The tuberculosis (TB) and human immunodeficiency virus (HIV) pandemics have proven to be a lethal
combination. Approximately 30% to 40% of people infected with HIV also have TB, making it the greatest
killer of HIV patients. The authors review the clinical presentation and imaging findings in these patients.
Mr. Scheel is a MD/PhD student at the University of
CaliforniaSan Diego, CA. Dr. Sartoris is deceased.
The World Health Organization estimates that approximately
one-third of the world's population is infected with tuberculosis
(TB). TB kills 3 million people each year and accounts for more
than 25% of preventable deaths in adults. Approximately 1% of the
world's population is infected with TB each year, resulting in more
deaths than malaria and AIDS combined. It is estimated that by
2020, another 1 billion people will be infected with TB, leading to
70 million deaths.
With the advent of antituberculous drugs 50 years ago, cases of
TB steadily declined until 1985, when the human immunodeficiency
virus (HIV) emerged as a global infectious disease.
HIV and TB pandemics, acting synergistically, have proven to be a
In most healthy individuals, the immune system effectively
attenuates the initial TB lung infection and prevents the spread of
the disease to extrapulmonary locations. The TB bacilli can lie
dormant in the lungs for many years until the host's immune system
is weakened, as in elderly patients and HIV-positive patients. In
fact, extra-pulmonary TB infections in patients with HIV is an
It is estimated that one-third of the TB cases over the last 5
years can be attributed to HIV. Of the estimated 31 million people
infected with HIV, 30% to 40% also have TB, making it the greatest
killer of HIV patients.
There has also been a rise in nontuberculosis mycobacterium among
HIV patients. Another cause of the increase in the number of TB
cases is an increase in foreign-born immigrants, accounting for
half of TB cases in industrialized nations.
Particularly disturbing has been the increasing frequency of
multiple drug-resistant tuberculosis (MDRTB) isolated from infected
individuals. MDRTB is defined as tuberculosis isolates having
resistance to both isoniazid (INH) and rifampin (RMP). In
HIV-negative patients, these infections are infrequent, and are
almost always the result of an irregular drug supply, inappropriate
prescription, or patient noncompliance. In patients who are
HIV-positive, MDRTB has also emerged as the result of inadequate
treatment; however, its spread to close-contacts has increased in
frequency at a more alarming rate than in non-infected individuals.
MDRTB cases have already exceeded 4% of clinical isolates in large
cities such as New York, NY; Newark, NJ; Dallas, TX; Oakland, CA;
and Sacramento, CA. In patients with HIV and MDRTB, the mortality
rates are 70% to 90% within months of infection; however, in
patients with HIV and drug-susceptible TB, case-fatality rates are
similar to those of other infections.
In 1994, the International Union Against Tuberculosis and Lung
Disease (IUATLD) began a global project on MDRTB surveillance in
order to measure its prevalence in several countries. During this
study, 50,000 TB cases from 35 countries, representing 20% of the
world's population, were studied. Each study evaluated resistance
against INH, RMP, ethambutol (EMB), and streptomycin (SM) as well
as whether the drug resistance was primary or acquired. Primary
drug resistance is defined as strains that were resistant prior to
drug treatment. The mean prevalence of MDRTB in this category was
1.4%, ranging from 0% in Kenya to 14.4% in Latvia. Kenya was the
only country studied that did not have a case of MDRTB. Acquired
drug resistant strains originate from case mismanagement; i.e.,
from patient noncompliance or insufficient drug therapy. This
category had a much higher mean prevalence at 13%, ranging from 0%
in Kenya to 54% in Latvia.
Individual countries have reported similar findings.
The clinical presentation of a dual HIV and TB infection may be
different from either of their individual symptomologies. With
increasing numbers of these two diseases in industrialized nations,
physicians need to be aware of how TB presents itself in
immunocompromised patients. In a patient not infected with HIV, TB
presents itself with general systemic signs and symptoms such as
weight loss, night sweats, fever, malaise, and anorexia, as well as
signs and symptoms specific to the site of infection. In the
presence of HIV, the clinical presentation depends on the status of
the patient's immune system. Many of the signs and symptoms of HIV
resemble those of TB, making it difficult to make an accurate
diagnosis. Usually, in patients initially infected with HIV,
pulmonary TB denotes stage 3 of the HIV infection and
extrapulmonary TB denotes stage 4.
Mycobacterium tuberculosis primarily infects the thoracic
(figure 1) and lumbar vertebrae (figure 2) and weight-bearing
joints (figures 3 and 4); however, it may involve any bone (figure
5), joint, tendon, bursa or fluid-containing structure (figure 6).
Approximately 10% to 15% of osteoarticular tuberculosis is
multifocal (figure 7), and in rare instances, aseptic-appearing
polyarthritis is associated with extra-articular TB. This condition
is believed by some to be Poncet's disease, a reactive arthritis,
because it responds to anti-TB therapy.
Imaging of tuberculosis
Radiology is still the most cost-effective test for TB, and
provides the most important information required for making a
diagnosis of spinal TB (figure 8). Radiography provides the
definitive proof of a TB infection, allowing the physician to begin
chemotherapeutic treatment immediately. In spinal cord cases,
radiography effectively shows the narrowing of the vertebral disc
spaces and involvement of the vertebral body (figure 9), eventually
leading to an anteriorly directed collapse with secondary kyphosis
CT and MRI are most useful in the preoperative evaluation of
patients requiring surgical treatment for TB, but are also more
accurate than plain radiography in showing vertebral involvement.
The anterior, posterior, and lateral soft-tissue features of spinal
infections were delineated equally with both. However, because the
coronal and sagittal planes can be obtained directly without loss
of spatial resolution, MRI is better at visualizing the extent of
Lateral CT scans are helpful in diagnosing spinal TB in areas of
the thoracic spine obscured by vertebral bodies. Axial CT is also
optimal for depicting posterior bony abnormalities, such as pedicle
Although both MRI and CT scans depict the extent of spinal cord
compressions equally, T1-weighted MRI is superior in its ability to
show the nature of the damage. MRI can also show the preserved
spinal tissue being compressed into the remaining vertebral body in
some cases. In a study by Hoffman et al,
MRI was performed on 11 paraplegic patients with TB and showed that
pus and granulation tissue from TB was the main cause of the spinal
cord compression in 8 of these patients. Resolution of the
compression and neurological lesions were also shown with MRI after
4 months of chemotherapy. Additionally, sagittal MRI was determined
to be the best method for visualizing the severity and nature of
the extradural compressions and a useful way to differentiate an
abscess from fibrous tissue.
Multiplanar MRI provides the most accurate demonstration of a TB
infection in the spinal cord, including extent of injury and the
spread of debris from the site of infection. The site of tissue
injury appears as a decrease in signal intensity on T2-weighted
images. This technique has been useful in the description of the
pathological tissue responsible for spinal cord compression by
allowing physicians to visualize liquid pus on T2-weighted images.
A worse outcome is indicated in patients with focal myelopathy in
the spinal cord, documented on T2-weighted images as discrete
abnormal lesions. Prolonged cord compression with resultant
ischemia has been observed in patients with a delayed presentation,
severe kyphosis, and marked neurological deficits. AR
The authors wish to thank G. Puri for her critical evaluation of
this paper. J.R. Scheel is supported by Research Fellowship F31
NS108060 from the National Institutes of Neurological Disorders and
1. WHO Report on the tuberculosis epidemic 1995. World Health
Organization Web site. Available at:
http://www.who.int/gtb/publications/ tbrep_95/index.htm. Accessed
2. Shafer RW, Kim DS, Weiss JP, Quale JM: Extrapulmonary
tuberculosis in patients with Human Immunodeficiency Virus
infection. Medicine 170:384-397, 1991.
3. Slutsker L, Castro KG, Ward JW, Dooley SW: Epidemiology of
extrapulmonary tuberculosis among persons with AIDS in the United
States. Clin Infect Dis 16:513-518, 1993.
4. Meier JL: Mycobacterial and fungal infections of bone and
joints. Curr Opin Reumatol 6:408-414, 1994.
5. Watters DA: Surgery for tuberculosis before and after human
immunodeficiency virus infection: A tropical perspective. Br J Surg
6. Straus WL, Ostroff SM, Jernigan DB, et al: Clinical and
epidemiologic characteristics of Mycobacterium haemophilum, an
emerging pathogen in immunocompromised patients. Ann Intern Med
7. Benson CA, Ellner JJ: Mycobacterium avium complex infection
and AIDS: Advances in theory and practice. Clin Infect Dis 17:7-20,
8. WHO publication: http://www.who.int/inf-fs /en/fact104.html.
Accessed July 1999.
9. Anti-tuberculosis drug resistance in the world: The
WHO/IUATLD Global Project on Antituberculosis Drug Resistance
Surveillance 1994-1997. Available on the World Health Organization
Web site at: http://www.who.int/gtb/publications/ dritw/index.htm.
Accessed July 1999.
10. CDC/NCHSTP Tuberculosis Surveillance Report:
http://www.cdc.gov/nchstp/tb/. Accessed July 1997.
11. Hameed K, Karim M, Islam N, Gibson T: The diagnosis of
Poncet's Disease. Br J Reumatol 32:824-826, 1993.
12. Hoffman EB, Crosier JH, Cremin BJ: Imaging in children with
spinal tuberculosis. J Bone Joint Surg 75-B(2):233-239, 1993.
13. Keat A: TB or not TB? That is the question. Br J Rheumatol