The Dose That Never Landed: Antidepressant Pharmacogenomics and the Enzymes That Decide

Key Takeaways:

• CYP2D6 and CYP2C19 metabolize most SSRIs and tricyclics, and genotype sorts patients into poor, intermediate, normal, rapid, and ultrarapid metabolizers — the same milligram dose can produce wildly different blood levels across those phenotypes.
• Predicted CYP2D6 ultrarapid metabolizer status ranges from roughly one percent to twenty-one percent depending on the population studied, and in one Ethiopian cohort twenty-nine percent carried duplicated or multiduplicated CYP2D6 genes.
• Pharmacogenomic-guided prescribing works, but modestly: in the 1,944-patient PRIME Care trial it sharply reduced drug-gene mismatches yet raised remission by only about 2.8 percentage points, an effect that had faded by week 24.

July is Minority Mental Health Month, and antidepressant pharmacogenomics deserves a place in that conversation for a strictly biochemical reason. Two liver enzymes, CYP2D6 and CYP2C19, clear most serotonin reuptake inhibitors and most tricyclics. Both genes are highly polymorphic, and the alleles a person inherits place them somewhere on a spectrum from poor metabolizer through intermediate, normal, and rapid, all the way to ultrarapid. A poor metabolizer given a standard dose of a CYP2C19-dependent drug can stack up plasma concentrations far above the intended range and collect side effects that get misread as intolerance. An ultrarapid metabolizer does the opposite: the liver strips the drug out faster than it accumulates, plasma levels never reach the therapeutic window, the patient feels nothing, and the chart gains the phrase “treatment-resistant.” The dose never landed. That is not the same problem as a depression that resists treatment.

Allele frequencies are not evenly distributed across human populations, and pretending otherwise does patients no favors. Pooled genotype data across more than sixty thousand subjects predict poor metabolizer status in as few as 0.4 percent and as many as 5.4 percent of people depending on the population, while predicted ultrarapid status ranges from about one percent to twenty-one percent. A classic study of healthy Ethiopian volunteers found that twenty-nine percent carried duplicated or multiduplicated functional CYP2D6 genes — a genuinely high ultrarapid burden, and one echoed across parts of North and East Africa and the Middle East. CYP2C19 runs the other direction: loss-of-function alleles are common in East and Southeast Asian populations, pushing more people toward poor metabolizer status and higher drug exposure. Here is the part that matters. These are population-level probabilities, not properties of a person. Ancestry shifts the prior; it never reads out an individual’s genotype. Treating a self-reported race as if it were a lab value is how the biology of mood gets flattened into a stereotype — and it is exactly what a real test replaces.

Does antidepressant pharmacogenomics testing actually work?

Modestly, and honesty about that matters more than enthusiasm. The Clinical Pharmacogenetics Implementation Consortium publishes actionable dosing guidance tied to CYP2D6 and CYP2C19 phenotype — reduce the starting dose or choose a different agent for poor metabolizers, consider an alternative drug not routed through the affected enzyme for ultrarapid metabolizers — and that guidance is well worth following once a genotype is in hand. But in PRIME Care, a pragmatic randomized trial of 1,944 patients with major depression, giving clinicians pharmacogenomic results dramatically cut the prescribing of drugs with predicted gene-drug interactions while improving remission by only a couple of percentage points, a benefit that was no longer significant by twenty-four weeks. So the correct claim is narrow and defensible: this is not a magic bullet, it is a way to stop wasting months on a molecule the patient’s liver was never going to let work. For someone who has already failed two or three antidepressants, a genotype and a careful look at the whole picture is a far better next move than a fourth guess at the same dose.

References:

1. Bousman, C. A., Stevenson, J. M., Ramsey, L. B., Sangkuhl, K., Hicks, J. K., Strawn, J. R., et al. (2023). Clinical Pharmacogenetics Implementation Consortium (CPIC) guideline for CYP2D6, CYP2C19, CYP2B6, SLC6A4, and HTR2A genotypes and serotonin reuptake inhibitor antidepressants. Clinical Pharmacology & Therapeutics, 114(1), 51-68.
2. Gaedigk, A., Sangkuhl, K., Whirl-Carrillo, M., Klein, T., & Leeder, J. S. (2017). Prediction of CYP2D6 phenotype from genotype across world populations. Genetics in Medicine, 19(1), 69-76.
3. Aklillu, E., Persson, I., Bertilsson, L., Johansson, I., Rodrigues, F., & Ingelman-Sundberg, M. (1996). Frequent distribution of ultrarapid metabolizers of debrisoquine in an Ethiopian population carrying duplicated and multiduplicated functional CYP2D6 alleles. The Journal of Pharmacology and Experimental Therapeutics, 278(1), 441-446.
4. Dorji, P. W., Tshering, G., & Na-Bangchang, K. (2019). CYP2C9, CYP2C19, CYP2D6 and CYP3A5 polymorphisms in South-East and East Asian populations: A systematic review. Journal of Clinical Pharmacy and Therapeutics, 44(4), 508-524.
5. Oslin, D. W., Lynch, K. G., Shih, M. C., Ingram, E. P., Wray, L. O., Chapman, S. R., et al. (2022). Effect of pharmacogenomic testing for drug-gene interactions on medication selection and remission of symptoms in major depressive disorder: The PRIME Care randomized clinical trial. JAMA, 328(2), 151-161.