viernes, 22 de octubre de 2010

Subsequent Vertebral Fractures following Spinal Fusion Surgery for Degenerative Lumbar Disease: A Mean Ten-year Follow-u

rom Spine

Subsequent Vertebral Fractures following Spinal Fusion Surgery for Degenerative Lumbar Disease: A Mean Ten-year Follow-up

Tomoaki Toyone, MD; Tomoyuki Ozawa, MD; Koya Kamikawa, MD; Atsuya Watanabe, MD; Keisuke Matsuki, MD; Takeshi Yamashita, MD; Ryutaro Shiboi, MD; Masato Takeuchi, MD; Yuichi Wada, MD; Kunimasa Inada, MD; Yasuchika Aoki, MD; Gen Inoue, MD; Seiji Ohtori, MD; Tadashi Tanaka, MD

Authors and Disclosures

Abstract and Introduction
Materials and Methods
Results
Discussion

Abstract and Introduction

Abstract

Study Design. Case-control study.
Objective. To assess the long-term prevalence of vertebral fractures after lumbar spinal fusion with instrumentation.
Summary of Background Data. The incidence of the adjacent and the nonadjacent, remote level subsequent vertebral fractures after lumbar spinal fusion is not well described in the literature.
Methods. The study is a retrospective analysis of 100 consecutive patients of 55 years of age or older with spinal fusion for degenerative diseases between L1 and S1, and instrumentation for less than 4 segments. Patients with prevalent vertebral fractures defined at the time of surgery, or patients with secondary causes of osteoporosis were excluded. Mean follow-up period was 10.2 years (range, 7–14 years). Acute vertebral fractures were determined by magnetic resonance imaging and lateral spine radiographs.
Results. Acute vertebral fractures were determined in 20 vertebrae in 14 (24%) of the 59 female patients, whereas 1 male patient (2%) had 1 vertebral fracture during the follow-up period. Eighteen of the 21 fractures occurred within 2 years of the spinal instrumentation surgery. Regarding time to fracture occurrence after surgery, adjacent level fractures occurred within 8 months, and remote level fractures occurred between 8 and 22 months after surgery.
Conclusion. Postmenopausal female patients who underwent lumbar spinal instrumentation surgery were susceptible to develop subsequent vertebral fractures within 2 years after surgery. The greater the number of spinal segments between the fracture and the instrumentation was, the longer the time after surgery.

Introduction

The use of spinal fusion surgery is rapidly increasing.^[1] The rational for spinal fusion includes prevention of painful motion and correction of deformity. Much of the increase in use has been seen in older adults, in association with laminectomy for spinal stenosis.^[2]

It is not surprising that fusion is associated with more complications than those with other types of spinal surgery. Common complications include instrumentation failure (occurring in about 7% of cases), bone graft donor site complications (11%), neural injuries (3%), pulmonary embolus (2%), infections (3%), and pseudarthrosis (15%).^[3]

Any abnormal process that develops in the mobile segment next to a spinal fusion is also known as adjacent segment diseases. In some articles, the term adjacent segment disease also refers to changes that occur at segments more proximal or distal than the one immediately next to a fusion. On the basis of a MEDLINE search using key words "adjacent," "transition zone," or "postoperative complication" combined with "spinal fusion," for the years between 1966 and 2002, Park et al reviewed 56 articles regarding adjacent segment diseases after lumbar or lumbosacral fusion.^[4] Common findings next to spinal fusion were disc degeneration, listhesis, instability, hypertrophic facet joint arthritis, herniated nucleus pulposus, and stenosis. Vertebral compression fractures were noted in only one article.^[5] Hart et al reported that proximal junctional acute collapse requiring revision surgery occurred in 2 of the 13 female patients older than 60 years who underwent lumbar fusions, and concluded that further efforts should be needed to determine more precisely the incidence of proximal junctional acute collapse and the effects of various risk factors on increasing this incidence.^[6]

This complication such as the incidence of the adjacent and the nonadjacent, remote level subsequent vertebral fractures after lumbar spinal fusion is not well described in the literature. Therefore, the purpose of this study was to assess the long-term prevalence of vertebral compression fractures after lumbar spinal fusion with rigid instrumentation.

Subsequent Vertebral Fractures Following Spinal Fusion Surgey: Materials and Methods

Authors and Disclosures

Physician Assistant Rating:

( 1 Vote )

Rate This Article:

Abstract and Introduction
Materials and Methods
Results
Discussion

Materials and Methods

We performed a retrospective analysis of 100 consecutive patients. Inclusion criteria were as follows: patients aged 55 years or older with spinal fusion between L1 and S1 and instrumentation for less than 4 segments for degenerative disease, surgery was performed between April 1993 and March 2002, and continuously followed for at least 7 years. Patients of age 55 years or older were selected so that premenopausal women should not be included in the study. Patients with prevalent vertebral fractures defined at the time of surgery were excluded. Hospital records were reviewed to exclude secondary causes of osteoporosis. Secondary causes of osteoporosis were those related to prescription of corticosteroids, endocrine disorders, neoplastic diseases, or gastrointestinal disorders. Patients whose bone mineral density (BMD) was less than 80% of the young adult mean values were also excluded. Furthermore, patients who were on medication for osteoporosis were excluded.

