Serum Concentration of Risperidone and Adverse Effects in Children and Adolescents

To the Editor

In December 2015, an article published in this journal by our research group assessed, among other information, the frequency of hyperprolactinemia in a sample of children and adolescents taking risperidone (Dos Santos Júnior et al. 2015). A second article focused on the frequency of overweight, obesity, and metabolic complications among the same participants (Dos Santos-Júnior et al. 2016). Nevertheless, there was no information regarding possible correlations/associations of these parameters with serum concentrations of risperidone, its active metabolite, 9-hydroxyrisperidone, or their combination. From the 120 previous participants, in an authors' further effort to summon them, 67 were reassessed and had these concentrations measured, together with new evaluations of the same previously studied parameters. All of them had, in the figure of themselves or their caregivers, signed an informed consent form approved by the University of Campinas Ethics Committee of Research.

All participants were using oral tablets of short-acting risperidone. Determinations of serum risperidone and 9-hydroxyrisperidone concentrations (high-performance liquid chromatography, with UV detection at a wavelength of 237 nm) were taken from patients using risperidone after 12 hours fasting and before they took the morning dose of the medication, using the method described by Khorana et al. (2011).

The following continuous parameters were evaluated: age; prescribed and weight-adjusted doses of risperidone; risperidone usage time; body mass index (BMI) z-scores; blood pressure z-scores; serum glucose and insulin, for the calculation of the Homeostatic Model Assessment for Insulin Resistance index (HOMA-IR); total cholesterol and its fractions; triglycerides; transaminases; and leptin. Discrete evaluated parameters were sex; diagnoses; use of other medications, including metabolic enzyme inhibitors or inducers of cytochrome P450 2D6 and 3A4 enzymes; and presence or absence of elevated BMI, abdominal obesity, hyperprolactinemia, total and low-density lipoprotein hypercholesterolemia, high-density lipoprotein hypocholesterolemia, hypertension, metabolic syndrome, altered HOMA-IR, and CYP2D6 genetic status (poor metabolizers: those with CT or TT genotypes; or extensive metabolizers: those with CC genotype).

All blood samples for laboratorial metabolic evaluations were collected together with those for the determination of risperidone and 9-hydroxyrisperidone serum concentrations. Blood collection for the determination of the CYP2D6 genetic status of the participants had occurred previously (Dos Santos Júnior et al. 2015; Dos Santos-Júnior et al. 2016). To determine possible correlations of the serum concentrations of the drug, its metabolite, and their combination with the continuous parameters, Spearman's correlation tests were used. Mann–Whitney tests were used to test for associations with the discrete parameters (statistic significance level: 5%).

Fifty-one males (76.1%) and 16 females (23.9%) were evaluated. The mean age was 13.8 ± 3 years old [95% CI: 13.1–14.5 years old; median: 13.8 years old]. Regarding CYP2D6 genetic status, 43 patients (64.2%) were extensive metabolizers and 24 (35.8%) were poor metabolizers. Eight main groups of diagnoses have been identified: (1) 36 individuals (53.7%) with conduct disorders (F91) and mixed disorders of emotions and conduct (F92); (2) 26 individuals (38.8%) with hyperkinetic disorders (F90); (3) 23 individuals (34.3%) with depressive disorders (F32, F33, F34, F38, F39); (4) 16 individuals (23.9%) with emotional disorders with onset specific to childhood (F93) or neurotic, stress-related, and somatoform disorders (F40–F48); (5) 14 individuals (20.9%) with mild (F70) or moderate (F71) mental retardation, with significant impairment of behavior requiring attention or treatment (F7x.1), or with other impairments of behavior (F7x.8); (6) 14 individuals (20.9%) with pervasive developmental disorders (F84); (7) 11 individuals (16.4%) with specific developmental disorders of speech and language (F80) or of scholastic skills (F81); and (8) 6 individuals (9%) with schizophrenia (F20) and schizotypal disorder (F21).

Eighteen patients (26.9%) were also using, besides risperidone, moderate/strong CYP2D6 inhibitors (Bazire 2014): imipramine (n = 7), sertraline in doses higher than 100 mg/day (n = 5), fluoxetine (n = 2), cimetidine (n = 1), both cimetidine and sertraline (n = 1), clomipramine (n = 1), and paroxetine (n = 1). Cimetidine and tricyclics are moderate CYP3A4 inhibitors (Bazire 2014) and 10 patients (14.9%) were also using such drugs. Three patients (4.5%) were using carbamazepine, which is an inducer of both 2D6 and 3A4 enzymes (Bazire 2014).

The mean risperidone usage time was 30.6 ± 30.1 months [95% CI: 23.2–37.9 months; median: 20 months]. The mean concentrations of risperidone, its metabolite, and the combination of both were 6.5 ± 4.9 ng/mL [95% CI: 5.3–7.7 ng/mL; median: 5.1 ng/mL], 13.6 ± 12.2 ng/mL [95% CI: 10.6–16.6 ng/mL; median: 12.8 ng/mL], and 20.1 ± 12.7 ng/mL [95% CI: 17.0–23.2 ng/mL; median: 18.9 ng/mL], respectively. These concentrations neither correlated with the continuous parameters nor associated with the discrete parameters, except when patients were under combined use of inhibitors or inducers of CYP2D6 and 3A4.

The serum concentration of risperidone was higher among those using combination of CYP-2D6 moderate/strong inhibitors (8.3 ± 3.8 ng/mL) than among those who were not using (5.8 ± 5.1 ng/mL—p = 0.005). It was also higher among those in combined use of CYP-3A4 moderate inhibitors (8.3 ± 3.2 ng/mL) than among those who were not (6.1 ± 5.1 ng/mL—p = 0.029). There were no associations between the combined use of CYP-2D6 or 3A4 inhibitors and the concentrations of 9-hydroxyrisperidone.

The serum concentration of risperidone was not associated with the combined use of carbamazepine, but the concentration of 9-hydroxyrisperidone was higher among those in combined use (27.7 ± 9.6 ng/mL) than among those who were not using (13.0 ± 12.0 ng/mL—p = 0.03). This study did not find associations between the serum concentrations of risperidone or its metabolite and different genotypes of the CYP2D6 gene. Clinical pharmacogenetic studies are important to introduce individualized medicine in psychiatric practice (Cabaleiro et al. 2014). Although there is a large amount of published data on CYP2D6 polymorphisms, remarking that the plasma concentrations, efficacy, and adverse effects of risperidone differ widely between genotypes, the clinical implications of associations between polymorphisms and response to treatment and adverse effects remain unclear (Cabaleiro et al. 2014).

Many patients in this study were using risperidone for a long time. Findings could be different if the sample was composed of individuals at the beginning of treatment, with blood collections occurring before achieving the steady serum concentrations of risperidone and 9-hydroxyrisperidone. Data on plasma concentrations of risperidone in adolescents are not consistent (Whitney et al. 2015). Although some authors address no associations/correlations with gender, clinical response, or adverse effects, there are others who report a possible dose-dependent relationship between serum prolactin and risperidone concentrations in adolescence (Whitney et al. 2015). Therapeutic drug monitoring of risperidone may be beneficial in circumstances such as assessing compliance, looking for associations with different pharmacokinetic profiles, or emergence of early side effects. Our findings do not provide enough support for the recommendation of this procedure as a routine for patients using risperidone, but corroborate its indication in any circumstance, in which cytochrome P450 inhibitors or inducers are coprescribed.