Seeding following percutaneous diagnostic and therapeutic approaches for hepatocellular carcinoma. What is the risk and the outcome?

Cancer Treatment Reviews (2007) xxx, xxx–xxx

ARTICLE IN PRESS

ava i lab le a t www.sc iencedi rec t . com

journal homepage: www.elsevierheal th .com/ journals /c t rv

TUMOUR REVIEW

Seeding following percutaneous diagnostic andtherapeutic approaches for hepatocellularcarcinoma. What is the risk and the outcome?Seeding risk for percutaneous approach of HCC

R. Stigliano a, L. Marelli a,c, D. Yu b,d, N. Davies b,e,D. Patch a, A.K. Burroughs a,*

a Liver Transplantation and Hepatobiliary Medicine Unit, Royal Free Hospital, Pond Street, NW3 2QG London, UKb Department of Radiology, Royal Free Hospital, Pond Street, NW3 2QG London, UK

Received 12 October 2006; received in revised form 30 March 2007; accepted 3 April 2007

03do

*

Nec

d

e

KEYWORDSSeeding;Risk;Hepatocellularcarcinoma (HCC);Needle core biopsy(NCB);Fine needle aspirationcytology (FNAC);Percutaneous ethanolinjection (PEI);Radiofrequencyablation (RFA)

05-7372/$ - see front mattei:10.1016/j.ctrv.2007.04.00

Corresponding author. TelE-mail addresses: [email protected].: +44 20 7274 6229; faTel.: +44 20 7794 0500x34Tel.: +44 20 7794 0500x38

Please cite this article in phepatocellular carcinoma.

r �c 2001

.: +44 [email protected](N. Daviex: +44 20150.767.

ress as: SWhat is

Summary Background: Tumour biopsy is usually considered mandatory for patient manage-ment by oncologists. Currently percutaneous ablation is used therapeutically for cirrhoticpatients with small hepatocellular carcinoma (HCC), not suitable for resection or waiting forliver transplantation. However malignant seeding is a recognized complication of both diagnos-tic and therapeutic procedures in patients with HCC. Although percutaneous therapy whetherwith or without biopsy of a suspected HCC nodule may minimize the risk of seeding, this has notbeen confirmed.Aim: To evaluate the risk of seeding, defined as new neoplastic disease occurring outside theliver capsule, either in the subcutaneous tissue or peritoneal cavity following needle biopsyand/or local ablation therapy (LAT).Methods: A literature search resulted in 179 events in 99 articles between January 1983 andFebruary 2007: 66 seedings followed liver biopsy, 26 percutaneous ethanol injection (PEI), 1microwave, 22 radiofrequency ablation (RFA), and 64 after combined biopsy and percutaneoustreatment (5 microwave; 33 PEI; 26 RFA).Results: In 41 papers specifying the total number of patients biopsied and/or treated, the med-ian risk of seeding was 2.29% (range 0–11%) for biopsy group; 1.4% (1.15–1.85%) for PEI when

7 Elsevier Ltd. All rights reserved.

7274 6229; fax: +44 20 7472 6226.(R. Stigliano), [email protected] (L. Marelli), [email protected] (D. Yu),s), [email protected] (D. Patch), [email protected] (A.K. Burroughs).7472 6226.

tigliano R et al., Seeding following percutaneous diagnostic and therapeutic approaches forthe risk and the outcome? ..., Cancer Treat Rev (2007), doi:10.1016/j.ctrv.2007.04.001

mailto:[email protected]






2 R. Stigliano et al.

ARTICLE IN PRESS

Please cite this article in press as: Shepatocellular carcinoma. What i

used with biopsy and 0.61% (0–5.56%) for RFA without biopsy, 0.95% (0–12.5%) for RFA withbiopsy and 0.72% (0–10%) for liver nodules (including non-HCC nodules) biopsied and ablated.Conclusion: Risk of seeding with HCC is substantial and appears greater with using diagnosticbiopsy alone compared to therapeutic percutaneous procedures. This risk is particularly rele-vant for patients being considered for liver transplantation.

�c 2007 Elsevier Ltd. All rights reserved.

Introduction

Hepatocellular carcinoma (HCC) is the fifth most commonmalignant disorder worldwide and causes nearly 1 milliondeaths a year. In Europe and USA the incidence of HCC isincreasing as a result of the high prevalence of hepatitis Cand is now the leading cause of death amongst cirrhotic pa-tients in Europe.1–3

The improved survival in patients treated for portalhypertensive related complications of cirrhosis has led topatients being at greater risk of developing HCC4. In addi-tion, the increased use of ultrasound surveillance programsfor HCC has led to its increased detection which will furtherincrease with developments in imaging technology5.Although diagnostic biopsy rates have diminished, biopsyis still used6 and is often considered mandatory for patientmanagement by oncologists. Furthermore liver biopsy isbeing reconsidered as an integral part of HCC management,as it may give prognostic information7–9 regarding malig-nant potential and may be required for allocation systemsin liver transplantation programmes. There has also beenincreasing use of percutaneous techniques to treat HCC:small tumors can be ablated either by percutaneous ethanolinjection (PEI), radiofrequency ablation (RFA) or othertechniques microwave, laser, cryotherapy. Statistical mod-eling has shown that such interventions are cost effectiveespecially in the transplant setting when the expected wait-ing time for liver transplantation is longer than 6 months.6

Percutaneous ablation is the best treatment option for pa-tients with small HCC who are not suitable for resection ortransplantation; in some Japanese centers this is offered asthe first therapeutic option.10–12 There are no randomizedcontrol trials (RCT) comparing local ablations to resectionbut recurrence rates are similar.6 Biopsy and percutaneousablative therapies may result in needle track seeding. Someclinicians have advocated using PEI or RFA after a biopsy ofthe suspected HCC nodule to minimize the risk of seeding.

PEI is usually performed under ultrasound guidance. It isthe best known and best studied approach13 and as it is inex-pensive and highly effective for small HCC and has a low rateof adverse effects. PEI involves multiple passes which poten-tially increases the risk of seeding. Some recurrences occur inthe vicinity of the treated nodule and are due to the presenceof microscopic satellites not included in the ablation zone.

Therefore RFA following a biopsy of a suspicious nodulehas been considered theoretically better than PEI in reduc-ing or removing the risk of seeding, because it results in alarger margin of destruction of normal parenchyma aroundthe HCC nodule, however this has not been proven. This po-tential advantage might be lost, due to the larger needlesused, as the risk of seeding is considered higher than usingPEI, although larger nodules are treated with RFA.

tigliano R et al., Seeding follos the risk and the outcome? ...,

In a recent review14 several mechanisms are suggested ascontributing to seeding after percutaneous RFA. Viabletumor cells may adhere to a biopsy needle or to electrodeduring its retraction. Tumor cells may also be carried intothe track with a little bleeding. Furthermore, cells may beforced into the track by sudden increase in intratumoralpressure which is frequently encountered during RFA. Final-ly, cells may be driven in, when saline is injected during orbefore RFA. Seeding has also been reported following lapa-roscopic surgery and microwave ablation,15 a series report-ing a 10% seeding rate in 20 patients.

This increased use of PEI, acetic acid, RFA and laser tech-niques, as well as the need of biopsy in some cases, hasincreased the likelihood of seeding as it is a known compli-cation of both diagnostic and therapeutic procedures in pa-tients with HCC. This is especially important in potentialcandidates for LT, as following percutaneous approachesseeding has been described after liver transplantation, sec-ondary just to pre-transplant diagnostic biopsy.16–18 HCCare particularly prone to seeding, as higher rates occur(0–5.1%)19 even when using fine needle aspiration cytology(FNAC), compared to seeding rates with other solid tumourssuch as pancreatic tumours 0.003–0.017% and with otherabdominal tumours 0–0.03%.20,21

What is the risk of seeding in HCC? Seeding of HCC wasfirst reported in 1983.22 Since then there have been only afew published prospective studies, several case reportsand small retrospective series. However the magnitude ofthe risk of seeding is currently unknown and it appears veryheterogenous. This is in part due to the lack of a standarddefinition of needle tract seeding and of the interval ofobservation required to diagnose it. Previous estimatesand reports of seeding are likely to underestimate currentrates as percutaneous techniques are increasingly used inpatients with well compensated cirrhosis, who will survivelonger as thus have longer periods to be at risk for seeding.Thus the real incidence of seeding with HCC is difficult toassess precisely. Some experts19,23 feel that there is asystematic underestimation of the risk of metastasis duringneedle biopsy, and it is possible that several seedingepisodes are not diagnosed or not reported thus leading toan underestimation of the risk of seeding in the literature.

Thus we have evaluated the literature concerning seed-ing in patients with HCC with the above considerations inmind.

Methods

A Pub Med search of both English and non-English literaturewas done using the keywords HCC, seeding, implantation,tract, complication, hazards, dissemination, spreading,deposits, adverse events, fine needle biopsy, fine needle

wing percutaneous diagnostic and therapeutic approaches forCancer Treat Rev (2007), doi:10.1016/j.ctrv.2007.04.001

Seeding following percutaneous diagnostic and therapeutic approaches for hepatocellular carcinoma 3

ARTICLE IN PRESS

aspiration. Abstracts from recent international meetingswere also analysed.

For the purposes of the study, we defined needle tractseeding of liver tumour as the development of new neoplas-tic disease outside the liver capsule, either in the subcuta-neous soft tissue or peritoneal cavity.24 The literaturesearch resulted 179 seeding reports from January 1983 toFebruary 2007. We specifically looked for HCC seeding afterbiopsy and or local ablation therapy (LAT) using PEI, aceticacid and hot saline injection, RFA, microwave, laser or cryo-therapy excluding other settings. Where documented wenoted the total number of the cohort which was biopsiedto define the denominator so as to establish a percentagerisk of seeding. We recorded the needle biopsy, size, type(aspiration/trucut), the approach (epigastic/intercostal),whether more than 1 cm of parenchyma was present be-tween the biopsied nodule and the liver capsule, the num-ber of passes, histology and differentiation of tumours andtreatment of lesion, interval to seeding, site of seeding, sizeof seeding, aFP at biopsy and at seeding time, histology ofseeding, treatment of seeding, outcome of patient, andwhether the seeding occurred before and/or after livertransplantation. We excluded the paper by Agoneseet al.25 because they reported only intraparenchymal seed-ing which occurred in nine cases of rapid unexpected spreadof HCC after radiofrequency thermal ablation.

Results

In total there were 99 papers describing seeding in 179cases: 66 followed liver biopsy, 26 followed PEI, 1 followedmicrowave, 22 followed RFA, and 64 followed a combinationof biopsy and percutaneous treatment (5 microwave; 33 PEI;26 RFA).

The cohort number biopsied and/or treated by percuta-neous approach was only reported in 41 papers. The riskof seeding in these groups is shown in Table 1, ranging from0.61% to 2.29% with a median of 0.95%.

Seeding following percutaneous diagnosticapproaches for HCC

A total of 41 papers were found describing 66 cases of seed-ing following biopsy: 39 cases belonged to series of patients

Table 1 Summary of series documenting seeding of HCC followincohort size is specified

Biopsy alone Biopsy + PEI RF

Series 14 4 5Total patients from series 2242 766 15No. seeding from series 43 11 19Risk of seeding mean 3.17% 1.4% 1.

Risk of seeding median 2.29% 1.4% 0.

PEI = percutaneous ethanol injection of proven HCC.RFA = radiofrequency ablation of proven HCC.LAT̂ = local ablation techniques mainly with biopsy of the nodule.

Please cite this article in press as: Stigliano R et al., Seeding followhepatocellular carcinoma. What is the risk and the outcome? ...,

and 27 were case reports (three abstracts not in Pub-Med17,22,26–48).

The risk of seeding was reported in 14 series ranging from0% to 11%, median value 2.29% (see Table 2). The size ofbiopsied nodule was reported in 35 cases out of 66, mediandiameter 40 mm (range 10–145 mm).

The serum concentration of aFP was specified in 15 re-ports, 7 had a normal value, and in 13 the median valuewas 317.9 ng/ml (range 4–26,000).

Liver specimen was obtained by needle core biopsy (NCB)or fine needle aspiration biopsy (FNAB). NCB was performedin 76% (n = 39: Sonopsy or Trucut or Chiba or Vim Silvermannneedle) and FNAB was performed in 24 (n = 12). Only 1 re-port did not specify the needle used, in only five caseswas the amount of parenchyma between the lesion andthe capsule specified as being greater than 1 cm. The ap-proach was specified in six cases: 1 subcostal; 3 intercostal;1 lateral; 1 epigastric. The number of passes was specifiedin 20 seedings; with a median value of 2 (range 1–3).

The histology of the biopsied HCC was reported in 29: se-ven were well differentiated HCC, five were moderate dif-ferentiated HCC and three were poor differentiated HCC;the degree of differentiation was not specified in 14 of 29HCC.

The sites of the seeding were thoracic wall in 17; subcu-taneous in the epigastrium in 3; subcutaneous and dia-phragm in 1; subcutaneous localisations (not specified ifthoracic wall) in 21; intercostal muscle in 2; rectus muscleof the abdomen in 1; and intraperitoneal in 3.

The histology of seeding was reported in only 17 reports:12 did not describe the differentiation, whereas three doc-umented moderate and two poor differentiation.

The treatment of seeding (n = 36) was resection in 22cases, chemoembolisation in 9, chemotherapy in 2, notreatment in 3. The aFP at diagnosis of seeding was re-ported in 13 cases; three were normal; two had raised val-ues without specifying titres, and two had raised values 26and 74 fold. In the other six cases, the median value ofaFP was 121 ng/ml (range 65–521).

The median time to diagnosis of seeding was 13 months(range 1–58). The outcome of patients was only given fora few: three were waiting for liver transplantation and itwas not possible to go ahead because of the seeding so theywere treated with palliative options.17 Thirteen patientshad died (five due to HCC spread, one accident, remainder

g biopsy or percutaneous therapeutic techniques in which the

A Biopsy + RFA Nodules and LAT̂ Total

8 10 4125 2218 3725 9783

24 28 12573% 2.5% 2.04% Overall 1.27%

(125/9783)

61% 0.95 0.72% 0.95&

ing percutaneous diagnostic and therapeutic approaches forCancer Treat Rev (2007), doi:10.1016/j.ctrv.2007.04.001

Table 2 Individual reports of seeding after percutaneous biopsy alone in series of HCC patients in which the size of the cohortbiopsied is reported

Author Year No. patients No. seeding Percentage with seeding

1 Huang62 1996 420 8 1.902 Kanematsu94 1997 50 2 43 Chapoutot95 1999 150 4 2.674 Dumortier16 2000 31 1 3.225 Kim96 2000 205 7 3.416 Takamory97 2000 59 3 5.087 Durand56 2001 122 2 1.648 Kosugu12 2004 372 6 1.619 Ng98 2004 91 1 1.0910 Shuto74 2004 480 5 1.0411 Wang99 2005 90 0 012 Saborido18 2005 26 2 7.6913 Herrero63 2006 18 2 1114 Maturen65 2006 128 0 0

Total number of patients 2242 39 2.29 Median3.17 MeanOverall 1.73(=39/2242 · 100)


ARTICLE IN PRESS

not specified) and 20 were alive. The median time of followup was 16 months (range 1–108).

