Preoperative retrolisthesis as a risk factor of postdecompression lumbar disc herniation

Free access

OBJECT

In this study, the authors aimed to identify specific risk factors for postdecompression lumbar disc herniation (PDLDH) in patients who have not undergone discectomy and/or fusion.

METHODS

Between 2007 and 2012, 493 patients with lumbar spinal stenosis underwent bilateral partial laminectomy without discectomy and/or fusion in a single hospital. Eighteen patients (herniation group [H group]: 15 men, 3 women; mean age 65.1 years) developed acute sciatica as a result of PDLDH within 2 years after surgery. Ninety patients who did not develop postoperative acute sciatica were selected as a control group (C group: 75 men, 15 women; mean age 65.4 years). Patients in the C group were age and sex matched with those in the H group. The patients in the groups were also matched for decompression level, number of decompression levels, and surgery date. The radiographic variables measured included percentage of slippage, intervertebral angle, range of motion, lumbar lordosis, disc height, facet angle, extent of facet removal, facet degeneration, disc degeneration, and vertebral endplate degeneration. The threshold for PDLDH risk factors was evaluated using a continuous numerical variable and receiver operating characteristic curve analysis. The area under the curve was used to determine the diagnostic performance, and values greater than 0.75 were considered to represent good performance.

RESULTS

Multivariate analysis revealed that preoperative retrolisthesis during extension was the sole significant independent risk factor for PDLDH. The area under the curve for preoperative retrolisthesis during extension was 0.849; the cutoff value was estimated to be a retrolisthesis of 7.2% during extension.

CONCLUSIONS

The authors observed that bilateral partial laminectomy, performed along with the removal of the posterior support ligament, may not be suitable for lumbar spinal stenosis patients with preoperative retrolisthesis greater than 7.2% during extension.

ABBREVIATIONSAUC = area under the curve; BPL = bilateral partial laminectomy; DM = diabetes mellitus; ISL = interspinous ligament; JOA = Japanese Orthopaedic Association; LSS = lumbar spinal canal stenosis; PDLDH = postdecompression lumbar disc herniation; ROC = receiver operating characteristic; ROM = range of motion; SSL = supraspinous ligament.

Abstract

OBJECT

In this study, the authors aimed to identify specific risk factors for postdecompression lumbar disc herniation (PDLDH) in patients who have not undergone discectomy and/or fusion.

METHODS

Between 2007 and 2012, 493 patients with lumbar spinal stenosis underwent bilateral partial laminectomy without discectomy and/or fusion in a single hospital. Eighteen patients (herniation group [H group]: 15 men, 3 women; mean age 65.1 years) developed acute sciatica as a result of PDLDH within 2 years after surgery. Ninety patients who did not develop postoperative acute sciatica were selected as a control group (C group: 75 men, 15 women; mean age 65.4 years). Patients in the C group were age and sex matched with those in the H group. The patients in the groups were also matched for decompression level, number of decompression levels, and surgery date. The radiographic variables measured included percentage of slippage, intervertebral angle, range of motion, lumbar lordosis, disc height, facet angle, extent of facet removal, facet degeneration, disc degeneration, and vertebral endplate degeneration. The threshold for PDLDH risk factors was evaluated using a continuous numerical variable and receiver operating characteristic curve analysis. The area under the curve was used to determine the diagnostic performance, and values greater than 0.75 were considered to represent good performance.

RESULTS

Multivariate analysis revealed that preoperative retrolisthesis during extension was the sole significant independent risk factor for PDLDH. The area under the curve for preoperative retrolisthesis during extension was 0.849; the cutoff value was estimated to be a retrolisthesis of 7.2% during extension.

CONCLUSIONS

The authors observed that bilateral partial laminectomy, performed along with the removal of the posterior support ligament, may not be suitable for lumbar spinal stenosis patients with preoperative retrolisthesis greater than 7.2% during extension.

Bilateral partial laminectomy (BPL) is a widely accepted procedure for the treatment of patients with lumbar spinal canal stenosis (LSS) without any instability or anterolisthesis. Although most LSS patients show significant neurological gain and good clinical courses after lumbar posterior decompression, including BPL,8,38 some patients experience a recurrence of acute limb pain in the relatively early postoperative period because of restenosis, instability, or disc herniation.5,8,11,17,19,20,26,28,36,38 In particular, postdecompression lumbar disc herniation (PDLDH) at the same level is the first or second most common reason for reoperations.8,11,17,19,21,36,38 By identifying PDLDH risk factors, it is possible to avoid reoperations at the same level, which are often complicated and even risky.35 However, previous reports have included patient cohorts that were too heterogeneous to appropriately elucidate significant PDLDH risk factors. In the present study, we aimed to identify the preoperative and intraoperative risk factors for PDLDH in homogeneous patient cohorts.

Methods

Patients

This study was reviewed and approved by our institution's ethics committee, and informed consent was obtained from the patients. Between September 2007 and April 2012, 493 LSS patients (326 men, 167 women) with an average age of 69.7 years (range 29–90 years) underwent BPL without concomitant discectomy or fusion. None of the patients had previously undergone lumbar surgery. This cohort did not include patients with a forward slip of greater than 11% and or an intervertebral range of motion (ROM) greater than 11°, based on the findings of preoperative flexion-extension lateral radiographs. Preoperatively, patients received conservative treatment, which did not sufficiently relieve their pain or intermittent claudication, or experienced significant palsy that required surgery.

As shown in Fig. 1, 319 patients were eventually enrolled in the study. PDLDH was defined as acute lower-limb pain that developed within 2 years postoperatively. Patients were included in a herniation group (H group, n = 18) if 1) newly herniated discs were observed postoperatively, and 2) MRI findings were consistent with acute lower-limb pain (Fig. 2). However, new, asymptomatic herniated discs were not considered in this definition.

FIG. 1.
FIG. 1.

Flow chart showing patient exclusion criteria. pts = patients.

FIG. 2.
FIG. 2.

Case 15. Sagittal MR images obtained preoperatively (A) and at 10 months after surgery (B), showing L4–5 disc herniation. The white circle indicates PDLDH. The patient experienced acute lower-limb pain after the development of PDLDH.

The candidates for the control group (C group) included 301 patients (189 men, 112 women) with an average age of 69.8 years (range 29–90 years). Generally, sampling 4 or 5 control subjects for each case provides sufficient statistical power.34 Therefore, we selected 5 control subjects for each H-group patient (a total of 90 patients in the C group).

