Percutaneous radiofrequency ablation (RFA) is a locoregional. Radiofrequency Ablation of Lung Tumors: Feasibility and Safety

Transcription

1 Radiofrequency Ablation of Lung Tumors: Feasibility and Safety GENERAL THORACIC Jacqui C. Zhu, MBBS, Tristan D. Yan, BSc, MBBS, PhD, Derek Glenn, MBBS, and David L. Morris, MD, PhD Departments of Surgery and Radiology, St. George Hospital, Sydney, Australia Background. Percutaneous image-guided radiofrequency ablation is being promoted as a novel technique with a low morbidity rate in the treatment of inoperable lung tumors. The purpose of this study was to assess the incidence and risk factors of various complications after radiofrequency ablation of pulmonary neoplasms. Methods. The clinical and treatment-related data regarding 129 consecutive percutaneous radiofrequency ablation treatment sessions for 100 patients with inoperable lung tumors were collected prospectively. Univariate and multivariate analyses were conducted to identify significant risk factors associated with postprocedural overall morbidity, pleuritic chest pain, hemoptysis, pneumothorax, pleural effusions, and chest drain requirement. Results. There was no postprocedural mortality. The overall morbidity rate was 43% (n 55 of 129). The most common adverse effect was pneumothorax, occurring in 32% (n 41 of 129) of treatment sessions. Other significant complications included pleuritic chest pain (18%, n 23 of 129), hemoptysis (7%, n 9 of 129), pleural effusions (12%, n 15 of 129), and chest drain insertion (20%, n 26 of 129). Both univariate and multivariate analyses identified more than two lesions ablated per session as a significant risk factor for overall morbidity, pneumothorax, and chest drain insertion, but not for pleuritic pain, hemoptysis, and pleural effusions. Length of the ablation probe trajectory greater than 3 cm was an additional independent risk factor for overall morbidity and pneumothorax. Hilar location of lung tumor/s was the only independent risk factor associated with the increased incidence of hemoptysis. Conclusions. Radiofrequency ablation for lung tumors can be considered as a safe and technically feasible procedure with acceptable incidence of complications. (Ann Thorac Surg 2009;87:1023 9) 2009 by The Society of Thoracic Surgeons Percutaneous radiofrequency ablation (RFA) is a locoregional therapy that has recently emerged as an alternative treatment option for nonsurgical candidates with lung tumors. A number of studies have reported the promising clinical safety and efficacy profiles of RFA application [1 16]. The main advantage of image-guided lung RFA is avoiding thoracotomy and its associated short hospital stay [1 4, 7, 10, 14]. However, the risk factors associated with its complications have rarely been delineated. The purpose of this study was to assess the incidence of complications of RFA for lung tumors, with a special reference to clarify risk factors for these complications. Patients and Methods Since November 2000, a total of 129 consecutive RFA procedures were performed for a cohort of 100 patients with primary and secondary lung tumors. These included 18 repeat, 3 third-time, and 3 fourth-time RFA procedures. Written informed consent was obtained from all patients before treatment. The study was approved by Accepted for publication Nov 10, Address correspondence to Prof Morris, Department of Surgery, St. George Hospital, Sydney, NSW, Australia; david.morris@ unsw.edu.au. the hospital ethics committee, and a waiver for individual patient consent for this retrospective review was also obtained from the ethics committee. In this study, we were mainly interested in the feasibility of performing percutaneous RFA for primary and secondary lung tumors. Feasibility was defined as the ability to successfully perform a percutaneous RFA without significant morbidity or perioperative mortality. Significant morbidities included pleuritic pain requiring opioid analgesia, hemoptysis, pneumothorax, pleural effusions, and chest drain insertion. Patient Selection Selection of patients to undergo lung RFA was often multifactorial, and the consensus on the treatment plans for these patients was obtained from a group of surgical oncologists, medical oncologists, and a radiologist at weekly multidisciplinary seminars. Specific inclusion criteria consisted of (1) age between 18 and 85 years; (2) patients with early stage non small cell lung carcinoma pulmonary metastases evaluated by thoracic surgeons at our institution and precluded from surgery because of poor pulmonary function with FEV 1 (forced expiratory volume in 1 second) less than 1 L, cardiac risk with New York Heart Association Class of 3 or greater, or poor performance status with Eastern Cooperative Oncology 2009 by The Society of Thoracic Surgeons /09/$36.00 Published by Elsevier Inc doi: /j.athoracsur

2 GENERAL THORACIC 1024 ZHU ET AL Ann Thorac Surg LUNG NEOPLASM 2009;87: Group performance of 2 or greater; and (3) patient refusing to undergo surgery. Exclusion criteria included the following: (1) diameter of metastases greater than 5 cm; (2) greater than 6 lesions per hemithorax; (3) lesions immediately adjacent to major pulmonary vessels; (4) lesions located immediately adjacent to the pulmonary hilum of the lung; (5) lesions immediately adjacent to major bronchi; (6) bleeding diatheses not responding to medical treatment (prothrombin time greater than 1.5 and platelets less than ); and (7) extrapulmonary disease. Radiofrequency Ablation Technique Before the procedure, all patients underwent clinical examinations, bone scans, and abdominal, pelvic, and chest computed tomography (CT) to assess the tumor and plan the procedure. Spirometry was a routine part of physiologic staging. Serial measurement of carcinoembryonic antigen was also obtained in patients with colorectal lung metastases. Positron emission tomography was used selectively at our institute. Two interventional radiologists performed all the RFA procedures. Patient positioning for the 129 treatment sessions was determined by tumor location, with supine and prone positions to achieve the shortest probe trajectory to access anterior and posterior tumors, respectively. All the procedures were performed with the RITA starburst XL electrode probe (RITA Medical, Mountain View, CA), which has a diameter of 14-gauge with nine deployable tines and is available in three lengths: 10, 15, and 25 cm. Only 10- and 15-cm probes were used owing to space limitations of the CT gantry. The electrode probe was able to create a maximal area of ablation of 5 cm. The generator used was the RITA 1500 generator (RITA Medical) with real-time recording as well as display of variables including temperature, power, and impedance. Grounding pads were placed on both thighs with the wider edge facing the RFA site to lessen the risk of skin burns. All patients were provided with intravenous access. Both conscious sedation and general anesthesia have been used at our institute. The first 100 procedures were performed under conscious sedation consisting of generous local anesthesia (Lignocaine 1%, Pfizer Pty Ltd, Perth, Australia) and intravenous sedation (Morphine and Midazolam [Hospira Australia Pty Ltd, Melbourne, Australia] on demand). The subsequent 29 procedures were performed under general anesthesia and positivepressure ventilation through endotracheal intubation. All procedures were performed using aseptic technique, and the RFA probe was advanced to the tumor under CT (Xpress SX; Toshiba, Tokyo, Japan) fluoroscopic guidance. After confirming that the needle tip was advanced to the desired location in the tumor to be treated, an ablation algorithm with staged electrode deployment was used as per manufacturer instructions. A target temperature of 90 C and a maximum power of 150 W were used. The target temperature was maintained for 15, 20, and 27 minutes to achieve ablation zones of 3, 4, and 5 cm, respectively. Whenever technically feasible, an ablation margin of at least 1 cm in all planes was achieved. For ablations greater than 5 cm, overlapping ablations were performed to ensure complete ablation. Additional ablations were performed by withdrawing the electrode and changing its trajectory within the pulmonary parenchyma without rebreaching the pleura to reduce the incidence of pneumothorax. To prevent tumor seeding, track ablations were routinely performed to cauterize the access track on the way out at the completion of each tumor ablation. Throughout the procedure, the patients were continuously monitored for oxygen saturation, pulse rate, blood pressure, and temperature. This is continued hourly for 6 hours after the procedure. Preprocedural or intraprocedural antimicrobial agents were not routinely administered for prophylactic reasons. Post-Radiofrequency Ablation Monitoring All patients were monitored in hospital overnight, and chest radiography was performed after RFA procedure and before discharge to exclude pneumothorax and pleural effusions. Patients with small asymptomatic pneumothorax were observed. Symptomatic patients or those with large pneumothoraces or effusions were managed with chest drain insertion and repeat radiograph until the pneumothorax had resolved. After discharge, all patients were followed up with chest CT performed at 1 week, 1 month, and every 3 months thereafter to monitor the treatment response (see Fig 