Josh P. Roberts has an M.A. in the history and philosophy of science, and he also went through the Ph.D. program in molecular, cellular, developmental biology, and genetics at the University of Minnesota, with dissertation research in ocular immunology.
Techniques for monitoring tumor biomarkers are finding their way into the clinic as important tools for patient stratification and prognosis. The biomarkers are ideally obtained in a relatively noninvasive manner—from blood or urine, for example—but could be derived from a biopsy, as well. The National Cancer Institute (NCI) website notes that “a number of tumor markers are currently being used for a wide range of cancer types” and lists about 35 such markers “currently in common use.”
These assays are found as either in vitro diagnostic devices (IVDs), or laboratory-developed tests (LDTs) that look at individual markers or panels of markers that collectively define a signature, which physicians can order. Here we look at how researchers are examining various biomarkers and using the information to further their understanding of disease states.
What’s being detected?
The NCI’s 35 common biomarkers include numerous normal and mutated genes, proteins and even cells found in tissue or bodily fluids. Most, like prostate-specific antigen (PSA), are normal proteins that are found in elevated levels in the blood of cancer patients. They may also be collections of protein antigens or genes found in tumors that together make up a “signature” indicating, for example, the most advantageous course of treatment. Mutated genes, translocations, copy number variations and other genetic abnormalities are also queried on the genetic, RNA and protein levels. CELLSEARCH®, the first and only clinically validated, FDA-cleared IVD system, collects circulating tumor cells (CTCs) or epithelial cells that stain positive for certain antigens thought to be shed by distal tumors to provide biomarker characterization.
Researchers are also looking at a host of other markers, ranging from cell-free nucleic acids, to epigenetic transcriptional control indicators such as DNA and chromatin methylation, to protein post-translational modifications (PTMs), that may indicate disease state, prognosis or therapeutic course.
Single or multiplex?
Assays have typically been designed to look at only a single, or perhaps two, biomarkers at one time, “largely due to the exponential complexity that comes when assessing and understanding the impact of multiple biomarkers for disease correlation,” notes Guy Afseth, director of oncology product marketing at Affymetrix. The advent of technologies such as next-generation sequencing (NGS) and microarrays (and the accompanying bioinformatics) has made it practicable to simultaneously interrogate multiple biomarkers—even the entire genome or transcriptome. Targeted panels—looking at, say, five to 50 genes or proteins—are often used to query sets of putative biomarkers.
That doesn’t mean the resulting clinical assay will require whole-genome sequencing. “As biomarkers are further narrowed and validated and move toward the clinic, the technologies typically move to lower plex,” notes Afseth. These can then be queried by PCR, immunohistochemistry (IHC) or in situ hybridization (ISH), for example, with multiple biomarker signatures using midplex assays as such as flow cytometry, Luminex or targeted NGS.
Tests of even higher complexity, such as Tissue of Origin Test (Cancer Genetics Incorporated) and the MMprofiler (Skyline Dx), have been developed based on microarrays, Afseth points out.
Now trending
In pre-clinical settings, researchers use every conceivable form of checking and validating targets “to build their case for taking new targets, with new drugs, into clinical trials,” including Western blotting, two-dimensional gel electrophoresis, ELISA, mass spectrometry (MS) and RNA-based PCR, says Nicola McCarthy, oncology program manager at Horizon Discovery. NGS is popular, she notes, but it doesn’t tell anything about protein expression levels.
And even the total amount of a protein in many cases may not mean as much as the level of the phosphorylated version that’s active in signal transduction, notes Daniel Braunschweig, global product manager for Bio-Rad Laboratory’s Bio-Plex products. Bio-Rad offers about 50 Luminex-based assays for cell signaling proteins and more than 100 nonphospho targets that can be related to cancer research, some of which can be found in multiplexed panels.
Anthony Couvillon, scientific marketing project manager at Cell Signaling Technology, sees a big trend in people looking at PTMs beyond phosphorylation, as well. “People are starting to appreciate the role of acetylation, methylation, ubiquitination, oxidation … even GlcNAcylation and ADP-ribosylation.”
IHC is a standard way of querying the presence of a biomarker in its morphological context, but it’s generally restricted to one or two colors. With tyramide-amplified multiplex IHC—which is “becoming increasingly appreciated as being a great diagnostic and stratification tool for tumors,” notes Couvillon—up to eight different antigens can be targeted. Multiplex IHC lets researchers study PD-L1 in the context of E Cadherin or other tumor markers, for example, while at the same time trying to understand whether the tumor has undergone the EMT transition, and whether Th or Treg cells have infiltrated the tumor, he explains. But the technique, he cautions, requires “exquisitely specific antibodies.”
RNA ISH enables researchers to visualize RNA biomarkers—including noncoding RNAs such as lncRNA and miRNA—quantitatively and in context, as well.
Advances in tissue culture, such as the ability to keep tumor slices viable to study interactions between the heterogeneous cells therein, are among the factors piquing Big Pharma’s increased interest in live imaging. As far as McCarthy is aware, this technique hasn’t made its way into the clinic, either in oncology clinical trials or to classify patients who are being treated, but it is “coming down the pipeline.”
And of course, “there is also much more interest now in the proteomics side, because it’s gotten cheaper and more accessible,” says McCarthy. MS provides “the most massively multiplex answer, but just for a single sample at a time,” Braunschweig points out.
From multiplex to multi-omics
For the past decade or so, researchers have been combining datasets from various types of analytes—transcriptomics with genomics with proteomics, for example, with a sprinkling of metabolomics and microbiomics thrown in for good measure—in what is called a multi-omics approach. “The meta-analysis lets you put all this data on top of each other, and out of that starts to come biomarkers. It’s a refined list of things that are changing in the tumor environment,” says Couvillon.
Image: Shutterstock Images