Abstract
Background
Posterior decompression is a common treatment for cervical OPLL but may accelerate intervertebral disc degeneration by disrupting posterior tension-band structures. The long-term behavior of intervertebral disc height (IDH) after surgery remains poorly understood.
Purpose
To evaluate the prevalence and radiographic predictors of long-term IDH loss following posterior cervical surgery for OPLL.
Methods
This single-center retrospective study included 108 patients who underwent posterior decompression for cervical OPLL between 2007 and 2020, with ≥ 3 years of follow-up. Total C3–C7 IDH was measured on midsagittal CT preoperatively and at final follow-up. Patients were grouped by IDH loss (> 1 mm vs. ≤ 1 mm). Radiographic and clinical parameters were compared. Multivariate logistic regression with AIC-based stepwise selection identified independent predictors. Model performance was assessed by AUC and F1 score.
Results
At a median follow-up of 61 months, 24.1% of patients exhibited IDH loss > 1 mm. The IDH group had a larger preoperative C0–2 Cobb angle (27.4° vs. 22.9°, P = 0.008), higher C2 slope (17.9° vs. 13.2°, P = 0.014), and greater total disc height (11.8 mm vs. 10.7 mm, P = 0.040). Multivariate analysis identified preoperative OALL at C5/6 (OR = 3.02, 95% CI 1.08–8.49, P = 0.036) and larger C0–2 Cobb angle (OR = 1.10, 95% CI 1.02–1.17, P = 0.009) as independent risk factors, whereas OPLL at C6 was protective (OR = 0.26, 95% CI 0.09–0.77, P = 0.015). The final model yielded an AUC of 0.74 and F1 score of 0.52.
Conclusions
Approximately one-quarter of patients developed significant IDH loss after posterior cervical surgery for OPLL. Larger cranio-cervical extension (C0–2 Cobb), OALL at C5/6, and absence of OPLL at C6 were associated with greater risk. Preoperative sagittal alignment and ossification patterns may help identify patients at risk for postoperative disc degeneration.
Keywords: OPLL, Intervertebral disc height, Posterior decompression, Cranio-cervical alignment, OALL, CT-based measurement
Introduction
Ossification of the posterior longitudinal ligament (OPLL) is a progressive disorder characterized by ectopic calcification of the posterior longitudinal ligament, most commonly involving the cervical spine [1]. Its prevalence is estimated at ≈ 1.9–4.3% in East-Asian populations and ≈ 0.1–1.7% in European and North-American cohorts [2, 3], and up to 86.7% of affected individuals eventually develop clinical symptoms [4]. Computed tomography (CT) is regarded as the imaging gold standard because it delineates the type and extent of ossification with high fidelity—particularly for small or localized lesions—whereas magnetic resonance imaging (MRI) is superior for evaluating spinal-cord compression but less reliable for visualizing the ossified mass itself [5].
When OPLL-related compression produces significant neurological deficits, surgical intervention is warranted [6]. Posterior decompression procedures such as laminectomy and laminoplasty are widely used; although they do not directly address the intervertebral disc, they disrupt key posterior tension-band elements—including the laminae, spinous processes, and facet joints—that are critical for maintaining cervical stability and disc height [7, 8]. Resection of these structures redistributes mechanical loads, elevates intradiscal pressure, accelerates degeneration, and can precipitate progressive intervertebral disc-height (IDH) loss [9]. Moreover, the ossified ligament itself modulates segmental biomechanics: segments in which the ossification bridges the disc space generally retain greater disc height and signal intensity, whereas non-bridging segments show more severe height loss and degeneration, suggesting that the ossified mass can act as an anterior-column buttress that limits motion [10]. Because postoperative IDH loss is implicated in neck pain, suboptimal functional recovery, and sagittal imbalance, a systematic analysis of the clinical and imaging factors that influence disc-height reduction after posterior cervical surgery for OPLL is still needed.
This study aimed to identify clinical and imaging factors associated with long-term IDH reduction after posterior cervical surgery for OPLL.
Materials and methods
Patients
This single-centre retrospective cohort study reviewed all consecutive patients diagnosed with cervical OPLL who underwent posterior cervical surgery at Peking University Third Hospital between January 2007 and August 2020. The study complied with the Declaration of Helsinki and was approved by the institutional review board of our institution (approval No. I IRB00006761-M2024489).
Inclusion criteria were: (1) availability of baseline cervical plain radiography, computed tomography (CT), and magnetic-resonance imaging (MRI); (2) a time interval ≥ 3 years between the initial diagnosis and the last follow-up; and (3) cervical radiography, CT, and MRI obtained at the final follow-up.
Exclusion criteria were: (1) severe cervical scoliosis that impeded accurate measurement of disc height; (2) a prior history of cervical spine surgery; and (3) receipt of any additional anterior cervical procedures or instrumented fusion surgery.
The patient-selection flowchart is illustrated in Fig. 1.
Fig. 1.
Flowchart of patient enrollment and group classification
Radiological and clinical data collection
Clinical data were collected for all patients, including age, sex, surgical approach, surgical levels, and follow-up duration. Cervical sagittal alignment and range of motion (ROM) parameters were measured on lateral radiographs in the neutral, maximal flexion, and maximal extension positions. Specifically, the following parameters were assessed: the C0–2 Cobb angle (the angle between the McGregor line and the inferior endplate of C2), the C2–7 Cobb angle (between the inferior endplates of C2 and C7), the T1 slope (the angle between the superior endplate of T1 and the horizontal line), the C2 slope (between the inferior endplate of C2 and the horizontal line), and the sagittal vertical axis (SVA; the horizontal distance between the vertical plumb line from the center of C2 and the posterosuperior corner of C7). In addition, C2–7 Cobb angles were measured in both flexion and extension positions to calculate cervical ROM: extension ROM = extension Cobb angle − neutral Cobb angle; flexion ROM = neutral Cobb angle − flexion Cobb angle; total ROM = extension Cobb angle − flexion Cobb angle.
