Jump to: Page Content, Site Navigation, Site Search,
You are seeing this message because your web browser does not support basic web standards. Find out more about why this message is appearing and what you can do to make your experience on this site better.
Published 10 June 2009, doi:10.1136/bmj.b1210
Cite this as: BMJ 2009;338:b1210
Marc Tebruegge, honorary clinical research fellow1, Nicole Ritz, honorary clinical research fellow1, Tom Connell, consultant in paediatric infectious diseases1, Nigel Curtis, professor and head of paediatric infectious diseases2
1 Infectious Diseases Unit, Department of General Medicine & Murdoch Childrens Research Institute, Royal Childrens Hospital Melbourne, Victoria 3052, Australia, 2 Department of Paediatrics, University of Melbourne, Royal Childrens Hospital Melbourne, Victoria 3052, Australia
N Curtis nigel.curtis{at}rch.org.au
A 2 year old girl presented to the accident and emergency department with high grade fever (temperature 39.2oC), cough, tachypnoea (respiratory rate 45 breaths/min), and lethargy. She had been unwell for six weeks with general malaise and intermittent fever. She had also been anorexic for three weeks and had lost about 7% of her body weight. She was born in Australia to parents from the South Pacific Islands. Her routine immunisations were up to date and her past medical history was unremarkable.
The patient underwent chest radiography (fig 1
).
|
Short answers
|
release assay are the most useful initial tests.
Long answers
1 Abnormalities on radiography
The finding of bilateral fine reticulogranular (miliary) shadowing on chest radiography is considered a cardinal diagnostic feature in miliary tuberculosis.1 Between 50% and 90% of affected patients show a "classic" miliary pattern, consisting of small discrete opacities with widespread distribution throughout the lung fields.2 3 4 The differential diagnosis for this classic appearance includes Pneumocystis jiroveci pneumonia, lymphocytic interstitial pneumonitis, and viral pneumonia. A non-miliary pattern—with asymmetrical nodules, coalescence of nodules, or a mottled appearance—is less common. Other features, such as hilar lymphadenopathy, pleural effusion, segmental consolidation, and (as in our case) pulmonary cavitation may also be present.2 3 5 A small proportion of patients with miliary tuberculosis develop parenchymal air leaks, which can result in pneumothorax, pneumopericardium, or pneumomediastinum.5 6 7 8 9
Other imaging modalities can reveal further foci of disseminated infection. In our patient, who was also irritable and lethargic at presentation, a cranial magnetic resonance imaging scan showed multiple inflammatory foci scattered throughout the brain, corresponding to cerebral miliary lesions (fig 3
).
|
Miliary tuberculosis is the result of haematogenous spread of Mycobacterium tuberculosis from a pulmonary or extrapulmonary focus, and it accounts for 1-5% of all cases of tuberculosis.2 Although predominately a disease of infants and young children, the reported incidence of miliary tuberculosis in adults in the United Kingdom has increased over the past few decades.12 Predisposing factors include pregnancy, malnutrition, diabetes mellitus, chronic renal failure, treatment with immunosuppressive drugs (such as high dose corticosteroids) or immunomodulatory drugs (such as infliximab or etanercept), malignancies, and HIV infection.2 3 13
The presenting features of miliary tuberculosis include fever, night sweats, malaise, anorexia, weight loss, cough, peripheral lymphadenopathy, and hepatosplenomegaly.1 2 14 The last two features are more commonly found in children.2 In addition, although rare, fundoscopic examination may reveal pathognomonic choroidal tubercles.
3 Other investigations
The diagnosis of active tuberculosis in children is most commonly based on the triad of an abnormal chest radiograph, a positive tuberculin skin test, and a history of contact with an infectious tuberculosis index case.15
The tuberculin skin test involves injecting purified protein derivative intradermally. The test result is read after 48-72 hours by measuring the diameter of induration around the injection site. Importantly, interpretation of this test may be confounded by several factors, including previous BCG immunisation, previous exposure to environmental mycobacteria, immunosuppression, and immunodeficiency.16 17 In common with other severe forms of tuberculosis, a large proportion of patients with miliary tuberculosis exhibit tuberculin anergy, resulting in false negative test results.2 18 In one study, only 26% of children with miliary tuberculosis had a positive tuberculin skin test at presentation.1 Therefore, although a negative result does not exclude active tuberculosis in this setting, a positive result in conjunction with typical radiographical findings supports the presumptive diagnosis.
Isolation of M tuberculosis confirms the diagnosis of tuberculosis and can provide important drug susceptibility data. However, tuberculosis is less often microbiologically confirmed in children than in adults. The decision to start antituberculous treatment usually needs to be made before confirmation of diagnosis by cultures, which may take several weeks to yield a positive result. Despite this, microscopy and culture of appropriate clinical specimens (such as sputum, gastric aspirates, bronchoalveolar lavage fluid, blood, and urine) should be done in all cases of suspected miliary tuberculosis.2 19 Notably, mycobacterial blood cultures are positive in only about a third of children with miliary tuberculosis.4
The latest guidelines from the National Institute for Health and Clinical Excellence also recommend performing a lumbar puncture in all patients with suspected miliary tuberculosis, on the basis of the fact that 10-30% of patients with this condition have evidence of tuberculous meningitis.2 20 Other invasive tests, such as bone marrow and liver biopsy, have relatively low yields.21
Interferon-
release assays, which have only relatively recently become commercially available, offer another means of supporting the diagnosis of tuberculosis. Two assays are currently licensed for use in the UK—QuantiFERON-TB Gold and T-Spot.TB. Both entail the detection of interferon-
released from T cells after stimulation of the patients blood with M tuberculosis specific antigens in vitro. Data on the use of these assays for the diagnosis of tuberculosis specifically in children are scarce,22 and some studies suggest that these assays perform less well in children than in adults.23 24 Limited evidence suggests, however, that the sensitivity of these assays for active tuberculosis in children is similar to or greater than that of the tuberculin skin test.25 Despite this fact, in a recent position statement released in 2008, the Health Protection Agency highlights that interferon-
release assays "should currently not be used as a routine diagnostic tool for active tuberculosis."26 Nevertheless, as with a positive tuberculin skin test in this setting, a positive interferon-
release assay result supports the presumptive diagnosis, although a negative test should not be used to exclude tuberculosis.
In conclusion, none of the currently available diagnostic tests for tuberculosis is sufficiently sensitive to reliably rule out active tuberculosis. Given that miliary tuberculosis is almost always fatal without specific treatment, clinicians should have a low threshold to start antituberculous treatment if the history, symptoms and radiographic findings suggest this diagnosis.2
Cite this as: BMJ 2009;338:b1210
Competing interests: None declared.
Provenance and peer review: Not commissioned; externally peer reviewed.