A team comprising researchers from Johns Hopkins University, Texas A&M University, and a Panamanian research institute has demonstrated how multiplex allele-specific PCR, or MAS-PCR, could prove an accurate and inexpensive tool for rapid detection of multidrug-resistant tuberculosis in developing countries.
Having vetted their method, the scientists are now attempting to translate it into a true point-of-care assay, a format that has proven challenging for PCR-based assays but that may be facilitated by the use of isothermal amplification techniques such as loop-mediated isothermal amplification, Petros Karakousis, a JHU researcher and corresponding author on the study, told PCR Insider this week.
Karakousis and colleagues at the JHU Schools of Medicine and International Health, Texas A&M, and Panama's Instituto Conmemorativo Gorgas de Estudios de la Salud, or ICGES, described their work in a paper published in July in PLoS One.
In their study, the researchers used phenotypic drug-susceptibility testing, the current gold standard for detecting MDR-TB, to identify 67 MDR-TB isolates and 31 drug-sensitive clinical isolates collected between 2002 and 2011 at ICGES.
Although phenotypic drug-susceptibility testing is the gold standard in terms of accuracy for identifying MDR-TB, it is an arduous and slow technique requiring significant laboratory expertise and infrastructure, making it generally unsuitable for quickly identifying and containing outbreaks in resource-limited areas of the world.
Several companies and groups have tried to address this issue by developing molecular assays for MDR-TB, and, in particular, real-time PCR-based methods have become increasingly popular. For instance, Cepheid's Xpert MTB/RIF, after receiving endorsement from the World Health Organization, is now well-established in many resource-poor areas of the world, making it the most successful such commercially available test. Meantime, researchers at the National University of Ireland Galway recently developed and validated a two-stage multiplex real-time PCR assay to identify all members of the Mycobacterium tuberculosis complex (PCR Insider, 5/24/2012).
However, many public health officials still bemoan the relatively high cost of real-time PCR-based methods as compared to cell culture-based methods. In addition, besides the NUI-Galway assay, there are few examples of real-time PCR assays that can detect the number of targets needed to capture all possible drug-resistant strains of M. tuberculosis — although Cepheid and collaborators are currently working on such a test.
As an alternative, the JHU-led team turned to MAS-PCR, an endpoint PCR-based technique previously described in a paper published in 2005 in Diagnostic Microbiology and Infectious Disease. The method as used by the JHU-led team simultaneously amplifies the five most common mutation hotspots that confer resistance to the first-line drugs isoniazid and rifampin, with the end products detected using capillary electrophoresis.
"Essentially this does something very similar [to the Cepheid assay], except that it doesn't require an expensive piece of equipment," Karakousis told PCR Insider. "One of the differences in this multiplex PCR is that it actually detects isoniazid resistance, as well, which the Cepheid test doesn't — it just looks for rifampin resistance."
In the majority of MDR-TB cases, that's OK, Karakousis noted, since 90 percent to 95 percent of rifampin-resistant M. tuberculosis isolates are also resistant to isoniazid, making detection of the former a surrogate for detecting MDR-TB.
However, "resistance to isoniazid, potentially in the absence of resistance to rifampin" can also be "useful clinical information," Karakousis said. "With Cepheid … if you get a positive result for TB but a negative for rifampin, then your interpretation is that it's a drug-susceptible isolate, when it may be in fact resistant to isoniazid alone," a situation that is occurring more frequently in several geographies, particularly developing countries, he added.
Karakousis and colleagues used whole-genome sequencing to confirm drug resistance mutations identified using MAS-PCR, and provided the frequency of the mutations, revealing that 70.1 percent of MDR strains had point mutations at codon 315 of the katG gene, 19.4 percent had mutations within the mabA-inhA promoter, and 98.5 percent had mutations at three hotspots within rpoB.
MAS-PCR detected each of these mutations, yielding 82.8 percent sensitivity and 100 percent specificity for isoniazid resistance, and 98.4 percent sensitivity and 100 percent specificity for rifampin resistance relative to phenotypic drug-susceptibility testing.
The researchers noted in their paper that "the high sensitivity and specificity of MAS-PCR in detecting MDR-TB, along with its low cost relative to other rapid molecular assays and ease of use make it an attractive alternative for rapid detection of MDR" relative to phenotypic drug-susceptibility testing.
"Further analysis of the available whole-genome sequences of these archived MDR-TB isolates could reveal novel mutations associated with [isoniazid and rifampin] resistance, as well as mutations that confer resistance to pyrazinamide and second-line drugs," they added. "These findings could then be used to further refine the MAS-PCR assay and improve detection of [drug resistance] in M. tuberculosis clinical isolates."
The assay also needs further refinement in terms of its ability to be employed at the point of care, Karakousis said.
"Ultimately we're trying to move toward a point-of-care test, so this still obviously requires some equipment, including a PCR machine and things like electricity," he said. "It's fairly crude right now, not real-time. It doesn't require a fancy piece of equipment, but at the same time there are some requirements for being able to determine the patterns on the gel."
Nevertheless he described the assay as "fairly straightforward, because we had undergraduate students doing this alongside more experienced people. It doesn't require a high degree of sophistication to interpret the results."
The researchers are now taking additional steps toward adapting their assay for point-of-care use, starting with eliminating the need for a thermal cycler by using isothermal amplification methods like loop-mediated isothermal amplification, or LAMP, an assay technology owned by Japan's Eiken Chemical but being explored for use in a variety of POC molecular assays.
"That still requires electricity, but a reaction could be done in a water bath a constant temperature," Karakousis said. He also noted that the group is exploring a collaboration with German research organization Fraunhofer, which has developed an optical sensor technique that is able to detect small amounts of DNA.
"This may even get around the amplification steps," Karakousis said. "It's more like a hybridization. They've done proof of principle [in] biodefense, like detecting trace amounts of dynamite. But they've also used it for things like sepsis, detecting small numbers of biomarkers in blood samples. So the question is, 'Can they detect DNA?' This potentially wouldn't require refrigeration because these oligos are fairly stable."
In the meantime, the research group hopes to begin a prospective study of its assay in Panama or other countries where MDR-TB is problematic.
In the work so far, "we already knew [the strains'] drug resistance patterns using conventional phenotypic testing," Karakousis said, but for the prospective study, "we would run this assay on patients coming in for a rapid result, but would also do traditional testing — which might take up to four weeks — then compare the results."