Numerous factors have been related to the risk of neo-plastic dissemination after invasive procedures: largerdiameter needles;49 more passes;49 superficial location ofthe tumour in the liver; intrinsic metastatic property ofthe tumour50 related to either or both tumour size oraggressivity and patients’ immunodepression51 resulting ina reduced tumour surveillance. An exponential increase ofseeding with increasing numbers of these risk factors hasbeen reported. For example, increasing the needle diame-ter by a factor of two, increased the seeding by a factorof 60, without improving diagnostic efficiency.52,53

In an animal model Ryd et al. showed that FNAC, consid-ered less invasive than biopsy, induced the implantation ofbetween 1000 and 100,000 cells along the needle tract.54

Leakage of tumour cells along the needle track may resultin local tumour implantation, and extravasation of tumourcells into the blood stream and lymphatics may cause dis-tant metastases as evaluated by mRNA expression of a-feto-protein in circulating mononuclear cells.55 Whether thiscould have clinical relevance in cirrhotic patients is notknown.

Fine needle aspiration could give a smaller risk of seedingthan biopsy. However although the specificity and the posi-tive predictive value of FNAC for focal liver lesions is veryhigh, the sensitivity ranges widely between 67% and93%53,56–60 and thus diagnostic accuracy is less than for his-tology.53,59 In addition cytology is less able to distinguishmalignant from non-malignant nodules when the nodule is2 cm or less.61,62 Nowadays biopsy is often advocated forthese small suspicious nodules. In addition biopsy is no bet-ter than the diagnostic accuracy of radiological investiga-tions which have up to 91% sensitivity.58 Similar results toFNAC have been achieved employing only radiological tech-niques (ultrasound, abdominal lipiodol-CT, MRI) and biolog-ical markers (AFP, CEA, CA19-9, PIVKA). Thus fine needleaspiration cannot substitute histology. However even

Please cite this article in press as: Stigliano R et al., Seeding follohepatocellular carcinoma. What is the risk and the outcome? ...,

though NCB rather than FNAC is the preferred method todiagnose HCC ‘‘histologically’’, the accuracy of NCB maynot be as high as is commonly perceived.

Recently Saborido et al.18 reported that in 27% ofnodules (size 4.6 ± 3.1 cm) biopsy was not diagnostic, andwas only suggestive of HCC in 15.4%. Herrero et al.63 re-ported similar results for nodules above 2 cm diameter:19.2% of biopsies were not diagnostic and 7.8% were falsenegatives.

In a recent series64 of fine needle biopsies, Caturelliet al. pointed out that for HCC nodules 2 cm or less in diam-eter, the diagnostic accuracy decreased with diminishingsize of nodules. The false negative cases were 10.6% andthe specimen was inadequate for diagnosis in 2.9%. Repeat-ing the biopsy did not improve the diagnosis that remainedin doubt in two-thirds of cases.

There are no RCT to evaluate the risk of seeding in thispopulation and we do not know if all the precautions toavoid seeding were applied.

Maturen et al. retrospectively evaluated a 6 year periodwith respect to the incidence of HCC seeding by applyinga different technique65 using a coaxial cutting needle. Noseedings were identified in 101 patients who had more than30 days of clinical and/or imaging follow-up after biopsy(mean, 410 days; range, 33–1.989).

They used a needle introducer that remains in positionduring multiple cutting needle passes, this may protectsthe normal tissue along the track and may reduce seeding.This would have particular importance for patients withstage I-II HCC, for whom liver transplantation can becurative.

Seeding following percutaneous therapeuticapproaches without biopsy for HCC

PEIWe found a total of 16 papers describing 26 cases of seedingfollowing therapeutic PEI without biopsy. There is only one



ARTICLE IN PRESS

cohort of patients66 with no case of seeding and 26 case re-ports.51,67–79

There is only one cohort of patients66 without seeding.The maximum diameter of nodule treated was specified

only in 10 reports with a median value of 32.5 mm range(10–50). The aFP was not available. No reports specifiedthat the amount of parenchyma between the lesion andthe liver capsule was more than 1 cm. The sites of seedingwere abdominal wall in two patients, cutaneous in 1; chestwall in 3; intraperitoneal in 6.

The size of seeding was reported in nine cases: mediansize was 35 mm (range 15–50). The median time from PEIto detecting seeding (n = 6) was 6 months (range 2–48).The histology of seeding was one poor, two moderate. Thetreatment was resection in 11 patients, PEI in 2; chemoem-bolisation treatment in 1. The outcome was only given in afew: five patients had died and seven were alive with a med-ian follow up (in 12) of up to 15 months (range 3–49) afterseeding.

Multiple intraperitoneal seeding (largest nodule 5 cm)was also reported after combined treatment by microwaveand PEI of a superficial HCC whose diameter was 40 mm(at time of biopsy the aFP was normal). aFP was 193 ng/ml at diagnosis of seeding. The histology showed poordifferentiation. Two months after detecting seeding,chemoembolization combined with cisplatin in the perito-neal cavity and resection was performed, but the patientdied after 3 month after surgery.

Seeding may occur because despite the use of smallerneedles and injecting alcohol down the track, multiplepasses are used.

RFASome authors suggest that the puncture-related seeding isusually restricted to poorly differentiated tumours and toperipheral tumors that cannot be approached through arim of non-tumoral liver.80,81 Therefore RFA of tumourshas been considered theoretically better than PEI in reduc-ing or removing the risk of seeding, because it results in alarger margin of destruction of normal parenchyma aroundthe HCC nodule, however this has not been proven.

A total of eight papers were found, describing 21 cases ofseeding followed RFA.

Five cohorts were reported with a total number of 1525patients and 19 seeding (Table 4), giving a mean value forrisk of seeding of 1.73% and a median value of 0.61% (range0–5.56). In addition two case reports82,83 documented seed-ing. Subcapsular location and poor differentiation have

Table 3 Individual reports of seeding after biopsy and percutanethe size of the cohort is specified

Authors Year No. patients

1 Ishii11 1998 3482 Shiina100 2005 1143 Ebara101 2005 270

Total 766


been associated with increased risk of peritonealseeding.80,81

The size of nodule treated was only specified in 2:3 and4 cm diameter. The aFP was documented in two cases48.6 and 1243 ng/ml. The number of sessions was specifiedonly in three cases: there was a single treatment session inone; two treatment sessions in the second and four sessionsbefore seeding in the third case. Site of seeding was oneperitoneal and one subcutaneous; in the others it was notspecified. The time to diagnosis of seeding was 8 and7 months in two cases. The outcome is given only in onecase: the patient was still alive after resection of theseeded HCC.

Seeding following combined percutaneousdiagnostic and therapeutic approaches for HCC

PEI after a biopsy has been advocated to minimize the riskof seeding because of the smaller needle, the effect of alco-hol, and perhaps because smaller nodules are trea-ted.13,50,52,53,84 Furthermore some authors suggests thatRFA following a biopsy of a suspicious nodule is theoreticallybetter than PEI in reducing or removing the risk of seeding,because it results in a larger margin of destruction of normalparenchyma around the HCC nodule, however this has notbeen proven.

There were 64 seedings reported after both biopsy andpercutaneous treatment (5 microwave; 33 PEI; 26 RFA).Unfortunately none of the authors specified the timing be-tween the biopsy and the treatment procedures but one as-sumes these were either combined procedures or veryclosely related in time.

The histological differentiation of the nodule was avail-able in 37 cases (16 well, 10 moderate, 11 poorly differen-tiated HCC). In 12, histology after seeding was alsoavailable: four well; five moderate; three poor differentia-tions. A worsening of differentiation was documented inonly one case.

