Imaging in spinal tuberculosis: A comprehensive radiologic review

INTRODUCTION AND EPIDEMIOLOGY

Tuberculosis (TB) remains a major global health burden, with an estimated 8–10 million new cases annually worldwide.^[1] Approximately 84% of cases are pulmonary TB, while 16% present as extra-pulmonary TB (EPTB).^[2] Although EPTB represents a smaller proportion overall, it contributes substantially to morbidity. Skeletal TB (STB) accounts for nearly 10% of EPTB, and the spine is the most frequently involved site, comprising about 50% of STB cases, making spinal TB (Pott’s disease) the most common form of osteoarticular TB.^[3]

India bears the highest global TB burden, contributing nearly one-quarter of worldwide cases. In India, EPTB constitutes around 16% of new TB cases, mirroring global trends.^[4] Overall, spinal TB constitutes approximately 1–2% of all TB cases globally, with higher estimates reported in endemic regions such as India.^[5] The thoracolumbar junction is most commonly affected,followed by the lumbar and cervical spine, underscoring the clinical and imaging relevance of this entity.

Hence, we attempted to review the current literature through this article and we have discussed the whole topic in question-and-answer format for easy understanding and clinical relevance.

PATHOPHYSIOLOGY

Spinal TB is caused by Mycobacterium TB, typically spreading hematogenously from a primary focus in the lungs or genitourinary tract to the vertebral cancellous bone.^[3] Arterial spread through segmental arteries often leads to paradiscal involvement of contiguous vertebrae, while venous spread through Batson’s plexus may result in central lesions and non-contiguous vertebral involvement.

Risk factors include advanced age, overcrowding, diabetes, drug abuse, and immunocompromised states. Although preventable and treatable, TB remains a global health concern due to rising human immunodeficiency virus (HIV) prevalence, poor living conditions in developing nations, and increased migration.^[6]

It predominantly affects individuals in their productive years, leading to significant socioeconomic burden. While there is no gender predilection, its incidence is rising among the elderly due to immunosuppression, comorbidities, and increased life expectancy.

HOW DO PATIENTS TYPICALLY PRESENT TO US?

Back pain is the most common presentation of spinal TB, with neurological deficits occurring in some cases due to spinal cord involvement. Constitutional symptoms are less common than in pulmonary TB.

WHAT IS THE BEST IMAGING MODALITY AND WHAT ARE THE MERITS AND DEMERITS OF EACH MODALITY?

Plain radiography

Although spinal radiographs are typically the first-line investigation in the evaluation of back pain, radiographic changes in spinal TB become apparent only in advanced stages typically when bone mineral loss has reached approximately one-third. It is only in these later stages that vertebral endplate irregularities, disc space narrowing^[7] [Figure 1a and b], and kyphotic deformities can be appreciated.^[4] However, radiographs are essential in the initial workup which serves as a baseline investigation, which are helpful for further frequent follow-ups. It is recommended to take AP and lateral radiographs.

Close

Figure 1:

(a) Antero-posterior and (b) lateral radiographs of lumbosacral spine demonstrating end plate sclerosis with irregularities and reduction in disc space involving L1- L2 (yellow arrow) (transitional vertebrae as per magnetic resonance imaging with lumbarized S1 vertebra) mimicking degenerative disease.

Export to PPT

Computed tomography (CT)

CT can detect vertebral changes earlier than plain radiographs and provides detailed assessment of bony destruction [Figure 2]. Bone destruction is better appreciated on CT and may present as fragmentary (47% of cases), osteolytic (34%), subperiosteal (30%), or localized destructive with sclerotic margins (10%)^[6] [Figure 3]. CT is also the modality of choice for detecting soft tissue calcifications [Figure 2] which are almost pathognomonic for TB. It is sometimes difficult to distinguish between a soft-tissue calcification versus destroyed bony fragments. CT is valuable for characterizing focal bony lesions and plays a key role in planning image-guided biopsies. It is not the primary imaging modality for spinal TB and is utilized selectively based on clinical indications.

Close

Figure 2:

(a) Computed tomography (CT) sagittal and (b) axial sections - bone window shows the value of CT in demonstrating the lytic pattern of involvement of L5 vertebral body with adjacent sclerosis (white arrow in b). CT can also be helpful in detecting para spinal soft-tissue involvement (white asterix in a). (c) CT axial section showing calcific foci in bilateral psoas muscles (yellow arrows) - which is pathognomonic in tuberculosis.

Export to PPT

Close

Figure 3:

Different patterns of involvement of vertebral body in tuberculosis spine which are observed in computed tomography scan.

Export to PPT

Magnetic resonance imaging (MRI)

MRI is the gold standard for evaluating spinal TB, with a reported sensitivity of ~96% and specificity of ~93%.^[8] It is strongly recommended by the infectious diseases society of America for suspected spinal infections.^[8] MRI enables early detection before overt bone destruction, accurately identifies skip lesions, and provides superior assessment of marrow involvement, soft-tissue disease, epidural extension, and neural compression [Figure 4]. Contrast-enhanced MRI further improves lesion conspicuity and characterization, with a favorable safety profile compared to iodinated contrast.^[9]

Close

Figure 4:

A 43-year-old male presented with backache and fever (a) magnetic resonance (MR) imaging showing classic example of Infective spondylodiscitis. Sagittal T1WI, (b) T2WI and (c) STIR MR images of LS spine demonstrates T1 hypo intensity, T2 and STIR hyperintensities involving L1 and L2 vertebral bodies, consistent with edema along with intervening disc involvement (yellow arrows in a-c). On radiograph the disease mimicked degenerative disc disease.

Export to PPT

Recommended MRI protocol

A comprehensive protocol should include sagittal and axial T1- and T2-weighted sequences, supplemented with Short Tau Inversion Recovery (STIR) for detecting marrow edema and skip lesions, especially during whole-spine screening [Figure 5]. Coronal sequences with a wide field of view extending to the lesser trochanter are valuable for assessing paraspinal and psoas abscesses. Fat-saturated T1-weighted and fat-saturated fluid-attenuated inversion recovery sequences are useful when contrast is contraindicated. Some centers additionally employ 3D T2-weighted sequences for improved spatial resolution and multiplanar reformats.

Close

Figure 5:

Various imaging planes and techniques used in spinal magnetic resonance imaging (MRI). (a) MRI thoracolumbar spine sagittal images of T2, (b) STIR, (c) post contrast fat saturated T1 and (d) Coronal STIR images showing collapse of the vertebral body with posterior convexity (yellow arrows in image a and d) with edematous changes involving the cord (yellow arrow in image b). The involved vertebral bodies (yellow asterisk in image c) show peripheral enhancement with central non enhancing areas. (d) Coronal STIR images show better depiction of para vertebral involvement of psoas muscle bilaterally more so on the right side (white arrow). STIR: Short tau inversion recovery

Export to PPT

Role of contrast

Contrast-enhanced MRI is helpful in differentiating cold abscesses from phlegmon – the former usually showing thin, smooth peripheral enhancement, and the latter demonstrating uniform enhancement [Figure 6].^[3,9] Contrast also improves characterization of spondylitis, discitis, epidural collections, and disease extent, and is useful in treatment follow-up.

Close

Figure 6:

A 45-year-old male presented with fever and low backache. Post contrast axial images of lumbar spine showing psoas abscess on right side (white arrow) and multiple tiny foci of vertebral abscesses showing central non-enhancing areas with peripheral rim enhancement represented (yellow arrowhead). Enhancing lymph node is also seen anteriorly (yellow arrow). No involvement of posterior elements (asterisk).

Export to PPT

Cold abscesses in spinal TB lack typical inflammatory signs and arise from the primary spinal focus. They characteristically spread along paths of least resistance, including fascial planes and neurovascular bundles. Depending on the spinal level, they may present as retropharyngeal abscesses in the cervical region, paraspinal or chest wall swellings in the thoracic spine, and swellings in the Petit’s triangle or groin in lumbar involvement.

Dynamic contrast-enhanced MRI may aid in differentiating spinal TB from metastases, as metastatic lesions more often show washout or plateau enhancement patterns.^[10]

Role of diffusion imaging

Diffusion-weighted imaging (DWI) is increasingly incorporated for lesion characterization. Imaging with multiple b-values (at least three) is recommended to reduce T2 shine-through. Tuberculous vertebrae typically demonstrate higher apparent diffusion coefficient (ADC) values (approximately 1.47 ± 0.25 × 10^-3 mm²/s) compared with normal vertebrae^[11] and some other pathologies; however, overlap with other infections or metastases can occur [Figure 7].^[12] Therefore, DWI and ADC values should always be interpreted in conjunction with clinical and conventional MRI findings.

Close

Figure 7:

A 48-year-old male presented with weakness, paraplegia and fever. Magnetic resonance imaging (a) diffusion-weighted imaging (DWI) done with b-value of 400 and (b) apparent diffusion coefficient (ADC) images of lumbar spine coronal view showing right psoas abscess (yellow arrows in a and b) and left side pelvic abscess (white arrows in a and b) with diffusion restriction (high signal on DWI and low signal on ADC images). Tuberculosis abscesses show variable amount of restricted diffusion. ADC values from the abscess ranged from –0.8 to 1.2 × 10^–3 mm²/s.

Export to PPT

WHAT ARE THE COMMON SITES FOR TB IN THE SPINE?

The thoracolumbar junction is the most commonly affected site in spinal TB, though cervical and sacral involvement is also not uncommon. There are four distinct patterns [Figure 8] of vertebral involvement in spinal TB^[13]:

1. Paradiscal type: This is the most common type of vertebral involvement, where the vertebral bodies adjacent to the intervertebral disc are affected due to arterial spread of infection.

2. Central type: In this pattern, the central vertebral body is involved as a result of the spread of infection through the valveless Batson’s para vertebral venous plexus.

3. Anterior type: Infection spreads beneath the anterior longitudinal ligament, resulting in sub ligamentous involvement of the vertebral body.

4. Posterior element involvement: This is rare and occurs through the spread of infection through the posterior external vertebral venous plexus or by direct spread to the posterior elements of the vertebrae

Close

Figure 8:

Illustrated diagram depicting the common locations of spinal tuberculosis.

Export to PPT

WHAT IS THE CLASSICAL IMAGING PATTERN OF SPINAL TB?

The median number of vertebrae affected in patients with spinal TB is three.^[14]

The classical imaging pattern of spinal TB often involves paradiscal lesions, as the disc is avascular and the paradiscal area is supplied by the end plates of the vertebrae on either side of the disc making this location a favorite site for TB [Figure 9].

Close

Figure 9:

(a) Sagittal T1WI, (b) T2WI, (c) Coronal STIR, and (d) Axial T2WI MR images of the thoraco-lumbar spine in a 37 year man showing classical pattern of tuberculosis with involvement of two adjacent lower thoracic vertebral bodies with intervening discitis (yellow arrows in a-c), associated posterior epidural collection (white arrows in a and b) causing cord compression, paraspinal collections extending along psoas (black arrows in c and d) and right posterior paraspinal collection (thin yellow arrow in d). MR: Magnetic resonance

Export to PPT

Based on some studies, isolated solitary vertebral body TB is seen in only 1.69% of the total proven cases of spinal TB^[15] and the incidence of multi-level noncontiguous vertebral TB was observed as high as 71%.^[6]

On MRI, these lesions appear as low-signal intensity on T1-weighted images and high-signal intensity on T2-weighted images, with involvement of the end plates of the adjacent vertebral bodies. This leads to narrowing of the disc, either due to destruction of the subchondral bone or through the subsequent herniation of disc material into the vertebral body.

