Diabetes Care 1

Melanie J. Davies,1,2 David A. D’Alessio,3 Management of Hyperglycemia Judith Fradkin,4 Walter N. Kernan,5 Chantal Mathieu,6 Geltrude Mingrone,7,8 in Type 2 Diabetes, 2018. Peter Rossing,9,10 Apostolos Tsapas,11 A Consensus Report by the Deborah J. Wexler,12,13 and John B. Buse14 American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD) https://doi.org/10.2337/dci18-0033 OSNU REPORT CONSENSUS 1Diabetes Research Centre, University of Leices- ter, Leicester, U.K. 2Leicester Diabetes Centre, Leicester General Hospital, Leicester, U.K. 3Departmentof Medicine,Duke UniversitySchool of Medicine, Durham, NC 4National Institute of Diabetes and Digestive The American Diabetes Association and the European Association for the Study of and Kidney Diseases, National Institutes of Health, Bethesda, MD Diabetes convened a panel to update the prior position statements, published in 5Department of Medicine, Yale School of Med- 2012 and 2015, on the management of type 2 diabetes in adults. A systematic icine, New Haven, CT 6 evaluation of the literature since 2014 informed new recommendations. These Clinical and Experimental Endocrinology, UZ Gasthuisberg, KU Leuven, Leuven, Belgium include additional focus on lifestyle management and diabetes self-management 7Department of Internal Medicine, Catholic Uni- education and support. For those with obesity, efforts targeting weight loss, versity, Rome, Italy including lifestyle, medication, and surgical interventions, are recommended. With 8DiabetesandNutritionalSciences,King’sCollege London, London, U.K. regards to medication management, for patients with clinical cardiovascular 9 – Steno Diabetes Center Copenhagen, Gentofte, disease, a sodium glucose cotransporter 2 (SGLT2) inhibitor or a glucagon-like Denmark peptide 1 (GLP-1) receptor agonist with proven cardiovascular benefit is recom- 10University of Copenhagen, Copenhagen, Den- mended. For patients with chronic kidney disease or clinical heart failure and mark 11 atherosclerotic cardiovascular disease, an SGLT2 inhibitor with proven benefitis Second Medical Department, Aristotle Univer- fi sity Thessaloniki, Thessaloniki, Greece recommended. GLP-1 receptor agonists are generally recommended as the rst 12Department of Medicine and Diabetes Unit, injectable medication. Massachusetts General Hospital, Boston, MA 13Harvard Medical School, Boston, MA 14Department of Medicine, University of North The goals of treatment for type 2 diabetes are to prevent or delay complications and Carolina School of Medicine, Chapel Hill, NC maintainquality oflife(Fig. 1).This requires control ofglycemia andcardiovascular risk Corresponding author: John B. Buse, jbuse@med factormanagement, regularfollow-up,and,importantly,apatient-centered approach .unc.edu. to enhance patient engagement in self-care activities (1). Careful consideration of M.J.D.andJ.B.B.wereco-chairsfortheConsensus patient factors and preferences must inform the process of individualizing treatment Statement Writing Group. D.A.D’A., J.F., W.N.K., goals and strategies (2,3). and D.J.W. were the writing group members for This consensus report addresses the approaches to management of glycemia in the American Diabetes Association. C.M., G.M., P.R.,andA.T.werewritinggroupmembersforthe adults with type 2 diabetes, with the goal of reducing complications and maintaining European Association for the Study of Diabetes. quality of life in the context of comprehensive cardiovascular risk management and This article is being simultaneously published in patient-centered care. The principles of how this can be achieved are summarized in Diabetes Care and Diabetologia by the American Fig. 1 and underpin the approach to management and care. These recommendations Diabetes Association and the European Associ- are not generally applicable to patients with monogenic diabetes, secondary diabetes, ation for the Study of Diabetes. or type 1 diabetes, or to children. © 2018 American Diabetes Association and European Association for the Study of Diabetes. Data Sources, Searches, and Study Selection Readers may use this article as long as the work is properly cited, the use is educational and not The writing group accepted the 2012 (4) and 2015 (5) editions of this position for profit, and the work is not altered. More infor- statement as a starting point. To identify newer evidence, a search was conducted on mation is available at http://www.diabetesjournals PubMed for randomized clinical trials (RCTs), systematic reviews, and meta-analyses .org/content/license. Diabetes Care Publish Ahead of Print, published online October 4, 2018 osnu Report Consensus 2 ibtsCare Diabetes

Figure 1—Decision cycle for patient-centered glycemic management in type 2 diabetes. care.diabetesjournals.org Davies and Associates 3

published in English between 1 January studies with other medication classes. laboratories (www.ngsp.org) (11). As 2014 and 28 February 2018; eligible The greatest absolute risk reduction with any laboratory test, HbA1c has lim- publications examined the effectiveness (ARR) comes from improving poor gly- itations (2). Because there is variability or safety of pharmacological or nonphar- cemic control, and a more modest re- in the measurement of HbA1c, clinicians macological interventions in adults with duction results from near normalization should exercise judgment, particularly type 2 diabetes mellitus. Reference lists of glycemia (6). The impact of glucose when the result is close to the threshold were scanned in eligible reports to iden- control on macrovascular complica- that might prompt a change in therapy. tify additional articles relevant to the sub- tions is less certain. Because the bene- HbA1c results may be discrepant from the ject. Details on the keywords and the fits of intensive glucose control emerge patient’s true mean glycemia in certain search strategy are available at https:// slowly, while the harms can be immedi- racial and ethnic groups, and in condi- doi.org/10.17632/h5rcnxpk8w.1. Papers ate, people with longer life expectancy tions that alter red blood cell turnover, were grouped according to subject, and have more to gain from intensive glu- such as anemia, end-stage renal disease the authors reviewed this new evidence to cose control. A reasonable HbA1c target (ESRD) (especially with erythropoietin inform the consensus recommendations. for most nonpregnant adults with suffi- therapy), and pregnancy, or if an HbA1c The draft consensus recommendations cient life expectancy to see microvascu- assay sensitive to hemoglobin variants were peer reviewed (see “Acknowledg- lar benefits (generally ;10 years) is is used in someone with sickle cell trait ments”), and suggestions incorporated around 53 mmol/mol (7%) or less (6). or other hemoglobinopathy. Discrep- as deemed appropriate by the authors. Glycemic treatment targets should be ancies between measured HbA1c and Nevertheless, though evidence-based, individualized based on patient prefer- measured or reported glucose levels the recommendations presented herein ences and goals, risk of adverse effects should prompt consideration that one are the opinions of the authors. of therapy (e.g., hypoglycemia and of these may not be reliable (12). weight gain), and patient characteris- Regular self-monitoring of blood glucose The Rationale, Importance, and tics, including frailty and comorbid (SMBG) may help with self-management Context of Glucose-Lowering conditions (2). and medication adjustment, particularly Treatment Atherosclerotic cardiovascular dis- in individuals taking insulin. SMBG plans Lifestyle management, including medical ease (ASCVD) is the leading cause of should be individualized. People with nutrition therapy (MNT), physical activ- death in people with type 2 diabetes diabetes and the health care team should ity, weight loss, counseling for smoking (7). Diabetes confers substantial inde- use the data in an effective and timely cessation, and psychological support, pendent ASCVD risk, and most people manner. In people with type 2 diabetes often delivered in the context of diabetes with type 2 diabetes have additional risk not using insulin, routine glucose mon- self-management education and support factors such as hypertension, dyslipide- itoring is of limited additional clinical (DSMES), are fundamental aspects of mia, obesity, physical inactivity, chronic benefit while adding burden and cost diabetes care. The expanding number kidney disease (CKD), and smoking. Nu- (13,14). However, for some individuals, of glucose-lowering treatmentsdfrom merous studies have demonstrated glucose monitoring can provide insight behavioral interventions to medica- the benefits of controlling modifiable into the impact of lifestyle and medica- tions and surgerydand growing infor- ASCVD risk factors in people with di- tion management on blood glucose and mation about their benefits and risks abetes. Substantial reductions in ASCVD symptoms, particularly when combined provides more options for people with events and death are seen when multi- with education and support. Novel tech- diabetes and providers, but can compli- ple ASCVD risk factors are addressed nologies, such as continuous or flash cate decision making. In this consen- simultaneously, with long-standing ben- glucose monitoring, provide more infor- sus statement, we attempt to provide efits (8,9). Comprehensive implementa- mation. However, in type 2 diabetes, an approach that summarizes a large tion of evidence-based interventions has they have been associated with only body of recent evidence for practi- likely contributed to the significant re- modest benefits (15). tioners in the U.S. and Europe. ductions in ASCVD events and mortality Marked hyperglycemia is associated seen in people with diabetes in recent Principles of Care with symptoms including frequent uri- decades (10). ASCVD risk management Consensus recommendation nation, thirst, blurred vision, fatigue, and in its many forms is an essential part c Providers and health care systems recurring infections. Beyond alleviat- of diabetes management that is beyond should prioritize the delivery of ing symptoms, the aim of blood glu- the scope of this statement, but physi- patient-centered care. cose lowering (hereafter, referred to cians should be aware of the impor- as glycemic management) is to reduce tance of multifactorial treatment in Providing patient-centered care that long-term complications of diabetes. type 2 diabetes (7). acknowledges multimorbidity, and is Good glycemic management yields sub- respectful of and responsive to individ- stantial and enduring reductions in on- Glucose Management: Monitoring ual patient preferences and barriers, set and progression of microvascular Glycemic management is primarily as- including the differential costs of thera- complications. This benefit has been sessed with the HbA1c test, which was the pies, is essential to effective diabetes demonstrated most clearly early in the measure studied in trials demonstrating management (16). Shared decision mak- natural history of the disease in studies the benefits of glucose lowering (2). The ing, facilitated by decision aids that using metformin, sulfonylureas, and in- performance of the test is generally show the absolute benefit and risk of sulin but is supported by more recent excellent for NGSP-certified assays and alternative treatment options, is a useful 4 Consensus Report Diabetes Care

strategy to arrive at the best treatment In type 2 diabetes, high-quality evi- care. Multiple factors contribute to course for an individual (17–20). Pro- dence has consistently shown that inconsistent medication use and treat- viders should evaluate the impact of DSMES is a cost-effective intervention ment discontinuation, including patient- any suggested intervention, including in the health care systems studied. perceived lack of medication efficacy, self-care regimens, in the context of DSMES significantly improves clinical fear of hypoglycemia, lack of access cognitive impairment, limited literacy, and psychological outcomes, improves to medication, and adverse effects of distinct cultural beliefs, and individual glycemic control, reduces hospital ad- medication (37). Medication adherence fears or health concerns given their missions, improves patient knowledge, (including persistence) varies across impact on treatment efficacy. and reduces the risk of all-cause mor- medication classes and careful consider- tality (22,26–31). The best outcomes ation of these differences may help im- DSMES are achieved in those programs with a prove outcomes (38). Ultimately, patient preference is a major factor driving the Consensus recommendation theory-based and structured curricu- choice of medication. Even in cases c All people with type 2 diabetes lum and with contact time of over where clinical characteristics suggest should be offered access to ongo- 10 h. While online programs may re- the use of a particular medication based ing DSMES programs. inforce learning, there is little evi- dence they are effective when used on the available evidence from clinical trials, patient preferences regarding DSMES is a key intervention to en- alone (27). route of administration, injection de- able people with diabetes to make vices, side effects, or cost may prevent informed decisions and to assume re- Consensus recommendation their use by some individuals (39). sponsibility for day-to-day diabetes c Facilitating medication adherence Therapeutic inertia, sometimes re- management. DSMES is central to es- should be specifically considered ferred to as clinical inertia, refers to tablishing and implementing the princi- when selecting glucose-lowering failure to intensify therapy when treat- ples of care (Fig. 1). DSMES programs medications. ment targets are not met. The causes usually involve face-to-face contact in of therapeutic inertia are multifacto- group or individual sessions with trained Suboptimal adherence, including poor rial, occurring at the level of the prac- educators, and key components are persistence, to therapy affects almost titioner, patient, and/or health care shown in Table 1 (21–25). While DSMES half of people with diabetes, leading system (40). Interventions targeting should be available on an ongoing basis, to suboptimal glycemic and cardiovas- therapeutic inertia have facilitated critical junctures when DSMES should cular disease (CVD) risk factor control as improved glycemic control and timely occur include at diagnosis, annually, well as increased risk of diabetes com- insulin intensification (41,42). For when complications arise, and during plications, mortality, hospital admis- example, multidisciplinary teams that transitions in life and care (22). sions, and health care costs (32–36). include nurse practitioners or pharma- DSMES programs delivered from Though this consensus recommenda- cists may help reduce therapeutic iner- diagnosis can promote medication tion focuses on medication adherence tia (43,44). A fragmented health care adherence, healthy eating, and physi- (including persistence), the principles system may contribute to therapeutic cal activity, and increase self-efficacy. are pertinent to all aspects of diabetes inertia and impair delivery of patient- centered care. A coordinated chronic — – Table 1 Key components of DSMES (21,23 25) care model, including self-management c Evidence-based support, decision support, delivery sys- c Individualized to the needs of the person, including language and culture tem design, clinical information sys- c Has a structured theory-driven written curriculum with supporting materials tems, and community resources and c Delivered by trained and competent individuals (educators) who are quality assured policies, promotes interaction between c Delivered in group or individual settings more empowered patients and better c Aligns with the local population needs prepared and proactive health care c Supports the person and their family in developing attitudes, beliefs, knowledge, and skills to teams (45). self-manage diabetes c Includes core content; i.e., diabetes pathophysiology and treatment options; medication RECOMMENDED PROCESS FOR usage; monitoring, preventing, detecting, and treating acute and chronic complications; GLUCOSE-LOWERING healthy coping with psychological issues and concerns; problem solving and dealing with special situations (i.e., travel, fasting) MEDICATION SELECTION: WHERE DOES NEW EVIDENCE FROM c Available to patients at critical times (i.e., at diagnosis, annually, when complications arise, and when transitions in care occur) CARDIOVASCULAR OUTCOMES TRIALS FIT IN? c Includes monitoring of patient progress, including health status, quality of life fi c Quality audited regularly In prior consensus statements, ef cacy in reducing hyperglycemia, along with DSMES is a critical element of care for all people with diabetes and is the ongoing process of facilitating the knowledge, skills, and ability necessary for diabetes self-care as well as activities tolerability and safety were primary that assist a person implementing and sustaining behaviors needed to manage their factors in glucose-lowering medication diabetes on an ongoing basis. National organizations in the U.S. and Europe have published selection. Patient preferences, glycemic standards to underpin DSMES. In the U.S., these are defined as DSMES “services,” whereas in targets, comorbidities, polypharmacy, Europe they are often referred to as “programs.” Nevertheless, the broad components are similar. side effects, and cost were additional care.diabetesjournals.org Davies and Associates 5

important considerations. For every in- Where the current evidence is strongest 0.74, 0.97; P 5 0.02; ARR 1.4%) for dividual, the choice of glucose-lowering for a specific medication within a class, it all-cause mortality (47). In the Trial to medication should be underpinned by is noted. The American Diabetes Asso- Evaluate Cardiovascular and Other Long- lifestyle management, DSMES, and the ciation’s (ADA) Standards of Medical term Outcomes with Semaglutide in patient-centered care principles outlined Care in Diabetes will align with this Subjects with Type 2 Diabetes (SUSTAIN in Fig. 1. document and will be updated to reflect 6) (n 5 3,297), semaglutide compared Figure 2 describes our new consensus new evidence as it emerges from ongoing with placebo demonstrated an ARR of approach to glucose lowering with med- clinical trials. 2.3% with HR 0.74 for MACE (95% CI 0.58, ications in type 2 diabetes. Because of the 0.95; P 5 0.02 for superiority) over 2.1 new evidence for the benefit of specific Consensus recommendation years, but the reduction in events ap- medications to reduce mortality, heart c Among patients with type 2 diabe- peared to be driven by the rate of stroke failure (HF), and progression of renal tes who have established ASCVD, rather than CVD death (48). The Exena- disease in the setting of established SGLT2 inhibitors or GLP-1 recep- tide Study of Cardiovascular Event Low- CVD, their use was considered compel- tor agonists with proven cardiovas- ering (EXSCEL) compared exenatide ling in this patient group. Thus, we rec- cular benefit are recommended extended-release with placebo over ommend that providers consider a as part of glycemic management 3.2 years in 14,752 participants with history of CVD very early in the process (Figs. 2 and 3). type 2 diabetes. While the medication of treatment selection. Other factors was safe (noninferior), the HR for MACE affect the choice of glucose-lowering ASCVD is defined somewhat differ- in the entire trial was 0.91 (95% CI 0.83, medications, particularly in the setting ently across trials, but all trials enrolled 1.0; P 5 0.06) not reaching the threshold of patient-centered care. In addition to individuals with established CVD (e.g., for demonstrated superiority versus pla- CVD, we recommend early consideration myocardial infarction [MI], stroke, cebo; ARR was 0.8% (49). All-cause death of weight, hypoglycemic risk, treatment any revascularization procedure) while was lower in the exenatide arm (ARR 1%, cost, and other patient-related factors variably including related conditions HR 0.86 [95% CI 0.77, 0.97]), but it was that may influence treatment selection compatible with clinically significant not considered to be statistically signif- (Figs. 2–6). atherosclerosis (e.g., transient ischemic icant in the hierarchical testing proce- attack, hospitalized unstable angina, dure applied. Lixisenatide, a short-acting Implications of New Evidence From amputation, congestive heart failure GLP-1 receptor agonist, did not demon- Cardiovascular Outcomes Trials New York Heart Association [NYHA] class strate CVD benefit or harm in a trial of The major change from prior consensus II–III, .50% stenosis of any artery, symp- patients recruited within 180 days of an reports is based on new evidence that tomatic or asymptomatic coronary artery acute coronary syndrome admission (50). specific sodium–glucose cotransporter disease documented by imaging, CKD Taken together, it appears that among 2 (SGLT2) inhibitors or glucagon-like with estimated glomerular filtration rate patients with established CVD, some GLP- peptide 1 (GLP-1) receptor agonists im- [eGFR] ,60mL min21[1.73]22).Most trials 1 receptor agonists may provide cardio- prove cardiovascular outcomes, as well also included a “risk factor only” group vascular benefit, with the evidence of as secondary outcomes such as HF and with entry criteria based on age and usually benefit strongest for liraglutide, favorable progression of renal disease, in patients thepresenceoftwoormorecardiacrisk for semaglutide, and less certain for ex- with established CVD or CKD. Therefore, factors (46). Trials were designed to eval- enatide. There is no evidence of cardio- an important early step in this new uate cardiovascular safety (i.e., statistical vascular benefit withlixisenatide. Adverse approach (Fig. 3) is to consider the pres- noninferiority compared with placebo), but effects for the class are discussed in the ence or absence of ASCVD, HF, and CKD, several showed ASCVD outcome benefit section “The Full Range of Therapeutic conditions in aggregate affecting 15–25% (i.e., statistical superiority compared with Options: Lifestyle Management, Medica- of the population with type 2 diabetes. placebo), including, in some cases, mortality. tion, and Obesity Management.” While the new evidence supporting the Among GLP-1 receptor agonists, lira- Among the SGLT2 inhibitors, em- use of particular medications in patients glutide, studied in the Liraglutide pagliflozin compared with placebo was who also have established CVD or are at Effect and Action in Diabetes: Evalua- studied in the Empagliflozin, Cardiovas- high risk of CVD is derived from large tion of Cardiovascular Outcome Results cular Outcome Event Trial in Type 2 cardiovascular outcomes trials (CVOTs) (LEADER) trial (n 5 9,340), demonstrated Diabetes Mellitus Patients (EMPA-REG demonstrating substantial benefits over an ARR of 1.9% with a hazard ratio (HR) OUTCOME) in 7,020 participants with 2–5 years, it is important to remember of 0.87 (95% CI 0.78, 0.97; P 5 0.01 for type 2 diabetes and CVD. With a median that each trial constitutes a single ex- superiority) for the primary composite follow-up of 3.1 years, the ARR was 1.6% periment. Within each drug class, results outcome of cardiovascular death, non- and the HR was 0.86 (95% CI 0.74, 0.99; have been heterogeneous. It is not clear fatal MI, and nonfatal stroke (major ad- P 5 0.04 for superiority) for the primary whether there are true drug-class ef- verse cardiac events [MACE]) compared composite end point of nonfatal MI, fects with different findings for individ- with placebo over 3.8 years. Each com- nonfatal stroke, and cardiovascular ual medications due to differences in ponent of the composite contributed to death. The ARR was 2.2% and the HR trial design and conduct, or whether the benefit, and the HR for cardiovascular was 0.62 (95% CI 0.49, 0.77; P , 0.001) there are real differences between death was 0.78 (95% CI 0.66, 0.93; P 5 for cardiovascular death (51). The ARR medications within a drug class due to 0.007; ARR 1.7%). The LEADER trial also was 2.6% and the HR was 0.68 (95% CI properties of the individual compounds. demonstrated an HR of 0.85 (95% CI, 0.57, 0.82; P , 0.001) for death from osnu Report Consensus 6 ibtsCare Diabetes

