sign in sign up
<<
Home , Insulin (medication)

1702 Care Volume 39, October 2016

Chantal Mathieu,1 Bernard Zinman,2 Efficacy and Safety of Joanna Udden´ Hemmingsson,3 Vincent Woo,4 Peter Colman,5 Added to Treatment in Erik Christiansen,6 Martin Linder,6 and Bruce Bode,7 for the ADJUNCT ONE : The ADJUNCT Investigators* ONE Treat-To-Target Randomized Trial Diabetes Care 2016;39:1702–1710 | DOI: 10.2337/dc16-0691 CLIN CARE/EDUCATION/NUTRITION/PSYCHOSOCIAL

OBJECTIVE To investigate whether liraglutide added to treat-to-target insulin improves gly- cemic control and reduces insulin requirements and body weight in subjects with type 1 diabetes.

RESEARCH DESIGN AND METHODS A 52-week, double-blind, treat-to-target trial involving 1,398 adults randomized 1Gasthuisberg Hospital, University of Leuven, 3:1 to receive once-daily subcutaneous injections of liraglutide (1.8, 1.2, or 0.6 mg) Leuven, Belgium 2Mount Sinai Hospital, University of Toronto, or placebo added to insulin. Toronto, Ontario, Canada 3Capio St. Gorans¨ Hospital and the Karolinska RESULTS Institute, Stockholm, Sweden 4 HbA1c level was reduced 0.34–0.54% (3.7–5.9 mmol/mol) from a mean baseline Health Sciences Centre, Winnipeg, Manitoba, of 8.2% (66 mmol/mol), and significantly more for liraglutide 1.8 and 1.2 mg com- Canada 5Royal Melbourne Hospital, Parkville, Victoria, pared with placebo (estimated treatment differences [ETDs]: 1.8 mg liraglutide Australia 20.20% [95% CI 20.32; 20.07]; 1.2 mg liraglutide 20.15% [95% CI 20.27; 20.03]; 6Novo Nordisk A/S, Bagsværd, Denmark 0.6 mg liraglutide 20.09% [95% CI 20.21; 0.03]). Insulin doses were reduced by the 7Atlanta Diabetes Associates, Atlanta, GA addition of liraglutide 1.8 and 1.2 mg versus placebo (estimated treatment ratios: Corresponding author: Chantal Mathieu, chantal 1.8 mg liraglutide 0.92 [95% CI 0.88; 0.96]; 1.2 mg liraglutide 0.95 [95% CI 0.91; [email protected]. 0.99]; 0.6 mg liraglutide 1.00 [95% CI 0.96; 1.04]). Mean body weight was signifi- Received 30 March 2016 and accepted 16 July 2016. cantly reduced in all liraglutide groups compared with placebo ETDs (1.8 mg 2 2 2 2 2 reg. no. NCT01836523, clinicaltrials liraglutide 4.9 kg [95% CI 5.7; 4.2]; 1.2 mg liraglutide 3.6 kg [95% CI 4.3; .gov. 22.8]; 0.6 mg liraglutide 22.2 kg [95% CI 22.9; 21.5]). The rate of symptomatic This article contains Supplementary Data online increased in all liraglutide groups (estimated rate ratios: 1.8 mg at http://care.diabetesjournals.org/lookup/ liraglutide 1.31 [95% CI 1.07; 1.59]; 1.2 mg liraglutide 1.27 [95% CI 1.03; 1.55]; suppl/doi:10.2337/dc16-0691/-/DC1. 0.6 mg liraglutide 1.17 [95% CI 0.97; 1.43]), and with ketosis in- *A complete list of the ADJUNCT ONE Investiga- creased significantly for liraglutide 1.8 mg only (event rate ratio 2.22 [95% CI tors can be found in the Supplementary Data 1.13; 4.34]). online. © 2016 by the American Diabetes Association. CONCLUSIONS Readers may use this article as long as the work is properly cited, the use is educational and not Liraglutide added to insulin therapy reduced HbA1c levels, total insulin dose, and for profit, and the work is not altered. More infor- body weight in a population that was generally representative of subjects with mation is available at http://www.diabetesjournals type 1 diabetes, accompanied by increased rates of symptomatic hypoglycemia .org/content/license. and hyperglycemia with ketosis, thereby limiting clinical use in this group. See accompanying article, p. 1693. care.diabetesjournals.org Mathieu and Associates 1703

