Ligamentum flavum hypertrophy significantly contributes to the severity of neurogenic intermittent claudication in patients with lumbar spinal canal stenosis

Abstract

Ligamentum flavum hypertrophy (LFH) is a known contributor to lumbar spinal canal stenosis (LSCS). However, the clinical significance and quantitative role of LFH compared to other components, such as disc bulging and facet hypertrophy, have not yet been examined. We investigated the correlation between the quantitative radiological factors, clinical symptoms, and outcomes in patients with LSCS. In total, 163 patients diagnosed with single-level (L4–L5) stenosis were included. The patients were divided into 2 groups according to claudication severity: >100 m for mild (n = 92) and < 100 m for severe (n = 71). The visual analog scale (VAS) was used to quantify back and leg pain, and the Oswestry Disability Index (ODI) and Short form-36 (SF-36) physical component summary (PCS) scores, and Macnab criteria were evaluated as clinical factors 6 months after treatment. We measured the baseline canal cross-sectional area, ligamentum flavum (LF) area, disc herniation area, dural sac area, fat area, and LF thickness using MRI. A comparative analysis was performed to evaluate the association between radiologic and clinical factors. Additionally, further comparative analyses between the types of surgeries were performed. Among various radiologic factors, the baseline LF thickness (odds ratio [OR] 1.73; 95% confidence interval [CI] 1.25–2.41) was the only major contributing factor to the severity of claudication in the multivariate logistic regression analysis. The types of surgery (decompression alone vs fusion) did not significantly differ in terms of their clinical outcomes, including back and leg VAS, ODI, SF-36 PCS, and satisfaction with the MacNab classification. LF thickness is a major factor contributing to claudication severity.

1. Introduction

Patients with lumbar spinal canal stenosis (LSCS) typically present with neurogenic intermittent claudication, with or without lower back pain, radicular leg pain, and lower extremity weakness.^[1,2] These clinical symptoms appear when the spinal canal is narrowed due to changes in the anatomical structures of the spine, compressing the nerves.^[2–6] When the space of the spinal canal is reduced as a result of ligamentum flavum hypertrophy (LFH), disc bulging, and facet hypertrophy, the nerves inside the dural sac become compressed.^[1,4,7–17] In the real-world practice and studies, LSCS severity is occasionally graded according to how narrow the spinal canal is. However, there is not only no unified measuring method for grading system nor this grading often does not correlate with the severity of clinical symptoms.^[18,19]

Although various quantitative measures, such as the anteroposterior diameter and cross-sectional area of the spinal canal, are frequently applied to LSCS, most of these measures have not reached a consensus on the radiologic grading of LSCS.^[20,21] In a recent systematic literature review, the authors identified 11 different semiquantitative or qualitative radiologic criteria for the diagnosis of LSCS.^{[3,6,22–31]} Unfortunately, these criteria exhibited marked variability in the severity of LSCS and intra- and inter-observer reliability. In addition, these measurement methods dealt only with the spinal canal itself; the radiological and clinical contributions of the ligamentum flavum, disc, fat, and facet, which constitute the canal, were not considered. Therefore, further quantitative analysis of the correlation between various anatomic structures of the spinal canal and symptoms is needed.

It is well known that surgical treatment is helpful for patients with severe LSCS.^[32,33] In addition, there is much evidence showing that sufficient clinical improvement can be expected only with simple decompression without fusion, especially when there is little or no instability.^[34–37] However, most of these studies compared only the surgical methods and clinical results, and the theoretical proof based on the anatomical structure remains insufficient.