Figure 2—Glucose-lowering medication in type 2 diabetes: overall approach. CV, cardiovascular; DPP-4i, dipeptidyl peptidase 4 inhibitor; GLP-1 RA, glucagon-like peptide 1 receptor agonist; SGLT2i, SGLT2 inhibitor; SU, sulfonylurea. care.diabetesjournals.org Davies and Associates 7

Figure 3—Choosing glucose-lowering medication in those with established ASCVD, HF, and CKD. CV, cardiovascular; DPP-4i, dipeptidyl peptidase 4 inhibitor; GLP-1 RA, glucagon-like peptide 1 receptor agonist; SGLT2i, SGLT2 inhibitor; SU, sulfonylurea. 8 Consensus Report Diabetes Care

Figure 4—Choosing glucose-lowering medication if compelling need to minimize weight gain or promote weight loss. GLP-1 RA, glucagon-like peptide 1 receptor agonist; T2DM, type 2 diabetes; SGLT2i, SGLT2 inhibitor; SU, sulfonylurea. care.diabetesjournals.org Davies and Associates 9

Figure 5—Choosing glucose-lowering medication if compelling need to minimize hypoglycemia. DPP-4i, dipeptidyl peptidase 4 inhibitor; GLP-1 RA, glucagon-like peptide 1 receptor agonist; SGLT2i, SGLT2 inhibitor. 10 Consensus Report Diabetes Care

Figure 6—Choosing glucose-lowering medication if cost is a major issue. DPP-4i, dipeptidyl peptidase 4 inhibitor; SGLT2i, SGLT2 inhibitor; SU, sulfonylurea. care.diabetesjournals.org Davies and Associates 11

any cause. Canagliflozin compared with that patients with clinical CVD not meet- hospitalized HF of 5.5 vs. 8.7 events per placebo was studied in the Canagliflozin ing individualized glycemic targets while 1,000 patient-years (55). Because HF was Cardiovascular Assessment Study treated with metformin (or in whom neither well characterized at baseline nor (CANVAS) Program (comprised of two metformin is contraindicated or not tol- as carefully adjudicated as it would have similar trials, CANVAS and CANVAS- erated) should have an SGLT2 inhibitor been in a trial specifically designed to Renal; n 5 10,142) in participants with or GLP-1 receptor agonist with proven evaluate HF outcomes, and because HF type 2 diabetes, 66% of whom had benefit for cardiovascular risk reduction was a secondary end point in the trials, a history of CVD. Participants were added to their treatment program. There further ongoing studies are required to followed for a median of 3.6 years. In are no clinical trial data that support conclusively address the issue. That said, the combined analysis of the two trials, prescribing an SGLT2 inhibitor or GLP- the significant reduction in hospitaliza- the primary composite end point of 1 receptor agonist with the intent of tion for HF demonstrated in the two MI, stroke, or cardiovascular death was reducing cardiovascular risk in patients study populations and the consistency reduced with canagliflozin (26.9 vs. with an HbA1c ,53 mmol/mol (,7%). across two independent trial programs 31.5 participants per patient-year with Limited data suggest that there is no suggesttousthattreatmentwithSGLT2 placebo; HR 0.86 [95% CI 0.75, 0.97]; heterogeneity in the cardiovascular ben- inhibitors in the setting of clinical HF may P 5 0.02) for superiority in the pooled efits of SGLT2 inhibitors or GLP-1 recep- provide substantial benefitandshould analysis, with consistent findings in tor agonists as a function of background be specifically considered in people with the component studies. Though there glucose-lowering therapy. Thus, back- type 2 diabetes and ASCVD and HF. wasatrendtowardbenefitforcardio- ground glucose-lowering therapy in pa- In the GLP-1 receptor agonist studies vascular death, the difference from tients with clinical CVD arguably is not LEADER, SUSTAIN 6, and EXSCEL, there placebo was not statistically significant pertinent in clinical decision making. was no significant effect on hospitaliza- in the CANVAS Program (52). For the However, dose adjustment or discontin- tion for HF with HR 0.86 (95% CI 0.71, SGLT2 inhibitors studied to date, it ap- uation of background medications may 1.06), 1.11 (95% CI 0.77, 1.61), and 0.94 pears that among patients with established be required to avoid hypoglycemia (95% CI 0.78, 1.13), respectively (47–49). CVD, there is likely cardiovascular benefit, when adding a new agent to a regi- Two short-term studies of liraglutide in with the evidence of benefitmodestly men containing insulin, sulfonylurea, or patients with reduced ejection fraction stronger for empagliflozin than canagli- glinide therapy, particularly in patients at suggested a lack of benefit in this setting flozin. Adverse effects for the class or near glycemic goals. Full efforts to (56,57). are discussed in the section “The Full achieve glycemic and blood pressure Among the recent cardiovascular Range of Therapeutic Options: Lifestyle targets and to adhere to lipid, antiplate- safety outcomes trials testing DPP-4 in- Management, Medication, and Obesity let, antithrombotic, and tobacco ces- hibitors, the Saxagliptin Assessment of Management.” sation guidelines (7) should continue Vascular Outcomes Recorded in Patients While the evidence of an ASCVD out- afteranSGLT2inhibitororGLP-1re- with Diabetes Mellitus–Thrombolysis comes benefit for GLP-1 receptor ago- ceptor agonist is added, as such efforts in Myocardial Infarction 53 (SAVOR- nists and SGLT2 inhibitors has been were integral to all studies that have TIMI 53) study evaluating saxagliptin demonstrated for people with estab- demonstrated cardiovascular benefit demonstrated a significant increased risk lished ASCVD, the evidence of benefit of these agents. of HF, with 3.5% risk of hospitalization beyond glucose lowering has not been for HF versus 2.8% for placebo (HR 1.27; demonstrated in those without ASCVD. Consensus recommendation 95% CI 1.07, 1.51; P 5 0.007) (58). In the Indeed, in subgroup analyses of these c Among patients with ASCVD in subsequent Examination of Cardiovascu- trials, lower-risk individuals have not whom HF coexists or is of special lar Outcomes with Alogliptin versus Stan- been observed to have an ASCVD benefit. concern, SGLT2 inhibitors are rec- dard of Care (EXAMINE) study of alogliptin While this may be due to the short time ommended (Figs. 2 and 3). there was no statistically significant dif- frame of the studies and the low event ference in HF hospitalization (3.9% vs. 3.3% rate in those without ASCVD, the finding Patients with type 2 diabetes are at with placebo) (59), and in the Trial Eval- is consistent across the reported trials. increased risk of HF (53). In the EMPA- uating Cardiovascular Outcomes with Overall, CVOTs of dipeptidyl peptidase REG OUTCOME and CANVAS CVOT stud- Sitagliptin (TECOS), the rate of hospital- 4 (DPP-4) inhibitors have demonstrated ies testing SGLT2 inhibitors, which izationforHFwas3.1%inbothsitagliptin- safety, i.e., noninferiority relative to pla- enrolled participants with ASCVD, .85% and placebo-treated patients (60). cebo, for the primary MACE end point, of participants did not have symptomatic but not cardiovascular benefit. HF at baseline. Yet, in both trials there Consensus recommendation The available evidence for cardiovas- was a clinically and statistically significant c For patients with type 2 diabetes cular event reduction in patients with reduction in hospitalization for HF for and CKD, with or without CVD, type 2 diabetes and clinical CVD is derived the SGLT2 inhibitor as compared with consider the use of an SGLT2 in- from trials in which the participants were placebo. In the EMPA-REG OUTCOME hibitor shown to reduce CKD pro- not meeting glycemic targets (HbA study with empagliflozin (54), the ARR 1c gression or, if contraindicated or $53 mmol/mol [$7%] at baseline). Fur- was 1.4%, and the HR 0.65 (95% CI not preferred, a GLP-1 receptor thermore, most (;70% across trials) 0.50, 0.85), and in the CANVAS Pro- agonist shown to reduce CKD pro- participants were treated with metfor- gram with canagliflozin, the HR was gression (Figs. 2 and 3). min at baseline. Thus, we recommend 0.67 (95% CI 0.52, 0.87), with a rate of 12 Consensus Report Diabetes Care

Patients with type 2 diabetes and kidney components (doubling of serum creati- emphasize foods of demonstrated health disease are at an increased risk for car- nine, ESRD, or renal death) did not con- benefit, that minimize foods of demon- diovascular events. A substantial number tribute substantially to the benefit. In strated harm, and that accommodate of participants with an eGFR of 30–60 mL the DPP-4 inhibitor CVOTs, the DPP-4 in- patient preference and metabolic needs, min–1 [1.73 m]–2 were included in EMPA- hibitors have been shown to be safe with the goal of identifying healthy di- REG OUTCOME, CANVAS, LEADER, and from a renal perspective, with modest etary habits that are feasible and sustain- SUSTAIN 6. An important finding in the reduction in albuminuria (64). able. Three trials of a Mediterranean studies was reduction of the primary eating pattern reported modest weight ASCVD outcome even among participants THE FULL RANGE OF THERAPEUTIC loss and improved glycemic control –1 with stage 3 CKD (eGFR 30–60 mL min OPTIONS: LIFESTYLE (66–68). In one of these, people with – [1.73 m] 2). For SGLT2 inhibitors, this con- MANAGEMENT, MEDICATION, new-onset diabetes assigned to a low- trasts with the glucose-lowering effect, AND OBESITY MANAGEMENT carbohydrate Mediterranean eating pat- which diminishes with declining eGFR. This section summarizes the lifestyle, tern were 37% less likely to require In addition to the primary cardiovas- medication, and obesity management glucose-lowering medications over 4 cular end points, most of the SGLT2 therapies that lower glucose or improve years compared with patients assigned inhibitor and GLP-1 receptor agonist other outcomes in patients with type 2 to a low-fat diet (HR 0.63 [95% CI 0.51, CVOTs reported benefit in renal end diabetes. A more comprehensive discus- 0.86]). A meta-analysis of RCTs in pa- points, albeit as secondary outcomes. sion of these issues is available elsewhere tients with type 2 diabetes showed that The renal outcome benefit has been (3,21,65). For more details on weight loss the Mediterranean eating pattern re- most pronounced and consistent for medications and metabolic surgery, see duced HbA1c more than control diets SGLT2 inhibitors. EMPA-REG OUTCOME the section “Obesity Management Beyond (mean difference 23.3 mmol/mol, 95% (empagliflozin) demonstrated an ARR Lifestyle Intervention.” Basic information CI 25.1, 21.5 mmol/mol [20.30%, 95% CI 6.1%, HR of 0.61 (95% CI 0.53, 0.70) about specific options in each category of 20.46%, 20.14%]) (69). Low-carbohydrate, for the composite outcome of new or therapyissummarizedinTable2. low glycemic index, and high-protein diets, worsening nephropathy (progression to Lifestyle interventions, including MNT and the Dietary Approaches to Stop Hyper- urine albumin/creatinine ratio .33.9 and physical activity, are effective and tension (DASH) diet all improve glycemic mg/mmol [.300 mg/g], doubling of se- safe for improving glucose control in type control, but the effect of the Mediterra- rum creatinine and ESRD, or death by 2 diabetes. For these reasons, they are nean eating pattern appears to be the ESRD). The most prevalent outcome com- recommended as first-line therapies from greatest (70–72). Low-carbohydrate diets ponent was the development of sustained the time of diagnosis and as cotherapy (,26% of total energy) produce substan- 2 albuminuria, but the other components for patients who also require glucose- tial reductions in HbA1c at 3 months ( 5.2 were each significantly reduced relative lowering medications or metabolic sur- mmol/mol, 95% CI 27.8, 22.5 mmol/mol to placebo (61). CANVAS (canagliflozin) gery. Lifestyle management should be part [20.47%, 95% CI 20.71%, 20.23%]) and reported an HR of 1.7 (95% CI 1.51, 1.91) of the ongoing discussion with individ- 6 months (4.0 mmol/mol, 95% CI 26.8, for regression of albuminuria and a 40% uals with type 2 diabetes at each visit. 21.0 mmol/mol [20.36%, 95% CI 20.62%, reduction in risk in the composite out- 20.09%]), with diminishing effects at 12 come of eGFR, ESRD, or renal death (5.5 Lifestyle Management and 24 months; no benefit of moderate vs. 9.0 participants per 1,000 patient- carbohydrate restriction (26–45%) was years; HR 0.60; 95% CI 0.47, 0.77) (52). Consensus recommendation observed (73). Vegetarian eating pat- c An individualized program of MNT Additional trials with primary renal end terns have been shown to lower HbA , should be offered to all patients. 1c points are ongoing in high-risk renal but not fasting glucose, compared with populations. The Evaluation of the Ef- nonvegetarian ones (74). Very recent tri- fects of Canagliflozin on Renal and Car- Medical Nutrition Therapy als of different eating patterns in type 2 diovascular Outcomes in Participants MNTcomprises education andsupportto diabetes have typically also included with Diabetic Nephropathy (CREDENCE) help patients adopt healthy eating pat- weight reduction, hindering firm con- trial examining canagliflozin in CKD with terns. The goal of MNT is to manage clusions regarding the distinct contribu- proteinuria has been stopped at a planned blood glucose and cardiovascular risk tion of dietary quality. interim analysis for achieving the primary factorstoreduceriskfordiabetes-related complications while preserving the plea- efficacy end point (62). Consensus recommendation In LEADER and SUSTAIN 6, the GLP-1 sure of eating (21). Two basic dimensions c All overweight and obese patients receptor agonist liraglutide was associ- of MNT include dietary quality and energy with diabetes should be advised of ated with an ARR of 1.5% and an HR of restriction. Strategies directed at each the health benefits of weight loss 0.78 (95% CI 0.67, 0.92) for new or wors- dimension can improve glycemic control. and encouraged to engage in a ening nephropathy (63), and semaglu- Dietary Quality and Eating Patterns. There is program of intensive lifestyle man- tide demonstrated an ARR of 2.3% no single ratio of carbohydrate, proteins, agement, which may include food and an HR of 0.64 (95% CI 0.46, 0.88) and fat intake that is optimal for every substitution. for new or worsening nephropathy (48). person with type 2 diabetes. Instead, Progression of albuminuria was the there are many good options and pro- Nonsurgical Energy Restriction for Weight most prevalent component of the com- fessional guidelines usually recommend Loss. If a patient wishes to aim for re- posite renal end point, whereas the other individually selected eating patterns that mission of type 2 diabetes, particularly care.diabetesjournals.org Table 2—Glucose-lowering medications and therapies available in the U.S. or Europe and specific characteristics that may guide individualized treatment choices in nonpregnant adults with type 2 diabetes Medications/therapies Class in class Primary physiological action(s) Advantages Disadvantages/adverse effects Efficacy Lifestyle Diet quality c Mediterranean type c Depends on diet c Inexpensive c Requires instruction Intermediate c DASH c No side effects c Requires motivation c Low carbohydrate c Requires lifelong behavioral change c Vegetarian c Social barriers may exist c Others Physical activity c Running, walking c Energy expenditure c Inexpensive c Risk of musculoskeletal injury Intermediate c Bicycling (including c Weight management c ↓ Fall risk by increasing balance/strength c Requires motivation stationary) c ↑ Insulin sensitivity c ? Improves mental health c Risk of foot trauma in patients with c Swimming c ↑ Bone density neuropathy c Resistance training c ↓ Blood pressure c Requires lifelong behavioral change c Yoga c ↓ Weight c Tai chi c Improves ASCVD risk factors c Many others Energy restriction c Individual energy c Energy restriction c Lowers glycemia c Requires motivation Variable, with potential for restriction with or c Weight management c Reduces need for diabetes and c Requires lifelong behavioral change very high efficacy; often without energy tracking c ↓ Hepatic and pancreatic other medications intermediate c Programs with fat c No serious side effects counseling c ↑ Insulin sensitivity c Improves ASCVD risk factors c Food substitution programs Oral medications Biguanides c Metformin c ↓ Hepatic glucose production c Extensive experience c GI symptoms High c Multiple other non-insulin- c No hypoglycemia c Vitamin B12 deficiency mediated mechanisms c Inexpensive c Use with caution or dose adjustment for CKD stage 3B (eGFR 30–44 mL min21 [1.73 m]22) c Lactic acidosis (rare) SGLT2 inhibitors c Canagliflozin c Blocks glucose reabsorption c No hypoglycemia c Genital infections Intermediate–high c Dapagliflozin by the kidney, increasing c ↓ Weight c UTI (dependent on GFR) c Empagliflozin glucosuria c ↓ Blood pressure c Polyuria c Ertugliflozin c ? Other tubulo-glomerular c Effective at all stages of T2DM c Volume depletion/hypotension/ effects with preserved glomerular function dizziness c ↓ MACE, HF, CKD with some c ↑ LDL-C agents (see text) c ↑ Creatinine (transient) c Dose adjustment/avoidance for renal

disease 13 Associates and Davies c ↑ Risk for amputation (canagliflozin) c ↑ Risk for fracture (canagliflozin)

Continued on p. 14 4CnessReport Consensus 14

Table 2—Continued Medications/therapies Class in class Primary physiological action(s) Advantages Disadvantages/adverse effects Efficacy

c ↑ Risk for DKA (rare) c Fournier’s gangrene (rare) c Expensive DPP-4 inhibitors c Sitagliptin c Glucose dependent: c No hypoglycemia c Rare urticaria/angioedema Intermediate c Vildagliptina ↑ Insulin secretion c Weight neutral c ↑ HF hospitalization (saxagliptin) c Saxagliptin ↓ Glucagon secretion c Well tolerated c Dose adjustment/avoidance for renal c Linagliptin disease depending on agent c Alogliptin c ? Pancreatitis c ? Arthralgia c ? Bullous pemphigoid c Expensive (U.S.); variable in Europe Sulfonylureas c Glibenclamide/glyburide c ↑ Insulin secretion c Extensive experience c Hypoglycemia High c Glipizide c ↓ Microvascular risk (UKPDS) c ↑ Weight c Gliclazidea c Inexpensive c Uncertain cardiovascular safety c Glimepiride c Dose adjustment/avoidance for renal disease c High rate of secondary failure TZDs c Pioglitazone c ↑ Insulin sensitivity c Low risk for hypoglycemia c ↑ Weight High c Rosiglitazoneb c Durability c Edema/heart failure c ↑ HDL-C c Bone loss c ↓ Triacylglycerols (pioglitazone) c ↑ Bone fractures c ↓ ASCVD events (pioglitazone: in a c ↑ LDL-C (rosiglitazone) poststroke insulin-resistant population and c ? Bladder cancer as secondary end point in a high-risk-of- c ? Macular edema CVD diabetes population) c Lower cost Meglitinides (Glinides) c Repaglinide c ↑ Insulin secretion c ↓ Postprandial glucose excursions c Hypoglycemia Intermediate–high c Nateglinide c Dosing flexibility c ↑ Weight c Safe in advanced renal disease with c Uncertain cardiovascular safety cautious dosing (especially repaglinide) c Frequent dosing schedule c Lower cost a-Glucosidase inhibitors c Acarbose c Slows carbohydrate c Low risk for hypoglycemia c Frequent GI side effects Low–intermediate c Miglitol digestion/absorption c ↓ Postprandial glucose excursions c Frequent dosing schedule c Nonsystemic mechanism of action c Dose adjustment/avoidance for renal c Cardiovascular safety disease c Lower cost Bile acid sequestrants c Colesevelamb c ? ↓ Hepatic glucose c No hypoglycemia c Constipation Low–intermediate production c ↓ LDL-C c ↑ Triacylglycerols c ? ↑ Incretin levels c May ↓ absorption of other

medications Care Diabetes c Intermediate expense Continued on p. 15 care.diabetesjournals.org

Table 2—Continued Medications/therapies Class in class Primary physiological action(s) Advantages Disadvantages/adverse effects Efficacy

Dopamine-2 agonists c Quick-release c Modulates hypothalamic c No hypoglycemia c Headache/dizziness/syncope Low–intermediate bromocriptineb regulation of metabolism c ? ↓ ASCVD events c Nausea c ↑ Insulin sensitivity c Fatigue c Rhinitis c High cost Injectable medications Insulins Long acting (basal) c Degludec (U100, U200) c Activates insulin receptor c Nearly universal response c Hypoglycemia Very high c Detemir c ↑ Glucose disposal c Theoretically unlimited efficacy c Weight gain c Glargine (U100, U300) c ↓ Glucose production c Once-daily injection c Training requirements c Frequentdoseadjustmentforoptimal efficacy c High cost Intermediate acting c Human NPH c Activates insulin receptor c Nearly universal response c Hypoglycemia Very high (basal) c ↑ Glucose disposal c Theoretically unlimited efficacy c Weight gain c ↓ Glucose production c Less expensive than analogs c Training requirements c Often given twice daily c Frequent dose adjustment for optimal efficacy Rapid acting c Aspart (conventional and c Activates insulin receptor c Nearly universal response c Hypoglycemia Very high fast acting) c ↑ Glucose disposal c Theoretically unlimited efficacy c Weight gain c Lispro (U100, U200) c ↓ Glucose production c ↓ Postprandial glucose c Training requirements c Glulisine c May require multiple daily injections c Frequent dose adjustment for optimal efficacy c High cost Inhaled rapid acting c Human insulin inhalation c Activates insulin receptor c Nearly universal response c Spirometry (FEV1) required before High powderb c ↑ Glucose disposal c ↓ Postprandial glucose initiating, after 6 months, and c ↓ Glucose production c More rapid onset and shorter duration annually than rapid-acting analogs c Contraindicated in chronic lung disease c Not recommended in smokers c Hypoglycemia c Weight gain c Training requirements c May require multiple inhalations daily

Continued on p. 16 15 Associates and Davies 6CnessReport Consensus 16

Table 2—Continued Medications/therapies Class in class Primary physiological action(s) Advantages Disadvantages/adverse effects Efficacy

c Frequent dose adjustment for optimal efficacy; limited options in dosing interval c High cost c Respiratory side effects (e.g., bronchospasm, cough, decline in FEV1) Short acting c Human regular (U100, c Activates insulin receptor c Nearly universal response c Hypoglycemia Very high U500) c ↑ Glucose disposal c Theoretically unlimited efficacy c Weight gain c ↓ Glucose production c ↓ Postprandial glucose c Training requirements c Less expensive than analogs c Frequent dose adjustment for optimal efficacy c May require multiple daily injections Premixed c Many c Activates insulin receptor c Nearly universal response c Hypoglycemia Very high c ↑ Glucose disposal c Theoretically unlimited efficacy c Weight gain c ↓ Glucose production c Fewer injections than basal/bolus c Training requirements before every meal c Frequent dose adjustment for optimal c Recombinant human analogs efficacy are less expensive c High cost (except human insulin premix) c Can lead to obligate eating GLP-1 RA Shorter acting c Exenatide c Glucose dependent: c No hypoglycemia as monotherapy c Frequent GI side effects that may be Intermediate–high c Lixisenatide ↑ Insulin secretion c ↓ Weight transient ↓ Glucagon secretion c Excellent postprandial glucose c Modestly ↑ heart rate c Slows gastric emptying efficacy for meals after injections c Training requirements c ↑ Satiety c Improves cardiovascular risk c Dose adjustment/avoidance in renal factors disease c Acute pancreatitis (rare/uncertain) c Very high cost Longer acting c Dulaglutide c Glucose dependent: c No hypoglycemia as monotherapy c GI side effects, including gallbladder High–very high c Exenatide extended- ↑ Insulin secretion c ↓ Weight disease release ↓ Glucagon secretion c ↓ Postprandial glucose excursions c Greater ↑ heart rate c Liraglutide c ↑ Satiety c Improves cardiovascular risk factors c Training requirements c Semaglutide c ↓ MACE with some agents (see text) c Dose adjustment/avoidance for some c ↓ Albuminuria with some agents agents in renal disease (see text) c Acute pancreatitis (rare/uncertain) Continued on p. 17 ibtsCare Diabetes care.diabetesjournals.org

Table 2—Continued Medications/therapies Class in class Primary physiological action(s) Advantages Disadvantages/adverse effects Efficacy

c Greater lowering of fasting c C-cell hyperplasia/medullary thyroid glucose vs. short-acting preparations tumors (rare/uncertain; observed in c Once-weekly dosing (except animals only) liraglutide, which is daily) c Very high cost Other injectables Amylin mimetics c Pramlintideb c ↓ Glucagon secretion c ↓ Postprandial glucose excursions c Hypoglycemia Intermediate c Slows gastric emptying c ↓ Weight c Frequent dosing schedule c ↑ Satiety c Training requirements c Frequent GI side effects c Very high cost Fixed-dose c Liraglutide/degludec c Combined activities of c Enhanced glycemic efficacy vs. components c Less weight loss than GLP-1 receptor Very high combination of GLP-1 c Lixisenatide/glargine components c Reduced adverse effects (e.g., GI, agonist alone RA and basal insulin hypoglycemia) c Very high cost analogs vs. components Weight loss medications c Lorcaserinb c Reduced appetite c Mean 3–9 kg weight loss vs. placebo c High discontinuation rates from side Intermediate c Naltrexone/bupropion c Fat malabsorption (orlistat) effects c Orlistat c ,50% achieve $5% weight loss c Phentermine/ c Drug-specific side effects topiramateb c Limited durability c Liraglutide 3 mg c High cost Metabolic surgery c VSG c Restriction of food intake (all) c Sustained weight reduction c High initial cost Very high c RYGB c Malabsorption (RYGB, BPD) c ↑ Rate of remission of diabetes c ↑ Risk for early and late surgical c Adjustable gastric band c Changes in hormonal and c ↓ Number of diabetes drugs complications c BPD possibly neuronal signaling c ↓ Blood pressure c ↑ Risk for reoperation (VSG, RYGB, BPD) c Improved lipid metabolism c ↑ Risk for dumping syndrome c ↑ Nutrient and vitamin malabsorption c ↑ Risk for new-onset depression c ↑ Risk for new-onset opioid use c ↑ Risk for gastroduodenal ulcer c ↑ Risk for hypoglycemia c ↑ Risk for alcohol use disorder

More details available in ADA’s Standards of Medical Care in Diabetesd2018 (3). Glucose-lowering efficacy of drugs by change in HbA1c: .22 mmol/mol (2%) very high, 11–22 mmol/mol (1–2%) high, 6–11 mmol/mol (0.5–1.5%) intermediate, ,6 mmol/mol (0.5%) low. aNot licensed in the U.S. for type 2 diabetes. bNot licensed in Europe for type 2 diabetes. BPD, biliopancreatic diversion; aisadAscae 17 Associates and Davies DKA, diabetic ketoacidosis; FEV1, forced expiratory volume in 1 s on pulmonary function testing; GI, gastrointestinal; GLP-1 RA, glucagon-like peptide 1 receptor agonist; HDL-C, HDL-cholesterol; LDL-C, LDL-cholesterol; RYGB, Roux-en-Y gastric bypass; VSG, vertical sleeve gastroplasty; T2DM, type 2 diabetes mellitus; UTI, urinary tract infection. 18 Consensus Report Diabetes Care

within 6 years of diagnosis, evidence- in reducing HbA1c by about 6.6 mmol/mol monotherapy, and the potential for based weight management programs are (0.6%) (81–84). Of these modalities, some some weight loss. Some studies have often successful. evidence suggests that aerobic exer- suggested a benefit for preventing The most effective nonsurgical strat- cise and the combination of aerobic ex- CVD (98), but this has not been supported egies for weight reduction involve food ercise and resistance training may be by the results of a recent meta-analysis substitution and intensive, sustained more effective than resistance training (99). However, metformin may lower risk counseling (e.g., 12–26 individual coun- alone (85), but this remains controver- for cardiovascular mortality compared seling sessions over 6–12 months). sial. When considering exercise interven- with sulfonylurea therapy (100). Rare Among adults with type 2 diabetes, tions, special considerations are required cases of lactic acidosis have been re- meal replacement (825–853 kcal/day for individuals with CVD, uncontrolled ported, usually in the setting of severe [3,450–3,570 kJ/day] formula diet for retinopathy or nephropathy, and severe illness or acute kidney injury. Therefore, 3–5 months) followed by gradual rein- neuropathy. A wide range of physical metformin should be omitted in the set- troduction of food and intensive coun- activity, including leisure time activities ting of severe illness, vomiting, or de- seling resulted in 9-kg placebo-adjusted (e.g., walking, swimming, gardening, jog- hydration. Metformin may result in lower weight loss at 1 year and high rates of ging, tai chi, and yoga) can significantly serum vitamin B12 concentration; there- diabetes remission (46% vs. 4%; odds reduce HbA1c (86–90). In general, super- fore, periodic monitoring and supple- ratio [OR] 19.7 [95% CI 7.8, 49.8]) com- vision of exercise and motivational strat- mentation is generally recommended paredwithbestusualpractice(75).In egies, such as monitoring using a step if levels are deficient, particularly in terms of intensive behavioral interven- counter, can improve the effect of exer- those with anemia or neuropathy (101). tions, the Action for Health in Diabetes cise on HbA1c compared with advice Because of its high efficacy in lowering (Look AHEAD) trial (76) randomized alone (84,91). The combination of dietary HbA1c, good safety profile, and low cost, 5,145 overweight or obese patients change for weight reduction and physical metformin remains the first-line medica- with type 2 diabetes to an intensive exercise improves hyperglycemia and tion for management of type 2 diabetes. lifestyle program that promoted energy reduces cardiovascular risk factors more restriction, incorporating meal replace- than dietary interventions or physical SGLT2 Inhibitors ments to induce and sustain weight loss, activity alone (92). SGLT2 inhibitors are oral medications along with increased physical activity that reduce plasma glucose by enhancing compared with standard diabetes edu- Medications for Lowering Glucose urinary excretion of glucose (102). The cation and support in the control group. Metformin glucose-lowering efficacy of these med- After 9.6 years, weight loss was greater Metformin is an oral medication that ications is dependent on renal function. in the intervention group (8.6% vs. 0.7% reduces plasma glucose via multiple mech- Initiation and continuation of SGLT2 in- at 1 year; 6.0% vs. 3.5% at study end; anisms. It is available as an immediate- hibitors are restricted by eGFR and re- both P , 0.05). HbA1c also fell in the release formulation that is typically quire intermittent monitoring of renal intervention group despite less use of administered twice a day and as ex- function (refer to European Medicines glucose-lowering medications. Cardio- tended-release formulations for once- Agency and U.S. Food and Drug Ad- vascular event rates were not reduced, daily or twice-daily administration. The ministration prescribing information but there were numerous other bene- formulations are equally effective with for current recommendations). These fits. In a 12-month trial, 563 adults with no consistent differences in side effect medications are of high efficacy in low- type 2 diabetes who were randomized profile (93). Dosages of immediate- ering glucose in the setting of normal to Weight Watchers compared with stan- release metformin start at 500 mg once renal function (51,52,103). All SGLT2 in- dard care had a 2.1% net weight loss or twice a day with meals and should be hibitors are associated with a reduction (24.0% vs. 21.9%; P , 0.001), a 5.3 increased as tolerated to a target dosage in weight and blood pressure. Alone or mmol/mol (23.5 vs. 11.8 mmol/mol; of 1,000 mg twice a day. The maximum with metformin, they do not increase the P 5 0.020) net absolute improvement in daily dose is 2,550 mg in the U.S. and risk for hypoglycemia. Empagliflozin and HbA1c (0.48% [20.32% vs. 10.16%]), and 3,000 mg in the European Union, though canagliflozin have cardiac and renal ben- a greater reduction in use of glucose- doses above 2,000 mg are generally efits in patients with established or at lowering medications (226% vs. 112%; associated with little additional efficacy high risk of ASCVD. Cardiac and renal P , 0.001) (77). Similar programs have and poorer tolerability (94). Gastroin- benefits have been demonstrated down resulted in a net 3-kg weight loss over testinal symptoms are common and to an eGFR of 30 mL min–1 [1.73 m]–2, 12–18 months (78–80). dose dependent, and may improve over though currently none of the SGLT2 time or with dose reduction. Metformin inhibitors have been approved for Physical Activity should not be used in patients with an use by regulators at an eGFR below –1 –2 –1 –2 Consensus recommendation eGFR ,30 mL min [1.73 m] and dose 45 mL min [1.73 m] (see the section c Increasing physical activity improves reduction should be considered when “Recommended Process for Glucose- – – glycemic control and should be en- the eGFR is ,45 mL min 1 [1.73 m] 2 Lowering Medication Selection: Where couraged in all people with type 2 (95–97). Caution should be taken when Does New Evidence From Cardiovascular diabetes. conditions are present that may re- Outcomes Trials Fit In?”) (51,52,61). The duce eGFR. Advantages of metformin classisassociatedwithincreasedriskfor Aerobic exercise, resistance training, and include its high efficacy, low cost, min- mycotic genital infections (mostly vagi- the combination of the two are effective imal hypoglycemia risk when used as nitis in women, balanitis in men) (51, care.diabetesjournals.org Davies and Associates 19

52,104,105). Case reports of diabetic receptor agonists reduce weight (110); Process for Glucose-Lowering Medica- ketoacidosis with SGLT2 inhibitors in the reduction ranges from about 1.5 kg tion Selection: Where Does New Evi- type 2 diabetes continue to raise con- to 6.0 kg over about 30 weeks of ther- dence From Cardiovascular Outcomes cern, though increased rates have not apy (110,117). Liraglutide and semaglutide Trials Fit In?”). been confirmed in large trials (102,106). have been shown to improve cardiovas- Therefore, the SGLT2 inhibitors should cular outcomes (47,48) (see the section Thiazolidinediones be used with caution and appropriate “Recommended Process for Glucose- Thiazolidinediones (TZDs) (pioglitazone patient education should be provided for Lowering Medication Selection: Where and rosiglitazone) are oral medications those with insulin deficiency. SGLT2 in- Does New Evidence From Cardiovascu- that increase insulin sensitivity and are hibitors have been associated with an lar Outcomes Trials Fit In?”). The most of high glucose-lowering efficacy (129– increased risk of acute kidney injury, common side effects of GLP-1 receptor 131). TZDs increase HDL-cholesterol dehydration, and orthostatic hypoten- agonists are nausea, vomiting, and di- (132,133), and pioglitazone has been sion; caution should be taken when arrhea, though these tend to diminish shown to reduce cardiovascular end SGLT2 inhibitors are used in combina- over time. GLP-1 receptor agonists have points (132,134–138) and hepatic steato- tion with diuretics and/or ACE inhibi- minimal risk for hypoglycemia, but may hepatitis (139), but without conclusive tors and angiotensin receptor blockers. increase the hypoglycemic potential evidence for benefit. TZDs are associated Canagliflozin has been associated with of insulin and sulfonylureas when com- with the best evidence among glucose- increased risk for lower-limb amputa- bined with those medications (118). lowering medications for glycemic du- fl tion (6.3 canagli ozin vs. 3.4 per 1,000 Contrary to early signals, GLP-1 receptor rability (140). However, these notable patient-years with placebo after 3.1 agonists do not seem to substan- benefits must be balanced with safety years; HR 1.97 [95% CI 1.41, 2.75]) (52). tially increase risk for pancreatitis, pan- concerns regarding fluid retention and Similarly, fracture risk has been reported creatic cancer, or bone disease (119). congestive heart failure (136,140,141), fl with canagli ozin (15.4 vs. 11.9 partici- They are associated with increased risk weight gain (132,136,140–142), bone pants with fracture per 1,000 patient- of gallbladder events (120). Semaglu- fracture (143,144), and, possibly, bladder years; HR 1.26 [95% CI 1.04, 1.52]) (52). It tide was associated with increased ret- cancer (145). Lower-dose therapy (e.g., is uncertain whether amputation and inopathy complications in the SUSTAIN pioglitazone 15–30 mg) mitigates weight fractures are class effects. 6 trial (HR 1.76, 95% CI 1.11, 2.78), largely gain and edema, but the broader ben- among those with baseline retinopa- efits and harms of low-dose TZD therapy GLP-1 Receptor Agonists thy who had rapid improvement of have not been evaluated. GLP-1 receptor agonists are currently glycemic control (48). While this ob- delivered by subcutaneous injection. servation remains unexplained, this is These medications stimulate insulin se- Sulfonylureas also a recognized effect of intensifica- Sulfonylureas are oral medications that cretion and reduce glucagon secretion tion of glycemic control with insulin. lower glucose by stimulating insulin se- in a glucose-dependent manner, im- cretion from pancreatic b-cells. They are prove satiety, and promote weight loss DPP-4 Inhibitors inexpensive, widely available, and have (107,108). Structural differences among DPP-4 inhibitors are oral medications that high glucose-lowering efficacy (146). GLP-1 receptor agonists affect dura- increase insulin secretion and reduce glu- tion of action, and their formulation cagon secretion in a glucose-dependent Sulfonylureas were used as part of the and dosing may affect efficacy for glu- manner. They have moderate glucose- glucose-lowering regimen in the UK Pro- cose-lowering and weight reduction as lowering efficacy (121,122). DPP-4 inhib- spective Diabetes Study (UKPDS) (147) well as side effect profile and cardio- itors are well tolerated, have a neutral andActioninDiabetesandVascular vascular effects (109). Dulaglutide, effect on weight, and have minimal risk of Disease: Preterax and Diamicron MR exenatide extended-release, and sema- hypoglycemia when used as monotherapy Controlled Evaluation (ADVANCE) (148) glutide are administered once weekly (123). When added to sulfonylurea ther- trials, which both demonstrated reduc- (108,109). Liraglutide and lixisenatide apy, however, the risk for hypoglycemia is tions in microvascular complications. Sul- are administered once daily, and exena- increased 50% compared with sulfonyl- fonylureas are associated with weight tide is available in a twice-daily formu- urea therapy alone (124). The recom- gain and risk for hypoglycemia and down lation. GLP-1 receptor agonists have high mended dose for each DPP-4 inhibitor titration of dose to reduce the risk of glucose-lowering efficacy, but with var- is determined and needs to be adjusted hypoglycemia results in higher HbA1c iation within the drug class (110,111). based on renal function; linagliptin is the (146,149,150). Sulfonylureas are known Evidence suggests that the effect may exception as it has minimal renal excretion. to be associated with a lack of durable be greatest for semaglutide once weekly, Rare but increased rates of pancreatitis effect on glucose lowering (144,151). The followed by dulaglutide and liraglutide, (125) and musculoskeletal side effects weight gain associated with sulfonyl- closely followed by exenatide once have been reported (126). CVOTs dem- ureas is relatively modest in large cohort weekly, and then exenatide twice daily onstrated the cardiovascular safety but studies and the incidence of severe hy- andlixisenatide(110,112–116).The short- no cardiovascular benefit of three DPP-4 poglycemia is lower than with insulin acting medications exenatide twice daily inhibitors (saxagliptin, alogliptin, and (152). Important differences among sul- and lixisenatide have greater postpran- sitagliptin) as well as imbalances regard- fonylureas affect both safety and effi- dial effects, at least after the meals with ing HF for saxagliptin and alogliptin cacy. Glibenclamide (known as glyburide which they are administered. All GLP-1 (127,128) (see the section “Recommended in the U.S. and Canada) has a higher risk 20 Consensus Report Diabetes Care

