Wild spine infections: Long-term outcomes and surgical strategies in refractory postoperative spinal infections

Abstract

Background:

Most postoperative spinal infections resolve with timely debridement and culture-directed antimicrobial therapy. A minority, however, progress to persistent infection with osseous destruction, mechanical instability, and repeated surgical failure. Evidence-based strategies for managing this high-risk salvage population remain limited. This study aimed to evaluate the long-term clinical, neurological, and radiological outcomes of patients undergoing definitive reconstructive surgery after multiple failed revision procedures for refractory postoperative spinal infection.

Methods:

We retrospectively reviewed patients who underwent two or more revision surgeries for refractory postoperative spinal infection with associated instability or deformity. Clinical outcomes were assessed with the visual analog scale (VAS), Oswestry disability index (ODI), and Frankel neurological grade; radiological outcomes focused on sagittal alignment. All patients underwent aggressive debridement followed by posterior, anterior, or combined reconstruction, with adjunctive antimicrobial therapy and hyperbaric oxygen. Given the limited sample size, statistical analyses were considered hypothesis-generating.

Results:

Thirteen patients (mean age, 47.7 years) with a mean follow-up of 64 months and a mean of 4.9 prior spinal surgeries were included. Mean VAS score decreased from 9.5 preoperatively to 1.7 at 1 year (p < 0.001), and mean ODI improved from 62.4 to 19.2 (p < 0.001). Neurological improvement occurred in 53.8% of patients, with complete recovery in 30.8%. Sagittal segmental alignment improved significantly after surgery, but the magnitude of radiological correction did not correlate with pain or functional outcomes.

Conclusion:

In refractory postoperative spinal infection requiring salvage reconstruction, restoration of mechanical stability, rather than radiological correction, is the principal determinant of clinical and neurological recovery. Early recognition and timely definitive reconstruction may improve long-term outcomes.

INTRODUCTION

Postoperative spinal infection is among the most challenging complications of spine surgery. Most cases resolve with prompt debridement and culture-directed antibiotic therapy, but a small subset follows an aggressive, treatment-resistant course marked by neurological impairment, mechanical instability, progressive vertebral destruction, and repeated surgical failure.[8,22]

We use the term “wild spinal infection” in this study as a descriptive label rather than a formal diagnostic entity, defined by the coexistence of three features: Failure of two or more revision procedures, ongoing infection with extensive osseous destruction or instability, and an inadequate clinical or biological response to prolonged antimicrobial treatment. The term is intended only to identify a high-risk salvage population in whom conventional, stepwise treatment strategies are often insufficient.[3,8,17,22]

This study characterizes the clinical features, microbiological spectrum, surgical management, and long-term outcomes of this complex subgroup, with the goal of informing salvage strategies for refractory postoperative spinal infection.

MATERIALS AND METHODS

Study design and patient selection

This retrospective study was conducted at a single tertiary referral center. Between January 2015 and December 2025, 282 patients underwent surgical treatment for postoperative spinal infection at our institution. Of these, 227 resolved with a single debridement procedure and did not meet the threshold of two or more revision surgeries; 23 had undergone their index surgery at an outside institution with incomplete perioperative or microbiological records; eight had incomplete clinical or radiological data precluding accurate classification; and 11 declined further treatment or were lost to follow-up before definitive management could be characterized. The remaining 13 consecutive patients met criteria for wild spinal infection, as defined below, and constitute the study cohort. We reviewed patients who underwent two or more revision procedures for refractory postoperative spinal infection, here classified as “wild spinal infection” and characterized by established pseudarthrosis, mechanical instability, severe vertebral destruction, or spinal deformity. Inclusion criteria were a history of prior spinal surgery, diagnosis of postoperative spinal infection established using a composite of clinical, laboratory, and radiological criteria, consistent with previously reported diagnostic frameworks for spinal implant and postoperative spinal infection.[3,5,7] Diagnosis required at least two of the following four criteria: (a) Clinical signs of infection, including wound drainage, dehiscence, local erythema, or fever >38°C; (b) elevated inflammatory markers (C-reactive protein [CRP] and/or erythrocyte sedimentation rate [ESR]); (c) radiological findings on magnetic resonance imaging (MRI) or computed tomography (CT) consistent with infection, including epidural or paraspinal collection, endplate destruction, or abnormal marrow enhancement; and (d) microbiological confirmation by intraoperative aerobic, anaerobic, or fungal culture. Microbiological confirmation was supportive but not mandatory for inclusion, given the well-documented occurrence of culture-negative disease in postoperative spinal infection.[3] The need for multiple revision surgeries for persistent or recurrent disease was also required for inclusion; consistent with this threshold of two or more revision procedures, the number of prior operations in the study cohort ranged from 2 to 9 [Table 1]. Patients successfully treated with a single debridement for early, superficial surgical site infection, defined according to Centers for Disease Control and Prevention and National Healthcare Safety Network criteria as infection involving only skin and subcutaneous tissue at the incision, occurring within 30 days of surgery, without involvement of deep soft tissue, the disc space, or the vertebral body, were excluded from the study. The aim of this study was not to compare refractory infections with uncomplicated postoperative infections, but to characterize a highly selected salvage population after failure of conventional treatment strategies.

Table 1:

Clinical characteristics, microbiological findings, and surgical management of patients with refractory postoperative spinal infection.

Clinical and radiological assessment

We recorded the number of prior surgical procedures, comorbidities, patient demographics, and index-procedure parameters. Pain severity and functional status were assessed prospectively with the visual analog scale (VAS) and Oswestry disability index (ODI) at 3 time points: Preoperatively, on postoperative day 10, and at 1-year follow-up. Neurological status was graded with the Frankel scale before surgery and at 1 year.

Radiological evaluation included anteroposterior and lateral radiographs, CT, and contrast-enhanced MRI to assess neural compression, spinal instability, and osseous damage. Sagittal segmental alignment (kyphosis/lordosis angle) was measured before and after surgery to quantify the effect of surgical correction. Sagittal segmental alignment was defined as the Cobb angle between the endplates of the vertebrae immediately adjacent to the operated segment(s), measured on standing or supine lateral radiographs. Anterolisthesis or retrolisthesis was not systematically quantified in this cohort.

Surgical strategy and multidisciplinary management

Surgical approach was guided by the location and extent of infection and the severity of mechanical or neurological compromise. Management decisions in cases of septic discitis or epidural abscess formation were individualized by the multidisciplinary team according to the extent of neural compression, the degree of mechanical instability, and the patient’s clinical status, rather than following a fixed institutional algorithm. Debridement continued until healthy, bleeding bone was reached. Spinal stability and sagittal alignment were restored using anterior, posterior, or combined instrumented fusion; instrumentation was placed even in contaminated fields when mechanical stability required it. Intraoperative bone and disc specimens were collected for aerobic, anaerobic, and fungal culture.

All patients received culture-directed antibiotic therapy postoperatively under infectious disease consultation, together with a standardized adjunctive hyperbaric oxygen regimen of 21 sessions of 100% oxygen at 2.4 atmospheres absolute for 90 min each.

Statistical analysis

Data were analyzed with IBM Statistical Package for the Social Sciences Statistics (version 26.0, IBM Corp., Armonk, NY). Continuous variables are expressed as mean ± standard deviation. Because of the small sample size and non-normal distribution, non-parametric tests were used. Longitudinal changes in VAS and ODI scores were analyzed with the Friedman test, with post hoc Wilcoxon signed-rank tests and Bonferroni correction for pairwise comparisons. Preoperative and postoperative neurological grades were compared with the Stuart-Maxwell test, and correlations between radiological correction and clinical outcome were assessed with Spearman’s rank correlation. p < 0.05 was considered statistically significant.

RESULTS

Patient profile

Thirteen patients with wild spinal infection were included [Table 1]: Nine men and four women, with a mean age of 47.7 ± 12.4 years. Diabetes mellitus was the most common comorbidity (n = 5), followed by Parkinson disease (n = 2), osteoporosis (n = 1), and intravenous drug use (n = 1). The index surgical indication was trauma in three patients, degenerative stenosis or disc herniation in seven, and isthmic spondylolisthesis in three. Infection was located in the lumbar spine in eight patients, the cervical spine in two, the thoracic spine in one, the cervicothoracic junction in one, and the thoracolumbar junction in one.

The index procedure was decompression or microdiscectomy alone in four patients and instrumented fusion in seven; one patient with a C6 burst fracture underwent anterior corpectomy and fusion with an autogenous iliac graft, and one patient underwent two-level vertebroplasty for T12 and L1 compression fractures. Patients had undergone a mean of 4.9 ± 2.2 prior surgeries before the onset of infectious deformity, and mean follow-up was 64 months (range, 5–112).

Clinical and radiological outcomes

Given the small sample size, the statistical comparisons in this section should be regarded as hypothesis-generating rather than definitive. All patients presented with axial or radicular pain and neurological impairment at admission; one had combined bowel and urinary incontinence, and two had urinary incontinence alone.

Preoperative Frankel grade was A in 1 patient (7.7%), C in 4 (30.8%), and D in 8 (61.5%); no patient was grade E. At 1 year, no patient remained grade C: Eight patients (61.5%) were grade D, 4 (30.8%) had improved to grade E, and 1 (7.7%) remained grade A. Overall, neurological status improved in 7 patients (53.8%), remained stable in 5 (38.5%), and did not improve in 1 (7.7%) [Table 2]. The change in Frankel grade from baseline to 1 year was statistically significant (Stuart-Maxwell test, p < 0.05).

Table 2:

Neurological status and postoperative neurological outcomes.

Pain and functional outcomes improved substantially. Mean VAS score fell from 9.5 ± 0.9 preoperatively to 3.4 ± 1.8 on postoperative day 10, and to 1.8 ± 0.9 at 1 year (Friedman test, p < 0.001). Mean ODI improved from 62.5 ± 14.8 preoperatively to 34.9 ± 17.2 on postoperative day 10, and to 19.2 ± 12.7 at 1 year (p < 0.001).

Sagittal segmental alignment also improved significantly, from a mean kyphotic angle of −14.9° ± 17.4° preoperatively to a lordotic angle of +4.8° ± 8.7° postoperatively (Wilcoxon signed-rank test, p = 0.00024). The magnitude of angular correction, however, did not correlate significantly with change in VAS (Spearman r = 0.30, p = 0.33) or ODI (Spearman r = 0.11, p = 0.72) [Table 3]. These results indicate that, although significant radiological realignment and clinical improvement were both achieved, angular correction was not the sole determinant of clinical outcome.

Table 3:

Changes in pain, functional status, and sagittal alignment after revision surgery.

Laboratory findings

At presentation, the mean leukocyte count was 8.4 × 10³/mm³ (range, 3.4–16.5), the mean ESR at 1 h was 65 mm/h (range, 6–110), and CRP was elevated in all patients (mean, 45 mg/dL; range, 4–140).

Surgical approach and complications

Surgical technique was determined by the site and extent of the lesion and the severity of deformity and instability; representative preoperative and postoperative imaging from one patient is shown in Figures 1 and 2. Eleven patients received posterior instrumentation, nine underwent anterior decompression and fusion, and three required osteotomy. Two patients had an intraoperative dural tear with cerebrospinal fluid leak; both healed without a postoperative leak.

Figure 1:

Preoperative radiological findings in a representative patient (Case 4) with refractory postoperative wild spinal infection. (A and B) Anteroposterior and lateral radiographs demonstrating prior instrumentation and loss of segmental alignment. (C and D) Sagittal magnetic resonance images demonstrating vertebral body destruction and an epidural/paraspinal collection. (E) Coronal magnetic resonance image at an adjacent level. (F and G) Coronal and sagittal computed tomography reconstructions showing the extent of posterior instrumentation and segmental malalignment. (H) Sagittal computed tomography, bone window, demonstrating vertebral body collapse adjacent to instrumentation. (I) Sagittal magnetic resonance image of the lower lumbar spine.

Figure 2:

Postoperative radiological images of the same patient (Case 4) after aggressive surgical debridement and combined reconstructive stabilization. (A and B) Postoperative anteroposterior and lateral radiographs demonstrating restored instrumentation and alignment. (C-E) Postoperative sagittal magnetic resonance and computed tomography images demonstrating resolution of the epidural/paraspinal collection and restored sagittal alignment. (F and G) Coronal and sagittal computed tomography reconstructions confirming the integrity and positioning of the long-segment posterior instrumentation. (H-J) Follow-up magnetic resonance images demonstrating interval resolution of infection-related signal changes. (K) Axial computed tomography confirming bilateral pedicle screw placement. (L) Intraoperative photograph of the surgical field after radical debridement, demonstrating the extent of exposure required to achieve source control, including intraoperative bleeding and the exposed instrumented construct. (M) Gross specimen of resected necrotic and infected tissue removed during debridement, shown alongside a surgical instrument for scale.

Microbiological findings and medical treatment

Methicillin-resistant Staphylococcus aureus was the predominant pathogen (n = 8), followed by methicillin-sensitive S. aureus (MSSA, n = 3); Streptococcus mitis was identified in one patient with a history of intravenous drug use. Within the MSSA group, one patient had concurrent tuberculosis, one developed superinfection with MSSA and Escherichia coli, and one developed superinfection with Aspergillus fumigatus after a year of antibiotic therapy.

Patients with pyogenic infection received intravenous vancomycin and rifampicin at standard doses for 6 weeks, with treatment response monitored by CRP, clinical evaluation, and radiological follow-up; itraconazole was added for 6 weeks in the patient with Aspergillus infection. The patient with tuberculous spondylitis received rifampicin (maximum 600 mg/day orally), isoniazid (maximum 300 mg/day orally), pyrazinamide (maximum 2 g/day), and streptomycin (1 g/day intramuscularly) for 2 months, followed by rifampicin and isoniazid for an additional 10 months. For pyogenic infections, culture-directed antibiotics were continued parenterally for 6–8 weeks and orally for 2–3 months, with duration adjusted according to CRP and ESR trends.

DISCUSSION

Postoperative spinal infection typically responds to conventional treatment, but the patients in this study followed a markedly different course.[8] Rather than resolving with debridement and culture-directed antibiotics, this subgroup developed persistent infection, progressive osseous destruction, mechanical instability, and repeated surgical failure,[1] suggesting that refractory postoperative spinal infection represents a distinct end-stage phenotype rather than an extension of uncomplicated surgical site infection.[23]

A defining feature of these wild spinal infections was their poor and delayed response to antibiotic therapy. Whereas typical postoperative infections show rapid clinical and biochemical improvement with appropriate antimicrobial treatment, these patients had sustained inflammatory activity despite prolonged, culture-directed regimens.[17] The resulting progressive osteolysis and soft-tissue compromise weakened the structural integrity of the spine, predisposing patients to instability, deformity, and implant failure.[6]

These findings suggest that refractory postoperative spinal infection reflects an interaction between biological and mechanical failure rather than a purely infectious process. Persistent inflammation and inadequate antimicrobial response appear to initiate the process,[10] but biological factors alone do not explain the severity of clinical deterioration.[24] Mechanical failure, manifested as progressive instability, loss of load-bearing capacity, and segmental collapse, appeared to be the dominant driver of pain, neurological compromise, and functional decline in this population.[9,19] Once mechanical integrity is lost, repeated limited debridement without definitive stabilization is likely to prolong the disease course and delay recovery.[13] Our results support the view that, while biological control of infection is necessary, restoration of mechanical stability is the critical threshold for clinical improvement in end-stage postoperative infection.[23]

Although surgery achieved significant correction of sagittal alignment, the magnitude of this correction did not correlate with improvement in pain or function.[14] This dissociation distinguishes salvage reconstruction from elective deformity surgery: In biologically compromised bone, an aggressive pursuit of optimal alignment may increase mechanical stress at the bone-implant interface and, in turn, the risk of implant failure or recurrent infection.[11] Restoration of adequate load-bearing stability, rather than radiological perfection, appeared to be the primary determinant of symptomatic and neurological recovery in this cohort. These findings suggest that the surgical goal in refractory postoperative spinal infection should be mechanical stability and biological control rather than radiological optimization,[20] since overcorrection in a biologically weakened environment may confer no additional benefit and could be counterproductive.[4]

Despite the severity of disease and the high number of prior surgeries, more than half of the patients had neurological improvement, with meaningful functional recovery in many. Given the advanced stage of disease at presentation, these results highlight the potential for neurological recovery once adequate decompression and stabilization are achieved,[2,21] and likely reflect the combined effect of devitalized tissue removal, mechanical stabilization, and multidisciplinary postoperative management.[12]

Hyperbaric oxygen was applied uniformly as an adjunctive biological treatment.[15] Although it may enhance tissue oxygenation and support wound healing, its independent contribution to clinical outcome could not be isolated in this study, and it should be regarded as a complementary measure rather than a primary determinant of outcome.[16] In this cohort, durable infection control and functional recovery appeared to depend primarily on radical debridement and stable reconstruction.[18]

This study has several limitations, including its retrospective design, small sample size, and absence of a control group; the findings should not be generalized to postoperative spinal infection as a whole. Radiological assessment in this study was limited to segmental angular alignment; translational parameters such as listhesis were not systematically captured, which may have led to underestimation of residual mechanical compromise in some patients. In addition, the absence of a standardized institutional protocol for the surgical management of septic discitis or epidural abscess may have introduced variability in the timing and extent of intervention across cases. Because the cohort represents a rare, highly selected, end-stage population in whom conventional treatment had already failed, our aim was not to compare outcomes with uncomplicated postoperative infection, but to characterize a salvage population and identify principles that may guide its management.

From a practical standpoint, persistent infection despite adequate antimicrobial therapy should raise suspicion of underlying mechanical failure. In refractory cases, repeated attempts at biological control without addressing instability may delay definitive treatment and worsen biomechanical collapse. Early recognition of this high-risk phenotype, with timely transition to aggressive salvage reconstruction, may improve long-term functional and neurological outcomes.

CONCLUSION

Refractory postoperative spinal infection represents an end-stage phenotype in which persistent biological activity leads to progressive mechanical failure. In this subgroup, restoration of mechanical stability, rather than the extent of radiological correction, appears to be the key determinant of clinical recovery. Early recognition of this phenotype, with timely transition to aggressive salvage reconstruction, may improve long-term functional and neurological outcomes.

Ethical approval:

The Institutional Review Board approval is not required as it is a retrospective study.

Declaration of patient consent:

The authors certify that they have obtained all appropriate patient consent forms. In the form, the patient has given consent for their images and other clinical information to be reported in the journal. The patient understands that the patient’s names and initials will not be published and due efforts will be made to conceal their identity, but anonymity cannot be guaranteed.

Financial support and sponsorship:

Nil.

Conflicts of interest:

There are no conflicts of interest.

Use of artificial intelligence (AI)-assisted technology for manuscript preparation:

The author(s) confirms that they have used artificial intelligence (AI)-assisted technology solely to improve the grammar, language, and readability of the manuscript. No AI-generated content, data, analyses, or scientific interpretations were included. The authors reviewed and approved the final manuscript and take full responsibility for its content.

Disclaimer

The views and opinions expressed in this article are those of the authors and do not necessarily reflect the official policy or position of the Journal or its management. The information contained in this article should not be considered to be medical advice; patients should consult their own physicians for advice as to their specific medical needs.