Modified Wiltse approach versus conventional posterior approach for treating single-segment spinal tuberculosis and infections: a retrospective study of 102 cases in Guangxi Province, China

Abu Moro; ShiAn Liao; Yue Fan; Yihan Wang; Wenfei Gu; Xinli Zhan; JinSong Yang

doi:10.31616/asj.2024.0440

Moro, Liao, Fan, Wang, Gu, Zhan, and Yang: Modified Wiltse approach versus conventional posterior approach for treating single-segment spinal tuberculosis and infections: a retrospective study of 102 cases in Guangxi Province, China

Clinical Study

Asian Spine Journal 2025;asj.2024.0440.

Online first: May 30, 2025

DOI: https://doi.org/10.31616/asj.2024.0440

Modified Wiltse approach versus conventional posterior approach for treating single-segment spinal tuberculosis and infections: a retrospective study of 102 cases in Guangxi Province, China

Abu Moro¹

, ShiAn Liao¹

, Yue Fan²

, Yihan Wang¹

, Wenfei Gu¹

, Xinli Zhan^1,^*

, JinSong Yang^1,^*

¹Department of Spine and Osteopathic Surgery, The First Affiliated Hospital of Guangxi Medical University, Nanning, China

²Department of Spine and Osteopathic Surgery, First People’s Hospital of QinZhou, QinZhou, China

Corresponding author: Xinli Zhan, Department of Spine and Osteopathic Surgery, The First Affiliated Hospital of Guangxi Medical University, Guangxi Province, Nanning City, P.R. China,
Tel: +86-0771-5350189, Fax: +86-771-5350031, E-mail: zhanxinli@stu.gxmu.edu.cn

Co-corresponding author: Jinsong Yang, Department of Spine and Osteopathic Surgery, The First Affiliated Hospital of Guangxi Medical University, Guangxi Province, Nanning City, P.R. China,
Tel: +86-0771-5350189, Fax: +86-771-5350031, E-mail: yang-jinsong@sr.gxmu.edu.cn

Received October 21, 2024 Revised January 31, 2025 Accepted February 3, 2025

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Study Design

Retrospective study.

Purpose

To compare the modified Wiltse with the conventional posterior approach for single-segment spinal tuberculosis/infections.

Overview of Literature

Spinal infections (such as spondylitis, discitis, and spondylodiscitis) are therapeutically challenging owing to their pathophysiological complexities and delayed diagnosis, particularly in developing countries. Despite treatment advancements, challenges persist, which include multidrug resistance and surgery-related complications.

Methods

A retrospective analysis of 102 patients who underwent single-segment spinal tuberculosis surgery was conducted to compare patients who underwent surgery using the conventional posterior with those using the modified Wiltse approach. The key parameters analyzed included surgical duration, blood loss, complication rates, intensive care unit stay, deformity correction, and neurological improvement.

Results

The modified Wiltse group exhibited lower intraoperative hemorrhage (184.72 mL vs. 365.91 mL, p<0.05) without significant differences in surgery duration and other tested outcomes. Both groups demonstrated significant improvement in thoracic kyphosis over time (p<0.01) and between the groups, with pairwise comparisons showing significant improvement from preoperative to postoperative and from preoperative to follow-up (p<0.01) but not from postoperative to follow-up, demonstrating postoperative kyphotic stability. Significant improvements in lumbar lordosis (p=0.047) and within the groups (p=0.002) were observed over time; however, pairwise comparisons did not reveal significant differences (p>0.05), demonstrating surgical intervention that focused on lordotic stability rather than correction in both groups. Significant neurological improvements were also observed, with the modified Wiltse group demonstrating superior recovery to the conventional posterior group from baseline.

Conclusions

The modified Wiltse approach offers a promising alternative to conventional techniques by reducing intraoperative blood loss while maintaining efficacy in deformity correction and neurological recovery. It is a viable option for patients who require reduced invasiveness and blood loss.

Keywords: Spinal tuberculosis; Spondylodiscitis; Surgical procedures; Muscle-sparing approac

Introduction

Spinal infections involve the pathological involvement of various spinal components, including vertebrae (spondylitis), intervertebral disks (discitis), and a combination of vertebrae and disks (spondylodiscitis), and cause epidural abscesses and septic facet joints, with an estimated annual prevalence of 2.2 per 100,000 people [1]. Among these conditions, spondylodiscitis is the most common, whereas spinal tuberculosis (TB, Pott’s disease) is particularly prevalent in developing countries [2]. Spinal TB is the oldest recorded form of spinal infection, with evidence dating back to 3400 BC [3]. It is one of the most life-threatening forms of spinal infection, leading to extensive bone destruction, deformities, paraplegia, and significant deterioration in a patient’s quality of life [4]. Despite advancements in medical and surgical interventions, treating spinal TB remains challenging because of factors such as multidrug resistance (MDR), risk of reinfection, and surgical complications [5].

