Percutaneous radiofrequency ablation of lung tumours: results in the mid-term

Percutaneous radiofrequency ablation of lung tumours: resultsin the mid-term§

Marcello Carlo Ambrogi a,*, Marco Lucchi a, Paolo Dini a, Franca Melfi a,Gabriella Fontanini b, Pinuccia Faviana b, Olivia Fanucchi a, Alfredo Mussi a

aCardiac and Thoracic Department, Via Paradisa 2, 56124 Pisa, ItalybDepartment of Surgery, Via Paradisa 2, 56124 Pisa, Italy

Received 21 February 2006; received in revised form 23 March 2006; accepted 30 March 2006

www.elsevier.com/locate/ejctsEuropean Journal of Cardio-thoracic Surgery 30 (2006) 177—183

Abstract

Introduction: Radiofrequency ablation (RFA) of lung tumours has recently received much attention for the promising results achieved. Herein,we report the 5 years experience of a single referral centre in Europe, with particular reference to the mid-term results. Methods: BetweenOctober 2001 and June 2005, we performed 88 RFAs of lung tumours, 9 of which were followed by surgical resection. The remaining 79 RFAs, theobject of this paper, were performed to treat 64 lesions in 54 patients: 39males and 15 females with amean age of 71.7 years (range of 51—89). Allpatients had clinical or pathological evidence of the neoplastic lesion, which was non-small cell lung cancer (NSCLC) in 40 cases and a metastasisin 24 cases. The mean size of the lesions was 2.4 cm (range of 1—5). Ten lesions were re-treated from one to as many as four times. The procedurewas always performed under local anaesthesia and conscious sedation. A generator of RF with max power output of 200 W was utilised togetherwith a needle with nine deployable electrodes, to achieve a target temperature of 90 8C that was maintained for 15—27 min according to the sizeof the lesions. Results: In all cases, except two, the procedure was technically successful. Morbidity consisted in 10 cases (12.7%) of partialpneumothorax, 1 haematoma of the chest wall and 1 pleural effusion. At a mean follow-up of 23.7 months (range of 6—50) we recorded a 61.9% ofcomplete responses, with a higher rate in the metastatic lesions (70.8%) and in those smaller than 3 cm (69.7%). Mean (median) overall survivaland local progression-free interval were 17.3 (28.9) months and 12.9 (24.1) months, respectively. Conclusions: Efficacy of RFA in the mid-termseems to settle at a promising level, with better results for metastatic lesions and, above all, for lesions smaller than 3 cm. Notwithstanding theseencouraging results, RFA remains an alternative local therapy only when surgery cannot be performed, especially in NSCLC.# 2006 Elsevier B.V. All rights reserved.

Keywords: Lung cancer; Pulmonary metastasis; Radiofrequency ablation; Percutaneous thermal ablation; Minimally invasive treatment

1. Introduction

Surgical resection still represents the mainstay oftreatment, which achieves the major potentiality of curein the case of localised non-small cell lung cancer (NSCLC)(stages I, II, IIIa) [1]. Chemotherapy and radiotherapy, aloneor associated, bring a modest therapeutic contribution,prevalently of the palliative kind [2].

The lung is then the second most frequent site ofmetastatic disease, and several studies reported that, inselected cases, surgical resection of pulmonary metastasisrepresents a viable option of treatment [3].

Nevertheless, surgical treatment is not always feasible.Patients with NSCLC, in fact, are frequently poor surgical

§ Presented at the joint 19th Annual Meeting of the European Association forCardio-thoracic Surgery and the 13th Annual Meeting of the European Societyof Thoracic Surgeons, Barcelona, Spain, September 25—28, 2005.* Corresponding author. Tel.: +39 050 995211; fax: +39 050 577239.E-mail address: [email protected] (M.C. Ambrogi).

1010-7940/$ — see front matter # 2006 Elsevier B.V. All rights reserved.doi:10.1016/j.ejcts.2006.03.067

candidates because of coexistent chronic obstructivebroncho-pneumopathy or other associated diseases. It isestimated that more than 20% of patients with stage I or IINSCLC would not undergo surgical resection [4]. On the otherhand, in patients with pulmonary metastasis, the number andlocation of the lesions could require a sacrifice of pulmonaryparenchyma, which is out of proportion with the aim of thistherapeutic option, often palliative. And, again, it should beconsidered that the costs/benefits ratio of the surgicaltreatment, in patents with a stage IV neoplasm, is not alwaysin proportion with the results in terms of survival and qualityof life.

