Local progression after radiofrequency ablation for pulmonary metastases


  • Erik M. von Meyenfeldt MD,

    1. Department of Surgery, Netherlands Cancer Institute-Antoni van Leeuwenhoek Hospital, Amsterdam, the Netherlands
    Search for more papers by this author
  • Warner Prevoo MD,

    1. Department of Radiology, Netherlands Cancer Institute-Antoni van Leeuwenhoek Hospital, Amsterdam, the Netherlands
    Search for more papers by this author
  • David Peyrot MD,

    1. Department of Radiology, Netherlands Cancer Institute-Antoni van Leeuwenhoek Hospital, Amsterdam, the Netherlands
    Search for more papers by this author
  • Nathalie Lai A Fat MD,

    1. Department of Surgery, Netherlands Cancer Institute-Antoni van Leeuwenhoek Hospital, Amsterdam, the Netherlands
    Search for more papers by this author
  • Sjaak J. A. Burgers MD, PhD,

    1. Department of Thoracic Oncology, Netherlands Cancer Institute-Antoni van Leeuwenhoek Hospital, Amsterdam, the Netherlands
    Search for more papers by this author
  • Michel W. Wouters MD,

    1. Department of Surgery, Netherlands Cancer Institute-Antoni van Leeuwenhoek Hospital, Amsterdam, the Netherlands
    Search for more papers by this author
  • Houke M. Klomp MD, PhD

    Corresponding author
    1. Department of Surgery, Netherlands Cancer Institute-Antoni van Leeuwenhoek Hospital, Amsterdam, the Netherlands
    • NKI-AVL, Department of Surgical Oncology, PO Box 90203, 1006 BE Amsterdam, the Netherlands
    Search for more papers by this author
    • Fax: (011) 31 20 5122554



Local treatment for pulmonary metastases is considered to be a reasonable treatment option in patients with oligometastatic disease. Percutaneous radio frequency ablation (RFA) has been reported as an alternative to surgery. Results of RFA for local control of pulmonary metastases were evaluated.


All consecutive patients treated with RFA for pulmonary metastases (2004-2009) were included. RFA was performed percutaneously under computed tomographic guidance. Follow-up was scheduled at 1, 3, and 6 months after treatment and every 6 months thereafter. Major outcome parameters were local and any-site progression, complications, and survival.


Ninety pulmonary metastases were treated, in 46 patients at 65 sessions. Many patients had recurrent metastases after previous surgery (n = 36 of 46). Pneumothorax occurred in 34% (chest drain in 25%) and major complications in 6%. After median follow-up of 22 months (range, 2-65 months), 25 local progressions occurred after RFA; the 2-year local progression rate per lesion was 35%. Overall survival at 3 years was 69%.


Notwithstanding its relatively low morbidity, follow-up after RFA for pulmonary metastases shows a considerable rate of local progression. The role of local ablation techniques for long-term disease control in oligometastatic disease is discussed. Cancer 2011. © 2011 American Cancer Society.

Local (surgical) treatment is an accepted treatment option for oligometastatic pulmonary disease of various origins. This is largely based on positive results of metastasectomy in retrospective studies and registry data, but no randomized trials are available.1-3 The decision to perform pulmonary metastasectomy is usually based on such criteria as: control of the primary tumor, oligometastatic disease, no (unresectable) extrapulmonary lesions, long disease-free interval, and/or lack of treatment alternatives.4 However, surgical treatment requires sufficient cardiopulmonary reserve and is associated with a risk of morbidity and relapse.

Percutaneous radiofrequency ablation (RFA) is a minimally invasive image-guided local treatment modality for cancer. The radiofrequency energy is applied through a needle electrode introduced into the target tissue, which induces heat through cell agitation, resulting in denaturation and cell death. This technique has been used for both patients with primary lung cancer and for patients with pulmonary metastatic disease. RFA may be an alternative to surgical resection of metastatic pulmonary lesions with less morbidity and less negative side effects. Several reports have suggested treatment efficacy comparable to open resection in a patient category deemed unfit for surgery.5-11 Procedure-related mortality has generally been reported to be low (0-2.6%).5-11 The main adverse event is pneumothorax requiring chest tube drainage in 10% to 38% of the procedures. Local progression rates after RFA range from 7% to 36% after follow-up of 1 to 2 years.5-11

In the Netherlands Cancer Institute, RFA treatment was introduced for treatment of pulmonary metastases in 2004. The aim of this study was to evaluate the contribution of this local ablation therapy for disease control in patients with pulmonary (oligo)metastatic disease.


