High Probability of Gene-Drug Interactions Associated with Medication Side Effects in Adolescent Depression: Results from a Randomized Controlled Trial of Pharmacogenetic Testing

Abstract

Introduction:

Combinatorial pharmacogenetic testing panels are widely available in clinical practice and often separate medications into columns/bins associated with low, medium, or high probability of gene-drug interactions. The objective of the Adolescent Management of Depression (AMOD) study was to determine the clinical utility of combinatorial pharmacogenetic testing in a double-blind, randomized, controlled effectiveness study by comparing patients who had genetic testing results at time of medication initiation to those that did not have results until week 8. The objective of this post hoc analysis was to assess and report additional outcomes with respect to significant gene-drug interactions (i.e., a medication in the high probability gene-drug interaction bin as defined by a proprietary algorithm) compared with patients taking a medication with minimal to moderate gene-drug interactions (i.e., a medication from the low or medium probability gene-drug interaction bin, respectively).

Methods:

Adolescents 13–18 years (N = 170) with moderate to severe major depressive disorder received pharmacogenetic testing. Symptom improvement and side effects were assessed at baseline, week 4, week 8, and 6 months. Patients were grouped into three categories based on whether the medication they were prescribed was associated with low, medium, or high risk for gene-drug interactions. Patients taking a medication from the low/medium gene-drug interaction bin were compared with patients taking a medication from the high gene-drug interaction bin.

Results:

Patients taking a medication from the high gene-drug interaction bin were more likely to endorse side effects compared with patients taking a medication in the low/medium gene-drug interaction bin at week 8 (p = 0.001) and 6 months (p < 0.0001). Depressive symptom severity scores did not differ significantly across the medication bins.

Conclusions:

This study demonstrates the utility of gene-drug interaction testing to guide medication decisions to minimize side effect burden rather than solely prioritizing the search for the most efficacious medication. (Clinical Trials Registration Identifier: NCT02286440).

Introduction

Timely identification and treatment of depression in childhood and adolescence is critical and could prevent residual, long-term depressive symptoms. Finding the most effective medication with the least number of side effects as quickly as possible for each individual patient has always been the goal. Unfortunately, about 40% of pediatric patients do not respond to their first trial of an antidepressant (Brent et al., 2008; Emslie, 2012). In addition, children and adolescents often experience more side effects than adults (Luft et al., 2018; Rynn et al., 2015), which can lead to early discontinuation of treatment (Akpinar et al., 2016; Al Maruf et al., 2019).

Pharmacogenetic testing, a personalized approach to choosing a medication based on the evaluation of high probability of gene-drug interactions, has been used in some medical specialities, such as optimizing medications in the treatment of cancer (Phillips et al., 2001; Schroth et al., 2009). Similarly, there have been early implementation efforts in the field of psychiatry (Bousman et al., 2023).

Combinatorial pharmacogenetic testing provides genetic results for gene variations for multiple cytochrome P450 genes (pharmacokinetics) and genes that affect neuronal function (pharmacodynamics), such as receptor and transporter genes. Combinatorial panels have decision support tools that provide recommendations on dosing strategy and medication choice, which are based on proprietary algorithms of potential gene-drug interactions.

These decision support tools often have medications divided into three categories/bins associated with low, medium, or high probability of gene-drug interactions: “use as directed,” “use with caution,” and “use with increased caution and with more frequent monitoring” (Greden et al., 2019; Hall-Flavin et al., 2012; Tanner et al., 2018).

The Adolescent Management of Depression (AMOD) study was the first double-blind, randomized, controlled effectiveness study evaluating the impact of pharmacogenetic testing in adolescents with depression (Vande Voort et al., 2022). The Genomics Used to Improve Depression Decisions (GUIDED) trial, which had a similar design to the AMOD study, included a large, prospective cohort of 1167 adult patients with depression (Thase et al., 2019).

Although neither trial showed a separation between treatment arms for those patients who had pharmacogenetic testing results versus those without at the time of medication initiation, the GUIDED trial did show improved outcomes when specifically evaluating those patients with high probability of gene-drug interactions (Thase et al., 2019; Vande Voort et al., 2022). However, to date, outcomes associated with gene-drug interactions by bin in this type of randomized, controlled effectiveness study have not been examined in adolescents with depression.

To address this gap, we conducted a post hoc analysis from the AMOD study to evaluate outcomes in those patients taking a medication associated with significant gene-drug interactions compared with patients taking a medication with minimal to moderate gene-drug interactions. We hypothesized that those patients taking a medication from the high gene-drug interaction bin would have more adverse events/side effects (AEs/SEs) and reduced improvement in depressive symptoms as measured by the Children's Depression Rating Scale-Revised (CDRS-R) (Poznanski et al., 1984) than those patients taking a medication from the low to medium gene-drug interaction bin. The primary outcome measures were the number of AEs/SEs and the continuous change in CDRS-R scores from baseline to week 8.

Methods

Participants

All participants were between the ages of 13–18 years and diagnosed with moderate to severe major depressive disorder. Patients 13–17 years old provided informed assent with parents providing informed consent. Patients 18 years old and their parents provided informed consent. A real-world sample of participants were recruited from 2015 to 2018 from multiple settings, including an inpatient psychiatric hospital, outpatient specialty clinic, and community primary care clinics in Rochester, Minnesota, and surrounding communities. Research procedures were approved by the Mayo Clinic Institutional Review Board.

Inclusion and exclusion criteria, study procedures, and specifics regarding the pharmacogenetic testing have been previously published (Vande Voort et al., 2022). Briefly, the major depressive episode was assessed via the Kiddie Schedule for Affective Disorders and Schizophrenia-Past and Lifetime (K-SADS-PL), and patients were included if they had a score of ≥40 on the CDRS-R. Patients were excluded if they scored ≥15 on the Young Mania Rating Scale (YMRS) (Young et al., 1978) or had a diagnosis of autism spectrum disorder, anorexia nervosa, bipolar disorder, schizophrenia spectrum disorder, current substance use disorder other than that involving nicotine, caffeine, or cannabis, or were pregnant.

Study procedures

As previously described (Vande Voort et al., 2022), all participants were assessed with the K-SADS-PL at the screening visit and provided a buccal swab for pharmacogenetic testing after assent/consent was signed. Participants were seen by a board-certified child and adolescent psychiatrist (J.L.V.V., S.S.O., J.S.) for a baseline visit (when medication was initiated), at 4 weeks, 8 weeks, and 6 months. Multiple assessments were completed at each timepoint by a trained rater and overseen by a board-certified child psychiatrist (P.E.C.).

Depressive symptom severity was assessed with the CDRS-R (primary outcome measure); Quick Inventory of Depressive Symptomatology-Clinician Rated, Adolescent Self Report, and Parent Report [QIDS-A17-CR, QIDS-A17-SR, QIDS-A17-SR(P), respectively] (Bernstein et al., 2010). Other outcome measures included the Children's Global Assessment Scale (CGAS) (Shaffer et al., 1983); YMRS; Frequency, Intensity, and Burden of Side Effects Ratings (FIBSER) (Wisniewski et al., 2006) and Safety Monitoring Uniform Research Form-Modified (SMURF-M) (Greenhill et al., 2004) to track the number of AEs/SEs.

The FIBSER has shown reliability and validity as a self-report measure in a population of adults treated for depression (Wisniewski et al., 2006), but this is yet to be validated in a child population to our knowledge. This 3-item measure asks participants over the past week to: (1) rate the frequency of side effects on a scale of 0–6; (2) rate the intensity of side effects on a scale of 0–6; and (3) rate the degree that side effects have interfered with day-to-day functioning on a scale of 0–6 (Wisniewski et al., 2006).

Although not yet validated, the SMURF was developed to assess possible adverse events in children/adolescents through the Research Units on Pediatric Psychopharmacology, which is a network sponsored through the National Institute of Mental Health aimed at investigating psychotropic medications in patients <18 years old (Greenhill et al., 2004). The SMURF-M includes a brief General Inquiry, a Drug-Specific Inquiry, and a comprehensive Body System Review (Greenhill et al., 2004). AEs/SEs were measured as a total symptom count.

Patients and parents were not aware of the pharmacogenetic testing results or if prescribed medications were associated with low, medium, or high probability gene-drug interaction bins until after all ratings were completed at the week 8 visit. The prescribing clinician was aware of the testing results for patients in one arm of the randomization.

Pharmacogenetic testing

As previously reported (Vande Voort et al., 2022), pharmacogenetic testing was completed by Assurex Health, Inc. (Mason, OH), and the panel utilized in this study included eight genes (CYP2D6, CYP2C9, CYP2C19, CYP3A4, CYP2B6, CYP1A2, SLC6A4, and HTR2A). Alleles for each gene evaluated in the combinatorial panel can be found in Supplementary Table S1. The decision support tools for this panel included 22 medications classified as antidepressants, 16 medications classified as antipsychotics, and 8 medications classified as anticonvulsants.

All medications were separated into one of three bins: low (“use as directed”), medium (“use with caution”), and high (“use with increased caution and with more frequent monitoring”) probability of gene-drug interactions. Dosing recommendations (such as “higher or lower doses may be required”) are found as footnotes for those medications associated with the high probability of gene-drug interactions. It should be noted that the company performing the testing may redefine some genotypes, and differences may exist between the testing results in this study and Clinical Pharmacogenetics Implementation Consortium (CPIC) phenotypes.

Statistical analysis

Demographics, clinical characteristics, and medications are summarized using means and standard deviations (SD) or frequencies and percentages. Comparisons of these across medication bins were made using chi-square tests, one-way analysis of variance, or Kruskal-Wallis tests. Repeated-measure longitudinal mixed models were used to test for differences across all three bins or between the high gene-drug interaction bin and combined low/medium gene-drug interaction bin for adverse events and symptom measures.

Combining low/medium gene-drug interaction bins and comparing outcomes with the high gene-drug interaction bin is consistent with prior studies (Greden et al., 2019) and ensures adequate statistical power. p-Values are shown unadjusted and at a Bonferroni-adjusted significance level of α = 0.05/4 = 0.0125 is used for significance in discussing results for four statistical tests conducted for two definitions of bins (three-level for separate bins and high vs. low/medium gene interaction bins) and two additional time points (8 weeks and 6 months). Analyses were conducted using SAS (version 9.4; Cary, NC).

Results

Participant description

There were 196 patients screened for this study. Of those, 170 were included in the secondary analysis (n = 60 taking a baseline medication from the low gene-drug interaction bin, n = 83 from the medium gene-drug interaction bin, and n = 27 from the high gene-drug interaction bin). Among them, 125 (73.5%) completed the 6-month visit (n = 43 taking a baseline medication from the low gene-drug interaction bin, n = 64 from the medium gene-drug interaction bin, and n = 18 from the high gene-drug interaction bin) with no significant difference by bins (p = 0.520). The mean age of participants was 15.4 years (SD 1.5, range 13–18), and 41 were male (24.1%). Table 1 provides the demographic characteristics in detail.

Table 1.

Summary of Demographics, Clinical Characteristics, and Medications Prescribed During the Study by Primary Medication Bin

	Low gene-drug interation		Medium gene-drug interaction		High gene-drug interaction		p
	n	Mean (SD) or %	n	Mean (SD) or %	n	Mean (SD) or %	p
Age, years	60	15.2 (1.5)	83	15.6 (1.5)	27	15.3 (1.4)	0.318
Sex, male (%)	60	12 (20.0)	83	25 (30.1)	27	4 (14.8)	0.177
Race, white (%)	58	48 (82.8)	80	69 (86.3)	27	23 (85.2)	0.851
Anxiety (%)	60	26 (43.3)	83	35 (42.2)	27	9 (33.3)	0.659
PTSD (%)	60	3 (5.0)	83	6 (7.2)	27	0 (0)	0.343
ADHD (%)	60	9 (15.0)	65	17 (20.7)	23	4 (14.8)	0.616
Dysthymia (%)	60	6 (10.0)	82	5 (6.1)	27	2 (7.4)	0.688
Cannabis use (%)	60	2 (3.3)	82	6 (7.3)	27	7 (25.9)	0.002
Substance use (%)	59	1 (1.7)	83	6 (7.2)	27	3 (11.1)	0.178
Number of past AD trials	60	1.4 (1.7)	83	1.3 (1.7)	26	1.4 (1.2)	0.577
Duration of illness, years	59	1.4 (1.7)	82	1.1 (1.3)	27	1.1 (0.7)	0.588
Age of onset, years	52	13.4 (1.9)	72	13.9 (2.5)	23	13.3 (1.7)	0.156
SSRI (%)	60	40 (66.7)	83	69 (83.1)	27	22 (81.5)	0.058
SNRI (%)	60	12 (20.0)	83	10 (12.1)	27	3 (11.0)	0.147
Atypical AD (%)	60	7 (11.7)	83	4 (30.8)	27	2 (15.4)	0.326
Antipsychotic (%)	60	1 (1.7)	83	0 (0)	27	0 (0)	—
Fluoxetine (%)	60	22 (36.7)	83	33 (39.8)	27	17 (63.0)	0.057
Escitalopram (%)	60	8 (13.3)	83	26 (31.3)	27	4 (14.8)	0.023

