Impact of CYP2D6 Genotype and Inhibitor Use on Risperidone Metabolism in Children: Functional Insights Into the *17 and *29 Alleles

ABSTRACT

Risperidone, an atypical antipsychotic, is increasingly prescribed in pediatric patients with psychiatric disorders. It is primarily metabolized by CYP2D6 to 9‐hydroxyrisperidone, an active metabolite associated with a higher risk of adverse effects. The CYP2D6*17 and *29 alleles are prevalent in individuals of African ancestry and less studied than variants found in European ancestry individuals. This knowledge gap contributes to differences in health outcomes in individuals of non‐European origin. The primary objective of this study is to replicate recently identified risperidone‐specific activity for the CYP2D6*17 and *29 alleles. During psychiatric hospitalization, CYP2D6 was genotyped as part of routine care. Remnant plasma specimens were analyzed for risperidone and 9‐hydroxyrisperidone by tandem liquid chromatography mass spectrometry in 161 patients administered risperidone. The log‐transformed metabolite‐to‐parent ratio was used to estimate CYP2D6 enzymatic activity. The effect of each CYP2D6 allele on activity was assessed with linear regression that included strong CYP2D6 inhibitor use. Patients were predominantly white, non‐Hispanic youth, ages 5–18 years old (mean 12.5 years), and 26.7% were prescribed concomitant CYP2D6 inhibitors. The frequency of *17 and *29 alleles were 15.5% and 8.3%, respectively, among Black patients (n = 42). After accounting for CYP2D6 inhibitors in the model, the activity of the *17 allele was > 4‐fold that of the *1 allele (p = 0.006) and the *29 allele was 10% that of the *1 allele (p = 0.021), comparable to alleles with no function. The *17 allele confers greater metabolic activity for risperidone than reflected in current pharmacogenetic guidelines. Using existing activity scores may underestimate metabolism in *17 carriers and overestimate it in *29 carriers. Incorporating substrate‐specific activity scores and dosing recommendations would support improved dosing.

Study Highlights

• What is the current knowledge on the topic?

  • ○
    The CYP2D6*17 allele appears to exhibit substrate‐specific effects, including increased metabolic activity for risperidone due to a variant in the binding pocket of the enzyme.
• What questions did this study address?

  • ○
    Can we replicate the previous study's finding of increased enzyme activity in CYP2D6*17 allele carriers and no function in *29 allele carriers? How does phenoconversion affect risperidone metabolism?
• What does this study add to our knowledge?

  • ○
    This study replicates a prior finding that CYP2D6*17 demonstrates increased metabolic activity for risperidone (> 4‐fold more than the *1 allele) and phenoconversion in *17 allele carriers was incomplete in the presence of strong CYP2D6 inhibitors. It also replicated the finding of very little function in *29 allele carriers.
• How might this change clinical pharmacology or translational science?

  • ○
    Using the current CYP2D6 activity value of 0.5 for the *17 and *29 alleles to translate to a metabolizer status for risperidone dosing could put patients at risk of side effects. In addition, it could reduce treatment effectiveness or prolong time‐to‐response if clinicians under‐dose individuals based on existing guidelines that classify *17 as a reduced function allele. Risperidone dosing recommendations based on CYP2D6 should specifically consider the substrate‐specific impact of the *17 and *29 alleles.

1. Introduction

Cytochrome P450 2D6 (CYP2D6) is a highly polymorphic gene encoding the hepatic enzyme that contributes to the metabolism of approximately 20%–25% of drugs. More than 100 allelic variants in the CYP2D6 gene have been identified [1, 2]. Currently, the Clinical Pharmacogenetics Implementation Consortium (CPIC) and the Dutch Pharmacogenetics Working Group (DPWG) recommend using CYP2D6 test results, if available, to optimize dosing for a multitude of drugs [3, 4]. CYP2D6 enzyme activity is calculated based on a scoring system where activity values of both inherited alleles are summed, resulting in an activity score that is intended to be used as a binning system rather than a percentage of activity [5]. Based on these activity scores, individuals are classified as poor metabolizers (PM), intermediate metabolizers (IM), normal metabolizers (NM), or ultrarapid metabolizers (UM) [6]. CPIC regularly reviews and updates these activity scores as new evidence emerges, which was the case for CYP2D6*9 and *41 in 2022 [7].

Drug‐induced phenoconversion resulting from CYP2D6 inhibitors such as fluoxetine, bupropion, and paroxetine can further complicate the prediction of phenotypes [8, 9]. Phenoconversion can transform genotype‐predicted NMs or UMs to become phenotypically PMs or IMs. For example, fluoxetine, a strong CYP2D6 inhibitor, is generally considered to functionally phenoconvert individuals to PMs regardless of genotype [9, 10]. Cannabis and CBD are also implicated as CYP2D6 inhibitors [11].

Risperidone, a second‐generation (atypical) antipsychotic, has increasingly been prescribed for managing a range of psychiatric disorders in children and adolescents. While most first‐generation antipsychotics primarily bind to D₂ receptors, atypical antipsychotics have a broader range of receptor affinities and are antagonists at 5‐HT_2A receptors [12]. Risperidone was first approved for schizophrenia and bipolar disorder in adults, and later received FDA approval for irritability associated with autism spectrum disorder in children aged 5 and older, schizophrenia in adolescents aged 13 and older, and bipolar I disorder in youth aged 10 and older. It is also frequently used off‐label to manage aggression, impulsivity, and severe behavioral dysregulation in ADHD, oppositional defiant disorder, and conduct disorder [13, 14].

Risperidone is hydroxylated by CYP2D6 into the metabolite 9‐hydroxyrisperidone (paliperidone), which is primarily excreted unchanged. Both risperidone and 9‐hydroxyrisperidone are pharmacologically active, and the total active moiety of risperidone + 9‐hydroxyrisperidone is used for therapeutic drug monitoring, guidelines, and efficacy evaluation [15]. Side effects can vary between adults and children and, in youth, include weight gain, hyperprolactinemia, sialorrhea, and drug‐induced parkinsonism and akathisia. These side effects may result in medication discontinuation or switching to alternative medications. Risperidone and paliperidone also have differences in their side effect profiles, with risperidone having a higher incidence of weight gain and paliperidone more likely to produce hyperprolactinemia [16, 17]. Further, adverse drug reactions are more common in CYP2D6 PMs compared to other metabolizer phenotypes [18].

Genetic variation in CYP2D6 and its impact on medication response and tolerability—particularly for antidepressants and antipsychotics—has historically been studied in adult populations of European ancestry despite Africa having the most genetically diverse population in the world [19]. As a result, activity scores are more likely to be misassigned for alleles that are underrepresented or insufficiently characterized, leading to inaccurate predictions of pharmacokinetics. This knowledge gap contributes to differences in treatment outcomes in individuals of non‐European ancestry [20]. In addition to these effects at the individual level, adverse drug reactions impose a significant burden on healthcare systems, increasing rates of hospitalization and overall healthcare costs [21].

Two alleles, CYP2D6*17 and *29, are common in individuals of African ancestry, with allele frequencies of 17%–20% and 9%–12%, respectively [17, 20]. Current CPIC guidelines assign an activity value of 0.5 to the *17 and *29 alleles for all substrates [22, 23], yet previous studies suggest higher activity values for risperidone. In 2009, the *17 allele was found to have normal activity [24], while more recent data suggest that *17 has an increased activity and *29 has no activity when considering only risperidone‐treated patients [25]. These recent findings further suggest substrate‐specific effects for these alleles and highlight the need for broader pharmacogenetic testing in historically underrepresented populations. The primary objective of this study is to replicate recently identified risperidone‐specific activity for the CYP2D6*17 and *29 alleles.

2. Materials and Methods

2.1. Sample Collection and Concentration Analysis

This study was approved by the Institutional Review Board at Cincinnati Children's Hospital Medical Center (CCHMC). As previously described, an alert system was established using the VigiLanz clinical surveillance platform to identify when a blood specimen was collected within 24 h of an risperidone administration in psychiatrically‐hospitalized youth [26]. Specimens were obtained in EDTA‐anticoagulated tubes during routine complete blood count testing as part of standard clinical care. After clinical assays were completed, residual blood was retrieved from the CCHMC clinical laboratory. Plasma was separated from the remnant specimens and stored at −80°C until analysis. Demographic and clinical variables (e.g., sex, age, race, ethnicity, weight, dose, hours since the last dose, and concomitant fluoxetine, paroxetine, or bupropion use) were collected from the electronic health record (Table 1). Cannabis use was confirmed by a positive urine drug screen during admission, or clinician documented cannabis use during the same admission as the sample was collected. Cannabidiol use was recorded according to the same criteria. Data was collected in REDCap [27, 28].

TABLE 1.

Cohort description.

	Total (N = 161)
Sex
Female	60 (37.3%)
Male	101 (62.7%)
Age (years)
Mean (SD)	12.5 (3.3)
Range	5.0–18.0
Weight (kg)
Mean (SD)	59.5 (23.0)
Range	19.5–133.3
Race
Black	42 (26.1%)
White	107 (66.5%)
Other	7 (4.3%)
Not available/unknown	5 (3.1%)
Ethnicity
Hispanic	7 (4.3%)
Not Hispanic	152 (94.4%)
Not available/unknown	2 (1.2%)
Dose (mg/day)
Mean (SD)	1.35 (1.1)
Range	0.125–6.0
Dose‐adjusted risperidone concentration (ng/mL)
BLOQ (N)	20
Mean (SD)	4.9 (13.7)
Range	0.07–149.7
Dose‐adjusted 9‐hydroxyrisperidone (ng/mL)
BLOQ (N)	2
Mean (SD)	7.1 (7.5)
Range	0.46–62.3
Dose‐adjusted active moiety (ng/mL)
BLOQ for either parent or metabolite (N)	21
Mean (SD)	12.6 (16.6)
Range	1.9–169.4
Hours since last dose
Median (IQR)	12.2 (10.9–15.9)
Range	0.017–85.0
Concomitant Cannabis
No	142 (88.2%)
Unsure	2 (1.2%)
Yes	17 (10.6%)
Concomitant Cannabidiol	4 (2.5%)
Concomtiant Fluoxetine	15 (9.3%)
Concomitant Bupropion	29 (18.0%)
Concomitant Paroxetine	2 (1.2%)

Abbreviations: BLOQ, below the limit of quantitation; IQR, interquartile range; SD, standard deviation.

2.2. Genotyping and Genotype‐Defined CYP2D6 Enzyme Activity Scores

As previously described, pharmacogenetic testing for CYP2D6 was conducted in the context of routine clinical care, with genomic DNA extracted from either whole blood or buccal swabs during the patient's initial psychiatric hospitalization [29]. Genotyping was conducted by the institutional Genetics and Genomics Diagnostic Laboratory. From 2013 to 2019, a TaqMan low‐density array (Applied Biosystems, Forest City, CA) in combination with long‐range PCR was employed to detect CYP2D6 alleles (*2, *3, *4, *5, *6, *7, *8, *9, *10, *11, *14, *15, *17, *18, *19, *20, *40, *41, *42, *44) and gene deletion/duplication. Beginning in 2019, the MassARRAY system (Agena Bioscience, San Diego, CA) was utilized, with an expanded panel assessing 22 CYP2D6 alleles (*2, *3, *4, *5, *6, *7, *8, *9, *10, *11, *12, *14, *15, *17, *18, *19, *20, *29, *36, *41, *69, *114), including copy number variation via long‐range PCR or the VeriDose CYP2D6 CNV panel [30]. The *1 allele was inferred when no known variant alleles were present. Activity values, activity scores, and genotype‐to‐phenotype assignments were based on CPIC guidelines, categorizing patients as PMs, IMs, NMs, or UMs for CYP2D6 [31]. All pharmacogenetic data, including genotypes and predicted phenotypes, were uploaded to the electronic health record. Phenoconversion was considered when a patient was concomitantly prescribed a strong inhibitor (fluoxetine, paroxetine, or bupropion). No patients were prescribed duloxetine, a moderate inhibitor of CYP2D6. In the presence of a strong inhibitor, the activity score was multiplied by 0 to generate a phenoconverted activity score and phenotype [8].

2.3. Quantification of Drug Concentrations

Plasma concentrations of risperidone and 9‐hydroxyrisperidone were analyzed by tandem liquid chromatography mass spectrometry as previously described [25]. Briefly, extracts were analyzed on a Waters Xevo TQ‐XS Triple Quadrupole Mass Spectrometer with a Waters Acquity FTN‐I (ultra‐performance liquid chromatography I Class Plus) in positive mode using a Waters Acquity ultra‐performance liquid chromatography BEH C18, 1.7 μm, 2.1 × 50 mm column. The calibration curve range of detection for the assay was 0.1–100 ng/mL for both analytes. When the concentration for one analyte was below 0.1 ng/mL, the concentration was set to 0.05 ng/mL (n = 19 for risperidone, n = 1 for 9‐hydroxyrisperidone) [32]. When the concentration for both analytes was below 0.1 ng/mL (n = 1), the sample was excluded. The total active moiety was calculated by summing the risperidone and 9‐hydroxyrisperidone concentrations.

2.4. Ratio‐Defined CYP2D6 Enzyme Activity

We used the metabolite:parent (9‐hydroxyrisperidone:risperidone) metabolic ratio to estimate CYP2D6 enzyme activity. We calculated the ratio‐defined CYP2D6 activity as previously described [25, 33]. Briefly, we set the median metabolic ratio of patients carrying the *1/*1 diplotype not taking an inhibitor to 1.0 and those carrying two no function alleles (i.e., poor metabolizers) to 0. Although it has normal function, we did not include *2 carriers with the *1/*1 patients to calculate activity because of discrepancies in its function across substrates [34, 35]. The ratio‐defined CYP2D6 activity for each patient was calculated according to the equation below:

$$\left(\mathrm{MR}-\mathrm{x}\right)/\left(\mathrm{y}-\mathrm{x}\right)$$

where MR is the individual patient's metabolite:parent ratio (9‐hydroxyrisperidone:risperidone), x is the median metabolic ratio of the poor metabolizers, and y is the median metabolic ratio of the patients with a *1/*1 diplotype that did not take CYP2D6 inhibitors. In order to log‐transform the activity, negative activity values were set to 0.0001 (n = 7 PMs, n = 1 pPM, and n = 1 NM). Throughout the manuscript, the terms “activity value” and “activity score” refer specifically to the CPIC‐defined term, while the activity we determined experimentally is referred to as the “ratio‐defined CYP2D6 activity.”

2.5. Statistical Analyses

We used linear regression to determine the effect of CYP2D6 inhibitors and alleles on risperidone metabolism. As our primary goal was to replicate prior findings for the activity of the *17 and *29 alleles, we performed analyses as described by Kehinde et al. [25]. Natural log‐transformed ratio‐defined CYP2D6 activity was used as the primary dependent variable and clinical covariates were evaluated as predictors. Age, sex, race, ethnicity, dose, and hours since the last dose did not significantly influence the log‐transformed ratio‐defined CYP2D6 activity in univariate analyses, but the concomitant prescription of a strong inhibitor did. Therefore, a linear regression model including all CYP2D6 alleles also included concomitant strong inhibitor use. The coefficient for each allele was exponentiated to determine the effect size and compared to the effect size of the *1 allele.

Based on the analysis by Kehinde et al. [25], we reassigned activity values for CYP2D6 *17 (from 0.5 to 2) and *29 (from 0.5 to 0) and these new activity values were used to generate revised CYP2D6 genotype‐defined activity scores (with and without phenoconversion) to assess the influence on the log of the ratio‐defined CYP2D6 activity.

Pairwise comparisons between metabolizer phenotypes were performed on the log‐transformed ratio‐defined CYP2D6 activity using a Wilcoxon rank‐sum exact test adjusted for multiple tests using the Benjamini‐Hochberg method. Analyses were performed in RStudio version 2025.05.1 with R version 4.4.0, and findings were considered statistically significant at p < 0.05.

3. Results

Samples were obtained from 161 pediatric patients receiving risperidone (Table 1). Patient ages ranged from 5 to 18 years old with a mean age of 12.5 years and were predominately white and non‐Hispanic. Strong CYP2D6 inhibitors were prescribed in 26.7% of patients. There were 17 (10.6%) and 4 (2.5%) patients using cannabis and CBD, respectively, at the time of sample collection. Among the CYP2D6 alleles identified in the total allelic count (N = 322), *1 (32.9%) and *2 (24.5%) were the most prevalent. Among Black patients, *17 and *29 were observed at frequencies of 15.5% and 8.3%, respectively (Table S1).

When evaluating the functional impact of the *17 allele, the effect size in the multivariable model was 4.2 relative to the *1 allele (p = 0.006) when accounting for concomitant CYP2D6 inhibitors (Table 2). The *29 allele had an effect size similar to that of the no function alleles *4 and *5 (p = 0.021). Including each individual allele in the model with the concomitant CYP2D6 inhibitors, 59.9% of the variability in log‐transformed ratio‐defined activity was explained. However, including the inhibitors and the current CYP2D6 activity score explained 35.4% of the variability (adjusted R²), while using the CYP2D6 activity score with an updated activity value of 2 for *17 and 0 for *29 (proposed by Kehinde et al., Table S2) explained 46.2% (adjusted R²) of the variability.

TABLE 2.

The estimated activity of each CYP2D6 allele calculated from the log‐transformed ratio of 9‐hydroxyrisperidone to risperidone in a multivariable linear regression model including concomitant inhibitor use (effect size for inhibitor −2.32, p < 0.001).

Allele	Count with detectable risperidone	Model coefficient	p	Effect size relative to *1	Current activity value
*1	91 (86%)	0.264	0.58	1.0	1.0
*2	71 (90%)	−0.206	0.67	0.6	1.0
*3	3 (75%)	−1.893	0.052	0.1	0.0
*4	41 (89%)	−2.054	0.0003	0.1	0.0
*5	18 (95%)	−1.331	0.04	0.2	0.0
*6	6 (100%)	−6.142	< 0.0001	0.0	0.0
*9	2 (100%)	1.110	0.38	2.3	0.25
*10	5 (100%)	0.024	0.98	0.8	0.25
*17	12 (67%)	1.708	0.006	4.2	0.5
*29	6 (86%)	−1.831	0.021	0.1	0.5
*41	22 (96%)	−1.04	0.06	0.3	0.25

Note: p value shows whether the allele is significant in the multivariable model. The effect size column represents the exponent of the model coefficient, divided by the effect size of the *1 allele.

There were 43 patients concomitantly prescribed a strong CYP2D6 inhibitor, so phenoconversion was considered to be complete, converting 16 IMs and 26 NMs to PMs. Strong inhibitors significantly increased the log‐transformed dose‐adjusted concentration of risperidone (p = 0.00001) and significantly reduced the ratio‐defined CYP2D6 activity (p = 0.0001), but did not significantly influence the log‐transformed dose‐adjusted concentration of the metabolite, 9‐hydroxyrisperidone (p = 0.15) or the total active moiety (p = 0.32). There was significantly higher ratio‐defined CYP2D6 activity in phenoconverted PMs compared to genotype predicted PMs (Figure 1, p = 2.5 × 10⁻⁶, Wilcoxon rank‐sum exact test adjusted for multiple tests), but lower than the activity of the IMs (p = 1.2 × 10⁻⁷). Cannabis use did not significantly affect the dose‐adjusted risperidone concentration, dose‐adjusted 9‐hydroxyrisperidone concentration, active moiety concentration, or ratio‐defined CYP2D6 activity (all p > 0.4).

FIGURE 1.

Log‐transformed ratio‐defined CYP2D6 activity shows incomplete phenoconversion. Medians are indicated with horizontal lines, error bars indicate interquartile range. Abbreviations: PM = genotype‐predicted PM; pPM = phenoconverted PM; IM = intermediate metabolizer; NM = normal metabolizer; UM = ultrarapid metabolizer; ns = not significant; ***, p < 0.001, pairwise Wilcoxon rank sum test with continuity correction.

Five patients with a *17 allele were concomitantly prescribed a strong CYP2D6 inhibitor and were considered to have a phenoconverted activity score of 0. In combination with their *17 allele, four of these patients had a no function allele and one patient had a *1 allele, but all appeared to have incomplete phenoconversion (Figure 2).

FIGURE 2.

CYP2D6*17 allele carriers taking strong CYP2D6 inhibitors (phenoconverted activity score of 0) have higher CYP2D6 activity than other phenoconverted PMs. Triangles indicate patients with a *17 allele, and circles indicate patients without a *17 allele. Activity score groups with at least 3 patients with a *17 allele were plotted separately. ****, p < 0.0001, Mann–Whitney test.

4. Discussion

The primary goal of our study was to replicate findings from previous studies demonstrating increased activity of the CYP2D6*17 allele and decreased activity of the *29 allele that is specific to risperidone [24, 25, 36]. In this separate cohort of children and adolescents, we were able to replicate these observations as well as demonstrate incomplete phenoconversion when a CYP2D6 inhibitor was concomitantly prescribed.

CYP2D6*17 and *29 alleles are predominantly observed in individuals of African ancestry. The observed frequencies for the CYP2D6*17 and *29 alleles were 15.5% and 8.3%, respectively, which were similar to those reported in previous studies (17% and 9%) [37, 38].

Our analyses reveal the activity of the *17 allele to be > four‐cofold higher than that of *1, a level slightly higher but consistent with a previous findings in a large Nigerian cohort [25] and other previous studies [10, 18, 24, 36]. This finding contrasts with current allele function determined by CPIC, which classifies *17 as a decreased function allele with an activity score of 0.5 for all substrates and is used in dosing recommendations for several other drugs (e.g., opioids, SSRIs, etc.) [3, 31, 39, 40]. This may even be an under‐representation of the speed of metabolism of risperidone, as five of the thirteen *17 allele carriers not on an inhibitor had risperidone concentrations that were undetectable within 24 h after a dose and were imputed to half the lower limit of quantification. Caution should be used when comparing the function of other alleles that were not as well‐represented in our study (e.g., *3, *9).

CPIC assigns allele function categories that are translated into metabolizer phenotypes to guide prescribing recommendations. However, the current activity scores do not account for substrate specificity. Instead, these scores are based on phenotypic data derived from a limited number of probe drugs (e.g., dextromethorphan, sparteine, and debrisoquine for CYP2D6) [41, 42, 43, 44]. Our results, along with the results of the Kehinde study [25] and the de Leon studies [24, 36] may compel reassessment of the activity scores of the CYP2D6*17 and *29 alleles for risperidone. More broadly, these findings highlight the need for future pharmacogenetic research and guideline development efforts to more carefully consider the specific drug or substrate when assigning activity scores, particularly for allele–substrate combinations that are not well‐represented in the literature.

In this study, over 25% of the patients were prescribed a concomitant CYP2D6 inhibitor, allowing us to evaluate the extent of phenoconversion. Patients prescribed a concomitant inhibitor had higher activity than the patients whose genotype predicted them to be PMs, indicating that the phenoconversion was incomplete, although it was lower than in the IMs.

While this study provides a valuable replication of a recently published study in a pediatric cohort, it has several limitations. First, it was conducted using a single‐center cohort, which may limit the generalizability of the findings. However, similar associations between CYP2D6*17 and risperidone concentrations have been reported in other populations, including studies by Cai et al. and Kehinde et al. [25, 36]. Second, despite focusing on alleles that are prevalent among individuals of African ancestry, most patients were white and non‐Hispanic. Additionally, the sample size for several alleles were small, which limits our power to detect significant effects. We did not account for adherence or dosing history of risperidone or CYP2D6 inhibitors, which could have introduced variability in the concentration measurements. While this study examined the impact of CYP2D6 alleles on risperidone metabolism, clinical outcomes such as therapeutic efficacy, side effect burden, or medication discontinuation were not assessed. Further, we used targeted genotyping of known variants; therefore, rare variants that affect CYP2D6 activity and risperidone metabolism via other enzymes (e.g., CYP3A5) that may be present were not identified. Finally, patients were included based on the availability of remnant blood samples collected during routine clinical care which may introduce selection bias, as individuals undergoing blood testing during an inpatient admission may differ systematically from the broader pediatric population prescribed risperidone, including those who do not require psychiatric hospitalization.

Importantly, despite these limitations, our findings remain interpretable and clinically meaningful. This study leveraged a large, well‐characterized pediatric inpatient cohort that reflects real‐world prescribing practices, dosing variability, and transdiagnostic clinical complexity, capturing the heterogeneity of illness severity, comorbidity, and concomitant medication use seen in routine care and providing the power to detect clinically relevant pharmacokinetic differences across CYP2D6 phenotypes. Importantly, our study accounted for phenoconversion and more than one quarter of patients (26%) were receiving CYP2D6 inhibitors, compared with only 1.9% in a prior pediatric risperidone study [25] thereby substantially enhancing real‐world validity. In addition, the sample included individuals with cannabis use, with approximately 10% of patients reporting cannabis use. Taken together, these features strengthen the generalizability of the findings and support their relevance to real‐world clinical decision‐making for risperidone‐treated pediatric patients.

Author Contributions

M.H.B., E.A.P., S.E.V., B.R., P.T., J.R.S., and L.B.R.: wrote the manuscript. M.H.B., E.A.P., S.E.V., J.R.S., and L.B.R.: designed the research. M.H.B., E.A.P., S.E.V., B.R., P.T., J.R.S., and L.B.R.: performed the research. M.H.B. and L.B.R.: analyzed data.

Funding

Financial support for this work was received from the Children's Mercy Research Institute (L.B.R.) and the Cincinnati Children's Center for Pediatric Genomics (L.B.R. and S.E.V.). J.R.S., is supported by the National Institutes of Health (NIH) Clinical and Translational Science Award (CTSA) program, grant 1UM1TR005265 through the National Center for Advancing Translational Sciences (NCATS). EAP is supported by a Eunice Kennedy Shriver National Institute of Child Health and Human Development training grant (T32HD069038). L.B.R., and J.R.S., are supported by R01HD099775 through the Eunice Kennedy Shriver National Institute of Child Health and Human Development.

Conflicts of Interest

L.B.R., has received research support from BTG Specialty Pharmaceuticals, unrelated to the current work. J.R.S., has received research support from the National Institutes of Health and PCORI and has also received material support from Myriad Genetics. Additionally, he receives royalties from Springer Publishing and Cambridge University Press, honoraria from the Neuroscience Education Institute, and serves as an author for UpToDate. J.R.S., has consulted for MindMed, AbbVie (Cerevel), Alkermes, Collegium, Otsuka, Vistagen, and Genomind. He will serve on the speakers bureau for AbbVie and Collegium, unrelated to the current work. All other authors declared no competing interests for this work.

Supporting information

Table S1: Allele frequency and activity values (AV) current and proposed.

Table S2: Diplotype frequency, current and revised activity score (AS). PM poor metabolizer, IM intermediate metabolizer, NM normal metabolizer, UM ultrarapid metabolizer.