| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology, Mater Misericordiae University Hospital, Eccles St.,
Dublin 7, Ireland.
2 Cappagh National Orthopaedic Hospital, Finglas, Dublin 11, Ireland.
Received January 28, 2005;
accepted after revision August 10, 2005.
Address correspondence to C. P. Cantwell
(ccanty{at}gofree.indigo.ie).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Radiofrequency ablation was performed in 10 patients with a clinical and radiologic diagnosis of osteoid osteoma. A cooled radiofrequency probe was inserted in the nidus. Twelve minutes of radiofrequency energy was applied from a 200-W radiofrequency generator in an impedance-control setting. MRI with multiplanar turbo spin-echo T1-weighted and STIR sequences was performed at 1, 7, and 28 days after the procedure in seven patients. The three remaining patients had follow-up imaging at 28 days only. The images were reviewed by two radiologists who categorized the imaging features and measured the marrow zone of signal alteration when visible. The size of the zone of marrow signal change produced by the radiofrequency technique was compared with published data for radiofrequency ablation with manual-control protocols.
RESULTS. A 1-mm band of homogeneous altered marrow signal distributed symmetrically parallel to the entire probe tract was seen earliest, at 1 day, in the femoral neck lesion treated with the 2-cm probe. The band was low signal on the T1 sequence and high signal on the STIR sequence, and the diameter of the zone was 27 mm. By 7 days, five of the seven treated bones showed a band of marrow signal alteration. By 28 days, all 10 treated bones had a band of marrow signal alteration. The interband distance at 90° to the probe measured on STIR images at 28 days was a mean of 20.9 mm (confidence interval, 16.1-25.7 mm [p < 0.05]; range ± measurement error, 10.5-35 ± 1.64 mm) with a 1-cm probe and 30.5 mm (measurement error, ± 0.78 mm) on T1 images without contrast material when a 2-cm exposed-tip probe was used. Higher-output generators with impedance-control software and internally cooled radiofrequency probes with longer exposed tips produce larger zones of marrow signal change than expected with manual-control protocols.
CONCLUSION. MRI allows detection of temporal marrow signal change after radiofrequency ablation. The marrow signal change with a high-energy delivery protocol is larger than manual-control protocols.
Keywords: ablation • bone • bone marrow • MRI • osteoma • radiofrequency ablation
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The extent and adequacy of therapy are difficult to monitor because there is no reported change in marrow density on CT. Reactive bone sclerosis may occur at the periphery of the treated lesion [2]. Follow-up CT at 6 months after radiofrequency may also show complete or partial ossification of the nidus in 73% of cases [3]. Radiographs after therapy show little evidence of local bone healing [4].
MRI is a noninvasive method of monitoring the extent of marrow change after radiofrequency ablation. The signal changes are accurate in animal models for determining the zone of thermal ablation at 7 and 14 days after therapy [2, 5].
With the rapid increase in the availability and use of modern radiofrequency equipment, there is limited information about the physical effect of the therapy in human bone.
We applied the manufacturer's recommended therapy protocol for bone using cooled probes and impedance-control energy delivery because we believed that a larger ablation zone would be produced and that the larger ablation zone would lead to a higher clinical success rate. We followed up on the patients' marrow signal changes over time in an effort to determine the characteristics and extent of signal alteration. The characteristics and accuracy of marrow signal changes in the zone of marrow ablation have been described in a pig bone model with histologic correlation [2, 5]. We assumed if the marrow signal changes were the same in human and pig marrow that the changes would accurately reflect the extent of therapy in humans.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
There were five male and five female patients (mean age, 21.8 years; age range, 11-30 years). Five lesions were in the femur, two of which were in the femoral neck, and five lesions were in the tibia. Treated lesions were 3-12 mm in diameter (mean, 7 mm).
One patient had recurrent symptoms after surgical resection with bone grafting 18 months before radiofrequency ablation. A second patient had a delay in the diagnosis of osteoid osteoma and had two bone marrow biopsies while under general anesthesia with an assumed diagnosis of lymphoma.