AD, antidepressant; ADHD, attention-deficit/hyperactivity disorder; PTSD, posttraumatic stress disorder; SD, standard deviation; SNRI, serotonin norepinephrine reuptake inhibitor; SSRI, selective serotonin reuptake inhibitor.

Anxiety disorders were the most prevalent comorbidity among the participants, and attention-deficit/hyperactivity disorder (ADHD) was the second most common comorbidity. The majority of the participants were prescribed a selective serotonin reuptake inhibitor (SSRI) with fluoxetine being most frequently prescribed.

Adverse events/side effects

Participants taking a baseline medication in the high gene-drug interaction bin were more likely to endorse AEs/SEs compared with those patients taking a medication in the low/medium gene-drug interaction bin at week 8 (p = 0.001) and 6 months (p < 0.0001) (Table 2). The number of patients reporting side effects across various organ systems is reflected in Supplementary Table S2.

Table 2.

Outcomes at Baseline, Week 8, and 6 Months and Comparison of Changes from Baseline by Bin

Outcome measure	Timepoint	Low gene-drug interaction	Medium gene-drug interaction	High gene-drug interaction	Three level ^a	Low/medium vs. high gene-drug interaction	Longitudinal p-value^b
Outcome measure	Timepoint	n = 43	n = 64	n = 18	Δ p	Δ p	Longitudinal p-value^b
AE/SE Total	Baseline	1.11 (2.01)	0.96 (2.14)	0.58 (1.39)	—	—	0.480
	8 Weeks	0.76 (1.48)	0.84 (1.52)	1.59 (2.28)	0.004	0.001
	6 Months	0.53 (1.28)	0.31 (0.87)	1.28 (1.81)	<0.0001	<0.0001
YMRS	Baseline	0.58 (1.01)	0.48 (1.20)	0.69 (1.38)			0.016
	8 Weeks	0.48 (1.13)	0.16 (0.69)	0.45 (1.01)	0.107	0.448
	6 Months	0.18 (0.62)	0.05 (0.21)	0.72 (1.27)	0.0003	0.0001
CDRS-R	Baseline	57.7 (8.8)	57.6 (8.2)	59.9 (7.5)	—	—	0.588
	8 Weeks	35.4 (10.2)	34.6 (10.7)	34.3 (9.2)	0.743	0.540
	6 Months	32.3 (10.9)	29.9 (10.2)	32.5 (9.8)	0.475	0.654
CGAS	Baseline	46.2 (5.7)	47.2 (6.2)	45.2 (4.7)	—	—	0.298
	8 Weeks	68.4 (11.9)	70.4 (13.4)	71.0 (13.3)	0.538	0.398
	6 Months	75.1 (12.6)	78.6 (14.0)	75.8 (14.2)	0.514	0.841
QIDS total	Baseline	16.9 (4.1)	16.1 (3.1)	16.7 (2.5)	—	—	0.686
	8 Weeks	9.6 (4.4)	10.3 (4.9)	9.9 (3.9)	0.443	0.738
	6 Months	9.0 (4.4)	8.1 (4.4)	9.0 (3.7)	0.613	0.808

Three-level categorical test using ANOVA or Kruskal-Wallis tests.

