Needle track seeding after percutaneous radiofrequency ablation of hepatocellular carcinoma: 14-year experience at a single centre

Abstract

Purpose: To determine the incidence, risk factors and prognosis associated with needle track seeding after percutaneous radiofrequency ablations (RFA) for hepatocellular carcinoma (HCC) with a long-term follow-up.

Materials and methods: A total of 741 HCC patients undergoing percutaneous RFA were retrospectively analysed. Mean follow-up interval was 34.3 ± 26.8 months. All seeding neoplasms were diagnosed by imaging modalities with or without pathological evaluation. Risk factors, including Child–Pugh grading, tumour size, number, location, serum alpha-fetoprotein (AFP) level, track number, biopsy before RFA and electrode type were performed by univariate analysis. Further therapy and survival of seeding after RFA were assessed. Survival analysis was analysed by Kaplan–Meier method.

Results: Twelve patients (12 tumours) were diagnosed as seeding. It corresponds to an incidence of 1.6% (12/741) per patient and 0.9% (12/1341) per tumour. Seeding developed an average of 14.0 ± 8.1 months (6–33 months). Significant risk factors included tumour >3 cm (p = 0.031), subcapsular tumour (p = 0.031), biopsy before RFA (p = 0.001) and non-cool-tip electrode (p = 0.034). Eight patients received local therapy and four cases only received systematic therapy for uncontrolled advanced hepatic tumour or distal metastasis. Of eight patients receiving local therapy, one patient had local recurrence 16 months later and other seven patients did not have local recurrence for 3–73 months. The cumulative survival rates after seeding were 55.6%, 27.8%, 9.3% at 1, 3 and 5 years, respectively.

Conclusion: Needle track seeding is a rare delayed complication after percutaneous RFA. Tumour >3 cm, subcapsular tumour, biopsy before RFA and non-cool-tip electrode are potential risk factors for seeding. Local therapies are effective methods for seeding patients.

Keywords:

• Liver malignancy
• radiofrequency ablation
• complication
• seeding

Introduction

Hepatocellular carcinoma (HCC) is the most common primary malignant tumours of the liver. Historically, curative therapies for HCC were surgical resection and transplantation [1]. However, in recent years, radiofrequency ablation (RFA) has become an alternative curative therapy for early HCC [2,3]. In addition, RFA also plays an important role in the treatment of unresectable tumours [4–6]. Even though RFA is a minimally invasive technique, various complications have been reported [7–10]. Among those, needle track seeding is a rare major complication. Most studies reported the occurrence rate of seeding was at a lower incidence of 0–5.1% [11–16] and the likely risk factors included poorly differentiated tumour, larger tumour, subcapsular tumour, a high serum alpha-fetoprotein (AFP) value, biopsy before RFA and repeated insertion [11–14,17]. However, risk factors for seeding remain poorly understood because of lower incidence. Furthermore, it is not clear whether seeding affect patients long-term outcome. In this study, we retrospectively summarised our 14 years’ experience on post-RFA needle track seeding with a large patients samples. We conducted this study to evaluate the incidence, risk factors and prognosis of needle track seeding associated with percutaneous RFA in HCC patients. Also, we collected the 1-, 3-, 5-year survival data after post-RFA needle track seeding to assess its effect on long-term outcomes of HCC patients.

Materials and methods

Patients

From January 2000 to December 2013, 772 consecutive HCC patients undergoing ultrasound-guided percutaneous RFA were retrospectively analysed in this study. The diagnosis of HCC was obtained by core needle biopsy or non-invasive criteria defined by American Association for the Study of Liver Diseases practice guidelines [18]. The treatment was approved by the institutional review board and written informed consents were obtained from all patients before treatment.

All patients met the following criteria for RFA: (1) tumours visible on ultrasonography or contrast-enhanced ultrasonography, (2) without direct invasion of adjacent organs, (3) the absence of portal venous thrombosis, (4) platelet count >50 × 10³/mm³ and prothrombin activity >50%, (5) for radical treatment, single tumour size ≤5 cm in maximum diameter, while multiple tumours with a maximum diameter ≤3 cm and tumour number ≤3, (6) for palliative treatment, patients with larger tumours or more tumour number were also included for the aim of reducing tumour burden and prolonging the survival.