Glycemic control is still frequently subop- 0.2, or 0.3 mL) once daily by subcutane- subjects with established type 1 diabe- timal in people with type 1 diabetes (1). ous injection. Subjects started with a tes with inadequate glycemic control. Maintaining strict glycemic control by in- dose of 0.6 mg liraglutide/0.1 mL pla- The secondary objective was to deter- tensive insulin therapy, although effec- cebo, and increased by 0.6 mg/0.1 mL mine the efficacy of liraglutide added tive, is challenging and often results in every other week until reaching their to insulin in reducing episodes of hypo- hypoglycemic events and weight gain target dose. Once reached, the liraglutide glycemia compared with placebo. (2). Adjunct therapies to insulin, like met- dose was maintained, with no dose re- Additional efficacy measures included formin, are being used off label to stabi- duction permitted. Subjects continued fasting plasma (FPG), self- lize glucose control, limit insulin dose, and insulin treatment in addition to liraglu- determined nine-point self-monitored reduce weight gain. Novel combination tide or placebo throughout the trial. The plasma glucose profiles (using a self- therapies using agents commonly pre- total daily insulin dose was initially re- measured blood glucose [BG] device scribed for show promise duced by 25% on the day of randomiza- calibrated to report plasma glucose), when added to insulin therapy in type 1 tion and by a further 10% on subsequent 1,5-anhydroglucitol (1,5-AG), and qual- diabetes (3). Among these therapies is days of dose escalation for a minimum of ity of life (treatment-related impact liraglutide, a -like peptide 1 re- 24 h. Subjects’ insulin dose was subse- measures-diabetes [TRIM-D] and Short ceptor (GLP-1RA), which en- quently increased as necessary, which Form [SF]-36 scores). hances insulin secretion and suppresses was in line with the treat-to-target de- Key safety assessments included ad- elevated levels of glucagon in a glucose- sign. Pretrial insulin treatment was con- verse events (AEs), hypoglycemic and dependent manner (4). In patients with sidered to be background hyperglycemic episodes, pulse, systolic type 2 diabetes, liraglutide, when used and was continued throughout the trial. and diastolic blood pressure, the labo- as an adjunct to insulin, has provided clin- ratory safety variables amylase and li- ically meaningful reductions in glycated Trial Population pase, and predefined medical events of Key inclusion criteria were as follows: hemoglobin (HbA )(5). special interest (deaths, acute coronary 1c clinically diagnosed type 1 diabetes Recently, small-scale studies and ret- syndrome, cerebrovascular event, pancre- $12 months prior to the screening visit, rospective analyses have suggested a atitis, neoplasm, thyroid disease, symp- treatment with basal or continu- potential for liraglutide in the treatment tomatic hyperglycemic episodes with ous subcutaneous insulin infusion (CSII) of people with type 1 diabetes. The ad- ketosis [plasma glucose .16.7 mmol/L $6 months, stable insulin treatment for ministration of liraglutide as an adjunct (.300 mg/dL)], and plasma ketone the last 3 months, BMI $20 kg/m2,and to insulin has consistently shown a re- level .1.5 mmol/L), severe hypoglycemic age 18–75 years. A notably wide range of duction in daily insulin dose and weight episodes, and medication errors concern- HbA levels(7.0–10%[53–86mmol/mol]) loss compared with control subjects, 1c ing trial products. The majority of medical was specified in order to include subjects and neutral or positive results in terms events of special interest were evaluated clinically representative of people with of HbA levels and rates of hypoglyce- by an independent external adjudication 1c type 1 diabetes. Importantly, and in con- mia (6–12). This study aims to address committee (EAC). Symptomatic hypogly- trast to most studies, the current trial did whether adding liraglutide to insulin cemic episodes were defined by Novo Nor- not exclude people with a history of se- therapy in a treat-to-target approach disk as severe according to the American vere hypoglycemia or hypoglycemia un- can safely improve glycemic control Diabetes Association (ADA) and a plasma awareness, or people with a history of in a population that is generally repre- glucose value ,3.1 mmol/L (56 mg/dL) ketoacidosis. sentative of people with type 1 dia- with symptoms consistent with hypogly- Key exclusion criteria included the fol- betes when studied in a randomized, cemia (13). lowing: prior use of GLP-1RA or dipep- double-blind fashion over a 52-week tidyl peptidase-4 inhibitors; the use of period. Statistical Analyses medication that interfered with glyce- Thesamplesizewasdeterminedto mic control, including all hypoglycemic RESEARCH DESIGN AND METHODS establish noninferiority in HbA level agents or steroids; history of acute or 1c Trial Design change after 52 weeks of treatment be- chronic pancreatitis; severely decreased The trial was a 52-week randomized, tween liraglutide 1.8 mg and placebo renal function (estimated glomerular placebo-controlled, double-blind, parallel- with 90% power, assuming a noninferior- filtration rate ,30 mL/min/1.73 m2, group, treat-to-target, phase 3 trial per- ity margin of 0.3%, a two-sided t test at level .50 ng/L at screening, formed at 177 centers in 17 countries. a5%significance level, an SD of 1.1%, personal/family history of medullary A screening visit (Week 22) was fol- and a true difference of 0%. Furthermore, thyroid carcinoma or multiple endocrine lowed by a randomization visit 2 weeks a dropout rate of 20% was assumed, and neoplasia syndrome type 2, or severe later (Week 0). Eligible subjects were strat- it was also assumed that none of the with- neuropathy. ified by HbA1c level and BMI (HbA1c ,8.5% drawn subjects would contribute to the and $8.5% [,69 and $69 mmol/mol], Outcome Measures analysis. The required sample size was BMI #27 and .27 kg/m2), and were cen- The primary objective of this study was found to be 1,395 subjects, 349 per arm, trally randomized by (Clinical to investigate the efficacy of liraglutide allowing for the expected dropout rate. Supplies Coordination) to one of six parallel- added to treat-to-target insulin therapy Continuous data were analyzed using a treatment groups in a 3:1 ratio to receive on glycemic control, and the reduction mixed model for repeated measurements either liraglutide (0.6, 1.2, or 1.8 mg) or a in total daily insulin dose and body (MMRM) using an unstructured covariance corresponding volume of placebo (0.1, weight loss compared with placebo in matrix and with treatment, stratification, 1704 Liraglutide Added to Adjusted Insulin in T1D Diabetes Care Volume 39, October 2016

