Latent (or inactive) tuberculosis (TB) infection occurs when an individual has dormant Mycobacterium tuberculosis organisms and no active disease. It can be diagnosed by a tuberculin skin test (TST) or by a blood test called an interferon-gamma release assay (IGRA); both are described below. HIV-infected persons with latent TB infection (LTBI) have a much higher risk of developing active TB (estimated at 10% per year) than the general population (estimated at 10% in a lifetime). The risk of an individual with LTBI developing active TB can be reduced 90% with treatment of LTBI. Hence, identifying and treating HIV-infected persons with LTBI is a high priority. Treatment of LTBI not only reduces the risk of disease to the individual but also reduces the risk of further TB transmission should the HIV/TB-coinfected person develop active pulmonary TB. Standard treatment with isoniazid (INH) is effective and safe.
Issues of concern regarding treatment of LTBI among HIV-infected persons include the following:
Ask about symptoms of active TB, including fever, cough, and weight loss; see chapter Mycobacterium tuberculosis. HIV-infected persons who have no symptoms of active TB and have not been treated previously for active or latent TB are eligible for LTBI treatment. When patients do have symptoms that could represent active TB, active TB must be evaluated and ruled out by appropriate diagnostic methods before initiating treatment (see "Assessment," below).
Persons who have had bacillus Calmette-Guérin (BCG) vaccine can be evaluated by an IGRA (preferably) or by a TST with correct interpretation. Immigrants from many countries will have had childhood BCG vaccination.
Health care providers should ask about a history of potential exposure to TB, because that might indicate infection with drug-resistant TB. Such risk may occur when the patient knows a source patient or when the exposure occurs in a setting with known drug resistance or in a location with ongoing TB transmission where others remain at risk of exposure.
Current U.S. Centers for Disease Control and Prevention (CDC) guidelines strongly recommend LTBI testing for all newly diagnosed HIV-infected persons. Repeat testing is recommended for patients whose CD4 lymphocyte count increases from low numbers to counts of >200 cells/µL, and annual testing is suggested for patients whose initial test result is negative.
Provisional recommendations from the CDC state that use of either the TST or any of the three currently available IGRAs is acceptable. The TST evaluates delayed hypersensitivity to antigens from a closely related Mycobacterium species whereas the IGRAs are in vitro tests of lymphocyte recognition and response to M. tuberculosis. A positive TST or IGRA result indicates previous contact between the patient and M. tuberculosis and it implies latent TB infection.
In general, all the licensed IGRA tests are considered interchangeable with the TST. Use of an IGRA is preferred for persons with previous BCG exposure and for patient groups at risk of not returning for TST readings. The TST is preferred for children under 5 years of age. The cost of an IGRA is higher than the cost of a TST, and individual clinics should compare the total costs of both testing approaches, including the costs involved in TST protocols of recalling patients who miss visits, repeating TSTs, and treating persons with prior BCG exposure who may have false-positive TST results.
The TST is administered as an intradermal injection of 0.1 mL (5 tuberculin units) of purified protein derivative (PPD), which raises a wheal in the skin. This is sometimes referred to as the Mantoux test. Multiple-puncture tests such as tine tests and the use of other strengths of PPD are considered unreliable. Anergy testing is not routinely recommended because a randomized controlled study of HIV-infected patients in the United States did not show an advantage to treating anergic, tuberculin-negative persons.
PPD tests are not designed for reading by the patient; a trained health care worker must measure the area of induration (not erythema) 48-72 hours after the test is administered. Induration of 5 mm or more is considered a positive result for HIV-infected persons, other immunosuppressed persons, anyone with recent TB exposure, and anyone with fibrosis on chest X ray that is consistent with previous TB. For HIV-uninfected health care workers, 10 mm of induration is a positive result; in various other populations, either 10 mm or 15 mm of induration may be considered positive. Care providers at many large HIV clinics find it challenging to ensure that their patients return for the PPD reading. One randomized study found that offering incentives (e.g., a fast-food coupon) plus counseling was more effective than counseling alone in obtaining return visits for PPD readings.
