Some adverse drug events are a predictable consequence of therapeutic, especially dosing failures. Idiosyncratic toxicities, on the other hand, are much less predictable and much more insidious. Pharmacogenomics is a powerful tool that can reveal some of these vulnerabilities before toxicity occurs or, if performed after the fact, can explain the cause of an adverse drug event that has already happened.
Clinical trials cannot detect all adverse drug events due to the limited population in which the drugs are being tested and the incomplete definition of an adverse safety outcome that would be clinically predictive. Since most studies do not store biological specimens, it is often impossible to investigate the cause of an adverse drug event once the trial has ended. Large biobanking projects present an opportunity to perform targeted pharmacogenomic testing in selected patients and discover biomarkers that are robust and predictive.
Drugs that produce chronic toxicity, such as antiretrovirals and antipsychotics, are manageable by regular monitoring drug levels in the blood. Another common strategy is regular monitoring of liver function tests to prevent drug-induced liver injury. Not all adverse drug events can be prevented by monitoring, however. Rare, idiosyncratic reactions may deter patients from getting an otherwise useful drug, especially when the population at risk cannot be accurately defined. In these instances, pharmacogenomic testing is especially valuable as it can detect patients likely to develop adverse events when taking a specific drug based on their genetic information.
Some drugs become toxic when co-administered with enzyme inhibitors. Terfenadine, a non-sedating antihistamine, is metabolized by CYP3A4 in the liver. Its co-administration with CYP3A4 inhibitors such as erythromycin or grapefruit juice can result in the prolonged QT interval and torsade de pointes. Similarly, a calcium channel blocker mibefradil was also withdrawn from the market. The reason was their potential to accumulate to dangerous levels when administered with a CYP3A4 inhibitor.
Thioridazine, a first-generation antipsychotic, was the first drug where a pharmacogenomic test was required. The drug was contraindicated in patients who tested as CYP2D6 poor metabolizers because of a risk of a potentially fatal arrhythmia. Thioridazine was withdrawn from the market in 2005.
Numerous actionable pharmacogenomic biomarkers are listed on FDA-approved drug labels. In most instances, it is up to the prescribing physician to decide whether a genomic test is appropriate. Currently, testing is only required for a handful of drugs: abacavir (HLA-B), carbamazepine (HLA-B), ivacaftor (CFTR), rasburicase (G6PD), siponimod (CYP2C9), tafenoquine (G6PD), divalproex sodium (POLG), eliglustat (CYP2D6), pimozide (CYP2D6), tetrabenazine (CYP2D6), valproic acid (POLG), velaglucerase alfa (GBA), carglumic acid (NAGS), pegloticase (G6PD) and primaquine (G6PD).
The list of drugs that require mandatory pharmacogenomic testing is only a tip of the imaginary iceberg. As the cost of genomic testing decreases, routine screening for genes responsible for drug toxicity should become a routine part of clinical practice.