Glucagon-like peptide-1 analogs against antipsychotic-induced weight gain: potential physiological benefits

Current interventions against antipsychotic-induced weight gain fall into three categories. The first category involves changing antipsychotic medication to a compound less prone to result in weight gain. This strategy is recommended in clinical guidelines and results in a modest slowing of antipsychotic-induced weight gain to approximately 1.9 kg [11]. Change of medication, however, may increase the risk of exacerbating psychotic symptoms [11]. The second category involves lifestyle intervention and/or cognitive therapy and appears to slowdown the SGA-induced weight gain by 2.6 kg after 3 to 4 months; 4.2 kg after 6 months; and 3.1 kg after 12 to 18 months of treatment [25]. Finally, adjuvant medical treatment, the third category, has been attempted with numerous drugs, including ephedrine, sibutramine, orlistat, nizatidine, cimetidine, naltrexone, amantadine, reboxetine, fluoxetine, topiramate, dextroamphetamine, d-fenfluramine, famotidine, fluvoxamine, phenylpropanolamine, rosiglitazone, metformin and the combination of metformin and sibutramine [26]. Metformin is by far the most investigated of these drugs and is associated with the most pronounced deceleration of antipsychotic-induced weight gain: 2.9 kg over 13 weeks compared with placebo [26]. Metformin, however, is considered a weight neutral intervention in the treatment of type 2 diabetes [27].

Notably, all current interventions are weight neutral at best. This is the case because a weight reducing intervention may mask a net weight gain (for example, a reported weight loss of 4 kg may be based on a 2 kg weight gain in the intervention group compared with a 6 kg weight gain in the placebo group [28]). Absolute, placebo-controlled weight loss has only been achieved by means of resource-demanding interventions comprising strict diets and intensive physical training among hospitalized patients [29] or during short treatment periods (≤12 weeks) [30].

In the general population, treatment of obesity often comprises a combination of diet, exercise, modification of daily routines and medication [31]. Currently, orlistat (a lipase inhibitor), sibutramine (a serotonin-norepinephrine reuptake inhibitor) and sympathomimetics are the only approved drugs with the indication obesity [32]. Bariatric surgery (for example, gastric bypass surgery) is the single most efficient way to achieve lasting weight loss. However, this intervention is not without risk and, therefore, only indicated in selected patients (for example, morbid obesity (BMI ≥40 kg/m²)) [33].

Thus, there is an exigent need for better treatment strategies as well as a deeper understanding of the physiological consequences of the weight gain and the metabolic side effects associated with antipsychotic drugs. One such strategy may involve the effects of currently available GLP-1 analogs.

GLP-1 is a gut peptide hormone synthesized in endocrine L cells in the intestinal mucosa. GLP-1 is secreted into circulation after food intake. In the pancreas, GLP-1 stimulates glucose-induced insulin secretion (an incretin hormone) and inhibits glucagon secretion, thereby substantially contributing to maintaining the glucose homeostasis [34]. Activation of both peripheral GLP-1 receptors and GLP-1 receptors in the central nervous system reduces appetite and food intake thereby ensuring that body weight is kept down. In accordance with these observations, GLP-1 analogs for the treatment of type 2 diabetes have been developed (generally lowering HbA1c approximately 1%). A recent meta-analysis of studies of obese patients (with and without type 2 diabetes) demonstrated that ≥20 weeks of treatment with a GLP-1 analog (compared to other antidiabetic drugs, including metformin) induced a weight loss of 3 kg [35]. Since previous antidiabetic treatment strategies are generally associated with weight gain, the weight reducing effect of GLP-1 analogs may be of vital importance in the treatment of type 2 diabetes.

Clinically, the degree of weight loss appears to be positively correlated with the dose of GLP-1 analog. A recent two-year prospective study (including non-diabetic patients with baseline BMI ≥30 randomized to treatment with the GLP-1 analog liraglutide (2.4 to 3.0 mg once-daily) or placebo) reported a weight loss of 7.8 kg (compared to a 2 kg weight loss in the placebo group) [36]. Moreover, preclinical studies have indicated that GLP-1 analog treatment can reduce lipid accumulation in the liver [37, 38].

Interestingly, a preclinical study has shown that olanzapine-induced weight gain and metabolic derangements in rats are reversed after treatment with the GLP-1 analog liraglutide [39].

At present, six different GLP-1 analogs have been subject to trials: exenatide (Eli Lilly), liraglutide (Novo Nordisk), albiglutide (GlaxoSmithKline), taspoglutide (Ipsen and Roche), lixisenatide (Sanofi-Aventis) and LY2189265 (Eli Lilly). Only exenatide and liraglutide have been approved by the U.S. Food and the Drug Administration and European Medicines Agency. GLP-1 analogs have the indication type 2 diabetes in combination with metformin and/or sulphonylurea, when treatment with these drugs is insufficient. Much academic and commercial effort is being put into investigating the possibility of extending the indication to obesity. Both exenatide and liraglutide are administered subcutaneously. Exenatide exists in two formulations: twice daily (Byetta®) and once weekly (Bydureon®, Eli Lilly and Company, Indianapolis, Indiana, USA). Liraglutide (Victoza®, Novo Nordisk A/S, Bagsvaerd, Denmark) is administered once daily.

The major side effects of exenatide and liraglutide are mild to moderate nausea and vomiting. These side effects are dose-dependent and decline over time and they do not explain the observed weight loss associated with GLP-1 analog treatment [40]. The incidence of treatment-associated hypoglycemia is low (European Medicines Agency).

Besides the effects of GLP-1 analogs on hyperglycemia and obesity (and possibly on hepatic lipid deposition), activation of central GLP-1 receptors may stimulate neural plasticity and prevent apoptosis, thereby suggesting involvement of GLP-1 in neuroprotection and learning [41, 42]. An animal model of Alzheimer's disease has been used to show that the GLP-1 analog liraglutide prevents memory impairment and increases both neurogenesis as well as synaptic plasticity in the hippocampus [43, 44]. In models of Parkinson's disease [45] and Huntington's chorea [46], dopaminergic neurons were protected and motor deficits improved when exposed to the GLP-1 analog exenatide.

It is well-established that schizophrenia is characterized by structural brain changes [47]. Already at first presentation of symptoms, reductions in total brain volume and hippocampus volume, along with increased ventricle volume, can be observed [48, 49]. Moreover, the volumes of the dopamine-sensitive striatal regions appear to be reduced before initiation of antipsychotic treatment [50]. Progressive cerebral volume loss, for example, enlarged ventricles, is associated with more negative symptoms, cognitive decline and a more severe course of illness [23]. Current antipsychotic drugs have a limited effect on these symptom dimensions, and interventions with convincing neurotrophic properties are still lacking in the treatment of schizophrenia [51].

Growing evidence suggests that overweight is also associated with structural brain changes and cognitive deficits [24]. Large cross-sectional structural magnetic resonance imaging studies of obese patients have shown both global gray matter volume reductions (independently of age) [52] and widespread regional structural abnormalities in areas implicated in cognitive functions, impulse control and reward processing (for example, prefrontal cortex and striatum) [53, 54]. These observations are supported by magnetic resonance spectroscopy and single-photon emission computed tomography studies suggesting that elevated BMI is associated with both neuronal abnormalities in frontal brain regions [55, 56] and impaired cognitive function [57]. In schizophrenia, reduced prefrontal blood perfusion also appears to be associated with more negative symptoms [58]. Besides similarities in gray matter changes in schizophrenia and obesity, also coinciding white matter abnormalities in corpus callosum and fronto-temporal tracts have been reported [59, 60].

Although brain abnormalities associated with obesity are already present in adolescents [61], it is still unclear if structural brain abnormalities precede the disorder as is the case in schizophrenia [62]. To our knowledge, weight loss has not been associated with changes in brain volume. In the available study addressing this issue, only global brain volume was assessed [63], thereby rendering the possibility that weight loss may still have a regional effect on brain volume. Nevertheless, some of the cognitive deficits associated with elevated BMI (that is, memory and attention/executive functioning) may be reversed after intentional weight loss [64].

Functional neuroimaging studies have further established the indications of a shared substrate for schizophrenia and obesity. For example, satiety appears to be controlled by the prefrontal cortex, which sends inhibitory inputs to the limbic and paralimbic regions, thereby suppressing hunger [65]. Studies using positron emission tomography have compared obese, previously obese and lean subjects and found abnormal cerebral blood flow in the insula and hippocampus in both obese and previously obese, but not in lean subjects [65]. This suggests that in obesity, disturbances in the satiety network are 'trait' rather than 'state' dependent. Interestingly, abnormal hippocampal-prefrontal connectivity appears to be present in first-episode schizophrenia, but also in subjects at high risk for developing schizophrenia [66].

More recently it has also been suggested that the regional disturbances in the responsivity of the reward system (for example, in the ventral striatum) may explain excessive food intake, thereby increasing the risk of developing overweight [67]. Of note, reward disturbances are also becoming increasingly recognized in schizophrenia patients as they even appear to precede the initiation of antipsychotic treatment [68].

First, encouraging and growing evidence supports that a sizable and enduring weight loss can be obtained by GLP-1 analog treatment in both diabetic and non-diabetic overweight or obese patients. Concurrently, an improved glycemic control is obtained. Second, a striking overlap appears to exist between key pathophysiological findings in schizophrenia and overweight/obesity. The extent to which this overlap is driven by the weight gain induced by antipsychotic medication or is caused by interactions from shared pathophysiological mechanisms warrants clarification. A shared substrate between schizophrenia and overweight will likely involve shared genetic risk factors [13]. With changes in body weight, physiological mechanisms may also come into play [64, 69]. As such, exploration of the cerebral effects of GLP-1 analogs in overweight/obese subjects as well as in schizophrenia patients with antipsychotic induced weight gain will help to elucidate this proposed overlap between the two disorders.

An apparent way forward is initiation of clinical studies addressing whether GLP-1 analog treatment can reduce body weight in patients with antipsychotic-induced weight gain. In fact, there is currently a study investigating whether the GLP-1 analog exenatide can induce weight loss in obese patients treated with olanzapine [70]. Since previously un-medicated schizophrenia patients are in particularly high risk of developing metabolic side effects [71], addressing whether or not GLP-1 analog treatment can prevent weight gain is also pertinent. Finally, the indications of the procognitive and neurotrophic effects of GLP-1 call for further investigation. If the progressive brain loss and cognitive decline associated with schizophrenia and overweight/obesity can indeed be ameliorated by GLP-1 treatment, this will have major implications for future treatment of schizophrenia, but also for the treatment of obesity. In psychiatry, particular priority should be given to the effects of early adjunctive GLP-1 treatment, that is, in first-episode schizophrenia patients.

Current data certainly bear the connotation of adjunctive treatment with GLP-1 analog as a potential 'magic bullet' for addressing hitherto unresolved challenges in schizophrenia. Table 1 summarizes the hypothetical implications of GLP-1 treatment in antipsychotic-induced weight gain.