IGRA tests performed on peripheral blood samples are available in the United States from two manufacturers. Although the tests are expensive, results are obtained without the patient having to make a return visit to the clinic, and false-positive readings following BCG vaccination do not occur. Both IGRA tests use specific antigens from M. tuberculosis to stimulate the patient's lymphocytes and quantify the production of interferon-gamma in response. QuantiFERON-TB Gold (QFT-G) is performed by incubating a heparinized blood sample for 16-24 hours with the synthetic MTB antigens along with positive and negative controls. The incubation must begin within 12 hours, so the need for proximity to a laboratory that performs the test imposes limits on access. The laboratory calculates the amount of interferon-gamma produced by the patient's cells, providing a quantitative result and an interpretation (positive, negative, or indeterminate). The QuantiFERON-TB Gold In Tube (QFT-GIT) provides the clinician with three prepared tubes (three antigens in one tube and positive and negative controls) into which patient blood samples are placed; these tubes may be held for up to 16 hours at room temperature, incubated at 37ºC for 16-24 hours, then centrifuged. After centrifugation, the same tubes can be held again at room temperature for up to 3 days or refrigerated for up to 4 weeks before they are analyzed. This provides much greater flexibility in sending specimens to distant or reference laboratories and it makes the technology useful for a much wider array of clinical sites. Again, quantitative results and an interpretation are provided. The T.SPOT TB (T-SPOT) test requires separation and processing of 4 aliquots of 250,000 peripheral blood mononuclear cells each, a step that requires a sophisticated laboratory setting. The number of cells that produce interferon in response to the two MTB antigens is compared with positive and negative controls; both a quantitative result and its interpretation (positive, negative, borderline, or indeterminate) are provided. Performing the test is very labor intensive, and it must take place within 8 hours of obtaining the sample (within 32 hours if using T-Cell Xtend reagent); thus T-SPOT is the least flexible of the three assays.
Limited data suggest that the QFT-GIT is the most specific test (lowest rate of false-positive results) and the T-SPOT is the most sensitive test (with the lowest rate of false-negative results). However, few studies have compared the IGRA assays, and there is no reliable "gold standard" for sensitivity or specificity for latent TB infection. Although these tests do not detect previous BCG exposure, they will yield positive results in all three TB-causing species of the Mycobacterium tuberculosis complex, M. tuberculosis, Mycobacterium bovis, and Mycobacterium africanum, and may yield positive results in persons with exposure to certain rare mycobacteria other than M. tuberculosis (Mycobacterium kansasii, Mycobacterium marinum, and Mycobacterium szulgai).
There is limited experience using IGRAs for persons with HIV infection and other immunosuppressed states. The overall frequency of positive and negative results is the same with QFT-GIT and TST. However, there is substantial discordance among HIV-infected patients: many with positive TST results will have negative IGRA results, and vice versa. No data are available to determine which test is more reliable. Indeterminate results with QFT-GIT are more frequent with patients who have lower CD4 lymphocyte counts. Some experts anticipate that the T-SPOT may be more consistent in persons with lower lymphocyte (including CD4 cell) counts because a fixed number of cells are used in the assay; however, lymphocyte function may be altered in any case. It appears that there are higher rates of positive results among HIV patients with the T-SPOT than with the other IGRAs. TSTs are known to yield false-negative results in persons with HIV-associated immunosuppression, yet despite this limitation, TSTs are recommended for use with all newly diagnosed HIV patients and for annual testing in certain patients.
A chest X ray should be obtained for all HIV-infected persons with positive TST or IGRA tests. Asymptomatic patients with negative chest X-ray results should be offered treatment for LTBI. Persons with symptoms consistent with pulmonary or extrapulmonary TB, and those with abnormal chest radiography, require further assessment. This assessment may include sending three separate sputum specimens collected on three consecutive days, or at least 8 hours apart, including at least one early morning specimen (using saline spontaneously) for acid-fast bacilli (AFB) stain, nucleic acid amplification test, and culture, or obtaining other specimens depending on the suspected site of extrapulmonary TB. If suspicion of TB is low, patients with negative sputum smears (or other biopsy or tissue samples) can begin LTBI treatment. If suspicion of active disease is high, treatment for active disease should be started while the culture results are pending (see chapter Mycobacterium tuberculosis). HIV-infected persons with significant exposure to someone with infectious TB should receive a full course of TB prophylaxis regardless of TST or IGRA results. Health departments can assist clinicians in assessing the degree of exposure for an HIV-infected patient and determining whether there is a need for a full course of treatment.
An HIV-infected person with fibrosis on a chest X ray that is consistent with previous TB and with no history of TB treatment (or a history of inadequate treatment) should be evaluated for active TB regardless of TST or IGRA results. If found not to have active TB, the patient should be treated for LTBI. If the patient is strongly suspected to have active TB, standard treatment for active TB should be given while culture analysis is under way. If cultures are negative, patients may be switched to LTBI treatment. Highly suspect but culture-negative patients may be given a 4-month total course of multidrug TB treatment. (See chapter Mycobacterium tuberculosis.)
