Pharmacotherapy for Obesity

A Field in Crisis?

Wiebke Fenske; Jennifer Parker; Stephen R Bloom


Expert Rev Endocrinol Metab. 2011;6(4):563-577. 

In This Article

Centrally Acting Agents Affecting the Homeostatic Regulation of Food Intake

A variety of potentially satiety-enhancing drugs have been developed targeting different hypothalamic circuits. Meal-induced hormonal and neuronal signals travel from the GI tract to the area postrema and nucleus tractus solitarius in the brainstem. From here, sensory input is transmitted to other centers (including the amygdale and nucleus accumbens). Dopaminergic, opioid and endocannabinoid signaling assign reward value to meals consumed. Inputs from these pathways appear to be integrated with circulating signals of nutritional state, such as fatty acids and the adipocyte hormone leptin, which are detected in the arcuate nucleus (Figure 1).[8] Leptin is known to stimulate the activity of neurons expressing pro-opiomelanocortin (POMC), while inhibiting neurons expressing neuropeptide Y (NPY).[9,10] POMC neurons in turn stimulate the release of α-melanocyte-stimulating hormone (α-MSH), activating the melanocortin receptor 4 (MC4R), leading to a reduction in food intake and increase in energy expenditure.[11] By contrast, activation of NPY-Y1 and Y5 receptors is known to cause increases in food intake and reduction in energy expenditure.[12] NPY-expressing neurons also release agouti-related peptide (AgRP), an endogenous antagonist of the MC4R.[13]

Many of the neuropeptide receptors expressed centrally are also expressed peripherally and thus actions of agonists or antagonists of these receptors cannot be assumed to induce weight loss by central mechanisms alone.

Melanocortin 4 Receptor Agonists

The melanocortin 4 receptor is known to be crucial to the regulation of appetite. MC4R-null mice are hyperphagic and obese[14] and MC4R deficiency is the most common monogenetic cause of obesity in humans.[15] At least in theory, MC4R agonists could induce weight loss both in MC4R-deficient individuals, who retain some functional MC4R receptors and the obese population more generally. However Merck's MK-0493, an orally active, selective MC4R agonist, caused no significant reduction in energy intake or bodyweight in Phase II clinical trials.[16] The dose of MK-0493 was limited by nausea and vomiting seen at high doses and efficacy was not achievable at doses below the threshold for these adverse effects. Merck also reported increases in blood pressure in animal studies using MK-0493 and the authors concluded that "taken together, our results suggest that an MC4R agonist is not likely to provide clinically meaningful reductions in bodyweight at doses that are well tolerated." Rhythm showed promising preclinical results in obese primates when treated with their peptide MC4R agonist (RM-493) in June 2010 and Rhythm say this peptide will shortly enter Phase I clinical trials. Merck's MK-0493 also seemed promising at the preclinical stage but was only tested in rodents, whereas RM-493 showed efficacy in primates and thus may be more predictive of the results of human trials.

Targeting the Serotonin System

Serotonergic signaling appears able to modulate the activity of NPY/AgRP and POMC-expressing neurons in the arcuate nucleus of the hypothalamus,[17] which have well-characterized roles in the regulation of appetite and energy expenditure. Anti-obesity treatments targeting the serotonin (5-HT) system are typically derived from β-phenethylamine, which has the same basic structure as the neurotransmitters dopamine, noradrenaline and adrenaline and mediate their effects by influencing noradrenergic, dopaminergic and serotonergic neurotransmission. Of the 14 recognized 5-HT receptor subtypes, the 5-HT1B, 5-HT2C and 5-HT6 receptors, which are distributed widely within the CNS, are of particular interest for the modulation of bodyweight.[18] Activation of these receptor subtypes, either directly using receptor agonists, or indirectly by increasing the availability of endogenous 5-HT, leads to reduced food consumption, whereas decreasing 5-HT receptor activation produces the opposite effect.

Serotonin-reuptake Inhibitors

Sibutramine (Meridia® Abbott), was approved by the FDA in 1997 as the first agent blocking the reuptake of noradrenaline and 5-HT on the market.[19,20] Cardiovascular concerns were reported quite early, but it was a further 13 years before a study of sufficient size and duration was completed affirming these concerns. The Sibutramine Cardiovascular Outcomes (SCOUT) trial, which analyzed 10,000 patients with pre-existing cardiovascular disease over a period of 3.4 years, reported a clearly higher cardiovascular risk in patients taking sibutramine compared with placebo (11.4 vs 10%; p = 0.02).[21] Based on these findings, sibutramine has undergone a revision process and was finally withdrawn from the European and US market last year.[205]

A second serotonin-reuptake inhibitor is currently in development. Tesofensine (NS2330) is described in clinical trials as an inhibitor of the reuptake of noradrenaline, dopamine and 5-HT,[22,23] which was initially developed for the treatment of Alzheimer's and Parkinson's diseases. Tesofensine was reported to cause some weight loss in obese patients being treated for both these diseases and hence the drug was subsequently assessed in dose-dependent weight loss studies, where it induced promising weight reductions of 2.1% (0.125 mg dose), 8.2% (0.25 mg), 14.1% (0.5 mg) and 20.9% (1 mg) after 14 weeks of therapy.[23] The authors emphasized that this made tesofensine of comparable efficacy to sibutramine.[23] In a Phase IIb dose-dependent trial over 24 weeks,[24] tesofensine also had beneficial effects on hyperlipidemia and glucose metabolism. However, dose-dependent adverse effects on blood pressure and heart rate were reported, and patients in the 1 mg group displayed increased anger and hostility. In fact, tesofensine appears to show the same selectivity in inhibiting noradrenaline and 5-HT uptake over dopamine uptake as does sibutramine and thus is likely to have a similar side-effect profile.[25,26] Although serious psychiatric adverse reactions were denied by the investigators, it will be interesting to see whether this is substantiated in Phase III trials.

Serotonin Receptor Agonists & Antagonists

Among the wide spectrum of serotonin (5-HT) receptors, three subtypes are of particular interest in the modulation of bodyweight: the 5-HT1B, 5-HT2C and 5-HT6 receptors. Evidence suggests that a 5-HT1B receptor agonist reduces food intake in rodents.[18] However, the 5-HT1B receptor agonists currently used in the treatment of migraine are associated with cardiovascular side effects such as chest pain and myocardial infarction,[27–29] precluding any attempt to further evaluate this target for anti-obesity therapy. Selective 5-HT6 receptor antagonists and partial agonists have demonstrated promising hypophagic effects and efficacious weight loss in rodent studies. The effects of these ligands are apparently behaviorally selective and they are well tolerated.[30] Currently, drugs targeting the 5-HT6 receptors are primarily considered as a potential treatment for Alzheimer's disease. Further characterization of the role of 5-HT6 receptors in energy homeostasis is desirable before they can be seriously considered as targets for anti-obesity agents.

Analogs of the 5-HT2C receptor are of special interest for obesity treatment. Fenfluramine and its more active isomer, dexfenfluramine, were widely prescribed in the 1980 and 1990s, but escape from their effect after a few months was a significant clinical problem and a rare, although usually fatal, complication was pulmonary hypertension. They induce weight loss via the formation of an active metabolite, D-norfenfluramine, a potent 5-HT2C agonist.[31] Consistent with rodent studies, the anorexigenic activity of dexfenfluramine in humans can be attenuated by the nonselective 5-HT2C receptor antagonist ritanserin.[32] Consequently, the combination of fenfluramine and phentermine was widely prescribed in the mid-1990s as a very effective appetite suppressant, leading to approximately 10% weight loss.[33] Phentermines are amphetamine-like compounds that act by releasing noradrenaline from presynaptic vesicles in the lateral hypothalamus. The combination was withdrawn from the market in 1997 after dexfenfluramine was shown to be associated with cardiac valvulopathy[34,35] and pulmonary hypertension.[36] Both side effects appear to be the result of the concomitant 5-HT2B receptor activation[31,37] and thus selective 5-HT2C agonists may be efficacious without the adverse effects of 5-HT2B agonism. The development of compounds with the required selectivity between 5-HT2C and other receptor subtypes, however, turned out to be difficult.

The first drug designed as a selective 5-HT2C receptor agonist with a functional selectivity of about 15–100-times over that for 5-HT2A and 5-HT2B receptors, was lorcaserin (APD-356). In a recent Phase III clinical study[38] 3182 obese patients were randomly assigned to receive 10 mg lorcaserin or placebo twice daily over 1 year. An average weight loss of 5.8 kg was reported in the lorcaserin group compared with 2.2 kg in the placebo group. After 2 years, patients who switched to placebo for the second year gained back the lost weight, whereas subjects who continued with lorcaserin regained some weight, leaving a placebo-subtracted weight loss of only 2 kg after 2 years. Overall, 50% of patients in the lorcaserin group failed to lose the required 5% bodyweight, and some of them lost nothing or even gained weight. Although lorcaserin was found not to cause any major mood-related adverse effects or cardiac valvulopathy in this 2-year study, its relative lack of efficacy compared with orlistat makes it unlikely to be a useful treatment for the majority of obese patients. Owing to this relative lack of efficacy and some remaining concerns about safety, the FDA expert panel voted against approval in October 2010, and the developers have been asked to provide additional evidence about the safety of lorcaserin before approval will be considered again.[206]

NPY Receptor Ligands

The neuropeptide NPY is thought to stimulate appetite by activation of the Y1 and Y5 receptors in the hypothalamus. As a result, antagonists at these receptors have been produced with the aim of blocking this orexigenic effect. Merck developed an orally available Y5 receptor antagonist MK-0557; however, in a 52-week multicenter, randomized double-blind placebo-controlled human trial, only a 1.6 kg placebo-subtracted weight loss was seen in the Y5 antagonist-treated group.[39] Numerous Y1 receptor antagonists have been developed, including BMS-193885 (Bristol-Myers Squibb), which showed promising effects on bodyweight in rats[40] but has not been tested in humans. The Y2 receptor is expressed presynaptically and activation of this receptor is thought to inhibit the release of NPY,[41] thus reducing food intake. PYY3–36 is an anorectic hormone, released from the gut postprandially in proportion to the calories consumed.[42] PYY is thought to mediate its effect by agonism of Y2 receptors in the arcuate nucleus of the hypothalamus[42] and also on the vagus nerve.[43,44] Obese individuals are characterized by lower levels of PYY3–36 and its anorexigenic effect appears to be preserved in obese patients, making PYY3–36 an attractive therapeutic target.[45] Several studies in humans have shown that intravenous infusions of a single dose[46] and graded infusions of PYY3–36[47] reduce appetite and food consumption by more than 30% in lean and obese subjects. However in a Phase II clinical trial, intranasally delivered PYY3–36, which produces a poorly sustained rapid spike in blood levels, proved to be ineffective at inducing weight loss at a low dose, and problems with nausea and vomiting were encountered at high doses.[48] Nausea is a frequent side effect of all appetite inhibitors, perhaps because it is merely a physiological extreme form of appetite inhibition, an inseperable high-dose effect. Nausea is a known side effect of pharmacological doses of PYY3–36[49] and thus the potential of Y2 agonists as treatments for obesity may be limited by this adverse effect. Nonetheless, Y2 agonists continue to be developed, and one peptide agonist RG7089 (Roche) reportedly entered Phase I clinical trials in 2009; however, the results have not yet been reported and this peptide is not currently listed in Roche's development pipeline.[207] In addition, Roche have reported preclinical studies showing PEGylated PYY3–36 (RO5095932) had beneficial effects on glucose homeostasis and weight loss in leptin receptor-deficient mice and this drug is currently in Phase I clinical trials in combination with metformin ( trial number NCT01017302).[208] If PYY3–36 acts in the CNS, it is unclear if pegylation limits access.

Neuropeptide Y and the NPY receptors are widely expressed in the brain and in the periphery[50] and have been implicated in roles as diverse as learning and memory[51] and excitation–contraction coupling in cardiomyocytes;[52] thus, any drugs targeting this system could prove to have unwanted side effects.

