It seems as if every medical journal, and even President Obama's "State of the Union" address earlier this year, mentions "precision" medicine, "targeted" therapies, or makes some other reference to using genomic tests in diagnosing or treating cancer patients.
This innovative approach takes into account an individual patient's molecular information to understand the biology of his or her disease. Genomic tests, or panels, can be used to customize treatment and predict prognosis and therapeutic response.1 Because cancer is a genetic disease, often caused by a combination of mutations working together, the potential impact of precision medicine in oncology can be huge, and therapeutics have already been developed against a variety of cancer-related mutations.2
But what do we mean by "genomic panels," and what can hematologist/oncologists do with the information they get back from testing labs? ASH Clinical News spoke with researchers specializing in genomics and genetics about techniques for gathering genomic data, the differences among sequencing platforms, and their applications in clinical practice.
Cancer Genetics 101
There are two basic types of genetic mutations that can cause cancer: somatic and germline.
- Somatic, or acquired, mutations are changes in DNA that occur after conception, only affect cells that grow from the mutated cell, and are not passed on to the next generation. Somatic mutations cause the majority of cancers. This can be spontaneous, arising randomly, or caused by environmental factors, such as extreme radiation exposure.
- Germline, or hereditary, mutations are alterations to DNA within all cells of the body, including germ cells (sperm and egg). People are born with these mutations, and they can be passed on to the next generation. Inherited cancers caused by germline mutations make up approximately 5 to 10 percent of malignancies.3,4
The Role of Genomic Panels in Precision Oncology
The development of high-throughput, massively parallel DNA sequencing technologies, or next-generation sequencing (NGS), has dramatically reduced the cost of sequencing over the last decade. These methods are rapidly providing unprecedented insight into the pathogenesis of hematologic malignancies by uncovering known and newly discovered genetic variants that have clear clinical relevance.
Further integration of NGS techniques into routine clinical care should have a major impact on diagnostic accuracy, risk assessment, and tailored clinical care – and may yield novel therapeutic approaches.5,6 Genome sequencing has also identified millions of single nucleotide polymorphisms (SNPs), the most common type of human genetic variation.7,8
"The difference among these platforms is the extent of genomic DNA being evaluated, or ‘genotyped,'" Susan Slager, PhD, associate professor of biostatistics at the Mayo Clinic College of Medicine in Rochester, Minnesota, told ASH Clinical News.
Whole-Genome Sequencing
Whole-genome sequencing (WGS) involves creation of an in vitro library from a patient sample by fragmentation of genomic DNA. The fragments are then amplified and sequenced using one of many NGS platforms and chemistries available. After alignment with a reference genome, the sequence is analyzed for variants.7
"Whole-genome sequencing describes techniques used to reveal the identity of every DNA base in an entire genome," said Rafael Bejar, MD, PhD, an assistant professor of hematology and oncology at UCSD Moores Cancer Center in San Diego. "In humans, that's over 3 billion bases!"
The purpose of WGS is to analyze all non-redundant sequences in the genome, explained John Welch, MD, PhD, assistant professor of medicine in stem cell biology at Washington University in St. Louis, Missouri. "It allows for identification of point mutations (single nucleotide variants, or SNVs), small insertions and deletions (indels), translocations, and copy number changes."
"WGS is useful for identifying unexpected structural changes," to a person's DNA, he added.
Whole-Exome Sequencing
As opposed to whole-genome sequencing, which identifies every bit of DNA, whole-exome sequencing is more efficient, focusing mainly on the parts of DNA that actually encode proteins. "Most of the human genome is made up of sequences that do not encode proteins," Dr. Bejar said. "These are mostly repetitive structural elements or regulatory regions. Only 1 percent of the DNA sequence in our genomes represents protein-coding regions. The protein-coding elements of genes are called exons. Exons are the areas examined by whole-exome sequencing, or WES," Dr. Bejar said.
WES is an NGS technique that uses an initial exome enrichment step.6 "WES analyzes only the regions of the genome that are captured efficiently by the exome capture reagent," Dr. Welch explained. "Although these reagents focus on capturing exons, they also analyze a leader sequence (5' UTR) and trailer sequence (3' UTR) [on either side of the relevant exon]. This can be useful for analyzing mutations that could alter expression patterns." In other words, WES also can identify mutations in DNA portions that immediately precede or follow protein-encoding exons, and thus could affect whether that exon is translated to a protein.
Exome data are useful in the identification of somatic mutations that may drive cancer. "WES is a useful platform when looking for known mutations that could be targeted with specific drugs," Dr. Welch said. He added that WES can detect indels but does not identify translocation events well. "Copy number analysis can be done but requires some adaption for the coverage gaps."
For both WGS and WES, a non-tumor, germline control sample from the patient should be simultaneously analyzed with each tumor sample to identify the tumor-specific somatic mutations. A skin biopsy can provide the germline control DNA.5,6 This enables comparisons between the tumor sample and control sample, to identify which genetic mutations are acquired and may have caused the cancer.
"There are a few genes that are recurrently mutated in hematologic malignancies with mutational hotspots (e.g., IDH1/2, KIT, FLT3, NPM1, etc.). These mutations can be identified without matched germline," Dr. Welch pointed out. "However, mutations in TET2, DNMT3A, TP53, and the cohesins occur at diverse positions across the gene. It is difficult, and nearly impossible, to determine whether a variant in one of these genes is a rare inherited SNP or a somatic mutation."
Single Nucleotide Polymorphism Genotyping
"Single nucleotide polymorphism, or SNP, genotyping looks at an even more limited number of spots in the genome," explained Dr. Bejar. "SNP genotyping techniques aim to determine which DNA base is present at specific genomic positions on the genomic map." These are then compared with a "normal" population of patients who don't have a known cancer, to determine if the identified base may play a role in cancer formation.
Most SNPs have no known effect on gene function or expression, but some are associated with specific traits or diseases and can be used as markers for the actual causative variants.7,8 "These techniques can also be used to estimate how many copies of a gene are present in each cell. Cancer cells often have loss or duplication of genomic regions, which can be detected using this approach," he added.
A variety of SNP genotyping methods have been developed, but genome-wide studies are performed using microarrays that detect the degree of similarity between target DNA and complementary probes immobilized on a small chip. NGS is also becoming a viable option for SNP genotyping.8
"Many of the SNP genotyping techniques can examine more than 1 million bases at a time," Dr. Bejar noted. "Still, this represents less than 1 percent of the territory covered by WES and less than 0.01 percent of that covered by WGS."
SNP genotyping does not identify genetic insertions, deletions, or translocations, and is generally not designed to detect activating mutations, except when the location of the mutation in the array is already known. In such cases, it can be used to quickly determine if specific oncogene mutations are present.
"This approach is less useful to detect mutations that inactivate proteins, like those encoded by tumor suppressor genes, since disruptive mutations can occur almost anywhere along the length of a gene," Dr. Bejar said.
Which Genomic Panel Should You Choose?
The goal of genomic studies is to improve diagnostic precision and patient outcomes, but establishing the best circumstances for their use remains a challenge.
"As always, the clinical picture is the most important factor in determining which test is appropriate," Dr. Welch said. "The physician must understand what question is being asked and then apply the best test to answer it." There are no cut-and-dried rules for when to apply a specific test, but the experts interviewed for this article provided some guidance about which test to use when.
WGS: Too Much Information to Handle?
WGS is most appropriate when structural variants or non-coding mutations are suspected, or when there is uncertainty about which chromosome position to evaluate. But it may be an information overload.
"Ideally, we would use WGS all of the time, as it is intended to capture all of the genomic DNA information," Dr. Slager noted. "However, the current infrastructure is not able to handle such a large number of samples with a quick enough turnaround for clinical practice (e.g., within 1-2 days for results)."
Dr. Slager also pointed out that, even when the infrastructure for WGS is in place, it might not be warranted for analysis of a specific genomic region known to be cancer-related, such as a gene or pathway. "Simpler genotyping panels would get the information easily and quickly without the overhead that comes with managing all the data generated from WGS technologies."
Dr. Bejar agreed that a more targeted approach is useful for quickly detecting a specific mutation, such as the JAK2 V617F mutation in a patient with polycythemia vera. "In that scenario, we really only want to know about one particular mutation and don't need to spend the time or money on a broader panel."
A laboratory must have specialized equipment to perform NGS, and physicians also must be trained to interpret the data and develop appropriate therapeutic strategies for a wide range of molecular subtypes.6,7 For this reason, "WGS is still primarily in the realm of research since we don't know how to interpret the overwhelming majority of what we find," Dr. Bejar said.
WES: An Unbiased Approach
"Even if we were to perform WGS on a patient or that patient's tumor, we would likely restrict our analysis to mutations whose implications are best understood," he added. "These are almost entirely in protein coding regions, which are much better covered by WES."
WES is an unbiased way to examine every protein coding region, which allows hematologists to locate mutations and polymorphisms that they were not initially looking for.
SNP Genotyping: A Happy Medium?
"At the moment, WGS is not well suited for routine clinical use. It's more expensive and often less sensitive than other approaches. This may change as sequencing becomes cheaper and our ability to interpret what we find grows," Dr. Bejar said. "Even WES is overkill in most clinical situations."
Cost also remains a significant factor when deciding which genomic test to use. WGS is the most expensive approach and not yet cheap enough for routine clinical use, with SNP genotyping as the most affordable option. He suggested that for most diseases, including cancers, targeted sequencing of specific genes is the sensible choice for obtaining high coverage at relatively low cost. "SNP genotyping arrays that look at SNPs across the genome remain useful in some clinical scenarios but are rapidly being eclipsed by the falling cost of WES and other targeted sequencing techniques."
The Present and the Future of Genomics
One challenge with the routine use of genomic profiling today is that the cost of computing is not falling as quickly as the cost of sequence production, so computing power is not sufficient to deal with all the raw data being generated by NGS.5,7 "We are entering a phase when the analysis costs more than the sequencing production," Dr. Welch pointed out.
"Most genomic panels in clinical use focus on the exons of a small number of target genes. In cancer, these are typically oncogenes and tumor suppressor genes that are recurrently mutated – a subset of which will have important clinical implications," Dr. Bejar said. "As the cost of targeted sequencing has fallen, this approach has become increasingly favored in the clinical setting," Dr. Bejar said.
Recently, at the 2014 ASH Annual Meeting, Janine Pichardo, BS, from the department of Pathology at Memorial Sloan Kettering Cancer Center in New York, and researchers demonstrated the clinical utility of this approach. Using a commercially available NGS-based gene panel targeting hundreds of cancer-related genes, investigators were able to diagnose a wide spectrum of hematologic malignancies known to be difficult to diagnose, subclassify, and risk-stratify with conventional methods.9
"Our study shows that a broad sequencing panel targeting single nucleotide variations, insertions, deletions, copy number alterations, and translocations may improve diagnostic accuracy in 10 to 15 percent of patients with hematologic malignancies," Ms. Pichardo said. "In our opinion, physicians treating hematologic malignancies should make every effort to integrate comprehensive targeted genomic profiling to the care of patients with hematologic malignancies."—By Amy Dear
References
1. Ciardiello F, Arnold D, Casali PG, et al. Delivering precision medicine in oncology today and in future-the promise and challenges of personalised cancer medicine: a position paper by the European Society for Medical Oncology (ESMO). Ann Oncol. 2014;25:1673-78.
2. Garraway LA, Verweij J, Ballman KV. Precision oncology: an overview. J Clin Oncol. 2013;31:1803-05.
3. Genetics Home Reference. Handbook: Help Me Understand Genetics. Lister Hill National Center for Biomedical Communications; U.S. National Library of Medicine;
National Institutes of Health; Department of Health & Human Services; 2015. Accessed April 23, 2015 from http://ghr.nlm.nih.gov/handbook.pdf.
4. American Cancer Society. Family Cancer Syndromes. Accessed April 23, 2015 from www.cancer.org/cancer/cancercauses/geneticsandcancer/heredity-and-cancer. Revised June 25, 2014.
5. Merker JD, Valouev A, Gotlib J. Next-generation sequencing in hematologic malignancies: what will be the dividends? Ther Adv Hematol. 2012;3:333-39.
6. Braggio E, Egan JB, Fonseca R, Stewart AK. Lessons from next-generation sequencing analysis in hematological malignancies. Blood Cancer J. 2013;3:e127.
7. Johnsen JM, Nickerson DA, Reiner AP. Massively parallel sequencing: the new frontier of hematologic genomics. Blood. 2013;122:3268-75.
8. Edenberg HJ, Liu Y. Laboratory methods for high-throughput genotyping. Cold Spring Harb Protoc. 2009; (11):pdb.top62.
9. Pichardo JD, Feldstein JT, Arcila M, et al. A comprehensive clinical next generation sequencing-based assay can impact hematopathologic diagnosis in a significant subset of patients with hematologic malignancies. Abstract #2984. Presented at the ASH Annual Meeting, December 7, 2014.
