This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Letters to the Editor: Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when Letters to the Editor are posted
Right arrow Alert me if a correction is posted
Services
Right arrow E-mail this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My File Cabinet
Right arrow Download to citation manager
Right arrow Rights and Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by LONSTEIN, J. E.
Right arrow Articles by WINTER, R. B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by LONSTEIN, J. E.
Right arrow Articles by WINTER, R. B.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us   Add to Facebook   Add to Technorati   Add to Twitter  
What's this?
The Journal of Bone and Joint Surgery 81:1519-28 (1999)
© 1999 The Journal of Bone and Joint Surgery, Inc.

Complications Associated with Pedicle Screws*

JOHN E. LONSTEIN, M.D.{dagger}, FRANCIS DENIS, M.D.{dagger}, JOSEPH H. PERRA, M.D.{dagger}, MANUEL R. PINTO, M.D.{dagger}, MICHAEL D. SMITH, M.D.{dagger} and ROBERT B. WINTER, M.D.{dagger}, MINNEAPOLIS, MINNESOTA

Investigation performed at the Twin Cities Spine Center, Minneapolis


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Background: The safety and the effectiveness of pedicle-screw instrumentation in the spine have been questioned despite its use worldwide to enhance stabilization of the spine. This review was performed to answer questions about the technique of insertion and the nature and etiology of complications directly attributable to the screws. Methods: We performed a retrospective review of all of the pedicle-screw procedures that were done by us from January 1, 1984, to December 31, 1993. We inserted 4790 screws during 915 operative procedures on 875 patients; 668 (76.3 percent) of the patients had a lumbosacral arthrodesis. The mean duration of follow-up was three years (range, two to five years). The accuracy of screw placement was assessed on intraoperative, immediate postoperative, and follow-up radiographs with use of a technique that was developed by one of us (F. D.); this technique has yet to be validated to determine the prevalence of various types of error. Results: Of the 4790 screws, 4548 (94.9 percent) had been inserted within the pedicle and the vertebral body. One hundred and thirty-four (2.8 percent) of the screws had perforated the anterior cortex, and this was the most common type of perforation. One hundred and fifteen (2.4 percent) of the screws were associated with complications that could be ascribed to the use of pedicle screws. The most common problem was late-onset discomfort or pain related to a pseudarthrosis or perhaps to the screws; this problem was associated with 1102 (23.0 percent) of the screws, used in 222 (24.3 percent) of the procedures. The symptoms necessitated removal of the instrumentation with or without repair of the pseudarthrosis. A pseudarthrosis was found during forty-six (20.7 percent) of the 222 procedures. Irritation of a nerve root occurred after nine procedures (1.0 percent) and was caused by eleven screws (0.2 percent); it was more commonly caused by medially placed screws. Three patients had residual neurological weakness despite removal of the screws. Twenty-five screws (0.5 percent), used in twenty procedures (2.2 percent), broke. The screws that broke were of an early design. A pseudarthrosis was found in thirteen of twenty patients who had broken screws. Sixteen of the twenty patients had an exploration; three of them were found to have a solid fusion, and thirteen were found to have a pseudarthrosis. The remaining four patients had evidence of a solid fusion on radiographs and had no pain. Conclusions: There are few problems associated with the insertion of screws, provided that the surgeon is experienced and adheres to the principles and details of the operative technique. Our review revealed a low rate of postoperative complications related to pedicle screws. The problem of late-onset pain may be related to the implants or to the stiffness of the construct; however, it is difficult to accurately identify its exact etiology.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
The complications associated with the use of pedicle screws have been reviewed in several studies, and the accuracy of insertion of the screws has been discussed in some reports as well. These studies have varied greatly with respect to the numbers of patients and devices, with most authors providing the number of patients but not the number of screws. The results have been presented in relation to either the number of procedures or the number of screws, but rarely in relation to both. Most of the series have involved a single diagnosis or procedure, usually a fracture of the spine or an arthrodesis of the lumbosacral region.

The number of patients in the studies of which we are aware ranged from twenty-one to 617; we know of only seven studies2-5,10,19,25 that involved more than 200 patients. Six of the studies2-5,19,25 were based on the experience of a single surgeon or group. Esses et al.10 reported on the largest series of which we are aware; they documented the results of a questionnaire sent to members of the American Back Society. Only ten5,7,10-12,21-23,25,29 of twenty-nine1,2,4-30 articles that we reviewed mentioned the total number of screws that were inserted; this number ranged from 116 to 3949. Only four of the series5,10,21,25 involved more than 1000 screws. The spinal instrumentation used in these studies included the Steffee VSP (variable screw placement), Cotrel-Dubousset, AO internal fixator, Luque, Zielke, Wiltse, and Roy-Camille systems.

Eight reports documented the rate of misplaced screws; this rate ranged from 1 to 9 percent of patients5,7,8,10,28 and from 1 to 18 percent of screws5,7,8,11,12,21,28.

The most common complication mentioned in these reports was screw breakage, with rates of 1 percent of 198 screws7 to 11 percent of 116 screws11 and of 0.6 percent of 170 patients14 to 25 percent of seventy-seven patients30. The rates of nerve-root irritation have ranged from 0 percent of 941 screws21 to 1 percent of forty screws and from 1.1 percent of 533 procedures5 to 10 percent of twenty-one patients18. The literature appears to combine different n values for patients and screws, making it difficult to analyze the data.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
With use of our computer database, we identified patients in whom we had inserted pedicle screws between January 1, 1984, and December 31, 1993; we then performed a retrospective review of their medical records and radiographs. The collected data were entered on a form that could be scanned into the database and were assessed with standard methods of evaluation and statistics. All of the patients were followed postoperatively for a minimum of two years (range, two to five years; mean, three years) to identify all problems, such as broken and bent screws and the need to remove screws. The medical records and radiographs were reviewed by a nurse-researcher and one of us, and no surgeon reviewed the cases of his own patients. The operative, postoperative, and follow-up reports were reviewed to identify any problems that had occurred during insertion of the screws and any postoperative complications related to the screws. Postoperative sensory or motor radiculopathy was not listed as a complication related to the screw when decompression had been performed as part of the arthrodesis and the screw was within the pedicle. Neurological symptoms associated with a misplaced screw were included as complications related to screws. When a decompression was performed, it was possible to assess malposition of the screw or a fracture of the pedicle wall because the pedicle was visible and palpable, and insertion of the screws was generally more accurate because the medial and inferior walls could be palpated.

The accuracy of screw placement within the pedicle was assessed intraoperatively, postoperatively, and at the time of follow-up with radiographs made in the coronal, sagittal, and oblique planes. Computed tomography is the most accurate method with which to assess screw placement. However, in our series, this modality was used only when there was a neurological deficit in the early postoperative period or when there was pain in the late postoperative period (particularly when the position of the screws could not be ascertained on radiographs). Computed tomography was thus performed on too few patients for us to be able to draw any conclusions on the basis of that modality. Multiple radiographs were usually necessary because of the marked variation in the projected image in the presence of multiple screws.

One of us (F. D.) developed a method to assess the accuracy of screw placement within the pedicle. However, no study has been performed to validate the accuracy and reliability of this method. The placement of the screw in the pedicle is studied on coronal (anteroposterior or posteroanterior) and sagittal radiographs centered on the pedicle. The junction of the pedicle and the vertebral body on the sagittal radiograph corresponds with the pedicle shadow on the coronal radiograph. The sagittal radiograph is used to assess the ratio of the screw length in the vertebral body to the total screw length (Fig. 1-A). The position of the screw is then assessed on the coronal radiograph. The portion of the screw that was identified at the junction of the pedicle and the vertebral body on the sagittal radiograph is identified on the coronal radiograph (Fig. 1-B). This part of the screw should lie in the center of the pedicle shadow on the coronal radiograph. The direction of the screw is defined by its position—that is, medial, lateral, superior, or inferior—relative to the pedicle shadow. The placement of each screw was assessed in this manner.



View larger version (29K):
[in this window]
[in a new window]
  		Fig. 1-A Illustrations showing the technique, developed by one of us (F. D.), for assessment of the position of the screw in the pedicle. The ratio of the screw length in the vertebral body to the total screw length is assessed on a radiograph made in the sagittal plane (Fig. 1-A). The ratio shown is 50 percent. The portion of the screw at the junction of the pedicle and the vertebral body is also identified on the sagittal radiograph, and then this position is identified on the radiograph made in the coronal plane (Fig. 1-B). This part of the screw, which here is the midpoint of the screw, should lie in the center of the pedicle shadow on the coronal radiograph.

 


View larger version (21K):
[in this window]
[in a new window]
  		Fig. 1-B Illustrations showing the technique, developed by one of us (F. D.), for assessment of the position of the screw in the pedicle. The ratio of the screw length in the vertebral body to the total screw length is assessed on a radiograph made in the sagittal plane (Fig. 1-A). The ratio shown is 50 percent. The portion of the screw at the junction of the pedicle and the vertebral body is also identified on the sagittal radiograph, and then this position is identified on the radiograph made in the coronal plane (Fig. 1-B). This part of the screw, which here is the midpoint of the screw, should lie in the center of the pedicle shadow on the coronal radiograph.

 
The position of the screw in relation to the longitudinal axis can also be evaluated with this technique, but only if the vertebral body is seen on a direct end-on radiograph made in the coronal plane; the vertebra must lie directly on the central x-ray beam. Because of the lordosis in the lumbosacral area, the Ferguson radiograph is used to obtain a coronal view of the fifth lumbar vertebral body. The superior-inferior position of the screw is best seen on the sagittal radiograph, but the screw must lie on the central beam, in the middle of the film cassette. This is achieved intraoperatively by careful positioning of the film cassette and x-ray tube and postoperatively by making radiographs of the lumbosacral spine rather than of the lumbar spine. In our study, oblique radiographs showed any penetration of the anterior cortex of the vertebral body by the screw. Any concerns about the position of the screw were reviewed with a second one of us who was not the operating surgeon. The data were entered onto custom-designed data-entry forms, which were then scanned into a database program and evaluated with the use of statistical software (SPSS, Chicago, Illinois).

