A Comparison of Outcomes between the Wiltse Approach with Pedicle Screw Fixation and the Percutaneous Pedicle Screw Fixation for Multi‐Segmental Thoracolumbar Fractures

Abstract

Objective

Multi‐segmental thoracolumbar fracture (MSF) generally refers to fractures occurring in two or more segments of the thoracolumbar spine. With the development of minimally invasive concept, there is little research on its application in the field of MSF. The purpose of this study is to compare two minimally invasive surgical techniques and determine which one is more suitable for treating patients with neurologically intact MSF.

Methods

We retrospectively analyzed the clinical data of 49 MSF patients with intact nerves who were admitted from January 2017 to February 2019. Among them, 25 cases underwent percutaneous pedicle screw fixation (PPSF), and 24 cases underwent Wiltse approach pedicle screw fixation (WAPSF). The operation time, number of fixed segments, blood loss, length of incision, postoperative ambulation time, accuracy of pedicle screw placement, facet joint violation (FJV), number of C‐arm exposures, as well as pre‐ and postoperative visual analogue scale (VAS), Oswestry disability index (ODI), local Cobb's angle (LCA), and percentage of anterior vertebral body height (PAVBH) were recorded for both groups. Paired sample t‐test was used for intra‐group comparison before and after surgery while independent sample t‐test was used for inter‐group comparison.

Results

The differences in the number of fixed segments, intraoperative bleeding, postoperative bed time, accuracy rate of pedicle screw placement, VAS, and ODI between the two groups were not statistically significant (p > 0.05). However, the operative time and total surgical incision length were significantly shorter in the WAPSF group than in the PPSF group (p < 0.05), and the FJV was significantly higher in the PPSF group than in the WAPSF group (p < 0.05). Also, the PPSF group received more intraoperative fluoroscopy (p < 0.05). The result of LCA and PAVBH in the WAPSF group were significantly better than in the PPSF group (p < 0.05).

Conclusions

Both PPSF and WAPSF were found to be safe and effective in the treatment of MSF without neurological deficits through our study. However, considering radiation exposure, FJV, vertebral height restoration, correction of kyphosis, and learning curve, WAPSF may be a better choice for neurologically intact MSF.

In this article, a comparison was made between percutaneous pedicle screw fixation and Wiltse approach pedicle screw fixation for the treatment of neurologically intact multi‐segment thoracolumbar vertebral fractures. The optimal surgical approach was determined and provided as a reference for clinical treatment.

Introduction

Among spinal fractures, the unique anatomical structure and biomechanical characteristics of the thoracolumbar spine have led to a high percentage of fractures in this area. ¹ , ² , ³ , ⁴ Due to the increasing prevalence of high‐energy injuries and improved diagnostic techniques, there has been a progressive increase in the reported incidence of its multi‐segmental thoracolumbar fractures (MSF), which is mostly determined by bone structure and generally refers to fractures occurring in two or more thoracolumbar vertebral segments (other than the spinous and transverse processes), and was first discussed by Kosven. ⁵ These fractures involve a large number of vertebrae, are often multiple injuries, and are easily confused with single‐segment fractures of the thoracolumbar spine and other diseases, with a high rate of missed and misdiagnosis. Early surgical intervention is currently advocated to restore the height of the injured vertebra, maintain the biomechanical stability of the spine, and early postoperative ambulation and reduce postoperative complications. Its traditional open surgery reveals a large field of view, traumatic bleeding, and extensive stripping of paravertebral tissues, which makes the paravertebral muscles lose nerve atrophy causing late low back pain and affecting patient prognosis and rehabilitation. ⁶ , ⁷ , ⁸ , ⁹

