| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Developments |
1 From the Department of Radiology (J.B.M., S.G.W., T.S., A.K., D.A.R.), Division of Oncology (Y.S, P.Y.D.), and Division of Neurosurgery (M.P.), Geneva University Hospital, Rue Micheli-du-Crest 24, CH-1211 Geneva 14, Switzerland; and the Interventional Neuroradiology Section, Johns Hopkins Medical Institutions, Baltimore, Md (P.G.). From the 2000 RSNA scientific assembly. Received August 2, 2002; revision requested October 7; revision received January 8, 2003; accepted February 28. Address correspondence to J.B.M. (e-mail: jean-baptiste.martin@dim.hcuge.ch).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMaterials and Methods ResultsDiscussionREFERENCES |
|---|
© RSNA, 2003
Index terms: Spine, secondary neoplasms, 32.331, 33.331 • Spine, vertebroplasty, 32.126, 33.126 • Spondylolysis, 32.423, 33.423
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and Methods ResultsDiscussionREFERENCES |
|---|
There are two main access routes for performing vertebroplasty in the thoracic or lumbar spine: the transpedicular route, which may be either bipedicular (9) or unipedicular (10), and the direct lateral route (6). The direct lateral route is usually used for the treatment of tumors if there is a vertebral metastatic infiltration that includes both the pedicles. When only one pedicle is affected, unipedicular needle placement into the unaffected and thus fluoroscopically visible pedicle is an alternative access route. However, treatment of the entire vertebral body, especially of the vertebral body adjacent to the contralateral lysed pedicle, may not be possible. Moreover, with these techniques the lysed pedicle is usually not treated, and pain arising from it is not diminished.
We recently used an alternative access route to perform percutaneous vertebroplasty of vertebral bodies with pedicle lysis: We placed the vertebroplasty needle via the diseased pedicle into the vertebral body. This route makes it possible to treat infiltrated vertebral bodies and infiltrated pedicles with percutaneous injection of PMMA cement. The feasibility of this technique, which is termed pediculoplasty, was recently described (11). The purpose of our study was to obtain preliminary results with use of this access route in consecutive patients with pedicle lysis.
|
Materials and Methods |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and Methods ResultsDiscussionREFERENCES |
|---|
The primary tumor categories were lung cancer (n = 11); breast cancer (n = 9); myeloma (n = 8); and other (n = 4), which included one case each of renal carcinoma, melanoma, epidermoid carcinoma, and neuroendocrine tumor. In the 32 patients, 87 vertebral bodies were treated with vertebroplasty. In addition to standard vertebroplasty, 47 of these vertebral bodies were treated with pediculoplasty. Four vertebral bodies were treated with pediculoplasty on both sides. Thus, a total of 51 pedicles were treated: 34 in the thoracic spine and 17 in the lumbar spine. Pedicles were treated in the thoracic spine up to the level of T2. In detail, the levels were (n = numbers of pedicles): T2 (n = 1), T4 (n = 1), T5 (n = 3), T6 (n = 4), T7 (n = 4), T8 (n = 6), T9 (n = 5), T10 (n = 4), T11 (n = 3), T12 (n = 3), L1 (n = 5), L2 (n = 1), L2 (n = 3), L4 (n = 2), and L5 (n = 6).
