treating hyperglycemia in type 2 diabetes - cleveland clinic journal

CLEVELAND CL IN IC JOURNAL OF MEDICINE VOLUME 69 • NUMBER 10 OCTOBER 2002 809

E ARE TAKING type 2 diabetes mellitus alot more seriously than in the past, and

treating it more aggressively.For starters, forget about the old term for

the disease, “non–insulin-dependent dia-betes.” Many patients with type 2 or adult-onset diabetes do need insulin. In fact, someneed basal insulin treatment with intermedi-ate-acting or long-acting insulins, while a fewmay need pre-meal insulin treatment to con-trol postprandial hyperglycemia in addition tobasal treatment.

We now have a menu of oral antihyper-glycemic drugs and new insulin preparationsthat, if used rationally, can greatly improvemetabolic control and decrease complicationsin most patients.

This review provides a practical perspec-tive on how to use the new drugs, discussinghow they work, their effect on components ofthe metabolic syndrome, their adverse effects,and how they can be used in combination.

■ CONSEQUENCES OF TYPE 2 DIABETES

Type 2 diabetes is one of the major problemsconfronting the health care system. By 2010,its prevalence will have increased approximate-ly 2.5-fold from 1990 figures.1 It is the leadingcause of new blindness2 and end-stage renaldisease,3 it accounts for slightly more than onehalf of lower-extremity amputations,4 and it isa major risk factor for cardiovascular disease.5

HAROLD E. LEBOVITZ, MD*

Professor of Medicine, Division of Endocrinology andMetabolism/Diabetes, State University of New York(SUNY) Health Sciences Center, Brooklyn

Treating hyperglycemia in type 2diabetes: New goals and strategies

REVIEW

■ ABSTRACT

To achieve glycemic goals in type 2 diabetes, one mustusually use combinations of oral agents or oral agents plusinsulin. This paper discusses the metabolic derangements oftype 2 diabetes, the different classes of antihyperglycemicdrugs, and strategies for using these drugs rationally.

■ KEY POINTS

The American Diabetes Association calls for a goalhemoglobin A1c level of 7.0% in type 2 diabetes; otherorganizations set the goal at 6.5%. Plasma glucose levelsthat correspond to a hemoglobin A1c level of 6.5% are afasting level less than 110 mg/dL and a 2-hour postprandiallevel less than 140 mg/dL.

Both fasting and postprandial glucose levels need to bemonitored and controlled, as do components of themetabolic syndrome such as insulin resistance,dyslipidemia, hypertension, and a procoagulant state.

Agents that decrease fasting plasma glucose levelsselectively (eg, sulfonylureas and metformin) or inconjunction with lowering postprandial glucose excursions(eg, repaglinide, pioglitazone, and rosiglitazone) lowermean hemoglobin A1c levels 1.5 to 2.0 percentage points.

Agents that primarily lower postprandial hyperglycemia arethe alpha-glucosidase inhibitors and nateglinide. Thesedecrease mean hemoglobin A1c levels by 0.5 to 1.0percentage points.

W

*The author has indicated that he has received grant or research supportfrom the Bristol-Myers Squibb, GlaxoSmithKline, and Novo Nordiskcorporations; serves as a consultant for the Amylin, AstraZeneca, Bayer,Bristol-Myers Squibb, GlaxoSmithKline, Janssen Pharmaceutica, Merck-Lipha, Novartis, Novo Nordisk, Pfizer, and Takeda corporations; is on thespeakers’ bureaus of all of the above plus the Aventis corporation; and is amajor stock shareholder in the Bayer, Bristol-Myers Squibb, andGlaxoSmithKline corporations.

on January 21, 2022. For personal use only. All other uses require permission.www.ccjm.orgDownloaded from

http://www.ccjm.org/

810 CLEVELAND CL IN IC JOURNAL OF MEDICINE VOLUME 69 • NUMBER 10 OCTOBER 2002

■ WHAT GOES WRONG

Beta cell dysfunctionIn type 2 diabetes, the normal physiologicrelationship between plasma glucose levelsand insulin secretion is impaired, due to dis-turbances in pancreatic beta cell function.These disturbances express themselves clini-cally in three major ways:• The normal increase in insulin secretionthat occurs after eating is delayed,6 resulting inearly, exaggerated postprandial hyperglycemiaand, potentially, late postprandial hypo-glycemia.7

• Not enough insulin is secreted to meet theneeds of insulin-sensitive tissues.8• The insulin-secreting function of betacells progressively declines.9

Insofar as possible, treatment strategiesneed to address all of these issues.

Insulin resistanceMost patients who develop type 2 diabeteshave insulin resistance, which precedes andprobably promotes the progression from nor-mal glucose tolerance to impaired glucose tol-erance and then to type 2 diabetes.10,11

The pancreas tries to compensate for

Insulinsecretogogueswork only ifenough betacells remain

TYPE 2 DIABETES DRUGS LEBOVITZ

Antihyperglycemic drugs for type 2 diabetes

DRUG DOSE MECHANISM OF ACTION

Insulin secretogoguesSulfonylureas (one or two divided doses) Close KATP channel

Glipizide 2.5–25 mg/day* and open Ca2+ channel of beta cellsGlipizide GITS 2.5–10 mg/day*Glyburide 1.25–15 mg/day*

Micronized form 1–12 mg/dayGlimepiride 1–8 mg/day

Rapid-acting Close KATP channelNateglinide 60–120 mg with each meal and open Ca2+ channel of beta cellsRepaglinide 0.5–4.0 mg with each meal

Basal insulins Highly variable Decrease hepatic glucose productionNPH Increase glucose uptakeLente Decrease lipolysisUltralenteGlargine

Pre-meal insulins Highly variable Increase postprandial glucose uptakeRegularLisproAspart

Insulin sensitizersBiguanide Increases insulin action in liver

Metformin 500–1,000 mg twice daily Decreases hepatic gluconeogenesisThiazolidinediones Increase insulin-mediated glucose

Pioglitazone 15–45 mg/day uptake in muscles and adipose tissueRosiglitazone 4–8 mg/day Increase adipogenesis

Alpha-glucosidase inhibitors Inhibit cleavage of oligosaccharidesAcarbose 25–100 mg with each meal in intestineMiglitol 50–100 mg with each meal

*These are the dose ranges that have been shown to be effective in the clinic. The manufacturers recommendhigher ranges: glipizide to 40 mg/day; glipizide GITS to 20 mg/day; glyburide to 20 mg/day.

T A B L E 1



insulin resistance by secreting more insulin,8but patients with the genetic predisposition todevelop type 2 diabetes cannot secreteenough.

In addition, insulin resistance either caus-es or is associated with a cluster of metabolicabnormalities known as the insulin resistancesyndrome, the metabolic syndrome, or the dys-metabolic syndrome.10,12 The components ofthis syndrome—eg, glucose intolerance, dys-lipidemia, hypertension, a procoagulant state,and a noninfective inflammatory state—arecardiovascular risk factors, and patients withthe syndrome have a higher risk of cardiovas-cular diseases.13,14

The third Adult Treatment Panel (ATPIII)15 of the National Cholesterol EducationProgram has provided clinical criteria for diag-nosing the metabolic syndrome. A patient canbe considered to have the metabolic syn-drome if he or she has any three of the follow-ing five criteria :• Increased waist circumference (> 40 inch-

es in men; > 35 inches in women)• Plasma triglycerides ≥ 150 mg/dL• Plasma high-density lipoprotein (HDL)

cholesterol < 40 mg/dL (men) or < 50mg/dL (women)

• Blood pressure ≥ 130/85 mm Hg• Fasting plasma glucose ≥ 110 mg/dL.


CCF©2002

FIGURE 1

■ Drug therapy for type 2 diabetesDrug therapy for type 2 diabetes aims to control blood sugar levels both in the basal (fasting) state and postprandially; rationalcombinations of agents with different mechanisms of action can be used

Insulin secretogogues close ATP-sensitive potassiumchannels in the beta cells of the pancreas, increasinginsulin production; slow-acting and rapid-acting agentsare available

Metformin, an insulin sensitizer, increases insulin action in theliver, lowering glucose production and fasting glucose levels. Itmay have a small effect in increasing insulin-mediated muscleglucose uptake

Insulin therapy often has twocomponents, an intermediate-acting or long-acting insulin givenat bedtime, and a rapid-actinginsulin given before meals

Alpha-glucosidase inhibitors slow digestion ofoligosaccharides, lowering postprandial glucose levels

Thiazolidinediones, another classof insulin sensitizer, enhance insulin-mediated glucose uptake in muscleand, to a lesser degree, in the liver

G L U C O S E




Treatment strategies for type 2 diabetesmust target insulin resistance, as improvementin insulin action allows endogenous insulin tobe more effective and reduces cardiovascularrisk factors.

■ GLUCOSE METABOLISM:FASTING AND POSTPRANDIAL

During fasting, glucose and insulin levelsdecline. The liver responds to low insulin lev-els by producing glucose in a process calledgluconeogenesis, which plays the major role incontrolling fasting plasma glucose levels.16,17

In contrast, postprandial glucose levels areregulated largely by uptake by musclecells.16,17

The insulin level necessary to turn off glu-cose production in the liver is significantlylower than the level necessary to drive glucoseinto muscle cells. Therefore, patients withtype 2 diabetes have a dissociation betweenthe regulation of fasting and postprandialhyperglycemia—specifically, they can havenormal or even low fasting glucose levels butstill have large excursions after meals.Continuous monitoring over 48 to 72 hourshas confirmed that patients with type 2 dia-betes cannot achieve target glycemic controlunless both fasting and postprandial glucoselevels are measured, and unless treatments aregiven to appropriately regulate each.

■ ANTIHYPERGLYCEMIC AGENTS

Diet and exercise are the first step of therapyfor type 2 diabetes; if these do not keep bloodsugar at goal levels, then antihyperglycemicagents are added.

Four major classes of antihyperglycemicscan be used (TABLE 1, FIGURE 1), either asmonotherapy or, more appropriately, in com-bination with one another:• Insulin secretogogues• Insulins• Insulin sensitizers• Alpha-glucosidase inhibitors.

■ INSULIN SECRETOGOGUES

Insulin secretogogues correct hyperglycemiaby stimulating insulin secretion—but only ifthe patient still has enough functioning betacells.

The mechanism uses the same final path-way as glucose itself.7,18 When plasma glucoselevels rise, potassium channels in the plasmamembrane of beta cells normally close; thesechannels are sensitive to intracellular adeno-sine triphosphate levels and so are designatedKATP. Closure of KATP channels causes adja-cent calcium channels to open. The resultantincrease in cytosolic calcium stimulatesinsulin granule secretion. In type 2 diabetes,the signal is deficient and delayed.

Sulfonylureasdecrease latebut not earlypostprandialhyperglycemia


Sulfonylureas vs rapid-acting insulin secretogogues(nateglinide and repaglinide)

SULFONYLUREAS RAPID-ACTING SECRETOGOGUES

Early meal-mediated insulin secretion No effect IncreaseLate meal-mediated insulin secretion Marked increase IncreaseEarly meal-mediated hyperglycemia Slight effect DecreaseLate meal-mediated hyperglycemia Marked decrease DecreaseFasting hypoglycemia Moderate occurrence Small occurrenceWeight gain Moderate SmallEffect on KATP channels Possible None

in myocardium and coronary arteriesDosing Once or twice daily With each mealApproximate cost $7.50–$26.00/month $75.00–$85.00/month

T A B L E 2



Insulin secretogogues close KATP channelsby binding to specific receptors on these chan-nels. Thus, they can supplement an insuffi-cient response to glucose. On the other hand,they have no primary effect on the compo-nents of the metabolic syndrome.

The major side effects of insulin secreto-gogues (and insulin replacement) are hypo-glycemia and weight gain.

SulfonylureasSulfonylurea drugs—eg, glipizide (Glucotrol),glyburide (Diabeta, Micronase), glimepiride(Amaryl)—increase the quantity of insulinsecreted in patients with type 2 diabetes, butdo not correct the delay in meal-mediatedinsulin secretion because their binding toreceptors on the KATP channel is slow andprolonged.19–21 Thus, they decrease late butnot early postprandial hyperglycemia.19–21 Infact, in the early stages of type 2 diabetes, sul-fonylurea therapy often leads to late postpran-dial hypoglycemia because of a markedincrease in late meal-mediated insulin secre-tion.

Rapid-acting secretogoguesTwo other available insulin secretogogues alsowork by binding to and closing KATP chan-nels, but they bind and detach more rapidly

than the sulfonylureas and have a shorter plas-ma half-life. Thus, they have a faster onsetand shorter duration of action.22–25

Nateglinide (Starlix) binds to the samesite as the sulfonylureas. Given either with orshortly before a meal, it rapidly increasesmeal-mediated insulin secretion, and thiseffect lasts 3 to 4 hours.26 Its primary effect isto reduce postprandial hyperglycemia.27

Repaglinide (Prandin) binds to a differentsite on the KATP channel, and its binding isslightly slower and more prolonged than thatof nateglinide.28,29 It greatly reduces postpran-dial hyperglycemia, but because its effects aremore prolonged, it also greatly reduces fastinghyperglycemia.30

Relative advantages and disadvantagesThe rapid-acting insulin secretogogues, giventhree times a day with meals, restore normalpostprandial glucose metabolism more closelythan do the sulfonylureas, which are givenonce or twice daily (TABLE 2). Moreover, unlikethe sulfonylureas, they interact little or not atall with KATP channels in cardiovascular tis-sues. (There has been some concern, thoughstill controversial, that causing closure ofKATP channels in the myocardium and coro-nary arteries might interfere with some pro-tection against acute ischemia.)


Nateglinide andrepaglinide willnot work if asulfonylureadid not

Characteristics of insulin preparations and insulin analogues

PREPARATION ONSET PEAK EFFECTIVE DURATION(HOURS) (HOURS) (HOURS)

Rapid-actingLispro (analogue) 0.25 1 3Aspart (analogue) 0.25* 1* 3–4*

Short-actingRegular (soluble) 0.5–1.0 2–3 3–6

Intermediate-actingNPH (isophane) 2–4 7–8 10–12Lente (zinc suspension) 2–4 7–8 10–12

Long-actingUltralente (zinc suspension) 4 Variable 18–20Glargine (analogue) 1–2 Flat, predictable 24*

*Occasionally less

T A B L E 3




Both types of insulin secretogogues act byclosing KATP channels; therefore, combining asulfonylurea with a rapid-acting insulin secre-togogue is likely to be of no benefit. Moreover,rapid-acting insulin secretogogues will notwork in patients in whom sulfonylureas do notwork.

■ INSULINS: BASAL AND BOLUS

As beta cell function deteriorates during thecourse of type 2 diabetes, insulin secretogogueseventually become ineffective.4,31 At thatstage, the patient needs insulin replacement(TABLE 3).

Basal insulin replacementAn early sign that the beta cells are becomingless responsive to insulin secretogogues is thatfasting plasma glucose levels begin to rise. Thereason: so much insulin is depleted duringmeals that little or none is left for secretionovernight to control hepatic glucose produc-tion.

This can be managed by giving neutralprotamine Hagedorn (NPH) or Lente insulinat bedtime (10 to 11 PM).32,33 Giving the doseat bedtime provides circulating insulin from 1to 8 AM and can be used to regulate plasmaglucose levels overnight.34 The dose is adjust-ed every 3 or 4 days until the fasting plasmaglucose level is 110 to 120 mg/dL.33

The absorption and effects of NPH andLente insulins vary substantially from day today, however. A better option may be insulinglargine (Lantus), which has less variability inits action and a longer duration of action. Itcan be given at 6 PM or bedtime to achieveovernight glycemic control.35

At this stage, most patients should con-tinue to take an oral agent in the daytimealong with insulin at night.33,35

Rapid-acting insulinsAs beta cell function deteriorates further, thepatient may need both a basal insulin to con-trol fasting blood sugar levels and boluses of arapid-acting insulin to cover postprandial lev-els.36 This goal can be best achieved by givinga long-acting insulin such as insulin glarginein the evening and a rapid-acting insulin suchas insulin lispro (Humalog) or insulin aspart

(NovoLog) before each meal (TABLE 3).37

If a patient needs a less-intensive regimen,for whatever reason, the regimen most oftenused is a mixture of a short-acting and anintermediate-acting insulin, given twice a day,ie, before breakfast and before the eveningmeal.38 Several premixed formulations areavailable.

■ INSULIN SENSITIZERS

Two classes of insulin sensitizers are currentlyavailable. They increase insulin sensitivity bydifferent mechanisms and have their primaryeffects in different tissues.

Metformin, a biguanideMetformin (Glucophage), a biguanide, hasbeen used to treat type 2 diabetes for morethan 40 years.

Metformin’s mechanism of action is stillunknown, despite extensive studies. Its prima-ry effect on glucose metabolism is to decreasethe exaggerated hepatic glucose productionthat causes fasting hyperglycemia.39 It does soby increasing insulin action in the liver, there-by reducing hepatic gluconeogenesis.40

Whether metformin affects glucoseuptake in the muscles and other peripheral tis-sues is controversial. Earlier studies41,42

showed modest increases in whole-bodyinsulin-mediated glucose uptake, but they didnot control for weight loss and improvementin glycemia. Later studies that controlled forthese confounding factors showed little or noeffect.43,44

Clinically, metformin decreases plasmainsulin levels, causes a small amount ofweight loss, and improves some of the com-ponents of the insulin resistance syndrome(TABLE 4).45,46 It lowers fasting hyperglycemiabut has little or no effect on postprandialhyperglycemia.45

Metformin may even reduce the inci-dence of coronary artery disease. Overweightpatients with type 2 diabetes who receivedmetformin in the United KingdomProspective Diabetes Study (UKPDS)47 had a39% lower incidence of myocardial infarctionand a 42% lower incidence of diabetes-relateddeaths compared with patients undergoingconventional treatment (ie, lifestyle modifica-

Adjust thebedtime insulindose every 3 or4 days until thefasting glucoseis 110 – 120




tion). These differences were highly statisti-cally significant and did not occur in over-weight patients treated with insulin or sul-fonylureas, although the improvement inglycemic control was similar.

Metformin contraindications. Metforminis contraindicated in patients with impairedrenal function (serum creatinine ≥ 1.5 mg/dLin men, ≥ 1.4 in women; or creatinine clear-ance < 60 mL/minute), symptomatic conges-tive heart failure being treated by drugs, orconfirmed acidosis. These all predispose tometformin-induced lactic acidosis, which,however, is rare in the absence of these fac-tors.46,48

Since radiographic contrast material canoccasionally cause impaired renal function,metformin should be stopped at the time ofsuch studies and restarted when it is clear thatrenal function has not been compromised.

Metformin side effects. The major prob-lems with metformin treatment are abdominaldiscomfort and diarrhea. These are dose-relat-ed and preclude full dosing in 10% to 20% ofpatients. Giving metformin with meals and

titrating the dose from 0.5 g/day up to 2.0g/day over a week or so decreases the severityof gastrointestinal side effects. The slow-release form of metformin seems to causefewer gastrointestinal side effects.

ThiazolidinedionesThe thiazolidinediones pioglitazone (Actos)and rosiglitazone (Avandia) decrease periph-eral insulin resistance by enhancing insulin-mediated glucose uptake by muscle.12,49–51

They have a lesser effect on insulin action inthe liver.

Thiazolidinediones act by binding to andactivating a specific transcription factor in thenucleus of the cell.50,51 When activated, thisfactor binds to specific genes and either stim-ulates or inhibits their transcription. Some ofthese genes regulate proteins involved in adi-pose tissue differentiation, lipid metabolism,and the intracellular insulin action cascade.

Thiazolidinediones improve insulinaction in patients with type 2 diabetes byapproximately 25% to 40%, depending on thepopulation studied.52,53 They also decrease


Metforminlowers fastinghyperglycemia,but notpostprandial

Effects of insulin sensitizers

EFFECT METFORMIN THIAZOLIDINEDIONES

Insulin sensitivity—liver Significant increase Small increaseInsulin sensitivity—muscle Controversial 25% to 40% increaseHyperinsulinemia Small decrease DecreasePlasma free fatty acids Small decrease 25% reductionDyslipidemia* Minimal effect Significant improvementProcoagulant state Decreased PAI-1† Decreased PAI-1Hypertension No effect 2–4 mm Hg reduction in

diastolic blood pressureVisceral adiposity Some decrease No effectBody weight Decrease or no change Weight gain of 2–3 kgEndothelial dysfunction Improved ImprovedEdema None 4% to 5% incidenceUrinary albumin excretion No primary effect Glycemia-independent reduction

*Dyslipidemia consists of increased plasma triglycerides, decreased plasma high-density lipoprotein cholesterol,and a shift in low-density lipoprotein cholesterol particle size from large and buoyant to small and atherogenic.Thiazolidinediones reverse these abnormalities.

†Plasminogen activator inhibitor-1

T A B L E 4




circulating free fatty acid levels approximately25%,52 and this is one of the mechanismsthought responsible for the decrease in insulinresistance.

A major effect of thiazolidinediones is thedifferentiation of stem cells into adipocytes.This occurs in subcutaneous adipose tissue, butnot in visceral adipose tissue,54 and accountsfor some of the weight gain associated with thi-azolidinedione therapy.

The mean decrease in hemoglobin A1cwith thiazolidinedione therapy is 1.5 percent-age points. These drugs also mitigate many ofthe features of the metabolic (insulin resis-tance) syndrome (TABLE 4),12,49 such as diabeticdyslipidemia and the procoagulant state.Therefore, it is reasonable to speculate thatthese drugs will reduce the incidence of car-diovascular diseases. A number of in vitro andsmall in vivo animal and human studies sup-port such a conclusion. Several large clinicaloutcome studies are under way to test thishypothesis definitively.

Moreover, thiazolidinediones might pre-serve beta cell function. The UKPDS9 demon-strated that neither sulfonylureas, metformin,nor dietary treatments had any effect on theprogressive deterioration of pancreatic betacell function. Studies in animal models ofinsulin-resistant diabetes55 and preliminarystudies in humans raise the possibility thatlong-term treatment with thiazolidinedionesmay slow such deterioration. Several long-term clinical trials are under way to examinethis.

Adverse effects of thiazolidinediones.The major adverse effects of the thiazolidine-diones are fluid retention and weight gain.48,49

As monotherapy, both rosiglitazone andpioglitazone cause a modest increase in plasmavolume and are reported to cause mild to mod-erate edema in 4% to 5% of patients. When athiazolidinedione is taken with a sulfonylurea,6% to 7% of patients develop edema; when itis taken with insulin, approximately 15% ofpatients develop edema.

Occasionally, congestive heart failuredevelops during thiazolidinedione treatment,but there are no published data to defineeither the prevalence or the relationship ofthis to drug therapy. It is not unreasonable tosuspect that an increase in plasma volume

might precipitate clinical congestive heartfailure in patients who have cardiovasculardisease and are in borderline compensation.Thiazolidinediones are not recommended inpatients with New York Heart Associationfunctional class III or IV heart failure.

Weight gain is due to an increase in sub-cutaneous adipose tissue and in fluid reten-tion. Monotherapy is associated with a weightgain of 1.6 to 3.5 kg in the first year. A slight-ly greater weight gain occurs when a thiazo-lidinedione is combined with a sulfonylurea orinsulin. Rarely and for unknown reasons, somepatients develop severe edema, gain 10 to 20kg, or both.

Hepatotoxicity was seen with troglita-zone, an earlier thiazolidinedione, and wasresponsible for its removal from the market,but it does not appear to occur with rosiglita-zone or pioglitazone.56 However, as a precau-tion, patients receiving thiazolidinedionesshould undergo baseline and periodic moni-toring of liver enzymes. These drugs shouldnot be given to patients with baseline alanineaminotransferase (ALT) levels 2.5 or moretimes the upper limit of the normal range.Persistence of ALT levels more than threetimes the upper limit of the normal range dur-ing therapy is an indication to stop the drug.

Role of insulin sensitizers in type 2 diabetesInsulin sensitizers improve hyperglycemia bydecreasing insulin resistance. Their uniquecontribution to the management of type 2 dia-betic patients is that they treat many of thecomponents of the metabolic syndrome (dys-lipidemia, procoagulant state, endothelial dys-function, inflammatory responses) and reducethe risk for cardiovascular complications. Formetformin, this has been shown to result infewer clinical cardiovascular events.

■ ALPHA-GLUCOSIDASE INHIBITORS

The alpha-glucosidase inhibitors acarbose(Precose) and miglitol (Glyset) competitive-ly inhibit digestion of oligosaccharides tomonosaccharides, so that glucose is slowlyabsorbed throughout the length of the smallintestine, rather than rapidly in the proximaljejunum.57,58 Advantages:• They specifically lower postprandial plas-


UKPDS: FewerMIs, deathswith metforminvs lifestylemodification



ma glucose levels• Their action is independent of and addi-

tive to all other forms of pharmacologictherapy

• They do not cause weight loss or gain• They are relatively nontoxic.57,58

When given to patients already takinginsulin, sulfonylureas, or metformin, thealpha-glucosidase inhibitors decrease maximalpostprandial plasma glucose levels by approx-imately 50 mg/dL and hemoglobin A1c byapproximately 0.5 percentage points.59,60

Alpha-glucosidase inhibitors must begiven with the start of each meal. They areeffective only if more than 40% of the caloriesin the diet come from complex carbohy-drates—a high content.57

If hypoglycemia occurs when an alpha-glucosidase inhibitor is used in combinationwith insulin or a sulfonylurea, glucose—notdisaccharides—must be given because of thedelayed digestion of complex carbohydrates.

Side effects. The major disadvantage ofalpha-glucosidase inhibitors is their gastroin-testinal side effects, which include abdominaldiscomfort, flatulence, and, occasionally, diar-rhea.59,60 These effects are due to carbohy-drate spilling into the colon, where it is fer-mented by bacteria. These side effects can beminimized by starting with very small doseswith the evening meal and titrating upwardvery slowly. This allows the distal jejunum andileum time to increase their normally low con-centrations of alpha-glucosidase enzymes, sothat the oligosaccharides that reach those seg-ments can be digested.

■ CHANGING GOALS OF THERAPY

The goals of therapy for patients with type 2diabetes have changed in the last severalyears.

Trials show that chronic vascular compli-cations increase in prevalence and severitywith the duration of type 1 and type 2 diabetesand with the magnitude of hyperglycemia,beginning at hemoglobin A1c levels above themean normal value (ie, > 5.2%).61,62 Thismeans that the ideal goal for glycemic controlshould be a hemoglobin A1c level as close to6.0% as possible without causing serious sideeffects. The American Diabetes Association

sets the goal as less than 7.0%; the EuropeanDiabetes Association and the AmericanAssociation of Clinical Endocrinologists setthe goal at 6.5% or lower.

Plasma glucose levels that correspond to ahemoglobin A1c level of 6.5% are a fastinglevel less than 110 mg/dL and a 2-hour post-prandial level less than 140 mg/dL.

The exact goal should be tailored to theindividual patient, based on his or her clinicalcharacteristics.

Blood pressure and lipid controlThe goal for blood pressure control in diabet-ic patients is 130/80 mm Hg or lower.63

The goals for the management of dyslipi-demia as defined by the ATP III of theNational Cholesterol Education Program15

are a low-density lipoprotein cholesterollevel lower than 100 mg/dL, with no specificgoals for plasma triglycerides or high-densitylipoprotein cholesterol. However, studiessuggest that a target plasma triglyceride levelof 150 mg/dL or lower and a plasma high-density lipoprotein cholesterol level of 40mg/dL or higher are desirable.

■ EFFECTS ON GLYCEMIC CONTROL

The effect of all antihyperglycemic agents onoverall glycemic control as measured by reduc-tion in hemoglobin A1c depends on the base-line hemoglobin A1c level.49 Interventionalstudies with rosiglitazone, metformin, andinsulin secretogogues have shown that theydecrease mean hemoglobin A1c by 1.5 to 2.0percentage points vs placebo when the base-line hemoglobin A1c is between 8.5% and9.5%, but only about 1 percentage point whenthe baseline is between 6.5% and 7.5%.48

Chronic plasma glucose levels over 300mg/dL have direct inhibitory effects on insulinsecretion and cause some impairment ofinsulin action.64 Any treatment of hyper-glycemia decreases glucose toxicity as it lowershyperglycemia.

Agents that have major effects indecreasing fasting plasma glucose levels (sul-fonylureas, repaglinide, metformin, pioglita-zone, rosiglitazone) lowered mean hemoglo-bin A1c levels 1.5 to 2.0 percentage points vsplacebo in the registration studies submitted


Diabetes goals:• A1c ≤ 6.5 or 7• BP ≤ 130/80• LDL < 100




to the US Food and Drug Administration(FDA).48

In the few comparative studies available(metformin vs glyburide, repaglinide vs gly-buride, rosiglitazone vs glyburide), theseagents appeared to be equally effective indecreasing hemoglobin A1c, though they havedifferent mechanisms.45,65 There are no directcomparative studies demonstrating the superi-ority of any single agent in decreasing hemo-globin A1c.

Agents that primarily lower postprandialhyperglycemia (alpha-glucosidase inhibitors,nateglinide) decreased mean hemoglobin A1c0.5 to 1.0 percentage points compared withplacebo in registration studies presented to theFDA.48

Insulin supplementation can lowerhemoglobin A1c to any desired level, but itsuse is limited by unacceptable rates of severehypoglycemia.

Monotherapy or combination therapy?The degree to which a single drug can achieveglycemic control is limited by its intrinsicantihyperglycemic activity, by the patient’sresidual beta cell function, and by its sideeffects. Consequently, monotherapy achievestarget glycemic control in type 2 diabeticpatients with only mild elevations of hemo-globin A1c (probably no higher than 7.0%).

The agent of choice for these patientsshould be one that does not cause hypo-glycemia, ie, metformin, a thiazolidinedione, oran alpha-glucosidase inhibitor. The concomi-tant presence of the metabolic syndrome woulddictate the use of an insulin sensitizer. If aninsulin secretogogue were to be used, a rapid,short-acting one would be preferable to mini-mize late hypoglycemia. Recent studies66 foundthat monotherapy with metformin decreasedthe rate of progression from impaired glucosetolerance to type 2 diabetes over 3 years by31%, and acarbose decreased it by 25%.

More severe elevations of hemoglobin A1care best treated with combinations of submax-imal doses of two or more agents, each ofwhich corrects hyperglycemia by a differentmechanism. Type 2 diabetic patients withsevere insulin deficiency require either bed-time insulin as a component of their treatmentor intensive insulin replacement therapy.

■ COMBINATION THERAPYOF TYPE 2 DIABETES

The most effective treatment of moderatehyperglycemia (hemoglobin A1c 7% to 8%) orsevere hyperglycemia (hemoglobin A1c > 8%)is a combination either of oral agents thathave different modes of action or of oralagents with insulin replacement.48

Combinations of oral agentsCombination oral therapy usually involvesan insulin sensitizer and an insulin secreto-gogue. The addition of the second agentusually lowers the mean hemoglobin A1clevel 1.0 to 1.4 percentage points more thanmonotherapy when the baseline value isbetween 8.5% and 9.5%.48 Data are avail-able on glyburide-metformin, repaglinide-metformin, nateglinide-metformin, sulfony-lurea-pioglitazone, sulfonylurea-rosiglita-zone, repaglinide-pioglitazone, andrepaglinide-rosiglitazone.48,67–70 From 30%to 50% of patients taking these regimenslower their hemoglobin A1c levels to 7.0%or less.

Some combinations can be given in asingle pill: eg, glyburide and metformin(Glucovance). The advantage of a singlepill is lower cost, a single copayment forthe patient, and better compliance. Thedisadvantage is the fixed ratio of the drugs,which rules out dosage adjustment of anindividual agent.

Combination therapy can start with sub-maximal doses of each agent, or after thedose of the first agent has been maximized.Few data are available to prove whichapproach is better, but starting with a combi-nation has the potential advantages of moreexpeditious control of glycemia and fewerside effects.

Metformin plus a thiazolidinedione hasbeen shown to improve glycemic control, low-ering hemoglobin A1c by 1.0 to 1.2 percentagepoints compared with maximum doses ofeither drug alone, when the mean baselinehemoglobin A1c level was between 8.5% and9.5%.67,70 This additive benefit in glycemiccontrol occurs because these agents improveinsulin resistance by different mechanisms andin different tissues.


Combinationregimens lowerA1c levels to7.0% or less in30%–50% ofpatients



Adding a thirdor fourth oral agentOne can consider adding a third or fourth oraldrug with yet another mechanism of action ifthe hemoglobin A1c level achieved with twoagents is 8.0% or less, since one may antici-pate a potential decrease of approximately 0.5to 1.3 percentage points, which could bringthe value to below 7.0%.

However, if the hemoglobin A1c on twoagents is above 8.0%, the likelihood thatadding a third oral agent will bring it to thetarget range is small, and adding bedtimeinsulin replacement is more likely to result intarget glycemic control.

The third agent could be an alpha-glu-cosidase inhibitor; if the patient has been tak-

ing an insulin secretogogue and an insulinsensitizer, the third agent could be anotherinsulin sensitizer.

Combining oral therapywith insulin replacementThe addition of an insulin sensitizer to fullinsulin replacement therapy lowers the hemo-globin A1c level by approximately 1.3 per-centage points while reducing the requiredinsulin dose by 20% to 25%.

The cost of combination therapy must beconsidered in addition to the potential bene-fits. The combination of three oral antihyper-glycemic agents is likely to cost more thanbedtime insulin replacement and a daytimeinsulin sensitizer.


■ REFERENCES1. Amos AF, McCarty DJ, Zimmet P. The rising global burden of dia-

betes and its complications: estimates and projections to the year2010. Diabet Med 1997; 14(suppl):7–85.

2. Aiello LP, Cavallerano JD, Gardner TW, et al. Diabetic retinopathy.Diabetes Care 1998; 21:143–156.

3. Centers for Disease Control and Prevention. 1999 DiabetesSurveillance Report. Chapter 8: End Stage Renal Disease. Availableat www.cdc.gov/diabetes/statistics/survl99/chap8/contents.htm.Accessed 5/8/02.

4. Mayfield JA, Janisse D, Reiber GE, Pogach LM, Sanders LJ.Preventive foot care in people with diabetes. Diabetes Care 1998;21:2161–2177.

5. Wingard DL, Barrett-Connor E. Heart disease and diabetes. In:Diabetes in America. 2nd ed. NIH Publication No. 95-1468.1995:429–448.

6. Bergstrom RW, Wahl PW, Leonetti DI, Fujimoto WY. Association offasting glucose levels with a delayed secretion of insulin after oralglucose in subjects with glucose intolerance. J Clin EndocrinolMetab 1990; 71:1447–1453.

7. Lebovitz HE. Insulin secretogogues: old and new. Diabetes Reviews1999; 7:139–153.

8. Kahn SE, Prigeon RL, McCulloch DK, et al. Quantification of therelationship between insulin sensitivity and beta-cell function inhuman subjects. Evidence for a hyperbolic function. Diabetes1993; 42:1663–1672.

9. UK Prospective Diabetes Study Group. Overview of 6 years thera-py of type 2 diabetes: a progressive disease (UKPDS 16). Diabetes1995; 44:1249–1258.

10. Haffner SM, Stern MP, Hazuda HP, Mitchell BD, Patterson JK.Cardiovascular risk factors in confirmed prediabetic individuals:does the clock for coronary heart disease start ticking before theonset of clinical diabetes? JAMA 1990; 263:2893–2898.

11. Weyer C, Bogardus C, Mott DM, Pratley RE. The natural history ofinsulin secretory dysfunction and insulin resistance in the patho-genesis of type 2 diabetes mellitus. J Clin Invest 1999;104:787–794.

12. Lebovitz HE. Insulin resistance: definition and consequences. ExpClin Endocrinol Diabetes 2001; 109(suppl 2):135–148.

13. Isomaa B, Almgren P, Tuomi T, et al. Cardiovascular morbidity andmortality associated with the metabolic syndrome. Diabetes Care2001; 24:683–689.

14. Haffner SM, Mykkänen L, Festa A, Burke JP, Stern MP. Insulin-resis-tant prediabetic subjects have more atherogenic risk factors than

insulin-sensitive prediabetic subjects. Implications for preventingcoronary heart disease during the prediabetic state. Circulation2000; 101:975–980.

15. Expert Panel on Detection, Evaluation, and Treatment of HighBlood Cholesterol in Adults. Executive summary of the thirdreport of the National Cholesterol Education Program (NCEP)Expert Panel on Detection, Evaluation, and Treatment of HighBlood Cholesterol in Adults (Adult Treatment Panel III). JAMA2001; 285:2486–2497.

16. van Haeften TW, Pimenta W, Mitrakou A, et al. Relative contribu-tions of beta-cell function and tissue insulin sensitivity to fastingand postglucose-load glycemia. Metabolism 2000; 49:1318–1325.

17. Dostou J, Gerich J. Pathogenesis of type 2 diabetes mellitus. ExpClin Endocrinol Diabetes 2001; 109(suppl 2):149–156.

18. Aguilar-Bryan L, Bryan J, Nakazaki M. Of mice and men: KATPchannels and insulin secretion. Recent Prog Horm Res 2001;56:47–68.

19. Clark HE, Matthews DR. The effect of glimepiride on pancreaticbeta-cell function under hyperglycaemic clamp and hyperinsuli-naemic, euglycaemic clamp conditions in non-insulin-dependentdiabetes mellitus. Horm Metab Res 1996; 28:445–450.

20. Ashcroft FM. Mechanisms of the glycaemic effects of sulfony-lureas. Horm Metab Res 1996; 28:456–463.

21. Shapiro ET, Van Cauter E, Tillil H, et al. Glyburide enhances theresponsiveness of the beta cell to glucose but does not correct theabnormal patterns of insulin secretion in non–insulin-dependentdiabetes mellitus. J Clin Endocrinol Metab 1989; 69:571–576.

22. Fuhlendorff J, Rorsman P, Kofod H, et al. Stimulation of insulinrelease by repaglinide and glibenclamide involves both commonand distinct processes. Diabetes 1998; 47:345–351.

23. Hu S, Wang S, Dunning BE. Tissue selectivity of antidiabetic agentnateglinide: study on cardiovascular and beta-cell KATP channels. JPharmacol Exp Ther 1999; 291:1372–1379.

24. Hu S, Wang S, Fanelli B, et al. Pancreatic beta-cell KATP channelactivity and membrane-binding studies with nateglinide: a com-parison with sulfonylureas and repaglinide. J Pharmacol Exp Ther2000; 293:444–452.

25. Hatorp V, Huang WC, Strange P. Pharmacokinetic properties ofrepaglinide in elderly subjects with type 2 diabetes. J ClinEndocrinol Metab 1999; 84:1475–1478.

26. Walter YH, Spratt DI, Garreffa S, McLeod JF. Mealtime glucose reg-ulation by nateglinide in type-2 diabetes mellitus. Eur J ClinPharmacol 2000; 56:129–133.

27. Hanefeld M, Dickinson S, Bouter KP, Guitard C. Rapid and short-acting mealtime insulin secretion with nateglinide controls both



prandial and mean glycemia. Diabetes Care 2000; 23:202–207.28. Guay DRP. Repaglinide, a novel, short-acting hypoglycemic agent

for type 2 diabetes mellitus. Pharmacotherapy 1998;18:1195–1204.

29. Damsbo P, Marbury TC, Clauson P, Windfeld K. A double-blindrandomized comparison of meal-related glycemic control byrepaglinide and glyburide in well-controlled type 2 diabeticpatients. Diabetes Care 1999; 22:789–794.

30. Jovanovic L, Dailey G III, Huang W-C, Strange P, Goldstein BJ.Repaglinide in type 2 diabetes: A 24-week fixed-dose efficacy andsafety study. J Clin Pharmacol 2000; 40:49–57.

31. UK Prospective Diabetes Study Group. Intensive blood-glucosecontrol with sulphonylureas or insulin compared with convention-al treatment and risk of complications in patients with type 2 dia-betes (UKPDS 33). Lancet 1998; 352:837–853.

32. Abraira C, Henderson WG, Colwell JA, et al. Response to intensivetherapy steps and to glipizide dose in combination with insulin intype 2 diabetes. VA feasibility study on glycemic control and com-plications (VA CSDM). Diabetes Care 1998; 21:574–579.

33. Yki-Jarvinen H, Ryysy L, Nikkila K, Tulokas T, Vanamo R, HeikkilaM. Comparison of bedtime insulin regimens in patients with type2 diabetes mellitus: a randomized, controlled trial. Ann InternMed 1999; 130:389–396.

34. Mudaliar S, Edelman SV. Insulin therapy in type 2 diabetes.Endocrinol Metab Clin North Am 2001: 30:935–982.

35. Yki-Jarvinen H, Dressler A, Zieman M, Study Group HOE 901/3002.Less nocturnal hypoglycemia and better post-dinner glucose con-trol with bedtime insulin glargine compared with bedtime NPHinsulin during combination therapy in type 2 diabetes. DiabetesCare 2001; 23:1130–1136.

36. Bolli G. Physiological insulin replacement in type 1 diabetes melli-tus. Exp Clin Endocrinol Diabetes 2001; 109(suppl 2):317–332.

37. Bolli G, Owens DR. Insulin glargine. Lancet 2000; 356:443–445.38. Skyler JS. Insulin treatment. In: Lebovitz H, editor. Therapy for

Diabetes Mellitus and Related Disorders. 3rd ed. Alexandria, Va:American Diabetes Association, 1998:186–203.

39. Stumvoll M, Nurjhan N, Perriello G, Dailey G, Gerich J. Metaboliceffects of metformin in non–insulin-dependent diabetes mellitus.N Engl J Med 1995; 333:550–554.

40. Hundal RS, Krssak M, Dufour S, et al. Mechanism by which met-formin reduces glucose production in type 2 diabetes. Diabetes2000; 49:2063–2069.

41. Bell PM, Hadden DR. Metformin. Endocrinol Metab Clin North Am1997; 26:523–537.

42. Campbell IW, Howlett HCS. Worldwide experience of metforminas an effective glucose-lowering agent: a meta-analysis. DiabetesMetab Rev 1995; 11(suppl):57–62.

43. Inzucchi SE, Maggs DG, Spollett GR, et al. Efficacy and metaboliceffects of metformin and troglitazone in type 2 diabetes mellitus.N Engl J Med 1998; 338:867–872.

44. Yu JG, Kruszynska YT, Mulford MI, Olefsky JM. A comparison oftroglitazone and metformin on insulin requirements in eugly-cemic intensively insulin-treated type 2 diabetic patients. Diabetes1999; 48:2414–2421.

45. DeFronzo RA, Goodman AM. Efficacy of metformin in patientswith non–insulin-dependent diabetes mellitus. N Engl J Med 1995;333:541–549.

46. Bailey CJ, Turner RC. Drug therapy: metformin. N Engl J Med1996; 334:574–579.

47. UK Prospective Diabetes Study Group. Effect of intensive bloodglucose control with metformin on complications in overweightpatients with type 2 diabetes (UKPDS 34). Lancet 1998;352:854–865.

48. Lebovitz HE. Oral therapies for diabetic hyperglycemia. EndocrinolMetab Clin North Am 2001; 30:909–933.

49. Lebovitz HE, Banerji MA. Insulin resistance and its treatment bythiazolidinediones. Recent Prog Horm Res 2001; 56:265–294.

50. Willson TM, Brown PJ, Sternbach DD, Henke BR. The PPARs: fromorphan receptors to drug discovery. J Med Chem 2000;43:527–550.

51. Willson TM, Lambert MH, Kliewer SA. Peroxisome proliferator-activated receptor gamma and metabolic disease. Annu RevBiochem 2001; 70:341–367.

52. Lebovitz HE, Dole JF, Patwardhan R, Rappaport EB, Freed MI.Rosiglitazone monotherapy is effective in patients with type 2diabetes. J Clin Endocrinol Metab 2001; 86:280–288.

53. Miyazaki Y, Ferrannini E, Mahankali A, et al. Improved glycemiccontrol and enhanced insulin sensitivity in type 2 diabetic subjectstreated with pioglitazone. Diabetes Care 2001; 24:710–719.

54. Adams M, Montague CT, Prins JB, et al. Activators of PPAR-gammahave depot-specific effects on human pre-adipocyte differentia-tion. J Clin Invest 1997; 100:3149–3153.

55. Finegood DT, McArthur MD, Kojwang D, et al. Beta-cell massdynamics in Zucker diabetic fatty rats; rosiglitazone prevents therise in net cell death. Diabetes 2001; 50:1021–1029.

56. Lebovitz HE, Kreider M, Freed MI. Evaluation of liver function intype 2 diabetic patients during clinical trials: evidence that rosigli-tazone does not cause hepatic dysfunction. Diabetes Care 2002;25:815–823.

57. Lebovitz HE. Alpha-glucosidase inhibitors as agents in the treat-ment of diabetes. Diabetes Reviews 1998; 6:132–145.

58. Raptis SA, Dimitriadis GD. Oral hypoglycemic agents: insulin secre-togogues, alpha-glucosidase inhibitors and insulin sensitizers. ExpClin Endocrinol Diabetes 2001; 109(suppl 2):265–287.

59. Chiasson JL, Josse R, Hunt J, et al. The efficacy of acarbose in thetreatment of patients with non–insulin-dependent diabetes melli-tus. Ann Intern Med 1994; 121:928–935.

60. Holman RR, Cull CA, Turner RC. A randomized double-blind trialof acarbose in type 2 diabetes shows improved glycemic controlover 3 years. Diabetes Care 1999; 22:960–964.

61. DCCT Research Group. The absence of a glycemic threshold forthe development of long-term complications: the perspective ofthe diabetes control and complications trial. Diabetes 1996;45:1289–1298.

62. Stratton IM, Adler AI, Neil HA, et al. Association of glycemia withmacrovascular and microvascular complications of type 2 diabetes(UKPDS 35): prospective observational study. BMJ 2000;321:405–412.

63. American Diabetes Association. Standards of medical care forpatients with diabetes mellitus. Diabetes Care 2002; 25(supplA):S33–S49.

64. Boden G, Ruiz J, Kim CJ, Chen Y. Effects of prolonged glucoseinfusion on insulin secretion, clearance and action in normal sub-jects. Am J Physiol 1996; 270:E251–E258.

65. Marbury T, Huang W-C, Strange P, Lebovitz HE. Repaglinide vs gly-buride: a 1-year comparison trial. Diabetes Res Clin Pract 1999;43:155–166.

66. Diabetes Prevention Program Research Group. Reduction in theincidence of type 2 diabetes with lifestyle intervention or met-formin. N Engl J Med 2002; 346:393–403.

67. Fonseca V, Rosenstock J, Patwardhan R, Salzman A. Effect of met-formin and rosiglitazone combination therapy in patients withtype 2 diabetes mellitus. A randomized controlled trial. JAMA2000; 283:1695–1702.

68. Horton ES, Foley J, Clinkingbeard C, Foley J, Mallow S, Shen S.Nateglinide alone and in combination with metformin improvesglycemic control by reducing mealtime glucose levels in type 2diabetes. Diabetes Care 2000; 23:1660–1665.

69. Horton ES, Whitehouse F, Ghazzi MN, Venable TC, TheTroglitazone Study Group, Whitcomb RW. Troglitazone in combi-nation with sulfonylurea restores glycemic control in patients withtype 2 diabetes. Diabetes Care 1998; 21:1462–1469.

70. Einhorn D, Rendell M, Rosenzweig J, Egan JW, Mathisen AL,Schneider RL. Pioglitazone hydrochloride in combination withmetformin in the treatment of type 2 diabetes mellitus: a ran-domized, placebo-controlled study. Clin Ther 2000;22:1395–1409.

ADDRESS: Harold E. Lebovitz, MD, 416 Henderson Avenue, Staten Island,NY 10310.





treating hyperglycemia in type 2 diabetes - cleveland clinic journal

Documents