|
 |
|
|||||||
|
|||||||||
Copyright ©ERS Journals Ltd 2007
From the authors
Depts of 1 Internal Medicine, 2 Pulmonology and Tuberculosis, and 3 Medical Microbiology and Immunology, Diakonessenhuis Utrecht, Utrecht, The Netherlands.
We would like to thank J-P. Janssens for his interest and comments on our recent paper 1. We fully agree that the decision to use the T-SPOTTM.TB (Oxford, Immunotec, Oxford, UK) to exclude active tuberculosis (TB) infection should not be taken lightly. We appreciate the opportunity to make some comments in reply to his letter.
First, he calculates the risk of suffering from active TB to be intermediate (0.25–0.75) in the patients described. His criteria for this risk estimation are not mentioned and could be arbitrary.
Secondly, because acid-fast bacteria were initially identified in all patients, one could easily say that the risk of active TB is high (>0.75). However, research in the Netherlands in 2005 demonstrated that 
600 (38%) out of 1,600 cultured isolates were Mycobacteria other than TB 2. This would decrease the a priori likelihood for M. tuberculosis (MTB) infection considerably.
Thirdly, in our case study we described four patients with a relatively high probability for active MTB infection (condition A) and the use of a test to confirm or exclude active MTB infection (condition B). As we noted in the paper, the sensitivity of the T-SPOTTM.TB test in our hospital, with our patient population, was 100%. To date, we have identified 33 patients with active TB. One patient had an indeterminate test result and the remaining 32 patients were all T-SPOTTM.TB positive. Thus, condition B has a probability of 1 and Bayes theorem should not be used if one of the conditional probabilities is 1 3.
J-P. Janssens correctly mentioned that the sensitivity of T-SPOTTM.TB for detecting active TB is not 100% in different studies. Therefore, we were cautious in our paper not to use this approach in severely immunocompromised patients and, furthermore, not to incorrectly interpret indeterminate results as negative. Case B was immunocompromised based on the use of steroids and azathioprine. However, steroids are not associated with indeterminate elispot results 4, and a positive elispot result has been demonstrated in a patient with chronic azathioprine treatment 5. We suggest each clinic must establish its own test performance before the T-SPOTTM.TB can be used to rule out active TB disease. Even then, we are reluctant to fully rely on negative T-SPOTTM.TB test results. Withholding therapy while awaiting culture results can only be justified if the patient’s condition is closely monitored and precautions to avoid further spreading of Mycobacteria are made.
Case A has been under observation for nearly 2 yrs and to date has not shown any signs of TB disease. In the other three cases atypical Mycobacteria (M. genavense, M. avium and M. malmoense) were cultured and treated.
Exclusion of active TB will not be feasible in countries with a high incidence of latent TB infection because the background T-SPOTTM.TB positive test results will be high and, as a consequence, the specificity to prove active disease will be low.
Infection control measures will always be undertaken the moment a patient is suspected to suffer from active pulmonary TB. In a hospital setting the patient will most likely be isolated from other patients, in an outpatient setting the patient will be told to stay at home and to not visit places with high numbers of people, such as pubs and bars, supermarkets or sport clubs. Furthermore, the stigma of suffering from TB is a real problem even in the 21st century. Emotional distress for the patient and their close contacts is not to be dismissed. In the Netherlands, the Municipal Health Authority will be notified when there is suspicion of active TB. The Authority will not wait for another 6–10 weeks before starting contact tracing. The first circle of contacts (household members and close friends) will be screened as soon as possible to identify a possible source patient or to diagnose other patients with active TB disease.
In conclusion, we underscore the general point made by J-P. Janssens that an interferon-
release assay should not be used lightly to exclude active tuberculosis. However, in a setting with: 1) low endemicity; 2) the possibility to follow the patient; 3) a proven track record of the facilitating laboratory; and 4) including the T-cell stimulation control test to detect nonresponsive (indeterminate) patients, such an approach is feasible.
REFERENCES
- van Leeuwen RM, Bossink AW, Thijsen SF. Exclusion of active Mycobacterium tuberculosis complex infection with the T-SPOT.TB assay. Eur Respir J 2007;29:605–607.
[Abstract/Free Full Text] - Dijkstra F, van Gageldonk-Lafeber AB, Brandsema P, et al. Respiratoire infectieziekten in het jaar 2005/2006. [Respiratory infections in the year 2005/2006.]. Infectieziekten Bulletin 2006;11:390–397.
- Hays WL. Statistics. 5th Edn. Oxford, Harcourt Brace College Publishers, 1994
- Ferrara G, Losi M, D'Amico R, et al. Use in routine clinical practice of two commercial blood tests for diagnosis of infection with Mycobacterium tuberculosis: a prospective study. Lancet 2006;367:1328–1334.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
- Richeldi L, Ewer K, Losi M, et al. Early diagnosis of subclinical multidrug-resistant tuberculosis. Ann Intern Med 2004;140:709–713.
[Abstract/Free Full Text]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |