, evidence- base updateThearnctee In tme seriesare indepcndcnlly researched and compiled by P~A ccmrrussicneu auloo/:; i:lf'l l}Ij'tlr revreweo.

t

• Have an understanding of the newclassesof medications that areavailable ormay soon be available totreat type 2 diabetes, including theirmodesof action, effectiveness andsafety.

• Consider the likelyplace of thesenew medications in light of currenttherapeutic options for type 2diabetes.

Learning objectives:Afterreading thisarticle, the readershould:

>

I

OOl ·n

By Dr Luke BereznickiandProfessor Gregory Peterson

N@w rnedlcatlons to treattype 2 diabetes.

~ PIiCii'iifac;st \tl.tne 27I~ 3 IMarch 2COl

IntroductionThe number of Australians with type 2 diabeteshas tripledsince 1981 and continues to increase. It is projected that1.6 mill ion Australians will have type 2 diabetes by 2030.'Effective treatment of hyperglycaemia is a priority, giventhat strict glycaemic control reduces the microvascularcomplicat ions of type 2 diabetes." Epidemiologicaldata fr om the UK suggests that improving glycaemiccontrol wi ll also reduce the risk of macrovascularcomplications (e.g. cardiovascular disease),' althoughthis is controvers ial and it is recognised that improvingglycaemic control is only one of a number of possiblestrategies to reduce the macrovascular risk associatedwith diabetes. Health professionals involved in themanagement of diabetes should focus on blood pressuremanagement, cholestero l loweri ng and the use of low­dose aspirin as means of reducing cardiovascu lar risk,as well as control of blood glucose.' The treatment ofhyperglycaemia in type 2 diabetes is complicated, andcombination hypoglycaemic therapy is often requiredto achieve and maintain target blood glucose levels.Unfortunately, recent Australian data suggests thattarget glycated haemoglobin IHbA"llevels are achievedin only 30-50% of type 2 diabetics who are managedin the primary care setting.··· The focus of this articleis to review the evidence for the latest medications toemerge in the battl e to manage hyperglycaemia in type2 diabetes,

A range of optionsThe pathogenesis of diabetes has been traditiona llycharacterised by absolute or relative lossof pancreatic B cellfunction and insulin deficiency or tissue resistance' Morerecently, it has become clear that addit ional pancreatic andgut hormones play an important role in glucose homeostasis.These hormones nowprovide additional therapeutic targets formedications to treat hyperglycaemia associatedwith diabetes.Table 1 li sts the main characteristics of medication classesthat are or may soon be available to treat type 2 diabetes.TheFood and DrugAdministration IFDAI approved the incretinmimetic exenatide in the US in 2005. l ast year exenatidegained approva l in Australia as adjunctive therapy for patientswho have not achieved adequate glycaemic control withmetformin. a sulfonylurea. or both. Pramlintide, an injectablesynthetic hormone that resembles human amylin, was alsoapproved by the FDA in 2005. In 2006, the FDA approved thefirst oral incretin enhancer, sitagliptin, for use as monotherapyor in combination with metformin or a thiazolidinedione fortype 2 diabetes . Sitagliptin IJanuvial was registered by theTherapeutic Goods Administration as a combination therapywith metformin, a sulfonylurea or a thiazolidedione inDecember 2007. The first inhaled insulin to market, Exubete,was withdrawn from the US marke t in October 2007 due topoor sales. However, other inhaled insulin produc ts are in the

\"boo 271 Nunncr 3 1MarCh 2OC6

advancedstage of clinical trials andare likely to beapprovedIn the US In the near future.

Incretin therapyThe incretin effect is the augmentat ion of glucose-stim l t d. I' . ua emsu In se~retlOn by intestinally derived peptides. which arereleased In the presence of glucose in the gastrointestinalt ract." This theory is based on the observation that an oraldose of glucose causes more insulin secretion than the sameamount given intravenously. Improved understanding of thiseffect hasled to the development of newantidiabetic agents.The incretin effect results primari ly from the actions of twopeptides, glucose-dependent insu linotropic polypeptide IGIP}and glucagonlike peptide 1 IGlP-11." lncretins are rapidlyinactivated by the enzyme dipeptidyl peptidase 4 IDPP4l,result ing in a short halt-lite.'? The action of this pathwayappear to be diminished in type 2 diabetes, making thepathway a target for novel pharmacologic agents." GlP-1, inaddition to potentiating glucose stimulated insulin secretion,also inhibits glucagon secretion, retards gastric emptyingand reduces appetite. In animal studies, GlP-1 stimulatesprolif eration of ~ cells and inhibits their apoptosis," anencouraging finding that is yet to be confirmed in humanstudies.