Imaging and Evaluation
At the time of the procedure, the presence of pedicle lysis was based in all patients on findings obtained from computed tomographic (CT) and/or magnetic resonance (MR) images. A pedicle was considered lytic if obvious signs of tumor invasion within the pedicle were evident; complete destruction of the pedicle did not have to be present. For the retrospective analysis, we assessed four imaging criteria in a standardized fashion. For all criteria, imaging evaluation was performed by consensus of two neuroradiologists who are part of the team of four operators who perform percutaneous vertebroplasty at our institution. First, CT images obtained prior to treatment were assessed (J.B.M., A.K.) to determine if the inner cortex of the pedicle to be injected was intact. The inner cortex was rated as intact if the cortical border was visible without any destruction in its entire course. In the majority of patients, CT scanning had been performed on an outpatient basis in outside institutions and with use of different scanning protocols. However, in all cases the CT scanning was performed with a maximum section thickness of 3 mm, and images reconstructed with a bone algorithm were available. Second, fluoroscopic images obtained for planning of the access route were assessed (J.B.M., S.G.W.) to determine if the pedicles to be treated were visible. A pedicle was rated as visible if, on anteroposterior images, the characteristic oval shape of the pedicle was clearly seen in all parts. If part of that shape was not visible or if the pedicle was not visible at all, the pedicle was rated as partially visible or not visible. Third, fluoroscopic images were assessed (J.B.M., S.G.W.) to determine if the cement filling was satisfactory. The filling was considered satisfactory if cement filled the metastasis and extended from the body through the affected pedicle. Fourth, CT images obtained immediately after treatment were assessed (J.B.M., A.K.) to determine if there was local extravasation of PMMA toward the spinal canal or into the neuroforamen. These images were obtained with a four–detector row CT scanner (MX 8000; Philips Medical Systems, Eindhoven, the Netherlands) with a collimation of 2–3 mm, pitch of 1, 1.6-mm reconstruction intervals, 120 kV, 330 mAs, and 180-mm field of view.
Technique
All interventional procedures were performed by one of four operators (J.B.M., S.G.W., T.S., A.K.) who had 2–7 years of experience performing percutaneous vertebroplasty. Approximately 150 patients undergo percutaneous vertebroplasty at our institution each year.
Patients were treated in an angiography suite equipped with a biplanar fluoroscopic unit (BV 3000; Philips Medical Systems, Best, the Netherlands). For the majority of procedures (n = 26), patients were administered general anesthetic. Each patient was placed in a prone position on the examination table, the patient’s skin was prepared and draped, and the access route through the lysed pedicle was defined on anteroposterior fluoroscopic images. If the pedicle was not visible or was only partially visible, the position of the pedicle was deduced from the position of the contralateral pedicle and the spinous process, which served as a landmark (Fig 1). If neither pedicle was visible, transpedicular access was achieved by aligning the needle parallel to the position of pedicles above and below the level to be treated.
|
|
After treatment of the vertebral body, which was performed in all patients in approximately 6 minutes, the needle was withdrawn stepwise through the pedicle, and the cement injection was controlled by using lateral fluoroscopy. If barium particles were observed to move perpendicular to the needle, which indicated potential extrapedicular leakage, the injection was immediately stopped (Fig 3). To inject preferentially into the outer part of the pedicle, the bevel was oriented externally (Fig 2). In most patients it was sufficient to deliver only the amount of cement contained in the lumen of the needle (0.7 mL); this was performed by introducing the stylet into the needle. This was performed in all patients within approximately 2 minutes. The filling was considered satisfactory if the cement filled the metastasis and extended from the body through the affected pedicle (Fig 1). In the four patients with severe bilateral pedicle lysis who underwent treatment of both sites, the needle placement, cement preparation, and injection were performed consecutively for one site and then for the other.
|
the worst pain of your life.