1 for sequential CT monitoring of an RFA-treated lesion). Data Collection The multiple clinical and treatment-related variables were collected from a prospectively maintained computerized database. The following patient variables were recorded: sex, age, the presence of pulmonary emphysema, and any history of surgery to the RFA-treated hemithorax. For the purpose of this study, pulmonary emphysema was defined as FEV 1 to forced vital capacity less than 0.70 and FEV 1 less than 80% predicted. Tumor variables collected were tumor size, number, location, and distribution. Tumor size was the diameter measured along the longest axis of the largest lesion ablated. Hilar location was regarded as the presence of a lesion located within 3 cm from the pulmonary hilum. Tumor located within 1 cm of pulmonary vessels and bronchi was defined as in close proximity to such structures. Distribution of lesions was characterized by involvement of one or more lobes and hemithoraces. Relevant ablation variables consisted of number of overlapping ablations used per tumor, postablation size, total ablation time, patient positioning, type of anesthesia, and length of aerated lung traversed by the electrode probe. The length of the electrode trajectory through the aerated lung to each lesion was measured on CT images and defined as the distance from the site of pleural puncture to the edge of the tumor along the electrode trajectory. The postablation size was the diameter measured along the longest axis of the postablation zone. For patients who had multiple lesions ablated at a single session, the number of overlapping ablations and the postablation size of the largest lesion was used in the analysis. The longest electrode trajectory was recorded

3 Ann Thorac Surg ZHU ET AL 2009;87: LUNG NEOPLASM 1025 Results Demographic Data These 100 patients underwent a total of 129 ablation sessions performed by two experienced interventional radiologists. The main reasons for RFA treatment are given in Table 1. The reasons leading to consideration for lung RFA were often multifactorial. The mean age of the study cohort at the time of the RFA was 65 8 years. There were 56 male patients. The mean and median number of tumors ablated per treatment session was and 2 (range, 1 to 7), respectively. The mean and median size of tumors ablated in all sessions was cm and 2.0 cm (range, 0.5 to 5 cm), respectively. The mean duration of each ablation session was hours. Six patients were treated for primary lung carcinoma while the remaining patients had pulmonary metastases. The most common extrathoracic primary cancer was colorectal cancer (n 68), followed by renal cell carcinoma (n 8), soft-tissue sarcoma (n 6), and miscellaneous tumors (n 12). All patients with secondary lung tumors had no evidence of other systemic diseases. GENERAL THORACIC Fig 1. Sequential computed tomography images of lung radiofrequency ablation. (A) A solitary pulmonary metastasis before treatment; (B) opacification during a lung radiofrequency ablation treatment; (C, D, E, and F) post radiofrequency ablation computed tomography images at 1, 3, 6, and 12 months, respectively, transforming from a zone of consolidation to a small fibrotic scar. and used in the analysis if multiple electrode trajectories were created. Statistical Analysis Repeat ablation sessions were considered to be independent events, and analysis was performed by using the total number of ablation sessions as the sampling unit (total of 129 sessions). Multiple clinical and treatmentrelated variables were subsequently divided into categorical variables. The univariate analysis used to determine the underlying risk factors contributing to various complications was performed using the 2 analysis or Fisher s exact test. Multivariate analysis using a binary logistic regression model was performed for various factors related to procedure-related complications to identify independent risk factors. A probability value less than 0.05 was considered to indicate significant difference for all analyses. Statistical analyses were all performed using a commercially available software program (SPSS for Windows, version 15.0, SPSS Inc, Chicago, IL). Morbidity and Mortality There was no procedure-related mortality. The mean and median hospital stay was 1 2 days and 1 day (range, 1 to 5, respectively). The overall morbidity rate for all treatment sessions was 43% (n 55 of 129), with some procedures complicated by multiple adverse effects. The most common complication was pneumothorax (32%, n 41 of 129), which was either identified during the procedure or within 24 hours on postprocedural chest radiograph. Twenty-one of the pneumothoraces detected were small and