Nine patients had day case procedures. The youngest patient was admitted overnight because of concerns about adequate analgesia.
Procedures
Informed consent was obtained in all cases. All procedures were performed
by the authors. The procedure was performed with the patient under general
anesthesia without muscle relaxation. The patient was positioned supine on the
CT table (Somatom, Volume Zoom 4, Siemens Medical Solutions). Two large
grounding pads were applied to the patient's skin and connected to the
radiofrequency generator.
An unenhanced axial CT image (1.5-mm slice thickness, 1:1 pitch, 120 kV, and 90 mAs) of the lesion was obtained, and a puncture site was marked on the skin using an indelible marker. The skin was prepared and draped. A 20-gauge spinal needle was inserted into the approximate path of the trocar. Another quick-check CT image (3-mm slice thickness, 120 kV, and 80 mAs) was obtained to confirm the correct level of puncture and trajectory.
Through a small skin incision, an 11-gauge Jamshidi needle was used to reach the cortex abutting the nidus. When the trocar was removed, it provided a protected tract for the nidus to be drilled with a 4.5-French bone biopsy needle (Ostycut, Angiomed) or a 1.6-mm nonthreaded K-wire (MicroAire). The drill was then removed, and the 4.5-French (1.5 mm) radiofrequency probe (Cool-tip, Valleylab) was inserted through the outer cannula into the lesion and guided through the defect created by the drill. A 2-cm exposed-tip probe was used in our first case, a femoral neck lesion. The MR images obtained 1 day after the procedure showed extensive marrow change, and all subsequent treatments used a 1-cm exposed-tip probe. The cannula was withdrawn and secured. A CT scan was obtained to confirm the position of the probe tip with the proximal extent of the exposed tip at the near extent of the nidus by measuring the exposed length from the probe tip. The radiofrequency probe was circulated with saline cooled with ice to near 0°C by a mechanical pump at a rate of 80 mL/min. A 200-W maximum output radiofrequency generator (RFG-3C, Radionics) was used. Twelve minutes of radiofrequency energy was delivered in the impedance-control mode.
The electrode was withdrawn from the lesion at the end of treatment, and the skin incision was covered with a self-adhesive dressing. The patient was allowed to bear full weight immediately. The patient was advised to avoid participating in a contact sport empirically for 6 weeks. Patients were discharged with a prescription for nonsteroidal antiinflammatory medication to be taken orally as required for 24-72 hr.
The procedure was technically successful if the probe was placed in the osteoid osteoma and 12 min of radiofrequency therapy was delivered.
Imaging Analysis
Imaging was performed with a 1.5-T MR scanner (Intera, Philips Medical
Systems). MRI was performed at 1, 7, and 28 days after the procedure. Turbo
spin-echo imaging and STIR imaging of the ablation zone were performed in the
axial plane in the first three patients only and in the coronal or sagittal
oblique plane in all patients. Table
1 lists the imaging parameters.
|
Two postfellowship radiologists interpreted the images from a parallel workstation. The interpretations were performed by the two reviewers independently and differences in interpretations were resolved by consensus. The reviewers were blinded to the time interval after therapy.
The cortical drill hole was identified and its location recorded. The soft-tissue, periosteal, cortical, osteoid osteoma, and marrow signal changes were noted at each of the time intervals. The signal was also recorded as low, high, or isointense relative to normal periosteum, cortex, and marrow. The marrow signal was recorded as heterogeneous or homogeneous. The marrow signal was described at the probe placement tract and adjacent to the tract in the zone of ablation.