With the continuous development of spinal surgery, minimally invasive procedures have gradually become mainstream, and the two techniques proposed by Wiltse ¹⁰ in 1968 and Foley ¹¹ in 2001 are the most commonly used minimally invasive techniques for the treatment of neurologically intact thoracolumbar fractures, with the main advantages of less trauma, less bleeding, less interference with the paravertebral muscles, lower complication rate, better prognosis, and better patient acceptance and satisfaction. Previous studies have also demonstrated the effectiveness and advantages of percutaneous pedicle screw fixation (PPSF) and Wiltse approach pedicle screw fixation (WAPSF), but they were applied in the field of minimally invasive surgery for single‐level thoracolumbar fractures with neurologically intact. ¹² , ¹³ , ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ It remains to be further studied whether these two minimally invasive techniques are feasible for treating MSF and what specific differences in effectiveness exist. Therefore, this study aims to explore two issues: (i) the clinical characteristics and surgical indications of MSF; (ii) the clinical efficacy of PPSF and WAPSF in treating neurologically intact MSF; and discuss which surgical approach is the optimal choice. This study provides a reference basis for the clinical treatment of neurologically intact MSF.

Materials and Methods

Inclusion and Exclusion Criteria

Retrospective study of 49 patients with MSF from January 2017 to February 2019, of whom 25 were treated with percutaneous pedicle screw fixation (PPSF) and 24 were treated with Wiltse approach pedicle screw fixation (WAPSF). Prior to surgery, the surgeon provides the patient and family with a detailed explanation of the advantages and disadvantages of conservative and surgical treatment. Ultimately, the patient was allowed to choose the treatment method independently. This retrospective study used the following inclusion criteria: (1) MSF confirmed by imaging; (2) MSF Sanyuan Tang Method ¹⁸ typing criteria of type IA (two adjacent segmental fractures), type IB (three or more adjacent segmental fractures), and type IIA (one normal segment with two fractures in between); (3) age 20–55 years; (4) fresh fractures with injury time <2 weeks; and (5) complete follow‐up data. Exclusion criteria: (1) pathological fracture caused by osteoporosis, tumor, tuberculosis; (2) single vertebral fracture; (3) with spinal nerve injury; (4) combined with infection; (5) combined with serious heart, lung, kidney‐brain, and other organ diseases that cannot tolerate surgery; and (6) with mental illness that cannot cooperate with treatment.

Clinical Outcomes

The number of fixed segments, length of incision, operative time, blood loss, number of C‐arm exposures, postoperative time to bed, and perioperative complications were compared between the two groups. Visual analogue scores (VAS), Oswestry disability index (ODI) scores, percentage of anterior vertebral body height (PAVBH, anterior vertebral body height/posterior vertebral body height), and local Cobb's angle (LCA) were assessed preoperatively, on postoperative day 3, and at the final follow‐up (PPSF: 11.92 ± 3.87 months; WAPSF: 12.36 ± 4.21 months), respectively. The accuracy of pedicle screw placement was based on postoperative computed tomography (CT) assessment of both positional accuracy and facet joint violation (FJV) according to the classification system described by Babu et al. ¹⁹ All data collection was done independently by the same research assistant.

Operative Techniques

PPSF Group

The patient is placed in the prone position after general anesthesia, and silicone pads are placed on the chest and hips to allow proper hyperextension of the spine and attempted postural repositioning prior to surgical correction, and the correct segment to be nailed is identified and located using C‐arm (Brivo OEC 850, GE, Fairfield, CT, USA) antero‐posterior position and lateral fluoroscopy, and the skin is marked at the point of projection of the vertebral arch, followed by skin disinfection and towel laying, and a 1.5 cm longitudinal incision is made along the marked point (adjust the length of the incision according to the patient's obesity). The tip of the puncture needle was tapped into the bony structure of the target segment and penetrated the cortex. If the puncture point and angle are correct on lateral fluoroscopy, the puncture needle is screwed into the posterior edge of the vertebral body in the lateral position, not exceeding the medial edge of the arch in orthogonal fluoroscopy, and a guide wire is placed. In the frontal and lateral views, ensure that the guide wire does not penetrate the medial wall of the pedicle or the anterior wall of the vertebral body, place the expansion sleeve step by step to protect the soft tissues, retain the outer expansion sleeve, and place a tap for tapping after the opening. Monitor the depth of the screw, when screwing into the screw still use the needle holder to hold the guide wire fixed to prevent deeper with the screw. Screwing into the vertebral body can remove the guide wire. The appropriate length of titanium rods can be pre‐bent according to the physiological curvature. After the installation of connecting rods, with the bracing device for reset, fluoroscopic reset is satisfactory after locking the nut. Break the tail of the nail, use 0.9% saline irrigation wound and built‐in materials. The incision was closed layer by layer with closed injection (Figure 1).

FIGURE 1.

Patient 1, female, 40 years old, suffered from L1 and L3 fractures due to a fall from height. The fracture type was classified as Sanyuan Tang type IIA. She underwent percutaneous pedicle screw fixation (PPSF) surgery for treatment. X‐ray of lumbar spine in anterior‐posterior position shows loss of height in L1 and L3 vertebral bodies (A); X‐ray of lumbar spine in lateral position shows collapse of the anterior edge height compared to adjacent vertebrae in L1 and L3 vertebral bodies (B); CT sagittal scan shows discontinuity of bone cortex in L1 and L3 vertebral bodies, with visible fracture fragments (C); MRI scan T1WI shows low signal in L1 and L3 vertebral bodies without nerve damage (D); MRI scan T2WI shows high signal in L1 and L3 vertebral bodies (E); intraoperatively, a guide needle was used to locate the puncture point of the spinous process from T12 to L4 (F); postoperative lumbar spine X‐ray anterior–posterior view shows satisfactory position of pedicle screws and appropriate length of titanium connecting rod (G); postoperative lumbar spine X‐ray lateral view shows good recovery of injured vertebral body height and satisfactory position of internal fixation (H).

WAPSF Group

The same preoperative procedure as used in the percutaneous approach was used for the WAPSF group. The C‐arm X‐ray marks the projection of the skin of the pedicle of the vertebra to be nailed, then the skin is disinfected and toweled, the lumbar dorsal fascia is incised longitudinally in the posterior median, and the articular and transverse processes are separated bluntly between the multifidus and longest muscles, and the lumbar spine is positioned using the “human” cresta or transverse process, and the thoracic spine is positioned using the upper edge of the root of the transverse process and the joint of the vertebral plate. A suitable pedicle screw is placed, and a pre‐curved connecting rod (consistent with the physiological curvature of the thoracolumbar spine at the fracture site) is inserted along the tunnel after satisfactory frontal and lateral fluoroscopy, and the rod is installed with the pedicle nail within the incision and slightly fixed. The wound and endophyte were flushed with 0.9% saline and the incision was closed layer by layer (Figure 2).

FIGURE 2.

Patient 2, male, 43 years old, suffered from T12‐L2 fractures caused by a traffic accident. The fracture type was classified as Sanyuan Tang type IB. He underwent Wiltse approach pedicle screw fixation (WAPSF) surgery for treatment. Thoracolumbar spine X‐ray in anterior–posterior position shows vertebral height collapse from T12 to L2 (A); lumbar spine X‐ray in lateral position shows loss of height at the anterior edge of T12‐L2 vertebrae compared to adjacent vertebrae (B); CT sagittal scan reveals discontinuity of bone cortex and visible fracture fragments in the vertebral attachments at T12‐L2 level (C); MRI scan T1WI shows low signal intensity in the T12‐L2 vertebral bodies with neurologically intact (D); MRI scan T2WI shows high signal in the T12‐L2 vertebral bodies (E); intraoperatively, the approach was through the multifidus and longissimus muscles gap, with internal fixation of pedicle screws from T11 to L3 vertebral bodies (F); postoperative lumbar anteroposterior x‐ray showed satisfactory position of pedicle screws and appropriate length of titanium connecting rod (G); postoperative lumbar lateral x‐ray showed good restoration of injured vertebral body height and satisfactory internal fixation position, with short pedicle screw fixation from T12 to L2 (H).

Postoperative Management

Patients in both groups were given antibiotics to prevent infection, acid suppressants to prevent stress ulcers and pain pumps, and individual braces. Encourage patients to engage in rehabilitation exercises with the support of orthopaedic brace 3 days after surgery, and adjust the amount of activity in real time according to their physical tolerance. Wear orthopaedic brace for 12 weeks after discharge. Follow up with patients at 3 days and 6 months postoperatively to inquire about their condition and review imaging results. Remove implants based on fracture healing status 1 year after surgery.