There were 59 women and 41 men, and the mean age at surgery was 67 (range, 55–80). Preoperative diagnosis was degenerative spondylolisthesis in 63, isthmic spondylolisthesis in 10, degenerative scoliosis in 19, and stenosis and other spondylotic conditions in 8 patients. Surgical procedure was posterolateral fusion in 57 and posterior lumbar interbody fusion (PLIF) in 43 patients, and all the patients underwent pedicle screw fixation. The level of instrumentation was L5–S1 in 13, L4–L5 in 38, L3–L4 in 9, L2–L3 in 2, L1–L2 in 1, L4–S1 in 12, L3–L5 in 11, L2–L4 in 2, L1–L3 in 1, L3–S1 in 4, L2–L5 in 5, and L1–L4 in 2 patients. Mean follow-up period was 10.2 years (range, 7–14 years).

Lateral spine radiographs obtained at baseline (just before surgery), at 3 and 6 months after surgery, and yearly after surgery to the final follow-up were used for morphometric vertebral fracture ascertainment. Six points were placed on each vertebra to define anterior, posterior, and mid heights. Prevalent vertebral fractures at baseline were defined as deformity on preoperative radiographs derived from quantitative morphometry. Asymptomatic subsequent vertebral fractures were defined as a new fracture with a decrease of more than 20% in any vertebral height (minimum, 4 mm) from baseline.^[7] Diagnostic criteria of symptomatic subsequent vertebral fractures was as follows: acute increase in back pain as a result of a fall from standing height or less or without any trauma, radiologic evidence of acute vertebral fractures determined by magnetic resonance imaging, showing geographic patterns of low-intensity-signal changes on T1-weighted images, and high-intensity-signal changes on T2-weighted images.^[8]

Statistical Analysis

Fisher exact or χ² tests were used for between-group comparisons. The Spearman correlation coefficient was calculated to test relationships between variables. The level of significance was set at P < 0.05. All statistical analyses were carried out using the statistical package SPSS 14.0 (SPSS Inc, Chicago, IL), Windows version.

Subsequent Vertebral Fractures Following Spinal Fusion Surgey: Results

Authors and Disclosures

Physician Assistant Rating:

( 1 Vote )

Rate This Article:

Results

Acute vertebral fractures were determined in 20 vertebrae in 14 (24%) of the 59 female patients, whereas 1 male patient (2%) had 1 vertebral fracture during the follow-up period. Subsequent vertebral fractures were recognized between T10 and L3, and they peaked at L1. Twenty of the 21 fractures were symptomatic and painful. One asymptomatic fracture was defined at a periodic medical check-up 8 years after surgery. The latest fracture occurred 11 years after surgery at the adjacent vertebra of the posterolateral fusion for degenerative scoliosis. With respect to the incidence of subsequent vertebral fracture, there were significant differences between female and male patients (P< 0.001). Although, PLIF, 3-level fusion, and low preoperative BMD were found to be frequent in patients with subsequent vertebral fractures, no variables such as surgical procedures (P = 0.15), the number of fusion levels (P= 0.30), preoperative and postoperative sagittal alignment (P = 0.44 and P = 0.39), and preoperative BMD (P = 0.12) were significantly related with the incidence of subsequent vertebral fractures.

Eighteen of the 21 fracture occurred within 2 years of the spinal instrumentation surgery (Table 1). Regarding time to fracture occurrence after surgery, adjacent level fractures occurred within 8 months, and remote level fractures occurred between 8 and 22 months after surgery (P < 0.001)(Figure 1). Subsequent fractures 2-levels away from the instrumentation level occurred between 8 and 12 months, fractures 3-levels away from the instrumentation level occurred between 8 and 20 months, and fractures 4-levels away from the instrumentation level occurred between 14 and 22 months after surgery (Table 2). The distance of the subsequent vertebral fracture from the instrumentation level was associated with time to subsequent fracture occurrence (adjacent with 2-levels, P = 0.01; 2-levels with 3-levels, P = 0.09; 3-levels with 4-levels, P = 0.06).

(Enlarge Image)

Discussion

Etebar and Cahill performed a retrospective analysis of 125 consecutive patients in whom instrumentation was placed to promote lumbar fusion for the treatment of degenerative instability.^[5]Eighteen of 125 patients developed symptomatic next-segment failure based on radiographic studies. The mean follow-up period was 44.8 months. Stress fracture of the adjacent vertebral body was observed in 5 patients, and they concluded that the risk appeared to be especially high in postmenopausal women. Recently, to investigate the morphologic features of adults after spinal deformity surgery, Watanabe et al reported 10 adult patients who underwent segmental spinal instrumented fusion and concluded that old age, osteopenia, preoperative comorbidities, and severe global sagittal imbalance were found to be frequent in patients with proximal junctional fracture.^[9] In the present study, PLIF, 3-levels fusion, and low preoperative BMD might be considered as a risk factor, but there was no significant difference in the incidence of subsequent vertebral fractures.

Published data on subsequent vertebral fractures after lumbar or thoracolumbar spinal instrumentation is scarce, but the literature for vertebroplasty and kyphoplasty reports a subsequent fracture rate of 12% to 52%.^[10–13]Fribourg et al undertook a retrospective review of 38 kyphoplasty patients.^[10] Within the mean follow-up period of 8 months, 10 patients sustained 17 subsequent vertebral fractures. They reported that 15 adjacent level fractures occurred within 60 days, whereas 2 remote level fractures occurred 181 and 559 days after kyphoplasty. Trout et al studied 432 patients who underwent vertebroplasty, and reported that 186 new vertebral fractures occurred in 86 patients.^[13] Adjacent level fractures occurred in a mean of 55 days after surgery, whereas remote level fractures occurred in a mean of 127 days after surgery. The study of kyphoplasty, the study of vertebroplasty, and our study concerning spinal fusion with rigid instrumentation have all come up with similar results showing that adjacent level vertebral fractures occurred sooner than remote level vertebral fractures. However, adjacent level vertebral fractures occurred within 2 months after kyphoplasty and vertebroplasty, whereas within 8 months after surgery in our study.

Several studies have been conducted addressing human BMD with respect to spinal fusion. An initial (6 months) decrease in the BMD at the level adjacent to the instrumented spine and fusion level has been observed. Bogdanffy et al studied 15 patients who underwent a combined anteroposterior L4–S1 spinal fusion as a first time operation, and assessed BMD by dual-energy radiograph absorptiometry.^[14] They reported that BMD at L3, 1 level above the fusion, and L2, 2 levels above, decreased significantly at 3 months and remained reduced at 6 months after surgery. This loss of BMD appears to be time-dependent and at a 1-year follow-up, the BMD at the level above the combined anterior-posterior instrumented fusion returned to or exceeded the preoperative level in 60% of the patients. The authors attributed these changes to postoperative immobilization and altered biomechanics secondary to the arthrodesis. An alternative explanation for the initial decrease in bone density may be related to Frost's concept of bone remodeling. This process has been termed as a regional accelerating phenomenon, which is reflected in the loss of bone density for up to 6 months to a year.

Only one study to our knowledge has evaluated vertebral BMD at longer periods of follow-up. Singh et al studied 7 patients who underwent a posterior lumbar spinal fusion with instrumentation.^[15] The BMD increased at a mean 10.8-year follow-up at all 3 adjacent cephalad levels when compared with a mean 4-year follow-up. There was a gradual decrease in BMD changes with increasing distance from the fusion level. The authors stated that the discectomy group did not demonstrate a significant change in vertebral BMD, supporting the theory of stress-concentration altering local bone metabolism resulting in bone deposition and a resultant increase in vertebral BMD.

These results, an initial decrease in the BMD at the level adjacent to the spinal instrumentation and fusion level and an increase at longer periods of follow-up, might explain a possible pathogenesis of subsequent vertebral fractures occurring within 2 years after spinal fusion surgery in this study. According to the results from the European Prospective Osteoporosis Study, the incidence increased markedly with age in both men and women.^[16] Our data also differs from this large population-based study.

In this investigation, we have intended to exclude several potential confounding factors, such as baseline BMD, secondary causes of osteoporosis, prevalent vertebral fractures, and medications for osteoporosis, which might reduce risk of fracture. Nevertheless, as for the frequency of fracture between male and female patients, female patients might suffer from osteoporosis more frequently in this population and that might explain the gender difference for the frequency of the subsequent fracture. We were not able to include bone markers into the analysis, which might be a potential problem of this study. To define the rate of subsequent vertebral fractures and their risk factors, further prospective investigations are crucial.

This is the first study to ascertain the influence of spinal instrumentation surgery for subsequent vertebral fractures in patients more than 55 years of age. Results of our study with a 7-year minimum follow-up suggest that postmenopausal female patients who underwent lumbar spinal instrumentation surgery were susceptible to develop subsequent vertebral fractures within 2 years after surgery. It could be concluded that the greater the number of spinal segments between the fracture and the instrumentation, the longer the time after surgery.