Product Notes and Case Studies

The Biomechanical Effect of Postoperative Hypolordosis in Instrumented Lumbar Fusion on Instrumented and Adjacent Spinal Segments

Shinji Umehara, MD*† · Michael R. Zindrick, MD *‡ · Avinash G. Patwardhan, PhD*
Robert M. Havey, BS · Lori A. Vrbos, MS* · Gary W. Knight, MS* · Suichi Miyano, MD*§
Marie Kirincic, MD‡ · Kiyoshi Kaneda, MD† · Mark A. Lorenz, MD*‡
SPINE Volume 25, Number 13, pp 1617–1624
©2000, Lippincott Williams & Wilkins, Inc.

Study Design

Change in lumbar lordosis was measured in patients that had undergone posterolateral lumbar fusions using transpedicular instrumentation. The biomechanical effects of postoperative lumbar malalignment were measured in cadaveric specimens.

Objectives

To determine the extent of postoperative lumbar sagittal malalignment caused by an intraoperative kneeling position with 90° of hip and knee flexion, and to assess its effect on the mechanical loading of the instrumented and adjacent segments.

Summary of Background Data

The importance of maintaining the baseline lumbar lordosis after surgery has been stressed in the literature. However, there are few objective data to evaluate whether postoperative hypolordosis in the instrumented segments can increase the likelihood of junctional breakdown.

Methods

Segmental lordosis was measured on preoperative standing, intraoperative prone, and postoperative standing radiographs. In human cadaveric spines, a lordosis loss of up to 8° was created across L4–S1 using calibrated transpedicular devices. Specimens were tested in extension and under axial loading in the upright posture.

Results

In patients who underwent L4–S1 fusions, the lordosis within the fusion decreased by 10° intraoperatively and after surgery. Postoperative lordosis in the proximal (L2–L3 and L3–L4) segments increased by 2° each, as compared with the preoperative measures. Hypolordosis in the instrumented segments increased the load across the posterior transpedicular devices, the posterior shear force, and the lamina strain at the adjacent level.

Conclusions

Hypolordosis in the instrumented segments caused increased loading of the posterior column of the adjacent segments. These biomechanical effects may explain the degenerative changes at the junctional level that have been observed as long-term consequences of lumbar fusion. [Key words: adjacent segment, biomechanics, fusion, instrumentation, junctional breakdown, lumbar lordosis, lumbar spine, sagittal alignment]
Spine 2000;25:1617–1624


With improved fusion rates from the use of internal fixation devices, other causes of postoperative pain and disability have become apparent. Breakdown of the adjacent level (transitional zone) has been identified as one such cause clinically.7,9,12,14,17,18 However, the mechanisms of transitional zone breakdown are not completely understood. Factors that have been implicated include rigid fixation, number of levels fused, health of the adjacent level, and postoperative lumbar sagittal malalignment. The altered mechanical environment after lumbar fusion with instrumentation is considered to induce or accelerate degenerative changes at the adjacent levels. Previous in vivo human and animal studies, ex vivo cadaveric studies, and finite element models have demonstrated that the motion and stresses in the adjacent segments (proximal or distal to the fusion) are altered in the presence of instrumentation and fusion.4,5,13,15,26,27

An instrumented fusion can alter the mechanics of the lumbar spine in at least two ways. First, the elimination of one or more mobile segments resulting from a rigid construct may induce abnormal motion and “stress concentration” at the adjacent mobile segment caused by an attempt by the patient to maintain the preinstrumentation range of motion. Second, a postoperative lumbar sagittal malalignment may induce nonphysiologic loading of the adjacent segments. In an attempt to maintain the normal sagittal posture, the patient with a loss of lumbar lordosis may increase stresses on the posterior column.

