While qualitative tests are still very much in use in clinical diagnostic labs today — finding the presence of a specific microbe, or determining whether someone harbors a specific gene variant for disease susceptibility, for instance — it's quantitative tests that have come of age, thanks mostly to advances in real-time PCR instrumentation.
Real-time instruments with multi-well plates have been on the market for years, but the trend toward multiplexed assays, or running more than one PCR reaction in the same tube, has graced the stage only recently. Whether it's through automation, improved sample throughput, or multiplexing capabilities, the increased technical capacity of many instruments are paving the road for their use in diagnostics.
It's no surprise that clinical labs have migrated to real-time PCR for their bread-and-butter tests. "All labs worth their salt that have any level of sophistication that are doing new age analytes are almost certainly doing them using real-time PCR," says Dan Farkas, executive director of the Center for Molecular Medicine in Grand Rapids, Mich. "It's just so much simpler, one platform, it's less interaction with the tests by the technologists. It virtually eliminates the possibility of amplicon contamination."
Over at the University of Pittsburgh Medical Center, Jeffrey Kant directs the division of molecular diagnostics, which runs a lab serving the academic medical center and clinical researchers. Kant sees "an evolution away from qualitative testing toward quantitative testing," he says. "For viruses or fungi, the feeling is a test that can quantitate the level of something that's present is superior to a test that can just tell you whether it's present or absent."
One of the biggest strengths of real-time PCR is its reproducibility. That certainly comes in handy in clinical and diagnostic labs, where the key to good patient therapy monitoring is the ability to track a quantitative measure from day to day. "It's the reproducibility, that if I use your test today, get a result, and if a year from now I have to look back on it, there's consistency and accuracy and reproducibility in that result that I get," says John McCune, Roche's manager of molecular diagnostic reagent marketing.
The difference between a clinical laboratory and research laboratory is "one of rigor and one of reproducibility" — where running a test the same way each time is the key to success, says Farkas.
The multiplex factor
Vendors are reaching toward diagnostics not only with higher-throughput machines, but also with multiplexed assays. Running these tests — whether to look for multiple gene variants, a panel of gene expression transcripts, or the presence of more than one microorganism — has certainly been made easier by improved automation. However, bottlenecks remain, mostly in the form of sample handling and data analysis.
Several vendors are at the forefront. Applied Biosystems recently licensed its technology to Abbott Diagnostics, which, in collaboration with Celera, created the m2000 system. This platform automates the extraction and purification of DNA and RNA from patient samples prior to running real-time PCR. While most PCR instruments offer 96-well plates, several can run 384 reactions simultaneously. Roche will launch a 1,536-well plate instrument later this year and Fluidigm will upgrade its BioMark to run 96 samples.
Today, these instruments are mostly used for gene expression analysis. In a move to multiplex this process, several companies have launched assays that can test for many genes simultaneously. Agendia's MammaPrint looks at the gene expression profile of 70 genes and uses PCR followed by hybridization to a standard microarray chip. Genomic Health's Oncotype DX uses PCR to quantify the expression levels of a panel of 21 genes.
Farkas sees cancer management as being the end goal of many of the multigene panel tests coming out, including MammaPrint, Oncotype DX, AviaraDx's suite of tests for cancer, and Pathwork Diagnostics' test. "These are, of course, PCR-based tests; but then they depend on an array — it might be a chip, in the case of Pathworks Diagnostics, it might be a real-time PCR panel, which is arguably an array, in the case of the Oncotype DX test," he says. "I think these are examples of the next evolutionary step in molecular diagnostics." He adds that these multivariate diagnostic tests are "a little bit ahead of the curve … but just a little bit."
Mikael Kubista, head of Sweden's TATAA Biocenter, says that these "multimarker diagnostics" will become routine in the next five to 10 years, especially since the newer PCR instruments are becoming "really suitable for expression profiling."
Experts are realizing that the biggest hurdle no longer lies with optimizing the assay itself. "What we see in Europe, the bottleneck is no longer the PCR, it's what happens before and what happens after the PCR," Kubista says.
Having solid procedures for sample management and quality assurance helps eliminate the possibility of introducing error, lending robustness and reproducibility to the assay. In fact, Kubista predicts that in the future, real-time assays will be sent out to commercial labs that specialize in running particular ones. "This is nothing that you simply just transfer from one lab to another."
Optimizing the performance of multiplexed reactions also poses the challenge of making sure all reactions perform the same for all targets. However, says Roche's McCune, "the challenges that you face aren't necessarily technological; the challenges now are, what do I do with that information?" he says. "What's the value of it and what does it mean?"
One of the bigger challenges to developing tests that reach beyond the low-hanging fruit of single-gene assays are the regulations surrounding their FDA approval. Tests that incorporate complex data analysis will require FDA clearance. While laboratory-developed tests based on analyte-specific reagents ordered from common manufacturers are subject to low-level regulation, the general feeling in the community is that "if you're going to target more than one analytic end result, you're now stepping into ground that FDA feels it probably should be more actively regulating," says Jeffrey Kant at Pittsburgh. "In theory you could develop something yourself, which you validated, but if it's complex and particularly if it involves data analysis through non-intuitive algorithms, then the FDA is saying, 'That's something we need to look at.'"
Even as real-time PCR is perfected for the clinic, new applications continue to emerge. Kubista runs courses that cover the most cutting-edge applications for real-time PCR, and he believes that these tests will become even more elaborate. Rather than the existing instruments being overly optimized, he says, "What we actually see is, rather than the instruments being niched, you have different instruments for different niche applications."
High resolution melt is one example of those niche applications. HRM is a post-PCR method that compares the dissociation characteristics of different amplification products. It can be used to detect methylated DNA and to probe for SNPs, all with extremely high resolution and accuracy.
"HRM analysis of methylation state is an extremely sensitive, reliable, and cost-effective method," says Brant Bassam, marketing manager at Corbett Life Science. "It can discriminate the proportional status of locus-specific methylation versus non-methylation and readily distinguish heterogeneous methylation status."
The technique is offered by several vendors, including ABI, Idaho Technologies, Roche, and Corbett. It's still in early days, though, driven mostly by the scientific research community. Though Corbett doesn't promote HRM for methylation analysis, Bassam says that because it's so sensitive and reproducible, it's already being used in clinical research settings. "The most exciting application is the use of clinical samples in diagnostic and research settings. HRM sensitivity is remarkable, and low proportions of methylation amongst normal and heterogeneous tissue can be reliably detected."