Several factors contribute to the difficulty in eradicating spinal TB. The increased global life expectancy and the increasing prevalence of immunocompromizing conditions, such as obesity, diabetes, pulmonary TB, and human immunodeficiency virus/acquired immunodeficiency syndrome, particularly in developing countries, exacerbates the difficulty of managing this condition [6]. Furthermore, delayed diagnosis, nonspecific symptoms, and poor penetration of drugs into vertebral tissues further complicate treatment efforts [7]. Spinal TB most commonly affects the anterior vertebral column, with 98% of documented cases involving this region [8]. A study suggested that the slow-flowing vascular plexuses in the paradiskal regions of the spongy metaphyseal bones of the anterior column contribute to this localization [9].

The primary treatment for spinal TB is antitubercular chemotherapy, which typically consists of a combination of first-line anti-TB drugs, including isoniazid, rifampicin, pyrazinamide, and ethambutol [10]. In cases with drug resistance or treatment failure, second-line agents such as fluoroquinolones, aminoglycosides, or newer agents such as bedaquiline and delamanid may be required [11,12]. Surgical interventions are often necessary and typically involve extensive debridement, spinal canal decompression, deformity correction, bone graft fusion, and internal fixation through anterior, posterior, or combined anterior–posterior approaches [13]. Owing to the involvement of the anterior column in most cases, surgeons frequently opt for anterior or combined approaches to ensure thorough infection removal and sufficient stabilization. However, these conventional techniques are associated with numerous complications, such as spinal instability, neurologic injury, cerebrospinal fluid leakage, and site-specific risks, particularly with anterior approaches [14,15]. In addition, reinfection and MDR further complicate treatment success, often resulting in treatment failure, multiple cycles of antitubercular chemotherapy, or reoperations, which place significant physical and financial burdens on the patients [16,17].

To address the challenges in treating spinal infections, the demand for surgical approaches that not only achieve effective outcomes but also minimize complications and maintain spinal stability is increasing. The modified Wiltse approach, first introduced in 1968 and refined by Wiltse and Spencer [18] in 1988, presents a viable alternative. This approach utilizes a paramedian incision and preserves the posterior osseous and ligamentous structures while allowing sufficient access for laminectomy, laminotomy, discectomy, and debridement. Sufficient access to the laminae, spinal canal, nerve roots, and intervertebral disks facilitates effective treatment. Importantly, the Wiltse approach, by minimizing the dissection of the paravertebral muscles and surrounding soft tissues, is associated with reduced postoperative complications and faster recovery times [19,20]. This balance in effectiveness and reduced morbidity makes it a valuable option in the surgical management of spinal infections.

This study compared the clinical efficacy and outcomes of the conventional posterior approach for single-segment spinal TB with the modified Wiltse approach. The aim was to evaluate the potential benefits of the Wiltse approach in treating single-segment spinal TB/infections, particularly in reducing complications and improving patient outcomes.

Materials and Methods

Ethical statement

This study was conducted in accordance with the ethical standards of the Declaration of Helsinki and was approved by the Institutional Review Board (IRB) and the ethical committee of The First Affiliated Hospital of Guangxi Medical University (approved IRB no., 2025-E0371). Informed consent was obtained from all participants before their inclusion in the study. To maintain patient confidentiality throughout the study, all data analyzed in the study were anonymized.

Patient demography and data collection method

A 1:2 propensity score matching was performed on a total of 152 patients with preoperatively diagnosed or radiographically suspected single-segment spinal TB/infection (Table 1, Fig. 3A). Among them, 114 underwent surgery through the conventional posterior approach, whereas 38 underwent surgery using the modified Wiltse approach. After the propensity score matching, the 66 cases from the conventional posterior thoracic/thoracolumbar/lumbar interbody fusion group were matched with the 36 cases from the modified Wiltse group. These were then subsequently categorized into the Wiltse group (WG) and the posterior group (PG), respectively.

In this study, all 102 cases were postoperatively confirmed by pathological studies to be caused by Mycobacterium tuberculosis. The inclusion criteria were as follows: (1) patients with confirmed or suspected spinal infection based on preoperative radiographic scans, later confirmed through postoperative pathology, in whom conservative management failed or who meet the surgical indications for urgent surgical intervention; (2) patients who underwent surgery at our institution between January 2018 and March 2024; and (3) patients with complete clinical data who agreed to participate in the study. Conservative management failure occurs when the patient exhibits (1) worsening clinical symptoms, (2) ongoing radiographic evidence of infectious foci, and (3) elevated biomarkers (erythrocyte sedimentation rate and C-reactive protein) despite completing 4–6 months of antitubercular chemotherapy.

Neurological deficits were classified according to the Frankel grading system, depending on the degree of sensory and motor functions below the lesion area: grade A, complete neurological deficit; grade B, preservation of some sensations without motor functions; grade C, sparing of sensation but no functional motor functions; grade D, preserved functional motor functions with sparing of sensation; and grade E, normal sensory and motor functions [21].