In this scenario, it is comprehensible that minimallyinvasive treatments often receive great interest, as ishappening for radiofrequency ablation (RFA). This methodhas been successfully used for the treatment of hepatocel-lular carcinoma [5], hepatic metastases [6], osteoid osteoma[7] and other solid tumours [8—10]. Thin metallic probes,similar to aspiration biopsy needles, are percutaneouslyinserted into the lesion using computed tomography (CT)

M.C. Ambrogi et al. / European Journal of Cardio-thoracic Surgery 30 (2006) 177—183178

scanning or ultrasound guidance. Radiofrequency energy isthen applied in order to achieve a temperature greater than60 8C (in most cases 90 8C). Thus, coagulative necrosis of thetumour is induced in a controlled manner.

More recently, this technique has been applied topulmonary tumours, too, with promising preliminary results[11—20]. All recent clinical trials reported, in fact, a goodlocal response, with an elevated tolerability and a very lowrate of complications. However, these trials have a shortfollow-up period and little is yet understood about theefficacy of RFA in the mid-to-long term.

The object of this paper concerns the experience of asingle referral centre regarding radiofrequency ablation oflung tumours, with particular reference to the results in theintermediate period.

2. Materials and methods

This study was conducted with the approval of the localethical committee for human research care.

The primary endpoint was local disease-free survival,while secondary endpoints included overall survival and sideeffects.

2.1. Radiofrequency generator and devices

We utilised a generator able to provide monopolarradiofrequency energy to perform coagulation and ablationof soft tissue (Model 1500, RITA Medical System, MountainView, CA, USA). This is an automatic apparatus with a maxpower output of 150 W, operating at 460 Hz. It has multipletemperature displays as well as impedance and powermonitoring. The energy was transferred into the tissue bymeans of a multitined expandable array (StarBrust XL, RITAMedical system). It consists of a 15-gauge needle cannulawith nine deployable electrodes which open flower-like up to5 cm (Fig. 1). Five electrodes are equipped with thermo-couples which allow continuous measurement of thetemperature inside the tissue. Two grounding pads wereapplied to each shaved leg to ground the current and to

Fig. 1. The electrode we utilised is constituted of a 15-gauge needle cannulawith nine electrodes that are deployable flower-like into the tumour up to5 cm.

reduce risks of skin heat injuries. Once the system waspowered up, the physician set the parameters of theablation: the mode of operation (we utilised ‘averagetemperature mode’, in which delivered power was auto-matically regulated to maintain the target temperature set);the target temperature; the ablation time at targettemperature; the maximum power delivery level, whichcan be modified at any time during the procedure.

2.2. Entry criteria

All cases were evaluated by a thoracic surgeon withadequate experience in thoracic oncology and interventionalradiology. Preoperative assessment of the patients was thatstandard to perform major pulmonary resection. It includesplain chest radiograph, CT scans of the chest and upperabdomen, lung function tests, cardiovascular fitness assess-ment. In addition, patients were required to meet thefollowing specific criteria for enrolment: contra-indication tosurgery or refusal; to be at least 18 years of age; preoperativehistological or clinical proof of lung malignancy; maximumdiameter of the tumour less than or equal to 5 cm. Criteriafor exclusion included: lesions located less than 1 cm frommajor blood vessels or airways; patients with a platelet countof less than 50 � 103/mL.

Main contra-indications to surgery were as follows:comorbidity, poor lung function (even for a limited resec-tion), previous repeated operations on the same lung,distribution and site of the lesions (which would haverequired an intolerable sacrifice of lung parenchyma).

Preoperative proof of lung malignancy was achievedsystematically by biopsy of the lesion in all the first 30patients of our series and then when there was a clinicaldoubt. Either new or increasing lung lesions at plain chestradiographs or CT scans, particularly in patients at risk forNSCLC or with a history of cancer, were considered clinicalproof of malignancy (in some cases supported by PET scanstoo).

Patients with metastatic disease were treated if theprimary cancer was cured or controlled, and after exclusionof other sites of disease (by PET scan in most cases).

All patients were required to sign a written informedconsent, after adequate explanation of risks and benefits ofthe procedure.

2.3. RFA procedure and follow-up

All the procedures were performed with the patientsunder conscious sedation (usually achieved with administra-tion of ketorolac 0.5—0.8 mg/kg, propofol 1—2 mg/(kg h)and remifentanil 0.1 mg/(kg/min)) and local anaesthesia(subcutaneous 1% xylocaine). Vital signs of the patient werenon-invasively monitored continuously. We always utilised CTguidance, enhanced by contrast media before and after theprocedure, in order to obtain more information abouteffectiveness. In all cases the target temperature was90 8C. It was maintained for a time which ranged from 15 to27 min according to the size of the tumour, which alsodetermined the deployment of the electrodes. These lastwere deployed gradually, starting from 2 cm and then 1 cmfor each step. When technically possible, we always pursued

M.C. Ambrogi et al. / European Journal of Cardio-thoracic Surgery 30 (2006) 177—183 179

Fig. 2. We always pursued the objective to completely encompass the tumourwith the thermal ablation area, as it is shown in the right side of this figure.

the objective of encompassing the tumour with an ablationzone at least 1 cm larger (Fig. 2). After the procedure, allpatients were transferred to the recovery room for observa-tion. Twenty-four hours later, after a chest radiograph toexclude complications (such as a pneumothorax), the patientswere discharged. Radiological follow-up included contrast-enhanced CT at 1, 3 and 6 months and then at 6 monthsintervals. CT scans were evaluated by two physicians, thesame thoracic surgeon who performed the procedure anda radiologist. First evaluation was done just after RFA with apost-procedural CT enhanced with contrast material. Aprocedure was considered technically successful when itwas performed according to the protocol, and the ablationzone completely covered the tumourwith lower enhancementthan before RFA. Effectiveness was calculated starting fromthe 1st month follow-up in terms of a local control of thedisease. Due to a frequent increase of the lesion immediatelyafter RFA, resulting from the sum of the tumour and theablation zone, at the 1st and 3rd month follow-up the maincriteria for complete response was reduction of the contrastmaterial enhancement, in respect to thatmeasuredbefore theprocedural. However, a value of the lesion’s density greaterthan 25 Hounsfield Unit (HU) was considered indicative ofpersistenceor relapse of thedisease.Thedimensional criteria,instead, became the main criteria from the 6th month on, inaccordance with the RECIST criteria [21]. An increase of thesize of the lesion in respect to its size pre-RFAwas considered apersistenceor relapse of the disease, and thereforea failure ofthe procedure. However, both criteria, the size and theenhancement, have been considered and integrated at alltimes, case by case, to determine whether progression hadoccurred. In five cases a PETscan was utilised to better assessrecurrence, or lack of recurrence, of the lesion.

In addition, the first 20 patients underwent pulmonaryfunction tests before RFA and then at 1 and 3 months afterablation.

2.4. Statistical analysis

Descriptive statistics for the series were generated withSTATISTICA software (StatSoft Inc., Tulsa, OK, USA; version6.0 for Windows). Comparisons of percentages and mean

values were performed by using Student’s t-test. Kaplan—Meier with log-rank analysis was used to calculate overallsurvival and local progression-free interval and to comparepatients. A value of p less than 0.05 was considered toindicate a statistically significant difference.

3. Results

BetweenOctober2001andJune2005weperformed88RFAsof lungmalignancies. In nine cases the patients underwent RFAfollowed by surgical resection, in accordance with two earlyresearch protocols (paper in press) with the object to definebiological effects of RFA on the lung and its tumours. Theremaining 79RFAs, theobject of this paper,were performedon54 patients. Ten patients were treated for two separatelesions, synchronous in eight cases (NSCLC in four cases andmetastases in six cases). In 10 patients, RFAwas repeated fromone to as many as four times to treat the same lesion. Thepatients were 39males and 15 femaleswith amean age of 71.7years (range of 51—89). Pathological and/or clinical diagnosisclassified 40 lesions as NSCLC (eight of these lesions weremetastasis from NSCLC, according to the criteria by Martiniand Melamed [22]) and another 24 as metastatic lesions fromextra-thoracic malignancies, mainly from colorectal cancer.Histology or cytology of the lesions was available in 35 cases; itassessed a primary lung cancer in 24 cases and a pulmonarymetastasis in 11 patients. Mean size of the lesions was 2.4 cmmeasured on CT scans (range of 1—5).