and country as fixed factors, and baseline smoker status, country, and C-peptide teaching and guidance during the trial. value as a covariate, all nested within a level (less than or greater than or equal The drop from baseline was largest and visit. Insulin dose, 1.5-AG, amylase, and to the lower limit of quantification [LLOQ; dose dependent in the liraglutide-treated lipase were logarithmically transformed 0.03 nmol/L]). subjects, but was partially lost by Week before analysis. The response consisted 52 (0.54%, 0.49%, 0.43% for liraglutide of all scheduled postbaseline measure- RESULTS 1.8, 1.2, and 0.6 mg and 0.34% for pla- ments obtained while the subjects were Subject Disposition cebo [5.9, 5.4, 4.8, and 3.8 mmol/mol, re- receiving treatment. A total of 1,789 adults with type 1 diabe- spectively]). This reduction in HbA1c level Binary data were analyzed by a logis- tes were screened, of whom 1,398 were was significantly larger for liraglutide tic regression model with treatment and randomized and 1,393 were exposed to 1.8 and 1.2 mg compared with placebo stratification as fixed factors, and base- trial product (Supplementary Fig. 1). (1.8 mg liraglutide estimated treatment line as a covariate. Missing data were difference [ETD]: 20.20 [95% CI 20.32; Baseline Characteristics 2 2 imputed from the MMRM used for the Subjects were broadly homogeneous 0.07], P= 0.0019; 1.2 mg liraglutide 0.15 2 2 analysis of HbA1c. Event data (hypogly- across treatment groups in terms of age, [95% CI 0.27; 0.03], P = 0.0164; 0.6 mg cemic/hyperglycemic episodes) were liraglutide 20.09 [95% CI 20.21; 0.03], P = gender, HbA1c level, and type 1 diabetes analyzed by a negative binomial regres- duration (Table 1 and Supplementary 0.1299). fi sion model with a log-link function and Table 1). Notably, the study population There were no signi cant differences in the logarithm of the exposure time as included ;37% of patients with HbA FPG and 1,5-AG levels for any liraglutide fi 1c offset, with treatment, strati cation, and level $8.5%, ;7% had a history of severe group compared with the placebo group fi country as xed factors, and the HbA1c hypoglycemia, and 6% had hypoglycemia (Supplementary Fig. 2 and Supplementary value at baseline as a covariate. Placebo unawareness. The majority of the study Table 2). The observed reduction in the groups were pooled in all analyses. population, ;83%, were C-peptide nega- mean of the nine-point self-monitored fi tive. The remaining 17% had C-peptide plasma glucose pro le from baseline to Subgroup Analyses levels below the normal range, which is Week 52 was somewhat lower for the Analyseswereundertakeninanattempt consistent with a diagnosis of type 1 di- liraglutide groups compared with the pla- to identify subgroups with an improved abetes, but above the LLOQ (0.03 nmol/L). cebo group (Supplementary Fig. 2 and treatment effect. Prespecified analyses Supplementary Table 2). used the baseline variables age, sex, dura- Efficacy The total insulin dose decreased from tion of type 1 diabetes, method of insulin Despite the treat-to-target design, a dif- baseline to Week 52 in the 1.8 mg administration, BMI, HbA1c level, hypogly- ferential effect on HbA1c levels was seen liraglutide (25%) and 1.2 mg liraglutide cemia unawareness status, severe hypo- for the different treatment groups (Fig. (22%)groups,butincreasedinthe glycemia within the last 12 months, renal 1A). All groups had their HbA1c level de- 0.6 mg liraglutide (4%) and placebo function, and geographic region. Post creased during the study, which is con- groups (4%) (Fig. 1B). The reduction in hoc analyses (specified after unblind- sistent with the treat-to-target protocol total insulin dose, primarily bolus insulin, ing) included baseline variables such as including insulin titration, and enhanced was significant for liraglutide 1.8 and

Table 1—Baseline characteristics Liraglutide Liraglutide Liraglutide 1.8 mg 1.2 mg 0.6 mg Placebo (N =346) (N = 346) (N = 350) (N =347) Age (years) 43.7 6 13.3 43.9 6 13.1 43.6 6 12.8 43.4 6 12.6 Female sex 181 (52.3) 179 (51.7) 186 (53.1) 180 (51.9) Duration of type 1 diabetes (years) 21.5 6 12.6 21.6 6 12.2 20.9 6 12.2 21.6 6 11.8 MDI total daily insulin dose (units), geometric mean (CV) 62.52 (49.54) 59.61 (49.82) 59.54 (42.66) 62.42 (46.09) CSII 115 (33.2) 100 (28.9) 71 (20.3) 96 (27.7) CSII total daily insulin dose (units), geometric mean (CV) 50.46 (46.31) 50.73 (43.87) 52.97 (43.74) 49.18 (59.94) Hypoglycemia awareness: yes 317 (91.6) 326 (94.2) 334 (95.4) 328 (94.5) Body weight (kg) 86.3 6 17.3 85.4 6 17.2 86.5 6 17.3 86.4 6 17.8 BMI (kg/m2)29.56 5.2 29.3 6 5.1 29.5 6 5.3 29.8 6 5.6

HbA1c (%) 8.14 6 0.740 8.16 6 0.779 8.18 6 0.738 8.15 6 0.728

HbA1c (mmol/mol) 65.5 6 8.1 65.7 6 8.5 65.9 6 8.1 65.5 6 8.0 (N =345) (N = 344) (N = 350) (N =345) Severe hypoglycemic episode within the last year? Yes 25 (7.2) 23 (6.7) 23 (6.6) 28 (8.1) (N =341) (N = 340) (N = 343) (N =340) FPG (mmol/L) 10.01 6 4.348 10.64 6 4.205 10.19 6 4.204 9.96 6 4.197 (N =343) (N = 345) (N = 343) (N =342) Fasting C-peptide ,LLOQ 285 (83.1) 289 (83.8) 269 (78.4) 291 (85.1) Values are reported as the mean 6 SD or N (%), unless otherwise indicated. CV, coefficient of variation; MDI, multiple dose injection. care.diabetesjournals.org Mathieu and Associates 1705

per kilogram of body weight returned to approximately baseline levels in all treat- ment groups after 52 weeks. Categorizing subjects by reduction in HbA1c level highlighted the fact that ap- proximately twice as many subjects re- ceiving liraglutide 1.8 mg had reductions in HbA1c level of .1% without severe hypoglycemia after 52 weeks compared with those receiving placebo. In addi- tion, nearly twice as many subjects re- ceiving liraglutide 1.8 and 1.2 mg had significant HbA1c reductions below the ADA target of 7.0% or HbA1c reduc- tions ,7% without severe hypoglycemia compared with those receiving placebo (Fig. 2). Treatment with liraglutide in addition to insulin significantly reduced body weight in a dose-dependent manner, with losses of 4.0, 2.7, and 1.3 kg for liraglutide 1.8, 1.2, and 0.6 mg, respectively. Subjects randomized to receive placebo gained 0.9 kg (ETD: 1.8 mg liraglutide 24.90 [95% CI 25.65; 24.16], P , 0.0001; 1.2 mg liraglutide 23.55 [95% CI 24.29; 22.81], P , 0.0001; 0.6 mg liraglutide 22.19 [95% CI 22.91; 21.47], P , 0.0001).