The most widely used surgical techniques for the treatment of LSCS are posterior decompression alone and decompression-with-fusion (fusion). The common goal of both surgeries is sufficient nerve decompression. The biggest difference between the 2 surgeries is the absence or presence of bony fusion and the decompression range. For decompression alone, only laminotomy and flavectomy are performed. In the case of fusion, facetectomy and discectomy are performed in addition to laminotomy and flavectomy, which can result in extensive decompression. Considering the 3 major components of LSCS (LFH, facet hypertrophy, and disc bulging), the decompression alone solves the LFH only, while the fusion deals with all 3. Spinal surgeons are well aware that there are many patients with LSCS in whom sufficient nerve decompression can be achieved with flavectomy; therefore, they often perform decompression-only surgery unless the deformity or instability is severe. However, as discussed above, few literature reports provide anatomical evidence for these clinical decisions.

This study quantified the structures involved in the pathophysiology of LSCS and investigated their contribution to the clinical symptoms. In addition, we analyzed the correlation between the clinical results and surgical methods by observing the differences in the structures removed by different surgeries.

2. Methods

2.1. Patient population

A retrospective review was performed on 163 patients who were diagnosed with single-level (L4–L5) stenosis at our institution between January 2015 and December 2019. The patients were followed up for at least 6 months after the initial treatment. The patients were divided into 2 groups according to claudication severity: >100 m for mild (n = 92) and < 100 m for severe (n = 71). A total of 130 patients with insufficient improvement following conservative management underwent surgery performed by 2 spine surgeons with more than 10 years of experience. Among them, 99 patients who underwent laminectomy and flavectomy were in the decompression-alone group, whereas the other patients were in the fusion group. Patients who refused surgery were followed up with nonsurgical procedures, including nerve blocks.

The inclusion criteria for this study were as follows: (1) single-level (L4–L5) central canal stenosis confirmed by magnetic resonance imaging (MRI); (2) formal MRI reports from our radiologists; and (3) neurogenic intermittent claudication with or without concomitant lower back pain, radicular leg pain, and lower extremity weakness. Single-level L4/5 was selected to control the level-dependent complexity and to focus on the quantitative role of the canal-composing structures. Each MRI report was measured by 2 observers to increase reliability. The following patients were excluded: (1) those with foraminal or lateral stenosis; (2) those with instability; (3) those with a history of lumbar surgery; (4) those with a history of psychological disorders; and (5) those with radiologic evidence of other pathologies correlated with symptoms such as prominent ruptured disc or infection.

2.2. Clinical analysis

Clinical factors, including symptoms and outcomes, were investigated using the patients’ medical records. The visual analog scale (VAS) score for back and leg pain, the Oswestry Disability Index (ODI), and the Short form-36 (SF-36) physical component summary (PCS) scores were obtained as clinical factors at baseline and 6 months after the initial nonsurgical procedure or surgery.^[38,39] A satisfactory outcome was defined as excellent or good based on the MacNab classification.^[40] The data on the patients’ baseline demographic characteristics, comorbidities, and clinical factors are summarized in Table 1.

Table 1.

Patient baseline demographics comparing the mild and severe claudication groups.

	Mild Claudication Group (n = 92)	Severe Claudication Group (n = 71)	P
Age, years	64.5 ± 11.0	66.5 ± 7.3	.295
Female sex, n (%)	48 (52.2)	31 (43.7)	.367
VAS back†	3.0 ± 3.0	4.8 ± 3.0	.105
VAS leg†	5.1 ± 2.8	6.0 ± 2.4	.337
ODI‡	29.8 ± 15.5	50.8 ± 15.9	.001*
SF-36 PCS score§	58.5 ± 19.4	30.7 ± 11.2	<.001*
Symptom duration, months	22.1 ± 24.4	23.7 ± 25.3	.746
Comorbidities
Hypertension, n (%)	72 (78.2)	53 (74.6)	.656
Diabetes, n (%)	47 (51.1)	38 (53.5)	.789
Smoking, n (%)	17 (18.5)	9 (12.7)	.453
BMI¶	24.7 ± 3.3	24.8 ± 2.4	.988
Surgery, n (%)	65 (70.7)	65 (91.5)	.002*
Decompression-alone, n (%)	33 (35.9)	36 (50.7)	.054
Fusion, n (%)	32 (34.8)	29 (40.8)	.395

Plus–minus values are means ± SD.