Median maximum diameter of the biopsied nodule was25 mm (range 8–84) reported for 31 nodules. The noduleof seeding had a median size of 25 mm (range 2–300). TheaFP titre was reported in four cases and only one reportspecified that the amount of parenchyma between the le-sion biopsied and treated and the liver capsule as more than1 cm. The site of seeding was chest wall in four cases, twocutaneous, one diaphragm, one pleural and 10 subcutane-ous. The interval between the percutaneous approach andseeding was 7 months (range 2–54) in 30 cases. The treat-

ous ethanol injection (PEI) therapy for HCC in series in which

No. seeding Percentage with seeding

4 1.152 1.755 1.85

11 1.4 Median1.43 MeanOverall 1.43 = 11/766 · 100)


Table 4 Individual reports of seeding after (Panel A) radiofrequency ablation (RFA) therapy and (Panel B) biopsy andradiofrequency ablation (RFA) therapy for HCC in which the size of the cohort is specified

Author Year No. patients No. seeding Percentage with seeding

Panel A1 Bolondi102 2001 63 0 02 Goldberg103 2001 450 1 0.223 Livraghi104 2001 330 2 0.614 Bonny105 2003 18 1 5.565 Akahane10 2005 664 15 2.26

Total 1525 19 0.61 Median1.73 MeanOverall 1.23 = 43/3480 · 100

Panel B1 Curley106 2000 110 0 02 Llovet81 2001 32 4 12.503 Liu107 2003 33 1 3.034 Poon108 2003 80 0 0.005 Livraghi109 2005 1314 12 0.916 Shiina100 2005 118 3 2.547 Pompili110 2005 30 0 0.008 Baldan111 2006 401 4 0.99

Total 2218 24 0.95 Median2.50 MeanOverall1.08 = 24/2218 · 100


ARTICLE IN PRESS

ment of seeding was resection for 20 patients, PEI for threepatients, radiotherapy combined with TAE for one patientand four patients were not treated. The outcome was givenonly in 25 reports: six patients had died and 19 patientswere still alive with a median follow up (in 14) of up to9.5 months (range 2–37) after treatment of seeding.

There are 10 series reporting cohorts in which liver nod-ules were treated by percutaneous techniques (1 biopsy, 1PEI, 1 PEI + RFA, 7 RFA) without specifying if these were alsoclinically or histologically hepatocellular carcinoma in allcases (Table 5). The risk of seeding in this heterogenousgroup (which included HCC) ranged from 0% to 10% with amedian value of 0.72%, i.e. lower than the reports whichspecified that all patients had HCC.

Although seeding after a combined biopsy and PEI cannotbeen distinguished as secondary to the percutaneous biopsyrather than to PEI itself, it is not universal practice to usePEI or (RFA) without previous biopsy.

In addition, it has been documented that the location ofthe nodule may not materially diminish the risk of seedingas Di Stasi et al.70 reported seven cases after PEI followingbiopsy and treatment of deeply located HCC, as well as fol-lowing biopsy of small HCC.

No correlation between seeding and the differentiationof HCC could be made in our review, and this cannot beknown about before biopsy. Tumour seeding has been re-ported after PEI for poorly differentiated HCC,70 as well asfor well differentiated ones using histological assessment.Whether larger volumes of ethanol injected during each ses-sion or the use of another agent, such as acetic acid wouldreduce the risk of seeding is not known. The only data avail-


able regard a case of parietal peritoneal seeding reported ina series of 108 patients (0.93%) after a single session PEI allwith large HCC85 suggesting size of the treated nodule maybe relevant.

There is only one prospective study by Riccardi et al.86

which assessed the rate of seeding after biopsy with initialPEI (subsequently repeated) during which alcohol injectionof the biopsy track was performed. Ultrasound follow-upwas used at the point of puncture (marked with a tattoo).It reported two cases of seeding in 29 patients (6.8%) within12 months. This could represent an accurate estimate ofseeding but it is clearly a higher estimate than the 1.4% re-ported in cohort studies when PEI followed a liver biopsy ofa suspicious nodule, subsequently confirmed to be HCC. Thisagain suggests that the literature underestimates the risk ofseeding.82 However considering RFA, even if discounting thehigh seeding rate of 12.5% reported by Llovet et al.81 be-cause of the superficial position of the nodules biopsied,the seeding rate ranged from 0% to 5.1%19 with a median va-lue of 0.61% when only RFA is performed87,88 and 0.95%when RFA follows the biopsy.

Discussion

Malignant seeding is a well known complication of both diag-nostic and therapeutic procedures in patients with HCC, butthe clinical issue is the magnitude of the risk of seeding andits consequences.

It seems that biopsy of HCC confers a greater risk ofseeding compared to percutaneous local ablation alone.


Table 5 Individual reports of seeding after radiofrequency (RFA) or PEI for liver nodules (not all specified as HCC histologically)in which the size of the cohort biopsied is reported

Author Year No. patients total No. seeding Percentage seeding Only RFA

1 Matsukawa15 1997 Microwave 20 2 102 Mueller112 b 1995 PEI 141 2 1.42 1.423 Wood113b 2000 RFA 84 0 0.00 0.004 Livraghi80 a 2003 RFA 2320 12 0.52 0.525 deBaere114 2003 RFA 226 1 0.44 0.446 Riccardi86 2004 RFA/PEI 239 2 0.84 0.847 Buscarini115 2004 RFA 166 1 0.60 0.608 Jaskolka24 2005 RFA 155c 6 3.90 3.909 Chen116 2006 RFA 332c 1 0.3 0.310 De Baere117 2006 RFA 42 1 2.38 2.38

3725 28 0.72 Median 0.60% Median

Total no. patients Total no. seedings 2.04 Mean 1.16 MeanOverall0.75 = 28/3725 · 100

a In this series, 12 cases of seeding were reported in 2320 focal liver lesions; four seedings definitely followed biopsy of HCC.b Not biopsied.c HCC.


ARTICLE IN PRESS

The local treatment of HCC may have a greater risk of seed-ing compared to other solid liver tumours (Tables 1–5), as inthe reports which included non-HCC liver tumours the risk ofseeding ranged from 0% to 10% with a median value 0.72%.

The consequences of seeding as well as the risk areimportant when considering biopsy and or treatment of sus-pected or proven HCC.

From the data available, when the number in each cohortbiopsied or treated was reported, we calculated an overallrisk of seeding of 1.22%. It ranges from a median of 0.99%when solely RFA treatment is performed, to 1.4% when onlyPEI treatment of HCC is performed.

In 1995 there was the first report of seeding in a patientwho underwent liver transplantation,17 followed a few yearslater by a second report.42 These reports raise the issue ofseeding in a group of patients in whom liver transplantationis undertaken for oncological cure. Seeding represents afailure of this treatment strategy. Further series of trans-planted patients for HCC have been published focusing onthe seeding issue18,63,89 which suggest this is not necessarilyan event so rare that it can be discounted. Four seedingswere reported in 84 patients (4.76%), and three more see-dings were reported after liver transplantation and docu-mented as case reports.15,16,42

In 1991 Roussel suggested some procedures to minimizeseeding if one is compelled to biopsy:90 (1) the needle biop-sies of solid masses should be performed by trained teams;(2) only when taking decisions about the patient’s manage-ment; (3) through normal parenchyma, whenever possible,respecting anatomical boundaries as to ensure that the nee-dle track is resected with the metastases; (4) always with afine non-cutting needle; (5) the sampling has to be done un-der suction, which must be maintained when withdrawingthe needle; (6) the sample quality has to be checked laterto keep the number of passes to the very minimum. How-ever these are difficult to achieve in cirrhotic patients,and have not been validated as substantially reducing therisk of seeding of HCC. It is not clear from the reports we


evaluated whether the optimal technique was used as sug-gested by Roussel and thus the risk of seeding cannot beattributed conclusively to differences in technique ofbiopsy. An important issue is how well the reported casesof seeding document the surveillance if any, for the occur-rence of seeding following percutaneous, diagnostic and/ortherapeutic procedures. We can add a further consider-ation: in how many cases was a diagnostic biopsy neededin the first place? In transplant candidates it may haveimportant consequences19 to misdiagnose a tumour nodulebut also it is notable that many cases of seeding were per-formed in patients with a diagnostic value of aFP of over200 ng/ml, 400 ng/ml91 and diameter >2 cm which usuallyhave pathognonomic hypervascularity on imaging.91,92

