou I, et al. Computed tomography-guided radiofrequency ablation of hepatocellular carcinoma: Treatment efficacy and complications. J Gastrointestin Liver Dis 2009;18:323-8.  Back to cited text no. 9
Kitamoto M, Imagawa M, Yamada H, Watanabe C, Sumioka M, Satoh O, et al. Radiofrequency ablation in the treatment of small hepatocellular carcinomas: Comparison of the radiofrequency effect with and without chemoembolization. AJR Am J Roentgenol 2003;181:997-1003.  Back to cited text no. 10
Curley SA, Izzo F, Delrio P, Ellis LM, Granchi J, Vallone P, et al. Radiofrequency ablation of unresectable primary and metastatic hepatic malignancies: Results in 123 patients. Ann Surg 1999;230:1-8.  Back to cited text no. 11
Goldberg SN, Hahn PF, Halpern EF, Fogle RM, Gazelle GS. Radio-frequency tissue ablation: Effect of pharmacologic modulation of blood flow on coagulation diameter. Radiology 1998;209:761-7.  Back to cited text no. 12
Hou YB, Chen MH, Yan K, Wu JY, Zhang H, Yang W, et al. Feasibility of improving radiofrequency ablation of hepatocellular carcinoma by percutaneously blocking tumor-feeding vessels. Zhongguo Yi Xue Ke Xue Yuan Xue Bao 2008;30:448-54.  Back to cited text no. 13
Lin ZY, Chen J, Deng XF. Treatment of hepatocellular carcinoma adjacent to large blood vessels using 1.5T MRI-guided percutaneous radiofrequency ablation combined with iodine-125 radioactive seed implantation. Eur J Radiol 2012;81:3079-83.  Back to cited text no. 14


  [Figure 1], [Figure 2], [Figure 3]

  [Table 1], [Table 2], [Table 3]


Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
Access Statistics
Email Alert *
Add to My List *
* Registration required (free)

  >Abstract>Introduction>Subjects and Methods>Results>Discussion>Article Figures>Article Tables
  In this article

 Article Access Statistics
    PDF Downloaded111    
    Comments [Add]    

Recommend this journal