Direct anterior decompression in patients with ossification of the posterior longitudinal ligament significantly relieves short-segment spinal cord high signal

Abstract

Background

In patients with ossification of the posterior longitudinal ligament of the cervical spine (OPLL), high spinal cord signal (HCS) is frequently observed in the spinal cord of the corresponding segment. However, studies on the differences in the improvement of high spinal cord signal due to different surgical approaches are limited. The aim of this study was to investigate the improvement of high spinal cord signal in long and short segments with different choices of surgical approaches.

Methods

In this study, we conducted a meticulous review of medical records for patients diagnosed with ossification of the posterior longitudinal ligament (OPLL). Demographic variables, including gender, age, and body mass index (BMI), were systematically recorded. We evaluated the severity of neurological impairment using the Japanese Orthopaedic Association (JOA) scores both preoperatively and at multiple postoperative follow-up points. Neurological assessments were complemented by serial magnetic resonance imaging (MRI) T2-weighted imaging (T2WI) to measure the extent of high-signal changes (HCS) in the spinal cord, and the alteration of the HCS was quantified by the SCR (the ratio between the signal intensity value of the HCS region and the signal intensity value of the normal spinal cord region at C7-T1).

Results

In the short-segment high signal change (HCS) group, comparisons of JOA score improvement (Recovery1) and HCS improvement (CR1) at 6 months postoperatively did not demonstrate significant differences between the surgical approaches (P > 0.05; Table 1). However, at the 2-year follow-up, patients who underwent anterior surgery exhibited significantly greater improvements in both JOA scores (Recovery2) and HCS (CR2), with statistical significance achieved (P < 0.05; Table 1). In contrast, in the long-segment HCS group, there was no significant difference between the anterior and posterior surgical approaches in terms of JOA improvement and HCS improvement at 6 months and 2 years postoperatively (P > 0.05; Table 2).

Conclusions

In patients with OPLL who present with spinal cord high signal, anterior surgery by resection of the ossified posterior longitudinal ligament and direct decompression is more conducive to regression of small spinal cord high signal and improvement of clinical neurological function if the extent of spinal cord high signal is small.

Introduction

Ossification of the posterior longitudinal ligament (OPLL) represents a prevalent and clinically significant pathology within spinal surgery, primarily implicated as a principal etiological factor in cervical spinal cord compression and subsequent paralysis [1, 2]. The pathogenesis of OPLL is characterized by heterotopic ossification of the posterior longitudinal ligament, wherein progressive accretion of ossified material incrementally diminishes the spinal canal's capacity, exerting direct compression on the spinal cord. This compression cascade culminates in extensive neurological deficits, encompassing sensory and motor impairments, and sphincter dysfunction, which may progress to quadriplegia as well as urinary and fecal incontinence—severely compromising patient quality of life and survival. Initially identified in Japan during the 1960s, OPLL exhibits a pronounced prevalence in East Asian populations, recorded at 4.3% [3, 4], and displays marked ethnospecific disparities. Contemporary enhancements in diagnostic imaging modalities have facilitated the increased detection of OPLL, with recent epidemiological studies indicating a prevalence rate between 20 to 34% among Asians over 65 years of age [5]. The clinical management of OPLL poses substantial challenges due to its insidious onset, extensive ossification, severe spinal cord compression, and limited spinal canal reserve, thus making surgical intervention particularly daunting.

Magnetic resonance imaging (MRI) is pivotal in the non-invasive evaluation of ossification of the posterior longitudinal ligament (OPLL), offering high-resolution insights into soft tissue structures and the extent of spinal cord compression and stenosis. T2-weighted imaging (T2WI) sequences are particularly instrumental, enabling detailed assessment of disc degeneration and spinal cord compression. In OPLL, the detection of localized high signal intensity on T2WI can indicate spinal cord injury, characterized by nonspecific pathologic changes such as edema, inflammation, vascular ischemia, glial cell proliferation, and myelin sheath disturbances [6, 7]. These pathologies typically manifest corresponding neurological symptoms and are more distinctly observed on T2WI compared to T1-weighted imaging (T1WI) [8].

Furthermore, the intensity of high signal changes (HCS) on preoperative T2WI has been found to inversely correlate with neurological prognosis post-surgery, as evidenced by multiple studies. A marked regression of HCS postoperatively is often associated with significant recovery from neurological dysfunction [9–12]. Variability in the extent of HCS among OPLL patients ranges from localized points to extensive areas, occasionally spanning beyond the length of one vertebra. This heterogeneity presents challenges in surgical decision-making, as the optimal surgical approach may vary significantly based on the extent of the lesion. Current research remains divided on the most effective surgical strategies tailored to these differing extents of spinal cord high signal lesions.

In the surgical management of ossification of the posterior longitudinal ligament (OPLL), both anterior and posterior approaches are employed, each demonstrating favorable clinical outcomes. However, the optimal surgical strategy—whether to prefer an anterior or posterior approach—continues to be a subject of robust debate within the medical community. Anterior surgery is generally advocated for cases characterized by significant vertebral canal occupancy or concurrent kyphotic deformity, as it directly removes the compressive mass and potentially corrects the kyphotic alignment. Conversely, posterior surgery is often favored for extensive long-segment OPLL, offering broader segmental decompression capabilities [13–18].

The evaluation of postoperative neurological recovery and long-term prognosis has yielded mixed results across studies, with many utilizing functional metrics such as the Japanese Orthopaedic Association scores (JOAs), alongside radiological assessments including canal narrowing ratio (CNR), cervical lordosis angle (Cobb), and sagittal vertical axis (SVA). Despite extensive research, a significant gap remains in understanding the correlation between surgical approach and the amelioration of intramedullary high signal conditions (HCS) post-surgery. This study aims to explore this relationship further, assessing the impact of both anterior and posterior surgical interventions on different extents of HCS in OPLL patients. The findings could provide crucial insights into refining surgical strategies and enhancing patient outcomes in this complex clinical arena.

Method

In this study, we conducted a meticulous review of medical records for patients diagnosed with ossification of the posterior longitudinal ligament (OPLL). Demographic variables, including gender, age, and body mass index (BMI), were systematically recorded. We evaluated the severity of neurological impairment using the Japanese Orthopaedic Association (JOA) scores both preoperatively and at multiple postoperative follow-up points. Neurological assessments were complemented by serial magnetic resonance imaging (MRI) T2-weighted imaging (T2WI) to measure the extent of high-signal changes (HCS) in the spinal cord, and the alteration of the HCS was quantified by the SCR (Fig. 1). In our study, MRI evaluations were performed using a [3.0 Tesla Siemens Magnetom Skyra scanner (Siemens Healthineers, Erlangen, Germany)].

Fig. 1.

Schematic representation of the SCR calculation method. The SCR is calculated as the ratio of the signal intensity value in the high signal region of the spinal cord to the signal intensity value in the normal spinal cord region of C7-T1 (e.g. signal intensity value A/ signal intensity value B = 2.0799 in the schematic)

Patients were stratified into two groups based on the preoperative length of the HCS: a long-segment group (HCS length greater than 10 mm) and a short-segment group (HCS length 10 mm or less). This categorization allowed for a nuanced analysis of the relationship between the extent of HCS and the outcomes following different surgical approaches. The study aims to elucidate the correlation between surgical strategy—whether anterior or posterior—and the postoperative improvement in HCS, providing valuable insights into tailored therapeutic interventions for OPLL.

Patient enrollment and case collection

Source of patients

In this retrospective study, we analyzed patients diagnosed with ossification of the posterior longitudinal ligament (OPLL) who underwent surgical interventions at the Cervical Spine Surgery Department of Shanghai Changzheng Hospital between January 1, 2014, and January 1, 2022. Each patient underwent both preoperative and postoperative MRIs, and were monitored over a follow-up period of approximately two years. Of the initial 317 reviewed cases, 90 met the stringent enrollment criteria outlined below (Fig. 2).

Fig. 2.

Schematic diagram of the measurement of CNR, C2–7 Cobb angle, SVA and mK-line INT

Inclusion criteria

Diagnosis: Patients must have a clinical and imaging-confirmed diagnosis of cervical ossification of the posterior longitudinal ligament (OPLL).

Imaging Confirmation: Presence of OPLL must be visible on cervical computed tomography (CT) scans or X-rays.

Extent of OPLL: The OPLL must extend over a length greater than that of one vertebral body in the sagittal plane of the cervical spine.

MRI Evaluations: Participants must have undergone magnetic resonance imaging (MRI) at our institution preoperatively, as well as at six months and two years postoperatively (Figs. 3 and 4).

Fig. 3.

Schematic representation of changes in short-segment HCS after anterior surgery. From left to right, images are shown preoperatively, 6 months postoperatively, and 2 years postoperatively. Anterior surgery group: preoperative SCR = 2.2792; 6 months postoperative SCR = 1.7826; 2 years postoperative SCR = 1.4971

Fig. 4.

Schematic representation of changes in long-segment HCS after posterior surgery. From left to right, images are shown preoperatively, 6 months postoperatively, and 2 years postoperatively. Posterior surgery group: preoperative SCR = 2.5962; 6 months postoperative SCR = 2.3946; 2 years postoperative SCR = 1.9456

Signal Change Ratio (SCR): There must be an increased signal intensity within the spinal cord at the area of OPLL compression on preoperative MRI T2-weighted imaging (T2WI), with an SCR of ≥ 1.20 (Fig. 5).

Fig. 5.

Schematic illustration of the presence of high spinal cord signal (HCS) in short and long segments in patients with OPLL

Follow-up Period: A minimum postoperative follow-up period of 24 months is required for inclusion.

Exclusion criteria

• Patients were excluded from the study if they had:

• Undergone previous cervical spine surgery;

• Absence of high signal change (HCS) on preoperative MRI T2WIs;

• Spinal pathology due to trauma, infection, or neoplastic processes.

Subsequent to the application of these criteria, patients were stratified into two cohorts based on the length of the spinal cord high signal (HCS) as observed on MRI: short-segment HCS (≤ 10 mm) and long-segment HCS (> 10 mm). These groups were further subdivided according to the surgical approach employed: anterior surgery group S1 (22 patients), posterior surgery group S2 (23 patients), anterior surgery group L1 (20 patients), and posterior surgery group L2 (25 patients). The anterior surgery group included anterior cervical discectomy and fusion (ACDF), anterior cervical discectomy and fusion (ACCF), and hybrid decompression fusion (HDF), and the posterior surgery group included laminoplasty (LP) and laminectomy and fusion (LF). We selected the surgical procedures based on several criteria, including patient anatomical characteristics, extent of spinal cord compression, and surgeon experience. Posterior surgery was chosen for patients with multilevel OPLL and preserved cervical alignment, whereas anterior surgery was used for patients with localized OPLL and significant kyphotic deformity. The ratio of posterior fusion within the posterior surgery group was 16.7% for laminectomy and fusion and 83.3% for laminoplasty. All implant materials are titanium or PEEK. All surgical procedures were performed by a single, experienced senior spine surgeon.

Measurement of imaging data

Core independent variables: Selection of the anterior OR posterior surgical routes

• Ending Variables: Improvement in HCS at 6 months and 2 years postoperatively (CR1 and CR2).

Covariate

• Canal narrowing ratio (CNR): An axial image indicating CNR calculation method, CNR = D2/D1 (Fig. 6).

• C2–7 Cobb angle: The C2–7 Cobb angle was defined as the angle between lines drawn tangential to inferior endplates of C2 and C7 (Fig. 6).

• Sagittal vertical axis (SVA): The C2–7 sagittal vertical axis (SVA) was defined as the horizontal distance between the vertical line from the center of the C2 vertebral body and the posterosuperior corner of the C7 vertebral body (Fig. 6).

• Modified K-line interval (mK-line INT): mK-line INT was defined as the minimum interval between the tip of the anterior compression factor of the spinal cord and the line connecting the midpoints of the cord at the level of the inferior endplates of C2 and C7 (Fig. 6).

Fig. 6.

Schematic diagram of the measurement of CNR, C2–7 Cobb angle, SVA and mK-line

Statistical analysis

Statistical analyses were conducted using SPSS software, version 26.0. A threshold of P < 0.05 was established to denote statistical significance between groups.

To evaluate the interobserver and intraobserver reliability, a subset of 10 patients was randomly selected from the study cohort. Imaging parameters were initially measured and subsequently remeasured one week later by both a radiologist and a well-trained spine surgeon with three years of specialized experience in musculoskeletal imaging. The intraclass correlation coefficient (ICC) was calculated to assess the repeatability of these imaging-based measurements.

All continuous variables were tested for normality using appropriate statistical tests, and data were confirmed to be normally distributed. Differences between the short-segment high signal change (HCS) and long-segment HCS groups were analyzed using the independent samples t-test for normally distributed continuous variables. For dichotomous variables, the chi-square test was employed, with significance set at P < 0.05.

Result

Excellent interobserver and intraobserver reliability can be observed in this study. In which, the ICC value of all imaging based parameters were > 0.8 in both interobserver and intraobserver reliability judgement.

Statistical analysis revealed no significant differences in the baseline characteristics across the study groups, including preoperative signal change ratio (SCR), canal stenosis ratio (CNR), C2-7 Cobb angle, sagittal vertical axis (SVA), modified K-line intervals (mK-line INT), and Japanese Orthopaedic Association (JOA) scores (P > 0.05; Tables 1 and 2).

Table 1.

Significant difference in short-segment group between patients with anterior and posterior approach cervical operations

	Anterior approach	Posterior approach	P-Values
CR1	0.131 ± 0.118	0.027 ± 0.236	0.07
CR2	0.24 ± 0.132	0.017 ± 0.222	0.000**
Recovery1	0.16 ± 0.126	0.125 ± 0.146	0.397
Recovery2	0.531 ± 0.18	0.279 ± 0.31	0.002**
Preoperative SCR	1.67 ± 0.369	1.72 ± 0.348	0.63
Preoperative JOAs	10.68 ± 1.59	10.78 ± 1.62	0.834
C2–7 Cobb angle	9.48 ± 10.8	14.24 ± 12.9	0.188
SVA	16.05 ± 7.8	17.94 ± 8.89	0.452
mK-line INT	3.7 ± 3.28	4.54 ± 2.22	0.319
Age	58.59 ± 5.64	61.17 ± 9.35	0.271
BMI	25.58 ± 4.72	26.95 ± 4.58	0.331
Sex (Male/Female)	10/12	7/16	0.299
Hypertension	12/10	14/9	0.668
Diabetes	16/6	19/4	0.425
Smoking history	19/3	16/7	0.175
Symptom duration	6.64 ± 4.03	7.96 ± 7.96	0.28

^* P < 0.05

^**P < 0.01

Table 2.

Significant difference in long-segment group between patients with anterior and posterior approach cervical operations

	Anterior approach	Posterior approach	P-Values
CR1	0.139 ± 0.117	0.04 ± 0.241	0.099
CR2	0.119 ± 0.147	0.066 ± 0.184	0.301
Recovery1	0.159 ± 0.132	0.133 ± 0.142	0.525
Recovery2	0.427 ± 0.157	0.355 ± 0.235	0.251
Preoperative SCR	1.693 ± 0.36	1.786 ± 0.35	0.389
Preoperative JOAs	10.7 ± 1.59	10.56 ± 1.76	0.783
C2–7 Cobb angle	10.02 ± 11.17	15.6 ± 12.91	0.133
SVA	15.99 ± 7.01	18.43 ± 8.65	0.313
mK-line INT	4.065 ± 3.535	5.08 ± 2.238	0.247
Age	59.3 ± 5.46	60.84 ± 10.21	0.547
BMI	24.93 ± 4.938	27.328 ± 4.457	0.094
Sex (Male/Female)	10/10	8/17	0.221
Hypertension	11/9	16/9	0.54
Diabetes	14/6	20/5	0.438
Smoking history	18/2	18/7	0.134
Symptom duration	8.3 ± 5.51	6.72 ± 4.54	0.297

^*P < 0.05

^**P < 0.01

In the short-segment high signal change (HCS) group, comparisons of JOA score improvement (Recovery1) and HCS improvement (CR1) at 6 months postoperatively did not demonstrate significant differences between the surgical approaches (P > 0.05; Table 1). However, at the 2-year follow-up, patients who underwent anterior surgery exhibited significantly greater improvements in both JOA scores (Recovery2) and HCS (CR2), with statistical significance achieved (P < 0.05; Table 1). In contrast, in the long-segment HCS group, there was no significant difference between the anterior and posterior surgical approaches in terms of JOA improvement and HCS improvement at 6 months and 2 years postoperatively (P > 0.05; Table 2).

Discussion

The progression of ossification of the posterior longitudinal ligament (OPLL) frequently results in spinal cord compression, eliciting a spectrum of clinical symptoms. Cervical spine surgery, encompassing both anterior and posterior approaches, is pivotal in treating cervical spondylotic myelopathy induced by OPLL. Various interbody fusion techniques such as anterior cervical discectomy and fusion (ACDF), anterior cervical corpectomy and fusion (ACCF), and hybrid decompression fusion (HDF) are employed for anterior surgeries, while laminoplasty (LP) and laminectomy with fusion (LF) are preferred for posterior interventions. The selection of surgical approach is influenced by several factors including the use of K-wires, the canal stenosis rate (CNR), and the extent of cervical deformity, yet the optimal surgical strategy for OPLL remains a subject of ongoing debate.

Magnetic resonance imaging (MRI) is integral to diagnosing cervical spondylotic myelopathy linked to OPLL. It effectively delineates spinal cord compression and identifies signal changes within the spinal cord. Notably, spinal cord compression manifests as high signal changes (HCS) on T2-weighted MRI (T2WI). The severity of these changes is quantifiable through the signal change ratio (SCR), comparing the intensity of localized high signals at areas of compression to normal signals at the C7-T1 level. Extensive research has underscored that the severity of preoperative intramedullary HCS is a critical prognostic marker for postoperative neurological function. Elevated HCS signal intensities often suggest a greater likelihood of irreversible neuronal loss. Conversely, mild HCS is associated with less severe neuropathy and a higher potential for recovery [8–11, 19, 20]. Postoperative regression of HCS is indicative of spinal cord recuperation and is predictive of symptomatic improvement. The SCR provides a quantitative analysis of changes in HCS post-surgery, serving as an objective imaging metric to assess symptomatic progress during the postoperative period. Thus, postoperative changes in HCS are instrumental in forecasting neurological recovery following cervical spine surgery.

Despite the extensive body of literature examining the qualitative aspects of preoperative high signal changes (HCS) in the spinal cord and their relationship with postoperative neurological outcomes, there remains a notable deficiency in quantitative analyses that delineate the postoperative progression of HCS of varying magnitudes [6, 9, 10, 21, 22]. This study aims to bridge this gap by conducting a rigorous quantitative analysis of postoperative changes in HCS, utilizing objective imaging metrics. We specifically assess the modifications in HCS across different sizes in patients with ossification of the posterior longitudinal ligament (OPLL) who have undergone various surgical interventions. The ultimate objective is to furnish novel insights that will aid in refining the selection of the most efficacious surgical strategies tailored to individual patient needs based on precise, quantifiable changes in spinal cord imaging.

Our analysis revealed no statistically significant differences in high signal change (HCS) recovery between anterior and posterior surgical approaches at both 6 months and 2 years postoperatively for patients in the long-segment HCS group (P > 0.05, Table 2). However, in the short-segment HCS group, while no significant differences were noted at 6 months, a significant improvement in HCS (CR2) and Japanese Orthopaedic Association (JOA) scores (Recovery2) was observed at 2 years in patients undergoing anterior surgery (P < 0.05, Table 1). This outcome suggests that anterior surgery, which entails direct decompression of the ossified material, may provide superior long-term neurological recovery in patients with localized segments of spinal cord high signal due to OPLL (Fig. 3).

Several previous studies have shown that the superior efficacy of anterior surgery in improving cervical OPLL patients can be attributed to its ability to directly decompress the spinal cord, restore local spinal cord perfusion, and enhance cervical stability and alignment. By directly resecting the compressive ossified mass, anterior surgery effectively alleviates focal mechanical stress, reducing ischemic damage and facilitating faster regression of HCS. Furthermore, the anterior approach allows for robust stabilization with interbody fusion, preventing secondary injury from micromotion and ensuring long-term structural integrity. Additionally, the capacity for correcting local kyphosis and restoring cervical lordosis minimizes residual cord contact, creating a more favorable biomechanical environment for neural recovery [6, 8, 9, 13, 19, 20, 22]. These combined factors contribute to the superior radiological and clinical outcomes seen in short-segment HCS patients treated with anterior approaches.

This study underscores the importance of tailored surgical strategies based on the extent and location of HCS in patients with OPLL. In instances of OPLL with localized HCS, despite extensive ossification, the critical compression typically arises from a confined segment of the ossified material. Anterior surgery, by directly removing this focal ossification, potentially mitigates the risk of spinal cord injury during cervical spine movements, given that posterior surgery leaves residual ossification at the vertebral body’s posterior margin.

