Article Text


Tuberculosis: epidemiology

Statistics from


1N. Javanshir, 1D. Shah, 1J. Ivanyi, 2H. J. Milburn. 1King’s College, University of London, London, UK, 2Guy’s and St Thomas’ NHS Foundation Trust, London, UK

Diagnosis of tuberculosis (TB) disease should ideally involve culture and sensitivity of the organism but in low income countries this is not practised routinely. The World Health Organization estimates that, globally, substandard detection occurs in 40% of patients, with many diagnosed on clinical suspicion or response to anti-TB drugs. TB produces a strong antibody response suitable for serodiagnostic assays which can be simple, inexpensive and rapid. Attempts to use serological tests for diagnosis have, however, proved disappointing.

Aim We are currently evaluating a new point of care rapid serological test based on lateral flow immunochromatography (TB-ST Rapid Test, Lionex, Germany) to determine sensitivity and specificity. This test will then be further assessed in a resource-poor setting (Kakamega and Eldoret, Kenya).

Methods The test is currently being evaluated in 50 patients with confirmed active TB (pulmonary and extrapulmonary), 50 with latent TB infection (LTBI) (household contacts with a Mantoux skin test response ⩾10 mm and absence of clinical or radiographic evidence of active disease) and 50 healthy non-infected controls. Two drops of whole blood from a finger prick plus three drops of diluent are applied to the well in the test plate. Within 5–15 min a negative result shows just a visible control band while a positive result shows an additional visible test band.

Results To date, 44 controls, 4 patients with LTBI and 12 with active TB have been tested. All controls and those with LTBI tested negative. Only 3/12 patients with known active TB tested positive (2 pulmonary, 1 lymph node, all smear positive). All 3 had not received any treatment for TB whereas the other 9 had all started therapy (range 3 days to 2 months), giving a sensitivity to date of 25% and specificity of 100%.

Conclusions The results to date indicate that this test is not suitable as a diagnostic aid in patients who have received any amount of TB medication. With further numbers it may prove helpful with diagnosis in untreated smear negative or extrapulmonary disease, although our positive results to date have all been in patients with large numbers of organisms.


H. S. R. Hosker, J. Anderson, A. Crabtree, P. Godwin. Airedale NHS Trust, Keighley, UK

Interferon-gamma release assays (IGRAs) have an established role in contact tracing, new entrant and occupational health screening for tuberculosis (TB) (NICE/HPA recommendations). Their role in supporting or ruling out TB diagnosis in suspected cases (when microbiology is unavailable) is less clear.

We retrospectively analysed results from 95 patients who had a T-spot test performed as part of their investigation for possible TB. Patients presented with a variety of symptoms and clinical findings; TB microbiology was either unavailable or cultures were ongoing at the time of testing. No patient had HIV infection or other significant immunosuppression.

29 (31%) of 95 suspected cases had a positive T-spot test. 11 (38%) were true positive based on subsequent positive cultures or subsequent response to TB therapy. 11 (38%) other positive results were probably due to latent TB (LTBI) from old TB many years ago or recent TB contact with no current evidence of active TB. Seven cases (24%) were false positives in the context of various other diagnoses (including two cases of atypical mycobacteria). Of the 16 cases subsequently found to have active TB (excluding probable LTBI cases), 5 (31%) had a negative T-spot result (false negative). Four of these had “systemic” extrapulmonary TB (2 miliary, 1 TB meningitis, 1 TB empyema) with likely generalised immune paresis. 11 of 12 (92%) cases of pulmonary or mediastinal TB had a positive T-spot result. Overall specificity was 90% and sensitivity was 81% in cases of suspected TB (table 1).

Abstract P80 Table 1

We conclude that, in cases of suspected TB, (a) a negative TB spot is useful in ruling out pulmonary TB but (b) patients with extrapulmonary TB often have a false negative result and (c) a positive T-spot might be due to co-existing LTBI or other pathology in some cases. The results of IGRAs in patients with suspected TB need to be interpreted with caution.


1D. W. Connell, 2R. Charif, 2N. Duncan, 3C. McCrudden, 4E. Harden, 4S. Seneviratne, 1O. M. Kon. 1Department of Chest and Allergy, St Mary’s Hospital, Imperial College Healthcare NHS Trust, London, UK, 2West London Renal and Transplant Centre, Hammersmith Hospital, Imperial College Healthcare NHS Trust, London, UK, 3Hammersmith Hospital, Imperial College Healthcare NHS Trust, London, UK, 4Department of Immunology, St Mary’s Hospital, Imperial College Healthcare NHS Trust, London, UK

Introduction and Hypothesis There are few studies on the use of interferon-gamma release assays (IGRAs) in the diagnosis of tuberculosis (TB) in patients with renal disease. We describe the real-life contact tracing of 62 patients from a general renal inpatient ward in which a staff member was found to have pulmonary TB. We hypothesised that IGRAs would diagnose more cases of TB in the cohort, and would be more associated with exposure to the index case than the tuberculin skin test (TST).

Methods A healthcare worker was found to have pulmonary TB after a period of work on a renal inpatient ward. Patients were invited for standard screening with a TST, and IGRA testing with the T-SPOT and Quantiferon (QFT) tests. Concordance between the tests was calculated with χ2 testing, as was the effect of length of exposure to the index case on the test results.

Results 61 patients were contact traced from the cohort of 62. There was a broad spectrum of ethnicities and modalities of renal replacement therapy (haemodialysis in 45%). 29 patients did not receive a completed TST despite standard follow-up by experienced TB nurses. 12 patients (19.6%) had a positive QFT, 13 patients (21.3%) had a positive T-SPOT and 5 patients (8.1%) had a positive TST. No patients had evidence of active TB disease. 25 patients had all three tests performed. Their results are summarised in fig 1. In this group there was a significant association between T-SPOT and QFT results (p<0.002, Kappa 0.694). There were no significant associations with TST and either IGRA (p>0.05, Kappa <0.37 for both). Length of exposure to the index case had no effect on test results (Mann-Whitney U test, p>0.1 for all tests).

Abstract P81 Figure 1

Twenty-five patients (41%) had all three tests. Of this group, four patients had a positive tuberculin skin test (TST), five had a positive Quantiferon (QFT) and eight had a positive T-SPOT.

Conclusions This is the first UK description of a real-life comparison between TST with IGRAs in a population with mixed kidney disease. It demonstrates the difficulty of performing the TST in these patients. The results of the IGRA tests were significantly associated with one another, suggesting they may be a more reliable test in this population for the diagnosis of TB.

View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.