Type 3 test of three-level categorical fixed-effect from model, including the corresponding outcome measure at all three timepoints.

AE/SE, adverse event/side effect; ANOVA, analysis of variance; CDRS-R, Children's Depression Rating Scale-Revised; CGAS, Children's Global Assessment Scale; QIDS, Quick Inventory of Depressive Symptomatology; YMRS, Young Mania Rating Scale.

Mean YMRS scores were consistently low throughout the study. However, at 6 months, the mean YMRS score was noted to be statistically higher for patients taking a medication from the high gene-drug interaction bin compared with those taking a medication from the low/medium gene-drug interaction bin (p = 0.0001).

Symptom improvement

At baseline, the mean CDRS scores for the low, medium, and high gene-drug interaction bins were 57.7 (SD 8.8), 57.6 (SD 8.2), and 59.9 (SD 7.5), respectively. At week 8, the mean CDRS scores for the low, medium, and high gene-drug interaction bins were 35.4 (SD 10.2), 34.6 (SD 10.7), 34.3 (SD 9.2), respectively. There were no statistically significant differences in the change from baseline to week 8 (p = 0.540) or 6 months (p = 0.654) on this primary outcome measure between the low/medium gene-drug interaction bin and high gene-drug interaction bin. Additional analyses showed no statistical differences in improvement regarding the CGAS or QIDS for patients taking a medication in the low/medium gene-drug interaction bin versus the high gene-drug interaction bin (Table 2).

Discussion

The original, primary analysis from the AMOD trial did not show a difference in depression improvement or the rate of AEs/SEs for those adolescents whose psychiatrist had access to the pharmacogenetic testing results at baseline (GENE arm) compared with those that had access to the results at week 8 (Treatment as Usual [TAU] arm) (Vande Voort et al., 2022). However, this result could have been diluted by patients in the TAU arm who coincidently received a medication with no gene-drug interactions.

The current study reported here went a step further and analyzed outcomes, such as rates of AEs/SEs and depression improvement, by medication bin, which represents those medications with low, medium, and high probability for gene-drug interactions. Results showed that patients taking a medication from the high gene-drug interaction bin had increased rates of AEs/SEs compared with those patients taking a medication from the low or medium gene-drug interaction bin at week 8 and 6 months.

Although knowing any individual's risk for the high probability of gene-drug interactions can only occur after the testing has been ordered, the results of this study suggest that side effect burden for youth struggling with depression may be minimized by having knowledge of an individual's pharmacogenetic profile.

Although pharmacogenetic testing results may provide helpful clinical utility regarding the minimization of side effects by avoiding medications with significant gene-drug interactions, this study did not find a difference in depression improvement, by either the CDRS or QIDS, for those patients taking a medication in the low, medium, or high gene-drug interaction bin. There was also no difference among the groups when looking at improvement via the CGAS.

Although there was a statistically significant finding regarding the YMRS, which showed patients taking a medication from the high gene-drug interaction bin had statistically higher scores at 6 months, this has uncertain clinical significance given the low YMRS scores overall throughout the study. To our knowledge, there is no established correlation in the literature between higher YMRS scores and individuals with depression experiencing pronounced side effect burdens. There were no patients in the study who had any emergence of mania/hypomania symptoms after initiation of an antidepressant.

The results of this study support recent recommendations and conceptualized frameworks of how pharmacogenetic testing can be implemented into clinical practice (Bousman et al., 2022; Ramsey et al., 2021). For example, rather than ordering pharmacogenetic testing and selecting a medication from the “low gene-drug interaction bin” with the expectation that depression improvement will occur, a more thoughtful approach of examining specific medication-gene pairs could prove more useful.

The CPIC Guidelines provide a growing body of evidence that specific gene-drug interactions [e.g., CYP2C19 metabolizer status and use of (es)citalopram] can impact drug concentrations and provide recommendations on dose adjustments to minimize side effect burden (Aldrich et al., 2019; Bousman et al., 2023; Hicks et al., 2015). In addition, Ramsey et al. (2021) suggested implementation recommendations that focused on appropriate diagnosis, choosing a medication based on current practice guidelines, and using pharmacogenetic testing results when those are available to the prescribing clinician.

