Oral Antidiabetic Agents - Luzimar Teixeira

Drugs 2005; 65 (3): 385-411 

REVIEW ARTICLE 0012-6667/05/0003-0385/$39.95/0 

© 2005 Adis Data Information BV. All rights reserved. 

Oral Antidiabetic Agents 

Current Role in Type 2 Diabetes Mellitus 

Andrew J. Krentz 1 and Clifford J. Bailey 2 

1 Southampton University Hospitals NHS Trust, Southampton, UK 

2 Life and Health Sciences, Aston Pharmacy School, Aston University, Birmingham, UK 

Contents 

Abstract ....................................................................................386 

1. Insulin Secretagogues ....................................................................390 

1.1 Sulphonylureas ......................................................................390 

1.1.1 Mode of Action ................................................................390 

1.1.2 Pharmacokinetics ..............................................................391 

1.1.3 Indications and Contraindications ...............................................391 

1.1.4 Efficacy .......................................................................393 

1.1.5 Adverse Events ................................................................393 

1.1.6 New Formulations of Sulphonylureas .............................................395 

1.2 Rapid-Acting Prandial Insulin Releasers .................................................395 

1.2.1 Mode of Action ................................................................395 



1.2.4 Efficacy .......................................................................396 

1.2.5 Adverse Events ................................................................396 

2. α-Glucosidase Inhibitors ..................................................................397 

2.1 Mode of Action .....................................................................397 

2.2 Pharmacokinetics ....................................................................397 

2.3 Indications and Contraindications .....................................................398 

2.4 Efficacy .............................................................................398 

2.5 Adverse Effects ......................................................................398 

3. Insulin Sensitisers ..........................................................................399 

3.1 Biguanides ..........................................................................399 

3.1.1 Mode of Action ................................................................399 



3.1.4 Efficacy .......................................................................402 

3.1.5 Adverse Effects ................................................................403 

3.2 Thiazolidinediones ...................................................................404 

3.2.1 Mode of Action ................................................................404 



3.2.4 Efficacy .......................................................................407 

3.2.5 Adverse Effects ................................................................407 

4. Summary and Conclusion .................................................................408

386 Krentz & Bailey 

Abstract 

Type 2 diabetes mellitus is a progressive and complex disorder that is difficult 

to treat effectively in the long term. The majority of patients are overweight or 

obese at diagnosis and will be unable to achieve or sustain near normoglycaemia 

without oral antidiabetic agents; a sizeable proportion of patients will eventually 

require insulin therapy to maintain long-term glycaemic control, either as monotherapy 

or in conjunction with oral antidiabetic therapy. The frequent need for 

escalating therapy is held to reflect progressive loss of islet β-cell function, 

usually in the presence of obesity-related insulin resistance. 

Today’s clinicians are presented with an extensive range of oral antidiabetic 

drugs for type 2 diabetes. The main classes are heterogeneous in their modes of 

action, safety profiles and tolerability. These main classes include agents that 

stimulate insulin secretion (sulphonylureas and rapid-acting secretagogues), 

reduce hepatic glucose production (biguanides), delay digestion and absorption of 

intestinal carbohydrate (α-glucosidase inhibitors) or improve insulin action (thiazolidinediones). 

The UKPDS (United Kingdom Prospective Diabetes Study) demonstrated the 

benefits of intensified glycaemic control on microvascular complications in newly 

diagnosed patients with type 2 diabetes. However, the picture was less clearcut 

with regard to macrovascular disease, with neither sulphonylureas nor insulin 

significantly reducing cardiovascular events. The impact of oral antidiabetic 

agents on atherosclerosis – beyond expected effects on glycaemic control – is an 

increasingly important consideration. In the UKPDS, overweight and obese 

patients randomised to initial monotherapy with metformin experienced significant 

reductions in myocardial infarction and diabetes-related deaths. Metformin 

does not promote weight gain and has beneficial effects on several cardiovascular 

risk factors. Accordingly, metformin is widely regarded as the drug of choice for 

most patients with type 2 diabetes. Concern about cardiovascular safety of 

sulphonylureas has largely dissipated with generally reassuring results from 

clinical trials, including the UKPDS. Encouragingly, the recent Steno-2 Study 

showed that intensive target-driven, multifactorial approach to management, 

based around a sulphonylurea, reduced the risk of both micro- and macrovascular 

complications in high-risk patients. Theoretical advantages of selectively targeting 

postprandial hyperglycaemia require confirmation in clinical trials of drugs 

with preferential effects on this facet of hyperglycaemia are currently in progress. 

The insulin-sensitising thiazolidinedione class of antidiabetic agents has potentially 

advantageous effects on multiple components of the metabolic syndrome; 

the results of clinical trials with cardiovascular endpoints are awaited. 

The selection of initial monotherapy is based on a clinical and biochemical 

assessment of the patient, safety considerations being paramount. In some circumstances, 

for example pregnancy or severe hepatic or renal impairment, insulin may 

be the treatment of choice when nonpharmacological measures prove inadequate. 

Insulin is also required for metabolic decompensation, that is, incipient or actual 

diabetic ketoacidosis, or non-ketotic hyperosmolar hyperglycaemia. Certain 

comorbidities, for example presentation with myocardial infarction during other 

acute intercurrent illness, may make insulin the best option. 

Oral antidiabetic agents should be initiated at a low dose and titrated up 

according to glycaemic response, as judged by measurement of glycosylated 

© 2005 Adis Data Information BV. All rights reserved. Drugs 2005; 65 (3)

Oral Antidiabetic Agents 387 

haemoglobin (HbA1c) concentration, supplemented in some patients by self 

monitoring of capillary blood glucose. The average glucose-lowering effect of the 

major classes of oral antidiabetic agents is broadly similar (averaging a 1–2% 

reduction in HbA1c), α-glucosidase inhibitors being rather less effective. Tailoring 

the treatment to the individual patient is an important principle. Doses are 

gradually titrated up according to response. However, the maximal glucose-lowering 

action for sulphonylureas is usually attained at appreciably lower doses 

(approximately 50%) than the manufacturers’ recommended daily maximum. 

Combinations of certain agents, for example a secretagogue plus a biguanide or a 

thiazolidinedione, are logical and widely used, and combination preparations are 

now available in some countries. While the benefits of metformin added to a 

sulphonylurea were initially less favourable in the UKPDS, longer-term data have 

allayed concern. When considering long-term therapy, issues such as tolerability 

and convenience are important additional considerations. 

Neither sulphonylureas nor biguanides are able to appreciably alter the rate of 

progression of hyperglycaemia in patients with type 2 diabetes. Preliminary data 

suggesting that thiazolidinediones may provide better long-term glycaemic stability 

are currently being tested in clinical trials; current evidence, while encouraging, 

is not conclusive. 

Delayed progression from glucose intolerance to type 2 diabetes in high-risk 

individuals with glucose intolerance has been demonstrated with troglitazone, 

metformin and acarbose. However, intensive lifestyle intervention can be more 

effective than drug therapy, at least in the setting of interventional clinical trials. 

No antidiabetic drugs are presently licensed for use in prediabetic individuals. 

In 1998, the results of the randomised, multicen- management plan that encompasses effective treattre 

UKPDS (United Kingdom Prospective Diabetes ment of hypertension and dyslipidaemia; [2-6] both 

Study) [1] provided firm evidence of the importance are commonly encountered in patients with type 2 

of long-term glycaemic control in middle-aged patients 

with newly diagnosed type 2 diabetes mellitus. 

Table I. Summary of main results of UKPDS (United Kingdom 

Compared with dietary manipulation alone, 

Relative 

Prospective Diabetes Study) glycaemic control study. 

risk (RR) reductions in clinical endpoints for patients randomised to 

intensified therapy in the form of oral antidiabetic intensive (i.e. sulphonylurea or insulin) vs conventional therapy (i.e. 

agents or insulin significantly reduced the development 

of microvascular complications (table I). [1] 

diet) 

Endpoints RR for Confidence Log-rank 

intensive interval 

This knowledge drives current clinical practice, in 

p-value 

therapy 

which treatment is directed to the attainment of 

Aggregate endpoints b 

near-normoglycaemia, i.e. glycosylated haemoglobin 

Diabetes-related endpoints 0.88 0.79, 0.99 0.029 

(HbA1c) concentrations of 6.5–7.0%. [2-4] Microvascular complications 0.75 0.60, 0.93 0.0099 

While such targets may be perceived as being unrealistic 

for many – perhaps most – patients, there is 

Single endpoints 

Sudden death 0.54 0.24, 1.21 0.047 

a broad consensus that chronic hyperglycaemia Retinal photocoagulation 0.71 0.53, 0.96 0.0031 

should be managed as well as is possible, weighing Cataract extraction 0.76 0.53, 1.08 0.046 

safety and quality-of-life considerations on an individual 

basis. It is important to bear in mind that 

a 95% Confidence interval for aggregate endpoints; 99% 

confidence interval for single endpoints. 

glycaemic control is just one aspect of an overall 

b As defined and ascertained in UKPDS 33. [1] 



Table II. Main results for intensive (n = 80) vs conventional (n = 80) treatment of patients with type 2 diabetes mellitus and microalbuminuria. 

Mean follow-up was 7.8 years [8] 

Outcomes Intensive (%) Conventional (%) Adjusted HR (95% CI) RRR (95% CI) NNT (95% CI) 

Composite endpoint 24 44 0.47 (0.22, 0.74) 5 (3, 19) 

nephropathy 24 47 61% (13, 83) 4 (3, 14) 

retinopathy 52 71 58% (14, 79) 5 (3, 35) 

autonomic neuropathy 36 64 63% (21, 82) 4 (2, 9) 

HR = hazard ratio; NNT = number needed to treat; RRR = relative risk reduction. 

diabetes and are regarded as important modifiable physical activity. The objective is always to improve 

risk factors for atherosclerosis, the principal cause metabolic control through reductions in bodyweight 

of premature mortality. Thus, a combined mul- – obesity being present in the majority of patients – 

tifactorial therapeutic approach is required to max- and other lifestyle measures that help improve insuimise 

the impact of lifestyle and drug therapy on lin sensitivity. However, it is recognised that even if 

chronic micro- and macrovascular complications. diet and exercise advice is successfully implement- 

Since management of chronic vascular and neuro- ed, the majority of patients will require pharmacopathic 

complications accounts for the majority of logical therapy in the medium- to long term. Thus, 

health service spending for diabetes, such an ap- only 25% of patients in UKPDS maintained a HbA1c 

proach is likely to be cost effective. [7] The Steno-2 level


obese, sedentary middle-aged patients. Failure to diabetes. We consider both well established drugs 

respond rapidly (i.e. within a week or two) to an oral and recent additions to the armamentarium. For each 

agent in a patient thought to be complying with the class of agents we present an outline of the mode of 

dietary advice usually signals the need for early use action, pharmacokinetics, indications and contrainof 

insulin. If a partial response is observed, dose dications, efficacy, safety and tolerability, current 

escalation is followed by step-wise addition of com- place in management and future prospects, includplementary 

drugs (figure 1). Insulin is usually re- ing role in prevention of type 2 diabetes. We have 

served for patients: (i) who fail to respond adequate- grouped the drugs according to their principal mode 

ly to a combination of oral agents; (ii) in whom of action: (i) those that increase insulin secretion 

control deteriorates despite logical and adequate (insulin secretagogues); (ii) drugs delaying the rate 

drug combinations; or (iii) for whom safety and of digestion and absorption of carbohydrates (αefficacy 

considerations favour its use as the drug of glucosidase inhibitors); and (iii) those with direct 

choice, for example during pregnancy, or in patients effects on insulin-responsive tissues (insulin-senwith 

severe hepatic or renal impairment. [10] Several sitising agents). This sequence should not be taken 

classes of oral antidiabetic agents are currently to imply a hierarchy in terms of efficacy or merit. 

available, the range of options having enjoyed a The recognition that type 2 diabetes is usually a 

welcome expansion in recent years. However, the progressive disease implies that drug dosages will 

evidence base and clinical experience vary consider- need to be increased or therapy moved to another 

ably not only between classes but also between stage in the treatment algorithm. [2,4] 

drugs drawn from the same class. As a result, pre- 

While this article primarily reflects current pracscribing 

decisions often appear to be made on rather 

tice in the UK, we have endeavoured to provide a 

subjective grounds, such as familiarity with a particreview 

that acknowledges important differences in 

ular drug; this practice may help to explain notable 

prescribing in other countries. A word about moniregional 

differences in prescribing. 

toring: assessing the response to antidiabetic therapy 

In the remainder of this article we focus on involves periodic – generally 3- to 6-monthly – 

treatment of hyperglycaemia in patients with type 2 measurement of HbA 1c . This approach, which is 

Aim 

Diagnosis 

Procedure 

Diet, exercise, weight control 

and health education 

Relieve symptoms, 

improve glycaemic 

control, enhance 

quality of life 

Oral agent monotherapy: 

metformin, sulphonylurea, meglitinide, 

thiazolidinedione 1 , acarbose 

Oral agent combination therapy 

(using two different classes) 

Move to next stage 

if there is inadequate 

control of glycaemia 

or inadequate relief 

of symptoms 

Insulin or insulin plus an oral agent 

Fig. 1. An algorithm for the treatment of type 2 diabetes mellitus. The progressive hyperglycaemia in type 2 diabetes requires a steppedcare 

approach with treatment being modified and added over time. Rapid progression to the next stage is recommended if the glycaemic 

target is not achieved. Late introduction of combinations of oral antidiabetic agents is often a prelude to insulin treatment. 1 Note that in 

Europe, thiazolidinediones and nateglinide have limited licenses. The α-glucosidase inhibitor miglitol is also available in some countries 

(reproduced from Krentz and Bailey, [4] with permission from the Royal Society of Medicine Press). 



recommended in the UK, can be usefully comple- nylureas with respect to the risks of weight gain and 

mented by self measurement of capillary blood glu- hypoglycaemia. Compared with older sulphocose 

in selected, empowered patients and in particu- nylureas, glimepiride is relatively expensive and 

lar clinical scenarios, for example in patients in clinical outcome data are not available, as they are 

whom iatrogenic hypoglycaemia is a concern. for the agents used in the UKPDS. The clinical 

relevance of theoretical, but much debated, effects 

1. Insulin Secretagogues 

of glimepiride on ischaemic preconditioning – 

whereby a brief episode of ischaemia protects the 

myocardium against the detrimental effects of sub- 

1.1 Sulphonylureas sequent and more severe interruption of perfusion – 

remain uncertain. The issues of the importance of 

Sulphonylureas have been extensively used for ischaemic preconditioning and the possible influthe 

treatment of type 2 diabetes for nearly 50 years. ence of different sulphonylureas continue to be de- 

They lower blood glucose concentrations primarily bated (see section 1.1.5). [14] 

by stimulating insulin secretion from the β cells of 

the pancreatic islets. By the 1960s several sulphonylureas 

were available, including tolbutamide, 

1.1.1 Mode of Action 

acetohexamide, tolazamide and chlorpropamide, ofproducing 

Sulphonylureas have direct effects on the insulin- 

fering a range of pharmacokinetic options. Howsulphonylurea 

islet β cells. The drugs bind to the β-cell 

ever, doubts about safety were raised in the 1970s. A 

receptor (SUR)-1, part of a transever, 

large US multicentre trial of antidiabetic therapy, membrane complex with adenosine 5′-triphosphatethe 

UGDP (University Group Diabetes Program) [11] sensitive Kir 6.2 potassium channels (KATP chan- 

reported apparent detrimental cardiovascular effects nels). [14,15] Binding of the sulphonylurea closes these 

of tolbutamide. The UGDP was heavily criticised KATP channels; this reduces cellular potassium ef- 

for perceived methodological failings and its find- flux favouring membrane depolarisation. In turn, 

ings were far from being universally accepted. Sub- depolarisation opens voltage-dependent calcium 

sequent observational and randomised clinical stud- channels, resulting in an influx of calcium that actiies 

using sulphonylureas have provided mixed evi- vates calcium-dependent proteins that control the 

dence, but a review of the available literature release of insulin (figure 2). When sulphonylureas 

provides little in the way of convincing evidence of interact with SUR1 in the β-cell plasma membrane 

cardiovascular toxicity. [12] Indeed, some studies they cause prompt release of pre-formed insulin 

have reported a decreased incidence of cardio- granules adjacent to the plasma membrane – the sovascular 

events in subjects with lesser degrees of called ‘first phase’ of insulin release. [16] Sulphoglucose 

intolerance who received sulphony- nylureas also increase the extended (‘second phase’) 

lureas. [12] The UKPDS investigators did not find any of insulin release that begins approximately 10 minincrease 

in risk of myocardial infarction among pa- utes later as insulin granules are translocated to the 

tients treated with sulphonylureas compared with membrane from within the β cell. [17] The protracted 

patients randomised to insulin as monotherapy. [1] stimulation of the ‘second phase’ of insulin release 

The Steno-2 Study, [8] has already been mentioned. involves the secretion of newly formed insulin granules. 

A succession of more potent so-called secondwhile 

The increased release of insulin continues 

generation sulphonylureas emerged in the 1970s and 

there is ongoing drug stimulation, provided 

1980s, for example glibenclamide (glyburide), the β cells are fully functional. Sulphonylureas can 

gliclazide and glipizide. The latest, glimepiride, was cause hypoglycaemia since insulin release is initiat- 

introduced in the late 1990s. [13] Glimepiride is a ed even when glucose concentrations are below the 

once-daily drug for which claims have been made normal threshold for glucose-stimulated insulin re- 

that it might offer advantages over other sulpho- lease (approximately 5 mmol/L). 



Glucose 

GLUT2 

Succinate esters 

Glucokinase 

Glucose 

metabolism 

ATP 

SUR1 

Kir 6.2 

K ATP 

channel 

Sulphonylureas 

Repaglinide 

Nateglinide 

Proinsulin 

biosynthesis 

Depolarisation 

PKA 

Ca 2+ -sensitive 

proteins 

Ca 2+ 

channel 

GLP-1 

Exenatide 

Adrenergic 

receptors 

α 2 -adrenoceptor 

antagonists 

Receptors 

cAMP 

PDE 

inhibitors 

Insulin 

Exocytosis 

Insulin 

Fig. 2. The insulin-releasing effect of sulphonylureas and other agents on the pancreatic islet β cell. Sulphonylureas bind to the sulphonylurea 

receptor (SUR)-1 located within the plasma membrane. This closes Kir 6.2 potassium channels which reduces potassium efflux, 

depolarises the cell and opens voltage-dependent calcium influx channels. Raised intracellular calcium brings about insulin release. 

According to the stimulus-secretion model, metabolism of glucose generates adenosine 5′-triphosphate (ATP) leading to closure of 

potassium channels, permitting the normal β cell to link insulin secretion closely to glucose concentration. Sulphonylureas may also 

enhance nutrient-stimulated insulin secretion by other actions on the β cell. Other secretagogues, e.g. repaglinide, nateglinide, also 

stimulate insulin secretion via the SUR-Kir 6.2 complex. Other agents, e.g. phosphodiesterase (PDE) inhibitors, glucagon-like peptide 

(GLP)-1 (7–36 amide), act via cyclic adenosine monophosphate (cAMP) and protein kinase A (PKA) to promote proinsulin synthesis 

(reproduced from Krentz and Bailey, [4] with permission from the Royal Society of Medicine Press). GLUT2 = glucose transporter-2. 

1.1.2 Pharmacokinetics liver, although metabolites and their routes of elimination 

The principal distinguishing feature between different 

vary considerably between compounds. 

sulphonylureas relates to their pharmacokineproteins 

Since all sulphonylureas are highly bound to plasma 

tic characteristics (table III). Duration of action vardrugs 

they have the potential to interact with other 

ies from 24 hours for 

sharing this binding, for example salicylates, 

chlorpropamide because of differences in (i) rates of sulphonamides and warfarin; displacement from cir- 

metabolism; (ii) activity of metabolites; and (iii) culating proteins has been implicated in cases of 

rates of elimination. [18] These properties have im- severe sulphonylurea-induced hypoglycaemia (table 

portant implications for the risk of hypoglycaemia 

IV). 

associated with various sulphonylureas, an issue that 1.1.3 Indications and Contraindications 

is further complicated by retarded release prepara- Sulphonylureas remain a popular choice as firsttions 

of some compounds. All sulphonylureas are line oral therapy for patients with type 2 diabetes 

well absorbed and most reach peak plasma concen- who have not achieved or maintained adequate glytration 

in 2–4 hours. They are metabolised in the caemic control using nonpharmacological measures. 



Table III. Pharmacokinetic properties of sulphonylureas [19] 

Sulphonylureas Daily dosage (mg) Duration of action a Activity of metabolites Main route of 

elimination 

First generation 

Chlorpropamide b 100–500 Long Active Urine >90% 

Tolbutamide c 500–2000 Short Inactive Urine ≈100% 

Second generation 

Glibenclamide (glyburide) 2.5–15 Intermediate to long Active Bile ≈50% 

Glimepiride 1–6 Intermediate Active Urine ≈80% 

Glipizide 2.5–20 Short to intermediate Inactive Urine ≈70% 

Gliquidone 15–180 Short to intermediate Inactive Bile ≈95% 

Gliclazide 40–320 d Intermediate Inactive Urine ≈65% 

a 

b 

c 

d 

Long >24h; intermediate 12–24h; short


be monitored by periodic measurement of HbA1c (or nylurea therapy generally has modest effects on 

fructosamine if HbA1c is not available). 

blood lipid profiles, although some studies have 

noted a small decrease in plasma triglyceride levels 

1.1.4 Efficacy – possibly linked to improved glycaemic control – 

The blood glucose-lowering efficacy of sulpho- and minor increments in high-density lipoprotein 

nylureas has been evaluated in many retrospective (HDL)-cholesterol. When a sulphonylurea is used in 

and prospective studies, and from decades of collec- combination with another antidiabetic agent, the 

tive worldwide clinical experience. When used as glucose-lowering efficacy of the sulphonylurea is 

monotherapy in patients inadequately controlled by approximately additive to the effect of the other 

nonpharmacological measures, sulphonylureas can agent. Once again, response is crucially dependent 

be expected to reduce fasting plasma glucose by an on the presence of adequate β-cell function. Early 

average of 2–4 mmol/L accompanied by a decrease use of such combination therapy is indicated when 

in HbA1c of 1–2%. [4,19,21] However, individual re- optimal titration of a single agent does not achieve 

sponses are variable. Since the hypoglycaemic ef- adequate glycaemic control. 

fect of sulphonylureas is attributable to increased The combination of two different types of agents 

insulin secretion, the effectiveness of these drugs is is more likely to achieve glycaemic targets, albeit 

dependent on adequate β-cell function. The afore- for a variable period of time. If combination therapy 

mentioned progressive β-cell failure that determines is started at a stage when hyperglycaemia is already 

the natural history of type 2 diabetes may require an marked (after ‘failure’ of monotherapy), then β-cell 

increased dosage of sulphonylureas if glycaemic depletion is likely to be advanced. Under these 

control deteriorates. Rapid and uncontrollable dete- circumstances, oral combination therapy is likely to 

rioration of glycaemic control during sulphonylurea offer limited benefit and the need for an early move 

therapy is sometimes termed ‘secondary sulphony- to insulin treatment is usually clear. Since there are 

lurea failure’. This phenomenon, which is some- occasional exceptions to this rule, a limited trial of 

thing of a misnomer, occurs in approximately combination oral therapy may be worthwhile. How- 

5–10% of patients per annum with suggestions of ever, the temptation to procrastinate unduly on 

differences in ‘failure’ rates between some com- transferring the patient to insulin treatment should 

pounds. [21,22] The inability to maintain acceptable be firmly resisted, not least since some patients 

glycaemic control is common to all sulphonylureas derive rapid symptomatic benefit from insulin therand 

is held to reflect an advanced stage of β-cell apy. Impending metabolic decompensation, with or 

failure, that is, it is a reflection of disease progres- without ketosis, mandates immediate insulin treatsion 

rather than a true failure of therapy. Individuals ment; more severe degrees of decompensation, for 

who have greater degrees of β-cell reserve usually example obtundation, dehydration, ketosis-assorespond 

well to sulphonylureas; early use of sulpho- ciated vomiting, necessitates emergency hospitalisanylureas 

as first-line monotherapy in these patients tion for treatment with intravenous insulin, fluids 

will produce better blood glucose lowering than late and electrolytes. 

intervention in patients with severely compromised 

β-cell function. 

1.1.5 Adverse Events 

The plasma insulin concentrations achieved Hypoglycaemia, usually subclinical or minor but 

during sulphonylurea therapy do not usually extend occasionally life threatening, is the most common 

beyond the range observed in the general non-diabe- and potentially most serious adverse effect of sultic 

population (including those with impaired glu- phonylurea therapy. [23] Patients receiving sulphocose 

tolerance), and suggestions that sulphonylurea- nylureas should receive instruction on the recogniinduced 

hyperinsulinaemia might increase the risk tion and prevention of hypoglycaemia and the 

of detrimental insulin-induced effects on the cardio- prompt actions they must take should warning 

vascular system remain unsubstantiated. [12] Sulpho- symptoms develop. Severe protracted hypogly- 



caemia is more likely with longer-acting sulpho- patients receiving insulin therapy is orders of magninylureas 

such as glibenclamide, with tolbutamide tude higher. However, this does not detract from the 

holding the lowest place in the hierarchy of risk (see importance of sulphonylurea-induced hypoglyalso 

section 1.1.6). Individuals with irregular eating caemia. Minor recurrent hypoglycaemia should 

habits (see section 1.2.3) or excessive alcohol con- prompt a reassessment of the choice of agent and 

sumption are at higher risk of sulphonylurea-in- consideration of an alternative secretagogue, for exduced 

hypoglycaemia. As mentioned in section ample a rapid-acting insulin releaser (see section 

1.1.3, hypoglycaemia is also more likely to occur in 1.2). The treatment schedule, the possibility of drug 

patients with satisfactory glycaemic control, as indi- interactions (table IV) and relevant features of the 

cated by an HbA1c concentration within, or just patient’s lifestyle, such as diet, meal patterns and 

above, the non-diabetic reference range. These pa- alcohol use, should be reviewed. Severe episodes of 

tients should always be questioned directly about sulphonylurea-induced hypoglycaemia mandate imrecent 

symptoms of hypoglycaemia, although their mediate admission to hospital: treatment with a connonspecific 

nature can raise problems of over-diag- tinuous intravenous infusion of dextrose may be 

nosis; self-monitoring of capillary blood glucose required for several days. There is a tendency for 

concentrations during suggestive episodes should hypoglycaemia to recur shortly after initial resuscihelp 

to clarify this issue, although uncertainties may tation with intravenous dextrose; the patient should 

not be completely dispelled. If there is continuing not be prematurely discharged after emergency 

doubt, a temporary reduction in dose is usually treatment. Where accumulation of chlorpropamide 

indicated. Estimates of the incidence of mild hypo- is suspected, renal elimination may be enhanced by 

glycaemia, that is, not requiring assistance from forced alkaline diuresis. The vasodilator diazoxide 

another individual, are often based on symptoms 

and the somatostatin analogue octreotide [24] have 

which have not necessarily been confirmed by conbeen 

used successfully to reversibly inhibit insulin 

temporaneous self-measurement of capillary blood 

secretion in severe sulphonylurea-induced hypoglyglucose. 

In the UKPDS, for example, about 20% of 

caemia, thereby reducing intravenous dextrose resulphonylurea-treated 

patients reported one or more 

quirements. These drugs should be regarded as poepisodes 

suggestive of hypoglycaemia annually; 

tentially useful adjuncts to intravenous glucose in 

other studies have suggested similar rates. [23] The 

some patients; octreotide avoids the adverse haemotiming 

of hypoglycaemia tends to reflect the 

dynamic effects of diazoxide, an obsolete antihyperpharmacokinetics 

of the sulphonylurea. Thus, glitensive 

agent that may pose a hazard in the elderly 

benclamide has a propensity to cause inter-prandial 

patient with compromised cardiovascular reflexes. 

hypoglycaemia whereas chlorpropamide tends to 

induce hypoglycaemia in the pre-breakfast period. Other adverse events of sulphonylureas include 

More severe hypoglycaemia (i.e. requiring assis- uncommon sensitivity reactions – usually cutaneous 

tance) occurred in about 1% of sulphonylurea-treatrare. 

– that are usually transient; erythema multiforme is 

ed patients annually in the UKPDS. In general, 

Fever, jaundice and blood dyscrasias are very 

lower rates (approximately 0.2–2.5 episodes per rare; some sulphonylureas can reportedly precipitate 

1000 patient-years) have been reported from adheyday, 

acute porphyria in predisposed individuals. In its 

verse event reporting to regulatory authorities or 

chlorpropamide was notorious for causing 

from physician-completed questionnaires. The morquantities 

unpleasant facial flushing after consuming small 

tality risk from severe sulphonylurea-induced hyporeported. 

of alcohol; photosensitivity has also been 

glycaemia has been calculated to be 0.014–0.033 per 

Chlorpropamide could also increase renal 

1000 patient-years. [23] Predictably, longer-acting sensitivity to antidiuretic hormone, occasionally 

high-potency agents, such as glibenclamide, appear causing water retention with hyponatraemia. In con- 

to carry the greater mortality risk. For comparison, trast, glibenclamide is credited with a mild diuretic 

the occurrence of severe hypoglycaemia induced in action. Weight gain is regarded as a class effect of 



sulphonylurea therapy, typically amounting to held to be equivalent to 80mg of unmodified glicla- 

1–4kg and stabilising after about 6 months. This zide. In a recent 6-month comparative multicentre 

weight gain, which is always unwelcome, is thought study, gliclazide MR was associated with approxito 

reflect the anabolic effects of increased plasma mately 50% reduction in episodes of minor hypoglyinsulin 

concentrations; some studies have suggested caemia compared with glimepiride, at similar levels 

that reduced loss of calories as glucose in the urine of glycaemic control; no episodes of severe hypomay 

account for the majority of the weight glycaemia were observed with either agent in this 

gain. [19,21] study. [26] 

The saga of the questionable cardiovascular safety 

of the sulphonylureas was given a nudge by the 1.2 Rapid-Acting Prandial Insulin Releasers 

discovery that cardiac muscle and vascular smooth 

Under experimental conditions the first phase of 

muscle express isoforms of the SUR2A and SUR2B. 

glucose-stimulated insulin secretion is diminished 

Sulphonylureas that contain a benzamido group (gliearly 

in the natural history of type 2 diabetes. The 

benclamide, glipizide, glimepiride) can bind to 

prompt physiological rise in plasma insulin in res- 

SUR2A and SUR2B, [15] whereas those without (e.g. 

ponse to meals is attenuated and its peak delayed. 

tolbutamide, chlorpropamide and gliclazide) show An initial surge of insulin release appears to be 

very little interaction with the cardiac and vascular 

particularly important for effective postprandial 

SUR receptors. The effects of the KATP channel 

suppression of hepatic glucose production; failure to 

opener nicorandil (an anti-anginal drug with cardisuppress 

endogenous glucose production exaceroprotective 

properties) are blocked by sulphobates 

postprandial hyperglycaemia. Because postnylureas 

that have a benzamido group. [15] The clinprandial 

hyperglycaemia contributes to elevated 

ical implications of these observations remain to be 

HbA1c levels it is a logical therapeutic target. Rapiddetermined. 

Although very high concentrations of 

acting prandial insulin releasers are available that 

sulphonylureas can cause contraction of cardiac and 

stimulate rapid, but short-lived, insulin secrevascular 

muscle, this is regarded as being unlikely to 

tion. [27,28] These agents are taken orally immediately 

be clinically significant effect at therapeutic drug 

before a meal. Derivatives of meglitinide, such as 

concentrations. Nonetheless, on the basis of adverse 

repaglinide and the phenylalanine derivative nategclinical 

experiences in high-risk patients, some high 

linide, are promoted as ‘prandial glucose regulaprofile 

authorities continue to advocate that sulphotors’; 

in fact, fasting hyperglycaemia is also imnylurea 

use be kept to a minimum in patients with 

proved to a lesser extent, particularly with repagliovert 

coronary artery disease. [25] 

nide. Clinical experience with these agents remains 

limited in most countries; these drugs are appreci- 

1.1.6 New Formulations of Sulphonylureas 

ably more expensive than most sulphonylureas, the 

Alterations to the formulation of some sulpholatter 

also having the reassurance of outcome data 

nylureas have been undertaken to modify the durafrom 

the UKPDS. 

tion of action. [4] For example, a micronised formulation 

of glibenclamide is available in the US that 1.2.1 Mode of Action 

increases the rate of gastrointestinal absorption, Benzamido prandial insulin releasers bind to the 

thereby enabling an earlier onset of action. A longer- SUR1 in the plasma membrane of the β cell at a site 

acting (‘extended release’) formulation of glipizide distinct from the sulphonylurea binding site (figure 

has also been introduced. A new (‘modified release’ 2). Since the KATP channel is closed when either the 

[MR]) formulation of gliclazide was launched in benzamido binding site or the sulphonylurea bindsome 

countries in 2002. This formulation has been ing site on the SUR1 is bound with its respective 

designed to produce an initially rapid, followed by agonist, there is no advantage in giving a prandial 

steady release of the drug to enable once-daily dos- insulin releaser in addition to a sulphonylurea. Howage. 

For the MR formulation of gliclazide, 30mg is ever, drugs are also in development that promote β- 



cell proinsulin synthesis and act via signalling path- daily dosages is a potential disincentive. Repagliways 

distinct from the KATP channel (figure 2). The nide should ideally be taken about 15–30 minutes 

short half-life of repaglinide results in enhancement before a meal. Starting with a low dose, for example 

of the first-phase and early second-phase of insulin 0.5mg before each main meal, the effect on glysecretion 

that is less sustained than that observed caemic control is monitored and the dosage titrated 

with sulphonylureas. [27-29] Theoretical benefits on up every 2 weeks to a maximum of 4mg before each 

cardiovascular outcomes from preferentially target- main meal; if a meal is not consumed the correing 

the postprandial period remain to be con- sponding dose of repaglinide should be omitted. If 

firmed. [28,29] It is unclear whether postprandial glycaemic targets are not met, consider early introhyperglycaemia 

per se is detrimental to the vascular duction of combination therapy (e.g. with metforendothelium 

or whether closely associated metabol- min). Unlike some sulphonylureas and metformin, 

ic disturbances, for example dyslipidaemia, are re- repaglinide is suitable for patients with moderate 

sponsible. Thus, the mechanism of the association renal impairment, although careful upward dose 

between post-challenge hyperglycaemia and mor- titration and close monitoring is still recommended. 

tality observed in the multicentre DECODE (Diabe- In contrast with the US, the UK license for nateglites 

Epidemiology: Collaborative analysis Of Diag- nide currently restricts use to combination therapy 

nostic criteria in Europe) study is uncertain. [30] Ran- with metformin in patients who do not achieve glydomised 

trials that are currently in progress should caemic targets with the latter drug as monotherhelp 

clarify this issue. 

apy. [29] In the US, nateglinide may also be used as 

monotherapy or combined with a thiazolidinedione. 

1.2.2 Pharmacokinetics 

Nateglinide should be used with caution in patients 

Repaglinide is rapidly and almost completely 

with hepatic disease. 

absorbed after oral administration, with peak plasma 

concentrations achieved in about 1 hour. [27] The 

1.2.4 Efficacy 

drug is rapidly metabolised in the liver to inactive 

Repaglinide (0.5–4mg taken about 15–30 minmetabolites, 

which are mainly excreted in bile. 

utes before meals) results in dose-dependent in- 

When taken about 15 minutes before a meal, repagcreases 

in insulin secretion with reduced postprandilinide 

produces a prompt insulin-releasing effect, 

al hyperglycaemia; a lesser reduction in fasting 

which is limited to a period of about 3 hours, roughhyperglycaemia 

is also observed. Overall reductions 

ly coinciding with the duration of meal digestion. 

in HbA1c are similar in magnitude to those observed 

Nateglinide has a slightly faster onset and shorter 

with sulphonylureas, that is 1–2%. Combined with 

duration of action, its binding to target receptors 

metformin, nateglinide reduces HbA 1c by up to 

lasting only seconds. A 60mg dose of nateglinide 

1.5%. 

taken 20 minutes before an intravenous glucose 

[28,29] 

tolerance test restored first-phase insulin release and 

1.2.5 Adverse Events 

lowered glucose concentrations. [28,29] 

The overall incidence of hypoglycaemic episodes 

1.2.3 Indications and Contraindications is lower with repaglinide than with sulphonyureas. 

Repaglinide may be used as monotherapy in pa- Sensitivity reactions, usually transient, can occur. 

tients inadequately controlled by nonpharmacologi- Increased plasma levels of repaglinide have been 

cal measures. Suitable candidates for rapid-acting reported when co-administered with gemfibrozil. A 

insulin releasers include individuals with irregular small increase in bodyweight can be expected in 

lifestyles wherein meals are unpredictable or patients starting repaglinide as initial monotherapy, 

missed. The lower risk of hypoglycaemia associated but there may be little change in weight among 

with its use makes repaglinide an attractive option patients switched from a sulphonylurea. Nateglinide 

for some elderly patients, particularly if other agents appears to have little effect on bodyweight when 

are contraindicated. However, the need for multiple combined with metformin. [29] 



Intestinal 

lumen 

Brush 

border 

α-Amylase 

Starch 

Maltose 

maltotriose 

dextrins 

Sucrose 

α-Glucosidase 

inhibitor (acarbose, 

miglitol, voglibose) 

Enterocyte 

Microvillus 

α-Glucosidase 

enzymes 

Villus 

Fig. 3. α-Glucosidase inhibitors (e.g. acarbose) competitively inhibit the activity of α-glucosidase enzymes in the brush border of small 

intestinal enterocytes (reproduced from Krentz and Bailey, [4] with permission from the Royal Society of Medicine Press). 

2. α-Glucosidase Inhibitors carbohydrate digestion until further along the intestinal 

tract, in turn causing glucose absorption to be 

Inhibitors of intestinal α-glucosidase enzymes 

delayed. The α-glucosidase inhibitors should be takretard 

the rate of carbohydrate digestion, thereby 

en with meals containing digestible carbohydrates, 

providing an alternative means to reduce postprannot 

monosaccharides; these drugs generally do not 

dial hyperglycaemia. [31] Acarbose, the first α- 

significantly affect the absorption of glucose. Since 

glucosidase inhibitor to be marketed, was introα-glucosidase 

inhibitors move glucose absorption 

duced in the early 1990s. Recently, two additional 

more distally along the intestinal tract they alter 

agents, miglitol and voglibose, have been introglucose-dependent 

release of intestinal hormones 

duced in some countries. [4] The α-glucosidase inhibthat 

enhance nutrient-induced insulin secretion. Reitors 

do not cause weight gain, can reduce postpranlease 

of gastric inhibitory polypeptide, which occurs 

dial hyperinsulinaemia and have lowered plasma 

triglyceride concentrations in some studies. [31] Their mainly from the jejunal mucosa, may be reduced by 

good safety record is a further advantage, but limited α-glucosidase inhibitors, whereas glucagon-like 

gastrointestinal tolerability has substantially limited peptide-1 (7–36 amide) secretion (mostly from the 

their use. The relatively high cost of α-glucosidase ileal mucosa) is increased. Overall, α-glucosidase 

inhibitors is another consideration that has influthrough 

the attenuated rise in postprandial glucose 

inhibitors reduce postprandial insulin concentrations 

enced prescribing. In the UK, acarbose use remains 

low. 

levels. [31] 

2.1 Mode of Action 

The α-glucosidase inhibitors competitively inhibit 

the activity of α-glucosidase enzymes in the 

brush border of enterocytes lining the intestinal villi 

(figure 3). High affinity binding prevents these enzymes 

from cleaving their normal disaccharide and 

oligosaccharide substrates into monosaccharides 

prior to absorption. This defers the completion of 

2.2 Pharmacokinetics 

Acarbose is absorbed only to a trivial degree 

(


2.3 Indications and Contraindications breast-feeding are traditionally regarded to be contraindications 

for all oral antidiabetic drugs, mainly 

An α-glucosidase inhibitor may be used as because of a lack of safety data rather than evidence 

monotherapy for patients with type 2 diabetes that is of detrimental effects. 

inadequately controlled by nonpharmacological 

measures. Because α-glucosidase inhibitors target 2.4 Efficacy 

postprandial hyperglycaemia, they can be a useful 

An α-glucosidase inhibitor can reduce peak confirst-line 

treatment in patients who have a combinacentrations 

of blood glucose and reduce interprandition 

of only slightly raised basal glucose concentraal 

troughs. Used as monotherapy to patients who 

tions and more marked postprandial hyperglycomply 

appropriately with dietary advice, an α- 

caemia. A recent multicentre clinical trial (STOPglucosidase 

inhibitor will typically reduce postpran- 

NIDDM [Study TO Prevent NonInsulin-Dependent 

dial glucose concentrations by 1–4 mmol/L. The 

Diabetes Mellitus]) confirmed the utility of acarbose 

incremental area under the postprandial plasma gluin 

preventing the transition from impaired glucose 

cose curve can be more than halved in some individtolerance 

to diabetes [32] (see section 2.4). Acarbose 

uals. There seems to be a ‘carry-over’ effect that 

can be used in combination with other antidiabetic 

may produce a reduction in basal glycaemia up to 

agents. When starting therapy with an α-glucosidase 

1 mmol/L. The decrease in HbA1c is usually about 

inhibitor it is said to be important to ensure that the 

0.5–1.0%, provided that a high dose of the drug is 

patient is taking a diet rich in complex carbohytolerated 

and dietary compliance is maintained. 

drates, as opposed to simple sugars. Acarbose 

[33] 

There may be a trivial alteration in the gastrointestishould 

be taken with meals, starting with a low dose, 

nal absorption of other oral antidiabetic agents when 

for example 50 mg/day, and slowly titrating up over 

used in combination therapy. In general, the extra 

several weeks. Monitoring of glycaemic control, 

benefit to glycaemic control achieved by addition of 

particularly postprandially, may be helpful. The 

an α-glucosidase inhibitor to another antidiabetic 

postprandial action of these agents would not be 

agent is additive. In the recently published multicenexpected 

to induce hypoglycaemia, at least when 

tre STOP-NIDDM trial acarbose reduced the risk of 

they are used as monotherapy. The maximum dosprogression 

from impaired glucose tolerance to type 

age of α-glucosidase inhibitors may be limited by 

2 diabetes (relative hazard 0.75; 95% CI 0.63, 0.90; 

gastrointestinal symptoms; this is certainly our exp 

= 0.0015). [32] This study randomised 1429 patients 

perience with acarbose (see section 2.5). Intuitively, 

with impaired glucose tolerance to acarbose 100mg 

patients experiencing gastrointestinal adverse efthree 

times daily or placebo, of whom data were 

fects with metformin may not be the best candidates 

available for a modified intention-to-treat analysis 

in whom to add an α-glucosidase inhibitor. A hisin 

1368 patients. Glucose tolerance was determined 

tory of chronic intestinal disease serves as a – largeusing 

a 75g oral glucose tolerance test. Intriguingly, 

ly theoretical – contraindication to acarbose and 

new cases of hypertension and major cardiac events, 

other agents in this class. High dosages of acarbose 

including overt and clinically silent myocardial incan 

occasionally increase liver enzyme concentrafarction, 

were also reduced by acarbose therapy. 

tions, and it is recommended that transaminase con- 

[34] 

The latter were not primary endpoints of the study, a 

centrations are measured at intervals in patients relimitation 

acknowledged by the investigators. 

ceiving the maximum dosage (200mg three times 

[34] 

The results of ongoing trials using acarbose and 

daily in the UK, a dosage rarely attained in practice 

other agents in this class are awaited. 

for the aforementioned reasons). If liver enzymes 

[35] 

are raised, the dosage of acarbose should be reduced 

to a level at which normal enzyme concentrations 

2.5 Adverse Effects 

are re-established. Alternative causes of hepatic dys- The most common problems with α-glucosidase 

function should be considered. Pregnancy and inhibitors are gastrointestinal adverse effects. In the 



guanidine derivatives in the 1920s. These early 

antidiabetic agents were all but forgotten as insulin 

became widely available and it was not until the late 

1950s that three antidiabetic biguanides were report- 

ed: metformin, phenformin and buformin. Phen- 

formin was withdrawn in many countries in the 

1970s because of a high incidence of lactic acidosis; 

buformin received limited use in a few countries, 

leaving metformin as the main biguanide on a global 

basis. Metformin is the only biguanide available in 

the UK and, since 1995, the US. [23,40] Extensive 

clinical experience with metformin has been com- 

plemented by favourable results from the UKPDS. 

Metformin also enjoys the accolade of being among 

the least expensive of the oral antidiabetic agents. 

STOP-NIDDM trial 31% of acarbose-treated patients 

compared with 19% on placebo discontinued 

treatment early. [32] If the dosage is too high (relative 

to the amount of complex carbohydrate in the meal), 

undigested oligosaccharides pass into the large bowel. 

[23] Carbohydrates fermented by the flora of the 

large bowel cause flatulence, abdominal discomfort 

and sometimes diarrhoea. This is most likely to 

occur during the initial titration of the drug and can 

sometimes be minimised by slow titration and by 

ensuring dietary compliance with meals rich in complex 

carbohydrate. In some patients the gastrointestinal 

symptoms may gradually subside with time, 

suggesting an adaptive response within the gastrointestinal 

tract. Hypoglycaemia is only likely to be 

encountered when an α-glucosidase inhibitor is used 

in combination with a sulphonylurea or insulin. [23] 

No clinically significant drug interations have been 

reported. However, agents affecting gut motility can 

potentially influence the efficacy and gastrointesti- 

nal effects of acarbose; cholestyramine may increase 

the glucose-lowering effect of acarbose. 

3. Insulin Sensitisers 

Insulin resistance is a prominent metabolic defect 

in most patients with type 2 diabetes. [36,37] Defective 

insulin action is not confined to glucose metabolism, 

subtle defects also being demonstrable in the regulation 

of other aspects of intermediary metabolism 

(e.g. lipolysis), using appropriate investigative techniques. 

Many cross-sectional and prospective studies 

have implicated insulin resistance in the pathogenesis 

of type 2 diabetes and the related metabolic 

syndrome of cardiovascular risk. [38] Therefore, defective 

insulin action at target tissue level is an 

attractive therapeutic target in type 2 diabetes. [39] 

The biguanides and, in particular, the thiazolidinediones 

act directly against insulin resistance, and so 

are regarded as insulin sensitising drugs. 

3.1 Biguanides 

The finding that Galega officinalis (goat’s rue or 

French lilac), historically used as a traditional treatment 

for diabetes in Europe, was rich in guanidine 

led to the introduction of several glucose-lowering 


Metformin has a variety of metabolic effects, 

some of which may confer clinical benefits that 

extend beyond glucose lowering (table V). However, 

the molecular mechanisms of metformin have 

yet to be fully identified. At the cellular level, met- 

formin improves insulin sensitivity to some extent, 

an action mediated via post-receptor signalling pathways 

for insulin. [41,42] Recent data have suggested 

that adenosine 5′-monophosphate-activated protein 

Table V. Metabolic and vascular effects of metformin 

Anti-hyperglycaemic action 

suppresses hepatic glucose output 

increases insulin-mediated glucose utilisation 

decreases fatty acid oxidation 

increases splanchnic glucose turnover 

Weight stabilisation or reduction 

Improves lipid profile 

reduces hypertriglyceridaemia 

lowers plasma fatty acids and LDL-cholesterol; raises HDLcholesterol 

in some patients 

No risk of serious hypoglycaemia 

Counters insulin resistance 

decreases endogenous or exogenous insulin requirements 

reduces basal plasma insulin concentrations 

Vascular effects 

increased fibrinolysis 

decreases PAI-1 levels 

improved endothelial function 

HDL = high-density lipoprotein; LDL = low-density lipoprotein; 

PAI-1 = plasminogen activator inhibitor-1. 



Metformin 

Intestine 

Fat 

↑ Anaerobic glucose 

metabolism 

↑ Glucose uptake 

and oxidation 

↑ Lactate 

↓ Fatty acids 

↓ Glyconeogenesis 

↓ Glycogenesis 

↓ Oxidation of FA 

↑ Glucose 

uptake and 

oxidation 

↑ Glycogenesis 

↓ Oxidation 

of FA 

Liver 

Muscle 

↓ Hepatic glucose 

production 

↑ Insulin-mediated 

glucose disposal 

↓ Blood glucose 

concentration 

Fig. 4. Actions of metformin. Inhibition of hepatic glucose production is regarded as the principal mechanism through which metformin 

lowers blood glucose (reproduced from Krentz and Bailey, [4] with permission from the Royal Society of Medicine Press). FA = fatty acids; 

↑ indicates increase; ↓ indicates decrease. 

nant glucose-lowering mechanism of action of met- 

formin is to reduce excessive rates of hepatic glu- 

cose production. Metformin reduces gluconeogene- 

sis by increasing hepatic sensitivity to insulin (figure 

4) and decreasing the hepatic extraction of certain 

gluconeogenic substrates (e.g. lactate). Hepatic 

glycogenolysis is also decreased by metformin. In- 

sulin-stimulated glucose uptake in skeletal muscle is 

enhanced by metformin. This involves an increase 

in the movement of insulin-sensitive glucose trans- 

porter molecules to the cell membrane; an increase 

in the activity of the enzyme glycogen synthase 

promotes synthesis of glycogen. Metformin also 

kinase (AMPK) is a possible intracellular target of 

metformin. [43] Through phosphorylation of key proteins, 

AMPK acts as a regulator of glucose and lipid 

metabolism and cellular energy regulation. [44] Since 

metformin lowers blood glucose concentrations 

without causing overt hypoglycaemia it is most appropriately 

classed as an anti-hyperglycaemic – as 

distinct from hypoglycaemic – agent. The clinical 

efficacy of metformin in patients with type 2 diabetes 

requires the presence of insulin. The drug does 

not stimulate insulin release and a small decrease in 

fasting insulin concentrations is typically observed 

in patients with hyperinsulinaemia. [21] The predomi- 



acts in an insulin-independent manner to suppress other class of oral antidiabetic agent or with insulin. 

oxidation of fatty acids and to reduce triglyceride The drug is contraindicated in patients with imlevels 

in patients with hypertriglyceridaemia. [19] paired renal function (i.e. serum creatinine 

This reduces the energy supply for hepatic gluco- >120–130 µmol/L, depending on lean body mass), 

neogenesis and has favourable effects on the glu- as a precaution against drug accumulation. Cardiac 

cose-fatty acid (Randle) cycle (in which fatty acids or respiratory insufficiency, or any other condition 

are held to compete with glucose as a cellular energy 

predisposing to hypoxia or reduced perfusion (e.g. 

source). [37] Glucose metabolism in the splanchnic 

hypotension, septicaemia) are further contraindicabed 

is increased by metformin through insulin-indetions, 

as well as liver disease, alcohol abuse and a 

pendent mechanisms. This may contribute to the 

blood glucose-lowering effect of the drug, and in 

history of metabolic acidosis. Metformin can be 

turn may help to prevent gains in bodyweight. Coland 

other exclusions are not present. A difficulty in 

used in the elderly, provided that renal insufficiency 

lectively, the cellular effects of metformin serve to 

counter insulin resistance and to reduce the putative practice is that significant renal dysfunction may be 

toxic metabolic effects of hyperglycaemia (glucose present without the aforementioned elevation of se- 

toxicity) and fatty acids (lipotoxicity) in type 2 rum creatinine. 

diabetes. 

The improvement in insulin sensitivity can cause 

ovulation to resume in cases of anovulatory polycys- 


tic ovary syndrome (PCOS) [an unlicensed applica- 

Metformin is a stable hydrophilic biguanide that 

tion of the drug in the absence of diabetes]. [45] 

is quickly absorbed and eliminated unchanged in the 

Metformin should be taken with meals or immedurine. 

It is imperative that metformin is only preiately 

before meals to minimise possible gastrointesscribed 

to patients with renal function that is suffitinal 

adverse effects. Treatment should be started 

cient to avoid accumulation of the drug. Renal clearwith 

500 or 850mg once daily, or 500mg twice daily 

ance of metformin is achieved more by tubular 

secretion than glomerular filtration, the only signif- (one tablet with the morning and evening meals). 

icant drug interaction being competition with cime- The dosage is increased slowly – one tablet at a time 

tidine, which can increase plasma metformin conof 

– at intervals of about 2 weeks until the target level 

centrations. There is little binding of metformin to 

glycaemic control is attained. If the target is not 

plasma proteins. Metformin is not metabolised, and attained and an additional dose produces no greater 

so does not interfere with the metabolism of co- effect, return to the previous dose and, in the case of 

administered drugs. Metformin is widely distribut- monotherapy, consider combination therapy by aded, 

high concentrations being retained in the walls of ding in another agent (e.g. a sulphonylurea, prandial 

the gastrointestinal tract; this provides a reservoir insulin releaser or thiazolidinedione). The maximal 

from which plasma concentrations are maintained. 

effective dosage appears to be about 2000 mg/day, 

Nevertheless, peak plasma metformin concentragiven 

in divided doses with meals, the absolute 

tions are short-lived: in patients with normal renal 

maximum being 2550 or 3000 mg/day in different 

function the plasma half-life (t 1 /2) for metformin is 

countries. Several single tablet combinations of a 

2–5 hours, and almost 90% of an absorbed dosage is 

eliminated within 12 hours. [40] 

sulphonylurea (usually glibenclamide) with a bigua- 

nide (metformin or phenformin) have been available 

3.1.3 Indications and Contraindications 

in some European countries and elsewhere for more 

Metformin is the therapy of choice for overformin 

than a decade. A slow-release formulation of metweight 

and obese patients with type 2 diabetes. [42] It 

and a fixed-dose combination of metformin 

can be equally effective in normal weight patients. with glibenclamide is available in the US 

Metformin can also be used in combination with any (Glucovance ® , Bristol-Myers Squibb Company, 



Princeton, NJ, USA) 1 and elsewhere (although not secretion. Indeed, the reduction of basal insulin conin 

the UK). A combined rosiglitazone/metformin centrations, notably in hyperinsulinaemic patients, 

(Avandamet ® , GlaxoSmithKline, Philadelphia, PA, should itself improve insulin sensitivity by relieving 

USA) [see section 3.2] preparation is also available the insulin-induced downregulation of insulin recepin 

some parts of the world. 

tor number and suppression of post-receptor insulin 

During long-term treatment with metformin it is pathways. [35] Bodyweight tends to stabilise or deadvisable 

to check (e.g. annually) for the developprovements 

crease slightly during metformin therapy. Small imment 

of contraindications, particularly an elevated 

in the blood lipid profile may be observ- 

serum creatinine concentration (yearly measurement ed in hyperlipidaemic patients; plasma concentra- 

of creatinine clearance posing practical difficulties). tions of triglycerides, fatty acids and low-density 

Metformin can reduce gastrointestinal absorption of lipoprotein (LDL)-cholesterol tend to fall, whereas 

cyanocobalamin (vitamin B12). While anaemia is cardioprotective HDL-cholesterol tends to rise. 

very rare, an annual haemoglobin measurement is These effects appear to be independent of the antiprudent 

in patients at risk of nutritional deficiencies. hyperglycaemic effect, although a lowering of trig- 

It is advised to stop metformin treatment temporariinsulin 

lyceride and free fatty acids is likely to help improve 

ly during use of intravenous radiographic contrast 

sensitivity and benefit the glucose-fatty acid 

media, surgery and any other intercurrent situation cycle. 

in which the exclusion criteria could be invoked. [46] In the UKPDS, overweight patients who started 

Substitution with insulin may be appropriate at such oral antidiabetic therapy with metformin showed a 

times. Metformin alone is unlikely to cause serious statistically significant 39% reduced risk of myocarhypoglycaemia, 

but hypoglycaemia becomes an is- dial infarction compared with conventional treatsue 

when metformin is used in combination with an ment (p = 0.01). [47] No clear relationship is evident 

insulin-releasing agent or insulin. 

between metformin dosage and decreased coronary 

artery events. This suggests that patients who can 

3.1.4 Efficacy only tolerate a low dosage of metformin may benefit 

The long-term blood glucose-lowering efficacy from continuing the drug, even when other agents 

of metformin is broadly similar to sulphonylureas. have to be added to optimise glycaemic control. The 

As monotherapy in patients who are not adequately decrease in myocardial infarction observed with 

controlled on nonpharmacological therapy, optimal- metformin therapy in the UKPDS was not attributaly 

titrated metformin therapy typically reduces fast- ble to more effective lowering of HbA1c or major 

ing plasma glucose by 2–4 mmol/L, corresponding effects on classic cardiovascular risk factors such as 

to a decrease in HbA 1c by approximately 1–2%. [40] plasma lipids. Consequently, other potentially 

The effect is dependent upon the presence of some vasoprotective effects of metformin have been inendogenous 

β-cell function, and is largely indepen- voked. Reported benefits of metformin on non-clasdent 

of bodyweight, age and duration of diabetes. sic cardiovascular risk factors (table V) include in- 

However, given the progressive nature of type 2 creased fibrinolysis and a reduced concentration of 

diabetes, re-assessment of dosage and consideration the anti-thrombolytic factor plasminogen activator 

of additional therapy are required to maintain gly- inhibitor-1 (PAI-1). [41,46] The mechanism of the 

caemic control in the long term. [4,21] Metformin has cardioprotective effects of metformin remains unseveral 

features that mark it out as a good choice for certain. Detracting somewhat from this generally 

first-line monotherapy. The anti-hyperglycaemic ac- favourable view was evidence of an initially greater 

tion of metformin means that it is unlikely to cause mortality when metformin was added to a sulphonysevere 

hypoglycaemia. This may be explained in lurea in a UKPDS substudy, [47] but longer-term folpart 

because metformin does not stimulate insulin low-up has shown the benefits of metformin to be 

1 The use of trade names is for product identification purposes only and does not imply endorsement. 



sustained. [48] The explanation may have been, at 

least in part, a spuriously low mortality rate in the 

comparator sulphonylurea monotherapy group. [47,49] 

The small number of events in this substudy adds to 

the uncertainty. 

Sulphonylurea plus metformin is a commonly 

used combination and it would be reassuring to have 

definitive safety data. Since each class as monotherapy 

appears safe from the cardiovascular perspec- 

tive, alternative explanations have been postulated 

to explain similar findings seen in observational 

studies. [49] One plausible confounder might be great- 

er cardiovascular risk attributable to more severe 

metabolic derangements in patients treated with the 

combination. Results from US trials and various 

large databases of follow-up with sulphonylurea 

plus metformin combination therapy have been re- 

assuring. [21,49,50] Additional well designed compara- 

tive studies of appropriate statistical power would be 

required to quantify the risk to benefit equation for 

combination treatment with sulphonylurea plus met- 

formin. However, recent results from the 5-year 

follow-up of UKPDS – with no further attempt to 

continue in randomised groups – show that the adverse 

impact of sulphonylurea plus metformin com- 

bination seen initially is no longer evident. [48] At this 

point, the aforementioned benefits observed on mor- 

tality and cardiovascular disease in overweight patients 

initially randomised to metformin monother- 

apy, while diminished, remained significant. 

Consistent with the action of metformin on insulin 

sensitivity, addition of metformin to patients 

receiving insulin therapy may necessitate a reduction 

of insulin dosage. Some patients also show an 

improvement in glycaemic control, although this is 

not always impressive. Metformin reduces the 

weight gain associated with insulin therapy and, by 

decreasing the insulin dosage, there may be a decrease 

in hypoglycaemic episodes. The regimen has 

usually involved once-daily bedtime long-acting 

(lente) insulin or twice-daily insulin suspension isophane 

with metformin at mealtimes. In the US Diabetes 

Prevention Program, metformin reduced the 

incidence of new cases of diabetes in overweight 

and obese patients with impaired glucose tolerance 

by 33% overall. This compares with an intensive 

regimen of diet and exercise, which reduced the risk 

by 58%. [51] Younger, more obese individuals show- 

ed the most response to the preventive effects of 

metformin. 

3.1.5 Adverse Effects 

Abdominal discomfort and other gastrointestinal 

adverse effects, including diarrhoea, are encountered 

fairly commonly during the introduction of 

metformin. Symptoms may remit if the dose is re- 

duced and re-titrated slowly, but about 10% of patients 

cannot tolerate the drug at any dose. The most 

serious feared adverse event associated with metfor- 

min is lactic acidosis; the occurrence is rare (about 

0.03 cases per 1000 patient-years), but the mortality 

rate is high. [14,38] Since the background incidence of 

lactic acidosis amongst type 2 diabetic patients has 

not been established, it is possible that a proportion 

of cases that have been attributed to the drug were 

caused by other factors; this remains an area of 

controversy. Most cases of lactic acidosis in patients 

receiving metformin are due to inappropriate pre- 

scription of the drug. [23,40,46,52] The leading contraindication 

is renal insufficiency. [52] Metformin increases 

glycolysis to lactate, particularly in the 

splanchnic bed. The situation will be aggravated by 

any hypoxic condition or impaired liver function. [53] 

Hyperlactataemia occurs in cardiogenic shock and 

other illnesses that decrease tissue perfusion, and 

metformin is often only an incidental factor in these 

cases. [54] In the absence of reliable data to the con- 

trary, metformin treatment should be stopped im- 

mediately in all cases of suspected or proven lactic 

acidosis, regardless of cause. Lactic acidosis is typi- 

cally characterised by a raised blood lactate concen- 

tration (e.g. >5 mmol/L), decreased arterial pH and/ 

or bicarbonate concentration with an increased ani- 

on gap ([Na + ] – [Cl – + HCO3 – ] >15 mmol/L). 

Presenting symptoms are often nonspecific, but fre- 

quently include hyperventilation, malaise and abdominal 

discomfort. Treatment should be commenced 

immediately without waiting to determine whether 

metformin is a cause; bicarbonate remains the ther- 

apy of choice but evidence of its efficacy is scanty. 

The value of haemodialysis in removing accumulat- 



ed metformin has been challenged by some authorities, 

but dialysis may nonetheless be helpful in optimising 

fluid and electrolyte balance during treatment 

with high-dose intravenous bicarbonates. [54] 

3.2 Thiazolidinediones 


Rosiglitazone and pioglitazone are rapidly, and 

nearly completely absorbed (1–2 hours to peak con- 

centration), although absorption is slightly delayed 

when taken with food. Both agents are extensively 

metabolised by the liver. Rosiglitazone is metabol- 

ised mainly to very weakly active metabolites with 

lesser activity that are excreted predominantly in the 

urine. The metabolites of pioglitazone are more 

active and excreted mainly in the bile. Metabolism 

of rosiglitazone is undertaken mainly by cyto- 


Stimulation of PPARγ is regarded as the principal 

mechanism through which thiazolidinediones enhance 

insulin sensitivity. PPARγ is expressed at 

highest levels in adipose tissue, and less so in 

muscle and liver. PPARγ operates in association 

with the retinoid X receptor. The resulting heterodimer 

binds to nuclear response elements, thereby 

modulating transcription of a range of insulin-sensitive 

genes, in the presence of necessary cofactors 

(figure 4). [55,57] Many of the genes activated or suppressed 

by thiazolidinediones are involved in lipid 

and carbohydrate metabolism (table VI). Stimulation 

of PPARγ by a thiazolidinedione promotes 

differentiation of pre-adipocytes with accompanying 

lipogenesis, effects that promote or enhance the 

local effects of insulin. Thiazolidinediones increase 

Table VI. Metabolic effects of thiazolidinediones [55] 

Adipose tissue Muscle Liver 

↑ Glucose uptake ↑ Glucose uptake ↓ Gluconeogenesis 

↑ Fatty acid uptake ↑ Glycolysis ↓ Glycogenolysis 

↑ Lipogenesis ↑ Glucose oxidation ↑ Lipogenesis 

↑ Pre-adipocyte ↑ Glycogenesis a ↑ Glucose uptake a 

differentiation 

a Inconsistent findings. 

↑ indicates increase; ↓ indicates decrease. 

Thiazolidinediones improve whole-body insulin 

sensitivity via multiple actions on gene regulation. 

These effects result from stimulation of a nuclear 

receptor peroxisome proliferator-activated receptorγ 

(PPARγ), for which thiazolidinediones are potent 

synthetic agonists. [55] The antidiabetic activity of 

thiazolidinediones was described in the early 1980s, 

troglitazone being the first of the class to become 

available for clinical use. Troglitazone was introduced 

in the US in 1997, only to be withdrawn in 

2000 because of cases of idiosyncratic hepato- 

toxicity resulting in fatalities. Troglitazone was 

available in the UK for only for a few weeks in 1997 

before being withdrawn by its distributor as reports 

of hepatotoxicity accumulated in other countries. To 

date, two other thiazolidinediones, rosiglitazone and 

pioglitazone, have not shown the hepatotoxicity that 

led to the demise of troglitazone. Rosiglitazone and 

pioglitazone were introduced in the US in 1999 and 

in Europe in 2000. [56] Combination preparations 

(e.g. thiazolidinedione plus metformin) are also 

available. 

glucose uptake via glucose transporter-4 in skeletal 

muscle, and some reports indicate that rates of glu- 

coneogenesis in the liver are reduced. Stimulation of 

lipogenesis via PPARγ reduces circulating non-es- 

terified fatty acid (NEFA) concentrations through 

cellular uptake and triglyceride synthesis (figure 5). 

The reduction in plasma NEFA concentrations is 

associated with increased glucose utilisation and 

reducing gluconeogenesis by reducing operation of 

the glucose-fatty acid cycle; reductions in ectopic 

lipid deposition in muscle and liver may contribute 

to the improvements on glucose metabolism. Thia- 

zolidinediones also reduce the production and ac- 

tivity of the adipocyte-derived cytokine tumour ne- 

crosis factor (TNF)-α. [55] The latter has been impli- 

cated in the development of impaired insulin action 

in muscle, [58] although the precise role of TNFα in 

human states of insulin resistance remains unclear. 

Reductions in plasma insulin concentrations and 

lowering of circulating triglycerides are additional 

indirect mechanisms that may help to improve 

whole-body insulin sensitivity. Thiazolidinediones, 

like metformin, are anti-hyperglycaemic agents and 

require the presence of sufficient insulin to generate 

a significant blood glucose-lowering effect. 



chrome P450 (CYP) 2C8, which is not a widely 

activated isoform of CYP. [59] Thus, rosiglitazone 

does not interfere with the metabolism of other 

drugs. Pioglitazone is metabolised in part by 

CYP3A4 but, to date, no clinically significant reductions 

in plasma concentrations of other drugs (e.g. 

oral contraceptives) has been reported. Although 

both thiazolidinediones are almost completely 

bound to plasma proteins, their concentrations are 

low and have not been reported to interfere with 

other protein-bound drugs. 

3.2.3 Indications and Contraindications 

In the US, rosiglitazone and pioglitazone are 

available for use as monotherapy in non-obese and 

obese patients with type 2 diabetes in whom diabe- 

tes is not adequately controlled by nonpharmacological 

measures. They can also be used in combina- 

tion with various other antidiabetic drugs and in 

combination with insulin. In Europe, rosiglitazone 

and pioglitazone can be used as monotherapy if the 

patient is contraindicated for or intolerant of metfor- 

min. Thiazolidinediones can be used in combination 

with metformin or a sulphonylurea. In Europe, combination 

with insulin remains a contraindication to 

thiazolidinediones. [60] Substituting a thiazolidine- 

dione for either a sulphonylurea or metformin in 

patients with inadequate glycaemic control is gener- 

ally of limited value and risks a temporary deterioration 

in glycaemic control because of the slow onset 

of action of thiazolidinediones. Having been disap- 

pointed with this experience, some UK diabetolo- 

gists have elected to use thiazolidinediones in com- 

bination with both a sulphonylurea and metfor- 

min. [60] The former strategy has met with variable 

success: some patients respond well, others show 

little response, requiring transfer to insulin. The 

combination of thiazolidinedione plus insulin can 

improve glycaemic control while reducing insulin 

dosages in obese patients, although peripheral oedema 

has been reported. [61] 

The main cautions to using thiazolidinediones are 

listed in table VII. Rosiglitazone and pioglitazone 

can cause fluid retention with increased plasma vol- 

ume, a reduced haematocrit and a decrease in 

haemoglobin concentration. Therefore, the risk of 

oedema and anaemia should be taken into account, 

and in Europe, use of thiazolidinediones in patients 

with any evidence of congestive heart disease or 

Thiazolidinedione Glucose Fatty acids 

GLUT-4 

FATP 

Glucose uptake 

and utilisation 

aP2, acyl- 

CoA synthase 

PPARγ 

RXR 

Lipogenesis 

and adipocyte 

differentiation 

Transcription of certain 

insulin-sensitive genes 

Lipoprotein 

lipase 

↑ Hydrolysis of 

circulating triglycerides 

in chylomicrons and VLDL 

Adipocyte 

Fig. 5. Mechanism of action of a thiazolidinedione on an adipocyte (reproduced from Krentz and Bailey, [4] with permission from the Royal 

Society of Medicine Press). aP2 = adipocyte fatty acid binding protein; CoA = coenzyme A; FATP = fatty acid transporter protein; GLUT-4 = 

glucose transporter-4; PPARγ = peroxisome proliferator-activated receptor-γ; RXR = retinoid X receptor; VLDL = very low-density lipoproteins; 

↑ indicates increase. 



Table VII. Cautions in the use of thiazolidinediones 

Active liver disease 

This remains a contraindication to the use of thiazolidinediones 

even though neither rosiglitazone nor pioglitazone have been 

associated with troglitazone-like hepatotoxicity. In fact, the latter 

drugs are under investigation as a potential treatment for nonalcoholic 

steatohepatitis. In 2004, the US FDA recommendation 

for 2-monthly monitoring of biochemical liver function tests was 

relaxed. Instead, periodic biochemical monitoring is now left to 

the supervising clinician’s discretion 

Heart failure 

The precise contraindications differ between countries. In Europe, 

current heart failure or a history of heart failure are 

contraindications to thiazolidinediones 

Insulin treatment 

Although rosiglitazone and pioglitazone are licensed in the US for 

use in combination with insulin, caution is required. Concerns 

about higher rates of heart failure underlie this concern. The 

European Agency for the Evaluation of Medicinal Products 

considers insulin therapy a contraindication to the use of 

thiazolidinediones 

Pregnancy and breast-feeding 

Thiazolidinediones are classified as pregnancy category C 

because of growth retardation in mid-to-late gestation in animal 

models. These drugs should only be used during pregnancy if the 

potential benefit justifies the potential risk to the fetus 

Polycystic ovary syndrome 

Thiazolidinediones can cause ovulation to recommence in women 

with hyperandrogenism and chronic anovulation; risk of 

pregnancy 

According to the EU license, rosiglitazone can be 

given at a dosage of 4 mg/day in combination with a 

sulphonylurea, increasing to 8 mg/day (either once 

daily or in divided doses) in combination with met- 

formin. Pioglitazone can be given as a once-daily 

dosage of 15mg, increasing to 30mg if necessary 

(maximum 45mg in the US and Europe). The thera- 

peutic response varies markedly between patients 

and it can be difficult to predict those most likely to 

respond. If no effect is observed after 3 months it is 

appropriate to consider the patient as a nonresponder 

and to stop the treatment. Rosiglitazone and piog- 

litazone can be used in the elderly, provided there 

are no contraindications. Both drugs may be used in 

patients with mild-to-moderate renal impairment, 

although the potential for oedema is a concern. In 

women with anovulatory PCOS the improvement in 

insulin sensitivity may cause ovulation to resume 

during thiazolidinedione therapy. A combination 

heart failure is contraindicated. The choice of which 

patients to exclude on the basis of cardiac status 

varies between the product labelling sheets in the 

US and Europe. Consensus guidelines from the 

American Heart Association and the American Diabetes 

Association have recently been published. [62] 

Patients treated with a combination of insulin plus 

thiazolidinedione appear to be at highest risk of 

oedema, although the absolute rate of cardiac failure 

is low despite the fact the diabetes is a major risk 

factor for this complication. [62] The guidelines urge a 

cautious approach and careful clinical monitoring, 

especially for patients likely to be at higher risk of 

cardiac failure. The haemogloblin concentration 

should be checked before starting a thiazolidinedione, 

bearing in mind that reductions of up to 1 g/ 

dL in haemoglobin concentration may occur during 

therapy. No adverse effects on blood pressure have 

been noted with the thiazolidinediones, even with 

the increase in plasma volume; on the contrary, there 

is some evidence for a modest blood pressure-lowering 

effect. [63] 

As a precautionary measure, liver function 

should be assessed by measuring serum ALT before 

starting therapy and subsequently at 2-monthly intervals 

(or, in the US, as judged necessary by the 

prescribing clinician) during the first year of treatment; 

thereafter, periodic monitoring of liver function 

is prudent. Pre-existing liver disease, the development 

of clinical hepatic dysfunction or elevated 

ALT levels >2.5 times the upper limit for the laboratory 

serve as contraindications to thiazolidinediones. 

However, as mentioned earlier, hepatotoxicity 

has not been a concern with either rosiglitazone or 

pioglitazone. Isolated cases of nonfatal hepatocellular 

damage have been reported; however, the issue is 

clouded by reports suggesting an intrinsically higher 

risk of liver failure in patients with type 2 diabetes. 

Nevertheless, precautionary monitoring of liver 

function remains advisable. When initiating therapy 

with rosiglitazone or pioglitazone, blood glucose 

monitoring and titration of drug dosage should be 

undertaken while bearing in mind that thiazolidinediones 

exert a slowly generated anti-hyperglycaemic 

effect that usually requires 2–3 months to 

reach maximum effect. 



preparation containing rosiglitazone plus metformin subcutaneous depots increase as new small, insulin- 

(Avandamet ® ; combining rosiglitazone/metformin sensitive adipocytes are formed. There are proviin 

strengths 1mg/500mg, 2mg/500mg, 4mg/500mg, sional data to suggest that thiazolidinediones exert a 

2mg/1000mg, although not all strengths are avail- range of effects on aspects of the metabolic synable 

in all countries). 

drome that might reduce the risk of atherosclerotic 

cardiovascular disease. [63,65] For example, thiazoli- 

3.2.4 Efficacy dinediones have been reported to downregulate 

Addition of rosiglitazone or pioglitazone to the PAI-1 expression. Thiazolidinediones have also 

treatment schedule of patients whose glycaemic been reported to decrease urinary albumin excretion 

control with a sulphonylurea or metformin is subop- to a greater extent than expected for the improvetimal 

has consistently resulted in significant reduc- ment in glycaemic control and to reduce circulating 

tions in HbA1c. As judged by the available literature, markers of chronic low-grade inflammation. 

these agents have similar glucose-lowering effects, Preclinical studies suggesting that treatment of glureducing 

HbA1c by around 0.5–1.5%. [64] However, cose-intolerant animals with a thiazolidinedione 

the participants in these clinical trials had known preserved β-cell function have yet to be confirmed 

diabetes of several years’ duration, the effects of in human studies. In insulin-resistant women with a 

thiazolidinediones being more apparent when β-cell history of gestational diabetes at high risk of type 2 

function is less impaired. While earlier use of thia- diabetes troglitazone reduced the incidence of newzolidinediones 

may be advantageous, the longer- onset diabetes. [66] Whether thiazolidinediones will 

term picture requires clarification. Estimates of in- prove more effective than conventional antidiabetic 

sulin sensitivity and β-cell function (based on ana- agents in reducing the decline in β-cell function in 

lysis of fasting glucose and insulin concentrations) patients with established type 2 diabetes remains to 

have indicated that both defects can be improved by be determined, although preliminary data in patients 

the addition of a thiazolidinedione. [64] The effects on who respond to the drugs have been encouraging. [67] 

plasma lipids and apoproteins have been the subject Also of considerable interest are the clinical impliof 

debate. Rosiglitazone can cause a small rise in the cations of the aforementioned effects of thiazoliditotal 

cholesterol concentration, which stabilises nediones on risk factors for cardiovascular disease. 

within about 3 months. This is accounted for by a These effects, allied to direct anti-atherogenic acrise 

in both the LDL-cholesterol and the HDL-cho- tions reported in animal studies, are presently being 

lesterol, leaving the LDL : HDL-cholesterol ratio studied in clinical trials with cardiovascular endand 

the total : HDL-cholesterol ratio little changed points. [68] 

or slightly raised. Pioglitazone generally appears to 

have little effect on total cholesterol, and has been 3.2.5 Adverse Effects 

shown to reduce triglyceride concentrations in sev- Rosiglitazone and pioglitazone are generally well 

eral studies. Both thiazolidinediones reduce the pro- tolerated. As noted in section 3.2.3, caution is adportion 

of the smaller, more dense (more atherogen- vised in heart disease; in the UK this includes a 

ic) LDL particles. [64] To date, no prospective com- history of cardiac failure, oedema, anaemia and liver 

parative studies of the two drugs have been reported function requiring intermittent monitoring in accorand 

the clinical implications of these changes are dance with the package labelling. If contraindicauncertain. 

[63] 

tions arise during treatment, monitoring should be 

Weight gain, similar in magnitude to sulphonylu- intensified and, if necessary, treatment discontinrea 

therapy (typically 1–4kg) and stabilising over ued. Hypoglycaemia may occur several weeks after 

6–12 months, has been observed during thiazo- adding a thiazolidinedione to a sulphonylurea; selflidinedione 

therapy. There is some evidence that the monitoring of blood glucose can be helpful in identidistribution 

of body fat is altered such that visceral fying the point at which the dosage of the sulphonyadipose 

depots are little changed or reduced, while lurea should be reduced. Since PPARγ is expressed 



by many tissues, albeit at a low level, we must await 

the verdict of time for any unforeseen effects of 

long-term stimulation with thiazolidinediones. For 

example, PPARγ activation in macrophages can reduce 

the production of some inflammatory cytokines 

and might increase transformation of monocytes 

to macrophages in the vascular wall. Stimulation 

of PPARγ in colon cells has been variously 

reported to increase and decrease division and 

differentiation of these cells in different animals and 

cell models; [69] thus, familial polyposis coli is a 

contraindication to thiazolidinediones on theoretical 

grounds. 

4. Summary and Conclusion 

The management of patients with type 2 diabetes 

has been given a firm evidence base in recent years 

through the results of randomised clinical trials, 

notably the UKPDS. An improved understanding of 

the pathogenesis and natural history of this complex 

metabolic disorder has facilitated the application of 

new therapeutic agents. Attainment and maintenance 

of near-normal glycaemic control, while 

minimising the risk of iatrogenic hypoglycaemia, is 

a central long-term objective of therapy; however, 

this is often difficult to achieve in practice. 

The following general principles should be ap- 

plied while using oral antidiabetic drugs. 

• Antidiabetic drug therapy must be considered 

carefully within the context of the overall care 

plan. This includes an assessment of which agent 

is most likely to achieve the therapeutic goals of 

the care plan, taking account of the accompanying 

medical and lifestyle circumstances and commitments 

of the patient. 

• Always check for contraindications. 

• For some classes of agents, e.g. sulphonylureas, 

duration of action and route of elimination will be 

important considerations if hypoglycaemia is 

likely, or if renal or liver disease raises concerns. 

Shorter-acting preparations are preferred for 

those at risk of hypoglycaemia and in the elderly. 

• Start with the lowest recommended dose and 

monitor response taking the mode of action into 

account. Sulphonylureas generally produce a rap- 

id improvement in glycaemic control (within 

days), whereas the maximal response to thiazoli- 

dinediones may take several weeks to become 

apparent. Maximal glucose-lowering effects are 

usually obtained at doses lower than the manu- 

facturer’s recommendations, e.g. 5–10 mg/day 

for glibenclamide. 

• If the glycaemic target is not achieved consider 

adding another class of agent at an early stage. 

Undertake the same evaluation and titration pro- 

cedure for the second agent. If a combination of 

two oral agents does not give adequate control, 

there may be some patients who will benefit from 

addition of a third differently acting oral therapy. 

Compliance generally deteriorates as the daily 

number of doses increases. 

• Inability to achieve adequate glycaemic control 

with a logical combination of oral therapies is 

likely to indicate that the natural history of the 

disease has progressed to a state of severe β-cell 

failure. In this situation it is usually necessary to 

switch to insulin therapy. Similarly, failure to 

respond to an oral agent (so-called primary fail- 

ure) or loss of control (secondary failure) usually 

reflects a severe degree of insulin deficiency and 

early need for insulin. All oral antidiabetic agents 

are contraindicated in type 1 diabetes and in 

major metabolic decompensation. Insulin may be 

required temporarily during intercurrent severe 

illness. 

Clinicians have a greater range of antidiabetic 

treatments to choose from than ever, but this has 

brought a new level of complexity to management. 

In addition, polypharmacy has become the norm for 

many patients with type 2 diabetes in recognition of 

the importance of treating hypertension and dyslipidaemia, 

both commonly encountered and modi- 

fiable cardiovascular risk factors. The main classes 

of oral antidiabetic drugs are broadly similar in their 

glucose-lowering capacity, at least in the short- to 

medium term. [70] Accordingly, the most appropriate 

therapy should be selected according to the clinical 

and biochemical characteristics of the patient, safety 

considerations always being a major consideration. 

The UKPDS has influenced prescribing in the UK 



On theoretical grounds, the thiazolidinediones 

appear promising, particularly with respect to poss- 

ible preservation of β-cell function and the potential 

for cardiovascular disease prevention. However, 

these agents have perhaps not entirely fulfilled early 

expectations of success, at least if judged in terms of 

their glucose-lowering abilities, which are no better 

than the conventional drugs. Part of this shortfall 

may be attributable to the complexity and hetero- 

Table VIII. Relative costs and frequency of prescriptions for oral 

antidiabetic drugs in the UK [71,72]a 

Drug 

Frequency of prescriptions 

Relatively inexpensive 

Biguanides 

Only metformin b in UK 

Sulphonylureas c 

Moderate 

α-Glucosidase inhibitors 

Sulphonylureas c 

Only acarbose in UK 

Relatively expensive 

Rapid-acting prandial insulin Repaglinide, nateglinide 

releasers 

Thiazolidinediones 

Rosiglitazone, pioglitazone 

a This classification attempts to take average effective 

maintenance dosages into account and may be regarded as 

an approximate guide to relative UK Drug Tariff prices. 

b Use of metformin has increased in recent years, this drug 

now being the most widely prescribed oral antidiabetic agent 

in the UK (49%); sulphonylureas lie close behind, gliclazide 

being the most popular agent accounting for 31% of spending 

on oral antidiabetic agents. Thiazolidinediones account for 

only 5% of all prescriptions but for 32% of the cost of all oral 

antidiabetic agents in the UK; these figures predate the 2003 

license amendment permitting limited prescription of 

thiazolidinediones as monotherapy in selected patients. 

Acarbose, the only α-glucosidase inhibitor in the UK, and the 

rapid-acting prandial insulin releasers repaglinide and 

nateglinide account for a small percentage (


References 

23. Krentz AJ, Ferner RE, Bailey CJ. Comparative tolerability 

profiles of oral antidiabetic agents. Drug Saf 1994; 11: 223-41 

1. UK Prospective Diabetes Study Group. Intensive blood-glucose 

control with sulphonylureas or insulin compared with convensevere 

refractory sulfonylurea-induced hypoglycemia with oc- 

24. Krentz AJ, Boyle PJ, Justice K, et al. Successful treatment of 

tional treatment and risk of complications in patients with type 

2 diabetes (UKPDS 33). Lancet 1998; 352: 837-53 

treotide. Diabetes Care 1993; 16: 184-6 

2. European Diabetes Policy Group. A desktop guide to type 2 

25. Wilson SH, Kennedy FP, Garratt KN. Optimisation of the 

diabetes mellitus. Diabetic Med 1999; 16: 716-30 

management of patients with coronary heart disease and type 2 

diabetes mellitus. Drugs Aging 2001; 18: 325-33 

3. Tuomilheto J. Controlling glucose and blood pressure in type 2 

diabetes. BMJ 2000; 321: 394-6 

26. Schernthaner G, Grimaldi A, Di Mario U, et al. GUIDE study: 

double-blind comparison of once-daily gliclazide MR and 

4. Krentz AJ, Bailey CJ. Type 2 diabetes in practice. London: 

glimepiride in type 2 diabetic patients. Eur J Clin Invest 2004; 

Royal Society of Medicine Press, 2001 

34: 535-42 

5. Adler AI. Cardiovascular risk reduction in diabetes: underemphasized 

and overdue: messages from major trials. Clin 

27. Landgraf R. Meglitinide analogues in the treatment of type 2 

Med 2001; 1: 472-7 

diabetes mellitus. Drugs Aging 2000; 17 (5): 411-25 

6. Beckman JA, Creager MA, Libby P. Diabetes and atherosclero- 

28. Dornhorst A. Insulotropic meglitinide analogues. Lancet 2001; 

sis: epidemiology, pathophysiology and management. JAMA 

358: 1709-15 

2002; 287: 2570-81 

29. Davies M. Nateglinide: better post-prandial glucose control. 

7. Gray A, Clarke P, Farmer A, et al. Implementing intensive 

Prescriber 2002; 13: 17-27 

control of blood glucose concentration and blood pressure in 30. Qiao Q, Tuomilehto J, Borch-Johnsen K. Post-challenge hypertype 

2 diabetes in England: cost analysis. United Kingdom glycaemia is associated with premature death and macrovascu- 

Prospective Diabetes Study (UKPDS) Group. BMJ 2002; 325: lar complications. Diabetologia 2002; 46 Suppl. 1: M17-21 

860-3 

31. Lebovitz HE. α-Glucosidase inhibitors as agents in the treat- 

8. Gaede P, Vedel P, Larsen N, et al. Multifactorial intervention ment of diabetes. Diabetes Revs 1998; 6: 132-45 

and cardiovascular disease in patients with type 2 diabetes. N 32. Chiasson JL, Josse RG, Gomis R, et al. Acarbose for the 

Engl J Med 2003; 348: 383-93 

prevention of diabetes mellitus: the STOP-NIDDM random- 

9. Turner RC, Cull CA, Frighi V, et al. Glycemic control with diet, ised trial. STOP-NIDDM Trial Research Group. Lancet 2002; 

sulfonylurea, metformin, or insulin in patients with type 2 359: 2072-7 

diabetes mellitus: progressive requirements for multiple theratrial 

33. Holman RR, Cull CA, Turner RC. A randomised double-blind 

pies. JAMA 1999; 281: 2005-12 

of acarbose in type 2 diabetes shows improved glycemic 

10. Evans AJ, Krentz AJ. Benefits and risks of transfer from oral control over 3 years (UK Prospective Diabetes Study 44). 

antidiabetic agents to insulin in type 2 diabetes. In: Krentz AJ, Diabetes Care 1999; 22: 960-4 

editor. Drug treatment of type 2 diabetes. Auckland: Adis 34. Chiasson J-L, Josse RG, Gomis R, et al. Acarbose treatment and 

Books, 2000: 85-101 

the risk of cardiovascular disease and hypertension in patients 

11. Seltzer HS. A summary of criticisms of the findings and conclu- with impaired glucose tolerance: the STOP-NIDDM trial. The 

sions of the University Group Diabetes Program. Diabetes STOP-NIDDM Trial Research Group. JAMA 2003; 290: 

1972; 21: 976-9 

486-94 

12. Krentz AJ. Sulfonylureas in the prevention of vascular compliprevention 

35. Scheen A. Is there a role for alpha-glucosidase inhibitors in the 

cations: from UKPDS to the ADVANCE study. Proceedings 

of type 2 diabetes mellitus? Drugs 2003; 63 (10): 

of the VIIIth European symposium on metabolism. Amster- 933-51 

dam: Elsevier Science International Congress Series 1253; 36. Reaven GM. Role of insulin resistance in human disease. Diabe- 

2003: 261-77 

tes 1988; 37: 1595-607 

13. Evans AJ, Krentz AJ. Glimepiride: a new sulphonylurea. Prescriber 

37. Krentz AJ. Insulin resistance. Oxford: Blackwell Science, 2002 

1999; 10: 51-8 

38. Ginsberg HN. Insulin resistance and cardiovascular disease. J 

14. Ashcroft FM, Gribble FM. ATP-sensitive K+ channels and Clin Invest 2000; 106: 453-8 

insulin secretion: their role in health and disease. Diabetologia 39. Campbell IW. Antidiabetic drugs past and future: will improv- 

1999; 42: 903-19 ing insulin resistance benefit cardiovascular risk in type 2 

15. Gribble FM, Reimann F. Pharmacological modulation of KATP diabetes mellitus? Drugs 2000; 60 (5): 1017-28 

channels. Biochem Soc Trans 2002; 30: 333-9 40. Bailey CJ, Turner RC. Metformin. N Engl J Med 1996; 334: 

16. Rorsman P, Renstrom E. Insulin granule dynamics in pancreatic 574-9 

beta cells. Diabetologica 2003; 46: 1029-45 

41. Kirpichnikov D, McFarlane SI, Sowers JR. Metformin: an up- 

17. Groop LC. Sulfonylureas in NIDDM. Diabetes Care 1992; 15: date. Ann Intern Med 2002; 137: 25-33 

1737-54 42. Cusi K, DeFronzo RA. Metformin: a review of its metabolic 

18. Rendell M. The role of sulfonylureas in the management of type effects. Diabetes Rev 1998; 6: 89-131 

2 diabetes. Drugs 2004; 64 (12): 1339-58 43. Zhou G, Myers R, Li Y, et al. Role of AMP-activated protein 

19. Bailey CJ, Day C. Antidiabetic drugs. Br J Cardiol 2003; 10: kinase in the mechanism of action of metformin. J Clin Invest 

128-36 2001; 108: 1167-74 

20. Yki-Järvinen H. Combination therapies with insulin in type 2 44. Winder WW, Hardie DG. AMP-activated protein kinase, a 

diabetes. Diabetes Care 2001; 24: 758-67 

metabolic master switch: possible roles in type 2 diabetes. Am 

21. DeFronzo RA. Pharmacologic therapy for type 2 diabetes mellitus. 

J Physiol 1999; 277: E1-E10 

Ann Intern Med 1999; 131: 281-303 

45. Lord JM, Flight IHK, Norman RJ. Metformin in polycystic 

22. Lebovitz HE. Insulin secretagogues: old and new. Diabetes ovary syndrome: systematic review and meta-analysis. BMJ 

Revs 1999; 7: 139-53 2003; 327: 951-5 



46. Howlett HCS, Bailey CJ. A risk-benefit assessment of metfor- 62. Nesto RW, Bell D, Bonow RO, et al. Thiazolidinedione use, 

min in type 2 diabetes mellitus. In: Krentz AJ, editor. Drug fluid retention, and congestive heart failure: a consensus statetreatment 

of type 2 diabetes. Auckland: Adis Books, 2000: 

ment from the American Heart Association and the American 

61-76 

Diabetes Association. Circulation 2003; 108: 2941-8 

47. UK Prospective Diabetes Study Group. Effect of intensive 

63. Parulkar AA, Pendergrass ML, Granda-Ayala R, et al. Nonhyblood-glucose 

control with metformin on complications in 

overweight patients with type 2 diabetes (UKPDS 34). Lancet 

poglycemic effects of thiazolidinediones. Ann Intern Med 

1998; 352: 854-65 

2001; 134: 61-71 

48. The Oxford Centre for Diabetes, Endocrinology and Metabol- 64. Yki-Järvinen H. Thiazolidinedions. N Engl J Med 2004; 351: 

ism: Diabetes Trials Unit. UK Prospective Diabetes Study 1106-18 

[online]. Available from URL: http://www.dtu.ox.ac.uk/ 65. Marten FMAC, Visseren FLJ, Lemay J, et al. Metabolic and 

ukpds/index.html [Accessed 2004 Nov 23] 

additional vascular effects of thiazolidinediones. Drugs 2002; 

49. Hermann LS, Lindberg G, Lindblad U, et al. Efficacy, effec- 62 (10): 1463-80 

tiveness and safety of sulphonylurea-metformin combination 

66. Buchanan TA, Xiang AH, Peters RK, et al. Preservation of 

therapy in patients with type 2 diabetes. Diabetes Obes Metab 

2002; 4: 296-304 

pancreatic beta-cell function and prevention of type 2 diabetes 

by pharmacological treatment of insulin resistance in high-risk 

50. Johnson JA, Majumdar SR, Simpson SH, et al. Decreased 

hispanic women. Diabetes 2002; 51: 2796-803 

mortality associated with sulfonylurea monotherapy in type 2 

diabetes. Diabetes Care 2002; 25: 2244-8 

67. Bell DSH. B-cell rejuvenation with thiazolidinediones. Am J 

Med 2003; 115 (8A): 20-23S 

51. Diabetes Prevention Program Research Group. Reduction of the 

incidence of type 2 diabetes with lifestyle intervention or 68. Roberts AW, Thomas A, Rees A, et al. Peroxisome proliferator 

metformin. N Engl J Med 2002; 346: 393-403 

activated receptor-γ agonists in atherosclerosis: current evidence 

and future directions. Curr Opin Lipidol 2003; 14: 

52. Sulkin T, Bosman D, Krentz AJ. Contraindications to metformin 

therapy in patients with NIDDM. Diabetes Care 1997; 20: 567-73 

925-8 

69. Schoonjans K, Auwerx J. Thiazolidinediones: an update. Lancet 

53. Holt HB, Krentz AJ. Metabolic emergencies in type 2 diabetes. 2000; 355: 1008-10 

In: Goldstein B, Müller-Wieland D, editors. Textbook of type 

70. Inzucchi SE. Oral antihyperglycemic therapy for type 2 diabe- 

2 diabetes. London: Martin Dunitz, 2003: 183-98 

tes. JAMA 2002; 287: 360-72 

54. Lalau J-D, Race J-M. Metformin and lactic acidosis in diabetic 

71. Kendall H. Trends in prescribing of drugs used to treat diabetes. 

humans. Diabetes Obes Metab 2000; 2: 131-7 

Prescriber 2003; 14 (24): 38-9 

55. Day C. Thiazolidinediones: a new class of antidiabetic drugs. 

72. Cuthbertson D, Leese G. Managing type 2 diabetes: oral an- 

Diabetic Med 1999; 16: 1-14 

tidiabetic drugs. Prescriber 2003; 14 (13): 47-53 

56. Krentz AJ, Bailey CJ, Melander A. Thiazolidinediones for type 

73. Bell DSH. Type 2 diabetes mellitus: what is the optimal treat- 

2 diabetes. BMJ 2000; 321: 252-3 

ment regimen? Am J Med 2004; 116 (5A): 23S-9S 

57. Rosen ED, Spiegelman BM. PPAR-γ: a nuclear regulator of 

metabolism, differentiation, and cell growth. J Biol Chem 

74. Kaufman FR. Type 2 diabetes mellitus in children and youth: a 

2001; 276: 37731-4 

new epidemic. J Pediatr Endocrinol Metab 2002; 15 Suppl. 2: 

737-44 

58. Fasshauer M, Paschke R. Regulation of adipocytokines and 

insulin resistance. Diabetologia 2003; 46: 1594-1603 

75. Huizar JF, Gonzalez LA, Alderman J, et al. Sulfonylureas 

attenuate electrocardiographic ST-segment elevation during a 

59. Baldwin SJ, Clarke SE, Chenery RJ. Characterisation of the 

cytochrome P450 enzymes involved in the in vitro metabolism myocardial infarction. J Am Coll Cardiol 2003; 42: 1017-21 

of rosiglitazone. Br J Clin Pharmacol 1999; 48: 424-32 

60. Bailey CJ, Day C, Krentz AJ. Nice timing for glitazones. Br J 

Diabetes Vasc Dis 2003; 3: 366-7 

61. Buch HN, Baskar V, Barton DM, et al. Combination of insulin 

and thiazolidinedione therapy in massively obese patients with 

type 2 diabetes. Diabetic Med 2002; 19: 572-4 

Correspondence and offprints: Dr Andrew J. Krentz, 

Mailpoint 47, Southampton General Hospital, Tremona 

Road, Southampton SO16 6YD, UK. 

E-mail: a.j.krentz@soton.ac.uk

Oral Antidiabetic Agents - Luzimar Teixeira

Create successful ePaper yourself

Delete template?

Save as template?