^*Statistically significant differences were observed between the groups. (P < .05).

^†The visual analog scale (VAS) ranged from 0 to 10, with lower scores indicating less severe symptoms.

^‡The Oswestry disability index (ODI) ranges from 0 to 100, with higher scores indicating greater disability related to back pain.

^§The short form-36 (SF-36) physical component summary (PCS) scores ranged from 0 to 100, with higher scores indicating better quality of life.

^¶Body mass index (BMI) is the weight in kilograms divided by the square of the height in meters.

2.3. Radiologic analysis

Two spine specialists blindly measured the baseline canal cross-sectional area, LF area, disc herniation area, dural sac area, fat area, and LF thickness using ImageJ software, version 1.53 (NIH, Bethesda, MD) (Fig. 1). The detailed measurement method has been described in previous studies.^[5,41] Baseline radiological measurements are summarized in Table 2.

Figure 1.

The method of measurement of radiologic factors taken from an axial image at the L4 to 5 intervertebral disc level. (A) disc herniation area, (B) LF area, (C) dural sac area, (D) fat area, (E) LF thickness, A + B + C + D: canal cross-sectional area.

Table 2.

Baseline radiologic measurements in both the groups.

	Mild Claudication Group	Severe Claudication Group	P
Canal cross-section area, mm2	350.29 ± 67.62	338.89 ± 61.41	.313
Ligamentum flavum area, mm2	145.82 ± 35.31	168.37 ± 29.55	<.001*
Disc herniation area, mm2	93.91 ± 36.71	85.62 ± 30.93	.169
Dural sac area, mm2	90.53 ± 39.14	71.47 ± 33.97	.004*
Fat area, mm2	20.04 ± 12.02	13.43 ± 9.97	.001*
Ligamentum flavum thickness, mm	4.99 ± 1.20	6.09 ± 1.39	<.001*

Plus–minus values are means ± SD.

^* Statistically significant differences were observed between the groups. (P < .05)

2.4. Statistical analysis

The student t-test was used to analyze the radiologic and clinical differences between the mild and severe claudication groups. Data are presented as the mean ± SD. For radiologic measurements, interobserver agreement was analyzed using kappa statistics. Multivariate logistic regression analysis was performed using multiple radiological factors as independent variables and claudication severity as the dependent variable. In addition, the Student t-test was performed to examine the radiologic differences between the types of surgery. Further comparison of the 6-month improvements in clinical outcomes and satisfaction in MacNab results (excellent and good) between surgery types was performed. Statistical significance was set at P < .05. Statistical assessments were performed using SPSS software (version 23.0; IBM Corporation, Armonk, NY).

3. Results

3.1. Baseline clinical characteristics in the mild and severe claudication group

There were no significant differences between the mild and severe claudication groups in terms of age, sex, comorbidities, and duration of symptoms (Table 1). In addition, no significant differences in back and leg VAS scores were found. However, the ODI (mild: 29.8 ± 15.5, severe: 50.8 ± 15.9, P = .001) and SF-36 PCS (mild: 58.5 ± 19.4, severe: 30.7 ± 11.2, P < .001) showed significant differences between the 2 groups. In addition, the total number of surgeries was significantly higher in the severe claudication group than that in the mild claudication group. Among the patients with mild claudication, 65 (70%) underwent surgery, while 65 (91 %) had severe claudication (P = .002). There were no significant differences in the composition of the surgical method (decompression alone vs fusion) between the 2 groups.

Table 2 provides the baseline radiologic measurements for the mild and severe claudication groups. The kappa coefficient for the interobserver radiologic measurement was 0.85, indicating nearly perfect agreement. LF area (P < .001), dural sac area (P = .004), fat area (P = .001), and LF thickness (P < .001) were significantly different between the 2 groups. However, the canal cross-sectional area (P = .313) and disc herniation area (P = .169) showed no significant differences.

