Efficacy and Safety of IDegLiraVersus Basal-Bolus InsulinTherapy in Patients With Type 2Diabetes Uncontrolled onMetformin and Basal Insulin: TheDUALVII Randomized ClinicalTrialDiabetes Care 2018;41:1009–1016 | https://doi.org/10.2337/dc17-1114

OBJECTIVE

In patientswith uncontrolled type 2 diabetes onbasal insulin, prandial insulinmay beinitiated.We assessed the efficacy and safety of initiating insulindegludec/liraglutidefixed-ratio combination (IDegLira) versus basal-bolus insulin.

RESEARCH DESIGN AND METHODS

A phase 3b trial examined patients with uncontrolled type 2 diabetes on insulinglargine (IGlar U100) 20–50 units/day and metformin, randomized to IDegLira orIGlar U100 and insulin aspart £4 times per day.

RESULTS

Glycated hemoglobin (HbA1c) decreased from 8.2% (66 mmol/mol) to 6.7%(50 mmol/mol) with IDegLira and from 8.2% (67 mmol/mol) to 6.7% (50 mmol/mol)with basal-bolus (estimated treatment difference [ETD] 20.02% [95% CI20.16,0.12];20.2 mmol/mol [95% CI21.7, 1.3]), confirming IDegLira noninferiority versusbasal-bolus (P < 0.0001). The number of severe or blood glucose–confirmed symptom-atic hypoglycemia events was lower with IDegLira versus basal-bolus (risk ratio 0.39[95% CI 0.29, 0.51]; rate ratio 0.11 [95% CI 0.08, 0.17]). Body weight decreased withIDegLira and increased with basal-bolus (ETD23.6 kg [95% CI24.2,22.9]). Fastingplasma glucose reductions were similar; lunch, dinner, and bedtime self-monitoredplasma glucose measurements were significantly lower with basal-bolus. Sixty-sixpercent of patients on IDegLira vs. 67.0% on basal-bolus achieved HbA1c <7.0%(53 mmol/mol). Total daily insulin dose was lower with IDegLira (40 units) thanbasal-bolus (84 units total; 52 units basal).

CONCLUSIONS

In patients with uncontrolled type 2 diabetes on IGlar U100 andmetformin, IDegLiratreatment elicited HbA1c reductions comparable to basal-bolus, with statisticallysuperior lower hypoglycemia rates and weight loss versus weight gain.

1NorthShore University HealthSystem, Evanston,IL2University of Chicago Pritzker School of Medi-cine, Chicago, IL3PrimeCare Medical Group, Houston, TX4La Rochelle Hospital, La Rochelle, France5Santojanni Hospital and Cenudiab, CiudadAutonoma de Buenos Aires, Argentina6Endocrine andMetabolic Consultants, Rockville,MD7Medical School, National and Kapodistrian Uni-versity of Athens, Athens, Greece8Novo Nordisk A/S, Søborg, Denmark9H. U. QuironSalud Madrid y Ruber Juan Bravo,Universidad Europea de Madrid, Madrid, Spain

Corresponding author: Liana K. Billings, lbillings@northshore.org.

Received 5 June 2017 and accepted 5 February2018.

Clinical trial reg. no. NCT02420262, clinicaltrials.gov.

This article contains Supplementary Data onlineat http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc17-1114/-/DC1.

This article is featured in a podcast available athttp://www.diabetesjournals.org/content/diabetes-core-update-podcasts.

© 2018 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 thework is notaltered.More infor-mation is available at http://www.diabetesjournals.org/content/license.

Liana K. Billings,1,2 Ankur Doshi,3

Didier Gouet,4 Alejandra Oviedo,5

Helena W. Rodbard,6

Nikolaos Tentolouris,7 Randi Grøn,8

Natalie Halladin,8 and Esteban Jodar9

Diabetes Care Volume 41, May 2018 1009

CLIN

CARE/ED

UCATIO

N/N

UTR

ITION/PSYC

HOSO

CIAL

Progression of type 2 diabetes means ad-equate blood glucose (BG) control withbasal insulin may deteriorate over time.By the timebasal insulin is initiated,patientson average have had a type 2 diabetes di-agnosis for 9.2 years and a glycated hemo-globin (HbA1c) level of 9.5% (80mmol/mol)(1). In one study, only 29% of patients couldmaintain HbA1c ,7.0% (53 mmol/mol)3 years after basal insulin initiation (2). Cur-rent guidelines recommend therapy inten-sification if HbA1c targets are not reachedafter 3–6months of basal insulin (3). How-ever,manypatients remainwithout therapyintensification (4) due to concerns over po-tential increased hypoglycemia risk, weightgain, and treatment complexity (5,6).Glucagon-like peptide 1 receptor ago-

nists (GLP-1RAs) are incretin mimeticsthat lower fasting plasma glucose (FPG)and postprandial glucose via stimulationof endogenous insulin and inhibition ofglucagon secretion, with weight-reducingeffects through appetite suppression (7).Using basal insulin with GLP-1RAs is a rec-ognized therapeutic option after failureto achieve glycemic control with basal in-sulin (3). Comparison of basal insulin andGLP-1RA given in separate injections ver-sus basal insulin alone showed a signifi-cantly greater HbA1c reduction and weightloss with the combination therapy, withsignificantly more hypoglycemic events(8). Comparison with basal-bolus showednoninferior HbA1c reduction with basalinsulin and GLP-1RA loose combination aswell as more desirable hypoglycemia ratesandweightmanagement (9). This evidencesupports combination therapies such asIDegLira, a fixed-ratio combination of in-sulin degludec (degludec) and liraglutide.IDegLira efficacy and safetywere estab-

lished in the Dual Action of Liraglutide andInsulinDegludec in Type2Diabetes (DUAL)clinical trial program in patients with un-controlled type 2 diabetes on oral antidia-betic drugs (DUAL I, IV, and VI) (10–12),GLP-1RAs (DUAL III) (13), and basal insu-lin (DUAL II and V) (14,15). DUAL VII as-sessed IDegLira efficacy and safety versusbasal-bolus.