If a well-defined band of high or low signal was identified, the signal intensity of the band was noted and the distance between the two bands from the point farthest from the probe placement tract perpendicular to the radiofrequency electrode trajectory was measured using a caliper on both STIR and T1-weighted images (Fig. 1).
|
|
|
Eight of the patients in our study group participated in a parallel clinical investigation of postprocedure pain, function, and satisfaction. The size of the interband distance in the marrow at 28 days and the postprocedure pain score, 24-hr pain score, and interval to being pain-free were tested for correlation.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CT performed before the procedure in one patient identified marked endosteal new bone formation and MRI showed low marrow signal on all imaging sequences (Figs. 2A and 2B). Another patient had previously undergone surgical resection and bone grafting and had associated marrow signal loss at the site of the resection. The marrow signal of one of the femoral neck lesions was increased on the STIR images secondary to a combination of two recent open bone marrow biopsies and the local effects of the osteoid osteoma.
|
|
|
|
|
There was minimal focal high signal in the soft tissue and periosteum peripheral to the drill hole on the STIR images in all patients. This was reflected by focal low signal on T1 images of the soft tissues. In one patient, extensive focal high signal on the STIR images was identified in the muscle and periosteum secondary to deep muscle ablation. The near cortex and osteoid osteoma were drilled, but the cortex adjacent to the medullary cavity was not completely drilled, resulting in protrusion of the exposed tip in the muscle in this patient.
MRI showed heterogeneous high signal on the STIR images and low signal on the T1 images in the 2-mm marrow probe placement tract in all patients deep in relation to the cortical drill hole. The bone marrow adjacent to the probe placement tract in five of the seven patients showed heterogeneous low signal on T1 images and heterogeneous local high signal on STIR images (Figs. 3A, 3B, 3C, 3D, and 3E). The two patients with extensive cortical thickening had no detectable change in marrow signal adjacent to the probe placement tract.
A 1-mm band of homogeneous low signal on the T1 sequence and homogeneous high signal on the STIR sequence of 27 mm in diameter distributed symmetrically parallel to the entire probe placement tract was seen only in the femoral neck lesion treated with the 2-cm probe at 1 day.
7 Days
Seven patients underwent imaging 7 days after the procedure. The cortical
drill hole could be identified in all patients on STIR images as a tract of
high signal and some T1 images as a tract of intermediate signal. Periosteal
high signal adjacent to the treated cortex remained visible on STIR images. No
change was identified on T1-weighted imaging of the periosteum, cortex, or
osteoid osteoma. In the complicated case with soft-tissue ablation, the
ablation zone was more defined on STIR and T1 imaging by 7 days.
MRI again showed heterogeneous high signal on the STIR images and low signal on the T1 images in the 2-mm marrow probe placement tract in all patients. The marrow signal adjacent to the probe placement tract was heterogeneously isointense on T1 images and STIR images in five of the seven cases in comparison with untreated marrow. Five of the seven marrow lesions showed a low-signal band in the marrow on T1 images and a high-signal band on STIR images parallel and distal to the probe placement tract (Figs. 3A, 3B, 3C, 3D, and 3E).
The two patients with extensive cortical thickening had only a local heterogeneous increase in marrow signal adjacent to the probe placement tract on the STIR images. No low-signal band in the marrow on T1 or high-signal band on STIR images parallel to the probe placement tract could be seen.
28 Days
Ten patients underwent imaging 28 days after the procedure. The cortical
drill hole could be identified in all patients on STIR images as a tract of
high signal and on some T1 images as a tract of intermediate signal.
Periosteal high signal adjacent to the treated cortex remained visible on STIR
images. No change was identified on T1-weighted imaging of the periosteum,
cortex, or osteoid osteoma.
MRI again showed heterogeneous high signal on the STIR images in the 2-mm probe placement tract in all patients. The marrow signal adjacent to the probe placement tract was heterogeneously isointense on T1 images and STIR images in all the patients in comparison with untreated marrow.
|
|
|
The size of the zone of marrow signal change produced by the radiofrequency technique used for our study was larger when compared with published data for radiofrequency ablation with manual-control protocols (Table 3).
|
Symptoms were completely relieved in all patients. There were no recurrences during the follow-up period (mean, 13.4 months; range, 3-27 months).
There was no significant correlation between the size of the marrow signal alteration on T1 or STIR images and postoperative pain, pain at 1 day, or the interval to pain resolution in the eight patients for whom both sets of data were available.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
The band of signal alteration of the marrow parallel and distal to the probe placement tract reflects the changes seen in studies of radiofrequency ablation in bone of a pig model. Aschoff et al. [5] in a series of experiments showed the ability of temporal MRI using T1 after contrast administration, T2, and STIR imaging at 0.2 T performed at 2, 7, and 14 days to accurately predict the zone of ablation in a pig femur with histologic correlation. Nour et al. [2] performed a similar experiment using porcine lumbar vertebrae and the same protocol as Aschoff et al. except the generator used for their study had a maximum output of 100 W.
Both Nour et al. [2] and Aschoff et al. [5] found the band was high signal on the T1 images with contrast material, T2, and STIR images. Aschoff and colleagues found good correlation between the measurements of lesion size on T1 images obtained after contrast administration and STIR images. Nour et al. found T2 and STIR images obtained at 2 weeks after the procedure to be more accurate in the vertebral body than T1 images obtained with contrast material (mean absolute difference ± SD, 0.72 ± 0.83 mm).
The marrow signal of these young pigs was, in general, higher on T2 and STIR images than on T1 images because of the relatively large amounts of water in hemopoietic marrow. After therapy, the lesions showed low signal on T2, T1, and STIR images in the marrow adjacent to the probe placement tract. Over the study period, the size of the ablation zone increased from 7 to 14 days.
In our experience, the marrow adjacent to the probe placement tract maintained signal. On T1 imaging without contrast material the peripheral band of the lesion had a low signal. The low signal on T1 imaging was mirrored by a high-signal band at the same site on STIR imaging. From imaging our first patient at 28 days, we found some enhancement of the peripheral low-signal band on T1 images after the administration of gadolinium but a low-signal band was still seen and correlated with the unenhanced T1 and STIR images in determining the therapy zone. We think that gadolinium enhancement is not necessary at a clinical magnetic field strength of 1.5 T.
No histologic correlation of the human marrow change is available, and we assume that the marrow signal changes in our patients reflect the thermal necrosis identified at histology in the pig model up to this marrow signal band. A potential weakness of this study is the lack of availability of preprocedural MRI for comparison because the cases were diagnosed on the basis of CT and clinical parameters.
The mean maximum osteoid osteoma diameter in this study was 7 mm (range, 3-12 mm). Osteoid osteomas rarely measure more than 1.5 cm. Ideally, the treatment zone should not be significantly larger than the treated lesion. However, the range of distances between the bands measured on STIR images at 28 days was 10.5-35 mm (measurement error, ± 1.64 mm) with a 1-cm probe and 30.5 mm (measurement error, ± 0.78 mm) on T1 images without contrast material when a 2-cm exposed-tip probe was used. If this marrow signal change reflects the degree of cortical therapy, then this far exceeds the thermal lesion necessary to adequately treat the lesion, especially if a 2-cm probe is used.
The ability to predict the extent of therapy allows appropriate planning. The range of marrow ablation with this technique and a 1-cm probe was large—between 10.5 and 30.5 mm. The marrow therapy size reflects the energy delivered, which is determined by the degree of insulation. Impedance-control software may increase this variability in ablation size by automatically determining the maximum current delivery at probe insertion and altering the current delivery depending on circuit resistance, which varies with the degree of cortical thickening secondary to the osteoid osteoma.
The pig models of radiofrequency ablation used smaller-output generators (100-150 W) and 2-cm noncooled probes with manual-control current delivery at a probe temperature of 90°C for 10 min in a magnetic field. They found smaller lesions with less variability in therapy. Aschoff et al. [5] found a mean pathologic lesion diameter of 15.4 mm (range, 11-18 mm) in the metaphysis of the femur. Nour et al. [2] reported a mean pathologic lesion diameter of 11.0 ± 1.8 mm (range not stated) in the vertebrae.
