Validation of a therapeutic window for RIS in children with ASD
发布日期:
2026-11-16
浏览次数:
0

Autism spectrum disorder (ASD) is characterised by deficits in social communication and repetitive, stereotyped behaviours [1]. Risperidone is the first-line atypical antipsychotic used to manage comorbid irritability and aggressive behaviour in people with ASD [2]. Guidelines recommend an initial dose of 0.01 mg/kg/day, with a maximum dose of 3.0 mg/day, for children with ASD. However, while risperidone is effective, it stimulates the metabolic system and often causes side effects such as weight gain, elevated prolactin levels and increased blood lipid levels [3]. The therapeutic window for therapeutic drug monitoring (TDM) in adults is 20–60 ng/mL  [4], while the reference ranges for children treated for psychotic and impulsive-aggressive symptoms are 9–33 ng/mL and 8–26 ng/mL, respectively[5, 6] . The SPACe research team established a population pharmacokinetic (popPK) model for risperidone and its active metabolite, 9-OH-risperidone. They correlated the sum of steady-state trough concentrations with weight gain and therapeutic efficacy [7] and calculated a therapeutic window of 3.5–7.0 ng/mL [8]. This study externally validated this model and therapeutic window using independent data.


 Methods and Results

Validation of a therapeutic window for RIS in children with ASD

Methods

Data were obtained from the German TDM-VIGIL study and the TDM routine service database at the University of Würzburg. The data included children with ASD, aged 6–18 years, who had reached a steady state and had complete data. Nonlinear mixed-effects modelling was used to externally validate the SPACe popPK model. Model performance was assessed using a goodness-of-fit plot and a visual prediction check (VPC) [9]; the upper cut-off was defined as an increase in BMI z-score from baseline [10], and the lower cut-off was defined as a Clinical Global Impression (CGI) score of 1–2 [11]. The upper and lower limits of the therapeutic window were determined via ROC curves.

Results

A total of 48 patients from the TDM-VIGIL study and 151 patients from the TDM service were included. Of these patients, data from 46 were suitable for further analysis. Only 10 patients had complete information and were eligible for analysis of the upper and lower limits. Model validation utilised data from 46 patients and 132 ritonavir and 9-OH-ritonavir concentration measurements (median 3.0 and 8.0 ng/mL, respectively), with a median dose of 1.50 mg. Using either GoF plots or VPC, the model performed better at predicting 9-OH-risperidone than risperidone itself: risperidone was often underestimated, while 9-OH-risperidone was sometimes overestimated (see Figure 1). The upper-limit analysis of ten cases with a median dose of 0.75 mg and a total steady-state trough concentration of 6.5 ng/mL showed that the total concentration was higher in the seven cases that experienced significant weight gain. However, this difference was not statistically significant (P = 0.86). The ROC curve yielded a cut-off point of 5.0 ng/mL (AUC = 0.74; Figure 2A). In the lower limit analysis of ten cases, the median dose was 0.63 mg, with a total trough concentration of 7.5 ng/mL. The total concentration was higher in six cases with a therapeutic response, though this difference was not statistically significant (P = 0.24). All but one of the cases were above the SPACe lower limit of 3.5 ng/mL and the ROC curve yielded a cut-off point of 5.0 ng/mL (AUC = 0.67; see Figure 2B). As the upper and lower cut-off points were both 5.0 ng/mL, further evaluation selected the point on the ROC curve furthest from the sensitivity/specificity line as the upper limit (8.0 ng/mL). Thus, the therapeutic window was defined as 5.0–8.0 ng/mL.

Validation of a therapeutic window for RIS in children with ASD

FIGURE 1  Visual predictive check of validation of the pharmacokinetics model stratified into a compartment for risperidone and 9-OHrisperidone. Visual predictive check showing the 50th percentile of the observations as the solid red line and 2.5th and 97.5th percentiles as the 2 dashed red lines and the 95% CI of the model simulated 50th (blue shaded area) and 2.5th and 97.5th percentiles (grey shaded areas).


Validation of a therapeutic window for RIS in children with ASD

FIGURE 2 Boxplots showing risperidone sum trough concentrations in (A) patients with BMI z-scores below and above the cut-off based on follow-up time, and (B) patients with and without treatment effectiveness. The red horizontal lines represent the cut-off values of the therapeutic window from the SPACe study.


Discussion

This study conducted an external validation of the SPACe study’s model and therapeutic window. Using GoF and VPC, it was found that the model predicted 9-OH-risperidone levels more accurately than risperidone levels, generally underestimating risperidone and overestimating 9-OH-risperidone. This is consistent with the study by Karatza et al.—risperidone was underestimated in all models, while the SPACe model effectively predicted 9-OH-risperidone [12]; Given the relatively accurate individual predictions, the model is suitable for therapeutic drug monitoring (TDM). The therapeutic window was ultimately set at 5.0–8.0 ng/mL; since the Youden index for both the upper and lower limits was 5.0 ng/mL, making it difficult to adhere to, the upper limit was adjusted to the second-best point on the ROC curve, suggesting that a lower value should be adopted to limit weight gain. This range is comparable to the previously reported 3.5–7.0 ng/mL; however, due to the extremely small sample size and few low-concentration data points (only one case in the lower limit analysis was below 5.0 ng/mL), it is not appropriate to modify the therapeutic window based on these findings. Instead, this should be viewed as validation of the finding that “the therapeutic window for children with ASD is significantly lower than that for adults and previous pediatric reference ranges” [4-6]. Limitations: The data were not designed for this analysis, requiring assumptions regarding blood sampling time and steady state; with only 10 cases per group, the ROC results must be interpreted with caution; TDM-VIGIL previously adjusted some doses based on the adult range of 20–60 ng/mL [4]. This study provides stronger evidence for the continued application of the SPACe model and therapeutic window; the ongoing SPACe 2: STAR study is currently using it to estimate trough concentrations and provide dosing recommendations based on the 3.5–7.0 ng/mL range [13]; when using TDM for dosing, attention should be paid to the model’s tendency to underestimate risperidone concentrations.


Validation of a therapeutic window for RIS in children with ASD

Validation of a therapeutic window for RIS in children with ASD

Validation of a therapeutic window for RIS in children with ASD


Reference

1. Sharma, S.R., X. Gonda, and F.I. Tarazi, Autism Spectrum Disorder: Classification, diagnosis and therapy. Pharmacol Ther, 2018. 190: p. 91-104.

2. Fung, L.K., et al., Pharmacologic Treatment of Severe Irritability and Problem Behaviors in Autism: A Systematic Review and Meta-analysis. Pediatrics, 2016. 137 Suppl 2: p. S124-35.

3. Persico, A.M., A. Ricciardello, and F. Cucinotta, The psychopharmacology of autism spectrum disorder and Rett syndrome. Handb Clin Neurol, 2019. 165: p. 391-414.

4. Hiemke, C., et al., Consensus Guidelines for Therapeutic Drug Monitoring in Neuropsychopharmacology: Update 2017. Pharmacopsychiatry, 2018. 51(1-02): p. 9-62.

5. Klampfl, K., et al., Serum concentrations, therapeutic response and side effects in children and adolescents with impulsive-aggressive symptoms during risperidone therapy. Pharmacopsychiatry, 2010. 43(2): p. 58-65.

6. Taurines, R., et al., Therapeutic drug monitoring in children and adolescents with schizophrenia and other psychotic disorders using risperidone. J Neural Transm (Vienna), 2022. 129(5-6): p. 689-701.

7. Kloosterboer, S.M., et al., Risperidone plasma concentrations are associated with side effects and effectiveness in children and adolescents with autism spectrum disorder. Br J Clin Pharmacol, 2021. 87(3): p. 1069-1081.

8. Hermans, R.A., et al., Therapeutic Drug Monitoring to Optimize Risperidone Treatment in Children with Autism Spectrum Disorder. Ther Drug Monit, 2024. 46(2): p. 259-264.

9. Bergstrand, M., et al., Prediction-corrected visual predictive checks for diagnosing nonlinear mixed-effects models. AAPS J, 2011. 13(2): p. 143-51.

10. van der Esch, C.C.L., et al., Risk factors and pattern of weight gain in youths using antipsychotic drugs. Eur Child Adolesc Psychiatry, 2021. 30(8): p. 1263-1271.

11. Busner, J. and S.D. Targum, The clinical global impressions scale: applying a research tool in clinical practice. Psychiatry (Edgmont), 2007. 4(7): p. 28-37.

12. Karatza, E., et al., External Evaluation of Risperidone Population Pharmacokinetic Models Using Opportunistic Pediatric Data. Front Pharmacol, 2022. 13: p. 817276.

13. Hermans, R.A., et al., The effect of therapeutic drug monitoring of risperidone and aripiprazole on weight gain in children and adolescents: the SPACe 2: STAR (trial) protocol of an international multicentre randomised controlled trial. BMC Psychiatry, 2022. 22(1): p. 814.

14. Hermans, R.A., et al., External validation of a therapeutic window for risperidone in children with autism spectrum disorder. Br J Clin Pharmacol, 2025. 91(10): p. 3009-3015.


Related Recommendations