In the 1970s, the drugs available for the treatment of obesity were either bulking agents, such as methylcellulose, or sympathomimetic agents. These are chemically related to amphetamine, although one of them, fenfluramine, had quite different pharmacological properties, being a serotonin (5HT) uptake inhibitor, whereas the other compounds acted on the noradrenergic system (1,2). None of the sympathomimetic agents are available in the UK now as a result either of potential for abuse or, in the case of fenfluramine, the finding that one isomer, dexfenfluramine, induced heart valve defects (3). Phentermine is still available in the USA.
These drugs were prescribed for short-term use only and analysis of many trials of such drugs by the US Food and Drug Administration (FDA) showed that they give a weight loss of approximately 0.5 lb (0.22 kg) more each week than placebo (4). It was thought that they produced their anti-obesity effect through a reduction in food intake, although there is little direct evidence that this was their only effect and an increase in metabolic rate also seems likely. Duration of use was restricted, based on the belief that once a new plateau body weight was reached, it showed that the drugs were no longer working. In fact, removal of the therapy usually resulted in the return to the original overweight body mass, indicating that the drugs were successfully maintaining a lower body mass.
In the 1970s and 1980s, there was little drug research for new anti-obesity drugs, largely because most companies did not perceive obesity as a major therapeutic opportunity and because of registration difficulties and a small number of patients. It was also thought that obesity was simply the result of gluttony and sloth. Exceptions were the Beecham Group (now GlaxoSmithKline) and ICI (now AstraZeneca), both of which had active research programmes aimed at producing exercise-mimetic drugs that would increase energy expenditure (5,6).
Between the 1970s and today, few drugs have been marketed. As already indicated, dexfenfluramine was withdrawn as a result of heart valve defects (3) and some cases of pulmonary hypertension. Topiramate, originally marketed as an anti-epileptic agent, was suspended while in phase III clinical trials because of its many side effects. Currently marketed agents are sibutramine (Reductil), a 5HT and noradrenaline re-uptake inhibitor, which acts centrally to reduce appetite but may have some thermogenic activity (7), and orlistat (Xenical), which is an inhibitor of pancreatic lipase and reduces fat absorption from the gut (8). Both these agents have some problems causing, respectively, slightly raised blood pressure and fatty stools with the possibility of anal leakage. Rimonabant, a cannabinoid CB-1 receptor antagonist (9), has recently (July 2006) been approved in Europe and is under review by the FDA.
Several drugs have entered phase III trials. Pramlintide – a peptide hormone amylin analogue marketed for treatment of type 1 and type 2 diabetes – has some weight loss effects (10), as does exendin – a hormone originally found in lizards and used in type 2 diabetes (11). Sertraline is currently marketed as an antidepressant but is being used in phase III trials in a higher dose as an anti-obesity drug (12). Phase III trials of leptin, a hormone only discovered a decade ago, showed low efficacy and side effects included the formation of antibodies (13). Ciliary neurotrophic factor (axokine) acts through the leptin signalling mechanism but in phase III trials was found to result in poor weight loss (14).
How were potential anti-obesity drugs discovered?
Every drug discovery scientist would like to design a drug that targeted a newly discovered protein that had been shown to be involved in energy balance (food intake/energy expenditure) or at least a protein that had not previously been shown to be involved in energy balance. However, currently marketed drugs and those in phase III trials have largely been discovered serendipitously. Dexfenfluramine is an active enantiomer of fenfluramine. Weight loss was a side effect of sibutramine, originally devised as a treatment for depression. Weight loss was also a side effect of the anti-epileptic drug topiramate and of ciliary neurotrophic factor, which was first used in the treatment of motor neurone disease. Orlistat was originally designed as a lipid-lowering agent and rimonabant was intended for central nervous system (CNS) indication. Leptin was discovered by the identification of a spontaneous mutation in a mouse model of obesity.
How effective are current treatments?
The FDA guidelines for registration of an anti-obesity therapy stipulate a weight loss of 5% more than placebo after 1 year in a significant number of subjects. In Europe, the guidelines require a 10% weight loss from baseline (i.e. including effects of diet, exercise, etc.).