Control Group Selection

The 90 patients in the C group were selected as follows. First, sex and decompression levels were matched with the H-group patients. Thereafter, patients were selected if their ages (in days) at the time of surgery were similar (within 2 years) to those of the H-group patients. Preliminarily, 5 or more control candidates were selected for each of the 18 H-group patients. However, for 2 patients in the H group, fewer than 5 patients were available for inclusion as control subjects. Therefore, to obtain at least 5 control subjects for each of these H-group patients, the matching age window was expanded to 7 years. Consequently, 7 and 18 control candidates, respectively, were selected for these 2 patients in the H group. After this initial matching, priority numbers were assigned to the control candidates according to their similarity with the H-group patients in terms of the number of decompression levels and the operation date. Finally, in cases in which the control subjects matched 2 or more H-group patients, those with higher priority numbers were exclusively selected. This selection method was not arbitrary and reduced confounding factors to the greatest extent possible. Sixty-eight of the control subjects with less than 2 years of face-to face follow-up did not have any evidence of acute lower-limb pain within 2 years postoperatively, as assessed through telephone interviews.

Surgical Procedures

BPL has been described previously.8 Briefly, to decompress the L4–5 levels, the lamina and spinous processes of the lower L-4 and upper L-5 vertebrae were removed using chisels and Kerrison clamps. In addition, the ligamentum flavum and supraspinous ligament (SSL) and interspinous ligament (ISL) of L4–5 were removed. Thereafter, BPL of other levels was performed, as necessary. Figure 3 shows a posterior view of a postoperative 3D CT image. The operations on H- and C-group patients were performed by 7 surgeons; 26 operations were performed by Surgeon A, 13 by Surgeon B, 28 by Surgeon C, 3 by Surgeon D, 25 by Surgeon E, 9 by Surgeon F, and 4 by Surgeon G.

FIG. 3.
FIG. 3.

3D CT image (posterior view) acquired after BPL at L4–5. The lamina and spinous processes of the lower L-4 and upper L-5 vertebrae were removed using a chisel and Kerrison clamp. The ligamentum flavum and SSL and ISL of L4–5 were removed. BPL was also performed at other levels.

Outcomes

All patient characteristics were recorded, including sex, age, body mass index, number of decompression levels, and distribution of decompression levels (Table 1). Vertebrae were counted from T-1 inferiorly to achieve accurate number assignment to the lumbosacral transitional vertebrae.23 Additional factors were used to compare the H and C groups, including preoperative diagnosis of diabetes mellitus (DM), preoperative Japanese Orthopaedic Association (JOA) score,22 operating time per level, and blood loss per level.

TABLE 1.

Patient characteristics*

CharacteristicH GroupCandidates for C Groupp ValueC Groupp Value
No. of patients1830190
M/F ratio15:3189:1120.12775:151.00
Mean age at op (range), yrs65.1 ± 5.8 (52–75)69.8 ± 9.2 (29–90)0.00965.4 ± 5.9 (46–77)0.882
Mean BMI (range), kg/m225.2 ± 4.1 (19.1–34.3)24.3 ± 3.4 (14.6–35.4)0.47325.0 ± 3.2 (18.5–35.4)0.944
DM/non-DM ratio4:1421:691.00
Mean preop JOA score (range)17.4 ± 4.4 (2–23)16.6 ± 4.7 (4–24)0.505
Mean op time per level (range), mins34.2 ± 9.8 (23–55)37.0 ± 10.0 (19–71)0.210
Mean blood loss per 1 level (range), g69.4 ± 47.0 (25–200)61.3 ± 39.6 (15–233)0.422
No. of decompression levels0.1130.480
  1310421
  2711836
  376426
  40126
  5131
Decompression level
  L1–21100.47830.523
  L2–35700.391321.000
  L3–4121820.456600.786
  L4–5172710.394881.000
  L5–S15620.550170.521

BMI = body mass index; — = not applicable.

p values in boldface indicate a statistically significant difference. Mean values are presented ± SD.

The p value resulting from a comparison between the H group and candidates for the C group.

The p value resulting from a comparison between the H and C groups.

Preoperative and postoperative anteroposterior radiographs were obtained in the neutral position. Preoperative and postoperative lateral radiographs were obtained in the maximally flexed and extended positions and in the neutral position.

On anteroposterior radiographs, lateral slippage and disc wedging angle were assessed. Lateral slippage was measured using the centroid method advocated by Freedman et al.12 On lateral radiographs, the percentage of vertebral slippage during flexion in a neutral position and during extension were assessed according to the method established by Dupuis et al.7; a positive value indicated slippage in the anterior direction. The intervertebral angle during flexion (the angle made by the endplates of the disc space), extension, and ROM, and in a neutral position were also assessed; a positive value indicated lordosis. In the neutral position, lumbar lordosis between L-1 and S-1 was also measured (Fig. 4). Postoperative data were obtained 1 year postoperatively in the C group and in H-group patients who developed PDLDH within 1 year postoperatively.

FIG. 4.
FIG. 4.

Radiographs and CT images demonstrating measurement of lateral slippage, disc wedging, intervertebral and facet angles, and disc height. A: Anteroposterior digital plain radiographs were used to determine lateral slippage and disc wedging angle. B: Lateral digital plain radiographs images were used to determine percentage of slippage and intervertebral angle. C: Disc height was measured on preoperative midsagittal CT images. D and E: Preoperative (D) and postoperative (E) axial CT scans. The facet angle was measured using the angles made by connecting the 2 end points of each facet on preoperative axial CT scan (at the superior endplate level of the lower vertebra) and a line connecting the 2 dorsal points of each facet joint. The right- and left-side angles were averaged (41.8° in this example). The extent of facet removal was calculated as a percentage based on the following formula: 100 × (preoperative facet length – postoperative facet length)/preoperative facet length. The right- and left-side percentages were averaged (14.5% in this example).

For 4 patients in the H group who underwent revision surgery, postoperative data were assessed immediately before revision surgery. For 5 patients in the H group who developed PDLDH 1–2 years postoperatively, postoperative data were assessed immediately after PDLDH diagnosis.

CT images were assessed for the following 4 parameters: disc height,3 facet joint angle,3 extent of facet removal,14 and facet degeneration.37 These parameters were measured according to previously reported methods3,14,37 (Fig. 4).