During the ten-year period of the study, 875 patients had 915 operative procedures that involved the insertion of pedicle screws; 605 of the procedures were performed during the last three years of the study (Fig. 2). The diagnoses varied. Most (668; 76.3 percent) of the patients had an arthrodesis of the lumbosacral region for the treatment of low-back problems (pseudarthrosis, a failed spinal operation, or degenerative disc disease) (484; 55.3 percent) or spondylolisthesis (184; 21.0 percent). The other diagnoses were idiopathic scoliosis in seventy-eight patients (8.9 percent), spinal trauma in seventy-three (8.3 percent), congenital scoliosis in twelve (1.4 percent), neuromuscular scoliosis in nine (1.0 percent), Scheuermann kyphosis in nine (1.0 percent), and miscellaneous diagnoses in twenty-six (3.0 percent). Three hundred and eighty (43.4 percent) of the patients had previously had one or more spinal operations; the number varied with the diagnosis. The group with low-back problems had the highest number of previous operations (270 patients; 55.8 percent of the group) (Fig. 3).



View larger version (22K):
[in this window]
[in a new window]
  		Fig. 2 Bar graph showing the number of procedures performed each year from 1984 to 1993.

 


View larger version (22K):
[in this window]
[in a new window]
  		Fig. 3 Bar graph showing the total number of patients who had each diagnosis as well as the number who had had a previous spinal procedure.

 
There were 484 female and 391 male patients, and the mean age at the time of the operation was 45.5 years (range, 3.4 to eighty-four years) (Fig. 4). Seven different instrumentation systems were used, with a total of 4790 screws inserted during the 915 operative procedures. The two most commonly used systems were the Cotrel-Dubousset system (3369 screws) and the Steffee VSP system (1227 screws), which together accounted for 95.9 percent of the screws used; the other five systems (AO, Cross, Harms, PWB [Puno-Winter-Byrd], and Texas Scottish Rite Hospital systems) made up the remaining 4.1 percent (Fig. 5). There was no difference in the percentage of complications related to the five less commonly used systems, and we did not find a learning-curve effect in relation to these five systems.



View larger version (25K):
[in this window]
[in a new window]
  		Fig. 4 Bar graph showing the age of the patients, by decade, at the time of the procedure.

 


View larger version (31K):
[in this window]
[in a new window]
  		Fig. 5 Pie chart showing the number of screws used with each of the different pedicle-screw systems. AO = AO internal fixator, CD = Cotrel-Dubousset, PWB = Puno-Winter-Byrd, VSP = Steffee variable screw placement, and TSRH = Texas Scottish Rite Hospital.

 
One to twelve screws (mean, five screws) were inserted during each procedure (Fig. 6). The level of insertion extended from the ninth thoracic to the second sacral vertebra, with only sixty-seven screws (1.4 percent) inserted in the thoracic spine. Most (3806; 79.5 percent) of the screws were inserted from the fourth lumbar to the first sacral vertebra (Fig. 7).



View larger version (22K):
[in this window]
[in a new window]
  		Fig. 6 Bar graph showing the number of procedures that were performed with each different number of screws.

 


View larger version (37K):
[in this window]
[in a new window]
  		Fig. 7 Bar graph showing the number of screws inserted at each vertebral level, from the ninth thoracic to the second sacral level.

 

Technique for Insertion of the Screws
The specific technique for insertion of the screws varied slightly among the six surgeons involved in the present study and was related to the experience of each surgeon. The patient was positioned prone and a cross-table lateral radiograph was made, centered on the site for the insertion of the pedicle screw, to provide information on the direction of the pedicle in the sagittal plane and, in some patients, to identify the vertebral level.

The segments to be stabilized were exposed to the tips of the transverse processes (and to the sacral alae in the lumbosacral area), and all of the soft tissue was cleared from the junction of the pedicle and the superior articular and transverse processes. The entry point to the pedicle was identified at the confluence of three anatomical landmarks, which often are marked by a ridge of bone; these landmarks consisted of a line along the axis of the transverse process, the pars interarticularis, and the superior articular process. Depending on the local anatomy and the vertebral level, the entry point could be more medial in relation to the superior articular process. A hole was made in the bone cortex with either a power dental burr or an awl. A cervical-spine curet or a pedicle probe was used to cannulate the pedicle. The direction of the insertion of the instrument was based on the level of the vertebra, the local anatomy, and the findings on the lateral radiograph. After the pedicle was cannulated, a ball-tip probe was used to carefully palpate all four quadrants of the pedicle for the presence of any defects of the osseous wall. Depending on the surgeon, his experience with the insertion of screws, the local anatomy, and the feel of the pedicle, posteroanterior and lateral radiographs were made either with pedicle markers or after insertion of the screws. The pedicle was tapped only if the surgeon determined that it was necessary on the basis of the ease of cannulation and the strength of the bone. Tapping of the pedicle was more commonly performed if a spinal arthrodesis had been done earlier and the bone was sclerotic. The transverse processes of the lumbar vertebrae or the alae of the sacrum were decorticated, and bone graft was laid in place before insertion of the screws.

When a laminectomy had been performed either before or at the time of the index procedure, the pedicle was palpated to identify the site of any break in the cortex and to provide an additional guide for the direction of cannulation of the pedicle. Also, the medial and inferior walls of the pedicle were inspected visually and palpated after both cannulation of the pedicle and insertion of the screws. When a decompressive laminectomy had not been performed and there was doubt as to the position of the pedicle, the screw placement, or the integrity of the pedicle walls, a laminotomy was done so that the integrity of the medial and inferior walls could be confirmed visually.

