Evaluation of a new rapid bacteriophage-based method for the drug susceptibility testing of Mycobacterium tuberculosis

It was estimated in 1991 that 1.7 billion people, about one-third of the world's population, had been infected with Mycobacterium tuberculosis, which in 1992 led to over 8 million cases of tuberculosis and about 2.7 million deaths. These numbers are likely to increase because of the influence of the pandemic of human immunodeficiency virus, which is a significant risk factor for the development of tuberculosis, and because of increasing concerns about drug resistance. Recently, there have been well-documented outbreaks of multidrug resistant (MDR) tuberculosis (strains resistant to isoniazid and rifampicin). Conventional culture-based drug susceptibility tests for M. tuberculosis are slow, taking one or more weeks to perform on a given isolate. Alternative genotypic tests have proved to be of some value in diagnosis and in detecting simple drug resistance genotypes such as resistance to rifampicin in which the majority of resistant genotypes map to a small specific region of the rpoB gene but have not been applied routinely to more complicated mechanisms of resistance such as that to isoniazid. We have developed a new phenotypic culture-based drug susceptibility assay, the PhaB (phage amplified biologically) assay, which addresses these limitations. It is based on the ability of viable M. tuberculosis bacilli to protect infecting mycobacteriophage from chemical inactivation and to support the replication of the infecting phage. When used to test the susceptibility to rifampicin the assay correctly assigned 1/1 (100%) rifampicin-resistant/isoniazid-sensitive isolates, and 35/37 (94.6%) rifampicin-sensitive isolates. When used to test the susceptibility to isoniazid, the PhaB assay correctly assigned 15 /17 (88.2%) isoniazid-resistant/rifampicin-sensitive isolates and 17 /21 (81%) isoniazid-sensitive isolates. The PhaB assay also correctly identified 8/8 (100%) MDR isolates. The PhaB assay correctly identified the drug susceptibility of the majority of isolates with the results obtained within 3-4 days. The simplicity and low cost of the assay should enable it to be performed in those parts of the world that have the least resources yet suffer most from the burden of M. tuberculosis infection. Although this evaluation was performed on cultured isolates, the sensitivity of the assay is such that it would be possible to perform it directly on patient samples, which could reduce the time taken for front-line drug sensitivity testing to 3-4 days from a total of