Indeed, the importance of maintaining the baseline lumbar lordosis after surgery has been stressed in the recent clinical literature. Sagittal alignment of the human spine and the postural changes in alignment have been studied radiographically.1,2,6,10 The differences in lumbar sagittal alignment when various spinal surgical tables are used also have been evaluated.8,16,22,25 For posterior lumbar spine surgery, the kneeling position on a spinal surgical frame has been advocated for enhanced exposure of the spinal elements and a decrease in abdominal pressure and blood loss.3,24 However, the kneeling position has been shown to cause a significant loss of lumbar lordosis intraoperatively.8,16,22,25 During fusion surgery, the application of rigid instrumentation to the intraoperative flat back may cause residual hypolordosis in the instrumented segment after surgery. To maintain normal posture and an optimal center of gravity after surgery, the segments proximal to the instrumentation may be required to compensate for the changes in spinal alignment. Therefore, in theory, postoperative lumbar malalignment may accelerate adjacent segment deterioration by loading the motion segment in a nonphysiologic fashion.

The loss of lordosis in the instrumented segments not only may affect the adjacent segments, but also the load on the posterior spinal implant may increase. In the case of instrumented lumbar segments with baseline lordosis, an extension moment may be counterbalanced by a compressive load in the posterior implant as well as tension loads in the anterior longitudinal ligament and the anterior anulus fibrosis. With a loss of lumbar lordosis in the instrumented segments, the tension in the anterior soft tissue structures may be diminished, thus increasing the compressive load on the posterior implant and its interface with the vertebrae.

This study investigated whether a change in the baseline lumbar sagittal alignment can significantly alter the mechanical loading of the lumbar spine. First, the intraoperative lumbar curvature and pre- and postoperative standing alignment were analyzed in patients who underwent instrumented posterolateral lumbar fusion to assess the extent of iatrogenic sagittal malalignment. Second, the lumbar malalignment was experimentally simulated in cadaveric specimens to assess its mechanical effect on the instrumented and adjacent motion segments.

This two-part study addressed the following questions:

  1. What is the extent of the intraoperative loss of lumbar lordosis (relative to the preoperative standing lordosis) induced by the intraoperative kneeling position on the surgical table with 90° of hip and knee flexion?
  2. What effect does the application of instrumentation in the intraoperative kneeling position have on postoperative lordosis of the instrumented segments and the segments adjacent to instrumentation?
  3. Can a loss of lordosis in the instrumented segments significantly increase the load on the posterior spinal implant?
  4. Can the postoperative changes in lumbar sagittal alignment cause significant changes in the mechanical loading of the adjacent segment?

Methods


Clinical Radiographic Analysis

To investigate the lordosis changes in patients who underwent instrumented posterolateral lumbar fusion surgery, the preoperative standing, intraoperative prone, and postoperative standing radiographs were analyzed retrospectively. The clinical record and radiographs for a consecutive series of 288 patients surgically treated by two senior orthopedic surgeons from January 1990 to July 1994 were reviewed. The inclusion criteria for the preoperative diagnosis required segmental instability, radiculopathy, or degenerative lumbar disease, and the procedures were considered first-time patient surgeries. Patients without appropriate sets of the preoperative standing, intraoperative prone, and postoperative standing radiographs were eliminated.

A total of 62 patients who had undergone posterolateral lumbar fusions using transpedicular instrumentation were included in this radiographic study. The patients were 24 women and 38 men ranging in age from 22 to 74 years (mean, 44.8 years). The spinal levels spanned by the instrumentation were L3–L4 in 2 patients, L3–S1 in 4 patients, L4–L5 in 18 patients, L4–S1 in 30 patients, and L5–S1 in 8 patients. The type of instrumentation was variable screw placement (DePuy AcroMed, Raynham, MA) in 37 patients, Wiltse (Advanced Spine Fixation Systems, Cypress, CA) in 19 patients, and Isola (DePuy AcroMed) in 6 patients. At surgery, patients were positioned with 90° of hip and knee flexion on the Andrews Spinal Surgery Frame (Orthopedic Systems Inc., Union City, CA).

On three separate occasions, three independent observers (two orthopedic surgeons and one general practitioner) digitized the vertebral bodies on the radiographs. The selection of the four apical points per vertebral body was observer dependent. The span of time between measurements was no less than 14 days and no more than 30 days. Intrarater reliability was found to be quite consistent, with coefficients ranging from 0.80 to 0.92. Interrater reliability, or the degree to which different raters, operating independently, assigned similar scores, ranged from a low of 0.39 to a high of 0.61. The reliability estimates for the two raters with an orthopedic background were largely similar (0.59–0.61), and data measured by these two raters were used for subsequent statistical analysis.