Anteroposterior and lateral radiographs were routinely made intraoperatively with the patient prone, postoperatively with the patient supine, and before discharge with the patient standing.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 

Position of the Screw
Of the 4790 screws, 4548 (94.9 percent) were judged to have been inserted wholly within the pedicle and the vertebral body. One hundred and thirty-four (2.8 percent) of the screws had perforated the anterior cortex (that is, they had penetrated the vertebral body and protruded anteriorly), and this was the most common type of perforation. The break in the integrity of the anterior cortex was often best seen on oblique radiographs made four to six months postoperatively to assess the quality of the fusion, and it reflected a so-called blind spot in the assessment of the lateral and posteroanterior position of the screw. Ninety-nine (74 percent) of the 134 screws that had penetrated the anterior cortex were in the sacrum, where anterior penetration was usually intentional to enhance screw fixation. Nineteen (0.4 percent) of the 4790 screws had penetrated the medial cortex of the pedicle; forty-nine (1.0 percent), the lateral cortex; eleven (0.2 percent), the superior cortex; and twenty-nine (0.6 percent), the inferior cortex.

Complications

Pain
The prevalence of complications directly related to the use of the pedicle screws was 2.4 percent (115 of 4790 screws). The most common postoperative problem was late-onset discomfort or pain due to a pseudarthrosis or perhaps to the instrumentation, necessitating removal of the instrumentation (with or without repair of the pseudarthrosis) and reinstrumentation. This complication was associated with 1102 (23.0 percent) of the screws, used in 222 (24.3 percent) of the procedures. A pseudarthrosis was found during forty-six of the procedures and a solid fusion, during the remaining 176.

Nerve-Root Irritation
Eleven screws (0.2 percent), used in nine procedures (1.0 percent), resulted in irritation of a nerve root. This complication was associated with seven of nineteen screws placed medially, three (10 percent) of twenty-nine screws placed inferiorly, and one (1 percent) of 134 screws placed anteriorly. The patients had radicular pain or weakness postoperatively. The diagnosis of a misplaced screw was made on the basis of postoperative radiographs or a myelogram followed by computed tomography. In seven patients (nine screws), exploration (performed on the first, second, third, seventh, tenth, twenty-fifth, or seventy-fifth day after insertion of the screw) showed that the nerve was displaced by the screw thread but was not transfixed or lacerated by the screw. Eight of the screws were removed, and the neurological problems and findings resolved after removal of seven of them (six of which had been placed medially and one, inferiorly). Even though the eighth screw, which had been placed inferiorly, was removed on the day of the operation, the patient still had marked residual weakness. One additional screw, which had been placed anteriorly in the sixth lumbar vertebra and was causing a definite radiculopathy, was replaced with a shorter screw, but the weakness persisted.

In two of the nine patients who had postoperative nerve-root irritation, the postoperative weakness was so minor that it was decided to observe the patient rather than to remove the screw. The mild weakness resolved in one patient, who had an inferiorly placed screw; however, the other patient, who had a medially placed screw, had persistent peroneal weakness on manual muscle-testing, with a normal gait. At the latest follow-up evaluation, three patients (0.3 percent of the patients, 0.3 percent of the procedures, and 0.1 percent of the screws) had residual neurological weakness.

Problems with Insertion of the Screw
Sixty-five (1.4 percent) of the screws, used in fifty-six (6.1 percent) of the procedures, cut out of the pedicle during insertion. Thirty-six of these screws, used in thirty-two procedures, were safely reinserted, and the redirected screw was identified as being within the pedicle. Twenty-seven screws, used in twenty-four procedures, could not be safely reinserted, and thus no screw was inserted at that level. Two screws that were seen to be positioned laterally on an intraoperative radiograph were secure and were therefore not removed. The data on intraoperative problems were obtained from the operative notes, when they included such information. However, we realize that such problems may not always be included in the operative notes. It is also possible that a screw that was not positioned completely within the pedicle will not be noted to be misplaced if there are no intraoperative or postoperative problems.

In addition, two screws, used in two procedures, were seen to be outside the pedicle on the postoperative radiographs; nonetheless, there were no neurological problems. One of these screws was believed to have been placed inferiorly in the fifth lumbar vertebra during an arthrodesis from the fourth lumbar to the first sacral vertebra. An exploration was performed immediately, and it showed the screw to be laterally placed; the screw was successfully reinserted into the pedicle. The second screw, which had been used in a repair of a lumbosacral pseudarthrosis, was seen to be laterally placed on the postoperative radiograph. An exploration was performed immediately, and the screw was removed.

Fracture of the Screw
Twenty-five (0.5 percent) of the screws, used in twenty (2.2 percent) of the procedures, fractured. These included thirteen (1.1 percent) of the 1227 Steffee VSP screws, eleven (0.3 percent) of the 3369 Cotrel-Dubousset screws, and one (2 percent) of the sixty-four PWB screws. Ten of the eleven broken Cotrel-Dubousset screws were of the closed sacral screw design. Screws fractured after twelve lumbosacral arthrodeses (eight done for the treatment of isthmic spondylolisthesis; two, for degenerative spondylolisthesis; and two, for degenerative disc disease), after five procedures for the treatment of a burst fracture (two posterior arthrodeses and three operations involving an anterior decompression and arthrodesis followed by a posterior arthrodesis and instrumentation), and after three operations in which an arthrodesis that had been previously performed for the treatment of scoliosis was extended to the sacrum. Of the twenty procedures, thirteen—six of the twelve lumbosacral arthrodeses, four of the five procedures for the treatment of a burst fracture, and all three procedures involving extension of the fusion to the sacrum—were followed by pseudarthrosis.