Disc degeneration and vertebral endplates were assessed using MRI. Pfirrmann classification was used to stratify disc degeneration into 5 grades, based on T2-weighted midsagittal MRI cuts of the lumbar spine.30 Vertebral endplate degenerative changes were classified as normal, or Type I, II, or III using the Modic scale.27

All the radiographic, CT, and MRI evaluations were performed at the level of PDLDH for the H group, as well as at the corresponding level for the C group, except in cases of lumbar lordosis.

Reproducibility Evaluation

To evaluate intraobserver and interobserver reliability, 20 of the 108 patients were selected randomly and underwent measurement twice, with a 2-week interval, by 1 investigator (S.T.) and by a second observer (K.T.). Both observers are orthopedic spine surgeons who were blinded to the subject information. The reproducibility of numeric data were evaluated using intraclass correlation coefficients, and that of categorized data were evaluated using kappa coefficients. The interobserver intraclass correlation coefficients were calculated using the average of 2 measurements. The interobserver kappa values were calculated using the first measurement. The intraobserver and interobserver reliabilities were fully acceptable (Table 2).24

TABLE 2.

Intraobserver and interobserver reliability

Radiographic FactorIntraobserverInterobserver
R1–R1 (95% CI)R2–R2 (95% CI)R1–R2
Plain radiography
  AP lateral slip0.962 (0.908–0.985)0.905 (0.780–0.961)0.826 (0.524–0.934)
  AP disc wedging angle0.880 (0.726–0.950)0.819 (0.604–0.924)0.772 (0.486–0.905)
  Percentage of slippage0.968 (0.924–0.987)0.922 (0.817–0.968)0.858 (0.681–0.941)
  Percentage of slippage (neutral)0.958 (0.899–0.983)0.903 (0.775–0.960)0.875 (0.716–0.948)
  Percentage of slippage (extension)0.976 (0.941–0.990)0.957 (0.897–0.983)0.840 (0.645–0.933)
  Intervertebral angle (flexion)0.923 (0.823–0.968)0.877 (0.720–0.949)0.937 (0.850–0.974)
  Intervertebral angle (neutral)0.889 (0.751–0.953)0.838 (0.642–0.932)0.880 (0.727–0.951)
  Intervertebral angle (extension)0.928 (0.834–0.970)0.921 (0.814–0.968)0.840 (0.537–0.933)
  Intervertebral ROM0.839 (0.643–0.933)0.816 (0.597–0.922)0.852 (0.667–0.939)
  Lumbar lordosis at L1–S10.965 (0.917-–0.986)0.979 (0.948–0.991)0.945 (0.867–0.978)
CT scanning
  Disc height0.962 (0.910–0.984)0.934 (0.843–0.973)0.966 (0.486–0.992)
  Facet angle0.985 (0.964–0.994)0.913 (0.797–0.964)0.898 (0.708–0.962)
  Extent of facet removal0.875 (0.722–0.947)0.840 (0.646–0.933)0.895 (0.757–0.958)
  Facet joint degeneration0.9300.9320.865
MRI
  Disc degeneration0.8570.6920.857
  Modic change1.000.6491.00

AP = anteroposterior; R1 and R2 = Reviewer 1 and 2.

Statistical Analysis

We used SPSS statistical software version 21.0 (IBM) for all statistical analyses. To compare the 2 groups using the Mann-Whitney U-test and Fisher's exact test, p values less than 0.05 were considered statistically significant. Multivariate logistic regression analysis with forward stepwise selection was performed using variables that showed p values less than 0.20 in the univariate analysis.

Receiver Operating Characteristic Curve

We evaluated the threshold for a potential PDLDH risk factor, described as a continuous numerical variable, using received operating characteristic (ROC) curve analyses, which plot sensitivity against the result of 1 minus the specificity.9 The area under the curve (AUC) was used to determine diagnostic performance. AUC values greater than 0.75 are generally considered to represent good performance.32 The cutoff value was determined using the Youden Index.10

Results

Patient-Group Demographics

A total of 18 patients who developed new lumbar disc herniation at the same level postsurgically were included in the H group (Fig. 2). Patients in the H group were significantly younger than the candidates for the C group (Table 1). After the strict selection of control subjects, 90 patients were included in the C group (Table 1).

The H and C groups were comparable in terms of sex ratio, age at the time of surgery, body mass index, DM prevalence, preoperative JOA score, operative time per level, blood loss per level, number of decompression levels, and distribution of decompression levels. The H-group patient characteristics are outlined in Table 3. Revision surgery was required in 4 H-group patients; 2 underwent discectomy and 2 underwent fusion surgery. The procedures and postoperative recoveries were uneventful. No patients in the C group underwent revision surgery.

TABLE 3.

Summary of the H group

Pt No.Age (yrs), SexHerniation LevelNo. of Decomp LevelsOnset of Herniation, YrsRevision Op (type)
170, MLt L2–330.07
252, MRt L2–330.15+ (discectomy)
366, MRt L2–331.93
462, MRt L3–420.48
564, MRt L3–430.24+ (fusion)
671, MRt L3–421.53
770, MRt L4–520.77
858, MLt L4–530.60
975, MRt L4–531.52
1069, MRt L4–510.55+ (fusion)
1162, MCent L4–521.96+ (discectomy)
1268, FRt L4–510.36
1362, MCent L4–520.12
1472, FRt L4–520.06
1567, MRt L4–510.84
1661, FRt L4–520.35
1758, MLt L4–531.98
1864, MLt L5–S151.91

Cent = central; decomp = decompression; pt = patient; − = patient underwent no revision surgery; + = patient underwent revision surgery.

Univariate Analysis of Pre- and Perioperative Radiologic Data

Table 4 lists the potential predictive variables for PDLDH. Univariate analysis revealed that patients in the H group were more likely to have preoperative retrolisthesis and lower lumbar lordosis than those in the C group; the differences were statistically significant. In both groups, the extent of facet removal was approximately 15%.

TABLE 4.