If these results are available for an individual patient, utilizing the U.S. Food and Drug Administration (FDA) labels and CPIC guidelines for dosing recommendations at the gene-drug level is recommended.

This type of approach can have meaningful clinical implications. For example, some antidepressant treatments, such as SSRIs, can have early-emerging side effects (i.e., gastrointestinal symptoms, insomnia, tiredness, fatigue, sedation, and activation) and/or persistent/late-emerging side effects (i.e., weight gain and sexual dysfunction) (Strawn et al., 2023).

Many of the early-emerging side effects can be transient in nature, but it is clinically useful to understand how those side effects correlate with elevated medication concentrations (separate from dose), which are impacted by pharmacokinetic metabolizer status. For example, patients with slower metabolism through CYP2C19 may have higher levels of “exposure” to escitalopram, which has correlated with increased side effects (e.g., activation) (Aldrich et al., 2019; Strawn et al., 2023; Strawn et al., 2020). Having this information can inform the clinician to reduce the titration schedule of the medication and/or the target dose to help minimize side effects and reduce the likelihood of early medication discontinuation.

Pharmacokinetic properties can also impact tolerability when patients struggle with medication nonadherence or choose to discontinue their medication under the guidance of their physician. For example, sertraline's short half-life can lead to withdrawal, flu-like symptoms (as compared with a mediation with a longer half-life, such as fluoxetine). A patient who is an ultrarapid metabolizer at CYP2C19, an enzymatic pathway for sertraline, may be even more susceptible to withdrawal side effects with missed medication doses as the maximum concentration of sertraline can quickly and dramatically decrease in their body (Strawn et al., 2023). This same premise can be applied when a patient and prescriber intentionally want to discontinue a medication.

When the prescriber is informed about the ultrarapid metabolizer status at CYP2C19, they may choose a taper that is slower in nature to avoid withdrawal symptoms (Strawn et al., 2023). Patients may be more likely to return to medications in the future when needed if they do not have a negative experience with medications initially.

By utilizing the type of systematic approach outlined by Ramsey et al. (2021), pharmacogenetic testings results can be implemented to promote medication choices that integrate pharmacokinetic data to reduce side effects and potentially mitigate the need for early medication discontinuation, which ultimately delays the treatment many youth desperately need.

Further, this type of approach may also reduce the likelihood that clinicians will be influenced solely by the bins associated with a medication. As noted in the original AMOD study, pharmacogenetic testing results with colored columns appeared to influence prescribing practices, and clinicians were more willing to deviate from prior evidence-based treatment algorithms if testing results were available to them (Vande Voort et al., 2022).

Strengths and limitations inherently exist in this study and were previously reported (Vande Voort et al., 2022). Additional limitations for this secondary analysis include the small number of patients (n = 18) who were taking a medication associated with the high probability of gene-drug interactions. This group was too small to then categorize by response and remission. Therefore, only the continuous change in CDRS and QIDS from baseline to week 8 or 6 months was utilized to evaluate the association between depression improvement and medication bin.

A second potential limitation could reside in potential bias when the prescribing psychiatrist knew the patient was taking a medication from a high gene-drug interaction bin. In this scenario, the prescriber may have implicitly engaged in more careful questioning regarding side effects, leading the patient to report more side effects on the rating scales. However, we do think this potential bias is mitigated to some degree by having the patient complete rating scales of side effects before meeting with the psychiatrist at each timepoint.

Nonetheless, this potential bias still exists as patients could be more cognizant of side effects from one timepoint to another and could report these side effects at later timepoints. A third limitation to this study is that the company performing the testing defines phenotypes differently than what is found through CPIC guidelines. For example, for CYP2C19, *1/*17 is classified as a normal metabolizer in this study, whereas it is classified as a rapid metabolizer phenotype per CPIC guidelines (Hicks et al., 2015).

A fourth limitation is the proprietary nature of the algorithm, and it is not common knowledge to the patient, clinician, or the researcher on how pharmacokinetic and pharmacodynamic genes impact medications getting grouped into various columns/bins.

