Home-based Intervention with Semaglutide Treatment of Neuroleptic-Related Prediabetes (HISTORI): protocol describing a prospective, randomised, placebo controlled and double-blinded multicentre trial

The life expectancy of subjects with schizophrenia spectrum disorders (SZD) (hereafter schizophrenia) is reduced by 10–20 years when compared with the general population.1–3 The excess mortality is partly explained by a 2–3 fold increase in cardiovascular disease (CVD).4 5 A nationwide register study from Sweden including more than 46 000 subjects with schizophrenia showed that CVD was the leading cause of death, causing more deaths than suicide and that death from CVD occurred 10 years earlier in schizophrenic subjects than in the general population.5

Another factor predisposing to the reduced life expectancy is type 2 diabetes (T2D), which develops in 10%–15% of all patients with schizophrenia.6 Compared with the general population, patients with schizophrenia are more likely to have diabetes, hypertension, dyslipidaemia, to be overweight and to smoke.7 Indeed, one-third of patients with schizophrenia have been reported to suffer from the metabolic syndrome (MetS).8

Antipsychotic treatment is vital to reduce schizophrenia-related symptoms9 and treatment with second generation anti-psychotic (SGA) drugs has demonstrated beneficial effects on mortality, suicide rates and hospitalisations.10 However, SGA treatment comes with unfortunate side effects, which typically include increased risks of developing obesity, MetS and pre-diabetes.11 12 Furthermore, SGA treatment per se leads to impaired glucose tolerance, diabetes, hyperlipidaemia and CVD.7 13

Gaining weight is associated with a reduced quality of life (QoL) in people with schizophrenia.14–19 In addition, weight gain and MetS are among the most frequently reported reasons for discontinuation of treatment in this group.20 Indeed, in a cohort of 304 subjects with schizophrenia, 94% reported distress over body weight and those who were obese appeared to be more than twice as likely to be non-adherent compared with those with a body mass index (BMI) below 25 kg/m².21 Therefore, it may be speculated that weight-reducing treatments may lead to an increased adherence to SGA treatment. However, to the best of our knowledge, the association between a glucagon-like peptide 1 (GLP-1) agonist-induced weight reduction, the severity of psychotic symptoms and the ability to adhere to SGA treatment has never been investigated in a randomised clinical trial (RCT).

Lifestyle and environmental factors such as smoking, poor diet and an insufficient capacity to exercise contribute to the risk of CVD.13 Behavioural lifestyle interventions aiming at improving diet and physical activity have indeed shown short-term beneficial effects, but long-term effects diminish over time compared with standard care.22 23 When lifestyle programmes fail to reduce the increased risk of CVD and T2D, metformin is standard treatment according to NICE guidelines (NICE 2012). In a meta-analysis of 19 RCTs (N=1279), the addition of metformin to ongoing antipsychotic treatment for an average of 3–4 months in subjects with severe mental illness reduced body weight but only by 0.61 kg.24 Therefore, alternative weight reducing drugs are required, for example, GLP-1 analogues.

The potential of GLP-1 analogues to improve the metabolic profile of patients with schizophrenia was recently successfully tested. In a randomised controlled trial, including 97 patients (placebo/active comparator: 50/47) with schizophrenia, pre-diabetes and a BMI ≥27 kg/m², 16 weeks of liraglutide (1.8 mg Victoza daily) reduced body weight, waist circumference and abdominal fat mass, LDL cholesterol, glucose levels and insulin resistance, without any worsening in the psychiatric disorder.25 The beneficial effect of GLP-1 analogue treatment in SGA-treated patients was later reconfirmed in a 2018 meta-analysis which, however, only included two additional and smaller trials of Exanatide (Bydureon), and only one of these studies was placebo controlled.26 Nevertheless, the review concluded that GLP-1 analogue treatment is effective and tolerable for SGA-related weight gain, but that more studies are required.27 28

Today, liraglutide has been replaced by semaglutide, which is more efficient as a weight losing agent in obese subjects.29 Furthermore, semaglutide is easier to administer than liraglutide as semaglutide requires weekly and not daily injections. This may be important as it opens for the possibility of administering semaglutide by mobile nurses performing weekly house calls, thereby improving adherence to treatment, which is known to be reduced in patients with schizophrenia.30 Importantly, in non-diabetic obese subjects, semaglutide is able to improve various cardiometabolic risk factors31 and to prevent new cardiovascular events (death from cardiovascular causes as well as non-fatal myocardial infarction and stroke) in subjects with pre-existing CVD.32 Therefore, semaglutide appears as a suitable treatment for obese schizophrenic patients.

The study will recruit patients from community mental health centres and private practitioners in the Region of Southern Denmark and the Region of Zealand, Denmark.

Patients with a clinical diagnosis of either schizophrenia, schizotypal disorder or schizoaffective disorder (ICD-10 diagnoses: F20.x; F21; F25.x), age between 18 and 60 years receiving treatment in the regional mental healthcare of participating centres, or by private practice, and fulfilling criteria as listed in. box 1

Adverse reactions will be evaluated weekly. At every 4-week visit, vital parameters will be recorded, including blood pressure, body weight and waist circumference. The safety parameters in the study include adverse reactions, vital parameters and liver and kidney function, which will be analysed at baseline, week 15 and week 30. If any side effects or changes in biochemical measures are observed, the trial physician or the physician responsible for the test will be informed.

All safety parameters will be recorded in an electronic case report form (eCRF) as soon as they have either been measured or reported by the study participant. The study will be monitored by Good Clinical Practice (GCP) and follow current GCP guidelines. The study protocol has been reviewed by GCP prior to inclusion of first participant, and subsequently, the study is monitored by GCP throughout the study period.

After drug administration, all potential side effects and other adverse events (AEs) observed will be reported in the eCRF. Semaglutide half-life is 1 week and thus expected to be completely eliminated after 6 weeks). Consequently, collection of AEs for all patients will be completed 6 weeks after the final injection by telephone calls. The nature of each AE will be assessed according to the severity and relationship of the experimental medicine or trial procedure and will be assessed by the primary investigator, who is a physician. If it is considered that the AE is related to the test drug or test procedure, it will be noted as a side effect in the eCRF. All severe AEs will be reported to the sponsor. Hospital admissions related to schizophrenia as soon as possible, other hospital admissions always within 24 hours of it coming to our attention. All severe adverse reactions (SARs) which, regardless of dose, result in death, are life-threatening, result in hospitalisation or prolongation of hospital stay in medical or surgical departments, result in significant or persistent disability or incapacity, or lead to congenital anomaly or malformation, will be considered in this trial to be unexpected (SUSAR).

Emergency unblinding will occur if the participants develop AEs that require knowledge of the treatment, if patient develop sign of overdose or if patient develop conditions that require treatment with the trial drug (semaglutide) or drugs with potential interactions with trial drug.

Primary outcomes for the study will be absolute changes in HbA1c. Changes in HbA1c will be evaluated after 15 weeks and after 30 weeks.

Assuming that 30 weeks of semaglutide reduces HbA1c by 0.2%, and that SD equals 0.35%, 65 subjects in each arm are required to obtain a power of 90%, using a two-sided significance level of 5%.

Drop-out from trials concerning pharmacological treatment for obesity is non-linear; the drop-out from the treatment group will occur earlier due to side effects, while in the placebo group dropouts occur later due to lack of efficacy.

All participants randomised in the study will be included in the assessment of endpoints. All data will be analysed based on the intention-to-treat population. Clinical results will be presented as mean, least squares mean, least square mean change from baseline or least square difference between groups with SE, SD or two-sided 95% CI as appropriate. Statistical tests will be conducted as two-sided tests with a 5% significance level.

A mixed effects linear regression analysis will be used to compare changes in glucose tolerance, HbA1c, from baseline to 30 weeks follow-up between the intervention and the placebo groups. Missing values will be imputed, primary analysis will be intention to treat. Supplementary per-protocol analysis will accompany the intention-to-treat analysis.

All changes in secondary outcomes from baseline to various follow-up times are analysed using mixed effects linear regression analyses for continuous outcomes and mixed effects logistic regression analyses for categorical outcomes. All analyses on continuous outcomes are adjusted for the baseline value of the outcome, and in case of significant imbalances in baseline patient characteristics between the two randomised groups, exploratory analyses including such variables are conducted. To correct for multiple testing of the secondary outcomes, Holm’s sequentially rejective multiple test procedure will be applied. Statistical analyses are performed using Stata software V. 16 (Statacorp). The hypothesis tests are two sided, and the 5% level of significance is considered.

The project is approved by the Regional Science Ethical Committee for the Region of Southern Denmark and the Danish Medicine Agency. Data processing in connection with the study is reported to the region’s list of ongoing research projects via the Executive Secretariat at Odense University Hospital (OUH). The study participants will not necessarily be able to benefit from participation in the study. Individual participants may be diagnosed with illness that may require treatment; in that case, they will be referred for relevant treatment. Participants will not have overt diabetes based on HbA1c. Therefore, following national guidelines, participants in this study are not candidates for standard treatment of diabetes and under normal circumstances they would not receive antidiabetic treatment or treatment that would reduce their risk of diabetes. However, metformin treatment of the MetS is accepted.

Amendments to the study protocol will be made to the Regional Science Ethical Committee for the Region of Southern Denmark and the Danish Medicine Agency.

The study will be conducted in accordance with the Declaration of Helsinki with amendments as well as in accordance with the General Data Protection Regulation, and the Danish Data Protection. In the case that biological material will be shipped to collaborators abroad, the Danish Protection Agency will be applied and Chapter V of the Danish Protection Act will be obeyed. It is mandatory that the clinical staff involved in the clinical trial has completed an official course in GCP.

Recruitment opens January 2022 with first patient, first visit, whereas last patient, last visit (LPLV) is scheduled for September 2024. Analysis of primary endpoints is planned to March 2025: Data will be published in EudraCT within 12 months from LPLV. Publication of the final results on primary endpoint is expected to be made public in March 2025.

Metabolic disturbances, obesity and diabetes development in SGA treated patients are a major problem in the clinical setting, but the underlying mechanism is poorly understood, even though there are theories focusing on an interaction of medication, genes, life style, diet and physical activity.8 13

GLP-1 analogues have shown remarkable effect in the treatment of diabetes by contributing to weight loss, lowering blood glucose levels and improving mortality from CVD.31 32 49 This is in contrast to life style interventions with physical activity and diet changes, as such initiatives have failed to induce robust long-term improvements of the metabolic disturbances.22 23 51

The main focus of this clinical trial is mitigation of metabolic aberrations due to SGA treatment and reversal of pre-diabetes. Hence, the primary outcome is the absolute reduction in HbA1c rather than weight loss.

Patients with severe mental illness experience many challenges due to schizophrenic symptoms and side effects to SGA treatment.11 12 Recruitment may be difficult in this patient population, and accordingly, we are collaborating with local psychiatric teams in both regions and with local consultants to create a secure environment. The study is designed to focus on adherence to study protocol and minimise the patient’s burden of study participation. The home visit by study nurses is likely to strengthen protocol adherence by assisting the participants and creating a secure and safe environment, eventually yielding higher reliability of data from interviews and questionnaires.

Semaglutide holds benefits compared with liraglutide when dealing with patients with weight gain caused by SGAs. Semaglutide requires weekly treatment only, which makes it easier to administer with help from a study nurse. In addition, the weight reducing effect of semaglutide is greater than liraglutide without compromising the cardioprotective effect.52 Hence, if semaglutide 1.34 mg/mL has a beneficial effect on weight, HbA1c, MetS and QoL, it could potentially be used in patients with schizophrenia in high risk of CVD to prevent premature death due to diabetes.

Our study will attempt to mitigate the metabolic side effects of SGA induced overweight in individuals with schizophrenia and pre-diabetes. Treatment with semaglutide improves glycaemic control, weight and clinical cardiovascular risk factors by reducing numbers of fatal and non-fatal myocardial infarction and non-fatal stroke within 104 weeks.53 We hypothesise that beneficial preclinical events are detectable by NAF uptake using PET-CT scans after 30 weeks of treatment. If so, this may contribute to a better understanding of the nature behind the increased cardiovascular risk within these patients.

We hope that our trial HISTORI will provide insight in how to benefit the cardiometabolic health and the QoL of schizophrenic patients—adding more and better years—and thereby change the HISTORY of this vulnerable group of patients. Additionally, if successful, HISTORI may pave the way for long-term, large-scale GLP-1 agonist interventions in schizophrenic patients with pre-diabetes, aiming to reduce the number of hardcore clinical cardiovascular endpoints, and thereby prolonging their life expectancy. Finally, a positive outcome may fuel initiatives to study the ability of GLP-1 agonists to prevent the development of the pre-diabetes, for example, by starting treatment shortly after treatment with SGAs are initiated in high-risk patients.