Radiographic assessment*

Radiographic FactorH GroupC Groupp Value
Plain radiography
  AP lateral slip, mm0.6 ± 0.7 (0–2.2)0.5 ± 0.9 (0–4.0)0.336
  AP disc wedging angle, °1.5 ± 1.6 (0–5.2)1.1 ± 1.6 (0–8.1)0.285
  Percentage of slippage (flexion)−2.8 ± 3.0 (−9.3 to 0)1.0 ± 4.5 (−17.1 to 10.9)<0.001
  Percentage of slippage (neutral)6.0 ± 3.3 (−11.4 to 0)−1.2 ± 4.4 (−12.9 to 9.9)<0.001
  Percentage of slippage (extension)−8.0 ± 3.5 (−14.4 to 0)−2.2 ± 4.5 (−18.1 to 8.4)<0.001
  Intervertebral angle (flexion), °1.7 ± 2.1 (−2.0 to 6.1)0.4 ± 3.6 (−7.2 to 10.5)0.064
  Intervertebral angle (neutral), °7.6 ± 3.8 (0.2–13.3)6.2 ± 4.4 (−5.3 to 20.2)0.137
  Intervertebral angle (extension), °10.1 ± 3.6 (1.5 to 15.5)8.2 ± 4.6 (−0.2 to 20.4)0.060
  Intervertebral ROM, °8.4 ± 3.7 (1.8–16.3)7.8 ± 3.4 (0.9–16.3)0.510
  Lumbar lordosis at L1–S1, °30.8 ± 12.1 (13.5–55.2)36.4 ± 11.4 (4.9–61.0)0.044
CT scanning
  Disc height, mm8.7± 2.4 (2.2–12.4)8.0 ± 2.8 (1.0–14.0)0.262
  Facet angle, °51.7 ± 10.3 (34.6–78.8)53.8 ± 10.9 (29.9–75.5)0.376
  Extent of facet removal, %15.0 ± 7.4 (4.0–30.3)15.0 ± 7.2 (1.3–35.8)0.869
  Facet joint degeneration0.126
    Grade 035
    Grade 1833
    Grade 2630
    Grade 3122
MRI
  Disc degeneration0.303
    Grade 3317
    Grade 41561
    Grade 5012
  Modic change0.839
    Normal1467
    Type I28
    Type II29
    Type III06

p values in boldface indicate a statistically significant difference.

Values are presented as mean ± SD (range) or the number of patients.

The p value resulting from the comparison between the H and C groups.

Multivariate Logistic Regression Analysis

To remove multicollinearity in multivariate logistic regression analysis, we omitted the percentage of slippage demonstrated during flexion and in the neutral position, which showed correlation coefficients greater than 0.7 with percentage of slippage during extension (0.821 and 0.881 by Spearman's correlation, respectively). Similarly, the intervertebral angle in the neutral position was omitted (Spearman's correlation 0.863, with the intervertebral angle during extension). Eventually, the following 5 variables were adopted as potential predictors: 1) percentage of slippage during extension, 2) intervertebral angle during flexion, 3) intervertebral angle during extension, 4) lumbar lordosis, and 5) facet joint degeneration. The analysis revealed that percentage of slippage during extension was the sole significant independent risk factor of PDLDH (p = 0.0001; OR 1.36; 95% CI 1.17–1.58).

ROC Curve Analysis

The AUC for retrolisthesis during extension, which was the negated value of the percentage of slippage during extension, for PDLDH was 0.849 (95% CI 0.755–0.943; p < 0.001). The cutoff value was 7.2%, per the Youden Index (Fig. 5). The average vertebral anteroposterior diameter in the 108 patients was 42.6 mm. Therefore, retrolisthesis of 7.2% corresponded to a value of 3.1 mm. A cutoff value of 7.2% indicated an OR of 23.7 (95% CI 6.8–82.5).

FIG. 5.
FIG. 5.

ROC curve for the prediction of PDLDH within 2 years. The closer the curve approximates the left upper corner, the better the test. The straight dotted line indicates the chance results. The dotted circle indicates the cutoff point, as determined by the Youden Index.

Comparison of Changes Between Pre- and Postoperative Radiographic Factors

Table 5 shows the changes in the absolute values, pre-and postoperatively, and radiographic factors for the patients. Patients in the H group demonstrated significant postoperative instability in lateral listhesis; disc wedging angle; percentage of slippage during flexion, extension, and in a neutral position; and the intervertebral angle at extension. Three patients in the H group developed de novo anterolisthesis postoperatively.

TABLE 5.

Changes between the pre- and postoperative radiographic factors*

Radiographic FactorH GroupC Groupp Value
AP lateral listhesis, mm0.7 ± 0.7 (0–2.3)0.3 ± 0.5 (0–4.0)0.004
AP disc wedging angle, °1.7 ± 1.7 (0–6.0)0.7 ± 0.8 (0–3.6)0.004
Percentage of slippage (flexion)4.0 ± 3.7 (0–14.0)1.5 ± 1.9 (0–9.4)0.001
Percentage of slippage (neutral)4.9 ± 5.4 (0–18.8)1.9 ± 2.4 (0–11.7)0.001
Percentage of slippage (extension)5.0 ± 6.3 (0.2–21.5)1.4 ± 1.6 (0–7.6)0.001
Intervertebral angle (flexion), °3.0 ± 2.6 (0.2–9.6)2.1 ± 1.6 (0.1–8.7)0.248
Intervertebral angle (neutral), °3.4 ± 3.1 (0.3–10.7)1.8 ± 1.3 (0–6.0)0.106
Intervertebral angle (extension), °2.9 ± 2.3 (0.2–7.1)1.6 ± 1.3 (0–4.7)0.036
Intervertebral ROM, °3.5 ± 3.5 (0.2–10.8)2.6 ± 2.1 (0–9.8)0.689
Lumbar lordosis at L1–S1, °4.8 ± 3.9 (0.6–14.8)4.3 ± 3.8 (0–16.9)0.499

p values in boldface indicate a statistically significant difference. Values are presented as mean ± SD (range).

The p value resulting from the comparison between the H and C groups.

Incidence of PDLDH Predictors for Each Surgeon

To evaluate the influence of the surgeon, the incidence of PDLDH for each surgeon was investigated. The incidence was not significantly different among the 7 surgeons (p = 0.394, Fisher exact test) (Table 6).

TABLE 6.

Incidence of postdecompression lumbar disc herniation by surgeon*

Study GroupSurgeon ASurgeon BSurgeon CSurgeon DSurgeon ESurgeon FSurgeon G
H group2171610
C group24122121984
Total26132832594

Data are presented as the number of occurrences. Difference among the 7 surgeons was not statistically significant (Fisher's exact test p = 0.394).

Discussion

Many authors have investigated the clinical and radiographic results of lumbar decompression surgery and have reported certain reoperation rates and etiologies (Table 7).5,6,8,11,13,17,19,26,29,36,38 In many papers, postoperative instability and lumbar disc herniation at the same level of decompression were the first or second most common reasons for reoperations.8,11,17,19,21,36,38 Given the risk of revision surgery, identifying the risk factors of PDLDH is important. However, the material or/and methods used in the previous reports were inappropriate for identifying the true risk factors of PDLDH. In many of those studies, the subjects comprised patients who did not undergo laminectomy or laminotomy alone but also underwent concomitant discectomy5,8,11,19,31,36 and/or fusion.11 In some studies, the mean follow-up period was longer than 5 years postoperatively.11,31,38 This appears to be a very long duration to extract sequelae of laminectomy or laminotomy, exclusively. Moreover, the laminectomy or laminotomy procedures varied among these studies, and some reports included several procedures.17,19 To determine the true PDLDH risk factors, we excluded patients undergoing concomitant discectomy and/or fusion during the index surgery. In addition, to minimize the aging effect, we excluded patients in whom PDLDH developed more than 2 years postoperatively. Moreover, to avoid confounding factors due to the procedures, subjects in our study included patients who underwent identical BPL surgeries.

TABLE 7.

Studies with 50 or more patients who underwent lumbar decompression surgery and included reoperation descriptions

Authors & YearOpPosterior Element PreservationNo. of PtsPDLDHs Requiring Additional Op*No. of PDLDHs Including Conservative Treatment*Postop Instability Requiring Fusion*Concomitant Discectomy*Concomitant PLFMean FU (yrs)
Present studyBPLNo3814 (1.0)18 (4.7)7 (1.8)NoNo2.1
Eule et al., 1999BPLNo1382 (1.4)ND2 (1.4)40 (29)No3.5
Yuzawa, 2011BPLNo631 (1.6)ND2 (3.2)NDNo5.0
Hopp & Tsou, 1988LNNo34414 (4.1)ND16 (5.5)50%NoND
Herno et al., 1993LNNo1082 (1.9)ND2 (1.9)NDNo12.8
Tuite et al., 1994LNNo32410 (3.1)ND0 (0.0)38No4.6
Fox et al., 1996LNNo1243 (2.4)ND7 (5.6)14 (11)Yes (n = 32; 11%)5.8
Fu et al., 2008LFYes760 (0.0)ND0 (0.0)NDNo3.3
Cavuşoğlu et al., 2007ULBDYes1000 (0.0)ND0 (0.0)13 (13)No5.4
Castro-Menéndez et al., 2009MELYes500 (0.0)ND2 (4.0)20 (40)No4.0
Pao et al., 2009MELYes530 (0.0)ND0 (0.0)NDNo1.3
Minamide et al., 2013MELYes310ND4 (1.3)3 (1.0)NDNo2.0

FU = follow-up; LF = laminoforaminotomy; LN = laminectomy; MEL = microendoscopic laminotomy; ND = not described; PLF = posterolateral fusion; ULBD = unilateral approach for bilateral decompression, under the microscope.

Data given as number (%) unless otherwise indicated.

Mainly laminectomy but including some other procedures.

The percentage is approximate; the actual value was not described.

We identified preoperative retrolisthesis as a risk factor of PDLDH within 2 years postoperatively. This result suggested that the presence of preoperative retrolisthesis makes patients more prone to the development of PDLDH. Through ROC curve analysis, the cutoff value of retrolisthesis was found to be 7.2%, or approximately 3 mm. This cutoff value provided an OR of 23.7 (95% CI 6.8–82.5). The 3-mm value dovetails with the definition of retrolisthesis described in previous papers that referred to retrolisthesis.2,31

To date, retrolisthesis has been considered to be of little clinical significance, and there is much less information available about retrolisthesis than there is regarding anterolisthesis. Heuer et al. investigated the lumbar biomechanics of human cadavers, using both angulation measurements and translational measurements as the response parameters, to study forward and backward bending.18 Their results showed that the angulation measurements were strongly related to the translational measurements. Although listhesis corresponds to intervertebral translation, we believe that lumbar biomechanical studies, using angulation measurements in forward and backward bending, can be used to understand the results of the present study. Moreover, we hypothesized a mechanism for the high frequency of PDLDH in patients with preoperative retrolisthesis.

Through biomechanical analyses, many authors indicated that forward bending of the lumbar spine is mainly restricted by the SSLs and ISLs15,16,25,33,39; the tensile strengths of these ligaments generate this resistance force. However, the spinous processes, facet joints, and intervertebral discs play important roles in restricting backward bending.1,15 Given that the intervertebral discs contribute to the restriction of backward bending, patients in the H group with preoperative retrolisthesis may be suggested to have had weak discs that could no longer restrain backward bending. Therefore, the spinous processes may compensate to resist backward bending. Thus, the H group showed significantly more postoperative instability than the C group. Although the determination of whether instability or PDLDH occurred first may be impossible, removing the inferior halves of the cranial spinous processes, the SSLs and ISLs, during surgery might trigger postoperative instability or/and disc herniation. Moreover, a finite element model study by Zander et al. showed that resection of the posterior bony or ligamentous elements influences the stresses and deformations in the intervertebral discs.39

As shown in Table 7, the incidence of PDLDH was lower for the minimally invasive surgeries in which the SSLs and ISLs were preserved than for conventional open decompressions. However, this comparison is overly simplified because some papers included patients in whom herniation recurred and some had different lengths of follow-up. Moreover, new surgical techniques tend to underestimate the true incidence of complications.4 Fusion surgeries for patients with a retrolisthesis greater than 7.2% might, however, be considered as overtreatment. Therefore, further studies are required to better understand how to prevent PDLDH.

This study has several limitations. First, we performed a case-control study with retrospective data collection. We believe that a nested case-control study with prospective data collection is better. However, in this study, control subjects were strictly selected without arbitrariness, making the comparisons between the H and C groups fully acceptable. Second, whole-spine radiographic evaluations were not performed with the patients in the standing position, which could provide additional information, such as that pertaining to sagittal balance. Although information regarding sagittal balance is desirable, the results of this study, nevertheless, may be helpful for making decisions to treat LSS patients. Third, although the H group strictly included patients who experienced postoperative pain remission, followed by the recurrence of lower-limb pain with consistent “new” herniation on MRI, there may be many factors that determine whether a disc herniation is symptomatic or not. This may lead to a Type 1 error.

Conclusions

This study revealed that preoperative retrolisthesis was the sole significant risk factor for PDLDH. Lumbar decompression surgery involving the removal of the posterior elements may not be suitable to prevent PDLDH. This is the first study to elucidate the significant risk factors for PDLDH.

References

  • 1

    Adams MADolan PHutton WC: The lumbar spine in backward bending. Spine (Phila Pa 1976) 13:101910261988

  • 2

    Berlemann UJeszenszky DJBühler DWHarms J: Mechanisms of retrolisthesis in the lower lumbar spine. A radiographic study Acta Orthop Belg 65:4724771999

  • 3

    Blumenthal CCurran JBenzel ECPotter RMagge SNHarrington JF Jr: Radiographic predictors of delayed instability following decompression without fusion for degenerative grade I lumbar spondylolisthesis. J Neurosurg Spine 18:3403462013

  • 4

    Carragee EJHurwitz ELWeiner BK: A critical review of recombinant human bone morphogenetic protein-2 trials in spinal surgery: emerging safety concerns and lessons learned. Spine J 11:4714912011

  • 5

    Castro-Menéndez MBravo-Ricoy JACasal-Moro RHernández-Blanco MJorge-Barreiro FJ: Midterm outcome after microendoscopic decompressive laminotomy for lumbar spinal stenosis: 4-year prospective study. Neurosurgery 65:100110A122009

  • 6

    Cavuşoğlu HKaya RATürkmenoglu ONTuncer CColak IAydin Y: Midterm outcome after unilateral approach for bilateral decompression of lumbar spinal stenosis: 5-year prospective study. Eur Spine J 16:213321422007

  • 7

    Dupuis PRYong-Hing KCassidy JDKirkaldy-Willis WH: Radiologic diagnosis of degenerative lumbar spinal instability. Spine (Phila Pa 1976) 10:2622761985

  • 8

    Eule JMBreeze RKindt GW: Bilateral partial laminectomy: a treatment for lumbar spinal stenosis and midline disc herniation. Surg Neurol 52:3293381999

  • 9

    Fischer JEBachmann LMJaeschke R: A readers' guide to the interpretation of diagnostic test properties: clinical example of sepsis. Intensive Care Med 29:104310512003

  • 10

    Fluss RFaraggi DReiser B: Estimation of the Youden Index and its associated cutoff point. Biom J 47:4584722005

  • 11

    Fox MWOnofrio BMOnofrio BMHanssen AD: Clinical outcomes and radiological instability following decompressive lumbar laminectomy for degenerative spinal stenosis: a comparison of patients undergoing concomitant arthrodesis versus decompression alone. J Neurosurg 85:7938021996

  • 12

    Freedman BAHorton WCRhee JMEdwards CC IIKuklo TR: Reliability analysis for manual radiographic measures of rotatory subluxation or lateral listhesis in adult scoliosis. Spine (Phila Pa 1976) 34:6036082009

  • 13

    Fu YSZeng BFXu JG: Long-term outcomes of two different decompressive techniques for lumbar spinal stenosis. Spine (Phila Pa 1976) 33:5145182008

  • 14

    Ghogawala ZBenzel ECAmin-Hanjani SBarker FG IIHarrington JFMagge SN: Prospective outcomes evaluation after decompression with or without instrumented fusion for lumbar stenosis and degenerative Grade I spondylolisthesis. J Neurosurg Spine 1:2672722004

  • 15

    Gillespie KADickey JP: Biomechanical role of lumbar spine ligaments in flexion and extension: determination using a parallel linkage robot and a porcine model. Spine (Phila Pa 1976) 29:120812162004

  • 16

    Hartmann FJanssen CBöhm SHely HRommens PMGercek E: Biomechanical effect of graded minimal-invasive decompression procedures on lumbar spinal stability. Arch Orthop Trauma Surg 132:123312392012

  • 17

    Herno AAiraksinen OSaari T: Long-term results of surgical treatment of lumbar spinal stenosis. Spine (Phila Pa 1976) 18:147114741993

  • 18

    Heuer FSchmidt HClaes LWilke HJ: Stepwise reduction of functional spinal structures increase vertebral translation and intradiscal pressure. J Biomech 40:7958032007

  • 19

    Hopp ETsou PM: Postdecompression lumbar instability. Clin Orthop Relat Res 227:1431511988

  • 20

    Iguchi TKanemura AKasahara KKurihara ADoita MYoshiya S: Age distribution of three radiologic factors for lumbar instability: probable aging process of the instability with disc degeneration. Spine (Phila Pa 1976) 28:262826332003

  • 21

    Iguchi TKurihara ANakayama JSato KKurosaka MYamasaki K: Minimum 10-year outcome of decompressive laminectomy for degenerative lumbar spinal stenosis. Spine (Phila Pa 1976) 25:175417592000

  • 22

    Izumida SInoue S: [Assessment of treatment for low back pain.]. J Jpn Orthop Assoc 60:3913941986. (Jpn)

  • 23

    Konin GPWalz DM: Lumbosacral transitional vertebrae: classification, imaging findings, and clinical relevance. AJNR Am J Neuroradiol 31:177817862010

  • 24

    Landis JRKoch GG: The measurement of observer agreement for categorical data. Biometrics 33:1591741977

  • 25

    Lee MJBransford RJBellabarba CChapman JRCohen AMHarrington RM: The effect of bilateral laminotomy versus laminectomy on the motion and stiffness of the human lumbar spine: a biomechanical comparison. Spine (Phila Pa 1976) 35:178917932010

  • 26

    Minamide AYoshida MYamada HNakagawa YKawai MMaio K: Endoscope-assisted spinal decompression surgery for lumbar spinal stenosis. J Neurosurg Spine 19:6646712013

  • 27

    Modic MTSteinberg PMRoss JSMasaryk TJCarter JR: Degenerative disk disease: assessment of changes in vertebral body marrow with MR imaging. Radiology 166:1931991988

  • 28

    Oertel MFRyang YMKorinth MCGilsbach JMRohde V: Long-term results of microsurgical treatment of lumbar spinal stenosis by unilateral laminotomy for bilateral decompression. Neurosurgery 59:126412702006

  • 29

    Pao JLChen WCChen PQ: Clinical outcomes of microendoscopic decompressive laminotomy for degenerative lumbar spinal stenosis. Eur Spine J 18:6726782009

  • 30

    Pfirrmann CWMetzdorf AZanetti MHodler JBoos N: Magnetic resonance classification of lumbar intervertebral disc degeneration. Spine (Phila Pa 1976) 26:187318782001

  • 31

    Shen MRazi ALurie JDHanscom BWeinstein J: Retrolisthesis and lumbar disc herniation: a preoperative assessment of patient function. Spine J 7:4064132007

  • 32

    Swets JA: Measuring the accuracy of diagnostic systems. Science 240:128512931988

  • 33

    Tai CLHsieh PHChen WPChen LHChen WJLai PL: Biomechanical comparison of lumbar spine instability between laminectomy and bilateral laminotomy for spinal stenosis syndrome—an experimental study in porcine model. BMC Musculoskelet Disord 9:842008

  • 34

    Taylor JM: Choosing the number of controls in a matched case-control study, some sample size, power and efficiency considerations. Stat Med 5:29361986

  • 35

    Tormenti MJMaserati MBBonfield CMGerszten PCMoossy JJKanter AS: Perioperative surgical complications of transforaminal lumbar interbody fusion: a single-center experience. J Neurosurg Spine 16:44502012

  • 36

    Tuite GFStern JDDoran SEPapadopoulos SMMcGillicuddy JEOyedijo DI: Outcome after laminectomy for lumbar spinal stenosis. Part I: Clinical correlations J Neurosurg 81:6997061994

  • 37

    Weishaupt DZanetti MBoos NHodler J: MR imaging and CT in osteoarthritis of the lumbar facet joints. Skeletal Radiol 28:2152191999

  • 38

    Yuzawa Y: The interspinous ligament should be removed for the decompression surgery with the case of lumbar spinal canal stenosis. Arch Orthop Trauma Surg 131:7537582011

  • 39

    Zander TRohlmann AKlöckner CBergmann G: Influence of graded facetectomy and laminectomy on spinal biomechanics. Eur Spine J 12:4274342003

Disclosures

Takeshi Fuji has served as a consultant for Daiichi-Sankyo and holds patents with Century Medical and Showa-Ikakogyo.

Author Contributions

Conception and design: Takenaka. Acquisition of data: Takenaka, Tateishi. Analysis and interpretation of data: Takenaka. Drafting the article: Takenaka. Critically revising the article: Hosono. Reviewed submitted version of manuscript: Mukai. Statistical analysis: Takenaka. Study supervision: Fuji.

If the inline PDF is not rendering correctly, you can download the PDF file here.

Article Information

INCLUDE WHEN CITING Published online December 11, 2015; DOI: 10.3171/2015.6.SPINE15288.

Correspondence Shota Takenaka, Orthopaedic Surgery, Japan Community Health-care Organization Osaka Hospital, 4-2-78 Fukushima, Osaka 553-0003, Japan. email: show@yb3.so-net.ne.jp.

© AANS, except where prohibited by US copyright law.

Headings

Figures

  • View in gallery

    Flow chart showing patient exclusion criteria. pts = patients.

  • View in gallery

    Case 15. Sagittal MR images obtained preoperatively (A) and at 10 months after surgery (B), showing L4–5 disc herniation. The white circle indicates PDLDH. The patient experienced acute lower-limb pain after the development of PDLDH.

  • View in gallery

    3D CT image (posterior view) acquired after BPL at L4–5. The lamina and spinous processes of the lower L-4 and upper L-5 vertebrae were removed using a chisel and Kerrison clamp. The ligamentum flavum and SSL and ISL of L4–5 were removed. BPL was also performed at other levels.

  • View in gallery

    Radiographs and CT images demonstrating measurement of lateral slippage, disc wedging, intervertebral and facet angles, and disc height. A: Anteroposterior digital plain radiographs were used to determine lateral slippage and disc wedging angle. B: Lateral digital plain radiographs images were used to determine percentage of slippage and intervertebral angle. C: Disc height was measured on preoperative midsagittal CT images. D and E: Preoperative (D) and postoperative (E) axial CT scans. The facet angle was measured using the angles made by connecting the 2 end points of each facet on preoperative axial CT scan (at the superior endplate level of the lower vertebra) and a line connecting the 2 dorsal points of each facet joint. The right- and left-side angles were averaged (41.8° in this example). The extent of facet removal was calculated as a percentage based on the following formula: 100 × (preoperative facet length – postoperative facet length)/preoperative facet length. The right- and left-side percentages were averaged (14.5% in this example).

  • View in gallery

    ROC curve for the prediction of PDLDH within 2 years. The closer the curve approximates the left upper corner, the better the test. The straight dotted line indicates the chance results. The dotted circle indicates the cutoff point, as determined by the Youden Index.

References

1

Adams MADolan PHutton WC: The lumbar spine in backward bending. Spine (Phila Pa 1976) 13:101910261988

2

Berlemann UJeszenszky DJBühler DWHarms J: Mechanisms of retrolisthesis in the lower lumbar spine. A radiographic study Acta Orthop Belg 65:4724771999

3

Blumenthal CCurran JBenzel ECPotter RMagge SNHarrington JF Jr: Radiographic predictors of delayed instability following decompression without fusion for degenerative grade I lumbar spondylolisthesis. J Neurosurg Spine 18:3403462013

4

Carragee EJHurwitz ELWeiner BK: A critical review of recombinant human bone morphogenetic protein-2 trials in spinal surgery: emerging safety concerns and lessons learned. Spine J 11:4714912011

5

Castro-Menéndez MBravo-Ricoy JACasal-Moro RHernández-Blanco MJorge-Barreiro FJ: Midterm outcome after microendoscopic decompressive laminotomy for lumbar spinal stenosis: 4-year prospective study. Neurosurgery 65:100110A122009

6

Cavuşoğlu HKaya RATürkmenoglu ONTuncer CColak IAydin Y: Midterm outcome after unilateral approach for bilateral decompression of lumbar spinal stenosis: 5-year prospective study. Eur Spine J 16:213321422007

7

Dupuis PRYong-Hing KCassidy JDKirkaldy-Willis WH: Radiologic diagnosis of degenerative lumbar spinal instability. Spine (Phila Pa 1976) 10:2622761985

8

Eule JMBreeze RKindt GW: Bilateral partial laminectomy: a treatment for lumbar spinal stenosis and midline disc herniation. Surg Neurol 52:3293381999

9

Fischer JEBachmann LMJaeschke R: A readers' guide to the interpretation of diagnostic test properties: clinical example of sepsis. Intensive Care Med 29:104310512003

10

Fluss RFaraggi DReiser B: Estimation of the Youden Index and its associated cutoff point. Biom J 47:4584722005

11

Fox MWOnofrio BMOnofrio BMHanssen AD: Clinical outcomes and radiological instability following decompressive lumbar laminectomy for degenerative spinal stenosis: a comparison of patients undergoing concomitant arthrodesis versus decompression alone. J Neurosurg 85:7938021996

12

Freedman BAHorton WCRhee JMEdwards CC IIKuklo TR: Reliability analysis for manual radiographic measures of rotatory subluxation or lateral listhesis in adult scoliosis. Spine (Phila Pa 1976) 34:6036082009

13

Fu YSZeng BFXu JG: Long-term outcomes of two different decompressive techniques for lumbar spinal stenosis. Spine (Phila Pa 1976) 33:5145182008

14

Ghogawala ZBenzel ECAmin-Hanjani SBarker FG IIHarrington JFMagge SN: Prospective outcomes evaluation after decompression with or without instrumented fusion for lumbar stenosis and degenerative Grade I spondylolisthesis. J Neurosurg Spine 1:2672722004

15

Gillespie KADickey JP: Biomechanical role of lumbar spine ligaments in flexion and extension: determination using a parallel linkage robot and a porcine model. Spine (Phila Pa 1976) 29:120812162004

16

Hartmann FJanssen CBöhm SHely HRommens PMGercek E: Biomechanical effect of graded minimal-invasive decompression procedures on lumbar spinal stability. Arch Orthop Trauma Surg 132:123312392012

17

Herno AAiraksinen OSaari T: Long-term results of surgical treatment of lumbar spinal stenosis. Spine (Phila Pa 1976) 18:147114741993

18

Heuer FSchmidt HClaes LWilke HJ: Stepwise reduction of functional spinal structures increase vertebral translation and intradiscal pressure. J Biomech 40:7958032007

19

Hopp ETsou PM: Postdecompression lumbar instability. Clin Orthop Relat Res 227:1431511988

20

Iguchi TKanemura AKasahara KKurihara ADoita MYoshiya S: Age distribution of three radiologic factors for lumbar instability: probable aging process of the instability with disc degeneration. Spine (Phila Pa 1976) 28:262826332003

21

Iguchi TKurihara ANakayama JSato KKurosaka MYamasaki K: Minimum 10-year outcome of decompressive laminectomy for degenerative lumbar spinal stenosis. Spine (Phila Pa 1976) 25:175417592000

22

Izumida SInoue S: [Assessment of treatment for low back pain.]. J Jpn Orthop Assoc 60:3913941986. (Jpn)

23

Konin GPWalz DM: Lumbosacral transitional vertebrae: classification, imaging findings, and clinical relevance. AJNR Am J Neuroradiol 31:177817862010

24

Landis JRKoch GG: The measurement of observer agreement for categorical data. Biometrics 33:1591741977

25

Lee MJBransford RJBellabarba CChapman JRCohen AMHarrington RM: The effect of bilateral laminotomy versus laminectomy on the motion and stiffness of the human lumbar spine: a biomechanical comparison. Spine (Phila Pa 1976) 35:178917932010

26

Minamide AYoshida MYamada HNakagawa YKawai MMaio K: Endoscope-assisted spinal decompression surgery for lumbar spinal stenosis. J Neurosurg Spine 19:6646712013

27

Modic MTSteinberg PMRoss JSMasaryk TJCarter JR: Degenerative disk disease: assessment of changes in vertebral body marrow with MR imaging. Radiology 166:1931991988

28

Oertel MFRyang YMKorinth MCGilsbach JMRohde V: Long-term results of microsurgical treatment of lumbar spinal stenosis by unilateral laminotomy for bilateral decompression. Neurosurgery 59:126412702006

29

Pao JLChen WCChen PQ: Clinical outcomes of microendoscopic decompressive laminotomy for degenerative lumbar spinal stenosis. Eur Spine J 18:6726782009

30

Pfirrmann CWMetzdorf AZanetti MHodler JBoos N: Magnetic resonance classification of lumbar intervertebral disc degeneration. Spine (Phila Pa 1976) 26:187318782001

31

Shen MRazi ALurie JDHanscom BWeinstein J: Retrolisthesis and lumbar disc herniation: a preoperative assessment of patient function. Spine J 7:4064132007

32

Swets JA: Measuring the accuracy of diagnostic systems. Science 240:128512931988

33

Tai CLHsieh PHChen WPChen LHChen WJLai PL: Biomechanical comparison of lumbar spine instability between laminectomy and bilateral laminotomy for spinal stenosis syndrome—an experimental study in porcine model. BMC Musculoskelet Disord 9:842008

34

Taylor JM: Choosing the number of controls in a matched case-control study, some sample size, power and efficiency considerations. Stat Med 5:29361986

35

Tormenti MJMaserati MBBonfield CMGerszten PCMoossy JJKanter AS: Perioperative surgical complications of transforaminal lumbar interbody fusion: a single-center experience. J Neurosurg Spine 16:44502012

36

Tuite GFStern JDDoran SEPapadopoulos SMMcGillicuddy JEOyedijo DI: Outcome after laminectomy for lumbar spinal stenosis. Part I: Clinical correlations J Neurosurg 81:6997061994

37

Weishaupt DZanetti MBoos NHodler J: MR imaging and CT in osteoarthritis of the lumbar facet joints. Skeletal Radiol 28:2152191999

38

Yuzawa Y: The interspinous ligament should be removed for the decompression surgery with the case of lumbar spinal canal stenosis. Arch Orthop Trauma Surg 131:7537582011

39

Zander TRohlmann AKlöckner CBergmann G: Influence of graded facetectomy and laminectomy on spinal biomechanics. Eur Spine J 12:4274342003

TrendMD

Metrics

Metrics

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 200 200 53
PDF Downloads 543 543 69
EPUB Downloads 0 0 0

PubMed

Google Scholar