Monitoring of Metabolic Adverse Events of Second Generation Antipsychotics in a Naive Pediatric Population (MEMAS)

Objectives The primary objective is to study selected factors that can influence the development of the SGA's metabolic AEs such as the main diagnosis for which the SGA is prescribed, comorbidities, type and dose of AP, metabolic family history (siblings, parents, parents' siblings, grandparents) and the patient's characteristics (age, height, ethnicity, weight, puberty status). We hypothesize that factors such as younger age of exposure, SGA type, higher SGA doses, lower BMI, non-white ethnic status, hospitalization status at baseline and longer treatment duration will be associated with greater weight gain and, potentially, more cardio-metabolic complications. The secondary objective is to evaluate the clinical adherence to CAMESA guidelines for monitoring of SGAs metabolic AEs in current practice. We hypothesize that the monitoring rates will be low.

Trial design The MEMAS study design is a multicenter, prospective, longitudinal observational study with repeated measures of metabolic monitoring up to 24 months of follow-up. Two recruiting centers have been selected. Recruitment started in January 2017 at CHU Sainte-Justine Hospital and in May 2018 at CIUSSS NIM including the Rivière-des-Prairies (HSM RDP) and Albert-Prévost Mental Health Hospitals (HSM AP). Patients have been included for up to 4 weeks after the initiation of SGA treatment (baseline). Patients will be assured a safe follow-up on their pharmacotherapy. Adherence to the proposed follow-up calendar by the CAMESA guidelines will allow for the detection and the early management of potential cardiometabolic AEs of SGAs. Assessments are performed at inclusion and during follow-up for anthropometric measures (AM), blood pressure (BP) and blood tests (BT) at baseline and 1, 2, 3, 6, 9, 12 and 24 months of follow-up. Participation in this study does not lead to any additional risk to current medical practices. Study participation will end when the patient reaches the end of their 24-month follow-up or earlier if the SGA treatment is discontinued. The prescription of SGA (including dose adjustments, end of treatment, switches and comedications) by the treating psychiatrist is clinically naturalistic. However, during this study, if measured parameters reach what is considered a critical value, the psychiatrist will be notified by the nurse or a member of the research team so that the patient can be referred to a specialist in order to have the proper intervention recommended by the CAMESA guidelines (Ho et al., 2011; Raffin et al., 2014).

Measures All anticipated measurements during the follow-up are : Demographic and clinical baseline data, Adherence to treatment, Anthropometric measures (AM), Blood pressure, Blood tests, and The "MEMAS questionnaire on lifestyle habits and stages of puberty".

Blood tests are performed on site for patients included at HSM AP or Sainte-Justine Hospital whenever possible. For patients included at HSM RDP, blood samples are performed in a local community service center (CLSC) or elsewhere, depending on available resources.

Data collection process In each recruitment center, nurses monitor the participant's studied variables according to the CAMESA calendar. Data is collected and coded from participant's medical records by a member of the research team. All data is anonymized and preserved in the participant's research file. All of the collected information will be kept confidential unless authorized by the participant or his caregiver or an exception from the law. The computerized data will be kept on a password protected file and the paper questionnaires will be kept in a locked space. The data collected will be kept for seven years after the end of the study. After this period of time, it will be safely deleted or destroyed by shredding. The parameters collected for the metabolic monitoring of each SGA-treated participant during follow-up were based on the recommendations of CAMESA guidelines as well as on other, more recent, clinical landmark studies (Nielsen et al., 2014; Pringsheim et al., 2011; Raffin et al., 2014; Rubin et al., 2015). This clinical research project does not require any further investigation compared to the standards of best practice.

Statistical analysis

1. Number of subjects required A number of 60 participants per recruitment center was calculated, for a total of 120 participants. The sample size has been estimated based on previous studies, with available data for four SGAs (risperidone, aripiprazole, quetiapine, olanzapine) and treatment duration up to 12 months (Correll et al., 2009; Findling et al., 2010; Marcus et al., 2011; Findling et al., 2013; Arango et al., 2014; Ronsley et al., 2015). For the estimation of the sample size, the BMI-z score was chosen as the primary dependent variable which is a good reflection of the metabolic changes related to weight gain. ANOVA one-way test was used with compared groups being olanzapine (O), risperidone (R), quetiapine (Q) and aripiprazole (A) with an expected distribution of subjects per group of 1:3:2:2 respectively, the power of 0.80 and an alpha of 0.05. The means to be compared were calculated using the BMI-z score at different times ((BMI-z 3 months + BMI-z 6 months + BMI-z 12 months) / 3); these means are as follow: 0.90 (O), 0.68 (R), 0.52 (Q) and 0.32 (A), with the standard deviation (SD) between 0.20 and 0.60. In the absence of an established size effect for the BMI-z score change, we varied the effect size (d) between 0.31 and 0.93, which resulted in an estimated sample size between 24 and 120 subjects. The sample size of our study was calculated as follows (Desu & Raghavarao, 1990): a one-way ANOVA study, sample sizes of 15, 45, 30, and 30 are obtained from the 4 groups whose means are to be compared. The total sample of 120 subjects achieves 80% power to detect differences among the means versus the alternative of equal means using an F test with a 0.05 significance level. The size of the variation in the means is represented by their standard deviation which is 0.19. The common standard deviation within a group is assumed to be 0.60. The effect size is 0.31. Participants will be recruited from the outpatient and inpatient mental health settings of CHU Ste-Justine and CIUSSS NIM. Thus, annually, the estimated number of patients newly treated with an SGA at CHU Sainte-Justine is approximately 30-40 at the Child and Adolescent Mental Health Outpatient Clinic, 5-10 at Gilles de la Tourette Syndrome Outpatient Clinic, 20 at Child and Adolescent Psychiatric Inpatient Unit (6-17 year old) and at CIUSSS NIM, 30-40 at the Child and Adolescent Psychiatric Inpatient Units (6-17 year olds) and 40 at the Outpatient Mental Health Clinics. Thus, considering a sample size of 120 subjects, we believe that the clinical reality will allow a realistic enrollment of approximately 60 subjects at CHU Sainte-Justine and 60 subjects at CIUSSS NIM.
2. Scheduled analyzes Some analyzes will be provided. Mean changes in AM (weight, BMI-z score, waist circumference), fasting glucose, lipids, and blood pressure will be calculated over time for each SGA group. The percentage of patients in each group who meet one of the following conditions will be calculated: an increase of at least 0.5 in BMI-z score or of more than 7% in weight from baseline over time; developing obesity, metabolic syndrome, hyperglycemia, type 2 diabetes or insulin resistance over time according to the previously mentioned criteria (Measures paragraph). Comparison of baseline values between groups will be considered; chi-square test will be used for categorical variables while a Mann-Whitney U test will be used for continuous variables. This method accounts for multiple comparisons. Comparison between SGAs use in mono- (single AP) vs poly-therapy (combination of two or more APs) will be done. Descriptive analysis will be performed for the sociodemographic variables using percentages, means, medians, ratios and frequencies. A separated analysis will be conducted for the subgroup of participants who received a pharmacological treatment in order to treat the SGAs' cardiometabolic AEs. Thus, only the data available before this pharmacological treatment will be included in the main analysis. Incidence of the metabolic complications will be calculated as the proportion of new-onset metabolic complications at each time point divided by the number of patients with available data. A separate analysis will evaluate the impact of the pharmacological treatment introduced to treat the SGA's cardiometabolic AEs; comparisons will be done with the remaining participants.