Prevalence and Risk Factors for Modic Changes in Symptomatic Cervical Ossification of the Posterior Longitudinal Ligament: A Retrospective Study of 203 Patients

Background

Modic changes (MC) are abnormal bone signals under the vertebral endplates on magnetic resonance imaging (MRI) of the spine, classified by Modic in 1988, suggesting vertebral endplates and bone marrow lesions near the vertebral body [1]. The magnetic resonance imaging T1 sequence and T2 sequence showed different signal changes in the 3 types of MC [2]. The 3 types of MCs are: Type 1, which shows a hypointense signal seen on T1 sequences and a hyperintense signal seen on T2 sequences; Type 2, which shows a hyperintense signal seen on T1 sequences and a hyperintense or isointense signal seen on T2 sequences; and Type 3, which shows a hypointense signal seen on T1 and T2 sequences [3]. Type 1 MC represents vascularized bone marrow and edema; Type 2 MC includes fatty change of the bone marrow; and Type 3 MC, which is rare, represents subchondral sclerosis. Types 1, 2, and 3 MC evolve and have been previously shown to be associated with advancing cervical vertebral degeneration [4]. Peterson et al first found that cervical spine MRI MCs were also present in 2007; 19 cases (16%) had MCs, and cervical spine MCs gradually became a focus of clinical research [5]. Yang et al analyzed 223 cases of single-segment nerve root neck type treated by surgery, and the imaging data of patients with spondylosis showed that 18% of patients had cervical spine MCs [6]. Kong et al found that in patients with cervical degenerative diseases, the incidence of cervical spine MCs was 12.34% [7]. Leboeuf-Yde et al proposed that heavy physical labor increased stress on the intervertebral disc, which eventually led to the occurrence and development of MC [8]. Mok et al reported that disc herniation can cause MC. Symptomatic cervical ossification of the posterior longitudinal ligament (OPLL) is attributed to heterotopic ossification of the cervical posterior longitudinal ligament, which results in spinal cord compression and neurological deterioration. Ossification of the posterior longitudinal ligament of the cervical spine directly compresses the nerve root or spinal cord, resulting in spinal cord involvement [9]. Cervical OPLL consists 4 types of cervical OPLL: continuous, segmental, mixed, and local. The age of onset of OPLL was 50–60 years old, and there was a significant sex difference, with a male-to-female ratio of approximately 2–3: 1 [10]. There have been few studies on whether ossification of the cervical posterior longitudinal ligament is associated with MC. Therefore, this retrospective study of 203 patients with ossification of the posterior longitudinal ligament aimed to evaluate factors associated with MC on cervical spine MRI.

Material and Methods

ETHICS STATEMENT:

This study was approved by the Ethics Committee of the Third Hospital of Hebei Medical University. Since this study was retrospective, the ethics committee agreed to sign a waiver of informed consent, and the ethics committee requires that the entire research process be kept strictly confidential and the collection of personal information such as the patient’s name and address is strictly prohibited.

INCLUSION AND EXCLUSION CRITERIA FOR STUDY PARTICIPANTS:

Patients presenting from July 2014 to April 2022 for symptomatic cervical ossification of the posterior longitudinal ligament, such as myelopathy, were included in our study. All patients were scanned using computed tomography (CT) and were divided into 4 types according to their scan results. The exclusion criteria were as follows: (1) patients with cervical OPLL without spinal symptoms; (2) cervical spine fractures; (3) cervical tumor; (4) previous cervical spine surgeries; (5) ankylosing spondylitis and immune disorders associated with OPLL, and (6) incomplete patient information and blank selection [11–13]. From this database, a total of 203 patients with symptomatic cervical ossification of the posterior longitudinal ligament were collected, and 43 patients were identified as having MC from July 2014 to April 2022 and were assessed for prevalence and risk factors.

CLASSIFICATION OF MC ON MRI:

All patients were examined using a 3.0-imager (Siemens Magnetic Symphony; Siemens, Berlin, Germany) to assess for the presence of MC. The magnetic resonance imaging T1 sequence and T2 sequence showed different signal changes in the 3 types of MC. According to the classification criteria specified by Modic et al, type 1 represents vascularized bone marrow and/or edema, with hypointensity on MRI T1 images and hyperintensity on MRI T2 images; type 2 represents proliferation of fatty tissue, and hyperintensity on MRI T1 image, MRI T2 images are hypointense or isointense; type 3 represents subchondral bone sclerosis, and both T1 and T2 MRI images are hypointense. Figure 1 shows an example of segmental OPLL with type 2 MC at the C5–6 levels on cervical MRI. The classification of MC on MRI was performed by 2 independent reviewers who obtained images without screening patient information, and disagreements between the 2 reviewers were finally decided by a third reviewer [14–17].

DATA COLLECTION PROCESS AND COLLECTION METHODS:

The demographic and clinical characteristics of patients with symptomatic cervical ossification of the posterior longitudinal ligament were divided into: age (years), sex (female and male), weight (<65 and ≥65 kg), body mass index (BMI; <25 and ≥25 kg/m2), neck pain (without neck pain and with neck pain), physical workload category (no heavy and heavy), smoking history (yes or no), drinking history (yes or no), combined diabetes (yes or no), combined hypertension (yes or no), and combined coronary heart disease (yes or no). The patient’s occupation was divided into sedentary, light physical labor, moderate physical labor, and heavy and very heavy physical labor. Patients with the first 3 workloads were defined as “non-heavy,” and patients with the last 2 workloads were defined as “heavy” [18].

The laboratory tests of patients with symptomatic cervical ossification of the posterior longitudinal ligament were divided into: white blood cell (WBC), neutrophil percentage (NEUT%), high-density lipoprotein (HDL), low-density lipoprotein (LDL), and triglyceride (TRIG); C2–7 Cobb angle (<20° and ≥20°); cervical alignment and range of motion/ROM (<40° and ≥40°); type of cervical ossification of the posterior longitudinal ligament (mixed, local, segmental, and continuous); k line [(+) and (−)]; and the relationship between MC and OPLL on sagittal CT images (same segment, adjacent segment, and different segment). The cervical spine was measured using SL radiography. C2–7 Cobb angle was defined as the angle formed by a straight line parallel to the lower endplate of the C2 vertebral body and a straight line parallel to the lower endplate of the C7 vertebral body. C2–7 range of motion (ROM) was determined by measuring the Cobb angle of cervical hyperextension and hyperflexion, in which hyperextension is α, hyperflexion is β (negative value when cervical kyphosis), and ROM=α+β. On the CT sagittal plane, OPLL was divided into 4 types: continuous, segmental, mixed and local. The K line is a virtual line connecting the midpoint of the anterior and posterior diameter of the spinal canal from C2 to C7. Ossification of OPLL beyond the K line is called the K line (−) or the K line (+). On the CT sagittal plane, MC and OPLL location were divided into the same segment, adjacent segment, and different segment [19,20]. We converted some continuous variables into binary variables (body weight ≥65 kg, body weight <65 kg; BMI ≥25 kg/m2, <25 kg/m2; C2–7 Cobb angle ≥20°, <20°; C2–7 ROM ≥40°, <40°) based on the research of Bai et al [21].

STATISTICAL ANALYSIS:

Statistical analyses were performed using SPSS version 25.0 (IBM, Armonk, NY, USA). Differences in the prevalence of MC were analyzed using Pearson’s chi-squared test or rank-sum test. The risk factors for MC were first determined using univariate logistic regression (P<0.05). Multivariate analysis was performed for the statistically significant risk factors (P<0.05). Further receiver operating characteristic (ROC) analysis displayed an area under the curve (AUC), and Youden index was used to determine the cutoff value of age for MC. Statistical significance was defined as a 2-tailed P<0.05. All charts were analyzed using GraphPad Prism 9 (GraphPad, Inc., USA).

Results

DEMOGRAPHIC CHARACTERISTICS OF MC:

In total, 203 patients with symptomatic cervical OPLL were included in our study. Among them, 43 (21.18%) patients had MC and 160 (78.82%) had no MC. The prevalence of MC in males and females was 18.88% and 15.38%, respectively, and there was no statistically significant difference between the groups with and without MC in terms of sex (χ2=0.063a, P=0.802). The prevalence of MC in patients weighing <65 and ≥65 kg was 9.52% and 22.53%, respectively, and there was no significant difference in weight between patients with and without MC (χ2=1.907a, P=0.167). The prevalence of MC <25 kg/m2 and ≥25 kg/m2 was 27.78% and 19.76%, respectively, and there was no significant difference in the BMI between patients with and without MC (χ2=1.140a, P=0.286). There was no significant difference in the prevalence of WBC, NEUT%, HDL, LDL, or TRIG between the 2 groups with or without MC. This study showed that there were significant differences in the distribution of MC patients according to neck pain (χ2=5.005a, P=0.025). The laboratory test results and other clinical data are presented in Table 1.