In all of the patients who had a burst fracture, the fracture of the screw permitted collapse of the burst fracture and, in one patient, the fusion became solid after the collapse. All thirteen patients with a pseudarthrosis had pain and were managed with exploration, repair of the pseudarthrosis, and reinsertion of the instrumentation. Three of the remaining seven patients who had a broken screw were found to have a solid fusion on exploration, and the instrumentation was removed; the other four patients had radiographic evidence of a solid fusion and had no low-back pain. In the last two years of the study, a new screw design and attention to biomechanical principles resulted in only one broken screw; that screw had been used in an arthrodesis that extended over a major portion of the spine to the sacrum.

Bending of the Screw
Six screws (0.1 percent), used in four procedures (0.4 percent), bent; four were closed Cotrel-Dubousset screws, and two were Steffee VSP screws. Two procedures for the treatment of spondylolisthesis resulted in a pseudarthrosis with a bent screw on one side and a broken screw on the other. The other two procedures were followed by fusion with the collapse allowed by bilateral bent screws; one was a lumbosacral arthrodesis with the bilateral bent screws at the first sacral level, and the other was a procedure for the treatment of a burst fracture of the second lumbar vertebra.

Fracture of the Pedicle
Three (0.1 percent) of the screws, used in two (0.2 percent) of the procedures, were associated with a pedicle fracture. One of these screws had been inserted in the third lumbar vertebra under direct visualization of the pedicle. A small crack was seen in the pedicle wall without evidence of any bone or screw thread impinging on the nerve. The screw was left in, and there were no postoperative symptoms. The other two screws had been used in a patient in whom a previous arthrodesis from the fourth lumbar vertebra to the sacrum had been extended to the third lumbar vertebra. Postoperatively, the patient had pain in the groin, and a myelogram and computed tomography scan raised suspicions that the screw in the third lumbar vertebra was in a medial position. Exploration and a laminotomy revealed that the third and fourth lumbar pedicles on the left had a two-to-three-millimeter break in the medial wall, with no screw thread visible at either level. The pedicle wall was smoothed, and the screws were not removed. The pain in the groin resolved postoperatively.

Dural Tear
A dural tear was seen in association with four (0.1 percent) of the screws, used in four (0.4 percent) of the procedures. In three of the procedures, the medial wall of the pedicle was perforated during sounding of the pedicle. Exploration revealed a dural tear, which was repaired during two of the procedures; however, repair was not possible during the third procedure. In the fourth procedure, the dural leak occurred with tapping of the pedicle, and additional probing revealed a small break in the medial wall. A small piece of Gelfoam (Upjohn, Kalamazoo, Michigan) was placed in the pedicle without inserting a screw at this level. None of these tears were associated with postoperative collection of spinal fluid or any wound problems.

Vascular and Visceral Problems
There were no vascular or visceral injuries associated with the pedicle screws in the present series.


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
Pedicle screws currently form an integral part of the armamentarium of the spinal surgeon. The screws provide secure fixation of the vertebra and can be used in the treatment of numerous disorders and deformities of the spine. However, there are still many controversies regarding the use of pedicle screws, and complications are still encountered. The complications are often related to the misuse of the device and the decision-making skills of the surgeon or they are directly related to the screws themselves. In the present study, we addressed this last issue and we demonstrated a very low complication rate when the screws had been inserted by experienced surgeons.

The exact location of a screw in a pedicle is difficult to determine; in experiments on cadavera, there was a high rate of failure in identifying the position on radiographs27. In that study, two surgeons inserted 124 screws from the eleventh thoracic to the first sacral vertebra in eight cadaveric specimens with the use of fluoroscopic control. The positions of the screws were assessed on radiographs, after which the specimen was transected and the true positions of the screws were noted. The positions of 21 percent of the screws were inaccurately predicted on the basis of the radiographs. In the present series, one screw that had initially been believed to be inferior on the basis of the postoperative radiograph and that was not associated with a neurological deficit was found, on exploration, to be lateral. The position of the screw on the film cassette made it difficult to diagnose the exact site of penetration. A high rate of penetration of the canal due to the metal artifact is seen on computed tomography scans; Gertzbein and Robbins12 found that 10 percent of 167 screws penetrated the canal by more than two millimeters. In the literature, the highest rates of misplaced screws were, to our knowledge, in that series and in the study by Luque21, who reported that, in his first-year experience, 18.3 percent of 327 screws were misplaced.

We found multiple radiographs to be necessary to adequately assess the position of the screws, especially with screws at multiple levels. Positioning in the medial-lateral direction was best assessed in the coronal plane, which showed the relationship of the screw to the pedicle shadow and the direction of the screw. The axial cuts of computed tomography scans, when available, gave additional information in this plane. The superior-inferior position was seen on the sagittal radiograph centered over the screw, and additional information was obtained from oblique radiographs. Penetration of the vertebral body anteriorly or anterolaterally was seen on sagittal and oblique radiographs. Some anterior penetrations are best seen on computed tomography scans.

We have not performed a study to validate our technique for assessing the accuracy of screw placement. In addition, routine computed tomography scans were not made for all of our patients; they were made only to verify the positions of the screws in the presence of back pain or radicular symptoms. Routine radiographs showed most of the screws to be definitely within the pedicle, and subsequent imaging studies confirmed this finding. Because of the multiple radiographs that were assessed for each patient, the possibility that all of the screws were not on the central beam, and the lack of routine computed tomography scans, it is probable that there were misplaced screws that were asymptomatic and were not detected. These screws would change the complication rates. However, we believe that they were not clinically relevant and do not alter the conclusions of our study.

In our series, 242 (5.1 percent) of the screws were not wholly contained within the pedicle or the vertebral body. One hundred and thirty-four (2.8 percent) of the screws penetrated the anterior vertebral cortex, the most common type of perforation. Most (ninety-nine; 74 percent) of these screws were in the sacrum, where slight penetration of the anterior cortex was deliberately performed to increase the screw purchase. It was impossible to determine from the records what percentage of the anterior penetrations were deliberate and what percentage were inadvertent. Only one anteriorly placed screw, which extended six millimeters anterior to the cortex of the sixth lumbar vertebra, caused nerve-root irritation, necessitating replacement of the screw.

The most common postoperative problem in our series was late-onset discomfort or pain, which either was due to a pseudarthrosis or perhaps was related to the instrumentation. This complication occurred after 24.3 percent (222) of the procedures and involved 23.0 percent (1102) of the screws. The rates of screw removal rarely have been noted in the literature. Steffee and Brantigan25, in the initial report on the use of the VSP system, removed the instrumentation from 24.0 percent of 250 patients. This rate was similar to the 26.0 percent rate reported by Blumenthal and Gill2 in their description of their initial experience with the Wiltse system, in 470 patients. Both rates are similar to that in the present series.

The rate of complications directly related to the pedicle screws was 2.4 percent (115 of the 4790 screws) in the present series. Difficulties or problems involved with insertion of pedicle screws have not been mentioned in the literature, to our knowledge. Sixty-five (1.4 percent) of the screws, used in fifty-six (6.1 percent) of the procedures, were noted to be malpositioned intraoperatively, and more than half of these screws could be reinserted within the pedicle. If a screw is not within the pedicle intraoperatively, reinsertion should be attempted, and a screw should not be inserted if the pedicle cannot be safely and securely entered.

Because neurological problems are the most feared complications of any spinal procedure, they are also the most commonly reported complications in the literature; however, they are usually reported in relation to the number of patients or procedures and rarely in relation to the number of screws that were inserted. No neurological problems were reported in three series; two of these series had less than thirty patients1,20 and one, reported by Daniaux et al.4, had 243 patients. The lowest reported rate of nerve-root irritation was approximately 1 percent, to our knowledge; specifically, Davne and Myers5 reported a rate of 1.1 percent after 533 procedures with the Steffee VSP system. They attributed half of the complications to the screws and half to the posterior lumbar interbody arthrodesis that was performed simultaneously. In the literature as a whole, the rate of neurological complications has ranged from 1.1 to 10 percent, with a wide range in the sizes of the series (twenty-one to 617 patients or procedures)1,4,5,10,12,14-16,18-20,23-26,28,29. In four series with at least 250 patients or procedures (533, 617, 440, and 250), the rate ranged from 1.1 to 4.7 percent5,10,19,25. Four reports gave the rate of nerve-root injury in relation to the number of screws. Gertzbein and Robbins12 reported a rate of 1 percent in a study involving 167 screws; Mac Millan et al.22, a rate of 0.4 percent in a study involving 268 screws; West et al.28, a rate of 0.3 percent in a study involving 783 screws; and Luque21, a rate of 0.3 percent in a study involving 1268 screws. The rate in our review, whether related either to the number of patients or to the number of screws, is the lowest reported, as far as we know.

Screws that are malpositioned medially or inferiorly are the ones that place the nerve at risk as it courses around the pedicle28. Our study confirmed this: ten of the eleven screws irritating a nerve root had been positioned either medially or inferiorly, and more medially malpositioned screws (seven of nineteen) than inferiorly malpositioned screws (three of twenty-nine; 10 percent) caused nerve-root irritation. One (0.7 percent) of the screws that perforated the anterior cortex caused nerve-root irritation, a rare occurrence with this screw position. Prompt diagnosis and removal of the screw gives the best chance of resolution of neurological weakness. Only one of eight patients from whom the screw had been removed had a persistent marked residual neurological deficit. Two others had mild residual weakness. One had an anteriorly placed screw that was replaced three days postoperatively, and the other had a medially placed screw with mild weakness that was kept under observation; the weakness failed to resolve, affecting function. Another patient, with an inferiorly placed screw, had a mild neurological deficit that resolved spontaneously.

Fracture or breakage of screws is well reported in the literature and has ranged from 1 to 11.2 percent of the screws inserted5,8,11,21,22,25,28. Most authors have reported the rate in relation to the number of patients in the series, with the rates ranging from 0.4 to 24.5 percent2-7,9,10,14,16-19,23,25,26,29,30. Screw breakage is related to three factors: the design of the screw, the presence of a pseudarthrosis, and use in burst fractures. More screws broke in the earlier years of our study, and only one broke in the last two years. This finding is similar to the high rates of breakage reported in the literature for the initial series of surgeons. Steffee and Brantigan25 reported that screws broke in 12.4 percent of 250 patients and that 2.5 percent of 1314 screws broke. Whitecloud et al.29 reported breakage in 18 percent of their first forty patients, and Zucherman et al.30 reported breakage in 25 percent of their first seventy-seven patients. Other than the high rates of breakage reported in relation to the initial experience of the surgeons, no other factors associated with breakage can be deduced from the literature.

