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


Patient Characteristics

Over half (57%) of the cohort were women and 55% were patients from medium or high prevalence TB countries (30–300 and >300 cases per 100,000 population, respectively, Table 1). 542 blood samples were collected from 520 patients (138 patients in the pilot study and 382 in the expanded phase, including 22 patients who had repeat samples tested after their first sample gave an indeterminate result). The immunospot assay yielded 452 negative results, 50 positive results, and 40 indeterminate results (Figure 1).

Figure 1.

Flow diagram of patients tested with the immunospot test and subsequent diagnosis and treatment of LTBI. TB, tuberculosis; LTBI, latent tuberculosis infection; Drug regimens are abbreviated to the letters for the drugs administered: H, isoniazid; R, rifampicin; Z, pyrazinamide.

Indeterminate and Borderline Results

After the first round of testing, 36 patients gave an indeterminate test. Twenty-two patients were retested, of whom 17 were negative, two positive, and four indeterminate for a second time. Thus, out of 542 tests, 40 (7.4%) were indeterminate. The reasons for an indeterminate test were: 15 samples had a low cell number (too few circulating T-cells to produce interferon-gamma); in 11 the positive control failed; nine had a high background in the negative control panel; three failed to separate cells after centrifugation; and CO2 flow failure affected test performance in two tests. Median CD4 count for the forty indeterminate test results was 503 cells/μl (IQR 401–606), compared to 456 cells/μl (IQR 310–610) for the 502 determinate (positive or negative) tests. Low CD4 count at test was not associated with an indeterminate result (p = 0.198).

In the United States, the immunospot test has a borderline category, in which those with 5–7 spots are placed.[4] Analysing our data by these criteria, five patients with a negative response would have been classified as borderline – none have since developed TB. A further 5, who were classified as reactive, would have been assigned borderline status, all of whom completed preventive treatment and did not develop TB.

Analysis of Positive and Negative Immunospot Results

A comparison of demographic and clinical characteristics of those with positive and negative results is shown in Table 2. Excluding the 10 with a borderline result did not change any of the significant differences between negative and positive individuals.

Sex, age and the proportion of individuals receiving HAART were similar between the two groups. Subgroup analysis of patients on HAART showed that the mean duration of antiretroviral exposure between positive and negative tests was not significantly different (4.1 vs. 3.5 years, P = 0.352).

A higher proportion of patients of black African ethnicity had a positive immunospot result compared to others (13.0% vs. 4.7%); however ethnicity was not significant predictor of test outcome in the regression model. Being from a high TB prevalence country (OR, 5.01; 95% CI, 1.75–14.36; P = 0.003) or medium TB prevalence country (OR, 3.73; 95% CI, 1.11–12.52; P = 0.033) were the only significant predictors of a positive immunospot test.

Investigation for Active TB

All fifty patients with positive immunospot tests were referred to the TB specialist. Forty seven (94%) patients had chest radiographs, 11 of which were abnormal (Table 3). Sputum samples were taken from 30 (60%) patients and included all with an abnormal chest radiograph; the remainder were unable to produce a sample despite the use of hypertonic saline to induce sputum. Twenty-five patients gave induced sputum samples, four had a productive cough, and one patient with abnormal chest radiography underwent bronchoalveolar lavage.

Three women were diagnosed with subclinical active TB. One individual without any symptoms and a normal chest radiograph had a positive immunospot and sputum culture grew Mycobacterium tuberculosis. Another had previously denied any symptoms, but when questioned more closely after the positive immunospot and abnormal chest x-ray, then reported symptoms and was also culture-positive. A third showed significant weight gain after treatment for TB and did not take any treatment for HIV during the time of the study. Thus, subclinical tuberculosis was detected in 0.6% (3/502) of those who tested and 6% (3/50) of those with a positive result. A positive result was encountered in a patient with M. xenopi infection who had no risk factors for exposure to TB. During the same period of study, two patients with symptoms and who were therefore not included in the screening cohort proved to have TB and had a positive immunospot test. During the study period, a total of 33 HIV-infected patients were under treatment for active tuberculosis at the hospital, of which nearly 10% (3/33) had been initially diagnosed through screening the asymptomatic HIV-infected population.

Latent TB Infection: Prevalence and Treatment

The prevalence of diagnosed latent TB infection was 9.2% (46/502 i.e. excluding three with active TB and one probably false positive from M. Xenopi, out of those patients with a determinate result.) (Figure 1). All except two were offered preventive treatment and 40 (87%) commenced treatment. Reasons for not starting preventive treatment included non-attendance at appointments (3), patient refusal due to low perceived risk of TB infection (2), and deferral due to multiple co-morbidities (1). Thirty-five patients chose to be managed in the TB clinic and five chose the HIV clinic. To date, 38 (95%) have completed treatment, one stopped 10 days before completing 6 months treatment with isoniazid and another stopped at 6 months although the CD4 count was 131 cells/μl.

Follow up of Incident Tuberculosis

The median duration of follow-up from immunospot testing was 35 months (range 27–69 months). No patient with a negative immunospot has developed TB. One patient who had a positive test and was treated with six months isoniazid was subsequently diagnosed with isoniazid-resistant pulmonary TB after two years (Table 3, Patient 10). The patient successfully completed tuberculosis treatment with rifampicin, pyrazinamide, ethambutol, moxifloxacin and streptomycin.