Among 772 patients, 31 patients who lost follow-up were excluded. Finally, 741 patients with 1341 HCC tumours were included in the study. There were 596 males and 145 females. The mean age of patients was 59.7 ± 11.5 years (24–88 years). The mean follow-up interval was 34.3 ± 26.8 months (1–161 months). According to Child–Pugh grading, there were 993 Child–Pugh A patients, 320 Child–Pugh B patients and 28 Child–Pugh C patients. The mean tumour size was 2.9 ± 1.4 cm (0.6–8.6 cm).

RFA procedure

Three kinds of RFA system were used in this study, including multitined expandable (RITA-1500 RFA system, Mountain View, CA), internally cool-tip Monopolar (Cool-tip RF ablation system, Boulder, CO) and internally cool-tip multipolar (CelonLab POWER system, Teltow, Germany) electrodes. Aloka SSD-4000, SSD-5500, α-10 (Tokyo, Japan) and GE Logic-9 (CT) ultrasound systems were used for scanning and guidance.

All RFA procedures were performed by two radiologists with more than 5 years of experience in ultrasound-guided interventional procedures. During the procedure, intravenous moderate sedation anaesthesia was induced with 2.5–5.0 mg of midazolam (Roche; Basel, Switzerland) and 50–100 μg of fentanyl (Fentaini, YiChang, China). Local infiltration anaesthesia was induced by 1% lidocaine (Liduokayin, Beijing, China). For tumours adjacent to the diaphragm, hepatic hilum or liver surface, an intravenous bolus of propofol (Diprivan, Macclesfield, United Kingdom) 1–2 mg was given to enhance anaesthesia.

Patients were placed in the supine or oblique position depended on tumour location. Electrodes were inserted into the tumour under ultrasound guidance. For tumours larger than 3 cm, overlapping ablations were used. During the procedure, vital sign of the patients was continuously monitored. After treatment, track ablation was performed. The track ablation mode was selected according to equipment instruction. The electrode was withdrawn step by step and each step was around 1–2 cm distrance until leave the patients skin. During the procedure, the temperature of needle tip was kept being above 70 °C.

Follow-up and diagnosis of needle track seeding

Treatment response was evaluated by contrast-enhanced computed tomography (CECT) or magnetic resonance imaging (MRI) at 1 month after RFA procedure. Then, serum AFP (alpha-fetoprotein) test, abdominal ultrasound and CECT or MRI were repeated every 3 months in the first year, and every 4–6 months in the following years.

Needle track seeding was defined as new neoplastic nodule occurring along the RFA needle track outside the liver. All neoplasms were demonstrated by imaging modalities, including ultrasonography, CT and MRI. Final diagnosis can be confirmed by pathological evaluation from biopsy or resection tissues. For those without pathological evaluation, the final diagnosis was based on imaging examination.

Statistical analysis

Continuous variables were expressed as mean ± SD. Comparisons of continuous variables were performed by Student’s t-test. Comparisons of categorical variables were performed by Fisher exact test. Risk factors, including Child–Pugh grading, tumour size, number, location, serum AFP level, track number, biopsy before RFA and electrode type were analysed with univariate analysis. To elucidate tumour location as a possible risk factor for seeding, we divided tumours into two groups: subcapsular tumour and deep tumour. The subcapsular tumour was defined as the tumour was located under the surface of liver and electrode was directly inserted into the tumour without passing through normal liver parenchyma. Survival analysis was analysed by Kaplan–Meier method. The level of significant was set at 0.05 for all tests. SPSS19.0 software (SPSS Inc., Chicago, IL) was used to perform the analysis.

Results

Incidence of needle track seeding

Twelve patients (12 tumours) were diagnosed as needle track seeding after RFA procedures (Figure 1). It corresponds to an incidence of 1.6% (12/741) per patient and 0.9% (12/1341) per tumour. Needle track seeding developed an average of 14.0 ± 8.1 months (6–33months) after RFA (Table 1). The site of seeding included abdominal wall in seven patients, peritoneal cavity in two patients and thoracic wall in three patients. Among these patients, four cases were diagnosed based on pathological examination and the other eight cases were diagnosed based on imaging findings. All neoplasms were firstly detected by ultrasonography. High-frequency ultrasonography and colour Doppler ultrasonography were very sensitive for the diagnosis of seeding. Ultrasonography demonstrated neoplastic nodules along the needle track with hyperechogenicity or hypoechogenicity. Colour Doppler ultrasound showed arterial or venous signal in 10 lesions and no blood signal in two lesions.

Figure 1. A 48-year-old man with well-differentiated hepatocellular carcinoma underwent percutaneous RFA. (A) Twenty-month follow-up ultrasonography shows a 10 mm seeding nodule (arrow) with hypo echo in abdominal wall. (B) CECT shows seeding nodule (arrow) and subcapsular ablation zone (arrowhead) in portal phase.

Table 1. Clinical feature of 12 HCC patients associated with neoplastic seeding.

No.	Differentiation of HCC	Interval between RFA (month)	Local therapy	Survival after seeding (month)
1	Poorly	12	None	3
2	Poorly	6	Resection	11
3	Poorly	24	Radiotherapy	21
4	Poorly	6	None	8
5	Poorly	8	None	7
6	Moderately	12	HIFU	27
7	Moderately	12	None	4
8	Well	8	HIFU	53
9	Moderately	15	Resection	36
10	Poorly	12	HIFU	34
11	Well	20	Resection	73^a
12	Moderately	33	HIFU	6^a

^a The patients were still alive.

HCC: hepatocellular carcinoma; HIFU: high-intensity focussed ultrasound therapy; RFA: radiofrequency ablation.

Risk factors analysis

The clinical feature of two groups was summarised in Table 2. There was significant difference between seeding and without seeding group in tumour >3 cm (p = 0.031), subcapsular tumour (p = 0.031), biopsy before RFA (p = 0.001) and non-cool-tip electrode (p = 0.034), while no significant difference in Child–Pugh grading, tumour number, serum AFP level and track number.

Table 2. Clinical feature of 741 HCC patients with 1341 tumours treated by percutaneous RFA.

Variables	Seeding (n = 12)	Without seeding (n = 1329)	p value
Child–Pugh grading			0.063
A	6	987
B	6	314
C	0	28
Tumour size			0.031
≤3 cm	4	851
>3 cm	8	478
Tumour number			0.555
Single	7	801
Multiple	5	528
Tumour location			0.031
Subcapsular	5	211
Deep	7	1118
Serum AFP			0.393
≤200 ng/ml	10	997
>200 ng/ml	2	332
Track number^a	2.50 ± 1.83	2.24 ± 1.37	0.514
Biopsy before RFA			0.001
Yes	12	718
No	0	611
Electrode type			0.034
Cool-tip	1	540
Non cool-tip	11	789

^a Values are expressed as mean ± SD.

AFP: alpha-fetoprotein; HCC: hepatocellular carcinoma; RFA: radiofrequency ablation.

Therapy and prognosis

In 12 seeding patients, eight patients without uncontrolled advanced hepatic tumour received local therapy, including resection (n = 3), high-intensity focussed ultrasound therapy (HIFU) (n = 4) and external radiotherapy (n = 1). One patient with HIFU had local recurrence 16 months later and other seven patients did not have local recurrence. Four patients with uncontrolled advanced hepatic tumour only received systematic therapy. The mean follow-up period of seeding nodules after detection was 23.6 ± 22.1 months (3–73 months). The overall survival rates after seeding were 55.6%, 27.8%, 9.3% at 1, 3 and 5 years respectively (mean survival: 25.6 ± 6.4 months). The survival rates in patients without uncontrolled advanced hepatic tumour who received local therapy at 1, 3 and 5 years were 85.7%, 42.9%, 14.3%, respectively, while the survivals in patients with uncontrolled advanced hepatic tumour were only 3–8 months.

Discussion

RFA has been considered as a relatively safe and effective technique for small HCC in recent years. Although as a minimally invasive approach, RFA is associated with some major complications including uncontrolled bleeding, bowel perforation, bile fistula and needle track seeding [7–10]. In the early stage, a small sample study by surgeons reported that needle track seeding occurred up to 12.5% [17]. However, most subsequent studies reported seeding occurred 0–5.1% [11–16]. In our centre, the incidence of seeding was 0.9% per tumour, which was at a low level of incidence compared with previous studies. One possible reason may be track ablation performed in our study. Track ablation can destroy potential tumour cell dissemination along the track during the procedure. According to previous studies, the incidence of seeding with track ablation was lower than those without track ablation (0.7–3.2% vs 5.1–12.5%) [11–17]. Therefore, to our knowledge, track ablation is encouraged as a routine technique when RFA is terminated.

Needle track seeding developed an average of 14 months (6–33 months) after RFA. Some studies reported seeding neoplasms were detected even 5 years after RFA [14,15]. This indicates that neoplastic seeding is a delayed complication usually occurring several months after RFA. Accordingly, seeding should be carefully evaluated during follow-up, especially in the first 3 years.

Previous studies reported that risk factors for seeding may include poorly differentiated tumour, larger tumour, subcapsular tumour, a high serum AFP value, biopsy before RFA and repeated insertion [11–14]. In the present study, Child–Pugh grading, tumour size, number, location, serum AFP level, track number, biopsy before RFA and electrode type were considered. The results showed that tumour size >3 cm, subcapsular tumour, biopsy before RFA and electrode type were potential risk factors for neoplastic seeding. Tumour size is an important factor for RFA effect. Complete necrosis is more likely to be achieved in tumour ≤3 cm than >3 cm [2]. Once incomplete necrosis occurred, activated tumour cells may disseminate along the needle track and finally become seeding neoplasm. During RFA procedure of subcapsular tumours, intratumoral pressure rapidly increased by heat may lead to dissemination of tumour cells into peritoneal cavity [14]. To prevent rapidly increased intratumoral pressure, gradually amplifying radiofrequency power may reduce the rate of seeding [14]. Moreover, electrode should not be inserted into subcapsular tumour directly if possible. Instead, a path that passed through normal liver parenchyma was recommended [13]. This approach created an intrahepatic path to reduce the risk of dissemination of tumour cells into peritoneal cavity.

Stigliano et al. reported the median risk of seeding was 0.61% (0–5.56%) for RFA without biopsy and 0.95% (0–12.5%) for RFA with biopsy [19]. Although track ablation was performed during RFA, biopsy before RFA was not in the same track of RFA electrode and the incidence of seeding increased. Therefore, it is suggested that biopsy should be avoided if HCC can be diagnosed by imaging examination and serum test.

Electrode type is another potential factor of seeding. Comparison of different electrode type was not reported in previous studies. Seeding in non-cool-tip electrode group is more than cool-tip electrode group. A possible explanation for this is that carbonisation occurs much more in non-cool-tip electrode and carbonisation tissue may adhere to the hook of electrode. When withdrawing the hooks, carbonisation tissue is clipped between the hooks, which increases the risk of dissemination of tumour cells, especially for the condition of repetitious electrode placement. However, the incidence of seeding in non-cool-tip group was still at a low level of 1.4% (11/800). Due to a small number of seeding cases, further studies are needed to confirm whether cool-tip electrode is superior.

As a major complication, neoplastic seeding seemed to affect patients’ prognosis. However, there were few studies reported impact of seeding on the patients’ survival. Imamura et al. reported the cumulative survival rates in the patients with seeding at 1, 2 and 3 years were 81.3%, 45.4% and 26.9%, respectively [14]. The authors considered seeding neoplasms themselves did not directly affect the patients’ survival. Shirai et al. reported the cumulative survival rates after seeding were 76.9% at 1 year, 41.0% at 3 years and 20.5% at 5 years [15]. In the present study, post-seeding survival rates were 55.6%, 27.8%, 9.3% at 1, 3 and 5 years, respectively, which showed poorer prognosis than the former studies. The reason might be that there were more advanced tumours included in our centre for palliative therapy.

The therapy of seeding includes resection, external radiotherapy, RFA and HIFU according to the previous studies [20–22]. In our study, three patients received resection, four received HIFU and one received external radiotherapy. About 87.5% (7/8) patients achieved local cure. The survival rate of those eight patients at 3 years was up to 42.9%. This result showed that the neoplastic seeding was a low-risk complication of percutaneous RFA of liver cancer and was considered acceptable in general after local therapy. Thus, once occurring track seeding without distant metastasis, seeding can be well controlled. However, patients with uncontrolled advanced hepatic tumour did not receive local therapy, and the survival was only 3–8 months. Therefore, local therapy should be considered, especially for those without uncontrolled advanced hepatic tumour.

The limitations of present study were as follows: (1) As a retrospective study, limited potential risk factors were analysed, and other factors which may result in seeding were not included, such as tumour differentiation. Further well-designed prospective studies are warranted to confirm. (2) Because of low incidence of seeding, a small number cases was another limitation. Thus, further multiple-centre studies or systems analysis are needed to well understand risk factors.

Conclusions

Needle track seeding is a rare delayed complication after percutaneous RFA. Tumour >3 cm, subcapsular tumour, biopsy before RFA and non-cool-tip electrode were potential risk factors for seeding. Careful follow-up should be done to detect seeding earlier. Local therapies are effective methods for seeding patients, especially those without uncontrolled intrahepatic recurrence.

Disclosure statement

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this article.

Additional information

Funding

This research was supported by Beijing Municipal Science and Technology Commission (Z161100000516222 and Z151100004015186).