[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
Views 1,232
Citations 0
Special Feature
May 2013

Picture of the Month—Diagnosis

JAMA Pediatr. 2013;167(5):484. doi:10.1001/jamapediatrics.2013.7b
Denouement and Discussion: Spinal and Scapular Tuberculosis

The chest radiograph shows marked scoliosis secondary to vertebral collapse but no pulmonary lesions or hilar lymphadenopathy to suggest presence of intrathoracic tuberculosis (TB). Magnetic resonance imaging of the spine revealed multiple hyperintense lesions of the vertebral bodies of T12 and L1 with a small adjacent epidural collection as well as wedge-shaped destruction of the vertebral body of T8 resulting in spinal gibbus deformity. Technetium Tc 99m bone scintigraphy, performed to assess whether further sites were involved, showed areas of increased radionuclide uptake in the left acromion and the thoracolumbar spine (not shown).

Tuberculosis was suspected based on the magnetic resonance imaging findings in conjunction with the history of birth and previous residence in a country with a high TB prevalence. The tuberculin skin test result was strongly positive (25-mm induration at 48 hours); both interferon-γ release assay (IGRA) results were also positive (QuantiFERON-TB Gold In-Tube and T-SPOT. TB). Biopsy of an affected vertebra showed histologic findings suggestive of mycobacterial infection, with presence of copious lymphocytes and epithelioid histiocytes as well as multinucleated giant cells. However, acid-fast bacilli were not present on microscopy and the results of mycobacteria-specific polymerase chain reaction tests were negative. Fully sensitive Mycobacterium tuberculosis was grown in cultures after 3 weeks of incubation.

The patient received rifampin, isoniazid, pyrazinamide, and ethambutol hydrochloride, which led to resolution of pyrexia and night sweats within 5 days. A spinal brace was fitted to address potential spinal instability. The shoulder pain resolved within 2 weeks and all symptoms had resolved by the 2-month review, including the cutaneous lesion, which was likely the result of a fistula from the primary focus in the scapula.

Extrapulmonary TB acounts for approximately 20% of the reported TB cases in the United States.1 Osteoarticular TB accounts for approximately 10% of extrapulmonary cases.1 The most common type of osteoarticular TB is spinal TB, also known as Pott disease.2

In children, the diagnosis of osteoarticular TB, similar to other forms of TB, is often challenging. The gold standard for the diagnosis of active TB remains the detection of the causative organism, M tuberculosis. This can be achieved by culture or molecular methods such as polymerase chain reaction. The major limitation of cultures is the significant delay in establishing a definitive diagnosis. Molecular techniques are faster, but they have suboptimal sensitivity in patients with paucibacillary disease. The sensitivity of polymerase chain reaction–based methods in osteoarticular TB ranges between 61% and 83%.3

Immunodiagnostic methods, including the tuberculin skin test and IGRAs, can support a presumptive diagnosis of active TB. However, the tuberculin skin test can yield false-positive results following BCG vaccination and in the context of nontuberculous mycobacterial disease. Interferon-γ release assays have better specificity but have suboptimal sensitivity (60%-80%) in children with active TB.4 Although IGRAs are exclusively licensed for the diagnosis of latent TB, in clinical practice they are frequently used as an adjunctive test in the investigation of suspected active TB. Unfortunately, discordance between IGRA and tuberculin skin test results and indeterminate IGRA results are considerably more common in children than adults.5,6

Radiological investigations can provide further valuable clues. Our patient had a wedge-shaped collapse of the T8 vertebral body. Notably, collapse of thoracolumbar vertebrae, which results in short-segment structural kyphosis, is a characteristic feature of spinal TB.

The treatment of osteoarticular TB typically comprises 3 or 4 active antituberculous drugs for 2 months followed by treatment with rifampin and isoniazid for 4 months, unless susceptibility testing reveals resistance to these drugs.2 However, there is ongoing controversy regarding the optimal treatment duration for osteoarticular TB, with some experts recommending extending treatment to a total of 9 or 12 months owing to concerns regarding inadequate penetration of antituberculous drugs into bony tissue.2

Return to Quiz Case.

Back to top
Article Information

Correspondence: Marc Tebruegge, MRCPCH, MD, NIHR Wellcome Trust Clinical Research Facility, University Hospitals Southampton NHS Foundation Trust, Tremona Road, Southampton SO16 6YD, United Kingdom (m.tebruegge@soton.ac.uk).

Accepted for Publication: December 20, 2012.

Author Contributions:Study concept and design: Tebruegge and Faust. Acquisition of data: Tebruegge and Sukhtankar. Analysis and interpretation of data: Tebruegge, Patel, and Faust. Drafting of the manuscript: Tebruegge. Critical revision of the manuscript for important intellectual content: All authors. Study supervision: Tebruegge, Patel, and Faust.

Conflict of Interest Disclosures: None reported.

Peto HM, Pratt RH, Harrington TA, LoBue PA, Armstrong LR. Epidemiology of extrapulmonary tuberculosis in the United States, 1993-2006.  Clin Infect Dis. 2009;49(9):1350-1357PubMedGoogle ScholarCrossref
Ritz N, Connell TG, Tebruegge M, Johnstone BR, Curtis N. Tuberculous dactylitis.  Eur J Clin Microbiol Infect Dis. 2011;30(11):1303-1310PubMedGoogle ScholarCrossref
Pandey V, Chawla K, Acharya K, Rao S, Rao S. The role of polymerase chain reaction in the management of osteoarticular tuberculosis.  Int Orthop. 2009;33(3):801-805PubMedGoogle ScholarCrossref
Kampmann B, Whittaker E, Williams A,  et al.  Interferon-gamma release assays do not identify more children with active tuberculosis than the tuberculin skin test.  Eur Respir J. 2009;33(6):1374-1382PubMedGoogle ScholarCrossref
Tebruegge M, Connell T, Ritz N, Bryant PA, Curtis N. Discordance between TSTs and IFN-gamma release assays.  Eur Respir J. 2010;36(1):214-216PubMedGoogle ScholarCrossref
Tebruegge M, Connell T, Curtis N. Tuberculosis in children.  N Engl J Med. 2012;367(16):1568-1569PubMedGoogle ScholarCrossref