Lab Manager | Run Your Lab Like a Business

INSIGHTS

3d illustration DNA molecules
One of the most successful applications of genomics to patient care is its use in cancer.
iStock, Design Cells

Genomics Brings Precision to Medicine

Genetic sequencing identifies overlooked approaches to therapy and treatment

Sherri Fraser, PhD

The exploding global market for genomics reflects the optimism industry experts and scientists share that new genomics approaches will change medicine forever. Recent uses of genomics to identify and treat rare diseases and cancer underscore this potential.

Genomics of cancer

The projected global market for genomics from 2021 to 2026
The global market for genomics is expected to exceed $30 billion US in 2022, with the expectation that it will continue to grow by at least 18 percent per year until 2026.
Lab Manager, Sources for statistics: Yahoo!Finance and Research and Markets

One of the most successful applications of genomics to patient care is its use in cancer1, 2.  Traditional chemotherapeutics with well-known outcomes have been identified that work well with specific genetic profiles, but typically focus on one biomarker for one drug. As Jordi Rodon, MD, associate professor in the Department of Investigational Cancer Therapeutics at UTMD Anderson Cancer Center, says, “What has changed with next-generation therapy is that now with one test, you can check multiple biomarkers, and that could guide you toward different therapies.”  


Genomics provides clinicians with information to use targeted therapies. As an example, targeted inhibitors for non-small-cell lung cancers are now available to patients whose cancers have aberrations in genes such as EGFR, ALK, or ROS1. Many such treatments use monoclonal antibodies to specifically target aberrant or fusion proteins identified in these cancers. These treatments are specific and targeted, which reduces the toxicity to the patient, dramatically improving their quality of life during treatment.

The development of resistance to therapeutics has always been a concern in the treatment of cancer patients. Genomics has also helped identify mutations leading to resistance. By understanding the genomic profile of the cancer and the patient, it may be possible to design therapeutic cocktails that forestall cancer resistance, again with the added benefit that cocktail treatments hit multiple pathways at lower doses, reducing therapy side effects.

Genomics of rare diseases

An infographic showing various next-generation sequencing choices and comparing their source, cost, and speed
An infographic comparing the source, cost, and speed of various NGS options.
Lab Manager

Genomics is also making a difference in the identification and management of rare diseases. The NIH defines a rare disease as affecting fewer than 200,000. Although rare, the burden of these diseases to patients, their families, and the health care system is substantial3. The cost of rare diseases is disproportionate compared to common conditions. Patients are regularly admitted for longer stays, with higher costs. They are often readmitted, and they face higher mortality. Most of these diseases have a genetic origin.

A timely and accurate diagnosis is one of the most important health care outcomes for any patient. “The only reason you ever go to a doctor is to figure out what's going on with you,” says Kym Boycott, MD, chief, Department of Genetics, Children’s Hospital of Eastern Ontario. Without a diagnosis, the prognosis and path-of-care are unknowns. In the case of epilepsy, Boycott explains, “Some of the genetic epilepsies are responsive to certain medications and not other medications. And in fact, some medications can actually make them worse.” Once a clinician knows what genetic pathways are impacted, they also gain insight into the long-term biological consequences. This improves the overall care for the patient by allowing a doctor to monitor for outcomes known to be impacted by the genes altered in the patient. 

Challenges and prospects for genomics

Diversity

The first reference genomes represented only white Europeans. Thankfully, this gap is being filled with dozens of new genome projects to improve the diversity of genomic databases4. In 2018, the National Human Genome Research Institute had listed at least 25 initiatives to enhance diversity in human genomics. 

Many variants in our DNA are harmless, but red flags appear when an unknown variant shows up during sequencing. As we expand the reference data to include diverse populations, many of the previously unidentified variants in one racial group will be found as “normal” in another. This helps provide a true understanding of the extent of normal variation in the human genome, which, in turn, helps reduce the number of rabbit holes bioinformaticians must explore when looking at someone’s genome.

Data collection, sharing, and presentation

The collection and sharing of data are some of the most important aspects of genomics but are also among the biggest challenges. Anytime a patient’s data is to be shared, consideration must be given to privacy and consent. How informed consent is collected poses great challenges to ensure consistency in global context. Storage and presentation of these data is also a challenge, considering different languages or variability of health care funding and administration in different countries. After 10 years of collecting cancer genomics data, the International Cancer Genome Consortium5 launched its next phase, ARGO (Accelerating Research in Genomics Oncology) to address these issues. Similar initiatives are occurring across the genomics spectrum.

Genomics tests can’t simply replace a standard-of-care

Medical care relies on standards-of-care. Genomics is rarely an option until after standard treatments have failed. Even though genomic sequencing may provide insights, it takes time before the use of these techniques is proven to provide outcomes as effective as known treatment options. Until that happens, patients are owed the best-known options first.  However, genomic testing is bound to improve the diagnostic journey for patients. A single comprehensive assessment will ultimately reduce the number of tests and the time to diagnosis. As studies show the positive impact of genomics, and by educating physicians about genomic tests, genomics will become part of a diagnostic toolkit.

Transcriptomics

Key limitations of DNA sequencing include the ability to identify gene fusions accurately or gene expression changes. Boycott and Rodon identify RNA sequencing as the future for genomics. The study of the transcripts synthesized from DNA is called transcriptomics. Cancer DNA sequencing, for example, suffers because tumor samples are always contaminated with non-malignant cells, which can overwhelm sequencing efforts. There are also recent promising results using transcriptome sequences obtained from single cells. The implications of this for studying tissue- and cell-specific differences are far-reaching. 

The other omics

Genomics isn’t just about DNA sequencing. It integrates the expression and interactions of genes. Other “omics,” in addition to transcriptomics, extend naturally from genomics. Proteomics studies the proteins present in cells and tissues. Epigenomics looks at the modifications to the DNA, which can be cell-specific and change DNA expression dramatically. Metabolomics studies show us the metabolites in tissue and cell samples. The expansion of genomics to these other omics will dramatically expand our understanding and treatment options for disease.

Machine learning (ML)

Considering how much data can be generated simply from DNA sequencing of a single person, the volume of data from the combined omics studies is profound. Currently, variants discovered in a genome are individually looked at by a human to determine their value for investigation. ML models promise to streamline that process and, hopefully, find results not noticed by the human researchers. 

Significant challenges still exist in the ML community, such as managing the differences in data sets used to train the ML model and the dataset in the real-world application. However, many studies are underway to improve the ability of ML models to accurately predict disease outcomes6.

References:

1.    Bossé, Y. & Amos, C. Cancer Epidemiol. Biomarkers Prev. (2018) 27:363

2.    Berger, M. & Mardis, E.  Nature Rev. Clin. Oncol. (2018) 15:353

3.    Navarrete-Opazo, A. et al. Genetics in Med. (2021) 23:2194

4.    Hindorff, L. et al. Nature Rev. Gen. (2018) 19:175

5.    The International Cancer Genome Consortium. Nature (2010) 464:993

6.    Huang, K., et al. Patterns 2 (2021) 2(10)