3.2. Quantitative contributors to the severity of claudication

The odds ratio of the radiologic factors for claudication severity was analyzed using multivariate logistic regression analysis. Multiple predictors were identified (Table 3). LF thickness (odds ratio [OR] 1.73; confidence interval [CI] 1.25–2.41) was a major predictor of claudication severity. The LF area also showed a positive correlation (CI 1.00–1.03); however, the disc herniation area (CI 0.98–1.00) and fat area (CI 0.91–0.99) were negatively correlated with the severity of claudication.

Table 3.

Multivariate logistic regression analysis of contributors to the severity of claudication.

	OR (95% CI)	P
Ligamentum flavum thickness	1.73 (1.25–2.41)	.001*
Ligamentum flavum area	1.02 (1.00–1.03)	.020*
Disc herniation area	0.99 (0.98–1.00)	.048*
Fat area	0.95 (0.91–0.99)	.023*

CI = confidence interval, OR = odds ratio.

^*Statistically significant differences were observed between the groups (P < .05).

3.3. Baseline radiologic measurements in the decompression-alone and fusion group

A significant contributor to LSCS was identified, and the correlation between radiologic factors and clinical outcomes was subsequently compared. The baseline radiological measurements comparing the decompression-alone and fusion groups are provided in Table 4. The dural sac area (decompression-alone: 74.84 ± 36.72 mm², fusion: 92.08 ± 41.27 mm², P = .043) showed significant differences between the 2 groups. However, the canal cross-sectional area, LF area, disc herniation area, fat area, and LF thickness were not significantly different between the 2 groups. Figure 2 describes the pre and postoperation MRI of each decompression-alone and fusion groups.

Table 4.

Baseline radiologic measurements comparing the decompression-alone and fusion group.

	Decompression-alone group (n = 99)	Fusion group (n = 31)	P
Canal cross-section area, mm2	351.38 ± 67.00	355.13 ± 69.01	.796
Ligamentum flavum area, mm2	161.77 ± 35.54	157.26 ± 34.81	.548
Disc herniation area, mm2	97.16 ± 38.00	86.62 ± 31.38	.155
Dural sac area, mm2	74.84 ± 36.72	92.08 ± 41.27	.043*
Fat area, mm2	17.61 ± 12.81	19.17 ± 12.24	.558
Ligamentum flavum thickness, mm	5.16 ± 1.15	5.17 ± 1.13	.948

Means ± SD.

^*Statistically significant differences were observed between the groups. (P < .05)

Figure 2.

The pre and post -operation MRI of each decompression-alone and fusion groups.

3.4. Clinical outcomes in the decompression-alone and fusion group

No significant differences were identified between the 2 groups 6 months after surgery, including changes in back VAS, leg VAS, ODI, and SF-36 PCS (Table 5). Sixty patients (60.6%) in the mild decompression group showed excellent or good improvement. Although 23 patients (74.2%) in the fusion group experienced satisfactory outcomes, there were no significant differences between the 2 groups.

Table 5.

Comparison of the 6-month improvements in the clinical factors and satisfaction according to the MacNab results (excellent and good) between the 2 groups.

	Decompression-alone group	Fusion group	P
VAS back	2.4 ± 3.4	3.1 ± 2.4	.598
VAS leg	3.8 ± 3.4	4.0 ± 2.9	.898
ODI	16.2 ± 21.9	15.3 ± 19.3	.917
SF-36 PCS score	20.2 ± 25.8	29.1 ± 32.1	.433
Satisfaction in MacNab results, n (%)	60 (60.6)	23 (74.2)	.488

Means ± SD.

ODI = Oswestry Disability Index, PCS = physical component summary, SF-36 = short form 36, VAS = visual analog scale.

4. Discussion

Spinal stenosis is a patho-anatomical condition commonly associated with the clinical symptom complex of neurogenic intermittent claudication.^[42] The main difference between spinal stenosis and lumbar disc herniation is claudication. Spinal stenosis is anatomically classified into central canal, lateral recess, neural foraminal, and extraforaminal stenosis. Each location presents slightly different clinical features. In this study, we focused on patients with central stenosis and claudication to reduce these variables.