Disc involvement, typically seen in the later stages of the disease, is demonstrated by reduced disc height, T2 hyper-intensity within the disc, loss of the T2 hypointense cleft, and contrast enhancement.

TB spondylodiscitis is often associated with marrow edema, epidural abscesses and large paraspinal abscesses.^[13] Typically, the soft-tissue component in TB is larger compared to other causes of spinal infections. The TB -related soft tissue involvement tends to have smooth, regular walls, and in some cases, calcifications may be seen which are pathognomonic of TB. Para vertebral soft-tissue involvement and abscess formation are more common in TB than in pyogenic infections [Figure 10]. These paraspinal abscesses may track through the iliopsoas compartment, extending into the retroperitoneum and pelvis.^[16]

Close

Figure 10:

A 49-year-old presented with backache and fever. Magnetic resonance imaging (a and b) coronal STIR images of lumbar spine showing bilateral psoas abscesses (yellow arrows in a and b) with spondylodiscitis (white arrow in a) in a patient showing classical features of tuberculosis.STIR: short tau inversion recovery

Export to PPT

Atypical presentations may include skip vertebral involvement, isolated cold abscesses without obvious or significant bony involvement, multifocal involvement, isolated disc involvement, isolated neural arch involvement, concentric vertebral collapse, or ivory vertebrae.^[3]

Differentiating tuberculous from pyogenic spinal involvement on MRI can be challenging. Discal involvement as the primary presenting feature is uncommon in TB because the disease lacks the proteolytic enzymes necessary for disc destruction.^[14] Loss of vertebral height, which leads to spinal deformity, is more often observed in the later stages of TB.

CAN TB HAVE SKIP LESIONS?

Batson’s peri vertebral venous plexus plays a significant role in the development of skip lesions. Skip lesions are defined as separate lesions occurring in at least two non-contiguous vertebrae, regardless of their location along the spine [Figure 11].

Close

Figure 11:

(a) Sagittal STIR MR image of lumbar spine and (b) T2W MR image of whole spine showing contiguous vertebral involvement at lower lumbar, sacral vertebrae showing marrow edema, disc edema, anterior sub-ligamentous spread (white arrow in a) and associated anterior epidural collections (thin yellow arrow in a), skip lesions involving mid dorsal and lumbosacral vertebrae (thick yellow arrows in b). MR: Magnetic resonance

Export to PPT

Statistical analysis shows that in approximately 40% of noncontiguous or skip lesions, the additional lesions are asymptomatic. These silent lesions may progress unnoticed, potentially leading to long-term morbidity.^[17]

Identifying skip lesions is crucial, as missing those on initial imaging can underestimate disease severity. These overlooked lesions may be more neurologically significant. If detected later during follow-up, they might be mistaken for new, drug-resistant TB lesions, when in fact they were present but missed earlier.^[18]

HOW CRUCIAL IS TO DETECT NEUROLOGICAL INVOLVEMENT IN CASES OF SPINAL TB?

Any pathology exerting mechanical pressure on the spinal cord, nerve roots such as abscesses, granulation tissue, sequestrum, subluxations, or distraction, can lead to neurological sequelae [Figure 12].

Close

Figure 12:

Magnetic resonance imaging sagittal T2W image shows tubercular spondylodiscitis, mild gibbus formation, soft tissue component together causing cord compression with associated mild myelopathy.

Export to PPT

The most common mode of spinal cord involvement in TB is believed to be hematogenous spread of the tuberculous bacilli from other parts of the body, with the thoracic cord being the most frequently affected. However, only 0.2% of central nervous system TB cases show evidence of spinal cord involvement.^[18] However, among the spinal TB cases, neurological system is involved in 23–76% of cases.^[19] Most of the cases, neurological symptoms are not directly related to infection rather due to extrinsic compression [Figure 13].

Close

Figure 13:

(a) Magnetic resonance imaging axial T2W (yellow arrow), (b) Coronal STIR (yellow arrows) and (c) single shot myelography images (yellow arrows) demonstrate ill-defined subpial, intra-dural and extra medullary cord surface rounded lesions at thoracic level and also at conus medullaris, with corresponding demonstration of the lesions in (c) MR myelography. This patient’s ophthalmic examination revealed typical choroid granulomas which resolved after anti-tubercular treatment. STIR: Short tau inversion recovery

Export to PPT

CAN TB PRESENT AS A SCHMORL’S NODE?

Schmorl’s node-like appearance is very common in TB as para discal involvement is the primary site for TB. Usually, these are associated with other vertebral body lesions rather than as an isolated presentation.

When Schmorl’s nodes are identified, it’s important to consider potential underlying pathologies. The most common cause of Schmorl’s node is degenerative. However, in cases where other vertebral bodies show marked para discal involvement with typical paraspinal abscesses, TB should be considered. Rarely pyogenic infection can present predominantly with Schmorl’s node. If the node is associated with sacroiliitis or Anderson’s lesions, ankylosing spondylitis could be a potential cause.

In some cases, with Schmorl’s nodes, where the disc involvement is very borderline and there is no significant end plate changes or paraspinal soft tissue involvement, it is better to follow up after 4–6 weeks than to intervene [Figure 14].

Close

Figure 14:

Various diseases showing schmorl’s node like appearance (a) - degenerative changes, (b) pseudomonas spondylodiscitis (c) ankylosing spondylitis and (d) tuberculosis (yellow arrows in a-d).

Export to PPT

CAN THERE BE ISOLATED PRE AND PARASPINAL SOFT TISSUE INVOLVEMENT?

Yes, TB can involve both para- and pre-spinal soft tissues alone [Figure 15].

Close

Figure 15:

Magnetic resonance imaging STIR images of spine: (a and b) Coronal and (c) para sagittal images showing enlarged bulky right psoas muscle with a right psoas abscess (yellow arrows in a-c). No obvious bony spinal involvement seen.

Export to PPT

CAN TB SPREAD SUB-LIGAMENTOUSLY AND SUB-PERIOSTEALLY?

Sub-ligamentous spread of the infection is one of the four primary patterns described in spinal TB. Infection through the anterior arterial arcade primarily affects the bone anteriorly, immediately adjacent to the disc. This infection may then spread to the adjacent vertebrae beneath the anterior longitudinal ligament [Figure 16].

Close

Figure 16:

(a) Magnetic resonance imaging (MRI) T2W image of cervical spine sagittal view showing spondylitis with sub ligamentous spread of the infection (yellow arrow) underneath the posterior longitudinal ligament causing cord compression. Spine also shows coarse anterior osteophytes. (b) MRI T1 post contrast fat-saturated images of thoracic spine sagittal view showing sub ligamentous spread of the infection (yellow arrow) posteriorly to involve two vertebral levels above. Few smaller vertebral lesions below the spondylodiscitis level represent multifocal areas of tubercular infection.

Export to PPT

WHAT ARE THE COMPLICATIONS OF SPINAL TB?

Spinal TB can lead to cord compression, often due to epidural granulation tissue and vertebral collapse. Progressive destruction may cause vertebra plana, kyphosis, or gibbus deformity. In some cases, there is extension to ligaments and soft tissues, further contributing to instability and neurological compromise. Some of the differentials of vertebra plana include trauma, osteoporosis, Langerhans cell histiocytosis, leukemia, and lymphoma.^[20]

CAN TB SPARE THE DISC?

Disc sparing is not common, but can be seen in some cases. Two to three contiguous vertebral bodies are involved with relative disc sparing, which can pose a diagnostic challenge in differentiating spinal TB from metastasis [Figure 17]. Disc involvement or sparing in spinal TB depends largely on vascularity. In some adults and older individuals, the disc is typically spared due to its sparse vascular supply. However, in children, where the disc has a rich vascular supply, it is more likely to be involved in the infection.^[6]

Close

Figure 17:

A 48-year-old, proved case of spinal tuberculosis. Magnetic resonance imaging (MRI) (a) STIR and (b) MRI T1W image of lumbosacral spine sagittal view showing involvement of L3, L4 vertebral bodies (yellow arrows) sparing the intervening disc STIR: Short tau inversion recovery.

Export to PPT

Non-contiguous multilevel spinal TB is relatively rare, with reported incidence rates varying widely in the literature, ranging from 1.1% to 71.4%.^[17] The “floating disc sign” may arise infrequently if there is severe spinal damage with disc sparing.^[19]

IS THE WHOLE SPINE SCREENING NECESSARY?

Whole spine screening is essential, as spinal TB often exhibits multifocal involvement. It also helps detect occult sites of neural compression that may influence clinical management. If contrast is administered, the whole spine screening must be repeated in post contrast study as well [Figure 18].

Close

Figure 18:

(a) Magnetic resonance imaging (MRI) T2W image of spine sagittal view showing involvement of two contiguous vertebral bodies of thoracic spine. (b) MRI T2W image of whole spine sagittal view shows involvement of thoracic spine (yellow arrow in a-d) with confidence; however the lesions in other locations are not clearly delineated on T2WI. This is a potential pitfall of T2WI alone and signifies the value of adding STIR sequence. (c) MRI T1W image of whole spine sagittal view showing involvement of lumbar spine (white arrow in c and d) in addition to the thoracic spine which is not clearly visualized in T2W images. Hence, (d) fat saturated images add an additional advantage to find small lesions. (d) MRI STIR image of whole spine sagittal view showing all the three lesions with another focus in the cervical spine (yellow arrowhead in c and d) apart from lesions in thoracic and lumbar spine. STIR: Short tau inversion recovery

Export to PPT

WHAT ARE THE TREATMENT ASPECTS RELEVANT TO IMAGING?

Most of the spinal TB patients are treated for 9 months. However, it varies from patient to patient and can extend up to 18 months. This variation comes due to clinical improvement and imaging evidence of response.

All the treatment aspects, response assessment have been covered in separate dedicated sections in this edition.

CAN PATIENTS WORSEN CLINICALLY AFTER STARTING TB TREATMENT?

Clinical non-improvement despite treatment can have several causes, including mechanical complications (e.g., collapse), immune reconstitution inflammatory syndrome, poor treatment compliance, an alternative diagnosis, or multidrug-resistant TB. Not all worsening indicates treatment failure [Figure 19].

Close

Figure 19:

Magnetic resonance imaging STIR image of thoracic spine sagittal view (a) shows cord compression (yellow arrow) and collapse of a lower dorsal vertebra (white asterisk in a and b) causing compressive myelopathy. After 15 days of treatment, there was further clinical worsening of the patient and imaging (b) revealed collapse of another vertebra (yellow asterisk) too. Patient eventually responded to antitubercular therapy and surgical decompression. STIR: Short tau inversion recovery

Export to PPT

Post-treatment imaging has been covered in detail in another chapter of spinal TB.

IS BIOPSY NECESSARY FOR DIAGNOSIS OF TB?

CT is the reference standard imaging modality for guiding percutaneous vertebral biopsy. The samples obtained are crucial for both microbiological and histopathological evaluation. Two types of samples are typically collected: Fine-needle aspiration cytology, which is also sent for microbiological analysis (including drug sensitivity testing), and a large core biopsy, which is used for histopathological evaluation. Both samples can undergo additional testing, including genotyping analysis and immunohistochemistry.

The sensitivity of CT-guided biopsy is approximately 76%. Among the cases that test positive, granulomatous inflammation is observed in about 84% of biopsies, serving as a key histological indicator. Although AFB smear positivity is relatively uncommon, with a yield of only 5.7%, AFB culture remains a valuable diagnostic tool, offering a higher yield of up to 17.3%.^[21]

In spinal TB, the advantage with Gene Xpert is that it is a rapid test and has high sensitivity of 91.18%, and specificity of 100%.^[22]

WHAT ARE THE VARIOUS APPROACHES FOR CT GUIDED VERTEBRAL BIOPSY?

Image-guided vertebral biopsies require a tailored approach based on spinal level, lesion accessibility, and surrounding anatomy. Selecting the appropriate technique optimizes safety and diagnostic accuracy. Table 1 outlines commonly used biopsy approaches and their specific indications [Figure 20a and b; Table 1].^[23]

Close

Figure 20:

(a) Images depicting various approaches in a CT guided vertebral biopsy. In dual type (fourth image) - Needle is passed through both bone (as represented in image) and soft tissue components (represented by orange colour) separately. (Concept courtesy: Dr. Dharmendra Kumar Singh). (b) CT guided biopsy - transpedicular approach in a lumbar vertebra. CT: Computed tomography

Export to PPT

WHAT TESTS DO YOU PERFORM FROM THE BIOPSY SAMPLE?

Biopsy is becoming mandatory even in tropical countries or endemic areas due to evidence-based practices and large number of imaging mimics. This is much more important in cases of atypical spinal infections, immunocompromised status, or when long-term follow-up is anticipated, biopsy remains the cornerstone for definitive diagnosis.

Accurate diagnosis of spinal TB relies on a combination of microbiological and molecular tests. Each test varies in turnaround time, sensitivity, and specificity. Table 2 summarizes commonly used diagnostic methods following tissue biopsy.^[24,25]

WHAT IS THE BEST APPROACH FOR DOUBTFUL OR BORDERLINE RADIOLOGICAL FINDINGS IN MRI?

In cases where imaging findings are inconclusive, such as subtle endplate changes, discal alterations without significant vertebral body involvement, without substantial para vertebral or paraspinal soft-tissue involvement, a follow-up approach is often advisable. This can be based on the clinical progression or after 6 weeks’ interval.

During follow-up, one may observe either resolution or progression of the imaging findings, which can help clarify the diagnosis. Based on the progression or lack thereof, further management can be determined, including whether to proceed with biopsy or initiate treatment provided the findings become more conclusive over the time [Figure 21].

Close

Figure 21:

(a) 51-year-old presents with low backache and stiffness. Magnetic Resonance Imaging (MRI) lumbar spine sagittal view - (a) T1WI, (b) T2WI revealed equivocal discitis like changes with subtle end plate changes (yellow arrows in a and b). (c) T1WI and (d) T2WI show a follow up after 3 months showed rapid increase in the end plate changes (yellow arrows in c and d)- This is consistent with infective cause rather than degenerative. Biopsy revealed as tuberculosis.

Export to PPT

HOW TO DIFFERENTIATE ACTIVE DISEASE FROM HEALED STAGE?

In spinal TB, the active stage shows abscesses, marrow edema, intense enhancement, and lytic lesions, while the healed stage is marked by reduced soft tissue, fatty marrow conversion, decreased enhancement, sclerosis, ankylosis, and new bone formation [Table 3 and Figure 22].

Close

Figure 22:

Magnetic resonance imaging (a) STIR, (b) T1-WI sagittal images of spine showing complete fatty conversion of marrow signal (yellow arrows a and b), which is a final indicator of complete healing. STIR: Short tau inversion recovery

Export to PPT

WHAT ARE THE MIMICKERS OF SPINAL TB?

Spinal TB can mimic several conditions which can be distinguished by collaboration of various key points [Table 4]. Although differentiation can be challenging in some cases, the key clues to help distinguish between tuberculous and pyogenic spondylitis is provided in Table 5.^[26]

Discussed elaborately in another chapter of Spinal TB mimics.

WHAT ARE THE RECENT ADVANCES IN THIS TOPIC?