of hypoglycemia compared with other of injection, and glycemic targets, has a dose of glargine for equivalent efficacy sulfonylureas (153). Glipizide, glimepiride, greater impact on the adverse effects of (165–167). and gliclazide may have a lower risk for insulin than differences among insulin Not all patients have their blood glu- hypoglycemia compared with other sul- formulations. cose adequately controlled with basal fonylureas (152,154). Adverse cardio- insulin. In particular, patients with higher vascular outcomes with sulfonylureas in Basal Insulin. Basal insulin refers to pretreatment HbA1c, higher BMI, longer some observational studies have raised longer-acting insulin that is meant to duration of disease, and a greater num- concerns, although findings from recent cover the body’s basal metabolic insulin ber of oral glucose-lowering medications systematic reviews have found no in- requirement (regulating hepatic glucose are more likely to require intensified crease in all-cause mortality compared production), in contrast to bolus or pran- therapy (168). with other active treatments (152). As dial insulin, which is meant to reduce newer-generation sulfonylureas appear glycemic excursions after meals. Basal Other Insulin Formulations. Short- and to confer a lower risk of hypoglycemia insulin is the preferred initial insulin rapid-acting insulin formulations admin- and have favorable cost, efficacy, and formulationinpatientswithtype2 istered at mealtime are generally used safety profiles, sulfonylureas remain a rea- diabetes. Options include once- or twice- to intensify basal insulin therapy in sonable choice among glucose-lowering daily administration of intermediate- patients not meeting glycemic targets. medications, particularly when cost is an acting NPH or detemir insulin and the Options include human regular insulin, important consideration. Patient educa- once-daily administration of glargine various analogs (aspart, glulisine, and tion and use of low or variable dosing with (U100 or U300) or degludec (U100 or lispro), formulations (faster insulin as- later generation sulfonylureas may be U200). Long-acting insulin analogs (de- part, lispro U200), biosimilars (lispro), used to mitigate the risk of hypoglyce- gludec [U100 or U200], glargine [U100 and insulins with different routes of mia.Greatestcautioninthisregardis and U300], detemir) have a modestly administration (inhaled). Rapid-acting warranted for people at high risk of lower absolute risk for hypoglycemia com- insulin analogs have a modestly lower hypoglycemia, such as older patients pared with NPH insulin, but cost more risk for hypoglycemia compared with and those with CKD. (157–160). However, in real-world set- human regular insulin but at a higher tings where patients are treated to con- cost. Various premixed formulations of Insulin ventional treatment targets, initiation of human and analog insulins are available Numerous formulations of insulin are NPH compared with determir or glargine andcontinuetobewidelyusedinsome available with differing durations of ac- U100 did not increase hypoglycemia- regions, though they tend to have an tion. Human insulins (NPH, regular [R], related emergency department visits or increased risk of hypoglycemia as com- and premixed combinations of NPH and hospital admissions (161). When com- pared with basal insulin alone (Table 2 R) are recombinant DNA-derived human paring human and analog insulins, cost and Fig. 7). insulin, while insulin analogs have been differences can be large while differ- designed to change the onset or duration ences in hypoglycemia risk are mod- Other Glucose-Lowering Medications of action. The main advantage of insulin est and differences in glycemic efficacy Other oral glucose-lowering medications over other glucose-lowering medications minimal. (i.e., meglitinides, a-glucosidase inhibitors, is that insulin lowers glucose in a dose- Degludec is associated with a lower colesevelam, quick-release bromocriptine, dependent manner over a wide range, to risk of severe hypoglycemia compared pramlintide) are not used commonly in almost any glycemic target as limited by with glargine U100 insulin when target- the U.S. and some are not licensed at all in hypoglycemia. Older formulations of in- ing intensive glycemic control in patients Europe. No major new scientific informa- sulin have also demonstrated reduction with long-standing type 2 diabetes at tion on these medications has emerged in microvascular complications and with high risk of CVD; absolute incidence in recent years. Their basic characteristics long-term follow-up, all-cause mortality, difference of 1.7% over 2 years (rate are listed in Table 2. and diabetes-related death (147,155). ratio 0.60; P , 0.001 for superiority; Beyond hypoglycemia, the disadvan- OR 0.73; P , 0.001 for superiority) Obesity Management Beyond Lifestyle tages of insulin include weight gain and (162). Biosimilar formulations are now Intervention the need for injection, frequent titration available for glargine with similar efficacy Medications for Weight Loss for optimal efficacy, and glucose moni- profile and lower cost (163). No insulin Several clinical practice guidelines rec- toring (156). has been shown to reduce risk for CVD ommend weight-loss medications as The effectiveness of insulin is highly (156), but data suggest that glargine an optional adjunct to intensive lifestyle dependent on its appropriate use; pa- U100anddegludecdonotincrease management for patients with obesity, tient selection and training; adjustment risk for MACE (162,164). particularly if they have diabetes of dose for changes in diet, activity, or Concentrated formulations of de- (169–171). Others do not (172). Several weight; and titration to acceptable, gludec (U200) and glargine (U300) are medications and medication combina- safe glucose targets. Formulations of available that allow injection of a re- tions approved in the U.S. or Europe intermediate- and long-acting insulin duced volume, a convenience for patients for weight loss have been found to have different timings of onset, durations on higher doses. Glargine U300 is asso- improve glucose control in people with of action, and risks of hypoglycemia. ciated with a lower risk of nocturnal diabetes (173,174). One glucose-lowering However, the way in which insulin is ad- hypoglycemia compared with glargine medication, liraglutide, is also ap- ministered, including the dose, timing U100 but requires a 10–14% higher proved for the treatment of obesity care.diabetesjournals.org Davies and Associates 21

Figure 7—Intensifying to injectable therapies. FRC, fixed-ratio combination; GLP-1 RA, glucagon-like peptide 1 receptor agonist; FBG, fasting blood glucose; FPG, fasting plasma glucose; max, maximum; PPG, postprandial glucose. 22 Consensus Report Diabetes Care

at a higher dose (175). Cost, side effects, use, including drug and alcohol use and preferredtoinitialcombination and modest efficacy limit the role of cigarette smoking (186). People with therapy. pharmacotherapy in long-term weight diabetes presenting for metabolic sur- management. gery also have increased rates of depres- In most patients, type 2 diabetes is a sion and other major psychiatric disorders progressive disease, a consequence gen- Metabolic Surgery (187). These factors should be assessed erally attributed to a steady decline of preoperatively and during follow-up. Met- Consensus recommendation insulin secretory capacity. The practical c abolic surgery should be performed in Metabolic surgery is a recommended impact of gradual loss of b-cell function is high-volume centers with multidisciplin- treatment option for adults with that achieving a glycemic target with ary teams that are experienced in the type 2 diabetes and 1)aBMI monotherapy is typically limited to sev- $ 2 $ 2 management of diabetes and gastrointes- 40.0 kg/m (BMI 37.5 kg/m eral years. Stepwise therapy (i.e., adding tinalsurgery.Long-termlifestylesupport in people of Asian ancestry) or 2) medications to metformin to maintain – 2 – and routine monitoring of micronutrient a BMI of 35.0 39.9 kg/m (32.5 HbA at target) is supported by clinical 2 and nutritional status must be provided 1c 37.4 kg/m in people of Asian an- trials (3). While there is some support for to patients after surgery (188,189). cestry) who do not achieve durable initial combination therapy due to the weight loss and improvement in PUTTING IT ALL TOGETHER: greater initial reduction of HbA1c than comorbidities with reasonable non- can be provided by metformin alone surgical methods. STRATEGIES FOR IMPLEMENTATION (190,191), there is little evidence that this approach is superior to sequential Metabolic surgery is highly effective in For an increasing number of patients, addition of medications for maintaining improving glucose control (176–178) presence of specific comorbidities (e.g., glycemic control or slowing the progres- and often produces disease remission ASCVD, HF, CKD, obesity), safety con- sion of diabetes. However, since the (179–182). The effects can be sustained cerns (e.g., risk of hypoglycemia), or health absolute effectiveness of most oral med- for at least 5 years (177,182). Benefits care environment (e.g., cost of medica- ications rarely exceeds an 11 mmol/mol include a reduction in the number of tions) mandate a specificapproachto (1%) reduction in HbA , initial combi- glucose-lowering medications needed to the choice of glucose-lowering medica- 1c nation therapy may be considered in achieve glycemic targets (178,179). tion. These are considered in Figs. 2–6. patients presenting with HbA levels Several clinical practice guidelines For patients not reaching their target 1c more than 17 mmol/mol (1.5%) above and position statements recommend HbA1c, it is important to re-emphasize their target. Fixed-dose formulations can consideration of metabolic surgery lifestyle measures, assess adherence, and improve medication adherence when as a treatment option for adults with arrange timely follow-up (e.g., within combination therapy is used (192), and type 2 diabetes and 1)aBMI$40.0 3–6 months) (Fig. 1). may help achieve glycemic targets more kg/m2 (BMI $37.5 kg/m2 in people of Initial Monotherapy rapidly (100). Potential benefits of com- Asian ancestry) or 2)aBMIof35.0– bination therapy need to be weighed 39.9 kg/m2 (32.5–37.4 kg/m2 in people Consensus recommendation against the exposure of patients to mul- of Asian ancestry) who do not achieve c Metformin is the preferred initial tiple medications and potential side ef- durable weight loss and improvement in glucose-lowering medication for fects, increased cost, and, in the case of comorbidities with reasonable nonsurgi- most people with type 2 diabetes. fixed combination medications, less flex- cal methods (65,183). Because baseline ibility in dosing. BMI does not predict surgical benefits on Metformin remains the preferred option glycemia or hard outcomes and the im- for initiating glucose-lowering medica- Choice of Glucose-Lowering provement in glycemic control occurs early tion in type 2 diabetes and should be Medication After Metformin through weight-independent mechanisms added to lifestyle measures in newly Consensus recommendation (183), metabolic surgery may be considered diagnosed patients. This recommenda- fi c The selection of medication added for those with a BMI of 30.0–34.9 kg/m2 tion is based on the ef cacy, safety, to metformin is based on patient (27.5–32.4 in people of Asian ancestry) tolerability, low cost, and extensive preference and clinical character- who do not achieve durable weight loss clinical experience with this medication. 5 istics. Important clinical charac- and improvement in comorbidities with Results from a substudy of UKPDS (n fi teristics include the presence of reasonable nonsurgical methods. 342) showed bene ts of initial treatment established ASCVD and other co- Adverse effects of bariatric surgery, with metformin on clinical outcomes morbidities such as HF or CKD; the which vary by procedure, include surgical related to diabetes, with less hypoglyce- risk for specific adverse medication complications (e.g., anastomotic or sta- mia and weight gain than with insulin or effects, particularly hypoglycemia ple line leaks, gastrointestinal bleeding, sulfonylureas (98). and weight gain; as well as safety, intestinal obstruction, the need for re- Initial Combination Therapy tolerability, and cost (Figs. 2–6). operation), late metabolic complications Compared With Stepwise Addition of (e.g., protein malnutrition, mineral defi- Glucose-Lowering Medication ciency, vitamin deficiency, anemia, hypo- As detailed in the “Medications for Low- Consensus recommendation glycemia), and gastroesophageal reflux ering Glucose” section, the glucose- c The stepwise addition of glucose- (184,185). Patients who undergo meta- lowering medications that can be added lowering medication is generally bolic surgery may be at risk for substance to metformin have distinct profiles of care.diabetesjournals.org Davies and Associates 23

action, efficacy, and adverse effects Intensification Beyond Two minimal benefits or if harm outweighs (100,193). The early introduction of basal Medications any benefit. In particular, ceasing or reduc- insulin is well established, in particu- ing the dose of medications that have Consensus recommendation lar when HbA levels are very high an increased risk of hypoglycemia is im- 1c c Intensification of treatment be- (.97 mmol/mol [.11%]), symptoms yond dual therapy to maintain portant when any new glucose-lowering of hyperglycemia are present, or there glycemic targets requires consider- treatment (lifestyle or medication) is is evidence of ongoing catabolism (e.g., ation of the impact of medication started (Fig. 7) (40). HbA1c levels below weight loss). This constellation of symp- side effects on comorbidities, as 48 mmol/mol (6.5%) or substantially be- toms can occur in type 2 diabetes but well as the burden of treatment low the individualized glycemic target suggest insulin deficiency and raise the and cost. should prompt consideration of stopping possibility of autoimmune (type 1) or or reducing the dose of medications with pancreatogenic diabetes in which insu- The lack of a substantial response to one risk of hypoglycemia or weight gain. lin would be the preferred therapy. or more noninsulin therapies should Addition of Injectable Medications While this remains the usual strategy raise the issue of adherence and, in those for patients when HbA1c levels are very with weight loss, the possibility that the Consensus recommendation high, SGLT2 inhibitors (194) and GLP-1 patient has autoimmune (type 1) or c In patients who need the greater receptor agonists (195) have demon- pancreatogenic diabetes. However, it glucose-lowering effect of an in- strated efficacy in patients with HbA1c is common in people with long-standing jectable medication, GLP-1 recep- levels exceeding 75 mmol/mol (9%), with diabetes to require more than two tor agonists are the preferred the additional benefits of weight reduc- glucose-lowering agents, often including choice to insulin. For patients tion and reduced risk of hypoglycemia. insulin. Compared with the knowledge with extreme and symptomatic hy- Evidence from clinical trials supports base guiding dual therapy of type 2 di- perglycemia, insulin is recom- the use of several of the SGLT2 inhibitors abetes, there is less evidence guiding mended (Fig. 7). and GLP-1 receptor agonists as add-on these choices (205). In general, inten- therapy for people with type 2 diabetes sification of treatment beyond two See the “Insulin” and “Basal Insulin” “ withan HbA1c .53 mmol/mol (.7%) and medications follows the same general sections in Medications for Lowering established CVD (48,51,52). However, principles as the addition of a second Glucose” for more medication details. since only 15–20% of patients with medication, with the assumption that the Patients often prefer combinations of type 2 diabetes conform to the charac- efficacy of third and fourth medications oral medications to injectable medications. teristics of patients in these trials, other will be generally less than expected. The range of combinations available with clinical features need to be considered No specific combination has demon- current oral medications allows many peo- in the majority when selecting second strated superiority except for those ple to reach glycemic targets safely. How- medications to add to metformin (Figs. that include insulin and GLP-1 receptor ever, there is currently no evidence that 2–6) (149,196–204). agonists that have broad ranges of gly- any single medication or combination has Sulfonylureas and insulin are associ- cemic efficacy. As more medications durable effects and, for many patients, in- ated with an increased risk for causing are added, there is an increased risk jectable medications become necessary hypoglycemia and would not be pre- of adverse effects. It is important to within 5–10 years of diabetes diagnosis. ferred for patients in whom this is a consider medication interactions and Evidence from trials comparing GLP-1 concern. Furthermore, hypoglycemia is whether regimen complexity may be- receptor agonists and insulin (basal, pre- distressing and so may reduce treatment come an obstacle to adherence. Finally, mixed, or basal-bolus) shows similar or adherence (Fig. 5). For patients prioritiz- with each additional medication comes even better efficacy in HbA1c reduction ing weight loss or weight maintenance increased costs, which can affect patient (212,213). GLP-1 receptor agonists (Fig. 4), important considerations include burden, medication-taking behavior, and have a lower risk of hypoglycemia and the weight reduction associated with medication effectiveness (193,205–211). are associated with reductions in body SGLT2 inhibitors and GLP-1 receptor While most patients require intensi- weight compared with weight gain with agonists, the weight neutrality of DPP-4 fication of glucose-lowering medica- insulin (212,214). Some GLP-1 receptor inhibitors, and the weight gain associ- tions, some require medication reduction agonists allow for once-weekly injec- ated with sulfonylureas, basal insulin, or discontinuation of medication, partic- tions, as opposed to daily or more often and TZDs. An important consideration ularly if the therapy is ineffective or is for insulin. Based on these consider- for society in general and for many exposing patients to a higher risk of side ations, a GLP-1 receptor agonist is the patients in particular is the cost of effects such as hypoglycemia or when preferred option in a patient with a def- medications; sulfonylureas, pioglita- glycemic goals have changed due to a inite diagnosis of type 2 diabetes who zone, and recombinant human insulins change in clinical circumstances (e.g., needs injectable therapy. However, the are relatively inexpensive, although development of comorbidities or even tolerability and high cost of GLP-1 re- their cost may vary across regions. healthy aging). A guiding principle is that ceptor agonists are important limita- Short-term acquisition costs, longer- for all therapies the response should be tions to their use. If additional glucose term treatment cost, and cost-effective- reviewed at regular intervals, including lowering is needed despite therapy with ness should be considered in clinical the impact on efficacy (HbA1c, weight) a long-acting GLP-1 receptor agonist, the decision making when data are available and safety; the therapy should be addition of basal insulin is a reason- (Fig. 6). stopped or the dose reduced if there are able option (215,216). 24 Consensus Report Diabetes Care

Alternatively, the addition of insulin to meta-analysis that studied the combi- a modestly greater reduction in HbA1c oral medication regimens is well estab- nation of either SGLT2 inhibitors or DPP- with basal-prandial regimens com- lished. In particular, using basal insulin 4 inhibitors with insulin, the SGLT2 pared with biphasic insulin regimens, in combination with oral medications is inhibitor–insulin combination was associ- but at the expense of greater weight effective, and has less hypoglycemia and ated with a greater reduction in HbA1c, gain (233–235). While still commonly weight gain than combinations using an advantage in terms of body weight used, we do not generally advocate premixed insulin formulations or pran- and no increase in the rates of hypo- premixed insulin regimens, particularly dial insulin (217). A standard approach glycemia (222,223). Depending on those administered three times daily, for optimizing basal insulin regimens is baseline HbA1c,glycemicprofile, and for routine use when intensifying insulin to titrate the dose based on a target fast- individual response, the insulin dose regimens (Fig. 7). ing glucose concentration, which is a sim- mayneedtobereducedtoprevent Continuous insulin infusion using in- ple index of effectiveness. Either NPH hypoglycemia when adding an SGLT2 sulin pumps may have a role in a small insulin or long-acting insulin analogs are inhibitor. minority of people with type 2 diabetes efficacious for controlling fasting glu- The combination of basal insulin and (236). cose, although basal analog formulations a GLP-1 receptor agonist has high effi- show reduced risks of hypoglycemia, cacy, with recent evidence from clinical Access and Cost particularly overnight, when titrated to trials demonstrating the benefits of Consensus recommendation the same fasting glucose target as NPH this combination to lower HbA1c and c Access, treatment cost, and in- insulin (157,218). limit weight gain and hypoglycemia surance coverage should all be fi Beyond Basal Insulin compared with intensi ed insulin reg- considered when selecting glucose- imens (224,225). Most data come from lowering medications. Consensus recommendation studies in which a GLP-1 receptor ago- c Patients who are unable to main- nist is added to basal insulin. However, The availability of glucose-lowering tain glycemic targets on basal in- there is evidence that insulin added to a medications, patient support systems, sulin in combination with oral GLP-1 receptor agonist can also effectively and blood glucose-monitoring devices medications can have treatment lower HbA1c, although some weight gain can differ worldwide, depending on a intensified with GLP-1 receptor ag- results (215). Fixed-ratio combinations region’s economy, culture, and health onists, SGLT2 inhibitors, or prandial of insulin and GLP-1 receptor agonists care system. Cost of and access to newer insulin (Figs. 7 and 8). are available and can decrease the medications and insulin remain impor- number of injections compared with tant issues throughout the world. Al- It has become common practice to ap- administering the medications sepa- though the economics of diabetes care proach insulin use in people with type 2 rately (226–228). is complex and broadly includes the costs diabetes by following the established A final approach to glycemic manage- to society of diabetic complications and paradigms developed for those with ment when basal insulin plus oral med- long-term outcomes, the cost of drugs type 1 diabetes. This includes multiple ications is insufficient to achieve HbA1c and the affordability of treatment are daily injections with doses of insulin targets is intensified insulin regimens often the primary basis for decision analogs before meals that are adjusted (Figs. 7 and 8). DSMES focused on insulin making. Within health care systems, var- based on ambient blood glucose and therapy is particularly helpful when in- ianceinmedicationcoverage is based on meal constituents. While this is reason- tensified insulin therapy is considered. different assessments of cost-effectiveness. able for people with type 2 diabetes Referral to a diabetes specialist team This results in huge disparities in the who are lean, insulinopenic, and sensi- should be considered in cases where cost of new and old glucose-lowering tive to exogenous insulin, it ignores the the provider is uncomfortable or unfa- medications in some countries, limiting substantial differences in pathophysiol- miliar with intensification, poor out- access to the full range of diabe- ogy between most people with type 2 comes continue despite intensification, tes therapies in large segments of the diabetes and type 1 diabetes. Most peo- or patients have other issues that com- population, and creating a two-tiered ple with type 2 diabetes are obese and plicate intensification. Intensified insulin system of treatment. Since glycemic insulin resistant, requiring much larger regimens include 1) one or more daily management remains a cornerstone of doses of insulin and experiencing lower injections of rapid- or short-acting insulin the prevention of diabetes complica- rates of hypoglycemia than those with before meals (prandial insulin) or 2) tions, these disparities raise questions type 1 diabetes. In patients with type 2 switching to one to three daily admin- of fairness, equity, and overall public diabetes, weight gain is a particularly istrations of a fixed combination of short- health. Nonetheless, the use of less problematic side effect of insulin use. and long-acting insulin (premixed or expensive agents, such as metformin, Recent evidence supports the effective- biphasic insulins) (229,230). When adding sulfonylureas, and human insulin, remain ness of combinations of insulin with prandial insulin, giving one injection with effective options (Figs. 2 and 6). Redou- glucose-lowering medications that do the largest meal of the day is a simple and bling lifestyle management efforts can not increase body weight. For example, safe approach (231). Over time, if glyce- also have great impact, but behavioral SGLT2 inhibitors can be added to insulin mic targets are not met with one dose of intervention and support can also be regimens to lower blood glucose levels prandial insulin daily, additional prandial costly, and socioeconomic barriers to without increasing insulin doses, weight injections can be added to other meals improving lifestyle are well described gain, or hypoglycemia (219–221). In a (232). Results of meta-analyses suggest (237). care.diabetesjournals.org Davies and Associates 25

Figure 8—Considering oral therapy in combination with injectable therapies. DKA, diabetic ketoacidosis; DPP-4i, dipeptidyl peptidase 4 inhibitor; GLP-1 RA, glucagon-like peptide 1 receptor agonist; SGLT2i, SGLT2 inhibitor; SU, sulfonylurea.

Emerging Technology to specific treatment recommendations KEY KNOWLEDGE GAPS There is an increasing call for the use of supported by real-time algorithms. Despite over 200 years of research on technology and telemedicine to improve Telemedicine incorporates multiple types lifestyle management of diabetes and patients’ health (238). Many types of of communication services, such as two- more than 50 years of comparative- inputs can be digitalized, such as blood way video, e-mail, texting, smartphones, effectiveness research in diabetes, glucose levels, time spent exercising, tablets, wireless monitors, decision sup- innumerable unanswerecd questions steps walked, energy ingested, medica- port tools, and other forms of telecom- regarding the management of type 2 tion doses administered, blood pressure, munication technologies. Results overall diabetes remain. In the context of and weight. Patterns in these variables suggest a modest improvement in glyce- our current consensus recommenda- can be identified by software, leading mic control (239,240). tions, the following is an incomplete 26 Consensus Report Diabetes Care

discussion of vexing issues that must be preference, and other patient-recorded approaches is enormous, evidence is des- addressed. outcome measures. Patients and other perately needed. Different models of Evolving areas of current investiga- stakeholders should have more input care are being implemented globally. tion will provide improvements in dia- into trial designs and outcomes. Prag- Defining optimal cost-effective approaches betes care and hold great hope for new matic designs will enhance generaliz- to care, particularly in the management treatments. ability of results and reduce cost. of patients (multimorbidity), is essential. Better application of “real-world ev- New questions arise from the recent c Implementation science. The tools idence” will complement randomized cardiovascular outcomes studies. Do available to prevent and treat diabetes trial evidence. the cardiovascular and renal benefits of are vastly improved. However, imple- c Drug development. New medications SGLT2 inhibitors and GLP-1 receptor ag- mentation of effective innovation has will require demonstration of broad onists demonstrated in patients with lagged behind. efficacy for glucose, comorbidities established CVD extend to lower-risk c Basic science. Our understanding of and/or complications, as well as safety patients? Is there additive benefitof the basic mechanisms of diabetes, the and tolerability to compete in the use of GLP-1 receptor agonists and development of complications, and marketplace. SGLT2 inhibitors for prevention of car- the treatment of both, though contin- c Complications. Steatohepatitis, HF, diovascular and renal events? If so, in uously advancing, has highlighted how nonalbuminuric CKD, chronic mental what populations? much we do not know. illness, and other emerging issues are Addressing these and other vital clin- c Personalized/precision medicine. complications in diabetes that may ical questions will require additional in- Though promising, these –omics and supplant classical microvascular and vestment in basic, translational, clinical, big data approaches addressing both macrovascular disease in importance and implementation research. More personal and environmental factors and impact. Understanding optimal time- and cost-efficient research para- and their interaction are largely un- diagnostic, screening, and treatment digms to address patient-centered end realized in diabetes care and will strategies is urgently needed. points will need to be developed through require large investments and coordi- regulatory reform and leveraging infor- nation to have impact. Other areas of importance include bet- matics and coordinated learning health c Informatics. The benefits and role ter segmentation of “type 2 diabetes,” as care systems. The increasing burden of of enhanced monitoring of glucose well as appropriate diagnosis of second- cardiorenal metabolic disease in terms and other variables leveraged with ary diabetes, which should allow more of incidence, prevalence, and cost is real-time informatics-based approaches informed individualization of care. Better an existential threat to society. Urgent to adapt treatment on an individual data on optimal approaches to diabetes attention to improve prevention and basis has great potential but has not management in frail and older adult treatment is of the essence. been elucidated. patients is urgently required considering The management of hyperglycemia in c Overweight/obesity. Current therapy the controversy around glycemic targets type 2 diabetes has become extraor- is clearly inadequate. Innovation in and the benefits and harms of specific dinarily complex with the number of methods and implementation would treatments from lifestyle management glucose-lowering medications now avail- transform diabetes prevention and to medications. Current approaches to the able. Patient-centered decision making care. Understanding the biology, psy- management of type 2 diabetes in ado- and support and consistent efforts to chology, and sociology of obesity to lescents and young adults do not seem to improve diet and exercise remain the identify pharmacological, behavioral, alter the loss of b-cell function and most foundation of all glycemic management. and political approaches to preventing individuals in this age-group quickly Initial use of metformin, followed by and treating this principal cause of transition to insulin therapy. Studies addition of glucose-lowering medications type 2 diabetes is essential. to guide optimal therapy in this emerging based on patient comorbidities and con- c Lifestyle management and DSMES. population with a terrifyingly high risk of cerns is recommended as we await an- Though the benefits of these ap- early disability is an immediate need. swers to the many questions that remain. proaches are clear, better paradigms There are enduring questions that on how to target, individualize, and continue to challenge guideline develop- sustain the effects are needed. ment. For example, does metformin pro- b fi Acknowledgments. The authors would like to c -Cell function. Preserving and en- vide cardiovascular bene t in patients acknowledge Mindy Saraco (Associate Director, hancing b-cell function is perceived with type 2 diabetes early in the natural Scientific & Medical Communication), Gedeon as the holy grail of diabetes and yet history of diabetes, as suggested by the Topacio (Finance & Project Manager, Research & effective techniques are inadequately UKPDS? Is metformin’s role as first-line Scientific Programs), and Erika Berg (Director, fi developed. medication management truly evidence- Scienti c & Medical Affairs) from the American Diabetes Association as well as Mary Hata (Ex- c Translational research. There is a huge based or a quirk of history? Though the ecutive Assistant) and Petra Niemann (Executive gap between the knowledge gained rationale for early combination therapy Assistant) from EASD for their help with the from clinical trials and application of targeting normal levels of glycemia in development of the consensus report and re- that information in clinical practice. early diabetes is seductive, clinical trial lated meetings/presentations. The authors fi fi would like to also acknowledge Mike Bonar This gap should be lled with prag- evidence to support speci c combina- (Creative Director) and Charlie Franklin (De- matic studies and other designs that tions and targets is essentially nonexis- sign Assistant) from the Leicester Diabetes Cen- include costs, measures of patient tent. As the cost implications for these tre, Leicester, U.K., who provided considerable care.diabetesjournals.org Davies and Associates 27

supportindraftingandamendingthe figures.The and grants from Merck and Ligand outside the 3. American Diabetes Association. 8. Pharma- authors also acknowledge Francesco Zaccardi submitted work. J.F. has nothing to disclose. J.F.’s cologic approaches to glycemic treatment: Stan- (PhD, Clinical Research Fellow, University of input into this consensus report is from her own dards of Medical Care in Diabetesd2018. Leicester, Leicester, U.K.) and David Kloecker perspective and the report does not reflect the Diabetes Care 2018;41(Suppl. 1):S73–S85 (Medical Student, University of Leicester) who view of the National Institutes of Health, De- 4. Inzucchi SE, Bergenstal RM, Buse JB, et al.; assisted with extracting PubMed articles and partment of Health and Human Services, or the Position Statement of the American Diabetes identifying relevant records by title and abstract; U.S.Government.W.N.K.hasnothingtodisclose. Association (ADA) and the European Associ- Francesco Zaccardi helped to define the initial C.M. reports grants and personal fees from Novo ation for the Study of Diabetes (EASD). Man- search strategy and prepare the Excel file. The Nordisk, grants and personal fees from Sanofi, agement of hyperglycaemia in type 2 diabetes authors acknowledge the invited peer reviewers grants and personal fees from Merck Sharp & a patient-centered approach. Position state- who provided comments on an earlier draft of Dohme, grants and personal fees from Eli Lilly ment of the American Diabetes Association this report: Amanda Adler (Addenbroke’sHospital, and Company, grants and personal fees from (ADA) and the European Association for the Cambridge, U.K.), Kare˚ I. Birkeland (University of Novartis, personal fees from Bristol-Myers Squibb, Study of Diabetes (EASD). Diabetologia 2012; Oslo, Oslo, Norway), James J. Chamberlain (St. personal fees from AstraZeneca, grants and 55:1577–1596 Mark’s Hospital, Salt Lake City, UT), Jill P. Crandall personal fees from Boehringer Ingelheim, per- 5. Inzucchi SE, Bergenstal RM, Buse JB, et al. (Albert Einstein College of Medicine, New York sonal fees from Hanmi Pharmaceuticals, grants Management of hyperglycaemia in type 2 di- City, NY), Ian H. de Boer (University of Washington, and personal fees from Roche Diagnostics, grants abetes, 2015: a patient-centred approach. Up- Seattle, WA), Stefano Del Prato (University of and personal fees from Medtronic, grants and date to a position statement of the American Pisa, Pisa, Italy), George Dimitriadis (Athens Uni- personal fees from Intrexon, grants and personal Diabetes Association and the European Associ- versity, Athens, Greece), Sean Dinneen (National fees from Abbott, and personal fees from UCB, ation for the Study of Diabetes. Diabetologia University of Ireland, Galway, Ireland), Vivian A. outside the submitted work. G.M. reports grants 2015;58:429–442 Fonseca (Tulane University, New Orleans, LA), and personal fees from Novo Nordisk, personal 6. Riddle MC, Gerstein HC, Holman RR, et al. Simon R. Heller (University of Sheffield, Sheffield, fees from Johnson & Johnson, and personal fees A1C targets should be personalized to maxi- U.K.), RichardI.G.Holt(UniversityofSouthampton, from Fractyl Inc., during the conduct of the study. mize benefits while limiting risks. Diabetes Care Southampton, U.K.), Silvio E. Inzucchi (Yale Uni- P.R. reports grants and nonfinancial and other 2018;41:1121–1124 versity, New Haven, CT), Eric L. Johnson (University support from Novo Nordisk, grants and other 7. American Diabetes Association. 9. Cardiovas- of North Dakota, Grand Forks, ND), Joshua support from AstraZeneca, other support from cular disease and risk management: Standards J. Neumiller (Washington State University, Spo- Bayer, other support from Boehringer Ingelheim, of Medical Care in Diabetesd2018. Diabetes kane,WA),KamleshKhunti(UniversityofLeicester, other support from Merck Sharp & Dohme, and Care 2018;41(Suppl. 1):S86–S104 Leicester, U.K.), Harald H. Klein (Ruhr University of other support from Eli Lilly, during the conduct of 8. Gæde P, Oellgaard J, Carstensen B, et al. Years Bochum, Bochum, U.K.), Line Kleinebreil (Hopitalˆ the study. A.T. reports nonfinancial support from of life gained by multifactorial intervention in national de Saint Maurice, Saint-Maurice, France), the European Association for the Study of Di- patients with type 2 diabetes mellitus and micro- Jose´ Manuel Fernandez-Real´ (Universitat de Gi- abetes during the conduct of the study; grants albuminuria: 21 years follow-up on the Steno-2 rona, Girona, Spain), Sally M. Marshall (Newcastle and other support from Boehringer Ingelheim, randomised trial. Diabetologia 2016;59:2298– University, Newcastle upon Tyne, U.K.), Manel grants and other support from Novo Nordisk, 2307 Mata-Cases (Institut Universitari d’Investigacioen´ other support from Novartis, grants and other 9. Khunti K, Kosiborod M, Ray KK. Legacy ben- AtencioPrim´ aria` Jordi Gol [IDIAP Jordi Gol], Bar- support from Sanofi, grants and other support efits of blood glucose, blood pressure and lipid celona, Spain), David R. Matthews (University of from AstraZeneca, grants from GSK, and grants control in individuals with diabetes and cardio- Oxford, Oxford, U.K.), David M. Nathan (Mas- and other support from European Foundationfor vascular disease: time to overcome multifactorial sachusetts General Hospital, Boston, MA), Michael the Study of Diabetes, outside the submitted therapeutic inertia? Diabetes Obes Metab 2018; A. Nauck (Diabetes Center Bochum-Hattingen, work. D.J.W. has nothing to disclose. J.B.B. has 20:1337–1341 St. Josef-Hospital, Ruhr-University, Bochum, provided consultation to Adocia, AstraZeneca, 10. Gregg EW, Sattar N, Ali MK. The changing Germany), Frank Nobels (OLV-Hospital, Aalst, Eli Lilly, GI Dynamics, Intarcia, MannKind, face of diabetes complications. Lancet Diabetes Belgium), Richard E. Pratley (Florida Hospital NovaTarg, Novo Nordisk, Senseonics, and vTv Endocrinol 2016;4:537–547 Diabetes Institute, Orlando, FL), Maria Jose Therapeutics with fees paid to the University 11. Little RR, Rohlfing CL, Sacks DB. Status of Redondo (Baylor College of Medicine, Houston, of North Carolina. He has received grant support HbA1c measurement and goals for improvement: TX), Michael R. Rickels (University of Pennsylvania, from AstraZeneca, Johnson & Johnson, Novo from chaos to order for improving diabetes care. Philadelphia, PA), Matthew C. Riddle (Oregon Nordisk, Sanofi, and vTv Therapeutics. He is a Clin Chem 2011;57:204–214 Health & Science University, Portland, OR), Julio consultant to Neurimmune AG. He holds stock 12. American Diabetes Association. 2. Classifi- Rosenstock (Diabetes and Endocrine Center, Dallas, options in Mellitus Health, PhaseBio, and Stabil- cation and diagnosis of diabetes: Standards of TX), Giorgio Sesti (Magna Graecia University of ity Health. He is supported by a grant from the Medical Care in Diabetesd2018. Diabetes Care Catanzaro, Catanzaro, Italy), Neil Skolnik (Abington National Institutes of Health (UL1TR002489). No 2018;41(Suppl. 1):S13–S27 Family Medicine, Jenkintown, PA), Krzysztof Strojek other potential conflicts of interest relevant to 13. Mannucci E, Antenore A, Giorgino F, Scavini (Silesian Medical University, Zabrze, Poland), Jennifer this article were reported. M. Effects of structured versus unstructured Trujillo (University of Colorado, Denver, CO), Author Contributions. All authors were re- self-monitoring of blood glucose on glucose Guillermo E. Umpierrez (Emory University, sponsible for drafting the article and revising it control in patients with non-insulin-treated Atlanta, GA), and Jennifer Wyckoff (University critically for important intellectual content. All type 2 diabetes: a meta-analysis of randomized of Michigan, Ann Arbor, MI). authors approved the version to be published. controlled trials. J Diabetes Sci Technol 2018;12: Funding. This activity was funded by the Amer- Data Availability. The details of the search 183–189 ican Diabetes Association and the European strategy, the results, and the classification 14. Young LA, Buse JB, Weaver MA, et al.; Association for the Study of Diabetes. for the included articles are available at Monitor Trial Group. Glucose self-monitoring Duality of Interest. M.J.D. reports personal fees https://dx.doi.org/10.17632/h5rcnxpk8w.1. in non–insulin-treated patients with type 2 di- and grants from Boehringer Ingelheim, Janssen, abetes in primary care settings: a randomized Novo Nordisk, and Sanofi and personal fees from References trial. JAMA Intern Med 2017;177:920–929 AstraZeneca,EliLilly,GileadSciencesLtd., Intarcia/ 1. Rodriguez-Gutierrez R, Gionfriddo MR, 15. Anjana RM, Kesavadev J, Neeta D, et al. A Servier, Merck Sharp & Dohme, Mitsubishi Ospina NS, et al. Shared decision making in multicenter real-life study on the effect of flash Tanabe Pharma Corporation, and Takeda Phar- endocrinology: present and future directions. glucose monitoring on glycemic control in pa- maceuticals International Inc. D.A.D’A. reports Lancet Diabetes Endocrinol 2016;4:706–716 tients with type 1 and type 2 diabetes. Diabetes personal fees from Eli Lilly, Merck, Novo Nordisk, 2. American Diabetes Association. 6. Glyce- Technol Ther 2017;19:533–540 and Intarcia and grants from Merck and Ligand mic targets: Standards of Medical Care in Dia- 16. American Diabetes Association. 3. Compre- during the conduct of the study; personal fees betesd2018. Diabetes Care 2018;41(Suppl. 1): hensive medical evaluation and assessment of fromEliLilly,Merck,NovoNordisk,andIntarcia S55–S64 comorbidities: Standards of Medical Care in 28 Consensus Report Diabetes Care

Diabetesd2018. Diabetes Care 2018;41(Suppl. mortality in type 2 diabetes patients: a system- interventions in primary care and community 1):S28–S37 atic review and meta-analysis. Endocrine 2017; settings. BMJ Open 2017;7:e015135 17. Kunneman M, Montori VM, Castaneda- 55:712–731 45. American Diabetes Association. 1. Improv- Guarderas A, Hess EP. What is shared decision 31. Chatterjee S, Davies MJ, Heller S, Speight J, ing care and promoting health in populations: making? (and what it is not). Acad Emerg Med Snoek FJ, Khunti K. Diabetes structured self- Standards of Medical Care in Diabetesd2018. 2016;23:1320–1324 management education programmes: a narrative Diabetes Care 2018;41(Suppl. 1):S7–S12 18. Breslin M, Mullan RJ, Montori VM. The review and current innovations. Lancet Diabetes 46. Cefalu WT, Kaul S, Gerstein HC, et al. Car- design of a decision aid about diabetes medications Endocrinol 2018;6:130–142 diovascular outcomes trials in type 2 diabetes: for use during the consultation with patients 32. Egede LE, Gebregziabher M, Echols C, Lynch where do we go from here? Reflections from a with type 2 diabetes. Patient Educ Couns 2008; CP. Longitudinal effects of medication nonad- Diabetes Care Editors’ Expert Forum. Diabetes 73:465–472 herence on glycemic control. Ann Pharmacother Care 2018;41:14–31 19. Mullan RJ, Montori VM, Shah ND, et al. The 2014;48:562–570 47. Marso SP, Daniels GH, Brown-Frandsen K, diabetes mellitus medication choice decision 33. Huber CA, Reich O. Medication adherence in et al.; LEADER Steering Committee; LEADER Trial aid: a randomized trial. Arch Intern Med 2009; patients with diabetes mellitus: does physician Investigators. Liraglutide and cardiovascular out- 169:1560–1568 drug dispensing enhance quality of care? Evi- comes in type 2 diabetes. N Engl J Med 2016;375: 20. Stacey D, Legar´ e´ F, Lewis K, et al. Decision dence from a large health claims database in 311–322 aids for people facing health treatment or screen- Switzerland. Patient Prefer Adherence 2016;10: 48. Marso SP, Bain SC, Consoli A, et al.; SUSTAIN- ing decisions. Cochrane Database Syst Rev 2017; 1803–1809 6 Investigators. Semaglutide and cardiovascular 4:CD001431 34. Iglay K, Cartier SE, Rosen VM, et al. Meta- outcomes in patients with type 2 diabetes. N Engl 21. American Diabetes Association. 4. Lifestyle analysis of studies examining medication adher- J Med 2016;375:1834–1844 management: Standards of Medical Care in ence, persistence, and discontinuation of oral 49. Holman RR, Bethel MA, Mentz RJ, et al.; Diabetesd2018. Diabetes Care 2018;41(Suppl. antihyperglycemic agentsin type 2 diabetes.Curr EXSCEL Study Group. Effects of once-weekly ex- 1):S38–S50 Med Res Opin 2015;31:1283–1296 enatide on cardiovascular outcomes in type 2 22. Powers MA, Bardsley J, Cypress M, et al. 35. McGovern A, Tippu Z, Hinton W, Munro N, diabetes. N Engl J Med 2017;377:1228–1239 Diabetes self-management education and sup- Whyte M, de Lusignan S. Systematic review of 50. Pfeffer MA, Claggett B, Diaz R, et al.; ELIXA port in type 2 diabetes: a joint position statement adherence rates by medication class in type 2 Investigators. Lixisenatide in patients with type 2 of the American Diabetes Association, the Amer- diabetes: a study protocol. BMJ Open 2016;6: diabetes and acute coronary syndrome. N Engl ican Association of Diabetes Educators, and the e010469 J Med 2015;373:2247–2257 Academy of Nutrition and Dietetics. Diabetes 36. Khunti K, Seidu S, Kunutsor S, Davies M. 51. Zinman B, Wanner C, Lachin JM, et al.; Care 2015;38:1372–1382 Association between adherence to pharma- EMPA-REG OUTCOME Investigators. Empagli- 23. Department of Health. Diabetes UK. Struc- cotherapy and outcomes in type 2 diabetes: a flozin, cardiovascular outcomes, and mortality tured patient education in diabetes: report meta-analysis. Diabetes Care 2017;40:1588– in type 2 diabetes. N Engl J Med 2015;373: from the Patient Education Working Group. 1596 2117–2128 Structured patient education in diabetes: report 37. Polonsky WH, Henry RR. Poor medication 52. Neal B, Perkovic V, Mahaffey KW, et al.; fromthe PatientEducationWorkingGroup,2005. adherence in type 2 diabetes: recognizing the CANVAS Program Collaborative Group. Cana- Available from http://webarchive.nationalarchives scope of the problem and its key contributors. gliflozin and cardiovascular and renal events .gov.uk/20130105204013/http://www.dh.gov.uk/ Patient Prefer Adherence 2016;10:1299–1307 in type 2 diabetes. N Engl J Med 2017;377: en/Publicationsandstatistics/Publications/ 38. McGovern A, Tippu Z, Hinton W, Munro N, 644–657 PublicationsPolicyAndGuidance/DH_4113195. Whyte M, de Lusignan S. Comparison of med- 53. Lehrke M, Marx N. Diabetes mellitus and Accessed 29 August 2018 ication adherence and persistence in type 2 heart failure. Am J Med 2017;130(6S):S40–S50 24. National Institute for Health and Clinical diabetes: a systematic review and meta-analysis. 54. Fitchett D, Zinman B, Wanner C, et al.; Excellence. (2011) Quality Standard for Diabetes Diabetes Obes Metab 2018;20:1040–1043 EMPA-REG OUTCOMEÒ trial investigators. Heart in Adults, Statements 2 and 3, 2011. Available 39. Lasalvia P, Barahona-Correa JE, Romero- failure outcomes with empagliflozin in patients from https://www.nice.org.uk/guidance/qs6. Alvernia DM, et al. Pen devices for insulin with type 2 diabetes at high cardiovascular risk: Accessed 29 August 2018 self-administration compared with needle and results of the EMPA-REG OUTCOMEÒ trial. Eur 25. Beck J, Greenwood DA, Blanton L, et al.; 2017 vial: systematic review of the literature and Heart J 2016;37:1526–1534 Standards Revision Task Force. 2017 National meta-analysis. J Diabetes Sci Technol 2016;10: 55. Radholm˚ K, Figtree G, Perkovic V, et al. standards for diabetes self-management educa- 959–966 Canagliflozin and heart failure in type 2 diabetes tion and support. Diabetes Educ 2017;43:449– 40. Khunti K, Davies MJ. Clinical inertia-time to mellitus: results from the CANVAS Program. 464 reappraise the terminology? Prim Care Diabetes Circulation 2018;138:458–469 26. Chrvala CA, Sherr D, Lipman RD. Diabetes 2017;11:105–106 56. Margulies KB, Hernandez AF, Redfield MM, self-management education for adults with 41. Furler J, O’Neal D, Speight J, et al. Supporting et al.; NHLBI Heart Failure Clinical Research type 2 diabetes mellitus: a systematic review insulin initiation in type 2 diabetes in primary Network. Effects of liraglutide on clinical stability of the effect on glycemic control. Patient Educ care: results of the Stepping Up pragmatic cluster among patients with advanced heart failure and Couns 2016;99:926–943 randomised controlled clinical trial. BMJ 2017; reduced ejection fraction: a randomized clinical 27. Pillay J, Armstrong MJ, Butalia S, et al. Be- 356:j783 trial. JAMA 2016;316:500–508 havioral programs for type 2 diabetes mellitus: a 42. Manski-Nankervis J-A, Furler J, O’Neal D, 57. Jorsal A, Kistorp C, Holmager P, et al. Effect of systematic review and network meta-analysis. Ginnivan L, Thuraisingam S, Blackberry I. Over- liraglutide, a glucagon-like peptide-1 analogue, Ann Intern Med 2015;163:848–860 coming clinical inertia in insulin initiation in on left ventricular function in stable chronic heart 28. ZhaoF-F,SuhonenR,KoskinenS,Leino-KilpiH. primary care for patients with type 2 diabetes: failure patients with and without diabetes Theory-based self-management educational 24-month follow-up of the Stepping Up cluster (LIVE)-a multicentre, double-blind, randomised, interventions on patients with type 2 diabetes: a randomised controlled trial. Prim Care Diabetes placebo-controlled trial. Eur J Heart Fail 2017;19: systematic review and meta-analysis of random- 2017;11:474–481 69–77 ized controlled trials. J Adv Nurs 2017;73:812–833 43. Tabesh M, Magliano DJ, Koye DN, Shaw JE. 58. Scirica BM, Bhatt DL, Braunwald E, et al.; 29. Odgers-Jewell K, Ball LE, Kelly JT, Isenring EA, The effect of nurse prescribers on glycaemic SAVOR-TIMI 53 Steering Committee and Inves- Reidlinger DP, Thomas R. Effectiveness of group- control in type 2 diabetes: a systematic review tigators. Saxagliptin and cardiovascular outcomes based self-management education for individu- and meta-analysis. Int J Nurs Stud 2018;78:37– in patients with type 2 diabetes mellitus. N Engl als with type 2 diabetes: a systematic review 43 J Med 2013;369:1317–1326 with meta-analyses and meta-regression. Diabet 44. Murphy ME, Byrne M, Galvin R, Boland F, 59. White WB, Cannon CP, Heller SR, et al.; Med 2017;34:1027–1039 Fahey T, Smith SM. Improving risk factor man- EXAMINE Investigators. Alogliptin after acute 30. He X, Li J, Wang B, et al. Diabetes self- agement for patients with poorly controlled type coronary syndrome in patients with type 2 di- management education reduces risk of all-cause 2 diabetes: a systematic review of healthcare abetes. N Engl J Med 2013;369:1327–1335 care.diabetesjournals.org Davies and Associates 29

60. Green JB, Bethel MA, Armstrong PW, et al.; meta-analysis. Cardiovasc Diagn Ther 2014;4: 89. Lee MS, Jun JH, Lim H-J, Lim H-S. A sys- TECOS Study Group. Effect of sitagliptin on 373–382 tematic review and meta-analysis of tai chi for cardiovascular outcomes in type 2 diabetes. N 75. Lean ME, Leslie WS, Barnes AC, et al. Primary treating type 2 diabetes. Maturitas 2015;80: Engl J Med 2015;373:232–242 care-led weight management for remission of 14–23 61. Wanner C, Inzucchi SE, Lachin JM, et al.; type 2 diabetes (DiRECT): an open-label, cluster- 90. Cui J, Yan J-H, Yan L-M, Pan L, Le JJ, Guo YZ. EMPA-REG OUTCOME Investigators. Empagliflo- randomised trial. Lancet 2018;391:541–551 Effects of yoga in adults with type 2 diabetes zin and progression of kidney disease in type 2 76. Wing RR, Bolin P, Brancati FL, et al.; Look mellitus: a meta-analysis. J Diabetes Investig diabetes. N Engl J Med 2016;375:323–334 AHEAD Research Group. Cardiovascular effects 2017;8:201–209 62. Jardine MJ, Mahaffey KW, Neal B, et al.; of intensive lifestyle intervention in type 2 di- 91. Qiu S, Cai X, Chen X, Yang B, Sun Z. Step CREDENCE study investigators. The Canagliflozin abetes. N Engl J Med 2013;369:145–154 counter use in type 2 diabetes: a meta-analysis and Renal Endpoints in Diabetes with Established 77. O’Neil PM, Miller-Kovach K, Tuerk PW, et al. of randomized controlled trials. BMC Med 2014; Nephropathy Clinical Evaluation (CREDENCE) Randomized controlled trial of a nationally avail- 12:36 study rationale, design, and baseline character- able weight control program tailored for adults 92. Franz MJ, Boucher JL, Rutten-Ramos S, istics. Am J Nephrol 2017;46:462–472 withtype2diabetes.Obesity(SilverSpring)2016; VanWormer JJ. Lifestyle weight-loss intervention 63. Mann JFE, Ørsted DD, Brown-Frandsen K, 24:2269–2277 outcomes in overweight and obese adults with et al.; LEADER Steering Committee and Investi- 78. Jebb SA, Ahern AL, Olson AD, et al. Primary type 2 diabetes: a systematic review and meta- gators. Liraglutide and renal outcomes in type 2 care referral to a commercial provider for weight analysis of randomized clinical trials. J Acad Nutr diabetes. N Engl J Med 2017;377:839–848 loss treatment versus standard care: a rando- Diet 2015;115:1447–1463 64. Deacon CF. A review of dipeptidyl peptidase- mised controlled trial. Lancet 2011;378:1485– 93. Aggarwal N, Singla A, Mathieu C, et al. 4 inhibitors. Hot topics from randomized con- 1492 Metformin extended-release versus immediate- trolled trials. Diabetes Obes Metab 2018;20 79. Leblanc ES, O’Connor E, Whitlock EP, release: an international, randomized, double- (Suppl. 1):34–46 Patnode CD, Kapka T. Effectiveness of primary blind, head-to-head trial in pharmacotherapy-na¨ıve 65. American Diabetes Association. 7. Obesity care-relevant treatments for obesity in adults: patients with type 2 diabetes. Diabetes Obes management for the treatment of type 2 diabe- a systematic evidence review for the U.S. Pre- Metab 2018;20:463–467 tes: Standards of Medical Care in Diabetesd ventive Services Task Force. Ann Intern Med 94. Garber AJ, Duncan TG, Goodman AM, Mills 2018. Diabetes Care 2018;41(Suppl. 1):S65–S72 2011;155:434–447 DJ, Rohlf JL. Efficacy of metformin in type II 66. Shai I, Schwarzfuchs D, Henkin Y, et al.; 80. Delahanty LM, Dalton KM, Porneala B, et al. diabetes: results of a double-blind, placebo- Dietary Intervention Randomized Controlled Improving diabetes outcomes through lifestyle controlled, dose-response trial. Am J Med Trial (DIRECT) Group. Weight loss with a low- change – a randomized controlled trial. Obesity 1997;103:491–497 carbohydrate, Mediterranean, or low-fat diet. (Silver Spring) 2015;23:1792–1799 95. Inzucchi SE, Lipska KJ, Mayo H, Bailey CJ, N Engl J Med 2008;359:229–241 81. Boule´ NG, Haddad E, Kenny GP, Wells GA, McGuire DK. Metformin in patients with type 67. Esposito K, Maiorino MI, Ciotola M, et al. Sigal RJ. Effects of exercise on glycemic control 2 diabetes and kidney disease: a systematic Effects of a Mediterranean-style diet on the need and body mass in type 2 diabetes mellitus: a review. JAMA 2014;312:2668–2675 for antihyperglycemic drug therapy in patients meta-analysis of controlled clinical trials. JAMA 96. Lalau J-D, Kajbaf F, Bennis Y, Hurtel-Lemaire with newly diagnosed type 2 diabetes: a ran- 2001;286:1218–1227 AS, BelpaireF, De Broe ME. Metformin treatment domized trial. Ann Intern Med 2009;151: 82. Chudyk A, Petrella RJ. Effects of exercise in patients with type 2 diabetes and chronic 306–314 on cardiovascular risk factors in type 2 diabetes: kidney disease stages 3A, 3B, or 4. Diabetes Care 68. Esposito K, Maiorino MI, Petrizzo M, a meta-analysis. Diabetes Care 2011;34:1228– 2018;41:547–553 Bellastella G, Giugliano D. The effects of a Med- 1237 97. Imam TH. Changes in metformin use in iterranean diet on the need for diabetes drugs 83. Yang Z, Scott CA, Mao C, Tang J, Farmer AJ. chronic kidney disease. Clin Kidney J 2017;10: and remission of newly diagnosed type 2 di- Resistance exercise versus aerobic exercise for 301–304 abetes: follow-up of a randomized trial. Diabetes type 2 diabetes: a systematic review and meta- 98. UK Prospective Diabetes Study (UKPDS) Care 2014;37:1824–1830 analysis. Sports Med 2014;44:487–499 Group. Effect of intensive blood-glucose control 69. Huo R, Du T, Xu Y, et al. Effects of Medi- 84. Balducci S, Zanuso S, Nicolucci A, et al.; with metformin on complications in overweight terranean-style diet on glycemic control, weight Italian Diabetes Exercise Study (IDES) Investiga- patients with type 2 diabetes (UKPDS 34). Lancet loss and cardiovascular risk factors among type 2 tors. Effect of an intensive exercise intervention 1998;352:854–865 diabetes individuals: a meta-analysis. Eur J Clin strategy on modifiable cardiovascular risk fac- 99. Griffin SJ, Leaver JK, Irving GJ. Impact of Nutr 2015;69:1200–1208 tors in subjects with type 2 diabetes mellitus: a metformin on cardiovascular disease: a meta- 70. Snorgaard O, Poulsen GM, Andersen HK, randomized controlled trial: the Italian Diabetes analysis of randomised trials among people Astrup A. Systematic review and meta-analysis and Exercise Study (IDES). Arch Intern Med with type 2 diabetes. Diabetologia 2017;60: of dietary carbohydrate restriction in patients 2010;170:1794–1803 1620–1629 with type 2 diabetes. BMJ Open Diabetes Res 85. Schwingshackl L, Missbach B, Dias S, Konig¨ J, 100. Maruthur NM, Tseng E, Hutfless S, et al. Care 2017;5:e000354 Hoffmann G. Impact of different training mo- Diabetes medications as monotherapy or 71. Ajala O, English P, Pinkney J. Systematic dalities on glycaemic control and blood lipids in metformin-based combination therapy for type 2 review and meta-analysis of different dietary patients with type 2 diabetes: a systematic re- diabetes: a systematic review and meta-analysis. approaches to the management of type 2 di- view and network meta-analysis. Diabetologia Ann Intern Med 2016;164:740–751 abetes. Am J Clin Nutr 2013;97:505–516 2014;57:1789–1797 101. Aroda VR, Edelstein SL, Goldberg RB, et al.; 72. Azadbakht L, Fard NRP, Karimi M, et al. 86. Qiu S, Cai X, Schumann U, Velders M, Sun Z, Diabetes Prevention Program Research Group. Effects of the Dietary Approaches to Stop Hy- Steinacker JM. Impact of walking on glycemic Long-term metformin use and vitamin B12 de- pertension (DASH) eating plan on cardiovascular control and other cardiovascular risk factors in ficiency in the Diabetes Prevention Program risks among type 2 diabetic patients: a random- type 2 diabetes: a meta-analysis. PLoS One 2014; Outcomes Study. J Clin Endocrinol Metab 2016; ized crossover clinical trial. Diabetes Care 2011; 9:e109767 101:1754–1761 34:55–57 87. Rees JL, Johnson ST, BouleNG.Aquatic´ 102. Zhang X-L, Zhu Q-Q, Chen Y-H, et al. Car- 73. Sainsbury E, Kizirian NV, Partridge SR, Gill T, exercise for adults with type 2 diabetes: a diovascular safety, long-term noncardiovascular Colagiuri S, Gibson AA. Effect of dietary carbo- meta-analysis. Acta Diabetol 2017;54:895– safety, and efficacy of sodium-glucose cotrans- hydrate restriction on glycemic control in adults 904 porter 2 inhibitors in patients with type 2 with diabetes: a systematic review and meta- 88. Pai L-W, Li T-C, Hwu Y-J, Chang SC, Chen LL, diabetes mellitus: a systemic review and meta- analysis. Diabetes Res Clin Pract 2018;139:239– Chang PY. The effectiveness of regular leisure- analysis with trial sequential analysis. J Am Heart 252 time physical activities on long-term glycemic Assoc 2018;7:e007165 74. Yokoyama Y, Barnard ND, Levin SM, control in people with type 2 diabetes: a sys- 103. Storgaard H, Gluud LL, Bennett C, et al. Watanabe M. Vegetarian diets and glycemic tematic review and meta-analysis. Diabetes Res Benefits and harms of sodium-glucose co- control in diabetes: a systematic review and Clin Pract 2016;113:77–85 transporter 2 inhibitors in patients with type 2 30 Consensus Report Diabetes Care

diabetes: a systematic review and meta-analysis. type 2 diabetes: a randomised, open-label, non- Pioglitazone 027 Study Group. Clin Ther 2000;22: PLoS One 2016;11:e0166125 inferiority study. Lancet 2008;372:1240–1250 1395–1409 104. Li D, Wang T, Shen S, Fang Z, Dong Y, Tang H. 117. Andreadis P, Karagiannis T, Malandris K, 131. Yki-Jarvinen¨ H. Thiazolidinediones. N Engl J Urinary tract and genital infections in patients et al. Semaglutide for type 2 diabetes mellitus: a Med 2004;351:1106–1118 with type 2 diabetes treated with sodium-glucose systematic review and meta-analysis. Diabetes 132. Kernan WN, Viscoli CM, Furie KL, et al.; IRIS co-transporter 2 inhibitors: a meta-analysis of Obes Metab 2018;20:2255–2263 Trial Investigators. Pioglitazone after ischemic randomized controlled trials. Diabetes Obes 118. Li Z, Zhang Y, Quan X, et al. Efficacy and stroke or transient ischemic attack. N Engl J Med Metab 2017;19:348–355 acceptability of glycemic control of glucagon-like 2016;374:1321–1331 105. Jabbour S, Seufert J, Scheen A, Bailey CJ, peptide-1 receptor agonists among type 2 di- 133. Hanefeld M, Marx N, Pfutzner¨ A, et al. Anti- Karup C, Langkilde AM. Dapagliflozin in patients abetes: a systematic review and network meta- inflammatory effects of pioglitazone and/or sim- with type 2 diabetes mellitus: a pooled analysis analysis. PLoS One 2016;11:e0154206 vastatin in high cardiovascular risk patients with of safety data from phase IIb/III clinical trials. 119. Storgaard H, Cold F, Gluud LL, Vilsbøll T, elevated high sensitivity C-reactive protein: the Diabetes Obes Metab 2018;20:620–628 Knop FK. Glucagon-like peptide-1 receptor ago- PIOSTAT Study. J Am Coll Cardiol 2007;49:290– 106. Tang H,Li D,Wang T,Zhai S, Song Y. Effectof nists and risk of acute pancreatitis in patients 297 sodium-glucose cotransporter 2 inhibitors on di- with type 2 diabetes. Diabetes Obes Metab 2017; 134. Saremi A, Schwenke DC, Buchanan TA, et al. abetic ketoacidosis among patients with type 2 19:906–908 Pioglitazone slows progression of atherosclero- diabetes: a meta-analysis of randomized con- 120. Monami M, Nreu B, Scatena A, et al. Safety sis in prediabetes independent of changes in trolled trials. Diabetes Care 2016;39:e123– issues with glucagon-like peptide-1 receptor cardiovascular risk factors. Arterioscler Thromb e124 agonists (pancreatitis, pancreatic cancer and Vasc Biol 2013;33:393–399 107. Thrasher J. Pharmacologic management cholelithiasis): data from randomized controlled 135. Nissen SE, Nicholls SJ, Wolski K, et al.; of type 2 diabetes mellitus: available therapies. trials. Diabetes Obes Metab 2017;19:1233–1241 PERISCOPE Investigators. Comparison of pioglita- Am J Med 2017;130(6S):S4–S17 121. Esposito K, Chiodini P, Maiorino MI, zone vs glimepiride on progression of coronary 108. Karagiannis T, Liakos A, Bekiari E, et al. Bellastella G, Capuano A, Giugliano D. Glycaemic atherosclerosis in patients with type 2 diabetes: Efficacy and safety of once-weekly glucagon-like durability with dipeptidyl peptidase-4 inhibi- the PERISCOPE randomized controlled trial. JAMA peptide 1 receptor agonists for the management tors in type 2 diabetes: a systematic review and 2008;299:1561–1573 of type 2 diabetes: a systematic review and meta- meta-analysis of long-term randomised controlled 136. Dormandy JA, Charbonnel B, Eckland DJA, analysis of randomized controlled trials. Diabe- trials. BMJ Open 2014;4:e005442 et al.; PROactive Investigators. Secondary pre- tes Obes Metab 2015;17:1065–1074 122. Aroda VR, Henry RR, Han J, et al. Efficacy of vention of macrovascular events in patients 109. Zaccardi F, Htike ZZ, Webb DR, Khunti K, GLP-1 receptor agonists and DPP-4 inhibitors: with type 2 diabetes in the PROactive Study Davies MJ. Benefits and harms of once-weekly meta-analysis and systematic review. Clin Ther (PROspective pioglitAzone Clinical Trial In macro- glucagon-like peptide-1 receptor agonist treat- 2012;34:1247–1258.e22 Vascular Events): a randomised controlled trial. ments: a systematic review and network meta- 123. Wu S, Chai S, Yang J, et al. Gastrointestinal Lancet 2005;366:1279–1289 analysis. Ann Intern Med 2016;164:102–113 adverse events of dipeptidyl peptidase 4 inhib- 137. Bach RG, BrooksMM, LombarderoM, et al.; 110. Htike ZZ, Zaccardi F, Papamargaritis D, itors in type 2 diabetes: a systematic review and BARI 2D Investigators. Rosiglitazone and out- Webb DR, Khunti K, Davies MJ. Efficacy and network meta-analysis. Clin Ther 2017;39:1780– comes for patients with diabetes mellitus and safety of glucagon-like peptide-1 receptor ago- 1789.e33 coronaryartery diseasein the BypassAngioplasty nists in type 2 diabetes: a systematic review and 124. Salvo F, Moore N, Arnaud M, et al. Addition Revascularization Investigation 2 Diabetes (BARI mixed-treatment comparison analysis. Diabetes of dipeptidyl peptidase-4 inhibitors to sulpho- 2D) trial. Circulation 2013;128:785–794 Obes Metab 2017;19:524–536 nylureas and risk of hypoglycaemia: systematic 138. Mahaffey KW, Hafley G, Dickerson S, et al. 111. Sorli C, Harashima S-I, Tsoukas GM, et al. review and meta-analysis. BMJ 2016;353:i2231 Results of a reevaluation of cardiovascular out- Efficacy and safety of once-weekly semaglutide 125. Tka´cˇ I, Raz I. Combined analysis of three comes in the RECORD trial. Am Heart J 2013;166: monotherapy versus placebo in patients with large interventional trials with gliptins indicates 240–249.e1 type 2 diabetes (SUSTAIN 1): a double-blind, increased incidence of acute pancreatitis in 139. Cusi K, Orsak B, Bril F, et al. Long-term randomised, placebo-controlled, parallel-group, patients with type 2 diabetes. Diabetes Care pioglitazone treatment for patients with non- multinational, multicentre phase 3a trial. Lancet 2017;40:284–286 alcoholic steatohepatitis and prediabetes or Diabetes Endocrinol 2017;5:251–260 126. Mascolo A, Rafaniello C, Sportiello L, et al. type 2 diabetes mellitus: a randomized trial. 112. Pratley RE, Aroda VR, Lingvay I, et al.; Dipeptidyl peptidase (DPP)-4 inhibitor-induced Ann Intern Med 2016;165:305–315 SUSTAIN 7 investigators. Semaglutide versus arthritis/arthralgia: a review of clinical cases. 140. Kahn SE, Haffner SM, Heise MA, et al.; dulaglutide once weekly in patients with type 2 Drug Saf 2016;39:401–407 ADOPT Study Group. Glycemic durability of diabetes (SUSTAIN 7): a randomised, open-label, 127. Nauck MA, Meier JJ, Cavender MA, Abd El rosiglitazone, metformin, or glyburide mono- phase 3btrial.Lancet Diabetes Endocrinol2018; Aziz M, Drucker DJ. Cardiovascular actions and therapy. N Engl J Med 2006;355:2427–2443 6:275–286 clinical outcomes with glucagon-like peptide-1 141. Home PD, Pocock SJ, Beck-Nielsen H, et al.; 113. Dungan KM, Povedano ST, Forst T, et al. receptor agonists and dipeptidyl peptidase-4 RECORD Study Team. Rosiglitazone evaluated Once-weekly dulaglutide versus once-daily lira- inhibitors. Circulation 2017;136:849–870 for cardiovascular outcomes in oral agent com- glutidein metformin-treated patientswith type 2 128. Li L, Li S, Deng K, et al. Dipeptidyl peptidase- bination therapy for type 2 diabetes (RECORD): diabetes (AWARD-6): a randomised, open-label, 4 inhibitors and risk of heart failure in type 2 a multicentre, randomised, open-label trial. Lan- phase 3, non-inferiority trial. Lancet 2014;384: diabetes: systematic review and meta-analysis cet 2009;373:2125–2135 [Lond Engl] 1349–1357 of randomised and observational studies. BMJ 142. Hanefeld M, Brunetti P, Schernthaner GH, 114. Buse JB, Nauck M, Forst T, et al. Exenatide 2016;352:i610 Matthews DR, Charbonnel BH; QUARTET Study once weekly versus liraglutide once daily in 129. Aronoff S, Rosenblatt S, Braithwaite S, Egan Group. One-year glycemic control with a sulfo- patients with type 2 diabetes (DURATION-6): JW, Mathisen AL, Schneider RL. Pioglitazone nylurea plus pioglitazone versus a sulfonylurea a randomised, open-label study. Lancet 2013; hydrochloride monotherapy improves glycemic plus metformin in patients with type 2 diabetes. 381:117–124 control in the treatment of patients with type 2 Diabetes Care 2004;27:141–147 115. Buse JB, Rosenstock J, Sesti G, et al.; LEAD-6 diabetes: a 6-month randomized placebo- 143. Viscoli CM, Inzucchi SE, Young LH, et al.; IRIS Study Group. Liraglutide once a day versus controlled dose-response study. The Pioglita- Trial Investigators. Pioglitazone and risk for exenatide twice a day for type 2 diabetes: a zone 001 Study Group. Diabetes Care 2000;23: bone fracture: safety data from a randomized 26-week randomised, parallel-group, multina- 1605–1611 clinical trial. J Clin Endocrinol Metab 2017;102: tional, open-label trial (LEAD-6). Lancet 2009; 130. Einhorn D, Rendell M, Rosenzweig J, Egan 914–922 374:39–47 JW, Mathisen AL, Schneider RL. Pioglitazone 144. Kahn SE, Zinman B, Lachin JM, et al.; 116. Drucker DJ, Buse JB, Taylor K, et al.; hydrochloride in combination with metformin Diabetes Outcome Progression Trial (ADOPT) DURATION-1 Study Group. Exenatide once in the treatment of type 2 diabetes mellitus: Study Group. Rosiglitazone-associated fractures weekly versus twice daily for the treatment of a randomized, placebo-controlled study. The in type 2 diabetes: an analysis from A Diabetes care.diabetesjournals.org Davies and Associates 31

Outcome Progression Trial (ADOPT). Diabetes 159. Freemantle N, Chou E, Frois C, et al. Safety Society. J Am Coll Cardiol 2014;63(25 Pt B):2985– Care 2008;31:845–851 and ef ficacy of insulin glargine 300 u/mL com- 3023 145. Lewis JD, Habel LA, Quesenberry CP, et al. pared with other basal insulin therapies in pa- 170. Garvey WT, Mechanick JI, Brett EM, et al.; Pioglitazone use and risk of bladder cancer and tients with type 2 diabetes mellitus: a network Reviewers of the AACE/ACE Obesity Clinical other common cancers in persons with diabetes. meta-analysis. BMJ Open 2016;6:e009421 Practice Guidelines. American Association of JAMA 2015;314:265–277 160. Russell-Jones D, Gall M-A, Niemeyer M, Clinical Endocrinologists and American College 146. Hirst JA, Farmer AJ, Dyar A, Lung TW, Diamant M, Del Prato S. Insulin degludec results of Endocrinology Comprehensive Clinical Prac- Stevens RJ. Estimating the effect of sulfonylurea in lower rates of nocturnal hypoglycaemia and tice Guidelines for Medical Care of Patients with on HbA1c in diabetes: a systematic review and fasting plasma glucose vs. insulin glargine: a Obesity. Endocr Pract 2016;22(Suppl. 3):1–203 meta-analysis. Diabetologia 2013;56:973–984 meta-analysis of seven clinical trials. Nutr Metab 171. Apovian CM, Aronne LJ, Bessesen DH, 147. UK Prospective Diabetes Study (UKPDS) Cardiovasc Dis 2015;25:898–905 et al.; Endocrine Society. Pharmacological man- Group. Intensive blood-glucose control with 161. Lipska KJ,Parker MM, MoffetHH, HuangES, agement of obesity: an Endocrine Society clinical sulphonylureas or insulin compared with con- Karter AJ. Association of initiation of basal insulin practice guideline. J Clin Endocrinol Metab 2015; ventional treatment and risk of complications in analogs vs neutral protamine hagedorn insulin 100:342–362 patients with type 2 diabetes (UKPDS 33). Lancet with hypoglycemia-related emergency depart- 172. Moyer VA; U.S. Preventive Services Task 1998;352:837–853 ment visits or hospital admissions and with Force. Screening for and management of obesity 148. Patel A, MacMahon S, Chalmers J, et al.; glycemic control in patients with type 2 diabetes. in adults: U.S. Preventive Services Task Force ADVANCE Collaborative Group. Intensive blood JAMA 2018;320:53–62 recommendation statement. Ann Intern Med glucose control and vascular outcomes in pa- 162. Marso SP, McGuire DK, Zinman B, et al.; 2012;157:373–378 tients with type 2 diabetes. N Engl J Med 2008; DEVOTE Study Group. Efficacy and safety of 173. Van Gaal L, Dirinck E. Pharmacological 358:2560–2572 degludec versus glargine in type 2 diabetes. N approaches in the treatment and maintenance 149. Monami M, Dicembrini I, Kundisova L, Engl J Med 2017;377:723–732 of weight loss. Diabetes Care 2016;39(Suppl. 2): Zannoni S, Nreu B, Mannucci E. A meta-analysis 163. Rosenstock J, HollanderP, BhargavaA, et al. S260–S267 of the hypoglycaemic risk in randomized con- Similar efficacy and safety of LY2963016 insulin 174. KheraR,PandeyA,ChandarAK,etal.Effects trolled trials with sulphonylureas in patients with glargine and insulin glargine (LantusÒ) in pa- of weight-loss medications on cardiometabolic type 2 diabetes. Diabetes Obes Metab 2014;16: tients with type 2 diabetes who were insulin- risk profiles: a systematic review and network 833–840 na¨ıve or previously treated with insulin glargine: meta-analysis. Gastroenterology 2018;154:1309– 150. Del Prato S, Camisasca R, Wilson C, Fleck P. a randomized, double-blind controlled trial (the 1319.e7 Durability of the efficacy and safety of alogliptin ELEMENT 2 study). Diabetes Obes Metab 2015; 175. Davies MJ, Bergenstal R, Bode B, et al.; compared with glipizide in type 2 diabetes 17:734–741 NN8022-1922 Study Group. Efficacy of liraglutide mellitus: a 2-year study. Diabetes Obes Metab 164. Gerstein HC, Bosch J, Dagenais GR, et al.; for weight loss among patients with type 2 2014;16:1239–1246 ORIGIN Trial Investigators. Basal insulin and diabetes: the SCALE diabetes randomized clinical 151. Mishriky BM, Cummings DM, Tanenberg cardiovascular and other outcomes in dysglyce- trial. JAMA 2015;314:687–699 RJ. The efficacy and safety of DPP4 inhibitors mia. N Engl J Med 2012;367:319–328 176. Schauer PR, Bhatt DL, Kirwan JP, et al.; compared to sulfonylureas as add-on therapy to 165. Riddle MC, Yki-Jarvinen¨ H, Bolli GB, et al. STAMPEDE Investigators. Bariatric surgery versus metformin in patients with type 2 diabetes: a One-year sustained glycaemic control and less intensive medical therapy for diabetesd3-year systematic review and meta-analysis. Diabetes hypoglycaemia with new insulin glargine 300 U/ml outcomes. N Engl J Med 2014;370:2002–2013 Res Clin Pract 2015;109:378–388 compared with 100 U/ml in people with type 2 177. Schauer PR, Bhatt DL, Kirwan JP, et al.; 152. KhuntiK,ChatterjeeS,GersteinHC,et al.Do diabetes using basal plus meal-time insulin: the STAMPEDE Investigators. Bariatric surgery ver- sulphonylureas still have a place in clinical prac- EDITION 1 12-month randomized trial, including sus intensive medical therapy for diabetesd tice? Lancet Diabetes Endocrinol 2018;Feb 28: 6-month extension. Diabetes Obes Metab 2015; 5-year outcomes. N Engl J Med 2017;376:641–651 pii:S2213-8587(18)30025-1 17:835–842 178. Ikramuddin S, Korner J, Lee W-J, et al. 153. Gangji AS, Cukierman T, Gerstein HC, 166. Yki-Jarvinen¨ H, Bergenstal R, Ziemen M, Lifestyle intervention and medical management Goldsmith CH, Clase CM. A systematic review et al.; EDITION 2 Study Investigators. New insulin with vs without Roux-en-Y gastric bypass and and meta-analysis of hypoglycemia and cardio- glargine 300 units/mL versus glargine 100 units/mL control of hemoglobin A1c, LDL cholesterol, and vascular events: a comparison of glyburide with in people with type 2 diabetes using oral systolic blood pressure at 5 years in the diabetes other secretagogues and with insulin. Diabetes agents and basal insulin: glucose control and surgery study. JAMA 2018;319:266–278 Care 2007;30:389–394 hypoglycemia in a 6-month randomized con- 179. Dixon JB, O’Brien PE, Playfair J, et al. Adjust- 154. Chan SP, Colagiuri S. Systematic review and trolled trial (EDITION 2). Diabetes Care 2014;37: able gastric banding and conventional therapy meta-analysis of the efficacy and hypoglycemic 3235–3243 fortype 2diabetes:arandomized controlledtrial. safety of gliclazide versus other insulinotropic 167. Riddle MC, Bolli GB, Ziemen M, Muehlen- JAMA 2008;299:316–323 agents. Diabetes Res Clin Pract 2015;110:75–81 Bartmer I, Bizet F, Home PD; EDITION 1 Study 180. Mingrone G, Panunzi S, De Gaetano A, et al. 155. Holman RR, Paul SK, Bethel MA, Matthews Investigators. New insulin glargine 300 units/mL Bariatric surgery versus conventional medical DR, Neil HA. 10-year follow-up of intensive versus glargine 100 units/mL in people with therapy for type 2 diabetes. N Engl J Med glucose control in type 2 diabetes. N Engl J type 2 diabetes using basal and mealtime insulin: 2012;366:1577–1585 Med 2008;359:1577–1589 glucose control and hypoglycemia in a 6-month 181. Sjostr¨ om¨ L, Peltonen M, Jacobson P, et al. 156. Erpeldinger S, Rehman MB, Berkhout C, randomized controlled trial (EDITION 1). Diabe- Association of bariatric surgery with long-term et al. Efficacy and safety of insulin in type 2 tes Care 2014;37:2755–2762 remission of type 2 diabetes and with microvas- diabetes: meta-analysis of randomised con- 168. Khunti K, Damci T, Husemoen LL, Babu V, cular and macrovascular complications. JAMA trolled trials. BMC Endocr Disord 2016;16:39 Liebl A. Exploring the characteristics of subopti- 2014;311:2297–2304 157. Owens DR, Traylor L, Mullins P, Landgraf W. mally controlled patients after 24 weeks of basal 182. Mingrone G, Panunzi S, De Gaetano A, et al. Patient-level meta-analysis of efficacy and hy- insulin treatment: an individualized approach to Bariatric-metabolic surgery versus conventional poglycaemia in people with type 2 diabetes intensification. Diabetes Res Clin Pract 2017;123: medical treatment in obese patients with type 2 initiating insulin glargine 100U/mL or neutral 209–217 diabetes: 5 year follow-up of an open-label, protamine Hagedorn insulin analysed according 169. Jensen MD, Ryan DH, Apovian CM, et al.; single-centre, randomised controlled trial. Lan- to concomitant oral antidiabetes therapy. Di- American College of Cardiology/American Heart cet 2015;386:964–973 abetes Res Clin Pract 2017;124:57–65 Association Task Force on Practice Guidelines; 183. Rubino F, Nathan DM, Eckel RH, et al.; 158. Goldman J, Kapitza C, Pettus J, Heise T. Obesity Society. 2013 AHA/ACC/TOS guideline Delegates of the 2nd Diabetes Surgery Summit. Understanding how pharmacokinetic and phar- for the management of overweight and obesity Metabolic surgery in the treatment algorithm macodynamic differences of basal analog insulins in adults: a report of the American College of for type 2 diabetes: a joint statement by inter- influence clinical practice. Curr Med Res Opin Cardiology/American Heart Association Task national diabetes organizations. Diabetes Care 2017;33:1821–1831 Force on Practice Guidelines and The Obesity 2016;39:861–877 32 Consensus Report Diabetes Care

