Emerging Treatments in Eating Disorders

Abstract

Eating disorders (EDs), including anorexia nervosa, bulimia nervosa, and binge-eating disorder, constitute a class of common and deadly psychiatric disorders. While numerous studies in humans highlight the important role of neurobiological alterations in the development of ED-related behaviors, the precise neural substrate that mediates this risk is unknown. Historically, pharmacological interventions have played a limited role in the treatment of eating disorders, typically providing symptomatic relief of comorbid psychiatric issues, like depression and anxiety, in support of the standard nutritional and psychological treatments. To date there are no Food and Drug Administration-approved medications or procedures for anorexia nervosa, and only one Food and Drug Administration-approved medication each for bulimia nervosa (fluoxetine) and binge-eating disorder (lisdexamfetamine). While there is little primary interest in drug development for eating disorders, postmarket monitoring of medications and procedures approved for other indications has identified several novel treatment options for patients with eating disorders. In this review, I utilize searches of the PubMed and ClinicalTrials.gov databases to highlight emerging treatments in eating disorders.

Electronic supplementary material

The online version of this article (doi:10.1007/s13311-017-0535-x) contains supplementary material, which is available to authorized users.

Introduction

Eating disorder (ED) research is informed by, but distinct from, 2 mature fields of study: the neurobiology of feeding and psychopathology. In the first circumstance, the neurobiological basis of feeding and body weight homeostasis are well understood in preclinical models, such as rodents. However, rodent models are notoriously difficult to model psychosocial stressors, such as Western ideals of thinness, which promote the development of EDs. In the other circumstance, human clinical studies are well-suited to the study of psychological processes contributing to the development of EDs, but often-times lack the mechanistic insight necessary for the development of therapeutic interventions.

One alternative approach is to study the action of approved medications to gain mechanistic insight into potential therapeutic targets. While no medications are Food and Drug Administration-approved for treatment of anorexia nervosa (AN), fluoxetine is approved for treatment of bulimia nervosa (BN) and lisdexamfetamine is approved for treatment of binge-eating disorder (BED). Preclinical studies have suggested that activation of serotonin-2C receptors on the midbrain dopamine neurons by fluoxetine or the stimulant d-fenfluramine reduces episodes of binge-like eating in mice [1]. As lisdexamfetamine has similar psychopharmacological properties as d-fenfluramine, it does suggest that both fluoxetine and lisdexamfetamine may target the dopaminergic system as a common biological substrate, perhaps through modulation of reward processing. Clinical studies testing this hypothesis have not yet been reported in humans and will be needed to confirm this hypothesis. Alternatively, several studies have found that elevated rates of impulsivity and behavioral disinhibition contribute to binge-eating episodes in patients with BN and BED [2–6]. Psychopharmacologic studies support a role for fluoxetine and lisdexamfetamine in treating binge-eating episodes by reducing impulsivity and behavioral disinhibition in this population [7–11], suggesting another potential pathway to target in the treatment of EDs.

In the current review, I will try to bridge the knowledge gap between basic neuroscience research and clinical studies of disordered eating to highlight novel therapeutic strategies. Even though there are several treatments under evaluation for various of aspects of EDs (e.g., correcting bone loss in AN), this review will focus on weight restoration in AN and decreasing binge eating in BN and BED.

Experimental Procedures

Search Criteria

The PubMed (https://www.ncbi.nlm.nih.gov/pubmed ) and ClinicalTrials.gov (https://clinicaltrials.gov/ ) databases were searched between 1 January 2017 and 8 March 2017. The search terms “anorexia nervosa treatment”, “bulimia nervosa treatment”, and “binge-eating disorder treatment” were utilized for the PubMed search, while “anorexia nervosa”, “bulimia nervosa”, and “binge-eating disorder” were used for ClinicalTrials.gov search. For the PubMed search, “anorexia nervosa treatment” 7460 articles were identified, for “bulimia nervosa treatment” 3360 articles were identified, and for “binge-eating disorder” 1860 articles were identified. Articles for Food and Drugs Administration-approved treatments and behavioral interventions were excluded from consideration. For the ClinicalTrials.gov search, 180 studies were identified for “anorexia nervosa”, 170 studies were identified for “bulimia nervosa”, and 392 studies for “binge-eating disorder” were identified. Studies without an intervention and studies in “unknown” or “terminated” status were excluded from consideration.

Weight Restoration in AN

Neuroendocrine

Starvation alters the levels of numerous circulating hormones. While most of these disturbances are corrected by refeeding, there has been interest in determining if pharmacologically targeting specific endocrine pathways could facilitate recovery from AN [12]. The “hunger” hormone, ghrelin, is one such candidate for the treatment of AN. Ghrelin is an acylated-peptide hormone synthesized in the enteroendocrine cells of the stomach [13], which is released by calorie restriction and stress to stimulate appetite and increase food intake [14, 15].

While ghrelin levels appear to be appropriately elevated in AN [16], several studies have found derangements in the regulation of behavioral responses by ghrelin. Whereas high circulating ghrelin levels are associated with activation of brain-reward regions measured by functional magnetic resonance imaging in a food-cue task, patients with AN displayed a negative relationship between ghrelin levels and brain activation in response to high-calorie foods [17]. Furthermore, underweight patients with AN demonstrated progressively decreased ghrelin levels after exposure to palatable food, whereas plasma ghrelin concentrations significantly increased after exposure to palatable food in healthy controls and weight-restored patients with AN, suggesting that dysregulation of ghrelin release may contribute to reduced motivation to consume palatable food in underweight patients with AN [18].

In a pilot study of female patients with restricting-type AN, intravenous ghrelin administration for 14 days increased calorie intake and decreased gastrointestinal (GI) complaints in 4 of 5 patients [19]. Intranasal administration of growth hormone-releasing peptide 2 (a ghrelin receptor agonist) 3 times daily before meals improved body weight and hypoglycemia in a severely emaciated patient with AN [20]. Finally, RM-131, a novel ghrelin receptor agonist, is currently in clinical trials for treatment of AN [21].

Several other endocrine factors are also under investigation for treatment of AN. Oxytocin is a hypothalamic neuropeptide that regulates food intake and levels of oxytocin have long been known to be low in patients with AN [22, 23]. Polymorphisms in the oxytocin receptor gene and methylation of the oxytocin receptor gene have also been associated with disordered eating in several initial studies [24–26], suggesting a potential link between oxytocin signaling and the development of ED behaviors. While a preliminary study of intranasal oxytocin found that a single dose of oxytocin did not increase short-term calorie intake in patients with AN [27], other studies report that intranasal administration of oxytocin altered attentional bias for fat shape and eating stimuli [28], and also decreased attentional vigilance to disgust in patients with AN [29], suggesting that it might have beneficial effects in other cognitive functions. Additionally, a clinical trial is currently underway to determine if intranasal oxytocin affects smell and food consumption in patients with AN [30]. Finally, androgen deficiency has been associated with hypometabolism of the brain and symptoms of depression, anxiety, and ED thoughts in women with AN [31, 32], and low-dose transdermal testosterone has been shown to improve depressive symptoms, spatial cognition, and lean body mass in initial small studies [33, 34]. Currently, the effect of transdermal testosterone on body mass index in patients with AN is being investigated in a phase II clinical trial [35].

Neuromodulation

Recent advances in the field of neuromodulation now offer the hope of nonpharmacological treatment options for an array of mental illnesses, including EDs and associated diagnoses, such as major depression and anxiety disorders. Transcranial direct current stimulation (tDCS) is a method for directly modulating the excitability of cortical regions using low-intensity electrical current. A pilot study of tDCS of the left dorsolateral prefrontal cortex (PFC) in 7 patients with AN found that the procedure was well tolerated and associated with a modest short-term benefits in 5 of 7 patients on eating scales, with a sixth patient reporting improvement in depressive symptoms without an effect on the eating scales [36]. These findings support a need for further studies on the potential efficacy of tDCS in AN; currently, 2 studies are registered as actively recruiting [37, 38].

