Molecular Diagnostics: Bringing Together Multi-Omics Approaches

Molecular diagnostics—roughly defined as assays looking at multiple biomarkers along with an algorithm to yield a single patient-specific result—is getting its fair share of attention at conferences and in the journals, as well as in the popular press. After all, to tell from a sample of blood why a child isn’t developing cognitively as expected, or whether a particular targeted cancer treatment will likely help or harm, is exciting. Improvements in technology have enabled researchers (and in some cases clinicians) to go beyond a single genetic or protein variant. In particular, next-generation sequencing (NGS) and mass spectrometry (MS) are allowing unprecedented numbers of parameters to be simultaneously interrogated, generating genomic, proteomic or even multi-omic signatures that can literally spell the difference between life and death.

Genomics

When discussing molecular diagnostic assays, it’s important to note at the outset that these are typically not for biomarker discovery as much as for looking at a collection of known marker candidates. The literature searches, genome-wide association studies (GWAS), functional assays linking genes or proteins or RNA expression with disease or therapeutic outcome—and more—have already been done. Researchers are now whittling down the collections of candidates, or validating a selection of candidates in larger cohorts or against different populations, to come up with a multivariate signature.

Clinical laboratories (and medical-device manufacturers) are creating tests based on these signatures to help with patient treatment decisions. 

There are a variety of methods to query genomic signatures from diverse sample types. PCR panels can be used to find single nucleotide variations (SNVs), insertions, deletions and some larger chromosomal abnormalities, for example. ARUP Laboratories, a nonprofit national reference laboratory, offers a PCR-based panel for cystic fibrosis (CF) that queries 165 variants of the CFTR gene from cellular DNA collected from at least one milliliter of whole blood. ARUP also offers several other CF tests based on multiplex ligation-dependent probe amplification (MLPA) and Sanger sequencing. Cytogenomic microarrays and NGS are available for other indications.

Unlike a strictly research lab, clinical reference labs “have to prove that we can do something—we have to validate whatever we offer to our patients,” explains Elaine Lyon, medical director for molecular genetics/genomics at ARUP laboratories and a professor of pathology at the University of Utah School of Medicine. ARUP and other CLIA labs are moving many assays over to NGS—especially those involving large numbers of variants, whether for a single gene or multiple genes.

At the same time, PCR is continuously evolving. Enzo’s AmpiProbe™ target amplification platform, for example, provides greater sensitivity from a smaller sample size; it does so by using Förster resonance energy transfer (FRET) to deliver a larger number of PCR cycles on almost any open qPCR platform “with virtually no background,” says Enzo Biochem’s vice president for corporate development David Goldberg. The New York State Department of Health-approved AmpiProbe Candidiasis™ assay, for example, is a multiplex assay to identify and discern the presence of five of the most common species of Candida from a single vaginal swab, “allowing the physician to really tailor the diagnostics and treatment to the patient’s condition.”

And in a recently published prospective trial of nonsquamous non-small-cell lung cancer patients, Bio-Rad’s droplet digital PCR (ddPCR ) platform was used to identify several mutations in the EGFR and KRAS genotype from circulating cell-free DNA (cfDNA). The study concluded that ddPCR assays offer the speed and “high specificity needed to select therapy and avoid repeat biopsies.”

Profiling the presence of particular mutations in tumor-derived cfDNA is emerging as a way to track and identify drug-resistant clones, says George Poste, Regents’ Professor and director of the Complex Systems Initiative at Arizona State University. “We’re building up an inventory of which particular substitution in the protein confers resistance against drug x or y, and in many instances that can be picked up at the level of a mutation in the genome,” he explains. Some people are using PCR, he says, but the big push is to use NGS.

NGS panels

“NGS is all about multiple markers … the ability to aggregate either individual genes, mutation types or expression targets into a single panel,” says Andy Felton, vice president of product management for clinical next-generation sequencing at Thermo Fisher.

NGS can be used to look at the entire genome, the regions coding for mature proteins (the exome) or a more targeted selection.

NGS is also used to query mRNA, microRNA and methylated DNA, among other more defined nucleic acid compartments, signatures of which are all being explored as potential diagnostics. Whole genome sequencing (WGS) and whole exome sequencing (WES) “are primarily research vehicles—the cost of doing these right now is prohibitive if you want to look at biomarkers,” says Felton.

That being said, WES is sometimes used for “diagnostic odysseys,” to look, for example, for possible causes of neurodevelopmental disorders.

Several vendors market targeted NGS panels. Thermo Fisher Scientific offers several NGS panels, including the Oncomine™ Comprehensive Assay, which “covers 143 genes and all mutation types—SNPs, indels, CNVs and gene fusions that are relevant to oncology,” as well as smaller, more disease-specific panels, notes Felton. Thermo Fisher currently markets panels “for research use only” (RUO).

An RUO marking, of course, may not prevent a CLIA lab from using them as clinical tests.

What about protein?

Protein signatures are also being used and further explored in molecular diagnostics. As in the genome and transcriptome, there are many assays that are utilized—from ELISA and Luminex-based immunoassays to multiplex immunohistochemistry (IHC ). But the NGS of the protein biomarker is mass spectrometry (MS).

When genes undergo alterations, there is often a commensurate downstream change in proteins and small molecules. Improvements in MS over the last five to 10 years have enabled researchers to find and measure increasingly minute levels of protein variants including, for example, different isoforms and post-translational modifications.

But unlike for NGS, Thermo Fisher does not currently offer any specific assays for MS. The company offers a portfolio of HPLCs and mass spectrometers approved for IVD (in vitro diagnostics) use, which have “the capability to measure multiple analytes,” says Brad Hart, strategic director, life sciences mass spec, clinical research at Thermo Fisher Scientific. “What happens is the researchers or lab – it could even be a large commercial clinical reference lab–use the instrument for their specific use and their specific LDT [laboratory developed test].  Today, we supply standards and calibrators to make sure the instrument is running properly, but no specific assays.”

And by using isobaric mass tags to label samples for protein analysis in clinical research, for example with Thermo Fisher Scientific’s Tandem Mass Tag™, it’s possible to interrogate up to 10 samples simultaneously in a single MS run. “This is one of the ways that MS is starting to approach the throughput of microarrays and NGS technologies,” Hart notes. Thousands of proteins in multiple samples can be measured and quantified in less than an hour.

Not alone

Of course, MS can measure more than just protein. It is the instrument of choice for metabolomics and some other ’omics, as well. Poste advocates a systems-biology approach, warning against “the grave danger of just generating a huge amount of data that can’t be translated into meaningful functional parameters.” Multi-omics, or panomics (studying the genome, epigenome, proteome, glycome, miRNAome “and all the other ’omes that are now part of this vernacular with regard to molecular profiling”), conducted by a broad, interdisciplinary team, is the key to understanding the “complete wiring diagram … whether it be in health or in disease,” Poste explains.

And because no one is going to be able to collect evidence for all the variations, even for a single 'ome, we need to share the data and make it available in a public database, says Lyon. “We are submitting genetic variants that we’ve identified to the NCBI’s ClinVar, along with the evidence we used to classify [the variant] as pathogenic, uncertain or benign. It’s becoming a central hub to see if other labs have identified the same variant, and if so, what they know about it.”