Despite these limitations, this study continues to provide guidance of how to conceptualize the use of pharmacogenetic testing and its implementation in clinical practice. Although data from the AMOD study are elusive in suggesting pharmacogenetic testing can help provide selection of the “most efficacious” medication for a particular individual, this study suggests that testing may help guide clinicians in avoiding medications that place patients at risk for increased side effects. This finding can have clinical merit as patients deserve to find a tolerable medication as quickly as possible to avoid constant switching of medications, which ultimately delays the timely treatment they desperately need.

Conclusion

Patients taking a medication that was associated with the high probability of gene-drug interactions were found to have increased rates of AEs/SEs at week 8 and 6 months compared with patients who were taking a medication with minimal to moderate gene-drug interactions. However, results did not show any association between depression improvement and the risk of gene-drug interactions. Utilizing pharmacogenetic testing results to evaluate individual gene-drug pairs may help mitigate the risk of side effects when selecting a medication for depression in adolescents.

Footnotes

Clinical Significance

Although the evidence remains elusive for combinatorial pharmacogenetic testing to guide clinicians to select the “best” medication for any single patient, this study does suggest that pharmacogenetic testing results may help guide clinicians away from medications that could be associated with increased side effects. By avoiding mediations that are prone to gene-drug interactions, patients may find medications that are more tolerable and have fewer side effects. In turn, patients will be more likely to continue treatment and refrain from medication switching, which ultimately allows patients to experience the response they are critically seeking.

Disclosures

Dr. Vande Voort was a site primary investigator for a multicenter study funded by the National Network of Depression Centers. Dr. Croarkin has received research support from the National Institutes of Health (NIH), National Science Foundation, and Brain and Behavior Research Foundation. Dr. Croarkin has received research grant support from Pfizer, Inc., and equipment support from Neuronetics, Inc., and MagVenture, Inc. He served as the primary investigator for a multicenter study funded by Neuronetics, Inc., and a site primary investigator for a study funded by NeoSync, Inc. Dr. Croarkin has served on advisory boards for Engrail Therapeutics, Myriad Neuroscience, and Sunovion. He served as a paid consultant for Procter & Gamble Company and Meta Platforms, Inc. Dr. Frye has received grant support through Assurex Health, Inc., Mayo Foundation, Milken Institute. He has received CME/Travel/Honoraria through Carnot Laboratories and American Physician Institute. His Financial Interest/Stock ownership/Royalties include Chymia, LLC. He has no Speakers' Bureau to disclose. All authors are currently employed or have been employed by Mayo Clinic, which had a prior financial interest in Assurex Health, Inc., and currently has a financial interest in OneOme and the technology referenced with this publication. All other authors report no biomedical financial interests or potential conflicts of interest.

Supplementary Material

Supplementary Table S1

Supplementary Table S2

References

Akpinar

, Cerit

, Talas

, et al. Agomelatine versus sertraline: An observational, open-labeled and 12 weeks follow-up study on efficacy and tolerability. Clin Psychopharmacol Neurosci, 2016; 14:351–356.

Al Maruf

, Greenslade

, Arnold

, et al. Antidepressant pharmacogenetics in children and young adults: A systematic review. J Affect Disord, 2019; 254:98–108.

Aldrich

, Poweleit

, Prows

, et al. Influence of CYP2C19 metabolizer status on escitalopram/citalopram tolerability and response in youth with anxiety and depressive disorders. Front Pharmacol, 2019; 10:99.

Bernstein

, Rush

, Trivedi

, et al. Psychometric properties of the Quick Inventory of Depressive Symptomatology in adolescents. Int J Methods Psychiatr Res, 2010; 19:185–194.

Bousman

, Mukerjee

, Men

, et al. Encountering pharmacogenetic test results in the psychiatric clinic. Can J Psychiatry, 2022; 67:95–100.

Bousman

, Stevenson

, Ramsey

, et al. Clinical Pharmacogenetics Implementation Consortium (CPIC) guideline for CYP2D6, CYP2C19, CYP2B6, SLC6A4, and HTR2A genotypes and serotonin reuptake inhibitor antidepressants. Clin Pharmacol Ther, 2023; 114:51–68.