Repetitive transcranial magnetic stimulation (rTMS) is another noninvasive approach that allows for stimulation or depression of selected cortical areas. Several initial studies have examined the efficacy of stimulation of the PFC by rTMS in the treatment of AN. A pilot study of 10 individuals with AN found that a single session of rTMS was well tolerated and improved feelings of fullness and anxiety [39]. A study of neuronavigated rTMS in 2 individuals with AN found that 19 to 20 sessions of high-frequency stimulation of the left dorsolateral PFC reduced ED symptomatology [40]. A study of 28 individuals with binge–purge subtype of AN or BN, reported that 20 to 30 sessions of high-frequency stimulation of the dorsomedial PFC resulted in reduced binge/purge behaviors in 16 of 28 participants [41]. A study of 5 patients with treatment-resistant AN found that 20 sessions of high-frequency stimulation of the left dorsolateral PFC resulted in significant improvement in eating pathology at 6-month follow in 3 of 5 patients [42]. Finally, a study of single-session stimulation of the left dorsolateral PFC in 60 patients with AN found a trend (p = 0.056) toward improvement in core ED symptoms [43], supporting the need for follow-up studies of multiple-session stimulation. Two studies are currently registered for the study of rTMS in AN [44, 45].

In addition to tDCS and rTMS, which are noninvasive, deep brain stimulation (DBS) is another neuromodulatory technique currently under investigation in the treatment of AN. DBS allows for direct modulation of the neuronal activity of deep structures in the brain by neurosurgical implantation of electrodes. DBS is an established therapy for the treatment of refractory cases of movement disorders and obsessive–compulsive disorder (OCD). Given the shared neural circuitry of AN and OCD, some have hypothesized that DBS may be effective for the treatment of AN as well [46, 47]. Several studies have provided support for this hypothesis by reporting a potential role for DBS in the treatment of severe, enduring cases of AN. Stimulation of the nucleus accumbens by DBS in 4 patients with severe AN found an improvement in body weight restoration with body weight > 85% at last follow-up in all individuals [48]. A case report of a women with AN who received DBS of the ventral striatum for severe comorbid OCD found improvements in the ED pathology in addition to reduction in OCD symptoms [49]. DBS of the subcallosal cingulate in 6 patients with severe, treatment-resistant AN found that 4 of 6 patients reported improvements in mood and anxiety symptoms and 3 of 6 patients reported improvements in body weight restoration and quality of life [50]. In the largest study to date, an open-label trial of 16 treatment-refractory patients with AN receiving DBS of the subcallosal cingulate found significant improvements in body mass index, depression, anxiety, and affective regulation at 1-year follow-up [51]. Currently, 3 clinical trials are underway to study the role of DBS in the treatment of AN [52–54].

Preclinical Models of Restricted Eating

While the risk of developing AN is highly heritable, the lack of genetically validated rodent models of voluntary food restriction have limited attempts to identify potential medications based on pathophysiology. Our group has recently identified the first 2 rare, highly penetrant genetic variants that increase the risk of developing an EDs [55]. Importantly, mice that are genetically manipulated to replicate the disease-associated human variants display several behavioral phenotypes relevant to EDs, including decreased willingness to work for a high-fat diet after being fasted overnight [56, 57]. These mouse lines with impaired behavioral responses to a negative energy state are the first monogenic models of AN and offer a unique opportunity for screening of novel pharmaceutical compounds. Based upon a bioinformatic analysis of potential genetic pathways affected in these mouse models, we identified rosiglitazone as a potential novel medication to treat symptoms of AN [58]. Of note, rosiglitazone has been associated with body weight gain when used to treat patients with diabetes mellitus and has also been studied as a potential treatment for major depression [59], which is frequently comorbid with AN. Esrra-null mice administered rosiglitazone display increased food intake, body weight, and adipose tissue compared with vehicle-treated littermate mice (data not shown). While further behavioral and neurobiological characterization of these mice is necessary before translational studies are initiated, it does demonstrate the potential benefit of the approach (Tables 1 and 2).

Table 1.