Additional expectations are no weight regain after peak weight loss, metabolic benefits on parameters such as glucose and lipids, weight loss to be predominantly fat loss, particularly visceral fat as this carries a higher cardiovascular risk, and safety, especially no significant cardiovascular or CNS side effects.
Results in clinical trials are generally better than those obtained in the out-patient setting. Compliance is a potential issue both in taking drugs (particularly orlistat because of anal leakage if fat intake is excessive) and in the necessary maintenance of dietary modification and exercise. The current drugs in the best situations achieve the theoretically reasonable rate of weight loss that can be achieved by reducing food intake (see Table 1).
Table 1. Predicted rates of weekly fat and weight loss for anorectic and thermogenic drugs*
Personal communication from Jon Arch, University of Buckingham.
Preservation of protein as shown with a thermogenic drug (15).
Energy deficit (kcal week−1)
Fat balance (g)
Lean tissue balance (g)
Body-weight change (g)
Table 1 also shows that only reducing food intake results in a reduction of lean tissue as well as fat tissue. As lean tissue has a greater contribution to basal metabolic rate than fat tissue, over time there will be a reduction in basal metabolic rate, so even if the reduction in food intake is maintained, a new energy balance will be achieved at the maximum weight loss without a further reduction in food intake. Maximum weight loss in clinical studies occurs within 6 months.
In contrast, increasing metabolic rate by exercise or thermogenic drugs (drugs to stimulate the safe increase in energy expenditure) tends to preserve lean tissue and may increase it. This results in all the weight loss being adipose tissue, and fat loss can occur without a change in body mass.
Regulators in the USA ask for 5% weight loss over a year relative to placebo. For a 100 kg subject, this is 5 kg. If one assumes a maintenance energy requirement of 2500 kcal d−1, the time to achieve this weight loss with a thermogenic agent that increases energy expenditure by 10% is 26 weeks, and with an anorectic drug it is 22 weeks. Experience with anorectic agents is about 22 weeks. Thus, patient expectations on rate of weight loss need to be managed.
The most likely effective treatment for obesity, whether via the help of drugs or not, is a combination of dietary restriction and increased exercise. This should produce the greatest maximum weight loss, will preserve muscle mass, including heart mass, and ensure that most of the weight loss is adipose tissue, which in turn is likely to have the best effect on metabolic parameters.
Origins of current and future obesity epidemic and influence on drug discovery
Understanding why the obesity epidemic is occurring potentially helps in the identification of new treatments by providing new molecular targets. There has been a huge hunt for obesity associated genes and to identify single nucleotide polymorphisms (SNPs). However, the genetic make-up of the world has not suddenly changed and hence if there is a genetic component then it is likely that it is the interaction between one or more variant genes with environmental factors such as high calorie-dense foods, high-fat foods, fructose syrup containing drinks and low levels of physical activity.
Current status of anti-obesity drug research – many targets, few drugs
There is an intense search within the pharmaceutical industry for new anti-obesity drugs. All the major companies are involved. Obesity is also a focus of the biotechnology sector and this sector is acting as a feeder industry in attempting to supply pharmaceutical companies with new anti-obesity pharmaceuticals. The opportunities for new therapies include better efficacy – that is, more than a 5 kg weight loss in the long term – or products with additional metabolic benefits, such as blood glucose control, reduction of blood triglycerides and total cholesterol beyond expectation from weight loss, or less lean tissue lost than with diet alone (typically 15% of weight loss). In addition, drugs that are safer or have less adverse effects than those currently available or which are selective for loss of central, abdominal, fat will also find a market.
The sources of new targets for anti-obesity drugs are many, including gene knockout mice showing obese or lean phenotype, differential gene expression between lean and obese individuals, or differential response to Western high-fat diets. Human and animal genetics, particularly where medical history is available (e.g. Decode in Iceland), are also likely to be useful. Large scale random mutagenesis studies are being conducted not only in mice but also in the fruit fly Drosophila and the flatworm C. elegans. Orphan receptors, i.e. receptors without a known function, matched to ligands using reverse pharmacology, are a source, as could be a better understanding of the signal transduction/neuronal pathways involved in energy balance.