The data were analyzed using SPSSx statistical software (SPSS, Inc., Chicago, IL). The amount and distribution of missing data, nonnormal variables, and outliers were evaluated. Required assumptions for repeated measures analysis of variance (ANOVA) also were reviewed. A conservative Greenhouse- Geisser estimate was used to adjust for the possibility of a positively biased F ratio. A Bonferroni correction was applied during the performance of subsequent post hoc comparisons. Analysis of covariance (ANCOVA) was used because of the nonrandomization of subjects in relation to the grouping of independent variables identified as gender, decade of age, and range of instrumentation on the pre-, intra-, and postoperative lordosis measurements. Although the possibility of surgeon differences and individual patient differences in height and weight is assumed, these data were not included in the formal analysis.

Biomechanical Experiments

Biomechanical tests were conducted on fresh human cadaveric specimens of the lumbosacral spine (L1–S1) with a mean age of 80 years (range, 69–90 years). All specimens were screened with plain radiographs. Specimens demonstrating metastatic disease, severe degenerative changes, bridging osteophytes, or significantly decreased segmental motion on manual testing were excluded from the study. The L1 and S1 vertebrae were secured in holding fixtures using bone cement and pins.

Figure 1. Experimental setup for biomechanical tests. A loading apparatus was attached to the L1 vertebra to test the spine specimen under axial loading in an upright posture (A) and under extension loading (B). The bottom segment (S1) was fixed to a six-component load cell. Markers with three infrared-emitting diode (IRED) targets each were attached to each vertebra for three-dimensional motion measurement. Intradiscal pressure was measured by a pressure transducer inserted into the L3–L4 intervertebral disc. Angular changes within the instrumented segments (L4 –S1) were monitored by an electronic clinometer, which allowed creation of a given lordosis change in the instrumented segments.

The mechanical response of the cadaveric specimens was measured under two test conditions: 1) axial loading in the upright posture and 2) extension loading. For the upright posture test, the L1 vertebra was allowed to displace in the axial direction while its position in the horizontal plane was fixed relative to the S1 segment (Figure 1A). This approximated the normal physiologic alignment of the lumbosacral spine in an upright posture.19 For the extension loading test, a loading apparatus was attached to the L1 vertebra, and deadweights were used to apply extension moments to the specimen (Figure 1B). In the extension loading test, the proximal end of the specimen was unconstrained. In both experiments, the bottom segment (S1) was fixed to a six-component load cell (MC3A- 6-1000, Advanced Mechanical Technology, Inc., Newton, MA) that monitored six orthogonal forces and moments to quantify the boundary conditions at S1.

A transpedicular device was designed to apply a controlled amount of distraction to the specimen to create a loss of lordosis in the lumbar segments. Two devices were applied bilaterally between the pedicle screws inserted in the L4 and S1 pedicles to simulate an L4–S1 instrumented fusion surgery. Each device was instrumented with strain gauges and calibrated so that forces across the device could be monitored (Figure 2).

Figure 2. A custom-made device was designed to apply a controlled amount of distraction to the specimen to create a loss of lordosis across the L4–S1 segments. The device was instrumented with strain gauges and calibrated so that forces across the device could be monitored. Two devices were applied bilaterally between the pedicle screws inserted in the L4 and S1 pedicles to simulate an L4–S1 instrumented fusion surgery.

Triaxial rosette strain gauges (FRA-2-3L, TX Measurements, Inc., College Station, TX) were mounted bilaterally on the L3 lamina near the pars interarticularis using established techniques (Figure 3).20 The strain data were reduced to the magnitude and direction of the principal strains, which indicated the relative load sharing through the posterior spinal column in the segment above the instrumentation. Intradiscal pressure at the disc proximal to instrumentation was measured by a pressure transducer (EPI-050-300S, Entran Devices, Inc., Fairfield, NJ) inserted into the L3–L4 intervertebral disc (Figure 1). The pressure data estimated the load transmitted through the anterior column of the spine. The segmental motion of each lumbar segment resulting from an applied load was recorded using an optoelectronic three-dimensional motion analysis system (WATSMART, Northern Digital, Inc., Waterloo, Ontario).21 Markers with three infrared-emitting diode (IRED) targets each were securely attached to each vertebra for the motion analysis (Figure 1).

Figure 2. A custom-made device was designed to apply a controlled amount of distraction to the specimen to create a loss of lordosis across the L4–S1 segments. The device was instrumented with strain gauges and calibrated so that forces across the device could be monitored. Two devices were applied bilaterally between the pedicle screws inserted in the L4 and S1 pedicles to simulate an L4–S1 instrumented fusion surgery.

Specimens were tested in the upright posture and in the extension loading mode. Loads were applied incrementally using deadweights. A maximum axial force of 392 N was applied in the upright posture while a maximum moment of 5 Nm was used for the extension loading test. These load values are within the range of physiologic loads experienced during activities of daily living.19 At each incremental load level, the load was held constant for 1 minute to allow for soft tissue relaxation before the data were acquired.21

The preceding tests were performed with the following treatments: intact specimen with no instrumentation, instrumentation across L4–S1 with in situ fixation, instrumentation with a 2° loss of lordosis (upright posture test only), instrumentation with a 4° loss of lordosis (extension mode only), and instrumentation with an 8° loss of lordosis (extension mode only). These parameters were chosen because in the upright posture test, the amount of distraction across L4–S1 is limited by the strength of the bone–implant interface. In preliminary experiments, it was noted that to induce a 4° loss of lordosis while maintaining the upright posture, a total distraction load in excess of 400 N was required, which caused loosening of the pedicle screws in some of the specimens. Therefore, for the upright posture test, the simulated loss of lordosis in the instrumented segments was kept at 2°. In creating the loss of lordosis, angular changes within the instrumented (L4–S1) and adjacent (L3–L4) segments were monitored by two electronic clinometers (AccuStar I, Lucas Control Systems Products, Hampton, VA; Figure 1). Test runs with the various treatments were performed in a randomized order. Eight lumbar specimens were used for the upright posture test, and six for the extension test.

In both the upright posture test and the extension loading test, the forces and moments at S1, loads across the distraction devices, strain data at the L3 laminae, and L3—L4 intradiscal pressure were recorded for 1 second for each incremental load and distraction level. The data were averaged over the 1-second duration and used in the subsequent analysis. The threedimensional position data were acquired with the WATSMART motion analysis system for a 1-second interval for each incremental load and distraction level. Segmental motions were calculated as a function of applied load and amount of distraction. The biomechanical data were reviewed for the required statistical assumptions and analyzed descriptively. Tests of multivariate analysis of variance (MANOVA) and ANOVA were used with corresponding post hoc tests. The alpha coefficient was adjusted to account for repeated testing.

Results


Clinical Radiographic Analysis

Over all 62 patients, the differences between preoperative standing, intraoperative prone, and postoperative standing lordosis were found to be statistically significant (P = 0.006). The total lumbar lordosis evaluated across L2–S1 was 52° preoperatively, 41° intraoperatively, and 47° postoperatively. In this particular sample, ANCOVA revealed no statistically significant effects of gender or decade of age on the changes in lumbar lordosis, nor was there significant interaction between these factors after adjustment for the covariates.

In the 30 patients who underwent L4–S1 fusion surgery, the lordosis within the fusion decreased by 10° intraoperatively and 10° postoperatively relative to the preoperative standing value (P < 0.001). After surgery, the lordosis in the proximal segments increased by 2° each at the L2–L3 and L3–L4 segments, as compared with the preoperative measures (P = 0.239). This trend of compensatory hyperlordosis in the proximal segments, although not statistically significant, was evident in the measurements of all raters.

In the 18 patients who underwent L4–L5 fusion surgery, the lordosis in the fusion segment decreased by 4° intraoperatively and 3° postoperatively relative to the preoperative standing value (P = 0.036). After surgery, the lordosis in the proximal segments increased by 2° each at the L2–L3 and L3–L4 segments, as compared with the preoperative values (P = 0.412). This trend of compensatory hyperlordosis in the proximal segments was noted in the measurements of all raters.

