Liquid Biopsies Coming of Age
Industrialization, validation at the forefront
The decision matrix for liquid biopsies is defined by the sample, method, the analysis platform, and the desired output. For typical platforms -- molecular assays like PCR or NGS -- sample acquisition and preparation is critical, as analytes often exist at extremely low concentrations within the sample, with seemingly overwhelming quantities of interferences.
In his thorough review of ctDNA in melanoma, Benoit Busser at Grenoble University Hospital (France) notes the diversity of cancer-related genes present in the plasma of melanoma patients. For these molecular targets and quantitation methods, which have been validated by comparison to DNA from solid tumor biopsies, plasma concentrations vary significantly by biomarker and stage of illness. Also significant are the implications for treatment efficacy and disease regression: “That biological recurrence precedes radiological signs of recurrence by several weeks is well established.”
Cancer patients typically have high levels of wild-type (normal) cell-free DNA, which interferes with conventional sequencing methods based on PCR, for example Sanger sequencing or pyrosequencing. These methods quantify ctDNA only when very high levels are present, which unfortunately only occurs in late-stage disease, or when isolation (or amplification) methods are sensitive and specific for mutant DNA.
Automation will be essential as liquid biopsies migrate from research labs to clinics as lab-developed tests or approved diagnostics. The obvious benefits are higher throughput, greater consistency, and elimination of human error, all of which have long been characteristics of industrialized laboratory diagnostics. The same goes for liquid biopsies, particularly at the sample collection and preparation step.
Biodesix (Boulder, Colorado, USA) has recently demonstrated a novel way to extract circulating RNA from plasma. To accomplish this, Biodesix collaborated with Norgen Biotek (Thorold, Ontario, Canada) which specializes in sample collection and preservation.
As molecular testing gains steam, automation will be increasingly important for maintaining consistency and achieving high-quality assay results. Biodesix had sought out Norgen to automate the Biodesix GeneStrat test platform without sacrificing performance.
GeneStrat is a liquid biopsy-type genomic assay that guides the use of targeted therapies in patients with lung cancer. The test quantifies genetic mutations and variations in circulating tumor DNA and RNA using highly sensitive droplet digital PCR developed by Bio-Rad. Treating physicians order the GeneStrat test for their lung cancer patients. The test delivers results for EGFR sensitizing, EGFR resistance, KRAS, and BRAF mutations, and EML4-ALK, ROS1 and RET fusion variants.
The specific improvement to GeneStrat involves automating a commercially available laboratory-developed test (LDT) for an mRNA fusion variant of EML4-ALK, a gene implicated in lung cancer. LDT is a specific regulatory designation for an in vitro diagnostic that is developed and used in one laboratory. The upgrade mechanizes the process for plasma RNA isolation and RNA concentration.
“We worked with Norgen Biotek to modify their Plasma/Serum Circulating and Exosomal RNA purification kit and RNA Clean-up and Concentration Micro-Elute kit,” says Gary Pestano, Ph.D., VP of Development and Operations. “The process was automated on the Hamilton microLAB Nimbus platform.”
To show how far liquid biopsies have come, Pestano describes GeneStrat as an alternative -- not to conventional biopsy, but to replace conventional tumor tissue genomic testing, or as a test option before tumor tissue becomes available.
GeneStrat is often used with another Biodesix assay, VeriStrat®, a blood-based proteomic test for patients with non-small cell lung cancer who test negative for EGFR mutations. “This strategy integrates genomic and proteomic biomarkers to inform tumor mutation status and measure a patient’s immune response to their tumor and the aggressiveness of their cancer,” Pestano adds.
Gene-amplification methods, most PCR-based, serve as front-end to NGS or as the main attraction in many liquid biopsies. The number of clinically relevant offshoots of PCR is growing, to the point where even the offshoots have offshoots.
Originally developed at the Dana-Farber Cancer Institute, COLD PCR was introduced in a Nature Medicine paper in 2008 and subsequently licensed by Transgenomic (New Haven, Connecticut, USA). Several groups, including Transgenomic, subsequently improved the method. ICE COLD PCR, one such variation, has several advantages over conventional PCR, specifically allele non-specificity, broad sample acceptance, suitability to all significant detection platforms, and up to a 500-fold improvement in sensitivity.
Exploiting these strengths, Precipio (New Haven, Connecticut, USA), which merged with Transgenomic in 2017, is commercializing ICE COLD PCR (ICP) for liquid biopsies. ICP preferentially enriches mutant DNA sequences over wild-type DNA through selective amplification, resulting in up to a 500-fold increase in sensitivity for identifying mutations associated with cancer, at sequence alteration detection as low as 0.1%. This sensitivity level is generally acknowledged as a standard for liquid biopsies for genetic testing, and can be applied for patient monitoring of treatment response, disease progression, and reoccurrence.
The significance of liquid biopsies, in terms of patient health and quality of care, requires examining how liquid and conventional biopsies are currently utilized.
To provide information on a patient’s status, surgeons preserve a portion of the biopsy taken from solid tumors for histologic and genomic examination. What they get is a day-one picture of the genetic footprint of that specific tumor. “As patients undergo therapy the tumor adapts, and the biology of the tumor may change,” notes Ilan Danieli, CEO of Precipio. “Because of those changes, the treatment may no longer work. Tumors stop shrinking and start growing. Re-biopsy is almost never done, so without any new genetic information, from that point onward physicians often operate in the dark.”
In part this is due to a holdover mentality from pre-genomic days, when the purpose of the biopsy was to classify a tumor’s origin and stage. Concerns about patient safety and subjecting them to a second (or third) invasive and dangerous procedure, in addition to cost, are in many cases, the main reasons why the initial biopsy remains the only source of that initial genetic information. Yet clinicians leave the full power of genomic analysis on the table when they eschew periodic genomic tests.
“The most significant value of ICE COLD PCR is that it eliminates the cost barrier to monitoring the patient’s current status by repeat-testing,” Danieli explains. “You don’t want to learn that chemo isn’t working the hard way, after several months of treatment, when a CT scan shows the tumor is growing, rather than shrinking, indicating the treatment stopped working. Liquid biopsies based on ICE COLD PCR allow you to test as often as you need, to catch that change early.”
According to Danieli the ICE COLD test costs as little as $150 to run, compared with the several thousand dollar price tag for competing liquid biopsies tests currently offered.
The underlying science of circulating tumor DNA (ctDNA) assays, or liquid biopsies, has outpaced the introduction of validated methods built on these technologies. A lack of adequate standards has been the most serious challenge.
SeraCare Life Sciences (Milford, Massachusetts, USA) has assembled a suite of clinical genomics validation and quality control tools, the most recent being SeraseqTM Myeloid DNA and Myeloid RNA fusion reference materials. Developers of liquid biopsies use these reference standards to ensure the sensitivity and robustness of next-generation sequencing (NGS) myeloid cancer assays.
Molecular analysis of myeloid cancers is shifting from conventional singleplex methods such as FISH and qPCR, to multiplexed NGS assays. Singleplex assays target one unique genetic sequence. As myeloid cancers may involve a diversity of mutations, singleplex assays are costly and time-consuming.
The Myeloid Mutation and Myeloid Fusion mix products cover a wide range of mutation types, from clinically important single-nucleotide variants to gene fusions.
The DNA mutation kit consists of a purified DNA mixture of actionable somatic mutations and a genomic background of highly characterized healthy individuals. “Actionable” mutations are genomic discrepancies physicians look for to detect specific tumor types. The gene mix is representative of mutated and wild-type genes found in typical patient samples, around 95% wild-type, 5% mutant. “We produce them to represent what a genotyping assay is likely to see in patients,” says Russell Garlick, Ph.D., Chief Scientific Officer.
Instead of specific DNA mutations, the RNA kit incorporates a mix of actionable RNA fusions. Users convert the same protocols they employ for patient samples, into complementary DNA, which they analyze through gene amplification, sequencing, etc.
By duplicating the range of meaningful genetic variants and relevant frequencies for both RNA and DNA, the products will help developers of liquid biopsies and other products to validate the analytical performance of their assays. These kits are, in a sense, manufactured preparations of typical patient samples for the genetic variants and fusions that users would look for in a liquid biopsy. They serve as standards -- and an independent means of validation -- for the sample preparation, analysis, and data systems end-users employ.
The kits address a standards gap for liquid biopsies with carefully designed preparations that are designed in silico for the analysis platform of interest, then synthesized and formulated precisely. Kits are customizable as well to specific targets.
“Let’s say you want to validate an assay for the rare FLT3 gene ITD, mutations in which are implicated in hematologic cancers. The gene, with its internal tandem duplications within exons, is difficult to detect but we can design the desired variant and put it into a mix,” Garlick says. “Users will be confident that if they can detect FLT3 ITD in our sample they can detect it in patients.”
Adopting liquid biopsies will take time as scientists, physicians, and reimbursement organizations reach an equilibrium point defined by the methods’ inherent potential and limitations, and cost. With instrumentation and informatics improving on a weekly basis, liquid biopsies nevertheless face one technical hurdle, which according to SeraCare’s Dr. Garlick is more biological than technical.
“Liquid biopsies are already sensitive and robust.” Detection limits of 0.5% for mutated genes could fall even further. “The challenge is whether tumors are shedding genetic material, and to a detectable degree, into the bloodstream. Are micrometastatic tumors, in particular, shedding enough material to be detectable?”
Improving sensitivity is only part of the answer, “a double-edged sword,” according to Garlick, as below about 0.1% frequency, biological and assay noise come to predominate the signals. “As the assays improve we’ll see false positives go down and true positives go up, but don’t expect a step change improvement in technical performance. Progress will be incremental."