The diagnosis of spine infection can be challenging on either a clinical basis or even with highly sensitive imaging studies such as magnetic resonance imaging (MRI). With back pain as a very ubiquitous patient complaint, it is not uncommon to encounter clinical situations in which the imaging study shows findings that resemble, but are not, spine infection. The diagnosis of spine infection cannot always be solely based upon imaging findings. It is in the context of this clinical management dilemma that this chapter intends to describe and analyze diagnostic entities that may mimic the imaging presentation of spine infection. In general, the radiologic differential diagnosis for spine infection includes degenerative disk disease, diskectomy, spondyloarthropathy, and, less commonly, trauma and neoplasm. The findings of some of these so-called spine infection mimics simulate the appearance of infection so closely that, in many instances, a spine biopsy is required to exclude the possibility of spine infection. It is important to understand that the clinical management of these entities is different than what would be seen with spine infection. Hence, a spine biopsy may be required in order to facilitate arriving at the correct clinical diagnosis.

This condition is encountered in the majority of spine imaging examinations that are performed in the adult population. The frequency and severity of imaging findings tend to increase with the age of the patient. Given the increased weight-bearing stress that is placed upon the lumbar spine, degenerative disk disease is often seen within the lower lumbar intervertebral disk spaces, including L4–L5 and L5–S1.One or more of the intervertebral disk space levels may be affected. Disk space narrowing at the affected level is usually present. Anterior spur formation may be seen particularly in elderly patients. A disk bulge, protrusion, or extrusion may also be present. Fibrovascular changes in the adjacent vertebral endplates result in varying amounts of T1 hypointensity and T2 hyperintensity within the affected endplates, the so-called type 1 change (Modic et al. 1988). These MR signal changes may occur in a horizontal band-like or triangular pattern. Because there is altered vascularity within the endplates, contrast enhancement is observed and is more pronounced on fat-suppressed T1-weighted MR images. Intradiskal post-contrast enhancement is distinctly absent. Another pattern (type 2) is due to fat deposition within the affected vertebral endplates resulting in T1 hyperintensity and T2 iso- or hypointensity with no associated contrast enhancement. The degenerative cascade ultimately progresses to osseous growth within the vertebral endplates that is manifested as sclerosis on radiographic or CT examinations and on MRI is hypointense on both T1 and T2 sequences (type 3 change). Thus, it is the type 1 reactive endplate change in degenerative disk disease that can at times be confused for spine infection, especially when the reactive change is exuberant (Fig. 10.1). The disk-endplate margin remains relatively intact in early reactive endplate changes which may help to distinguish this entity from spondylodiskitis. A focal diffusion pattern, or “claw” sign, on MR diffusion-weighted sequences is reported to be present in type 1 reactive endplate change (Fig. 10.1) (Patel et al. 2014). Some authors postulate that disk infection with Propionibacterium acnes bacteria, a low-virulence skin microbe, may play a role in the development of these reactive endplate changes and have reported improvement in low back pain and disability following treatment with antibiotics in a double-blind randomized clinical trial (Albert et al. 2013). In the presence of aggressive reactive endplate changes in patients with degenerative disk disease, if there is a reasonable clinical suspicion for spine infection, then the possibility of a spine biopsy may be considered.

A Schmorl’s node is a focal prolapse of disk material (annulus fibrosis and/or nucleus pulposus) through the superior or inferior vertebral endplate into the vertebral body. Schmorl’s nodes have a bimodal age distribution and are seen in young adolescents and patients in their sixth decade of life and beyond (Stabler et al. 1997). Males are affected much more frequently (76% of cases) than females (Mattei and Rehman 2014). The etiology of this acquired condition is variable and includes trauma, degenerative disk disease, and metabolic and neoplastic disorders. Schmorl’s nodes can be found anywhere along the spinal axis, but are most frequently encountered within the lower thoracic spine and the upper lumbar spine. They are most commonly located within the mid-portion of vertebral body (Mattei and Rehman 2014). When trauma is the precipitating event, an axial loading mechanism is responsible for Schmorl’s node formation (Wagner et al. 2000a). The other etiologic entities are thought to weaken the vertebral endplate, subsequently leading to intraosseous extension of the disk material. Schmorl’s nodes can be associated with focal axial back pain that localizes to the vertebral body level of the Schmorl’s node. Alternatively, they may be incidental findings that are seen during spine imaging examinations performed for other clinical indications.

Regardless of the etiology and the patient’s clinical presentation, Schmorl’s node imaging presentation on MR examinations can be quite striking and may even simulate infection. The prolapsed disk material elicits an inflammatory response in the surrounding bone (Fig. 10.2). This manifests as a zone or band of T1 hypointensity and T2 hyperintensity. There may be alternating zones of inflammation with Schmorl’s node at the center, which may create a target-like appearance. Since the vertebral endplates are vascularized structures, the inflammatory reaction surrounding Schmorl’s node may create a vascular response at the margin of Schmorl’s node. This zone of vascular proliferation will show enhancement on contrast-enhanced MRI studies (Stabler et al. 1997). To a certain extent, there is an overlap in this type of reactive endplate process and what is seen in the degenerative disk reactive endplate change cascade. MR may show Schmorl’s nodes with surrounding fatty replacement or sclerotic reaction (which can also be seen on plain radiographs or CT examinations). Additionally, the two processes, degenerative disk disease and Schmorl’s node formation, may be superimposed. But when there is prominent signal alteration and enhancement within and about the affected disk and Schmorl’s node, the MR imaging findings may mimic spondylodiskitis.

Fortunately, in the case of Schmorl’s nodes, there are a few imaging findings that may indicate the diagnosis. First, the preponderance of pathologic findings in Schmorl’s node formation is within the affected vertebral endplate and the adjacent portion of the vertebral trabeculae and marrow. This is where the corresponding MR signal and enhancement findings will be encountered with Schmorl’s node formation. In contrast, disk space infection shows signal and enhancement abnormalities within the disk. A target-like pattern of alternating zones of signal change with peripheral enhancement adjacent to the vertebral endplate as seen on multiplanar MR images may also indicate the presence of Schmorl’s node. The focal location within the mid-portion of the vertebral endplate may also suggest the presence of Schmorl’s node. Lastly, Schmorl’s nodes can occur at multiple vertebral levels within the same patient and may also show some variation in MR signal and enhancement characteristics. This constellation of imaging findings with respect to MR signal and enhancement pattern, location, and lesion multiplicity may strongly suggest the diagnosis and obviate the requirement for spine biopsy.

The MR imaging appearance in patients who have recently undergone a diskectomy or disk injection procedure can have features that overlap those that are seen with disk space infection (Fig. 10.3). Patients, who have undergone other spine surgical procedures whether open or percutaneous, may also demonstrate postsurgical imaging findings that can overlap with those of spine infection (Fig. 10.4). The diskectomy procedure is being performed on a disk that has already been injured; fissuring of the annulus fibrosis is associated with protrusion or extrusion of disk material. The herniated disk fragment is resected, a mechanical excision that further disturbs the normal architecture of the disk and vertebral endplate. Therefore, it is expected that the disk signal on T1- and T2-weighted images will be altered as compared to nearby normal disks. Additionally, contrast-enhanced MRI will show focal enhancement at the surgical site. Two thin bands of linear enhancement that extend within the disk annulus parallel to the vertebral endplates, with or without enhancement in the endplates, have been reported in 20% of asymptomatic post-diskectomy patients (Ross et al. 1996). It is when this pattern of signal and enhancement change starts to extend beyond the surgical site that the possibility of infection should be considered (Bittane et al. 2014; Boden et al. 1992).