Molecular test for TB works best combined with sputum smear

A rapid molecular test for tuberculosis (TB) endorsed last year by the World Health Organization turns out to be most accurate and economic when it's used in combination with sputum microscopy, a new study from South Africa shows.

Dr. Keertan Dheda from Groote Schuur Hospital in Capetown and colleagues reported online November 10 in the European Respiratory Journal that performing sputum microscopy on all suspected TB cases, with the Xpert MTB/RIF reserved for use when smears are negative, is more effective than using either test alone -- and $115 cheaper than using Xpert MTB alone.

That finding is in line with WHO's December 2010 recommendation that the point-of-care Xpert MTB/RIF (Cepheid, USA) be used as a follow-on to sputum microscopy in isolated tuberculosis. The organization also recommends it be used as a stand-alone first test when TB is suspected in HIV patients, and for multi-drug resistant (MDR) TB.

The new report, however, helps fill a gap highlighted in a recent WHO global consultation on the rapid implementation of the Xpert MTB/RIF. Attendees at that conference called for more data to inform the use of other tests along with the Xpert MTB, say the authors of the current study.

The Xpert MTB/RIF is a real time PCR test that uses molecular beacon technology for TB diagnosis; it can also detect rifampicin resistance. Importantly, it can be used in resource limited settings.

The research team studied 480 patients with suspected tuberculosis from a HIV prevalent region of South Africa, including 130 with HIV infection. They collected two "spot" sputum samples; one was sent for the culture and smear microscopy, while the other was analyzed with the Xpert MTB/RIF. X-rays were performed where feasible.

One hundred forty-nine patients had a positive smear or culture indicating definite TB. Another 182 patients without definite TB but with clinical and radiological features were classified as "probable tuberculosis."

The sensitivity, specificity, and negative predictive value were 68.5%, 99.1%, and 87.5%, respectively, for the stand-alone sputum smear and 77.2%, 95.5%, and 90.3% for the Xpert MTB/RIF.

However, if the Xpert MTB/RIF were used only in patients with a negative sputum smear, the sensitivity, specificity, and negative predictive value of the combination rose to 81.9%, 94.9%, and 92.1%.

The WHO recommends using the Xpert MTB/RIF as the first-line test for diagnosing TB in HIV patients. The authors of the new paper say the specificities of smear, Xpert MTB/RIF, and the combination were similar in HIV-positive patients, but the sensitivities were lower.

Dr. Frank Cobelens from the University of Amsterdam, who has studied the Xpert TB but was not involved in the new study, advised in an email to Reuters Health that "if a clinical suspicion of TB remains despite a negative test (which may happen in e.g. severely immunosuppressed HIV patients who may have TB with low bacterial loads) additional tests would be warranted."

The cost per case of TB detected was $401 for the combination and $516 for MTB/RIF used alone.

Chest x-rays that were done in 324 patients ruled out tuberculosis with a sensitivity and negative predictive value of 100%. X-rays were particularly useful when the Xpert MTB/RIF was negative -- but of course, they added to the costs of diagnosis. At any rate, the authors note, the lack of equipment and technicians in remote areas limit the use of x-rays.

As for the role of Xpert MTB/RIF in detecting rifampicin resistance, Dr. Stephen Lawn from the Desmond Tutu HIV Center in Cape Town, who was not involved in the current study, had some comments for Reuters Health. "The Xpert assay gives two read-outs, the first relating to the presence or absence of TB and the second relating to the presence or absence of rifampicin resistance," he said in an email. "With regard to the latter, there are some reports of false-positive rifampicin resistance (as I reported in a PLoS Med paper) and confirmatory testing is needed for this. Thus, Xpert is a very good screen for rifampicin resistance but requires follow up confirmation with another test."

Xpert MTB/RIF is "more accurate than the traditional sputum smear microscopy, and offers rapid information on drug resistance," Dr. Madhukar Pai from McGill University, Montreal, Canada, said in an email to Reuters Health.

But Dr. Pai, who was not involved with the study, puts the discussion into perspective: "Most resource-poor countries currently offer only sputum smears for TB diagnosis. In addition, chest x-rays are often used. A small fraction of patients may get cultures for TB diagnosis and drug resistance. Molecular tests such as Xpert are rarely available, in part because of high costs."

The authors themselves concluded, "Detailed cost-effectiveness analyses are required."

The research team did not respond to requests for comment.

By C. Vidya Shankar MD

Eur Respir J 2011.

Medscape Today

http://www.medscape.com/

http://www.medscape.com/viewarticle/755709

To subscribe to the Weekly Newsletter of new posts, enter your email here:


Published: Dec. 22, 2011, 3:48 a.m.

Last updated: Dec. 22, 2011, 4:49 a.m.

Tags: None

Print Share