Modified Wiltse approach

Following standard orthopedic skin preparation, a midline incision is made to establish the initial surgical access. Bilateral paravertebral incisions, approximately 3–4 cm (roughly two finger breadths) lateral to the spinous processes of the affected vertebrae, are then made along the latissimus dorsi muscle. Then, blunt dissection is performed along the natural cleavage plane between the iliocostalis and longissimus muscles to expose the multifidus muscle, which is then carefully dissected to reveal the facet joints (Fig. 1A). To facilitate adequate exposure, handheld navy retractors or Gelpi/Meyerding retractors are used. The laminae and facet joints of the affected vertebrae are systematically denuded using a pituitary rongeur. Depending on the case, a hemilaminectomy or a laminotomy+lateral facetectomy is performed (Fig. 1B, C).

Similar to the conventional transforaminal lumbar interbody fusion (TLIF) approach, hemilaminectomy provides direct access to the spinal cord, nerve roots, and the affected vertebrae or disks for thorough debridement (Fig. 1D). In this study, laminotomy+lateral facetectomy combined with extensive discectomy was performed in cases of limited spinal infections where the spinal canal was not involved (Fig. 1B, D). The aim is to preserve spinal stability while preventing contamination and infection spread to the spinal canal, subdural space, or central nervous system. Conversely, hemilaminectomy is reserved for cases where access to the spinal canal is necessary because of the tuberculous involvement of the canal (Figs. 1D, 2A–C).

Following hemilaminectomy, appropriately sized titanium mesh can be inserted after thorough debridement and bone grafting (Fig. 2B, C). For cases involving laminotomy, bone grafting (or cage placement) is performed as needed (Fig. 2D–I). If paraspinal abscesses are present in the psoas or quadratus muscles, the modified Wiltse approach facilitates easier access for drainage by needle aspiration or blunt dissection (Fig. 1E, Supplement 1).

Following meticulous debridement of the infectious foci, the area is irrigated with hydrogen peroxide and then copious saline rinsing. Note that hydrogen peroxide should not be used in cases of inadvertent dural tears. Bone grafting is performed using autograft and/or allograft material, presoaked in powdered streptomycin. The procedure is then completed with routine fluoroscopy-guided internal fixation.

Indications and contraindications

Indications for conventional posterior and modified Wiltse approaches

Early- to intermediate-stage spinal TB/infections with minimal vertebral body destruction but without significant lordosis/kyphotic deformities.

Spinal canal stenosis: Limited spinal canal stenosis with ≤30% encroachment or cases where the infection foci compress only the nerve roots without significant compromise of the spinal canal may benefit from the modified Wiltse approach. For cases with 30%≤ spinal canal stenosis <50%, conventional posterior procedures with total laminectomy may be more appropriate.

Patient-specific factors: Patients whose conditions could present challenges for instrumented fusion, such as obesity including those who may require a combined anterior–posterior approach could benefit from the modified Wiltse approach because of its minimally invasive nature and ease of instrumentation properties.

Localized paraspinal abscess: Patients with a localized paraspinal abscess in the psoas major/minor muscles without significant anterior or middle column involvement may undergo surgery either through the conventional posterior or the modified Wiltse approach. The modified Wiltse approach could be preferred owing to its limited disruption of spinal and musculoskeletal stability.

Risk mitigation: Patients with single-segment spinal TB who meet the above criteria, including those with a history of prior surgeries using either approach, could benefit from the modified Wiltse or conventional posterior approach to avoid re-exploration of previously accessed areas with dense scar tissues. Patients with comorbidities, or those at high risk of significant hemorrhage who require immediate surgical intervention, could benefit from the modified Wiltse approach, which has a lower risk of hemorrhage and complications.

Contraindications for conventional posterior and modified Wiltse approaches

Extensive spinal instability and severe deformities: Extensive disruption of spinal stability, such as infection foci involving the anterior and middle columns, or cases with significant vertebral body destruction with pronounced lordotic/kyphotic deformities, may exceed the capabilities of both the conventional posterior or modified Wiltse approach. In such cases, to achieve optimal debridement and deformity correction, an anterior or combined anterior–posterior approach (anterior debridement+modified Wiltse/conventional posterior approach) will be more appropriate.

Cervicothoracic junction spinal TB: Given the unique anatomical challenges at the cervicothoracic junction, implementing the modified Wiltse approach will be challenging. Such patients are more likely to benefit from an anterior or combined approach.

Statistical analysis

Statistical analysis was performed using IBM SPSS Statistics for Windows ver. 27.0 (IBM Corp., Armonk, NY, USA), GraphPad Prism ver. 10.2.3 (GraphPad Software, Boston, MA, USA; www.graphpad.com), and RStudio ver. 2024.04.2+764 (Posit Team, RStudio: Integrated Development Environment for R. Posit Software; PBC, Boston, MA, USA; http://www.posit.co). Cumulative link mixed models (CLMMs) in RStudio were used to analyze the changes in Frankel grades over time. Repeated-measures analysis of variance (RMANOVA) was used to analyze temporal changes in kyphosis and lordosis over the reviewed periods.

Results

Patient demographics

The general demographic data of the reviewed cases are shown in Table 1. Of the 102 patients enrolled in this study, 57 were male, and 45 were female. The average age of all the patients reviewed was 53.12±15.27 years. Among the patients, 53 had thoracic lesions, 43 had thoracolumbar lesions, and six had lumbosacral lesions (Table 1, Fig. 3B).