Glucagon-like Peptide-1 Receptor Agonists

Centrally, the glucagon-like peptide (GLP)-1 receptor is expressed in the paraventricular, arcuate and dorsomedial nuclei in the hypothalamus and the nucleus tractus solitarius, area postrema and parabrachial nucleus in the brainstem.[53] GLP-1 is an agonist at this receptor and is released postprandially from intestinal L cells, as well as from preproglucagon neurons in the nucleus tractus solitarius, and has been shown to inhibit food intake when administered peripherally or centrally.[54–56] GLP-1 also acts as an incretin, increasing glucose-stimulated insulin release and inhibits glucagon release and gastric emptying. Chronic subcutaneous infusion of GLP-1 to patients with Type 2 diabetes mellitus can induce weight loss and improved glucose homeostasis,[57] making the GLP-1 receptor an attractive target for anti-obesity agents. As GLP-1 itself is rapidly cleared from the circulation, analogs of this hormone have been developed that are resistant to dipeptidyl peptidase-IV, the primary enzyme responsible for GLP-1 degredation. The first available mimetics were exenatide (Byetta®; Lilly)[58–61] and liraglutide (Victoza®; Novo Nordisk), which caused improved glycemic control and dose-dependent weight loss in Phase III clinical trials in patients with Type 2 diabetes.[62–64] These agents have been approved for the treatment of Type 2 diabetes but are not currently licensed to treat obesity itself.

A once-weekly suspension of exenatide as a poly-lactide-glycolide microsphere suspension with 3% peptide content has recently been developed, aiming at sustained glycemic control along with a more comfortable standard once-weekly dosing.[65] In the first 30-week randomized trial, long-acting exenatide led to slightly greater improvements in fasting blood glucose and HbA1C than did exenatide given twice daily.[65] Similar benefits with long-acting compared with twice-daily exenatide were shown on bodyweight (3.7 ± 0.5 kg weight loss in 30 weeks [65]and 2.6 ± 0.2 kg weight loss in 26 weeks[66]), blood pressure and triglycerides.[65,67,68]

Another GLP-1 receptor agonist, taspoglutide (Roche), was also designed for once weekly administration. Despite its initially promising results in preclinical and clinical trials,[69,70] Roche discontinued Phase III trials in September 2010 due to diverse gastrointestinal events and hypersensitivity issues. It is not yet known whether these are indicative of the entire class of GLP-1 receptor agonists. Whether long-acting exenatide may succeed twice-daily exenatide as a treatment for type II diabetes will depend on the results of specifically designed trials for accrual of valid safety data.

The most common side effects observed with all GLP-1 mimetics are nausea and vomiting at high doses.[71,72] Thyroid C-cell abnormalities have been reported in preclinical studies for exenatide and liraglutide. Exenatide showed an increased incidence of benign C-cell adenomas, while liraglutide was associated with an increase in C-cell carcinomas in rats and female mice administered the highest liraglutide dose tested.[73] These findings could not be verified in nonhuman primates at an exposure up to 60-fold that of the maximum recommended human dose.[74]

In order to assess the potential role of liraglutide on C-cell function in humans, calcitonin levels were monitored among patients enrolled in liraglutide clinical trials. These data showed no differences in calcitonin levels between liraglutide-treated patients and control groups.[75] The discrepancy between rodent and human data may be due to a species differences in the sensitivity of rodent C cells versus primate C cells to activation by GLP-1 agonists or due to higher dose levels used in the preclinical studies. In view of the rodent data, liraglutide is contraindicated in patients with a history of medullary thyroid carcinoma or multiple neoplasia Type 2.[76]

Oxyntomodulin is also an endogenous GLP-1 receptor agonist. Like GLP-1, oxyntomodulin reduced food intake in rats when administered centrally or peripherally but, unlike GLP-1, oxyntomodulin also increases energy expenditure.[77–80] This is thought to be because oxyntomodulin is also a glucagon receptor agonist. In humans, intravenous infusion of oxyntomodulin reduces food intake,[81] while repeated subcutaneous injection increases energy expenditure and causes weight loss in obese volunteers.[82,83] A peptide analog of oxyntomodulin (PF-05212389) was developed by Thiakis and bought by Wyeth/Pfizer in 2008 who are putting this agent through Phase I trials. A pegylatedoxyntomodulin analog has also been produced, which has caused weight loss and improved glucose tolerance in preclinical studies.[84] As GLP-1 receptor agonists, it is possible that oxyntomodulin analogs will prove to have the same propensity to induce nausea and vomiting as the currently available GLP-1 receptor agonists.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.