The World Health Organization endorsed the rapid implementation of the GeneXpert MTB/RIF test for rapid tuberculosis detection, multidrug-resistant TB (MDR-TB) and TB in HIV infected individuals in December of 2010 – taking time to diagnosis down to 100 minutes, whereas other tests can take up to three months to deliver results. At that time, however, there was no data available on the efficacy of the test among children.
Mark Nicol from the University of Cape Town reported on his ongoing work to evaluate Gene Xpert in children at the International Union against Tuberculosis and Lung Disease in Lille, France this week. Published in the Lancet, his study “Accuracy of Xpert MTB/RIF test for the diagnosis of pulmonary tuberculosis in children admitted to hospital in Cape Town, South Africa: a descriptive study,” focused on children with a median age of 19 months. Shockingly, 10 percent of the children already had a prior history of tuberculosis. It was clear from the study that GeneXpert was a more effective diagnostic tool than smear microscopy, which only detected one-third of the active TB cases, and chest X-ray, which over-diagnosed TB in the children in the study. GeneXpert also seemed to work better in HIV-infected children than those without HIV infection.
GeneXpert had a lower ability to detect smear-negative TB than has been documented in adult studies and generally, detected about two-thirds of the cases diagnosed through culture – the gold standard of TB diagnosis. A single GeneXpert test was successful in diagnosing only 57 percent of the cases confirmed with culture, less sensitive than it has been shown to be in adults. This suggests that one GeneXpert test may not be inadequate for children, although the high specificity of this diagnostic can assure the clinician that a positive GeneXpert test will virtually always mean the presence of active TB disease.
Notably, 65 percent of the children placed on TB treatment in this prospective study had both a negative culture and a negative GeneXpert test.
According to Nichol, GeneXpert should not replace culture where culture is available, but can be very useful in rapid diagnosis and also very valuable in instances where MDR-TB is suspected. It is important to evaluate GeneXpert in children with less severe illness, and Nicol and others are now conducting a study in the community in Cape Town. So far the results do not look dramatically different. Nicol also noted the need to continue to evaluate the performance of GeneXpert in a larger group of HIV-infected pediatric patients.