Latent and Subclinical Tuberculosis in HIV Infected Patients

A Cross-sectional Study

Meaghan M Kall; Katherine M Coyne; Nigel J Garrett; Aileen E Boyd; Anthony T Ashcroft; Iain Reeves; Jane Anderson; Graham H Bothamley


BMC Infect Dis. 2012;12(7) 

In This Article


This paper has shown that screening for TB amongst those living with HIV is feasible. Screening showed that about 10% of our HIV clinic population were latently infected with TB, and three cases of active sub-clinical TB were detected. Screening among those with HIV is as effective as contact tracing, a standard procedure in TB control programs.[14] Since our study, the UK National Institute for Clinical Excellence (NICE) has issued guidelines which recommend IGRA testing for people with HIV (with CD4 200–500 cells/μl, and alongside a Mantoux test if CD4 < 200 cells/μl) and, if the test is positive, performing a clinical assessment to exclude TB and consider treating latent TB infection.[20] Our study supports these recommendations, although it is notable that we did not observe the test result to be sensitive to CD4 count.

Various algorithms have been used to detect active TB in those with HIV infection.[21,22] Symptom screening has been advocated. However, people living with HIV in South Africa and Zimbabwe also have high rates of active TB even when they are asymptomatic.[23,24] In this study, one patient did not admit to any symptoms until a positive test provoked further direct inquiry. Chest radiography is of limited usefulness and is complicated by atypical presentations and even a normal appearance in 14–23% of patients with culture-positive pulmonary disease and HIV co-infection.[25,26] A study in Tanzania found that of asymptomatic HIV-infected adults (CD4 >200, no TB symptoms and a normal chest radiograph), 5% had active TB diagnosed on positive sputum cultures.[27] Sputum examination is used, especially in areas of high TB incidence to detect active TB in those with HIV infection. A recent meta-analysis found that intensified case finding with microbiological investigation (sputum smear or culture) in all patients, irrespective of symptoms, detected an additional four TB cases per 100 HIV-infected individuals screened[28] and has been especially effective in areas of high TB incidence.[29] This was our rationale for collecting induced sputum even when history, examination, and chest radiograph were unremarkable. However, only 60% of those without sputum were able to provide a sample after nebulised hypertonic saline in this setting. In an area of low incidence of TB, a combination of symptoms, previous exposure to TB, previous use of HAART, weight <60 kg, CD4 count < 250/μl,[30] sputum smear and culture with or without an interferon-gamma release assay were the best option. However, we should note that two patients with active TB who denied any symptoms were only investigated further because of a positive immunospot test (Table 3), and therefore would have not been identified by the suggested algorithms. A systematic review of the interferon-gamma release assays in HIV/TB co-infection has noted a sensitivity of 80%.[3]

The rate at which those with a positive immunospot and HIV might develop TB is unknown. The importance of treating latent TB infection in those with HIV infection is suggested by earlier studies with the tuberculin skin test.[31] The interferon-gamma release assays were more accurate than the tuberculin skin test in predicting both progression to active disease and, in those with a negative test, the absence of progression.[32] A recent systematic review has suggested that these tests may be less useful in low or middle-income countries.[33] However, using the less sensitive tuberculin skin test in a low TB incidence country (United States), the rate of reactivation was 2.3 per 100 person years in HIV-infected individuals compared to 0.070 per 100 person years in those without HIV infection.[34] The WHO recommends that national HIV programs should provide TB preventive treatment for HIV-infected individuals with LTBI, provided that active TB has been excluded.[1] HAART significantly reduces the likelihood of TB in populations that have been exposed to infection.[35]

Treatment of LTBI with isoniazid monotherapy in HIV-infected patients reduces the incidence of active TB in individuals both on and off antiretroviral therapy.[15,36] The optimum duration of isoniazid is debated but the WHO recommends six months, which is a trade-off between efficacy and expected completion rates.[37] In our study, patients not on HAART were offered three months isoniazid plus rifampicin, which is equivalent to 6–12 months isoniazid in terms of efficacy and safety.[38,39] Preventive treatment was acceptable to patients in our study, with an uptake of 90% and adherence of 95% among those who started therapy. This is considerably better than reported completion rates of 45% in New York City,[40] and 54% in LTBI treated TB contacts in the same area of London[16] and suggests that in this population, isoniazid preventive therapy fulfils the requirements set for its implementation.[41]

Isoniazid-resistant TB occurred in one patient within two years of completing preventive treatment with isoniazid monotherapy. It is not known whether resistance was present in the original latent strain, or developed during preventive treatment, or whether the patient was newly infected with a resistant TB strain. Concerns have been expressed regarding the possibility that the use of isoniazid monotherapy in preventive treatment might increase the risk of isoniazid resistant strains of TB if the disease develops subsequently. However, a systematic review reported an overall relative risk of 1.45 (CI 0.85–2.47) comparing isoniazid-resistant strains in those given preventive treatment and placebo controls.[42] The results were the same when studies of HIV-infected and HIV-uninfected people were considered separately. Recent data from South Africa show that TB developing after recent isoniazid preventive therapy has a prevalence of drug resistance similar to background levels.[36] However, in vitro experiments have suggested that isoniazid resistance is more likely due to the use of the drug[43] and that the rate of mutation in latent TB may be the same as in active disease.[44] The latter paper recommends the use of a two-drug regimen for preventive treatment.[31,38]