Introduction 1 1 3 4 1 3 5 Materials and methods 1 Table 1 Bone metastasis classification, with regard to the primary malignant lesion and the site of the skeleton involved Site of primary neoplasm origin n Site of metastasis n Colon 13 Pelvis 15 Breast 7 Ribs 6 Prostate 2 Sacrum 5 Lung 4 Femur 3 Renal 2 Spine 3 Thyroid 1 Scapula 2 Skin melanoma 1 Tibia  1 Total 30 Total 34 2 Table 2 Criteria for the selection of patients to undergo radiofrequency ablation Patient selection criteria Brief Pain Inventory (BPI) score above 4 Lesions not responding to chemotherapy and/or radiation therapy (completion of therapy at least 3 weeks before the radiofrequency ablation session) Chemotherapy-associated complications that halted this treatment Lesions adjacent to structures sensitive to irradiation Patients with life expectancy greater than 2 months who were not eligible for surgical treatment  Patients who preferred this treatment over the other alternatives Physical examination was performed by the oncologist and in collaboration with the radiologist performing the ablation. Pain was assessed with the Brief Pain Inventory (BPI) The use of analgesics was recorded the day before the procedure. Before the procedure blood cell count and blood clotting analysis were performed. Minimal requirements were: platelet (PLT) count >50,000/ml (normal range, 150,000–350,000/ml); prothrombin time (PT), international normalised ratio (INR) <1.3 (normal range, 0.8–1.2); and partial thromboplastin time (PTT) <34 s (normal range, 25–34 s). The procedure was performed under conscious sedation (administration of 3 mg of bromazepam PO and 50 mg of pethidine hydrochloric acid intramurally, 45 min prior to the procedure) and was trained in regular breathing and breath-holding (suspended respiration) before the procedure. He/she was placed in the appropriate position (prone, supine, or lateral, depending on the site of the lesion) and a scan of the desired area with a 5-mm slice thickness was performed, using a Picker 5000® (Philips Medical Systems, Amsterdam, The Netherlands). At least one of two staff radiologists with extensive experience in biopsies and tumour ablations was involved in all ablations. The lesion’s exact location and depth, in relation to the overlying skin, was determined on CT. The skin was then prepared with povidone iodine (10%) solution. Local anaesthesia (15 ml of 2% lidocaine hydrochloride solution) was administered. Radiofrequency ablation was performed with a RITA Model 1500® electrosurgical generator (RITA Medical Systems, Mountain View, CA, USA) and a seven-array, 2- to 3-cm multitined electrode for lesions smaller than 3 cm (20 out of 34), or a nine-array multitined electrode for larger lesions (14 out of 34). The electrode tip was inserted to approximately 1 cm from the centre of the target. The electrodes were then deployed slowly, taking into account the need to ablate the lesion–bone interface. The net ablation time was ~15 min at an energy level of 90–110 W, with the target goal temperature set to 80–110°C. During the procedure the infusion port of the electrode was flushed with a 2% lidocaine hydrochloride solution in order to reduce patient discomfort and to decrease tissue overheating and vaporisation. The number of electrode placements, individual (per electrode) and total ablation times, the total energy delivered to the target and the lesion temperatures achieved were recorded. After each session a dual-phase spiral CT examination with intravenous contrast medium was performed in order to assess response, as confirmed by low lesion attenuation values and lack of contrast enhancement. Patients were hospitalised and observed for 24 h. Analgesics were administered if required. Before patient discharge the pain was re-evaluated with the BPI score. Post-ablation assessment was completed with telephone interview after one, four and eight week. The BPI score and the use of analgesics were recorded again. Results 3 Table 3 Lesion characteristics andtreatment Number of lesions Size of lesion (cm) Number of electrode placements Time of radiofrequency energy deposition (min) 15 <3 1 5 5 3 1 5 7 4 2 7 3 5 2 8 1 6 3 9 1 8 4 10 1 9 4 10 1 14 5 15 Total 34    1 p t 2 3 4 Fig. 1 a b c  Fig. 2 a b  Fig. 3 a b  Fig. 4 a  Prior to RF ablation 27 out of 30 patients received opioids or an opioid/NSAID combination. The remaining 3 patients received NSAIDs. One week after treatment 5 out of 30 patients were treated with a combination of NSAID/low-dose opioids. Six out of thirty used NSAIDs. After 4 and 8 weeks only 3 out of 30 patients received any medication NSAIDs. One patient died during the 8-week follow-up for reasons not related to RF ablation. Discussion In patients with cancer, pain originating from bone metastases can be difficult to treat. A number of treatment options are available, including NSAIDs, opioids, and adjuvant drugs medications, radiation therapy, chemotherapy, hormonal therapy, radiopharmaceutical therapy, surgery and vertebroplasty. Medication represents the first line of treatment. NSAIDs and adjuvant drugs represent basic medication, potentially followed by NSAID/low-dose opioid combinations, and finally increasing the opioid dose. 6 6 7 8 9 10 11 12 In our patients, we observed a considerable reduction of pain and improvement of the quality of life, as measured by the BPI score. 9 10 The reduction of the procedural time is limited by both the time needed to achieve the optimal target temperature, and the size of the lesion, because more than one deployment is required in larger bone metastases. Although a few patients reported mild discomfort during the ablation, none of the sessions was forced to stop owing to considerable patient distress. There were none of the possible adverse effects, including infection, haemorrhage, neurological complications, skin burns, or the so-called post-ablation syndrome (low-grade fevers ≤100°F [37.8°C]), myalgias, and malaise for up to 1 week after the procedure). Pain reduction was fast and occurred within the first 24 h for some and during the first week in the majority of the patients. This appears to be a fairly well-tolerated procedure and the combination of conscious sedation and local anaesthesia is adequate for its needs. 13 In our study, there were no lesions threatening the stability of the spine. 14 15 16 17 The follow-up period for this study was 8 weeks, a period that we believed was sufficient to demonstrate that RFA provides effective palliation. There is, however, a need for randomised prospective studies, to evaluate the method and to compare it with other treatment modalities, such as radiation therapy. Continued follow-up is warranted to determine the long-term efficacy of this interventional approach. In conclusion, imaged-guided RF ablation of painful bone metastases is promising. It appears to be effective, safe and well tolerated by patients.