On CT images, morphological characteristics of OPLL were assessed, including the maximum thickness (b, the largest anteroposterior diameter of the ossified lesion), the anteroposterior diameter of the spinal canal at the site of maximum OPLL thickness (a), the canal occupation ratio (b/a), the maximum craniocaudal length of the lesion, the minimum available canal area (i.e., the smallest cross-sectional area of the spinal canal in sagittal or axial planes before surgery), and the OPLL volume. Furthermore, each vertebral and intervertebral disc level was evaluated for the presence of OPLL and ossification of the anterior longitudinal ligament (OALL), and the OPLL morphology was classified into continuous, segmental, mixed, or localized types according to conventional criteria. For disc-level OPLL, “connection” was defined as the presence of ossification at the intervertebral level that either formed a bony bridge with adjacent vertebral lesions or extended cranially or caudally beyond half the height of the adjacent vertebral body [11].
Spinal cord changes were evaluated on MRI by assessing the presence of high signal intensity (ISI) on T2-weighted images.
Measurement of intervertebral disc height and grouping criteria
The IDH of each cervical segment was measured on midsagittal CT images. For disc segments C3/4 to C6/7, both anterior and posterior heights were assessed, with the average of these values denoting the segmental IDH (Fig. 2). The total C3–C7 IDH was calculated by summing the heights of C3/4, C4/5, C5/6, and C6/7 segments. The final result is the average of the measurements taken by the two physicians.
Fig. 2.
Representative sagittal reconstructed CT images demonstrating the measurement of intervertebral disc height (IDH) from C3 to C7. a Preoperative image showing clearly demarcated disc spaces with a total C3–7 IDH of 12.30 mm. b Postoperative image showing decreased intervertebral spaces with a total C3–7 IDH of 10.20 mm. The postoperative reduction exceeded 1 mm, meeting the criteria for the IDH group
IDH loss was defined as the difference between preoperative and final follow-up total C3–C7 IDH. Patients were categorized into an IDH group (with total C3–C7 IDH loss > 1 mm) and a NIDH group (with total C3–C7 IDH loss ≤ 1 mm) based on final follow-up changes.
Statistical analyses
All statistical analyses were conducted using Python 3.10 with the Pandas, NumPy, SciPy, Statsmodels, and Matplotlib libraries. The distribution of continuous variables was assessed visually. Non-normally distributed variables were presented as medians with interquartile ranges (IQR) and compared using the Mann–Whitney U test, while normally distributed variables were analyzed using the independent-samples t-test. Categorical variables were summarized as frequencies and percentages, and intergroup comparisons were performed using the chi-square test or Fisher’s exact test, as appropriate.
To identify independent risk factors for intervertebral disc height loss, multivariate logistic regression was performed using variables with P < 0.10 in univariate analysis as candidate predictors. A stepwise backward elimination procedure was applied based on the Akaike Information Criterion (AIC) to determine the optimal model, regardless of the individual P-values. All tests were two-tailed, and P < 0.05 was considered statistically significant unless otherwise indicated.
Results
Patient characteristics
A total of 108 patients diagnosed with cervical OPLL who underwent posterior cervical decompression surgery were included. The mean total IDH from C3 to C7 was 11.00 ± 2.31 mm preoperatively and 10.53 ± 2.33 mm postoperatively, with a mean IDH loss of 0.48 ± 0.76 mm. A total of 26 patients (24.1%) were classified into the IDH loss group. Figure 3a, b illustrate the segmental distribution of OPLL and OALL before and after surgery. OPLL was predominantly located at the vertebral levels, whereas OALL was more frequently observed at the intervertebral disc levels. In both preoperative and postoperative assessments, OPLL was most commonly distributed at the C4, C5, and C6 vertebrae, while OALL primarily involved the C4/5, C5/6, and C6/7 disc spaces. Table 1 summarizes cohort demographics and surgical details, and presents both preoperative and postoperative radiographic parameters. Table 2 presents the IDH values at each C3–C7 level, showing a significant postoperative reduction in IDH across all segments.
Fig. 3.
Segmental distribution of ossification lesions at each cervical level before and after surgery. a Distribution of ossification of the posterior longitudinal ligament (OPLL) across cervical vertebral levels (C2–C7). The OPLL lesions were most frequently located at the C4, C5, and C6 vertebral levels both preoperatively and postoperatively. b Distribution of ossification of the anterior longitudinal ligament (OALL) across intervertebral disc levels. OALL was predominantly observed at the C4/5, C5/6, and C6/7 disc levels
Table 1.
Baseline and postoperative characteristics of patients
| Variable | Sub-variable | Pre-op | Post-op |
|---|---|---|---|
| Age (years) | 54.31 ± 9.25 | ||
| Sex | Male | 70 (64.8%) | |
| Female | 38 (35.2%) | ||
| Follow-up duration (months) | 61.00 (47.00, 89.25) | ||
| Upper surgical level | 1 | 3 (2.8%) | |
| 2 | 29 (26.9%) | ||
| 3 | 73 (67.6%) | ||
| 4 | 3 (2.8%) | ||
| Lower surgical level | 5 | 1 (0.9%) | |
| 6 | 9 (8.3%) | ||
| 7 | 98 (90.7%) | ||
| Number of surgical levels | 3 | 1 (0.9%) | |
| 4 | 9 (8.3%) | ||
| 5 | 69 (63.9%) | ||
| 6 | 26 (24.1%) | ||
| 7 | 3 (2.8%) | ||
| Reoperation | Yes | 8 (7.4%) | |
| OPLL maximum thickness (mm) | 5.60 ± 1.84 | 6.32 ± 1.97 | |
| AP diameter at thickest point (mm) | 10.63 ± 1.03 | ||
| Occupying ratio of OPLL | 0.53 ± 0.16 | ||
| Maximum OPLL length (mm) | 58.70 ± 28.78 | 69.62 ± 30.37 | |
| OPLL volume (mm3) | 1622.17 (751.54, 3047.50) | 2662.34 (1161.29, 5156.80) | |
| OPLL index | 5.00 (3.00, 8.00) | 7.00 (4.00, 9.00) | |
| OPLL classification | Continuous | 15 (13.9%) | 29 (26.9%) |
| Segmental | 56 (51.9%) | 37 (34.3%) | |
| Mixed | 30 (27.8%) | 36 (33.3%) | |
| Localized | 7 (6.5%) | 6 (5.6%) | |
| C0–2 Cobb angle (°) | 24.00 ± 7.49 | 26.26 ± 8.01 | |
| C2–7 Cobb angle (°) | 7.89 ± 9.26 | 8.30 (2.35, 14.47) | |
| T1 slope (°) | 25.32 ± 8.07 | 25.40 (20.38, 31.92) | |
| C2 slope (°) | 14.30 ± 8.60 | 14.20 (10.55, 22.15) | |
| Sagittal vertical axis (mm) | 24.30 (16.80, 32.88) | 27.50 (20.35, 37.42) | |
| △SVA (mm) | −3.50 (−9.10, 0.95) | ||
| C2–7 Cobb angle in extension (°) | 18.45 ± 9.70 | 14.80 (7.72, 20.05) | |
| C2–7 Cobb angle in flexion (°) | −15.05 (−21.52, −9.53) | −5.95 (−11.15, 0.12) | |
| Extension ROM (°) | 9.10 (5.65, 14.05) | 5.40 (3.35, 7.43) | |
| Flexion ROM (°) | 22.70 (16.95, 29.95) | 11.60 (7.75, 17.00) | |
| Total cervical ROM (°) | 32.40 ± 12.78 | 17.45 (13.57, 22.38) | |
| High-signal intensity in spinal cord | Grade 0 | 21 (19.4%) | 33 (30.6%) |
| Grade 1 | 76 (70.4%) | 61 (56.5%) | |
| Grade 2 | 11 (10.2%) | 14 (13.0%) | |
| Total C3–7 disc height (mm) | 11.00 ± 2.31 | 10.53 ± 2.33 | |
| C3–7 disc height loss (mm) | 0.48 ± 0.76 |
Continuous variables are presented as mean ± standard deviation (SD) or median with interquartile range (IQR) where appropriate. Categorical variables are presented as number and percentage. Variables assessed only at baseline, or without distinction between preoperative and postoperative status (e.g., age, sex, surgical levels, follow-up duration, reoperation), are presented in the Pre-op column for clarity. OPLL, ossification of the posterior longitudinal ligament; ROM, range of motion; SVA, sagittal vertical axis.
Table 2.
Segmental disc heights before and after surgery
| Segment | Preoperative (mean ± SD) | Postoperative (mean ± SD) | Mean difference | Statistic | P-value |
|---|---|---|---|---|---|
| C3–4 | 2.65 ± 0.64 | 2.54 ± 0.65 | 0.10 | 1695.0 | 0.014 |
| C4–5 | 2.75 ± 0.74 | 2.61 ± 0.74 | 0.14 | 1265.0 | < 0.001 |
| C5–6 | 2.73 ± 0.76 | 2.66 ± 0.78 | 0.07 | 1808.0 | 0.041 |
| C6–7 | 2.88 ± 0.88 | 2.72 ± 0.89 | 0.16 | 1376.5 | < 0.001 |
| C3–7 total | 11.00 ± 2.31 | 10.53 ± 2.33 | 0.48 | 1173.5 | < 0.001 |
Values are presented as mean ± standard deviation (SD). Paired comparisons were performed using the Wilcoxon signed-rank test. “Mean Difference” refers to postoperative minus preoperative disc height. C3–7 total refers to the sum of disc heights from C3/4 to C6/7.
Factors associated with postoperative disc height reduction
Based on the extent of postoperative C3–C7 intervertebral disc height reduction, patients were categorized into an IDH group (n = 26) and an NIDH group (n = 82). Compared with the NIDH group, the IDH group exhibited a significantly greater total disc height loss (1.55 ± 0.34 mm vs. 0.13 ± 0.48 mm, p < 0.001).
Several preoperative radiographic parameters also significantly differed between the two groups. Patients in the IDH group had a significantly larger preoperative C0–2 Cobb angle (27.37 ± 6.93° vs. 22.93 ± 7.38°, p = 0.008), a higher C2 slope (17.90 ± 8.34° vs. 13.16 ± 8.41°, p = 0.014), and a greater preoperative total C3–7 disc height (11.81 ± 2.55 mm vs. 10.74 ± 2.19 mm, p = 0.040).