A high percentage of the broken screws were related to a pseudarthrosis. Twelve of nineteen patients who had a broken screw had a pseudarthrosis. Broken screws were seen after arthrodeses that had been performed for scoliosis were extended to the sacrum and after four of the five procedures for the treatment of burst fractures. When a broken screw is seen, a pseudarthrosis should be suspected until it can be ruled out. Routine radiographs in the coronal, sagittal, and oblique planes with the central x-ray beam centered on the level of the breakage should be used and should be supplemented with thin (one-millimeter) computed tomography cuts with coronal and sagittal reconstruction when necessary. When there is doubt about whether a patient with pain in the area of the broken pedicle screw has a pseudarthrosis, the fusion should be explored and the instrumentation should be removed. Exploration is not necessary when there is a broken screw but the patient has no pain and has a radiographically solid fusion.

The burst fractures with broken screws were interesting, as they all had anterior collapse with loss of anterior column support. Only one went on to fuse; the other four had a pseudarthrosis. When burst fractures are treated with pedicle screws, the anterior support is important for the biomechanical integrity of the construct.

Bent screws are unusual and have rarely been reported. Gurr and McAfee13 reported bent screws after 3 percent of 100 procedures. A rate of 2.6 percent (seven) of 268 screws in fifty patients was reported by Mac Millan et al.22, and a rate of 0.5 percent of 198 screws in forty-eight patients was reported by Esses7. In our series, four patients had a total of six bent screws. Two had bent screws on both sides of a solid fusion, and the other two, who had a bent screw on one side and a broken screw on the other, had an associated pseudarthrosis. Thus, when a bent or broken screw is seen, a pseudarthrosis should be suspected.

Fracture of the pedicle has also been rarely reported, with a rate of 0.4 percent of 470 patients reported by Blumenthal and Gill2, 2.5 percent of 243 patients reported by Daniaux at al.4, and 2.7 percent of 617 patients (based on the American Back Society survey) reported by Esses et al.10. The rate of postoperative fracture of the pedicle in our series was very low; it occurred after only two procedures (0.2 percent) and involved only three screws (0.1 percent). These fractures were in addition to the intraoperative fractures of the pedicle associated with insertion of the screws and with the fractures associated with extrusion of the screws from the pedicle.

In the present study, we found that a very low percentage (5.1 percent) of the pedicle screws were inserted outside the pedicle and very few patients (three; 0.3 percent) had permanent nerve-root injury due to screw insertion.


   Footnotes
 
*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.

