TB Test Combination Rules Out Active Infection

By John Gever
LONDON, March 4 -- Combining new and old tests for latent tuberculosis infection can effectively rule out active TB, researchers here said. When results on the traditional tuberculin skin test and an experimental immunologic blood test are both negative, patients almost certainly do not have active disease, reported Ajit Lalvani, M.D., of Imperial College London, and colleagues in the March 4 issue of Annals of Internal Medicine. In a prospective study involving 389 adults with moderate to high clinical suspicion of tuberculosis, the investigators found a likelihood ratio for active disease of 0.02 (95% CI 0 to 0.09) when results on both tests were negative.
Point out that the original ELISpot assay is not yet approved in the U.S. and that the enhanced version is not available in any market.
Dr. Lalvani predicted in an interview that the findings would be "practice-changing."
The experimental test is an enhanced version of a diagnostic called ELISpot first developed about seven years ago in Dr. Lalvani's lab.
ELISpot is now commercially available in Europe under the name T-Spot.TB as a test for latent TB infection. FDA-approval is pending.
Both the original and the enhanced versions of ELISpot detect gamma-interferon released from T cells in a patient blood sample exposed in vitro to Mycobacterium tuberculosis antigens.
The antigens used in the original assay are derived from early secretory antigenic target-6 and culture filtrate protein-10. The enhanced version, dubbed ELISpot-Plus, adds additional peptides from another M. tuberculosis region, known as Rv3879c, to the antigen mix.
In the clinical study, patients with suspected TB underwent testing with the tuberculin skin test, the original ELISpot, and ELISpot-Plus.
Most patients were of South Asian or black ethnicity and were evaluated at two hospitals in England. Of the 389 patients, active TB was confirmed or considered highly probable in 194 on the basis of clinical examination and/or microbiological culture. In 154, active TB was ruled out. The rest were clinically indeterminate.
About one-quarter of all patients in the study had double-negative results on tuberculin skin testing and ELISpot-Plus, of whom only one actually did have active TB.
In 27% of the sample, ELISpot-Plus and tuberculin skin testing gave discordant results, which were of no diagnostic value, the researchers said.
Double-negative results with the original ELISpot and tuberculin skin testing were associated with a likelihood ratio of 0.04 (95% CI 0.02 to 0.12) for active infection.
Dr. Lalvani and colleagues found that both ELISpot assays in combination with tuberculin skin testing were highly sensitive for active infection.
In patients with culture-confirmed or highly probable clinical diagnoses of TB, positive results with both the original ELISpot and tuberculin skin testing had a sensitivity of 97% (95% CI 93% to 99%). Positive results on both ELISpot-Plus and tuberculin skin testing had a sensitivity of 99% (95% CI 95% to 100%).
However, ELISpot-Plus by itself had a specificity of only 69% among patients in whom active TB had been excluded by other tests. This was poorer than the specificity of tuberculin skin testing (81%; P=0.03), Dr. Lalvani and colleagues found.
In 121 patients with positive results on both tests, 15 were determined not to have active TB, for a specificity of 88%.
Similar specificity results were found with the original ELISpot combined with tuberculin skin testing.
Because of the imperfect specificity, Dr. Lalvani stressed that double-positive results on tuberculin skin testing and the ELISpot assays could not be used to diagnose active infection.
On the other hand, he and his colleagues said they could still be helpful in clinical evaluation.
"Double-positive results may help guide decisions about early initiation of presumptive treatment in severe disease while awaiting culture results and in extrapulmonary disease, in which culture is frequently negative," they wrote.
Dr. Lalvani added that double-negative results should not be the last word in making a diagnosis. "The results have to be taken in the overall clinical context," he said.
"If it's a very high pre-test probability [of active TB], then I think you have to go with the overall clinical picture, rather than a given blood test result."
Nevertheless, he compared the clinical potential of tuberculin skin testing plus ELISpot-Plus to that of D-dimer testing to rule out deep vein thrombosis.
In an accompanying commentary, Dick Menzies, M.D., of McGill University in Montreal, said the study has important strengths. They included its prospective design and careful clinical evaluation of all patients.
On the other hand, he said, it was limited by the fact that tuberculin skin testing was missing in 17% of patients and that the now-discontinued Heaf method for tuberculin testing was used in another 27% of patients.
He also pointed out that 21% of cases had no microbiological confirmation.
More importantly, he questioned the clinical utility of a 0.02 likelihood ratio for active infection with the combination of tests.
"The post-test probability after a negative result on both of these tests may still exceed some physicians' threshold probability for starting treatment of patients with a high pre-test probability of active tuberculosis infection," he wrote.
He noted that 95% of TB infections occur in populations where the disease is endemic. "In these populations, the search for a rapid accurate test must go on," Dr. Menzies concluded.
Both he and Dr. Lalvani said it would be important to confirm the study's findings in additional trials.
Dr. Lalvani said the ELISpot-Plus assay does not have a commercial sponsor. The study was supported by the Wellcome Trust, the Sir Halley Stewart Trust, and the U.K. Department of Health. Dr. Lalvani and two co-authors of the study reported past or current relationships with Oxford Immunotec Ltd., which markets the T-Spot.TB test. Dr. Menzies reported no potential conflicts of interest.

Additional source: Annals of Internal MedicineSource reference: Dosanjh D, et al "Improved diagnostic evaluation of suspected tuberculosis" Ann Intern Med 2008; 148: 325-36.

Additional source: Annals of Internal MedicineSource reference: Menzies D, "Using tests for latent tuberculous infection to diagnose active tuberculosis: can we eat our cake and have it too?" Ann Intern Med 2008; 148: 398-99.