Combined Percutaneous Radiofrequency Ablation and Ethanol Injection of Renal Tumours: Midterm Results

Transcription

1 european urology 52 (2007) available at journal homepage: Kidney Cancer Combined Percutaneous Radiofrequency Ablation and Ethanol Injection of Renal Tumours: Midterm Results Nicos I. Fotiadis a, *, Tarun Sabharwal a, Jose P. Morales a, Dominic J. Hodgson b, Tim S. O Brien b, Andreas Adam a a Department of Interventional Radiology, Guy s and St Thomas Hospital, London, UK b Department of Urology, Guy s and St Thomas Hospital, London, UK Article info Article history: Accepted March 19, 2007 Published online ahead of print on March 28, 2007 Keywords: Ablation Ethanol Percutaneous Radiofrequency Renal tumours Abstract Objectives: To evaluate the safety and efficacy of combined percutaneous, image-guided, radiofrequency (RF) ablation and ethanol injection of renal tumours, and to present our midterm results. Methods: Since February 2002, 27 consecutive patients (22 men, 5 women; age range: yr; mean: 69) with 28 renal tumours (mean diameter: 2.87 cm) were treated with combined percutaneous RF and ethanol ablation, and were prospectively evaluated. Twenty-five patients were considered nonsurgical candidates because of comorbid conditions (16 patients) or had previous nephrectomy (9 patients), and 2 had refused surgery. Thirty-three ablation sessions were performed, with computed tomography (26 sessions), ultrasound (6), or combined magnetic resonance imaging/fluoroscopic guidance in 1. Absolute ethanol (0.5 3 ml; mean: 1.7) was injected into the tumour immediately before treatment with radiofrequency. Mean follow-up period was 18.6 mo (range: 3 56). Results: Twenty-seven of the 28 tumours were completely ablated with either one (21 tumours) or two treatment sessions (6 tumours). One patient with residual disease refused further treatment. Only three minor complications, including a subcapsular haematoma and two patients with loin pain, occurred; all three patients were treated conservatively. None of the complications was related to the ethanol injection. During the follow-up period, no evidence of local recurrence or metastatic disease was seen. Creatinine levels have not changed significantly in any of the patients following ablation. Conclusions: Combined use of percutaneous RF and ethanol ablation is a safe and effective alternative treatment for selective patients with renal tumours. # 2007 European Association of Urology. Published by Elsevier B.V. All rights reserved. * Corresponding author. Department of Interventional Radiology, First Floor, Lambeth Wing St Thomas Hospital, Lambeth Palace Road, London SE1 7EH, UK. Tel ; Fax: address: nicos.fotiadis@kcl.ac.uk (N.I. Fotiadis) /$ see back matter # 2007 European Association of Urology. Published by Elsevier B.V. All rights reserved. doi: /j.eururo

2 778 european urology 52 (2007) Introduction 2. Materials and methods The incidence of renal cell carcinoma (RCC) is increasing, and approximately 36,160 new cases were discovered in the United States in 2005 [1]. This increase has been documented for all stages of renal cancer, but widespread use of cross-sectional imaging has led to the detection of a greater number of small, early tumours [2,3]. Such tumours are often associated with a better prognosis and, although radical nephrectomy still remains the standard of care for localised RCC, nephron-sparing approaches are increasingly used. It has been well established that open partial nephrectomy has equivalent 5-yr survival rates with radical nephrectomy in properly selected patients [4]. However, while preserving renal mass, open partial nephrectomy does not decrease hospital stay or surgical procedure time. This shortcoming has led to the development of new minimally invasive modalities with different success rates, including laparoscopic partial nephrectomy, cryotherapy, ablation with high-intensity focused ultrasound, ethanol ablation, and radiofrequency ablation (RFA) [5 7]. RFA effectively destroys living tissue by inducing excitation of ions, generating frictional forces and heat [8]. RFA has been effective in destroying solid tumours such as unresectable neoplasms in the liver, lung, bone, brain, breast, prostate, and kidney [9]. It is a well-tolerated, minimally invasive procedure, which decreases morbidity, length of hospitalization, and cost, and helps to preserve normal parenchyma [9]. Percutaneous ethanol injection (PEI) has been extensively used in the past to ablate liver tumours, especially small hepatocellular carcinomas [10]. It has direct cytotoxic effect and causes marked cell necrosis through membrane lysis and protein denaturation [10]. Several experimental [11,12] and clinical [13,14] studies in hepatic tumours