Tuberculosis diagnosis--time for a game change.

The effective treatment of tuberculosis is a lifesaving intervention. The global scale-up of tuberculosis therapy has averted 6 million deaths over the past 15 years, making it one of the greatest public health interventions of our lifetime.1 Unfortunately, by the time most patients are treated, they have already infected many others.2 This failure to interrupt transmission fuels the global epidemic so that every year there are more new cases of tuberculosis than in the previous year.1 National tuberculosis programs are particularly challenged by multidrug-resistant tuberculosis. Globally, fewer than 2% of the estimated cases of multidrug-resistant disease are reported to the World Health Organization (WHO) and managed according to international guidelines. The vast majority of the remaining cases are probably never properly diagnosed or treated, further propagating the epidemic of multidrug-resistant tuberculosis. The situation is further worsened by the epidemic of human immunodeficiency virus (HIV), especially in Africa. For decades there has been little effort to improve techniques for diagnosing tuberculosis.3,4 Consequently, tuberculosis tests are antiquated and inadequate. The most widely used test (smear microscopy) is 125 years old and routinely misses half of all cases. These inadequacies are particularly problematic since such tests are generally performed in underfunded and dysfunctional health care systems.4,5 The problem is exacerbated by the widespread use of inaccurate and inappropriate diagnostic tools, such as serologic assays, in many countries.6 Fortunately, in the past few years, several improved tuberculosis tests have received WHO endorsement for widespread use.6,7 In this issue of the Journal, Boehme and colleagues8 describe a new automated nucleic acid–amplification test that may allow a relatively unskilled health care worker to diagnose tuberculosis and detect resistance to a key antibiotic within 90 minutes. This test and others that are likely to follow have the potential to revolutionize the diagnosis of tuberculosis. Thus, in the coming years, rapid diagnosis and targeted treatment will provide the greatest opportunity for stopping the tuberculosis epidemic. In a large, well-conducted, multicountry study, Boehme et al. evaluated an automated tuberculosis assay (Xpert MTB/RIF) for the presence of Mycobacterium tuberculosis (MTB) and resistance to rifampin (RIF). With a single test, this assay identified 98% of patients with smear-positive and culture-positive tuberculosis (including more than 70% of patients with smear-negative and culturepositive disease) and correctly identified 98% of bacteria that were resistant to rifampin.8 The assay has several critical advantages over conventional nucleic acid–amplification tests, which have been licensed for nearly 20 years and yet have not had a substantial effect on tuberculosis control. The MTB/RIF assay is simple to perform with minimal training, is not prone to cross-contamination, requires minimal biosafety facilities, and has a high sensitivity in smear-negative tuberculosis (the last factor being particularly relevant in patients with HIV infection).8 However promising these findings, issues involving the MTB/RIF assay may limit its global utility. These issues include its high cost, limitations in testing only for rifampin resistance, a platform that detects a relatively small number of mutations, and inability to indicate which patients are “sputum smear–positive” for reporting purposes, infection-control intervention, and treatment monitoring.