Time of RFA at target temperature was 19 min on average(range of 15—27). The mean time to reach the targettemperature was 4 min (range of 2—8), which includes thetime necessary to reach the target temperature again afterdeployment of the electrodes.

In all cases, except two, the procedure was technicallysuccessful. In one patient, probably due to the density of thetumour, we did not succeed in inserting the needle into thelesion, which was pushed against the oesophagus and inferiorcava vein. In this case we renounced the RFA, and this patientwas excluded from further evaluation in the study. In anotherpatient the procedure was interrupted 5 min before theprogrammed time due to a high impedance level.

All the procedures were tolerated well. Five patients,with lesion near or in contact with the pleural surface,experienced pain related to RFA which required thedeepening of sedation and that, however, cleared up in afew minutes after the end of the procedure.

We had nomortality, whilst morbidity affected 12 patients(15.2% of the procedures). Complications consisted of 10cases of partial pneumothorax (12.7%), 6 of which requiredpleural drainage, 1 pleural effusion, which was resolved withmedical treatment, and 1 chest wall haematoma.

Mean post-procedural hospital stay was 1.3 days (range of1—4).

3.1. Response rates and survival analysis

At a mean follow-up period of 23.7 months (median 24,range of 6—50), the overall radiological complete responserate was 61.9% (39/63 lesions). It seemed to be better forlesions smaller than 3 cm (69.7% vs 50%) and for metastases

M.C. Ambrogi et al. / European Journal of Cardio-thoracic Surgery 30 (2006) 177—183180

Fig. 3. Overall survival analysis with Kaplan—Meier method.Fig. 5. Analysis of survival by Kaplan—Meier method comparing lesions smallerthan 3 cm versus those equal to or greater.

Fig. 6. Analysis of local progression-free interval by Kaplan—Meier methodcomparing lesions smaller than 3 cm versus those equal to or greater.

from extra-thoracic malignancies (70.8% vs 56.4%), even ifneither reached statistical significance.

Thirty-one of 54 (60%) patients were alive, and amongthese 24 (77%) were locally disease-free (LDF); of thepatients who were deceased 15 (65%) were LDF. Distantrecurrence (mediastinal, pulmonary or other sites) occurredin 21 (39%) patients (5 of them were LDF). Most of thesepatients underwent some adjuvant therapy (CT and/or RT).Mean overall survival and local progression-free interval (LPI)were 17.3 and 12.9 months, respectively. They were relatedto the size of the tumour, both showing a difference betweenlesions smaller than 3 cm and those equal or greater than3 cm. Survival was 19.7 months for lesions smaller than 3 cmand 12.1 months for those greater ( p = .02), while LPI was15.8 and 6.6 months, respectively ( p = .002).

The Kaplan—Meier curves represent the overall actuarialsurvival in Fig. 3, whilst they compare survival betweenNSCLC and metastasis in Fig. 4, and between lesions greateror smaller than 3 cm in Fig. 5. Median overall survival is 28.9months; it decreases to 18.9 for NSCLC, and is not reached formetastases (without statistical significance), whilst itincreases to 30.5 months in cases of tumours smaller than3 cm versus 14.9 months for lesions equal to or larger than

Fig. 4. Analysis of survival by Kaplan—Meier method comparing primary versussecondary lung tumours.

3 cm (still without statistical significance, p = .08). Medianoverall LPI is 24.1 months, without statistical significancebetween NSCLC and metastasis, whilst it is statisticallysignificant when we compare tumours equal to or larger than3 cm versus those smaller ( p = .04) (Fig. 6).

3.2. Lung function tests

Forced volume capacity (FVC) and forced expiratoryvolume in the first second (FEV1) before RFA were 2.91 and1.97 L, respectively, on average (range of 1.82—4.38 and0.9—3). At 1 month from RFA they resulted minimallydecreased, then, at 3 months follow-up, they almostreturned to a value equal to that which was pre-RFA (Table 1).

Table 1Results of the lung function tests before and after RFA

Before RFA At 1 month At 3 months

FVC (range) (L) 2.91 (1.82—4.38) 2.63 (2—4.09) 2.8 (1.8—4.4)FEV1 (range) (L) 1.97 (0.9—3) 1.71 (0.81—3.19) 1.86 (0.72—3.56)

M.C. Ambrogi et al. / European Journal of Cardio-thoracic Surgery 30 (2006) 177—183 181

4. Discussion

Currently, many papers report promising clinical results inperforming percutaneous thermal ablation of primary and/orsecondary lung tumours [11—20]. But, notwithstanding thatin recent years a great amount of data on the clinical andradiological response of RFA in lung tumours has accumu-lated, its efficacy in the mid-to-long term remains poorlyunderstood.

The first percutaneous RFA of a lung tumour is reported byDupuy et al. [11] in 2000. He treated three patients, in whomthe main purpose of utilizing RFAwas palliation, with resultswhich were technically successful and uneventful. Sincethat paper, a number of manuscripts have described thefeasibility and safety of the procedure. In particular, aninternational study survey, reported by Steinke et al. [15] in2004, stressed, with almost 500 RFA of lung tumourscollected, how percutaneous pulmonary RFA seems to bea minimally invasive tool for local control with negligiblemortality, low morbidity, short hospital stay and gain inquality of life.

Unlike feasibility and safety, efficacy seems not to be soeasily assessable. Contrary to surgical resection, wherecancer is removed, in fact, treatment response to RFA isvaluable only with radiological follow-up. After RFA, anablation zone is generally visible at CT scan as a ground glassopacity encompassing the target tumour. In most cases, theablation zone involves a larger area than the original tumour,especially just after RFA and for the 1st months. As aconsequence, in this period, the dimensional criteria are notvery useful. This is the reason that some authors utilise CTdensitometry protocols to help evaluate for persistent orrecurrent disease [13]. Other centres, instead, have beenusing a modification of the RECIST criteria to assess tumourprogression at follow-up [11]. In our experience we utilisedboth criteria, privileging enhancement after contrastmaterial injection during the first 3 months after RFA, and,afterwards, the dimensional criteria in accordance with theRECIST. However, both radiological features of the lesions,the size and the density, were contemporarily evaluatedevery time and in all patients, together with other features,in order to assess treatment response. Moreover, whennecessary, a PET scan was performed to clear doubts, and todecide either when to proceed to further treatment (oftenwe repeated RFA) or to wait. Perhaps, a more useful tool inthe follow-up of these patients could be PET-CT, as recentlyreported by Griffo et al., in an abstract at the 3rd EACTS/ESTS Joint Meeting, held in Leipzig on 13th September 2004[23].

Table 2A comparison of results in the literature with particular reference to length of follo

Author No. of lesions Tumour type Lesion siz

Suh et al. [13] 19 NSCLC/Mts 0.5—7.4Herrera et al. [12] 33 NSCLC/Mts 2—16Yasui et al. [18] 99 NSCLC/Mts 0.3—8Akebosi et al. [16] 54 NSCLC/Mts 0.7—6Steinke et al. [17] 40 Mts 0.3—4.2Lee et al. [14] 32 NSCLC/Mts 0.5—12Fernando et al. [19] 21 NSCLC 1.2—4.5Our experience 63 NSCLC/Mts 1—5

Another issue, on which attention is currently focused, isthe response to RFA treatment of lung tumours in the mid-to-long term. Most of the papers, in fact, report preliminaryresults in the short period. In Table 2 are reported thecomplete response rates and the follow-up period of themostsignificant studies reported in the literature. It seems thatthat the longer the follow-up is, the worse is the completeresponse rate. So it will be interesting to see in the mid-to-long period on what overall rate of local pulmonary tumoureradication the RFA treatment will settle. Obviously this isnot the only parameter which affects efficacy of RFA. Manyothers will be taken into consideration, with particularattention to the RFA protocol and the technology utilised.Herrera et al. [12] treated amixed cohort of 18 patients (5 viathoracotomy and 13 percutaneously) with primary andmetastatic lung tumours. They utilised a needle electrodewith multiple tines deployable into the tumour for 2 or 4 cm.At a mean follow-up period of 6 months, they reported a 55%complete or partial response rate and a 17% stable diseaserate in lesions with a mean diameter of 5.3 cm. As happenedin our experience, the response rates seemed to be better forsmaller lesions (66% for lesions smaller than 5 cm vs 33% forlesions larger than 5 cm). Another important issue whicharose in this paper was the contraindication to treat centrallesions by RFA, due to the risk of fatal complications, ashappened in one case in their experience. For this reason, ourprotocol, as well as that of other authors [13,17], providesthat lesions must be at least 1 cm away from bigger airwaysand great vessels (i.e., those ilar or para-ilar). More recently,Fernando et al. [19] from the same institute reported anupdate of their experience selecting patients with primaryNSCLC. Despite a longer follow-up period (14 months), theresults were better than those of their preliminaryexperience (63% of complete ablation vs 55%). It seems thatprimary NSCLC is more responsive to RFA than is metastasis,which is in contrast with our findings. In our experience, infact, even if there is no statistical significance, metastaseshave a better complete response rate than primary lungtumours. An explanation of this finding could be found in thedifferent biological behaviour of NSCLC and lung metastases,with the seconds generally well delimitated from thesurrounding lung parenchyma, on the contrary of primarylung cancer that is often speculate and microscopicallyinvading the lymphatic vessels around the macroscopiclesion. Moreover, in our experience, the better survival incase of metastases from extra-thoracic malignancies may bethe result of a very accurate selection of the patients thatwere required to have the primary cancer cured and no othersite of the disease.

w-up and rate of complete ablations

e (cm) Complete response rate (%) Mean follow-up (months)

75 4.555 691 759 9.265 1238 12.563 1463.5 23.7

M.C. Ambrogi et al. / European Journal of Cardio-thoracic Surgery 30 (2006) 177—183182

Like Herrera et al. and Fernando et al., Akebosi et al. [16]also achieved, in a series of 54 primary and secondary lungtumours RFA, an overall complete necrosis rate of 59%, andreported a better response in smaller lesions. Theydistinguished lesions smaller than or equal to 3 cm fromlesions greater than 3 cm, and found a statistically significantdifference in the complete response rate between the twogroups (69% vs 39%, p < .05), which is similar to that of ourseries (69.7% vs 50%, p = NS). These findings have also beenconfirmed by Lee et al. [14]. With an overall completenecrosis rate of 38% at a mean follow-up of 12.5 months, theyreported that tumour size was highly significant forpredicting the efficacy of RFA. Lesions smaller than 3 cmhad a 100% complete necrosis rate, those measuring 3—5 cmwere completely ablated in 38% of cases ( p < .05) and lesionsgreater than 5 cm got only a 8% complete necrosis rate. Theyutilised both single electrode and multiple deployableelectrode needles, according to the size of the lesion, inmost of the cases with several insertions into the lesion. Inour series, also survival was better in patients with smallerlesions. Even if there is a natural more favourable outcome ofpatients with small peripheral tumours, the increased localcontrol of these lesions may have contributed too. Anotherpaper of particular interest is that by Steinke et al. [17], whoreported their experience on thermal ablation of lungmetastasis from colorectal cancer. They followed a RFAprotocol similar to that of our experience, utilizing the samedevices and obtaining similar results. For lesions whose sizeranged from 0.3 to 4.2 cm, CT scan follow-up at 12 monthsshowed disappearance, reduction or stability in 65% of cases,which is comparable with our overall 61.9% (70.8% in cases ofmetastasis from extra-thoracic malignancies, which weremainly from colorectal cancer).

Currently, to our knowledge, there is no other study thatreports the effect of RFA on pulmonary function. In ourexperience, spirometry performed before and after RFA atseveral intervals shows a slight reduction of FVC and FEV1 inthe 1st month (however not statistically significant), which ispromptly recovered at the 3rd month.

In conclusion, if the feasibility and safety of RFA in lungtumours have been assessed, its efficacy still remains to bedetermined, above all in the mid-to-long period. So, well-designed clinical trials with a long-term follow-up arerequired to confirm the short-term of the literature and ourmid-term results, before RFA can enter into clinicalpractice. From the first experiences, as it was expected,it seems to be more effective for lesions smaller than 3 cmand, in our experience, for metastases. Certainly, it isnecessary to underline the purpose of RFA, which regardsonly the local treatment of the tumour, with all its limits. Infact, if compared with surgical resection, despite asuccessful RFA, a higher recurrence rate is expected,locally and in the mediastinum. And, just as wedgeresection is considered a compromise operation in respectto lobectomy for NSCLC, RFA may be considered acompromise procedure in respect to surgical resectionfor pulmonary tumours (primary or metastatic) in high-riskpatients. In future, it would be of interest to compare RFAwith other emerging alternative therapies (i.e., stereo-tactic radiosurgery), respect to quality of life, tumourcontrol and survival.

References

[1] Van Schil PE. Surgery for non-small cell lung cancer. Lung Cancer2001;34(Suppl. 2):S127—32.

[2] Zimmermann FB, Bamberg M, Molls M, Jeremic B. Radiation therapy alonein early stage non-small cell lung cancer [review]. Semin Surg Oncol2003;21:91—7.

[3] Pastorino U, Buyse M, Friedel G, Ginsberg RJ, Girard P, Goldstraw P,Jonhston M, McCormack P, Pass H, Putnam JB. Long-term results of lungmetastasectomy: prognostic analyses based on 5206 cases. J ThoracCardiovasc Surg 1997;113:37—49.

[4] Bach PB, Cramer LD, Warren JL, Begg CB. Racial differences in thetreatment of early stage non-small cell lung cancer. New Engl J Med1999;341:1198—2205.

[5] Furley SA, Izzo F, Ellis LM, Vauthey JN, Vallone P. Radiofrequency ablationof hepatocellular cancer in 110 patients with cirrhosis. Ann Surg2000;232:381—91.

[6] Lencioni R, Goletti O, Armillotta N, Paolicchi A, Moretti M, Cioni D, DonatiF, Cicorelli A, Ricci S, Carrai M, Conte PF, Cavina E, Bartolozzi C. Radio-frequency thermal ablation of liver metastases with a cooled-tip elec-trode needle: results of a pilot clinical trial. Eur Radiol 1998;8:1205—11.

[7] Rosenthal DI, Springfield DS, Gebhardt MC, Rosenberg AE, Mankin HJ.Osteoid osteoma: percutaneous radio-frequency ablation. Radiology1995;197:451—4.

[8] Wood BJ. Feasibility of thermal ablation of lytic vertebral metastaseswith radiofrequency current. Cancer J 2002;8:26—9.

[9] Gervais DA, McGovern FJ, Wood BJ, Goldberg SN, McDouglas WS, MuellerPR. Radiofrequency ablation of renal cell carcinoma: early clinicalexperience. Radiology 2000;217:665—72.

[10] Anzai Y, Lufkin R, DeSalles A, Hamilton DR, Farahani K, Black KL. Pre-liminary experience with a MR-guided thermal ablation of brain tumor.Am J Neuroradiol 1995;16:39—48.

[11] Dupuy DE, Zagoria RJ, Akerley W, Mayo-SmithWW, Kavanagh PV, Safran H.Percutaneus radiofrequency ablation of malignancies in the lung. Am JRoentgen 2000;174:57—9.

[12] HerreraLJ, FernandoHC, PerryY,GoodingWE,BuenaventuraP,ChristieNA,Luketich JD. Radiofrequency ablation of pulmonary malignant tumors innon surgical candidates. J Thorac Cardiovasc Surg 2003;125(4):929—37.

[13] Suh RD, Wallace AB, Sheehan RE, Heinze SB, Goldin JG. Unresectablepulmonary malignancies: CT-guided percutaneous radiofrequency abla-tion—preliminary results. Radiology 2003;229:821—9.

[14] Lee JM, Jin GY, Goldberg SN, Lee YC, Chung GH, Han YM, Lee SY, Kim CS.Percutaneous radiofrequency ablation for inoperable non-small cell lungcancer and metastasis: preliminary results. Radiology 2004;230:125—34.

[15] Steinke K, Sewell PE, Dupuy D, Lencioni R, Helmberger T, Kee ST, JacobAL, Glenn DW, King J, Morris DL. Pulmonary radiofrequency ablation—aninternational study survey. Anticancer Res 2004;24:339—43.

[16] Akeboshi M, Yamakado K, Nakatsuka A, Hataji O, Taguchi O, Takao M,Takeda K. Percutaneous radiofrequency ablation of lung neoplasms:initial therapeutic response. J Vasc Interv Radiol 2004;15:463—70.

[17] Steinke K, Glenn D, King J, Clark W, Zhao J, Clingan P, Morris DL.Percutaneous imaging-guided radiofrequency ablation in patients withcolorectal pulmonary metastases: 1-year follow-up. Ann Surg Oncol2004;11:207—12.

[18] Yasui K, Kanazawa S, Sano Y, Fujiwara T, Kagawa S, Mimura H, Dendo S,Mukai T, Fujiwara H, Iguchi T, Hyodo T, Shimizu N, Tanaka N, Hiraki Y.Thoracic tumors treated with CT-guided radiofrequency ablation: initialexperience. Radiology 2004;231:850—7.

[19] Fernando HC, De Hoyos A, Landreneau RJ, Gilbert S, Gooding WE,Buenaventura PO, Christie NA, Belani C, Luketich JD. Radiofrequencyablation for the treatment of non-small cell lung cancer in marginalsurgical candidates. J Thorac Cardiovasc Surg 2005;129:639—44.

[20] Ambrogi MC, Lencioni R, Fontanini G, Lucchi M, Fanucchi O, Melfi F, Dini P,Mussi A. Percutaneous radiofrequency ablation of primary NSCLC. LungCancer 2005;49(Suppl. 2):S3.

[21] Padhani AR, Ollivier L. The RECIST (Response Evaluation Criteria in SolidTumors) criteria: implication for diagnostic radiologists. Br J Radiol2001;74:983—6.

[22] Martini N, Melamed MR. Multiple primary lung cancer. J Thorac Cardio-vasc Surg 1975;70:606—12.

[23] Ferrante G, Griffo S, Romano M, Salvatore M, Servillo G, Bianco R, GentileM, Mainolfi C. CT-guided radiofrequency treatment of lung metastases:efficacy and follow-up with CT-PET. Abstract at the 3rd EACTS/ESTS JointMeeting, Leipzig, 13 September 2004.

M.C. Ambrogi et al. / European Journal of Cardio-thoracic Surgery 30 (2006) 177—183 183

Appendix A. Conference discussion

Dr P. Macchiarini (Barcelona, Spain): You said that the indication aresurgical contraindication, but you still have nine patients that for some reasonwere operated thereafter, so could you explain on that.

Dr Lucchi: Those nine cases were part of the first protocol. We explained tothe patients what we were going to do and they accepted. Those were notsurgical contraindication, but we did it to verify that there was a sense to theradiofrequency ablation. Otherwise we stuck with our study.

Dr G. Friedel (Gerlingen, Germany): And in those patients you did, in all,did you find a complete response histologically?

Dr Lucchi: No, only six of nine, that means about 60%.

Dr R. Rami-Porta (Barcelona, Spain): My question is technical. Theduration of treatment ranged from 15 to 27 min. So how did you determine theduration of the heating?

Dr Lucchi: The length of the treatment is determined on the size of thetumour, on the diameter, and on the temperature that is monitored by theinstrument. The range 15—27 min is the time at the target temperature of 90 8C.To this time you have to add that necessary to reach the target temperature.

Dr D. Branscheid (Grosshansdorf, Germany): I’d like to read something outof your abstract. The last sentence, ‘must be considered an alternative localtherapy only when surgery cannot be performed.’

I missed that here in the conclusions. And I think it’s highly experimental.We are just finding out that complete resection, surgical resection ofmetastasis, gives a better survival. We have fought to use laser resection, andwe do not say that laser is better than conservative.

Might it probably be that you meant with your conclusion that could be anadditive to our strategy in metastasis, perhaps in the future, when we do notcome along with laser and with conventional resections?

Dr Lucchi: I agree with your consideration. We are all thoracic surgeons, sosurgery remains the first choice in most of the cases; also for metastasis, whenoncologist propose patients to us.

In any case, there are some patients who also for metastasis are notanymore indicated to surgical operation. And in reason of a so low morbidity,and maybe, as I observed, good results, also radiofrequency ablation may havea role. Moreover, in some cases, I think that hyperthermia, induced byradiofrequency ablation, may enhance the power of radiotherapy and also ofchemotherapy.

Dr S. Mattioli (Bologna, Italy): I am interested in the last sentence of yourconclusions. Would you apply first radiofrequency and secondly radiotherapy,stereotactic radiotherapy, or vice versa?

Dr Lucchi: What we know is that hyperthermia is the best radio-sensitizing.So there could be a rationale to do radiofrequency ablation and thenradiotherapy.