Tillotson et al. [4] showed in 1989 the effects of use of a 5-mm exposed-tip radiofrequency probe in a dog model. Energy was delivered from a radiofrequency generator (RF-5, Radionics) under manual control keeping the lesion temperature at 80°C for up to 4 min and produced marrow lesions of between 9 and 13 mm in diameter. The radial zones of cortical, periosteal, and medullary ablation were the same unless the probe had been eccentrically placed. This work formed the scientific basis for the adequacy of therapy with radiofrequency ablation of osteoid osteoma.
Subsequently, radiofrequency energy was applied in a clinical series using manual control to keep the lesion at 80-90°C at the probe tip for 4-6 min. Radiofrequency ablation of osteoid osteoma with the 5-mm exposed-tip probe has a primary success rate of 76-95% [3, 6-12].
All the procedures were clinically successful in our series. In the parallel investigation of postoperative pain in patients treated with this radiofrequency energy delivery protocol, the mean (± SD) postprocedure pain score was rated at 6.9 ± 3.06 (95% confidence interval) on a numeric rating scale from zero (no pain) to 10 (worst pain ever). In another series of 13 patients with similar follow-up who were treated with a 4-min manual-control protocol from a low-output generator with a noncooled 5-mm exposed-tip probe, the patients scored their postprocedure pain on the same scale as that used in our series and rated their pain at a mean of 3.3 [9]. There was significantly more postprocedure pain in our study group. The interval to being pain-free was 5.6 days in our series of 11 patients, which is prolonged in comparison with the published data with a 4-min manual-control protocol from a low-output generator with a noncooled 5-mm exposed-tip probe. This may reflect larger zones of ablation involving the periosteum, cortex, and marrow.
Since the first animal experiments and clinical application of radiofrequency ablation in bone in 1989 there have been considerable changes in the system of energy delivery. The goal has been ablation of an increased amount of tissue in the field of a soft-tissue tumor. Modern radiofrequency generators have potential outputs four times that used in the original experiments. Impedance-control software in radiofrequency generators increases the size of the zone of ablation. The treatment time has also been prolonged.
Probe design has changed. The commercially available cooled probe has a minimum exposure of 1 cm. The diameter has increased from 0.9 mm with a 5-mm exposed-tip noncooled probe to 1.5 mm with a cooled probe. They can be internally cooled by water mechanically pumped through the probe at near 0°C to alter the thermal profile of the lesion and allow a larger zone of ablation [13]. The cooled probe is now a single-use item because of the design of the internal cooling tubing, which increases the cost in comparison with the 5-mm noncooled probe that can be sterilized after each use.
The zone of marrow signal change in our series is larger than that produced in any of the animal experiments published about therapy of normal bone. In the therapy of osteoid osteoma, this thermal lesion is too big.
None of the patients in our study group experienced fracture, but the increased energy delivery by modern radiofrequency equipment and the size of the zone of ablation have an undetermined potential for bone weakening. The effect of cellular death in cortical bone and the subsequent weakening from healing are unknown. The fibroconnective matrix and inorganic matrix remain intact except for a small access drill hole. The drill hole we used was only 1.6 mm, which is smaller than the access used in other clinical series with the 5-mm exposed-tip probes.
MRI allows the detection of temporal marrow signal change after radiofrequency ablation. The characteristic pattern of MR signal changes in bone marrow after radiofrequency ablation can determine the extent and adequacy of marrow therapy when the question arises. The optimum time for imaging is 28 days after treatment. Long active-tip cooled probes and impedance-control software can deliver large zones of marrow signal alteration. Preliminary data indicate that increased therapy with the 1-cm probe leads to improved clinical success, but the protocol needs to be optimized.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  E. Liapi and J.-F. H. Geschwind Transcatheter and Ablative Therapeutic Approaches for Solid Malignancies J. Clin. Oncol., March 10, 2007; 25(8): 978 - 986. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Peyser, Y. Applbaum, A. Khoury, M. Liebergall, and K. Atesok Osteoid Osteoma: CT-Guided Radiofrequency Ablation Using a Water-Cooled Probe Ann. Surg. Oncol., February 1, 2007; 14(2): 591 - 596. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
