Involvement of L5-S1 level as an independent risk factor for adverse outcomes after surgical treatment of lumbar pyogenic spondylitis: A multivariate analysis

Abstract

Purpose:

To identify the independent risk factors for adverse outcomes and determine the effect of L5-S1 involvement on the outcome of surgical treatment of lumbar pyogenic spondylitis (PS).

Methods:

A retrospective analysis was performed for all consecutive patients who underwent surgery for lumbar PS between November 2004 and June 2020 at a single institution. The patients were divided into two groups based on the outcomes: good and adverse (treatment failure, relapse, or death). Treatment failure was defined as persistent or worsening pain with C-reactive protein (CRP) reduction less than 25% from preoperative measurement or requiring additional debridement. Relapse was defined as the reappearance of symptoms and signs with an elevated white blood cell count, erythrocyte sedimentation rate, and CRP after the first period of treatment. Binary logistic regression analyses were performed to identify the independent risk factors for adverse outcomes.

Results:

Twenty-four (21.2%) of the 113 patients were classified as having adverse outcomes: treatment failure, relapse, and death occurred in 15, 7, and 2 patients, respectively. The involvement of L5-S1 (adjusted odds ratio [aOR] = 6.561, P = 0.004), Methicillin-resistant Staphylococcus aureus (MRSA) infection (aOR = 6.870, P = 0.008), polymicrobial infection (aOR = 12.210, P = 0.022), and Charlson comorbidity index (CCI; P = 0.005) were identified as significant risk factors for adverse outcomes.

Conclusion:

Involvement of L5-S1, MRSA, polymicrobial infection, and CCI were identified as independent risk factors for adverse outcomes after surgical treatment of lumbar PS. Because L5-S1 is anatomically demanding to access anteriorly, judicious access and thorough debridement are recommended in patients requiring anterior debridement of L5-S1.

Keywords

adverse outcome infectious spondylitis L5-S1 pyogenic spondylitis surgical debridement

Introduction

The incidence of spinal infection has been rising recently, possibly due to the increased life expectancy of patients with chronic disease, better diagnostic rate, intravenous drug use, and an increase in health-care-related infection and spinal surgery.^1
–3 Spinal infection represents 2–7% of all cases of osteomyelitis,^1,4 and the incidence has been reported to range from 2 to 24 cases per million annually.^1,3,5

The etiology of spinal infection, or infectious spondylitis, comprises pyogenic, granulomatous (tuberculous, brucellar, or fungal), or parasitic infection.^3,6 Pyogenic spondylitis (PS), the most common type of infectious spondylitis,⁷ is a broad term encompassing spondylodiscitis, vertebral osteomyelitis (VO), epidural abscess, and facet arthropathy caused by bacterial species.^3,8 The diagnosis and management of PS remains a challenge due to the insidious onset, non-specific symptoms, and debilitating course.⁹ PS is potentially life-threatening as elderly patients with a precarious state of health or chronic debilitating disease are easily affected.^9,10 There have been several studies on the outcome or prognosis of the entire PS.^11,12 However, there is a paucity of literature on the prognosis and risk factors for adverse outcomes after surgical intervention for PS. Patients requiring surgical intervention for PS are generally more vulnerable to adverse outcomes than those cured by conservative treatment because they are more likely to have progressive neurological deficits, impending biomechanical instability, or failed antibiotic therapy. It is more critical to recognize those at a higher risk of adverse outcomes in this surgical population.

The L5-S1 level is anatomically difficult to access anteriorly due to the variation of major vascular structures and larger overlap of the iliac crest.^13,14 Additionally, infectious conditions would cause adhesion between the annulus and iliac vessels, making the identification and dissection of anatomical structures difficult and iliac vessels prone to injury during the anterior approach. Therefore, we aimed to determine whether the involvement of the L5-S1 segment affects the outcome and identify independent risk factors for adverse outcomes after surgical intervention of PS in the lumbar spine, the most common site for PS.

Materials and methods

Patients and surgical treatment

This retrospective study was approved by the institutional review board of our institution, and a waiver of consent was obtained. We reviewed the medical records of patients who underwent surgery for PS between November 2004 and June 2020 at a single institution, a tertiary care referral center. Patients with granulomatous (tuberculous, brucellosis, or fungal) spondylitis, cervical or thoracic infectious spondylitis, or incomplete medical record documentation were excluded. Four surgeons from a single orthopedic department performed surgical treatment for PS. Surgical indications included neurologic deficits or uncontrollable pain, significant progressive bone destruction causing mechanical instability, and failure to respond to conservative treatment.^1,10 The detailed surgical plan including the surgical approach was determined based on the location and extent of infection and can be divided into four techniques: (1) anterior debridement/decompression and fusion (anterior only); (2) posterior debridement/decompression with or without fusion with instrumentation (posterior only); (3) anterior debridement/decompression and fusion, with simultaneous posterior instrumentation with or without fusion, (single-stage combined); and (4) anterior debridement/decompression and fusion, with sequential posterior instrumentation with or without fusion (two-stage combined).¹⁵ Regarding the patients with remnant implants at the time of PS diagnosis, the surgeons removed the implants in case of screw loosening or the infection occurring more than 3 months after the index procedure, and reinserted implants in case of the risk of pseudarthrosis or loss of correction.^16
–18 Antibiotics were administered for at least 6 weeks according to the recommendation of an infection specialist in the internal medicine department.

Data collection

The medical records were reviewed for demographic data, comorbidities, preoperative laboratory data (white blood cell [WBC] count, erythrocyte sedimentation rate [ESR], and C-reactive protein [CRP]), microbiological and histological data, involved vertebral levels, preoperative neurologic status, radiographic characteristics (presence of paravertebral and/or epidural abscess), the number of levels operated, and surgical procedures.

The American Society of Anesthesiologists (ASA) physical status classification system and Charlson comorbidity index (CCI) were collected to assess the burden of comorbidities. ASA grades for each patient were obtained from the preoperative anesthesia record in our electronic medical record system, which was documented by the anesthesiologist at our institution. The CCI grade was classified into three categories according to previous studies: 0, 1, and 2+.¹⁹

Diagnosis of pyogenic spondylitis

In this study, PS was defined when all the following criteria were met: compatible symptoms or signs, including fever, neurological signs, or pain in the affected area; radiologic evidence on magnetic resonance imaging; and biological evidence, including laboratory data showing infectious conditions and/or isolation of microorganisms from an image-guided percutaneous or surgical tissue culture, or blood culture performed at the time of illness.^20
–22 When common resident skin flora grew in the tissue or blood culture, it was considered as a true pathogen only when the same organism was identified in two or more separate specimens.^20,21 In patients with negative culture results, PS was diagnosed when all the following criteria were satisfied: no histologic evidence of granulomatous inflammation or fungal infection; negative results of a prolonged cultures for Mycobacterium tuberculosis and the polymerase chain reaction for Mycobacterium complex; and intraoperative findings consistent with pyogenic infection such as purulent discharge.^22,23 Polymicrobial infection was defined as infection caused by at least two distinct organisms identified from culture studies.²³ Moreover, we defined postoperative PS as patients with demonstrated infection appearing within 1 year after spinal surgery at the same location or who had residual implants at the time of diagnosis.²¹

Outcome assessment

Outcome was determined based on changes in symptoms and signs, laboratory data, and overall mortality. The patients were divided into the good outcome and adverse outcome groups. Patients with treatment failure, relapse, or infection-related deaths were assigned to the adverse outcome group. Treatment failure was defined as persistent or worsening pain with reduction of CRP less than 25% from preoperative measurement, or requiring additional surgical debridement despite thorough surgical debridement and administration of appropriate antibiotics for more than 1 month according to consultation answers from an infection specialist. Relapse was defined as the reappearance of symptoms and signs with an elevated WBC, ESR, and CRP after completion of the first period of treatment.^21,24 Only in-hospital deaths related to infection or consequent complications were regarded as infection-related deaths in this study.²⁵

Statistical analysis

Continuous variables were compared using the Student’s t-test or Wilcoxon rank sum test, and categorical variables were compared using the χ² test or Fisher’s exact test. Univariate analyses were performed for all variables using binary logistic regression analysis. Multivariate binary logistic regression was applied to identify independent risk factors associated with adverse outcomes using variables with a P-value < 0.1, by univariate analysis. To manage any issues with multicollinearity, we used a stepwise method and applied an entry condition of P < 0.05, removal condition of P > 0.1, and the final model included only significant variables with a P-value < 0.05. Adjusted odds ratios (aORs) and 95% confidence intervals were calculated for all variables. All statistical analyses were performed using the SPSS system for Windows the SAS system for Windows, version 9.4 (SAS institute, Cary, NC, USA). Statistical significance was set at P < 0.05.

Results

Patients’ clinical characteristics

A total of 113 consecutive patients who underwent surgical intervention for pyogenic lumbar spondylitis were identified (63 male; mean age: 63.5 ± 13.1 years). Of these patients, pyogenic infection involved the L1-2 segment in 11 patients (9.7%), L2-3 in 21 patients (18.6%), L3-4 in 26 patients (23.0%), L4-5 in 56 patients (49.6%), and L5-S1 in 25 patients (22.1%). Twenty-four patients (21.2%) were classified as having adverse outcomes: treatment failure, relapse, and death occurred 15, 7, and 2 patients, respectively. The average follow-up duration was 34.5 ± 29.5 months.

Microbiological results

Among the 113 patients, microorganisms were identified in 97 cases (85.8%). The most commonly identified organism was Staphylococcus aureus (n = 29, 25.44%), followed by coagulase-negative Staphylococci (n = 22, 19.47%). Methicillin-resistant Staphylococcus aureus (MRSA) was significantly more common in patients with adverse outcomes than in those with good outcomes (29.17% vs. 8.99%, P = 0.017). All microbiological study results are described in Table 1.

Table 1.

Identified micro-organisms in patients with good and adverse outcome.

	Total n = 113 (%)	Good outcome n = 89 (%)	Adverse outcome n = 24 (%)	P-value*
Gram-positive species	77 (66.38)	56 (62.22)	21 (80.77)	0.100
Staphylococcus aureus	29 (25.44)	21 (23.6)	8 (32)	0.439
Methicillin-sensitive	14 (12.39)	13 (14.61)	1 (4.17)	0.295
Methicillin-resistant	15 (13.27)	8 (8.99)	7 (29.17)	0.017
Coagulase-negative staphylococci	22 (19.47)	17 (19.1)	5 (20.83)	1.000
Methicillin-sensitive	3 (2.65)	3 (3.37)	0 (0)	1.000
Methicillin-resistant	19 (16.81)	14 (15.73)	5 (20.83)	0.548
Streptococcus species	13 (11.5)	9 (10.11)	4 (16.67)	0.469
Streptococcus pneumoniae	1 (0.88)	1 (1.12)	0 (0)	1.000
Streptococcus agalactiae	3 (2.65)	1 (1.12)	2 (8.33)	0.114
Viridans streptococci	6 (5.31)	5 (5.62)	1 (4.17)	1.000
Other streptococci	3 (2.65)	2 (2.25)	1 (4.17)	0.515
Enterococcus species	13 (11.5)	9 (10.11)	4 (16.67)	0.469
Enterococcus faecalis	7 (6.19)	4 (4.49)	3 (12.5)	0.164
Enterococcus faecium	1 (0.88)	1 (1.12)	0 (0)	1.000
Enterococcus raffinosus	1 (0.88)	1 (1.12)	0 (0)	1.000
Gram-positive bacilli	4 (3.54)	3 (3.37)	1 (4.17)	1.000
Gram-negative species	46 (40.35)	36 (40.45)	10 (40)	1.000
Escherichia coli	7 (6.19)	4 (4.49)	3 (12.5)	0.164
Enterobacter cloacae	2 (1.77)	2 (2.25)	0 (0)	1.000
Klebsiella pneumoniae	4 (3.54)	3 (3.37)	1 (4.17)	1.000
Pseudomonas aeruginosa	5 (4.42)	4 (4.49)	1 (4.17)	1.000
Serratia marcescens	1 (0.88)	1 (1.12)	0 (0)	1.000
Salmonella	1 (0.88)	1 (1.12)	0 (0)	1.000
Burkholderia cepacia	3 (2.65)	3 (3.37)	0 (0)	1.000
Stenotrophomonas maltophilia	2 (1.77)	2 (2.25)	0 (0)	1.000
Other Gram-negative bacilli	3 (2.65)	2 (2.25)	1 (4.17)	0.515
Anaerobes	2 (1.77)	1 (1.12)	1 (4.17)	0.381
Culture-negative	16 (14.16)	13 (14.61)	3 (12.5)	1.000

*P-value derived by Fisher’s exact test.

Comparison of patients with good and adverse outcomes

When the patients with good (n = 89) and adverse outcomes (n = 24) were compared, patients with adverse outcomes had a significantly higher proportion of PS at L5-S1 (41.67% vs. 16.85%, P = 0.009), MRSA infection (29.17% vs. 8.99%, P = 0.017), polymicrobial infection (20.83% vs. 2.25%, P = 0.005), and preoperative paresis (41.67% vs. 21.35%, P = 0.043) (Table 2). Additionally, preoperative WBC (×10³/mm³) was significantly higher in patients with adverse outcomes than in those with good outcomes (10.05 ± 3.91 vs. 7.95 ± 3.00, P = 0.017). Regarding the burden of comorbidities, patients with adverse outcomes had significantly higher ASA class (P = 0.016) and CCI scores (P < 0.001).

Table 2.

Comparison of demographics and clinical characteristics between patients with good and adverse outcome.

	Good outcome n = 89	Adverse outcome n = 24	P-value
Age	63.33 ± 13.89	64.06 ± 9.79	0.863
Sex (n, %)			0.094
Male	46 (51.69)	17 (70.83)
Female	43 (48.31)	7 (29.17)
Number of levels operated	1.24 ± 0.58	1.21 ± 0.41	0.815
Preoperative laboratory data
WBC (×10³/mm³)	7.95 ± 3.00	10.05 ± 3.91	0.017*
ESR (mm/h)	60.52 ± 25.84	72.39 ± 31.41	0.103
CRP (mg/dL)	5.87 ± 6.16	7.52 ± 6.20	0.147
Abscess (n, %)
Epidural abscess	55 (61.80)	15 (62.5)	0.950
Paravertebral abscess	53 (59.55)	18 (75)	0.165
Surgical procedures (n, %)			0.162
Anterior only	22 (24.72)	2 (8.33)
Posterior only	18 (20.22)	3 (12.5)
Single-stage combined	24 (26.97)	8 (33.33)
Two-stage combined	25 (28.09)	11 (45.83)
Location of infection (n, %)
L1-2	11 (12.36)	0 (0)	0.117
L2-3	19 (21.35)	2 (8.33)	0.236
L3-4	20 (22.47)	6 (25)	0.794
L4-5	45 (50.56)	11 (45.83)	0.681
L5-S1	15 (16.85)	10 (41.67)	0.009*
MRSA infection (n, %)	8 (8.99)	7 (29.17)	0.017*
Polymicrobial infection (n, %)	2 (2.25)	5 (20.83)	0.005*
Negative culture	13 (14.61)	3 (12.5)	1.000
Postoperative pyogenic spondylitis (n, %)	33 (37.08)	11 (45.83)	0.435
Presence of implants	13 (14.61)	7 (29.17)	0.097
CCI (n, %)			<0.001*
0	46 (51.69)	2 (8.33)
1	35 (39.33)	15 (62.5)
2+	8 (8.99)	7 (29.17)
ASA (n, %)			0.016*
1	17 (19.1)	1 (4.17)
2	52 (58.43)	11 (45.83)
3	20 (22.47)	12 (50)
Preoperative paresis (n, %)	19 (21.35)	10 (41.67)	0.043*

ASA, American Society of Anesthesiologists Physical Status Classification System; CCI, Charlson Comorbidity Index; CRP, C-reactive protein; ESR, erythrocyte sedimentation rate; MRSA, Methicillin-resistant Staphylococcus aureus; WBC, white blood cell counts.

* Statistically significant (P < 0.05).

Logistic regression analysis: univariate and multivariate analyses

Univariate analysis identified preoperative WBC (crude odds ratio [cOR] = 1.207, P = 0.008), PS at L5-S1 (cOR = 3.524, P = 0.012), MRSA infection (cOR = 4.169, P = 0.014), polymicrobial infection (cOR = 11.447, P = 0.005), CCI (P = 0.003), ASA (P = 0.027), and preoperative paresis (cOR = 2.632, P = 0.048) (Table 3). Using the stepwise method, PS at L5-S1 (aOR = 6.561, P = 0.004), MRSA infection (aOR = 6.870, P = 0.008), polymicrobial infection (aOR = 12.210, P = 0.022), and CCI (P = 0.005) were identified by the multivariate analysis as significant risk factors for adverse outcomes after surgical treatment (Table 4).

Table 3.

Univariate analysis of risk factors for adverse outcome after surgical treatment of lumbar pyogenic spondylitis.

	Crude odds ratio	95% CI	P-value
Age	1.004	0.969–1.041	0.808
Sex	2.270	0.858–6.009	0.099
Number of levels operated	0.908	0.384–2.150	0.827
Preoperative laboratory data
WBC	1.207	1.051–1.385	0.008*
ESR	1.017	0.996–1.037	0.107
CRP	1.041	0.972–1.115	0.249
Abscess
Epidural abscess	1.049	0.414–2.662	0.919
Paravertebral abscess	2.077	0.751–5.742	0.159
Location of infection
L1-2	0.139	0.007–2.776	0.197
L2-3	0.335	0.072–1.553	0.162
L3-4	1.150	0.430–3.285	0.794
L4-5	0.827	0.335–2.043	0.681
L5-S1	3.524	1.318–9.418	0.012*
MRSA infection	4.169	1.332–13.052	0.014*
Polymicrobial infection	11.447	2.064–63.499	0.005*
Negative culture	0.835	0.218–3.206	0.793
Postoperative pyogenic spondylitis	1.436	0.557–3.239	0.436
Presence of implants	0.385	0.082–1.798	0.225
CCI			0.003*
1 vs. 0	9.855	2.114–45.942	0.004*
2+ vs. 0	20.121	3.526–114.808	0.001*
ASA			0.027*
2 vs. 1	3.596	0.432–29.932	0.237
3 vs. 1	10.200	1.200–86.692	0.033*
Preoperative paresis	2.632	1.011–6.851	0.048*

ASA, American Society of Anesthesiologists Physical Status Classification System; CCI, Charlson Comorbidity Index; CI, confidence interval; CRP, C-reactive protein; ESR, erythrocyte sedimentation rate; MRSA, Methicillin-resistant Staphylococcus aureus; WBC, white blood cell counts.

* Statistically significant (P < 0.05).

Table 4.

Multivariate analysis of risk factors for adverse outcome after surgical treatment of lumbar pyogenic spondylitis.

	Adjusted odds ratio	95% CI	P-value
Pyogenic spondylitis at L5-S1	6.561	1.843–23.357	0.004*
MRSA infection	6.870	1.655–28.519	0.008*
Polymicrobial infection	12.210	1.424–104.663	0.022*
CCI			0.005*
1 vs. 0	10.710	2.019–56.801	0.005*
2+ vs. 0	23.139	3.285–162.984	0.002*

CCI, Charlson Comorbidity Index; CI, confidence interval; MRSA, Methicillin-resistant Staphylococcus aureus.

* Statistically significant (P < 0.05).

Discussion

While conservative treatment, comprising antibiotic therapy and immobilization of the affected segment with bed rest and/or orthosis, is considered the first treatment choice for PS,^9,10 surgical intervention might be required in some cases. Accordingly, it is reasonable to distinguish the surgical treatment of PS from conservative treatment with respect to the outcomes and prognosis, since the indications of these treatments are different, and surgery would also be required when conservative treatment alone appears to be insufficient. Therefore, we only included patients managed by surgical treatment to improve the homogeneity of the study population and specify the surgery outcome.

The definitions of PS have varied in previous studies.^20,21,26 While the diagnosis of spinal infection can be confirmed by microbiological studies, several studies have reported culture-negative cases of spinal infection.^22,27,28 Sixteen cases (14.16%) in this cohort had negative cultures, which was lower than that reported in previous studies.^8,22 While patients with negative culture are reportedly less likely to have pyrexia, lower WBC, ESR, and CRP, culture-negativity had no significant impact on the final outcomes in this study, which is consistent with the results of previous studies.^22,28

There have been several studies on the outcomes of PS and related factors.^21,22 In this study, independent risk factors for adverse outcomes after surgical intervention of PS in the lumbar spine were involvement of the L5-S1 segment, MRSA infection, polymicrobial infection, and CCI. To our knowledge, this is the first study to suggest that L5-S1 infection is an independent risk factor for poor outcomes of spinal infection. In previous studies, several anatomical variations of vascular structures have been reported around the L5-S1 level, and these great vessels might block anterior access to the disc space.^13,29 Moreover, it is difficult to secure a working window because the iliac vessels, peritoneal contents, and nerves are prone to injury during the approach due to adhesion in the infectious condition, although wide exposure of infectious lesions is necessary for thorough debridement.

Among the 25 patients with L5-S1 involvement in the current cohort, 9 patients had facet joint infection and/or epidural abscess and underwent only posterior debridement. While only 1 (11.1%) of these 9 patients had adverse outcomes; however, 9 (56.3%) of 16 patients who underwent anterior debridement had adverse outcomes (P = 0.027). These results indicate that the presence of spondylodiscitis necessitating anterior debridement at the L5-S1 level can be associated with adverse outcomes after surgical management. Therefore, spine surgeons should secure a sufficient working window during the anterior approach to the L5-S1 level for meticulous debridement, and assistance by a vascular surgeon should be requested whenever necessary (Figure 1).

Figure 1.

A 40-year-old male patients who underwent anterior debridement and iliac bone graft with vascular surgeon’s assistance with sequential posterior instrumentation (two-stage combined) surgery for L5-S1 spondylodiscitis due to the failure to respond to antibiotics for more than 1 month. (A) Preoperative sagittal T2-weighted MR image showing intradiscal fluid-like signal and bone marrow edema involving L5 and S1. (B) Preoperative axial T2-weighted MR image revealing the adhesion between iliac vessels and infectious tissue. (C, D) Postoperative T2-weighted MR image showing partial improvement of pyogenic spondylitis. (E, F) The last follow-up X-ray showing well-maintained status without structural change.

Staphylococcus aureus was the most commonly isolated microorganism, and MRSA was significantly associated with the adverse outcome of PS in this study, consistent with the reported literature.^8,21,26 Kim et al.²¹ suggested more aggressive management, including prolonged use of antibiotics, debridement, and implant removal, to achieve favorable outcomes in pyogenic VO caused by MRSA. However, MRSA was still associated with adverse outcomes of PS despite surgical management and appropriate use of antibiotics. Meanwhile, DiPaola et al.³⁰ recommended multiple debridements to eradicate MRSA infections. In our study population, 5 of 15 patients with MRSA infection had multiple debridements and eventually achieved resolution of infection, although we designated these multiple debridement cases as adverse outcomes.

We identified 7 (6.19%) polymicrobial cases in this study population, which is comparable to the reported incidence ranging from 4.4% to 24.4%.^8,22,31 Polymicrobial infection has been associated with increased mortality in bloodstream infections (BSIs) and high rates of treatment failure in postoperative spinal infections.^17,18,32,33 However, Hadjipavlou et al.⁸ found that polymicrobial infection was not associated with the outcome of hematogenous PS. In our study, it was identified as a risk factor for poor outcomes after surgical treatment. Wille et al.¹⁷ suggested that antibiotic therapy of polymicrobial infections often requires the use of multiple antibiotics, thereby increasing the risk of drug toxicity and causing the interruption of adequate treatment. Additionally, several studies have reported that polymicrobial BSIs are associated with anaerobes or high-risk bacteria, such as Pseudomonas aeruginosa, Klebsiella pneumoniae, and Enterobacter faecalis, and excess mortality.^32
–34 In our cohort, 6 (85.7%) of the 7 polymicrobial cases were associated with these microorganisms. However, since there are some discrepancies in the study population and definitions of adverse outcomes in the literature, the direct comparison of these results is limited. A well-designed study regarding polymicrobial infection and outcomes of PS should be conducted in the future.

Comorbidity has been reported to be associated with impaired immunity and infection outcomes.^35,36 Yoon et al.²² suggested comorbidity as a significant factor in determining the outcome of pyogenic VO. Although the ASA and CCI were originally developed to assess the burden of comorbidities and predict mortality for various medical conditions, they have also been applied to evaluate the outcome and likelihood of complications following spine surgery.^37,38 The CCI was identified as a risk factor for adverse outcomes by multivariate analysis in our study. In addition, ASA was identified as a risk factor by univariate analysis, but not in multivariate analysis. These results imply that not only the comorbidity status, but also the management of those would have an important role in the management of PS.

In several previous studies, relapse rates of postoperative vertebral osteomyelitis (PVO) ranged 0–4%, and there were no significant differences in treatment failure rates between postoperative and native vertebral osteomyelitis (NVO).^24,39 In contrast, Kim et al.²¹ reported that the treatment failure and relapse rates of PVO were significantly higher than those of NVO, although more patients with PVO had surgery compared to those with NVO. However, while this report included both conservatively and surgically managed patients, the present study only included patients who underwent surgical debridement and showed no significant difference in the outcome between postoperative and native lumbar PS. Aggressive surgical debridement with or without implant removal has been suggested as an important factor for the successful treatment of postoperative spinal infections.^18,21,40 Consistent with these reports, our results suggest that aggressive surgical debridement should be considered in the management of postoperative PS. Future systematic study to standardize the detailed treatment strategy for postoperative spinal infections.

Our study has several limitations. First, due to its retrospective nature, there may be selection bias and confounding factors that have not been considered. Second, it was difficult to recognize the occurrence of relapse beyond our follow-up period in patients who were thought to have completely recovered from PS. Third, because the decision for surgery and surgical plan depended on each surgeon’s judgment, there might have been a slight discrepancy between surgeons even though there have been commonly agreed indications of surgery for PS. Finally, a clear consensus regarding the diagnosis and outcome of PS is yet to be established. Despite these limitations, to the best of our knowledge, this is the first study to analyze the outcome of the surgical population of the lumbar PS and propose the involvement of L5-S1 as a risk factor for adverse outcomes of PS.

Conclusion

Involvement of L5-S1, MRSA infection, polymicrobial infection, and CCI were identified as independent risk factors for adverse outcomes after surgical management of lumbar PS. Because the L5-S1 level is anatomically difficult to access anteriorly, judicious access and meticulous debridement are recommended in PS patients requiring anterior debridement of L5-S1.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Sung Cheol Park

Sam Yeol Chang

References

Cheung

Luk

. Pyogenic spondylitis. Int Orthop 2012; 36: 397–404.

Carragee

. Pyogenic vertebral osteomyelitis. JBJS 1997; 79: 874–880.

Guerado

Cerván

AM.

Surgical treatment of spondylodiscitis. An update. Int Orthop 2012; 36: 413–420.

Tyrrell

Cassar-Pullicino

McCall

. Spinal infection. Eur Radiol 1999; 9: 1066–1077.

Pandita

Paul

Yadav

, et al. Evaluation of challenges in diagnosis of spontaneous subacute pyogenic spondylodiscitis in immunocompetent patients: experiences from a tertiary care center. Asian Spine J 2019; 13: 621–629.

Kaufman

Kaplan

Litman

. Infectious agents in spinal epidural abscesses. Neurology 1980; 30: 844–850.

Yee

Samartzis

Wong

Y-W

, et al. Infective spondylitis in Southern Chinese: a descriptive and comparative study of ninety-one cases. Spine 2010; 35: 635–641.

Hadjipavlou

Mader

Necessary

, et al. Hematogenous pyogenic spinal infections and their surgical management. Spine 2000; 25: 1668–1679.

Rutges

Kempen

van Dijk

, et al. Outcome of conservative and surgical treatment of pyogenic spondylodiscitis: a systematic literature review. Eur Spine J 2016; 25: 983–999.

10.

Zarghooni

Röllinghoff

Sobottke

, et al. Treatment of spondylodiscitis. Int Orthop 2012; 36: 405–411.

11.

O’Daly

Morris

O’Rourke

. Long-term functional outcome in pyogenic spinal infection. Spine 2008; 33: E246–E253.

12.

Butler

Shelly

Timlin

, et al. Nontuberculous pyogenic spinal infection in adults: a 12-year experience from a tertiary referral center. Spine 2006; 31: 2695–2700.

13.

Kotani

Ikeura

Tokunaga

, et al. Single-level controlled comparison of OLIF51 and percutaneous screw in lateral position versus MIS-TLIF for lumbosacral degenerative disorders: clinical and radiologic study. J Orthop Sci 2020. Online ahead of print..

14.

Choi

Rhee

Ruparel

. Assessment of great vessels for anterior access of L5/S1 using patient positioning. Asian Spine J 2020; 14: 438–444.

15.

Okada

Miyamoto

Uno

, et al. Clinical and radiological outcome of surgery for pyogenic and tuberculous spondylitis: comparisons of surgical techniques and disease types. J Neurosurg Spine 2009; 11: 620–627.

16.

Chen

S-H

Lee

C-H

Huang

K-C

, et al. Postoperative wound infection after posterior spinal instrumentation: analysis of long-term treatment outcomes. Eur Spine J 2015; 24: 561–570.

17.

Wille

Dauchy

F-A

Desclaux

, et al. Efficacy of debridement, antibiotic therapy and implant retention within three months during postoperative instrumented spine infections. Infect Dis 2017; 49: 261–267.

18.

Maruo

Berven

. Outcome and treatment of postoperative spine surgical site infections: predictors of treatment success and failure. J Orthop Sci 2014; 19: 398–404.

19.

Harris

Reichmann

Bono

, et al. Mortality in elderly patients after cervical spine fractures. J Bone Joint Surg Am 2010; 92: 567–574.

20.

Bae

Kim

C-J

Kim

, et al. Concordance of results of blood and tissue cultures from patients with pyogenic spondylitis: a retrospective cohort study. Clin Microbiol Infect 2018; 24: 279–282.

21.

Kim

Bae

Kim

S-E

, et al. Comparison of pyogenic postoperative and native vertebral osteomyelitis. Spine J 2019; 19: 880–887.

22.

Yoon

Chung

Kim

K-J

, et al. Pyogenic vertebral osteomyelitis: identification of microorganism and laboratory markers used to predict clinical outcome. Eur Spine J 2010; 19: 575–582.

23.

Abdul-Jabbar

Berven

, et al. Surgical site infections in spine surgery: identification of microbiologic and surgical characteristics in 239 cases. Spine 2013; 38: E1425–E1431.

24.

Jiménez-Mejías

de Dios Colmenero

Sánchez-Lora

, et al. Postoperative spondylodiskitis: etiology, clinical findings, prognosis, and comparison with nonoperative pyogenic spondylodiskitis. Clin Infect Dis 1999; 29: 339–345.

25.

Park

K-H

Cho

O-H

Lee

Y-M

, et al. Therapeutic outcomes of hematogenous vertebral osteomyelitis with instrumented surgery. Clin Infect Dis 2015; 60: 1330–1338.

26.

Park

K-H

Cho

O-H

Lee

, et al. Optimal duration of antibiotic therapy in patients with hematogenous vertebral osteomyelitis at low risk and high risk of recurrence. Clin Infect Dis 2016; 62: 1262–1269.

27.

Luzzati

Giacomazzi

Danzi

, et al. Diagnosis, management and outcome of clinically-suspected spinal infection. J Infect 2009; 58: 259–265.

28.

Bhagat

Mathieson

Jandhyala

, et al. Spondylodiscitis (disc space infection) associated with negative microbiological tests: comparison of outcome of suspected disc space infections to documented non-tuberculous pyogenic discitis. Br J Neurosurg 2007; 21: 473–477.

29.

Molinares

Davis

Fung

. Retroperitoneal oblique corridor to the L2–S1 intervertebral discs: an MRI study. J Neurosurg Spine 2016; 24: 248–255.

30.

DiPaola

Saravanja

Boriani

, et al. Postoperative infection treatment score for the spine (PITSS): construction and validation of a predictive model to define need for single versus multiple irrigation and debridement for spinal surgical site infection. Spine J 2012; 12: 218–230.

31.

Sapico

Montgomerie

. Pyogenic vertebral osteomyelitis: report of nine cases and review of the literature. Rev Infect Dis 1979; 1: 754–776.

32.

Pavlaki

Poulakou

Drimousis

, et al. Polymicrobial bloodstream infections: epidemiology and impact on mortality. J Glob Antimicrob Resist 2013; 1: 207–212.

33.

C-H

Hsein

Y-C

Y-L

, et al. Clinical predictors and outcome impact of community-onset polymicrobial bloodstream infection. Int J Antimicrob Agents 2019; 54: 716–722.

34.

Sancho

Artero

Zaragoza

, et al. Impact of nosocomial polymicrobial bloodstream infections on the outcome in critically ill patients. Eur J Clin Microbiol Infect Dis 2012; 31: 1791–1796.

35.

Castle

Uyemura

Rafi

, et al. Comorbidity is a better predictor of impaired immunity than chronological age in older adults. J Am Geriatr Soc 2005; 53: 1565–1569.

36.

Yoshikawa

Ouslander

. Infection management for geriatrics in long-term care facilities. Boca Raton, FL: CRC Press, 2002.

37.

Shen

Silverstein

Roth

. In-hospital complications and mortality after elective spinal fusion surgery in the united states: a study of the nationwide inpatient sample from 2001 to 2005. J Neurosurg Anesthesiol 2009; 21: 21–30.

38.

Whitmore

Stephen

Vernick

, et al. ASA grade and Charlson Comorbidity Index of spinal surgery patients: correlation with complications and societal costs. Spine J 2014; 14: 31–38.

39.

Rawlings

III Wilkins

Gallis

, et al. Postoperative intervertebral disc space infection. Neurosurgery 1983; 13: 371–376.

40.

Mok

Guillaume

Talu

, et al. Clinical outcome of deep wound infection after instrumented posterior spinal fusion: a matched cohort analysis. Spine 2009; 34: 578–583.