Intra- and postoperative outcomes

The WG had a shorter average surgery duration (133.17±83.06 minutes) than the PG (138.32±62.60 minutes), although the difference was not significant (Table 2). The WG exhibited significantly smaller intraoperative hemorrhage (184.72±54.64 mL) than the PG (365.91±336.75 mL, p<0.05). No significant differences in intraoperative antibiotic doses, length of stay (LOS), bone fusion time, follow-up duration, or hospitalization costs were observed between the two groups. In addition, although none of the WG patients required intensive care unit (ICU) observation, this difference was not significant.

The incidence of complications (intraoperative and postoperative) was comparable between the groups. In the PG, complications included one case of hemorrhagic anemia, two of wound dehiscence, one of delayed wound healing, and an incident of incision site infection. In the WG, one patient each had wound dehiscence, incision site infection, and delayed wound healing (Supplement 2).

Temporal changes in the Cobb angle

The RMANOVA results for the temporal changes in thoracic kyphosis and lumbar lordosis are shown in Tables 3 and 4. A significant difference in the thoracic kyphosis Cobb angles was observed over time: preoperative, postoperative, and follow-up (p<0.001). However, the pattern of temporal changes in the Cobb angle between the two groups (time×group) was not significant (p=0.163). The pairwise comparison revealed a significant difference in the overall average changes in the thoracic kyphosis Cobb angle between the two groups (p=0.028). In addition, significant improvements in Cobb angle changes were observed between preoperative and postoperative, as well as between preoperative and follow-up (both p<0.001); however, the changes between postoperative and follow-up were not significant (p=0.068) (Tables 3, 4).

The analysis of lumbar lordosis changes revealed that the Cobb angle changes over time were significant (p=0.047). The interaction between the temporal changes in the Cobb angle in the two groups was also significant (p=0.002). However, the pairwise comparison revealed that the overall average changes in the lumbar lordosis Cobb angle between the two groups were not significant (p=0.067). Likewise, no significant differences were observed between preoperative and postoperative, between postoperative and follow-up, and between preoperative and follow-up (all p>0.05) (Tables 3, 4).

Frankel grade neurological outcomes

The distribution of the Frankel grades across the time points reviewed in this study is shown in Fig. 3C. The CLMM analysis results revealed significant improvement in Frankel grades from preoperative to postoperative (log-odds, 4.3547; p<0.001) and from preoperative to follow-up (log-odds, 5.9986; p<0.001) in both groups. The improvements were more pronounced at follow-up, as indicated by the higher log-odds coefficient, when compared with postoperative. The WG had significantly lower baseline Frankel grades than the PG (log-odds, −0.2371; p<0.001) but exhibited greater improvement during both postoperative (log-odds, 0.5475; p<0.001) and follow-up (log-odds, 1.0785; p<0.001) than the PG (Table 5).

Discussion

The Wiltse approach, which utilizes an intermuscular avascular plane, has been reported to reduce operation time and excessive blood loss. A study demonstrated that it is an effective surgical option for single-segment thoracic TB in older patients [22]. In the present study, the estimated intraoperative hemorrhage was significantly lower in the WG. Although the surgery duration was lower in the WG than in the PG, the difference was not significant. The use of an anatomically avascular plane may explain why limited hemorrhage and reduced need for blood transfusions were observed in the WG in this study.

Antibiotic resistance poses a major challenge within the healthcare industry and remains a problem in critical environments with high usage, such as perioperative and intensive care settings. The perioperative use of antibiotics is essential for the successful prevention of infections, including surgical site infections, despite the ongoing risk of drug resistance [23]. Therefore, most institutions require strict adherence to institutional guidelines for rational antibiotic use to prevent the risk of MDR [24]. In our institution, intraoperative use is restricted to every 3 hours, which is roughly the duration of single-segment instrumentation and fusion. In this study, intraoperative antibiotic usage in the WG was comparable with that of the PG. These findings were consistent with the observed comparable surgery duration between the two groups.

The local application of streptomycin powder, which can induce bactericide by inhibiting ribosomal protein synthesis in M. tuberculosis and Gram-negative bacteria, was reported to help maintain bone viability, prevent further bone damage, and reduce surgical site infections [25]. From the results of this study, although the bone fusion time was not significantly different, the WG had the shortest average duration for bone fusion (7.00±1.79 months vs. 7.64±1.30 months). In this study, although the results confirmed the previously reported data, the sample size was relatively small; thus, more studies are warranted. Moreover, although reports have indicated the benefits of local streptomycin in preventing surgical site infections, three cases of surgical wound infection occur in the WG.

The enhanced recovery after surgery (ERAS) protocol has become an integral component of early postoperative recovery. The unique characteristics of the modified Wiltse approach align well with the intraoperative and postoperative aspects of ERAS protocols. The reduced vascular damage and muscle-sparing nature of the technique was reported to significantly improve postoperative recovery times [26]. In this study, the WG demonstrated a LOS comparable with that of the PG. Although the modified Wiltse approach resulted in a significant reduction in the need for postoperative ICU observation and had no reported intraoperative complications compared with the PGs, the difference was not significant. The use of hemilaminectomy or laminotomy with lateral facetectomy in the WG and the utilization of avascular planes, as opposed to the total laminectomy typically employed in conventional TLIF, could explain these observations. From our results, the modified Wiltse approach enhanced recovery and reduced the burden of postoperative care; however, the difference was not significant, which may have been due to the influence of our relatively small sample size.

Treatment for spinal infections is often challenging because of the heterogeneity of causative organisms and their complex pathophysiology. The high risk of reinfections, postoperative infections, and MDR often results in extended treatment cycles, increasing the treatment cost and placing a significant financial burden on the patients and their families. Therefore, any treatment that can reduce the overall cost is beneficial. From our results, the cost is comparable between the WG and PG, and this may be attributed to the high need for pedicle screws, which increased the operational costs, even though locally sourced screw-rod systems were used.

The modified Wiltse approach has been reported to achieve deformity correction and restoration of physiological curvature comparable with the traditional midline approach for fractures, scoliosis, and other degenerative conditions [19,20,22]. In this study, significant improvement in the thoracic kyphosis Cobb angles were observed from preoperative to postoperative and from preoperative to follow-up but not from postoperative to follow-up. These observations indicate that the abnormal preoperative kyphosis in both groups was successfully corrected during surgery and that the changes thereafter were stable. In contrast, although significant temporal changes were observed in the lumbar lordosis Cobb angles, pairwise comparison of the individual changes across specific time points (preoperative vs. postoperative; postoperative vs. follow-up; preoperative vs. follow-up) were not, indicating stable changes over time. These observations may reflect the anatomical, biomechanical, and clinical characteristics of thoracic and lumbar vertebrae diseases. The thoracic vertebrae are anatomically and biomechanically more stable than the lumbar vertebra because of the additional support from the ribs [27]. Furthermore, thoracic vertebral diseases usually have unspecific clinical manifestations and may mimic other diseases such as cervical or lumbar disorders [28], which can lead to initial diagnostic confusion and missed opportunities for early treatment. As a result, thoracic diseases have an increased risk of significant progression before definitive diagnosis and treatment can be established. Consequently, by the time patients with thoracic conditions, such as spinal TB/infection, seek surgical intervention, the disease may have caused significant bone destruction with deformities that would require corrective procedures. Conversely, lumbar TB/infections often immediately and significantly affect patients’ quality of life owing to its load-bearing and mobility demand. Thus, earlier reports prompting interventions focused on stabilization rather than deformity correction.

Improvement in neurological symptoms is one of the critical determinants of postoperative patient satisfaction in spine surgery, and studies have reported that patient satisfaction is closely related to the achievement of clinical improvements in pain, neurological symptoms, and disabilities [29,30]. In this study, both surgical approaches resulted in significant improvement in the Frankel neurological grades from preoperative to postoperative to follow-up, with patients reporting more pronounced improvements at follow-up. However, the WG demonstrated superior improvements in baseline neurological symptoms than the PG. These observations align well with the surgical techniques employed in both groups. Because the modified Wiltse approach allows for targeted debridement of the infectious foci with minimal disruption to the spinal cord and nerve roots, it reduces the risk of nerve or spinal injury, facilitating faster postoperative neurological recovery. In contrast, the conventional posterior approach often involves total laminectomy with zygapophyseal joint manipulations, which can pose higher risk of nerve and/or spinal cord injury, potentially delaying neurological recovery.

Several limitations should be mentioned. Drug-resistant cases were excluded from the study owing to their prolonged diagnostic path, thus limiting the generalization of the findings for all cases of single-segment spinal TB/infections. In addition, the relatively small sample size may have increased the risk of bias. Therefore, to validate and strengthen our findings, a larger multicenter study may be needed.

Conclusions

In this study, the modified Wiltse approach was compared with the conventional posterior approach for the surgical treatment of single-segment spinal TB and spinal infections. From our results, the modified Wiltse approach demonstrated significantly reduced intraoperative hemorrhage but comparable operation time, complication rates, bone fusion time, hospitalization cost, and ICU observation requirements. Both approaches achieved comparable thoracic and lumbar physiological curvature correction; however, the modified Wiltse approach demonstrated superior baseline neurological symptom improvement. These findings highlight that the modified Wiltse approach is a viable surgical option for the surgical treatment of single-segment spinal TB/infection, particularly in cases requiring minimal surgical invasiveness and blood loss, especially in cases of spinal TB/infection with limited spinal canal involvement.

Key Points

Modified Wiltse approach demonstrated significantly lower intraoperative blood loss (184.72 mL vs. 365.91 mL, p<0.05)
Both approaches demonstrated comparable surgical outcomes in terms of length of stay, surgery duration, intensive care unit observation, postoperative complications, and bone fusion time.
Both approaches demonstrated comparable postoperative improvements in thoracic kyphosis and lumbar lordosis over time.
The modified Wiltse approach resulted in significantly better improvements in Frankel neurological grades over time.

Notes

Conflict of Interest

No potential conflict of interest relevant to this article was reported.

Acknowledgments

The authors would like to express their sincerest gratitude to our colleagues at the Department of Spine and Osteopathic Surgery of the First Affiliated Hospital of Guangxi Medical University for their contribution to this project. We would also like to acknowledge Giovanna Fransesca Posumah for the commissioned artwork used in this study (Fig. 1B–D). This project would not have been successful without the collective effort of everyone involved. The data that support the findings of this study are available from the corresponding author upon reasonable request.

Funding

This work was supported by the Guangxi Natural Science Foundation (no., 2017GXNSFAA198127).

Author Contributions

Conceptualization: XLZ. Methodology: SAL, XLZ, JSY. Data acquisition: AM, YF, YHW, WFG. Data analysis: AM, WFG. Investigation: YHW. Supervision: XLZ. Obtaining funding: JSY. Project administration: JSY. Writing–original draft: AM. Critical revision: AM, SAL, YF, XLZ. Writing–review & editing: SAL, YF, XLZ, JSY. Final approval of the manuscript: all authors.

Supplementary Materials

Supplementary materials can be available from https://doi.org/10.31616/asj.2024.0440.

Supplement 1. Blunt drainage of a psoas abscess.

asj-2024-0440-Supplementary-1.pdf

Supplement 2. Anonymised patient data.

asj-2024-0440-Supplementary-2.pdf

Fig. 1

Modified Wiltse procedure. (A) Blunt dissection along the avascular plane revealing the facet joints, multifidus muscle, spinous process, and laminae. (B) Laminotomy approach showing the laminotomy and lateral facetectomy performed to access the infective foci behind the nerve root. (C) Hemilaminectomy approach showing: the nerve roots, ligamentum flavum and spinal cord are protected prior to discectomy and thorough debridement of infected lesion. (D) Debridement procedure for the laminotomy approach showing the protection of the nerve root and spinal cord during discectomy and removal of the infected lesion. (E) Needle aspiration of a psoas abscess. (B–D): Artwork by Giovanna Fransesca, commissioned by the author, and used with permission.

Fig. 2

Preoperative, postoperative, and follow-up images of the Wiltse group. (A–C) Preoperative (A), immediate postoperative (B), and follow-up (C) radiographic scans of a patient who underwent a hemilaminectomy procedure. (D–F) Preoperative (D), immediate postoperative (E), and follow-up (F) radiographic scans of a patient with lumbosacral tuberculosis who underwent laminotomy with lateral facetectomy. (G–I) Preoperative (G), postoperative (H), and follow-up (I) radiographic scans of a patient who underwent laminotomy with lateral facetectomy for lumbar tuberculosis.

Fig. 3

Propensity score matching, infectious foci distribution, and temporal changes in Frankel grades. (A) 1:2 propensity score matching (PSM). (B) Distribution of infectious foci in the two groups. (C) Preoperative (Preop), postoperative (Postop), and follow-up Frankel grade distribution. PG, posterior group; WG, Wiltse group.

Table 1

General patient data and propensity score matching

Variable	PG (n=66)	WG (n=36)	Total (n=102)	Statistic	SMD	p-value
Age (yr)	53.03±15.63	53.28±14.82	53.12±15.27	t=−0.078	0.017	0.938
Gender				χ²=0.218		0.641
Male	38 (57.58)	19 (52.78)	57 (55.88)		0.096
Female	28 (42.42)	17 (47.22)	45 (44.12)		−0.096
Preop Cobb’s				Z=−0.599	0.094	0.549
Thoracic kyphosis (°)	29.93±7.24	31.86±4.86	30.40±6.74
Lumbar lordosis (°)	31.57±6.20	30.77±7.05	31.19±6.56
Preop Frankel				Z=−0.469	−0.068	0.639
A	3	1	4
B	0	1	1
C	12	7	19
D	20	12	32
E	31	15	46
Lesion distribution
Thoracic (T4–T11/12)	40 (60.61)	13 (36.11)	53 (51.96)
Thoracolumbar (T12–L3/4)	24 (36.36)	19 (52.78)	43 (42.16)
Lumbosacral (L4/5–S1)	2 (3.03)	4 (11.11)	6 (5.88)

Values are presented as mean±standard deviation, number (%), or number, unless otherwise stated.

PG, posterior group; WG, Wiltse group; SMD, standardized mean difference.

Table 2

Hospitalization, preoperative, intraoperative, postoperative, and follow-up analysis

Variable	PG	WG	F	t-value	χ²	p-value
Length of stay (day)	13.15±7.117	12.97±7.821	0.462	0.117		0.907
Duration of surgery (min)	138.32±62.60	133.17±83.06	0.352	0.353		0.725
Estimated intraoperative blood loss (mL)	365.91±336.75	184.72±54.64	22.53	4.267		<0.001^*
Follow-up duration (mo)	15.39±10.27	15.44±10.56	0.028	−0.023		0.981
Bone fusion (mo)	7.64±1.30	7.00±1.79	6.720	1.882		0.065
Intraoperative antibiotic doses (every 3 hr)	1.11±0.310	1.14±0.351	0.934	−0.488		0.627
Cost of hospitalization (¥)	76,289.23±20,647.69	76,612.56±29,476.27	3.078	−0.065		0.949
Intensive care unit observation					1.686	0.194
No	63	36
Yes	3	0
Complications					0.018	1.0
No	61	33
Yes	5	3

Values are presented as mean±standard deviation or number, unless otherwise stated.

PG, posterior group; WG, Wiltse group.

^* p<0.05 (Statistically significant).

Table 3

Results of RMAOVA for temporal changes in Cobb’s angle

Source	Type III ∑²	F (df)	p-value	Partial η²	Observed power^a)
Thoracic kyphosis
Within-subject effects
Time	1,152.894	56.246 (1.072)	<0.001^*	0.524	1.00
Time×group	40.863	1.994 (1.072)	0.163	0.038	0.293
Between-subject effects
Group	353.314	5.138 (1)	0.028^*	0.092	0.604
Lumbar lordosis
Within-subject effects
Time	67.438	3.947 (1.132)	0.047^*	0.077	0.527
Time×group	170.881	10.0 (1.132)	0.002^*	0.175	0.901
Between-subject effects
Group	185.537	3.511 (1)	0.067	0.070	0.451

RMAOVA for Cobb’s angle (lordosis): Mauchly’s W=0.233, p<0.001, Greenhouse-Geisser η=0.566. RMAOVA for Cobb’s angle (kyphosis): Mauchly’s W=0.135, p<0.001, Greenhouse-Geisser η=0.536.

RMAOVA, repeated measures analysis of variance; df, degrees of freedom.

^* p<0.05 (Statistically significant results).

^a) Computed using α (alpha)=0.05

Table 4

Pairwise comparisons for time effects in Cobb’s angle

Pairwise comparison for time effects	MD	SE	p-value (Bonferroni)	95% CI
Pairwise comparison for time effects	MD	SE	p-value (Bonferroni)	Lower bound	Upper bound
Thoracic kyphosis
WG vs. PG	3.465	1.529	0.028^*	0.396	6.533
Preop vs. postop	−6.396	0.853	<0.001^*	−8.507	−4.284
Postop vs. follow-up	−0.461	0.196	0.068	−0.946	0.024
Preop vs. follow-up	−6.857	0.895	<0.001^*	−9.072	−4.641
Lumbar lordosis
WG vs. PG	2.251	1.201	0.067	−0.166	4.668
Preop vs. postop	−1.173	0.704	0.307	−2.921	0.575
Postop vs. follow-up	−0.434	0.209	0.130	−0.953	0.085
Preop vs. follow-up	−1.607	0.715	0.088	−3.381	0.168

MD, mean difference; SE, standard error; CI, confidence interval; WG, Wiltse group; PG, posterior group; Preop, preoperative; Postop, postoperative.

^* p<0.05 (Statistically significant results).

Table 5

Results of CLMM for Frankel neurological grades

Frankel grades CLMM predictor	Estimate	SE	Z-value	p-value
Time postop	4.3547	0.0039	1,115.93	<0.001^*
Time follow-up	5.9986	0.0045	1,326.19	<0.001^*
Group WG	−0.2371	0.0044	−53.45	<0.001^*
Time postop: group WG	0.5475	0.0039	140.29	<0.001^*
Time follow-up: group WG	1.0785	0.0045	238.44	<0.001^*

CLMM for Frankel group: model formula: Frankel~Time×group+(1|ID). Random effects: variance of random intercept=9.641, standard deviation=3.105. Fixed effects: log-odds coefficients (estimates) for time, group, and their interaction.

CLMM, cumulative link mixed model; SE, standard error; WG, Wiltse group; Postop, postoperative.

^* p<0.05 (Statistically significant results).

References

1. Tsantes AG, Papadopoulos DV, Vrioni G, et al. Spinal infections: an update. Microorganisms 2020;8:476.

2. Nagashima H, Yamane K, Nishi T, Nanjo Y, Teshima R. Recent trends in spinal infections: retrospective analysis of patients treated during the past 50 years. Int Orthop 2010;34:395–9.

3. Taylor GM, Murphy E, Hopkins R, Rutland P, Chistov Y. First report of Mycobacterium bovis DNA in human remains from the Iron Age. Microbiology (Reading) 2007;153(Pt 4): 1243–9.

4. Jin D, Qu D, Chen J, Zhang H. One-stage anterior interbody autografting and instrumentation in primary surgical management of thoracolumbar spinal tuberculosis. Eur Spine J 2004;13:114–21.

5. Kumar V, Neradi D, Sherry B, Gaurav A, Dhatt SS. Tuberculosis of the spine and drug resistance: a review article. Neurosurg Rev 2022;45:217–29.

6. Dunn RN, Ben Husien M. Spinal tuberculosis: review of current management. Bone Joint J 2018;100–B:425–31.

7. Wang P, Liao W, Cao G, Jiang Y, Rao J, Yang Y. Characteristics and management of spinal tuberculosis in tuberculosis endemic area of Guizhou province: a retrospective study of 597 patients in a teaching hospital. Biomed Res Int 2020;2020:1468457.

8. Jain AK, Dhammi IK. Tuberculosis of the spine: a review. Clin Orthop Relat Res 2007;460:39–49.

9. Garg RK, Somvanshi DS. Spinal tuberculosis: a review. J Spinal Cord Med 2011;34:440–54.

10. Jiang G, Qin W, Du X, et al. The distribution pattern of first-line anti-tuberculosis drug concentrations between the blood and the vertebral focus of spinal tuberculosis patients. J Clin Med 2022;11:5409.

11. Yadav M, Jain AK, Singhal R, Chadha M, Arora VK, Bhargava A. Incidence and patterns of drug resistance in patients with spinal tuberculosis: a prospective, single-center study from a tuberculosis-endemic country. Indian J Orthop 2023;57:1833–41.

12. Li Y, Sun F, Zhang W. Bedaquiline and delamanid in the treatment of multidrug-resistant tuberculosis: promising but challenging. Drug Dev Res 2019;80:98–105.

13. Bezer M, Kucukdurmaz F, Aydin N, Kocaoglu B, Guven O. Tuberculous spondylitis of the lumbosacral region: long-term follow-up of patients treated by chemotherapy, transpedicular drainage, posterior instrumentation, and fusion. J Spinal Disord Tech 2005;18:425–9.

14. Kato S. Complications of thoracic spine surgery: their avoidance and management. J Clin Neurosci 2020;81:12–7.

15. Ikard RW. Methods and complications of anterior exposure of the thoracic and lumbar spine. Arch Surg 2006;141:1025–34.

16. Mohan K, Rawall S, Pawar UM, et al. Drug resistance patterns in 111 cases of drug-resistant tuberculosis spine. Eur Spine J 2013;22(Suppl 4): 647–52.

17. Jain AK, Dhammi IK, Modi P, Kumar J, Sreenivasan R, Saini NS. Tuberculosis spine: therapeutically refractory disease. Indian J Orthop 2012;46:171–8.

18. Wiltse LL, Spencer CW. New uses and refinements of the paraspinal approach to the lumbar spine. Spine (Phila Pa 1976) 1988;13:696–706.

19. Chang J, Cao J, Huang Z, Wang B, Sui T, Cao X. Comparison of the modified Wiltse’s approach with spinal minimally invasive system and traditional approach for the therapy of thoracolumbar fracture. J Biomed Res 2020;34:379–86.

20. Kieser DC, Thakar C, Cunningham G, Vidakovic H, Hammer N, Nnadi C. The value of a modified Wiltse approach for deformity correction in neuromuscular scoliosis. Int J Spine Surg 2020;14:170–4.

21. Khorasanizadeh M, Yousefifard M, Eskian M, et al. Neurological recovery following traumatic spinal cord injury: a systematic review and meta-analysis. J Neurosurg Spine 2019;30:683–99.

22. Zeng H, Liao S, Liang Y, Moro A, Chen G, Qin H. Wiltse transforaminal thoracic interbody fusion approach for the treatment of single‐segment thoracic spinal tuberculosis in elderly patients with osteoporosis: a retrospective study of 20 cases. Exp Ther Med 2023;25:135.

23. Murphy CV, Reed EE, Herman DD, Magrum B, Beatty JJ, Stevenson KB. Antimicrobial stewardship in the ICU. Semin Respir Crit Care Med 2022;43:131–40.

24. Wang Q, Cao M, Tao H, et al. Evidence-based guideline for the prevention and management of perioperative infection. J Evid Based Med 2023;16:50–67.

25. Hikmawati D, Maulida HN, Putra AP, Budiatin AS, Syahrom A. Synthesis and characterization of nanohydroxyapatite-gelatin composite with streptomycin as antituberculosis injectable bone substitute. Int J Biomater 2019;2019:7179243.

26. Sun S, Chu H, Wu Z, et al. Clinical application of a modified Wiltse approach in middle and lower thoracic vertebrae: a case-control study of thoracic fracture patients. BMC Musculoskelet Disord 2024;25:701.

27. Brasiliense LB, Lazaro BC, Reyes PM, Dogan S, Theodore N, Crawford NR. Biomechanical contribution of the rib cage to thoracic stability. Spine (Phila Pa 1976) 2011;36:E1686–93.

28. Ando K, Imagama S, Kobayashi K, et al. Clinical features of thoracic myelopathy: a single-center study. J Am Acad Orthop Surg Glob Res Rev 2019;3:e18.00090.

29. Sivaganesan A, Khan I, Pennings JS, et al. Why are patients dissatisfied after spine surgery when improvements in disability and pain are clinically meaningful? Spine J 2020;20:1535–43.

30. Rampersaud YR, Canizares M, Perruccio AV, et al. Fulfillment of patient expectations after spine surgery is critical to patient satisfaction: a cohort study of spine surgery patients. Neurosurgery 2022;91:173–81.