Lumbar spondylolysis is a stress fracture of the pars interarticularis. We have evaluated the site of origin of the fracture clinically and biomechanically.
Ten adolescents with incomplete stress fractures of the pars (four bilateral) were included in our study. There were seven boys and three girls aged between 11 and 17 years. The site of the fracture was confirmed by axial and sagittal reconstructed CT. The maximum principal tensile stresses and their locations in the L5 pars during lumbar movement were calculated using a three-dimensional finite-element model of the L3-S1 segment.
In all ten patients the fracture line was seen only at the caudal-ventral aspect of the pars and did not spread completely to the craniodorsal aspect. According to the finite-element analysis, the higher stresses were found at the caudal-ventral aspect in all loading modes. In extension, the stress was twofold higher in the ventral than in the dorsal aspect.
Our radiological and biomechanical results were in agreement with our clinical observations.
Spondylolysis is an important cause of low back pain in children and adolescents, especially in athletes or patients with dystonic cerebral palsy.1 The condition, which presents with low back pain during exercise, is a defect of the pars interarticularis and occurs in about 6% of the population.2,3 Furthermore, spondylolisthesis occasionally presents in the affected lumbar spine during the period of rapid growth.2–7
It is likely that the pathogenesis of spondylolysis is a stress fracture.8–12 Sairyo et al13 have reported that the movements which may contribute most to its aetiology are extension and axial rotation.
If spondylolysis is a stress fracture and not an acute injury, it is likely that it appears as a fracture line which is first seen at one cortex and not circumferentially. However, the literature on the earliest radiological findings in the development of spondylolysis is sparse. Our hypothesis was that the fracture line originates at the site of maximum tensile stress and that this correlates with the clinical findings. In this study, therefore, we reviewed adolescents aged 17 years and under with early spondylolysis in order to understand the initial event in the cortex of the pars interarticelaris. Also, by using finite-element analysis, we have evaluated the tensile stresses and their positions in the pars to correlate the stresses with clinical observations.
Patients and Methods
We assessed ten patients with spondylolysis who were aged between 11 and 17 years (seven boys, three girls; Table I⇓). They were all very active in sport. Four had bilateral spondylolysis, making a total of 14 defects. The inclusion criterion was incomplete spondylolysis whereby the fracture line did not cross the entire cortex of the pars interarticularis on CT scans. Therefore, based on these findings, the location of the defect was identified. In all cases, we also carried out MRI to assess any associated high signal change in the adjoining pedicle on the T2-weighted image.11
In a biomechanical study, an experimentally validated three-dimensional (3D) non-linear finite-element model of the intact ligamentous L3-S1 segment was used (Fig. 1a⇓).10,11,14–17 The geometrical data of the L3-S1 movement segment were obtained from CT scans (transverse slices 1.5 mm thick) of a normal cadaveric specimen of the lumbar spine. When stacked sequentially, digitised cross-sectional data generated the model using the commercial software ABAQUS 6.8 (SIMULIA inc, Providence, Rhode Island). The cortical and cancellous regions were constructed with 3D hexagonal elements (C3D8), which were assigned different material properties. The facet joints were simulated by 3D gap contact elements representing the physiological nature of the cartilaginous articular surface. These elements transferred force between nodes along a single direction as a specified gap between them closed as they came closer to each other due to movement of the respective surfaces. The elements were only able to transmit compression. The intervertebral disc was modelled as a composite of a solid matrix with embedded fibres orientated at ±30° to the horizontal (via the *REBAR parameter in ABAQUS) in concentric rings around a pseudofluid nucleus. “*REBAR” command in ABAQUS allowed us to simulate the fibres embedded in annulus matrix.18 The hydrostatic properties of the nucleus were simulated by C3D8 hexagonal elements which were assigned a very low stiffness (1 MPa) and near incompressibility (Poisson’s ratio, n = 0.4999). All seven major spinal ligaments were represented as elastic truss elements and assigned non-linear material properties. Natural changes in ligament stiffness (initially low stiffness at low strains followed by increasing stiffness at higher strains) was simulated through the designation of the ‘hypoelastic’ material, which allowed the definition of the axial stiffness as a function of axial strain. Three-dimensional two-node truss elements (T3D2) were used to construct the ligaments. Stress distributions in various structures were analysed in flexion, extension, lateral bending and axial rotation in response to axial compression of 400 N and a moment of 10.6 Nm. The maximum principal tensile stresses and their locations in the pars interarticularis were computed (Fig. 1⇓).