As expected, the severe claudication group had worse radiological and clinical features than the mild claudication group. Several studies have proposed setting an upper limit for a normal LF thickness of <4 mm.^[15,43] Abbas et al,^[44] on the other hand, reported that the borderline between normal and pathologic LF thickness should not be set to 4 mm because the LF thickness is an age-dependent and sex-independent phenomenon. Based on our radiologic data, the mean LF thickness of the mild group and the severe group was 4.99 mm and 6.09 mm, respectively. The baseline LF thickness was the only major contributing factor to claudication severity in the multivariate logistic regression analysis among various radiologic factors. The LF area was a weak predictor of claudication severity. Interestingly, disc herniation area, and fat area were weakly inversely correlated to the claudication severity, which was similar to the results of a previous study.^[5]

In a recent literature review, several authors proposed that decompression-alone surgery, rather than fusion, is sufficient for LSCS surgical treatment.^[45–54] Because we empirically known that LF removal is highly effective in treating spinal stenosis, LF removal often has been the target of LSCS surgical treatment. However, the quantitative evidence regarding the contribution of LFH to the clinical symptoms of LSCS compared to other contributors, such as facet hypertrophy and disc bulging, is lacking. Therefore, the most effective surgical option remains controversial as there is no pathophysiological evidence.^[36,47,49] Herein, we also studied the outcomes by the surgical method based on the radiologic significance of LF as a main contributor of LSCS. Except for the dural sac area, the 2 groups had no significant differences from baseline. Moreover, the types of surgery (decompression alone vs fusion) did not have much influence on their clinical outcomes, including the back and leg VAS, ODI, and SF-36 PCS. Similar to the previous studies, we also observed that performing decompression alone had satisfactory outcomes compared to fusion as a surgical treatment option for LSCS. Considering that fusion surgery removes all 3 elements of LSCS, the LF, disc, and facet, while the decompression-alone surgery removes only the LF, these results suggest that LF removal is crucial for LSCS treatment. This is consistent with the finding that LF thickness was positively correlated with the severity of clinical symptoms before treatment. Taken together, the results of this study provide scientific evidence that LFH plays an important role in LSCS symptoms and that removal of the LF is crucial for surgical treatment.

4.1. Limitations

Although quantitative evidence for decompression alone on LSCS was identified in this study, there are some limitations. In our study, only those with single-level L4–L5 LSCS was selected, which does not represent the entire spectrum of LSCS. Further analysis should be performed at each level, as spinal stenosis at multiple levels is more common than single-segment stenosis. Moreover, since cases with instability were excluded, the results cannot be generalized to LSCS patients with spondylolisthesis. Despite these limitations, to the best of our knowledge, this is the first report to analyze the quantitative role of stenosis-composing factors on the severity of claudication.

5. Conclusion

LF thickness is a major factor contributing to claudication severity.

Author contributions

Conceptualization: Junseok W Hur, Joohyun Kim

Data curation: Joohyun Kim, Junseok W Hur, Woo-Keun Kwon, Subum Lee, Hyunwook Cho, Dong Uk Jin

Formal analysis: Joohyun Kim, Junseok W Hur

Funding acquisition: Junseok W Hur

Resources: Junseok W Hur, Jang-Bo Lee, Jung-Yul Park

Supervision: Junseok W Hur, Subum Lee, Jang-Bo Lee, Jung-Yul Park, Eui Yub Jung

Writing: Joohyun Kim, Junseok W Hur, Woo-Keun Kwon, Hyunwook Cho, Subum Lee, Jang-Bo Lee, Jung-Yul Park, Dong Uk Jin, Eui Yub Jung

Acknowledgements

We would like to thank Dr Jennifer J.K. Bai and Editage (www.editage.co.kr) for English language editing.