Biomechanical Experiments

The load across the distraction devices showed a significant increase caused by a loss of lordosis in the instrumented segments. This was demonstrated in both the upright posture test and the extension loading test. In the upright posture test, the load across the distraction devices increased by up to 192 N when a 2° loss of lordosis was created across the instrumented segments (P = 0.000). In the extension loading test, with an extension moment of 5 Nm, the load across the distraction devices increased by 31 N (a 41% increase relative to the load for baseline lordosis) and 76 N (a 101% increase) because of a 4° and 8° loss of lordosis, respectively (P = 0.023) (Table 1).

The posterior shear force at the proximal segments needed to maintain an upright posture increased significantly with a loss of lordosis in the instrumented segments (Table 2).A2° loss of lordosis in the instrumented segments increased the posterior shear force by 98% (from 29 N to 57.4 N) (P < 0.05). With axial loading of 392 N in the upright posture, the increase in the posterior shear force resulting from a 2° loss of lordosis was 38%
(P < 0.05).

The posterior element strain at the adjacent level (L3) under extension loading increased with a loss of lordosis in the instrumented segments (P50.013). The strain at the L3 lamina caused by an extension moment of 5 Nm was 42% larger for the instrumented construct, with a 4° loss of lordosis relative to instrumentation with baseline lordosis (Table 3). This trend of increasing posterior element strain also was observed for the construct, with an 8° loss of lordosis. However, the measured values of intradiscal pressure in the adjacent (L3–L4) disc did not show significant changes resulting from a loss of lordosis in the instrumented segments in either the upright posture test or the extension loading test (P = 0.39 and P = 0.27 for the upright posture and extension loading tests, respectively).

The immediate effect of instrumentation on the flexibility of the proximal segments (L3–L4 and L2–L3) in extension loading was not observed. The sagittal motion at the L3–L4 and L2–L3 motion segments under an extension moment of 5 Nm ranged from 4° to 5°, and no consistent trends or significant effects were noted as a result of instrumentation across L4–S1 with or without a loss of lordosis in the instrumented segments (P = 0.95 and P = 0.09 for L3–L4 and L2–L3, respectively).

Discussion

Numerous clinical studies have reported on the effect of fusion on adjacent segments. Increased incidence of disc degeneration, segmental instability, osteoarthritis, and stenosis have been noted at the junctional level. Lehmann et al14 reported that patients who underwent lumbar fusion had higher pain levels than the general population and more radiographic incidence of instability and stenosis. Frymoyer et al7 compared radiographic findings of patients who underwent fusion or nonfusion 10 or more years after lumbar disc surgery. Total lumbar mobility (flexion– extension) was greater in the nonfusion group than in the fusion group. However, the mobility in proximal segments was greater in the fusion group, suggesting a compensatory increase in range of motion. These authors also reported that acquired spondylolysis at the upper end of the fusion was found in 2.5% of the fusion group. Stokes et al23 reported radiographic findings suggesting altered stress states in motion segments above the fusion site. Flexion, lateral bending, axial rotation, and their coupled motions were greater than normal when expressed as a proportion of the total motion in the mobile segments.

Biomechanical alterations of the segments adjacent to lumbar fusion have been studied. Quinnell and Stockdale17 investigated the influence of floating lumbar spinal fusions on the remaining lumbar spine. They performed independent fusions of L3–L4 and L4–L5 and compared them with the nonfused condition. Their results suggest that there is additional loading of the disc immediately below the fusion, as compared with nonfused conditions. Lee and Langrana13 evaluated the rigidity of three different types of spinal fusions and their effects on adjacent segments. The best method was considered to be bilateral–lateral fusion because it appeared to provide good rigidity to the fused segment and had the least effect on the adjacent segment. Posterior fusion resulted in the highest stress in the adjacent segments. An additional study from the same group reported resultant changes in lumbosacral spine stiffness with regard to combined compression–torsion loads after three types of fusion.27 Their findings suggest that compression and torsion loads produce significantly increased stress at levels adjacent to the fusion area.

The current study investigated whether a postoperative change in the normal lumbar sagittal alignment can significantly alter the mechanical loading of the instrumented and adjacent segments. The underlying premise of this study was that postoperative hypolordosis in the instrumented segments may cause compensatory hyperlordosis in the proximal mobile segments because of the patient’s attempts to maintain a normal sagittal posture and alignment. This in turn would lead to abnormal loading of the adjacent segment during activities of daily living. The current authors also reasoned that a loss of lordosis in the instrumented segments would decrease the tension in the anterior soft tissue structures, thereby increasing the load on the posterior spinal implant under extension loading of the lumbar spine.

First, the intraoperative lumbar curvature and preoperative and postoperative standing alignment were analyzed in patients who underwent posterolateral lumbar fusions with transpedicular instrumentation to assess the extent of iatrogenic sagittal malalignment. Second, the lumbar malalignment was simulated in cadaveric specimens to assess its mechanical effect on the instrumented and adjacent motion segments.

To conserve the total number of samples needed for cadaveric experiments, repeated tests were performed on each specimen. Although repeated measures designs have the advantage of enhancing precision and decreasing the necessary sample size requirement, this study was limited by the potential for carryover effects. Therefore, several precautions were taken to minimize the effects of repeated tests on measured responses. The tests were performed in a randomized order, and care was taken to prevent dehydration of the tissue by wrapping the specimen in saline-soaked gauze.

Previous studies have suggested that the ex vivo experimental conditions do not fully replicate the in vivo physiologic conditions, and may affect motion segment responses.11 However, this is unlikely to influence substantially the results concerning the relative effects of malalignment because data corresponding to the different values of lumbar lordosis were acquired under the same experimental conditions.

The specimens available for this study came from an older population. Ideally, this study should be performed with normal young cadaveric specimens. However, this material generally is unavailable. Therefore, the fixation strength of the bone screw in the pedicle was a limiting factor in the ability to simulate large magnitudes of hypolordosis across instrumented segments. As noted previously, the simulation of a loss of lordosis of 4° or more caused loosening of the bone screw in the pedicle in some of the specimens. However, a significant effect on the mechanics of the spine was observed even with a 2° loss of lordosis. The intent was to show that, regardless of specimen age, a change in the mechanics affects loading at the junctional level.

The effects of hypolordotic instrumented segments on the mechanical loading of adjacent segments are likely to depend on the number of spinal levels spanned by the instrumentation, and on whether mobile segments are available distally and proximally to compensate for the loss of lordosis in the instrumented segments. The authors chose to study the effects from loss of lordosis experimentally across L4–S1 because the majority of patients (30 out of 62) in their clinical radiographic analysis had instrumented fusions across these levels.

Our clinical radiographic analysis showed that intraoperative lumbar lordosis significantly decreased with the kneeling position on the spinal surgical frame with 90° of hip and knee flexion. The current results are consistent with those of previous studies measuring lumbar lordosis resulting from various positions on different spinal surgical positioning devices.8,16,22,25 Furthermore, the current study documents the observation that application of rigid instrumentation to the intraoperative flat back causes residual postoperative hypolordosis in the instrumented segments. Jackson and McManus10 reported that a loss of lumbar lordosis may be associated with compensatory changes in the lumbopelvic angulation. Although the current authors did not examine the change in lumbopelvic alignment, they noted a consistent trend of compensatory hyperlordosis in the mobile segments proximal to the hypolordotic instrumented segments.

The load across the distraction devices increased significantly with increasing loss of lordosis in the instrumented segments. The mechanics of this phenomenon may be explained as follows. In instrumented segments with baseline lordosis, the applied extension moment may be counterbalanced by the loads in the posterior implant and the anterior soft tissue structures including the anterior longitudinal ligament and the anterior portion of the anulus fibrosis. With a loss of lordosis in the instrumented segments, the tension in the anterior soft tissue structures may decrease, thereby increasing the implant load needed to balance the extension moment. This phenomenon also was observed in the upright posture test. To maintain an upright posture in the presence of a loss of lordosis, the posterior shear force on the proximal segments increased. This increased the extension moment on the lumbar spine and led to an increased loading of the posterior implant. The increased loading of the posterior implant caused by a hypolordotic alignment of the instrumented segments is likely to place the bone-implant interface at an increased risk of loosening because of repetitive extension loading induced during activities of daily living.

The hypolordosis in the instrumented segments caused a significant change in the mechanical loading of the adjacent segment. In the upright posture test, in which the anteroposterior position of the L1 vertebra was held fixed relative to the sacrum, the posteriorly directed shear force in the proximal segments increased significantly with a loss of lordosis in the instrumented segments. In the extension loading experiment, wherein the lumbar spine was subjected to an extension moment of 5 Nm, the strain in the L3 lamina increased when lordosis was decreased in the instrumented (L4–S1) segments, indicating an increased loading of the posterior column in the segment above the instrumentation. These biomechanical effects of postoperative sagittal malalignment on the loading of the adjacent segment may contribute to the degenerative changes at the junctional level reported as long-term consequences of lumbar fusion.

To the authors’ knowledge, this study is the first to document that lumbar malalignment can significantly increase the loading on the posterior column and the posteriorly directed shear force at the adjacent segment. It remains to be seen whether this altered mechanical loading environment is responsible for inducing or accelerating the breakdown of the adjacent levels. Currently, the results of this study suggest that intraoperative reproduction of preoperative standing lordosis may reduce the risk of junctional breakdown in patients undergoing posterolateral lumbar fusion with instrumentation.

Key Points

  • Change in lumbar lordosis was measured in patients who had undergone posterolateral lumbar fusions using transpedicular instrumentation, and the biomechanical effects of postoperative lumbar malalignment were measured in cadaveric specimens.
  • In patients who underwent L4–S1 fusions in the kneeling position with 90° of hip and knee flexion, the postoperative lordosis in the fusion decreased while the lordosis in the proximal segments increased as compared with the preoperative measures.
  • Hypolordosis in the instrumented segments increased the load across the posterior transpedicular devices, the posterior shear force, and the lamina strain at the adjacent level.
  • Hypolordosis in the instrumented segments caused increased loading on the posterior column of the adjacent segments. These biomechanical effects may explain the observed degenerative changes at the junctional level as long-term consequences of lumbar fusion.

Acknowledgment

This article is dedicated to the memory of Dr. Shinji Umehara who passed away on November 23, 1996. The authors thank Mark J. Sartori for experimental support, Brian K. Dunlap and John Schaefer for fabrication of experimental apparatus, and Wendy K. Flentge and Tracy L. Vincenti for assistance in radiographic analysis.


References

  1. Andersson GBJ, Murphy RW,O¨ rtengren R, Nachemson AL. The influence of backrest inclination and lumbar support on lumbar lordosis. Spine 1979;4: 52–8.
  2. Bernhardt M, Bridwell KH. Segmental analysis of the sagittal plane alignment of the normal thoracic and lumbar spines and thoracolumbar junction. Spine 1989;14:717–21.
  3. Böstman O, Hyrkäs J, Hirvensalo E, Kallio E. Blood loss, operating time, and positioning of the patient in lumbar disc surgery. Spine 1990;15:360–3.
  4. Chow DHK, Luk KDK, Evans JH, Leong JCY. Effects of short anterior lumbar interbody fusion on biomechanics of neighboring unfused segments. Spine 1996;21:549–55.
  5. Dekutoski MB, Schendel MJ, Ogilvie JW, Olsewski JM, Wallace LJ, Lewis JL. Comparison of in vivo and in vitro adjacent segment motion after lumbar fusion. Spine 1994;19:1745–51.
  6. Fernand R, Fox DE. Evaluation of lumbar lordosis: A prospective and retrospective study. Spine 1985;10:799–803.
  7. Frymoyer JW, Hanley EN, Howe J, Kuhlmann D, Matteri RE. A comparison of radiographic findings in fusion and nonfusion patients ten or more years following lumbar disc surgery. Spine 1979;4:435–40.
  8. Guanciale AF, Dinsay JM, Watkins RG. Lumbar lordosis in spinal fusion: A comparison of intraoperative results of patient positioning on two different operative table frame types. Spine 1996;21:964–9.
  9. Hayes MA, Tompkins SF, Herndon WA, Gruel CR, Kopta JA, Howard TC. Clinical and radiological evaluation of lumbosacral motion below fusion levels in idiopathic scoliosis. Spine 1988;13:1161–7.
  10. Jackson RP, McManus AC. Radiographic analysis of sagittal plane alignment and balance in standing volunteers and patients with low back pain matched for age, sex, and size. Spine 1994;19:1611–18.
  11. Keller TS, Holm SH, Hansson TH, Spengler DM. The dependence of intervertebral disc mechanical properties on physiologic conditions. Spine 1990;15: 751–61.
  12. Lee CK. Accelerated degeneration of the segment adjacent to a lumbar fusion. Spine 1988;13:375–7.
  13. Lee CK, Langrana NA. Lumbosacral spinal fusion: A biomechanical study. Spine 1984;9:574–81.
  14. Lehmann TR, Spratt KF, Tozzi JE, et al. Long-term follow-up of lower lumbar fusion patients. Spine 1987;12:97–104.
  15. Nagata H, Schendel MJ, Transfeld EE, Lewis JL. The effect of immobilization of long segments of the spine on the adjacent and distal facet force and lumbosacral motion. Spine 1993;18:2471–9.
  16. Peterson MD, Nelson LM, McManus AC, Jackson RP. The Effect of Operative Position on Lumbar Lordosis. Presented at the 29th annual meeting of the Scoliosis Research Society, Portland, Oregon, 1994.
  17. Quinnel RC, Stockdale HR. Some observations of the influence of a single lumbar floating fusion on the remaining lumbar spine. Spine 1981;6:263–7. 18. Schlegel JD, Smith JA, Schleusener RL. Lumbar motion segment pathology adjacent to thoracolumbar, lumbar, and lumbosacral fusions. Spine 1996;21: 970–81.
  18. Schultz AB, Miller JAA. Biomechanics of the human spine. In: Mow VC, Hayes WC, eds. Basic Orthopaedic Biomechanics. New York: Raven Press, 1991: 350.
  19. Shah JS, Hampson WGJ, Jayson MIV. The distribution of surface strain in the cadaveric lumbar spine. J Bone Joint Surg [Br] 1978;60:246–51.
  20. Slosar PJ, Patwardhan AG, Lorenz M, Havey R, Sartori M. Instability of the lumbar burst fracture and limitations of the transpedicular instrumentation. Spine 1995;20:1452–61.
  21. Stephens GC, Wilber RG, Jackson RP, Yoo JU. Comparison of Lumbar Sagittal Alignment Produced by Different Operative Positions. Presented at the 29th Annual Meeting of Scoliosis Research Society, Portland, Oregon, 1994.
  22. Stokes IAF, Wilder DG, Frymoyer JW, Pope MH. Assessment of patients with low back pain by biplaner radiographic measurement of intervertebral motion. Spine 1981;6:233–40.
  23. Sundén G, Wallöe A, Wingstrand H. A new device to reduce intra-abdominal pressure during lumbar surgery. Spine 1986;11:635–6.
  24. Tan SB, Kozak JA, Dickson JH, Nalty TJ. Effect of operative position on sagittal alignment of the lumbar spine. Spine 1994;19:314–18.
  25. Weinhoffer SL, Guyer RD, Herbert M, Griffith SL. Intradiscal pressure measurements above an instrumented fusion: A cadaveric study. Spine 1995;20:526– 31.
  26. Yang SW, Langrana NA, Lee CK. Biomechanics of lumbosacral spinal fusion in combined compression–torsion loads. Spine 1986;11:937–41.
From the *Department of Orthopaedic Surgery and Rehabilitation, Loyola University Medical Center, Maywood, Illinois, and Biomechanics Laboratory, Rehabilitation Research and Development Center, Department of Veterans Affairs, Edward Hines, Jr., Hospital, Hines, Illinois, the †Department of Orthopaedic Surgery, Hokkaido University School of Medicine, Sapporo, Japan, ‡Hinsdale Orthopedic Associates, S.C., Hinsdale, Illinois, and the §Department of Orthopaedic Surgery, Sapporo Medical University, Sapporo, Japan. Supported by the Orren Baab Research & Education Fund and Hinsdale Hospital Foundation, Hinsdale, Illinois, and the Rehabilitation Research and Development Center, Department of Veterans Affairs, Edward Hines, Jr., Hospital, Hines, Illinois.
Acknowledgment date: June 18, 1999.
Acceptance date: October 8, 1999.
Device status category: 11.
Conflict of interest category: 15.