Statistical Analysis

The statistical analyzes were conducted using SPSS version 26 (SPSS Inc., Chicago, IL, USA), and continuous measures were expressed as mean ± standard deviation (⁻ x ± s); comparisons between time points within groups were made using repeated measures ANOVA analysis of variance, Shapiro–Wilk normality test was applied to the data of each observation index in the two groups, and paired t‐test was applied for comparison of data before and after surgery within the same group conforming to normal distribution, and for non‐normally distributed data, the Wilcoxon Sign Rank test was applied; the t‐test for two independent samples was applied to the comparison of data between two groups, and the Mann–Whitney U‐test was applied to the non‐normally distributed data; the categorical variables data were expressed as numbers or percentages, and the χ ² ‐test or Fisher test was used. A p‐value of less than 0.05 is considered as statistically significant.

Results

General Population Information

In the PPSF, there were 16 males and nine females, with a mean age of 36.92 ± 8.36 years (range 22–52 years) and with a mean body mass index (BMI) of 22.74 ± 2.83 kg/m². The final follow‐up time was 11.92 ± 3.87 months, ranging from 8 to 15 months. In the WAPSF, there were 19 males and five females, with a mean age of 37.88 ± 7.98 years (range 21–54 years) and with a mean body mass index (BMI) of 23.09 ± 2.73 kg/m². The final follow‐up time was 12.36 ± 4.21 months, ranging from 9 to 18 months. Causes of injury: 18 cases of fall from height injury, 24 cases of traffic accident injury, and seven cases of crush injury. Fracture type: 27 cases of type IA, 14 cases of type IB, and eight cases of type IIA. Distribution of injured vertebrae: T₈1, T₉2, T₁₀13, T₁₁21, T₁₂24, L₁27, L₂22, L₃3, L₄/L₅, one each, a total of 115 vertebrae involved. Combined non‐spinal fracture injuries: 12 cases of extremity fractures, three cases of thoracic injuries, three cases of abdominal organ injuries, two cases of cranial injuries, and one case of pelvic fracture. The demographic information of all the 49 patients is summarized in Table 1. There was no significant difference between the two groups in terms of age, gander, BMI, and fracture type (p > 0.05).

TABLE 1.

Description of the patient population.

Characteristic	PPSF group	WAPSF group	t/χ 2 values	p‐values
No. of cases	25	24	‐	‐
Age.mean ± SD, year	36.92 ± 8.36	37.88 ± 7.98	−0.409	0.685
Gander (Male/Female)	16/9	19/5	1.380	0.240
BMI	22.74 ± 2.83	23.09 ± 2.73	−0.439	0.663
Fracture type			0.676	0.796
IA	15	12
IB	6	8
IIA	4	4

Abbreviations: BMI, body mass index; PPSF, percutaneous pedicle screw fixation; SD, standard deviation; WAPSF, wiltse approach pedicle screw fixation.

Clinical Outcomes

The clinical results of the two groups are presented in Tables 2 and 3. As shown in Table 2, the operative time (99.67 ± 14.373 min vs. 116.84 ± 16.809 min) and length of incision (10.96 ± 2.319 cm vs. 16.48 ± 3.154 cm) was significantly shorter in the WAPSF group (p < 0.01). There was no significant difference in blood loss (137.38 ± 15.461 mL vs. 129.08 ± 20.408 mL), number of fixed segments (4.84 ± 0.688 vs. 4.63 ± 0.711), or early postoperative ambulation time (2.20 ± 0.707 days vs. 2.13 ± 0.741 days) (p > 0.05). The number of C‐arm exposures was significantly reduced in the WAPSF group (9.88 ± 2.455 vs. 20.16 ± 3.891, p < 0.01). The accuracy rate of pedicle screw placement was 94.12% (224/238) in the PPSF group and 95.08 (232/244) in the WAPSF group (p > 0.05). FJV was significantly higher in the PPSF group (19.33% vs. 8.20%, p < 0.01). There was no intraoperative vascular or nerve injury in all patients. In Table 3, there were no significant differences in VAS scores and ODI scores between the two groups at the preoperative, 3 day postoperative, and final follow‐up time points. However, the result of LCA and PAVBH in the WAPSF group were significantly better than those in the PPSF group (Figure 3). No infections, delayed wound healing, pneumonia, deep vein thrombosis, implantation failure, or other complications were reported during the follow‐up period.