Conversely, in patients presenting with extensive HCS, attributed to widespread ossification, hypertrophied ligaments, and hyperplastic bone formations, our findings indicate no significant postoperative difference in HCS improvement between the two surgical techniques. This may be due to the complex interplay of extensive pathological changes that do not favor one surgical approach over another for significant HCS mitigation. These insights are crucial for optimizing surgical interventions in OPLL, enhancing patient outcomes by aligning surgical strategies with specific pathological and anatomical considerations.

This study, while providing valuable insights, is subject to several limitations. As a retrospective analysis with a relatively small cohort, the findings are potentially susceptible to biases associated with single-operator involvement, which might influence case selection, data measurement, and surgical level determination. Additionally, the study did not detect significant differences in complication rates between anterior and posterior cervical spine surgeries; however, this observation may be limited by the small sample size of reported complications. The follow-up period, predominantly short-term, restricts our ability to assess the long-term evolution of high signal changes (HCS) in the spinal cord, which would be more informatively evaluated over a 5- to 10-year timeframe. Moreover, the current study lacks detailed analysis regarding the lateral extent and precise anatomical localization of HCS-induced pathological alterations within specific spinal cord regions, such as the central gray column or lateral and posterior white columns.

Future studies would benefit from employing higher resolution magnetic resonance imaging (MRI) technologies to delineate more precisely the pathological impacts of HCS, particularly in areas like the anterior horn or posterior roots of the spinal cord. Such advancements would enhance our understanding of the nuanced relationships between surgical interventions and long-term neurological outcomes in patients with ossification of the posterior longitudinal ligament (OPLL).

Conclusion

In patients with OPLL who present with spinal cord high signal, anterior surgery by resection of the ossified posterior longitudinal ligament and direct decompression is more conducive to regression of small spinal cord high signal and improvement of clinical neurological function if the extent of spinal cord high signal is small.

Abbreviations

BMI

Body mass index

HCS

High cord signal

OPLL

Ossification of the posterior longitudinal ligament

MRI

Magnetic resonance imaging

SCR

The signal change ratio between the localized high signal and normal spinal cord signal at the C7-T1 levels.

CR1

The regression of high cord signals at 6 months postoperatively (i.e., CR1= (Preoperative SCR - SCR at 6 months postoperatively)/ Preoperative SCR)

CR2

The regression of high cord signal at 2 years postoperatively (i.e., CR2= (Preoperative SCR - SCR at 2 years postoperatively)/ Preoperative SCR)

CNR

Canal narrowing ratio

SVA

Sagittal vertical axis

mK-line INT

Modified K-line interval

JOAs

Japanese Orthopedic Association score

ACDF

Anterior cervical discectomy and fusion

ACCF

Anterior cervical discectomy and fusion

HDF

Hybrid decompression and fusion

LP

Laminoplasty

LF

Laminectomy and fusion

Recovery1

Degree of JOAs recovery at 6 months postoperatively (i.e., Recover1= (JOAs at 6 months postoperatively - Preoperative JOAs)/ (17- Preoperative JOAs))

Recovery2

Degree of JOAs recovery at 2 years postoperatively (i.e., Recover2= (JOAs at 2 years postoperatively - Preoperative JOAs)/ (17- Preoperative JOAs))

Authors’ contributions

Conception and design: Yang Liu and Zichuan Wu; Acquisition of data: Hanlin Song, Zichuan Wu, Aochen Xu, Min Qi, and Chen Xu; Data analysis and interpretation: Baifeng Sun, and Min Qi; Statistical analysis: Hanlin Song, Zichuan Wu, and Xuhong Zhang; Manuscript Preparation: Zichuan Wu, Xuhong Zhang; Manuscript revision and modification: Min Qi and Yang Liu.

Funding

This work is funded by National Natural Science Foundation of China (82172470, 81972090).

Data availability

All the data of the manuscript are presented in the paper.

Declarations

Ethics approval and consent to participate

Approval for the current study protocol was obtained from the ethics committees of Shanghai Changzheng Hospital (2021SL004). The norms on which the study is based are in accordance with the "Declaration of Helsinki".

Consent for publication

Not Applicable.

Competing interests

The authors declare no competing interests.