RESEARCH DESIGN AND METHODS

Trial Design and ParticipantsDUAL VII was a phase 3b, multinational(Supplementary Table 1), open-label,two-arm parallel, randomized trial in pa-tients with type 2 diabetes conducted at89 sites in 12 countries from July 2015 toOctober 2016. The trial lasted 32 weeks,

with 2-week screening, 26-week treat-ment, and 4-week follow-up periods. Itwas conducted in accordance with the In-ternational Conference onHarmonisationGuideline for Good Clinical Practice (16)and the Declaration of Helsinki (17).

Patients included were $18 years oldwith type 2 diabetes, HbA1c 7.0–10.0%(53–86 mmol/mol), BMI #40 kg/m2,and on stable daily doses of insulin glar-gine 100 units/mL (IGlar U100) 20–50units and metformin$1,500 mg or max-imum tolerated dose for .90 days priorto screening. Exclusion criteria includedtreatment with any medication for dia-betes or obesity other than stated in theinclusion criteria during 90 calendar daysbefore screening, anticipated initiationor change in concomitant medicationsknown to affect weight or glucosemetab-olism, and renal impairment (estimatedglomerular filtration rate ,60 mL/min/1.73 m2) (Supplementary Table 2).

A treat-to-target approach ensured op-timal insulin titration and improved glyce-mic control in all patients.

Randomization and MaskingPatients were randomized 1:1 using acentralized allocation via an interactivevoice web response system.

Procedures/InterventionsPatients were randomized to eitherIDegLira (insulin degludec 100 units/mLand liraglutide 3.6 mg/mL, in a 3-mLprefilled FlexTouch pen for subcutaneousinjection) or IGlar U100 (insulin glargine100 units/mL solution, in a 3-mL prefilledSolostar pen for subcutaneous injection)and insulin aspart (IAsp) (100 units/mLsolution in a 3-mL prefilled FlexPen forsubcutaneous injection). Metformin wascontinued at pretrial doses.

IDegLira was administered once dailyatany time, independentofmeals, repeatedapproximately the same time each day. Pa-tients were initiated on 16 units (16 unitsdegludec/0.58 mg liraglutide) and titratedtwice weekly, aiming for a mean prebreak-fast self-monitored plasma glucose (SMPG)target range of 4.0–5.0 mmol/L (72–90 mg/dL) (Supplementary Table 3A).The maximum dose of IDegLira was 50units (50 units degludec/1.8mg liraglutide).IGlar U100 was administered once dailyaccording to local labeling and was initi-ated at a dose equivalent to the pretrialdose, titrated using the same algorithm asIDegLira. IAsp was initiated from the day of

randomization at a starting dose of 4 unitsper main meal (#4 times per day) andtitrated twiceweekly, aiming for ameanpre-prandial and bedtime SMPG target range of4.0–6.0 mmol/L (72–108 mg/dL) (Supple-mentary Table 3B). The principal investi-gator at each study site could adjust thetitrationper clinical judgment. Therewereno protocols given to patients regardingmeal size or constituents.

Patients who prematurely discontin-ued treatment had telephone contactsevery 4 weeks to collect information onserious adverse events (AEs) (death, life-threatening experience, inpatient hospi-talization or prolongation of existinghospitalization, persistent/significant dis-ability/incapacity, congenital anomaly,event that jeopardizes the patient requir-ing medical/surgical intervention, andsuspicion of transmission of infectiousagents via trial product), major cardiovas-cular events, and antidiabetic medica-tion. These patients were also invitedto the week-26 visit for data collectionon these parameters, as well as HbA1c. Asmall number of patients completed theweek-26 visit as a telephone contact, withno HbA1c collected.

Outcome MeasuresThe primary end point was change inHbA1c from baseline to week 26 of treat-ment. Confirmatory secondary end pointsincluded number of treatment-emergentsevere or BG-confirmed symptomatic hy-poglycemic episodes during 26 weeks oftreatment and change in body weightfrom baseline after 26 weeks of treat-ment. Hypoglycemic events were definedas either severe according to the Ameri-can Diabetes Association (ADA) classifica-tion (requiring assistance of anotherperson to take corrective actions) (18)or symptomatic BG-confirmed (,3.1mmol/L [56 mg/dL] accompanied by gly-copenic symptoms). Supportive second-ary efficacy end points included total,basal, and bolus daily insulin doses; re-sponders for HbA1c ,7.0% (53 mmol/mol)and#6.5% (48mmol/mol) after 26weeksof treatment; and proportion of patientsachieving these HbA1c targets withoutsevere or BG-confirmed symptomatic hy-poglycemia in the last 12 weeks and/orweight gain. Hypoglycemia rates duringthe last 12 weeks were included as thiswas considered a maintenance period withmore stable insulin doses and comparablehypoglycemia rates. Other glycemia-related

1010 IDegLira Versus Basal-Bolus Insulin: DUAL VII Diabetes Care Volume 41, May 2018

secondary end points included changes inFPG and mean difference in the nine-pointSMPG profile.Supportive secondary safety end points

included number of treatment-emergentAEs and treatment-emergent nocturnalsevere or BG-confirmed symptomatichypoglycemic episodes (severe or BG-confirmed symptomatic hypoglycemiaoccurring between 0001 h and 0559 hinclusive) (Supplementary Fig. 1) andchanges from baseline in blood pressure,heart rate, laboratory assessments of bio-chemistry, hematology, and fasting lipidprofile.

