Background

Schmorl´s nodes (SN) are herniation of the disc material through a defect in the bony endpalte into the vertebral body described by Schmorl, a German pathologist [1]. their dirct cause is unknown, but it was assumed to be due to weakness of the intervertebral disc [2], degeneration of the spine due to exessive loads [3,4,5], or due to autoimmune reaction [6]. More recently, some researchers found a significant correlation between the morphology of the verterbral body and the presence of SN [7].

These nodes present, most commonly, as incidental findings in patients who did lumbar MRI fir various complaints [8], however, they can be the cause of chronic lower back pain [9].

There is a wide range (3.8–76%) in the reported prevalence of the Schmorl’s nodes in the general population [10,11,12].

The Schmorl node may be well detected by plain radiography, computed tomography (CT), and bone scintigraphy; however, MRI is the modality of choice for the diagnosis of Schmorl nodes, as they are best seen on the sagittal MRI sequences [13]. The nodes usually exhibit the same signal characteristics as the adjacent disc, with a thin rim of sclerosis at the margins [14].

Previous studies have reported a positive correlation between SNs and lumbar disc degeneration disease (DDD) and the severity of DDD [15, 16].

DDD is best evaluated by MRI, and this imaging modality is generally considered the most sensitive technique for examining disc degeneration. DDD was staged as the following according to Schneiderman grading system [12, 17, 18].

Stage 0: There is no signal change in the disc.

Stage1: There is a mild decrease in signal intensity of the disc on T2 sequences.

Stage 2: The disc is markedly hypointense with preserved disc height.

Stage 3: Decreased disc signal intensity and disc height.

Associated MRI features of DDD include disc bulging, herniation, and Modic change (MC) [19].

Disc herniation is seen as the focal of disc material beyond the edges of the apophyseal ring [20]. MC is seen as altered signal intensity at the vertebral endplate on both T1W and T2W images. Three types of Modic change were identified. Type I change show decreased signal intensity on T1-weighted images and increased signal intensity on T2-weighted due to fibro vascular replacement. Type II changes correlate with fatty marrow replacement and show increased signal intensity on T1-weighted images and hypointense signal on T2-weighted images. Type III changes correlate with dense bone (sclerosis) that are represented by decreased signal intensity on both T1- and T2-weighted images [21,22,23]. The aim of this study is to reveal the prevalence of SN in adults with lower back pain and to evaluate its relationships with lumber DDD.

Methods

Subjects

Three hundred-twenty four (324) adult patients referred by physicians to MRI unit to do lumbosacral MRI were included in this cross-sectional analytic study.

Inclusion criteria

Adult (more than 18 years old) patients with lower back pain were examined at the MRI department at the Azadi Teaching Hospital, Kirkuk city, Iraq, over the period from September 2016 to February 2018.

Exclusion criteria

Those who had a history of spinal surgery or back trauma, spinal infection, and malignant diseases were excluded from the study.

The clinical data were obtained from the documentations of the clinicians.

Imaging

All lumbosacral spine MRI examinations were done using 1.5T unit (Philips Achieva, Netherlands, 2010) with a dedicated lumbar coil, and the imaging protocol was as follows:

  1. 1.

    T1-weighted sagittal turbo spin echo (TSE) with 8 ms echo time (TE) and 500 ms repetition time (TR).

  2. 2.

    T2-weighted sagittal TSE with 100 ms TE and 4000 TR.

  3. 3.

    T2-weighted axial TSE with 120 ms TE and 4000 TR, and

  4. 4.

    Myelography with 1000 ms TE and 8000 TR.

All images were interpreted by a specialized radiologist with a minimum of 7 years’ experience. Each lumbar level was assessed for disc degeneration scoring and presence of Schmorl’s nodes, which was seen as a localized lesion in the vertebral endplate with or without sclerotic rim (Fig. 1). The presence of MC and disc herniation was also checked at each disc level.

Fig. 1
figure 1

A 48-year-old man with 6 months of lower back pain. T2-weighted sagittal magnetic resonance turbo spin echo (a, b) (echo time = 100 ms, repetition time = 2363 ms) showing SN in the lower end plate of L3 vertebra in a and in the lower endplate of L2 and upper endplate of L3, associated with score 1 degeneration at L2–L3 disc in b.

Full size image

Statistical analysis

The SN frequency according to age and gender, and its distribution were assessed. The relation between the presence of SN and the scores of disc degeneration, disc herniation, and MC at each single level were calculated using chi-square test. A P value level of less than 0.05 was required for significance. SPSS software, version 17, was used for the statistical analyses.

Result

The demographic criteria of the study sample was as follows: 206 females and 118 males, female to male ratio was 1.7:1, overall mean age was 45.5 ± 12.48 years, their mean weight was 79.2 ± 101 kg, and the mean height was 167.1 ± 6 cm (Table 1).

Table 1 Demographic criteria of the study sample
Full size table

The SN was seen in 22.2% of the patients (41% at one level and 59% in more than one). Twenty-eight percent of males and 18.8% of females had SN; males were affected significantly more than females (P = 0.03) (Table 2).

Table 2 Gender of the SN-affected patients
Full size table

The average age of SN-affected patients was 55 years, 5.5% of them were under 40 years old, 52.8% were between 40 and 59 years old, and 41.7% were over 60 years old. The prevalence of SN was the least at a younger age group (40 years old) and significantly increased as age advanced (P = 0.0001) (Table 3).

Table 3 Average age of SN-affected patients
Full size table

The total number of the discs of 324 patients was 1944; SN was seen in 144 discs. The single disc level was seen in 41% of patients, and more than one disc level in 59%. L1–L2 was the most affected disc level (23.6%) followed by, in a descending order, L2–L3, L3–L4, L4–L5, T12–L1, and L5–S1 (20.8%, 18.1%, 16.7%, 12.5%, and 8.3% respectively) (Fig. 2).

Fig. 2
figure 2

Bar chart showing the overall prevalence of Schmorl’s nodes (SN) by vertebral level

Full size image

The degeneration scores of SN-affected levels were as follows: score zero was seen in 54 patients (37.5%), score 1 in 52 patients (36.1%), score 2 in 20 patients (13.9%), and score 3 in 18 patients (12.5%). The degree of disc degeneration was not affected by the presence of SN, as the P = 0.76 at all scores (Table 4).

Table 4 The degeneration scores of SN-affected levels
Full size table

The SN-affected lumbar levels had also a disc bulge/herniation in 20.8% (30 levels). There was no statistically significant relation between the SN and disc bulge/herniation, as the P = 0.62 (OR = 1.4) (Table 5).

Table 5 Significant relation between the SN and disc bulge/herniation
Full size table

The MC and SN were seen together at 22 disc levels (13.9%). There was a significant relation between MC and SN as P = 0.00001 (OR = 7.9) (Table 6).

Table 6 Significant relation between MC and SN
Full size table

Discussion

The epidemiological data shows 3.8–76% of the general population was diagnosed with SN [11], which is broadly in line with the data from the present study that is showing almost one quarter of the population (22.2%) had SN.

The wide difference in SN prevalence may have different causes, such as different definitions of SN used and different spinal column regions involvement in studies, and some others suggested genetic influence as manifested by variations in ethnic distribution [5, 12].

In the current study, the number of men with SNs was more than women, which in line with previous studies [2, 3, 24]; the research of Dar G. and his collogues considered this high affection among males was due to their larger body size, and taller vertebral bodies and discs both make more mechanical stress on the endplates [5], whilst another study suggested genetic determinants for the male predilection [12].

The prevalence of SN was significantly increased as age advanced, comparable with previous studies that suggested weakened aged cartilaginous endplates and reduced bone density in older age groups; both factors may play a role in pathogenesis of SNs [25,26,27]. The higher lumber levels were affected more, as L1–L2 was the most affected disc (23.6%) in the current study; this was consistent with most of the papers [8, 24, 27]. This could be due to higher mechanical stresses, and the special anatomical features of this part of spine make it more prone to damage by torsional and axial body loads [2].

Several studies did not show significant relation between SNs and disc degeneration nor with disc bulge/herniation, like the current study. As Sonne-Holm S. and his colleagues’ assessed lumbar spine using radiographs in healthy adults [11], whilst Hilton RC and his colleagues studied post-mortem spines, they did not find a significant relation between the SN and DDD at the lumbar region.

Other studies assessed the MRI images of different sample criteria, including healthy twin females [12] and healthy adults [24], a paper analysed discography in adults with back pain [16], and another assessed CT scans of lumber spine [25]; these manuscripts showed a significant relation between SNs and disc degeneration.

This controversy about SN and DDD association might be related to different factors, such as different sample criteria, different radiological modalities used, and other spine region’s involvement with the lumber spine in the studies.

According to this paper, we prefer the theory of endplate osteonecrosis as a cause of SNs, rather than disc degeneration, as a study examined the surgical specimens of SNs proposed that the SNs are the end result of ischemic death of bone beneath the endplate and the herniation of the disc into the body of the vertebra is a secondary phenomenon. The hypothesis of microtrauma is also preferred as Burke et al. found more SN in American soldiers [26]; these minor traumas cause herniation of nucleus pulposus through developmental weak points in the endplates [28]. Also, the developmental models revealed that SNs are already present during skeletal maturation prior the beginning of degeneration [2]. Moreover, the disc degeneration mostly occurs in the lower lumbar levels in reverse to SNs which occurs in the upper lumbar levels.

Modic change was significantly associated with SNs in this study like another study done by Tobias et al .[8]; this result was probably due to disruption of the endplates and the herniation of disc material initiating inflammatory change and edema resembling MC as seen on MRI images.

Conclusion

SN is a frequent MRI finding in the lumbar spine, especially in males, and mostly occurs in the 40–59 years age group. SN is related to vertebral endplate bony pathology rather than discal pathology.