INTRODUCTION
The K-line is a useful measure for evaluating cervical curvature and informing surgical strategy in patients with cervical ossification of the posterior longitudinal ligament (OPLL) [1]. Cervical laminoplasty (LP) generally results in favorable neurological outcomes in patients with K-line (+) and multilevel OPLL involving more than 3 levels, whereas poor outcomes have been reported in K-line (-) patients [1]. However, even when LP is performed in patients with preoperative K-line (+) in a neutral position, some patients still experience suboptimal recovery.
Dynamic mechanical factors are significant contributors to the development and progression of the cervical myelopathy associated with OPLL [2,3]. Although the importance of static assessments of the K-line has been reported, this does not fully reflect dynamic cord compression during neck movement. In particular, K-line status changes according to neck flexion position and has been shown to correlate with clinical outcomes [4]. Patients who are K-line (+) in the neutral position but become K-line (-) during flexion (FK-line [-]) exhibit poor outcomes after LP [4,5]. This discrepancy indicates that flexed radiographs provide additional clinical value in surgical planning by identifying dynamic cord compression that is not apparent in the neutral position.
Although a few studies have proposed that laminectomy with fusion (LF), which restricts cervical flexion and stabilizes sagittal alignment in FK-line (-) OPLL patients [6], has potential benefits, no study has examined how instrumentation fixation affects the surgical outcomes of FK-line (-) OPLL patients. Furthermore, no comparison of surgical outcomes between LP and LF in patients with preoperative FK-line (-) has been conducted.
Among patients with cervical myelopathy caused by multilevel OPLL, this retrospective study compares clinical and radiological outcomes after LP and LF according to FK-line status. The aim was to analyze the relationship between changes in cervical parameters and neurological outcomes, especially focusing on preoperative FK-line (-) patients. Additionally, the study examines how changes in cervical parameters after surgery predict neurological recovery in FK-line (-) OPLL patients.
MATERIALS AND METHODS
1. Patient Demographics
This study was approved by the Institutional Review Board (IRB No. 2-2024-0541).
This retrospective multicenter study analyzed patients with OPLL from 3 university-affiliated hospitals. Between January 2017 and December 2023, 426 patients with multilevel OPLL (involving more than 3 vertebral levels) underwent posterior decompression surgery, either LP or LF. Among them, 69 patients were excluded due to insufficient follow-up (39 patients), the absence of initial dynamic radiographs (5 patients), or other confirmed neurological disorders (25 patients), such as Parkinson disease, cerebral infarction, or hemorrhage. Additionally, 8 patients who underwent combined anterior-posterior decompression were excluded due to small sample size. Ultimately, data for 349 patients with a minimum follow-up of 2 years were analyzed (Fig. 1). OPLL was confirmed by plain radiographs and computed tomography (CT). Cervical myelopathy was diagnosed based on radiographic evidence of cord compression and upper motor neuron signs on neurological examination. Exclusion criteria included trauma, infection, tumor, prior cervical surgery, circumferential fusion, and other neurological disorders.
2. Radiologic Assessments
The radiologic assessments were the C2 slope (C2S), C2–7 Cobb angle (CA), C2–7 sagittal vertical axis (SVA), T1 slope (T1S), C2–7 range of motion (ROM), canal occupying ratio (COR), and OPLL type. The radiographic measurements were defined as follows: C2S, the angle between the lower endplate of C2 and the horizontal plane; CA, the Cobb angle between the lower endplates of C2 and C7; SVA, the distance from the centroid of C2 to the posterior-superior corner of C7; and T1S, the angle between the upper endplate of T1 and the horizontal plane in a standing lateral view radiograph (Fig. 2A) [7]. ROM was defined as the difference between the extension CA and flexion CA. Dynamic extension reserve (DER) was calculated as the difference between the extension and neutral C2–7 CA, representing extension function [7]. OPLL was classified preoperatively on CT into segmental, continuous, mixed, or localized types [8]. COR was calculated as the maximum thickness of the OPLL divided by the anteroposterior diameter of the spinal canal at the most affected level. The K-line was drawn by connecting the midpoints of the spinal canal at C2 and C7 on lateral radiographs to determine whether the OPLL crossed (K-line [-]) or did not cross (K-line [+]) that line. We assessed the K-line in the neck-flexed position (FK-line) and indicated that the OPLL mass never crossed the K-line posteriorly (FK-line [+]) or crossed it (FK-line [-]). FK-line tilt was defined as the angle between the FK-line and a perpendicular line to the horizon during flexion. The FK-line distance was defined as the shortest distance between the FK-line and the posterior border of the vertebral body line, as shown on an OPLL lateral neck flexion radiograph in Fig. 2B. Spinal cord signal intensity was graded on T2-weighted magnetic resonance imaging (MRI) at the site of maximal compression as follows: grade 0, none; grade 1, faint and indistinct; and grade 2, intense and well-defined [2,9]. The radiological parameters in the cervical neutral and flexion states were measured and analyzed using software (ZeTTA PACS, TaeYoung Soft Co., Ltd., Korea). Radiographic measurements were collected from the database and independently reviewed by 2 surgeons (CKL and JJS). Discrepancies were resolved through consensus discussion. The representative radiographic figure demonstrates the FK-line measurement technique using annotated lines and arrows to enhance reproducibility and clarity (Fig. 3).
