FLICKR, ALEX PROIMOSWhen will whole human genome sequencing hit the long-sought-after $1,000 mark? How will scientists bear the interpretation challenges? Can a patient be trusted with their own genomic data?
As the din of debate on the utility, price, and ethics of human genomics grows louder, the techniques are already being incorporated into everyday routine operations in health care. Clinical genetics labs—which continue to make next-generation sequencing (NGS) technologies higher throughput and more accurate for a lower cost—have already launched a first foray of lab-developed tests that use genomics data to diagnose a variety of diseases. Labs estimate cost savings of 10 to 50 times over complex genetic tests, with turn-around times of days and weeks rather than weeks and months.
In the NGS world, everyone marvels at the dramatic fall in sequencing costs: it is 100 million times cheaper than it was only 10 years ago when...
Other costs associated with sequencing are not falling as quickly, however. If you truly account for every dollar—the library preparation, target enrichment, sequencing machine depreciation, reagents, lab personnel, power, building costs, data transfer, computation, storage, backup, and time for clinical geneticists to analyze the data—an exome quoted at $700 actually costs a total of $2,882. That’s at a sequencing coverage of about 30x, measuring each position on the exome 30 times, on average. Interestingly, if you more than triple the coverage to 100x—a reasonable depth for clinical use—the price only goes up 15 percent to $3,309. This demonstrates that the cost of exome sequencing is not driven by sequencing depth, but by the challenge of data interpretation.
That requires software. The software used to analyze NGS data in the clinic must meet higher standards than in research environments. No false negatives. Few false positives. Intuitive reports that physicians can understand. But many early stage companies are stepping up to the challenge, developing solutions in this space, and a few products are already on the market.
So with both the software and the clinical geneticists ready, NGS-based diagnostic tests are already here. One concern regarding their implementation, however, lies in preparing patients for a look inside their DNA. What if they learn of a genetic predisposition to a disease they don’t have? What if a genetic disease is found that has implications for relatives? What if the relatives didn’t want to know? Can patients really handle their own genomes? To address this concern, clinical geneticists have carefully crafted informed consent processes and offer genetic counseling to help patients interpret their data.
Another critique relates to overall knowledge of the genome. What if we find 3 million variants in a given genome, but only 30 of them are clinically actionable? What is our obligation regarding the millions of incidental findings?
Patients won’t benefit from seeing a list of 2.9 million variants of unknown or no significance. Indeed, labs are reporting on what is clinically actionable and saving the rest. They view their obligation as an ongoing process, however, continuing to send opted-in patients alerts when genetic variations found in their genome become characterized. Patients can further speed this advance by allowing their genomes to be used in research.
With these safeguards in place, innovators in leading diagnostic labs around the world are launching more and more gene and exome panels as lab-developed tests. Meanwhile, patients are ready for genomic diagnostics and are migrating to doctors who make this medical advance available to them. As a result of these trends, patients are getting answers more quickly, insurers are paying less, and behind it all, the clinical genome is growing.
Richard Resnick is the CEO of GenomeQuest, whose GQ-Dx genomic decision support system enables molecular laboratories to produce routine, actionable diagnostic reports from NGS data.