The CT scans in all ten patients showed an incomplete stress fracture, compatible with early spondylolysis according to the classification of Fujii et al.9 In the sagittal reconstructions, it was confirmed that the pars defect was present only in the caudal-ventral aspect (Fig. 2⇓). The high signal change on MRI has been reported to be the indicator of the early stage11 and in all our cases, this was confirmed.
In the biomechanical study the maximum principal stresses at the ventral wall of the L5 pars interarticularis were 2.5, 59.5, 10.7, 5.0, 14.5 and 33.8 MPa for flexion, extension, ipsilateral bending, contralateral bending, ipsilateral rotation and contralateral rotation, respectively. Corresponding stresses at the dorsal wall were 9.6, 26.7, 10.1, 1.5, 2.2 and 28.2 MPa. The stress magnitude at the ventral aspect was higher than that at the dorsal aspect during all lumbar movements except flexion (Fig. 3⇓). The highest stress was found in extension (Fig. 4a⇓). Similar values were observed in axial rotation (Fig. 4b⇓) and in extension the stress magnitude was twofold higher in the ventral than in the dorsal aspect.
Our data suggest that spondylolysis in children and adolescents always occurs as a stress fracture at the caudal-ventral aspect of the pars because tensile stresses are higher in the ventral than in the dorsal aspect. The likely initiating event will occur when the subject is undertaking repeated extension and/or axial rotation manoeuvres during sport. This agrees with the clinical observations in all our patients.
Therefore, once it begins, the defect progresses in a craniodorsal direction and, without treatment, becomes circumferential over approximately two months (Fig. 5⇓). At the initial consultation with one patient (case 3, Table I⇑), we recommended that she cease all sport and wear a brace. She did not take this advice and continued in sport. Two months later, she re-attended with severe low back pain. CT confirmed the progression of the fracture line from the caudal-ventral aspect to the craniodorsal direction (Fig. 5⇓). Early spondylolysis can heal completely with conservative treatment.9–12 Therefore, it is very important to diagnose the disorder correctly before it progresses to pseudarthrosis.
When forces causing compression, bending or twisting exceed the normal range, the bone will give way, and because it is stronger in compression it will fail earlier in tension. When a bending load is applied to a beam (Fig. 6⇓), loads are set up or induced in the beam. The top section is compressed and becomes shorter, whereas the bottom section is in tension and stretches. Tensile failure occurs at the underside of the beam, initially producing small tensile cracks which progress across the beam, eventually reaching the top side so that the beam fails. Our results indicate that the magnitude of the bending stresses at the caudal-ventral aspect of the pars create the highest tensile stress during extension and axial rotation and suggest that spondylolysis progresses from the ventral aspect of the pars.
For early diagnosis, bone scintography is usually advised19,20 and, more recently, Sairyo et al11 have proposed that high signal change of the adjoining pedicle on the T2-weighted MR scan can be a good indicator of an early-stage defect. However, the most reliable method is CT.9 From our findings, we propose that an axial slice through the caudal aspect of the pars and sagittal reconstruction CT are the most important methods for the diagnosis of early-stage spondylolysis (Fig. 7⇓).
No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.
- Received May 28, 2009.
- Accepted March 23, 2010.
- © 2010 British Editorial Society of Bone and Joint Surgery