Statistical AnalysesThe trial was powered for the primaryobjective of confirming noninferioritywith respect to HbA1c change from base-line, using a one-sided Student t test withthe following assumptions: a mean treat-ment difference of 0.0%, an SD fromHbA1c change from baseline of 1.0%, anoninferiority margin of 0.30%, and upto 15% of randomized patients excludedfrom the per-protocol analysis set. Sam-ple sizewas set to 250 per treatment arm,ensuring a power of at least 85% for con-firming the primary objective in the per-protocol analysis set.All efficacy end points were summa-

rized using the full analysis set (FAS),and safety end points were summarizedusing the safety analysis set (SAS). All sta-tistical analyses of efficacy and safety endpointswere basedon the FAS (analysis setdefinitions are detailed in SupplementaryTable 4).For the primary end point, noninferior-

ity was considered confirmed if the upperbound of the two-sided 95% CI for esti-mated mean between-treatment differ-ence in change from baseline in HbA1cwas ,0.30%. All postbaseline HbA1c

on-treatment measurements obtainedat scheduled visits were analyzed usinga linear mixed normal model using an un-structured residual covariance matrix forHbA1cmeasurementswithin the samepa-tient. The model included treatment,visit, and region as fixed factors and base-line HbA1c as covariate, with interactionsbetween visit and all factors and covari-ates included. In order to control theoverall type I error on a 5% level withregard to confirmatory secondary endpoints, a hierarchical testing procedurewas used (Statistical Analyses Detail inSupplementary Data).

A negative binomial model with a log-link function and the log of the time pe-riod in which a hypoglycemic episode isconsidered treatment emergent as off-set was used to analyze the numberof treatment-emergent severe or BG-confirmed symptomatic hypoglycemicepisodes, including treatment and regionas fixed factors. Risk of hypoglycemia ver-sus no hypoglycemia was analyzed usinga generalized linear regression modelwith a log link, including treatment andregion as fixed factors.

A mixed model for repeated mea-surements (MMRM) with an unstruc-tured covariance matrix was used forthe continuous confirmatory and sup-portive secondary end points, includingtreatment, visit, and region as fixed fac-tors and corresponding baseline valuesas covariates. Interactions between visitand all factors and the covariate werealso included in the model. Insulin dosewas analyzed using a compound symme-try covariance matrix and included IGlarU100 dose at screening as an additionalcovariate.

A number of sensitivity analyses, in-cluding reference-based multiple im-putation, were performed to assessthe robustness of primary and confirma-tory secondary analyses (SupplementaryTable 5).

The nine-point SMPG profile was ana-lyzed using a linear mixed-effect modelfitted to data at week 26. The model in-cluded treatment, region, time (withinnine-point profile), and interactions be-tween treatment and time and betweentime and region as fixed factors and pa-tient as a random effect, wheremeasure-ments within patients were assumedcorrelated using a compound symmetrycovariance matrix.

Responder end points were analyzedusing a logistic regression model withtreatment and region as fixed factorsand baseline HbA1c with or without base-line body weight as covariates. Missingresponse data were imputed from theMMRM analysis of the primary end pointchange in HbA1c from baseline with orwithout the confirmatory secondary endpoint change in body weight. AEs weresummarized descriptively.

RESULTS

PatientsOverall, 672 patients were screened and506 were randomized, with 99.2% of

those randomized to IDegLira and 98.0%of those randomized to basal-boluscompleting the trial. Furthermore, 94.4%and 91.7% completed 26weeks of treat-ment with IDegLira and basal-bolus,respectively (Fig. 1). Baseline character-istics were similar between the two arms(Table 1).

Primary End PointAfter 26weeks of treatment, mean HbA1cdecreased from 8.2% (66 mmol/mol) atbaseline to 6.7% (50 mmol/mol) at endof trial (EOT) with IDegLira and from 8.2%(67 mmol/mol) to 6.7% (50 mmol/mol)with basal-bolus. The estimated treatmentdifference (ETD)was20.02%(95%CI20.16,0.12) (20.2 mmol/mol [95% CI21.7, 1.3])(Fig. 2A), confirming noninferiority ofIDegLira versus basal-bolus (P , 0.0001).In total, 489 patients (244 and 245 in theIDegLira and basal-bolus arms, respec-tively) contributed to the primary anal-ysis. All preplanned sensitivity analyseswere in agreement with the primaryanalysis (Supplementary Tables 5 and 6and Supplementary Fig. 2).

Confirmatory Secondary End PointsDuring 26 weeks of treatment, 19.8% ofpatients on IDegLira experienced one ormore severe or BG-confirmed symptom-atic hypoglycemic episodes vs. 52.6%with basal-bolus treatment, correspond-ing to a 61% lower risk with IDegLira com-pared with basal-bolus (estimated riskratio 0.39 [95% CI 0.29, 0.51], P ,0.0001). Patients on IDegLira experienced129 episodes in total (rate of 1.07 epi-sodes per patient year of exposure[PYE)] versus patients on basal-boluswho experienced 975 episodes in total(rate of 8.17 episodes per PYE). This re-sulted in an 89% lower rate of severe orBG-confirmed symptomatic hypoglyce-mic episodes with IDegLira versus basal-bolus (estimated rate ratio 0.11 [95% CI0.08, 0.17], P , 0.0001), confirming thesuperiority of IDegLira versus basal-bolus.Figure 2B demonstrates an increasingdivergence in the cumulative hypoglyce-mia rate from randomization until EOT.Severe hypoglycemic episodes occurredin 1.2% of patients on IDegLira through-out the treatment period (three episodes)at a rate of 0.03 episodes per PYE vs. 1.6%of patients on basal-bolus (10 episodes)at a rate of 0.08 episodes per PYE. Therewas a 34% lower risk (estimated risk ratio0.76 [95% CI 0.17, 3.35], P = 0.7173)

care.diabetesjournals.org Billings and Associates 1011

and 72% lower rate (estimated rate ratio0.28 [0.04, 1.98], P = 0.2034) of severehypoglycemia in IDegLira compared withbasal-bolus. Over 26 weeks of treatment,observed mean body weight decreasedby 0.9 kg with IDegLira from 87.2 kgand increased with basal-bolus by 2.6 kgfrom 88.2 kg (ETD 23.6 kg [95%CI24.2,22.9], P, 0.0001) (Fig. 2C), con-firming the superiority of IDegLira overbasal-bolus.All preplanned sensitivity analyses, includ-

ing reference-based multiple imputation,were in agreement with conclusions fromconfirmatory analyses (SupplementaryTables 5 and 6 and Supplementary Fig. 2).

Supportive Secondary End PointsTotal daily insulin dose increased toa mean EOT dose of 40 and 84 units withIDegLira and basal-bolus, respectively (ETD244.5 units [95% CI248.3,240.7], P ,

0.0001) (Fig. 2D). Mean basal insulin dosewith basal-bolus increased from34unitsin week 1 to 52 units at EOT vs. a meanEOT dose of 40 units with IDegLira (cor-responding to 40 units degludec/1.44 mgliraglutide) (ETD 212.6 units [95% CI214.9, 210.3], P , 0.0001). Total dailybolus insulin dose in the basal-bolus armat EOT was 32 units of bolus insulin. After26 weeks of treatment, 86 patients onIDegLira were on the 50-unit maximumdose,withameanHbA1c atweek26of7.0%(53 mmol/mol); 57% of these patientsachieved HbA1c ,7.0% (53 mmol/mol),with a mean HbA1c at week 26 of 6.4%(46 mmol/mol). Those who did not achievean HbA1c ,7.0% on 50 units IDegLirahad a mean HbA1c at week 26 of 7.8%(61 mmol/mol).

By week 26, ;90% of patients onbasal-bolus reported taking at least threeinsulin injections per day, including the

once-daily basal injection (SupplementaryFig. 3) versus the once-daily single injec-tion with IDegLira.

Patients on IDegLira had a lower rate ofnocturnal hypoglycemia compared withbasal-bolus. During the 26 treatmentweeks, 4.8% of patients on IDegLira expe-rienced one or more nocturnal severe orBG-confirmed symptomatic hypoglyce-mic episodes (16 episodes) at a rate of0.13 episodes per PYE vs. 19.4% of pa-tients on basal-bolus (198 episodes) at arate of 1.66 per PYE. There was a 75%lower risk and 92% rate reduction in noc-turnal hypoglycemia with IDegLira versusbasal-bolus (estimated risk ratio 0.25[95% CI 0.13, 0.45], estimated rate ratio0.08 [95% CI 0.04, 0.17]; both P ,0.0001). No patients on IDegLira experi-enced any nocturnal severe hypoglycemicepisodes versus one episode in the basal-bolus arm.

Similar proportions of patients in eachtreatment arm achieved an HbA1c,7.0%(53 mmol/mol): 66% in IDegLira and67% in basal-bolus (Fig. 3). Likewise,49.6% and 44.6% of patients on IDegLiraand basal-bolus, respectively, achievedHbA1c #6.5% (48 mmol/mol) (ETDs notstatistically significant for both targets)(Supplementary Fig. 4). More patientsreached the triple composite HbA1c tar-gets (,7.0% [,53 mmol/mol]) withouthypoglycemic episodes in the last 12weeks of treatment and without weightgain on IDegLira versus basal-bolus (oddsratio [OR] 10.39 [95% CI 5.76, 18.75], P,0.0001) (Fig. 3). Odds for HbA1c tar-gets ,7.0% (53 mmol/mol) without hy-poglycemic episodes in 26 weeks oftreatment and without weight gain were

Table 1—Baseline characteristics

Characteristic IDegLira IGlar U100 + IAsp

FAS, n 252 254

Male, n (%) 110 (43.7) 117 (46.1)

Age (years) 58.6 (9.0) 58.0 (8.6)

Weight (kg) 87.2 (16.0) 88.2 (17.2)

BMI (kg/m2) 31.7 (4.4) 31.7 (4.5)

Duration of diabetes (years) 13.2 (7.0) 13.3 (6.8)

HbA1c (%) 8.2 (0.8) 8.2 (0.8)

HbA1c (mmol/mol) 66 (8) 67 (9)

FPG (mmol/L) 8.5 (2.7) 8.3 (2.5)

FPG (mg/dL) 153.5 (47.8) 149.3 (45.6)

Daily insulin dose (units) 34 (10.7) 33 (10.4)

Daily metformin dose (mg) 2,049 (456.0) 2,091 (458.3)

Values are mean (SD) unless otherwise stated.

Figure 1—Patient disposition. Treatment completers are patients who completed the week-26 visit without premature permanent discontinuation ofrandomized treatment. Trial completers are treatment completers and patients who completed the week-26 visit after premature permanent discon-tinuation.

1012 IDegLira Versus Basal-Bolus Insulin: DUAL VII Diabetes Care Volume 41, May 2018

higher with IDegLira versus basal-bolus(OR 12.56 [95% CI 6.46, 24.45], P ,0.0001) (Fig. 3).The observed mean FPG decreased

with IDegLira by 2.4 mmol/L and by1.9 mmol/L with basal-bolus (ETD 20.31mmol/L [95% CI20.67, 0.05], P = 0.0936)(Fig. 2E).

The nine-point SMPGprofile decreasedwith both treatments at all time points(Fig. 2F) but to a larger extent with basal-bolus than with IDegLira. The mean ofthe nine-point SMPG decreased from9.7 to 7.2 mmol/L with IDegLira andfrom 9.7 to 6.6 mmol/L with basal-bolus(ETD 0.57 mmol/L [95% CI 0.31, 0.83],

P, 0.0001; in favor of basal-bolus). Basal-bolus insulin treatment resulted in sta-tistically significantly lower postlunch,dinner, postdinner, and bedtime SMPGreadings (Fig. 2F). The mean prandial in-crements decreased in both arms at allthreemeals butwere only statistically sig-nificantly different in favor of basal-bolus

Figure 2—Results forHbA1c (A), severeorBG-confirmedsymptomatic hypoglycemia (B), change inbodyweight (C), daily total insulindose (D), FPG (E), andnine-point SMPG profiles (F). A: Mean observed values with error bars (SEM) based on FAS. B: Mean cumulative frequency based on the SAS. Severe orBG-confirmed symptomatic describes an episode that is severe according to the ADA classification or BG-confirmed by a plasma glucose value,3.1 mmol/L(,56mg/dL)with symptoms consistentwithhypoglycemia.C: Least squaremean (LSMean) valueswith error bars (SE of estimated LSMean) basedon FAS,using MMRM. D: Mean observed values with error bars (SEM) based on the SAS. E: Mean observed values with error bars (SEM) based on FAS. F: Meanobserved values with error bars (SEM) based on FAS. *Statistically significant difference between EOT means in favor of IGlar U100 + IAsp vs. IDegLira.

care.diabetesjournals.org Billings and Associates 1013

at the evening meal (ETD 0.49 mmol/L[95% CI 0.06, 0.92], P = 0.0261).

Clinical Observations and LaboratoryValuesA greater decrease inmean systolic bloodpressure was observed with IDegLira(24.5 mmHg) versus basal-bolus (21.2mmHg) (ETD 23.70 [95% CI 25.68,21.72], P = 0.0003), but no statisticallysignificant difference was observed be-tween the two arms in terms of changesin diastolic blood pressure. There was agreater difference in increase in heart ratewith IDegLira (3.5 bpm) versus basal-bolus(1.3 bpm) (ETD 2.73 [95% CI 1.30, 4.16], P =0.0002). Small but statistically significantdifferences were seen between the twoarms for various lipid profile parameters(Supplementary Table 8).

AEsAEs occurred in similar proportions of pa-tients in both treatment arms: 59.1% ofIDegLira patients at a rate of 4.1 eventsper PYE vs. 56.9% of patients on basal-bolus at a rate of 3.5 events per PYE(Supplementary Table 7). The majorityof AEs were nonserious, and there wasno clustering of events: 13 and 11 seriousAEswith IDegLira and basal-bolus, respec-tively (Supplementary Table 7). Therewere no fatal AEs. The most common AEwith IDegLira was nausea, with 11.1% vs.1.6% of patients on basal-bolus reporting

one ormore events. At all time points, thereported nausea rate was ,3% in eithertreatment arm (Supplementary Fig. 5).The most common AE on basal-boluswas nasopharyngitis, with 11.9% vs.4.8% of patients on IDegLira reportingone or more events. There was one con-firmed adjudicated cardiovascular eventin the IDegLira arm, confirmed as unsta-ble angina pectoris. There was one con-firmed benign neoplasm (a colorectalpolyp) with IDegLira and no confirmedneoplasms on basal-bolus. There wereno confirmed events of pancreatitis orthyroid disease or major cardiovascularevents in the trial.

CONCLUSIONS

This is the only trial thus far that com-pares a fixed-ratio combination of basalinsulin and GLP-1RA with basal-bolus.With once-daily versus multiple daily in-jections, IDegLira was noninferior tobasal-bolus with respect to HbA1c reduc-tion and statistically superiorwith respectto lower hypoglycemia rate and change inbodyweight in patients on IGlarU100 andmetformin with uncontrolled BG.

Other trials have investigated combi-nations of insulin and GLP-1RA given sep-arately. In the BEGIN: VICTOZA ADD-ONtrial, addition of liraglutide to degludec inseparate injections resulted in a signifi-cantly greater reduction in HbA1c versus

addition of IAsp at one meal to degludec,with weight loss and a reduced risk ofhypoglycemia (19). Addition of once-weekly albiglutide to IGlar U100 resultedin similar HbA1c reductions versus IGlarU100 and prandial insulin lispro, withweight loss and a reduced risk of hypo-glycemia (9). These results are mirroredin this trial. Furthermore, the fixed com-bination achieved these outcomes withonly one injection per day (vs. #5 in thebasal-bolus arm) and lower rates of nau-sea, likely attributable to the ability toslowly titrate the IDegLira dose in con-trast to the separate injections (9,19).

An insulin-sparing effect was demon-strated with IDegLira versus basal-bolus.Patients achieved similar HbA1c reduc-tions with a mean daily dose of;13 unitsof basal insulin less with IDegLira andnearly 45 units of total daily insulin less.

Despite being an efficacious glucose-lowering therapy, basal-bolus treatmentis associated with a higher rate of hypo-glycemia versus other diabetes therapies(20). Although resulting in similar HbA1creductions, IDegLira was confirmed tobe statistically superior to basal-bolus,with an 89% lower rate of severe orBG-confirmed symptomatic hypoglycemicepisodes and 92% lower rate of nocturnalhypoglycemia. The striking difference inthe cumulative incidence of hypoglycemiawas noted early in the treatment period,

Figure 3—Patients achieving composite outcomes with HbA1c,7.0% (53 mmol/mol) at week 26. ORs and P values are based on logistic regression. NS, notsignificant. *Statistically significant in favorof IDegLira.†SevereorBG-confirmedsymptomatichypoglycemiabasedonepisodesduring the last12weeksof treatment.‡Severe or BG-confirmed symptomatic hypoglycemia basedon hypoglycemic episodes during a patient’s entire 26weeks of treatment (post hoc analysis).

1014 IDegLira Versus Basal-Bolus Insulin: DUAL VII Diabetes Care Volume 41, May 2018

with an immediate split between thetwo curves at randomization and continu-ing divergence toward EOT. This differ-ence is likely to be explained by the lowertotal daily insulin dose with IDegLira;the glucose-stimulated insulin secretionpromoted by the GLP-1RA componentof IDegLira (7), which reduces the riskof low BG (21); and the lower rate of hy-poglycemia with degludec versus IGlarU100 (22).Weight gain is another concern when

initiating basal-bolus treatment, and in-creasing doses of insulin can promoteweight gain (3,6). Here, IDegLira was con-firmed to be statistically superior to basal-bolus in terms of change in body weight.This is likely to be due to the weight-reducing and insulin-sparing effects ofliraglutide mitigating the risk of weightgain with insulin (3).There were no unexpected safety or

tolerability issues identified with IDegLirain this trial, and the safety profile was asexpected based on results from previousDUAL andmonocomponent trials, includ-ing the LEADER (Liraglutide Effect andAction in Diabetes: Evaluation of Cardio-vascular Outcome ResultsdA Long TermEvaluation) and DEVOTE (Trial ComparingCardiovascular Safety of Insulin DegludecVersus Insulin Glargine in Subjects WithType 2 Diabetes at High Risk of Cardiovas-cular Events) cardiovascular outcomes tri-als (23,24).The main limitation in this trial was its

open-label design, which may have af-fected reporting of AEs and other results.However, it was necessary to avoid addi-tional placebo injections with IDegLira,which were deemed an unacceptableburden to patients.Trial design meant patients treated

with basal-bolus continued IGlar U100,whereas those treated with IDegLira trans-ferred from IGlar U100 to degludec asa basal insulin. The comparison of differentbasal insulins may be seen as a limita-tion as they have different pharmacoki-netic properties (25). Notably, degludechas an efficacy similar to IGlar U100 interms of HbA1c lowering (24). However,IGlar U100 was the chosen comparatorbecause it was the most prescribed basalinsulin at the time of study design.Additionally, this trial used more strin-

gent titration targets than some cur-rent guidelines. The titration targetfor preprandial SMPG (4.0–6.0 mmol/L[72–108 mg/dL]) with IAsp was lower

than the ADA-recommended target of4.4–7.2 mmol/L (80–130 mg/dL) (20).Thus, higher doses of preprandial insulinmay have been used in this trial, contrib-uting to the difference in total daily insu-lin dose, weight, and hypoglycemia atEOT between arms. Nonetheless, a sim-ilar significant lowering of nocturnal hy-poglycemia was observed with IDegLiraversus basal-bolus when fasting treat-ment targets were lower than ADA-recommended targets, but there wasno difference in FPG levels at EOT betweentreatment arms.

Medication compliance may be higherin this study than might be expected inthe real world, where compliance de-creases as the number of insulin injectionsincreases (26). This trial contains real-world features, including the choice ofbasal comparison, no specific meal plan,and open-label design. Real-world stud-ies are warranted to determine whetherthe results are replicated in clinicalpractice.

ConclusionIDegLira provides an efficacious intensifi-cation option with noninferior glycemiccontrol versus basal-bolus in patients withtype 2 diabetes on 20–50 units of basalinsulin with no renal impairment andHbA1c levels of 7.0–10.0%. Comparedwith basal-bolus, IDegLira offers an alter-native well-tolerated treatment withfewer injections, doses taken indepen-dently of meals, lower total daily insulindose, reduced monitoring, weight loss,and reduced rate of hypoglycemic epi-sodes. This alternative might help over-come the inertia that currently leavesmany patients with poor glycemic control.

Acknowledgments. The authors thank the in-vestigators, research coordinators, and patientsin the trial as well as Bue F. Ross Agner and JakobLanger (both of Novo Nordisk A/S) for theirreview and input to themanuscript. The authorsalso thankVictoriaStoneandGermanicusHansa-Wilkinson (Watermeadow Medical, an AshfieldCompany, Oxford, U.K.) for providing medicalwriting and editorial support.Funding and Duality of Interest. This trial wasfunded by Novo Nordisk A/S in accordance withGood Publication Practice (GPP3) guidelines(www.ismpp.org/gpp3). L.K.B. has participatedin advisory panels for Novo Nordisk and is onthe speakers’ bureau for Novo Nordisk. A.D. hasparticipated in research trials as a principal inves-tigator for Novo Nordisk, AstraZeneca, Eli Lilly andCompany, Pfizer, and Sanofi. D.G. has receivedresearch support from Novo Nordisk, Eli Lilly

and Company, Boehringer Ingelheim, Medtronic,Johnson & Johnson, and Sanofi. A.O. has receivedresearch support fromNovoNordisk and is on thespeakers’ bureau for Novo Nordisk. H.W.R. hasparticipated in advisory panels for AstraZeneca,Boehringer Ingelheim, Eli Lilly and Company,Janssen, Merck, Novo Nordisk, Sanofi, andRegeneron; is a consultant for AstraZeneca,Boehringer Ingelheim, Eli Lilly and Company,Janssen, Merck, Novo Nordisk, Sanofi, and Re-generon; has received research support fromAstraZeneca, Bristol-Myers Squibb, BoehringerIngelheim, Eli Lilly and Company, Janssen, Lexicon,Merck, Novo Nordisk, Sanofi, and Regeneron; andis on the speakers’ bureau for AstraZeneca, Eli Lillyand Company, Merck, Novo Nordisk, and Sanofi.N.T. has participated in advisory panels forMerck,AstraZeneca, Sanofi, Novo Nordisk, ELPEN, Eli Lillyand Company, Boehringer Ingelheim, and Novar-tis and has received research support fromMerck,Eli Lilly andCompany,NovoNordisk, Sanofi, Pfizer,AstraZeneca, Janssen, Cilag, GlaxoSmithKline, andNovartis. R.G. is an employee of Novo Nordisk.N.H. is an employee of Novo Nordisk. E.J. hasparticipated in advisory panels for Janssen, Eli Lillyand Company, Merck, and Novo Nordisk; hasreceived research support from Janssen, Eli Lillyand Company, Merck, Novo Nordisk, Pfizer, andSanofi; and is on the speakers’ bureau for Janssen,Eli Lilly and Company, Merck, and Novo Nordisk.Author Contributions. L.K.B., A.D., D.G., A.O.,and H.W.R. participated in drafting the manu-script or revising it critically for important in-tellectual content. N.T., R.G., N.H., and E.J. madesubstantial contributions to conception and de-sign, and/or acquisition of data, and/or analysisand interpretation of data and participated indrafting themanuscript or revising it critically forimportant intellectual content. All authors gavefinal approval of the versionof themanuscript tobe submitted and any revised version. NovoNordisk was involved in the trial design andprotocol development, provided logistical sup-port, and obtained the data, which were evalu-ated jointly by the authors and the sponsor.L.K.B. is theguarantorofthisworkand,assuch,hadfull access to all the data in the study and takesresponsibility for the integrity of the data and theaccuracy of the data analysis.Prior Presentation. Parts of this study werepresentedorally at the 77thScientific Sessionsofthe ADA, San Diego, CA, 9–13 June 2017, and as aposter at the 53rd Annual Meeting of the Euro-pean Association for the Study of Diabetes, Lis-bon, Portugal, 11–15 September 2017.

References1. Harris SB, Kapor J, Lank CN,WillanAR, HoustonT. Clinical inertia in patients with T2DM requiringinsulin in family practice. Can FamPhysician 2010;56:e418–e4242. Dale J, Martin S, Gadsby R. Insulin initiation inprimary care for patients with type 2 diabetes:3-year follow-up study. Prim Care Diabetes2010;4:85–893. Inzucchi SE, Bergenstal RM, Buse JB, et al.;American Diabetes Association (ADA); EuropeanAssociation for the Study of Diabetes (EASD).Management of hyperglycemia in type 2diabetes:a patient-centered approach: position statementof the American Diabetes Association (ADA) and

care.diabetesjournals.org Billings and Associates 1015

the European Association for the Study of Diabe-tes (EASD). Diabetes Care 2012;35:1364–13794. Blak BT, Smith HT, HardsM, Curtis BH, Ivanyi T.Optimization of insulin therapy in patients withtype 2 diabetes mellitus: beyond basal insulin.Diabet Med 2012;29:e13–e205. Khunti K, Millar-Jones D. Clinical inertia to in-sulin initiation and intensification in the UK: a fo-cused literature review. Prim Care Diabetes 2017;11:3–126. PeyrotM, Barnett AH,Meneghini LF, Schumm-Draeger PM. Insulin adherence behaviours andbarriers in the multinational Global Attitudes ofPatients and Physicians in Insulin Therapy Study.Diabet Med 2012;29:682–6897. Nyenwe EA, Jerkins TW, Umpierrez GE,Kitabchi AE. Management of type 2 diabetes:evolving strategies for the treatment of patientswith type 2 diabetes. Metabolism 2011;60:1–238. Ahmann A, Rodbard HW, Rosenstock J, et al.;NN2211-3917 Study Group. Efficacy and safety ofliraglutide versus placebo added to basal insulinanalogues (withorwithoutmetformin) in patientswith type 2 diabetes: a randomized, placebo-controlled trial. Diabetes Obes Metab 2015;17:1056–10649. Rosenstock J, Fonseca VA, Gross JL, et al.; Har-mony 6 Study Group. Advancing basal insulin re-placement in type 2 diabetes inadequatelycontrolled with insulin glargine plus oral agents:a comparison of adding albiglutide, a weeklyGLP-1 receptor agonist, versus thrice-daily pran-dial insulin lispro. Diabetes Care 2014;37:2317–232510. RodbardHW,BodeBW,Harris SB, et al. Safetyandefficacyof insulindegludec/liraglutide (IDegLira)added to sulphonylurea alone or to sulphonylureaand metformin in insulin-naıve people with type 2diabetes: the DUAL IV trial. Diabet Med 2017;34:189–196

11. Gough SC, Bode B, Woo V, et al.; NN9068-3697 (DUAL-I) trial investigators. Efficacy andsafety of a fixed-ratio combination of insulin de-gludec and liraglutide (IDegLira) compared withits components given alone: results of a phase 3,open-label, randomised, 26-week, treat-to-targettrial in insulin-naive patients with type 2 diabetes.Lancet Diabetes Endocrinol 2014;2:885–89312. Harris SB, Kocsis G, Prager R, et al. Safety andefficacy of IDegLira titrated once weekly versustwice weekly in patients with type 2 diabetes un-controlled on oral antidiabetic drugs: DUAL VIrandomized clinical trial. Diabetes Obes Metab2017;19:858–86513. Linjawi S, Bode BW, Chaykin LB, et al. Theefficacy of IDegLira (insulin degludec/liraglutidecombination) in adults with type 2 diabetes inad-equately controlledwith a GLP-1 receptor agonistand oral therapy: DUAL III randomized clinicaltrial. Diabetes Ther 2017;8:101–11414. Buse JB, Vilsbøll T, Thurman J, et al.; NN9068-3912 (DUAL-II) Trial Investigators. Contribution ofliraglutide in thefixed-ratio combination of insulindegludec and liraglutide (IDegLira). Diabetes Care2014;37:2926–293315. Lingvay I, Perez Manghi F, Garcıa-HernandezP, et al.; DUAL V Investigators. Effect of insulinglargine up-titration vs insulin degludec/liraglutideon glycated hemoglobin levels in patients with un-controlled type 2 diabetes: theDUAL V randomizedclinical trial [published correction appears in JAMA2016;315:2125]. JAMA 2016;315:898–90716. International Conference on Harmonisation.ICH Harmonised Tripartite Guideline: Guidelinefor Good Clinical Practice E6 (R1), step 4 [articleonline], 1996. Available from https://www.ich.org/fileadmin/Public_Web_Site/ICH_Products/Guidelines/Efficacy/E6/E6_R1_Guideline.pdf. Ac-cessed 3 March 201717. WorldMedical Association.WorldMedical As-sociationDeclarationof Helsinki: ethical principles

for medical research involving human subjects.JAMA 2013;310:2191–219418. Seaquist ER, Anderson J, Childs B, et al. Hypo-glycemia and diabetes: a report of a workgroup ofthe American Diabetes Association and the Endo-crine Society. Diabetes Care 2013;36:1384–139519. MathieuC,RodbardHW,CariouB, et al.; BEGIN:VICTOZAADD-ON (NN1250-3948) study group. Acomparison of adding liraglutide versus a singledaily dose of insulin aspart to insulin degludec insubjects with type 2 diabetes (BEGIN: VICTOZAADD-ON). Diabetes Obes Metab 2014;16:636–64420. American Diabetes Association. Standards ofMedical Care in Diabetesd2017. Diabetes Care2017;40(Suppl. 1):S1–S13521. Moghissi ES. Treating patients with diabetesof long duration: GLP-1 receptor agonists and in-sulin in combination. J AmOsteopath Assoc 2014;114(Suppl. 2):S22–S2922. WyshamC, Bhargava A, Chaykin L, et al. Effectof insulin degludec vs insulin glargine U100 on hypo-glycemia inpatientswith type2diabetes: theSWITCH2 randomized clinical trial. JAMA 2017;318:45–5623. Marso SP, Daniels GH, Brown-Frandsen K,et al.; LEADER Steering Committee; LEADER TrialInvestigators. Liraglutide and cardiovascular out-comes in type 2 diabetes. N Engl J Med 2016;375:311–32224. Marso SP, McGuire DK, Zinman B, et al.;DEVOTE Study Group. Efficacy and safety of de-gludec versus glargine in type 2 diabetes. N Engl JMed 2017;377:723–73225. Haahr H, Heise T. A review of the pharmaco-logical properties of insulin degludec and theirclinical relevance. Clin Pharmacokinet 2014;53:787–80026. Claxton AJ, Cramer J, Pierce C. A systematicreview of the associations between dose regi-mens and medication compliance. Clin Ther2001;23:1296–1310

1016 IDegLira Versus Basal-Bolus Insulin: DUAL VII Diabetes Care Volume 41, May 2018