A detailed comparison of demographic, surgical, and radiographic characteristics between the IDH and NIDH groups is presented in Table 3.
Table 3.
Comparison of baseline and radiographic characteristics between patients with and without postoperative disc height loss
| Variable | Sub-variable | NIDH group (n = 82) | IDH group (n = 26) | Statistic | p-value |
|---|---|---|---|---|---|
| Age (years) | 54.02 ± 8.57 | 55.19 ± 11.27 | −0.559 | 0.577 | |
| Sex | Male | 52 (63.4%) | 18 (69.2%) | 0.77 | 0.644 |
| Female | 30 (36.6%) | 8 (30.8%) | |||
| Follow-up duration (months) | 61.00 (47.00, 80.75) | 63.50 (47.25, 91.50) | 1041 | 0.86 | |
| Upper surgical level | 1 | 2 (2.4%) | 1 (3.8%) | 1.426 | 0.699 |
| 2 | 23 (28.0%) | 6 (23.1%) | |||
| 3 | 54 (65.9%) | 19 (73.1%) | |||
| 4 | 3 (3.7%) | 0 (0.0%) | |||
| 5 | 0 (0.0%) | 1 (3.8%) | 3.718 | 0.156 | |
| 6 | 6 (7.3%) | 3 (11.5%) | |||
| 7 | 76 (92.7%) | 22 (84.6%) | |||
| Number of surgical levels | 3 | 0 (0.0%) | 1 (3.8%) | 3.937 | 0.415 |
| 4 | 6 (7.3%) | 3 (11.5%) | |||
| 5 | 54 (65.9%) | 15 (57.7%) | |||
| 6 | 20 (24.4%) | 6 (23.1%) | |||
| 7 | 2 (2.4%) | 1 (3.8%) | |||
| Pre-op OPLL at C2 | 25 (30.5%) | 6 (23.1%) | 0.684 | 0.62 | |
| Pre-op OPLL at C2/3 | 28 (34.1%) | 9 (34.6%) | 1.021 | 1 | |
| Pre-op OPLL at C3 | 50 (61.0%) | 12 (46.2%) | 0.549 | 0.255 | |
| Pre-op OPLL at C3/4 | 30 (36.6%) | 12 (46.2%) | 1.486 | 0.489 | |
| Pre-op OPLL at C4 | 64 (78.0%) | 22 (84.6%) | 1.547 | 0.583 | |
| Pre-op OPLL at C4/5 | 34 (41.5%) | 15 (57.7%) | 1.925 | 0.178 | |
| Pre-op OPLL at C5 | 67 (81.7%) | 22 (84.6%) | 1.231 | 1 | |
| Pre-op OPLL at C5/6 | 39 (47.6%) | 8 (30.8%) | 0.49 | 0.174 | |
| Pre-op OPLL at C6 | 62 (75.6%) | 15 (57.7%) | 0.44 | 0.088 | |
| Pre-op OPLL at C6/7 | 21 (25.6%) | 7 (26.9%) | 1.07 | 1 | |
| Pre-op OPLL at C7 | 43 (52.4%) | 9 (34.6%) | 0.48 | 0.123 | |
| Pre-op OALL at C2 | 82 (100.0%) | 26 (100.0%) | 0 | 1 | |
| Pre-op OALL at C2/3 | 10 (12.2%) | 5 (19.2%) | 1.714 | 0.349 | |
| Pre-op OALL at C3 | 5 (6.1%) | 3 (11.5%) | 2.009 | 0.396 | |
| Pre-op OALL at C3/4 | 8 (9.8%) | 3 (11.5%) | 1.207 | 0.724 | |
| Pre-op OALL at C4 | 5 (6.1%) | 2 (7.7%) | 1.283 | 0.673 | |
| Pre-op OALL at C4/5 | 18 (22.0%) | 10 (38.5%) | 2.222 | 0.123 | |
| Pre-op OALL at C5 | 9 (11.0%) | 4 (15.4%) | 1.475 | 0.509 | |
| Pre-op OALL at C5/6 | 36 (43.9%) | 17 (65.4%) | 2.414 | 0.072 | |
| Pre-op OALL at C6 | 10 (12.2%) | 5 (19.2%) | 1.714 | 0.349 | |
| Pre-op OALL at C6/7 | 31 (37.8%) | 11 (42.3%) | 1.206 | 0.818 | |
| Pre-op OALL at C7 | 5 (6.1%) | 5 (19.2%) | 3.667 | 0.058 | |
| Pre-op OPLL maximum thickness (mm) | 5.66 ± 1.86 | 5.41 ± 1.81 | 0.594 | 0.554 | |
| Pre-op AP diameter at thickest point (mm) | 10.60 ± 1.09 | 10.71 ± 0.79 | −0.486 | 0.628 | |
| Occupying ratio of OPLL | 0.53 ± 0.17 | 0.50 ± 0.15 | 0.893 | 0.374 | |
| Pre-op maximum OPLL length (mm) | 60.34 ± 28.91 | 53.55 ± 28.31 | 1.049 | 0.297 | |
| Minimum available canal area (mm²) | 115.70 (83.67, 152.55) | 129.65 (100.15, 157.80) | 876 | 0.173 | |
| Pre-op C0–2 Cobb angle (°) | 22.93 ± 7.38 | 27.37 ± 6.93 | −2.709 | 0.008 | |
| Pre-op C2–7 Cobb angle (°) | 8.79 ± 9.05 | 5.06 ± 9.55 | 1.808 | 0.073 | |
| Pre-op T1 slope (°) | 25.24 ± 8.10 | 25.57 ± 8.15 | −0.177 | 0.86 | |
| Pre-op C2 slope (°) | 13.16 ± 8.41 | 17.90 ± 8.34 | −2.507 | 0.014 | |
| Pre-op sagittal vertical axis (mm) | 23.85 (16.73, 31.32) | 28.25 (18.10, 36.80) | 890.5 | 0.209 | |
| Pre-op C2–7 Cobb angle in extension (°) | 18.87 ± 9.73 | 17.13 ± 9.67 | 0.792 | 0.43 | |
| Pre-op C2–7 Cobb angle in flexion (°) | −14.90 (−19.65, −8.97) | −17.50 (−22.80, −10.62) | 1277 | 0.13 | |