Safety Overall, the rates of all AEs increased dose dependently in the liraglutide groups compared with the placebo group (1.8 mg, 7.7 events per patient-year of exposure [PYE]; 1.2 mg, 6.0 events/PYE; 0.6 mg, 5.3 events/PYE; and placebo, 4.8 events/PYE) (Table 2). Gastrointestinal disorder AEs were reported more frequently in the liraglu- tide groups compared with the placebo group and increased dose dependently (1.8 mg, 2.7 events/PYE; 1.2 mg, 1.9 events/PYE; 0.6 mg, 1.3 events/PYE; pla- cebo, 0.76 events/PYE) with the most frequently reported complaint being (1.8 mg, 0.97 events/PYE; 1.2 mg, Figure 1—A: Change in HbA1c (%) by treatment week. On-treatment analysis. Group mean estimates (6SEM) are from an MMRM with treatment, stratification, and country as fixed 0.69 events/PYE; 0.6 mg, 0.47 events/PYE; factors and baseline value as a fixed covariate, all nested within visit. B: Change in total daily placebo, 0.18 events/PYE). insulin dose by treatment week. On-treatment analysis. Group mean estimates (6SEM) are from The proportion of subjects with AEs an MMRM on the log-transformed postbaseline responses with treatment, stratification, leading to premature discontinuation and country as fixed factors and baseline value as a fixed covariate, all nested within visit. Group mean estimates are adjusted according to the observed baseline distribution and back- of trial product increased dose depen- transformed to the original scale. Error bars are calculated on the log scale before being back- dently (14.7%, 12.6%, and 4.9%, respec- transformed to original scale. C: Change in fasting body weight (in kg) by treatment week. tively, for liraglutide 1.8, 1.2, and 0.6 mg) On-treatment analysis. Group mean estimates (6SEM) are from a MMRM with treatment, compared with placebo (3.4%), and the stratification, and country as fixed factors and baseline value as a fixed covariate, all nested majority were gastrointestinal AEs. Most within visit. BL, baseline; Lira, liraglutide. of the discontinuations occurred within the initial 8–12 weeks of the trial. The 1.2 mg compared with placebo (estimated liraglutide 0.95 [95% CI 0.91; 0.99], P = proportion of subjects reporting serious treatment ratio: 1.8 mg liraglutide 0.92 0.0148; 0.6 mg liraglutide 1.00 [95% CI AEs (SAEs) was comparable among treat- [95% CI 0.88; 0.96], P , 0.0001; 1.2 mg 0.96; 1.04], P = 0.9615). Units of insulin ment groups (Table 2). 1706 Liraglutide Added to Adjusted Insulin in T1D Diabetes Care Volume 39, October 2016

confirmed by the EAC (Supplementary Table 3).

Hypoglycemic Episodes More than 90% of subjects experienced a hypoglycemic episode (BG ,3.9 mmol/L [70 mg/dL], by ADA definition), and ;82% of subjects experienced symptom- atic hypoglycemic episodes (Novo Nor- disk definition [Table 2], which is used in this study unless otherwise stated). The rate of symptomatic hypoglyce- mic episodes was observed to be higher in liraglutide-treated subjects (1.8 mg, 16.5 events/PYE; 1.2 mg, 16.1 events/ PYE; 0.6 mg, 15.7 events/PYE; placebo, 12.3 events/PYE). The rate of symptom- atic hypoglycemic episodes was sig- nificantly higher for liraglutide 1.8 and 1.2 mg compared with placebo (estimated rate ratio [ERR]: 1.8 mg liraglutide 1.31 [95% CI 1.07; 1.59], P = 0.0081; 1.2 mg liraglutide 1.27 [95% CI 1.03; 1.55], P = 0.0219; 0.6 mg liraglutide 1.17 [95% CI 0.97; 1.43], P = 0.1079). In contrast, the number of hypoglycemic events assessed by the EAC and con- firmed as severe was observed to be lower in all liraglutide-treated groups compared with the placebo group (45, 31, and 40 events, respectively, in the liraglutide 1.8, 1.2, and 0.6 mg groups, compared with 57 events in the placebo group) (Table 2). However, there was no statistically significant difference in the number of EAC-confirmed severe hypo- glycemic episodes for any treatment comparison (ERR: 1.8 mg liraglutide 0.85 [95% CI 0.48; 1.49], P = 0.5693; 1.2 mg liraglutide 0.61 [95% CI 0.33; 1.12], P = 0.1088; 0.6 mg, 0.64 [95% CI 0.36; 1.13], P = 0.1250). Novo Nordisk–defined asymptomatic hypoglycemic episodes are those episodes with plasma glucose levels ,3.1 mmol/L (56 mg/dL) and no symptoms of hypogly- cemia (Table 2).

Figure 2—HbA1c responders. A: Responders reducing HbA1c level to ,7%. B: Responders re- Considering the total number of severe , . ducing HbA1c to 7% without severe hypoglycemia. C: Responders reducing HbA1c level by 1% or BG-confirmed hypoglycemic episodes, without severe hypoglycemia. On-treatment; proportion of responders with missing data ;21.7%, 19.6%, and 18.2%, respectively, imputed by MMRM. EOR, estimated odds ratio. in the liraglutide 1.8, 1.2, and 0.6 mg groups, and 27.1% in the placebo group Mean (geometric) lipase levels were levels were initially ;50 units/L and were asymptomatic. The distribution of approximately 24 units/L at baseline and increased in all liraglutide groups com- asymptomatic hypoglycemic episodes increased in all liraglutide groups com- paredwithplacebo,butwithnoclear did not stem from unbalanced randomi- pared with placebo groups, but with no dose dependency (10%, 10%, and zation of trial subjects with a history of clear dose dependency (35%, 36%, and 10%, respectively, for liraglutide 1.8, hypoglycemia unawareness or severe hy- 32%, respectively, for liraglutide 1.8, 1.2, and 0.6 mg, respectively, and 3% poglycemia (Table 1). 1.2, and 0.6 mg, and 1% for placebo). forplacebo).Onecaseofpancreatitis Nocturnal hypoglycemia (onset be- Similarly, mean (geometric) amylase in the liraglutide 0.6 mg group was tween 0:01 and 5:59 A.M., both inclusive) care.diabetesjournals.org Mathieu and Associates 1707

Table 2—Safety: AEs and SAEs Liraglutide 1.8 mg Liraglutide 1.2 mg Liraglutide 0.6 mg Placebo N % RN% RN% RN% R Participants, N 347 348 350 348 Exposure years 278.9 286.3 315.5 296.6 AEs 313 90.2 7.7 302 86.8 6.0 298 85.1 5.3 275 79.0 4.8 SAEs 29 8.4 0.14 36 10.3 0.17 35 10.0 0.15 38 10.9 0.18 Leading to premature treatment discontinuation 51 14.7 0.30 44 12.6 0.29 17 4.9 0.08 12 3.4 0.06 Hypoglycemic episodes All 329 94.8 50.2 322 92.5 49.4 334 95.4 45.4 321 92.2 42.7 Symptomatic* 290 83.6 16.5 285 81.9 16.1 277 79.1 15.7 276 79.3 12.3 Severe or BG confirmed** 306 88.2 21.1 300 86.2 20.0 294 84.0 19.2 290 83.3 16.9 ADA classification 329 94.8 50.1 322 92.5 49.3 333 95.1 45.4 321 92.2 42.6 Severe 28 8.1 0.17 22 6.3 0.11 32 9.1 0.13 37 10.6 0.19 Documented symptomatic 309 89.0 34.4 303 87.1 34.0 311 88.9 32.6 302 86.8 27.7 Asymptomatic 295 85.0 15.2 272 78.2 14.8 276 78.9 12.2 278 79.9 14.0 Probable symptomatic 40 11.5 0.30 31 8.9 0.28 39 11.1 0.34 40 11.5 0.39 Pseudo-hypoglycemia 18 5.2 0.08 21 6.0 0.12 20 5.7 0.11 22 6.3 0.35 Hyperglycemic episodes All 307 88.5 33.5 293 84.2 30.9 309 88.3 29.5 312 89.7 34.7 Episodes with ketosis*** 39 11.2 0.28 26 7.5 0.15 22 6.3 0.17 24 6.9 0.12 Gastrointestinal disorders All 237 68.3 2.7 212 60.9 1.9 175 50.0 1.3 116 33.3 0.76 Nausea 172 49.6 0.97 142 40.8 0.69 112 32.0 0.47 42 12.1 0.18

EAC-confirmed SAEs N Events N Events N Events N Events

Neoplasms**** 9 9 3 3 6 6 4 4 Benign 7 7 3 3 6 6 3 3 Malignant 2 2 0 0 0 0 1 1 Pancreatitis 0 0 0 0 1 1 0 0 Cardiovascular disease ACS events 0 0 1 1 1 1 1 1 Stroke 0 0 0 0 0 0 0 0 Transient ischemic attack 0 0 1 1 0 0 0 0 EAC-confirmed severe hypoglycemia 28 45 22 31 32 40 38 57 EAC-confirmed DKA 3 3 1 1 3 4 0 0 ACS, acute coronary syndrome; Events, number of events; N, number of participants experiencing at least one event; %, percentage of participants experiencing at least one event; R, event rate per year of exposure. *Symptomatic hypoglycemia was defined as hypoglycemic episodes that are either severe according to ADA classification of hypoglycemia (13) or accompanied by a plasma glucose value of ,3.1 mmol/L (56 mg/dL), with symptoms consistent with hypoglycemia. **Severe or BG confirmed: hypoglycemic episodes that are either severe according to ADA classification of hypoglycemia (13) or an episode accompanied by a plasma BG value of ,3.1 mmol/L (56 mg/dL); Novo Nordisk asymptomatic hypoglycemia is the difference between “severe or BG-confirmed” hypoglycemia and “symptomatic” hypoglycemia. ***Plasma ketone level .1.5 mmol/L. ****EAC-confirmed malignant neoplasms were all skin cancers (two malignant melanomas; one basal cell carcinoma) using ADA classification of hypoglycemia (13). showed broadly similar patterns to di- and 6.3%, respectively, for liraglutide were adjudicated as being diabetic keto- urnal hypoglycemia. The observed rates 1.8, 1.2, and 0.6 mg, and 6.9% in the acidosis (DKA); 3, 1, and 4 events, respec- were higher for liraglutide compared placebo group. Higher rates were ob- tively, in the liraglutide 1.8, 1.2, and with placebo, but without any apparent served for the liraglutide groups com- 0.6 mg groups, and 0 in the placebo group dose dependency. paredwiththeplacebogroup(1.8 (Table 2). In all but one case (observed in a mg liraglutide, 0.28 events/PYE; 1.2 subject who had been susceptible to nau- Hyperglycemic Episodes mg liraglutide, 0.15 events/PYE; 0.6 mg sea and vomiting during the treatment Overall, ;90% of subjects reported hy- liraglutide, 0.17 events/PYE; and pla- escalation period), there was a clinically perglycemic episodes (plasma glucose cebo, 0.12 events/PYE). The treatment relevant condition, such as concomitant .16.7 mmol/L [.300 mg/dL]) (Table 2). difference was statistically significant for infection, pump malfunction, and post- The observed rates were 33.5, 30.9, liraglutide 1.8 mg compared with pla- operative insufficient insulin treatment, and 29.5 events/PYE for 1.8, 1.2, and cebo (ERR 2.22 [95% CI 1.13; 4.34], P = that triggered the episode or a medical 0.6 mg liraglutide, respectively, and 0.0205). All symptomatic hyperglycemic history of multiple DKA events. In addition, 34.7 events/PYE for placebo. The pro- episodes with ketosis (plasma ketone as shown in Supplementary Table 4, DKA portion of subjects with hyperglycemic level .1.5 mmol/L) were evaluated by was seen across a broad range of HbA1c val- episodes with ketosis (plasma ketone the EAC to confirm that they were keto- ues at the time of the DKA event, and all level .1.5 mmol/L) was 11.2%, 7.5%, acidosis events. A total of eight events DKA cases occurred in C-peptide–negative 1708 Liraglutide Added to Adjusted Insulin in T1D Diabetes Care Volume 39, October 2016

subjects (i.e., C-peptide level less than (1.8 mg liraglutide: ETD 4.58 [95% CI adjunctive therapy to insulin. The im- LLOQ). 2.82; 6.34], P . 0.0001; 1.2 mg liraglutide: proved glycemic control obtained with 4.85, ETD 3.09 [95% CI 1.35; 4.83], P = intensive insulin therapy must always Subgroup Analyses 0.0005; 0.6 mg liraglutide 4.71, ETD 2.95 be balanced against associated weight Subgroup analyses identified no notably [95% CI 1.27; 4.63], P = 0.0006). There gain, as overweight and obesity are in different findings, with the exception of were no significant treatment differences turn associated with C-peptide positivity (Supplementary for any of the liraglutide groups compared and cardiovascular risk factors in a pop- Tables 5–10 and Supplementary Fig. 3). with the placebo group in the SF-36 overall ulation with type 1 diabetes (16). There- In the current trial, 239 of 1,373 subjects physical and mental scores. fore, the weight loss associated with with baseline C-peptide measurements liraglutide is considered to be a benefit were C-peptide positive (value above CONCLUSIONS for patients with type 1 diabetes, but or equal to the LLOQ [0.03 nmol/L]) ADJUNCT ONE is the first 52-week, may be an issue in those where weight (Supplementary Table 1). Baseline double-blind, randomized controlled clin- is subnormal. The current trial shows characteristics of the C-peptide–positive ical trial to investigate the safety and that these beneficial findings come at subjects showed a shorter duration efficacy of adding liraglutide to a contin- the expense of a higher rate of symptom- of type 1 diabetes and higher base- uously adjusted treat-to-target insulin atic hypoglycemia and more episodes of line HbA values (Supplementary 1c dose in people with type 1 diabetes with hyperglycemia with ketosis compared Table 1). For both the 1.8 and 1.2 mg inadequate glycemic control at baseline. with insulin treatment alone. This finding liraglutide dose groups, reductions in The current trial demonstrated that has not been consistently reported be- HbA level from baseline were greater 1c in a population that is generally repre- fore in people with type 1 diabetes when for C-peptide–positive subjects (0.83% sentative of people with type 1 diabetes, liraglutide was added to therapy, possi- and 0.71%, respectively) than for sub- liraglutide added to insulin treatment bly reflecting the glycemic control and jects who were C-peptide negative results in a modest dose-dependent lower insulin dose in liraglutide-treated (0.47% and 0.44%, respectively). For reductioninHbA level despite a re- subjects, as well as difficulties in accu- the lowest liraglutide dose group 1c duced total daily insulin requirement rately adjusting bolus insulin and food (0.6 mg), HbA reductions were similar 1c for the two highest doses of liraglutide. intake (6,9). A more flexible insulin titra- for C-peptide–positive and C-peptide– All doses of liraglutide reduced body tion may have mitigated the risk of hy- negative patients (0.42% and 0.44%, weight, which is consistent with obser- poglycemia, though this would have to respectively). C-peptide–positive sub- vations in people with type 2 diabe- be balanced with the constraints of the jects also experienced a lower rate of tes (14). Though some of the efficacy treat-to-target design required for wider symptomatic hypoglycemia and very markers (HbA and 1,5-AG levels) risk-benefit analysis. Notably, rates of se- few episodes (two, one, zero, and one 1c driftedmoretowardbaselineinthe vere hypoglycemia were not higher in events, respectively, in the liraglutide liraglutide groups compared with the liraglutide-treated subjects compared 1.8, 1.2, and 0.6 mg, and placebo groups) placebo group, body weight remained with placebo, which in part could be of hyperglycemia with ketosis (plasma more constant, and insulin dose did due to the fact that liraglutide did ketone level .1.5 mmol/L) compared not increase despite the deterioration not compromise the glucagon counter- with C-peptide–negative participants. in HbA and 1,5-AG levels. The exact regulatory response to hypoglycemia The interaction between treatment ef- 1c mechanisms underlying these observa- (45 mg/dL) in type 1 diabetes patients fect and C-peptide positivity was not sta- tions remain unclear. (8). Post hoc analyses, though not conclu- tistically significant for either HbA level 1c Previous investigations of the effi- sive, also indicated a trend toward a lower or symptomatic hypoglycemia for any cacy of liraglutide added to insulin in rate of symptomatic hypoglycemia, al- dose group (Supplementary Tables 6 and people with type 1 diabetes have been most no episodes with hyperglycemia 8, respectively). inconclusive; some trials (8,15) found no with ketosis, and a greater reduction in For all liraglutide dose groups, esti- treatment difference compared with HbA level in C-peptide–positive subjects mated mean reductions in HbA level 1c 1c the placebo group, whereas others re- compared with C-peptide–negative sub- were greater for patients with a baseline ported reductions in BG and HbA , jects. The current study, importantly, did HbA level ,8.5% of those with a base- 1c 1c which to some extent could be due to not exclude people with a history of severe line HbA level of $8.5% (1.8 mg 0.57% 1c liraglutide reducing postprandial gluca- hypoglycemia and hypoglycemia unaware- vs. 0.46%; 1.2 mg 0.54% vs. 0.40%; gon secretion and glucose excursions. In ness, which distinguishes it from typical tri- 0.6 mg 0.44% vs. 0.41%). However, the the current trial, the placebo-corrected als in type 1 diabetes. This subgroup did not interaction between treatment effect reduction in HbA level reported with respond differently from the overall popu- and baseline HbA level was not statis- 1c 1c the highest dose of liraglutide (1.8 mg) lation, pointing to a hypoglycemia risk tically significant for any dose group was modest (0.2% [2.4 mmol/mol]). when combining liraglutide with insulin (Supplementary Table 10). A reduction in placebo-adjusted body in a broad population of people with Quality of Life weight of up to 4.9 kg was achieved for long-standing type 1 diabetes. Patient-reported outcomes were gener- subjects randomized to receive liraglu- The increases in mean amylase and ally positive; the TRIM-D total score was tide, which was comparable to that re- lipase reported in the liraglutide treat- significantly higher for all liraglutide cently reported (12) with liraglutide in ment groups did not have any clinical groups, particularly for diabetes manage- overweight patients with type 1 diabe- implications and are similar to obser- ment,comparedwiththeplacebogroup tes (26.8 kg). This is noteworthy for an vationswiththisdrugclassintype2 care.diabetesjournals.org Mathieu and Associates 1709