IMAGING FEATURES OF THE MC:

The number of vertebral segments with MC was 50, and the numbers of type 1, type 2, and type 3 MC were 3 (6.00%), 45 (90.00%), and 2 (4.00%), respectively. The most common site of MC was C5–6 (40.00%), followed by C2–3 (20.00%), C6–7 (14.00%), C3–4 (14.00%), C4–5 (12.00%), and C1–2 (0.00%) (Figure 2). There were 136 vertebral segments with OPLL, and the number of mixed, local, segmental, and continuous were 54 (39.71%), 5 (3.68%), 64 (47.06%) and 13 (9.56%), respectively. The OPLL was most frequent at C5–6 (25.00%), followed by C4–5 (23.53%), C3–4 (20.59%), C6–7 (15.44%), C2–3 (11.03%), and C1–2 (4.41%) (Figure 3). The relationship between the vertebral body segment with MC and the vertebral body segment with OPLL was defined as same segment, adjacent segment, and non-adjacent segment, and the number of same segment, adjacent segment, and non-adjacent segment were 34 (79.07%), 6 (13.95%), and 3 (6.98%), respectively. Local type and MC share the same segment (100.00%), followed by segmental type (77.27%), mixed type (75.00%), and continuous type (75.00%) (Figure 4). OPLL and MC share the same segment in type 3 MC, with a ratio of 100.00%, followed by type 2 MC (79.49%), and type 1 MC (66.67%) (Figure 5).

RISK FACTORS FOR DEVELOPING THE MC:

According to univariate logistic analysis, multiple characteristics were significantly related to MC. These factors included age (OR=1.05, 95% CI: 1.01–1.09, P=0.011), and neck pain (OR=2.76, 95% CI: 1.10–6.91, P=0.030). Multivariate logistic analysis showed that OPLL type was significantly associated with MC. These items were: age (OR=1.05, 95% CI: 1.01–1.09, P=0.013), and neck pain (OR=2.67, 95% CI: 1.04–6.83, P=0.041) (Table 2).

We selected MCs as the state variable and age as the test variable. The area under ROC curve between MC and age was shown in Figure 6 (AUC=0.657, P=0.002). Based on the ROC curve, the age corresponding to the optimal Youden index (0.349) was 60.50 (sensitivity: 0.67, specificity: 0.68). It could be assumed that 60.5 years was the critical age for MC, with a higher probability of MC for ≥60.5 years and a lower probability of MC for <60.5 years.

Discussion

OPLL was first reported by Key in 1838. It is generally believed that OPLL is mainly found in East Asians, especially the Japanese, but now it has been found in Europe and the United States [22,23]. European and American scholars believe that OPLL is a subtype of diffuse idiopathic skeletal hyperosteosis (DISH) [24–26]; but there have been few studies on cervical posterior longitudinal ligament ossification and MC.

MC is a common imaging manifestation on MRI, suggesting lesions of the vertebral endplate and bone marrow. In our study, we found that the most common type of MC in patients with cervical posterior longitudinal ligament ossification was type 2 MC. Peterson et al reported that type 1 MC was the most common type in patients with symptomatic cervical spondylosis [5], while Mann et al and Hayashi et al found that type 2 MC was the most common [27,28]. The conclusions of this study are similar to those of Hayashi et al. Li et al found that patients in the total disc replacement group had more significant improvement in neck pain, perhaps because more inflammatory disc tissue was removed, which is related to the effect of inflammation on cervical MCs [29]. This study found that inflammatory factors were not the influencing factors of MC in patients with cervical posterior longitudinal ligament ossification. Kong et al found that in patients with cervical spine degeneration, cervical MCs were associated with chronic neck pain, and the neck pain score of patients with type 1 MCs was significantly higher than that of patients with type 2 MCs [30]. This study found that neck pain is an influencing factor of MC in patients with cervical posterior longitudinal ligament ossification, and it was not possible to compare the differences in neck pain between the different MCs types due to the small number of type 1 MCs and type 3 MCs medical records. An et al found that for patients with cervical kyphosis associated with MCs, the range of motion between single cervical spine segments was significantly reduced [31]. Hayashi et al found that MC can lead to cervical stiffness, loss of mobility, and severe instability [32], and that cervical ROM was not an influencing factor of MC in patients with ossification of the posterior longitudinal ligament. Li et al found that the onset of MC in either the cervical or lumbar spine was age-related, and that MC increased with age, and also found that older age (≥60.5 years) was significantly associated with the occurrence of MC in cervical OPLL [33].

Our study had several limitations. First, it did not assess whether the severity and duration of neurological symptoms in the selected population affected MC. Second, it did not determine whether the selected population had cervical disc herniation and whether it affected MC. Finally, further research is needed to validate our results.

Conclusions

We found that older age and neck pain were significantly associated with development of cervical spine MC. Clinicians should pay attention to the possible occurrence of MC in patients aged ≥60.5 years with symptomatic cervical OPLL.