| Pre-op Extension ROM (°) | 8.75 (4.83, 13.80) | 10.25 (7.88, 14.65) | 882 | 0.187 | |
| Pre-op Flexion ROM (°) | 22.50 (17.92, 29.47) | 23.15 (14.72, 30.20) | 1086.5 | 0.886 | |
| Pre-op Total cervical ROM (°) | 31.59 ± 12.54 | 34.92 ± 13.45 | −1.159 | 0.249 | |
| Pre-op High-signal intensity in spinal cord | 0 | 16 (19.5%) | 5 (19.2%) | 0.069 | 0.966 |
| 1 | 58 (70.7%) | 18 (69.2%) | |||
| 2 | 8 (9.8%) | 3 (11.5%) | |||
| Pre-op OPLL classification | Continuous | 13 (15.9%) | 2 (7.7%) | 3.134 | 0.371 |
| Segmental | 39 (47.6%) | 17 (65.4%) | |||
| Mixed | 25 (30.5%) | 5 (19.2%) | |||
| Localized | 5 (6.1%) | 2 (7.7%) | |||
| Post-op OPLL classification | Continuous | 22 (26.8%) | 7 (26.9%) | 0.42 | 0.936 |
| Segmental | 29 (35.4%) | 8 (30.8%) | |||
| Mixed | 27 (32.9%) | 9 (34.6%) | |||
| Localized | 4 (4.9%) | 2 (7.7%) | |||
| Pre-op total C3–7 disc height (mm) | 10.74 ± 2.19 | 11.81 ± 2.55 | −2.079 | 0.04 | |
| Post-op total C3–7 disc height (mm) | 10.61 ± 2.29 | 10.26 ± 2.50 | 0.664 | 0.508 | |
| C3–7 disc height loss (mm) | 0.13 ± 0.48 | 1.55 ± 0.34 | −13.956 | < 0.001 | |
| Connection | C3-4 | 14 (17.1%) | 4 (15.4%) | 0.883 | 1 |
| C4-5 | 13 (15.9%) | 3 (11.5%) | 0.692 | 0.757 | |
| C5-6 | 15 (18.3%) | 2 (7.7%) | 0.372 | 0.353 | |
| C6-7 | 7 (8.5%) | 3 (11.5%) | 1.398 | 0.701 | |
| Number of connected disc segments | 0 | 52 (63.4%) | 18 (69.2%) | 3.662 | 0.454 |
| 1 | 18 (22.0%) | 6 (23.1%) | |||
| 2 | 7 (8.5%) | 0 (0.0%) | |||
| 3 | 3 (3.7%) | 2 (7.7%) | |||
| 4 | 2 (2.4%) | 0 (0.0%) | |||
Continuous variables are presented as mean ± standard deviation (SD) or median with interquartile range (Q1, Q3) as appropriate. Categorical variables are presented as number and percentage. Statistical tests used include the Mann–Whitney U test, Wilcoxon signed-rank test, and chi-square test depending on the type and distribution of the variable.
OPLL, ossification of the posterior longitudinal ligament; OALL, ossification of the anterior longitudinal ligament; ROM, range of motion; SVA, sagittal vertical axis.
Multivariate analysis of risk factors for postoperative disc height loss
Variables with a p-value < 0.10 in the univariate analysis were entered into a multivariate logistic regression model. Stepwise backward selection based on the AIC was used to identify the optimal set of predictors.
The final model demonstrated that the presence of preoperative OALL at C5/6 (odds ratio [OR] = 3.02, 95% CI 1.08–8.49, p = 0.036) and a larger C0–2 Cobb angle (OR = 1.10, 95% CI 1.02–1.17, p = 0.009) were independently associated with an increased risk of postoperative disc height loss. Conversely, the presence of OPLL at the C6 level was independently associated with a reduced risk (OR = 0.26, 95% CI 0.09–0.77, p = 0.015) (Table 4). Figure 4a shows the corresponding forest plot of the multivariate model.
Table 4.
Multivariate logistic regression analysis of risk factors for postoperative C3–7 disc height loss
| Variable | Coef. | P-value | OR | 95% CI lower | 95% CI upper |
|---|---|---|---|---|---|
| Const | −3.15 | < 0.001 | 0.04 | 0.01 | 0.28 |
| Pre-op OALL at C5/6 | 1.11 | 0.036 | 3.02 | 1.08 | 8.49 |
| Pre-op OPLL at C6 | −1.34 | 0.015 | 0.26 | 0.09 | 0.77 |
| Pre-op C0–2 Cobb angle (°) | 0.09 | 0.009 | 1.10 | 1.02 | 1.17 |
Multivariate logistic regression analysis was performed using variables with p < 0.1 in the univariate analysis. Odds ratios (ORs) and 95% confidence intervals (CIs) are reported. Preoperative OALL at C5/6 and a larger C0–2 Cobb angle were identified as significant risk factors for postoperative disc height loss, whereas the presence of OPLL at C6 was associated with a protective effect.
Fig. 4.
a Forest plot of independent predictors for postoperative disc height loss > 1 mm at C3–7. b ROC curve of the final logistic regression model. a Forest plot shows odds ratios (ORs) and 95% confidence intervals (CIs) for variables selected by AIC-based stepwise logistic regression. Red markers indicate risk factors (OR > 1); blue markers indicate protective factors (OR < 1). Asterisks denote significance levels (* p < 0.05, ** p < 0.01, *** p < 0.001). b Receiver operating characteristic (ROC) curve for the final model (AUC = 0.74), indicating moderate discrimination. The optimal cutoff was determined using the Youden index
The predictive performance of the model was moderate, with an AUC of 0.74 (95% CI 0.64–0.84). The model achieved a sensitivity of 0.73 (95% CI 0.47–0.90), a specificity of 0.65 (95% CI 0.43–0.87), a precision of 0.40 (95% CI 0.26–0.61), and an F1 score of 0.52 (95% CI 0.37–0.65) (Table 5). The ROC curve is displayed in Fig. 4b.
Table 5.
Predictive performance of the final logistic regression model
| Metric | AUC | Sensitivity | Specificity | Precision | F1 |
|---|---|---|---|---|---|
| Value with 95% CI | 0.74 (0.64, 0.84) | 0.73 (0.47, 0.9) | 0.65 (0.43, 0.87) | 0.40 (0.26, 0.61) | 0.52 (0.37, 0.65) |
The performance of the final multivariate logistic regression model was evaluated using receiver operating characteristic (ROC) analysis. The model achieved an area under the curve (AUC) of 0.74 (95% CI 0.64–0.84). Sensitivity, specificity, precision, and F1 score with 95% confidence intervals were estimated using bootstrapping (1000 iterations). The optimal cutoff point was determined according to Youden’s index.
Subgroup analyses by OPLL subtype and ligamentous connection
To further explore potential factors associated with disc height loss, we performed subgroup analyses according to OPLL subtype and the presence of ligamentous connection at each disc level. As shown in Table 6, patients with continuous or mixed OPLL exhibited a tendency toward smaller postoperative disc height loss and a lower proportion of cases with loss > 1 mm compared with those with segmental or localized OPLL, although these differences did not reach statistical significance.
Table 6.
Comparison of postoperative disc height loss between different OPLL subtypes
| Group | Segmental/localized (n = 63) | Continuous/mixed (n = 45) | P-value |
|---|---|---|---|
| Height loss (mm) | 0.58 ± 0.78 | 0.32 ± 0.70 | 0.115 |
| Height loss > 1 mm | 19 | 7 | 0.128 |
Values are presented as mean ± standard deviation (SD) or number of cases, as appropriate. p-values were calculated using the Mann–Whitney U test for continuous variables and the chi-square test for categorical variables.
Table 7 demonstrates the relationship between ligamentous connection and segmental disc height loss. At C3/4, C4/5, and C5/6 levels, disc height loss was consistently smaller in connected segments than in unconnected ones (e.g., C3/4: 0.01 ± 0.17 mm vs. 0.12 ± 0.34 mm), whereas at C6/7, the difference was minimal and slightly reversed. Although not statistically significant, these findings suggest that ligamentous connection may contribute to maintaining disc height following surgery.
Table 7.
Comparison of preoperative disc heights and postoperative disc height loss between connected and unconnected segments
| Segment | Pre-op height (connected) | Pre-op height (unconnected) | p-value | Height loss (connected) | Height loss (unconnected) | p-value |
|---|---|---|---|---|---|---|
| C3/4 | 2.36 ± 0.57 | 2.70 ± 0.64 | 0.041 | 0.01 ± 0.17 | 0.12 ± 0.34 | 0.159 |
| C4/5 | 2.59 ± 0.90 | 2.77 ± 0.71 | 0.371 | 0.12 ± 0.26 | 0.14 ± 0.32 | 0.832 |
| C5/6 | 3.06 ± 0.50 | 2.67 ± 0.78 | 0.049 | 0.01 ± 0.42 | 0.08 ± 0.27 | 0.339 |
| C6/7 | 3.23 ± 0.58 | 2.85 ± 0.90 | 0.191 | 0.19 ± 0.33 | 0.16 ± 0.38 | 0.780 |
This table compares the preoperative disc height and the postoperative disc height loss (defined as preoperative minus postoperative average disc height) between segments with and without ligamentous connection at each cervical disc level. Data are presented as mean ± standard deviation. Statistical comparisons were performed using the independent samples t-test.
Discussion
This study systematically investigated the determinants of postoperative C3–C7 intervertebral disc height reduction in patients with cervical OPLL undergoing posterior decompression. Approximately one-quarter of patients experienced clinically relevant disc height loss (>1 mm, ≈ 9–10% of baseline), consistent with thresholds used in prior degeneration studies [12]. Multivariate analysis identified preoperative OALL at C5/6 and a larger C0–2 Cobb angle as independent correlates of postoperative IDH reduction, whereas the presence of OPLL at C6 was associated with a protective effect. Subgroup analyses additionally suggested that continuous/mixed OPLL and ligamentous connection were associated with smaller disc height loss, although these associations did not reach statistical significance. Together, these findings highlight the importance of sagittal alignment and ossification distribution in influencing postoperative disc behavior [13–16].
