Introduction

Fracture-dislocation of fifth lumbar vertebra (L5) is a rare injury pattern and usually results from high-energy trauma. The direction of dislocation is mostly anterior. Posterior fracture-dislocation of L5 is exceedingly rare injury pattern [1,2,3,4,5,6,7,8,9,10,11,12,13] (Table 1). To the best of authors’ knowledge, atleast nine closed [2, 5, 7,8,9,10,11,12,13] and three open cases [3, 5, 6] of posterior fracture-dislocation of L5 have been reported in the literature. Majority of the reported cases had varying degrees of neurological deficits that improved after surgical decompression. We report two cases of traumatic posterior fracture-dislocation of L5 with comminuted fracture of its body. A review of literature is also presented with special emphasis on fracture morphology and possible mechanism of injury involved in the cases. We suggest that such cases may be considered under separate sub-type in existing Aihara type 5 [1].

Table 1 Published cases of posterior fracture-dislocation of the fifth lumbar vertebra.

Case presentation

Case 1: a 17-years-old girl fell from a height of ~25 feet and experienced severe pain in the lower back, weakness of both lower limbs, and inability to pass urine. She was referred to our institute after 17 days of sustaining the trauma. On examination, a step could be felt in the lower back as spinous process of L4 vertebra was prominent. Neurological examination is summarized in Table 2. We classified her as grade A as per ASIA impairment scale [14]. Radiographs revealed posterior fracture-dislocation of L5 with burst fracture of the body (Fig. 1A) and bilateral calcaneum fractures. Non-contrast computed tomography (NCCT) of lumbosacral spine revealed posterior fracture-dislocation of L5 with comminuted fracture of its body (Fig. 1B). There was presence of bilateral facet joints dislocations; fractures of left pedicle; left lamina; and right pars interarticularis (Fig. 2A, B). The L4 body was intact and displaced posteriorly with respect to sacral vertebrae (Figs. 1A, B and 2C). MRI scan additionally revealed severe compression on the dural sac (Fig. 3A, B).

Table 2 Description of cases reported in present study.
Fig. 1: Radiological features of 17-year-old girl who had a fall from a height (Case 1).
figure 1

A Lateral view radiograph of lumbosacral spine showing posterior fracture-dislocation of L5; B NCCT mid-sagittal view showing comminuted fracture-dislocation of L5. The smaller anteroinferior L5 vertebral body fragment is maintaining the relationship with sacrum, whereas the larger comminuted posterosuperior L5 body fragment is seen retropulsed.

Fig. 2: Detailed NCCT findings of Case 1.
figure 2

A NCCT axial view showing comminuted fracture of L5. Retropulsion of vertebral body fragments into the vertebral canal, fractures of left pedicle and spinous process are apparent. B Coronal section showing facetal dislocation. C Volume rendered image of lumbosacral spine and pelvis, as viewed from antero-lateral aspect from the right side, showing posterior fracture-dislocation of L5, right-sided facet joint dislocation, and fracture of right pars interarticularis.

Fig. 3: MRI findings of Case 1.
figure 3

A T2-weighted mid-sagittal section showing posterior fracture-dislocation of L5 and resultant compression of dural sac by retropulsed fragments. B Axial section showing severe compression of neural elements with associated prespinal, paraspinal, and ventral epidural collection indenting the thecal sac (possibly representing fracture-hematoma or cerebrospinal fluid).

She underwent surgery via posterior midline approach. Identification of the different parts of posterior elements was difficult because of comminution. There was circumferential loss of dura at the level of L5 resulting in CSF leak. Cauda equina was found severely compressed due to the retropulsed L5 body fragment. The roots were found contused and tethered between the retropulsed L5 body fragment and the lamina of L4 superiorly and S1 inferiorly. S1 nerve root was found avulsed. Comminuted L5 vertebral body fragments were removed from either side of cauda equina and nerve roots. The screws could not be placed in L5 vertebra as posterosuperior body fragment was small and left pedicle was fractured. Anteroinferior part of L5 body, that was lying in alignment with S1, was left in situ. Intervening discs between L4-L5 and L5-S1 were found ruptured and, thus removed. Inferior end-plate of L4 vertebra and superior end-plate of S1 vertebra were prepared by removing disc cartilage. Short segment pedicle screw fixation was done at L4 and S1 vertebrae. Anterior reconstruction was performed with a titanium cage filled with autologous bone graft as (1) McCormack score [15] was 8; and (2) considerable amount of bone from L5 vertebral body had to be removed during decompression. Despite our best efforts, full correction of lordosis could not be achieved probably because of delayed presentation. Supplemental posterolateral fusion was also performed.

Postoperative period was uneventful. She showed minor neurological improvement in the form of flickering of right great toe at 2 weeks. She was mobilized out of the bed on wheelchair with a modified brace (lumbosacral corset incorporated in Taylor’s brace). She was able to walk with the help of a quadruped walker at 6 months follow-up. At 5 years, she was able to walk independently without pain and had regained clinical bladder control (ASIA impairment scale: C). Detailed neurological status is shown in Table 2. Radiographs showed well-seated implants though inadequate lordotic correction was apparent. The cage was close to spinal canal proximally and one of the sacral pedicle screws was longer (Fig. 4A). NCCT revealed anterior fusion mass between L4 and S1 (Fig. 4B).

Fig. 4: Follow-up imaging of Case 1 at 5 years.
figure 4

A Follow-up lateral view radiograph at 5 years showing fixation construct and interbody cage L4-S1. Though the upper-posterior margin of the cage is closer to the posterior aspect of L4 vertebral body and one of the sacral pedicle screws was longer, the implant construct maintained its original placement. B Follow-up midsagittal view of NCCT showing evidence of anterior fusion mass at 5 years.

Case 2: a 28-year-old gentleman was hit by a three-wheeler from behind as he was riding scooter. He presented to our institute the next day with low back pain, weakness in left foot, and difficulty in passing urine. Tenderness was present at L4-L5 level. Detailed neurological examination is given in Table 2. We classified him as grade A as per ASIA impairment scale [14]. Radiograph showed posterior fracture-dislocation of L5 with burst fracture (Fig. 5A). NCCT revealed the details of injury patterns (Fig. 5B–D) as “empty facet sign” suggestive of its dislocation was apparent on the right side (Fig. 5C). MRI scan showed compression on the dural sac, and dislocated right facet joint (Fig. 6A, B). He underwent surgery under general anesthesia in prone position via posterior midline approach. Intraoperatively, we had a breach of S1 pedicle screw on the right side, so we put S2 sacral alar-iliac screws. We did short segment fixation between L4 and S2 without anterior reconstruction because (1) adequate lordotic correction could be achieved without much bone resection; and (2) McCormack score [15] was 6 (Fig. 7A, B). Postoperative period was uneventful. At 2 years of follow-up, he had functional recovery in ankle dorsiflexion and plantarflexion. He had partial sensory loss over L5 and S1 dermatome on the left side with pressure sore over lateral malleolus (Table 2) (ASIA impairment scale: D).

Fig. 5: Radiological features of 28-year-old man who had a road-side vehicular accident (Case 2).
figure 5

A Lateral view radiograph of the lumbosacral spine showing posterior fracture-dislocation of L5. B NCCT mid-sagittal view showing comminuted fracturedislocation of L5. The smaller anteroinferior L5 vertebral body fragment is maintaining the relationship with sacrum, whereas the larger comminuted posterosuperior L5 body fragment is seen retropulsed. C Axial view showing “empty facet sign” on the right side suggestive of facetal dislocation (blue arrow). Retropulsion of vertebral body fragments into the vertebral canal, fractures of left pedicle, and spinous process are apparent. D Coronal view showing rotatory dislocation of L5 as facet joint is not seen on the right side.

Fig. 6: MRI findings of Case 2.
figure 6

A T2-weighted MRI mid-sagittal view showing posterior fracture-dislocation of L5. Retropulsed L5 vertebral body fragments are seen compressing the dural sac. B Axial view showing compression of neural elements on right side. Additionally, “empty facet sign” on the right side suggestive of facetal dislocation.

Fig. 7: Follow-up imaging of Case 2 at 2 years.
figure 7

A 2 years follow-up anteroposterior view, and B lateral view radiograph showing well placed and maintained L4 pedicle screws and S2 sacral alar-iliac screws.

A written informed consent was obtained from both the patients/parents authorizing the treatment, photographic documentation, and radiographic examination. They were also informed that data concerning the case may be submitted for publication and they agreed.

Discussion

Regarding mechanism of injury

Fracture-dislocations of L5 vertebra usually result from high-energy trauma and represent unstable spinal injury. Majority of the reported cases had severe neurological deficit at the time of presentation [2,3,4,5,6,7,8, 10,11,12]. Such unstable injuries may require combined anterior and posterior stabilization [1, 3, 4, 9, 12, 13]. Flexion-distraction as well as hyperextension and shear have been postulated to cause these injuries [7, 9, 16]. These mechanisms have been suggested on the basis of case reports alone owing to the rarity of the entity and a consensus in the literature is lacking in relation to the mechanism of injury [7]. Though Steinitz et al. considered this injury to have flexion-distraction mechanism [9], hyperextension of the spine as the mode of injury is more widely accepted [7, 15]. Denis and Burkus [16] reported 12 cases of thoracic and lumbar spine fracture-dislocations following forceful hyperextension injury to the spine (lumberjack paraplegia). They found disruption of the anterior longitudinal ligament and anterior annulus. They suggested that the associated frequent fractures of posterior elements were secondary to compression element of the hyperextension injury. We suggest hyperextension and axial compression injury to the spine as the possible mechanism of injury in our patients. Torso would have exposed to axial compression first due to sudden deceleration following the fall. Continued force would have concurrently produced hyperextension and shear forces in addition to the axial compression. Thus, it seems that the axial compression in hyperextension position could have produced this injury pattern.

Regarding the possibility of placing these injuries in a separate sub-type

Aihara et al. [1] presented seven cases of fracture-dislocation of L5 and reviewed 50 other such cases from the existing literature and classified the fracture-dislocation of L5 into five types: Type 1 (unilateral lumbosacral facet dislocation with/without facet fracture and intact contralateral facet); Type 2 (bilateral lumbosacral facet dislocation with/without facet fracture); Type 3 (unilateral lumbosacral facet dislocation and contralateral lumbosacral facet fracture); Type 4 (dislocation of the body of L5 with bilateral fracture of the pars interarticularis-acute spondylotic spondylolisthesis); Type 5 (dislocation of the body of L5 with fracture of the body and/or pedicle with/without injury of the lamina and/or facet).

Posterior fracture-dislocations were not classified separately but placed in the system depending upon the geometry of injury pattern, even though the mechanism of injury is essentially different. They did mention that 4 out of total 57 cases had posterior fracture-dislocations, only one of these cases had belonged to their series that was described in their study with partial details. Since both the cases in our report had fracture of L5 vertebral body in addition to facetal fracture/dislocation, we suggest inclusion of these cases as a separate sub-type in Aihara type 5. A literature review detailing 13 published cases of posterior fracture-dislocation of L5 is presented (Table 1). Out of these, at least three were open fracture-dislocation [3, 4, 6] and one of these was managed with wound care alone [6].

Ver et al. [17] has described a new classification for traumatic lumbar spondylolisthesis which included complex fracture-dislocation of L5 vertebrae as Type 6. They recommended posterior instrumentation and fusion involving additional levels superiorly or inferiorly with possible anterior interbody fusion. Type 6 lesions had more severe neurological deficit with lower fusion rates suggesting poorer prognosis.

Cebesoy et al. [2] presented a similar case of retrolisthesis of L5 over S1. They followed the case for 16 years after the surgical decompression and fusion and suggested that cauda equina has a good healing potential even after such devastating injuries. As reported in other cases, 5 cases out of total 8 showed neurological improvement after surgical decompression [2, 4, 5, 7, 13]. The clinical results of our cases seem to be in tune with these published cases.

Conclusion

Posterior fracture-dislocation of L5 needs to be identified as a separate sub-type in existing Aihara classification. These injuries usually have higher chances of associated injuries and neurological deficit. Spinal decompression may result in neurological improvement.