have shown that the combined use of PEI and RFA has resulted in larger areas of coagulation than each method of treatment alone. There are no published studies on the use of PEI in RCC. Although there are several reports of use of RFA in treatment of RCC with promising preliminary and midterm results [15 23], to the best of our knowledge there is no study to evaluate the combined percutaneous ethanol and radiofrequency ablation of renal tumours. The purpose of our study was to evaluate the safety and efficacy of this technique, and to present our midterm results in 27 consecutive patients with 28 renal neoplasms with mean followup of 18.6 mo Patients This prospective study was undertaken since February 2002 with formal approval from the local institutional review board. Written informed consent was obtained from all patients. Twenty-seven consecutive patients (22 men, 5 women; mean age: yr; median: 70; range: 39 84) with 28 solid renal masses were referred for percutaneous imaging guided radiofrequency ablation. All patients had been examined by a urologic surgeon and, after discussion at a multidisciplinary meeting, were considered nonsurgical candidates because of previous nephrectomy (9 patients) or comorbid conditions (16 patients), or because some had refused surgery (2 patients). None had radiologic or laboratory evidence of metastatic disease or tumour infiltration beyond the kidney. The level of haemoglobin, electrolytes, blood-urea-nitrogen, and creatinine, as well as white blood cell and platelet counts were measured, and coagulation studies were carried out in all patients Tumour characteristics Of the 27 patients, 17 had pathologic confirmation of the diagnosis from imaging-guided biopsy performed at the time of the radiofrequency ablation. Histologic examination demonstrated 14 RCCs, 2 oncocytomas, and 1 angiomyolipoma. One patient had a biopsy, but the tissue was not adequate for a diagnosis. The remaining 9 patients (1 of whom had 2 tumours) did not have pathologic confirmation of the renal lesion, and the diagnosis was made on the basis of CT criteria, including average density more than 20 HU, enhancement more than 20 HU after contrast administration, absence of fat, and contour deformation [24]. The mean size of treated tumours was 2.87 cm (range: cm). In keeping with the classification suggested by Gervais et al in 2000 [18], we treated 16 exophytic tumours, 7 parenchymal, and 5 mixed tumours (Table 1). Five patients had previously undergone a total nephrectomy. Two had had a partial nephrectomy and were treated with ablation for tumours in the contralateral kidney. One patient had a partial nephrectomy and was treated for residual tumour; another patient had a left total and a right partial nephrectomy and was treated for a residual tumour in the right kidney. One patient was treated on his only functioning kidney, and one tumour was in a horseshoe kidney Technique The procedure was performed under ultrasound (Sonoline; Siemens, Erlangen, Germany) (n = 6) or CT (Brilliance 40- channel; Phillips) guidance (n = 26). In a single case (described below) magnetic resonance imaging and fluoroscopic guidance were used. We stopped using ultrasound guidance after the ninth procedure, and all subsequent procedures were performed under CT guidance, because the echogenic area resulting from the injection of ethanol made it more difficult to achieve accurate placement of the RFA electrode. One lesion,

3 european urology 52 (2007) Table 1 Patient data Patient Sex Age Histology Diameter (cm) Exophytic Parenchymal Mixed 1 M 78 No M 75 RCC F 66 Angiomyolipoma * M 62 No M 61 No F 84 No M 61 RCC M 79 No F 65 No M 72 Nondiagnostic F 62 Oncocytoma M 69 RCC M 82 RCC M 76 No M 70 RCC M 77 No M 59 No M 69 RCC M 81 RCC M 73 RCC F 82 RCC M 72 Oncocytoma M 65 RCC M 45 RCC M 39 RCC M 67 RCC M 76 RCC 2 + RCC = renal cell carcinoma. * Patient 4 had two tumors on the same kidney. in a patient with impaired renal function, was not clearly visualised on either unenhanced CT or ultrasound; the tumour was marked with a metallic coil placed under MRI guidance. The coil was subsequently used to target the tumour under fluoroscopic guidance [25]. The procedures were performed under local anesthesia (20 ml lignocaine 1%) conscious sedation (intravenous midazolam 4 10 mg), and analgesia (intravenous fentanyl mg) in 28 sessions and under general anesthesia in 5. Premedication with 100 mg of meperidine intramuscular injection preceded by 10 mg of maxolon intravenous was given 1 h before the procedure. No patient received prophylactic antibiotics before or after the procedure. In 18 cases a percutaneous biopsy with an 18-G cutting needle was performed, followed by insertion of a 22-G spinal needle. In the remaining cases, the initial part of the procedure was the placement of the 22-G needle. When using CT, rapid two-dimensional reconstructions along the plane of insertion of the needle were obtained to determine accurately the position of the tip of the needle in relation to the tumour and adjacent tissues; then the absolute ethanol was injected (mean volume: 1.7 ml; range: 0.5 3) (Fig. 1a) after aspiration to ensure that the tip of the needle was not in a blood vessel or in the renal collecting system. Subsequently the same technique was used for insertion of the radiofrequency electrode (Fig. 1b). Radiofrequency ablation started 4 5 min after the ethanol was injected. We used a Cool-Tip electrode (Radionics, Burlington, MA, USA) in all cases. A cluster Cool-Tip electrode was used in 13 sessions and a single Cool-Tip electrode in 20 sessions (Table 2). At conclusion of the procedure, the maximum tip temperature and the 1-min cool-down temperature were noted as indicators of the heat deposition. In 16 sessions we performed only one ablation; in 15 sessions we had to reposition the probe and do two overlapping ablations. In two other patients we had to do three overlapping ablations. In one patient who had two tumours, two overlapping ablations were carried out for the larger one, followed by repuncture and a single ablation for the smaller tumour. After the procedure, the blood pressure and heart rate were monitored for 24 h. Two patients were treated as day cases and were monitored for only 6 h Follow-up Patients were followed up with contrast-enhanced CT, except one with preexisting severely deranged renal function, who was followed-up with contrast-enhanced MRI. A three-phase CT (unenhanced, arterial, nephrographic) was performed the day after the procedure to check for complications and to assess the technical success of the procedure. If there was any suspicion of residual disease, the next CT was performed in 1 mo; otherwise it was performed at 3 mo and then at 6-mo

4 780 european urology 52 (2007) Fig. 1 (a) Computed tomography (CT) scan section with patient in prone position and a 22-G Chiba needle in situ within the large renal tumor (orange arrow), prior to ethanol injection. (b) Following the injection of ethanol, the Cool-Tip triple electrode is in position. (c) Contrast-enhanced CT scan with patient in supine position showing complete coagulation necrosis of the tumor (white arrow). intervals. Routine haematologic and biochemical studies were obtained. The treated tumours were assessed for residual enhancement and size changes from all follow-up studies as described by Gervais et al [18]. Residual disease was defined as persistent enhancement of >10 HU in any part of the tumour after ablation, on the 1-mo or 3-mo follow-up study. Recurrent disease was defined as new tumour enhancement after at least one imaging study had demonstrated complete absence of enhancement [26] Statistical analysis Statistical Software for the Social Sciences, version 12.0 for Windows (SPSS Inc, Chicago, IL, USA) was used for conducting the statistical analysis. Independent sample t test and paired sample t test were used, whenever appropriate, to assess whether the means of two variables differed statistically. Fisher exact test and chi-square test were used, whenever appropriate, to compare the categoric variables. Pearson correlation coefficients were calculated to assess the correlation of specific continuous variables. 3. Results 3.1. Early results A total of 53 ablations were performed during 33 ablation sessions (mean: 1.6 per session). Twentyseven of the 28 tumours were completely ablated at either one (21 tumours) or two visits (6 tumours). One 81-year-old patient with several comorbid conditions and a large mixed type tumour showed a small area of residual disease in the follow-up CT. He decided he did not want further treatment and preferred to have the tumour followed up with

5 european urology 52 (2007) Table 2 Treatment parameters Patient Ethanol injection (ml) RF time (min) Follow-up (mo) Guidance No. of treatments Electrode (Cool-Tip) CT 1 Single US 1 Triple US 1 Single 4 * 0.5/0.5 15/6 32 US/US 1 Single CT 1 Single CT 2 Triple 3 12 CT Single US 2 Triple 1 30 CT Triple US 2 Triple 1 12 CT Triple US 2 Triple 2 12 CT Single CT 1 Single CT 1 Single CT 1 Single CT 2 Triple 2 9 CT Triple CT 2 Triple 3 24 CT Single MRI X-ray 1 Single CT 1 Single CT 1 Single CT 1 Triple CT 1 Triple CT 1 Triple CT 1 Single CT 1 Single CT 1 Single CT 1 Single CT 1 Single CT 1 Single CT 1 Single RF = radiofrequency; CT = computed tomography; US = ultrasound; MRI = magnetic resonance imaging. * Patient 4 had two tumors on the same kidney. imaging. The mean treatment time with radiofrequency was 16.8 min (range: 7 36). Postprocedure complications included one subcapsular haematoma, which was managed conservatively with no need for transfusion, and transient loin pain in two patients, which resolved spontaneously in a few days. The pain was probably caused by transient trauma to the intercostal or lumbar nerves in the affected dermatome, as previously described [15]. There were no complications related to the ethanol injection Late results Imaging follow-up ranged from 3 to 56 mo (mean: 18.6). After the initial procedure, there was complete coagulation with lack of enhancement on CT (<10 HU) in 20 patients at 1 mo (Fig. 2a and b). Seven patients showed residual enhancing tumour at follow-up CT and six of them were retreated. In all of these patients, complete coagulation was achieved during the second treatment session. No lesion needed more than two treatment sessions for complete coagulation. Tumours requiring a second session were significantly larger (mean: 4.1 cm; maximum diameter range: 3.1 6) than masses that did not require a second treatment (mean: cm; range: ; p < independent sample t test). All tumours smaller than 3.1 cm were completely ablated after a single treatment session. Tumour size was the only significant prognostic factor determining initial technical success. There was no significant association between the location of the tumour (exophytic, parencymal, or mixed) and initial technical success ( p = chi-square test). The ablation zone was significantly increased in the first follow-up CT compared with the diameter of the tumour. (From a mean 2.87 cm diameter to 3.8 cm; 37.8% increase). During a mean follow-up of 18 mo, the ablation zone involuted gradually to a mean diameter of 2.6 cm. The primary technical success rate [27] of combined ethanol and RF ablation of renal masses,

6 782 european urology 52 (2007) Fig. 2 (a) Exophytic renal tumor, midpole left kidney (white arrow). (b) Postablation scan demonstrating coagulation necrosis with no evidence of residual tumor. omitting the three patients with benign disease, was 72% (18 of 25 tumours), and the secondary technical success rate, including tumours that were ablated after identification of residual tumour, was 96% (24 of 25 tumours). There were no significant changes in the creatinine level in any of the patients following ablation, including the nine patients (33%) who had undergone previous radical or partial nephrectomy. The mean baseline plasma creatinine level was mmol/l. The mean postablation creatinine level at a mean follow-up of 18.6 mo was mmol/l ( p = >0.05 paired sample t test). During the follow-up period (3 56 mo; mean: 18.6), one patient died at the age of 82 yr from causes unrelated to his renal disease after 46 mo of followup. All the other patients are well with no evidence of tumour recurrence or metastatic disease. 4. Discussion Although radical nephrectomy is considered the standard treatment for localised RCC, the increasing incidence of small incidentally detected tumours [24] has generated interest in nephron-sparing surgical techniques. Open partial nephrectomy has been shown to be as effective as radical nephrectomy in appropriately selected patients [4]. However, less invasive methods of treatment, such us cryotherapy, high-intensity focused ultrasound coagulation, ethanol ablation, and RFA [5 7,15 23,26], are often preferred by patients. The use of thermal ablation for the treatment of human renal neoplasms was first described in 1997 by Zlotta et al [17]. Since then several reports of RCC treated with percutaneous RFA have been published. Gervais et al [22] in the largest published series, with 100 tumours and mean follow-up of 2.3 yr, concluded that RFA of RCC is an effective method of treatment for exophytic RCC tumours up to 5.0 cm in diameter (all 67 exophytic tumours were completely ablated). Tumours larger than 3.0 cm extending to the renal sinus are more difficult to treat but can be ablated successfully. Zagoria el at [21] reported that effective ablation was achieved in 20 of 22 patients (91%) in a single session of therapy. However, they found that tumours larger than 3 cm in diameter were difficult to eradicate at the first attempt, usually requiring a second RFA session. Similarly, Mayo-Smith et al [15] reported that the average size of tumour requiring a second RFA was 3.5 cm. Gervais et al [22] concluded that tumours larger than 5.8 cm are unlikely to be treated completely with radiofrequency ablation alone. In this study, tumour size was the only significant prognostic factor determining initial clinical success. Tumour location (exophytic or central) had no influence on tumour eradication. This finding can be partly attributed to the small number of tumours with a central component in this study (5 of 28, 17%).

7 european urology 52 (2007) The number of ablations per session is less than any other study in the literature. Of course, this study was not designed as a formal comparison between RFA and ethanol plus RFA ablation methods. Nevertheless, it is possible that there is a synergistic effect of alcohol injections and radiofrequency ablation. The effect of combining RFA with absolute ethanol injection has been studied in different experimental models including rabbit livers by Lee et al [11], a rat breast tumour model by Goldberg et al [12], and porcine kidneys by Rehman et al [28], who all demonstrated that this form of combined therapy increases the area of coagulation compared with either single therapy, with no significant difference in complication rates. There are also clinical studies [13,14] demonstrating the effectiveness of combined RFA and ethanol injection in large liver tumours. The mechanism of the synergistic effect of these two ablation methods is not yet clear. A possible explanation is the elimination of perfusion-mediated tissue cooling by ethanol induced thrombosis of small vessels in the treated tissue prior to RFA heating [11 14,29,30]. An alternative mechanism is improved thermal conduction through previously coagulated tissue by ethanol [11]. It is possible that ethanol is warmed by subsequent RFA and that it may spread through areas that survived initial heating, such as perivascular cuffs [11]. To avoid the unpredictable distribution of the alcohol in renal tissue [28], we injected ethanol into the centre of the tumour and not at its periphery, to secure maximum distribution within the neoplasm rather than in normal renal parenchyma. Only three minor complications were noted; none of them related to ethanol injection. The small amount of ethanol injected (1.7 ml per session) and careful aspiration to avoid injecting it into a vessel or the collecting system have probably contributed the low complication rate. Moreover all three complications happened under ultrasound guidance when the hyperechogenic area caused by the injection of ethanol led to difficulties in subsequent precise placement of the RFA electrode. Multidetector CT guidance with rapid two-dimensional reconstructions improved the accuracy of electrode placement. Furthermore the synergetic effect of alcohol injection and RFA led to fewer ablations per session, requiring fewer relocations of the electrode, which may have contributed to the low rate of complications. This study has limitations, including its small size, lack of a control group, variable length of follow-up, and lack of pathologic confirmation in some of the patients. Nevertheless, we believe that our findings suggest that the combined use of PEI and RFA is a safe and effective alternative treatment for patients with renal tumours. 5. Conclusions Percutaneous, image-guided, combined use of RF and ethanol ablation is a safe, easily applicable, minimally invasive, and very effective treatment for patients with small renal tumours. Intermediate results are promising, but further follow-up is needed to assess the long-term efficiency of the method. Prospective controlled randomised trials comparing RFA (with and without ethanol injection) with nephron-sparing surgery should be performed to determine the role of ablative techniques in the treatment of small exophytic renal tumours. Conflicts of interest No disclosure from any of the authors. References [1] American Cancer Society. Cancer Facts and Figures Available from: CancerFacts&FiguresTM.pdf. Accessed 1 Aug [2] Jayson M, Sander H. Increased incidence of serendipitously discovered renal cell carcinoma. Urology 1998;51: [3] Homma Y, Kawabe K, Kitamura T, et al. Increased incidental detection and reduced mortality in renal cancer: recent retrospective analysis at eight institutions. Int J Urol 1995;2: [4] Fergany AF, Hafez KS, Novick AC. Long term results of nephron sparing surgery for localized renal cell carcinoma: 10-year follow up. J Urol 2000;163: [5] Jeschke K, Peschel R, Wakonig J, et al. Laparoscopic nephron-sparing surgery for renal tumors. Urology 2001;58: [6] Gill IS, Remer EM, Hasan WA, et al. Renal