Sulfonylureas are insulin secretogogues, since they control blood glucose levels by directly stimulating first-phase insulin secretion in the pancreatic β cells. Through the concerted efforts of GLUT2 (the high K_m glucose transporter) (22), glucokinase (the glucose sensor) (23), and glucose metabolism, these cells are responsible for sensing and secreting the appropriate amount of insulin in response to a glucose stimulus (24). Mitochondrial glucose metabolism leads to ATP generation and increases the intracellular ratio of ATP/ADP, which results in the closure of the ATP-sensitive potassium channel (K_ATP; a 140 kDa membrane protein) on the plasma membrane. Closure of this channel depolarizes the membrane and triggers the opening of voltage-sensitive calcium channels, leading to the rapid influx of calcium. Increased intracellular calcium causes an alteration in the cytoskeleton, and stimulates translocation of insulin-containing secretory granules to the plasma membrane and the exocytotic release of insulin.

The K_ATP channel is comprised of two subunits, both of which are required for the channel to be functional. One subunit contains the cytoplasmic binding sites for both sulfonylureas and ATP, and is designated as the sulfonylurea receptor type 1 (SUR1). The other subunit is the potassium channel, which acts as the pore-forming subunit (25). Either an increase in the ATP/ADP ratio or ligand binding (by sulfonylureas, meglitinides) to SUR1 results in the closure of the K_ATP channel and insulin secretion. Studies comparing sulfonylureas and non-sulfonylurea insulin secretogogues have identified several distinct binding sites on the SUR1 that cause channel closure. Some sites exhibit high affinity for glyburide and other sulfonylureas, while other sites exhibit high affinity for the non-sulfonylurea secretogogues (vide infra).

The over clinical efficacy of sulfonylureas in patients with type 2 diabetes is related to the pre-treatment levels of fasting plasma glucose and HbA_1C. The higher the fasting glucose level, the greater the effect will be. In patients with a pre-treatment glucose level of approximately 200 mg/dl (11.1 mmol/l), sulfonylureas typically will reduce glucose by 60-70 mg/dl (3.3-3.9 mmol/l) and HbA_1C by 1.5-2% (Table 3). The most responsive patients are those who exhibit mild-to-moderate fasting hyperglycemia (<200-240 mg/dl; <12.2-13.3 mmol/l), along with adequate residual β-cell function (evidenced by elevated fasting C-peptide). When used at maximally effective doses, results from well-controlled clinical trials have not indicated a superiority of one 2nd generation sulfonylurea over another. Similarly, 2nd generation sulfonylureas exhibit similar clinical efficacy compared to the 1st generation agents. The principal advantage of glimepiride and Glucotrol XL compared to other agents is the once daily dosing regimen. Approximately 10-20% of patients will exhibit a poor initial response to sulfonylureas (primary failures). While these patients are typically those who have severe fasting hyperglycemia (>280 mg/dl; >15.5 mmol/l) and reduced fasting C-peptide levels, these tests are not specific enough to help decide on the usefulness of a sulfonylurea for an individual patient. In addition, treatment with sulfonylureas results in the eventual loss of therapeutic effectiveness (secondary failure) in the range of 3-10% per year (7).

The major side effect from sulfonylurea treatment is hypoglycemia. This side effect is really just an extension of the therapeutic objective. Mild hypoglycemic events occur in approximately 2-4% of patients and severe hypoglycemic reactions that require hospitalization occur at a frequency of 0.2-0.4 cases per 1000 patient-years of treatment (26). In light of this, initiation of treatment with sulfonylureas should be at the lowest recommended dose. An additional undesirable effect of sulfonylurea therapy (as is also the case with insulin therapy) is weight gain. In the UKPDS, sulfonylurea treatment caused a net weight gain of 3 kg, which occurred during the first 3-4 years of treatment and then stabilized (27). In contrast, weight gain in response to insulin therapy increased progressively for the duration of the study. As mentioned above, chlorpropamide is associated with hyponatremia (SIADH) and an alcohol flushing reaction (disulfiram-Antibuse reaction). All the agents can cause intrahepatic cholestasis. Rarely maculopapular or urticarial rashes occur.

In renal failure, the dose of the sulfonylurea agent will require adjustment based on glucose monitoring. The half-life of insulin is extended in renal failure and thus there is an increased risk for hypoglycemia. This risk is typically manifest with fasting hypoglycemia.

Similar to sulfonylureas, meglitinides are insulin secretogogues, since they control blood glucose levels by directly stimulating first-phase insulin secretion in the pancreatic β cells. Receptor-binding studies performed in βTC-3 cells identified a high-affinity repaglinide (K_D = 3.6 nmol/l) site having lower affinity for glyburide (14.4 nmol/l), and one high-affinity glyburide (25 nmol/l) site having lower affinity for repaglinide (550 nmol/l)(Figure 1.4)(30). Repaglinide is approximately 5 times more potent than glyburide in stimulating insulin secretion. Unlike glyburide (and other sulfonylureas), repaglinide does not stimulate insulin secretion in vitro in the absence of glucose. Rather, it enhances glucose-stimulated insulin secretion especially at 180 mg/dl (10 mmol/l) glucose.

The mechanism of action of nateglinide (Figure 1.5) also involves the binding to and closure of the K_ATP channel resulting in membrane depolarization, an influx of calcium, and insulin exocytosis (31). The kinetics of interaction of nateglinide with the K_ATP channel are distinct compared to both rapaglinide and sulfonylureas, and accounts for its rapid insulinotropic effects. The onset of action of nateglinide is similar to that of glyburide but three-fold more rapid than that of rapaglinide (28). When nateglinide is removed from the K_ATP channel, its effect is reversed twice as quickly as glyburide and five times more quickly than rapaglinide. Thus, nateglinide initiates a more rapid release of insulin that is shorter in duration compared to rapaglinide (28), despite having an in vivo pharmacokinetic profile that is similar (32;33).

The efficacy of rapaglinide, when used as a monotherapy, is similar to sulfonylureas (34). Repaglinide treatment of patients with type 2 diabetes reduced fasting plasma glucose by approximately 60 mg/dl and HbA_1C by 1.7% (35). In a double-blind placebo-controlled study, rapaglinide had similar effects on lowering HbA_1C (0.5-2%) and fasting plasma glucose (65-75 mg/dl; 3.6-4.2 mmol/l) compared to glyburide (36). Repaglinide is also efficacious when used in combination with either metformin or troglitazone (a thiazolidinedione withdrawn from the market). In patients treated with rapaglinide and metformin, HbA_1C was decreased from 8.3% to 6.9% and fasting plasma glucose by 40 mg/dl (2.2 mmol/l) (37). Although lowered, the changes observed in subjects treated with either repaglinide or metformin monotherapy were not significant for HbA_1C (0.4 and 0.3% decrease, respectively), or fasting plasma glucose (9 mg/dl (0.5 mmol/l) increase and 5.4 mg/dl (0.3 mmol/l) decrease, respectively). Significant increases in body weight occurred in the both repaglinide and combined therapy groups (2.4 ± 0.5 and 3.0 ± 0.5 kg, respectively).

The combination therapy of rapaglinide and troglitazone showed a significant reduction in mean HbA_1C values (1.7%) that was greater than with either type of monotherapy (38). Repaglinide monotherapy resulted in a reduction of HbA_1C values that was significantly greater than troglitazone (0.8% vs. 0.4%). In addition, combination therapy was more effective in reducing fasting plasma glucose (80 mg/dl) than either repaglinide (43 mg/dl) or troglitazone (46 mg/dl) monotherapies. Repaglinide is also efficacious when used in combination with other available thiazolidinediones, rosiglitazone (Avandia®) and pioglitazone (Actos®) (39).

The efficacy of nateglinide when used as a monotherapy is similar to sulfonylureas and repaglinide (29;40). However, several therapeutically attractive features distinguish nateglinide from repaglinide and sulfonylureas. Nateglinide produces a more rapid post-prandial increase in insulin secretion, and its duration of response is shorter than that of glyburide (41;42). Thus, the risk of post-absorptive hypoglycemia should be lower than with either sulfonylureas or rapaglinide.

The efficacy of nateglinide treatment has been evaluated alone and in combination with metformin in patients with type 2 diabetes (43). In this randomized double-blind study, patients with an HbA_1C level between 6.8 and 11.0% during a 4-week placebo run-in received 24 weeks’ treatment with 120 mg nateglinide before meals (n = 179), 500 mg metformin three times a day (n = 178), combination therapy (n = 172), or placebo (n = 172). At the study conclusion, HbA_1C and fasting plasma glucose were significantly reduced from baseline with nateglinide [0.5% and 12.6 mg/dl (0.7 mmol/l), respectively] and metformin [0.8% and 28.8 mg/dl (1.6 mmol/l), respectively], but was increased with placebo [0.5% and 7.2 mg/dl (0.4 mmol/l), respectively]. Combination therapy was additive [HbA_1C, 1.4% and glucose, 43.2 mg/dl (2.4 mmol/l)]. Although only preliminary data are available, nateglinide also appears effective when used in combination with thiazolidinediones (29;39).

In 1-year trials, the most common adverse events reported in repaglinide recipients (n = 1,228) were hypoglycemia (16%), upper respiratory tract infection (10%), rhinitis (7%), bronchitis (6%) and headache (9%). The overall incidence of hypoglycemia was similar to that recorded in patients receiving glibenclamide, glipizide or gliclazide (18%; n = 597); however, the incidence of serious hypoglycemia appears to be slightly higher in sulphonylurea recipients. Weight gain does occur in patients treated with repaglinide, but the magnitude is significantly less compared to treatment with glyburide. In patients switched from sulfonylureas to repaglinide, no weight gain was observed; in drug-naïve patients, repaglinide-treatment increased body weight by approximately 3% (6lb) (36;44).

The clinical trials of nateglinide carried out to date have found the drug to be safe and well tolerated. Dosage regimens ranging from 60 to 240 mg have been evaluated. The most common adverse effects are nausea, diarrhea, dizziness, and lightheadedness. The incidence of mild hypoglycemia is lower with nateglinide than for rapaglinide and no reports of severe hypoglycemia, consistent with the mechanism of action of nateglinide. In the clinical studies carried out to date, there have been no reports of any increase body weight gain.

An elevated rate of basal hepatic glucose output is the primary determinant of elevated fasting blood glucose levels in patients with type 2 diabetes (48). The primary effect of metformin (Figure 1.6) is the suppression of basal hepatic glucose production, thereby reducing fasting plasma glucose (46;49;50). Despite the large number of studies both in vitro and in humans that have established this mode of action, the molecular target of metformin action still awaits identification. Metformin does not stimulate insulin secretion; in contrast, metformin reduces fasting plasma insulin and improves whole-body insulin-stimulated glucose metabolism (insulin sensitivity) (46;50). While it is possible that the beneficial effect of metformin on insulin sensitivity is mediated directly, a more likely explanation is that it is secondary to a reduction in hyperglycemia, triglycerides, and free fatty acids.

Recent in vitro and in vivo evidence has shown that metformin activates the AMP-activated protein kinase (AMPK) (51), a major cellular regulator of lipid and glucose metabolism (52). As a result, acetyl-CoA carboxylase activity was reduced, fatty acid oxidation induced (due to decreased malonyl-CoA), and the expression of lipogenic enzymes along with SREBP-1, a key lipogenic transcription factor, suppressed (51). The use of a novel AMPK inhibitor indicated that AMPK activation was required for the inhibitory effect of metformin on glucose production in hepatocytes. In isolated rat skeletal muscles, metformin stimulated glucose uptake coincident with AMPK activation. These results are intriguing in that they implicate the activation of AMPK as a unified explanation for the beneficial effects of metformin.

A large number of well-controlled clinical studies have established that metformin monotherapy consistently reduces fasting plasma glucose by 60-70 mg/dl (3.3-3.9 mmol/l) and HbA_1c by 1.5-2.0% (9;47;50). Thus, the efficacy of metformin is in the same range as that observed for monotherapy treatment with sulfonylureas. Similar to the sulfonylurea treatment, the overall magnitude of response to metformin is directly related to the starting fasting plasma glucose concentration. Metformin also reduces fasting plasma insulin, triglycerides, and free fatty acids (50). Unlike sulfonylurea treatment, metformin monotherapy is not associated with weight gain and even promotes a modest weight loss. When used in combination with other oral agents or insulin, weight gain is not observed. Metformin is the only oral agent that when used as a monotherapy has been reported to reduce the risk of developing macrovascular complications (53).

In patients with type 2 diabetes in the US, the mean HbA_1c is approximately 10% and fasting plasma glucose approximately 200-240 mg/dl (50). In patients of this sort, monotherapy with either metformin or a sulfonylurea generally will decrease plasma glucose to <140 mg/dl (<7.8 mmol/l) in about 25-30% of patients. In contrast, combined metformin and sulfonylurea therapy increases the percentage of patients who achieve this level of control to approximately 60-70% (50). When added to a sulfonylurea, the effects of both agents are additive, consistent with their different mechanisms of action. Interestingly, in patients that no longer responded to sulfonylurea treatment (secondary failures) and were removed from treatment, addition of metformin had minimal effects (54). Thus, in these patients, sulfonylurea treatment was still eliciting an effect, emphasizing the need to continue treatment with both agents.

The additive effect of metformin and sulfonylurea therapy is illustrated in Figure 4. As shown, there was no change in glucose levels when the sulfonylurea was changed to metformin. However, when the metformin was added, there was a dramatic decrease in plasma glucose. In fact, this pattern is seen in virtually all studies comparing two oral agents. This concept is illustrated in Figure 5. When a patient is on drug A and they are changed to drug B, no improvement in glucose control will be seen. However, if drug B is added to drug A, there is an improvement. This concept can often be extended by the addition of drug C, drug D, etc.

Figure 4. Effect of metformin on fasting plasma glucose when given as add-on therapy to glyburide. Data show the effects on fasting plasma glucose (mean change from baseline) in patients with type 2 diabetes continuing on glyburide therapy, switched to metformin, or given metformin as add-on therapy to glyburide. Data from .

Figure 5. Efficacy when oral agents are used as add-on therapy. When a patient is on drug A and they are changed to drug B, C, or D, often no improvement in glucose control will be seen. However, if drug B is added to drug A, there is an improvement. This concept can often be extended by the addition of drug C, drug D.

The most common side effects of metformin are gastrointestinal disturbances (abdominal discomfort, diarrhea), which occur in approximately 20-30% of patients. These effects are generally transient, and can be minimized or avoided by careful dose titration. The incidence of lactic acidosis is rare and occurs with a frequency of 3 cases per 100,000 patient-years. However, metformin is contraindicated in patients with impaired renal function and liver disease.

The risk of lactic acidosis can be minimized when the following are considered:

The primary effects of pioglitazone (Figure 1.7) and rosiglitazone (Figure 1.8) are the reduction of insulin resistance and improvement of insulin sensitivity, resulting in a reduction of fasting plasma glucose, insulin, and free fatty acids (10;63-66). Unlike other existing anti-diabetic medications that possess a very rapid onset of activity, pioglitazone and rosiglitazone exhibit a characteristic delay from 4-12 weeks in the onset of their therapeutic benefits. This is likely related to their mode of action, which involves the regulation of gene expression (63;67;68). Pioglitazone and rosiglitazone are selective agonists for the peroxisome proliferator-activated receptor γ (PPARγ), a member of the superfamily of nuclear hormone receptors that function as ligand-activated transcription factors (69;70). The PPAR family, which also includes PPARα and PPARδ, functions as receptors for fatty acids and their metabolites (e.g. eiconasoids) and, consequently, plays a critical physiological role the regulation of glucose, fatty acid, and cholesterol metabolism. PPARα is the receptor for the fibrate class of lipid-lowering drugs, and PPARδ is involved in the regulation of high-density lipoprotein metabolism (70;71).

The structure-activity relationship between PPARγ agonists and their glucose lowering activity in vivo has been established (72). In the absence of ligand, PPARs bind as heterodimers with the 9-cis retinoic acid receptor (RXR) and a multi-component co-repressor complex to a specific response element (PPRE) within the promoter region of their target genes (70;73). Once PPAR is activated by ligand, the co-repressor complex dissociates allowing the PPAR-RXR heterodimer to associate with a multi-component co-activator complex resulting in an increased rate of gene transcription. The target genes of PPARγ include those involved in the regulation of lipid and carbohydrate metabolism (74-76).

It does not appear that rosiglitazone and pioglitazone improve insulin sensitivity and glucose disposal by direct effects on either liver or muscle. PPARγ is expressed chiefly in adipose tissue, and its expression in liver and skeletal muscle is low (77;78). Thus, it is more likely that the primary effects of these drugs are on adipose tissue, followed by secondary benefits on other target tissues of insulin (79). The ability of rosiglitazone and pioglitazone to decrease circulating free fatty acids could lead to an improvement in insulin action in the periphery (63;64;80;81). More recently, PPARγ agonists have been reported to increase the expression and circulating level of adiponectin (Acrp30), an adipocyte-derived protein with insulin sensitizing activity (82;83), in diabetic rodents (79) and in patients with type 2 diabetes (84). Recognition of the importance of PPARγ in the overall regulation of carbohydrate and lipid metabolism along with growing realization that the adipocyte is an endocrine organ (85;86) suggests that investigations in this area will intensify, and perhaps uncover additional mechanisms by which rosiglitazone and pioglitazone improve insulin sensitivity and glucose disposal.

The major side effects of this class of drugs are edema, weight gain, decreased hematocrit and hemoglobin, and elevated (but reversible) alanine aminotransferase activity. Unlike troglitazone, idiosyncratic hepatic reaction does not appear to be a problem with rosiglitazone or pioglitazone. The edema ranges from bothersome trace to anasarca. The mechanism of the edema production is not known. Clinically, diuretics have minimal effect on reducing the edema. While there are no published studies on this subject, it does appear that the edema is dose dependent. Weight gain may be a modest 2-4 pounds to >20 lbs (100). Due to their mechanism of action, the risk of hypoglycemia with rosiglitazone or pioglitazone monotherapy is low. Mild to moderate hypoglycemia has been reported during combination therapy with sulfonylureas or insulin (10;101).

α-Glucosidase inhibitors are competitive, reversible inhibitors of pancreatic α-amylase and membrane-bound intestinal α-glucosidase hydrolase enzymes. Acarbose, the first α-glucosidase inhibitor discovered, is a nitrogen-containing pseudotetrasaccharide (Figure 1.9), while miglitol is a synthetic analog of 1-deoxynojirimycin (Figure 1.10). The mechanism of action of these inhibitors is similar but not identical. They bind competitively to the oligosaccharide binding site of the α-glucosidase enzymes, thereby preventing enzymatic hydrolysis. Acarbose binding affinity for the α-glucosidase enzymes is: glycoamylase > sucrase > maltase > dextranase (102;104). Acarbose has little affinity for isomaltase and no affinity for the β-glucosidase enzymes, such as lactase (102). Miglitol is a more potent inhibitor of sucrase and maltase that acarbose, has no effect on α-amylase, but does inhibit intestinal isomaltose (102).

Clinical trials conducted to date have established that the antihyperglycemic effectiveness of acarbose and miglitol is less than 50% than that of either sulfonylureas or metformin. When used as monotherapy, acarbose primarily affects post-prandial glucose levels, which is reduced by 40-50 mg/dl after meal (9;12;102;103;105;106). In most studies, α-glucosidase inhibitors have no significant effects on either fasting insulin or whole body insulin sensitivity in patients with type 2 diabetes. However, there is some evidence that acarbose and voglibose, a structural analog of miglitol in clinical development in Japan, reduces post-prandial hyperinsulinemia in glucose intolerant individuals (107;108). Some but not all studies have reported small decreases in fasting or post-prandial triglycerides (102). Since the mechanism of action of α-glucosidase inhibitors is different from other oral agents, their effects on glycemic control are additive when used in combination. As summarized by Lebovitz (102), addition of acarbose to sulfonylurea therapy decreases HbA_1c by 0.85%; addition of acarbose to metformin therapy decreases HbA_1c by 0.73%; and addition of acarbose to insulin therapy decreases HbA_1c by 0.54%. As for monotherapy, the predominant improvement is on post-prandial hyperglycemia. Treatment with α-glucosidase inhibitors appears to have a lower rate of secondary failures characteristic of sulfonylurea and metformin therapy.

The major side effects of the α-glucosidase inhibitors are related to gastrointestinal disturbances. These occur in approximately 25-30% of diabetic patients, and include flatulence, diarrhea, bloating, and abdominal discomfort. These side effects can often be minimized by careful dose titration, and sometimes diminish with time. Acarbose is contraindicated in patients with inflammatory bowel disease, cirrhosis, or elevated plasma creatinine (>177 μmol/l). This class of drugs is associated with dose-dependent hepatotoxicity, and serum transaminase levels require monitoring for patients receiving high doses (>200 mg three times daily). Transaminase elevations, which are often asymptomatic, are reversible upon cessation of treatment. Hypoglycemia does not occur in patients on α-glucosidase inhibitor monotherapy. If hypoglycemia occurs while a patient is taking an α-glucosidase inhibitor simultaneously with a sulfonylurea, insulin or a meglitinide, the recommended action is oral administration of pure glucose, dextrose or milk.

Glucagon-like peptide 1 (GLP-1) is an insulinotropic hormone secreted by L-cells of the small intestine. GLP-1 has several important biological actions including the stimulation of insulin secretion in a glucose-specific manner, inhibition of gastric emptying, suppression of glucagon secretion, and central anorexic activity (114;115). Patients with type 2 diabetes exhibit reduced levels of active GLP-1 (amino acids 7-36) along with an impaired GLP-1 response to a glucose load, and parenteral administration of GLP-1 has been shown to reduce fasting and post-prandial glycemia in patients with type 1 and type 2 diabetes (114;115). Although it possesses multiple effective clinical activities, administration of GLP-1 is not an ideal approach since it cannot be administered orally. Furthermore, endogenous (and exogenously administered) GLP-1 has undesirable pharmacokinetics; after it is secreted, it is rapidly cleaved and inactivated [plasma half-life < 1 min; (116)] by the enzyme dipeptidyl peptidase IV (DPP-IV). Thus, inhibition of DPP-IV has been suggested as a feasible alternative to elevate circulating GLP-1 levels, and circumvent the limitations of GLP-1 administration (116-123). Not surprisingly, there is intense activity directed towards the identification and characterization of small molecule inhibitors of DPP-IV for the treatment of impaired glucose tolerance and type 2 diabetes (124-128). This approach has been convincingly validated in vivo and in the clinic.

The initial, orally active DPP-IV inhibitor, NVP-DPP728, was identified and characterized in vitro and in vivo (118;119;129-131). Inhibition of DPP-IV by NVP-DPP728 resulted in a significantly amplified early phase of the insulin response to an oral glucose load in obese fa/fa rats, and restoration of glucose excursions to normal (129). In contrast, DPP-IV inhibition produced only minor effects in lean FA/? rats. Inactivation of GLP-1 (7-36) amide was completely prevented by DPP-IV inhibition suggesting that the effects of this compound on oral glucose tolerance were mediated by increased circulating concentrations of GLP-1 (7-36) amide.

In DPP-IV(+) [but not in DPP-IV(-)] transgenic rats fed either standard chow or a high-fat diet, NVP-DPP728 significantly suppressed glucose excursions after glucose challenge by inhibiting the plasma DPP-IV activity, associated with the stimulation of early insulin secretion (130). NVP-DPP728 also improved the glucose tolerance after an oral glucose challenge by potentiating the early insulin response by inhibition of plasma DPP-IV activity in aged DPP-IV(+) Wistar and F344 rats (131). In contrast, NVP-DPP728 did not affect the glucose tolerance after an oral glucose challenge in aged DPP-IV(-)F344 rats (131). Taken together, these results indicate that treatment with NVP-DPP728 ameliorates glucose tolerance in vivo by the direct inhibition of plasma DPP-IV activity, and presumably the subsequent increase in endogenous GLP-1 action.

The clinical activity of this compound as a monotherapy was reported in patients at an early stage of type 2 diabetes (132). Compared with placebo, NVP-DPP728 at 100 mg (tid; n =31) significantly reduced fasting glucose by 18 mg/dl (1.0 mmol/l), prandial glucose excursions by 21.6 mg/dl (1.2 mmol/l), and mean 24-h glucose levels by 18 mg/dl (1.0 mmol/l). Similar reductions were seen in the 150-mg (bid; n = 32) treatment group. Mean 24-h insulin was significantly reduced by 26 pmol/l in both groups. In the combined active treatment groups, HbA_1c was significantly reduced from 7.4 ± 0.7% to 6.9 ± 0.7%. Laboratory safety and tolerability were good in all groups. These results provided clinical proof of concept that inhibition of DPP-IV is a feasible approach for the treatment of type 2 diabetes in the early stage of the disease.

The appetite-stimulating actitivity of Cannabis sativa has long been known through traditional recreational and medicinal use (147;148). The active molecule, delta-9-tetrahydrocannabinol (THC), was initially purified in 1964 (149), but the discoveries of cannabinoid receptors and their endogenous ligands did not occur until more recently, within the last 15 years (150). The endocannabinoid system is composed of two distinct G-protein-coupled cannabinoid receptors (CB₁ and CB₂), two major endogenous ligands (anandamide and 2-arachidonyl glycerol), and the enzymes responsible for ligand synthesis (N-acylphosphatidylethanolamine-specific phospholipase D and diacylglycerol lipase) and degradation (fatty acid amide hydrolase and monoacylglycerol lipase). CB₁ receptors are widely distributed on central and peripheral neurons, and also present in adipose tissue, liver, muscle, pancreas, heart, vasculature, and the gastrointestinal tract (150;151). The distribution of CB₂ receptors is more limited, where they are mainly associated with the immune system and hematopoisesis (150;151). Ligand stimulation of the receptors results in decreased adenylate cyclase activity and cAMP production, along with a reduction in intracellular calcium.

The endocannabinoid system has recently been identified as playing a significant role in the regulation of food intake and energy metabolism (150). There is some evidence that the endocannabinoid system is consituitively active in certain regions of the brain that are known to control energy balance, suggesting that tonic activation favors energy intake, storage, and weight gain (152). In rodents, activation of CB₁ (but not CB₂) receptors by endogenous ligands or THC stimulates feeding (153-155), while CB₁-deficient mice are protected against diet-induced obesity (156;157). In human obesity, it has been reported that fatty acid amide hydrolase, the enzyme that degrades anandamide, is reduced (158), and that a missense mutation is associated with increased body mass index (159). Taken together, these data have provided the scientific rationale for the idea that pharmacological antagonists of CB₁ might offer a promising approach to treat obesity.

Rimonabant (SR 141716; Accomplia™) is a selective CB₁ receptor antagonist in late Phase 3 trials for obesity and smoking cessation (160-163). Animal studies have convincingly demonstrated that this compound exerts a beneficial effect on reducing hunger, caloric intake, and body weight, and for increasing satiety (163). In humans, rimonabant is proving equally effective, as judged by the clinical data reported thusfar. In the RIO (Rimonabant in Obesity)-Europe study, patients with body-mass index 30 kg/m² or greater, or body-mass index greater than 27 kg/ m² with treated or untreated dyslipidaemia, hypertension, or both, were randomised to receive double-blind treatment with placebo (n = 305), 5 mg rimonabant (n = 603), or 20 mg rimonabant (n = 599) once daily in addition to a mild hypocaloric diet (600 kcal/day deficit) (164). The primary efficacy endpoint was weight change from baseline after 1 year of treatment in the intention-to-treat population. Weight loss at 1 year was significantly greater in patients treated with rimonabant 5 mg (-3.4 kg; P = 0.002 vs. placebo) and 20 mg (-6.6 kg; P < 0.001 vs. placebo) compared with placebo (-1.8 kg). Significantly more patients treated with rimonabant 20 mg than placebo achieved weight loss of 5% or greater (P < 0.001), and 10% or greater (P < 0.001). Rimonabant (20 mg) produced significantly greater improvements than placebo in waist circumference, HDL-cholesterol, triglycerides, and insulin resistance, and prevalence of the metabolic syndrome. The effects of rimonabant 5 mg were less impressive with respect to these parameters. Rimonabant was generally well tolerated with mild and transient side effects. Thus, rimonabant (20 mg), combined with a hypocaloric diet over 1 year, promoted significant decrease of bodyweight and waist circumference, and improvement in cardiovascular risk factors.

The RIO-North America study compared the efficacy and safety of rimonabant with placebo, in conjunction with diet and exercise, for sustained changes in weight and cardiometabolic risk factors over 2 years (165). The study design was a randomized, double-blind, placebo-controlled trial of 3045 obese (body mass index > or =30) or overweight (body mass index > 27 and treated or untreated hypertension or dyslipidemia) adult patients at 64 US and 8 Canadian clinical research centers. After a 4-week single-blind placebo plus diet (600 kcal/d deficit) run-in period, patients were randomized to receive placebo, 5 mg rimonabant, or 20 mg rimonabant once daily for 1 year. Rimonabant-treated patients were re-randomized to receive placebo or continued to receive the same rimonabant dose while the placebo group continued to receive placebo during year 2. The primary out measures was body weight change over year 1, and prevention of weight regain during year 2. Additional efficacy measures included changes in waist circumference, plasma lipid levels, and other cardiometabolic risk factors. At year 1, the completion rate was 309 (51%) patients in the placebo group, 620 (51%) patients in the 5 mg rimonabant group, and 673 (55%) patients in the 20 mg rimonabant group. Compared with the placebo group, the 20 mg of rimonabant group produced greater mean reductions in weight (-6.3 kg vs -1.6 kg; P < 0.001), waist circumference (-6.1 cm vs -2.5 cm; P < 0.001), and level of triglycerides (percentage change, -5.3 vs 7.9; P < 0.001), and a greater increase in level of high-density lipoprotein cholesterol (percentage change, 12.6 vs 5.4; P < 0.001). Patients who were switched from the 20 mg of rimonabant group to the placebo group during year 2 experienced weight regain while those who continued to receive 20 mg of rimonabant maintained their weight loss and favorable changes in cardiometabolic risk factors. Use of different imputation (a simulation-based approach to the statistical analysis of incomplete data) methods to account for the high rate of dropouts in all 3 groups yielded similar results. Rimonabant was generally well tolerated; the most common drug-related adverse event was nausea (11.2% for the 20 mg of rimonabant group vs 5.8% for the placebo group). In this multicenter trial, treatment with 20 mg/d of rimonabant plus diet for 2 years promoted modest but sustained reductions in weight and waist circumference and favorable changes in cardiometabolic risk factors. However, the trial was limited by a high drop-out rate, and longer-term effects of the drug will require further study.

The RIO-Lipids study examined the effects of rimonabant on metabolic risk factors, including adiponectin levels, in high-risk patients who were overweight or obese and had dyslipidemia (166). Overweight or obese patients (n = 1036; body-mass index: 27 to 40) with untreated dyslipidemia (triglyceride levels >1.69 to 7.90 mmol per liter, or a ratio of cholesterol to high-density lipoprotein (HDL)-cholesterol of >4.5 among women and >5 among men) were randomly assigned to double-blinded therapy with either placebo or rimonabant at a dose of 5 mg or 20 mg once daily for 12 months in addition to a hypocaloric diet (600 kcal/d deficit). The rates of completion of the study were 62.6 percent, 60.3 percent, and 63.9 percent in the placebo group, the group receiving 5 mg of rimonabant, and the group receiving 20 mg of rimonabant, respectively. The most frequent adverse events resulting in discontinuation of the drug were depression, anxiety, and nausea. As compared with placebo, rimonabant at a dose of 20 mg was associated with a significant (P < 0.001) mean weight loss (-6.7kg), reduction in waist circumference (-5.8 cm), increase in HDL-cholesterol (10.0%), and reduction in triglycerides (-13.0%). Rimonabant at a dose of 20 mg also resulted in an increase in plasma adiponectin levels (58%; P < 0.001), an increase that was partly independent of weight loss alone. Thus, rimonabant significantly reduces body weight and waist circumference and improves the profile of several metabolic risk factors in high-risk patients who are overweight or obese and have an atherogenic dyslipidemia.

The results of the RIO-Diabetes trial are also available. In this trial, 1045 patients with type 2 diabetes received either 5 mg or 20 mg of rimonabant once daily or placebo for 1 year in addition to a hypocaloric diet (600 kcal/d deficit) (American Diabetes Association Meeting 2005; http://66.102.7.104/search?q=cache:yYTESm82sCkJ:www.diabetes.org/uedocuments/Rimonabant.pdf+rio-diabetes&hl=en&gl=us&ct=clnk&cd=2). Two-thirds of patients were receiving metformin, with the remainder on a sulfonylurea. Results from the lower dose of the drug were not clinically impressive. Patients receiving 20 mg rimonabant lost an average of 5.3 kg compared with 1.45 kg in the placebo group, with waist circumference reduced by 5.2 cm and 1.9 cm, respectively (P< 0.001, for both parameters). In addition, HbA_1c levels were reduced by 0.6% (P < 0.001) from a mean baseline value of 7.3% with 20 mg rimonabant, compared with an increase of 0.1% among patients taking placebo. In all, 43% of the 20 mg rimonabant patients achieved an HbA_1c level below 6.5% (the target treatment goal recommended by the International Diabetes Federation and the American Association of Clinical Endocrinologists) compared with 21% of those receiving placebo (P < 0.001). Among patients whose baseline HbA_1c was higher than 7% (the target treatment goal recommended by the American Diabetes Association), it was reported that ~53% of patients receiving rimonabant 20 mg lowered their HbA_1c beyond this by the end of the study, compared with ~27% of those receiving placebo (P < 0.001). Statistical analysis revealed that 0.3% of the HbA_1c improvement was attributable to weight loss, while 0.4% was an independent effect attributable to the rimonabant. The reported side effects were mainly mild and transient, with nausea most frequent – seen in 12.1% of patients on the higher dose – and dizziness experienced by 9.1%, mostly in the first few weeks. Thus, if confirmed in larger studies of longer duration, the weight loss reported with rimonabant in patients with type 2 diabetes could have important clinical implications, as current anti-hyperglycemic therapies (i.e. insulin, sulfonylureas, thiazolidinediones) are often associated with weight gain. This weight gain works against the benefits of treatment, and reduces the overall improvement in the risk for cardiovascular disease.