asymptomatic, requiring no chest tube drainage. Pleural effusions occurred in 12% (n 15 of 129) of treatment sessions. Six of these 15 cases of pleural effusions required chest tube insertion. The pleural fluids drained from these 6 patients were serous or transudative in nature, and we have not experienced hemothorax as a complication after RFA. The mean duration of chest tube insertion for both pneumothoraces and pleural effusions was days. Other common side effects included pleuritic pain requiring opioid analgesia (18%, n 15 of 129) and hemoptysis (9%, n 7 of 129). All patients with chest pain responded well to short-term analgesia, and no patients reported having persistent procedure-related pleuritic chest pain. All cases of hemoptysis were self-limiting, with no patients requiring blood transfusion or bronchoscopy to achieve hemostasis. Although postprocedural fever within 1 week of RFA procedure (body temperature 38 C) was noted in 10% of cases (n 13 of 129), pneumonia requiring antimicrobial treatment occurred only in 4% of cases (n 5 of 129). There was one case of large hydropneumothorax, which was managed successfully with chest tube insertion for 7 days. Furthermore, we encountered 1 patient who experienced a 4-2-cm partial-thickness cutaneous burn at the site of the thigh grounding pad attachment. This required no debridement and healed completely in 2 months. Most patients were asymptomatic after treatment and were observed for 24 hours with serial chest radiographs before discharge from the hospital. Risk Factors Both univariate and multivariate analysis of 20 potentially significant factors was performed to determine risk

4 GENERAL THORACIC 1026 ZHU ET AL Ann Thorac Surg LUNG NEOPLASM 2009;87: Table 1. Patient Characteristics Variable Result Age (y) Mean 65 8 Median (range) 69 (30 83) Sex (n/n) Male 56/100 Female 44/100 Tumor type (n/n) NSCLC 6/100 Metastatic 94/100 Tumor stage (n/n) NSCLC stage IA 3/6 NSCLC stage IB 2/6 NSCLC stage IIA 1/6 Previous therapy (n/n) Chemotherapy 28/100 Radiotherapy 19/100 Lung resection 21/100 Liver resection 39/100 Main reasons for no surgery (n/n) Poor lung function 35/100 Severe cardiac risk 18/100 Poor performance status 36/100 Refusal of surgery 24/100 Length of hospital stay (days) Mean 1 2 Median (range) 1 (1 5) n number of patients; N total number of patients; NSCLC non small cell lung carcinoma. factors for overall morbidity, pleuritic chest pain, hemoptysis, pneumothorax, pleural effusions, and chest drain insertion for the management of pneumothorax and pleural effusions. These factors consisted of (1) demographic factors including age ( 65 years; n 68 versus 65 years; n 61) and sex (male; n 76 versus female; n 53); (2) clinical factors including previous lung surgery (yes; n 30 versus no; n 99), pulmonary emphysema (yes; n 16 versus no; n 113), tumor diagnosis (colorectal metastases; n 91 versus noncolorectal lesions; n 32), number of pulmonary lesions treated ( 2 lesions; n 97 versus 2 lesions; n 32), largest pulmonary lesion treated ( 2 cm; n 80 versus 2 cm; n 49), location of lesion (hilar; n 28 versus peripheral; n 101), proximity of lesion to bronchus (yes; n 25 versus no; n 107), proximity of lesion to pulmonary vessels (yes; n 27 versus no; n 107), distribution of lesions (bilateral; n 33 versus unilateral; n 96), and number of lobes treated (1 lobe; n 76 versus 1 lobe; n 53); and (3) treatment factors including size of ablation zone ( 3 cm; n 42 versus 3 cm; n 87), number of overlapping ablations (1; n 105 versus 1; n 24), total ablation time ( 2 hour; n 64 versus 2 hours; n 65), length of probe trajectory ( 3 cm; n 78 versus 3 cm; n 51), type of anesthesia (general anesthesia; n 29 versus sedation; n 100), patient position (supine; n 65 versus prone; n 64), repeat RFA (yes; n 29 versus no; n 100), and treatment period (first 3 years; n 65 versus last 3 years; n 64). Overall Morbidity Four risk factors significantly associated with overall morbidity were identified by univariate analysis, including greater than 2 lesions ablated (p 0.001), bilateral distribution of lesions (p 0.024), total ablation time of 2 hours or greater (p 0.033), and length of probe trajectory greater than 3 cm. Multivariate analysis further recognized number of lesions ablated (p 0.001; odds ratio [OR], ; 95% confidence interval [CI], to ) and length of probe trajectory (p 0.030; OR, 2.895; 95% CI, to 7.584) as two independent risk factors for increased frequency of overall morbidity. Pain Three factors significantly associated with pleuritic chest pain requiring opioid analgesia were identified by univariate analysis, including greater than 2 lesions ablated (p 0.002), bilateral distribution of disease (p 0.001), and total ablation time of 2 hours or greater (p 0.020). Multivariate analysis suggested no independent risk factors for increased frequency of procedure related pleuritic chest pain. Hemoptysis One factor significantly associated with hemoptysis was identified by univariate analysis: hilar tumor location (p 0.023). Multivariate analysis further recognized hilar location of tumor (p 0.040; OR, ; 95% CI, to ) as an independent risk factor for increased frequency of hemoptysis after RFA. Pneumothorax Five risk factors significantly associated with pneumothorax were identified on univariate analysis, including more than 2 lesions ablated (p 0.001), bilateral distribution of lesions (p 0.009), more than 1 lobe ablated (p 0.001), total ablation time of 2 hours or greater (p 0.002), and length of probe trajectory greater than 3 cm (p 0.033). Multivariate analysis further recognized number of lesions ablated (p 0.001; OR, ; 95% CI, to ) and length of probe trajectory (p 0.042; OR, 3.108; 95% CI, to 9.255) as two independent risk factors for increased frequency of pneumothorax. Pleural Effusions Four risk factors significantly associated with pleural effusions were identified on univariate analysis including more than 2 lesions ablated (p 0.001), bilateral distribution of lesions (p 0.009), more than 1 lobe ablated (p 0.011), total ablation time of 2 hours or greater (p 0.025), and repeat RFA procedures (p 0.022). Multivariate analysis suggested no independent risk factors for increased frequency of pleural effusions.

5 Ann Thorac Surg ZHU ET AL 2009;87: LUNG NEOPLASM Chest Drain Placement Two risk factors significantly associated with chest drain insertion were identified on univariate analysis, including more than 2 lesions ablated (p 0.001), bilateral distribution of lesions (p 0.007), and total ablation time of 2 hours or greater (p 0.047). Multivariate analysis further recognized number of lesions ablated (p 0.001; OR, ; 95% CI, to ) as an independent risk factor for increased frequency of chest drain insertion. Comment The safety and minimal invasiveness of percutaneous RFA for unresectable lung tumors have been suggested in a number of series [1 16]. However, only five studies have analyzed risk factors associated with RFA-related complications [2, 15, 17 19]. Our series demonstrated no postprocedural mortality. The worldwide reported overall morbidity ranged from 15.5% to 55.6% [1 19]. At our institute, our overall morbidity in 129 treatment sessions was 43%. Pleuritic chest pain requiring opioid analgesia was a relatively common adverse effect of RFA procedures. Pain on univariate analysis was associated with increased number of tumors, bilateral distribution of tumors, and total ablation time. All such factors are related to increased puncturing of the chest wall and the pleura, hence increasing the probability of postprocedural pain. All the cases of hemoptysis in our cohort of patients were self-limiting, lasting a few days. No patients required transfusion or surgical intervention. Hilar location was the only significant risk factor for hemoptysis. Hilar location may indicate proximity to major pulmonary vessels and increase the risk of intraparenchymal hemorrhage during tumor ablation. Pneumothorax was been consistently shown to be a very common complication [1 19]. The incidence of pneumothorax has been inconsistently reported, ranging from 4.5% to 61.1% [2 6, 8 15, 17 19]. These mixed results could be explained by the fact that the diagnostic definition of pneumothorax differs at different treatment centers [2 6, 8 15, 17 19]. Although some studies reported frequency of radiologically detected pneumothoraces irrespective of size, some centers only reported large or clinically significant pneumothoraces [2 6, 8 15, 17 19]. At our center, pneumothorax was the most common complication, occurring in 32% of RFA procedures. Our relatively high incidence could be attributable to our radiologic diagnosis (rather than clinical diagnosis) of pneumothorax on our routine postprocedural chest radiographs. Only 20 cases of pneumothorax (16%) required definitive chest drain insertion, and the duration of chest drain insertion was short, with an average drainage of 2 days. The risk factors associated with pneumothorax in reported literature have also been conflicting [2, 15, 17 19]. Risk factors identified include (1) patient factors: male sex, no previous lung surgery, and presence of emphysema; (2) tumor factors: number of lesions ablated and lower and middle lobe involvement; and (3) protocol 1027 factors: RFA treatment period, length of aerated lung transversed by electrode, and number of ablation positions [2, 15, 17 19]. Our study also demonstrated the number of tumors greater than 2 and length of lung parenchyma traversed by probe trajectory greater than 3 cm as independent risk factors associated with increased incidence of pneumothorax. Number of tumors greater than 2 was a particularly significant risk factor for pneumothorax with an odds ratio of when compared with tumor numbers ablated per session being less or equal to 2. This is only intuitive, as increased pleural punctures have been created, as would the risk of pneumothorax be greater. Furthermore, increased probe trajectory length may represent increased technical difficulty of reaching the tumor and increased need for repositioning of the ablation probe when compared with the more easily accessible tumors. We have altered our anesthetic regimen from conscious sedation to general anesthesia for better patient comfort and intraprocedural analgesic control. It is interesting to note at this point at our institution that general anesthesia with positivepressure ventilation was not a risk factor for increased development of pneumothorax in both univariate and multivariate analysis. Pleural effusions after ablation were also a common complication and occurred after 12% of treatment sessions. The pathogenesis of pleural effusions after RFA has been hypothesized to be initiated by the conductance of heat over the pleura during the tumor ablation. The majority of pleural effusions were small in amount and asymptomatic, with only six moderate to massive pleural effusions requiring chest tube drainage. Univariate analysis identified number of lesions ablated greater than 2, ablation of bilateral hemithoraces, number of lobes ablated greater than 1, and total ablation time greater than 2 hours as risk factors for the development of pleural effusions. This is understandable when such tumor- and treatment-related factors would all potentially predispose the pleura to longer exposure of heat irritation and hence the development of effusions. Chest tube placement for pneumothorax and pleural effusions after RFA procedures occurred in 20% of procedures. This incidence of chest tube drainage is again inconsistent, ranging from 3.3% to 38.9% in different reported series [2 7, 10, 12, 14, 15, 17 19]. This inconsistency is largely explained by the differences in the management approach of RFA-related pneumothoraces [2 7, 10, 12, 14, 15, 17 19]. Some centers recommend treating pneumothoraces with aspiration only, whereas others, including our institution, prefer insertion of a chest tube with an underwater seal for drainage [2 7, 10, 12, 14, 15, 17 19]. Number of lesions ablated per session greater than 2 that increased the risk for both pneumothorax and pleural effusions was identified on multivariate analysis as the only significant risk factor for the higher prevalence of chest tube drainage. In conclusion, our first 129 procedures for these 100 patients have demonstrated that RFA for unresectable lung tumors is technically feasible. The low complication rate, short chest tube insertion duration and hospital stay GENERAL THORACIC

6 GENERAL THORACIC 1028 ZHU ET AL Ann Thorac Surg LUNG NEOPLASM 2009;87: coupled with recently reported promising survival results suggest that this therapy may be a useful alternative for nonsurgical candidates. Increased understanding of the potential risk factors associated with common adverse events may facilitate better prevention and prompt recognition and management of complications. References 1. Fernando HC, De Hoyos A, Landreneau RJ, et al. Radiofrequency ablation for the treatment of non-small cell lung cancer in marginal surgical candidates. J Thorac Cardiovasc Surg 2005;129: Yan TD, King J, Sjarif A, et al. Learning curve for percutaneous radiofrequency ablation of pulmonary metastases from colorectal carcinoma: a prospective study of 70 consecutive cases. Ann Surg Oncol 2006;13: Suh RD, Wallace AB, Sheehan RE, et al. Unresectable pulmonary malignancies: CT-guided percutaneous radiofrequency ablation preliminary results. Radiology 2003;229: Lee JM, Jin GY, Goldberg SN, et al. Percutaneous radiofrequency ablation for inoperable non-small cell lung cancer and metastases: preliminary report. Radiology 2004;230: Akeboshi M, Yamakado K, Nakatsuka A, et al. Percutaneous radiofrequency ablation of lung neoplasms: initial therapeutic response. J Vasc Interv Radiol 2004;15: Yasui K, Kanazawa S, Sano Y, et al. Thoracic tumors treated with CT-guided radiofrequency ablation: initial experience. Radiology 2004;231: Gadaleta C, Catino A, Ranieri G, et al. Radiofrequency thermal ablation of 69 lung neoplasms. J Chemother 2004; 16: Kang S, Luo R, Liao W, et al. Single group study to evaluate the feasibility and complications of radiofrequency ablation and usefulness of post treatment position emission tomography in lung tumours. World J Surg Oncol 2004;2: Belfiore G, Tedeschi E, Brillantino C, et al. Palliation of unresectable lung cancer with CT-guided radiofrequency ablation: clinical and imaging findings at one-year. Int J Clin Invest 2006;2: VanSonnenberg E, Shankar S, Morrison PR, et al. Radiofrequency ablation of thoracic lesions: part 2, initial clinical experience technical and multidisciplinary considerations in 30 patients. AJR Am J Roentgenol 2005;184: Kishi K, Nakamura H, Kobayashi K, et al. Percutaneous CT-guided radiofrequency ablation of pulmonary malignant tumors: preliminary report. Intern Med 2006;45: Ambrogi MC, Lucchi M, Dini P, et al. Percutaneous radiofrequency ablation of lung tumours: results in the mid-term. Eur J Cardiothorac Surg 2006;30: Lagana D, Carrafiello G, Mangini M, et al. Radiofrequency ablation of primary and metastatic lung tumors: preliminary experience with a single center device. Surg Endosc 2006;20: de Baere T, Palussiere J, Auperin A, et al. Midterm local efficacy and survival after radiofrequency ablation of lung tumors with minimum follow-up of 1 year: prospective evaluation. Radiology 2006;240: Hiraki T, Tajiri N, Mimura H, et al. Pneumothorax, pleural effusion, and chest tube placement after radiofrequency ablation of lung tumors: incidence and risk factors. Radiology 2006;241: Thanos L, Mylona S, Pomoni M, et al. Percutaneous radiofrequency thermal ablation of primary and metastatic lung tumors. Eur J Cardiothorac Surg 2006;30: Yamagami T, Kato T, Hirota T, et al. Pneumothorax as a complication of percutaneous radiofrequency ablation for lung neoplasms. J Vasc Interv Radiol 2006;17: Gillams AR, Lees WR. Analysis of the factors associated with radiofrequency ablation-induced pneumothorax. Clin Radiol 2007;62: Okuma T, Matsuoka T, Yamamoto A, et al. Frequency and risk factors of various complications after computed tomography-guided radiofrequency ablation of lung tumors. Cardiovasc Intervent Radiol 2008;31: INVITED COMMENTARY Zhu and associates [1] give us a realistic account of the risk and rate of complications related to radiofrequency ablation (RFA) procedures anticipated among clinicians skilled in these approaches. Important messages related to lesion selection for RFA based on size, depth, and the relative relationship of the lesion to pulmonary hilar bronchovascular structures are also illuminated. The authors commentary on patient selection for this compromise RFA approach to the otherwise resectable peripheral malignant pulmonary nodule is also appreciated. This brings me to my concerns with the use of RFA among marginal candidates for anatomic pulmonary resection and the growing acceptance of these approaches as more favorable alternatives to sublobar resection among patients with impairment in cardiopulmonary reserve or with lung metastases from remote primary malignancies. The American College of Surgeons Oncology Group (ACOSOG) is presently evaluating the role of RFA for highrisk patients with stage IA non-small cell lung cancers in a phase II trial schema (Z4033). Simultaneously open within ACOSOG is another trial (Z4032) evaluating the outcomes of similarly high-risk patients with stage IA disease randomized to sublobar resection alone or sublobar resection with intraoperative brachytherapy [2]. Primary end points of these trials are related to local control of the patients tumors and procedurally related morbidity. These are very important studies that will objectively determine some of the outcomes of these approaches to the high-risk patient with resectable early-stage lung cancer. Thoracic surgeons and our medical colleagues should patiently await the results of these trials before making the leap of faith toward applying these image-guided ablative approaches preferentially for the management of such malignant pulmonary parenchymal targets. Zhu and colleagues work involves the use of RFA for 100 patients with a variety of malignant lung nodules, only 6 of which were primary lung cancer. In the largest group of patients treated, the pulmonary metastases were from colorectal primary cancers. Patient selection criteria for performing the ablation of these metastatic lesions are not delineated in their work, and one is left to assume that the authors were directing these RFAs to patients with oligometastases with intent to clear all potential disease demonstrable by computed tomography of the chest. Interestingly, this concept of RFA management of metastatic pulmonary 2009 by The Society of Thoracic Surgeons /09/$36.00 Published by Elsevier Inc doi: /j.athoracsur