Rev. paul. pediatr. vol.34 número3

REVISTA

PAULISTA

DE

PEDIATRIA

www.rpped.com.br

REVIEW

ARTICLE

Insulin

therapy

in

patients

with

cystic

fibrosis

in

the

pre-diabetes

stage:

a

systematic

review

Mariana

Zorrón

Mei

Hsia

Pu

_,

_Flávia

_Corrêa

_{Christensen-Adad,}

Aline

Cristina

Gonc

¸alves,

Walter

José

Minicucci,

José

Dirceu

Ribeiro,

Antonio

Fernando

Ribeiro

FaculdadedeCiênciasMédicas,UniversidadedeCampinas(Unicamp),Campinas,SP,Brazil

Received1August2015;accepted1December2015 Availableonline21June2016

KEYWORDS Cysticfibrosis; Insulin;

Diabetesmellitus

Abstract

Objective: Toelucidatewhetherinsuliniseffectiveornotinpatientswithcysticfibrosisbefore

thediabetesmellitusphase.

Datasource:Thestudy wasperformedaccordingtothePrismamethodbetweenAugustand

September2014,usingthePubMed,Embase,LilacsandSciELOdatabases.Prospectivestudies

publishedinEnglish,PortugueseandSpanishfrom2002to2014,evaluatingtheeffectofinsulin

onweightparameters,bodymassindexandpulmonaryfunctioninpatientswithcysticfibrosis,

withameanageof17.37yearsbeforethediabetesmellitusphasewereincluded.

Datasynthesis: Eightarticleswereidentifiedthatincluded180patientsundergoinginsulinuse.

Sample sizerangedfrom4to54 patients, withameanage rangingfrom12.4to28 years.

Thetypeoffollow-up,timeofinsulinuse,thedoseandimplementationschedulewerevery

heterogeneousbetweenstudies.

Conclusions: Therearetheoreticalreasonstobelievethatinsulinhasabeneficialeffectinthe

studiedpopulation.Thedifferentmethodsandpopulationsassessedinthestudiesdonotallow

ustostatewhetherearlyinsulintherapyshouldorshouldnotbecarriedoutinpatientswith

cysticfibrosisprior tothediagnosis ofdiabetes. Therefore,studieswithlargersamples and

insulinusestandardizationarerequired.

©2016SociedadedePediatriadeS˜aoPaulo.PublishedbyElsevierEditoraLtda.Thisisanopen

accessarticleundertheCCBYlicense(http://creativecommons.org/licenses/by/4.0/).

PALAVRAS-CHAVE Fibrosecística; Insulina; Diabetesmelito

Insulinoterapiaempacientescomfibrosecísticanafasedepré-diabetes:umarevisão sistemática

Resumo

Objetivo: Elucidarseainsulinaéeficazounãoempacientescomfibrosecísticaantesdafase

dediabetes.

∗_{Correspondingauthor.}

E-mail:marianazorron@yahoo.com.br(M.Z.Pu).

http://dx.doi.org/10.1016/j.rppede.2016.03.006

Fontesdedados: OestudofoifeitodeacordocomométodoPrismaentreagostoesetembrode

2014,nasbasesdedadosPubMed,Embase,LilacseSciELO.Foramincluídosestudosprospectivos

publicadoseminglês,portuguêseespanholde2002a2014queavaliaramoefeitodainsulina

nosparâmetrospeso,índicedemassacorporalefunc¸ãopulmonar,empacientescomfibrose

cística,commédiade17,37anos,antesdafasedediabetes.

Síntesedosdados: Foramidentificadosoitoartigosque incluíram180indivíduos submetidos

ao usodeinsulina.Otamanhodasamostrasvariou de4a54pacientes, idademédiaentre

12,4e28anos.Otipodeacompanhamento,otempodeusodeinsulina,adoseeocronograma

deimplementac¸ãoforammuitoheterogêneosentreosestudos.

Conclusões: Existemrazõesteóricasparaseacreditarqueainsulinatenhaumefeitobenéfico

napopulac¸ãoestudada.Osdiferentesmétodosepopulac¸õesencontradosnãopermitemafirmar

seaterapiaprecocecominsulinadeveounãoserfeitaempacientescomfibrosecística

pre-viamenteaodiagnósticodediabetes.Portanto,sãonecessáriosestudoscomamostrasmaiores

euniformidadedeusodainsulina.

©2016SociedadedePediatriadeS˜aoPaulo.PublicadoporElsevierEditoraLtda.Este ´eum

artigoOpenAccesssobumalicenc¸aCCBY(http://creativecommons.org/licenses/by/4.0/).

Introduction

Cysticfibrosis-relateddiabetes(CFRD)isthemostcommon comorbidityinpatientswithcysticfibrosis(CF)andaffects 20%ofadolescentsand40---50%ofadultswithCF.1

Glucose disorders in CF patients typically begin with an intermittent postprandial hyperglycemia, followed by oralglucoseintolerancewithoutfastinghyperglycemiaand finallydiabeteswithfastinghyperglycemia.2,3

Insulindeficiencyandresultinghyperglycemiaaffectlung disease.3---5_{Insulinisahormonewithanaboliceffectsandits}

deficiencymayhaveanegativeclinicalimpactonpatients considered‘‘prediabetic’’.6_{Increasedserumglucoselevels}

(≥144mg/dL)mayhaveanadverseeffectonlungfunction. Furthermore,increasedglucoseinthebronchialtreefavors thegrowthofrespiratorypathogens.5_{Thereisstillalossof}

leanbodymassduetothecatabolicstatecausedbyinsulin deficiency,whichleadstoaconsumptionoffatandproteins andalsoaffectspulmonaryfunction.7

Therefore,insulindeficiencypromotesaclinical deteri-orationinthispopulationandnotonlyanabnormalglucose metabolism,whichmaybeenhancedbyearlyintervention withinsulin.6_{Bothdiabetesandglucoseintolerancereduce}

thelifeexpectancyofCFpatients;insulinistheonly treat-mentthatimprovesclinicaloutcomes.8_{Earlytreatmentwith}

insulinmayreducethemorbidityandmortalityofthe under-lyingdisease.9,10

Moreover,CF patients’classificationusingtheoral glu-cose tolerance test (OGTT) in intolerant and diabetic patients is based on criteria derived from epidemiologi-cal studies in non-CF subjects, it raises doubts whether theseconventionaldiagnostic limitswouldbeappropriate orrelevantforCFpatients.11_{Thus,theuseofconventional}

glucoseevaluationtestsintheCFpopulationcould under-estimate the number of patients with abnormal glucose metabolism, and, consequently, this group could benefit fromearlyinterventionwithinsulin,inglucoselevelsbelow those considered abnormal in populations without cystic fibrosis.12

Toourknowledge,thereisnosystematicreviewofearly initiationof insulintherapyin CFpatients.Therefore,the aimofthisstudywastoidentifytheeffectsofthis interven-tionandcontributetoclinicalpracticeandfuturestudies.

Method

ThesearchprocesswasdevelopedaccordingtothePrisma method(PreferredReportingItemsfor SystematicReviews andMeta-Analyses).13 _{The search wasconducted between}

August and September 2014 in the following electronic databases: PubMed, Lilacs, SciELO, and Excerpta Medica Database(Embase).

The following terms and descriptors (Medical Subjects Headings ---MeSH)were usedfor thesearch:‘cystic fibro-sis’,‘earlyinsulin’,‘insulin’,‘bodymassindex’,‘impaired glucosetolerance’,and‘therapy’;incombinations:‘cystic fibrosis and early insulin’, ‘cystic fibrosis and insulin and bodymassindex’,‘cysticfibrosisandearlyinsulin’,‘cystic fibrosisandinsulinandbodymassindex’,‘impairedglucose toleranceandcysticfibrosisandinsulinandtherapy’.

Studiespublished between2002 and 2014were identi-fiedthroughelectronicsearchbytwoindependentreviewers who evaluatedthe titles and abstracts of articles. Refer-ences of selected articleswere alsoreviewed in orderto identifystudies not foundin the surveyedbases. Discrep-anciesbetween reviewerswere discussedand resolvedby consensus.ThedateofthefirstsearchwasAugust28,2014, andthelast,September22,2014.

140mg/dL at any time, except at baseline and 120min; or postprandial glucose random or above 200mg/dL; or impairedglucosetolerance(IGT)diagnosisbyADAcriteria.14

Exclusioncriteriawere:(I)non-originalarticles,suchas letters,conference proceedingsandeditorials;(II)studies evaluated only CFRD without other types of disorders of glucose.

The extracted data were: study design; sample size; population characteristics; follow-uptime;type of insulin therapy (includingdose andregimeused); andeffects on weight,bodymassindex,andlungfunction.

Results

The initial search identified 508 articles, of which 111 wereselectedbasedontitlesandabstracts.Referencesof selectedpaperswerealsoreviewedandanadditionalstudy wasincluded.Ofthese,80wereidentifiedasduplicatesand removed; thus, 32 articleswere read in full, of which 24 wereexcludedbytheexclusioncriteria.Thefinalselection consistedofeightitems(Fig.1).Characteristicsofthestudy resultsaresummarizedinTable1.

Samplesizeoftheincludedstudiesrangedfrom4to54 patients,withmeanagefrom12.4to28years.Investigators andsubjectswerenotblindtothetreatmentassignmentin anyofthestudies.

Typeoffollow-up,insulin time,dose, and implementa-tionschedulewereveryheterogeneous,whichcanbeseen inTable1.Threestudiesusedcontrolgroupstocomparethe effectsofinsulin.Moranetal.15_{selectedcorresponding}

con-trolswhounderwentothertypesofintervention(repaglinide orplacebo),whileMinicuccietal.16_{usedcontrolswithIGT}

andKolouˇskováetal.17 _{usedcontrolswithnormalOGTTby}

ADAcriteria(NGT).Intheselast twostudies, controlsdid notundergopharmacologicalinterventions.

Inclusioncriteria for studies werevery heterogeneous. Mozzilloetal.18_{usedthefollowinginclusioncriteria:nouse}

ofsystemiccorticosteroidsandnoexacerbationoflung dis-ease.Minicuccietal.16 _{includedpatientswithatleastone}

ofthe following conditions:(I) bodymass indexBMI<10th percentile(p10_{);(II)lossofoneBMIpercentileforageand}

sexinthe previous year;(III) forcedexpiratory volumein onesecond (FEV1)≤80%of predicted;and (IV)decreased FEV1≥10%inthepreviousyear.Lungfunctiondeterioration andweightlosswerealsocriteriaforinclusionofsubjectsin thestudybyDobsonetal.6_{Incontrast,Moranetal.}15_chose

tointerveneinamoreclinicallystablegroupofpatientsand usedthefollowinginclusioncriteria:(I)endoflineargrowth; (II)weightstabilityinthelastthreemonths;(III)absenceof acuteinfectioninthelasttwomonths.Exclusioncriteriafor thisstudywere:(I)useoforalorintravenouscorticosteroids inthelastsixmonths;(II)fastinghyperglycemiainthe previ-ousyear;(III)liverdysfunction;(IV)pregnancy.Earlyinsulin deficiency,diagnosedbyintravenousglucosetolerancetest (IVGTT)and/orhighlevelsofglucoseinOGTT,wasusedas inclusioncriteriain thestudiesby Kolouˇskováetal.17 _and

Hameedetal.19

Five studies evaluated the effects of insulin in BMI of CF patients.15---18,20 _{Bizzarri et al.,}20 _{Moran et al.,}15 _and

Kolouˇskováet al.17 _{demonstrateda significant increase in}

BMIafterinsulinintervention.Mozilloetal.18 _{found a}

sig-nificant increase in BMI only in patients with initial BMI Z-score<−1.AlthoughMoranetal.15_{identifyimprovements}

in BMI in the group as a whole, in particular IGT group

Identification

Deleted duplicate articles (n=80)

Screening Additional records identified through_{references (n=1)}

Eligibility

Excluded full-text articles read in full (n=24)

Studies that include only CFRD* (n=7)

Studies that did not use insulin (n=11)

Retrospective studies (n=1) Letters, conference proceedings or editorial (n=5)

Included

Full-text articles assessed for eligibility (n=31)

Full-text articles read in full (n=32)

Articles included (n=8) Selected abstracts (n=111)

Records identified through database search

(n=508)

Table1 Characteristicsofincludedstudies.

Population Intervention Results

Sample characteristics

Meanfollow-up

(months)

Meanage(years)

(min---max)

Insulintypeand

averagedose

BMI FEV1

Dobsonetal.6 _NormalOGTT

glucoseand postprandial >200mg/dL

3 20.25

(15---23)

NPH6-8UI/d ouUR 5UI/d+7UI/d (70/30)

NA NSg

Casereport(n=4)

Bizzarrietal.20 _{IGT 16.8 18.2}

(9.2---27.8)

Glargine 0.3UI/kg/d

SI SI

Clinicaluncontrolled trial(n=6)

p=0.026 p=0.027

Mozzilloetal.18 _{2:CFRDFH+ 12 12.4}

(2.6---19)

Glargine 0.23UI/kg/d

SId _SI

Clinicaluncontrolled trial(n=22)

7:CFRDFH−

9:IGT p=0.017 p=0.01

4a

Moranetal.15 _{23:CFRDFH− 12 28±9 Aspart}

0.5UI:15gCHO

SIe _NS

Randomized controlledtrial (n=30)

7:severeIGTb _p=0.02

Kolouˇskováetal.17 _17:CFRDFH−

11:IGT

36 15.3c

(11.1---20.6)

NPH

0.12---0.25UI/kg/d

SIf _SIh

Randomized controlledtrial (n=28)

p<0.05 p=0.03

Drummondetal.12 _{24:CFRD 69.36 27.64}

(16---52)

UR,premixed, basaland basal-bolus

NA NS

Clinical

uncontrolledtrial (n=54)

18:IGT

12:NGT

Hameedetal.19 _{2:CFRDFH+ 9.6 12.5}

(7.2---18.1)

Detemir 0.13UI/kg/d

NA SI

Clinical uncontrolled trial(n=18)

4:CFRDFH−

6:IGT p=0.007

6:NGT

Minicuccietal.16 _{IGT 18 18}

(11---53)

Glargine 0.1---0.15UI/kg/d

NS NS

Randomized controlled

multicenterclinical trial(n=18)

CFRDFH+,CFRD withfastinghyperglycemia; CFRDFH−, CFRDwithout fastinghyperglycemia; NGT,normal glucosetolerance; CHO, carbohydrate;FEV1, forcedexpiratoryvolumeinone second;UR,ultrafastinsulin; NA,datanotevaluated; NS,notsignificant;SI, significantincrease.

a_{OneormoreofOGTTvalues>140mg/dL(betweenT30andT90).} b _{OGTT>200mg/dLatanytimeand180---199mg/dL120min.} c _Median.

d _{OnlyingroupwithinitialZ-score<−1.} e _{OnlyCFRDFH−group.}

f _{Onlyinsulindeficiencygroup.}

g_{Increasesuggested,butwithoutstatisticalevaluation.}

h _{Comparedwithcontrolgroup(therewasnostatisticaldifferenceintragroup).}

theydid not noticea significant increase in this parame-ter.Dobsonetal.,6_{Drummondetal.,}12_{andHameedetal.}19

chosetoassessbodyweight.Hameedetal.19_andDrummond

etal.12 _{foundsignificantweightgainafterinsulin}

interven-tion,whileDobsonetal.6_{suggestedthistrend,asdatawere}

notstatisticallyevaluated.

FEV1wastheonlyclinicalparameterassessedbyall stud-ies.Bizzarrietal.,20 _{Mozzilloetal.,}18 _{andHameedetal.}19

foundasignificantincreaseinFEV1aftertheuseofinsulin. Kolouˇskováetal.17_{foundthat,attheendoffollow-up,}

inter-ventiongrouphadhigherFEV1comparedtocontrolgroup.

Dobsonetal.6_{showedanapparentincreaseinthis}

param-eterwiththeuseofinsulin,asitwasonlyacasereport.In studiesbyMoranetal.,15_{Drummondetal.}12 _andMinicucci

etal.,16 _{FEV1remainedunchangedaftertheintervention.}

But Drummondetal.12 _{evaluatedseparatelyonly patients}

diagnosedwithIGTandfoundasignificantreductioninFEV1 declinerateinpatientsusinginsulin.Hameedetal.19

eval-uated separately only the early insulin-deficient patients (excludingpatientswithCFRD)andalsofoundinthisgroup a significant increase in FEV1. Moran et al.15 _{reported a}

comparedtoplacebo,butthisstabilitywasnotstatistically significant.

Mozzilloetal.18 _{found areducednumberofpulmonary}

exacerbations (as compared to the previous year), while Bizzarrietal.20 _{foundnochangesin thenumberof}

hospi-talizationsfor exacerbations. The fourpatients evaluated byDobsonetal.6_{showedanincreaseinforcedvital}

capac-ity(FVC)withtheuseof insulin.Hameed etal.19 _{found a}

significantimprovementinFVCaftertheintervention. In the results found by Bizzarri etal.,20 _{therewas no}

significantchangeinthelevelsofglycosylatedhemoglobin (HbA1c)afterinsulin,whereasthegroupofpatients evalu-atedbyMinicuccietal.16_{showedasignificantreductionin}

HbA1cwiththeuseofinsulin.

Frequentepisodesofhypoglycaemiawerereportedonly by Drummond et al.12 _{In the other studies cited in this}

review,the adverse effects of insulintherapy were infre-quentandwelltolerated.

Discussion

There are few published papers on the use of insulin in patients with cystic fibrosis prior to overt diabetes. Most arelimitedtoasinglecenterandmainlytoadults.Toour knowledge, this is the firstsystematic review to examine the benefitsand risksof insulinuse in CFpatients before thediagnosisofdiabetes.

Glucose intolerance indicates the presence of insulin deficiency,whichleadstoaproteinconsumptionand nega-tiveclinical/nutritionalimpact.Therefore,earlytreatment of insulincan have apositive effectin CFpatients in the prediabeticphase.20

The study results make sense when considering the pathophysiology of the evolution to CFRD. Initially, there isaninsulindeficiencythatgeneratesaproteincatabolism andglycemicexcursions,withconsequentdifficultygaining and maintaining weight and worsening of lung func-tion. Therefore, the introduction of insulin at this stage would likely prevent the catabolic effects of insulin deficiency.

Current dataare clear about theinsulin treatment for patientswithCFRDwithorwithoutfastinghyperglycemia,14

buttherearenoconsistentresultstodeterminewhetherthis treatmentshouldalsobeusedforthosewithotherglucose disorders,asitisnotwelldefinedwhatareglucosedisorders inthisspecificpopulation.Moreover,thereisdoubtwhether thecut-offvaluesforthediagnosisofCFRDandIGTarevalid for CF becausetheyare basedon population without the disease.

Anegativeimpactoftheprediabeticphaseisdescribedin nutritionalstatusandpulmonaryfunctionofCFpatients,6,21

itsuggeststhatinsulinshouldbestartedbeforethediagnosis of CFRD by the current methods available, as insulin has anaboliceffects and thesepatients have few side effects (hypoglycemia).Theonlystudyreportingfrequentepisodes ofhypoglycemia wasthestudy byDrummondetal.,12 _but

theyincluded severalinsulintherapy regimens inpatients withCFRD,IGT,andNGT,andtherewasnodescriptionofthe insulintypeordoseinstructionforeachgroup,whichmaybe relatedtothedifferenceinthefrequencyofhypoglycemia seenbetweenstudies.

Althoughmoststudieshaveasmallsamplesizeandused more than one type of insulin, only Moran et al.15 _and

Minicucci et al.16 _{reported no positive effects with early}

insulintherapy, but thesestudies have somepeculiarities describedbelow,suggesting thatearlyinitiation ofinsulin therapyinCFpatientscouldbebeneficial.

Minicuccietal.16 _{reportednoclinicalimprovementwith}

the use of glargine in CF patients with IGT (ADA crite-ria).The authors assumed that participation in the study made patients more aware of their change in glucose metabolism, leading to better nutritional behavior. Most otherstudies6,17---20 _{showedpositiveresults,buttheinsulin}

dosesusedwerehigher.

Mozzilloetal.18 _{foundthat after12monthsof therapy}

withinsulin,BMIcurve(CentersforDiseaseControl---CDC) improvedinpatientswithbaselineZ-scorebelow−1,which isinagreementwiththestudybyKolouˇskováetal.17 _that

alsoreportedimprovementinBMI,regardlessofthebaseline Z-score.

Kolouˇskováetal.17_{demonstratedthatinsulin}

administra-tionhaspositiveeffectsonleanbodymassduetoprotein catabolismreversal.However,controlgroupalsoshoweda tendencytowardsthatimprovement,probablyduetobetter nutritionalorientation, asinboth groups therewasa rec-ommendationfor increasedcaloricintakeupto120---150% of the daily needs. According to the authors, the results supporttheconceptthatinsulindeficiency,assessedinthis studybyusingIVGTTandOGTT,leadstoclinical deteriora-tioninCFpatientsandthatearlyinitiationofinsulintherapy couldberecommendedearlierthanitiscurrentlyaccepted (CFRD).

Dobson et al.,6 _{Bizzarri et al.,}20 _{and Hameed et al.}19

foundweightimprovementinpatientsusinginsulin. Moranetal.15_{reportedthatinsulinreversesweightlossin}

patientswithCFRDwithoutfastinghyperglycemia,butnotin patientswithsevereIGT.However,thestudypopulationwas inadulthoodandthereseemstobeabetterresponsewhen theseindividualsarein childhoodandadolescence. More-over,thegroupinquestionhadsevereIGT(OGTT≥200mg/dL atanytimeand120min between180and199mg/dL),and inotherstudies,patientswerepreviouslyselectedinsevere IGTphase,whichmayexplainthedifferencesfoundinthe results.

Bizarri et al.20 _{found improved FEV1 with no}

reduc-tioninpulmonaryexacerbations.Mozzilloetal.18 _reported

increasedFEV1andreducedpulmonaryexacerbationswith theuseof insulin.Hameedetal.19 _{demonstrateda fallin}

FEV1beforestartingtreatmentandimprovementafterthe introductionofinsulin.Kolouˇsková etal.17 _andDrummond

etal.12_{assessedFEV1comparedtountreatedsubjectsand}

identifiedadeclineinlungfunctionincontrolgroup,which wasnot seen in insulin-treated group. Moran etal.15 _and

Minicuccietal.16 _{foundnoimprovementinFEV1withearly}

useof insulin.However,in thestudy by Minicuccietal.16

there was a 10% decrease in FEV1 in the year prior to theintervention,whichmaybeabiasbecauseevenCFRD patientshavenosuchdeclineinthisparameter.Kolouˇsková etal.17 _{andDrummondetal.}12_{,whohadamoreconsistent}

numberofcases,foundthatFEV1waslowerintheuntreated groupcomparedtothosetreatedwithinsulin,which con-firmstheresultsdescribedbyDobsonetal.6_Bizarrietal.20_,

Dobson et al.6 _{suggest an improvement in pulmonary}

function(FEV1andFVC)withtheuseof insulin.However, theirsamplesizewassmall(n=4),selectedbyconvenience, and had nocontrol; the reevaluation occurred in a short period of time (3 months) without insulin standardiza-tion (more than one type of insulin was used) and there was no statistical analysis, probably due to the sample size.

Hameed et al.19 _{assessed height separately and found}

nodifferencesafterinsulintherapyinitiation.Thisis prob-ably due to the result seen in another study by Bizzarri etal.22_{TheysuggestthatinthedevelopmentofCFRDthere}

isalreadysubstantialandirreversibleimpairmentofheight becausemostpatientswithCFdevelopdiabetesatpuberty, thesametimethatthegrowthspurtoccurs.22

The evidence, favorable or unfavorable,to the use of insulinbeforeovertdiabetesinCFpatientsremains incon-clusive,withlittleknowledgeaboutlong-termresults.There arefewprospectivestudiesontheuseofinsulinbeforeovert diabetes in patients with CF and they include population withdifferenttypesofglucosedisorders,withoutagegroup delimitation,follow-uptime,andinsulintype,dosage,and implementationschedule.Onlytwostudies15,16_were

multi-centerandonlythreewerecontrolled.15---17_{Moreover,itwas}

notpossibletoassesstheeffectofimportantvariables,such asnutritionalroutine,andastandarddefinitionofglycemic disorders.

When analyzing the results of studies regarding the anaboliceffectsofinsulin,therearetheoreticalreasonsto believethatinsulinhasabeneficialeffectonthepopulation studied.However,theadditionofatreatmentfordiabetes inamultipledrugtreatmentregimeniscomplicated,which makesthisdecisionevenmorecontroversial.

Multicentertrialsinrandomizedpediatricpatients,with adequate nutritional support, type and standard doses of insulin(enoughtopromoteanabolism),areneededto deter-mine if treatment is justified or not. Keep in mind that placebo-controlledtrialsaredifficulttoperformduetothe factthatinsulinisaninjectablemedicine.

Conclusion

Thedifferentmethods andcase seriesusedin thestudies donotallowaffirmingthatearlyinsulintherapyshouldbe appliedin patientswith CFand glucosedisorders.To this end,studieswithlargersamples,dietstandardization,age group,anduniformityofinsulinuseareneeded.

Funding

This study did not receive funding. One of the coauthors receives doctoral fellowship from Fapesp (Fundac¸ão de Amparo à Pesquisa do Estado de São Paulo, process n◦ 2014/00611-2)

Conflicts

of

interest

Theauthorsdeclarenoconflictsofinterest.

References

1.MoranA, DunitzJ, NathanB,SaeedA,Holme B,ThomasW. Cysticfibrosis-relateddiabetes:currenttrendsinprevalence, incidence,andmortality.DiabetesCare.2009;32:1626---31. 2.LombardoF,DeLucaF,RosanoM,SferlazzasC,LucantoC,Arrigo

T,etal.Naturalhistoryofglucosetolerance,beta-cellfunction andperipheralinsulinsensitivityincysticfibrosispatientswith fastingeuglycemia.EurJEndocrinol.2003;149:53---9.

3.MoranA,BeckerD,CasellaSJ,GottliebPA,KirkmanMS,Marshall BC,etal.Epidemiology,pathophysiology,andprognostic impli-cationsofcysticfibrosis-relateddiabetes:atechnicalreview. DiabetesCare.2010;33:2677---83.

4.RolonMA,BenaliK,MunckA,NavarroJ,ClementA, Tubiana-RufiN,etal.Cysticfibrosis-relateddiabetesmellitus:clinical impactofprediabetesandeffectsofinsulintherapy.Acta Pae-diatr.2001;90:860---7.

5.BrennanAL,GyiKM,WoodDM,JohnsonJ,HollimanR,Baines DL,etal.Airwayglucoseconcentrationsandeffectongrowth of respiratory pathogens in cystic fibrosis. J Cyst Fibros. 2007;6:101---9.

6.DobsonL,HattersleyAT,TileyS,ElworthyS,OadesPJ,Sheldon CD.Clinicalimprovement incysticfibrosis withearlyinsulin treatment.ArchDisChild.2002;87:430---1.

7.RafiiM,ChapmanK,StewartC,KellyE,HannaA,WilsonDC, etal.Changesinresponsetoinsulinandtheeffectsofvarying glucosetoleranceonwhole-bodyproteinmetabolisminpatients withcysticfibrosis.AmJClinNutr.2005;81:421---6.

8.AlexanderS,AlshafiK,AndersonAK,Balfour-LynnI, SiânBent-ley, Buchdahl R, et al. Clinical guidelines: care of children with cystic fibrosis --- 2014. 6th ed; 2014. Available from: www.rbth.nhs.uk/childrencf[cited19.10.15].

9.LanngS,ThorsteinssonB,NerupJ,KochC.Diabetesmellitus incysticfibrosis:effectofinsulintherapyonlungfunctionand infections.ActaPaediatr.1994;83:849---53.

10.MerjanehL, He Q, Long Q, Phillips LS, Stecenko AA. Dispo-sition index identifies defective beta-cell function in cystic fibrosissubjectswithnormalglucosetolerance.JCystFibros. 2015.Availablefrom:http://www.sciencedirect.com/science/ article/pii/S1569199314001532[cited08.11.14].

11.Lek N, Acerini CL. Cystic fibrosis related diabetes mellitus ---diagnosticand managementchallenges.CurrDiabetesRev. 2010;6:9---16.

12.DrummondRS,RossE,BicknellS,SmallM,JonesG.Insulin ther-apyinpatientswithcysticfibrosisrelateddiabetesmellitus: benefit,timingofinitiationandhypoglycaemia.PractDiabetes Int.2011;28:177---82.

13.LiberatiA,AltmanDG,TetzlaffJ,MulrowC,GøtzschePC, Ioan-nidisJP,etal.ThePrismastatementforreportingsystematic reviews and meta-analyses of studies that evaluate health-care interventions: explanation and elaboration. PLoS Med. 2009;6:e1000100.

14.MoranA, BrunzellC,Cohen RC, KatzM,Marshall BC,Onady G, et al. Clinical care guidelines for cystic fibrosis-related diabetes:apositionstatementoftheAmericanDiabetes Asso-ciationandaclinicalpracticeguidelineoftheCysticFibrosis Foundation,endorsedbythePediatricEndocrineSociety. Dia-betesCare.2010;33:2697---708.

15.Moran A, Pekow P, Grover P, Zorn M, Slovis B, Pilewski J, etal.InsulintherapytoimproveBMIincysticfibrosis-related diabeteswithoutfasting hyperglycemia:resultsofthecystic fibrosisrelateddiabetestherapytrial.DiabetesCare.2009;32: 1783---8.

17.KolouˇskováS,ZemkováD,BartoˇsováJ,SkalickáV, ˇSumníkZ, VávrováV,etal.Low-doseinsulintherapyinpatientswithcystic fibrosisandearly-stageinsulinopeniapreventsdeteriorationof lungfunction:a3-yearprospectivestudy.JPediatrEndocrinol Metab.2011;24:449---54.

18.MozzilloE,FranzeseA, ValerioG, SepeA,DeSimoneI, Maz-zarellaG,etal.One-yearglarginetreatmentcanimprovethe courseoflungdisease inchildrenand adolescents with cys-ticfibrosisandearlyglucosederangements.PediatrDiabetes. 2009;10:162---7.

19.HameedS, Morton JR, Field PI, Belessis Y, Yoong T, KatzT, etal.Oncedailyinsulindetemirincysticfibrosiswithinsulin deficiency.ArchDisChild.2012;97:464---7.

20.BizzarriC,LucidiV,CiampaliniP,BellaS,RussoB,CappaM. Clin-icaleffectsofearlytreatmentwithinsulinglargineinpatients withcysticfibrosisandimpairedglucosetolerance.JEndocrinol Invest.2006;29:RC1---4.

21.ToféS,MorenoJC,MáizL,AlonsoM,EscobarH,BarrioR. Insulin-secretion abnormalities and clinicaldeterioration related to impairedglucosetoleranceincysticfibrosis.EurJEndocrinol EurFedEndocrSoc.2005;152:241---7.