Efficacy and safety of dulaglutide in patients with type 2 diabetes: a meta-analysis and systematic review

A meta-analysis was conducted to assess the clinical efficacy and safety of dulaglutide in patients with type 2 diabetes mellitus (T2DM). Medline, Embase, Cochrane Library and www. clinicaltrials. gov (up to February 15th, 2015) were searched. Randomized controlled trials comparing dulaglutide to other drugs for T2DM were collected. Twelve RCTs were included, and the overall bias was low. As the monotherapy, compared with control (placebo, metformin and liraglutide), dulaglutide resulted in a significant reduction in HbA1c (WMD, −0.68%; 95% CI, −0.95 to −0.40), FPG (WMD, −0.90 mmol/L; 95% CI, −1.28 to −0.52), a similar risk of hypoglycemia (7.8% vs. 10.6%), less body weight loss (WMD, 0.51 kg; 95% CI, 0.27 to 0.75). As an add-on intervention with oral antihyperglycemic medication (OAM) and insulin, compared with control (placebo, sitagliptin, exenatide, liraglutide and glargine), dulaglutide lowered HbA1c (WMD, −0.51%; 95% CI, −0.68 to −0.35) and body weight significantly (WMD, −1.30 kg, 95% CI, −1.85 to −1.02) notably, and elicited a similar reduction in FPG (WMD, −0.19 mmol/L; 95% CI, −1.20 to 0.82), an similar incidence of hypoglycemia (24.5% vs. 24.5%). This meta-analysis revealed the use of dulaglutide as a monotherapy or an add-on to OAM and lispro appeared to be effective and safe for adults with T2DM.

Methodological quality. The risk of bias was estimated according to the Cochrane Collaboration's risk of bias tool; the data were shown in Fig. 2(A,B). Random sequence generation was clear in eight RCTs 8-15 but can not be obtained in the other four [16][17][18][19] . Similarly, allocation concealment was described explicitly in four RCTs but not in the other six 8,[15][16][17][18][19] . The overall risk of bias in the included studies was low. Eli Lilly and Company funded all the trials. Intent-to-treat analysis was used to evaluate all randomized subjects who accepted at least one dose of the study treatment.
Dungan et al. 14 compared dulaglutide (1.5 mg) once weekly with liraglutide (1.8 mg) once daily using metformin as the background treatment. The mean reduction in HbA1c was − 1.42% and − 1.36%, respectively; there was no significant difference between groups (WMD, − 0.06%; 95% CI, − 0.20 to 0.08). The proportion of patients that achieved HbA1c target < 7.0% was 68% and 55%, and HbA1c target ≤ 6.5% was 55% and 51% in the dulaglutide and liraglutide groups respectively. There were no significant differences between the two groups.
When used as an add-on therapy to OAM and lispro, dulaglutide increased the risk of nausea (  Compared with control (placebo, sitagliptin, exenatide, liraglutide and glargine), as an add-on therapy to OAM and lispro, 0.75 mg QW, 1.0 mg QW, and 1.5 mg QW dulaglutide revealed reduction in HbA1c of 0.33%, 1.08%, and 0.43% (Fig. 4); except the changes of bodyweight was − 1.16 kg, − 1.22 kg, and − 1.45 kg (see Supplementary  Fig. S2  Pancreatitis. Seven cases of pancreatitis were reported in the included studies. Of these, two were considered to be related to the study drug 8 , two were in a sitagliptin arm 11 , one was in a placebo arm 9 , one in 1.5 mg dulaglutide arm, one in AWARD-2 16 , and one was in the placebo group during the sitagliptin period 11 . One patient who had no signs and symptoms of pancreatitis before the study was diagnosed with chronic pancreatitis ~7 months after treatment in the 1.5 mg dulaglutide arm 13 . No cases of adjudicated pancreatitis were reported during one study 12 , and no information was available from two trials 10 .

Discussion
This study was a meta-analysis to assess the efficacy and safety of the GLP-1 receptor agonist dulaglutide, which was prescribed to subjects with type 2 diabetes with or without other hypoglycemic drugs.
Bariatric surgery has been reported to ameliorate type 2 diabete with serious obesity. A retrospective trial evaluated the clinical efficacy of bariatric surgery vs liraglutide in patients with severely obese type 2 diabetic patients Scientific RepoRts | 6:18904 | DOI: 10.1038/srep18904 and found bariatric surgery lowered body weight and improved metabolic control than liraglutide significantly 24 . In the included studies, no evidences were found to compare dulaglutide with bariatric surgery directly. Compared with liraglutide, dulaglutide once weekly leads to the similar reduction of HbA1c, we speculate that bariatric surgery reduce HbA1c obviously than dulaglutide but it should be verified with clinical trials.
Hypoglycemia is a challenge and obstacle for the treatment of diabetes. When administered as a monotherapy, the risk of hypoglycemia was similar between the dulaglutide and control groups (placebo, metformin and liraglutide) (7.8% vs. 10.6%, respectively). When administered as an add-on therapy with OAM and lispro dulaglutide also did not increasethe risk of hypoglycemia (24.5% vs. 24.5%) compared with control (placebo, sitagliptin, exenatide, and liraglutide).
T2DM increases morbidity and mortality, mainly due to cardiovascular and cerebrovascular disease. The specific risk factors include weight gain, hypertension, and hyperlipidaemia. Therefore, antidiabetic drugs should be beneficial for cardiovascular and cerebrovascular diseases and also lower blood glucose. Obesity is associated with an increased risk for the development of hypertension, diabetes, dyslipidemia, and cardiovascular disease 25 . Weight gain is observed commonly in patients that use antidiabetic drugs, such as sulphonylureas, thiazolidinediones, glinides, and insulin; therefore, the correct drug should be selected carefully. GLP-1 receptor agonists (liraglutide, albiglutide, exenatide long-acting release, or exenatide [26][27][28][29][30], have favorable weight profiles in patients with type 2 diabetes mellitus. When administered as a monotherapy, dulaglutide did not reduce weight compared with control (placebo, metformin and liraglutide). However, when dulaglutide is added to an OAM and lispro, it lowered bodyweight by 1.30 kg compared with control (placebo, sitagliptin, exenatide, and liraglutide and glargine).
Of the included trials, three studies reported a significant dose-dependent reduction in HbA1c (dulaglutide 0.5 mg, 1.0 mg, and 1.5 mg 9 ; dulaglutide 0.25 mg, 0.5 mg, and 0.75 mg 15 ), FPG (dulaglutide 0.5 mg, 1.0 mg, and 1.5 mg 9 ) and weight loss (LY2189265 0.5/1.0 mg, 1.0/1.0 mg, and 1.0/2.0 mg 8 ). There were no dose-effect relationships with 0.75 mg QW, 1.0 mg QW, and 1.5 mg QW dulaglutide, whether administered as a monotherapy or as an add-on to another OAM and lispro, and control (placebo, metformin, sitagliptin, exenatide, liraglutide and glargine) regarding the reduction in HbA1c, FPG, the incidence of hypoglycemia, and the number of patients that experienced nausea and vomiting, except weight change. Therefore, additional RCTs are needed to clarify the dose-dependent effects of dulaglutide in the reduction in HbA1c and other parameters.
The risk of pancreatitis caused by incretin-based drugs used to treat diabetes is controversial, and the focus of international debates 32,33 . The FDA has made persistent efforts to estimate the risk of pancreatitis associated with incretin-based mimetic drugs. Current evidence does not support the view that incretin mimetic drugs are associated with an increased risk of pancreatitis 34,35 . In the current analysis the incidence of pancreatitis was rare in the dulaglutide arms; therefore, a causal relationship cannot be confirmed. And now, no evidences are found consuimg alcohol increases pancreatitis in the use of dulaglutide.
Up to date, safety and utility of dulaglutide were estimated in subjects with age more than 50 years, whether it shows the similar safety and utility in youth onset type 2 DM or people with age less than 50 years is unknown and needed to be evaluated in the future.
While most of the included studies on dulaglutide have been carried out on patients with HbA1c around 8-8.5%, there is some gap on the efficacy and safety. Mean duration of T2DM in the included studies was from about 3 to 12 years, which maybe one of the causes. With the extension of the duration, beta cell function decreases, this may lower dulaglutide efficacy because it reduces glucose through stimulation of islet cell secreting insulin. Second, doses of dulaglutide are various: 0.1 mg to 1.5 mg, that can lead to different efficacy and safety. Moreover, body weight, about 72 to 98 kg in our article, is associated with insulin sensitivity negatively, which can influence the efficacy of antidiabetic drug such as GLP-1 receptor agonists, insulin, metformin, and thiazolidinedione.
GLP-1 receptor agonists as the new agents for the treatment of T2DM, cost-benefit is an important topic. A retrospective cohort study estimated the cost effectiveness of treating patients to glycemic goal with Liraglutide versus Exenatide in a real-world clinical background and found that total diabetes related pharmacy costs per patient with the goal of HbA1c < 7% were lower when Liraglutide was used than Exenatide ($3,108 vs. $3,354; P < 0.0001) 36 . Up to now, no data explore the cost consideration versus benefit of use of dulaglutide and further researches are needed to estimate this scope of dulaglutide.
Limitations. Two of the included studies were Abstracts; therefore, some data could not be extracted and so some important information was lost. Moreover, only one trial provided data regarding renal function (serum creatinine < 1.5 mg/dL [males] or < 1.4 mg/dL [females]). Unlike endogenous GLP-1, dulaglutide is resistant to degradation by DPP-4; it is a large molecule with a delayed absorption rate and slower renal clearance. However, it remains unclear whether dulaglutide can be used in subjects with impaired renal function, and this should be evaluated. In addition, only a small number of articles reported fasting and postprandial blood glucose levels; therefore, we could not estimate postprandial changes after the administration of dulaglutide.

Conclusions
This meta-analysis evaluated the efficacy and safety of dulaglutide for the treatment of type 2 diabetes. First, compared with control, monotherapy dulaglutide exhibited beneficial effects regarding the control of HbA1c, had a similar risk of hypoglycemia and gastrointestinal disorders, and a less body weight reduction. Furthermore, when used as an add-on therapy to OAM and lispro, dulaglutide lowered HbA1c and bodyweight, brought similar risk Scientific RepoRts | 6:18904 | DOI: 10.1038/srep18904 of hypoglycemia and gastrointestinal disorders compared with control. Moreover, there were no dose-dependent relationships with 0.75 mg QW, 1.0 mg QW, and 1.5 mg QW dulaglutide either as a monotherapy or as an add-on intervention regarding HbA1c, FPG, the incidence of hypoglycemia, and the number of patients that experienced nausea, vomiting except weight loss, compared with control. In addition, further particularly long-term studies are needed to fully appraise the benefit/risk profiles of dulaglutide, which will help determine the superiority of dulaglutide.

Methods
Outcomes measures of efficacy and safety. The primary outcome of efficacy was the change in HbA1c from baseline to the end of the trials. The secondary endpoints were the percentage of participants achieving HbA1c value < 7% or ≤ 6.5%, fasting plasma glucose, and bodyweight. The safety and tolerability events were defined as any adverse events, symptomatic hypoglycemia, gastrointestinal disorders, and suspected pancreatitis.
Eligibility criteria. All randomized clinical trials (RCTs) that lasted at least 12 weeks and analyzed dulaglutide as a monotherapy or as an add-on to other hypoglycemic drugs compared with placebo or other active drugs were included. The included subjects were nonpregnant adults with type 2 diabetes that had been diagnosed according to American Diabetes Association (1997) or World Health Organization criteria (1999). Reviews, letters, case reports, non-human studies, and trials that lasted less than 12 weeks were excluded. Our article is meta-analysis, which is not related to ethics.
Search strategy. Medline (via PubMed), Embase (via OVID), the Cochrane Library and www.clinicaltrials. gov were searched until February 15 th , 2015. The search results were limited to studies performed in humans, and did not restrict the language. The search terms used were "dulaglutide" and "LY2189265", which were adjusted to comply with the relevant provisions in each database.
Data extraction. Two reviewers performed data extraction independently according to the inclusion and exclusion criteria. The extracted data included study design, baseline characteristics, interventions, efficacy outcomes, safety, and tolerability. Any discrepancies between the two reviewers were resolved by consensus in the presence of a third reviewer when necessary.

Risk of bias.
The criteria used to assess the risk of bias of RCTs were from the Cochrane Collaboration's risk of bias tool, and included random sequence generation, allocation concealment, blinding of outcome assessors, selective outcome reporting, and other items such as the funding source of the studies. Studies were graded as low, high, or unclear risk of bias. The two authors assessed the risk of bias and came to a consensus, and consulted a third reviewer to resolve any persistent disagreements.
Data synthesis and analysis. Continuous data were analyzed using mean differences (MDs) to express effect size, and relative risk (RR) was used to express dichotomous data. An inconsistency index (I 2 ), which is used to evaluate the heterogeneity of treatment effects, < 25%, 25-50%, and > 50% were considered as low, moderate, and high heterogeneity, respectively. A fixed-effects model was used for analysis if I 2 < 50%. In studies in which heterogeneity was identified, we searched for the sources of heterogeneity and subgroup analysis was considered or a random effects model was used. All analyses were performed using Review Manager V.5.2 statistical software.