NEW YORK – It has become apparent that combining DNA and RNA sequencing is beneficial for researchers seeking to understand more about the biology of pediatric cancers and for clinicians looking to pair patients with targeted treatments. But the necessary breadth of sequencing is something researchers haven't yet agreed on: While some say that whole-genome and whole-transcriptome sequencing are necessary, others argue that whole-exome sequencing or targeted RNA panels may suffice, at least in certain circumstances.
At the Institute for Genomic Medicine at Nationwide Children's Hospital, Co-Executive Director Elaine Mardis and her colleagues use a protocol that combines DNA and RNA sequencing in the clinical cancer care setting in order to help patients. The NCH cancer genomic profiling protocol includes tumor and germline whole-exome sequencing, with enhanced coverage of cancer-associated genes and copy-number probes across chromosomes at 250-fold coverage; RNA sequencing; targeted RNA-seq assays, including ArcherDx assays to look for driver fusions; and liquid biopsy testing to monitor disease changes.
"In terms of the breadth of sequencing that we should be using, I'm fairly comfortable with the enhanced exome [approach], because copy number is a consideration, including copy-number probes so that we have the ability to look at chromosomal amplification and deletions," Mardis said. "There are a lot of indications already from clinical trials in kids about chromosome-arm or whole-chromosome events that you'd want to be able to pick up because they're often prognostic — they tell us whether to treat the cancer more or less aggressively."
She further noted that RNA-seq is important not only for detecting fusions, but also for looking at immune infiltration and analyzing whether there are indications of enhanced PD-1 or PD-L1 expression in the cells. "I think that's only going to become a richer set of data analysis as we go forward," she added.
At Memorial Sloan Kettering Cancer Center, pediatrics department chair Andrew Kung uses the Memorial Sloan Kettering-Integrated Mutation Profiling of Actionable Cancer Targets (MSK-IMPACT) assay — a hybridization capture-based NGS assay for targeted deep sequencing of all exons and selected introns of nearly 500 cancer genes — to sequence pediatric cancer patients that come to the center for treatment. But unlike Mardis, Kung wants to go further and sequence the whole genomes of pediatric cancer patients.
"There are classes of [genomic] changes and [oncogenic] drivers that are only visible through whole-genome sequencing," he said. "Moreover, to whatever extent we haven't yet reached the point of comprehensively cataloguing all the potential drivers and seeing the structural changes that can drive pediatric cancers, I think that the opportunity to be completely comprehensive in approach is one that, as a field, we're compelled to move forward with."
For Jinghui Zhang, chair of the department of computational biology at St. Jude Children's Research Hospital, whole-genome sequencing is an integral part of pediatric cancer research. In a September 2018 paper in Nature Communications, Zhang and her colleagues evaluated the efficacy of combining whole-genome, whole-exome, and transcriptome sequencing on tumors and normal tissue from 78 pediatric cancer patients. They found that the three-platform sequencing approach had a positive predictive value of between 97 percent and 99 percent for somatic SNVs, 99 percent for indels, and 91 percent for structural variations. They also reported 240 pathogenic variants across all cases with 98 percent sensitivity, while combined WES and RNA-seq testing achieved only 78 percent sensitivity. These results, the authors noted, emphasized the need for incorporating WGS in pediatric oncology testing.
According to Zhang, this has been a driving principle behind the way St. Jude approaches pediatric cancer research since 2010, when the hospital began the Pediatric Cancer Genome Project in collaboration with Washington University School of Medicine in St. Louis, in an effort to identify the genetic changes underlying some of the deadliest childhood cancers. The teams sequenced more than 600 childhood cancer patients as part of that initiative.
"If you don't perform whole-genome sequencing, you're not getting [the whole picture], because structural variation breakpoints occur in intronic regions," Zhang said, noting that WGS has been useful for performing more complete mutation signature analysis beyond those present in the exome, and including intronic variations and SNV indels.
Zhang is also not convinced by the argument that WGS should only be applied to patients who present with rare cancers or cancers that are understudied. "The problem with that argument is that you need to know which patients are 'standard' patients," she said, explaining that WGS of all patients helps define the commonly occurring characteristics so that uncommon or unexpected features can be recognized when they show up in patients.
For Tim Triche, co-director for the Center of Personalized Medicine at Children's Hospital Los Angeles (CHLA), the context of the testing is important when trying to determine what modality would be most useful. His center routinely uses WES, WGS, and RNA-seq — employing both whole-transcriptome sequencing and poly(A) RNA-seq testing for coding genes — in order to study pediatric cancers. But from a clinical perspective, the most useful results for Triche have been from a targeted panel that he and his team at CHLA have developed in partnership with Thermo Fisher Scientific called the Oncomine Childhood Cancer Research Assay, or OncoKids.
He finds other, more comprehensive testing methods to be important research tools for identifying new cancer defects that can be clinically useful. "Once having found them, however, it becomes very useful to focus attention, particularly clinically, on the features that are recurring and clinically relevant to the disease," he noted, adding that currently in cancer, clinical utility is concentrated within a smaller subset of genomic markers.
In Triche's view, the number of variants of unknown significance dredged up through WGS, and even WES, can quickly bury the clinically useful findings under vast amounts of possibly useless data. "A more useful, practical alternative is to cluster in a panel the features that are generally agreed in the childhood cancer community to be relevant to the diagnosis, … prognosis, and even potential [identification of] therapy for childhood cancer," he said.
Moreover, many of the features that are important in childhood cancer are seen at the RNA level rather than the DNA level. So, while WGS is an incredibly important tool for research, it must be combined with total RNA-seq to cover the landscape of clinically relevant biomarkers within a given tumor. This is more expensive, Triche observed, and adds to the challenge of analyzing what could be a large amount of irrelevant data.
"Typically, the tools that are used to map the reads to the human genome don't really see things like gene fusion,," Triche said. "Well, one of the most important features in childhood cancer is gene fusions, where pieces of two genes are fused together. Those fusion genes are often lost, unless your mapping algorithm is particularly sensitive to detecting these sorts of things and the type of sequencing you've done picks them up reliably."
In an effort to create a tool with more clinical utility, his team at CHLA focused on advancing a targeted panel. The OncoKids assay was developed as an offshoot of Thermo Fisher's original Oncomine Cancer Research panel for adult cancers, which is being used in the National Cancer Institute's Molecular Analysis for Therapy Choice (MATCH) trial.
Seeing the obvious virtues of Oncomine's incorporation of the most common, clinically relevant features of adult cancers — especially those for which there might be a therapeutic match — Triche approached Thermo Fisher with the idea to develop a pediatric version of the assay.
"The intent from the outset was to not restrict it to any one type of cancer. It was developed specifically to deal with the entire spectrum of childhood cancer," and drew on many of the genes already interrogated by the Oncomine assay since variations in some of these genes sometimes do show up in pediatric cancers, Triche said.
Through an iterative process involving pediatric oncologists, surgeons, and a variety of pathologists from CHLA, as well as scientists from Thermo Fisher, the developers also ensured that the test would be able to gauge molecular features that may show up more readily in childhood cancers. For example, the test can assess the full coding regions of 44 cancer predisposition loci, tumor suppressor genes, and oncogenes; hotspots for mutations in 82 genes; and amplification events in 24 genes; and 1,421 gene fusions that have been shown to be clinically relevant in a variety of childhood cancers.
"The number one design goal was that they had to be features that occurred frequently enough [in pediatric cancer] that one could reliably use the panel to make decisions clinically about the diagnosis for the patient and the many subtypes, [as well as] the prognosis," he noted."Many of the subtypes carry prognostic information, which in the world of childhood cancer therefore dictates treatment."
Since the original panel was developed, new features of importance have been identified and incorporated into the panel, he added. It now includes 203 unique genes and thousands of fusion drivers, according to Thermo Fisher. Triche anticipated that the panel will continue to be enhanced as new features of importance are identified through techniques like WES, WGS, and total RNA-seq.
Still, approximately 20 percent of CHLA patients don't receive any useful information from the targeted OncoKids panel, and those are the patients for whom WGS and whole RNA-seq are more likely to be informative, in Triche's view.
St. Jude's Zhang noted that while some oncologists may prefer a more tiered approach of trying out enriched exome- or capture-based assays first before proceeding to WGS, this requires a longer timeline, which cancer patients may not have. It may also require more sample material, which may not be an issue with blood cancers, but is a major hurdle in solid tumors. "If you have utilized the sample for this [exome] assay, do you still have enough left for other tests?" Zhang posited. "Each institution really has to adjust for their own needs."
In addition to sample quantity, sample quality poses another diagnostic challenge. Standard practice after a biopsy is to fix a tumor sample in formalin and embed it in paraffin, which can limit the number of assays one can perform on the sample. Triche pointed out that performing WGS on FFPE tissues can sometimes result in skewed data and false positives. By the same token, total RNA-seq is more difficult with FFPE tissues because the RNA is fragmented. And extracting reliable information about genome-wide methylation patterns — which can be very useful in pediatric cancers — can also be negatively affected by the presence of formalin.
This is another mark in the favor of panels such as OncoKids, in Triche's view, because they have been developed with suboptimal sample sources in mind. The test can be performed on small amounts of input sample and is compatible with a variety of preservation methods. In November 2018, Triche and his colleagues published a study in the Journal of Molecular Diagnostics describing the panel, in which they noted that it used low input amounts of DNA and RNA — only 20 nanograms of each — and was compatible with FFPE and frozen tissue, bone marrow, and peripheral blood.
Ultimately, when deciding what NGS approach to use, the most important thing to keep in mind, according to Triche, is that childhood cancer is fundamentally different than adult cancers. "Adult cancer is a disease of aging and accumulated mutations over a lifetime. The older you get, the more mutations you have, and sooner or later, one or more of those mutations is going to give a selective growth advantage to one of the tissues in your body. And voilà, one day we call it cancer. But it's a process that plays out over many years. That's an impossible scenario and explanation for childhood cancer," he said.
In contrast, cancer in a child should be viewed as development gone awry. "Wilms tumor is trying to make a kidney. It doesn't do a very good job, but it actually is very much like developing kidney. Neuroblastoma is like your developing adrenal gland. Medulloblastoma is like your primitive brain trying to make brain," he said. "Almost all of these tumors are malignant analogues of a normal tissue. They're developmental."
This fundamental aspect of pediatric cancers may be useful in guiding oncologists as to when to move beyond targeted panels and use more comprehensive genomic profiling tools, such as WGS, RNA-seq, and epigenomic profiling, which can pick up the developmental features that are dysregulated. In contrast, using WES to identify genetic mutations that are not important in development may not be significantly useful in childhood cancers. "To my way of thinking, whole-genome is part of a three-legged stool where you look at RNA, DNA, and epigenetic modifications," Triche said.