As with any treatment of TB, adherence to the regimen is required for success. Treatment regimens for LTBI in the United States and other high-income nations are presented in Table 1.
|Drug||Dosage (taken orally)||Frequency||Duration (minimum number of doses for completion)|
|INH*||Adults: 300 mg Children: 5 mg/kg||Daily||9 months or 270 doses within 12 months|
|INH*||Adults: 900 mg Children: 10-20 mg/kg||BIW (DOT)**||9 months or 76 supervised doses within 12 months|
|Exposure to INH-Resistant TB or INH Intolerance|
|Rifampin#||Adults: 600 mg Children: 10-20 mg/kg||Daily||4 months or 120 doses in 6 months If used in children, 6 months is recommended|
|Exposure to Multidrug-Resistant TB|
|Treatment options depend on sensitivity of the organism.||Seek expert advice from public health authorities and those experienced in treatment of multidrug-resistant TB. Treatment may be postponed until sensitivity test results are available or may be based on resistance pattern of index case, if known.|
For treatment of LTBI in persons exposed to multidrug-resistant TB, there are no data on efficacy of various regimens in preventing progression to disease. Experts agree that it is important to follow those with presumed latent MDR TB infection for a minimum of 2 years following the exposure.
INH may cause liver toxicity and it should be used cautiously in patients with active alcohol use, liver disease, or chronic hepatitis B or C. INH is contraindicated for use in patients with acute hepatitis or decompensated liver disease. Before INH use, baseline liver and renal function should be tested. Routine monthly clinical monitoring for fever, fatigue, anorexia, nausea, vomiting, abdominal pain, jaundice, peripheral neuropathy, and rash should be performed. Alanine aminotransferase (ALT) should be monitored monthly in HIV-infected patients who have a risk of hepatitis. If patients develop abnormalities in liver transaminases while taking INH (ALT or aspartate aminotransferase >3 times the upper limit of normal [ULN] with symptoms, or >5 times above the ULN in the absence of symptoms), the INH should be withheld. Obtain expert consultation before treating patients with abnormal liver function or advanced liver disease.
Rifampin may cause liver and bone marrow toxicity. Before rifampin use, obtain baseline liver and renal function tests and a complete blood count. Follow-up is the same as for INH use.
Rifampin and rifabutin have important interactions with certain antiretroviral drugs, and dosage adjustments or treatment modifications may be required. Rifampin reduces the blood levels of nonnucleoside reverse transcriptase inhibitors (NNRTIs), protease inhibitors (PIs), the integrase inhibitor raltegravir, and the CCR5 antagonist maraviroc. Rifampin can be used by persons taking efavirenz although many experts recommend increasing the efavirenz dosage to 800 mg daily. Coadministration of rifampin and maraviroc is not recommended, but may be possible with appropriate dosage adjustment of maraviroc; consult with an expert. Rifampin should not be used with nevirapine, etravirine, or PIs. (In some cases, the adverse pharmacokinetic effect of rifampin on PIs may be overcome with large doses of ritonavir, but consultation with a specialist should be obtained before this approach is undertaken; rifampin should not be used in combination with ritonavir-boosted saquinavir because of high rates of hepatic toxicity.)
No data are available on the use of rifabutin for treatment of LTBI. Nevertheless, rifabutin may be considered in place of rifampin for patients taking antiretroviral combinations that include NNRTIs (other than efavirenz) or PIs (other than ritonavir alone). In these cases, the dosages of both rifabutin and the antiretroviral agent usually require adjustment. (See Table 3 in chapter Mycobacterium tuberculosis for details on dosage adjustments).
Rifampin decreases the blood concentrations of estrogens, anticonvulsants, hypoglycemic agents, and many other drugs. Review all medications a patient is taking before initiating rifampin and make adjustments as necessary. (See Table 12: Clinically significant drug-drug interactions involving the rifamycins.)
HIV-infected pregnant women with positive LTBI test results and no evidence of active TB should receive standard prophylaxis as soon as possible, even during the first trimester.
The preferred prophylaxis during pregnancy is a 9-month isoniazid regimen (with pyridoxine, as above). Alternative agents, such as rifampin or rifabutin, should be used with caution because of limited experience. Neonates born to women who received rifampin during pregnancy should be given vitamin K (10 mg) to reduce the risk of hemorrhagic disease. Pyrazinamide generally is avoided during pregnancy because of lack of information regarding fetal effects.