As shown in Table 1, of the 22 antidepressants evaluated, 2 were virtually always (>90%) in the green bin: desvenlafaxine and levomilnacipran; 8 of 22 antidepressants were almost never (≤10.5%) in the green bin: citalopram, duloxetine, escitalopram, fluoxetine, fluvoxamine, mirtazapine, paroxetine, and sertraline. Vilazodone appeared in the “green bin” 84.2% of the time. Paroxetine appeared in the red bin 68.4% of the time, which was the highest percentage for any of the antidepressants or antipsychotics in the red bin.
As shown in Table 2, of the 16 antipsychotics evaluated, 4 were virtually always (>90%) in the green bin: asenapine, lurasidone, paliperidone, and ziprasidone; 2 of the 16 antipsychotics were almost never (10.5%) in the green bin: chlorpromazine and thioridazine.
Given these results, one could predict which drugs are found in the “green bin” >90% of the time simply on the basis of knowledge of their oxidative drug metabolism without knowing the results of the patient’s genetic testing. As expected, drugs such as desvenlaxine and ziprasidone, which do not rely on CYP enzyme-mediated metabolism, were in the “green bin” 100% of the time. Medications that are metabolized approximately equally by multiple CYP and non-CYP enzymes and hence have little variability in blood level based on genetics were found in the “green bin” between 84.2% and 94.7% of the time and included levomilnacipran, vilazodone, asenapine, lurasidone, and paliperidone. For example, vilazodone is metabolized by several CYP enzymes and undergoes metabolism through non-CYP enzyme pathways, limiting its reliance on CYP-mediated drug metabolism. Hence, there is little risk with these medications that the blood level will be higher or lower than anticipated given the nature of their metabolism.
Drugs that are more reliant on CYP enzymes for their metabolism generally appeared in the “yellow” and “red” bins. The greater the reliance on hepatic metabolism and the narrower the therapeutic index, the more likely a drug would appear in the “red bin” versus the “yellow bin.” For example, amitriptyline appeared in the “red bin” 47.4% of the time. As amitriptyline relies heavily on CYP 2D6 for its metabolism, its blood levels vary widely depending upon an individual’s CYP 2D6 genotype. In addition, tricyclic antidepressants have a narrow therapeutic index, and genotypic CYP 2D6 poor metabolizers (ie, individuals who are genetically deficient in CYP 2D6 activity) are at risk for potentially toxic blood levels of such medications. For this reason, one could predict amitriptyline would frequently appear in the “red bin” because of its heavy reliance on this polymorphic CYP enzyme.
Therefore, if a patient did not respond to initial treatment with an antidepressant or had adverse effects related to that agent, providers could use information about the metabolic pathways of potential alternative medications along with the probability that the patient is deficient at specific metabolic pathways based on their ethnicity to choose a medication that has a lower risk of variability in blood level. Such an approach may be more cost-effective than genetic testing and may perhaps also be more practical, given that based on the results of this study, such logic would be consistent with the predictability of the genetic testing results.
Broadening the discussion to review the earlier published data on the effectiveness of GeneSight testing helps put the findings of this paper into perspective. Previously published studies on genetic testing had methodological flaws, including being underpowered or having an open-label design, which could bias the results in favor of the utility of pharmacogenetic testing. Keeping these limitations in mind, 2 open-label studies found statistically significant reduction in symptoms in patients whose treatment was guided by GeneSight testing compared with treatment as usual.3,4 Conversely, a double-blind randomized controlled trial found a trend for greater improvement in the GeneSight group but the differences were not statistically significant, although this study but may have been underpowered to detect a difference.5 Subsequently, GeneSight conducted a large (N=1200 patients) double-blind randomized clinical trial which included patients with treatment-resistant depression.6 This study found statistically significant higher rates of response and remission (P=0.01 and P<0.01, respectively) and greater reduction in symptoms in patients whose treatment was guided by GeneSight testing compared with those whose treatment was not; however, the number needed to treat was small (18 for response and 13 for remission at 8 wk).7
Two important caveats should be considered when examining results from this latest GeneSight study. First, the study did include patients with treatment-resistant depression and those patients may simply not be responsive to currently marketed antidepressants targeted at a biogenic amine mechanism of action regardless of how drug selection is performed, as was shown in the largest National Institute of Mental Health funded study on antidepressants, the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) study.8 Second, the small but positive results in this study and the open-label studies may have occurred, in part, because treatment as usual may not necessarily be guided by a knowledge of the differences in oxidative drug metabolism of the drugs involved, and hence may be inferior to treatment guided by genetic testing. The results presented in this column suggest that a study comparing drug selection based on oxidative drug metabolism versus drug selection guided by pharmacogenetic testing would yield comparable results because of the high predictability of which drugs are classified as “green bin” versus “yellow bin” versus “red bin.” Consistent with this finding, in a 2018 statement, the American College of Neuropsychopharmacology stated that “While the genomics revolution may eventually justify genetic testing in psychiatric patients for clinically meaningful genetic variations, such scientifically supported and responsibly administered testing lies in the future for these patients.”9
Although pharmacogenetic testing will almost undoubtedly prove critical in customizing treatment in psychiatry, the critical question is whether that time is now.10 On the basis of the results presented in this column, pharmacogenetic testing in terms of guiding the selection of specific antidepressant and antipsychotic medications may not yet have reached the threshold of clinical utility.
The authors acknowledge Maggie Searight, MS4, for help with data entry.
1. Rosenblat JD, Lee Y, McIntyre RS, et al. Does pharmacogenomic testing improve clinical outcomes for major depressive disorder? A systematic review of clinical trials and cost-effectiveness studies. J Clin Psychiatry. 2017;78:720–729.
3. Hall-Flavin DK, Winner JG, Allen JD, et al. Utility of integrated pharmacogenomic testing to support the treatment of major depressive disorder in a psychiatric outpatient setting. Pharmacogenet Genomics. 2013;23:535–548.
4. Hall-Flavin DK, Winner JG, Allen JD, et al. Using a pharmacogenomic algorithm to guide the treatment of depression. Transl Psychiatry. 2012;2:e172.
5. Winner JG, Carhart JM, Altar CA, et al. A prospective, randomized, double-blind study assessing the clinical impact of integrated pharmacogenomic testing for major depressive disorder. Discov Med. 2013;16:219–227.
7. Greden JF, Parikh SV, Rothschild AJ, et al. Combinatorial pharmacogenomics to guide treatment selection for major depressive disorder: congruency matters. Poster presented at the Annual Meeting of the American Psychiatric Association, New York, NY, May 2018.
8. Preskorn SH. Results of the STAR*D study: implications for clinicians and drug developers. J Psychiatr Pract. 2009;15:45–49.
10. Preskorn SH, Hatt CR. How pharmacogenomics (PG) are changing practice: implications for prescribers, their patients, and the healthcare system (PG series part I). J Psychiatr Pract. 2013;19:142–149.
Keywords:Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.
pharmacogenetic tests; drug metabolism; cytochrome P450 enzymes; antidepressants; antipsychotics; blood levels; GeneSight test