In univariate analysis, a higher C2 slope, larger C0–2 Cobb angle, and greater preoperative total C3–C7 disc height were associated with postoperative disc height loss. These findings are consistent with prior studies indicating that increased C2 slope and cranio-cervical extension reflect sagittal malalignment, elevating subaxial flexion moments and intradiscal loading once posterior stabilizers are disrupted [17, 18]. A larger C0–2 Cobb angle likely represents cranio-cervical compensation for reduced subaxial lordosis, shifting mechanical demand to the C3–C7 discs. Although both C2 slope and total disc height were significant in univariate analysis, they lost significance after adjustment, reflecting overlap with other parameters: C2 slope closely parallels cranio-cervical alignment and shares explanatory information with C0–2 Cobb [19], while total disc height appears to act as a surrogate for baseline disc mobility rather than an independent driver. Thus, C0–2 Cobb emerged as the more robust alignment marker in the multivariate model.
Beyond alignment, ossification patterns also influenced postoperative disc behavior. OALL at C5/6 independently increased the risk of height loss, possibly by stiffening a highly mobile level and predisposing it to collapse, whereas OPLL at C6 exerted a protective effect, acting as a form of “biological fusion.” These results are directionally consistent with prior work showing that continuous or bridging ossification stabilizes motion segments and mitigates postoperative degeneration [20, 21]. In our subgroup analyses, patients with continuous or mixed OPLL and segments with ligamentous connection generally exhibited less disc height reduction, though not statistically significant, suggesting that ossification distribution and ligamentous bridging may help buffer against postoperative collapse.
The clinical implication of our findings is that certain preoperative features—particularly a larger C0–2 Cobb angle, OALL at C5/6, and the absence of OPLL at C6—may serve as radiographic risk markers for postoperative disc height reduction. Identifying these features could help guide patient counseling and justify closer imaging follow-up in higher-risk individuals. The potential protective role of OPLL at C6 may be partly explained by biomechanical factors: increased stiffness of the posterior longitudinal ligament and other cervical ligaments has been shown to reduce segmental ROM, particularly at mobile levels such as C5–C6 [22], while coupled-motion studies demonstrated relatively high strain at C6–C7, suggesting that ossification at this level might redistribute loads and attenuate disc stress [23]. Importantly, our aim was not to build a predictive model but to explore associations; the moderate performance of the regression model (AUC 0.74) should therefore be interpreted as supportive evidence rather than a clinically applicable tool.
Several limitations should be noted. This was a single-center retrospective study, raising the possibility of selection bias. Functional outcomes were not analyzed, so the clinical impact of disc height reduction remains uncertain. Surgical heterogeneity, including the type and extent of posterior decompression, may also have influenced results. In addition, posterior procedures in general produce only modest disc height reduction, suggesting that our findings should be viewed as preliminary. Finally, OPLL progression was not directly assessed, although it may contribute to long-term disc degeneration. Larger, multicenter prospective studies incorporating both radiographic and clinical outcomes are required to validate these observations and clarify the complex interplay between sagittal alignment, ossification, and postoperative disc changes.
Conclusion
This study identified radiographic correlates of postoperative C3–C7 disc-height loss after posterior decompression for cervical OPLL. A larger preoperative C0–2 Cobb angle and the presence of OALL at C5/6 were independently associated with greater height reduction, whereas OPLL at C6 conferred a protective effect. Continuous/mixed OPLL morphology and ligamentous connection showed directionally smaller losses but without statistical significance. These findings underscore the biomechanical relevance of cranio-cervical alignment and ossification distribution, suggesting value for preoperative risk discussion and targeted follow-up; our aim was factor identification rather than prediction, and larger multicenter studies are needed to confirm these associations.
Acknowledgements
Not applicable.
Author contributions
S.Q., R.Q. wrote the main manuscript text. S.Q., R.Q. and W.Z. prepared Figs. 1, 2, 3 and 4. S.Q., R.Q, W.Z. and R.Y. contributed to data analysis and interpretation. N.L. supervised the study and revised the manuscript critically. All authors reviewed and approved the final manuscript.
Funding
This work was supported by the National Natural Science Foundation of China (Grant Number 82371921) and Proof of Concept Program of Zhongguancun Science City and Peking University Third Hospital (Grant Number HDCXZHKC2022202).
Data availability
The data that support the findings of this study are available on request from the corresponding author, Ning Lang, upon reasonable request.
Declarations
Ethics approval and consent to participate
The study complied with the Declaration of Helsinki and was approved by the institutional review board of our institution (approval No. I IRB00006761-M2024489). Given the retrospective design of this study, informed consent from participants was waived in accordance with ethical guidelines.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Siyuan Qin MD, Ruomu Qu MD, Weili Zhao MD and Ruixin Yan contributed equally to this work.
