Infectious diseases of the spine (IDS) encompass a range of musculoskeletal infections caused by pathogenic microorganisms in the intervertebral disc, vertebral body, spinal canal, or the surrounding tissues. These infections account for approximately 2%–7% of all musculoskeletal infections in the body [1–3]. Recent advancements in diagnostic methods for IDS have significantly improved the detection rate of microorganisms [4,5]. A combination of diagnostic modalities such as clinical radiology, microbial culture, molecular assays—including metagenomic next generation sequencing (mNGS)—and histopathology are increasingly being employed in clinical practice. Additionally, improved biopsy techniques, including the use of semirigid grasping forceps, enable the collection of larger specimen volumes, further increasing diagnostic efficacy [6]. Accurate etiological diagnosis of IDS is indispensable in controlling disease progression through personalized anti-infection protocols.

As population ages, specialist surgeons increasingly face the challenge of managing patients with IDS who often have poorly controlled comorbidities, including type 2 diabetes, hypertension, and renal insufficiency, along with complications such as heart disease and cerebral infarction. These patients typically require a prolonged period of management to stabilize their underlying conditions, thereby reducing anesthesia and perioperative risks. Emergent open surgery inevitably increases perioperative mortality [7]. Therefore, in IDS cases with neurological deficits warranting urgent surgery, a difficult decision arises regarding whether to prioritize preserving neurological function or proceeding with safe surgical intervention. This dilemma highlights the critical need for developing safe and effective debridement and decompression procedures tailored to patients with compromised systemic health and neurologic deficit.

Minimally invasive techniques have revolutionized surgical practice [8,9], with percutaneous endoscopy technology undergoing rapid advances over the past 2 decades. Spinal endoscopy, in particular, has become a widely accepted modality in clinical settings [10,11], offering numerous benefits, including reduced tissue damage, local anesthesia feasibility, enhanced perioperative safety, and effective drainage and irrigation of paravertebral and epidural spaces. Kim et al. [12] demonstrated the effectiveness of endoscopic treatment for thoracic spinal myelopathy in a 73-year-old patient with multiple comorbidities, including a severe kyphotic deformity (120°) resulting from spinal tuberculosis. This successful outcome underscores the potential of spinal endoscopic techniques as a viable alternative to traditional open decompression and fixation surgeries. We believe that endoscopic techniques may be particularly beneficial in managing IDS patients with multiple comorbidities and neurologic deficits. However, there is a lack of evidence regarding the efficacy and safety of spinal endoscopy in patients with IDS complicated by neurological deficits and multiple comorbidities. This study aimed to address this knowledge gap by investigating the effectiveness and safety of uniportal endoscopic decompression and debridement (UEDD) in this complex patient population.

This retrospective study was conducted at two-center hospital with the approval of the Institutional Review Board (IRB) of Jiangbei Branch of Southwest Hospital, 958th Hospital of the PLA Army (2024-KY003). Due to the retrospective nature of the study, the need to obtain the informed consent was waived by the IRB of 958th Hospital of the PLA Army. The American Society of Anesthesiologists (ASA) physical status classification system was used to assess comorbidities associated with IDS. The inclusion criteria consisted of the following conditions. Patients with ASA grades III and IV were included to evaluate the safety of spinal endoscopic technique. Additionally, the neurological function assessment developed by the American Spinal Cord Injury Association (ASIA) served as an inclusion criterion for patients with ASIA grade D or lower. Patients who underwent spinal endoscopy for non-infectious indications were excluded. An electronic database search was conducted to collect data regarding demographic characteristics, clinical manifestations, radiological manifestations, laboratory tests, microbiological cultures, histopathological examinations, and mNGS test results.

Thirty-two consecutive IDS patients with neurological deficits and multiple comorbidities who underwent UEDD surgery were included in this study. The mean age of patients in this series was 67.8 years, and the majority were male. Type 2 diabetes was the most common comorbidity. Twenty-five patients were classified as ASA grade III and seven patients as ASA grade IV. Prior to admission, 21 patients had received occasional antibiotic therapy, six patients had undergone only symptomatic treatment with oral non-steroidal anti-inflammatory drugs, and five patients had received diagnostic antituberculosis therapy. The demographic and clinical characteristics of these patients are summarized in Table 1.

In our series, infections in the lumbar spine were the most prevalent, followed by the thoracic spine. Specifically, the L3–4 and L4–5 segments were most affected in these regions. Severe back pain was the most common symptom. Epidural compression was predominantly observed on the ventral side, accounting for 56.3% of cases. Radiological assessments revealed a mean preoperative kyphotic angle of 35.2° in the thoracic focal segment, and a mean Cobb angle of 4.7° in the lumbar segment. The extent of vertebral bone destruction in all patients was less than half (Table 2).

The mean operation time was 151.4 minutes (range, 60–320 minutes), the mean length of hospital stay was 39.3 days (range, 15–83 days), and the mean duration of drainage was 11.8 days (range, 7–16 days). Only two patients had postoperative paresthesia of the lower limbs (Table 3).

Microbiological identification revealed definitive causative agents in 27 patients. Five patients did not receive an etiological diagnosis and were treated with empiric broad-spectrum antibiotics, typically a combination of vancomycin and a fluoroquinolone. The treatment outcome was favorable, with 28 patients meeting the criteria for being cured. Among patients with detectable bacteria, 24 individuals (88.9%) were successfully cured. Conversely, among patients with no detectable bacteria, 4 (80%) were cured. Statistical analysis revealed a significant difference in cure rate between the group with detected pathogens and group with undetected pathogen (80% versus 88.9%; χ²=19.36, p=0.000). Three patients experienced inadequate infection control, although continued antibiotic administration halted the spread of the infection. One of these cases lacked a definitively identified causative organism, while two cases of tuberculosis developed multi-drug resistance due to poor medication compliance.

Neurological deficits improved over time, as assessed by the ASIA grading system, with detailed results presented in Table 3. Pain levels, as measured by the VAS pain scores, consistently decreased from a mean of 7.9 preoperatively to 1.06 at the final follow-up. ODI demonstrated similar trends, with differences reaching statistical significance (p<0.05). Serum CRP levels decreased from an average preoperative value of 37.75±34.75 to 24.76±18.41 mg/L at 1 month postoperatively, 10.41±5.28 mg/L at 6 months postoperatively, and 4.13±2.14 mg/L at the last follow-up. ESR levels decreased more gradually over time (Table 4).

Tissue cultures were positive in 11 patients, while mNGS results were positive in 21 patients. Notably, there were two false-positive tissue culture results compared to four false-positive mNGS results. The average time to obtain a positive result from tissue culture was significantly longer, at 95.74±35.47 hours, compared to 41.72±6.81 hours for mNGS (p<0.05). Mycobacterium tuberculosis was the most common causative pathogen identified, followed by Staphylococcus aureus. Detailed strain information is provided in Table 5.

After drain removal, two patients experienced poor wound healing at the local drainage site, but eventually achieved healing following repeated dressing changes. One patient developed a pulmonary infection accompanied by pleural effusion, which responded well to anti-infective treatment. Drug-related hepatic dysfunction occurred in two cases of tuberculosis and one case of Aspergillus terreus infection. In these cases, we modified the antibiotic regimen and administered hepatoprotective therapy to manage the adverse effects.

The key finding of this study was that UEDD combined with subsequent antibiotic therapy achieved satisfactory efficacy (87.5% cure rate) in IDS patients with neurological deficits and multiple comorbidities, consistent with previous literature reports [14]. Moreover, UEDD offered a dual therapeutic-diagnostic advantage over traditional percutaneous biopsy during the initial admission phase, enabling simultaneous debridement, neural decompression, and targeted biopsy in a single intervention. This integrated approach facilitated timely neurological recovery through effective decompression of infected and inflamed tissues, and also accelerated etiological diagnosis through direct visualization and precise specimen collection. By overcoming the limitations of conventional biopsy methods, especially in cases requiring urgent intervention for neurological preservation, UEDD represents a major advancement in the management of spinal infections, combining therapeutic efficacy with diagnostic precision to optimize patient outcomes.

Despite its growing acceptance, spinal endoscopy in patients with IDS remains a topic of debate [15]. A major concern is the limited scope of endoscopic debridement, and that the use of irrigated saline can potentially facilitate the spread of infection. However, this study demonstrates that endoscopy can achieve therapeutic outcomes comparable to open surgery. This may be attributable to several factors. First, anti-infective antibiotic therapy is the cornerstone of IDS treatment, and a sensitive antibiotic regimen is crucial for successful treatment. In this study, emergency UEDD played a crucial role in improving neurological function by decompressing the spinal cord, while also enabling the collection of samples for comprehensive testing to identify pathogenic microorganisms. Subsequently, we formulated personalized antibiotic regimens based on the drug sensitivity testing results. Consistent with our findings, Pola et al. [16] identified microbiological diagnosis as a primary predictor of treatment success in their multifactorial analysis of 207 patients with pyogenic spondylodiscitis. The therapeutic benefits of UEDD extend beyond microbiological diagnosis. This minimally invasive procedure effectively removes necrotic tissue, while irrigation with normal saline and catheter utilization facilitate the discharge of local inflammatory substances and abscesses. By reducing the bacterial load and eliminating bacterial biofilm, UEDD helps reinstate the physiological healing process of acute wounds. Furthermore, local antibiotic injection through the drainage tube can provide additional benefits once a sensitive antibiotic regimen is established, enhancing the overall treatment efficacy. The current study observed no instances of iatrogenic spinal infection associated with standardized saline irrigation protocols, aligning with the established safety profiles in a prior systematic review [17]. This favorable outcome may be attributed to the implementation of two key procedural safeguards in our spinal infection debridement protocol: maintained patency of irrigation outflow channels through real-time pressure monitoring (target <30 mm Hg) and extended postoperative drainage (mean duration, 11.8 days) using negative pressure systems.

The management of IDS primarily centers on targeted antimicrobial therapy, while surgical intervention remains a critical adjunctive approach in selected cases. However, the optimal management strategy for IDS patients presenting with multiple comorbidities remains a subject of ongoing debate within the spine surgery community, particularly regarding the optimal balance between conservative pharmacotherapy and invasive surgical procedures. Jin et al. [18] investigated the risk factors for treatment failure in intervertebral discitis, comparing conservative and surgical approaches. Their study identified epidural abscesses, cervical or thoracic spine involvement, and multi-segment spinal involvement as risk factors for conservative treatment failure. Giampaolini et al. [19] proposed a criterion for surgical intervention in pyogenic spondylodiscitis, recommending surgical debridement after 4 weeks of conservative treatment if inflammatory markers do not significantly decline (ESR <50 mm/hr and CRP <2.7 mg/dL). The universal surgical indications for IDS include neurological deficits, and imaging signs of instability, deformity, or abscesses. Conventional anterior or posterior debridement procedures have been employed for surgically indicated IDS cases, demonstrating favorable outcomes in historical series [20–22]. However, our cohort comprised high-risk patients classified as ASA III or IV, a population with documented perioperative mortality rates ranging from 1.82% to 23.0% [23]. Notably, our series achieved zero perioperative mortality. The UEDD technique may offer a preferential therapeutic option for IDS management particularly in patients with compromised systemic conditions. The minimally invasive nature of UEDD, combined with contemporary diagnostic advancements, makes it an attractive alternative to conservative or open surgical procedures for high-risk patients.

Furthermore, UEDD offers a significant advancement over traditional percutaneous biopsy methods, boasting a dual capacity for neural decompression and precise endoscopic delineation of infected tissue margins. These technical advantages, unattainable through standard needle biopsy techniques, enable more effective infection management. In our clinical experience, patients undergoing UEDD exhibited marked postoperative improvement in axial back pain and radicular lower extremity symptoms. Subsequent pain resolution and functional recovery paralleled effective infection control achieved via continuous drainage and targeted antibiotic therapy.

There is no clear consensus regarding the spinal stability assessment, particularly concerning the ambiguous distinction between stability and instability during spinal infections [24]. Inflammatory stimulation often leads to localized tension in the paravertebral muscles, resulting in insensitivity to mechanical pain symptoms. In the present study, the degree of vertebral collapse was minimal, with destruction of the vertebral body being less than 50% in 23 patients, and no vertebral invasion in nine patients. Notably, none of the cases exhibited radiological signs of subluxation or translation. Consequently, after excluding mechanical symptoms, the highest SINS score recorded among patients was 12. This finding indicates the crucial role of mechanical pain symptoms in stability assessment. Therefore, in cases of epidural abscess with uncertain spinal stability, aggressive internal fixation combined with fusion surgery may be considered imprudent. Such surgical interventions not only extend the operative time and trauma but also increase perioperative risks. Moreover, in the absence of identified pathogenic bacteria, internal fixation can increase the risk of biofilm formation, particularly by S. aureus, which is notorious for readily generating biofilms [25]. This can exacerbate the local inflammatory microenvironment, promote osteoclast differentiation, complicate bone fusion, and potentially lead to pseudarthrosis formation [26,27]. A significant advantage of staged surgery is that the first stage involves endoscopic debridement and decompression. This approach provides several benefits, including restoring spinal cord function, enabling microbiological diagnosis, facilitating drainage of local inflammatory factors, and alleviating pain. In the second stage, a systematic reassessment of spinal stability is conducted for an objective evaluation of mechanical symptoms. Subsequent fusion surgery further helps alleviate the local inflammatory microenvironment, creating conditions conducive to bone fusion and promoting osteoblast differentiation. However, in cases with significant spinal stability loss, open debridement, bone graft fusion, and internal fixation may be necessary. This typically involves situations where vertebral collapse exceeds 50%, accompanied by kyphosis greater than 30°, and spinal subluxation or translation, particularly in the cervicothoracic and thoracolumbar junction regions.

Recent studies have documented satisfactory outcomes of endoscopic treatment for spondylodiscitis. Abreu et al. [15] reviewed 342 patients with spondylodiscitis who underwent endoscopic treatment, noting a failure rate of 0 to 33%. They concluded that endoscopic debridement may be an effective and safe method for treating refractory spondylodiscitis. Yamada et al. [28] reported a cure rate of 82.8% in 64 patients with spondylodiscitis who underwent endoscopic surgery combined with negative pressure drainage. Their analysis identified large abscess cavities and multi-level intervertebral infections as risk factors for failure. In our series, patients with a definitive etiological diagnosis exhibited significantly superior therapeutic outcomes compared to those with undetermined pathogens (88.9% versus 80%), suggesting a crucial role of precise microbiological identification in optimizing treatment efficacy for IDS. This observation highlights the clinical significance of targeted pathogen detection in improving infection resolution rates through individualized antimicrobial stewardship. Furthermore, our series achieved an etiological diagnosis rate of 84.3%, surpassing previously reported rates of 71.1% by Duan et al. [14] and 76% by Polo et al. [29]. This diagnostic superiority may be attributed to several key factors. Primarily, the implementation of mNGS significantly enhanced the detection of pathogenic microorganisms. This advanced technique enables comprehensive identification of suspected microbial species by performing high-throughput sequencing of all nucleic acids in clinical specimens. Unlike conventional culture methods, mNGS even detects residual nucleic acids from non-viable organisms, making its effectiveness independent of pathogen viability. A retrospective cohort analysis by Yin et al. [30] demonstrated the superiority of mNGS over conventional culture methods in detecting pathogens in IDS cases. Their study revealed significantly higher pathogen detection rates using mNGS (88.42%) compared to conventional culture methods (43.16%, p<0.05), with mNGS demonstrating a diagnostic sensitivity of 86.44% and specificity of 92.00% [30]. Additionally, biopsy specimen quality appears critical in microbiological identification. Polo et al. [29] reported sensitivity rates of 60.00% for open surgical biopsies versus 46.27% for CT-guided biopsies. Our UEDD technique, analogous to open surgical biopsy in terms of specimen acquisition quality, permits endoscopic-guided retrieval of diagnostically superior tissue samples, as illustrated in Fig. 1R. Notably, our culture-based pathogen detection rate (34.3%) proved lower than that reported in literature. While Staphylococcus species predominated in the aforementioned studies [14,29], M. tuberculosis was the primary causative organism in our cohort. This observed discrepancy in culture sensitivity likely reflects the more stringent culture requirements and prolonged incubation periods inherent to M. tuberculosis isolation. Additionally, physiological saline irrigation during specimen collection may reduce bacterial load in cultured samples, potentially impacting conventional culture sensitivity. Our findings further substantiate the diagnostic superiority of mNGS, which exhibits both reduced processing time and broader pathogen detection spectrum compared to conventional methods. This has important implications for the early implementation of personalized antibiotic regimens. However, clinicians should be aware of the inherent risk of false-positive mNGS results, as documented in our prior investigation [4]. Therefore, positive mNGS findings should be interpreted in conjunction with radiographic characteristics, clinical manifestations, and serological parameters to optimize diagnostic accuracy.

Some limitations of this study should be acknowledged. First, as a retrospective analysis, it has an inherently low level of evidence and lacks a control group, limiting our ability to demonstrate advantages over traditional open debridement. Second, the relatively small sample size and minimal complication rates precluded further multivariate analysis to identify potential risk factors for treatment failure. The relatively short follow-up period may also have confounded the results, as infections could recur after antibiotic discontinuation, particularly in patients with intracellular infections, like tuberculosis. Additionally, technical bias may have been introduced due to the involvement of different surgeons at the two centers. Finally, we were unable to calculate the sensitivity and specificity of mNGS due to the lack of data on non-spondylodiscitis cases diagnosed using this method.

This study demonstrated that UEDD is a viable alternative to traditional open surgery for managing IDS in high-risk patients. In this study, UEDD achieved a clinical cure rate of 87.5% among individuals presenting with neurological deficits (ASIA grade D or lower) and multiple comorbidities (ASA class III–IV). Furthermore, UEDD offered a dual therapeutic-diagnostic advantage during the initial admission phase, enabling simultaneous debridement, neurological decompression, and targeted biopsy in a single intervention. Moreover, mNGS played a significant role in rapid diagnosis and extensive pathogen coverage, outperforming traditional tissue culture.