184. Maggard-Gibbons M, Maglione M, Livhits on the incidence of cardiovascular events of 208. LukashevichV, DelPrato S,AragaM, Kothny M, et al. Bariatric surgery for weight loss and the addition of pioglitazone versus sulfonylureas W. Efficacy and safety of vildagliptin in patients glycemic control in nonmorbidly obese adults in patients with type 2 diabetes inadequately with type 2 diabetes mellitus inadequately con- with diabetes: a systematic review. JAMA 2013; controlled with metformin (TOSCA.IT): a rando- trolled with dual combination of metformin and 309:2250–2261 mised, multicentre trial. Lancet Diabetes Endo- sulphonylurea. Diabetes Obes Metab 2014;16: 185. Mingrone G, Bornstein S, Le Roux CW. crinol 2017;5:887–897 403–409 Optimisation of follow-up after metabolic sur- 197. Wang Z, Sun J, Han R, et al. Efficacy and 209. Hong AR, Lee J, Ku EJ, et al. Comparison of gery. Lancet Diabetes Endocrinol 2018;6:487– safety of sodium-glucose cotransporter-2 inhib- vildagliptin as an add-on therapy and sulfonyl- 499 itors versus dipeptidyl peptidase-4 inhibitors as urea dose-increasing therapy in patients with 186. Conason A, Teixeira J, Hsu C-H, Puma L, monotherapy or add-on to metformin in patients inadequately controlled type 2 diabetes using Knafo D, Geliebter A. Substance use following with type 2 diabetes mellitus: a systematic re- metformin and sulfonylurea (VISUAL study): a bariatric weight loss surgery. JAMA Surg 2013; view and meta-analysis. Diabetes Obes Metab randomized trial. Diabetes Res Clin Pract 2015; 148:145–150 2018;20:113–120 109:141–148 187. Dawes AJ, Maggard-Gibbons M, Maher AR, 198. Kawalec P, Mikrut A, Łopuch S. The safety 210. Moses RG, Kalra S, Brook D, et al. A et al. Mental health conditions among patients of dipeptidyl peptidase-4 (DPP-4) inhibitors or randomized controlled trial of the efficacy and seeking and undergoing bariatric surgery: a sodium-glucose cotransporter 2 (SGLT-2) inhib- safety of saxagliptin as add-on therapy in patients meta-analysis. JAMA 2016;315:150–163 itors added to metformin background therapy with type 2 diabetes and inadequate glycaemic 188. Mechanick JI, Kushner RF, Sugerman HJ, in patients with type 2 diabetes mellitus: a sys- control on metformin plus a sulphonylurea. Di- et al. (2009) American Association of Clinical tematic review and meta-analysis. Diabetes Metab abetes Obes Metab 2014;16:443–450 Endocrinologists, The Obesity Society, and Amer- Res Rev 2014;30:269–283 211. Moses RG, Round E, Shentu Y, et al. A ican Society for Metabolic & Bariatric Surgery 199. Leiter LA, Yoon K-H, Arias P, et al. Canagli- randomized clinical trial evaluating the safety medical guidelines for clinical practice for the flozin provides durable glycemic improvements and efficacy of sitagliptin added to the combi- perioperative nutritional, metabolic, and non- and body weight reduction over 104 weeks nation of sulfonylurea and metformin in patients surgical support of the bariatric surgery patient. versus glimepiride in patients with type 2 di- with type 2 diabetes mellitus and inadequate Obesity (Silver Spring) 2009;17(Suppl. 1):S1–S70, abetes on metformin: a randomized, double- glycemic control. J Diabetes 2016;8:701–711 v blind, phase 3 study. Diabetes Care 2015;38: 212. Singh S, Wright EE Jr, Kwan AYM, et al. 189. Mechanick JI, Youdim A, Jones DB, et al. 355–364 Glucagon-like peptide-1 receptor agonists com- (2013) Clinical practice guidelines for the peri- 200. Del Prato S, Nauck M, Duran-Garcia´ S, et al. pared with basal insulins for the treatment of operative nutritional, metabolic, and nonsurgical Long-term glycaemic response and tolerability type 2 diabetes mellitus: a systematic review and support of the bariatric surgery patient–2013 of dapagliflozin versus a sulphonylurea as add- meta-analysis. Diabetes Obes Metab 2017;19: update: cosponsored by American Association of on therapy to metformin in patients with type 2 228–238 Clinical Endocrinologists, The Obesity Society, diabetes: 4-year data. Diabetes Obes Metab 213. Levin PA, Nguyen H, Wittbrodt ET, Kim SC. and American Society for Metabolic & Bariatric 2015;17:581–590 Glucagon-like peptide-1 receptor agonists: a Surgery. Obesity (Silver Spring) 2013;21(Suppl. 201. Ridderstrale˚ M, Andersen KR, Zeller C, Kim systematic review of comparative effectiveness 1):S1–S27 G, Woerle HJ, Broedl UC; EMPA-REG H2H-SU trial research. Diabetes Metab Syndr Obes 2017;10: 190. Phung OJ, Sobieraj DM, Engel SS, Rajpathak investigators. Comparison of empagliflozin and 123–139 SN. Early combination therapy for the treatment glimepiride as add-on to metformin in patients 214. Abd El Aziz MS, Kahle M, Meier JJ, Nauck of type 2 diabetes mellitus: systematic review with type 2 diabetes: a 104-week andomized, MA. A meta-analysis comparing clinical effects and meta-analysis. Diabetes Obes Metab 2014; active-controlled, double-blind, phase 3 trial. of short- or long-acting GLP-1 receptor agonists 16:410–417 Lancet Diabetes Endocrinol 2014;2:691–700 versus insulin treatment from head-to-head 191. Abdul-Ghani MA, Puckett C, Triplitt C, et al. 202. Zhang Y, Hong J, Chi J, Gu W, Ning G, Wang studies in type 2 diabetic patients. Diabetes Initial combination therapy with metformin, W. Head-to-head comparison of dipeptidyl Obes Metab 2017;19:216–227 pioglitazone and exenatide is more effective peptidase-IV inhibitors and sulfonylureas – a 215. Eng C, Kramer CK, Zinman B, Retnakaran R. than sequential add-on therapy in subjects with meta-analysis from randomized clinical trials. Glucagon-like peptide-1 receptor agonist and new-onset diabetes. Results from the Efficacy and Diabetes Metab Res Rev 2014;30:241–256 basal insulin combination treatment for the Durability of Initial Combination Therapy for Type 2 203. Foroutan N, Muratov S, Levine M. Safety management of type 2 diabetes: a systematic Diabetes (EDICT): a randomized trial. Diabetes and efficacy of dipeptidyl peptidase-4 inhibitors review and meta-analysis. Lancet 2014;384: Obes Metab 2015;17:268–275 vs sulfonylurea in metformin-based combination 2228–2234 192. Vijayakumar TM, Jayram J, Meghana therapy for type 2 diabetes mellitus: systematic 216. Aroda VR, Bailey TS, Cariou B, et al. Effect of Cheekireddy V, Himaja D, Dharma Teja Y, review and meta-analysis. Clin Invest Med 2016; adding insulin degludec to treatment in patients Narayanasamy D. Safety, efficacy, and bioavail- 39:E48–E62 with type 2 diabetes inadequately controlled ability of fixed-dose combinations in type 2 di- 204. Chen K, Kang D, Yu M, et al. Direct head- with metformin and liraglutide: a double-blind abetes mellitus: a systematic updated review. to-head comparison of glycaemic durability of randomized controlled trial (BEGIN: ADD TO GLP-1 Curr Ther Res Clin Exp 2017;84:4–9 dipeptidyl peptidase-4 inhibitors and sulphonyl- Study). Diabetes Obes Metab 2016;18:663–670 193. Mearns ES, Saulsberry WJ, White CM, et al. ureas in patients with type 2 diabetes mellitus: a 217. Aschner P, Sethi B, Gomez-Peralta F, et al. Efficacy and safety of antihyperglycaemic drug meta-analysis of long-term randomized controlled Insulin glargine compared with premixed insulin regimensaddedtometforminandsulphonylurea trials. Diabetes Obes Metab 2018;20:1029– for management of insulin-na¨ıve type 2 diabetes therapy in type 2 diabetes: a network meta- 1033 patients uncontrolled on oral antidiabetic drugs: analysis. Diabet Med 2015;32:1530–1540 205. Zaccardi F, Dhalwani NN, Dales J, et al. the open-label, randomized GALAPAGOS study. 194. Henry RR, Murray AV, Marmolejo MH, Comparison of glucose-lowering agents after J Diabetes Complications 2015;29:838–845 Hennicken D, Ptaszynska A, List JF. Dapagliflozin, dual therapy failure in type 2 diabetes: a sys- 218. Rosenstock J, Fonseca V, Schinzel S, Dain metformin XR, or both: initial pharmacother- tematic review and network meta-analysis of MP, Mullins P, Riddle M. Reduced risk of hypo- apy for type 2 diabetes, a randomised controlled randomized controlled trials. Diabetes Obes glycemia with once-daily glargine versus twice- trial. Int J Clin Pract 2012;66:446–456 Metab 2018;20:985–997 daily NPH and number needed to harm with NPH 195. Buse JB, Peters A, Russell-Jones D, et al. Is 206. Downes MJ, Bettington EK, Gunton JE, to demonstrate the risk of one additional hypo- insulin the most effective injectable antihyper- Turkstra E. Triple therapy in type 2 diabetes; glycemic event in type 2 diabetes: evidence glycaemic therapy? Diabetes Obes Metab 2015; a systematic review and network meta-analysis. from a long-term controlled trial. J Diabetes 17:145–151 PeerJ 2015;3:e1461 Complications 2014;28:742–749 196. Vaccaro O, Masulli M, Nicolucci A, et al.; 207. Lee CMY, Woodward M, Colagiuri S. Triple 219. Tang H, Cui W, Li D, et al. Sodium- Thiazolidinediones Or Sulfonylureas Cardiovas- therapy combinations for the treatment of type glucose co-transporter 2 inhibitors in addi- cular Accidents Intervention Trial (TOSCA.IT) 2 diabetes – a network meta-analysis. Diabetes tion to insulin therapy for management of study group; Italian Diabetes Society. Effects Res Clin Pract 2016;116:149–158 type 2 diabetes mellitus: a meta-analysis of care.diabetesjournals.org Davies and Associates 33

randomized controlled trials. Diabetes Obes 226. Gough SCL, Bode BW, Woo VC, et al. One- 233. Giugliano D, Chiodini P, Maiorino MI, Metab 2017;19:142–147 year efficacy and safety of a fixed combination of Bellastella G, Esposito K. Intensification of insulin 220. Rosenstock J, Jelaska A, Zeller C, Kim G, insulin degludec and liraglutide in patients with therapy with basal-bolus or premixed insulin Broedl UC, Woerle HJ; EMPA-REG BASALTM trial type 2 diabetes: results of a 26-week extension regimens in type 2 diabetes: a systematic review investigators. Impact of empagliflozin added on to a 26-week main trial. Diabetes Obes Metab and meta-analysis of randomized controlled to basal insulin in type 2 diabetes inadequately 2015;17:965–973 trials. Endocrine 2016;51:417–428 controlled on basal insulin: a 78-week random- 227. Rosenstock J, Aronson R, Grunberger G, 234. Wang C, Mamza J, Idris I. Biphasic vs basal ized, double-blind, placebo-controlled trial. Di- et al.; LixiLan-O Trial Investigators. Benefits of bolus insulin regimen in type 2 diabetes: a abetes Obes Metab 2015;17:936–948 LixiLan, a titratable fixed-ratio combination of systematic review and meta-analysis of random- 221. Rosenstock J, Jelaska A, Frappin G, et al.; insulin glargine plus lixisenatide, versus insulin ized controlled trials. Diabet Med 2015;32:585– EMPA-REG MDI Trial Investigators. Improved glargine and lixisenatide monocomponents in 594 glucose control with weight loss, lower insulin type 2 diabetes inadequately controlled on oral 235. AnyanwaguU,MamzaJ,GordonJ, Donnelly doses, and no increased hypoglycemia with agents: the LixiLan-O randomized trial. Diabetes R, Idris I. Premixed vs basal-bolus insulin regimen empagliflozin added to titrated multiple daily Care 2016;39:2026–2035 in type 2 diabetes: comparison of clinical out- injections of insulin in obese inadequately con- 228. Aroda VR, Rosenstock J, Wysham C, et al.; comes from randomized controlled trials and trolled type 2 diabetes. Diabetes Care 2014;37: LixiLan-L Trial Investigators. Efficacy and safety of real-world data. Diabet Med 2017;34:1728–1736 1815–1823 LixiLan, a titratable fixed-ratio combination of 236. Aronson R, Reznik Y, Conget I, et al.; OpT2- 222. Yang W, Cai X, Gao X, Chen Y, Chen L, Ji L. insulin glargine plus lixisenatide in type 2 di- mise Study Group. Sustained efficacy of insulin Addition of dipeptidyl peptidase-4 inhibitors to abetes inadequately controlled on basal insulin pump therapy compared with multiple daily insulin treatment in type 2 diabetes patients: a and metformin: the LixiLan-L randomized trial. injections in type 2 diabetes: 12-month data meta-analysis. J Diabetes Investig 2018;9:813– Diabetes Care 2016;39:1972–1980 from the OpT2mise randomized trial. Diabetes 821 229. Garber AJ, Wahlen J, Wahl T, et al. Attain- Obes Metab 2016;18:500–507 223. Min SH, Yoon J-H, Hahn S, Cho YM. Com- ment of glycaemic goals in type 2 diabetes with 237. Kelly S, Martin S, Kuhn I, Cowan A, Brayne C, parison between SGLT2 inhibitors and DPP4 once-, twice-, or thrice-daily dosing with biphasic Lafortune L. Barriers and facilitators to the inhibitors added to insulin therapy in type 2 insulin aspart 70/30 (the 1-2-3 study). Diabetes uptake and maintenance of healthy behaviours diabetes: a systematic review with indirect com- Obes Metab 2006;8:58–66 by people at mid-life: a rapid systematic review. parison meta-analysis: SGLT2 or DPP4 inhibitor 230. Linjawi S, Lee B-W, Tabak O,¨ Lovdahl¨ S, PLoS One 2016;11:e0145074 with insulin. Diabetes Metab Res Rev 2017;33: Werther S, Abusnana S. A 32-week randomized 238. Gee PM, Greenwood DA, Paterniti DA, e2818 comparison of stepwise insulin intensification of Ward D, Miller LM. The eHealth Enhanced 224. Wysham CH, Lin J, Kuritzky L. Safety and biphasic insulin aspart (BIAsp 30) versus basal-bolus Chronic Care Model: a theory derivation ap- efficacy of a glucagon-like peptide-1 receptor therapy in insulin-na¨ıve patients with type 2 di- proach. J Med Internet Res 2015;17:e86 agonist added to basal insulin therapy versus abetes. Diabetes Ther 2018;9:1–11 239. Faruque LI, Wiebe N, Ehteshami-Afshar A, basal insulin with or without a rapid-acting in- 231. Raccah D, Huet D, Dib A, et al. Review of et al.; Alberta Kidney Disease Network. Effect of sulin in patients with type 2 diabetes: results basal-plus insulin regimen options for simpler telemedicine on glycated hemoglobin in di- of a meta-analysis. Postgrad Med 2017;129: insulin intensification in people with type 2 di- abetes: a systematic review and meta-analysis 436–445 abetes mellitus. Diabet Med 2017;34:1193–1204 of randomized trials. CMAJ 2017;189:E341– 225. Maiorino MI, Chiodini P, Bellastella G, 232. Rodbard HW, Visco VE, Andersen H, Hiort E364 Capuano A, Esposito K, Giugliano D. Insulin LC, Shu DH. Treatment intensification with step- 240. Lee SWH, Chan CKY, Chua SS, Chaiyakunapruk and glucagon-like peptide 1 receptor agonist wise addition of prandial insulin aspart boluses N. Comparative effectiveness of telemedicine combination therapy in type 2 diabetes: a sys- compared with full basal-bolus therapy (FullSTEP strategies on type 2 diabetes management: a tematic review and meta-analysis of randomized Study): a randomised, treat-to-target clinical systematic review and network meta-analysis. controlled trials. Diabetes Care 2017;40:614–624 trial. Lancet Diabetes Endocrinol 2014;2:30–37 Sci Rep 2017;7:12680