Beyond Metformin: SafetyConsiderations in the Decision-Making Process for Selecting aSecond Medication for Type 2Diabetes ManagementRef lections From a Diabetes CareEditors’ Expert ForumDiabetes Care 2014;37:2647–2659 | DOI: 10.2337/dc14-1395

The trend toward personalized management of diabetes has focused attention onthe differences among available pharmacological agents in terms of mechanismsof action, efficacy, and, most important, safety. Clinicians must select from thesefeatures to develop individualized therapy regimens. In June 2013, a nine-memberDiabetes Care Editors’ Expert Forum convened to review safety evidence for sixmajor diabetes drug classes: insulin, sulfonylureas (SUs), thiazolidinediones (TZDs),glucagon-like peptide-1 receptor agonists, dipeptidyl peptidase-4 inhibitors, andsodium glucose cotransporter 2 inhibitors. This article, an outgrowth of the forum,summarizes well-delineated and theoretical safety concerns related to these drugclasses, as well as the panelists’ opinions regarding their best use in patients withtype 2 diabetes. All of the options appear to have reasonably wide safety marginswhen used appropriately. Those about which we know the mostdmetformin, SUs,insulin, and perhaps now also TZDsdare efficacious in most patients and can beplaced into a basic initial algorithm. However, these agents leave some clinical needsunmet. Selectingnext steps is amore formidableprocess involving neweragents thatare understood less well and for which there are unresolved questions regarding riskversus benefit in certain populations. Choosing a specific agent is not as important asimplementing some form of early intervention and advancing rapidly to some formof combination therapy as needed. When all options are relatively safe given thebenefits they confer, therapeutic decision making must rely on a personalized ap-proach, taking into account patients’ clinical circumstances, phenotype, pathophys-iological defects, preferences, abilities, and costs.

Today, there are more therapy options for managing type 2 diabetes than everbefore. Primary care and specialty clinicians and the patients they advise benefitfrom having a wide range of interventions from which to choose in developingdiabetes management plans. However, this abundance also means that therapeuticdecision making has become increasingly challenging.Recommendations published in 2012 by the American Diabetes Association (ADA)

and the European Association for the Study of Diabetes (EASD) (1) set forth a flexible

1Pennington Biomedical Research Center, Louisi-ana State University, Baton Rouge, LA2University of North Carolina School of Medicine,Chapel Hill, NC3Department of Clinical and Experimental Med-icine, University of Pisa School of Medicine, Pisa,Italy4Newcastle University, Newcastle upon Tyne, U.K.5Icahn School ofMedicine atMt. Sinai, New York,NY6Diabeteszentrum Bad Lauterberg, Bad Lauter-berg, Germany7Diabetes Unit, Department of Internal Medi-cine, Hadassah HebrewUniversity Hospital, Jeru-salem, Israel8Dallas Diabetes and Endocrine Center at Medi-cal City and University of Texas SouthwesternMedical Center, Dallas, TX9Division of Endocrinology, Diabetes, and ClinicalNutrition, Oregon Health & Science University,Portland, OR

Corresponding author: William T. Cefalu, william.cefalu@pbrc.edu.

This article contains videos online at http://dx.doi.org/10.2337/dc14-1395.

© 2014 by the American Diabetes Association.Readers may use this article as long as the workis properly cited, the use is educational and notfor profit, and the work is not altered.

William T. Cefalu,1 John B. Buse,2

Stefano Del Prato,3 Philip D. Home,4

Derek LeRoith,5 Michael A. Nauck,6

Itamar Raz,7 Julio Rosenstock,8 and

Matthew C. Riddle9

Diabetes Care Volume 37, September 2014 2647

EDITO

RS'

EXPER

TFO

RUM

treatment algorithm that begins, in mostcases, with lifestyle intervention andmet-formin therapy. The algorithm progressesto dual and triple therapy and, through apatient-centered, individualized decision-making process, to numerous and in-creasingly complex combination therapyoptions involving various classes of oraland injectable medications. Recent con-sensus guidelines from the American As-sociation of Clinical Endocrinologists(AACE) (2) described a similar algorithmwith rather aggressive A1C criteria forinitiating dual therapy. Both sets ofguidelines encourage consideration ofindividual patients’ characteristics,needs, and preferences.This trend toward a more personal-

ized approach has focused attentionon the relative differences among avail-able pharmacological agents in termsof mechanisms of action, efficacy, and,perhaps most important, safety. It ison the basis of these differences thattreatment decisions for individual pa-tients must be made. To further this dis-cussion, we convened a nine-memberDiabetes Care Editors’ Expert Forum inJune 2013 to review the latest safetyevidence for six of the major diabetesdrug classesdinsulin, sulfonylureas(SUs), thiazolidinediones (TZDs), glucagon-like peptide-1 (GLP-1) receptor agonists,dipeptidyl peptidase-4 (DPP-4) inhibitors,and sodium glucose cotransporter 2(SGLT-2) inhibitors (Videos 1 and 2, avail-able at http://dx.doi.org/10.2337/dc14-1395). This article summarizesboth well-characterized and theoreticalsafety concerns related to thesedrug clas-ses, aswell as our opinions regarding theirmost efficacious use in patients with type2 diabetes. We also provide, in Table 1, alist of key topics for discussion with pa-tients who may be considering the use ofagents in these classes.

SAFETY CONSIDERATIONS FOR THEMAJOR DRUG CLASSES FOR TYPE 2DIABETES

InsulinSince its discovery in 1922, insulin hasbeen the essential treatment for type 1diabetes and has become an importantkeystone of treatment for type 2 diabe-tes. Given the natural history of type 2diabetes and the fact that insulin secre-tory deficits progress throughout thedisease process, insulin is eventually re-quired in the majority of cases. With

numerous insulin formulations nowavailable, one can design a therapy re-gimen that closely mimics normal physi-ology, offering efficacy while minimizinghypoglycemia (3). However, insulin of-ten is initiated very late in the courseof the disease (4).

Documented Safety Issues

Several concerns, both real and per-ceived, may explain why insulin initiationis often delayed. These include increasedrisks for hypoglycemia andweight gain, aswell as the misperception held by somethat insulin may not be appropriate in adisease not considered to be character-ized by insulin deficiency, but rather byhyperinsulinemia and insulin resistance.

Although hypoglycemia is the singlegreatest drawback to insulin therapy,the development of basal insulin ana-logs has provided better opportunitiesfor its safe and effective implementa-tion. A 2007 Cochrane analysis (5) re-viewed data on the long-acting insulinanalog glargine and revealed significantreductions of ;15% in the overall riskfor hypoglycemia and ;35% in the riskfor nocturnal hypoglycemia comparedwith NPH insulin in patients with type2 diabetes. This risk reduction appearsto be limited to the long-acting analogformulations, however. A 2006 Cochraneanalysis (6) comparing data on rapid-acting insulin analogs to those of regu-lar human insulin in type 2 diabetesfound little difference in hypoglycemiarates (weighted mean difference in over-all hypoglycemic episodes/patient/month of –0.2 [95% CI –0.5 to 0.1] foranalogs vs. regular insulin).

The risk for hypoglycemia also ap-pears to differ based on the type of in-sulin regimen used. For example, in the4-T (Treating To Target in Type 2 Diabe-tes) study (7), the use of basal analoginsulin was associated with fewer hypo-glycemic events, whereas premixed bi-phasic insulin formulations carried agreater hypoglycemia risk, and the riskwas higher still in regimens involvingprandial analog insulin (median events/patient/year of 1.7, 3.0, and 5.7, respec-tively; P , 0.001 for the overall compar-ison). Of interest, the different insulinformulations are also associated with dif-ferent risks for weight gain, with greatergains resulting from the use of prandialanalog and premixed formulations andless from basal insulin (7).

Potential Safety Issues

Although concerning to patients andclinicians, the risks for hypoglycemiaand weight gain may be managedthrough careful selection of insulin for-mulations and regimens, attention topatients’ clinical circumstances, and ap-propriate patient education regardinghypoglycemia prevention, detection,and treatment, as well as the impor-tance of lifestyle measures to controlweight.

Does insulin use pose a potentiallymore serious concern with regard tothe atherogenic effects of chronic hy-perinsulinemia in type 2 diabetes?Most of the data on this issue havecome from post hoc analyses suggest-ing that insulin may increase the riskfor cardiovascular events (8). However,well-designed, prospective studiesdmost notably the ORIGIN (OutcomeReduction With Initial Glargine Inter-vention) trial (9), in which patientswith short-duration type 2 diabetesand a high risk for cardiovascular dis-ease (CVD) were treated with basal in-sulin for .6 yearsdhave found thatinsulin therapy had a neutral effect oncardiovascular events compared withroutine comparator therapies. The riskfor severe hypoglycemia was relativelylow in glargine-treated subjects but stillgreater than in the standard group (1.0vs. 0.3% per year). Interestingly, the riskfor cardiovascular events was increasedin subjects with severe hypoglycemiaand not in those withmild hypoglycemiaregardless of treatment, but it was sig-nificantly higher in the standard groupthan in the glargine group (10).

The ORIGIN trial also helped to easeanother concern: the potential of long-acting insulin analogs to increase pa-tients’ risks for developing various formsof cancer. Although previous studies(11–13) reported increased cancer ratesamong patients using long-acting insu-lin, ORIGIN investigators found a neutraleffect of glargine on the risk for neopla-sia, both overall and in terms of severalspecific types of cancer.

In the ORIGIN trial, insulin use in peo-ple with prediabetic hyperglycemia wasassociated with reduced or delayed con-version to overt diabetic hyperglycemia.Along with this observation, evidence isaccumulating in support of the possibil-ity of inducing remission of new-onsettype 2 diabetes through early insulin

2648 Safety of Medications for Type 2 Diabetes Diabetes Care Volume 37, September 2014

Table 1—Topics for discussion with patients (or as part of structured education programs) regarding potential adversereactions to glucose-lowering pharmacotherapy

Problems to address Topics for patient education

Insulin Late initiation of insulin treatment Overcoming resistance to insulin therapyCareful balance of risks and benefits

Necessity for glucose monitoring Individualized self-monitoring plan (how frequently, whattimes of day)

Dependence on integrity of insulin preparation andascertainment of proper timing of administrationinto subcutaneous tissue

Proper insulin storageProper injection technique

Hypoglycemia Prevention of conditions potentially leading tohypoglycemia

Early recognition of hypoglycemiaSelf-treatment of hypoglycemiaTreatment of hypoglycemia through proxies

Weight gain Nutritional strategies to prevent weight gain or reducebody weight

Alternatives to progression to multiple injectionsafter failing “bedtime” insulin treatment

Instead of meal-time insulin:Add a GLP-1 receptor agonistAdd a TZDAdd an SGLT-2 inhibitor

SUs Hypoglycemia See hypoglycemia recommendations for insulinRisk for hypoglycemia recurrence after initially successfultreatment; necessity for continued surveillance

Weight gain See weight gain recommendations for insulinConcerns regarding cardiovascular risks Balanced discussion of current evidence (pro: observational

studies; con: UKPDS, ADVANCE)

TZDs Advantage of best record for durability Discussion of the concept of durability and its importancefor the individual patient

Advantage of potential cardiovascular benefit Discussion of cardiovascular outcomes with pioglitazoneWeight gain See weight gain recommendations for insulinRisk for fluid retention and related adverse effects(edema, congestive heart failure, anemia)

Warning signsScreening measuresAppropriate dosing

Increased risk for fractures Discussion of individual susceptibility for fracturesPreventive measuresAppropriate dosing

Increased risk for bladder cancer Discussion of this as an unresolved issueInformation about the potential quantitative impactScreening and surveillance measures

Incretin-basedtherapies

Nausea, vomiting, diarrhea (GLP-1 receptor agonists only) Rare and transient naturePossibility that drug needs to be withdrawn in aminority ofpatients

Potential pharmacotherapy for side effectsInjection site reactions and nodules (GLP-1 receptoragonists only)

Information about nature of this side effect(immunological responses potentially related toantibody formation)

Possibility that repeated episodes may suggest the need todiscontinue this treatment

Increased risk for hospitalization for heart failure (?) Information on recognition of symptomsClinical significance of study findings are undeterminedCaution for those at high risk

Increased risk for acute pancreatitis (?) Discussion of this as an unresolved issueEarly signs and symptoms of pancreatitis, behavioraladvice in such a case (seek medical advice, discontinuetreatment)

Advice for alternative treatment in the case of pastepisodes of pancreatitis

Increased risk for medullary thyroid carcinoma (?) Information about the low likelihood in the face of therarity of this disease

Advice for alternative treatment in the case of a personalor family history or with a given genetic background(multiple endocrine neoplasia syndrome type 2)

Continued on p. 2650

care.diabetesjournals.org Cefalu and Associates 2649

therapy. A recent meta-analysis (14)synthesized data from studies of short-term intensive insulin therapy in newlydiagnosed patients to determine its ef-fects on insulin sensitivity and b-celldysfunction, the two main pathophysio-logical defects responsible for hypergly-cemia and diabetes progression. Thisanalysis found that early introductionof insulin was associated with an im-provement in both conditions.Thus, although insulin is not without

risks, its use in appropriate patients atthe appropriate time does offer signifi-cant benefit. It not only improves short-term glycemic control, but also may be aviable strategy for altering the courseof the disease, enabling patients toachieve a state of remission during whichnormoglycemia can be sustained forsome time without further treatment.

Insulin: Rethinking When and How

Because most patients with type 2 dia-betes continue to secrete some amountof endogenous insulin even in the latestages of the disease, initial insulin ther-apy usually involves a basal-only regi-men aimed at suppressing overnighthepatic glucose production, therebylowering glucose levels during sleepand betweenmeals. In this regard, prev-alent fasting hyperglycemia is perhapsthe best indicator of the need for basalinsulin, even when it occurs early in thedisease. Other situations, such as inter-mittent comorbidities, surgical inter-ventions, and pregnancy, also may callfor insulin initiation, at least in the shortterm.Newly diagnosed patients usually be-

gin diabetes treatment with metformin

monotherapy (or monotherapy with anSU or incretin-based agent) and prog-ress to some form of combination oralagent therapy before starting basalinsulin. However, adding basal insulinimmediately after metformin may beappropriate in some cases; likewise,forgoing insulin in favor of a second orthird oral agent may be reasonable(although likely more costly) for somepatients. A notable exception to thestandard treatment algorithm is fornewly diagnosed patients with extreme-ly poor glycemic control and a high A1C,for whom immediate insulin therapyis recommended to ameliorate glucosetoxicity and related symptoms (1,2).Such patients often can substantially re-duce or discontinue insulin in favor of oralagents after glycemic control stabilizes.

SUsAs the first available oral agents for glu-cose lowering, SUs have a 60-year re-cord of use, second only to that ofinsulin. Thus, there is a substantial da-tabase from which to draw conclu-sions regarding their efficacy andsafety (15–19). These agents stimu-late endogenous insulin release in anonglucose-mediated manner by clos-ing ATP-sensitive potassium channelslocated on pancreatic b-cells (20).They are widely used in the U.S. andaround the world, accounting for ;25%of newly initiated oral therapies for di-abetes (21). In the U.S., they are consid-ered one of several second-line optionsafter metformin for most people withtype 2 diabetes and a viable first-line al-ternative for patients who cannot takemetformin (1,2).

Drugs in this class can be divided intotwo groups: historical agents that are nolongerwidely used (including carbutamide,acetohexamide, chlorpropamide, tol-butamide, and tolazamide) and cur-rently used agents (including glyburide[also known as glibenclamide], gliclazide,glipizide, and glimepiride). Most of thehistorical SUs have adverse effects thathave limited their use (22–24), and ithas been suggested that the routineuse of glyburide should also be re-stricted (25). Glyburide interferes withcardiac ischemic preconditioning (26);compared with the other modern SUs,it may be associated with higher mor-tality rates when coadministered withmetformin (27) and after hospitaliza-tion for myocardial infarction (MI)(22), and it causes more hypoglycemia(28).

To our way of thinking, the ideal anti-hyperglycemic agent would be easy toadminister, unlikely to cause symptom-atic side effects that pose barriers toadherence, inexpensive, reliably effica-cious, and safe. By such standards, it canbe argued that the remaining modernSUs do well (although they do leavesome clinical needs unmet). Glimepirideand extended-release preparations ofglipizide and gliclazide can be givenonce daily and rarely cause symptomaticside effects other than hypoglycemia.They appear to be as effective as ormore effective than other oral agentsin terms of A1C reduction, life expec-tancy, and quality-adjusted life-years(29,30). And SUs are remarkably inex-pensive compared with newer oralagents (31).

Table 1—Continued

Problems to address Topics for patient education

SGLT-2 inhibitors Genital infections (Candida and other fungi) Signs and symptomsPreventive measures (hygiene)Consider other treatments after repeated occurrence

Urinary tract infections (bacterial) Signs and symptoms (including those of more severe,ascending infections [urosepsis])

Preventive measures (hygiene)Consider other treatments after repeated occurrence

Negative fluid balance Information about potential consequences (too greata drop in blood pressure, impairment of kidney function)

Elevated LDL cholesterol Impact on overall cardiovascular riskTreatment options (statins, target values, careful dose-finding)

Risk for bladder cancer Discussion of this as an unresolved issueInformation about the potential quantitative impactScreening and surveillance measures

2650 Safety of Medications for Type 2 Diabetes Diabetes Care Volume 37, September 2014

Documented Safety Issues

Hypoglycemia and weight gain are themain well-documented safety issues re-lated to SUs (19,32). Starting an SU typ-ically leads to a weight gain of ;2 kgdepending on prior A1C level and hypo-glycemia rates two to three times higherthan with other agents (29). Althoughsignificant, these risks usually can bemanaged with appropriate patient se-lection, attention to dosing, and ade-quate patient education. Patients withcompromised renal functioning are par-ticularly susceptible to hypoglycemiafrom SUs; hence, their use in this popu-lation should be avoided. Allergy andother idiosyncratic effects are rare.

Potential Safety Issues

Concerns about the potential effects ofSUs on cardiovascular risk, possibly re-lated to interference with cardiac ische-mic preconditioning (33), have existedsince at least 1970, with the controversialresults from the University Group Diabe-tes Program (UGDP) (34) suggesting thattolbutamide might be associated with anincreased risk for cardiovascular mortal-ity. Attention to this issue subsided some-what with the emergence of other SUs(35), but findings from retrospective anal-yses of associations between oral diabe-tes drugs and cardiovascular outcomeshave been inconsistent (36–40).Recent trials and analyses have been

more reassuring. In 1998, the UK Pro-spective Diabetes Study (UKPDS) group(41) reported no difference in the ratesof MI or diabetes-related death amongsubjects receiving chlorpropamide, gly-buride, or insulin. Researchers in theADVANCE (Action in Diabetes and Vas-cular Disease: Preterax and DiamicronMR Controlled Evaluation) trial (42)reached a similar conclusion in 2008,reporting no significant differencesbetween intensive glucose control in-volving gliclazide and other drugs as re-quired and conventional care in majormacrovascular events, death from car-diovascular causes, or death from anycause. The RECORD (Rosiglitazone Eval-uated for Cardiovascular Outcomes inOral Agent Combination Therapy forType 2 Diabetes) study (43) similarlyfound no significant differences in therisk for cardiovascular events amongcombination therapy with rosiglitazoneand metformin, rosiglitazone and SU,or metformin and SU. Although not

designed to evaluate cardiovascular out-comes, the ADOPT (A Diabetes OutcomeProgression Trial) study (44), whichcompared the durability of effective-ness of monotherapy with metformin,rosiglitazone, or glyburide, reportednom-inally lower rates of cardiovascular eventsin patients taking glyburide than in thosetaking either rosiglitazone or metformin.

A 2013 meta-analysis (45) of 62 trialsreporting major cardiovascular eventswith SUs versus various comparatorshas provided perhaps the best synthesisof data to date. This analysis found anoverall odds ratio (OR) for major car-diovascular events with SU treatmentversus comparators of 1.08 (95% CI0.86–1.36), thus detecting no signal forcardiovascular risk. However, the au-thors urged cautious interpretation oftheir results given limitations of trialquality and potential underreporting ofinformation on cardiovascular eventsand mortality.

Further evidence in support of SUscomes from modern epidemiologicalstudies such as a recent retrospectivehealth system database analysis fromAlberta, Canada (46), which sought todetermine whether patients taking SUsat the time of acute coronary syndromewere more likely to have poor out-comes. Its results indicated an adjustedOR for using versus not using an SU of1.06 (95% CI 0.89–1.26). One possibleexplanation for the lack of an apparentassociation between the use of SUs andhigher cardiovascular risk in recentstudies is the decline in the use of gly-buride in favor of SUs with a lower pro-pensity to cause hypoglycemia andwithout significant effects on ischemicpreconditioning.

SUs: a Proven and Still Valuable Option

As new drug classes have been intro-duced, promotional efforts have sug-gested that SUs are an outmoded classto be replaced by newer agents. How-ever, we do clearly know the efficacy ofthese agents, just as we are aware oftheir limitations of hypoglycemia andweight gain. Objectively, one could ar-gue that, given the wealth of clinical ex-perience, new safety concerns are notlikely to emerge. Although poor durabil-ity of effectiveness has been a majorcriticism, participants assigned to stan-dard therapy in the ORIGIN trial, treatedmainly with metformin and an SU,

maintained glycemic control (averageA1C 6.5%) for 6 years (9). SUs also offerthe advantages of ease of administra-tion, good tolerability, and low cost.

Still, clinicians prescribing SUs musttake care to help patients avoid hypogly-cemia. Appropriate patient selection isimportant, especially when consideringSUs for patients who are elderly or frail,have a history of hypoglycemia or hypo-glycemia unawareness, or have renaldysfunction or other conditions or co-morbidities likely to place them at highrisk. Glyburide should rarely be consid-ered because of its greater tendency tocause hypoglycemia.

TZDsTZDs have been a source of both enthu-siasm and controversy since troglita-zone, the first agent in the class, wasapproved in the U.S. in 1997 to addressinsulin resistance. This ushered in a newparadigm of treatment and was fol-lowed by rosiglitazone and pioglitazonein 1999 (47). TZDs are synthetic ligandsthat activate PPAR-g nuclear receptorsin adipose tissue, skeletal muscle, andthe liver (48). They act primarily to im-prove insulin sensitivity and reduce he-patic glucose production and are, todate, the only class of agents specificallytargeting insulin resistance, which is oneof the primary defects in type 2 diabetesand other insulin-resistant states suchas impaired glucose tolerance and poly-cystic ovary syndrome.

The main appeal of TZDs is their du-rability of effect in lowering A1C (49–52), as well as their potential to alterthe natural progression of diabetes in away that cannot be achieved with otheroral agents (43,44). However, seekingthis goal would necessitate initiatingTZD therapy early in the course of thedisease, making consideration of safetyissues surrounding this drug class partic-ularly important.

Documented Safety Issues

Troglitazone was withdrawn from theU.S. market in 2000 because of concernsregarding hepatotoxicity (53). In themid-2000s, several articles were pub-lished suggesting that rosiglitazonewas associated with excess cardiovascu-lar events, specifically MIs. As a result,the use of rosiglitazone was tightly re-stricted in 2011 (54). However, theRECORD study (43), reported in 2009,had documented no increased risk for

care.diabetesjournals.org Cefalu and Associates 2651

cardiac adverse events. A recent U.S.Food and Drug Administration (FDA)–mandated readjudication of the RECORDfindings (55), which confirmed no in-crease in overall cardiovascular risk withrosiglitazone, prompted the FDA to lift itsrestrictions on this agent but maintainspecific warnings related to increasedrisks for congestive heart failure andbone fractures (56).For now, pioglitazone remains the

most widely used TZD, and research sug-gests that its safety profile is less con-troversial than other agents in this class.Although the PROactive (ProspectivePioglitazone Clinical Trial in Macrovas-cular Events) trial (57), a cardiovascularoutcomes randomized controlled trial(RCT), failed to meet statistical sig-nificance for a primary end point com-posed of numerous macrovascularoutcomes, a subanalysis of a secondaryend point encompassing MI, stroke, andcardiovascular death found a modestbut statistically significant reduction.Other well-recognized concerns re-

lated to the TZDs include weight gain,fluid retention leading in some cases toedema or congestive heart failure, andincreased risk for bone fractures, oftenat peripheral sites, which is perhaps oneof the greatest concerns with this class(58,59). Mechanistic studies haveshown that there is a theoretical possi-bility that TZDs could alter the differen-tiation of mesenchymal stem cellstoward the formation of fat and awayfrom the formation of bone. Both RCTs(60) and pharmacoepidemiological anal-yses (61) have revealed an increasedabsolute risk for bone fractures of ;1%in men and 3% in women.Fluid retention usually can be man-

aged through patient counseling andcareful patient selection and safely miti-gated with thiazide diuretics. Althoughthe potential for heart failure has beenan important safety consideration, indi-vidual studies and meta-analyses sug-gest that there is a small increase inthe absolute risk for heart failure requir-ing hospitalization but no increase infatal heart failure (62).TZDs also have been associated with

an increased incidence of macularedema in retrospective, observationalstudies with an approximately twofoldincreased risk (63). However, this hasnot been consistently observed in RCTs(64). Certainly, patients treated with

TZDs should have annual dilated fundu-scopic exams and should seek attentionfor more than transient visual changes.

Other common TZD side effects in-clude anemia, rare instances of in-creased creatine phosphokinase (butnot serious myositis), variable changesin lipids, and possible hepatic effects.The latter necessitate the avoidance ofthese agents in patients with substantialliver disease, although some studieshave suggested that TZDs actually maybe beneficial in the setting of fatty liverdisease (58,59,65); furthermore, be-cause TZDs are not excreted throughthe kidney, they are potentially usefulfor patients with chronic kidney disease.Finally, because the risk for hypoglyce-mia may be increased when TZDs areused in combination with insulin or in-sulin secretagogues, doses of these lat-ter agents should be reduced in suchcircumstances.

Potential Safety Issues

As accumulating evidence lessens con-cern about cardiovascular risk, the mainremaining potential, but as yet un-proven, safety issue related to TZDsis a possible link to increased rates ofbladder cancer with pioglitazone. Thisissue arose because of imbalances de-tected in the development of bladdertumors in preclinical animal studies ofpioglitazone and of experimental drugswith dual PPAR-g and PPAR-a activity(58,66). Imbalances in the number of blad-der cancer cases have also been reportedin some RCTs, most notably PROactive(57), which found a nonsignificant in-crease in bladder tumors in pioglitazone-treated patients. The most robustpharmacoepidemiological study to date(67), which is being conducted at therequest of the FDA, detected no overallsignal for bladder cancer risk at 5 years,but did suggest a modest increase in riskonly in patients with long exposure to oron high doses of pioglitazone. However,in more recent analyses, the statisticalsignificance of these findings disap-peared by 8 years of follow-up and thecancers detected were almost all in anearly stage (68).

Further study will be required to fullyelucidate this potential risk. For now,prescribing information for pioglitazonesuggests that it not be used in patientswith a history of bladder cancer and thatall patients should be instructed to seek

medical attention for any urinary symp-toms that may develop (58).

TZDs: Proceed With Caution

Until the issues enumerated above areput to rest, TZDs most likely will be con-sidered as third- or fourth-line optionsafter combinations of other agents. Insuch situations, the balance of benefitto risk is still quite strongly in favor ofthese drugs.

Incretin-Based Therapies: GLP-1Receptor Agonists and DPP-4 InhibitorsGLP-1 is a gut-derived incretin hormonethat stimulates insulin secretion in aglucose-dependent manner, suppressesglucagon secretion similarly, slows gas-tric emptying, reduces appetite, andmay expand b-cell mass (69). The devel-opment of two drug classes that actthrough stimulation of GLP-1 receptorsto enhance incretin actiondGLP-1 re-ceptor agonists (incretin mimetics) andDPP-4 inhibitors (incretin enhancers)dhas been an important advancement inthe pharmacological treatment of type 2diabetes (70) because these agents ef-fectively control glycemia without caus-ing hypoglycemia or weight gain.

GLP-1 receptor agonists are peptidesthat mimic native GLP-1, binding to itsreceptors to elicit the same effects, butat much higher pharmacological levelsthan the physiological profiles. Agentsin the class currently available in theU.S. include exenatide twice daily, exena-tide once weekly (QW), liraglutide oncedaily, and albiglutide QW, all of whichare administered through subcutaneousinjection. Other agents in developmentinclude dulaglutide QW and semaglutideQW, as well as lixisenatide once daily,which is available in Europe.

DPP-4 inhibitors act to suppress theproteolytic enzyme (i.e., DPP-4) thatnormally degrades endogenous GLP-1and thereby increase the concentrationof intact, biologically active GLP-1 andaugment its interaction with receptors.DPP-4 inhibitors available globally and inthe U.S. include alogliptin, linagliptin,saxagliptin, and sitagliptin. Vildagliptin isalso globally available except in the U.S.,and other DPP-4 inhibitors are availablein Japan. All are administered orally.

The mechanisms of incretin-basedtherapies, and the clinical trials demon-strating the efficacy and safety profileson which our understanding of them isbased, have been extensively reviewed

2652 Safety of Medications for Type 2 Diabetes Diabetes Care Volume 37, September 2014

elsewhere (70–73). Briefly, agents inboth classes have been shown to lower,to varying degrees, A1C, fasting plasmaglucose, and postprandial glucose.Whereas GLP-1 receptor agonists alsoslow the rate of gastric emptying in dif-ferent degrees based on their pharma-cokinetic profiles and can cause a senseof satiety leading to reduced food intakeand moderate weight loss, DPP-4 inhib-itors do not slow the rate of gastric emp-tying and are weight-neutral. Neithertype of therapy increases the risk forhypoglycemia except when associ-ated with SUs or insulin. Clinical trialsof agents in both classes also haveshown improvements in some surro-gate markers of pancreatic b-cell func-tion in type 2 diabetes. Both also mayyield modest improvement in lipid pro-files and systolic blood pressure levels,although short-term studies have shownonly neutrality for cardiovascular events(74,75).Current guidelines (1,76) recommend

both types of incretin-based therapiesfor use as monotherapy (mostly in pa-tients for whommetformin is not an op-tion) and in combination with otheragents (most often metformin) if, forexample, treatment priorities includereducing the risk for hypoglycemia andcontrolling body weight.

Documented Safety Issues

The most common treatment-relatedadverse effects of GLP-1 receptor ago-nists are gastrointestinal in nature andinclude nausea, vomiting, and diarrhea,which are usually mild and tend to sub-side over time but in some patients maybe intermittent and lead to eventual dis-continuation (77,78). A dose-titrationstrategy has been found to reduce theincidence of nausea (79), but ,5% ofpatients in clinical trials need to discon-tinue treatment because of such sideeffects (77,78). However, discontinua-tion rates in clinical practice are greater,mainly as a result of gastrointestinal in-tolerance without the type of supportsystem typically available in clinical tri-als and perhaps disenchantment whenthese agents are initiated mainly inhopes of achieving weight loss.Other documented but infrequent

concerns with GLP-1 receptor agonistsinclude injection site reactions (80) and,particularly with exenatide QW, the de-velopment of transient small nodules

around injection sites related to the vis-cous nature of the formulation (81).

DPP-4 inhibitors have been shown tohave a very good safety and tolerabilityprofile similar to that of placebo (72,82).Unlike GLP-1 receptor agonists, DPP-4inhibitors are not associated with gas-trointestinal adverse events. Early signsthat their use may be associated withmore upper respiratory tract and uri-nary tract infections have not beenconfirmed (82,83). Evidence to date sug-gests that DPP-4 inhibitors do not af-fect cardiovascular risk with ;2 yearsof exposure in high-risk populations(82).

Severe hypoglycemia has not beenobserved in trials of any incretin-basedmonotherapy (77,84–90), but mild tomoderate hypoglycemia has occurredin 0–12% of patients, depending onthe agent studied (77,84–86,88–91). Ahigher hypoglycemia rate has beendocumented when such agents areused in combination with SUs or insulin(78,92–99); hence, decreasing the dos-age of these concomitant agents is rec-ommended.

Potential Safety Issues

Low-frequency findings of as yet un-known clinical significance have raisedadditional questions regarding thelong-term safety of incretin-based ther-apies (100,101). The results of ongoingprospective studies will help to addressthese issues.

Heart Failure. The SAVOR (Saxagliptin As-sessment of Vascular Outcomes Re-corded in Patients with Diabetes Mellitus)study (74) found an increased rate ofhospitalization for heart failure com-pared with standard care, althoughoverall cardiovascular events, includingheart failure, did not increase. Datafrom the EXAMINE (Examination ofCardiovascular Outcomes with Alogliptinversus Standard of Care in PatientsWith Type 2 Diabetes Mellitus andAcute Coronary Syndrome) trial (75)were consistent with this observationbut even less robust. Whether the ef-fect is real and, if so, the extent of itsclinical significance remains uncertainpending the completion of other stud-ies (102,103). In the meantime, cautionseems indicated in people with or athigh risk for heart failure.

Acute Pancreatitis. Cases of pancreatitishave been reported in animals (104–

107) and humans (108) treated withGLP-1 receptor agonists or DPP-4 inhib-itors. The question of whether suchcasesmay potentially be caused by treat-ment with incretin-based therapies re-mains unanswered.

Animal studies have been inconsis-tent, although most preclinical modelscannot reflect the human pancreatitismodel. Some have described histologi-cal changes consistent with damage tothe exocrine pancreas with exenatide(104,105) and sitagliptin (107), but notwith liraglutide (106). Other studies inmice have documented improvementof experimentally induced acute pancre-atitis with exenatide (109) and an anti-inflammatory cytokine response withliraglutide (110).

Small imbalances in the number ofcases of acute pancreatitis reported inseveral of the clinical development pro-grams of incretin-based therapies ledthe FDA to require a warning regardingthe possibility of developing acute pan-creatitis and a recommendation to avoidthese agents in people with a historyof pancreatitis (111). Retrospective, ob-servational studies of acute pancreatitisfrom incretin-based therapies havefound ORs near 1, suggesting no in-creased risk. However, these ratioshave wide CIs due to the small numberof cases (112–116). Furthermore, thesetypes of reports, obtained from insur-ance claims data, electronic medical re-cords, or prescription databases, are notprospectively designed to answer spe-cific questions regarding the safety oftherapeutic interventions. Thus, anyfindings can only be regarded as eitherhypothesis-generating or merely sug-gestive, but certainly cannot be viewedas definitive. A single study reportingthat patients taking exenatide or sita-gliptin had a tenfold higher likelihoodof developing pancreatitis was basedon data from the FDA Adverse Event Re-porting System, which is susceptible toreporting bias (117). A case-controlstudy looking at the rate of hospitaliza-tion for pancreatitis found a higher ORamong patients taking “incretin-basedtherapies,” but yielded no significantfindings for either GLP-1 receptor ago-nist or DPP-4 inhibitor therapy when an-alyzed separately (118). The possibilityexists that a combination of gastrointes-tinal symptoms and spontaneously ele-vated lipase activity (as is typical for a

care.diabetesjournals.org Cefalu and Associates 2653

type 2 diabetic population [119]) weremisdiagnosed as pancreatitis in at leastsome of these cases.As recently recommended in this

journal (120), patient-level safety datafrom the multiple ongoing, cardiovascu-lar outcomes RCTs of incretin-basedtherapies should be combined to pro-vide sufficient statistical power formore conclusive meta-analyses. Suchefforts should yield more definitive an-swers regarding pancreatitis and incre-tin-based therapies in the coming years.

Chronic Pancreatitis and Pancreatic

Cancer. Concerns have been raised regard-ing the potential role of incretin-basedtherapies in inducing chronic pancreati-tis, which, in the long run, could promotepreneoplastic lesions, thus raising therisk for pancreatic cancer (101).Whereassome animal studies have shown histo-logical changes indicative of chronic pan-creatitis with exenatide (104,105,107),a study of liraglutide noted pancreatitisas a rare finding unrelated to dose andwith similar numbers in placebo-treatedrats, mice, and monkeys (106). Recentlyreported tissue analysis from organ do-nors who had type 2 diabetes and tookincretin-based agents found pancreaticabnormalities; these reports, and pub-lished reviews of them, included com-ments on methodological issues, aswell as preexisting conditions that mayhave predisposed to these findings(121–123).To date, there have been no reported

cases of clinically identifiable chronicpancreatitis or pancreatic cancer arisingafter initiation of incretin-based thera-pies. However, given the relatively shorttime these agents have been in use andthe typically slow development of pan-creatic carcinomas (124), it is too earlyto know.

Medullary Thyroid Carcinoma. One studyhas shown that exposure to long-actingGLP-1 receptor agonists increased thy-roid C-cell hyperplasia, adenomas, andmedullary thyroid carcinomas in miceand rats (125). It is important to note,however, that such abnormalities alsooccur spontaneously in these species,especially in male rats, in which medul-lary thyroid carcinoma also developedwith placebo. Rodent C-cell lines producecyclic AMP and secrete calcitonin. Sim-ilar human cell lines do not show sucheffects (125), and long-term high-dose

GLP-1 receptor agonist treatment intype 2 diabetic or obese humans did notelicit elevations in plasma calcitonin(126). Rodent C cells have considerablymore GLP-1 receptors than their humancounterparts (125). Thus, GLP-1 stimula-tion probably does not provoke prolifer-ative responses in human C cells, and nosuch cases have been reported. Medul-lary thyroid carcinoma is extremely rarein humans (127).

Incretins: Safety Signals Still Require

Vigilance

Although the known and anticipatedbenefits of incretin-based therapies ap-pear to be substantial, the potentialrisks for serious rare events remain con-troversial and in need of further elu-cidation through long-term studies.Safety concerns related to alterationsof the exocrine pancreas and thyroid,while deserving of further study, arenot yet firmly substantiated. An excel-lent debate on this topic was publishedin a previous issue of Diabetes Care(100,101). The FDA and the EuropeanMedicines Agency (EMA) carefully scru-tinized preclinical and clinical data onthis topic and recently came to a similarconclusion that no changes in recom-mendations are necessary until morefirm data are available (128).

SGLT-2 InhibitorsSGLT-2 inhibitors are the newest class ofmedications and, as such, have the leastavailable research and clinical data re-garding their effective use and adverseeffects. The mechanisms and efficacy ofthese agents have been reviewed else-where (129–131). The main advantageof SGLT-2 inhibitors is a completely dif-ferent mechanism of action; they workprimarily to lower the renal threshold toglucose, leading to increased glucoseexcretion and decreased plasma glucoselevels. Because this mode of action isnot dependent on insulin secretion,agents in this class can be considered foruse in combination with other glucose-lowering agents throughout the courseof type 2 diabetes and potentially canbe a useful add-on therapy to insulin intype 1 diabetes.

The efficacy of SGLT-2 inhibitors ap-pears to be similar to that of other anti-hyperglycemic agents; they significantlyreduce fasting and postprandial glucoselevels, leading to A1C reductions of;0.5–1.0%. These agents also induce

mild osmotic diuresis and a net loss ofcalories, yielding a slight reduction inblood pressure and a net weight loss.Because SGLT-2 inhibitors have no ef-fects on glucose-dependent endoge-nous insulin secretion and do notcompletely halt glucose reabsorption,they carry a low risk for severe hypogly-cemia (132–134).

Two SGLT-2 inhibitorsdcanagliflozinand dapagliflozindare currently avail-able in the U.S. and elsewhere; bothare administered in once-daily oral tab-lets (135,136).

Empagliflozin was recently approvedby the EMA, and several other SGLT-2inhibitors are available in Japan.

Documented Safety Issues

Genital mycotic infections and urinarytract infections have been the mostcommonly reported adverse events as-sociated with SGLT-2 inhibition to date(137–140). These occurrences are usu-ally mild to moderate and responsive totreatment, and they rarely result in dis-continuation of therapy (141,142). Stud-ies have also shown that some adverseevents related to osmotic diuresis (e.g.,polyuria) are greater with SGLT-2 inhib-itors than with placebo. In addition,there have been some unusual labora-tory parameters, such as changes inhemoglobin, plasma magnesium, andblood urea nitrogen levels and, interest-ingly, increases in both HDL and LDL cho-lesterol. Furthermore, studies havedocumented some volume-related ad-verse events, such as hypotension andpostural dizziness, which require furtherstudy and could have implications forthe use of these agents, particularly inthe elderly population (141,142). Mildhypoglycemia has been documentedwith concurrent use of SGLT-2 inhibitorsand insulin or insulin secretagogues(134,143). Hence, reductions in the dos-ages of such agents are recommendedwhen used in conjunction with SGLT-2inhibitors (135,136).

Potential Safety Issues

Although canagliflozin was the firstSGLT-2 inhibitor to receive FDA ap-proval, a similar agent, dapagliflozin,was approved earlier in Europe and inthe U.S. in 2014 (144). Dapagliflozin’ssafety and effectiveness were evaluatedin 16 clinical trials involving 9,400 patientswith type 2 diabetes taking the agent asmonotherapy or in combination with

2654 Safety of Medications for Type 2 Diabetes Diabetes Care Volume 37, September 2014

other diabetes pharmacotherapies.These trials showed improvement inA1C and found that the most commonside effects were genital fungal infec-tions and urinary tract infections. How-ever, because of a numerical imbalancein bladder cancers seen in dapagliflozinusers, the agent is not recommendedfor patients with active bladder cancer(144). Similar studies of canagliflozinhave not confirmed a bladder cancereffect; taken together, these studieshave shown no conclusive increasedrisks for bladder or breast cancer(141,142).Perhaps the greatest concern with

this class of medications is simply thefact that they have not been studiedlong enough to reach definitive conclu-sions about their long-term safety, ei-ther in the general population or inspecific subgroups such as the elderly.Recent studies of volume-related eventsin the elderly appeared to indicate thatthese agents remain well tolerated andbeneficial even in that high-risk popula-tion (141), although further study isneeded. Additional research is alsoneeded to determine the possible ef-fects of SGLT-2 inhibition on long-termcardiovascular risk factors and to bettercharacterize any possible increased risksfor breast or bladder cancer with long-term use. Large cardiovascular out-comes trials are under way to satisfyFDA requirements for cardiovascularsafety; these trials will also provide ad-ditional data regarding renal safety andany cancer-related concerns.

SGLT-2 Inhibitors: Much Potential, Many

Unanswered Questions

Through their unique focus on the kid-ney, SGLT-2 inhibitors have turned a con-dition once viewed as indicative of poorglycemic controldglucosuriadinto ameans of achieving decreased plasmaglucose concentrations. These agentscan potentially benefit any patients withdiabetes who have adequate renal func-tion. In this regard, some studies havesuggested favorable effects in patientswith type 1 diabetes (145,146).In addition to their favorable effects

on glucose andweight, SGLT-2 inhibitorsalso correct the excessive activity of so-dium reabsorption that is found in type 2diabetes and that contributes to hy-pertension, CVD, and other long-termconsequences. In theory, then, these

agents may be particularly useful in pa-tients who are obese and hypertensiveand who have some degree of estab-lished CVD.

As with the incretin-based classes,SGLT-2 inhibitors suffer from a lackof long-term clinical experience andresearch data, leaving numerous unan-swered questions regarding their long-term safety. Although themost frequentadverse effectsdgenital mycotic andurinary tract infectionsdappear to bemore an issue of patient tolerancethan of safety, we await more data re-garding possibly significant metabolicand other side effects, as well as furtherelucidation of potential cancer risks.

CONCLUSIONS

Although no pharmacological agent iswithout some risk, all of the options dis-cussed above appear to have wide mar-gins of safety when used appropriately.Many years of clinical experience withthe older agents (i.e., insulin, SUs, andTZDs) and a rapidly growing under-standing of the newer ones (i.e., GLP-1receptor agonists, DPP-4 inhibitors, andSGLT-2 inhibitors) leave us better posi-tioned than even a few years ago tohelp patients achieve and maintainglycemic control.

The question remaining, then, is howto determine the most appropriate rolefor any given agent in individual pa-tients. Luckily, the agents and drugclasses about which we have the mostknowledgedmetformin, SUs, insulin,and perhaps now TZDsdare quite suc-cessful in most patients and can beplaced into a basic algorithm such asthose recommended by ADA/EASD andAACE (1,2) and prescribed with someconfidence. Selecting next steps as apatient’s diabetes progresses remainsa more formidable process involvingnewer agents that are understood lesswell and for which there are unresolvedquestions regarding risk versus benefit incertain populations.

Shedding more light on the best usesof these agents will require greater ef-fort along two fronts. First, additionalevidence is needed regarding which pa-tients aremost and least likely to benefitfrom each pharmacological option. Re-search studiesmust identify not only theoverall effects of these agents on clinicaloutcomes, but also the high- and low-risk subgroups for each in terms of

physiology, demographics, comorbid-ities, and other factors. Second, as em-phasized in the ADA/EASD guidelines(1), clinicians must solicit and respectthe needs and desires of patients aspartners in the decisionmaking requiredto most effectively manage diabetes.

Perhaps the most important messageis that the selection of one agent overanother is not as important as imple-menting some form of early interven-tion and advancing rapidly to someform of combination therapy as needed.Studies such as ORIGIN (9) have demon-strated that even the most conventionalregimen can have excellent durabilitywhen started early. The classes ofagents that have been available the lon-gest are well-proven, cost-effective, andsensible options for many patients. Forother patients, initiation of an injectableagent (early insulin in combination withmetformin alone or with another oralagent or early treatment with a GLP-1receptor agonist) may be the preferredchoice.

When all options are relatively saferelative to the benefits they confer,therapeutic decision making reliesmore than ever on a personalized ap-proach taking into account patients’clinical circumstances, phenotype, spe-cific pathophysiological defects, prefer-ences, abilities, and costs. Regardless ofthe specific therapy selected, the over-arching goal should be to safely achieveglycemic control at the earliest possiblestage with the least risk for adverseevents, thereby increasing the likeli-hood of long-term durability of controland avoidance of complications in thefuture.

Acknowledgments. Writing and editing sup-port services for this article were provided byDebbie Kendall of Kendall Editorial in Richmond,VA. The Editorial Committee recognizes the workof the journal and contributions such as this ExpertForum would not be possible without the dedi-cated work and continued support from manyindividuals. Specifically, the planning, logistics, andfunding of themeeting and the incredible editorialsupportwould not have beenpossiblewithout thetireless effort of Chris Kohler and his staff at theADA publications office. In addition, the EditorialCommittee would like to thank Lyn Reynolds andher staff in the ADA editorial office for support inthis regard. The Editorial Committee would alsolike to thank Anne Gooch at the PenningtonBiomedical Research Center for her valuableassistance in helping to organize the ExpertForum.

care.diabetesjournals.org Cefalu and Associates 2655

Duality of Interest. W.T.C. has served as aconsultant to or received research fundingthrough his institution from AstraZeneca,GlaxoSmithKline, Halozyme Therapeutics, IntarciaTherapeutics, Johnson & Johnson, Lexicon,MannKind, Novo Nordisk, Nutrition 21, Procter& Gamble, Sanofi, and Shire Development. J.B.B.is a consultant to PhaseBio Pharmaceuticals andhas received payments, reimbursement fortravel, and stock options for that effort. Healso has been an investigator and/or consultantwithout any direct financial benefit under con-tracts between his employer (the University ofNorth Carolina) and the following companies:Amylin Pharmaceuticals, Andromeda, Astellas,AstraZeneca, Bayhill Therapeutics, BoehringerIngelheim, Bristol-Myers Squibb, Catabasis,Cebix, CureDM, Diartis Pharmaceuticals, ElcelyxTherapeutics, Eli Lilly, Exsulin, Genentech, GI Dy-namics, GlaxoSmithKline,HalozymeTherapeutics,F. Hoffmann-La Roche, Intarcia Therapeutics,Johnson & Johnson, Lexicon, LipoScience, Macro-Genics, Medtronic MiniMed, Merck, Meta-bolic Solutions Development Co., Metabolon,Metavention, Novan, Novo Nordisk A/S, NovellaClinical, Orexigen Therapeutics, Osiris Therapeu-tics, Pfizer, PhaseBio Pharmaceuticals, Quest Di-agnostics, Rhythm Pharmaceuticals, Sanofi,Spherix, Takeda, Tolerx, TransPharma Medical,TransTech Pharma, Veritas, and Verva. S.D.P.has served as a consultant or speaker for orhas received research funding fromAstraZeneca,Boehringer Ingelheim, Bristol-Myers Squibb, EliLilly, GlaxoSmithKline, Intarcia Therapeutics,Janssen, Merck, Novartis, Novo Nordisk, RochePharmaceuticals, Sanofi, and Takeda. P.D.H. hasreceived (or institutions with which he is associ-ated have received) funding for his educational,advisory, and research activities from AntriaBio,AstraZeneca/Bristol-Myers Squibb, BoehringerIngelheim, Eli Lilly, GlaxoSmithKline, Hanmi,Janssen/Johnson & Johnson, Merck, Novo Nor-disk, Roche Diagnostics, Roche Pharmaceuticals,Sanofi, Skyepharma, and Takeda. D.L. has servedas a consultant or received research fundingfrom AstraZeneca/Bristol-Myers Squibb, Janssen,Merck, and Sanofi. M.A.N. has served onadvisory boards for, received honoraria or con-sulting fees from, or received research fundingfrom Amylin, AstraZeneca, Berlin-Chemie, Boeh-ringer Ingelheim, Bristol-Myers Squibb, DiartisPharmaceuticals, Eli Lilly, GlaxoSmithKline,F. Hoffmann-La Roche, Intarcia Therapeutics,Janssen, MannKind, Merck, MetaCure, Novartis,Novo Nordisk, Roche Pharmaceuticals, Sanofi,Takeda, Versartis, and Wyeth Research. I.R. hasserved on the advisory boards or speakers’ bu-reaus of or served as a consultant for Androm-eda, AstraZeneca/Bristol-Myers Squibb, Eli Lilly,Medscape, Merck, Novartis, Novo Nordisk, andSanofi. He is a stock shareholder in Insuline.J.R. has served on advisory boards, receivedhonorarium or consulting fees, or receivedresearch funding from Amylin, AstraZeneca,Boehringer Ingelheim, Bristol-Myers Squibb,Daiichi Sankyo, Eli Lilly, GlaxoSmithKline, Halo-zymeTherapeutics, Intarcia Therapeutics, Janssen,Johnson & Johnson, Lexicon, MannKind, Merck,Novartis, Novo Nordisk, Pfizer, Roche, Sanofi,and Takeda. M.C.R. has received honoraria forspeaking or consulting and/or research fundingthrough his institution from Amylin/Bristol-Myers

Squibb/AstraZeneca, Elcelyx, Eli Lilly, Sanofi, andValeritas. This duality of interest has been re-viewed andmanaged by Oregon Health & ScienceUniversity. No other potential conflicts of interestrelevant to this article were reported.

References1. Inzucchi SE, Bergenstal RM, Buse JB, et al.;American Diabetes Association (ADA); Euro-pean Association for the Study of Diabetes(EASD). Management of hyperglycemia in type2 diabetes: a patient-centered approach: posi-tion statement of the American Diabetes Asso-ciation (ADA) and the European Association forthe Study of Diabetes (EASD). Diabetes Care2012;35:1364–13792. Garber AJ, Abrahamson MJ, Barzilay JI, et al.;American Association of Clinical Endocrinolo-gists. AACE comprehensive diabetes manage-ment algorithm 2013. Endocr Pract 2013;19:327–3363. Home P, Riddle M, Cefalu WT, et al. Insulintherapy in people with type 2 diabetes: oppor-tunities and challenges? Diabetes Care 2014;37:1499–15084. Zilov AV, Wenying Y, Gonzalez-Galvez G,et al. Prevalence of complications of diabetesin people with type 2 diabetes: data from Asia,Europe and Latin America from the A1chievestudy. Diabetes 2011;60(Suppl. 1):A6565. Horvath K, Jeitler K, Berghold A, et al. Long-acting insulin analogues versusNPH insulin (humanisophane insulin) for type 2 diabetes mellitus.Cochrane Database Syst Rev 2007;2:CD0056136. Siebenhofer A, Plank J, Berghold A, et al.Short acting insulin analogues versus regularhuman insulin in patients with diabetes melli-tus. Cochrane Database Syst Rev 2006;2:CD0032877. Holman RR, Thorne KI, Farmer AJ, et al.; 4-TStudy Group. Addition of biphasic, prandial, orbasal insulin to oral therapy in type 2 diabetes.N Engl J Med 2007;357:1716–17308. Gamble JM, Simpson SH, Eurich DT,MajumdarSR, Johnson JA. Insulin use and increased risk ofmortality in type 2 diabetes: a cohort study. Di-abetes Obes Metab 2010;12:47–539. Gerstein HC, Bosch J, Dagenais GR, et al.;ORIGIN Trial Investigators. Basal insulin and car-diovascular and other outcomes in dysglycemia.N Engl J Med 2012;367:319–32810. Mellbin LG, Ryden L, Riddle MC, et al.;ORIGIN Trial Investigators. Does hypoglycaemiaincrease the risk of cardiovascular events? A re-port from the ORIGIN trial. Eur Heart J 2013;34:3137–314411. Yang YX, Hennessy S, Lewis JD. Insulin ther-apy and colorectal cancer risk among type 2 di-abetes mellitus patients. Gastroenterology2004;127:1044–105012. Bowker SL, Majumdar SR, Veugelers P,Johnson JA. Increased cancer-related mortalityfor patients with type 2 diabetes who use sulfo-nylureas or insulin. Diabetes Care 2006;29:254–25813. Currie CJ, Poole CD, Gale EA. The influence ofglucose-lowering therapies on cancer risk in type2 diabetes. Diabetologia 2009;52:1766–177714. Kramer CK, Zinman B, Retnakaran R. Short-term intensive insulin therapy in type 2 diabetesmellitus: a systematic review andmeta-analysis.Lancet Diabetes Endocrinol 2013;1:28–34

15. Lebovitz HE, Feinglos MN. Sulfonylureadrugs: mechanism of antidiabetic action andtherapeutic usefulness. Diabetes Care 1978;1:189–19816. Reaven GM. Effect of glipizide treatment onvarious aspects of glucose, insulin, and lipid me-tabolism in patients with noninsulin-dependentdiabetes mellitus. Am J Med 1983;75:8–1417. Lebovitz HE. Clinical utility of oral hypogly-cemic agents in the management of patientswith noninsulin-dependent diabetes mellitus.Am J Med 1983;75(Suppl. 5B):94–9918. Kennedy DL, Piper JM, Baum C. Trends inuse of oral hypoglycemic agents 1964-1986. Di-abetes Care 1988;11:558–56219. Melander A, Lebovitz HE, Faber OK. Sulfo-nylureas. Why, which, and how? Diabetes Care1990;13(Suppl. 3):18–2520. Bryan J, Crane A, Vila-Carriles WH, BabenkoAP, Aguilar-Bryan L. Insulin secretagogues, sul-fonylurea receptors and K(ATP) channels. CurrPharm Des 2005;11:2699–271621. Desai NR, Shrank WH, Fischer MA, et al.Patterns of medication initiation in newly diag-nosed diabetes mellitus: quality and cost impli-cations. Am J Med 2012;125:e1–e722. Zeller M, Danchin N, Simon D, et al.; FrenchRegistry of Acute ST-Elevation and Non-ST-Elevation Myocardial Infarction investigators.Impact of type of preadmission sulfonylureason mortality and cardiovascular outcomes in di-abetic patients with acute myocardial infarction.J Clin Endocrinol Metab 2010;95:4993–500223. Goldner MG, Knatterud GL, Prout TE. Ef-fects of hypoglycemic agents on vascular com-plications in patients with adult-onset diabetes.3. Clinical implications of UGDP results. JAMA1971;218:1400–141024. Adler AI, Stratton IM, Neil HA, et al. Associa-tion of systolic blood pressure with macrovascu-lar and microvascular complications of type 2diabetes (UKPDS 36): prospective observationalstudy. BMJ 2000;321:412–41925. Riddle MC. More reasons to say goodbye toglyburide. J Clin Endocrinol Metab 2010;95:4867–487026. Lee T-M, Chou T-F. Impairment of myocar-dial protection in type 2 diabetic patients. J ClinEndocrinol Metab 2003;88:531–53727. MonamiM, Luzzi C, Lamanna C, et al. Three-year mortality in diabetic patients treated withdifferent combinations of insulin secretagoguesand metformin. Diabetes Metab Res Rev 2006;22:477–48228. Holstein A, Plaschke A, Egberts E-H. Lowerincidence of severe hypoglycaemia in patientswith type 2 diabetes treated with glimepirideversus glibenclamide. Diabetes Metab Res Rev2001;17:467–47329. Hirst JA, Farmer AJ, Dyar A, Lung TWC,Stevens RJ. Estimating the effect of sulfonylureaon HbA1c in diabetes: a systematic review andmeta-analysis. Diabetologia 2013;56:973–98430. Zhang Y, McCoy RG, Mason JE, Smith SA,Shah ND, Denton BT. Second-line agents for gly-cemic control for type 2 diabetes: are neweragents better? Diabetes Care 2014;37:1338–134531. Canaglifozin (Invokana) for type 2 diabetes.Med Lett Drugs Ther 2013;55:37–3932. Campbell IW. Comparing the actions ofolder and newer therapies on body weight: to

2656 Safety of Medications for Type 2 Diabetes Diabetes Care Volume 37, September 2014

what extent should these effects guide the se-lection of antidiabetic therapy? Int J Clin Pract2010;64:791–80133. Cleveland JC Jr, Meldrum DR, Cain BS,Banerjee A, Harken AH. Oral sulfonylurea hypo-glycemic agents prevent ischemic precondition-ing in human myocardium. Two paradoxesrevisited. Circulation 1997;96:29–3234. University Group Diabetes Program. A studyof the effects of hypoglycemia agents on vascularcomplications in patients with adult-onset diabe-tes. Diabetes 1970;19(Suppl. 2):747–83035. American Diabetes Association. Americandiabetes association policy statement: theUGDP controversy. Diabetes Care 1979;2:1–336. Rytter L, Troelsen S, Beck-Nielsen H. Preva-lence and mortality of acute myocardial infarc-tion in patients with diabetes. Diabetes Care1985;8:230–23437. Aronow WS, Ahn C. Incidence of new coro-nary events in older persons with diabetes mel-litus and prior myocardial infarction treatedwith sulfonylureas, insulin, metformin, anddiet alone. Am J Cardiol 2001;88:556–55738. Danchin N, Charpentier G, Ledru F, et al.Role of previous treatment with sulfonylureasin diabetic patients with acute myocardial in-farction: results from a nationwide French reg-istry. Diabetes Metab Res Rev 2005;21:143–14939. Halkin A, Roth A, Jonas M, Behar S.Sulfonylureas are not associated with increasedmortality in diabetics treated with thrombolysisfor acute myocardial infarction. J ThrombThrombolysis 2001;12:177–18440. Davis TM, Parsons RW, Broadhurst RJ,Hobbs MS, Jamrozik K. Arrhythmias and mortal-ity after myocardial infarction in diabetic pa-tients. Relationship to diabetes treatment.Diabetes Care 1998;21:637–64041. UK Prospective Diabetes Study (UKPDS)Group. Intensive blood-glucose control with sul-phonylureas or insulin compared with conven-tional treatment and risk of complications inpatients with type 2 diabetes (UKPDS 33). Lan-cet 1998;352:837–85342. Patel A, MacMahon S, Chalmers J, et al.;ADVANCE Collaborative Group. Intensive bloodglucose control and vascular outcomes in pa-tients with type 2 diabetes. N Engl J Med2008;358:2560–257243. Home PD, Pocock SJ, Beck-Nielsen H, et al.;RECORD Study Team. Rosiglitazone Evaluatedfor Cardiovascular Outcomes in Oral Agent Com-bination Therapy for Type 2 Diabetes (RECORD):a multicentre, randomised, open-label trial.Lancet 2009;373:2125–213544. Kahn SE, Haffner SM, Heise MA, et al.;ADOPT Study Group. Glycemic durability of ro-siglitazone, metformin, or glyburide mono-therapy. N Engl J Med 2006;355:2427–244345. Monami M, Genovese S, Mannucci E. Car-diovascular safety of sulfonylureas: a meta-analysis of randomized clinical trials. DiabetesObes Metab 2013;15:938–95346. Nagendran J, Oudit GY, Bakal JA, Light PE,Dyck JRB, McAlister FA. Are users of sulphonyl-ureas at the time of an acute coronary syn-drome at risk of poorer outcomes? DiabetesObes Metab 2013;15:1022–102847. Hirsch I. First, do no harm. Clin Diabetes2000;18:97–99

48. Hannele Y-J. Thiazolidinediones. N Engl JMed 2004;351:1106–111849. Mazzone T, Meyer PM, Feinstein SB, et al.Effect of pioglitazone compared with glimepirideon carotid intima-media thickness in type 2 di-abetes: a randomized trial. JAMA 2006;296:2572–258150. Hanefeld M, Pfutzner A, Forst T, Lubben G.Glycemic control and treatment failure withpioglitazone versus glibenclamide in type 2 di-abetes mellitus: a 42-month, open-label, obser-vational, primary care study. Curr Med Res Opin2006;22:1211–121551. Tan MH, Baksi A, Krahulec B, et al.; GLALStudy Group. Comparison of pioglitazone andgliclazide in sustaining glycemic control over2 years in patients with type 2 diabetes. Diabe-tes Care 2005;28:544–55052. Nissen SE, Nicholls SJ, Wolski K, et al.;PERISCOPE Investigators. Comparison of pi-oglitazone vs glimepiride on progression of cor-onary atherosclerosis in patients with type 2diabetes: the PERISCOPE randomized controlledtrial. JAMA 2008;299:1561–157353. U.S. Food and Drug Administration. Rezulinto be withdrawn from the market [press re-lease] March 21, 2000. Available from http://www.fda.gov/ohrms/dockets/ac/00/backgrd/3634b1a_tab6c.htm. Accessed 19 November201354. U.S. Food and Drug Administration. FDADrug Safety Communication: updated risk evalu-ation and mitigation strategy (REMS) to restrictaccess to rosiglitazone-containing medicines in-cluding Avandia, Avandamet, and Avandaryl.May 18, 2011. Available from http://www.fda.gov/Drugs/DrugSafety/ucm255005.htm. Accessed19 November 201355. Mahaffey KW, Hafley G, Sickerson S, et al.Results of a reevaluation of cardiovascular out-comes in the RECORD trial. Am Heart J 2013;166:240–249.e156. Tucker ME. FDA lifts rosiglitazone prescrib-ing restrictions. Medscape Medical News. 25November 2013. Available from http://www.medscape.com/viewarticle/814964. Accessed17 February 201457. Dormandy JA, Charbonnel B, Eckland DJA,et al.; PROactive Investigators. Secondary pre-vention of macrovascular events in patientswith type 2 diabetes in the PROactive study:a randomised controlled trial. Lancet 2005;366:1279–128958. Takeda Pharmaceuticals America. ActosPrescribing Information. Deerfield, Ill., TakedaPharmaceuticals America, 201359. GlaxoSmithKline. Avandia PrescribingInformation. Research Triangle Park, NC,GlaxoSmithKline, 200860. Kahn SE, Zinman B, Lachin JM, et al.; Diabe-tes Outcome Progression Trial (ADOPT) StudyGroup. Rosiglitazone-associated fractures intype 2 diabetes: an analysis from A DiabetesOutcome Progression Trial (ADOPT). DiabetesCare 2008;31:845–85161. Aubert RE, Herrera V, Chen W, Haffner SM,Pendergrass M. Rosiglitazone and pioglitazoneincrease fracture risk in women and men withtype 2 diabetes. Diabetes Obes Metab 2010;12:716–72162. Lago RM, Singh PP, Nesto RW. Congestiveheart failure and cardiovascular death in

patients with prediabetes and type 2 diabetesgiven thiazolidinediones: a meta-analysis ofrandomised clinical trials. Lancet 2007;370:1129–113663. Idris I, Warren G, Donnelly R. Associationbetween thiazolidinedione treatment and riskof macular edema among patients with type 2diabetes. Arch InternMed 2012;172:1005–101164. Ambrosius WT, Danis RP, Goff DC Jr, et al.;ACCORD Study Group. Lack of associationbetween thiazolidinediones and macularedema in type 2 diabetes: the ACCORD eyesubstudy. Arch Ophthalmol 2010;128:312–31865. Sanyal AJ, Chalasani N, Kowdley KV, et al.;NASH CRN. Pioglitazone, vitamin E, or placebofor nonalcoholic steatohepatitis. N Engl J Med2010;362:1675–168566. Cohen SM. Effects of PPARgamma and com-bined agonists on the urinary tract of rats andother species. Toxicol Sci 2005;87:322–32767. Lewis JD, Ferrara A, Peng T, et al. Risk ofbladder cancer among diabetic patients treatedwith pioglitazone: interim report of a longitudi-nal cohort study. Diabetes Care 2011;34:916–92268. Lewis JD, Strom BL, Bilker W, et al. Cohortstudy of pioglitazone and bladder cancer in pa-tients with diabetes: fourth interim analysis (8-year) report with data from January 1997 toDecember 31, 2010 [Online Report, 2012].Available from http://general.takedapharm.com/Trial-Disclosure/01-03-TL-OPI-524-8-year-Interim-Report.pdf. Accessed 13 March 201469. Drucker DJ. The biology of incretin hor-mones. Cell Metab 2006;3:153–16570. Drucker DJ, Nauck MA. The incretin system:glucagon-like peptide-1 receptor agonists anddipeptidyl peptidase-4 inhibitors in type 2 dia-betes. Lancet 2006;368:1696–170571. Nauck MA. Incretin-based therapies fortype 2 diabetes mellitus: properties, functions,and clinical implications. Am J Med 2011;124(Suppl.):S3–S1872. Deacon CF. Incretin-based treatment oftype 2 diabetes: glucagon-like peptide-1 recep-tor agonists and dipeptidyl peptidase-4 inhibi-tors. Diabetes ObesMetab 2007;9(Suppl. 1):23–3173. Cobble M. Differentiating among incretin-based therapies in the management of patientswith type 2 diabetes mellitus. Diabetol MetabSyndr 2012;4:874. Scirica BM, Bhatt DL, Braunwald E, et al.;SAVOR-TIMI 53 Steering Committee and Investi-gators. Saxagliptin and cardiovascular outcomesin patients with type 2 diabetes mellitus. N EnglJ Med 2013;369:1317–132675. White WB, Cannon CP, Heller SR, et al.;EXAMINE Investigators. Alogliptin after acutecoronary syndrome in patients with type 2 di-abetes. N Engl J Med 2013;369:1327–133576. Rodbard HW, Jellinger PS, Davidson JA,et al. Statement by an American Association ofClinical Endocrinologists/American College ofEndocrinology consensus panel on type 2 diabe-tes mellitus: an algorithm for glycemic control.Endocr Pract 2009;15:540–55977. Garber A, Henry R, Ratner R, et al.; LEAD-3(Mono) Study Group. Liraglutide versus glimepir-ide monotherapy for type 2 diabetes (LEAD-3Mono): a randomised, 52-week, phase III, double-

care.diabetesjournals.org Cefalu and Associates 2657

blind, parallel-treatment trial. Lancet 2009;373:473–48178. Nauck M, Frid A, Hermansen K, et al.; LEAD-2 Study Group. Efficacy and safety comparisonof liraglutide, glimepiride, and placebo, all incombination with metformin, in type 2 diabe-tes: the LEAD (liraglutide effect and action indiabetes)-2 study. Diabetes Care 2009;32:84–9079. Buse JB, Rosenstock J, Sesti G, et al.; LEAD-6Study Group. Liraglutide once a day versusexenatide twice a day for type 2 diabetes:a 26-week randomised, parallel-group, multina-tional, open-label trial (LEAD-6). Lancet 2009;374:39–4780. Buse JB, Garber A, Rosenstock J, et al. Lira-glutide treatment is associated with a low fre-quency and magnitude of antibody formationwith no apparent impact on glycemic responseor increased frequency of adverse events: re-sults from the Liraglutide Effect and Action inDiabetes (LEAD) trials. J Clin Endocrinol Metab2011;96:1695–170281. DruckerDJ, Buse JB, Taylor K, et al.; DURATION-1 Study Group. Exenatide once weekly versus twicedaily for the treatment of type 2 diabetes: a rando-mised, open-label, non-inferiority study. Lancet2008;372:1240–125082. Williams-Herman D, Engel SS, Round E,et al. Safety and tolerability of sitagliptin in clin-ical studies: a pooled analysis of data from10,246 patients with type 2 diabetes. BMCEndocr Disord 2010;10:783. Richter B, Bandeira-Echtler E, Bergerhoff K,Lerch CL. Dipeptidyl peptidase-4 (DPP-4) inhib-itors for type 2 diabetes mellitus. Cochrane Da-tabase Syst Rev 2008;2:CD00673984. Rosenstock J, Sankoh S, List JF. Glucose-lowering activity of the dipeptidyl peptidase-4inhibitor saxagliptin in drug-naive patients withtype 2 diabetes. Diabetes Obes Metab 2008;10:376–38685. Moretto TJ, Milton DR, Ridge TD, et al. Effi-cacy and tolerability of exenatide monotherapyover 24 weeks in antidiabetic drug-naive pa-tients with type 2 diabetes: a randomized,double-blind, placebo-controlled, parallel-group study. Clin Ther 2008;30:1448–146086. Nelson P, Poon T, Guan X, Schnabel C,Wintle M, Fineman M. The incretin mimetic ex-enatide as a monotherapy in patients with type2 diabetes. Diabetes Technol Ther 2007;9:317–32687. Hanefeld M, Herman GA, Wu M, Mickel C,SanchezM, Stein PP; Sitagliptin Study 014 Inves-tigators. Once-daily sitagliptin, a dipeptidylpeptidase-4 inhibitor, for the treatment of pa-tients with type 2 diabetes. Curr Med Res Opin2007;23:1329–133988. Scott R, Wu M, Sanchez M, Stein P. Efficacyand tolerability of the dipeptidyl peptidase-4inhibitor sitagliptin as monotherapy over 12weeks in patients with type 2 diabetes. Int JClin Pract 2007;61:171–18089. Vilsbøll T, Zdravkovic M, Le-Thi T, et al.Liraglutide, a long-acting human glucagon-likepeptide-1 analog, given as monotherapy signif-icantly improves glycemic control and lowersbody weight without risk of hypoglycemia inpatients with type 2 diabetes. Diabetes Care2007;30:1608–161090. Del Prato S, Barnett AH, Huisman H,Neubacher D, Woerle HJ, Dugi KA. Effect of

linagliptin monotherapy on glycaemic controland markers of b-cell function in patients withinadequately controlled type 2 diabetes: a ran-domized controlled trial. Diabetes Obes Metab2011;13:258–26791. Raz I, Hanefeld M, Xu L, Caria C, Williams-Herman D, Khatami H; Sitagliptin Study 023Group. Efficacy and safety of the dipeptidylpeptidase-4 inhibitor sitagliptin as monotherapyin patients with type 2 diabetes mellitus. Diabe-tologia 2006;49:2564–257192. Rosenstock J, Brazg R, Andryuk PJ, Lu K,Stein P; Sitagliptin Study 019 Group. Efficacyand safety of the dipeptidyl peptidase-4 inhibi-tor sitagliptin added to ongoing pioglitazonetherapy in patients with type 2 diabetes:a 24-week, multicenter, randomized, double-blind, placebo-controlled, parallel-group study.Clin Ther 2006;28:1556–156893. Zinman B, Hoogwerf BJ, Duran Garcıa S,et al. The effect of adding exenatide to a thia-zolidinedione in suboptimally controlled type 2diabetes: a randomized trial. Ann Intern Med2007;146:477–48594. Zinman B, Gerich J, Buse JB, et al.; LEAD-4Study Investigators. Efficacy and safety of thehuman glucagon-like peptide-1 analog liraglu-tide in combination with metformin and thia-zolidinedione in patients with type 2 diabetes(LEAD-4 Met1TZD). Diabetes Care 2009;32:1224–123095. DeFronzo RA, Hissa MN, Garber AJ, et al.;Saxagliptin 014 Study Group. The efficacy andsafety of saxagliptin when added to metformintherapy in patients with inadequately con-trolled type 2 diabetes with metformin alone.Diabetes Care 2009;32:1649–165596. Chacra AR, Tan GH, Apanovitch A,Ravichandran S, List J, Chen R; CV181-040 In-vestigators. Saxagliptin added to a submaximaldose of sulphonylurea improves glycaemic con-trol compared with uptitration of sulphonylureain patients with type 2 diabetes: a randomisedcontrolled trial. Int J Clin Pract 2009;63:1395–140697. Charbonnel B, Karasik A, Liu J, Wu M,Meininger G; Sitagliptin Study 020 Group. Effi-cacy and safety of the dipeptidyl peptidase-4inhibitor sitagliptin added to ongoing metfor-min therapy in patients with type 2 diabetesinadequately controlled with metformin alone.Diabetes Care 2006;29:2638–264398. DeFronzo RA, Ratner RE, Han J, Kim DD,Fineman MS, Baron AD. Effects of exenatide(exendin-4) on glycemic control and weightover 30 weeks in metformin-treated patientswith type 2 diabetes. Diabetes Care 2005;28:1092–110099. Buse JB, Henry RR, Han J, Kim DD, FinemanMS, Baron AD; Exenatide-113 Clinical StudyGroup. Effects of exenatide (exendin-4) on gly-cemic control over 30 weeks in sulfonylurea-treated patients with type 2 diabetes. DiabetesCare 2004;27:2628–2635100. Nauck MA. A critical analysis of the clinicaluse of incretin-based therapies: the benefits byfar outweigh the potential risks. Diabetes Care2013;36:2126–2132101. Butler PC, Elashoff M, Elashoff R, GaleEAM. A critical analysis of the clinical use ofincretin-based therapies: are the GLP-1 thera-pies safe? Diabetes Care 2013;36:2118–2125

102. Green JB, Bethel MA, Paul SK, et al. Ratio-nale, design, and organization of a randomized,controlled Trial Evaluating CardiovascularOutcomes with Sitagliptin (TECOS) in patientswith type 2 diabetes and established cardio-vascular disease. Am Heart J 2013;166:983–989.e7103. Rosenstock J, Marx N, Kahn SE, et al. Car-diovascular outcome trials in type 2 diabetesand the sulphonylurea controversy: rationalefor the active-comparator CAROLINA trial.Diab Vasc Dis Res 2013;10:289–301104. Nachnani JS, Bulchandani DG, Nookala A,et al. Biochemical and histological effects ofexendin-4 (exenatide) on the rat pancreas. Dia-betologia 2010;53:153–159105. Gier B, Matveyenko AV, Kirakossian D,Dawson D, Dry SM, Butler PC. Chronic GLP-1receptor activation by exendin-4 induces expan-sion of pancreatic duct glands in rats and accel-erates formation of dysplastic lesions andchronic pancreatitis in the Kras(G12D) mousemodel. Diabetes 2012;61:1250–1262106. Nyborg NC, Mølck AM, Madsen LW,Knudsen LB. The human GLP-1 analog liraglutideand the pancreas: evidence for the absence ofstructural pancreatic changes in three species.Diabetes 2012;61:1243–1249107. Matveyenko AV, Dry S, Cox HI, et al. Ben-eficial endocrine but adverse exocrine effects ofsitagliptin in the human islet amyloid polypep-tide transgenic rat model of type 2 diabetes:interactions with metformin. Diabetes 2009;58:1604–1615108. Ahmad SR, Swann J. Exenatide and rareadverse events. N Engl J Med 2008;358:1970–1971; discussion 1971–1972109. Tatarkiewicz K, Smith PA, Sablan EJ, et al.Exenatide does not evoke pancreatitis andattenuates chemically induced pancreatitis innormal and diabetic rodents. Am J PhysiolEndocrinol Metab 2010;299:E1076–E1086110. Koehler JA, Baggio LL, Lamont BJ, Ali S,Drucker DJ. Glucagon-like peptide-1 receptoractivation modulates pancreatitis-associatedgene expression but does not modify the sus-ceptibility to experimental pancreatitis in mice.Diabetes 2009;58:2148–2161111. Parks M, Rosebraugh C. Weighing risksand benefits of liraglutidedthe FDA’s reviewof a new antidiabetic therapy. N Engl J Med2010;362:774–777112. Garg R, Chen W, Pendergrass M. Acutepancreatitis in type 2 diabetes treated with ex-enatide or sitagliptin: a retrospective observa-tional pharmacy claims analysis. Diabetes Care2010;33:2349–2354113. Dore DD, Seeger JD, Arnold Chan K. Useof a claims-based active drug safety surveillancesystem to assess the risk of acute pancreatitiswith exenatide or sitagliptin compared to met-formin or glyburide. Curr Med Res Opin 2009;25:1019–1027114. Dore DD, Bloomgren GL, Wenten M, et al.A cohort study of acute pancreatitis in relationto exenatide use. Diabetes Obes Metab 2011;13:559–566115. Wenten M, Gaebler JA, Hussein M, et al.Relative risk of acute pancreatitis in initiators ofexenatide twice daily compared with other anti-diabetic medication: a follow-up study. DiabetMed 2012;29:1412–1418

2658 Safety of Medications for Type 2 Diabetes Diabetes Care Volume 37, September 2014

116. Pendergrass M, Chen W. Association be-tween diabetes, exenatide, sitagliptin and acutepancreatitis. Diabetes 2010;59(Suppl. 1):A160117. Elashoff M, Matveyenko AV, Gier B,Elashoff R, Butler PC. Pancreatitis, pancreatic,and thyroid cancer with glucagon-like peptide-1-based therapies. Gastroenterology 2011;141:150–156118. Singh S, Chang H-Y, Richards TM, WeinerJP, Clark JM, Segal JB. Glucagonlike peptide 1-based therapies and risk of hospitalization foracute pancreatitis in type 2 diabetes mellitus:a population-based matched case-controlstudy. JAMA Intern Med 2013;173:534–539119. Steinberg W, Rosenstock J, De Vries H,Bloch Thomsen A, Svendsen CB, Wadden TA.Elevated serum lipase activity in adults withtype 2 diabetes and no gastrointestinal symp-toms. Gastroenterology 2012;142(Suppl. 1):S93–S94120. Cefalu WT, Rosenstock J, Henry RR, RiddleM. Signals and noise in drug safety analyses: theincretin therapy debate provides the rationalefor revamping epidemiologic pharmacovigi-lance. Diabetes Care 2013;36:1804–1806121. Butler AE, Campbell-Thompson M, GurloT, Dawson DW, Atkinson M, Butler PC. Markedexpansion of exocrine and endocrine pancreaswith incretin therapy in humans with increasedexocrine pancreas dysplasia and the potentialfor glucagon-producing neuroendocrine tu-mors. Diabetes 2013;62:2595–2604122. Bonner-Weir S, In’t Veld PA, Weir GC.Reanalysis of study of pancreatic effects of in-cretin therapy: methodological deficiencies. Di-abetes Obes Metab. 8 January 2014 [Epubahead of print]123. In’t Veld P, De Munck N, Van Belle K, et al.Beta-cell replication is increased in donor or-gans from young patients after prolonged lifesupport. Diabetes 2010;59:1702–1708124. Yachida S, Jones S, Bozic I, et al. Distantmetastasis occurs late during the genetic evolu-tion of pancreatic cancer. Nature 2010;467:1114–1117125. Bjerre Knudsen L, Madsen LW, AndersenS, et al. Glucagon-like peptide-1 receptor ago-nists activate rodent thyroid C-cells causingcalcitonin release and C-cell proliferation. Endo-crinology 2010;151:1473–1486126. Hegedus L,Moses AC, ZdravkovicM, Le ThiT, Daniels GH. GLP-1 and calcitonin concentra-tion in humans: lack of evidence of calcitoninrelease from sequential screening in over 5000

subjects with type 2 diabetes or nondiabeticobese subjects treated with the human GLP-1analog, liraglutide. J Clin Endocrinol Metab2011;96:853–860127. Aschebrook-Kilfoy B, Ward MH, SabraMM, Devesa SS. Thyroid cancer incidence pat-terns in the United States by histologic type,1992–2006. Thyroid 2011;21:125–134128. Egan AG, Blind E, Dunder K, et al. Pancre-atic safety of incretin-based drugsdFDA andEMA assessment. N Engl J Med 2014;370:794–797129. Musso G, Gambino R, Cassader M, PaganoG. A novel approach to control hyperglycemia intype 2 diabetes: sodium glucose co-transport(SGLT) inhibitors: systematic review and meta-analysis of randomized trials. AnnMed 2012;44:375–393130. Andrianesis V, Doupis J. The role of kidneyin glucose homeostasisdSGLT2 inhibitors, anew approach in diabetes treatment. ExpertRev Clin Pharmacol 2013;6:519–539131. Nisly SA, Kolanczyk DM, Walton AM.Canagliflozin, a new sodium-glucose cotrans-porter 2 inhibitor, in the treatment of diabetes.Am J Health Syst Pharm 2013;70:311–319132. Komoroski B, Vachharajani N, Feng Y, Li L,Kornhauser D, Pfister M. Dapagliflozin, a novel,selective SGLT2 inhibitor, improved glycemiccontrol over 2 weeks in patients with type 2diabetes mellitus. Clin Pharmacol Ther 2009;85:513–519133. List JF, Woo V, Morales E, Tang W,Fiedorek FT. Sodium-glucose cotransport inhibi-tion with dapagliflozin in type 2 diabetes. Dia-betes Care 2009;32:650–657134. Wilding JP, Norwood P, T’joen C, BastienA, List JF, Fiedorek FT. A study of dapagliflozinin patients with type 2 diabetes receivinghigh doses of insulin plus insulin sensitizers:applicability of a novel insulin-independenttreatment. Diabetes Care 2009;32:1656–1662135. Pharmaceuticals J. Invokana PrescribingInformation. Titusville, NJ, Janssen Pharmaceut-icals, 2013136. Bristol-Myers Squibb. Farxiga PrescribingInformation. Princeton, NJ, Bristol-Myers Squibb,2014137. List J, Ley S, Ptaszynska A, et al. Character-ization of genital infections in the setting ofpharmacologically induced glucosuria. Diabetes2011;60(Suppl. 1):A270

138. Parikh SJ, Johnsson KM, Ptaszynska A,Schmitz BG, Sugg JE, List JF. Characterizationof urinary tract infections in the setting of phar-macologically induced glucosuria. Diabetes2011;60(Suppl. 1):A270139. Inagaki N, Kondo K, Iwasaki T, et al.Canagliflozin, a novel inhibitor of sodium glu-cose co-transporter 2 (SGLT2) improves glyce-mic control and reduces body weight inJapanese type 2 diabetes mellitus (T2DM). Di-abetes 2011;60(Suppl. 1):A274140. Rosenstock J, Aggarwal N, Polidori D,et al.; Canagliflozin DIA 2001 Study Group.Dose-ranging effects of canagliflozin, a sodium-glucose cotransporter 2 inhibitor, as add-on tometformin in subjects with type 2 diabetes. Di-abetes Care 2012;35:1232–1238141. Leiter LA, Cefalu WT, de Bruin TWA,Gause-Nilsson I, Sugg J, Parikh SJ. Dapagliflozinadded to usual care in individuals with type 2diabetes mellitus with preexisting cardiovascu-lar disease: a 24-week, multicenter, random-ized, double-blind, placebo-controlled studywith a 28-week extension. J Am Geriatr Soc. 2June 2014 [Epub ahead of print]142. CefaluWT, Leiter LA, Yoon KH, et al. Efficacyand safety of canagliflozin versus glimepiride inpatients with type 2 diabetes inadequately con-trolled with metformin (CANTATA-SU): 52 weekresults from a randomised, double-blind, phase 3non-inferiority trial. Lancet 2013;382:941–950143. Wilding JPH, Woo V, Pahor A, Sugg J,Langkilde A, Parikh S. Sustained effectivenessof dapagliflozin over 48 weeks in patients withtype 2 diabetes poorly controlled with insulin.Diabetologia 2010;53(Suppl. 1):S348–S349144. U.S. Food and Drug Administration. FDAapproves Farxiga to treat type 2 diabetes[press release]. 8 January 2014. Available fromhttp://www.fda.gov/NewsEvents/Newsroom/PressAnnouncements/ucm380829.htm. Ac-cessed 18 February 2014145. Henry RR, Rosenstock J, Chalamandaris A-G,Kasichayanula S, Bogle A, Griffen SC. Exploring thepotential of dapagliflozin in type 1 diabetes:phase 2a pilot study [abstract]. Presented at the73rd Scientific Sessions of the American DiabetesAssociation, Chicago, IL, 21–25 June 2013146. Perkins BA, Cherney DZ, Partridge H, et al.Sodium-glucose cotransporter 2 inhibition andglycemic control in type 1 diabetes: results of an8-week open-label proof-of-concept trial. Dia-betes Care 2014;37:1480–1483

care.diabetesjournals.org Cefalu and Associates 2659