Incretin mimetics: GLP-1 analoguesExenat ide was the first incretin mimetic resistant to DPP4degradation approved bythe FDA." Unlike GlP-1 , exenatide isnot rapidly inactivated, allowing it to be administered twice­daily. Exenatide is administered by subcutaneous injectionbefore the morning and evening meals. However, a once­weeklyadministered lonq-actinq formulation of exenatide hasrecently been tested." l irag lutide, another Gl P-l analogue,may be administered once daily. It is likely that the FDA willconsider approving this product in 2008.

A systematic review and meta-analysis of GlP-1 analogueslexenatide and liraglutide} was conducted recently." Itincluded eight published trials In = 3,1 39 adult participants}in which a GlP-1 analogue was used for type 2 diabetes.GLP-1 analogues were added to existing inadequate therapyIlifestyle or oral hypoglycaemics} and compared with adouble-blind, injectable placebo. metformin. or open-labelsubcutaneous insulin Iglargine or biphasic aspartl.' Anothersmall study In = 45) was included in which a lonq-actinqformulation of exenatide was compared to placebo in patientstaking metformin. The duration of GlP-l analogueuse in thesestudies ranged from 15 to 52 weeks. Glp ·1 analogue therapyresulted in a statistically significant reduct ion in HbA" frombaseline compared to placebo (weighted mean difference-0.97%,95% confidence interval lell -1.1 3% to ·081%). Inopen-label studies comparing exenatide with subcutaneousinsulin there was no difference in HbA" . Patients receivingexenatide were more likely to achieve target HbA" levels

.... : . ~~ .

,,-evidence- base update111e arucles III Illi~ series ale Ilidependently rr.sl1a rclled and compiled bl' PSA commissioned authors and pe~ r reviewed.

Table 1. Characteristics of currently avai lable blood glucose lowering medications (modified from Heine et al. 2006)l3

Medication(sl Delivery Reduction Main mode of action Benefits Side effects and limitationsin HbA!C(%)

Metforrnin Om l 1.5 Lowers production of hepat ic No weight qain; Gastro intestinal complaints: lactic i

qlucnse cheap acidosis (very rare)Sulfonylureas Oral 1.5 Stimulates insulin secretion Cheap Hypoglycaemia; weight gain(gibenclamide,glimepiride,gliclazide. glipizide)Thiazolirlinediones Oral 05 -1.5 Improve insulin sensitivity Increases HDL Fluid retention, which may lead(pioglitazone, levels to heart fa ilure; weight gain:rosiqlitazone] increased risk of CVD; expensive

~ - g l u co s i d a se Oral 0.5-0.8 Retard intestinal absorption of No weight gain Gastrointestinal side effects;inhihitors qlucoss mu lti ple daily dosing required;(acarbose) expansive: low potency

Meglitinides Oral 1-1.5 Stimulate insulin secretion Short-acting; Need to be taken at meal time;(repaglinidel less risk of expensive

hypoglycaemiaDipeptirlyl Oral 0.6-0.9 Stimulate insulin secretion Low risk of Limited experience; expensivepeptidase-a (DPP-4) hypoglycaemiainhibitors"Glucagon- Subcutaneous 0.8-1.1 Stimulates insulin secretion; Weight loss Need to be injected;like peptide-l injection suppressesglucagon: retards gastrointestinal side effects:analogues gastric emptying andreduces limited experience: expensive(exr.natide) energy intakePramlintide Subcutaneous 0.5-1 .0 Retards gastric emptying and Weight loss Needs to be injected before meals;

injection reduces energy intake gastrointestinaI side-effects;expensive; experience limited

Subcutaneous Subcutaneous > 2 Stimulatesperipheral glucose Reduces severe Weight gain; hypoglycaemia;insulin injection uptake and inhibits glucose hyperglycaernia; needs to be injected: monitoring

output cheap; extensive requirementsexperience

Inhaled insulin" Inhaled 0.5-1.0 Stimulates peripheral glucose No injecti ons Multiple daily dosing; monitoringuptake andinhibits glucose required requirements; lOlly term pulmonaryoutput effects unknown; expensive and

experience limited

"Not curren tly approved for use in Australia.

compared to those taking placebo (45% versus 10%.respective ly). but there was no difference between exenatideand insulin in comparative studies. Exenatide was superior toinsu lin at reducing postprandial glycaemia. while there wasno difference in fasting plasma glucose. One key advantage ofthe GLP-1 analogues comparedto their comparators is weightloss (weighted meandifference -2.37 kg; 95% CI -3.95 to -0.78kg). Weight loss was more pronounced when exenatide wascompared to insulin (weighted mean difference -4.76 kg; 95%CI -6.03 to -3.49 kg ) than with other agents. Weight loss wasprogressive. rlose dependentand did not appear to plateau atweek 30.

Severe ilypoglycaemia (requiring intervention) was rare withGLP-1 analogues and has only been reported in patients also

receiving sulfonylureas. The riskof hypoglycaemia wassimilarbetween exenat ide and insulin (approximately 2% in eachgroup), Nausea and vomiting were themost commonly reportedadverse events with exenatide compared to a comparator.with nausea and vom iting occurring in up to 57% and 17%of pa tients treated with exenatide. respect ively. Nausea wasmost common early in the course of therapy and declinedthereafter. In the comparative study with insulin glargine. 19%of patients using exenatidedropped out of the trial. comparedto 10% of patients using the insulin." Gastrointostinal adversereactions were common and contributed to these withdrawals.Diarrhoea was also more common in patients receivingGLP- l analogue therapy. Exenatide therapy commonly (in upto 67% of patients) results in the formation of antibodies tothe molecule. as it is not identical to human GLP-1 . However,

248 Ph-eirmcicist loIum", 27 INumber ':; I lvlarcll 20.1';

, evide nc e - bas e upda teThe articles ill this series arc inrlenendenlly researchedand compt ren hy PSA commissioneo authors and peer reviewel!.

....~-

250 )

this has not been associated with ilny effect on outcomes oradverse events." It is recommended thatexenatide is initiatedat a dosage of 5mcg twice-da ily to improve tolerability andincreased to a maximum dosage of 1mcg twice-dai ly."

Incretin enhancers: DPP4 inhibitors(the 'gliptins')Given that GLP-1 analogoes require injection, cons iderableeffort has been made to develop oral agents that target theincretin pathway. Inhibition of DPP4 extends the half-lifeof native incretins. prolong ing their action." A systematicreview and meta-analysis of DPP4 inhibitions lsitagliptin andvildagliptinl was cond ucted recsntlv." DPP4 treatment wascompared to placebo as monotherapy or as add-on therapy tooral hypoglycaemic agents or insulin in 13 randorn ised trials(n = 4,780). The durat ion of these trials ranqed from 12 to 52weeks. The incretin enhancers lowered HbA1t compared withplacebo with similar effectiveness as monotherapyor as add­on therapy (weighted mean difference -0.74%; 95% CI -0.85%to -0.62%l. Sitagliptin and vildagliptin have not been direc tlycompared, but seem to be similarly effec tive in lowering HbA'tcompared to placebo , In four trials (n = 3,053) comparinga DPP4 inhibitor to other hypog lycaemic agents (glipizide,metformin or a thiazolidenedioneL the DPP4 inhibitors wereslightly less effective at lowering HbA

1, (weighted mean

difference -0.21%; 95% CI 0.02% to 0.39%). Patients trea tedwith DPP4 inhibitors were more likely to ach ieve target HbA1tlevels compared to placebo (43% versus 17%, respectively).Compared to placebo, DPP4 inhibitors were associated witha small increase in weight (weighted mean difference 0.5 kg;95% CI 0.3 to 0.7 kg). Th is was main ly because meLforminwas assoc iated with an average 2.2 kg weight loss comparedto vildag liptin. However, sitagliptin and vildag liptin wereassociated with weight loss compared to glipizide andthiazolidenediones, respectively. Severe hypoglycaemia wasrare with GPP4 inhibitors, and there was no difference inthe incidence of mild to moderate hypog lycaemia betweenGPP4 inhibitors and comparators, DPP4 inhibitors were welltolerated, with no increased risk of gastrointestinal adverseeffects (nausea, vomiting, abdominal pain and diarrhoeal.Adverse effects noted significan tly more frequently with DPP4inhibitors compared to comparators included nasopharyngitis,urinary tract infection and headac he."

Amylin analogue: pramlintideAmylin is produced by the ~ cells of the pancreas and issecreted together with insulin in response to meals. Its ro leis to complement the action of insulin by regulating therate of glucose entry into the circulation following a mea l."It achieves this by slowing gastric emptying, suppress inginappropriate postpra ndial glucagon secretion and regulatingfood intake." Amylin concentrations are deficient in patientswith type 1 diabetes who are also de ficient in insulin." Nativeamyl in exhibits poor solubility and a tendency to aggregate,

Pharmacist

and is not su itable for clinical use. Pramlintide is a synthetic,soluble ana logue of amylin with similar mechan ism s of actionthat raqulata the appearanc e of glucose in the circulationfollowing meals." In the US, pramlintide is indicated asan adjunct to mealtime insulin in patients with type 1 and2 diabetes. In clinical studies, pramlintide improved post­prandial glucose control when added to insu lin therapy inpeople with type 1 or 2 diabetes and was also associatedwith weight loss." In people with type 2 diabetes who requireinsulin therapy, pram lintide therapy has been shown to reduceHbA1c and body weiqht." In one 52 week study. patientstreated with pram lintide (75 or 150mcg tds) had a reductionof HbA't of 1% compared to 0.5% IP < 0.01) for the placeboqroup." Additionally, patients in each of the pramlint idedosage groups had significant decreases in mean body weightcompared to placebo. Other studies have reported similarresults in type 2 diabetics treated with a range of doses ofpramlintide (60mcg tds, 90mcg tds or 120 mcg bd).2011Nauseawas more than twice as likely in patients treated withpramlintide than with placebo in these studies, although it didnot increa se drop-out rates. In peop le with type 1 diabetes,pra rn lintide reduced insulin requirements by 7-8%, withoutincreasing the risk of hypoglycaemia compared to placebo."In the majority of studies, pramlintide was administeredby injection, 15 minutes prior to meals and sepa rated fromthe insu lin injection. However, some smaller studies havedemonstra ted that combining insu lin and pramlintide in thesame injection does not attenuate the therapeutic effects ofpramlintide."

Inhaled insulinThe major drawback to traditiona l insulin therapy is its needfor injection. This has led to attempts to develop a su itablealternative that can be administered by a more desirableroute for diabetic .patients. Pharmacokinetic studies havedemonstrated that inhaled insu lin is similar to the rapid-onsetinsu lin analogues (Iispro and aspart) but possesses a slightlylonger duration of action." It is therefore regarded as arapid -acting insul in and is suitable for control of postprandialhyperglycaemia. Inhaled insulin has been compared withsubcutaneous insulin regimens in patients with type 1 andtype 2 diabetes, and with ora l hypoglycaemics in those withtype 2 diabetes." The combination of inhaled insu lin, takenbefore each mea l, and ultralente at night resulted in similarglycaemic control as a combination of lente and regular insulintwo to three times daily among patients with type 1 and type2 diabetes. Patients receiving the inhaled insu lin regimenhad slightly lower ra tes of hypoglycaemia.26,2' In people withtype 2 diabetes, the addition of inhaled insulin to existingora l therapywas shown to be more effective over 3-6 mun hsthan adding a second oral hypoglycaemic medication. H Inhaledinsulin, however, is consistently associated with a higherrisk of hypog lyca emia than is associated with oral aqents."In clinica l trials, patien ts were gen era lly more satisfied withinhaled insulin than with subcutaneous insulinz617.29

Volume 27 INumoer 3 I M.s.rc 2008

Subrn . lOur answers on" ne al www.psa .org.au ,1Ilel receive automatic feedback

'-.---

Unfortunately, at present, inhaled insulin therapy is limited by

a number of drawbacks. These include expense, the potential

for a reversible decline in lung function associated with its

use, a theoretical risk of formation Dr promotion of cancer with

chronic use and tho possibility that practical issues such as

smokinqand the presence ot upper respiratory tract infections

may affect the degree of insulin absorption and the risk of

hypo- or hyperglycaemia ,. It is also short-actinq and would

not replace the need fora basal insolin. Inhaled insulin is more

suitable for patients with HbA" levels that remain elevated

after fastingqlucose levels have been controllsc with a basal

insulin.

The tachnoloqv required for the adm inistrat ion ot inhaled

insulin is also more demanding than for other inhaled

med ications, where such a high degree of precision in dosing

is not required. In the case of Exubera, this resulted in

difficulties in dosing, The dosage of Exubera was measured

10 milligrams, not units, andinitial dosingwas based on body

weight rather than thecarbohydrate content of meals. Most of

the devices required to inhale insulin are relatively large and

awkward, requiring time and skill to master, although theywill

hopa fullv improve with furtherdevelopment."

Place of newer agents in the treatment of

type 2 diabetes

A recent systematic review provides some evidence on the

relative safety and effectiveness of older oral medications

(metformin and sulfonvlureas ] in comparison with newer

oral agents (thiazolidinediDnes, n-qlucosidase inhibitors and

meqlitinides]." The review round that compared with the

newer, more expensiveagents, older medications have simi lar

or superiDr effects on glycaemic control. lipids and other

intermediate endpoints. This findingsupports the status Df the

newer agents as add-on therapy to metformin, a sulfonylurea

or both in type 2 diabetes."

The roles Df exenatide and sitagliptin in the treatment of

type 2 diabetes are unclear. At present, if opt imum therapy

with oral hypoglycaemics dDes not cDntrol the disease,

subcutaneDus insulin is the next step." Exenatide may have

similar effects on HbA" to insuliu glargine," or twice-daily

insulin aspart," but it causes more adverse effects. The ideal

patient for whDm exenatide could be considered is obese,

with elevated glucoseconcentrations in spite of therapy with

oral hypoglycaemics" The effects of pramlintide on blood

glucose and body weight are more modest compared with

those of exenatide. Pramlintide may have a greater role in type

2 diabetic patients who have longstanding disease and are

more insulin·deficient. as exenatide requires ~ cell function

to ach ieve its therapeutic actiolls, whereas pramlintide does

not." It is likely that the practica l drawbacks associated

with inha led insulin will be overcome in time, but they are

signif ica nt at present. This is perhaps most evident in the

decision tDremove the first inhaled insulin from the market

_me 27 I t>I.rnt€r aIMatcn =

ill the US recently due to poorsales. Newly developed inhaled

insu lin products may prove more succossful. The new classes

of hypoglycaemic agents will need continuod evaluation in

terms of long-term efficacy and safety to fully determine their

role among thewell-established therapies for type 2 diabetes.

Dr LukeBer91nicki and Prutessor Gregory Peterson. Unit for Medication

Outcomes Besearch and Education.School of Pharmacy. Un iversity of Tasmania.

References1. W. I S.R\:)'jcG,lSecnA,~ OILQOCal jllt"t:llo«e at OOr.lo'!S: t nJrna ltl$ fCftIll:,.,.. 2000 and

!.Oiecocr~!of 2{OO. carctescare Ikiy 2004,ZJ:l047-53

1. U~l KifQ:Iom PUl$pOC1JYtlD~ Study (\.I(PDSj Grow, ~1t!!lSiYe lllro'l -qlllCO$~ CIJ'~I~ With

~eas II'~11D:l\?iIf1':ll WIltl (lQWllMalln'UllnIIll anti risk orcomplicalms II pallt!l1tS

with!'J1lC 2 lla:X.1eI ' t:PDS 331. IJ(,I'f~ 08.llIl0!S SluJy"~~ UOCCl~ 12

1995;352 8375.1.3. tn"'-1 KJlO;1oIn Pn:3P(.'ClM! O1ibIItIt> Sll.'<lJ~C~. Elkd atIM<IbMlDIOod-glllCO!ic

tullrol W1itl lTll!ilormin00 cQIT1Jli:aIiOns in uverwe:gntIXItl!'ll\1S l'iJt/1 lyflll2 ClaDll!l!slJ<PDS 341,lJ(

1"l C\SpC'elMt~ SIudy JI<POS) GruloP· 1..a'lCct. $l!1'l 12 19'J!:1 ;352:854-1X1

4.. S1/'ifkl\ 1Ll,Ida AI. fbII Ho\ et i1L As.sDc- ~ gIrraPrnia d1 maaovol'iCliar and n~

~d~pc l~~J5I: ~~dlserViIlicl't1t m.4'j. BW.AuQ 12

;JOOO:321:405-12.5. Mel1CN,RIIWJR,l UlUrO. et al. Jo'lt;'h1ciarl att.11J.!n5 andpraclc.esand p;!1it.ll1 awarer ll!SS at1fle

~ CD«(IIicalmatcianelei. J .AlP ColCM:klI. 2002.;40:11l11-81.

6. Brp' \'i ~..I\, a.shoIm 1lJ, et aI. Dab&!eS~: eeee ICI (na:;h tIMtoPr.J:llS.P?

Me!lJ Aust.~ 1B 2006;18:5;X15-9

7. l<ernI: r~. BillT U. l mrnct pl. "I aI. Gll.In.e, "PKL illl lllOOO l:~" w~1i illAustra."lll AO.n!

witt1 typll 2 OIaOI:lltlS: the1999-2000A1 lSOiAD. lJialll!t'*i care..Il1!2005:21:1:1490·2.

1:I VCf1!l TV, Phit-.I'IG, I'tll __ PJ·M.:lfI:tlI'i!!llCfl1uulalmcsa i~r ,1!SWlth type 2 Clii;D:!IKS: latgetirll 1I1e

1Q.rearillUlllIIe n" l100I1ll"-,teartttY-'I!.e. /'oIcl:l J A.4 JlIl1 8 2007,l86:ti22..t.

9 Eoerral Si, DalStM' T, Fnas J'. PrarrmIide IIlhc eeeurem01 diabetes WJ CIin Przl. Dec

2IX6. tJO: 1647·53-10. Amcti f1( llIu J, PittJ5AG. Elltacy J.'ld ;J tt;1yd incn'!ill lherapy In~ 2 C>3lJclIlS: S,'jSlCffi.1UC

revie'N allIl l1lf.ta·a rllJ~ls . JJ.MA. jj 112007:29lI:194-206.

11. [)rujo;a-DJ, Na:.l:tMA. The iru~trlS)'5l IllIt ltlCllPHN! 1lt'tJtO!- 1 reccpkJ lIQCllIS,5 ;n1~

p..~ imili1ln111)0& 2~. L:n::lt. NcN 11 2006:368:1lJlti-705

12. ()u:kef DJ.1blgic ldilXls a1Il1hnpculi: plJ".erDaI d lie~~ Noll CIn

Pracl !.lnx:rirlaMelotl. tQ 2005:1:21·31.

13 KimU. M~l, ZhuarJJ D,et at E!ftctsot uv:e-'~I~rJg 01 a D Ilj-~ »eease

1orm~1OO II~ M~ cUllroi :nl llOO)' WllIgIl_1S\i)jX.1'i witt! tvDe 2eeercsDillIe'.e5(JI!. JlIl2l:)J] ,30;148/·93

14.~ RI. Van liaallF. JQt!'l$ O. e a.~ \llnIJS r.!'*' gII.yine I'IlXJlitJllli 'MI;h~

1XlI1trtlIltlI1 l1,:lI! 2l~~.1l Iarmni(~ mal . Am .-.em t.'lod. Del: 18 2Q05;IU 55')-fiQ

15. AnOIl'jIllOOS,~ <tuo;s: l!.er1il~dt: , AIls! Proo:l"tler. 2007;3(1:1 64·65.

16. -noogAA. AIIl'jin's~ andts roIc il diaOOtoC$. rolJ'T CJpln Enoo:MOI OJ;b:!:es.

1997;4-26" 90.17, KIlJ;le!' tf, GIo:rn" ItlA. Pr;n6n!e Q Ihl nattto:r' a llSIJI1 'r~~leSmeIl15. rabOlIl l,;

aodIl:'Iil'w d cfir*;9I datl.~. 2004:64:1<119-32.

18.~ RJ. fll:ll.itidc andprnmlililkle: new gkJCOsc .lI1ow.rinq aocnls lor treatillg Iiatletr:s II ItUS

Cew~ J t.'eJ. May?OCl6}3: 477·84.

19, RlIloer II:.wartlL Fnlman us. el ill~ II... if'! will fle alTl)inaI:Iak:Que l:JarnIrJU;lr:

!Y:s 10.. t:ormIlI!d~ II~ aM welQllllU~ il F.I!lJIIl-.iJea!Qd !il.QeCtswit1 t\tll:

2lJabeles DQ~res la1wl .2002:4:51 bl .

'D. tlo\lOllel' PA., l evy P, ffilman M,r!Ial.llrnml,nt« as ~l aa,JncI lo 1'IsIJ11l1t1llf3ll'l~ lCtlg ­

term~rnic and\\~: CCW'11111 in Pabt:l_s WlIfl~ 2 dia0etc3: a I·yearlandomizlldcorrror.o.l

Irlal.~~ taf'!, 200J;2ti7M·9021. 1tAtr'der PA.~ rx;, f\qJIes JA,et ar. Ftlt:et of pmmIifIlicle 011 _qn 11~ nlmCSe

fls ll llH rc;d!'ll fWJC 2 lkc.llc_ patents. CIles Res, :.'OO4:12:661.f.l1

27. 'M1 ltct1cll fiC F,KiuQ9r Of. Fioerran M, '" at Ar.. 1Clom1lCll $lI1l1vamll({lI lll - 1abC1e.!1!I'lSiOn~

t!lC kJ'lO·lcrmefta:y uf~Iidl! asan~ III in!Un 1llC~ fl l'jPC 1 cliab;!l'*i, [)m,!~es

Ct'e. 2OQ"1:24 n04·30n ~C. fftrra'MS.Slrttlel S. cr ;t~fi~ar.:I """"" l(lCIlrniJ:inQ. Am J

~~ PlD'm. 2mi;62 816 'l1.

24. McfIdtIIlfI.YuH OK. Jnh;;ll!lI i!sutl: where11'''W9 and wre-e~ '"' QO? Med JAusl

2OU7;186:390·91,25, {'.oJQIa L t auJ. p~w IIG. MI!Ia-olWjsis:: ,;r:teacy alld saletyII Wla:~ rwil ltJe~ '" a1lJlt, W11h

~ I'l'Il!li1B.Am~II 1.100. 2006;145:1;65-75

26 . Q.IaUmT. BemQer A.ftta:n:xl N.l,et iL fll'cacy1I'd safE'tY or......ed insuIrI P.J:.eral (:I'JI'llIlatl:

'Mlh Sl.lOCUt" Te:IUS . t>lJn\I'lcf...,in ripe 1~ r....:Its 01a 6·montl1. r.lIGltI1IZ8cl,~IIW

llial. DiatJt,les carn. ?OO4:21;"622·21.27. I-tll\;nlt!l PI\. 61PJle L Howol R. ctIII. Et!icacy lind saletyol lnhaJe(l insuln~l compared

'III\Il~ Ilsl*l lhefapy 10 pali!rU Wl!;h~ 2 /Ialle!t's; ,l!QItS It 6-mcdl\,~.

~.1rallJab=!eS Car~. 2OOIl ;1r235Q 62.

28.~ J. l man B,~ lJ. el aLWUlllJ W'ISl.*i~OYCS~ CQ'$m v.ren:sWStJll.1OO

ItYor aOt!e'J 10ornI COOlDIlaIm \t1eraoy in t)'1lo' 2 ciabe:es-a ""1dOtTlIlC'd. coo:rolled II~ Ann l'l1!!!11

Mea.. 2(KlS;143;549--58.29, Ri::J:;l:I '6IOdIJ.~l X .Boblef 8. t1. al Palltnt sallSfaetOl ;n1~ClY1lrOIllfI~

1 yotaf w(h~ea lftd\ ([xJbfal lf1 pIU!fl:i Wiln P''P@ I Of t)po2 daDell!S lRJletesCai'l!

200421 .131S.2330. !loIeI1S. Folctman L Vawi J. ctaI.S;stcmali::review: Il wt; MalM: "II~ a.""o'l SIllelyOflI'a!

nte!l:catlOlis 1or~2~"ti m&Mu:;.Aru1 .,tlJ'n Med. Sop 18 2007:1.:1:386·99.

Jl . RACGP.ClIabe!es mar'lill}:ltlelll lI !JlIIfl pradt,,: G~ fa"!'jIIlt 2 diatlelts 13111 ed.GcJttlk<J\

1iSW:~AuslriJI;a, 2007.32. Kau:kMA,lUanS. Kin 0, el a.A~ oIll'llCa-mly tW!l"atl(lc!l'll bipI'raslC~ aspart

In catenl5witt! \W! 2 l!'abell!SWhO~~ slb:lptimal'l COlIII' ' . Wllll slli/cwly!\lft'i] andmethJmin: a

lJ)fl·illliOOril'j :l!1 l<tf, OiaIJeltio!laa 21m :50:Zi9-67,

33. t'enl RJ. Damant M, I.'bartfll JC. it :ll M:¥'.agement ofIl)tlcnjyl:aerl'.aa n lyJ te2 d<beIl!S: ll10 lin:!

d J!£U'!mI bUll? 8M./.Dl!c 9 2OOi:333:1200-".