Transpedicular Approach for Corpectomy and Circumferential Arthrodesis in Traumatic Lumbar Vertebral Body Burst Fractures: A Retrospective Analysis of Outcome in 35 Patients
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.314521
Source of Support: None, Conflict of Interest: None
Keywords: Circumferential arthrodesis, corpectomy, posterior transpedicular approach, single stage posterior only approach, traumatic lumbar burst fracturesKey Message: Transpedicular approach for circumferential arthrodesis in TLBFs is a challenging approach. However, the sense of familiarity amongst spine surgeons, lesser approach-related morbidity, shorter operative time and results comparable to anterior/combined approaches make TA an attractive and a viable approach for TLBF.
Traumatic lumbar fractures account for ~ 44% of all traumatic vertebral fractures. Traumatic lumbar burst fractures (TLBFs) may cause canal compromise and neural compression from fractured fragments as well as kyphotic deformity. The goal of surgical management in fractures requiring surgical intervention is to achieve adequate decompression of spinal canal, correction of kyphotic deformity and spinal stabilization. This requires corpectomy followed by anterior column reconstruction along with pedicle screws and rod (PSR) insertion to achieve circumferential arthrodesis.
Transpedicular approach (TA) that includes transpedicular corpectomy and anterior column reconstruction via a posterior only approach, combined with PSR fixation, can achieve canal decompression and circumferential arthrodesis via single posterior approach.,,,,,,,,,,,,,,, TA has mostly been described for non-traumatic vertebral pathologies including primary or metastatic vertebral tumors, especially of thoracic spine where the thoracic nerves can be sacrificed with impunity,,,,,,,, However, transpedicular corpectomy for TLBFs is technically challenging because of the indispensable lumbar nerve roots.,,,,,
In this retrospective study, we present our experience in managing TLBFs using TA. The surgical technique, results of surgery, complications, patient outcome and pertinent literature are discussed in detail.
This was a retrospective study conducted at a tertiary care trauma center. We included 35 patients with TLBFs who underwent circumferential arthrodesis via TA over a period of 5 years. Patient's casualty/in-patient records, radiological investigations (MRI and NCCT), operation notes and out-patient records were scrutinized to collate data. Demographic details, pre- and postoperative neurological status, preoperative deformity, intraoperative details, duration of hospital stay, deformity correction and bony fusion at last follow-up were analyzed.
A detailed preoperative neurological examination was done and recorded. Patient's neurological status was then scored using ASIA scoring system. All patients underwent preoperative spine CT scan to look for type of fracture and the extent of canal compromise. Cobb's angle, defined as the angle between lines drawn along superior endplate of unaffected vertebral body (VB) rostral to deformity, and inferior endplate of unaffected VB caudal to deformity, was used to quantify deformity. Most patients also underwent preoperative MRI for assessing the presence of disc prolapse and injury/compression of spinal cord/nerve roots. All patients included in the study either had post-traumatic kyphosis or a fracture fragment impinging on the cord. Thoracolumbar Injury Classification and Severity (TLICS) was used to decide between conservative or surgical management.
After positioning the patient prone, posterior bony elements were exposed and bilateral laminectomies of fractured vertebra were performed to decompress thecal sac. Pedicle screws were inserted under C-arm/O-arm guidance and a rod was placed on right side and distracted to open up the space for corpectomy. The superior and inferior facets surrounding the pedicle of VB to be resected were drilled to expose the pedicle and nerve root. Initially, pedicle on left side was removed followed by corpectomy using high-speed drill, curettes and rongeurs [Figure 1]. Posterior longitudinal ligament was preserved to prevent thecal sac damage from drill/instruments. Majority of VB could be removed through unilateral transpedicular route, the remainder was removed through contralateral pedicle. Resection of cortical bone immediately behind great vessels was not done to avoid vascular injury., The intervertebral discs (both rostral and caudal to the resected VB) were thoroughly removed. The cartilaginous endplate was meticulously removed while exercising due caution to avoid gross violation of bony endplates. The cavity was then thoroughly washed with saline to remove any loose disc fragments. The exiting nerve roots at corpectomy level were preserved (black arrow heads in [Figure 1]a and [Figure 1]b). Anterior reconstruction was performed by inserting titanium expandable/mesh cages filled with autologous bone. For safe insertion of cage, roots were mobilized and cage was inserted along the axis of nerve roots and then turned in situ to lie along the long axis of spine. The collapsed cage was positioned with caudal portion of cage in contact with endplate of caudal vertebra, while rostral portion of cage telescoped superiorly with expansion. Mesh cage was cut in appropriate size and positioned by placing inferior end of cage against superior endplate of VB below, then the cage was rotated and superior aspect of cage was gently tapped against inferior endplate of vertebral body above [Figure 2]. After cage placement, compression was applied across posterior pedicle screw fixation so that the cage snugly fits and remains in contact with end plates on either side. However, caution was exercised to avoid overzealous compression which could increase the risk of cage subsidence. The posterior bony elements of vertebrae above and below were then decorticated with drill to achieve raw bony surface and then covered with locally harvested bone fragments. The removal of all three columns at fracture site and distraction achieved by placing an inter-body cage made it feasible to achieve restoration of height and deformity correction [Figure 3] and [Figure 4].