Brent

, Emslie

, Clarke

, et al. Switching to another SSRI or to venlafaxine with or without cognitive behavioral therapy for adolescents with SSRI-resistant depression: The TORDIA randomized controlled trial. JAMA, 2008; 299:901–913.

Emslie

. The psychopharmacology of adolescent depression. J Child Adolesc Psychopharmacol, 2012; 22:2–4.

Greden

, Parikh

, Rothschild

, et al. Impact of pharmacogenomics on clinical outcomes in major depressive disorder in the GUIDED trial: A large, patient- and rater-blinded, randomized, controlled study. J Psychiatr Res, 2019; 111:59–67.

10.

Greenhill

, Vitiello

, Fisher

, et al. Comparison of increasingly detailed elicitation methods for the assessment of adverse events in pediatric psychopharmacology. J Am Acad Child Adolesc Psychiatry, 2004; 43:1488–1496.

11.

Hall-Flavin

, Winner

, Allen

, et al. Using a pharmacogenomic algorithm to guide the treatment of depression. Transl Psychiatry, 2012; 2:e172.

12.

Hicks

, Bishop

, Sangkuhl

, et al. Clinical Pharmacogenetics Implementation Consortium (CPIC) guideline for CYP2D6 and CYP2C19 genotypes and dosing of selective serotonin reuptake inhibitors. Clin Pharmacol Ther, 2015; 98:127–134.

13.

Luft

, Lamy

, Delbello

, et al. Antidepressant-induced activation in children and adolescents: Risk, recognition and management. Curr Probl Pediatr Adolesc Health Care, 2018; 48:50–62.

14.

Phillips

, Veenstra

, Oren

, et al. Potential role of pharmacogenomics in reducing adverse drug reactions: A systematic review. JAMA, 2001; 286:2270–2279.

15.

Poznanski

, Grossman

, Buchsbaum

, et al. Preliminary studies of the reliability and validity of the children's depression rating scale. J Am Acad Child Psychiatry, 1984; 23:191–197.

16.

Ramsey

, Namerow

, Bishop

, et al. Thoughtful clinical use of pharmacogenetics in child and adolescent psychopharmacology. J Am Acad Child Adolesc Psychiatry, 2021; 60:660–664.

17.

Rynn

, Walkup

, Compton

, et al. Child/adolescent anxiety multimodal study: Evaluating safety. J Am Acad Child Adolesc Psychiatry, 2015; 54:180–190.

18.

Schroth

, Goetz

, Hamann

, et al. Association between CYP2D6 polymorphisms and outcomes among women with early stage breast cancer treated with tamoxifen. JAMA, 2009; 302:1429–1436.

19.

Shaffer

, Gould

, Brasic

, et al. A children's global assessment scale (CGAS). Arch Gen Psychiatry, 1983; 40:1228–1231.

20.

Strawn

, Mills

, Poweleit

, et al. Adverse effects of antidepressant medications and their management in children and adolescents. Pharmacotherapy, 2023; 43:675–690.

21.

Strawn

, Mills

, Schroeder

, et al. Escitalopram in adolescents with generalized anxiety disorder: A double-blind, randomized, placebo-controlled study. J Clin Psychiatry, 2020; 81(5):20m13396.

22.

Tanner

, Davies

, Voudouris

, et al. Combinatorial pharmacogenomics and improved patient outcomes in depression: Treatment by primary care physicians or psychiatrists. J Psychiatr Res, 2018; 104:157–162.

23.

Thase

, Parikh

, Rothschild

, et al. Impact of pharmacogenomics on clinical outcomes for patients taking medications with gene-drug interactions in a randomized controlled trial. J Clin Psychiatry, 2019; 80(6):19m12910.

24.

Vande Voort

, Orth

, Shekunov

, et al. A randomized controlled trial of combinatorial pharmacogenetics testing in adolescent depression. J Am Acad Child Adolesc Psychiatry, 2022; 61:46–55.

25.

Wisniewski

, Rush

, Balasubramani

, et al. Self-rated global measure of the frequency, intensity, and burden of side effects. J Psychiatr Pract, 2006; 12:71–79.

26.

Young

, Biggs

, Ziegler

, et al. A rating scale for mania: Reliability, validity and sensitivity. Br J Psychiatry, 1978; 133:429–435.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB