| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 All authors: Mallinckrodt Institute of Radiology, Washington University School
of Medicine, 510 S Kingshighway Blvd., Campus Box 8131, St. Louis, MO
63110.
2 Present address: Lancaster Radiology Associates, Lancaster, PA.
3 Present address: Advanced Diagnostic Imaging, Belleville, IL.
Received April 19, 2006;
accepted after revision August 1, 2006.
Address correspondence to L. A. Gilula.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. A retrospective study was conducted to review 354 consecutive percutaneous vertebroplasty procedures on 694 compression fractures. Patients were excluded from consideration if they were treated for metastatic compression fracture or if they were treated at more than a single vertebral body level. Sixty-five patients met the inclusion criteria. Preprocedure radiographs and MR images were reviewed with specific attention to the presence or absence of intravertebral gas or fluid. Images obtained at the procedure also were reviewed for the presence or absence of an intravertebral cleft. Imaging findings were correlated with subjective pain scores immediately, 2 weeks, 1 month, 3 months, 6 months, 1 year, and 2 years after the procedure.
RESULTS. Thirty-one (48%) of the 65 patients had evidence of a fracture cleft. Twenty-seven patients had opacification of an intravertebral fracture cleft at percutaneous vertebroplasty, and four patients had an intravertebral cleft on preprocedure imaging but did not have cleft opacification. Thirty-four (52%) of the patients had no evidence of a fracture cleft and had only a trabecular pattern of opacification. Although there was a trend toward a greater failure rate in patients with a filled cleft, there was no statistically significant difference in subjective pain scores between the groups.
CONCLUSION. Pain relief with vertebroplasty is similar in patients with and those without intravertebral fracture clefts. Because of the small number of unfilled fracture clefts in our population, the true incidence of post-percutaneous vertebroplasty pain in patients with an un-filled cleft remains uncertain.
Keywords: interventional radiology • spine • vertebroplasty
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
|
Lane et al. [20] reported a trend toward greater pain relief in patients with clefts opacified at percutaneous vertebroplasty. Complete filling of intravertebral fracture clefts with PMMA cement has been advocated to maximize stabilization of the vertebral body [20]. Anecdotal evidence at our institution supports the theory that stabilization of a vertebral fracture cleft with cement may relieve pain. A few patients who have undergone percutaneous vertebroplasty at other institutions have had pain relief after repeated percutaneous vertebroplasty at the same vertebral level at our institution when an intravertebral cleft formerly not filled was found to fill at retreatment. The purpose of this study was to determine whether patients who underwent filling of intravertebral fracture clefts had more favorable outcome of percutaneous vertebroplasty than patients without clefts.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Patients were excluded from consideration for this study if they were treated for compression fractures secondary to metastatic disease or multiple myeloma. To simplify evaluation of postprocedure pain, patients were also excluded from consideration if they were treated at more than a single vertebral level or if they were subsequently treated for new painful compression fractures. After the exclusions, 65 patients (15 men, 50 women; age range, 31-96 years; mean age, 72 ± 12 [SD] years), were selected for inclusion in the study (Table 1).
|
All available preprocedure radiographs, CT scans, and MR images were reviewed by two of the authors with specific attention to the presence or absence of an intravertebral cleft. Cases of disagreement or question were resolved by consensus of all medically trained authors. On radiography and CT, intravertebral cleft was defined as a linear, well-demarcated focus of intravertebral fluid or gas attenuation (Fig. 1A, 1B, 1C). On MRI, intravertebral cleft was defined as a linear well-demarcated focus of T2 prolongation similar to that of adjacent CSF (Fig. 2). Signal void on T2- and T1-weighted images, which is characteristic of gas, was also considered an intravertebral cleft. Fluoroscopic spot radiographs of the spine immediately before and during percutaneous vertebroplasty were reviewed. An intravertebral cleft was defined on fluoroscopy as a linear well-demarcated focus of fluid or gas attenuation within the vertebral body as shown on routine radiography (Fig. 3A). Postprocedure fluoroscopic images were reviewed to classify the pattern of cement opacification. Immediate dense filling of a geographic intravertebral cavity was classified as cleft opacification (Figs. 3B and 3C). Less dense opacification tracking along intravertebral trabeculae in a nongeographic distribution was characterized as trabecular opacification. Patients were separated into three groups: those who had cleft opacification, those who had no indication of a cleft on preprocedure imaging or during fluoroscopy and had no cleft opacification, and those who had a fracture cleft on preprocedure imaging or during fluoroscopy but did not have cleft opacification at percutaneous vertebroplasty.
|
|
|
|
Percutaneous Vertebroplasty Procedure
All procedures were performed in the presence of a board-certified
musculoskeletal radiologist. Most of the procedures were performed by one of
the authors. The patients were interviewed before the procedure to determine
the exact site of pain. Clinical examinations were performed with fluoroscopy
to aid in precise pain localization. The patients were placed in the prone,
slightly oblique position on the fluoroscopy table. With increasing
experience, positioning the patient to produce hyperextension of the spine to
try to increase vertebral height became routine. Percutaneous vertebroplasty
was performed under strict sterile conditions with fluoroscopic guidance. The
procedure was performed in the usual manner, described in detail by Shimony et
al. [39]. Cement was injected
until it reached the posterior one fourth of the vertebral body or until there
was leakage outside the vertebral body. If there was pre-dominant filling of
only one side of the vertebral body, a second needle was used to enter the
contralateral pedicle. The procedure was repeated to achieve filling of most
of the vertebral body.
Outcome Evaluation
Patients were asked to rate their pain level immediately before the
procedure using a visual analog scale of 0-100, zero meaning no pain and 100
indicating the worst possible pain. At discharge after percutaneous
vertebroplasty, the visual analog scale level was rechecked. The pain level
was then evaluated with a follow-up telephone questionnaire approved by the
local institutional review board (Appendix
1). The calls were made by a research assistant not involved in
the vertebroplasty procedure 2 weeks, 1 month, 3 months, 6 months, 1 year, and
2 years after the procedure. The patients were asked whether pain was absent,
improved, the same, or increased compared with the pain before the procedure.
The follow-up questionnaire was completed in this manner because many patients
had difficulty using the visual analog scale in person or over the
telephone.
APPENDIX 1 : Vertebroplasty Follow-Up Questionnaire
This study did not have a completely longitudinal structure because many of the patients had incomplete data and because the telephone follow-up approved by the institutional review board started well after vertebroplasty had been introduced at our institution. Sixteen patients had pain scores for all six follow-up times, six patients had pain scores for five of the times, four patients had pain scores for four times, nine patients had pain scores for three times, 14 patients had pain scores for two times, 10 patients had pain scores for one time, and six patients had no pain scores for any of the follow-up times. Of the six patients with no follow-up information, five had died and one was lost to follow-up. The five deaths were not related to vertebroplasty.
Statistical Analysis
Data were analyzed with contingency tables of cleft outcome versus reported
pain by time. Because the sample size was relatively small, patterns were
tested for statistical significance after combination of the pain categories
gone and better and the categories same and worse. Patterns were tested for
statistical significance with Fisher's exact tests. Analysis was performed
with JMP 5.0 (SAS Institute).
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Overall, there was a large decrease in numeric pain score immediately after the procedure, from a mean of 69.7 to a mean of 15.8. The differences in pain scores between cleft outcomes were trivial both before (filled cleft mean, 64.8; unfilled cleft mean, 60.0; no cleft mean, 74.6; p = 0.23, analysis of variance) and immediately after treatment (filled cleft mean, 14.2; unfilled cleft mean, 18.8; no cleft mean, 16.6; p = 0.92, analysis of variance). Most of the patients also reported relief of pain on follow-up (Table 2).
|
Although there was a trend toward less pain relief in patients with filled clefts compared with patients without clefts, this difference did not approach statistical significance (p > 0.15 in all cases, Fisher's exact test). Only four patients had unfilled clefts, and they had a maximum of three reports at any follow-up time point, so it is impossible to generalize from these data beyond observing that there were no reports of worse pain.
In the cases of 21 (78%) of the 27 patients with cleft opacification at percutaneous vertebroplasty, fluoroscopic spot views obtained during the procedure before injection of cement showed an intravertebral cleft (Table 3). Preprocedure MR images were available for review for 23 of the 27 patients, and the images of 13 (57%) of the patients showed a cleft. Preprocedure radiographs of the spine were available for review for 25 patients, and these images showed only 11 (44%) of the patients had a cleft. Preprocedure radiographs and MR images were available for review for 22 patients, and only seven (32%) of these patients were found to have a fracture cleft on both studies. CT examinations available for review for eight patients showed that five (62.5%) of the patients had clefts. Three patients with cleft opacification at percutaneous vertebroplasty did not have an intravertebral cleft on any preprocedure images or during fluoroscopy.
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Our data indicate a trend toward less pain relief in patients with a filled cleft compared with patients with no cleft. These findings may be explained by the preferential filling of intravertebral clefts during percutaneous vertebroplasty. As mentioned in the literature, vertebral fracture clefts often easily opacify without specific redirection of the needle into the cleft [40]. If there is predominant filling of the cleft, it is possible that the remaining vertebral body will remain unsupported and untreated, causing further pain, especially if there is further collapse of the unfilled part of the treated vertebral body. It has been suggested that filling the vertebral cleft alone may not be adequate in some patients [41]. Wagner and Baskurt [41] described one case in which a vertebral body refractured with anterior extrusion of cement after filling of a large intravertebral cleft during percutaneous vertebroplasty. Of our four patients who had a cleft on preprocedure imaging but did not have cleft opacification, none returned with increased pain. Although it is difficult to generalize from so few patients, it seems that some clefts that do not initially opacify with cement may be stabilized effectively without filling of the cleft with PMMA.
Studies [12, 20] have shown that preprocedure imaging is not sensitive in detection of all clefts seen at percutaneous vertebroplasty, and our data support this finding. MRI depicted only 57% (13/23) of clefts opacified at percutaneous vertebroplasty. The most sensitive technique in detection of clefts was fluoroscopy performed at the procedure, which showed 78% (21/27) of clefts. This success is likely secondary to widening of the cleft with extension of the spine during positioning for percutaneous vertebroplasty, a maneuver that was performed on our patients with increasing frequency as our experience with vertebroplasty increased. Fracture clefts have been shown to change in appearance with changing position [10]. As suggested by McKiernan and Faciszewski [12], the insensitivity of imaging may also be related to the time it takes for cleft margins to become more defined after fracture.
Limitations of this study stemmed from the retrospective nature of this type of analysis. The study was not completely longitudinal because many patients had incomplete data. The lack of data at some of the earlier time points was related to the use of a questionnaire approved by the institutional review board well after the introduction of vertebroplasty. In addition, many of our patients experienced difficulty with use of the visual analog scale over the telephone during follow-up interviews. We elected to use a simpler scale of pain score, which has not been proven in clinical studies. Patients were asked to remember pain that they had had as long as 2 years previously, a difficult task in the best of circumstances. Finally, in an attempt to simplify outcome, our population was limited to a subset of patients treated at only a single vertebral level. Patients who returned for subsequent fracture treatments also were excluded. These exclusions may have produced a population with a particularly favorable post-percutaneous vertebroplasty outcome, masking potential differences. These exclusions also decreased the total number of patients, limiting the power of statistical analysis. Because of the small number of unfilled fracture clefts in our population, the true incidence of persistent pain after nonfilling of a fracture cleft is unknown.
In summary, intravertebral fracture clefts have long been described in the literature and are often identified in patients with intractable back pain who undergo percutaneous vertebroplasty. Preprocedure MRI and preprocedure radiography are not sensitive in consistent detection of clefts seen at the procedure, and most commonly clefts are visualized during fluoroscopy before cement injection. Often the clefts easily opacify with PMMA without direction of the needle into the cleft. There is no reported difference in outcome among patients who have filled clefts compared with those without clefts, although we identified a tendency toward more frequent failure in patients with a filled cleft than in those with no cleft.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  N. Tanigawa, S. Kariya, A. Komemushi, T. Tokuda, M. Nakatani, R. Yagi, and S. Sawada Cement Leakage in Percutaneous Vertebroplasty for Osteoporotic Compression Fractures With or Without Intravertebral Clefts Am. J. Roentgenol., November 1, 2009; 193(5): W442 - W445. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. E. Rad, L.A. Gray, and D.F. Kallmes Significance and Targeting of Small, Central Clefts in Severe Fractures Treated With Vertebroplasty AJNR Am. J. Neuroradiol., August 1, 2008; 29(7): 1285 - 1287. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