Synopsis of emerging treatments for weight restoration in Anorexia Nervosa

Intervention	Classification	Proposed mechanism of action	Supporting evidence
RM-131	Gut neuroendocine	Appetite stimulating ghrelin receptor agonist	Successful Phase II clinical trials proof-of-concept trial on 22 patients with AN
Oxytocin	Pituitary hormone	Affects social and feeding behaviors	Changes in attention to food/shape stimuli, clinical trial on-going
Testorterone	Sex hormone	Corrects hypometabolism in brain of underweight patients	Increases lean body mass in two RCTs
tDCS	Neuromodulation	Modulates excitation of cortical regions	Improvement in eating and depressive symptoms in 5/7 patients
rTMS	Neuromodulation	Stimulation or depression of cortical regions	Improvement in eating symptoms in 4 case series, trend toward benefit in 1 RCT
DBS	Neuromodulation	Stimulation of subcallosal cingulate	Several case series, one open-label trial of 16 patients
Rosiglitazone	PPARg-agonist	Increases mitochondrial biogenesis in brain	Increases food intake and body weight in one rodent model of restriction
Omega-6 fatty acids	Nutraceutical	Reduction in inflammation	Clinical study on-going
Cannabinoid signaling	Neurotransmission	Stimulate appetite, reduce inflammation	Clinical study on-going
Lactobacillus reuteri	Probiotic	Normalizes bowel flora, improves gut function	Clinical study on-going

Abbrevations: tDCS transcranial direct current stimulation, rTMS repetitive transcranial magnetic stimulation, DBS deep brain stimulation, PPARg peroxisome proliferator-activated receptor gamma, RTC randomized controlled trial

Table 2.

Synopsis of emerging treatments for binge-eating episodes in Bulimia Nervosa and Binge-Eating Disorder

Intervention	Classification	Proposed mechanism of action	Supporting evidence
CCK	Gut neurohormone	Increases satiety	Reduced post-prandial levels in BN, clinical study of CCK receptor agonist on-going
GLP-1	Gut neurohormone	Increases satiety	Reduced post-prandial levels in BN, liraglutide FDA-approved for weight loss, clinical study on-going in BN and BED
PYY	Gut neurohormone	Increases satiety	Reduced post-prandial levels in BN
Topiramate	Antiepileptic	Suppresses appetite	Mono-therapy associated with weight loss in obesity, combination with phenteramine reduced binge-eating in a case series, clinical study on-going
Zonisamide	Antiepileptic	Suppresses appetite	Reduces binge-eating in BED in 1 RTC and in BN in 1 open-label pilot study
Naltrexone, nalmefene	Mu-opioid antagonist	Suppresses appetite	Naltrexone-bupropion combination FDA approved for weight loss, nalmefene improves BED in a case report, clinical studies on-going
tDCS	Neuromodulation	Modulates excitation of cortical regions	Positive effects 1 case series in BED and 1 RCT in BN
rTMS	Neuromodulation	Stimulation or depression of cortical regions	Positive effects in 4 small studies, 1 failed RCT in BN
Lorcaserin	Neurotransmission	Agonist of serotonin 2C receptor	FDA approved for weight loss, reduces binge-episodes in mouse model of binge-eating
Erythromycin	Antibiotic	Prokinetic agent	Clinical study on-going
Memantine	Neurotransmission	Modulates signaling on NMDA receptor	Clinical study on-going
Baclofen	Neurotransmission	GABA-B receptor agonist	Clinical study on-going

Abbreviations: CCK cholecystokinin, GLP-1- glucagon-like peptide 1, PYY polypeptide YY, tDCS transcranial direct current stimulation, rTMS repetitive transcranial magnetic stimulation, RTC randomized controlled trial

Additional Approaches

Genetic variants in the epoxide hydrolase 2 gene have been associated with the risk of developing AN [60]. The epoxide hydrolase 2 protein catalyzes the conversion bioactive epoxides on polyunsaturated fatty acids into corresponding diols and is important for a number of metabolic pathways, including cholesterol synthesis [61], suggesting that omega-6 fatty-acid supplementation may affect the clinical course of AN [62]. Several other diverse approaches are also currently under consideration as well, including cannabinoid receptor signaling [63, 64], probiotics [65], and fish oil [66].

Binge Eating in BN and BED

Neuroendocrine

While not all patients with BN and BED are overweight, efforts to find new treatments for obesity have contributed to the understanding of appetite regulation with important implications for binge eating. One particularly important area of research is in satiety hormones, which are secreted by the GI tract in response to nutrients to help terminate meals. Since impairments in satiety have been associated with overeating in clinical populations, targeting satiety responses may represent a new target for treatment of binge-eating.

While numerous satiety hormones have been identified, most research has focused on cholecystokinin (CCK), glucagon-like peptide 1 (GLP-1), and polypeptide YY (PYY) [67]. CCK is released by L cells of the intestine in response to lipids and proteins, and CCK has numerous postprandial functions, including decreasing appetite, slowing gastric emptying, and increasing gall bladder stimulation [68]. Postprandial levels of CCK have been reported to be blunted in patients with BN [69, 70]. One clinical study is currently investigating the role of novel CCK receptor agonist in BN [71].

GLP-1 is a cleavage product of the glucagon gene and is secreted primarily by the enteroendocrine L cells of the intestines in response to many nutrients, including monosaccharides, peptides and amino acids, and fatty acids [72]. GLP-1 has numerous functions, including slowing of gastric emptying, increasing insulin release, and suppressing appetite. Pharmacologic agonists of GLP-1 receptor signaling have been approved for the treatment of diabetes and, more recently, chronic weight management [73]. Several studies have reported blunted postprandial levels of GLP-1 in patients with BN [74, 75], although it is unclear if these differences directly contribute to the development of disordered eating or are a consequence of it. One study is currently registered to examine the potential benefits of the GLP-1 agonist liraglutide on binge eating in BN and BED [76].

PYY is also synthesized by enteroendocrine L cells in the intestine in response to multiple nutrients and has similar effects on reducing GI track motility, increasing insulin secretion, and suppressing appetite [77]. Several studies have examined the PYY levels in patients with EDs with mixed results. Some studies have found a blunted increase in PYY levels after test meals in patients with BN [78–80], while no difference in PYY levels were observed in female patients with BED [81]. Furthermore, there was a negative correlation between postprandial PYY levels and binge frequency in women [80], although it is again unclear if altered PYY levels contribute to the development of disordered eating behaviors or are a consequence of it. While several clinical trials are investigating the use of intranasal PYY in the treatment of obesity, no trials are currently registered to examine PYY agonists in the treatment of EDs.

Antiepileptic Drugs

Several medications originally developed for the treatment of seizures have been found to affect body weight. Topiramate modulates the activity glutamate and γ-aminobutryic acid receptors and voltage-dependent sodium channels, although the precise mechanism by which it causes weight loss is not well understood [82]. While topiramate alone promotes weight loss in patients with metabolic disorders, such as type 2 diabetes, it was associated with a significantly increased risk of serious adverse events, including liver failure, serious injuries or falls, kidney stones, and viral infection with leukopenia, making it less appealing as a monotherapy for treatment of obesity [83]. More recently an extended release formulation of topiramate and the stimulant phentermine has been approved for the treatment of obesity primarily through appetite suppression [84]. A small study found that the topiramate/phentermine combination reduced binge-eating in 2 patients with BED, supporting a need for larger studies [85]. One study is currently testing topiramate/phenteramine in the treatment of BED and BN (NCT02553824). In addition to topiramate, another antiepileptic drug, zonisamide, has been associated with weight loss in clinical populations [86]. Furthermore, 2 small trials of zonisamide in patients with BED with obesity [87] and with BN [88] found significant reductions in ED behaviors, although the discontinuation rate was very high in both studies. No current clinical trials are registered for zonisamide in BED or BN.

Opioid Antagonists

While inhibition of mu-opioid receptor signaling has long been known to decrease “hedonic” feeding in rodents [89], most clinical trials attempting to use the mu-opioid antagonist naltrexone in the treatment of obesity and binge eating have been disappointing [90–92]. However, in recent years the use of naltrexone as part of combination therapies has been more widely investigated as researchers have come to appreciate redundancies in feeding systems [93]. One such combination, naltrexone with the antidepressant bupropion, has been approved for the treatment of obesity demonstrating a 3% to 7% decrease in body weight primarily through a reduction in calorie intake [94]. As depression is frequently comorbid with BED, the naltrexone–bupropion combination is especially attractive to clinicians hoping to minimize polypharmacy in their patients. Additionally, other formulations of opioid antagonists, including intranasal naloxone and nalmefene, have shown some beneficial effects in BED [95, 96]. Currently, 1 clinical trial is actively investigating the efficacy of naltrexone–bupropion in BED [97] and 1 trial examining intranasal naloxone in BED [98].

Neuromodulation

Similar to AN, several research groups are currently investing nonpharmacological approaches to the treatment of binge eating. A recent study found that a single session of tDCS of the dorsolateral PFC reduced urges to binge in 39 patients with BN [99], while a similarly designed study reported reduction in urges to binge in 30 patients with BED [100]. Several preliminary studies of women with BN suggested that rTMS of the left dorsolateral PFC [101–103] and dorsomedial PFC [104] reduced food cravings, but a randomized double-blinded study of 47 women with BN found no clinical benefit of rTMS to the left dorsolateral PFC [105].

Several registered clinical trials are currently following up on the potential for neuromodulation to treat BED. One group investigating low-frequency (1 Hz) rTMS of the dorsomedial PFC is also studying high frequency (20 Hz) stimulation of the dorsomedial PFC in BN [106]. A separate study is investigating 1-session 10-Hz stimulation of the dorsolateral PFC in BN [107]. Another group is measuring the effect of 10-Hz stimulation of the dorsolateral PFC on binge-eating episodes in patients with BED [108]. Direct current stimulation of the PFC is also currently under investigation in the treatment of BN and BED [109].

Preclinical Models of Binge-Like Eating

At present there are no monogenic models of BED as no specific genomic variants have been linked to the risk of developing BN or BED. However, there are behavioral models that are useful as preclinical models of binge-like eating. One model that has been particularly informative involves repetitive limited access to palatable food, such as high-fat food, that produces binge-like bouts of eating (up to 50% of the daily calorie intake within 2 h) in both rats and mice [110, 111]. One important feature of this model is that it does not require forced calorie restriction or stress to induce the binge-like episodes. This model has identified several potential novel therapies, including DBS of the nucleus accumbens [112, 113], the serotonin 2C receptor agonist lorcaserin [105], and the nociceptin receptor antagonist LY2940094 [114].

Additional Approaches

Erythromycin, used for its nonantibiotic effect to stimulate GI motility in patients, is currently being investigated in patients with BN to determine if it reduces binge/purge frequency [115]. Memantine, which modulates the glutamatergic N-methyl-D-aspartate receptor, is being tested for efficacy in reducing binge/frequency and body dysmorphia in patients with BN [116]. Finally, a clinical trial is testing the potential of the γ-aminobutryic acid-B receptor agonist baclofen to decrease binge frequency in BN and BED [117].

Conclusions

While EDs have historically been under studied relative to their burden on society, new insights from related fields have identified several potential therapeutic interventions. Research from the field of feeding and body weight homeostasis has found several endocrine factors that that can stimulate (e.g., ghrelin) or suppress (e.g., GLP-1, PYY) appetite. These factors are now being tested for the treatment of restricted and binge eating, respectively. Central nervous system drugs originally approved for other indications, such as antiepileptic drugs and opioid antagonists, may reduce binge-eating episodes. Neuromodulation techniques developed for treatment of mood and anxiety disorders show promising results as an adjunctive treatment for AN, BN, and BED. Finally, the explosion in genetic research has led to novel rodent models of ED-like behaviors that allow for rapid preclinical screening of compounds. It is now incumbent upon funding agencies to evaluate rigorously the efficacy of these emerging therapies.

Electronic supplementary material

Below is the link to the electronic supplementary material.