Radiological Diagnosis of Spinal Tuberculosis (Pott Disease)

Radiological Hallmarks

The classic imaging pattern of spinal tuberculosis consists of destruction of two or more contiguous vertebrae with involvement of their opposing endplates, spread of infection along the anterior longitudinal ligament, and disc involvement that may be accompanied by paraspinal “cold” abscesses; a less common presentation is spondylitis without disc involvement. 1

Magnetic resonance imaging of the spine performed with and without intravenous contrast is the preferred diagnostic test, achieving approximately 96 % sensitivity, 94 % specificity, and 92 % overall accuracy for detecting spinal tuberculosis. 2

T1‑weighted MRI sequences are more sensitive than T2‑weighted sequences for visualizing inflammatory changes within vertebral bodies in tuberculous spondylitis. 1
Plain radiographs lag behind clinical symptoms by 3–6 weeks and only reveal bone destruction in advanced disease, limiting their usefulness for early diagnosis. [1][2]
Non‑contrast computed tomography provides excellent detail of bony structures but fails to detect marrow edema and soft‑tissue disease; therefore it should be used as an adjunct to, not a replacement for, MRI. 2

Step 1: Obtain MRI of the spine with and without IV contrast as the first‑line imaging study in any patient with subacute back pain who has epidemiologic risk factors for tuberculosis. 2
Step 2: Perform a tuberculin skin test (PPD) or an interferon‑γ release assay— the latter offering higher sensitivity, especially in immunocompromised individuals—when characteristic radiologic findings are present. 1
Step 3: If MRI cannot be performed (e.g., due to contraindications), acquire a combined gallium‑67/Tc‑99m bone scan, which provides roughly 78 % sensitivity and 81 % specificity for spinal infection. [1][3]

Tuberculous spondylitis typically shows early relative preservation of intervertebral disc height, subligamentous spread of disease, larger paraspinal abscesses, and involvement of multiple vertebral levels, distinguishing it from pyogenic infection. 1
In contrast, pyogenic spondylitis more often presents with rapid disc destruction, smaller abscesses, and single‑level involvement. 1

MRI is the imaging modality of choice for diagnosing spinal tuberculosis, with a sensitivity of 96%, specificity of 94%, and accuracy of 92% 4, 5
In tuberculous spondylitis, T1-weighted sequences appear more sensitive than T2-weighted sequences in demonstrating inflammatory processes in vertebral bodies 6
Epidural enhancement combined with abnormal lab values is predictive of positive biopsy results for spondylodiscitis 7

The American College of Radiology recommends MRI without and with IV contrast as the first imaging choice in patients with suspected spinal TB 5, 6
If MRI cannot be performed, consider a combination spine gallium/Tc99 bone scan, CT scan, or PET scan 6
In patients with suspected TB spine, perform tuberculin skin test (PPD) or interferon-γ release assay, especially in those from endemic regions 6
MRI findings often lag behind clinical improvement based on clinical and laboratory parameters 7
Diffusion-weighted imaging sequences may help differentiate infectious spondylitis from reactive (Modic type 1) vertebral endplate changes 7

Destruction of 2 or more contiguous vertebrae and their opposed endplates 6
Spread along the anterior longitudinal ligament 6
Disc infection with or without a paraspinal mass or mixed soft tissue fluid collection 6
Less commonly, spondylitis without disc involvement 6