Pretreatment evaluation was performed on the day prior to the procedure, and follow-up evaluation was performed immediately prior to the discharge of the patient from the hospital, which was typically 2–5 days after vertebroplasty. Two experienced oncologists (Y.S., P.Y.D.) together administered the clinical pain evaluation to all patients before and after vertebroplasty. Any amount of pain relief was considered positive. The same oncologists (Y.S., P.Y.D.) reviewed a patient’s chart if clinical complications related to the percutaneous vertebroplasty treatment had occurred (eg, pulmonary embolism). |
Results |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and Methods ResultsDiscussionREFERENCES |
|---|
In 27 (84%) of 32 patients, CT images obtained prior to treatment were available for review. In these 27 patients, 42 pedicles were treated. In the remaining five patients, in whom a total of nine pedicles were treated, either CT images were not retrievable or only cross-sectional MR imaging had been performed. Thus, 42 (82%) of all 51 treated pedicles were reviewed on preprocedural CT images. In 23 (55%) of the 42 pedicles, signs of destruction of the medial cortex were present. At the retrospective analysis of the visibility of the vertebral pedicles on fluoroscopic images, 18 (35%) of the 51 treated pedicles had been completely visible, and 19 (37%) and 14 (27%) pedicles had been partially visible or not visible, respectively. In all pedicles treated, the filling was considered satisfactory. On postoperative CT images, extravasation of PMMA cement toward the spinal canal was observed in areas adjacent to three (6%) of 51 pedicles treated. Leakage of PMMA into a neuroforamen was seen next to one treated pedicle (2%) (Fig 4). In all cases, the amount of PMMA cement was minimal, and none of the patients experienced an increase in pain level or a clinical complication.
|
|
Discussion |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and Methods ResultsDiscussionREFERENCES |
|---|
Pedicle lysis is common in metastatic disease of the spine, and the absence of the usually well-visualized pedicle on conventional radiographs represents a well-known sign in the detection of osteolytic lesions. In our experience, the integrity of the pedicle was not a limiting factor for the direct access route: Even severely lysed pedicles, which could not be directly visualized during anteroposterior fluoroscopy, were treated safely. We consider knowledge of the polymerization time of the cement mixture, which is necessary for performing all percutaneous vertebroplasty procedures safely, to be a key point to performing pedicle reconstruction without complications. When the mixture is prepared as in our study, we know that the polymerization time is 8 minutes when working at room temperature (21°C) (5,7,8). In all of our cases, the cement was injected into the pedicle 6–7 minutes after its preparation and after treatment of the vertebral body. At this point, the cement begins to become very viscous and thus the likelihood of cement leakage into adjacent structures or vessels becomes unlikely as long as the material is delivered cautiously. However, fluoroscopic guidance throughout cement filling of the pedicle is mandatory. A lateral view was therefore used in our study to monitor the needle retrieval and the cement delivery into the pedicle. Why was this view used? First, since pediculoplasty was performed after vertebroplasty, detection of medial extravasation of PMMA toward the spinal canal would have been impaired on anteroposterior images because of superimpositions of the pedicle and the vertebral body. Second, since the syringe filled with PMMA was directly connected to the needle, anteroposterior imaging during cement injection would not have been possible because of the exposure of the operator’s hand. This could have been avoided by using an extension line between the syringe and the needle. We did not use such a system, as we preferred to have a simultaneous manual control on both the needle—placed in a fragile pedicle—and the syringe. In our experience, however, this sole viewing direction can be considered sufficient in controlling cement leakage for the following reasons: First, the injection was performed after filling of the vertebral body during retrieval of the needle; the delivered cement, which was very viscous at this time, primarily filled the canal that was created by the withdrawn needle. Second, a potential venous leakage into the epidural venous plexus or the periradicular veins could occur. These veins are not oriented parallel to the pedicle. Therefore an ascending or descending movement of barium particles in relation to the pedicle, which could be visualized in the lateral view, would be observed. Third, we immediately stopped the delivery of PMMA if the barium particles took a preferential direction that was not along the axes of the pedicle, which indicates the possibility that there is an extrapedicular leakage. By taking these precautions and taking into consideration that only a small amount of PMMA (less than 1 mL) was injected into the pedicle, we believe that the risk of an epidural PMMA leakage is very low and, if it occurred, would result in a leakage of only a small amount of PMMA. In our series of 51 pedicles treated with that technique, we encountered extrapedicular leakage of PMMA adjacent to only four pedicles, and the amount of cement was minimal in all cases.
We recommend the use of beveled needles, which allow applied forces to be directed: If the bevel is directed medially during passage through the pedicle, the needle tends to deviate externally. This helps to avoid injury of the spinal canal. Furthermore, the delivery of cement direction can be influenced by the bevel orientation (eg, if the bevel is oriented externally during pedicle reconstruction, the cement is delivered preferentially to the external part of the pedicle). In a recent study (12), it was shown that venography does not significantly increase the safety of percutaneous vertebroplasty performed by experienced operators. However, for treatment of tumors that may manifest abnormal vessel architecture, we find that antecedent venography prior to PMMA injection is useful to detect potential epidural leakage sites and to select the starting time of cement delivery (eg, if severe venous leakage is seen at venography, the cement is allowed to polymerize for a longer time before injection to increase the viscosity). However, the need for this procedure has yet to be proved. Further technical improvements of this access route might include venographic guidance during positioning of the needle, not only when the needle is placed in the vertebral body, but also when it is placed in the pedicle during the access. In our study, the vast majority of patients were administered general anesthetic, as in most patients multiple vertebral levels were treated. It may be favorable to perform percutaneous vertebroplasty with pediculoplasty with use of local anesthetic so that the patients could comment on radicular pain, which would indicate potential cement leakage. The type of anesthetic must be chosen on an individual basis after considering patient comfort, ease of performing procedure, and safety.
There are well-recognized limitations to this retrospective analysis. The most important is that we cannot isolate the efficiency of pediculoplasty for pain treatment; this is because in all of our patients the pedicle reconstruction was performed as an adjunct to stabilization of a vertebral body. The results of this study evince the need for further evaluation of the utility of pediculoplasty. Prospective studies comparing patients treated with pediculoplasty with those who are treated only with PMMA injection into the vertebral body are needed, and both procedural safety considerations and clinical outcomes should be taken into account. Furthermore, it would be of interest to investigate the effectiveness of pediculoplasty for pain treatment in those patients with vertebral metastatic disease in whom lytic areas in the vertebral body are absent.
Although the clinical effect of pediculoplasty cannot be well assessed on the basis of our preliminary data, we strongly believe that the technical aspect of pediculoplasty merits its consideration in all centers performing vertebroplasty in patients with tumors. First, the method proved to be safe in all applied cases. Second, and more important, the access route used for pediculoplasty allows treatment of vertebral metastasis, which might not be reached by using the established techniques and thus can be considered as an alternative to the established methods. Moreover, it might be possible that the assumed, yet not clinically proved, antitumoral exothermic effect of PMMA (13,14) applied as well to an affected pedicle is favorable for the long-term outcome.
In our experience of 51 pedicles treated in 32 patients, percutaneous PMMA injection for treatment of vertebral metastatic disease through a lysed pedicle proved to be safe. This technique allows, in contrast to other access routes, the direct treatment of the lysed pedicles.
|
FOOTNOTES |
|---|
Author contributions: Guarantor of integrity of entire study, J.B.M.; study concepts and design, J.B.M.; literature research, A.K.; clinical studies, J.B.M., S.G.W., Y.S., T.S., P.Y.D.; data acquisition, J.B.M., S.G.W., Y.S., P.Y.D.; data analysis/interpretation, J.B.M., S.G.W., Y.S., T.S., M.P., P.Y.D.; manuscript preparation, J.B.M., S.G.W., A.K., D.A.R.; manuscript definition of intellectual content, D.A.R.; manuscript editing, S.G.W., D.A.R.; manuscript revision/review, S.G.W., P.G., D.A.R.; manuscript final version approval, all authors
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and Methods ResultsDiscussionREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  G. Christoforidis, D. Dang, and J. Gabriel Catheter-directed percutaneous transpedicular c2/c3 vertebroplasty in a patient with fibrous dysplasia using seldinger technique. AJNR Am. J. Neuroradiol., September 1, 2006; 27(8): 1738 - 1740. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. M. Barragan-Campos, J.-N. Vallee, D. Lo, E. Cormier, B. Jean, M. Rose, P. Astagneau, and J. Chiras Percutaneous Vertebroplasty for Spinal Metastases: Complications Radiology, January 1, 2006; 238(1): 354 - 362. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
