W. Gregory Feero, MD, PhD, Maine Dartmouth Family Medicine Residency
Earlier this fall, the Blue Cross Blue Shield Technology Evaluation Center produced a report [PDF 210.52 KB] evaluating the clinical use of exome sequencing in the diagnosis of rare diseases. That this report was even generated is remarkable, as it marks an appreciable level of penetration by exome sequencing into clinical care. A decade ago, sequencing and interpreting even the small portion of the human genome that the exome represents would have been cost-prohibitive. Now, clinical grade exome sequencing with interpretation can be obtained for around $5-10,000. This is still a relatively big price tag but it is well within an order of magnitude of what insurers cover for many diagnostic tests. The Blue Cross report signals that sequencing with next-generation, high-throughput technologies has progressed in the minds of many from strictly a research endeavor into the realm of clinical reality.
For a person with an undiagnosed, potentially genetic condition, exome sequencing may provide a definitive diagnosis where no other technology can. This benefits the patient, who has likely experienced a diagnostic odyssey, incurring both emotional and physical costs. A diagnosis can also provide families with much needed insight into the magnitude of the risk to other family members. Though a diagnosis may occasionally lead directly to therapeutic insights, this is likely the exception rather than the rule. From an economic perspective, achieving a definitive diagnosis has the potential to prevent unnecessary spending on additional testing. The potential for improved care at a lower cost is in everyone’s interest in the current healthcare environment, but is clearly a major motivator for health insurers considering coverage of exome sequencing. An exploration of economic factors affecting the integration of genomics into healthcare was the subject of a July 2012 Institute of Medicine Genomics Roundtable meeting. A report summarizing the meeting is available.
Low-cost exome and genome sequencing technologies, as a group, represent disruptive technologies that in time could supplant- existing molecular diagnostic approaches. There is considerable momentum in the genomics community to this end. Nevertheless, after attending the October 2013 meeting of the American Society of Human Genetics, I feel there is ample reason to remain cautious about overstating the clinical utility of these tests and to recognize that they do not as of yet replace but rather augment more traditional diagnostic approaches. Three issues that directly affect clinical care need to be carefully considered: the analytic performance of the sequencing platform, assigning meaning to discovered variants, and the level of preparation of the medical community to effectively deal with findings that address the immediate medical question as well as “incidental” findings.
First, next generation sequencing technologies are novel and still evolving. Each technology has strengths and limitations in its ability to detect different variations in the genome. The analytic validity and capacity to reliably identify disease-associated variant varies across the genome and are dependent on the effective use of the sequencing instrumentation and subsequent analysis. There are regions of the genome that cannot be sequenced reliably. It is incumbent on the clinical laboratory to understand, identify, and effectively communicate these limitations to the ordering physician and when indicated, offer an alternate method for variant detection. If a variant is detected, confirmatory testing using an alternate method (usually traditional Sanger sequencing) is typically conducted to minimize false positives. In contrast, a negative result from sequencing is less informative in the context of rare, undiagnosed conditions and in most instances will not rule out the presence of a disease-associated variant In 2011, the CDC convened a national workgroup to establish elements of a quality management system to assure the analytical validity and compliance of next generation sequencing tests with existing regulatory and professional quality standards. The workgroup identified and addressed gaps in quality practices that could compromise the quality of both clinical laboratory services and translational efforts needed to advance the implementation and utility of sequencing in practice. Similar documents were developed by a number of professional groups within the US and Europe. Recently, the College of American Pathology has included evaluation of these new sequencing technologies as part of their accreditation program.
Second, assuming that the analytic accuracy of a given sequencing approach is high enough and a variant is detected, substantial challenges remain in assigning meaning to the variants detected. Databases of pathogenic mutations are available for a number of conditions, but the full spectrum of genetic diversity is uncharted for most disorders and traits. Furthermore, there is growing evidence that a substantial fraction of what we think we know regarding the relationship among genotypes and phenotypes is incorrect or incomplete. Many of the databases that exist were developed for research rather than clinical applications and are not curated for medical use. Prediction of pathogenicity using computational algorithms that incorporate predicted protein structure, regulatory regions of the genome, and sequence conservation across evolution is still a work in progress. Laboratory professionals are aware of these limitations and each takes an individualized approach to addressing them. At this time, there are no standard algorithms that have been adopted for variant interpretation. Consequently, assigning pathogenicity remains as much an art as a science, and an interpretation of the potential pathogenicity of a given variant by one group of experts might differ from that of another.
Finally, available evidence gives us little reason to feel comfortable that the broader medical community is prepared to deal appropriately with all the potential primary and “incidental” findings arising from clinical genome sequencing. Consensus seems to be emerging that it is important to return certain incidental findings in some situations. However, there is not full agreement regarding which findings to return, under what circumstances they should be returned, who will take the responsibility for explaining the findings, and who will assume the downstream management of patients with such results. These issues are immensely complex, emotionally charged, and may cause liability issues. Major medical centers with substantial infrastructure for handling such findings are working out the details, but outside of such centers there is definitely cause to think very carefully about the potential consequences of incidental findings.
The introduction of exome and whole genome sequencing technologies into health care offers unparalleled opportunities for improvements in diagnosis for genetic conditions. Proponents take a “glass half full” perspective and are working to rapidly apply the technology to care. Others take a “glass half empty” position, arguing that the technology is not yet mature, and that further refinements are needed before exome or whole genome sequencing is applied outside of research protocols. Both positions have valid points supporting them. I would argue for a third perspective. Clinicians considering the use of genome sequencing should first carefully consider, with the aid of colleagues having the requisite expertise, not only if the sequencing glass is half-full or half empty but whether or not it is the right tool for the job. In some cases, exome or whole genome sequencing will have a reasonable likelihood of being informative while offering unique advantages over more traditional technologies. In others clinical situations, it may be prudent to stick with proven, if unglamorous, technologies. Making that distinction will remain a considerable challenge for clinicians and the healthcare system for the foreseeable future.
We appreciate your input and feedback on this topic.