3. Assessment of Clinical Outcomes
Clinical outcomes were evaluated using a patient-reported visual analogue scale (VAS) for neck pain and the Japanese Orthopaedic Association (JOA) score [10]. Data were assessed preoperatively and at least 24 months postoperatively. The JOA recovery ratio was calculated as: JOA recovery ratio (%)=(postoperative JOA score–preoperative JOA score)/(17 [total score]–preoperative JOA score).
4. Surgical Procedures
The surgical policy involved assessing the patient’s comorbidities [11]. LP was preferred in patients with preserved preoperative lordosis, whereas LF was performed in patients with kyphotic alignment greater than 10° to minimize further loss of lordosis. When OPLL extended to C7 or T1, fusion was extended to the lower levels across the cervicothoracic junction. LP was performed through a midline approach with bilateral muscle detachment, removal of the spinous process, creation of unilateral gutters, and fixation using posterior cervical titanium miniplates (Centerpiece, Medtronic Sofamor Danek, USA), which maintained the “door” in the open position. LF involved laminectomy with posterior fusion using lateral mass screws (C1 and C3–6) and pedicle screws (C2 and C7), with autologous bone grafting. Postoperatively, all patients wore either a Miami collar or a Philadelphia brace for approximately 3 months. Follow-up included clinical and radiographic evaluations 1, 3, and 6 months postoperatively and every 6 months thereafter.
5. Statistical Analysis
Continuous variables are reported as the mean±standard deviation or percentage. Normally distributed data were compared using Student t-test or analysis of variance, while non-normally distributed data were compared using the Mann-Whitney U-test or Kruskal-Wallis test. For categorical variables, cross-tabulation analysis was conducted to compare distributions between groups. The chi-square test was used to assess statistical significance, and Fisher exact test was applied when the expected frequency for any category was less than 5. Correlations between the recovery ratio and radiologic or clinical factors were assessed using Pearson’s correlation and analysis of variance, as appropriate. Multivariate stepwise regressions were performed to identify independent predictors of neurological recovery. Statistical significance was set at p<0.05. All analyses were conducted using MedCalc ver. 23.2.8 (MedCalc Software Ltd., Belgium).
RESULTS
A total of 349 patients (257 males and 92 females) underwent posterior cervical decompression, with 238 treated by LP and 111 by LF, for multilevel cervical OPLL. There were 226 patients with neutral K-line (+) and 123 with neutral K-line (-). Among the 226 patients classified as neutral K-line (+), 107 (47.35%) were FK-line (-) on flexion radiographs (Supplementary Fig. 1). Ultimately, 119 patients were in the FK-line (+) group, and 230 were in the FK-line (-) group. The mean follow-up period was 43.2±18.7 (range, 25–64) months, and the average age at surgery was 57.32±10.17 (range, 34–82) years. Baseline demographic and clinical characteristics stratified by FK-line status are summarized in Table 1.
