What is pharmacogenomics?

Pharmacogenomics is the study of how genes influence people’s response to medications. This field continues to grow as basic science research is translated to clinical medicine. As a result, IMO Health customers are increasingly asking for assistance in this area where precise terminology is vital.

Translating a genetic result into actionable prescribing insights requires significant contributions from and communications between disparate and sometimes disconnected laboratory, pharmacy, and informatics entities within the healthcare ecosystem. Therefore, ensuring the unambiguous exchange of genetic information is paramount. At IMO Health, we are working to help clients integrate complex, standardized genetic nomenclature with terms commonly used in laboratory and clinical settings. We then map the normalized terms to standard code sets, as appropriate, to facilitate the transmission of discrete data elements.

How does pharmacogenomics work?

From a terminology standpoint, the basic building blocks of the pharmacogenomics workflow are genotypes, which reflect a patient’s unique genetic makeup at the level of specific gene, and phenotypes, which reflect a patient’s predicted clinical response to a medication given that genotype. For instance, many pharmacogenomics panels test for the gene CYP2D6, which produces a protein product that is important for metabolizing several drugs in the liver, including the breast cancer drug tamoxifen. A patient may have a CYP2D6 genotype of *1/*5 (they inherited one normal and one non-functioning gene), which predicts a phenotype known as a CYP2D6 intermediate metabolizer (some residual protein function but not as much as the “standard” patient). By properly embedding this individualized knowledge as structured data in the EHR, one can suggest dose adjustments or alternative drugs that are tailored to this individual.

Standardizing to a single clinical terminology

At IMO Health, our goal is to represent each of these genotype and phenotype terms in our database in a canonical form that can serve as a bridge to internationally recognized standardized code sets such as SNOMED® and LOINC. Of the two types of terms, genotypes tend to be more challenging to standardize. This is due, in part, to the highly granular nature of genetic mutations: they can be expressed at a level as small as one substitution among the roughly three billion base pairs that make up the human genome.

Furthermore, various “legacy” terminologies exist, some of which are unique to specific genes. The cytochrome P450 genes, such as CYP2D6 mentioned above, tend to use the “star” system to describe variants, where the official HGNC gene name is addended with an asterisk and number. The G6PD gene has many variants associated with adverse reactions to many medications. These variants have traditionally been named after geographic locales, in some cases posing a challenge to integrating test results into an actionable data set. For instance, the G6PD Mediterranean variant, Caligari variant, and Birmingham variant all refer to the exact same mutation but have different names since they were discovered simultaneously in different locations.

Ideally, genotypes would be normalized to one standardized nomenclature system. In this sense, the Human Genome Variant Society (HGVS) offers hope for creating a uniform system that can be used across different genes. However, adoption of the HGVS nomenclature is far from universal. Furthermore, HGVS terms are typically not “human readable” to those not intimately familiar with its structure. For example, NM_001042351.2:c.563C>T reflects the G6PD variant just mentioned. Therefore, we turn to authoritative organizations in the field, such as the Clinical Pharmacogenetics Implementation Consortium (CPIC) and PharmGKB, when determining our “canonical” terms at IMO Health and crafting them in the most clinician-friendly language possible.

Phenotypes, in contrast, are not as numerous and are better represented in standardized code sets already heavily utilized in health IT. SNOMED CT concepts cover a large percentage of the phenotypes endorsed by CPIC and PharmGKB. Since we already map IMO Health content to SNOMED CT, we have the infrastructure in place to support our clients who want to take advantage of the discrete data representation of pharmacogenomic phenotypes by SNOMED CT.

Like the rest of clinical terminology, pharmacogenomics nomenclature continues to evolve. We will continue to update our content and explore supporting new code sets if it will encourage the uptake of pharmacogenomics for our customers.

To learn more about the importance of precise clinical terminology, download our white paper, A different kind of Rosetta Stone: The pivotal role of clinical interface terminology in healthcare.