Skip to main content
Premium Trial:

Request an Annual Quote

Pharmacogenomic Incidental Findings in Exome, Genome Data May Be Clinically Useful


NEW YORK (GenomeWeb) – Researchers at the National Institutes of Health have found that incidental pharmacogenomic findings can be extracted from SNP chip, exome sequencing, and genome sequencing data and may be clinically useful, suggesting that such information should be considered for return to patients along with other types of secondary findings.

In a study published in Genetics in Medicine last week, NIH researchers led by Murat Sincan analyzed exome sequencing and SNP array data from more than 1,000 participants in the NIH Undiagnosed Diseases Program (UDP) and found that each carried at least one pharmacogenetic finding, and that nine had PGx variants associated with medications they had already been prescribed.

Separately, Les Biesecker, a researcher at the National Human Genome Research Institute, reported at the European Society of Human Genetics annual meeting in Barcelona last month that his team has identified pharmacogenetic variants from exome and genome sequencing data of five participants in the NHGRI's ClinSeq study, comparing the results with those from widely used pharmacogenetic arrays.

For their study, Sincan and his colleagues looked for variants that are annotated in the Pharmacogenomics Knowledgebase (PharmGKB) in SNP chip data from 1,101 participants in the UDP who are part of 308 families, and in exome sequencing data from a subset of 645 of these from 158 families. Because patients in the UDP were admitted to the NIH Clinical Center for a week, the researchers had access to the medication history of 359 participants, allowing them to look for PGx findings with immediate relevance to their healthcare.

Two years ago, they had already published a study on secondary findings in the 56 genes recommended for analysis by the American College of Medical Genetics and Genomics in the UDP cohort. But pharmacogenetic variants, which are not currently included in the ACMG guidelines, could also be actionable and clinically useful. "Being actionable is a very important aspect of these secondary findings — if you find something that you can't do anything about, that doesn't really have a lot of impact on the person's health," Sincan, the senior author of the study, told GenomeWeb. "But when you are considering prescribing a drug to the patient and they have a mutation that affects either the efficacy or the toxicity of this drug, then that's an immediately useful, high-impact piece of information that is really important in the delivery of healthcare to that patient."

The SNP chip data yielded 395 different variants of interest, including 19 variants associated with 17 drugs that ranked in the top two PharmGKB categories, 1A and 1B, meaning there is substantive evidence for their clinical relevance. The exome data contained 388 different variants of interest, including 21 contained in PharmGKB categories 1A and 1B. Overall, each individual had at least one PGx finding.

Focusing their attention on the 359 participants for whom they had medication records, the researchers found that none of them had PharmGKB 1A or 1B variants for medications they had already been prescribed. But when the scientists also considered variants in categories 2B and 3, they found five pharmacologically relevant variants in nine individuals.

Three of these variants, each occurring in one individual, appear to play a role in the efficacy of the drugs escitalopram, carbamazepine, and morphine, respectively. The other two variants — found in six individuals — have been implicated in the efficacy of acetylsalicylic acid.

"We realized these are not necessarily life-altering cases where if the pharmacogenomic variant was not taken into consideration, the patient's health would be at severe risk," Sincan said. "But still, this shows us that even in a small cohort like this — and in a situation where the patients were admitted for a screening visit, not treated with intensive therapeutics — we found a good number of matches between their genomic variation and the drugs that they were given in a short amount of time."

The researchers did include a one-sentence summary of the PharmGKB variant with the associated drug information into the report that went into the UDP database, but none of the clinical teams with access to this information chose to return a category 2 or 3 variant to a participant, both because the consent stated that only variants with relevance to the test indication were to be returned and because current ACMG guidelines do not include pharmacogenetic variants.

"Integrating the pharmacogenomic variants into routine clinical care is a bigger conversation than just being able to identify these variants," Sincan said. "It has psychosocial dimensions, ethical dimensions, and technical dimensions. How do we integrate these findings into the electronic medical record and the clinical decision support systems, so when the clinicians are ordering these drugs, they are presented with meaningful alerts and decision support?"

Nonetheless, he said, he and his team hope that the findings pave the way for further discussions, "so we can move forward towards that goal of integrating these pharmacogenomic variants in a meaningful way into routine clinical care."

"I personally believe that the variants, especially high-impact variants, should be returned to the patients," Sincan said. "But more importantly, they should be part of the electronic medical record, similar to the allergy list that you have." Known drug allergies are usually listed in EMRs, he said, and likewise, clinicians should have access to pharmacogenetic information that indicates, for example, drug toxicity.

Since conducting the study, Sincan has left the UDP to become a staff scientist and genomic program administrator at the National Institute of Dental and Craniofacial Research, so he does not know whether the UDP plans to report PGx variants to its participants going forward. A request for comment to the director of the UDP program remained unanswered before deadline.

In the meantime, other groups have also been exploring whether PGx variants can be discovered in exome and genome data. In a workshop at the ESHG annual meeting in Barcelona last month, Les Biesecker from the NHGRI reported on a pilot project that analyzed data from ClinSeq participants for pharmacogenetic variants, an approach he called "opportunistic pharmacogenetic screening."

He and his colleagues looked at five individuals from the ClinSeq cohort who had undergone exome and genome sequencing and were also tested with a widely used pharmacogenetic microarray, which covers about 2,000 variants in 231 genes and which he said is the current gold standard for pharmacogenetic testing.

Overall, they identified 203 PGx variants that have strong recommendations from the Clinical Pharmacogenetics Implementation Consortium (CPIC). Genome sequencing data picked up 99 percent of these, and exome sequencing data 64 percent, while the arrays only found 30 percent of the variants. 

One of the largest barriers to the widespread adoption of pharmacogenetic testing, Biesecker said during the workshop, is its turnaround time, because physicians would like to make immediate decisions on what drugs to prescribe, and at what dose, and do not want to wait for test results.

Opportunistic screening for PGx variants in clinical sequencing data could help with that, he said, because it makes PGx information that could be used at a later stage "instantly available at marginal additional cost." 

Also, he said, by providing additional clinically useful information, such an approach could potentially increase the cost-effectiveness of diagnostic exome or genome sequencing tests.

The Scan

Study Tracks Off-Target Gene Edits Linked to Epigenetic Features

Using machine learning, researchers characterize in BMC Genomics the potential off-target effects of 19 computed or experimentally determined epigenetic features during CRISPR-Cas9 editing.

Coronary Artery Disease Risk Loci, Candidate Genes Identified in GWAS Meta-Analysis

A GWAS in Nature Genetics of nearly 1.4 million coronary artery disease cases and controls focused in on more than 200 candidate causal genes, including the cell motility-related myosin gene MYO9B.

Multiple Sclerosis Contributors Found in Proteome-Wide Association Study

With a combination of genome-wide association and brain proteome data, researchers in the Annals of Clinical and Translational Neurology tracked down dozens of potential multiple sclerosis risk proteins.

Quality Improvement Study Compares Molecular Tumor Boards, Central Consensus Recommendations

With 50 simulated cancer cases, researchers in JAMA Network Open compared molecular tumor board recommendations with central consensus plans at a dozen centers in Japan.