What is the role of PET CT?

Fludeoxyglucose (FDG) PET/CT and Technetium-99 bisphosphonate scans are valuable in detecting abnormal metabolic activity and multifocal involvement in spinal TB but struggle to reliably differentiate infectious from non-infectious or malignant lesions due to overlapping imaging features. Gallium-67, especially when combined with Technetium-99, improves specificity for infection by identifying both bone and soft tissue involvement.^[27] Despite standardized uptake value (SUV) >2.5 being suggestive of malignancy, TB lesions can show very high SUV values, up to 21.0, complicating diagnosis.^[28] Dual-time-point imaging can aid differentiation, as TB lesions often show FDG washout, whereas malignancies show increased uptake.^[29]

PET/CT also supports therapy monitoring and biopsy targeting by assessing lesion activity, extent, and metabolic response using SUV metrics and volumetric parameters.^[30] While not TB-specific, integrated PET/CT when paired with contrast-enhanced CT enhances diagnostic confidence [Figure 23]. Future developments, such as TB-specific radiotracers, hold promise for improving specificity and more reliably distinguishing TB from malignancy and other inflammatory conditions.^[31]

Close

Figure 23:

(a) Axial view of positron emission tomography (PET) scan, (b) Sagittal view of Fusion image, (c) Sagittal view of CT spine, (d) Sagittal view of STIR sequence in magnetic resonance imaging spine, (e) Diffusion-weighted imaging image of spine and (f) axial view of PET scan. (a and b) FDG avid (c) lytic destructive lesion involving body of L1-L2 (yellow arrows) vertebrae with loss of disc height with spondylodiscitis on (d) STIR images with (e) restricted diffusion along the discal margins (yellow arrow in e), (f) Increased FDG uptake seen in several mediastinal and para tracheal nodes (yellow arrows in a-f). STIR: Short tau inversion recovery, FDG: Fluorodeoxyglucose

Export to PPT

PET/CT may underperform MRI in detecting spinal TB, particularly in cold abscesses and non-FDG-avid lesions. SUVmax values vary widely due to biological, technical, and disease extent-related factors. Increased FDG uptake post-treatment may reflect tissue regeneration, not active disease. Standardized PET/CT protocols and further research are essential to define its role alongside MRI in STB.^[32]

While PET-CT provides valuable functional information for assessing disease activity, extent, and treatment response in spinal TB, MRI remains the first-line imaging modality due to its superior soft-tissue contrast, ability to detect early marrow and discal involvement, and detailed evaluation of neural and paraspinal complications.

What is the role of DWIBS?

Whole body DWIBS has a potential role in spinal TB and cervical lymph nodal TB, to identify the dissemination of the disease and assessment of disease burden. More research and comparative studies with PET CT scan may add further information and pave way for future directions [Figure 24].^[33]

Close

Figure 24:

(a) Axial diffusion-weighted imaging image and (b) coronal diffusion weighted imaging with whole body screening. In this case of tuberculosis, with vertebral body involvement, whole body screening revealed involvement of multiple para-tracheal and mediastinal lymph nodes (yellow arrows in a and b).

Export to PPT

CONCLUSION

Spinal tuberculosis most commonly presents as spondylodiscitis and remains an important cause of morbidity even in the present era, particularly in endemic regions. Imaging plays a crucial role in early diagnosis, assessment of disease extent, and monitoring of treatment response. MRI remains the modality of choice because of its excellent ability to detect marrow edema, disc involvement, paraspinal and epidural collections, neural compression, and skip lesions. Careful radiologic evaluation is essential to differentiate infective spondylodiscitis from other inflammatory spondylodiscopathies, as these entities often show overlapping imaging features. Dual-energy CT may further aid in differentiating spinal osteolytic infections from metastases. As several spinal infections can mimic tuberculosis, image-guided biopsy is becoming increasingly common to establish a definitive diagnosis and guide appropriate management.