1. Clinical and Radiographic Characteristics According to FK-Line Status
Age did not differ significantly between the FK-line (+) and FK-line (-) groups; however, the FK-line (-) group had a significantly higher proportion of male patients (p=0.014) and a higher number of operated levels (p=0.037). FK-line (-) patients exhibited a smaller FK-line distance, larger C2S, lower T1S, and reduced C2–7 CA, indicating a more kyphotic alignment than FK-line (+) patients. The FK-line (-) group demonstrated a significantly higher COR (p=0.009), suggesting more severe spinal cord compression. Additionally, those with FK-line (-) showed greater cervical flexion; however, C2–7 SVA, FK-line tilt angle, ROM, and DER did not differ significantly between groups. Preoperative JOA scores were similar, but postoperative recovery ratios were significantly higher in the FK-line (+) group (p=0.011). The distribution of OPLL subtypes did not differ significantly between the groups (Table 1).
2. Comparison of LP and LF in FK-Line (-) Patients
In the FK-line (-) subgroup, 133 patients underwent LP, and 97 underwent LF. Patients who underwent LF were significantly older (p<0.001), more often male (p=0.007), and had more operative levels (p<0.001) than those in the LP group. The LF group demonstrated significantly greater posterior cord shift, reflected by a larger increase in the FK-line distance (p=0.021), and more effective correction of the C2–7 flexion angle (p= 0.001). These radiologic improvements corresponded with better neurological recovery in the LF group, evidenced by a greater JOA score improvement (p<0.001) and a higher recovery ratio (p=0.023), despite a lower preoperative functional baseline. ROM reduction was more pronounced in the LF group. Operative time and intraoperative blood loss were significantly higher in the LF group than in the LP group. However, the incidence of postoperative complications (C5 palsy, hematoma, dural tear, and revision) did not differ significantly between groups (Table 2). Additionally, among the 123 patients with neutral K-line (-), 74 underwent LP, often due to comorbidities or surgical preference. These patients had a lower mean JOA recovery rate compared to those who underwent LF, as shown in Supplementary Table 1.
3. Multivariate Regression Analysis of Cervical Parameters for Neurologic Improvement
The stepwise multivariate regression analyses identified postoperative FK-line distance, postoperative C2–7 flexion Cobb angle, and preoperative DER as independent predictors of neurological recovery (Table 3). Increased postoperative FK-line distance was positively associated with the JOA recovery ratio (β=3.06, p=0.001), indicating that greater posterior cord shift contributed to better outcomes. Higher DER was positively associated with recovery (β=0.68, p=0.039), and increased postoperative C2–7 kyphosis during flexion was negatively associated with recovery (β=-0.89, p=0.011).
4. Changes in Cervical Parameters After LP and LF in FK-Line (-) Patients
In FK-line (-) patients, LF resulted in a significantly greater increase in FK-line distance than LP (p=0.003), indicating a more pronounced posterior cord shift. Although changes in C2S, C2–7 CA, and T1S tended to favor LF over LP, suggesting a potential advantage of LF in improving sagittal alignment, those differences did not reach statistical significance. Representative cases are illustrated in Figs. 4 and 5.
DISCUSSION
This study investigated the clinical significance of FK-line status and compared the surgical outcomes of LP and LF in patients with multilevel cervical OPLL, focusing on those with FK-line (-). FK-line (-) patients exhibited a smaller FK-line distance, lower C2–7 CA, higher COR, larger C2S, and lower T1S, indicating a more kyphotic alignment and greater cervical flexion angle than FK-line (+) patients. Neurological recovery and sagittal alignment were poorer in FK-line (-) patients following posterior decompression. Among FK-line (-) patients, LF achieved better correction of kyphotic alignment, greater posterior cord shift, and superior neurological outcomes compared with LP. Additionally, multivariate analysis identified postoperative FK-line distance, postoperative C2–7 flexion angle, and preoperative DER as independent predictors of neurological recovery.
Dynamic factors, such as neck flexion and extension, influence the progression of myelopathy in OPLL by altering the sagittal diameter of the spinal canal and increasing anterior cord compression [2,3,12]. Takeuchi et al. [4] reported poor clinical outcomes after LP in OPLL patients with K-line (+) in the neutral position and K-line (-) in the neck flexion position. Nori et al. [6] also observed higher incidences of residual spinal cord compression and poorer outcomes in FK-line (-) patients after LP. Unlike prior reports [4,6], our study directly compared LP and LF in the FK-line (-) subgroup and identified specific radiologic predictors of recovery, reinforcing the clinical relevance of FK-line status in surgical planning.
During flexion, the cervical spinal cord moves more anteriorly and elongates longitudinally, resulting in increased ventral compression and mechanical stress [13]. An anterior shift of the spinal cord during flexion places pressure on the ventral gray matter, including the anterior horn laminae and adjacent corticospinal tracts near the anterior median fissure. This compression may disrupt axonal integrity and compromise microvascular perfusion in the anterior spinal artery territory, resulting in ischemia and motor pathway injury, which likely contributes to poor postoperative neurological recovery [3]. These pathophysiological changes are supported by MRI signal alterations and diffusion tensor imaging evidence of microstructural damage [14]. Specifically, in cervical kyphotic alignment, the ventral OPLL mass exacerbates anterior spinal cord compression, leading to insufficient decompression and affecting neurological recovery after posterior decompression [12,13]. Our findings suggest that FK-line (-) patients exhibited lower T1S, lower C2–7 CA, and a larger C2S than FK-line (+) patients. FK-line (-) patients exhibited increased cervical kyphosis during flexion and diminished lordosis during extension, indicating a flexion-dominant cervical alignment. Postoperative JOA scores and recovery ratios were lower in FK-line (-) patients than FK-line (+) patients, consistent with previous findings [4-6]. Furthermore, FK-line (-) patients exhibited a greater cervical flexion angle, smaller FK-line distance, and higher COR than FK-line (+) patients. Based on those observations, it is reasonable to predict that neurological recovery in FK-line (-) patients might be improved by increasing cervical lordosis, decreasing the cervical flexion angle, and expanding the FK-line distance.
Both LP and LF are effective posterior surgical procedures for multilevel cervical OPLL, but they produce significantly different outcomes in patients with FK-line (-). Our findings indicate that LF leads to superior neurological improvement compared with LP in FK-line (-) patients. Posterior decompression procedures, such as LP and LF, indirectly achieve ventral decompression by allowing posterior cord shift [15,16]. Baba et al. [17] reported that, as measured by the posterior shift score of the cord, significant neurological improvement is associated with posterior cord migration after LP. The extent of spinal cord shift influences spinal cord biomechanics, with LF resulting in lower stress and strain on the spinal cord than LP [18]. Additionally, the greater spinal cord drift achieved with LF compared to LP suggests its superior ability to maximize decompression in severe degenerative cervical myelopathy [19]. However, other studies reported that the extent of the cord shift does not consistently predict clinical outcomes [20,21]. Itoh and Tsuji emphasized that a 4-mm expansion of the spinal canal is optimal for achieving favorable outcomes following LP, and a 3-mm posterior shift of the spinal cord is required to attain a recovery rate exceeding 50% [19,22,23]. In the literature, the posterior shift of the cervical cord after LF averaged around 2.7 mm, compared with 1.7–2.2 mm after LP [19,24]. In this study, the increase in FK-line distance was greater in the LF group (mean, 5.37 mm) than the LP group (3.32 mm), which likely explains the superior neurological recovery observed in the LF group.
One complication following posterior decompression surgery is C5 nerve palsy, attributed to spinal cord shifting and the resultant traction on the spinal cord and/or nerve roots. The incidence has been reported to range from 0% to 13.6% [25,26]. In this study, 5 patients who underwent LP (n=2) and LF (n=3) developed C5 palsy, which resolved within 7 months of rehabilitation. Radcliff et al. [27] reported that patients who developed C5 palsy after LF had significantly greater spinal cord drift at the C5 level (5.1 mm) compared to those without palsy (2.4 mm). Other studies noted that although LF resulted in a greater degree of cord shift than LP (2.65 mm vs. 2.20 mm), the difference in C5 palsy incidence was not statistically significant [19]. Our findings are consistent with prior literature, suggesting that while increased cord shift may contribute to C5 palsy, most cases are transient and recover with rehabilitation.
