Harnessing Genomics in the Rare Diseases Realm

Interpretation of Data

Dr. Lee

Brendan Lee, MD, PhD, chair of molecular and human genetics at Baylor College of Medicine, Houston, said it’s true that the cost of genome and exome sequencing has plummeted, but that is just one part of the problem.

“The cost of sequencing is dropping enormously, but it is ultimately the interpretation of that sequence which is the challenge,” he said.

The results of whole exome sequencing are often not straightforward, pointing to a single diagnosis, Dr. Lee said. A study by a colleague at his center on whole exome sequencing (WES) of adults found that WES yielded a diagnosis rate of 17.5%, with a “dual” diagnosis involving “blended phenotypes” 7.1% of the time.³

Interpretation of WES data changes over time as new discoveries are made.

At Dr. Lee’s center, the diagnostic yield of WES at one time point was 25% in one cohort of 250, and at a later time point it was 36%. In a larger cohort of 2,000, the yield rose from 25% to 30% when re-analyzed later, he said.^4,5

“This is why, in fact, our lab tries in a structural fashion to integrate re-analysis in our pipeline so that when one orders a test, there is … an interval subsequently when there will be a re-analysis,” Dr. Lee said. “And this is not an easy task given the nature of genomic interpretation.”

Exome sequencing of infants in intensive care produced a diagnosis 36.7% of the time, and in 52% of those cases the diagnosis affected medical care, either by redirection of care, referral of a subspecialist or in diet or medication.⁶

The bottom line question, Dr. Lee said, is how do we functionalize the genome?

He works with the Undiagnosed Diseases Network, in which patients apply and, if accepted, are seen at one of 12 regional U.S. sites. About 40% of patients are accepted for evaluation, and the diagnostic yield is 35%—even though many of the patients had exome sequencing done previously.

Dr. Lee said the Undiagnosed Diseases Network is not just out to make new diagnoses. It’s “also a laboratory in thinking about how this can be implemented more broadly.”

Challenges of Individualized Genomic Medicine

The promise is enormous, he said, but real challenges stand in the way of perfecting individualized genomic medicine. One is the knowledge itself, he said. “The knowledge base is evolving, and therefore this changes our interpretation.”

As individualization increases, the evidence base for it is bound to shrink, Dr. Lee said. “Sometimes there’s an inverse correlation with individualization and personalized medicine and what we view as evidence.”

“There are payer issues. And transformation of data into information requires the education of physicians and direct counseling of patients. And then there’s, of course, the regulatory oversight aspect.”