TABLE 2.

Clinical findings.

Variable	PPSF group	WAPSF group	t/Z/χ 2 ‐values	p‐values
Operation time, mean ± SD, min	116.84 ± 16.809	99.67 ± 14.373	3.836	<0.01
Blood loss, mean ± SD, mL	129.08 ± 20.408	137.38 ± 15.461	−1.599	0.117
Length of incision, mean ± SD, cm	16.48 ± 3.154	10.96 ± 2.319	6.962	<0.01
No. of fixed segments	4.84 ± 0.688	4.63 ± 0.711	1.076	0.287
No. of C‐arm exposures	20.16 ± 3.891	9.88 ± 2.455	11.012	<0.01
Early postoperative ambulation time, mean ± SD, day	2.20 ± 0.707	2.13 ± 0.741	−0.615	0.539
Accuracy rate of pedicle screw placement (%)	94.12% (224/238)	95.08% (232/244)	0.220	0.639
Facet joint violation rate (%)	19.33% (46/238)	8.20% (20/244)	12.631	<0.01

Abbreviations: PPSF, percutaneous pedicle screw fixation; SD, standard deviation; WAPSF, wiltse approach pedicle screw fixation.

TABLE 3.

Comparison of VAS, ODI, LCA, and PAVBH between the two groups.

Variable		PPSF group	WAPSF group	t/Z values	p‐values
VAS	Preoperative	7.52 ± 0.714	7.67 ± 0.637	−0.758	0.453
	Postoperative	2.44 ± 0.821	2.71 ± 0.751	−0.991	0.322
	Final follow‐up	2.40 ± 0.500	2.25 ± 0.442	−1.108	0.268
ODI (%)	Preoperative	42.88 ± 4.136	43.13 ± 4.377	−0.201	0.841
	Postoperative	11.60 ± 1.633	12.00 ± 2.022	−0.763	0.449
	Final follow‐up	10.24 ± 1.562	10.46 ± 1.351	−0.522	0.604
LCA (°)	Preoperative	20.68 ± 3.660	21.63 ± 3.573	−0.914	0.365
	Postoperative	8.76 ± 1.128	7.71 ± 1.781	2.480	0.017*
	Final follow‐up	10.32 ± 1.464	9.33 ± 1.404	2.406	0.020*
PAVBH (%)	Preoperative	62.64 ± 5.656	61.58 ± 6.317	0.617	0.540
	Postoperative	87.12 ± 3.909	90.21 ± 5.082	−2.390	0.021*
	Final follow‐up	85.04 ± 3.446	88.08 ± 4.872	−2.533	0.015*

Abbreviations: LCA, local Cobb's angle; ODI, oswestry disability index (version 2) on a 0% to 100% scale; PAVBH, percentage of anterior vertebral body height; PPSF, percutaneous pedicle screw fixation; VAS, visual analog score; WAPSF, wiltse approach pedicle screw fixation.

FIGURE 3.

First, both groups showed significant improvement after surgery compared to before (p < 0.05), but there was no significant difference between the two groups after surgery (p > 0.05). Second, both groups showed significant improvement after Oswestry disability index (ODI) surgery compared to preoperative status (p < 0.05), but there was no significant difference between the two groups postoperatively (p > 0.05). Third, both groups showed a significant reduction in local Cobb's angle (LCA) after surgery (p < 0.05), and the postoperative comparison between the two groups indicated that Wiltse approach pedicle screw fixation (WAPSF) was significantly better than percutaneous pedicle screw fixation (PPSF) (p < 0.05). Fourth, both groups showed significant improvement in percentage of anterior vertebral body height (PAVBH) after surgery compared to before (p < 0.05). The comparison between the two groups postoperative indicated that WAPSF was significantly better than PPSF (p < 0.05).

Discussion

Clinical Features and Surgical Indications

In this study, it was found that thoracolumbar junction region (T10‐L2) fractures accounted for as high as 93.04% (107/115), with high‐energy injuries such as traffic accidents or falls from heights accounting for 85.71% (42/49) and combined complex injuries accounting for 42.86% (21/49), similar to previous reports. Upper vertebral body fractures are often more severe than lower vertebral body fractures due to the high‐energy nature of MSF injuries, which often involve non‐spinal injuries. Therefore, it is necessary to pay attention to comprehensive and careful examination in clinical diagnosis and avoid delayed treatment due to misdiagnosis or missed diagnosis. Due to the different characteristics and mechanisms of injury in MSF compared with single‐segment thoracolumbar spine fractures, personalized treatment strategies should be considered comprehensively when treating this type of fracture. ²⁰ , ²¹ , ²² , ²³

For the classification of MSF, Korres et al. ²⁴ referred to spinal fractures of more than two consecutive segments as adjacent and at least one normal segment between two fractured vertebrae as non‐adjacent, also known as multiple‐level noncontinuous spinal fractures (MNSF). Calenoff et al. ²⁵ divided the non‐contiguous type into three types, A, B, and C, according to the different locations of the primary injury and secondary loss: type A is a primary fracture at the C5‐C7 segment causing a secondary fracture at the T12 and lumbar levels; type B is a primary fracture at the T12 segment and lumbar level causing a secondary fracture at the cervical level; type C is a primary fracture at the T12‐L2 segment causing a secondary fracture at the L4‐L5 segment; and type C is a primary fracture at the T12‐L2 segment causing a secondary fracture at the L4‐L5 level. Henderson et al. ²⁶ divided the non‐adjacent types into five zones: C1 to C3, C4 to C7, T1 to T10, T12 to L2, and L3 to L5, and classified them into mild and severe injuries depending on the structure of the injury. There is a lack of unified standard for further classification, so this study uses the typing method proposed by Tang Sanyuan ¹⁸ on MSF in China, which covers the fracture types proposed by the aforementioned scholars and typing on the basis of excluding pathological fractures caused by tuberculosis and tumors, which is more detailed than the original classification, simple to use and more in line with clinical practice, and can guide the selection of specific surgical strategies.

Due to the powerful and multi‐level vertebral body involvement of MSF‐induced injuries, it can cause significant damage to spinal stability and result in additional kyphotic deformity for adjacent multiple‐segment compression fractures of type IA or IB. This may lead to neurological symptoms and should be considered an unstable fracture. Conservative treatment may result in spinal kyphosis and neurological symptoms in later stages, affecting appearance and social labor ability. Most MSF patients are young adult males who cannot tolerate the pain caused by fractures or back braces, nor can they accept a severe decline in quality of life due to long‐term bed rest. Surgical stabilization treatment is recommended for all unstable injury segments, especially when two injury sites are close together, so that patients can start early activity and reduce the occurrence of spinal kyphosis. ²⁷ , ²⁸ , ²⁹ Therefore, compared with single‐segment thoracolumbar fractures, the indications for MSF surgery should be appropriately relaxed. For surgical timing related to spinal cord injuries themselves, surgery should be performed as soon as possible after basic life support has been ensured for the patient to relieve pressure on the spinal cord nerves caused by fracture dislocation during surgery using strong internal fixation with screws or rods to restore biomechanical stability of spine as much as possible while avoiding missing the best opportunity for recovery of spinal cord nerves which will benefit early postoperative rehabilitation exercises leading toward improved quality of life. ³⁰ , ³¹ , ³²

Minimally Invasive Surgery (PPSF and WAPSF) for MSF

Once an MSF patient needs surgical treatment, regardless of whether there is spinal cord injury or not, posterior approach vertebral pedicle screw fixation and reduction has become a widely recognized method. ³³ , ³⁴ Currently, traditional open surgical approaches such as laminectomy for decompression are still the most common choice involving nerve injury. However, for patients with intact neurological function, these approaches have larger incisions and more severe surgical trauma, which invade paraspinal muscles and severely affect patients' quality of life in the later stage with significant back pain and limited mobility. ⁹ Kim et al. ³⁵ found that the degree of muscle atrophy of the lower back in percutaneous internal fixation was significantly higher than that in traditional open surgery. Minimally invasive treatment techniques can better protect the deepest muscle groups in the lumbar region (multifidus), which are dominated by sagittal plane rotation. In addition, their minimally invasive procedures have similar corrective effects to traditional posterior surgery, but also have low invasiveness, minimal trauma, low bleeding, good prognosis, high patient acceptance and satisfaction, and are increasingly being used in clinical practice. ⁹ , ¹² , ³⁶ , ³⁷ , ³⁸ Currently, there are two main minimally invasive treatment techniques for the treatment of neurologically intact thoracolumbar fractures: PPSF and WAPSF. Sheng et al. ¹³ suggested that during WAPSF surgery, the rod could be placed in the muscle gap between the longest dorsal muscle and the multifidus under direct vision, which would reduce the area of atrophy of the multifidus.

The percutaneous approach uses special instruments to insert screws after continuous expansion, while the Wiltse approach inserts screws while preserving the paraspinal muscles. Both approaches effectively reduced the iatrogenic injury of the paraspinal muscles, which also explained why there were no statistical differences between the two groups in blood loss, postoperative ambulation time, VAS, and ODI scores. This is because the two groups had minimally invasive surgery, less surgical trauma, and faster postoperative recovery. In the case of no difference in the number of fixed segments between the two groups, the operation time of the PPSF group was significantly longer than that of the WAPSF group, because the number of fluoroscopies during the percutaneous approach was significantly more than that of the Wiltse approach. Studies ¹⁴ , ¹⁵ , ³⁹ reported that percutaneous internal fixation relied on intraoperative fluoroscopy, and doctors and patients also received higher doses of radiation. Surgeons are advised to wear a variety of protective clothing, but heavy lead clothing not only increases the burden on the operator, but also affects the operation. This technique may have a steeper learning curve and requires extensive surgical experience. ¹² , ⁴⁰ , ⁴¹ During the operation, the Wiltse approach can be approached from the muscle space between the multifidus muscle and the longissimus muscle under direct vision to directly observe the anatomical structure of the spine and find the pedicle screw insertion point, which improves the accuracy of screw placement and reduces the number of times of fluoroscopy, and we think that this technique is quick to learn and easy to master. In addition, this approach can be completed with conventional spinal surgery instruments, while the percutaneous approach requires special instruments and equipment.

The accuracy of pedicle screw placement and FJV were evaluated by CT postoperatively. The results showed that the accuracy of pedicle screw placement in PPSF group and WAPSF group were 94.12% and 95.08%, respectively (p > 0.05). Although the results were not different, the WAPSF group seemed to perform better intraoperatively, possibly due to the placement of the pedicle screw under direct vision. However, this study found that the FJV rate in the PPSF group was significantly higher than that in the WAPSF group (19.33% vs. 8.20%, p < 0.01). Ohba et al. ⁴² reported that the incidence of FJV in percutaneous screw placement was 30.5%. Teles et al. ⁴³ reported that PPSF and traditional open surgery FJV were 28% and 12.3%. Zou et al. and Sheng et al. ¹³ , ¹⁶ compared the Mini‐Open Wiltse approach and the percutaneous approach and found that the FJV rates were 9.2% versus 22.4% and 5.9% versus 19.7%, respectively. This is similar to what our study found. This result may be because, unlike the pedicle screw placement under direct vision in the WAPSF group, the PPSF group needs to be determined by the tactile sensation of the puncture needle under repeated fluoroscopy guided by the C‐arm, which inevitably increases the incidence of FJV.

For the selection of MSF fixation segments, Cho et al. ⁴⁴ believed that when decompression is not required, it is recommended to minimize the number of fusion fixation segments in multiple thoracolumbar fractures, so as to limit the spinal activities and postoperative adjacent segments. The impact of complications is minimized. In this study, for 41 patients with MSF without nerve injury, a total of 41 cases of type IA and type IB fractures were fixed at the upper and lower adjacent segments of the injured vertebra, and the posterior long‐segment internal fixation system was used. For the other eight cases of non‐adjacent type II A fractures, since the injured vertebra was separated by a normal vertebra, four to five groups of pedicle screws were inserted during the operation, and the same treatment method as type I fracture was used for the long segment. In‐segment fixation system, the postoperative correction effect was satisfactory. During the follow‐up period, there was no internal fixation loosening or fracture, no kyphotic deformity of the spine and no obvious loss of vertebral body height. During the operation, according to the integrity of the pedicle, short pedicle screws were placed and fixed in the injured vertebra to enhance the stability of the spine to achieve correction and firm fixation. This study showed that LCA and PAVBH in both groups were significantly improved compared with preoperatively. Fan et al. ¹⁷ compared WAPSF and PPSF and found that the vertebral body height and Cobb angle in the WAPSF group were significantly better than those in the PPSF group, which is similar to the results of this study. This may be attributed to the fact that PPSF is affected by skin and subcutaneous fat extrusion during the reduction process, while WAPSF has a better reduction effect in the intermuscular space through stretching between levels.

Strengths and Limitations

This study has the following advantages: (i) a comprehensive description of the clinical characteristics of multi‐segmental thoracolumbar fractures; (ii) summary of surgical indications for MSF, which differs from the clinical management principles for single‐segment thoracolumbar fractures; (iii) a variety of clinical observation indicators were applied to evaluate the differences in clinical efficacy between PPSF and WAPSF in treating neurologically intact MSF, and the optimal surgical method was selected. Meanwhile, this study also had important limitations: (i) this study is a clinical study with a single center and small sample size, which may cause bias in the research results; (ii) there may be some bias in the collection of imaging data; (iii) this article does not study whether the two minimally invasive procedures are suitable for MSF type IIB (fracture involving two normal segments with two or more normal segments in the middle interval) and type IIC (fracture involving three or more normal segments with one or more normal segments in the middle interval). For the safety and feasibility of normal segments above, a large number of clinical studies are still needed to further explore its specific clinical efficacy.

Conclusions

MSF is more common in young adults and is often caused by high‐energy trauma, often accompanied by injuries to other parts of the body besides the spine, and is most common in the thoracolumbar junction region (T10‐L2). Compared to single‐segment thoracolumbar vertebral fractures, its surgical indications should be appropriately relaxed. Early surgery uses screw‐rod reduction fixation to restore spinal stability. Various clinical results have shown that PPSF and WAPSF have certain clinical advantages in treating neurologically intact MSF. Both surgical methods are safe and effective, but in some aspects such as radiation exposure, FJV, vertebral height reduction, kyphosis correction and learning curve, WAPSF is superior to PPSF. Therefore, it is recommended to use WAPSF for the treatment of neurologically intact MSF.

Author Contributions

Dingjun Hao designed the study and critically revised the manuscript. Yadong Zhang carried out the statistical analyses and drafted the manuscript. Wentao Wang contributed in designing the study, and in drafting and critically revising the manuscript. Yadong Zhang and Lulu Bai were responsible for the data collection and measurement of radiographic data. The author(s) read and approved the final manuscript.

Funding Information

This work was supported by the National Natural Science Foundation of China (grant number 81830077).

Conflict of Interest Statement

The authors declare no conflicts of interest.

Ethics Statement

This study was approved by the ethics committee of Xi'an Honghui Hospital, and the informed consent was waived by the ethics committee as it is a retrospective study. We confirm that all methods were performed in accordance with the relevant guidelines and regulations.

Data Availability Statement

The data which analyzed during the study are stored in our hospital and are available from the corresponding author on reasonable request.