Single vs Double Index Screws in Thoracolumbar Burst Fractures: A Retrospective Analysis

Abstract

Study Design

Retrospective study.

Objectives

To assess the radiological, clinical, and neurological outcome of patients who had thoracolumbar burst fractures and had undergone short-segment posterior stabilization with single-index screw (SIS) or double-index screws (DIS) at the fractured vertebra.

Methods

We included patients with AO type A and B with thoracolumbar injury classification and severity score (TLICS) ≥ 5, load-sharing classification (LSC) score ≥ 7, and a follow-up period of more than 2 years. Of 94 patients, 45 had DIS fixation, and 49 had SIS fixation. All patients’ pre-operative, immediate post-operative, and final follow-up visit X-rays were analyzed for Cobb’s angle (CA), vertebral wedge angle (VWA), and vertebral body compression ratio (VBCR) in a neutral lateral view and transverse spinal area (TSA) in axial CT section. Clinical and neurological outcomes were assessed using the Visual Analog Scale (VAS) for back pain, the Oswestry Disability Index (ODI), and the American Spinal Injury Association (ASIA) Impairment Scale (AIS) at their final follow-up visit.

Results

The radiological parameters between both groups showed significant differences in terms of CA, VWA, and VBCR at the latest follow-up (SIS: 12.5°, 15°, 80.1% vs DIS: 8.04°, 11.2°, 87.1%) (p < 0.05). VAS and ODI were better in the DIS group but were statistically insignificant (p > 0.05).

Conclusions

Short-segment posterior stabilization with DIS in the fractured vertebra resulted in radiologically better maintenance of CA, VWA, and VBCR compared to SIS. There were no significant differences in terms of clinical (VAS and ODI), neurological outcome (AIS), and spinal canal remodeling (TSA).

Introduction

Thoracolumbar vertebral body burst fractures account for more than half of thoracolumbar spinal fractures [1]. The goal of treating these fractures is the restoration of spinal stability and alignment, kyphosis correction, and neurological recovery. The optimal method to achieve these goals remains an area of debate. Different authors variably favor a variety of treatment options ranging from non-operative to surgical decompression and stabilization. The surgical options include posterior, anterior, or combined approaches using open or minimally invasive techniques.

With the advent of modern implants and instrumentation, short-segment fixation with index screws in the fractured vertebra became the standard surgical method for treating thoracolumbar burst fractures [2, 3]. However, there is no clear consensus on whether a single- or double-index screw is ideal for management of thoracolumbar burst fractures. Further, there is very limited research available on this comparison. Thus, the current study is aimed to assess the radiological, clinical, and neurological outcome of patients who had thoracolumbar burst fractures and had undergone posterior stabilization with single-index screw (SIS) or double-index screws (DIS) at the fractured vertebra.

Materials and Methods

We conducted this study in compliance with the principles of the Declaration of Helsinki. The study’s protocol was reviewed and approved by the Institutional Review Board (IRB No. PHPL/ACAD-2023-2/11). It was a retrospective analysis of all patients who underwent short-segment posterior-only pedicle screw stabilization with SIS or DIS for thoracolumbar burst fractures from 2018 to 2022 in a tertiary spinal care center. The study adhered to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines [4].

Study Participants

The inclusion criteria were as follows: AO type A and B with thoracolumbar injury classification and severity score (TLICS) [5] ≥ 5, load-sharing classification (LSC) score [6] ≥ 7, and a follow-up period of more than 2 years. We treated 237 thoracolumbar fractures during the study period. Of these, 94 patients met our criteria and were included in our study. Of 94 patients, 45 had double-index screw fixation, and 49 had single-index screw fixation. We excluded patients with pathological fractures, osteoporotic fractures (fractures resulted from trivial fall or fracture in patients with DEXA score ≥ − 2.5), ≥ 2 vertebra fractures, TLICS ≤ 4, LSC score ≤ 6, patients with < 2 years of follow-up period, patients who had post-operative surgical site infection, and patients with AO type C injury.

