Treatment Guideline for Patients with Native Culture-negative Pyogenic Vertebral Osteomyelitis

With the development of evidence-based guidelines for the diagnosis and antibiotic treatment of patients with pyogenic vertebral osteomyelitis, various nonsurgical and surgical treatment strategies have achieved favorable results. However, sufficient administration of appropriate antibiotics is a prerequisite for treatment success, which cannot be guaranteed in patients with culture-negative pyogenic vertebral osteomyelitis. Unfortunately, previous studies on culture-negative pyogenic vertebral osteomyelitis were limited by small patient groups, short follow-up periods, varied treatment protocols, and inconsistently defined clinical endpoints.

Among patients treated according to our center's treatment protocol for culture-negative pyogenic vertebral osteomyelitis, which included patients treated without surgery, with surgery but without spinal instrumentation, as well as patients treated with instrumented surgery, (1) what proportion of patients in each treatment group experienced recurrence (or persistence) of infection, complications of treatment, and death; and (2) what factors were independently associated with recurrent or persistent infection after treatment under this algorithm?

This was a retrospective evaluation of a treatment protocol in use at one center from 2008 to 2020. During that time, we treated 183 patients for culture-negative pyogenic vertebral osteomyelitis. The diagnosis was based on clinical, laboratory, and radiological features excluding disease that presents similar features to pyogenic vertebral osteomyelitis. For those patients, our protocol included three possible approaches: nonsurgical treatment, including the use of empirical antibiotics (nonoperative group, n = 82); spinal decompression without instrumentation (noninstrumented group, n = 41); and spinal decompression with instrumentation (instrumented group, n = 60). The indications for each treatment during the period remained relatively consistent. Nonsurgical treatment including empirical antibiotics was applied to all patients according to the specified antibiotic protocol. Spinal decompression without instrumentation was employed when a patient presented substantial or aggravating neurologic deficits or intractable pain from an abscess. Instrumentation was added in patients with mechanical instability before or after spinal decompression. Minimum follow-up to be included in this study was 1 year, and 91% (252 of 277) of patients were accounted for at that time, with no differential loss to follow-up among the study groups (12, five, and four patients missing from the three groups, respectively, at 1 year). Chart review was performed to ascertain the proportion of patients in each group who experienced recurrence, complications, and death. Factors associated with recurrence were assessed across the entire cohort using a multivariable logistic model. We analyzed 1-year recurrence and mortality rates using the Kaplan-Meier method, and their 95% confidence intervals were calculated using the method by Hosmer and Lemeshow.

At 1 year, the recurrence-free survival rate was 87% (95% confidence interval 78% to 93%) in the nonoperative group, 87% (95% CI 72% to 94%) in the noninstrumented group, and 91% (95% CI 80% to 96%) in the instrumented group. The proportion of patients who experienced a major medical complication in each group was 12% (10 of 82), 10% (4 of 41), and 8% (5 of 60), respectively. At 1 year, the survival rate with patient death as the endpoint was 95% (95% CI 88% to 98%) in the nonoperative group, 95% (95% CI 82% to 99%) in the noninstrumented group, and 97% (95% CI 87% to 99%) in the instrumented group. After controlling for potentially confounding variables including age, medical comorbidities, and anatomical involvement of infection, the following factors were independently associated with increased odds of infection recurrence or persistence: higher Charlson Comorbidity Index (CCI) score (odds ratio 1.6 per point on the CCI [95% CI 1.2 to 2.1]; p = 0.004) and the presence of a psoas abscess (OR 4.7 [95% CI 1.6 to 13.9]; p = 0.005).

Among patients with negative initial nonoperative culture results, spinal decompression and abscess drainage can be used in those with substantial or aggravating neurological deficits or intractable pain caused by an abscess, while additional early spinal instrumentation can be applied upon consideration of their medical comorbidities and the presence of a psoas abscess when mechanical instability is present before or after the spinal decompression. Reasonable clinical results can be expected, regardless of the results from subsequent operative cultures. However, our study results should be replicated by other centers, and further studies that consider individual differences such as bone mineral density and include patients with previous spinal instrumentation or recurrent infection should be performed to establish a more comprehensive treatment protocol.

Level IV, therapeutic study.