These multiple approaches have yielded a huge number of molecular targets. For example, well over 1000 knockout or transgenic mouse models have been associated with an effect in energy balance resulting in either obesity, leanness, or resistance to developing obesity as a result of feeding diets high in fat (16). The interaction of the tonic homeostatic (adipose tissue) regulation of energy balance with episodic (gut), hormonal and hedonic (brain) signals gives rise to potential drug targets as illustrated in Fig. 1.
The large number of potential drug targets clearly presents a problem in target selection. The large number of targets also shows the potential plasticity in the energy balance system, indicating that treatment of obesity might require a combination of drugs acting at different points in the regulation of energy balance.
One way to separate the targets is on the tissue site of action, i.e. the brain, the gut and peripheral tissues (Table 2).
Table 2. Pros and cons of target types
Brain penetration adds to the challenges for synthetic chemists.
Leptin is a natural feedback regulator of body fat mass, but:
exogenous leptin is ineffectual, and endogenous production gives supramaximal level in most obese subjects;
there is resistance because of a transport defect or a down-regulation of response.
Activators of the brain biogenic amine receptors and transporters have a history of efficacy but are not without safety issues.
Brain neuropeptide receptors:
no compound has reached phase III and many neuropeptides have roles outside energy balance.
Other brain neurotransmitters:
cannabinoid CB1 blockers similar to rimonabant provide hope but current studies suggest efficacy will not break the 5 kg barrier by a significant amount.
Natural hormones or modified peptides of natural hormones are used to reduce hunger/enhance satiety:
built-in efficacy and safety.
They have to be injected, but long-acting treatments (1 dose month−1) are being developed.
Nausea tends to be a problem and this could limit use of 1 dose month−1 treatment.
Nearly all targets seem to affect energy balance by increasing energy expenditure, and low energy expenditure (relative to body mass) is at least a partial cause of obesity.
Selective loss of fat relative to lean tissue.
Increased fat oxidation improves insulin action, giving benefits in preventing/treating type 2 diabetes.
Mild exercise does not elicit compensatory increase in food intake, so drugs to increase energy expenditure may not invoke counter-regulatory response.
Drugs to increase mitochondrial activity but uncoupling drugs have safety concerns.
There are currently two drugs in phase III studies. Rimonabant (Sanofi-Aventis) is a cannabinoid CB1 antagonist, which has completed phase III trials in the USA. In a study of 3045 patients, mean weight loss over a year on a 20 mg d−1 dose was 9 kg, as against 2 kg for placebo, and waist circumference decreased by 9 cm compared with 2 cm for placebo. However, there was a 40% dropout rate (17). Rimonabant is believed to act centrally to regulate food intake, but may also interact with receptors in fat tissue. Certainly, animal studies suggest that weight loss is not just due to reduced food intake. ATL-962, now named cetilistat (Alizyme), is a pancreatic lipase inhibitor related to orlistat that blocks fat digestion. It does not appear to be more efficacious but may have less side effects. Two peptides currently marketed in the USA by Amylin Corporation for diabetes – the amylin mimetic Symlin and the glucagon-like peptide-1 analogue exendin-4 – have shown anti-obesity properties in animals and human trials might be expected.
Popular obesity targets at an earlier stage are shown in Table 3.
Table 3. Some popular anti-obesity targets
PPAR, peroxisome proliferator activated receptor.
CB-1 receptor antagonists – many compounds as followers to rimonabant are being developed.
Melanocortin-4 receptor agonists – problems of achieving receptor selectivity and side-effect issues.
Melanin-concentrating hormone-1 receptor antagonist – evidence of efficacy in animal models.
Serotonin-2c receptor agonists in clinical studies – theoretically, should have efficacy of dexfenfluramine without side effects.
Gut and other peptide targets
Peptide YY3-36– natural anorexic/satiety agent.
Oxyntomodulin – natural anorexic/satiety agent.
Small-molecule leptin mimetic – oral-dose therapy with potential to access the brain and activate leptin receptor.
PPAR delta activators – cause mitochondrial biogenesis and increase oxidative muscle mass.
11β-hydroxysteroid dehydrogenase-1 inhibitors – block production of stress hormone cortisol and exerts anti-obesity and anti-diabetic effects.
Current cutting-edge research
A potentially important finding is that events in utero and in early life can programme subjects for life with respect to regulation of energy balance. Rodent studies have shown that dietary manipulation of the dams and in early life of the offspring can determine the body composition in adult life. Moreover, this can be modified by the administration to the dams of the hormone leptin (18). This has important implications in our understanding of what factors determine the set point for body energy regulation. Several findings in human studies suggest that the findings in rodents may well be mirrored in humans, for example, the comparative effects of breast milk vs. formula feeds and the result of rapid catch-up growth in babies. Thus, it seems that there is a critical period in the pre-weaning phase when nutritional control and possibly hormonal supplementation with leptin could have a lifelong effect in programming the offspring to resist obesity. Research will need to identify whether there are other stages in life, for example puberty, when programming occurs.
A second area of research is the systematic evaluation of botanicals from various parts of the world to identify extracts and ultimately chemical entities with potential to treat obesity and its metabolic consequences. One extract is the sap of the Hoodia cactus, which has powerful anorectic activity (19). Such chemicals might be drugs in their own right or serve as precursors for chemical modification into drugs, as has been done recently with Artemisia for malaria treatment. However, possibly the more important issue with botanicals is that botanical extracts could be produced locally in developing countries, where the brunt of the future obesity problem will fall, as a cheap alternative to international pharmaceutical products.
The impact of research in the next 25 years
During the next 10 years, several new drugs will probably come forward from major pharmaceutical companies. It is unlikely that any of them alone will be any more successful in treating obesity than present ones in terms of absolute weight loss, but they will probably have less side effects. However, if coupled with an ability to identify an individual’s SNP profile, targeted medicine should allow the identification of individuals most likely to benefit. As such methodology will be relatively expensive (and the more patients pay, the more successful the weight loss), this should lead to greater compliance. It is also likely that combination therapy (again tailored to genetic make-up) will become the norm and this could result in greater weight loss than monotherapy.
The theoretical possibility of supplementing formula feeds with leptin or giving breastfed babies supplements that would alter programming, so that obesity in later life is resisted is likely to promote an ethical debate, but is it any different from giving vaccinations to prevent infectious disease?
The next 10 years are likely to reveal a greater understanding of the control of mitochondrial biogenesis and why mitochondrial loss occurs. Such understanding will be driven by not only obesity research, but also research on other age-related diseases including type 2 diabetes and cognitive impairment. Mitochondria are the energy powerhouses of cells and there is a growing belief that it is defects in mitochondrial metabolism that lie at the heart of defective fat oxidation and low energy expenditure, which underlie the development of obesity (20). This research drive will be coupled with research to identify drugs to modify mitochondrial metabolism.
In the next 25 years, there will be research to identify the mechanism by which early life programming determines the set point of energy balance and how the numerous brain circuits are integrated and regulated so that energy expenditure and energy intake are matched. This research will take into account the impact of various nutrient, tonic and hedonic signals. This understanding could allow reprogramming at critical points, better use of drugs, the design of diet and exercise regimes, and the improved use of behavioural psychology to elicit improved treatments.
Conflict of Interest Statement
• The author has a patent application covering formulations and use of leptin in milk to protect against the development of obesity.
• The author is a consultant on obesity and diabetes to numerous pharmaceutical companies and is on the scientific advisory board of Inpharmatica (UK), Carex (France) and Probiodrug (Germany).
• The author is a shareholder in GlaxoSmithKline and Phytopharm.
I thank Professor J. Arch for allowing me to use Fig. 1 and for considerable help in the preparation of this review.