Surgical Procedure

All patients had undergone short-segment posterior stabilization using C-arm guidance under general anesthesia. The decision to use a single- or double-index screw was made using a pre-operative computed tomography (CT) scan based on the integrity of the pedicles of the fractured vertebra. If both pedicles of the fractured vertebra were intact, double-index screws were applied, and a single-index screw was applied if either of the pedicles was intact. No attempts of impaction or removal of retropulsed fragments from the spinal canal were made. No patients underwent interbody fusion. All patients underwent similar post-operative rehabilitation with thoracolumbosacral orthosis following surgery for up to 3 months, and early ambulation with rehabilitation was carried out in both groups.

Clinical Assessment

The demographic variables collected include age, gender, time to surgery, operating time, intra-operative blood loss, hospital length of stay, and neurological status using the American Spinal Injury Association (ASIA) Impairment Scale (AIS) [7] were documented. Clinical outcomes at the final follow-up visit were assessed using the Visual Analog Scale (VAS) for back pain and Oswestry Disability Index (ODI) [8] at the final follow-up visit.

Radiological Assessment

All patients had pre-operative X-rays, CT, and magnetic resonance imaging (MRI) scans of the injured level. Pre-operative scores of TLICS and LSC were calculated using these pre-operative imaging and patient charts.

X-rays were taken immediately after the procedure, at 3rd, 6th, and 12th month. At the final follow-up visit, both X-ray and CT scans were taken. The pre-operative, immediate post-operative, and final follow-up X-ray images were analyzed using Cobb’s angle (CA), vertebral wedge angle (VWA), and vertebral body compression ratio (VBCR) in a neutral lateral view. Figure 1 demonstrates the calculation of all these radiological parameters. The pre-operative and final follow-up visit CT scans were assessed, to look for spinal canal remodeling, which was calculated using transverse spinal area (TSA) [9] (Fig. 2).

Fig. 1.

Schematic diagram of radiological parameters measured on a lateral neutral radiograph. Vertebral body compression ratio (VBCR): [2 × BH 2/(BH 1 + BH 3)] × 100. CA Cobb angle, VWA vertebral body wedge angle, BH body height

Fig. 2.

Transverse spinal area (TSA = a*b.) measurements were obtained at the level of maximal obstruction, which included the pedicles and lamina. This ensured that measurements were made of a bony canal and excluded the need to extrapolate canal boundaries involving soft tissues

Statistical Analysis

All statistical analyses were performed with R language and environment (version 4.3.1). Descriptive statistics were presented as mean ± standard deviation (SD) or mean with range. For categorical variables (gender and AIS), frequencies were assessed. For inferential analysis, a paired t test was used to evaluate differences (reduction or increase) in CA, VCR, VBCR, and TSA post-op and at the final follow-up within groups. The radiological variables (CA, VWA, VBCR, and TSA) between SIS and DIS groups were compared using a two-independent t test. Statistical significance was set at p < 0.05.

Results

Demographic Results

The mean age in the SIS group was 43.7 years (range 19–58 years) and the DIS group was 48.5 years (range 17–55 years). The mean follow-up period in both groups was 29.5 months (range 24–96 months). We operated within 72 h in most patients (84.4%) with the maximum interval between injury and surgery being 34 days. There was no statistically significant difference (p > 0.05) between the SIS and DIS groups in terms of age, gender, pre-operative or post-operative neurological status, mean time to surgery, operating time, intra-operative blood loss, and follow-up period (Table 1).

Table 1.

Patient demographic data

	SIS (n = 49)	DIS (n = 45)	p value
Age (years), mean (range)	43.7 (19–58)	48.5 (17–55)	> 0.05
Gender (M/F), n	36/13	30/15
Time to surgery (days), mean ± SD	1.9 ± 1.5	1.4 ± 2.1
Operating time (min), mean ± SD	105 ± 35	120 ± 45
Intraoperative blood loss (mL), mean ± SD	290 ± 75	310 ± 80
Hospital length of stay (days), mean ± SD	3.8 ± 1.5	4.1 ± 1.3
Follow-up (months), mean (range)	28.7 (24–72)	30.2 (24–96)
ASIA impairment scale (AIS), n
A	8	11
B	5	3
C	10	10
D	13	9
E	13	12

Radiological Outcome

Pre-operative, immediate post-operative, and final follow-up X-rays were evaluated for CA, VWA, and VBCR. The mean pre-operative CA was 29.8° and 26.8° in the SIS and DIS groups respectively. In the immediate post-operative period, it improved to 8.95° and 6.38°. At the final follow-up, CA in the SIS and DIS group was found to be 12.5° and 8.04°, with loss of 3.55° and 1.66° (p = 0.016), respectively, in SIS and DIS groups. The mean pre-operative VWA in SIS and DIS groups was found to be 37.5° and 35.8° which improved to 12.7° and 10.1° in the immediate post-operative period. At the final follow-up, VWA in the SIS and DIS group was 15° and 11.2° with loss of 2.3° and 1.1° (p = 0.004) in the SIS and DIS group, respectively. The mean pre-operative VBCR in the SIS and DIS groups was 57.3% and 60.7%, which corrected to 85.2% and 89.3%, respectively, in the immediate post-operative period. At the latest follow-up, a loss of 5.1% in the SIS group and 2.2% in DIS (p = 0.006) was noted. Figure 3 shows an illustrative case of L1 AO type A4 with TLICS 7 with LSC 8, who had undergone double-index screw fixation with maintained correction of radiological parameters throughout the follow-up period. The pre-operative values of TSA in the SIS and DIS group were 156.3 and 162.6 mm² which improved to 242.4 and 268.5 mm² at the latest follow-up (p = 0.215). The comparative results of radiological parameters between the SIS and DIS groups are depicted in Table 2.

Fig. 3.

Patient with L1 AO type A4 with TLICS 7 with LSC 8 underwent double-index screw fixation. Pre-op (A), immediate post-op (B), and at 36 month follow-up (C) showing the maintained correction of radiological parameters throughout the follow-up period

Table 2.

The comparative results of radiological parameters between the SIS and DIS groups

Radiological parameters	SIS (N = 49)	DIS (N = 45)	p value
	Mean (SD)	Mean (SD)
Cobb’s angle (CA)
Pre-op	29.8 (7.97)	26.8 (8.31)	0.581
Post-op	8.95 (3.50)	6.38 (2.96)	0.246
At latest follow-up	12.5 (5.30)	8.04 (3.96)	0.028
Difference between post-op and latest follow-up	3.55 (2.69)	1.66 (2.27)	0.016
Vertebral wedge angle (VWA)
Pre-op	37.5 (7.93)	35.8 (8.87)	0.680
Post-op	12.7 (3.34)	10.1 (3.45)	0.342
At latest follow-up	15.0 (3.10)	11.2 (3.45)	0.010
Difference between post-op and latest follow-up	2.3 (1.10)	1.1 (1.07)	0.004
Vertebral body compression ratio (VBCR)
Pre-op	57.3 (12.4)	60.7 (12.0)	0.846
Post-op	85.2 (14.0)	89.3 (7.00)	0.540
At latest follow-up	80.1 (13.6)	87.1 (7.19)	0.012
Difference between post-op and latest follow-up	5.1 (2.29)	2.2 (1.37)	0.006
TSA
Pre-op (mm2)	156.3	162.6	0.352
At latest follow-up (mm2)	242.4	268.5	0.215

The p-value in bold letters (p < 0.05) indicates statistically significant values

Clinical Outcome

At the final follow-up, in the SIS group, the mean VAS for back pain was 2.5 and the mean ODI was 15.4%. In the DIS group, the mean VAS for back pain was 2 and the mean ODI was 12.5% (Table 3).

Table 3.

Clinical outcomes at final follow-up

Clinical outcome at final follow-up	SIS	DIS	p value
Visual analog scale, mean (range)	2.5 (1–7)	2 (0–5)	0.146
Oswestry Disability Index, mean (range)	15.4% (10.3%–29.6%)	12.5% (10%–23.5%)	0.162

Discussion

We found that CA, VWA, and VBCR are well maintained throughout their follow-up period in the DIS group compared to the SIS group. Though the clinical outcomes were slightly better in the DIS group, it was insignificant. Further, there was no difference between both groups concerning spinal canal remodeling and neurological recovery. So far, there is only one study that compared SIS and DIS in thoracolumbar fractures and they concluded that both groups had similar radiological and clinical outcomes [10]. They evaluated the efficacy of unilateral versus bilateral pedicle fixation at the fracture level in treating thoracolumbar fractures with mild-to-moderate instability in 46 patients with LSC 5-7. The drawbacks of this study include–inclusion of only mild thoracolumbar burst fractures (LSC 5-7), short follow-up period (minimum 12 months), and assessed only CA and anterior body height compression in follow-up X-rays.

Treatment of unstable thoracolumbar burst fractures is always surgical stabilization. However, there is an area of wide research with the dilemma of whether to approach from anterior or posterior and if posterior, long- or short-segment posterior stabilization. Many spine surgeons now prefer posterior stabilization due to its familiarity, ease of approach, and fewer complications as compared to the anterior approach. Gurr and McAfee reported that long-segment posterior stabilization provided adequate stiffness to unstable burst fractures of the thoracolumbar region with effective correction of kyphosis [11]. The disadvantage of such long-segment posterior stabilization is extensive paraspinal muscle dissection and high blood loss, which might further damage the already injured spinal cord due to hypotension and oxygen deprivation caused by high blood loss. Other disadvantages include longer procedure time, and immobilizing longer motion segments, thus leading to further complications and increased hospital stay and costs.

The improvements in the instrumentation and advent of pedicle screws favored the short-segment posterior stabilization of unstable thoracolumbar burst fractures surmounting all the disadvantages of long-segment posterior stabilization. However, initially, short-segment posterior stabilization involves the insertion of pedicle screws at the adjacent levels to the fractured vertebra only. This resulted in post-operative implant failure and progressive kyphosis attributed to inadequate anterior column support [12, 13]. Further, the literature review showed that short-segment pedicle fixation alone (not including fractured vertebra in the construct) led to a 9–54% incidence of implant failure and re-kyphosis in the long-term, and 50% of the patients with implant failure had moderate-to-severe pain [14, 15]. This led to the addition of a pedicle screw in the fractured vertebra (index/intercalary/intermediate/fractured vertebra screw). The results of the addition of an index screw in the posterior stabilization have proven to be safer, feasible, and successful, evading the post-operative complications of traditional short-segment posterior stabilization without index screws [3, 16–19]. A systematic review and meta-analysis, comparing posterior short-segment fixation with four screw construct and adding intermediate screws at fracture level, concluded that adding intermediate screws at fracture level results in less post-operative pain, better radiological outcomes, and less implant failure at the cost of a longer operation time and higher blood loss [20]. However, to date, there is no clear consensus on whether single or double screws at the fractured vertebra are adequate.

The addition of index screws in the fractured vertebra has proven to be biomechanically superior as reported by Anekstein et al. [21] in the animal cadaveric model and Mahar et al. [22] in the human cadaveric model. Norton et al. conducted a biomechanical study of instrumentation in vertebral body cadaveric models that simulates an L1 axial load injury pattern [23] and found that a six-screw construct with screws in the fractured level is more rigid than a four-screw construct that skips the injured vertebral body. The mean stiffness in flexion–extension was increased by 31% (p < 0.03) with the addition of the two pedicle screws in L1. Deng et al. [19] analyzed the effect of the intermediate pedicle screws and their insertion depth on sagittal balance and functional outcomes of lumbar fracture and concluded that both four-pedicle screw fixation and six-pedicle screw fixation were effective in treating a lumbar fracture. They further concluded that six-pedicle screw fixation with short intermediate pedicle screws showed better radiographic and functional outcomes after surgery. However, using long and thick pedicle screws in fractured vertebrae resulted in better maintenance of posterior body height at the final follow-up [24]. We used longer screws in all cases extending up to the anterior column to increase the amount of bone contact with better fixation and provide anterior column support, thus decreasing the post-operative anterior column collapse. The benefits of index screws in fractured vertebrae are—it gives good three-point fixation, reduce the parallelogram effect, and avoid stretching of normal intervertebral disks, and it dispenses the stress of the pedicle screw connection.

McCormack et al. [6] reported that certain thoracolumbar fractures might need supplemental anterior reconstruction based on the severity of comminution, separation of fracture fragments, and amount of kyphosis correction. He suggested anterior reconstruction in patients with a score ≥ 7 to prevent posterior instrumentation failure. Earlier following his research, many authors supported his inference. However, Scholl et al. in 2008 reported that McCormack et al. load-sharing classification was not a predictor of posterior instrumentation failure [25]. This might be because McCormack used older-generation implants which do not withstand adequate stress. In the current era of evolving concepts of fracture fixation in spine, the relevance of LSC in the management of unstable burst fractures is questionable [26]. Our study adds to the existing literature by doing no supplemental anterior reconstruction even when the LSC score is ≥ 7 and there was no significant post-operative kyphosis which required surgical intervention.

The resorption of retropulsed fragments in the spinal canal was also noted thus improving the TSA at the fractured level at the final follow-up CT scan in both the groups. We attribute that this is due to indirect decompression achieved by ligamentotaxis. The clinical outcome was also better in DIS group compared to SIS group, but it was found to be statistically insignificant. Neurological recovery was seen in both groups up to 2 grades in the AIS scale without any significance between the two groups. These findings are congruent with previous studies [27, 28]. Though theoretically there was increased intra-operative blood loss and operating time in the DIS group, it does not affect the overall outcome of the patient.

Strengths and Limitations

The strength of our study includes a higher sample size with no differences in demographic characteristics among both groups. This helps in reducing the bias and providing support for the accuracy of the results of this study. Notably, the limitations of our study include those inherent to its retrospective database design, such as missing data, coding errors, and lack of granularity. Second, a few patients had undergone decompressive laminectomy due to neurological deficit, and its association with radiological and clinical outcomes was not studied. Further, the effect of pedicle screw dimensions and rod contouring was not considered. Also, this study does not include pertinent information regarding the variations in device manufacturing. Finally, outcome analysis between specific fracture patterns was not done. Despite these limitations, our study has provided important insight regarding the short-segment stabilization using index screws for unstable thoracolumbar fractures which were lacking in the literature.

Conclusion

Fixation of unstable thoracolumbar fractures with short-segment posterior stabilization with double-index screws (DIS) in the fractured vertebra resulted in radiologically better maintenance of CA, VWA, and VBCR compared to single-index screw (SIS). Though the VAS and ODI were better in the DIS group, it was found to be statistically insignificant. There was no difference in spinal canal remodeling and neurological recovery in both groups.

Author Contributions

G.K and S.R—Conceptualization, writing—original draft, and visualization. S.R, S.K and N.D.R— Supervision, methodology, and investigation. C.M and G.K—Formal analysis and data curation. V.T and B.R.J—Resources and validation. C.M and N.D.R—Writing—review, editing, and software. V.T and G.K—Visualization, and writing—original draft. B.R.J and S.K—Project administration and conceptualization.

Funding

No funding was acquired in support of this work.

Data Availability

All data generated or analyzed during this study are included in this published article.

Data Availability Statement

All collected data are available for this study. Data will be provided upon request.

Code Availability

All statistical analyses were performed with R studio version 4.3.1.

Declarations

Conflict of interest

The authors have no conflicts of interest to declare that are relevant to the content of this article.

Ethical Approval

Ethics approval was obtained by our institution’s Institutional Review Board (IRB No. PHPL/ACAD-2023-2/11).

Informed Consent

All authors have read the final version of the manuscript and agreed to its publication.

Consent to Participate

Consent to participate was obtained from all the patients.

Consent for Publication

Consent to publish was obtained from all the patients.