CCPP Module 29, Section 5: Overview of PharmGKB, Genetic Databases, and Reporting Tools
Module 29: Applied Pharmacogenomics in Collaborative Practice

Section 5: Overview of PharmGKB, Genetic Databases, and Reporting Tools

A hands-on tour of the essential digital tools for the modern clinical pharmacist. You will learn how to effectively use the Pharmacogenomics Knowledgebase (PharmGKB) and other key databases to find the latest evidence, interpret complex genetic reports, and stay current in this rapidly evolving field.

SECTION 29.5

Overview of PharmGKB, Genetic Databases, and Reporting Tools

Your Digital Toolkit for Evidence-Based PGx Practice.

29.5.1 The “Why”: From Relying on Reports to Commanding the Data

You have now built a formidable foundation in pharmacogenomics. You understand the science, the vocabulary, the clinical indicators for testing, and the operational hurdles of implementation. The final pillar of expertise is information mastery. The field of PGx is not static; it is one of the most rapidly evolving areas of medicine. New gene-drug associations are discovered, the function of variants is clarified, and clinical guidelines are updated continuously. A PGx test result that is interpreted today may have new clinical implications five years from now based on emerging evidence.

Therefore, your practice cannot be solely dependent on the summary interpretation provided on a commercial lab report. While these reports are a helpful starting point, they represent a snapshot in time and may vary in quality and depth. To function as a true collaborative practice expert, you must possess the skills to go directly to the primary sources of evidence yourself. You must be able to independently verify an interpretation, find the underlying data that supports a clinical guideline, and stay abreast of new developments that could impact your patient’s care long after the initial test was performed. This is the essence of being a lifelong learner and an evidence-based practitioner in the 21st century.

This section is your guided, hands-on tour of the essential digital armory of a pharmacogenomics expert. We will move beyond simply mentioning these tools to conducting a masterclass on how to use them effectively and efficiently. Our primary focus will be the Pharmacogenomics Knowledgebase (PharmGKB), the world’s most comprehensive repository of PGx evidence. We will also explore other critical resources like the CPIC website, dbSNP, ClinVar, and the FDA’s pharmacogenomics table. By the end of this section, you will no longer see a genetic report as a final, static document. You will see it as a key that unlocks a wealth of dynamic, continuously updated information that you can command to provide the highest level of care for your patients throughout their lives.

Pharmacist Analogy: The Evolution of a Drug Information Expert

Think about the resources you use to answer a complex drug information question. Your evolution as a PGx information master will mirror your journey as a traditional drug information pharmacist.

  • Level 1: The Package Insert (The Commercial Lab Report): When you were a student, the first place you looked for information was the drug’s package insert. It’s a useful, FDA-approved summary. It gives you the basics of dosing, side effects, and major warnings. A commercial PGx lab report is like a package insert. It provides a good, compliant summary of the key findings and top-level recommendations. It’s an essential starting point, but it’s not the whole story.
  • Level 2: The Drug Compendium (CPIC and PharmGKB Clinical Annotations): As you gained experience, you graduated to using comprehensive drug compendia like Lexicomp, Micromedex, or Facts & Comparisons. These resources aggregate data from multiple primary sources, grade the evidence, and present it in a structured, clinically focused format. This is the role of the CPIC guidelines and the high-level Clinical Annotations on PharmGKB. They are curated, evidence-based resources designed for clinicians to make decisions. This is where most of your daily work will be done.
  • Level 3: The Primary Literature (PubMed and Raw Databases): For the most complex and novel questions, you know you have to go to the source: the primary clinical trials and case reports found on PubMed. You need to be able to critically evaluate the original study that led to the recommendation in Lexicomp. In the world of PGx, the “primary literature” includes not only PubMed but also the raw genetic databases. When you need to understand the function of a specific rare variant not covered by CPIC, or explore the population frequency of an allele, you need to know how to navigate resources like dbSNP and ClinVar.

A true expert doesn’t just read the summary; they know how to find and evaluate the evidence that the summary is built on. This section teaches you how to move seamlessly between all three levels of PGx information, from the lab report to the clinical guideline to the primary data, making you a self-sufficient and authoritative expert.

29.5.2 PharmGKB Masterclass: Your Primary Evidence Library

The Pharmacogenomics Knowledgebase (PharmGKB.org) is the single most important repository of curated PGx information in the world. It is funded by the National Institutes of Health (NIH) and managed by Stanford University. Its mission is to collect, curate, and disseminate knowledge about the impact of human genetic variation on drug response. It is not a clinical guideline generator like CPIC; it is the comprehensive, meticulously organized library of evidence from which guidelines are built. Proficiency in navigating PharmGKB is a non-negotiable skill.

Core Concepts of PharmGKB Curation

PharmGKB is not just a search engine; it is a knowledgebase. This means that teams of PhD-level scientific curators read the primary literature (clinical trials, cohort studies, case reports) and manually extract the key findings, mapping them to specific genes, drugs, and variants in a standardized way. This expert curation is what makes it so powerful. When you look up a drug-gene pair, you are not just getting a list of papers; you are getting a synthesized summary of the current state of evidence.

Practical Tutorial: A Guided Tour of the PharmGKB Homepage

Let’s begin by dissecting the main entry points into the database. Navigate to www.pharmgkb.org.

  1. The Search Bar: This is your primary tool. You can enter a drug name (e.g., “warfarin”), a gene symbol (e.g., “CYP2C9”), or an rsID for a specific SNP (e.g., “rs9923231”). The search is intuitive and will provide auto-completed suggestions.
  2. Main Navigation Tabs:
    • Drugs: Allows you to browse an alphabetical list of all drugs in the database.
    • Genes: Allows you to browse an alphabetical list of all genes.
    • Clinical Annotations: This takes you to a searchable list of all curated summaries of gene-drug associations, which we will explore in detail.
    • Prescribing Info: A collection of PGx information found in FDA drug labels and recommendations from other international regulatory bodies.
    • Pathways: Curated diagrams illustrating the pharmacokinetic and pharmacodynamic pathways for key drugs, highlighting the role of relevant genes.
  3. Key Resources on the Homepage: The main page often highlights new guideline releases from CPIC and other organizations and provides links to important pages, like the “CPIC Guideline Genes” list.

Deep Dive: The Clinical Annotation

The Clinical Annotation is the heart of PharmGKB’s value for a practicing clinician. It is a concise, evidence-based summary of a specific drug-gene interaction, complete with a Level of Evidence score. When you search for “clopidogrel,” for example, the top results will be Clinical Annotations for the `clopidogrel-CYP2C19` interaction.

Masterclass Table: Understanding PharmGKB Levels of Evidence

PharmGKB assigns a Level of Evidence to each Clinical Annotation to help you quickly assess the strength of the data supporting that association. Understanding this hierarchy is crucial for critical appraisal.

Level Description Criteria Clinical Implication
Level 1A Association is found in a CPIC or other major clinical practice guideline. High-level evidence, well-established, and supported by a professional body. The association is considered clinically actionable. Highest Priority. This information should be used in clinical practice. The clopidogrel-CYP2C19 interaction is a classic example.
Level 1B Association is supported by at least one well-designed study in a large cohort and has a clear biological basis. Strong evidence, but not yet incorporated into a major guideline. Often a precursor to a future CPIC guideline. Clinically Actionable. Warrants strong consideration in clinical decision-making, especially if other options have failed.
Level 2A Association is found in a meta-analysis or replication study with a large sample size. Moderate evidence. The association has been replicated, suggesting it is likely real, but the clinical effect size may be small or variable. Use with Caution. Can be used as a secondary piece of information to help explain a patient’s response, but may not be strong enough to guide therapy on its own.
Level 2B Association is found in at least one well-designed study. Moderate evidence from a single study. Needs replication before it can be considered robust. Informational. Good for hypothesis generation but generally not strong enough for clinical action without more supporting data.
Level 3 Association is found in a study that is not well-replicated or has limitations. Preliminary or conflicting evidence. The association is plausible but not yet well-supported. Not Clinically Actionable. This is the realm of ongoing research. Do not use this information to guide prescribing.
Level 4 Case report, non-significant study, or in-vitro evidence only. Weakest level of evidence. Purely speculative or hypothesis-generating. Not Clinically Actionable.
Practical Tutorial: Analyzing a Clinical Annotation
  1. Search for “ondansetron” on PharmGKB. Click on the top result, which should be the Clinical Annotation for ondansetron and CYP2D6.
  2. Examine the Header: At the top, you will see the key information: Drug (ondansetron), Gene (CYP2D6), and critically, the Level of Evidence (Level 1A). This immediately tells you the association is well-established and actionable.
  3. Read the Annotation Text: The annotation will provide a concise summary. For example: “CYP2D6 ultrarapid metabolizers may have decreased concentrations of ondansetron… This may lead to an increased likelihood of failure of ondansetron therapy.”
  4. Check the “Prescribing” Tab: Within the annotation page, there is a “Prescribing” tab. This will show you the specific text from the CPIC guideline, such as “For CYP2D6 ultrarapid metabolizers… consider an alternative 5-HT3 antagonist not metabolized by CYP2D6, such as granisetron.”
  5. Review the Literature: Click on the “Literature” tab to see a table of all the primary research papers that the curators reviewed to create this annotation, with direct links to the PubMed abstracts. This allows you to trace the evidence back to its source.

29.5.3 The Extended Toolkit: Other Essential Databases

While PharmGKB and CPIC are your day-to-day workhorses, a true expert should be familiar with the other major databases that provide complementary information. These are often the sources that PharmGKB curators use themselves.

The National Center for Biotechnology Information (NCBI) Databases

The NCBI, part of the NIH, hosts a suite of fundamental genetic databases that are the primary repositories for raw genetic data.

  • dbSNP (Database of Single Nucleotide Polymorphisms): This is the definitive public archive for short genetic variations. Every validated SNP is assigned a unique Reference SNP (rs) number, or rsID.
    Clinical Utility: When a lab report or a research paper mentions a specific SNP (e.g., `VKORC1 -1639G>A`), it will often list its rsID (`rs9923231`). You can search this rsID in dbSNP to find detailed information about the variant, including its exact chromosomal location, the nucleotide change, and its observed allele frequencies in different global populations (e.g., you can see that the ‘A’ allele is much more common in Europeans and Asians than in Africans).
  • ClinVar: While dbSNP archives variation, ClinVar archives the interpretation of that variation in relation to human health. It aggregates information from clinical testing labs, research studies, and expert panels, assigning a clinical significance to variants.
    Clinical Utility: ClinVar is broader than just pharmacogenomics; it covers variants associated with Mendelian diseases, cancer, etc. However, it is an excellent resource for checking the interpretation of a variant. You can search by gene or rsID. It uses a standardized terminology for interpretation (e.g., “Benign,” “Likely Benign,” “Uncertain Significance,” “Likely Pathogenic,” “Pathogenic”). For PGx, you will often find entries describing a variant as having “drug response” implications.

The FDA Table of Pharmacogenomic Biomarkers in Drug Labeling

This is a critical regulatory resource. The FDA maintains a publicly available, searchable table that lists all the drugs with pharmacogenomic information in their FDA-approved labeling. This is the source of truth for what the US regulatory body considers an established biomarker.

Practical Tutorial: Using the FDA PGx Table
  1. Find the Table: Search for “FDA Table of Pharmacogenomic Biomarkers”. The official table is hosted on the FDA website.
  2. Understand the Columns: The table lists the Drug, Therapeutic Area, Label Section with PGx Info (e.g., Boxed Warning, Warnings and Precautions, Clinical Pharmacology), and the specific Biomarker (gene).
  3. Use Case: A provider asks if there is any PGx information for the new oncology drug, tepotinib. You can quickly search the FDA table for “tepotinib.” You will find it listed, with the biomarker `MET`. The table indicates that the PGx information is in the “Indications and Usage” section. This tells you that the drug is only indicated for patients with a specific `MET` genetic variant, making genetic testing a prerequisite for its use. This is a powerful and fast way to check for regulatory-endorsed PGx associations.

29.5.4 Putting It All Together: From Lab Report to Actionable Recommendation

Let’s synthesize everything we’ve learned in this module with a final, comprehensive case study that takes you from the moment you receive a PGx report to the moment you deliver a confident, evidence-based recommendation.

The Case

Mr. Chen is a 68-year-old Chinese American male with a new diagnosis of Major Depressive Disorder and neuropathic pain. His provider is considering starting him on amitriptyline. Due to a family history of medication sensitivity, a PGx panel was ordered. You have just received the report.

Sample PGx Lab Report (Abbreviated)
Gene Genotype Predicted Phenotype Lab Interpretation Summary
CYP2D6 *10/*10 Intermediate Metabolizer Significantly reduced metabolism of CYP2D6 substrates. Increased risk of side effects from drugs like TCAs and some SSRIs. Reduced analgesic effect from codeine/tramadol.
CYP2C19 *1/*2 Intermediate Metabolizer Reduced metabolism of CYP2C19 substrates. Increased exposure to drugs like citalopram and PPIs. Reduced activation of clopidogrel.
CYP2C9 *1/*1 Normal Metabolizer Normal metabolism of CYP2C9 substrates such as warfarin and phenytoin.
DPYD *1/*1 Normal Metabolizer Normal DPD enzyme activity. Standard risk for toxicity with 5-FU/capecitabine.

Your Pharmacist Interpretation Workflow: A Step-by-Step Tutorial

The Clinical Question: Is amitriptyline a safe and effective choice for Mr. Chen?
  1. Step 1: Identify Relevant Genes.

    My first step is to determine which genes on this report are relevant to amitriptyline. I don’t rely on memory. I go to PharmGKB.org and search for “amitriptyline.” I look at the pathway diagram and the top-level clinical annotations. PharmGKB confirms that amitriptyline is metabolized to its active metabolite, nortriptyline, primarily by CYP2C19. Both the parent drug and the active metabolite are then cleared from the body primarily by CYP2D6. Therefore, both genes are highly relevant.

  2. Step 2: Verify the Phenotype Interpretation.

    The lab report calls a `CYP2D6 *10/*10` genotype an “Intermediate Metabolizer.” Is this correct? I go to cpicpgx.org, navigate to the “Genes” tab, and click on “CYP2D6.” I open the “Allele Definition” table. I confirm that `*10` is classified as a “Decreased Function” allele with an activity score of 0.5. I then open the “Diplotype-to-Phenotype” table. For a diplotype with a total activity score of 1.0 (0.5 + 0.5), CPIC assigns the phenotype of “Intermediate Metabolizer.” The lab report is consistent with the CPIC definition. I repeat this process for `CYP2C19 *1/*2` and confirm its IM status as well.

  3. Step 3: Consult the CPIC Guideline for an Actionable Recommendation.

    Now that I have the verified phenotypes, I consult the CPIC guideline for Tricyclic Antidepressants. I find the Dosing Recommendations table. I look for the row corresponding to a “CYP2D6 Intermediate Metabolizer” and/or “CYP2C19 Intermediate Metabolizer.”

    The guideline states for a CYP2D6 IM: “Consider a 25% reduction of the recommended starting dose… monitor for response and side effects.” It gives a similar recommendation for a CYP2C19 IM regarding tertiary amines like amitriptyline. Since Mr. Chen has reduced function in both key metabolic pathways, he is at a substantially elevated risk of accumulation and toxicity.

  4. Step 4: Synthesize and Deliver the Recommendation.

    I am now ready to contact the provider with a confident, multi-layered, evidence-based recommendation.

    Your Script for the Provider

    “Hi Dr. Evans, this is the pharmacist calling about Mr. Chen’s PGx results. We received the report, and it provides a clear reason for his family history of medication sensitivity and helps us optimize his new therapy.

    “The results show he is both a CYP2D6 and a CYP2C19 Intermediate Metabolizer. Amitriptyline is heavily metabolized by both of these enzymes. The CPIC guidelines recommend a dose reduction for reduced function in either pathway; because Mr. Chen has reduced function in both, he is at a very high risk of accumulating the drug and experiencing severe anticholinergic side effects, especially given his age.

    “Therefore, I recommend we avoid amitriptyline as a first-line agent. A much safer choice would be an agent that is not a primary substrate of either enzyme, such as desvenlafaxine or vortioxetine. This would give us the best chance of achieving an effective dose without causing problematic toxicity. Would you like me to put in an order for desvenlafaxine to start?”