Some clinicians have advocated using PEI or RFA after abiopsy of the suspected HCC nodule to minimize the riskof seeding. We found a shorter median time to seeding inRFA (7 months range 4–18 months) compared to PEI(11 months range 2–46 months) groups. From the dataavailable the median diameter of treated HCC was32.5 mm in the PEI; 30 mm15–75 in RFA after biopsy and40 mm in the biopsied group, so that despite a larger diam-eter nodule in the biopsied group seeding occurred earlier inthe treated group. The shorter interval to seeding with RFA,may be related to larger needles used and usually the largersize nodules which was treated.14 It is clear that well com-pensated cirrhotics with small nodules 2 cm diameter or lesswithout characteristic diagnostic imaging (i.e. no hypervas-cularity) can be managed, even if waiting for liver trans-plantation, with delayed diagnosis by a re-imaging every2–3 months.19 During this interval some patients with HCCwill achieve diagnostic imaging criteria91 and this will elim-inate them from consideration of a biopsy. These patients,providing the nodule remains single and is no more than4 cm diameter, or if more than one nodule develops thereare less than three nodules 63 cm diameter, will be at al-most zero risk of recurrence following liver transplanta-tion.93 This would be a safer strategy than to biopsy.



ARTICLE IN PRESS

Tumours larger than 2 cm diameter with characteristichypervascularity in cirrhotic patients do not need a biopsy,as diagnostic imaging criteria, are very sound and will im-prove.56 Conversely when nodules are less than 2 cm diam-eter, histology may not be sufficiently accurate and a risk ofseeding remains.

Therefore liver biopsy of HCC must be considered poten-tially hazardous due the risk of seeding, and also limited inits diagnostic accuracy, especially when the suspicious nod-ule may be 62 cm diameter. Combining a percutaneoustreatment with biopsy appears to reduce the risk of seedingcompared to biopsy alone.

There is current research evaluating several tissue mark-ers, as to whether they can improve the histological diagno-sis of HCC and/or give sufficiently reliable information onthe potential invasiveness of the tumour and thus prognosis.In this context the risk of seeding must be weighed againstpotential better outcomes resulting from knowing moreabout prognosis and/or diagnosis. Ideally randomized stud-ies should be performed in centres which currently routinelybiopsy nodules to assess if this extra information affectsoutcomes, as the risk of seeding will remain. The diagnosticvalue of biopsy with new tissue markers can be assessed ver-sus treating suspicious nodules, without prior biopsy, eitherwhen the nodule is first diagnosed, or following changes onimaging during follow up in cirrhotic patients.

In conclusion, this review of the risk of seeding in HCCsuggests seeding is clinically substantial. We believe thecurrent evidence suggests biopsy of HCC in cirrhotic pa-tients should not be performed in routine clinical practiceand be confined to research protocols.

References

1. Fattovich G, Stroffolini T, Zagni I, Donato F. Hepatocellularcarcinoma in cirrhosis: incidence and risk factors. Gastroen-terology 2004;127(5 Suppl 1):S35–50.

2. Benvegnu L, Gios M, Boccato S, Alberti A. Natural history ofcompensated viral cirrhosis: a prospective study on theincidence and hierarchy of major complications. Gut2004;53(5):744–9.

3. Fattovich G, Pantalena M, Zagni I, Realdi G, Schalm SW,Christensen E. Effect of hepatitis B and C virus infections onthe natural history of compensated cirrhosis: a cohort study of297 patients. Am J Gastroenterol 2002;97(11):2886–95.

4. Vong S, Bell BP. Chronic liver disease mortality in the UnitedStates, 1990–1998. Hepatology 2004;39(2):476–83.

5. Sangiovanni A, Del Ninno E, Fasani P, De Fazio C, Ronchi G,Romeo R, et al. Increased survival of cirrhotic patients with ahepatocellular carcinoma detected during surveillance. Gas-troenterology 2004;126(4):1005–14.

6. Llovet JM, Burroughs A, Bruix J. Hepatocellular carcinoma.Lancet 2003;362(9399):1907–17.

7. Chignard N, Bergenfeldt M. Proteomics for hepato-cellular carcinoma marker discovery. Gastroenterology2004;127(127):S120–5.

8. Lee JS, Thorgeirsson SS. Functional and genomic implicationsof global gene expression profiles in cell lines from humanhepatocellular cancer. Hepatology 2002;35(5):1134–43.

9. Lee JS, Chu IS, Heo J, Calvisi DF, Sun Z, Roskams T, et al.Classification and prediction of survival in hepatocellularcarcinoma by gene expression profiling. Hepatology2004;40(3):667–76.


10. Akahane M, Koga H, Kato N, Yamada H, Uozumi K, Tateishi R,et al. Complications of percutaneous radiofrequency ablationfor hepato-cellular carcinoma: imaging spectrum and man-agement. Radiographics 2005;25(Suppl 1):S57–68.

11. Ishii H, Okada S, Okusaka T, Yoshimori M, Nakasuka H,Shimada K, et al. Needle tract implantation of hepatocellularcarcinoma after percutaneous ethanol injection. Cancer1998;82(9):1638–42.

12. Kosugi C, Furuse J, Ishii H, Maru Y, Yoshino M, Kinoshita T,et al. Needle tract implantation of hepatocellular carcinomaand pancreatic carcinoma after ultrasound-guided percutane-ous puncture: clinical and pathologic characteristics and thetreatment of needle tract implantation. World J Surg2004;28(1):29–32.

13. Livraghi T, Giorgio A, Marin G, Salmi A, de SI, Bolondi L, et al.Hepatocellular carcinoma and cirrhosis in 746 patients: long-term results of percutaneous ethanol injection. Radiology1995;197(1):101–8.

14. Mulier S, Mulier P, Ni Y, Miao YDB, Marchal G, De Wever I,et al. Complications of radiofrequency coagulation of livertumors. Br J Surg 2002;89:1206–22.

15. Matsukawa T, Yamashita Y, Arakawa A, Nishiharu T, Urata J,Murakami R, et al. Percutaneous microwave coagulationtherapy in liver tumors. A 3-year experience. Acta Radiol1997;38(3):410–5.

16. Dumortier J, Lombard-Bohas C, Valette PJ, Boillot O, ScoazecF, Berger F, et al. Needle tract recurrence of hepatocellularcarcinoma after liver transplantation. Gut 2000;47(2):301.

17. McPeake J, Williams R. Liver transplantation for hepatocellu-lar carcinoma. Gut 1995;36(5):644–6.

18. Saborido BP, Diaz JC, Los Galanes SJ, Segurola CL, de UseraMD, Garrido MD, et al. Does preoperative fine needle aspira-tion-biopsy produce tumor recurrence in patients followingliver transplantation for hepatocellular carcinoma? TransplantProc 2005;37(9):3874–7.

19. Stigliano R, Burroughs AK. Should we biopsy each liver masssuspicious for HCC before liver transplantation?–no, pleasedon’t. J Hepatol 2005;43(4):563–8.

20. Fornari F, Civardi G, Cavanna L, Di Stasi M, Rossi S, Sbolli G,et al. Complications of ultrasonically guided fine-needleabdominal biopsy. Results of a multicenter Italian study andreview of the literature. The Cooperative Italian Study Group.Scand J Gastroenterol 1989;24(8):949–55.

21. Smith EH. Complications of percutaneous abdominal fine-needle-biopsy. Rev Radiol 1991;178:253–8.

22. Sakurai M, Okamura J, Seki K, Kuroda C. Needle tractimplantation of hepatocellular carcinoma after percutaneousliver biopsy. Am J Surg Pathol 1983;7(2):191–5.

23. Denton KJ, Cotton DW, Nakielny RA, Goepel JR. Secondarytumour deposits in needle biopsy tracks: an underestimatedrisk? J Clin Pathol 1990;43(1):83.

24. Jaskolka JD, Asch MR, Kachura JR, Ho CS, Ossip M, Wong F,et al. Needle tract seeding after radiofrequency ablation ofhepatic tumors. J Vasc Interv Radiol 2005;16(4):485–91.

25. Angonese C, Baldan A, Cillo U, D’Alessandro A, De Antoni M,De Giorgio M, et al. Complications of radiofrequency thermalablation in hepatocellular carcinoma: what about ‘‘explo-sive’’ spread? Gut 2006;55(3):435–6.

26. Dohmen K, Omori F, Yamano Y, Nagano M, Mizoguchi M, IwataY, et al. A case of primary hepatocellular carcinoma with skinmetastasis by aspiration biopsy. Nippon Shokakibyo GakkaiZasshi 1992;89(8):1538–42.

27. Hamazaki K, Matsubara N, Mori M, Gochi A, Mimura H, Orita K,et al. Needle tract implantation of hepatocellular carcinomaafter ultrasonically guided needle liver biopsy: a case report.Hepatogastroenterology 1995;42(5):601–6.

28. Isobe H, Imari Y, Sakai H, Sakamoto S, Nawata H. Subcuta-neous seeding of hepatocellular carcinoma following fine-



ARTICLE IN PRESS

needle aspiration biopsy. J Clin Gastroenterol 1993;17(4):350–2.

29. John TG, Garden OJ. Needle track seeding of primary andsecondary liver carcinoma after percutaneous liver biopsy.HPB Surg 1993;6(3):199–203.

30. Ka MM, Dangou JM, Fall B, Pouye A, Ndiaye MF, Diop TM, et al.Tumor seeding of the abdominal wall after fine needlecytologic puncture of the liver. Apropos of a case. AnnGastroenterol Hepatol (Paris) 1995;31(4):221–5.

31. Konno K, Ishida H, Hamashima Y, Komatsuda T, Sato M, FuruyaT, et al. Color Doppler findings of tumor seeding after US-guided liver tumor biopsy. Abdom Imaging 1999;24(4):401–3.

32. Liu YW, Chen CL, Chen YS, Wang CC, Wang SH, Lin CC. Needletract implantation of hepatocellular carcinoma after fineneedle biopsy. Dig Dis Sci 2007;52(1):228–31.

33. Lundstedt C, Stridbeck H, Andersson R, Tranberg KG, Andren-Sandberg A. Tumor seeding occurring after fine-needle biopsyof abdominal malignancies. Acta Radiol 1991;32(6):518–20.

34. Nanjoh H, Asanuma Y, Satoh T, Omokawa S, Koyama K,Hashimoto M, et al. A case of hepatocellular carcinomaimplanted at the chest wall by ultrasonic guided liver biopsy.Nippon Shokakibyo Gakkai Zasshi 1992;89(3):653–6.

35. Navarro F, Taourel P, Michel J, Perney P, Fabre JM, Blanc F,et al. Diaphragmatic and subcutaneous seeding of hepatocel-lular carcinoma following fine-needle aspiration biopsy. Liver1998;18(4):251–4.

36. Onodera H, Oikawa M, Abe M, Chida N, Kimura S, Satake K,et al. Cutaneous seeding of hepatocellular carcinoma afterfine-needle aspiration biopsy. J Ultrasound Med 1987;6(5):273–5.

37. Ozaki I. A case of hepatocellular carcinoma with a seedingaround the celiotomy scar resulting from aspiration biopsy.Gastroenterol Surg.

38. Park YM, Lee CD, Yoon KH, Han NI, Cho HM, Yook KS, et al. Acase of subcutaneous seeding of hepatocellular carcinomaafter fine needle aspiration biopsy. Kor J Int Med 1989;4(1):96–100.

39. Postovsky S, Elhasid R, Ben Arush MW, Berant M, Hayari L, OferA. Local dissemination of hepatocellular carcinoma in a childafter fine-needle aspiration. Med Pediatr Oncol 2001;36(6):667–8.

40. Quaghebeur G, Thompson JN, Blumgart LH, Benjamin IS.Implantation of hepatocellular carcinoma after percutaneousneedle biopsy. J R Coll Surg Edinb 1991;36(2):127.

41. Scheele J, Altendorf-Hofmann A. Tumour implantation fromneedle biopsy of hepatic metastases. Hepatogastroenterology1990;37:335–7.

42. Schotman SN, De Man RA, Stoker J, Zondervan PE, IjzermansJN. Subcutaneous seeding of hepatocellular carcinoma afterpercutaneous needle biopsy. Gut 1999;45(4):626–7.

43. Sood A, Midha V, Sood N. Needle track metastasis of primaryliver carcinoma after percutaneous needle aspiration. Am JGastroenterol 2002;97(3):765–6.

44. Tabata T, Usami M, Ohmiya H, Ohyanagi H, Saitoh Y, Ueda K,et al. A case report of cutaneous seeding of hepatocellularcarcinoma following ultrasound guided fine needle aspirationbiopsy. Nippon Shokakibyo Gakkai Zasshi 1990;87(5):1253–7.

45. Takahashi H, Konishi M, Nakagohri T, Inoue K, Takahashi S,Tanizawa Y, et al. Aggressive multimodal treatment forperitoneal dissemination and needle tract implantation ofhepatocellular carcinoma: a case report. Jpn J Clin Oncol2004;34(9):551–5.

46. Takayasu K, Moriyama N, Muramatsu Y, Makuuchi M, HasegawaN, Okazaki N, et al. The diagnosis of small hepatocellularcarcinomas: efficacy of various imaging procedures in 100patients. AJR Am J Roentgenol 1990;155(1):49–54.

47. Yamada N, Shinzawa H, Ukai K, Wakabayashi H, Togashi H,Takahashi T, et al. Subcutaneous seeding of small hepatocel-


lular carcinoma after fine needle aspiration biopsy. J Gastro-enterol Hepatol 1993;8(2):195–8.

48. Evans GH, Harries SA, Hobbs KE. Safety of and necessity forneedle biopsy of liver tumours. Lancet 1987;1(8533):620.

49. Smith EH. The hazards of fine-needle aspiration biopsy.Ultrasound Med Biol 1984;10:629.

50. Goletti O, Chiarugi M, Buccianti P, Macchiarini P. Subcutane-ous implantation of liver metastasis after fine needle biopsy.Eur J Surg Oncol 1992;18:636–7.

51. Shimada M, Maeda T, Saitoh A, Morotomi I, Kano T. Needletrack seeding after percutaneous ethanol injection therapy forsmall hepatocellular carcinoma. J Surg Oncol 1995;58(4):278–81.

52. Bedenne L, Mottot C, Courtois B, Meny B, Cauvin JM, Hillon P,et al. Is the Tru-Cut needle more efficient than the fineneedle in the diagnosis of hepatic lesions? Comparative studyof 45 echography-guided punctures. Gastroenterol Clin Biol1990;14(1):62–6.

53. Cochand-Priollet B, Chagnon S, Ferrand J, Blery M, Hoang C,Galian A. Comparison of cytologic examination of smears andhistologic examination of tissue cores obtained by fine needleaspiration biopsy of the liver. Acta Cytol 1987;31(4):476–80.

54. Ryd W, Hagmar B, Eriksson O. Local tumour cell seeding byfine-needle aspiration biopsy. A semiquantitative study. ActaPathol Microbiol Immunol Scand [A] 1983;91(1):17–21.

55. Louha M, Nicolet J, Zylberberg H, Sabile A, Vons C, Vona G,et al. Liver resection and needle liver biopsy cause hematog-enous dissemination of liver cells. Hepatology 1999;29(3):879–82.

56. Durand F, Regimbeau JM, Belghiti J, Sauvanet A, Vilgrain V,Terris B, et al. Assessment of the benefits and risks ofpercutaneous biopsy before surgical resection of hepatocel-lular carcinoma. J Hepatol 2001;35(2):254–8.

57. Herszenyi L, Farinati F, Cecchetto A, Marafin C, de Maria N,Cardin R, et al. Fine-needle biopsy in focal liver lesions: theusefulness of a screening programme and the role of cytologyand microhistology. Ital J Gastroenterol 1995;27(9):473–8.

58. Kinkel K, Lu Y, Both M, Warren RS, Thoeni RF. Detection ofhepatic metastases from cancers of the gastrointestinal tractby using noninvasive imaging methods (US, CT, MR imaging,PET): a meta-analysis. Radiology 2002;224:748–56.

59. Kuo FY, Chen WJ, Lu SN, Wang JH, Eng HL. Fine needleaspiration cytodiagnosis of liver tumors.Acta Cytol 2004;48(2):142–8.

60. Torzilli G, Minagawa M, Takayama T, Inoue K, Hui AM, KubotaK, et al. Accurate preoperative evaluation of liver masslesions without fine-needle biopsy. Hepatology 1999;30(4):889–93.

61. Tsai YY, Lu SN, Changchien CS, Wang JH, Lee CM, Eng HL,et al. Combined cytologic and histologic diagnosis of livertumors via one-shot aspiration. Hepatogastroenterology2002;49(45):644–7.

62. Huang G, Sheu JC, Yang PM, et al. Ultrasound-guided cuttingbiopsy for the diagnosis of hepatocellular carcinoma—a studybased on 420 patients. J Hepatol 1996;25:334–8.

63. Herrero Ji, Sangro B, Quiroga J, Inarrairaegur M, Pardo F,Rotellar F, et al. Is it necessary the pathological confirmationof hepatocellular carcinoma before liver transplantation. JHepatol 2006;C99. Ref. type: Abstract.

64. Caturelli E, Solmi L, Anti M, Fusilli S, Roselli P, Andriulli A,et al. Ultrasound guided fine needle biopsy of early hepato-cellular carcinoma complicating liver cirrhosis: a multicentrestudy. Gut 2004;53(9):1356–62.

65. Maturen KE, Nghiem HV, Marrero JA, Hussain HK, Higgins EG,Fox GA, et al. Lack of tumor seeding of hepatocellularcarcinoma after percutaneous needle biopsy using coaxialcutting needle technique. AJR Am J Roentgenol2006;187(5):1184–7.



ARTICLE IN PRESS

66. Castroagudin JF, Delgado M, Villanueva A, Bustamante M,Martinez J, Otero E, et al. Safety of percutaneous ethanolinjection as neoadjuvant therapy for hepatocellular carci-noma in waiting list liver transplant candidates. TransplantProc 2005;37(9):3871–3.

67. Arrive L, Vurgait A, Monnier-Cholley L, Lewin M, Balladur P,Poupon R, et al. Long-term follow-up after neoplastic seedingcomplicating percutaneous ethanol injection for treatment ofhepatocellular carcinoma. Eur Radiol 2002;12(1):74–6.

68. Cardinale L, Castiglione V. Subcutaneous neoplasm seedingalong the needle track. A rare complication following thepercutaneous alcoholization of a hepatocarcinoma. RadiolMed (Torino) 1995;89(4):557–9.

69. Casella G, Cacopardo E, Rovere G, Buda CA, Cascinu S, BaldiniV. Cutaneous seeding after ultrasound-guided percutaneousethanol injection for treatment of hepatocellular carcinoma.J Clin Ultrasound 2001;29(6):354–8.

70. Di Stasi M, Buscarini L, Livraghi T, Giorgio A, Salmi A, de SI,et al. Percutaneous ethanol injection in the treatment ofhepatocellular carcinoma. A multicenter survey of evaluationpractices and complication rates. Scand J Gastroenterol1997;32(11):1168–73.

71. Lee MC, Huang YL, Yang CH, Kuo TT, Hong HS. Cutaneousseeding of hepatocellular carcinoma due to percutaneousethanol injection and masquerading as a pyogenic granuloma.Dermatol Surg 2004;30(3):438–40.

72. Miki K, Makuuchi M, Takayama T, Matsukura A, Minagawa M,Kubota K, et al. Peritoneal seeding of hepatocellular carci-noma after ethanol injection therapy. Hepatogastroenterol-ogy 2000;47(35):1428–30.

73. Nagaoka Y, Nakayama R, Iwata M. Cutaneous seeding follow-ing percutaneous ethanol injection therapy for hepatocellularcarcinoma. Int Med 2004;43(3):268–9.

74. Shuto T, Yamamoto T, Tanaka S, Kanazawa A, Takemura S,Tanaka H, et al. Resection of needle-tract implantation afterpercutaneous puncture for hepatocellular carcinoma. J Gas-troenterol 2004;39(9):907–8.

75. Yamanaka N, Okamoto E, Fujihara S, Kato T, Fujimoto J,Oriyama T, et al. Do the tumor cells of hepatocellularcarcinomas dislodge into the portal venous stream duringhepatic resection? Cancer 1992;70(9):2263–7.

76. Zerbey AL, Mueller PR, Dawson SL, Hoover Jr HC. Pleuralseeding from hepatocellular carcinoma: a complication ofpercutaneous alcohol ablation. Radiology 1994;193(1):81–2.

77. Braune C, Widjaja A, Bartels M, Bleck JS, Flemming P, MannsMP, et al. Surgical removal of a distinct subcutaneousmetastasis of multilocular hepatocellular carcinoma 2 monthsafter initial percutaneous ethanol injection therapy. Z Gas-troenterol 2001;39(9):789–92.

78. Nakano H, Sasaki Y, Imaoka S, Furukawa H, Ishikawa O, KabutoT, et al. A case of hepatocellular carcinoma implantation tochest wall by percutaneous ethanol injection therapy. Gan ToKagaku Ryoho 1994;21(13):2350–3.

79. Yano S, Nakamura K, Yamane K, Kakinuma T, Asahina A,Tamaki K. Subcutaneous metastasis following percutaneousethanol injection therapy for hepatocellular carcinoma. ActaDerm Venereol 2001;81:213–4.

80. Livraghi T, Solbiati L, Meloni MF, Gazelle GS, Halpern EF,Goldberg SN. Treatment of focal liver tumors with percuta-neous radio-frequency ablation: complications encountered ina multicenter study. Radiology 2003;226(2):441–51.

81. Llovet JM, Vilana R, Bru C, Bianchi L, Salmeron JM, Boix L,et al. Increased risk of tumor seeding after percutaneousradiofrequency ablation for single hepatocellular carcinoma.Hepatology 2001;33(5):1124–9.

82. Shakado S, Saitsu H, Watanabe J, Takemoto R, Fukumori K,Akasu I, et al. A case of intrahepatic dissemination ofhepatocellular carcinoma after radiofrequency ablation ther-


apy with expandable needle electrode. Nippon ShokakibyoGakkai Zasshi 2002;99(12):1493–7.

83. Decker GA, Gores GJ, Roberts LR. Tumor seeding complicatingradiofrequency ablation of hepatocellular carcinoma. J Hep-atol 2003;38(5):692.

84. Smith EH. Complications of percutaneous abdominal fine-needle biopsy. Rev Radiol 1991;178(1):253–8.

85. Livraghi T, Benedini V, Lazzaroni S, Meloni F, Torzilli G,Vettori C. Long term results of single session percutaneousethanol injection in patients with large hepatocellular carci-noma. Cancer 1998;83(1):48–57.

86. Riccardi L, Pompili M, Guzzo Pirozzo M, Elia F, Arancio F,Gismant M, et al. Prospective study of neoplastic seeding inthe abdominal wall. XVI European Congress of Ultrasound inMedicine and Biology. Giornale Italiano di Ecografia2004;126(S 2):A119.

87. Roussel F, Dalion J, Benozio M. The risk of tumoral seeding inneedle biopsies. Acta Cytol 1989;33(6):936–9.

88. Saborido BP, Diaz JC, Los Galanes SJ, Segurola CL, de UseraMD, Garrido MD, et al. To biopsy or not to biopsy? LiverTransplant Worldwide 2006;12:895–6.

89. Pompili M, Mirante VG, Rondinara G, Fassati LR, Piscaglia F,Agnes S, et al. Percutaneous ablation procedures in cirrhoticpatients with hepatocellular carcinoma submitted to livertransplantation: assessment of efficacy at explant analysis andof safety for tumor recurrence. Liver Transplant 2005;11(9):1117–26.

90. Roussel F. Fine needle cytology should replace large boreneedle biopsy in the diagnosis of abdominal tumors. Gastro-enterol Clin Biol 1991;15(4):358–9.

91. Bruix J, Sherman M. Management of hepatocellular carci-noma. Hepatology 2005;42(5):1208–36.

92. Bruix J, Sherman M, Llovet JM, Beaugrand M, Lencioni R,Burroughs AK, et al. Clinical management of hepatocellularcarcinoma. Conclusions of the Barcelona-2000 EASL confer-ence. European Association for the Study of the Liver. JHepatol 2001;35(3):421–30.

93. Grasso A, Stigliano R, Morisco F, Martines H, Quaglia A, DhillonAP, et al. Liver transplantation and recurrent hepatocellu-lar carcinoma: predictive value of nodule size in a retrospec-tive and explant study. Transplantation 2006;81(11):1532–41.

94. Kanematsu M, Hoshi H, Takao H, Sugiyama Y. Abdominal walltumor seeding at sonographically guided needle-core aspira-tion biopsy of hepatocellular carcinoma. AJR Am J Roentgenol1997;169(4):1198–9.

95. Chapoutot C, Perney P, Fabre D, Taourel P, Bruel JM, LarreyD, et al. Needle-tract seeding after ultrasound-guided punc-ture of hepatocellular carcinoma. A study of 150 patients.Gastroenterol Clin Biol 1999;23(5):552–6.

96. Kim SH, Lim HK, Lee WJ, Cho JM, Jang HJ. Needle-tractimplantation in hepatocellular carcinoma: frequency and CTfindings after biopsy with a 19.5-gauge automated biopsy gun.Abdom Imaging 2000;25(3):246–50.

97. Takamory R, Wong LL, Dang C, Wong L. Needle tractimplantation from hepatocellular cancer: Is needle biopsy ofthe liver always necessary? Liver Transplant 2000;6:67–72.

98. Ng KK, Poon RT, Lo CM, Liu CL, Lam CM, Ng IO, et al. Impactof preoperative fine-needle aspiration cytologic examinationon clinical outcome in patients with hepatocellular carcinomain a tertiary referral center. Arch Surg 2004;139(2):193–200.

99. Wang CW, Lin ZY, Chuang WL, Wang LY, Yu ML, Chen SC, et al.Safety of fine-needle aspiration in patients with small hepa-tocellular carcinoma. Hepatol Res 2005;31(1):31–5.

100. Shiina S, Teratani T, Obi S, Sato S, Tateishi R, Fujishima T,et al. A randomized controlled trial of radiofrequency abla-tion with ethanol injection for small hepatocellular carci-noma. Gastroenterology 2005;129(1):122–30.



ARTICLE IN PRESS

101. Ebara M, Okabe S, Kita K, Sugiura N, Fukuda H, Yoshikawa M,et al. Percutaneous ethanol injection for small hepatocellularcarcinoma: therapeutic efficacy based on 20-year observa-tion. J Hepatol 2005;43(3):458–64.

102. Bolondi L, Gaiani S, Celli N, Piscaglia F. Tumor disseminationafter radiofrequency ablation of hepatocellular carcinoma.Hepatology 2001;34(3):608-1.

103. Goldberg SN, Solbiati L. Tumor dissemination after radiofre-quency ablation of hepatocellular carcinoma. Hepatology2001;34(3):609-1.

104. Livraghi T. Tumor dissemination after radiofrequency ablationof hepatocellular carcinoma. Hepatology 2001;34(3):608–9.

105. Bonny C, Abergel A, Gayard P, Chouzet S, Ughetto S, Slim K,et al. Radiofrequency ablation of hepatocellular carcinomain patients with cirrhosis. Gastroenterol Clin Biol 2002;26(8–9):735–41.

106. Curley SA, Izzo F, Lee M, Vauthey JN, Vallone P. Radiofre-quency ablation of hepatocellular carcinoma in 110 patientswith cirrhosis. Ann Surg 2000;232(3):381–91.

107. Liu C, Frilling A, Dereskewitz C, Broelsch CE. Tumor seedingafter fine needle aspiration biopsy and percutaneous radio-frequency thermal ablation of hepatocellular carcinoma. DigSurg 2003;20(5):460–3.

108. Poon RT, Ng KK, Lam CM, Ai V, Yuen J, Fan ST. Radiofrequencyablation for subcapsular hepatocellular carcinoma. Ann SurgOncol 2004;11(3):281–9.

109. Livraghi T, Lazzaroni S, Meloni F, Solbiati L. Risk of tumourseeding after percutaneous radiofrequency ablation for hepa-tocellular carcinoma. Br J Surg 2005;92(7):856–8.

110. Pompili M, Mirante VG, Rondinara G, Fassati LR, Piscaglia F,Agnes S, et al. Percutaneous ablation procedures in cirrhotich


patients with hepatocellular carcinoma submitted to livertransplantation: assessment of efficacy at explant analysis andof safety for tumor recurrence. Liver Transplant 2005;11(9):1117–26.

111. Baldan A, Marino D, De Giorgio M, Angonese C, Cillo U,D’Alessandro A, et al. Percutaneous radiofrequency thermalablation for hepatocellular carcinoma. Aliment PharmacolTher 2006;24(10):1495–501.

112. Mueller PR, Charboneau JW, Chezmar JL, Unger EC, Botet JF,Kane RA. North American experience with percutaneousalcohol ablation in 141 patients: complication and overview.Radiology 1995;197, 200. Ref. type: Abstract.

113. Wood TF, Rose DM, Chung M, Allegra DP, Foshag LJ, Bilchik AJ.Radiofrequency ablation of 231 unresectable hepatic tumors:indications, limitations, and complications. Ann Surg Oncol2000;7(8):593–600.

114. de Baere T, Risse O, Kuoch V, Dromain C, Sengel C, Smayra T,et al. Adverse events during radiofrequency treatment of 582hepatic tumors. AJR Am J Roentgenol 2003;181(3):695–700.

115. Buscarini E, Buscarini L. Radiofrequency thermal ablation withexpandable needle of focal liver malignancies: complicationreport. Eur Radiol 2004;14(1):31–7.

116. Chen MH, Wei Y, Yan K, Gao W, Dai Y, Huo L, et al. Treatmentstrategy to optimize radiofrequency ablation for liver malig-nancies. J Vasc Interv Radiol 2006;17(4):671–83.

117. de Baere T, Rehim MA, Teriitheau C, Deschamps F, Lapeyre M,Dromain C, et al. Usefulness of guiding needles for radiofre-quency ablative treatment of liver tumors. Cardiovasc Inter-vent Radiol 2006;29(4):650–4.


Documents

Seeding following percutaneous diagnostic and therapeutic approaches for hepatocellular carcinoma. What is the risk and the outcome?