Vertebral osteomyelitis represents approximately 2% to 7% of all reported cases of osteomyelitis. This disease involves the vertebrae and the associated disk spaces and is caused by many types of microorganisms, with Staphylococcus aureus being responsible for the majority of vertebral osteomyelitis.1–3 Although vertebral osteomyelitis remains uncommon, the incidence of this disease is rising due to an increasing number of patients with predisposing factors such as advanced age, diabetes mellitus, chronic renal or liver disease, intravenous drug use, human immunodeficiency virus (HIV) infection, chronic corticosteroid use, chemotherapy, and severe trauma.4,5
The treatment of vertebral osteomyelitis can be challenging for spine surgeons. Most cases of vertebral osteomyelitis are managed initially by needle aspiration biopsy from the intervertebral disk space or vertebral bone, which has a sensitivity ranging from a low of 30% to 50% to a high of 70% to 90% when used to identify the pathogen.6 In cases without neurologic deficits and no marked kyphotic deformity or instability, spinal infections can be managed without surgical intervention.1 Even if the majority of patients with vertebral osteomyelitis respond to medical treatment, up to 40% of patients suffering from this disease eventually will require surgical intervention.1,7–10 Surgery is indicated in cases with failure of prolonged medical management, delayed treatment or complications of sepsis, neurologic impairment, residual vertebral destruction leading to early or late spinal instability, or segmental kyphosis with intractable pain.1,6,7,9–15
The purpose of this study was to evaluate the clinical outcomes of patients treated for vertebral osteomyelitis using instrumented versus noninstrumented (decompression) surgical procedures. Specifically, the following criteria were compared: (1) mortality rates, (2) infection clearance rates, (3) clinical outcomes measured by Oswestry Disability Index (ODI), (4) mean length of stay (LOS), and (5) baseline differences between the 2 cohorts.
Materials and Methods
The study population included all patients who had a discharge diagnosis of spinal osteomyelitis and were treated at the authors' institution between January 2004 and March 2012. Patients were identified using International Classification of Diseases, 9th Revision (ICD-9) codes that included osteomyelitis (730.28, 730.08, 730.2, and 730.00), diskitis (722.90, 722.91, 722.92, and 722.93), and epidural abscess (324.1 and 234.9) of the spine. Inclusion criteria involved appropriate initial imaging, laboratory results, evaluation by the orthopedic department, and no treatment from an outside institution prior to admission. Patients who were treated nonoperatively were excluded.
A total of 104 patients (62 men and 42 women) met these criteria and were included in the study. Mean patient age was 59 years (range, 20–93 years). Indications for laminectomy were epidural abscesses requiring drainage and decompression. All patients underwent follow-up for a mean of 43 months (range, 24–72 months). Patient demographics are summarized in Table 1. Inpatient hospital charts and medical records were reviewed. Clinical information and patient demographics including age, gender, use of instrumentation during surgery, and clearance of infection, as well as mortality directly related to the osteomyelitis, were recorded. Appropriate review board approval for this study was obtained.
Patients were stratified by the type of surgical treatment received to either the noninstrumented (decompression) group or the instrumented group. Noninstrumented surgeries included laminectomy and debridement; instrumented surgeries included anterior spine instrumentation, posterior spine instrumentation, or combined anterior and posterior spine instrumentations. Noninstrumented surgeries were all decompressions and did not include noninstrumented fusions. Titanium cages were used for all corpectomies. Allografts included crushed cancellous bone and demineralized bone matrix. Location and surgical approach are summarized in Table 2.
Location of Osteomyelitis and Surgical Approach
The decision to perform the type of treatment option was according to the on-call spine surgeon's evaluation of the patient's radiographs, neurologic presentation, and medical comorbidities. Decompression cases did not require instrumentation based on the radiographic and clinical evaluations. Patients who had spinal instability were treated with instrumented surgery if possible. Overall, 57 patients received surgical instrumentation and 47 received noninstrumented surgery.
The size of infection was defined as the total volume of infected bone and soft tissues measured on coronal, sagittal, and axial magnetic resonance images (MRI). The MRIs were used to calculate the volume of infection by multiplying the length, width, and height of the total area of infection on sagittal and coronal planes and dividing it by 2. A volume of 8 cm3 was used as the cutoff for a large infection.
Mortality was defined as mortality directly related to the osteomyelitis itself within 1 year of presentation. Two experienced fellowship-trained spine surgeons and 1 musculoskeletal fellowship-trained radiologist reviewed all of the clinical and radiographic data. In all cases, consensus was reached if the patient cleared the infection.
All data were recorded using Excel (Microsoft Corporation, Redmond, Washington). Statistical analysis was performed using t test, Fisher exact test, and odds ratio (OR) to compare infection clearance rate, mortality rate, and clinical outcomes for patients who had received instrumented and noninstrumented surgery. A P value of less than .05 was used as the threshold for significance.
Kaplan-Meier survival analysis demonstrated no significant difference in the 1- and 5-year mortality rates due to vertebral osteomyelitis between the instrumented and noninstrumented cohorts (89% vs 86%; P=.17) (Figure). The 1-year and overall mortality rate due to vertebral osteomyelitis in the instrumented cohort was 9% (n=5 of 57), which was lower but not significantly different from the 17% rate (n=8 of 47) in the noninstrumented group (P=.21). When evaluating the overall mortality due to any reason other than osteomyelitis, there also was no significant difference in the 5-year mortality rates between the instrumented and noninstrumented cohorts (18% vs 13%; OR, 1.3; 95% confidence interval [CI], 0.45–3.9; P=.58). For the Charleston comorbidity index, greater than 6 points was associated with a higher mortality OR due to vertebral osteomyelitis (OR, 4.12; 95% CI, 0.49–34.5; P=.19).
Kaplan-Meier survival analysis of mortality due to vertebral osteomyelitis (VO) in instrumented and noninstrumented patients.
There were no significant differences in the clearance of infection between patients who were treated with surgical instrumentation and those who were treated without instrumentation (OR, 1.55; 95% CI, 0.61–3.9; P=.35). The clearance of infection in the cohort of patients who underwent surgical instrumentation was 54% (n=31 of 57); this rate was higher than in patients who underwent decompression alone but not significantly different from the clearance rate of 42.5% (n=20 of 47) in the noninstrumented group (P=.35) (Table 3).
There were no significance differences in mean ODI score between the instrumented and noninstrumented groups (P=.32). In the instrumented group, mean ODI score improved from 62 points (range, 12–94 points) preoperatively to 40 points (range, 0–80 points) postoperatively. In the noninstrumented group, mean ODI score improved from 75 points (range, 50–100 points) preoperatively to 45 points (range, 12–94 points) postoperatively.
Mean LOS was significantly higher in the instrumented group (P=.02). Mean LOS in the instrumented cohort was 19 days (range, 3–75 days). This was significantly longer than the mean LOS of 13 days (range, 2–63 days) in the noninstrumented group (P=.02) (Table 3).
Overall, patients in the instrumented surgery group had more severe cases of vertebral osteomyelitis as they had a higher OR of neurologic deficiency (OR, 4.6; 95% CI, 2.3–9.3; P<.01) and instability (OR, 2.8; 95% CI, 1.1–6.9; P=.03) compared with patients in the noninstrumented group. Patients in the noninstrumented group had significantly smaller infections (29%, n=11) than patients in the instrumented group (51%, n=19) (OR, 0.39; 95% CI, 0.5–1.0; P=.047]. Overall, the instrumentation averaged 3.7 levels of surgery, with 61% undergoing a corpectomy. The mean number of corpectomies in the decompression cohort was 1.84 levels. The thoracolumbar instrumented cases averaged 2.7 levels, with 45% having corpectomies averaging 2.2 levels.
The choice of surgical techniques, instrumentation, and the decision to stage the instrumentation portion of the surgery when treating vertebral osteomyelitis are matters of controversy.1 Options include anterior or posterior approaches, to instrument or not to instrument, and single-stage or 2-stage surgery. The decision regarding the surgical approach and technique should always be guided by the neurologic status and mechanical spinal instability.1,16 In addition, whether instrumentation affects the infection recurrence rate is another consideration. The reported infection recurrence rates after nonsurgical treatment have ranged from 0% to 25%, whereas recurrence rates ranging from 2% to 18% have been reported after the surgical treatment of vertebral osteomyelitis.7,8,10,16,17 However, there is a paucity of published studies that have compared clinical outcomes of surgical noninstrumented cases to instrumented cases. The purpose of this study was to evaluate the efficacy of different surgical treatment modalities for vertebral osteomyelitis including clearance of infection rate, clinical outcomes, and mortality rate.
There were several limitations of this study. A prospective study could have reduced potential selection and information biases. Broader quality of life measures such as patient activity or satisfaction were not evaluated. Clearance of infection was determined as resolution on imaging and serologic tests, which may not be the most accurate method for assessing clearance of infection. Neurologic outcomes in patients with preoperative neurologic deficit were not reported. Immune status, HIV status, smoking history, and the types of organisms for the patients in this study were not evaluated. However, the authors believe the outcomes are valuable because this was a large series of surgical cases with a study design superior to other studies since it compared instrumented to noninstrumented surgical cases, as opposed to comparing surgical cases to nonoperative cases.
No significant differences were noted in the mortality rate between the instrumented and noninstrumented cohorts. It would appear with the more aggressive osteomyelitis cases occurring in the instrumented cohort, a greater insult to the patient occurs with the metabolic demand and sepsis from the osteomyelitis. Many of the patients in the instrumented cohort had radical debridement, corpectomy, or both radical debridement and corpectomy. Therefore, it is possible that eradicating a large portion of the infected bone and soft tissues in the instrumented cases resulted in the patients having good outcomes. Thus, it is encouraging that even in appropriately addressed cases of severe osteomyelitis with surgery and instrumentation, clinical outcomes are comparable to less severe cases treated without instrumentation.
These findings are comparable to the report by Rayes et al,18 in which 47 patients (32 men and 15 women) with vertebral osteomyelitis underwent instrumentation procedures. In their study, mean LOS was 25 days (range, 9–78 days), and the mortality rate was 17% at a mean follow-up of 22 months (range, 1–80 months); however, there was no comparison cohort. Although the 1-year mortality of 9% in the cohort in the current study may be interpreted as high, many of the patients were noncompliant and elderly with multiple comorbidities. In addition, the mean LOS in the current study was high as these patients often had multiple medical comorbidities, and their hospital stay was prolonged due to medical management of their comorbidities.
In the current study, instrumented surgery for vertebral osteomyelitis resulted in similar clearance rates to noninstrumented cases. These outcomes are comparable to previously reported studies.7,8,10,16,17 In their case series of instrumented vertebral osteomyelitis, Rayes et al18 reported 2 recurrent infections (4.3%). The current study attempted to validate the safety of instrumentation in regard to recurrence of vertebral osteomyelitis, and it is believed that the radical surgical debridement in the instrumented cohort led to low infection recurrence rates. Therefore, instrumentation helped stabilize the spine with acceptable recurrence rates of the infection.
A higher number of patients who had mechanical instability or neurologic deficits underwent surgical instrumentation in the current study. One of the primary reasons for instrumentation in vertebral osteomyelitis is to stabilize an unstable spine due to the pervasive nature of the osteomyelitis. Another reason is iatrogenic instability as a result of the radical surgical debridement and corpectomy. The prevalence of preoperative neurologic deficits in the instrumented cohort is multifactorial. With severe collapse and spinal instability, as well as sagittal (coronal) misalignment, the involved vertebral levels have greater impingement of the theca sac and consequent neurologic deficits.
Surgical instrumentation to stabilize the spine, even in more severe cases of vertebral osteomyelitis, appeared to result in comparable clinical outcomes to decompression alone. Although patients who underwent instrumentation had more medical comorbidities and a higher rate of preoperative neurologic deficits and spinal instability, it is encouraging that they achieved similar clinical outcomes and mortality rates as the noninstrumented patients with less severe disease. In appropriately selected cases, adding spinal instrumentation to the surgical treatment of vertebral osteomyelitis does not appear to compromise patient outcome.
- Pola E, Logroscino CA, Gentiempo M, et al. Medical and surgical treatment of pyogenic spondylodiscitis. Eur Rev Med Pharmacol Sci. 2012; 16(suppl 2):35–49.
- Tyrrell PN, Cassar-Pullicino VN, McCall IW. Spinal infection. Eur Radiol. 1999; 9(6):1066–1077. doi:10.1007/s003300050793 [CrossRef]
- Stabler A, Reiser MF. Imaging of spinal infection. Radiol Clin North Am. 2001; 39(1):115–135. doi:10.1016/S0033-8389(05)70266-9 [CrossRef]
- Cottle L, Riordan T. Infectious spondylodiscitis. J Infect. 2008; 56(6):401–412. doi:10.1016/j.jinf.2008.02.005 [CrossRef]
- Carragee EJ. Pyogenic vertebral osteomyelitis. J Bone Joint Surg Am. 1997; 79(6):874–880.
- Hsieh PC, Wienecke RJ, O'Shaughnessy BA, Koski TR, Ondra SL. Surgical strategies for vertebral osteomyelitis and epidural abscess. Neurosurg Focus. 2004; 17(6):E4. doi:10.3171/foc.2004.17.6.4 [CrossRef]
- Kuklo TR, Potter BK, Bell RS, Moquin RR, Rosner MK. Single-stage treatment of pyogenic spinal infection with titanium mesh cages. J Spinal Disord Tech. 2006; 19(5):376–382. doi:10.1097/01.bsd.0000203945.03922.f6 [CrossRef]
- Hadjipavlou AG, Mader JT, Necessary JT, Muffoletto AJ. Hematogenous pyogenic spinal infections and their surgical management. Spine (Phila Pa 1976). 2000; 25(13):1668–1679. doi:10.1097/00007632-200007010-00010 [CrossRef]
- Ruf M, Stoltze D, Merk HR, Ames M, Harms J. Treatment of vertebral osteomyelitis by radical debridement and stabilization using titanium mesh cages. Spine (Phila Pa 1976). 2007; 32(9):E275–E280. doi:10.1097/01.brs.0000261034.83395.7f [CrossRef]
- Korovessis P, Repantis T, Iliopoulos P, Hadjipavlou A. Beneficial influence of titanium mesh cage on infection healing and spinal reconstruction in hematogenous septic spondylitis: a retrospective analysis of surgical outcome of twenty-five consecutive cases and review of literature. Spine (Phila Pa 1976). 2008; 33(21):E759–E767. doi:10.1097/BRS.0b013e318187875e [CrossRef]
- Ogden AT, Kaiser MG. Single-stage debridement and instrumentation for pyogenic spinal infections. Neurosurg Focus. 2004; 17(6):E5. doi:10.3171/foc.2004.17.6.5 [CrossRef]
- Deininger MH, Unfried MI, Vougioukas VI, Hubbe U. Minimally invasive dorsal percutaneous spondylodesis for the treatment of adult pyogenic spondylodiscitis. Acta Neurochir (Wien). 2009; 151(11):1451–1457. doi:10.1007/s00701-009-0377-3 [CrossRef]
- Strowitzki M, Vastmans J, Vogel M, Jaksche H. Complex 360°-reconstruction and stabilization of the cervical spine due to osteomyelitis. Eur Spine J. 2011; 20(suppl 2):S248–S252. doi:10.1007/s00586-010-1645-x [CrossRef]
- Acosta FL Jr, Chin CT, Quinones-Hinojosa A, Ames CP, Weinstein PR, Chou D. Diagnosis and management of adult pyogenic osteomyelitis of the cervical spine. Neurosurg Focus. 2004; 17(6):E2. doi:10.3171/foc.2004.17.6.2 [CrossRef]
- Pee YH, Park JD, Choi YG, Lee SH. Anterior debridement and fusion followed by posterior pedicle screw fixation in pyogenic spondylodiscitis: autologous iliac bone strut versus cage. J Neurosurg Spine. 2008; 8(5):405–412. doi:10.3171/SPI/2008/8/5/405 [CrossRef]
- Lee MC, Wang MY, Fessler RG, Liauw J, Kim DH. Instrumentation in patients with spinal infection. Neurosurg Focus. 2004; 17(6):E7. doi:10.3171/foc.2004.17.6.7 [CrossRef]
- Fang D, Cheung KM, Dos Remedios ID, Lee YK, Leong JC. Pyogenic vertebral osteomyelitis: treatment by anterior spinal debridement and fusion. J Spinal Disord. 1994; 7(2):173–180. doi:10.1097/00002517-199407020-00012 [CrossRef]
- Rayes M, Colen CB, Bahgat DA, et al. Safety of instrumentation in patients with spinal infection. J Neurosurg Spine. 2010; 12(6):647–659. doi:10.3171/2009.12.SPINE09428 [CrossRef]
|Demographic||Instrumented Group||Noninstrumented Group||P|
|Total patients, No.||62||42||-|
|Age, mean (range), y||57.5 (29.7–85.0)||59 (21.0–82.5)||.59|
|Men to women, ratio, No.||1:6||1:2||.54|
|Comorbidity index, mean (range)||3.7 (0–10)||3.1 (0–8)||.34|
|Preoperative neurologic deficiency||63%||27%||.01|
Location of Osteomyelitis and Surgical Approach
|Location and Surgical Approach||Instrumented Group||Noninstrumented Group|
|Anterior spine instrumentation||58%||-|
|Posterior spine instrumentation||13%||-|
|Anterior and posterior spine instrumentationa||29%||-|
|Outcome||Instrumented Group||Noninstrumented Group||P|
|Nonosteomyelitis-related 5-year mortality||18%||13%||.58|
|Clearance of infection||54%||42.5%||.35|
|Preoperative ODI score, mean (range), points||62 (12–94)||75 (50–100)||.0175|
|Postoperative ODI score, mean (range), points||40 (0–80)||45 (20–76)||.32|
|Length of stay, mean (range), d||19 (3–75)||13 (2–63)||.02|