{dagger}Twin Cities Spine Center, 913 East 26th Street, Suite 600, Minneapolis, Minnesota 55404.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Aebi, M.; Etter, C.; Kehl, T.; and Thalgott, J.: The internal skeletal fixation system. A new treatment of thoracolumbar fractures and other spinal disorders. Clin. Orthop., 227: 30-43, 1988.[Medline]
  2. Blumenthal, S., and Gill, K.: Complications of the Wiltse pedicle screw fixation system. Spine, 18: 1867-1871, 1993.[Medline]
  3. Cotrel, Y.; Dubousset, J.; and Guillaumat, M.: New universal instrumentation in spinal surgery. Clin. Orthop., 227: 10-23, 1988.[Medline]
  4. Daniaux, H.; Seykora, P.; Genelin, A.; Lang, T.; and Kathreim, A.: Application of posterior plating and modifications in thoracolumbar spine injuries. Indications, technique, and results. Spine, 16(3S): 125-S133, 1991.
  5. Davne, S. H., and Myers, D. L.: Complications of lumbar spinal fusion with transpedicular instrumentation. Spine, 17(6S): 184-S189, 1992.
  6. Dick, W.: The "fixateur interne" as a versatile implant for spine surgery. Spine, 12: 882-900, 1987.[Medline]
  7. Esses, S. I.: The AO spinal internal fixator. Spine, 14: 373-378, 1989.[Medline]
  8. Esses, S. I., and Bednar, D. A.: The spinal pedicle screw: techniques and systems. Orthop. Rev., 18: 676-682, 1989.[Medline]
  9. Esses, S. I.; Botsford, D. J.; Wright, T.; Bednar, D.; and Bailey, S.: Operative treatment of spinal fractures with the AO internal fixator. Spine, 16(3S): 146-S150, 1991.[Medline]
  10. Esses, S. I.; Sachs, B. L.; and Dreyzin, V.: Complications associated with the technique of pedicle screw fixation. A selected survey of ABS members. Spine, 18: 2231-2238, 1993.[Medline]
  11. Gertzbein, S.; Robbins, S.; Schwartz, M.; Chow, M.; Fazl, M.; and Rowed, D.: Preliminary results of the AO internal fixator for spinal fractures. Orthop. Trans., 13: 752-753, 1989.
  12. Gertzbein, S. D., and Robbins, S. E.: Accuracy of pedicular screw placement in vivo. Spine, 15: 11-14, 1990.[Medline]
  13. Gurr, K. R., and McAfee, P. C.: Cotrel-Dubousset instrumentation in adults. A preliminary report. Spine, 18: 510-520, 1988.
  14. Guyer, D. W.; Wiltse, L. L.; and Peek, R. D.: The Wiltse pedicle screw fixation system. Orthopedics, 11: 1455-1460, 1988.[Medline]
  15. Hehne, H.-J.; Zielke, K.; and Böhm, H.: Polysegmental lumbar osteotomies and transpedicle fixation for correction of long-curved kyphotic deformities in ankylosing spondylitis. Report of 177 cases. Clin. Orthop., 258: 49-55, 1990.
  16. Horowitch, A.; Peek, R. D.; Thomas, J. C., Jr.; Widell, E. H., Jr.; DiMartino, P. P.; Spencer, C. W., III; Weinstein, J.; and Wiltse, L. L.: The Wiltse pedicle screw fixation system. Early clinical results. Spine, 14: 461-467, 1989.[Medline]
  17. Hsu, K.; Zucherman, J.; and Krag, M.: Complications of transpedicle spine fixation. In The Lumbar Spine, edited by S. W. Wiesel, J. N. Weinstein, H. N. Herkowitz, J. Dvorák, and G. R. Bell. Ed. 2, vol. 2, pp. 1203-1216. Philadelphia, W. B. Saunders, 1996.
  18. Kinnard, P.; Ghibely, A.; Gordon, D.; Trias, A.; and Basora, J.: Roy-Camille plates in unstable spinal conditions. A preliminary report. Spine, 11: 131-135, 1986.[Medline]
  19. Louis, R.: Fusion of the lumbar and sacral spine by internal fixation with screw plates. Clin. Orthop., 203: 18-33, 1986.
  20. Lowe, T. G.: Zielke pedicular instrumentation and fusion for instability of the lumbosacral spine. Orthop. Trans., 12: 665-666, 1988.
  21. Luque, E. R.: Complications of interpeduncular correction and fixation. Orthop. Trans., 12: 238-239, 1988.
  22. Mac Millan, M. M.; Cooper, R.; and Haid, R.: Lumbar and lumbosacral fusions using Cotrel-Dubousset pedicle screws and rods. Spine, 19: 430-434, 1994.[Medline]
  23. Matsuzaki, H.; Tokuhashi, Y.; Matsumoto, F.; Hoshino, M.; Kiuchi, T.; and Toriyama, S.: Problems and solutions of pedicle screw plate fixation of lumbar spine. Spine, 15: 1159-1165, 1990.[Medline]
  24. Roy-Camille, R.; Saillant, G.; and Mazel, C.: Internal fixation of the lumbar spine with pedicle screw plating. Clin. Orthop., 203: 7-17, 1986.
  25. Steffee, A. D., and Brantigan, J. W.: The variable screw plate spinal fixation system. Report of a prospective study of 250 patients enrolled in Food and Drug Administration clinical trials. Spine, 18: 1160-1172, 1993.[Medline]
  26. Thalgott, J. S.; LaRocca, H.; Aebi, M.; Dwyer, A. P.; and Razza, B. E.: Reconstruction of the lumbar spine using AO DCP plate internal fixation. Spine, 14: 91-95, 1989.[Medline]
  27. Weinstein, J. N.; Spratt, K. F.; Spengler, D.; Brick, C.; and Reid, S.: Spinal pedicle fixation: reliability and validity of roentgenogram-based assessment and surgical factors on successful screw placement. Spine, 13: 1012-1018, 1988.[Medline]
  28. West, J. L., III; Ogilvie, J. W.; and Bradford, D. S.: Complications of the variable screw plate pedicle screw fixation. Spine, 16: 576-579, 1991.[Medline]
  29. Whitecloud, T. S., III; Butler, J. C.; Cohen, J. L.; and Candelora, P. D.: Complications with the variable spinal plating system. Spine, 14: 472-476, 1989.[Medline]
  30. Zucherman, J.; Hsu, K.; White, A.; and Wynne, G.: Early results of spinal fusion using variable spinal plating system. Spine, 13: 570-579, 1988.[Medline]

Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Facebook Facebook   Add to Technorati Technorati   Add to Twitter Twitter    What's this?


This article has been cited by other articles:


Home page
 J Am Acad Orthop SurgHome page
E. Cuartas, A. Rasouli, M. O'Brien, and H. L. Shufflebarger
Use of All-pedicle-screw Constructs in the Treatment of Adolescent Idiopathic Scoliosis
J. Am. Acad. Ortho. Surg., September 1, 2009; 17(9): 550 - 561.
[Abstract] [Full Text] [PDF]

Home page
 RadioGraphicsHome page
P. M. Young, T. H. Berquist, L. W. Bancroft, and J. J. Peterson
Complications of Spinal Instrumentation
RadioGraphics, May 1, 2007; 27(3): 775 - 789.
[Abstract] [Full Text] [PDF]

Home page
 JBJSHome page
D. A. Bednar, T. A. Einhorn, and P. C. McAfee
Postoperative Pain in Research Animals, and Use of Interbody Fusion Cages
J. Bone Joint Surg. Am., February 1, 2001; 83(2): 293-b - 293.
[Full Text]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Letters to the Editor: Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when Letters to the Editor are posted
Right arrow Alert me if a correction is posted
Services
Right arrow E-mail this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My File Cabinet
Right arrow Download to citation manager
Right arrow Rights and Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by LONSTEIN, J. E.
Right arrow Articles by WINTER, R. B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by LONSTEIN, J. E.
Right arrow Articles by WINTER, R. B.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us   Add to Facebook   Add to Technorati   Add to Twitter  
What's this?