Imaging characteristics and CT sensitivity for pyogenic spinal infections.

Background/objective: Contrast-enhanced magnetic resonance imaging (MRI) is the preferred imaging modality for diagnosing pyogenic spinal infection (PSI), but it is not always available. Our objective was to describe pyogenic spinal infection imaging characteristics in patients presenting to a community emergency department (ED) and estimate the computed tomography (CT) sensitivity for these infections.

Methods: We examined the MRI reports from a cohort of 88 PSI patients whom we enrolled in a prospective cohort study and report the prevalence of each PSI type (spinal epidural abscess/infection, vertebral osteomyelitis/discitis, septic facet, and paravertebral abscess/infection) according to contemporary nomenclature. In a 14 patient subcohort who underwent both CT and MRI studies, we report the sensitivity for each PSI from a post hoc blinded overread of the CT imaging by a single neuroradiologist.

Results: Of the 88 PSI patients, the median age was 55 years, and 31% were female. The PSI prevalence included: spinal epidural abscess/infection (SEA) in 61(69%), vertebral osteomyelitis/discitis (VO/D) in 54 (61%), septic facet (SF) in 15 (17%), and paravertebral abscess/infection (PVA) in 53 (60%). Of the SEAs, 82% (50/61) were associated with other spinal infections, while 18% (11/61) were isolated SEAs. The overall CT sensitivity in a masked overread was 79% (11/14) for any PSI, 83% (10/12) for any infection outside the spinal canal, and only 18% (2/11) for SEA.

Conclusion: Patients found to have vertebral osteomyelitis/discitis, septic facet, and paravertebral infections frequently had a SEA coinfection. CT interpretation by a neuroradiologist had moderate sensitivity for infections outside the spinal canal but had low sensitivity for SEA.