Patient Selection

Patients were treated at the Netherlands Cancer Institute-Antoni van Leeuwenhoek Hospital, a reference oncology center in the Netherlands, from 2004 to 2009. All consecutive patients were included. Pretreatment workup for pulmonary metastases included physical status assessment, chest radiography, bronchoscopy, diagnostic computed tomography (CT) scan (Somatom Sensation Open, Siemens, Erlangen, Germany) fluorodeoxyglucose-positron emission tomography (PET) scan in PET-avid tumors (Gemini TF 16-slice CT, Philips, Amsterdam, the Netherlands), cytology (if relevant), and appropriate laboratory and pulmonary function tests. Management of patients was discussed in the multidisciplinary thoracic oncology team with representatives from thoracic and medical oncology, surgical oncology, pathology, radiology, radiotherapy, and nuclear medicine. Patient selection for RFA was based on clinical information and the IRLM System of Prognostic Grouping.1, 4 Surgical resection was the preferred treatment for patients with good prognosis, that is, solitary pulmonary metastasis and a disease-free interval of >36 months.1 Patients with pulmonary metastases were eligible for RFA if they had limited (recurrent) pulmonary metastatic disease with peripheral location(s). In some cases, only progressive lesions were treated (other lesions including scarring and rest lesions after previous treatment being stable). Exclusion criteria for RFA treatment were: >5 lesions, recent pneumonia, inadequate medical condition, and irreversible coagulation disorders, including vital indication for anticoagulant medication.

If RFA was a treatment option, the attending specialist and the radiologist performing the RFA explained the procedure and potential complications, and eligible patients gave informed consent. RFA was performed within 4 weeks after the diagnostic CT imaging. The protocol review board of the Netherlands Cancer Institute-Antoni van Leeuwenhoek hospital approved the treatment protocol.


Two experienced interventional radiologists performed the RFAs. Patients received intravenous cefuroxime as antibiotic prophylaxis for 24 hours. Most patients had epidural anesthesia with conscious sedation to relieve (post)procedural pain. Three patients had general anesthesia. A beanbag allowed comfortable positioning of the patient on the scanning table with optimal location and approach to the lesion(s). In patients with multiple metastatic lesions, a maximum of 3 lesions were treated during 1 RFA session. An anesthesiologist and a nurse were present for cardiovascular, respiratory, and pain monitoring, and a radiology technician assisted with the CT scan. In the period 2004 to 2006, LeVeen Needle Electrodes (Boston Scientific Corporation, Natick, Mass) were used. Once electrode position was achieved, the electrode was deployed and attached to a Radio Therapeutics RF generator (RF 3000, Radio Therapeutics Corporation, Sunnyvale, Calif). After 2006, The Cool-tip RF Tissue Ablation System was used as well (Covidien, Boulder, Colo). Before ablation, 4 rectangular grounding electrode pads were placed on the patient's thighs, to lessen the risk of skin burn. First, a planning CT scan was obtained (3-5 mm sections, 1 mm collimation, 1.0 pitch) through the whole thorax. Next a dedicated planning scan was performed through the lesion. Needle size was chosen for a projected ablation zone of at least 1 cm larger than the diameter of the tumor, except for tumors >4 cm, as the largest needle used was 5 cm. The path of the needle tract was evaluated in advance, to avoid large vessels, bronchi, blebs, and fissures. After disinfecting the skin with iodine alcohol, the RFA needle was introduced under CT fluoroscopy. Ablation was done according to the manufacturer's algorithms. Ablation was ended after CT control when a rim around the lesion was seen. Ablation time varied between 12 and 25 minutes. It took longer to ablate larger lesions. Depending on the size of the tumor, ablation was performed in 1 to 6 positions. In all cases, CT imaging was used to change position of the needle and to monitor the effects of the ablation. Treatment was considered successful when roll-off was reached and the tumor was surrounded by a sufficient margin of ground glass opacity.12 A post-treatment monitoring CT scan was performed to evaluate the treatment result and check for potential complications (eg, pneumothorax). A chest drain was inserted in case of a large (>3 cm), progressive, or symptomatic pneumothorax. In most patients, epidural anesthesia was discontinued on the first postprocedural day. A chest x-ray was performed before discharge from the hospital.


Follow-up visits were planned at 1, 3, and 6 months and every 6 months thereafter. Radiological follow-up of the treatment area was done with CT at the same intervals.

Data Collection and Analysis

The following data were collected and entered in a concurrent database: age, sex, performance status (World Health Organization/Eastern Cooperative Oncology Group), weight loss, pulmonary function, comorbidity, histology, tumor characteristics, treatment-related morbidity and mortality, local progression of disease, and survival. Treatment-related factors were size of the lesion, contact with vessels that were ≥3 mm in diameter, and tumor location.12, 13 Tumor size was measured as the maximum size on transverse, sagittal, or coronal planes in lung window setting. Location was split up into groups, with parenchymal, pleural, or central localization.

CT scan evaluation was done by a radiologist and a chest physician. Signs of progression were growth beyond postablation densities/scarring and (increased) enhancement of lesions. Any new or growing lesion in the ablation zone or within 1 cm from this zone was considered to be local progression. Date of recurrence was chosen as the first CT image showing recurrence.

As expected in this patient population, patients were not only at risk for local progression but also for recurrent disease at other sites. Progression was defined as either appearance of new lesions at any site or growth in previously stable lesions. Patients with widespread recurrent disease, in whom chest CT follow-up was discontinued, were censored in the analysis of local progression.

For statistical analysis, SPSS version 17.0 software (SPSS Inc., Chicago, Ill) was used. The Kaplan-Meier method was used to estimate survival functions for progression or survival rates. Recurrence rates per session were established as well. For recurrence reporting, both absolute data and Kaplan-Meier estimates of recurrence percentages were given. Means were reported with 95% confidence intervals, medians with 95% confidence intervals, or range. The log-rank test was performed to compare survival data. Two-sided chi-squared and Fisher exact test were used to evaluate prognostic factors. A P value of <.05 was considered significant.


From 2004 to 2009, 90 pulmonary metastases were treated with RFA in 46 consecutive patients. The 90 lesions were treated in 65 sessions. Median age was 57 years (range, 32-78 years). Many patients were treated for recurrent pulmonary metastases (n = 36 of 46, 78%). Tissue diagnosis was available from biopsy (68%) or other metastatic site histology. Table 1 shows the patient characteristics and data on previous treatment.

Table 1. Patient Characteristics (n = 46)
CharacteristicNo. (%)
  1. RFA indicates radiofrequency ablation; WHO, World Health Organization; DFI, disease-free interval.

  2. RFA 2004-2009; 90 lesions in 65 procedures in 46 patients.

SexMen, 19/46 (41) Women, 27/46 (59)
Age at time of RFA, median y [range]57 [32-78]
WHO performance class at time of RFA, %
 032/46 (70)
 114/46 (30)
First presentation with metastasis10/46 (22)
Primary tumor histology
 Sarcoma12/46 (26)
 Colorectal cancer14/46 (30)
 Renal cell carcinoma4/46 (9)
 Breast cancer3/46 (7)
 Melanoma3/46 (7)
 Miscellaneous10/46 (21)
Previous interventions for metastasis before study period
 010/46 (22)
 120/46 (44)
 210/46 (22)
 ≥36/46 (12)
Number previous thoracic interventions for metastasis before study period
  024/46 (52)
  114/46 (30)
  ≥28/46 (18)
DFI at time of RFA, median mo [range]9 [0-84]
Previous systemic therapy21/46 (46)

Tumor characteristics are summarized in Table 2. Electrode position was successfully obtained, and roll-off was achieved according to the algorithms in all 90 tumors. Ablation was followed by ground glass opacity around the ablated tumor (Fig. 1). Patients were only considered to be tumor free after the RFA session in the absence of other untreated (mostly contralateral) lesions.

Figure 1.

(A) A preablation computed tomographic image shows a rectal cancer metastasis in the left upper lobe; tines of the needle are visible in the tumor (prone position). (B) Changes are seen 3 days after radiofrequency ablation, with a wide margin of ground glass opacity surrounding the tumor (supine position). (C) The image shows the same lesion 6 months postablation. There is no contrast enhancement (supine position). (D) The image shows the same lesion 18 months postablation; a shrinking scar in the lung parenchyma is seen without contrast enhancement (supine position).

Table 2. Metastases Characteristics (n = 90)
CharacteristicNo. (%)
  1. Radiofrequency ablation 2004-2009; 90 metastases in 65 procedures in 46 patients.

Size of lesion, median mm [range]16 [5-80]
 <2 cm62 (69)
 2-3 cm14 (16)
 3-5 cm8 (9)
 >5 cm6 (6)
Location of lesion
 Central1 (1)
 Parenchymal50 (56)
 Pleural36 (40)
 Other3 (3)
Direct contact with vessels ≥3 mm diameter26/90 (29)

Pneumothorax occurred after 22 (34%) of 65 procedures, of which 16 (25%) needed to be treated by chest tube drainage. Four patients had major complications: 3 patients were treated for pneumonia, and 1 had a hematothorax. Three patients had minor complications: parenchymal bleeding around the ablation zone or atelectasis (Table 3). Other postinterventional symptoms were pleural effusion, hemoptysis, pain, and coughing. No anesthetic complications occurred. One delayed treatment-related death occurred; this patient died after 3 months, after several infectious events and pulmonary deterioration.

Table 3. Treatment Characteristics and Outcome
Characteristic/OutcomeNo. (%)
  • RFA indicates radiofrequency ablation; CI, confidence interval; KM, Kaplan-Meier.

  • RFA 2004-2009; 90 metastases in 65 procedures in 46 patients.

  • a

    Some patients had more >1 complication.

Metastases treated per patient
 123/46 (50%)
 216/46 (35%)
 ≥37/46 (15%)
Number of RFA sessions per patient
 133/46 (72%)
 210/46 (22%)
 ≥33/46 (6%)
Number of metastases treated per RFA session
 146/65 (71%)
 213/65 (20%)
 36/65 (9%)
Mortality1/65 (1.5%)
Post-treatment hospital stay, median d [range]2 [1-11]
Readmission <30 days5/65 (8%)
Complicationsa31/65 (48%)
 Uncomplicated sessions36/65 (55%)
 Pneumothorax22/65 (34%)
 Major4/65 (6%)
 Other minor5/65 (8%)
Follow-up after RFA, median mo [range]22 [2-65]
Progression-free survival, 65 procedures, %
 1 year33
 3 year11
Estimated time to recurrence, median mo {95% CI}4 {2.7-5.3}
Local progression25/90
Patients considered tumor free after first RFA30/46 (65%)
Overall survival, 46 patients, %
 1 year84%
 3 year69%

Median follow-up after RFA was 22 months (range, 2-65 months). Analysis per procedure showed a median progression-free survival estimate of 4 months (95% confidence interval [CI], 2.7-5.3); 1- and 3-year progression free survival rates per procedure were 33% and 11% (Fig. 2, Table 3). For patients with all lesions treated (n = 36), 1- and 3-year progression-free survival rates were 49% and 20%, respectively. Median estimated time to progression in this group was 10 months (95% CI, 5.7-14.3).

Figure 2.

Kaplan-Meier graph depicts any-site recurrence after the radiofrequency ablation (RFA) procedure.

After 25 (38%) of 65 RFA sessions, local tumor progression occurred. Evaluation per lesion showed local progression in 25 of 90 (Kaplan-Meier estimate, 22% at 1 year, 35% at 2 years) (Fig. 3). To avoid underreporting, patients with widespread recurrent disease, in whom chest CT follow-up was discontinued, were censored in the analysis of local progression. Tumor size >30 mm was associated with a higher chance of local progression, although this was not statistically significant (47% vs 24%; P = .07). Contact with vessels with a ≥3 mm diameter, pleural/parenchymal localization, unilateral or bilateral localization, previous thoracic interventions, disease-free interval ≥18 months, and primary tumor type were not identified as significant prognostic factors for local progression in this analysis.

Figure 3.

Kaplan-Meier graph depicts local progression of ablated lesions.

Twelve patients died during follow-up. One death was procedure-related, as mentioned previously. Death in the other patients was disease-related. Estimated median overall survival for the complete cohort was 55 months (95% CI, 26-84). Eighty-four percent were alive after 1 year and 69% after 3 years. Figure 4 shows a significant overall survival difference between patients considered to be tumor free after their first RFA and patients with residual lesions (79% vs 49% at 3 years, P = .011). At the end of follow-up, 7 (15%) patients were considered to be without recurrence.

Figure 4.

Kaplan-Meier graph depicts survival after the first radiofrequency ablation (RFA) procedure in patients with or without residual lesions (P = .011).


Long-term follow-up of patients after RFA for pulmonary metastases shows considerable local progression rates. Although RFA is a relatively safe technique with low morbidity and mortality,5-11 its place in the treatment of pulmonary metastases needs to be discussed.

Oligometastatic disease is a controversial concept. Many will argue that in fact it represents limited detectable metastatic disease with more widespread occult microscopic metastatic deposits. The process of cell colonization and proliferation in another organ is considered biologically highly inefficient. Alternatively, the host may be immunologically competent to control metastatic (out)growth.14

Despite the lack of randomized trials, local treatment of oligometastatic disease is associated with favorable survival, although recurrent metastatic disease is common. Previously reported series of pulmonary metastasectomies have shown favorable 5-year overall survival rates. In the combined group (all primary tumors) of the International Registry of Lung Metastases, complete resection of oligometastatic disease resulted in a 5-year survival rate of 36%.1 Only a proportion of these patients have remained disease free (5-year disease-free survival, <20%), but even repeat metastasectomy is associated with favorable survival.15-18 Whether this is because of local treatment of the metastasis or because of favorable tumor biology, which happens to coincide with resectability, still remains to be proven.19

In recent years, less invasive techniques, such as RFA and stereotactic radiotherapy,20 have emerged as potential alternatives for local treatment of pulmonary tumors, with possibly lesser side effects and morbidity. Local tumor control, however, should be adequate as well. Although many patients will face recurrence anywhere in the lungs or at other distant sites after treatment of pulmonary metastases, local control should be the best possible.

Data on local progression are difficult to compare, because of differences in definitions. A recent review of stereotactic radiotherapy reported widely varying local control rates from 67% to 96% after median follow-up of 18 months.20 With various inclusion criteria, endpoints (eg, primary tumor origin, tumor size and number, definition of local progression), and follow-up, reported local progression rates after RFA range from 7% to 36%, with a median 12- to 27-month period of follow-up.5-11 The definition of local progression is of major importance in judging the efficacy of RFA. In our study, we decided to define local progression as growing lesions in the ablation zone or within 1 cm from this zone. Evaluation per lesion showed local progressions at 25 of 90 lesions after median follow-up of almost 2 years (Kaplan-Meier estimate, 35%).

Significantly better local tumor control has been reported for tumors <2.5 to 3.5 cm.7, 8, 12, 21 Simon et al and Yamakado et al reported local progression rates for tumors <3 cm of 20% (colorectal metastases) and up to 75% for tumors >3 cm (various primary tumors) after a median follow-up of 25 to 27 months.7, 10 Our data are consistent with these previous reports. Some pathomorphological studies might explain the high local progression rates. In a small cohort study by Clasen et al, RFA for lung malignancies was followed by metastasectomy and pathomorphological examination. Tumor ablation was found to be complete in 91% of the tumors of up to 47 mm diameter. Failure occurred when no sufficient ablation margin was reached, because of an adjacent large vessel (>3 mm).13 A similar study by Nguyen et al found >80% nonviability in 88% of tumors, but complete tumor cell necrosis in only 38% of tumors. Complete tumor cell necrosis was found in smaller (<2 cm) lesions.22

The local progression rate in our study is high (35%), but within the reported range. Although local progression rates are considerable, and recurrence rates at any site are even higher (70% within 1 year), survival rates in our series are still acceptable. An overall survival rate of 69% after 3 years is surprisingly good, emphasizing the finding that patients with (even recurrent) oligometastatic disease have a relatively good prognosis. Increasingly incorporating this technique in a multimodality strategy with systemic treatment might explain these results.23 Any survival benefit of RFA or any other local treatment modality (surgery, radiotherapy) can only be evaluated in controlled studies, which are lacking.19

Although this study is limited by lack of a control group and describing a highly selected and heterogeneous population, some careful conclusions may be drawn. RFA for pulmonary oligometastatic disease is associated with considerable local progression rates. Therefore, this technique should be further improved, and hopefully new developments such as microwave ablation will be able to induce tumor necrosis more efficiently.20, 24 Long-term disease control in patients with (recurrent) oligometastatic disease, however, is possible, and further studies are needed to define the role of local ablation techniques, stereotactic radiotherapy, and surgery.


The authors made no disclosures.