References
- 1.Smith ZA, Buchanan CC, Raphael D, Khoo LT. Ossification of the posterior longitudinal ligament: pathogenesis, management, and current surgical approaches. 2011. 10.3171/2011.1.FOCUS10256 [DOI] [PubMed]
- 2.Le HV, Wick JB, Van BW, Klineberg EO. Ossification of the posterior longitudinal ligament: pathophysiology, diagnosis, and management. J Am Acad Orthop Surg. 2022;30:820–30. 10.5435/JAAOS-D-22-00049. [DOI] [PubMed] [Google Scholar]
- 3.Belanger TA, Roh JS, Hanks SE, et al. Ossification of the posterior longitudinal ligament. Results of anterior cervical decompression and arthrodesis in sixty-one North American patients. J Bone Jt Surg Am. 2005;87:610–5. 10.2106/JBJS.C.01711. [DOI] [PubMed] [Google Scholar]
- 4.Sasaki E, Ono A, Yokoyama T, et al. Prevalence and symptom of ossification of posterior longitudinal ligaments in the Japanese general population. J Orthop Sci. 2014;19:405–11. 10.1007/s00776-014-0552-0. [DOI] [PubMed] [Google Scholar]
- 5.Abiola R, Rubery P, Mesfin A. Ossification of the posterior longitudinal ligament: Etiology, Diagnosis, and outcomes of nonoperative and operative management. Global Spine J. 2016;6:195–204. 10.1055/s-0035-1556580. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Sun N, Jiang C, Liu Y. Surgical options for ossification of the posterior longitudinal ligament of the cervical spine: a narrative review. J Orthop Surg Res. 2024;19:707. 10.1186/s13018-024-05215-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Gondar R, Jesse CM, Schär RT, et al. Preservation of the posterior interspinous ligamentary complex in posterior and transforaminal lumbar interbody fusion. Surgeries. 2023;4:600–10. 10.3390/surgeries4040058. [Google Scholar]
- 8.Merrill RK, Clohisy JC, Albert TJ, Qureshi SA. Concepts and techniques to prevent cervical spine deformity after spine surgery: A narrative review. Neurospine. 2023;20:221–30. 10.14245/ns.2244780.390. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Kumaresan S, Yoganandan N, Pintar FA, et al. Finite element modeling of cervical laminectomy with graded facetectomy. Clin Spine Surg. 1997;10:40. [PubMed] [Google Scholar]
- 10.Luo X, Sun K, Zhu J, et al. Analysis of intervertebral disc degeneration in patients with ossification of the posterior longitudinal ligament. Quant Imaging Med Surg. 2022;12:1919928–1911928. 10.21037/qims-21-154. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.S An, Lee J-B, Lee S, et al. True continuous segment of ossification of posterior longitudinal ligament is protective against postoperative early kyphosis progression after laminoplasty. Neurosurgery. 2023. 10.1227/neu.0000000000002773. [DOI] [PubMed] [Google Scholar]
- 12.Park Y-K, Moon HJ, Kwon TH, Kim JH. Long-term outcomes following anterior foraminotomy for one- or two-level cervical radiculopathy. Eur Spine J. 2013;22:1489–96. 10.1007/s00586-013-2712-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Teo AQA, Thomas AC, Hey HWD. Sagittal alignment of the cervical spine: Do we know enough for successful surgery? J Spine Surg. 2020;6:124–35. 10.21037/jss.2019.11.18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Patel PD, Arutyunyan G, Plusch K, et al. A review of cervical spine alignment in the normal and degenerative spine. J Spine Surg. 2020;6:106–23. 10.21037/jss.2020.01.10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Lee J-K, Hyun S-J, Yang SH, Kim K-J. Reciprocal changes following cervical realignment surgery. Neurospine. 2022;19:853–61. 10.14245/ns.2249938.469. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Liu T, Zhang J, Deng L, et al. Comparison of radiological and clinical outcomes of cervical laminoplasty versus lateral mass screw fixation in patients with ossification of the posterior longitudinal ligament. BMC Musculoskelet Disord. 2024;25:337. 10.1186/s12891-024-07385-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Akgün MY, Ateş Ö, Tepebaşili MA, et al. Clinical parameters of laminoplasty and laminectomy with fusion in the treatment of cervical spondylosis and analysis of postoperative sagittal balance. Turk J Med Sci. 2023;53:1458–64. 10.55730/1300-0144.5713. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Liu Z, Wang Z, Zhang P, et al. Relationship between spinocranial angle and clinical outcomes after laminoplasty in patients with ossification of the posterior longitudinal ligament. Front Surg. 2022;9:1045085. 10.3389/fsurg.2022.1045085. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Ye J, Rider SM, Lafage R, et al. Spinopelvic sagittal compensation in adult cervical deformity. J Neurosurg Spine. 2023;39:1–10. 10.3171/2023.2.SPINE221295. [DOI] [PubMed] [Google Scholar]
- 20.Mun HW, Lee JJ, Shin HC, et al. Comparative analysis of outcomes and kyphotic risk factors after cervical laminoplasty in 2 different ossification of the posterior longitudinal ligament groups and cervical spondylotic myelopathy. Neurosurgery. 2024;97:320–31. 10.1227/neu.0000000000003299. [DOI] [PubMed] [Google Scholar]
- 21.Yu Z, Lv J, Wang Z, et al. Reduced vertebral Hounsfield unit values in the bridging group of ossification-Related segments in patients with ossification of the posterior longitudinal ligament of the cervical spine. World Neurosurg. 2025;195:123638. 10.1016/j.wneu.2024.123638. [DOI] [PubMed] [Google Scholar]
- 22.Wang K, Deng Z, Wang H, et al. Influence of variations in stiffness of cervical ligaments on C5-C6 segment. J Mech Behav Biomed Mater. 2017;72:129–37. 10.1016/j.jmbbm.2017.05.005. [DOI] [PubMed] [Google Scholar]
- 23.Huang H, Nightingale RW, Dang ABC. Biomechanics of coupled motion in the cervical spine during simulated whiplash in patients with pre-existing cervical or lumbar spinal fusion. Bone Joint Res. 2018;7:28–35. 10.1302/2046-3758.71.BJR-2017-0100.R1. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data that support the findings of this study are available on request from the corresponding author, Ning Lang, upon reasonable request.