diabetes (17–19). Because pancreatic enrollment, compared with ,1% of sub- and AstraZeneca. B.Z. has served as a consultant volume and enzyme secretion are often jects treated with liraglutide in the cur- for Abbott, AstraZeneca, Boehringer Ingelheim, Eli fi decreased in subjects with type 1 diabe- rent trial for the same period (25). Lilly, Merck, Novo Nordisk, and Sano and has received grant support from AstraZeneca, tes, elevations from baseline for an in- Paradoxically, a major strength of the Boehringer Ingelheim, and Novo Nordisk. J.U.H. dividual subject within the defined ADJUNCT ONE trial, namely, the broad has served on advisory panels for Novo Nordisk normal range need to be interpreted in subject recruitment, may also be a lim- and Sanofi Diabetes; has received lecture fees this context (20–22). itation. Widening the subject popula- from Novo Nordisk, Sanofi Diabetes, and ; fl hasserved as a consultant for Novo Nordisk; Although hyperglycemic events oc- tion may better re ect the clinical and has received grant support from InfuCare- curred in an approximately equal pro- spectrum of diabetes but may also Dexcom. V.W. has received honoraria for speaking portion of subjects across all treatment mask the subgroups in which liraglutide and for participation on advisory boards and in groups and the rate of events was lower adjunct to insulin would produce the clinical trials from Novo Nordisk, Eli Lilly, Sanofi, in the liraglutide-treated groups than most benefit. This is most clearly illus- Boehringer Ingelheim, and AstraZeneca. P.C. has served on advisory panels for Novo Nordisk, for the placebo group, the higher rate trated by the post hoc analyses, which Eli Lilly, and Sanofi Diabetes and has received of hyperglycemia with ketosis may suggest better treatment effects in sub- lecture fees from Novo Nordisk, Eli Lilly, and have two explanations. First, it could jects with residual C-peptide levels at Sanofi Diabetes. E.C. is an employee and share- be linked, in part, to the nausea associ- the start of the trial. This is both consis- holder of Novo Nordisk A/S. M.L. is an employee ated with liraglutide treatment; 41% of tent with the mechanism of action of of Novo Nordisk A/S. B.B. has served as consul- tant for and owns stocks in Aseko; has served subjects treated with liraglutide and liraglutide and in line with a recent re- on a speakers’ bureau for Merck; and has served 12.1% randomized to receive placebo port (26) in people with type 2 diabetes on a speakers’ bureau and as a consultant for reported nausea, the most frequent AE. highlighting a better clinical response to Valeritas. His employer/institution has received It is possible that the dose-dependent in- GLP-1RAs in those subjects with residual research support from Abbott, AstraZeneca, crease in nausea may alert subjects to C-peptide levels. Dexcom, Lexicon, Roche, and Senseonics; his in- stitution has received speakers’ bureau and con- check ketones more frequently, resulting In conclusion, the current trial dem- sultant fees and research support from Janssen, in a parallel dose-driven reporting of hy- onstrated that adding liraglutide to in- Medtronic, Novo Nordisk, and Sanofi;andhisin- perglycemia with ketosis. Second, and sulin therapy for a population generally stitution has received speakers’ bureau and re- more probably, the increased incidence representative of people with type 1 di- search support from Boehringer Ingelheim/Lilly fl fl and GlaxoSmithKline. No other potential con icts of hyperglycemia with ketosis re ects abetes resulted in better glycemic con- of interest relevant to this article were reported. the effect of lowering the insulin dose in trol, less insulin, greater body weight Author Contributions. C.M., E.C., and B.B. the liraglutide groups, which in some pa- loss, and a greater proportion of subjects contributed to the study design and contact and data collection and analysis. B.Z., J.U.H., tients may then be below that required to achieving the ADA target of HbA1c ,7% suppress peripheral lipolysis. Lowering (53 mmol/mol). This dose-dependent ef- V.W., and P.C. contributed to study contact and data collection and analysis. M.L. contributed to the insulin dose below a critical level in fect was accompanied by a higher rate of the data analysis. All authors were involved in patients with type 1 diabetes will result in symptomatic hypoglycemia and hypergly- the writing of the manuscript and approved the ketone production, especially when car- cemic episodes with ketosis, limiting the final version of the manuscript. C.M. is the bohydrate levels are reduced as a result clinical use of GLP1-RAs in people with guarantor of this work and, as such, had full of decreased carbohydrate intake. Similar type 1 diabetes. access to all the data in the study and takes responsibility for the integrity of the data and observations have been made for so- the accuracy of the data analysis. dium/glucose cotransporter 2 inhibitors Prior Presentation. Parts of this study were added to insulin therapy, where carbohy- presented in abstract form at the American Acknowledgments. The authors thank all sub- Association of Clinical Endocrinologists 25th drate levels may be reduced as a result of jects and ADJUNCT ONE investigators involved Annual Meeting and Clinical Congress, Orlando, urinary excretion (23,24). in the study. The authors also thank Salvatore FL, 25–29 June 2016. Very few cases of DKA were ob- Calanna and Thomas Jon Jensen of Novo Nordisk served in this trial, all of which were in A/S for their review and input to the manuscript and Daniel Hayward of Novo Nordisk A/S for References the liraglutide-treated groups, though providing medical writing assistance during the 1. Miller KM, Foster NC, Beck RW, et al.; T1D were not dose dependent. Information preparation of this manuscript. Submission sup- Exchange Clinic Network. Current state of type 1 on a history of DKA was not systemati- port was provided by Watermeadow Medical, an diabetes treatment in the U.S.: updated data cally collected from subjects before Ashfield Company, part of UDG Healthcare plc, from the T1D Exchange clinic registry. Diabetes – study enrollment, and no attempt to ex- funded by Novo Nordisk. Care 2015;38:971 978 Duality of Interest. This study was sponsored 2. The Diabetes Control and Complications Trial clude subjects prone to DKA was made by Novo Nordisk A/S (NN9211-3919). The spon- Research Group. Adverse events and their asso- (these steps were taken in agreement sor (Novo Nordisk) was involved in the study ciation with treatment regimens in the Diabetes with the regulatory authorities in order design and protocol development, reviewed Control and Complications Trial. Diabetes Care to recruit a clinically informative trial the manuscript for scientific accuracy, and pro- 1995;18:1415–1427 population). The incidence of DKA in vided statistical support. C.M. has served on 3. Bode BW, Garg SK. The emerging role ad- advisory panels for Novo Nordisk, Sanofi Aventis, junctive noninsulin antihyperglycemic therapy the current trial should also be consid- Merck Sharp & Dohme Ltd., Eli Lilly, Novartis, in the management of type 1 diabetes. Endocr ered in the context of wider reported Bristol-Myers Squibb, AstraZeneca LP, Pfizer, Pract 2016;22:220–230 baseline rates of DKA in the type 1 di- Jansen Pharmaceuticals, and Hanmi; has re- 4.HolstJJ,DeaconCF,VilsbøllT,KrarupT, abetes population; almost 5% of pa- ceived research support from Novo Nordisk, Madsbad S. Glucagon-like peptide-1, glucose ho- Sanofi Aventis, Merck Sharp & Dohme Ltd., meostasis and diabetes. Trends Mol Med 2008; tients with type 1 diabetes in the T1D Eli Lilly, and Novartis; and has served on 14:161–168 Exchange Clinical Registry reported a speakers’ bureaus for Novo Nordisk, 5. Holst JJ, Vilsbøll T. Combining GLP-1 receptor DKA event in the 12 months before their Aventis, Merck Sharp & Dohme, Eli Lilly, Novartis, with insulin: therapeutic rationales and 1710 Liraglutide Added to Adjusted Insulin in T1D Diabetes Care Volume 39, October 2016

clinical findings. Diabetes Obes Metab 2013;15: double-blind, placebo-controlled trial. Lancet 19. Kapitza C, Nosek L, Jensen L, Hartvig H, 3–14 Diabetes Endocrinol 2016;4:221–232 Jensen CB, Flint A. , a once-weekly 6. Frandsen CS, Dejgaard TF, Holst JJ, Andersen 13. Seaquist ER, Anderson J, Childs B, et al. Hy- human GLP-1 analog, does not reduce the bio- HU, Thorsteinsson B, Madsbad S. Twelve-week poglycemia and diabetes: a report of a work- availability of the combined oral contraceptive, treatment with liraglutide as add-on to insulin in group of the American Diabetes Association and ethinylestradiol/levonorgestrel. J Clin Pharma- normal-weight patients with poorly controlled the Endocrine Society. Diabetes Care 2013; col 2015;55:497–504 type 1 diabetes: a randomized, placebo-controlled, 36:1384–1395 20. Foo Y, Rosalki SB, Ramdial L, Mikhailidis D, double-blind parallel study. Diabetes Care 2015; 14. Perry CM. Liraglutide: a review of its use in Dandona P. Serum isoamylase activities in di- 38:2250–2257 the management of type 2 diabetes mellitus. abetes mellitus. J Clin Pathol 1980;33:1102– 7. Crisci I, Aragona M, Politi KS, Daniele G, Del Drugs 2011;71:2347–2373 1105 Prato S. GLP-1 receptor agonists in type 1 dia- 15. Kuhadiya ND, Dhindsa S, Ghanim H, et al. 21. Fonseca V, Berger LA, Beckett AG, Dandona betes: a proof-of-concept approach. Acta Diabe- Addition of liraglutide to insulin in patients P. Size of in diabetes mellitus: a study tol 2015;52:1129–1133 with type 1 diabetes: a randomized placebo- based on ultrasound. Br Med J (Clin Res Ed) 8. Pieber TR, Deller S, Korsatko S, et al. Counter- controlled clinical trial of 12 weeks. Diabetes 1985;291:1240–1241 regulatory responses to hypoglycae- Care 2016;39:1027–1035 22. Junglee D, De Albarran R, Katrak A, mia in subjects with type 1 diabetes following 16. Purnell JQ, Zinman B, Brunzell JD; DCCT/ Freedman DB, Beckett AG, Dandona P. Low pan- 4 weeks of treatment with liraglutide adjunct to EDIC Research Group. The effect of excess creatic lipase in insulin-dependent diabetics. insulin: a randomized, placebo-controlled, double- weightgainwithintensive diabetes mellitus J Clin Pathol 1983;36:200–202 blind, crossover trial. Diabetes Obes Metab 2015; treatment on cardiovascular disease risk fac- 23. Rosenstock J, Ferrannini E. Euglycemic di- 17:742–750 tors and atherosclerosis in type 1 diabetes mel- abetic ketoacidosis: a predictable, detectable, 9. Kuhadiya ND, Malik R, Bellini NJ, et al. Liraglutide litus: results from the Diabetes Control and and preventable safety concern with SGLT2 in- as additional treatment to insulin in obese patients Complications Trial/Epidemiology of Diabetes In- hibitors. Diabetes Care 2015;38:1638–1642 with type 1 diabetes mellitus. Endocr Pract 2013;19: terventions and Complications Study (DCCT/EDIC) 24. Kitabchi AE, Umpierrez GE, Miles JM, Fisher 963–967 study. Circulation 2013;127:180–187 JN. Hyperglycemic crises in adult patients with 10. Kielgast U, Krarup T, Holst JJ, Madsbad S. Four 17. Emoto M, Terauchi Y, Ozeki A, Oura T, diabetes. Diabetes Care 2009;32:1335–1343 weeks of treatment with liraglutide reduces insu- Takeuchi M, Imaoka T. A 1-year safety study of 25. Weinstock RS, Xing D, Maahs DM, et al.; lin dose without loss of glycemic control in type 1 in Japanese patients with type 2 di- T1D Exchange Clinic Network. Severe hypogly- diabetic patients with and without residual beta- abetes on a single oral hypoglycemic agent: cemia and in adults with cell function. Diabetes Care 2011;34:1463–1468 an open-label, nonrandomized, phase 3 trial. type 1 diabetes: results from the T1D Exchange 11. Varanasi A, Bellini N, Rawal D, et al. Liraglutide Endocr J 2015;62:1101–1114 clinic registry. J Clin Endocrinol Metab 2013;98: as additional treatment for type 1 diabetes. Eur J 18. Sun F, Wu S, Guo S, et al. Impact of GLP-1 3411–3419 Endocrinol 2011;165:77–84 receptor agonists on blood pressure, heart rate 26. Jones AG, McDonald TJ, Shields BM, et al.; 12. Dejgaard TF, Frandsen CS, Hansen TS, et al. and hypertension among patients with type 2 PRIBA Study Group. Markers of b-cell failure Efficacy and safety of liraglutide for overweight diabetes: a systematic review and network predict poor glycemic response to GLP-1 recep- adult patients with type 1 diabetes and insuffi- meta-analysis. Diabetes Res Clin Pract 2015; tor agonist therapy in type 2 diabetes. Diabetes cient glycaemic control (Lira-1): a randomised, 110:26–37 Care 2016;39:250–257