cryoablation: outcome at 3 years. J Urol 2005;173: [7] Aron M, Gill IS. Minimally invasive nephron-sparing surgery (MINSS) for renal tumours, part II: probe ablative therapy. Eur Urol 2007;51: [8] Goldberg SN, Gazelle GS, Mueller PR. Thermal ablation therapy for focal malignancy: a unified approach to underlying principles, techniques, and diagnostic imaging gyidance. AJR 2000;174: [9] Gazelle GS, Goldberg SN, Solbiati L, et al. Tumor ablation with radio-frequency energy. Radiology 2000;217: [10] Shiina S, Tagawa K, Niwa Y, et al. Percutaneous ethanol injection therapy for hepatocellular carcinoma: results in 146 patients. AJR 1993;160:

8 784 european urology 52 (2007) [11] Lee JM, Lee YH, Kim YK, et al. Combined therapy of radiofrequency ablation and ethanol injection of rabbit liver: an in vivo feasibility study. Cardiovasc Intervent Radiol 2004;27: [12] Goldberg SN, Kruskal JB, Oliver BS, et al. Percutaneous tumor ablation: increased coagulation by combining radio-frequency ablation and ethanol instillation in a rat breast tumor model. Radiology 2000;217: [13] Kurokohchi K, Watanabe S, Masaki T, et al. Combined use of percutaneous ethanol injection and radiofrequency ablation for the effective treatment of hepatocellular carcinoma. Int J Oncol 2002;21: [14] Shankar S, vansonneberg E, Morrison PR, et al. Combined radiofrequency and alcohol injection for percutaneous hepatic tumor ablation. AJR 2004;183: [15] Mayo-Smith WW, Dupuy DE, Parikh PM, et al. Imagingguided percutaneous radiofrequency ablation of solid renal masses: techniques and outcomes of 38 treatment sessions in 32 consecutive patients. AJR 2003;180: [16] Pavlovich CP, Walther MM, Choyke PL, et al. Percutaneous radio frequency ablation of small renal tumours: initial results. J Urol 2002;167:10 5. [17] Zlotta AR, Wildschutz T, Raviv G, et al. Radiofrequency interstitial tumor ablation (RITA) is a possible new modality for treatment of renal cancer: ex vivo and in vivo experience. J Endourol 1997;11: [18] Gervais DA, McGovern FJ, Wood BJ, et al. Radiofrequency ablation of renal cell carcinoma: early clinical experience. Radiology 2000;217: [19] McGovern FJ, Wood BJ, Goldberg N, et al. Radiofrequency ablation of renal cell carcinoma via image guided needle electrodes. J Urol 1999;161: [20] Gervais DA, McGorvern FJ, Arellano RS, et al. Renal cell carcinoma: clinical experience and technical success with radio-frequency ablation of 42 tumors. Radiology 2003;226: [21] Zagoria RJ, Hawkins AD, Clark PE, et al. Percutaneous CT-guided radiofrequency ablation of renal neoplasms: factors influencing success. AJR 2004;183: [22] Gervais DA, McGovern FJ, Arellano RS, et al. Radiofrequency ablation of renal cell carcinoma, part I: indications, results and role in patient management over a 6-year period and ablation of 100 tumors. AJR 2005;185: [23] Hwang JJ, Walters MM, Pautler SE, et al. Radiofrequency ablation of small renal tumors: Intermediate results. J Urol 2004;171: [24] Zagoria RJ. Imaging of small renal masses: a medical success story. AJR 2000;175: [25] Adam A, Hatzidakis A, Hamady M, et al. Percutaneous coil placement prior to radiofrequency ablation of poorly visible hepatic tumours. Eur Radiol 2004;14: [26] Mouraviev V, Joniau S, Van Poppel H, et al. Current status of minimally invasive ablative techniques in the treatment of small renal tumours. Eur Urol 2007;51: [27] Goldberg SN, Grassi CJ, Cardella JF, et al. Image guided tumour ablation: standardization of terminology and reporting criteria. J Vasc Interv Radiol 2005;16: [28] Rehman J, Lahman J, Lee D, et al. Needle based ablation of renal parenchyma using microwave, cryoablation, impedance- and temperature-based monopolar and bipolar radiofrequency, and liquid and gel chemoablation: laboratory studies and review of the literature. J Endourol 2004;18: [29] Festi D, Monti F, Casanova S, et al. Morphological and biochemical effects of intrahepatic alcohol injection in the rabbit. J Gastroenterol Hepatol 1990;5: [30] Rossi S, Garbagnati F, Lencioni R, et al. Percutaneous radiofrequency thermal ablation of nonresectable hepatocellular carcinoma after occlusion of tumor blood supply. Radiology 2000;217: