Responsiveness of routine diagnostic tests for vertebral osteomyelitis may be influenced by the infecting organism

Abstract

Background ContextVertebral osteomyelitis (VO) becomes increasingly more prevalent as people age, and it is a condition seen frequently by referral center spine surgeons. It can take as long as 6 months for a proper diagnosis to be made. Staphylococcus aureus (S. aureus) is the most common isolated organism in up to 80% of the affected population. The clinical presentation of vertebral osteomyelitis is typically non specific (back pain), which can make timely diagnosis challenging. Fever is often absent. Serum C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), white blood cell count (WBC) and serum polymorphonuclear leukocyte percentage (PMN%) are traditionally used as first-line tests because of their perceived sensitivity to help diagnose vertebral osteomyelitis. It is not known whether these test values are affected by the infecting organism. PurposeTo determine whether individual first-line diagnostics differed based on infecting organism and whether certain organisms are associated with lower lab values. Additionally, this study sought to determine if VO caused by lower virulent (eg, culture-negative and nonpyogenic organisms) could contribute to delays in treatment due to lack of elevated biomarkers. Study Design/SettingSingle-center retrospective cohort study. Patient SampleWe reviewed clinical data of 133 patients (60% male) diagnosed with VO from 2015–2019 in a US Midwest academic hospital. Outcomes MeasuresPrimary outcome measures included the maximum temperature upon presentation, serum C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), white blood cell count (WBC), and percentage neutrophils during the hospital admission. MethodsInclusion criteria were adult patients diagnosed with vertebral osteomyelitis who underwent blood culture and/or biopsy prior to treatment. All patients enrolled in the study were diagnosed with VO and confirmed via magnetic resonance imaging (MRI). MRI findings associated with VO included destruction of endplates, increased signal in vertebral bodies, and the surrounding disc on T2-weighted imaging were confirmed. The patients had laboratory work up and clinical follow up regardless of positive culture or negative culture. The mean peak inflammatory marker levels were compared among organisms with student's t test. Demographics, comorbidities, and CCI were collected and multivariable logistic regression models were used. Receiver operating characteristic curve analysis was performed to delineate separate, optimal cut offs for CRP, ESR, WBC, and PMN% for patients with culture positive osteomyelitis ResultsPatients’ average age was 57.0±13.7 years with a mean BMI of 30.5±9.70 kg/m2, and a mean Charleston Comorbidity Index (CCI) of 3.17±2.35. Staphylococcus aureus and antibiotic resistant organisms (MRSA and VRE) demonstrated a higher mean CRP and ESR than culture negative, fungal and TB cases. Staphylococcus aureus, antibiotic resistant organisms, and coagulase negative Staphylococcus demonstrated a higher mean WBC than culture negative as well as fungal and TB cases. Staphylococcus aureus, antibiotic resistant organisms, coagulase negative Staphylococcus, and Streptococcus species had a higher mean peak PMN%, than culture negative as well as fungal and TB case. Temperature did not correlate with a diagnosis of osteomyelitis. ConclusionsSerum laboratory markers in the diagnosis of VO appear to be influenced by the infecting organism type. Laboratory values in patients diagnosed with VO with culture negative or non-pyogenic organisms are lower compared to antibiotic resistant and S. aureus organisms. Fever did not correlate with a diagnosis of VO.