When the C2–7 kyphotic angle exceeds 20°, biomechanical studies have shown a rapid increase in intraspinal pressure [12]. Local kyphosis exceeding 13° on preoperative alignment has been associated with the highest risk of poor recovery after LP [28]. Funaba et al. reported that the postoperative C2–7 kyphotic angle in neck flexion is associated with poor outcomes [29,30]. Our study demonstrates that the preoperative cervical flexion angle is associated with FK-line status and that the postoperative C2–7 flexion angle correlates significantly with neurological recovery, differentiating outcomes between LP and LF. LF achieved more effective correction of the C2–7 flexion angle, resulting in improved posterior cord shift and sagittal alignment, thereby enhancing neurological recovery. These findings provide radiological and clinical evidence explaining the differences in outcomes between LP and LF in patients with high COR and FK-line (-), a relationship previously reported but not mechanistically elucidated [31]. Additionally, we found that, among patients with neutral K-line (-), those who underwent LF showed greater neurological improvement than those who underwent LP. All patients with neutral K-line (-) maintained FK-line (-) on flexion. These findings suggest that fusion surgery may be more appropriate to achieve decompression and stabilization, especially when kyphosis is present.
This study suggests that LF should be considered over LP in patients with FK-line (-) status, a high COR, and preoperative kyphotic alignment—particularly when posterior decompression without stabilization may be insufficient to achieve ventral cord decompression or correct sagittal imbalance. Patients with a greater cervical flexion angle and limited extension function may be at increased risk of inadequate decompression and postoperative sagittal malalignment if managed with LP [7]. In such cases, LF provides superior correction of flexion kyphosis, greater posterior cord shift, and better restoration of alignment, contributing to improved neurological recovery.
Postoperative loss of cervical lordosis is associated with poor outcomes, possibly as a result of injury to the posterior musculoligamentous complex [32,33]. Recently, DER, which indicates the functional reserve of posterior structures, has attracted attention as a potential risk factor for decreased cervical lordosis [7,34]. Lower DER has been shown to correlate with greater loss of cervical lordosis and poorer outcomes after LP [35]. However, some researchers have argued that DER is not a significant predictor of postoperative loss of lordosis, suggesting instead that parameters such as higher C2–7 SVA and preoperative lordosis are more influential [36]. Our findings support that DER serves as an independent predictor of postoperative neurological outcomes by reflecting cervical extension function.
Assessing the FK-line status provides a practical evaluation of dynamic anterior cord compression and kyphotic alignment. Our findings underscore the importance of both dynamic and structural sagittal parameters in predicting surgical outcomes. In FK-line (-) patients, the greater increase in FK-line distance following LF suggests more effective spinal cord decompression and sagittal correction, potentially contributing to improved neurological recovery, in patients with preoperative kyphosis and high COR. However, LF was associated with longer operative times, increased blood loss, and a more pronounced reduction in neck ROM compared to LP.
This study has several limitations. First, its retrospective design carries inherent risk of selection bias, particularly due to differing surgical indications for LP and LF. Second, surgeon preference might have influenced the surgical methods. However, unlike previous studies [4,6], we directly compared LP and LF in FK-line (-) patients in a large multicenter cohort, providing stronger evidence for surgical decision-making in this challenging population. Third, radiological assessments were limited to standard static and flexion-extension radiographs, without MRI-based confirmation of cord shift. Fourth, the mean followup period was 43.2 months, so longer-term studies are needed to assess the durability of the outcomes. Further prospective studies with long-term follow-up are warranted to validate our findings and to establish standardized criteria for surgical selection based on FK-line status.
CONCLUSION
FK-line status is a clinically significant radiological parameter that reflects dynamic sagittal alignment during neck flexion and serves as a valuable predictor of neurological recovery following posterior decompression in patients with cervical OPLL. FK-line (-) patients exhibited more kyphotic alignment, greater cervical flexion, and poorer neurological recovery than FK-line (+) patients. Among FK-line (-) patients, LF was more effective than LP in promoting posterior cord shift, improving sagittal alignment, and enhancing neurological recovery. Postoperative FK-line distance, C2–7 flexion angle, and preoperative extension reserve independently predicted recovery. These findings highlight the importance of incorporating FK-line status and dynamic sagittal parameters into surgical planning to optimize outcomes in cervical OPLL.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Full text via PMC
Download Citation
Supplement
Print