A detailed postoperative neurological examination was done and scored using ASIA scoring at each follow-up. Neurological status was categorized as improved, stable or worsened based upon ASIA grade change. Postoperative NCCT was done in all patients before discharge and then during follow-up. Cobb's angle was measured on postoperative and follow-up NCCT. Also CT scans were analyzed for fusion, subsidence of cage or implant failure. Bony fusion was evaluated using radiographic criteria i.e., formation of bridging trabecular bone and absence of a dark halo around the implant or implant fracture.,
A total of 35 patients underwent circumferential arthrodesis via TA during study period. The age range of patients was 13-58 years with an average of 28.4 years. There were 21 males and 14 females. The mode of injury was road traffic accident in 29, fall from height in 3 and fall of a heavy object in 3 patients. L1 was the most common vertebra involved in 18 (51%) patients followed by L2 in 9 (25%) patients [Table 1].
Preoperative neurological status: [Table 1]
Amongst 35 patients with available follow-up, 8 (22.2%) belonged to ASIA A with complete loss of motor/sensory and bowel/bladder function below the level of injury. Twenty (57%) patients had incomplete neurological deficits (ASIA C and D) while 2 (5%) patients had only sensations intact with paraplegia (ASIA B). Only 5 (13.9%) patients were neurologically intact (ASIA E) pre-operatively. All the patients had a TLICS score >=4 and thus were managed surgically.
Preoperative Cobb's angle was calculated in all patients. Most of the patients presented with kyphotic deformity of 10-20° and the mean preoperative kyphotic deformity was 13.97°.
Operative parameters and Perioperative Complications: [Table 1]
The mean duration of surgery was 386 minutes (range: 240-660 min) and mean blood loss was 1352 mL (range: 500–5000 mL). The patient with 5000 mL blood loss had excessive bleeding from the pravertebral venous plexus as well as from the fractures VB. Short segment (one segment cephalad and one caudal to level of corpectomy ) fixation was done in 23 patients while 12 patients underwent long segment (>=2) fixation. The decision of short/long segment fusion was based on surgeon's preference. The expandable cage was used in 21 and mesh cage in 14 patients. Iatrogenic durotomies occurred in four patients while dural tears secondary to trauma were found in nine patients. Preexistent nerve root injury attributable to trauma was encountered in three patients and none suffered iatrogenic injury.
Postoperatively, two patients had surgical wound infection with dehiscence, which required debridement and resuturing. There was no perioperative mortality or postoperative neurological deterioration. The mean postoperative hospital stay was 5.92 (2–22) days [Table 1].
Neurological status at last follow-up: [Table 2]
Out of 35 patients, 27 (77.1%) showed neurological improvement after surgery. Amongst these 27 patients, 17 (48.6%) had improvement by one ASIA grade, 8 (22.8%) improved by 2 grades and 2 (5.7%) patients improved by 3 grades. The improvement in motor functions was observed in 26 (74.2%) patients. Nineteen (54.3%) patients had preoperative bladder and bowel disturbances, which improved in 5 (14.3%) patients. One patient regained bladder sensations but remained incontinent. Three patients with preoperative ASIAA and 5 with ASIAE score remained neurologically same, while none of the patients showed neurological deterioration. The median ASIA score improved from 3 to 5.
Status of deformity at last follow-up
The cobb's angle of lordotic spine after deformity correction was marked as negative to determine extent of correction. The mean preoperative and postoperative cobb's angles were 13.97° and –3.57°, respectively, thus achieving a mean cobb's angle correction of 17.54°. Over a mean follow-up of 39.1 months, there was a loss of lordosis to some extent in some patients with cage subsidence seen in 8 patients. Out of these, 6 were observed in patients with expandabale cage and two were observed in patients with mesh cage. The mean Cobb's angle during follow-up was 1.23° thus resulting in loss of 4.8°.
Fusion: Over a mean follow-up of ~39.1 months, 91.4% patients achieved bony fusion.
During follow-up scans, eight patients had cage subsidence of >2.5 mm [Figure 4c] with an average subsidence of 7.2 mm and 1 patient had a broken pedicle screw. The subsidence was measured on CT scans and was defined as the sum of subsidence of superior and inferior ends of cage into VB. These patients complained of mild backache which was controlled with analgesics and none of them required revision surgery.
Though optimal treatment for TLBFs, in absence of neurological deficits or significant instability, still remains a matter of debate,,,, surgical decompression and fixation is usually recommended for patients with significant deformity and in those with neurological deficits due to canal compromise., Surgical management in these patients requires corpectomy and spinal canal decompression followed by circumferential arthrodesis. The choice of surgical approach (anterior/combined vs posterior), however, remains a matter of debate.,,,,,,,,,
Anterior approach, first described by Burns in 1933, has been used by many authors for various pathological conditions of spine.,,,,,,,,,, Though, anterior approach has certain advantages including direct access and decompression of spinal canal, straight forward graft/cage placement, correction of kyphotic deformity, no para-vertebral muscle dissection, no risk of nerve root injury or cerebrospinal fluid leakage,,,,, it is a relatively more invasive approach that requires traversing body cavities with unavoidable handling of viscera and great vessels. Such an extensive surgery becomes especially demanding in those who are obese, have co-morbidities and those who have had previous abdominal surgeries.,,,,,,,, It is also associated with late visualization of neural elements and requires two separate incisions for combined anterior-posterior approach, each with additive risks and prolonged hospital stay.,,,,,,,, The most common complications reported with anterior approach are vascular injuries (range 1-16%) followed by thromboembolic events (0 to 12%).,, Other complications causing substantial morbidity include abdominal hernias and possible impotence or retrograde ejaculation in male patients. Vahldiek and Punjabi and Khodadadyan-Klostermann et al. have recommended that anterior column stabilization with cage placement and anterior plating is insufficient to provided required stability following corpectomy in thoracic and lumbar region. Thus, anterior approach may also require additional posterior approach for achieving 360° fusion.,,,
Posterior approach for lumbar interbody fusion was first described by Cloward but it did not become popular due to high complication rates. Later, modifications to include pedicle screw fixation resulted in improved rates of arthrodesis with decreased rates of graft extrusion., However, a high failure rate of standalone posterior instrumentation via PA and later need of augmentation by AA in a significant number of patients is an important argument in favor of AA and against posterior approach.,,, TA combines the advantages of both anterior and posterior approaches to achieve circumferential arthrodesis from posterior route only, thereby obviating the need for an additional AA and also offers familiarity, less invasiveness, early visualization of neural elements and decreased morbidity of posterior approach.,,,,,,,,,,,,,, TA for thoracic corpectomy and interbody cage placement has been a well-described approach and can be performed with little difficulty as thoracic nerve roots can be easily sacrificed.,,,,,,, The real difficulty is in performing this approach in lumbar spine where one can not sacrifice the roots and available working corridors are thus very narrow. This is the main reason why TA has not gained popularity amongst surgeons for lumbar spinal pathologies. However, recently, there has been a growing interest in this approach for managing various pathologies of lumbar region as well,,,,,,,,,,,,,, and one of the major reasons for this has been the availability of various types of interbody cages including nondistractable mesh cages and expandable ones., The nondistractable mesh cages are more difficult to be inserted as compared with expandable cages. In this study, successful mesh cage placement was performed in 15 patients. Pre-insertion slight distraction across corpectomy level helped in insertion of appropriate sized cage in all cases [Figure 2]. However, an important encouraging advancement has been the development of expandable titanium cages which are easier to be inserted in compressed state through narrow corridors bounded by lumbar nerve roots and can be expanded once placed in desired position and orientation.,,, In this study 21 patients underwent expandable cage placement. In our study, we found no significant difference in terms of fusion rates, subsidence or construct failure between the groups of patients undergoing mesh/expandable cage placement.
In this study, we achieved a kyphosis correction of 17.54 degree (from 13.97° to -3.57°), which is comparable to what has been described in literature.,,,,,,,,,,,,,,,, Wang et al. reported a kyphosis correction from 26° preoperatively to 12.4° postoperatively amongst 14 patients with lumbar fractures undergoing surgery via TA.
Sciubbia et al. successfully used TA to achieve average kyphosis correction of 15.9° (from 28° to 12.1° at a 16-month follow-up). Recently, Choi et al. described TA for circumferential arthrodesis in 11 patients with traumatic lumbar fractures. They achieved a mean kyphosis correction of 7.7°(9.2° to 16.9°) after surgery. These results are comparable to those achieved by anterior approaches by various authors.,,,,,,,,
In this study, we observed a loss of deformity correction from -3.57° to 0.52° (loss of 4.22°) during 29.1 month follow-up. The loss of kyphosis correction has been consistently observed in various studies in literature. Wang et al. also reported a loss of kyphosis correction from 12.4° to 13.3° at 31 months follow-up after single-stage thoracolumbar corpectomies and placement of expandable cages in 28 patients. Similarly, Choi et al. also observed a loss of kyphosis angle from 16.9° to 15.1° over a 12-month follow-up period. Similarly, loss of kyphosis correction has also been observed by authors using anterior approaches.,,
In our study, neurological improvement was observed in 28 (77.8%) patients while none deteriorated after surgery. There was an average improvement in ASIA score from 2.97 to 4.17 over a period of 29.1 months. Wang et al. reported an improvement in ASIA score from 3.7 to 4.5 over 31 months in their study of 28 patients with traumatic fractures operated via posterior approach. Similarly, Sasani et al. and Hofstetter et al. noted improvement from 4.0 to 4.4 and 3.7 to 4.2 in their respective studies of 14 and 67 patients with a follow-up period of 24 and 14 months, respectively.
The mean estimated blood loss (EBL) and mean operative time in this study were 1352 mL and 393.8 min, respectively, which are comparable to that reported in literature,, Danisa et al. compared the results of surgery in thoracolumbar burst fractures, operated via anterior, posterior or combined approaches. They concluded that posterior approach was associated with a statistically significant decrease in operative time, blood loss and need of blood transfusion as compared with anterior and combined approaches. On the other hand, no statistically significant difference was seen with respect to degree of kyphosis correction or improvement in neurological functions. Similarly, Lu DC et al. while comparing anterior, posterior and combined approaches concluded that morbidity is lesser and hospital stay shorter in patients undergoing surgery via posterior only approach. The anterior-posterior approach was associated with longer mean surgical times (729 vs 450 minutes), increased EBL (3154 vs 1857 mL) and higher complication rates (41 vs. 29%). Similarly results have been reported by others as well.
The literature as well as the results of this study clearly demonstrate the feasibility of circumferential arthrodesis via TA in lumbar region, as well as results comparable to those achieved with anterior or combined anterior and posterior approaches. In addition TA is associated with either similar or lesser complications as compared with anterior or combined approaches. TA also offers the advantage of a sense of familiarity to neurosurgeons and spine surgeons.
Our study demonstrates that it is feasible to place an adequately sized interbody cage using posterior only approach for anterior column reconstruction and without injuring nerve roots even in the lumbar region. The key is to perform wide dissection, adequate mobilization of lumbar roots and distraction across corpectomy level, thus creating a corridor for cage insertion.
The this study is one of the largest study describing corpectomy and circumferential arthrodesis via TA in patients with TLBFs. Our study is also important because it highlights the ability of TA in achieving circumferrential arthrodesis via single posterior only approach thereby avoiding the complications and risks associated with anterior and combined approaches and adds to the relatively scarce literature on the subject.
The retrospective nature of the study is the limiting factor of our study.
The transpedicular approach although technically difficult can be used for deformity correction and anterior column reconstruction in TLBFs, without sacrificing nerve roots. The sense of familiarity with posterior approach amongst spine surgeons, lesser approach-related morbidity, shorter operative time, less blood loss and the results comparable to anterior or combined approaches make single-stage TA an attractive and viable alternate option for management of TLBFs and adds to the armamentarium of spine surgeons.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2], [Figure 3], [Figure 4]
[Table 1], [Table 2]