sign in sign up
<<
Home , Acarbose, Dysglycemia, Impaired fasting glucose, Insulin analog, Insulin degludec, Insulin glargine

Diabetes Care Volume 40, March 2017 291

Basal Can Be Controlled, Matthew C. Riddle but the Prandial Problem

PersistsdIt’stheNextTarget! CARE IN PERSPECTIVES Care 2017;40:291–300 | DOI: 10.2337/dc16-2380

Both basal and postprandial elevations contribute to the hyperglycemic exposure of diabetes, but current therapies are mainly effective in controlling the basal compo- nent. Inability to control postprandial limits success in maintaining overall glycemic control beyond the first 5 to 10 years after diagnosis, and it is also related to the weight gain that is common during therapy. The “prandial problem”dcomprising abnormalities of glucose and other metabolites, weight gain, and risk of hypoglycemiaddeserves more attention. Several approaches to prandial abnormalities have recently been studied, but the patient populations for which they are best suited and the best ways of using them remain incompletely defined. Encouragingly, several proof-of-concept studies suggest that short-acting - like 1 or the can be very effective in controlling postprandial hyperglycemia in in specific settings. This article reviews these topics and proposes that a greater proportion of available resources be directed to basic and clinical research on the prandial problem.

Widespread self-testing of glucose and, more recently, continuous glucose mon- itoring (CGM) have drawn attention to the daily patterns of blood glucose in both type 1 and type 2 diabetes. Seeing these patterns has allowed separate consideration of (basal) hyperglycemia and after-meal (postprandial) increments of glucose above basal levels. In entirely healthy individuals, fasting plasma glucose is rarely .100 mg/dL (5.5 mmol/L). Peak values after meals are ,140 mg/dL (7.8 mmol/L) and return quickly to basal levels. In the early stages of , high basal glucose (described as ) is the main abnormality for some people, whereas for others hyperglycemia after meals or an oral glucose load (impaired glucose tolerance) is more evident. These differences reflect varying mechanisms, may predict specific responses to initial treatments, and deserve further study. However, both basal and postprandial hyperglycemia are routinely present later in the natural history of type 2 diabetes, in part because sustained hyperglycemia can impair both the secretion and the various Division of Endocrinology, Diabetes & Clinical actions of insulin. Both are always present in . In the last 25 years we Nutrition, Oregon Health & Science University, have become very successful in managing basal hyperglycemia, but postprandial hy- Portland, OR perglycemia and its associated abnormalities remain untamed. Corresponding author: Matthew C. Riddle, riddlem@ ohsu.edu. STUDIES OF BASAL GLYCEMIC THERAPIES IN TYPE 2 DIABETES Received 7 November 2016 and accepted 10 In untreated type 2 diabetes, basal hyperglycemia is usually quantitatively greater than December 2016. further elevation of glucose after meals, especially when A1C is .8.0% (64 mmol/mol) © 2017 by the American Diabetes Association. (1–4). This is fortunate because most treatments for type 2 diabetes are more effective Readers may use this article as long as the work is properly cited, the use is educational and not in controlling basal hyperglycemia. , , , and for profit, and the work is not altered. More infor- basal (human NPH and insulins glargine, detemir, and degludec) all have rel- mation is available at http://www.diabetesjournals atively modest effects on postprandial hyperglycemia. The same is true for most of the .org/content/license. 292 The Prandial Problem Diabetes Care Volume 40, March 2017

newer agents, including dipeptidyl pepti- 12,500 patients with impaired fasting accompanying the use of glargine 100 dase 4 inhibitors, sodium–glucose co- glucose, impaired glucose tolerance, or units/mL described above suggests that transporter blockers, and longer-acting type 2 diabetes treated with no more individuals who have low BMI and low glucagon-like peptide 1 (GLP-1) receptor than one oral agent (and also selected basal insulin requirements might be agonists. Consequently, in a mixed popu- for high cardiovascular risk) were ran- most likely to have more stable glycemic lation of type 2 diabetes, glycemic expo- domized to either stepwise oral therapy control with a longer-acting insulin (12). sure from basal hyperglycemia is typically (in most cases metformin and ) When the risk of is reduced, dominant when A1C is high, and when or added to any prior oral further titration of dosage during long- control is improved by treatment, the therapy and titrated to specific targets. term follow-up might favor more frequent contribution from postprandial hypergly- For the 88% of patients in ORIGIN with attainment of A1C targets, but this expec- cemia becomes more prominent (5). Even overt diabetes at enrollment, the mean du- tation has been difficult to verify (17,18). when dietary or pharmacological treat- ration of diabetes was 4 years and median Long-acting GLP-1 receptor agonists ment maintains A1C values below the A1C was 6.6% (49 mmol/mol) (11). After are an alternative to basal insulin for pa- 7.0% (53 mmol/mol) target, residual post- 7 years of treatment A1C was 6.6% with tients with high risk of hypoglycemia and prandial hyperglycemia commonly limits the regimen based on oral therapy (with can be combined with basal insulins to attainment of A1C ,6.5%, the upper end ;10% needing to add insulin) and 6.3% provide further improvement of glycemic of the normal range (4,6). Persistence of (45 mmol/mol) with the basal insulin regi- control beyond what is possible with ei- postprandial hyperglycemia is most evi- men. Safety findings were reassuring even ther component alone (19). They have a dent when long-acting (basal) insulin is in this high-risk population and similar be- favorable effect on weight but, like basal added to one or more oral agents and tween the regimens except for a threefold insulin, only a modest effect on incre- systematically titrated to a fasting glucose greater frequency of hypoglycemia in the ments of glucose after meals (20,21). target. An analysis of self-measured glu- basal insulin arm. In summary, with an array of oral and cose profiles in type 2 diabetes no longer Hypoglycemia is associated with mor- injectable agents to control basal hyper- well controlled with oral agents, pooled bidity and mortality and is the leading glycemia, as well as the recent addition from six studies, found that 79% of the limitation of basal insulin therapy. In of basal insulins with improved profiles glucose elevation .100 mg/dL was due ORIGIN, younger age, lower BMI, and of action, type 2 diabetes can usually be to basal hyperglycemia (7). Six months lower attained dosage of glargine as basal managed well for the first decade after after initiation of basal insulin, 34% of insulin were independent predictors of diagnosis. In this time frame, the leading the hyperglycemic burden was basal and nonsevere hypoglycemia (12). In contrast, barriers to maintaining A1C ,7.0% are 66% was postprandial. In studies of this and consistent with other studies, events delay in diagnosis, clinical inertia in ad- kind, average fasting glucose after optimi- requiring assistance by another person vancing therapy (22), and difficulty iden- zation of basal insulin generally ranges were associated with older age and evi- tifying patients at greatest risk for between 100 and 130 mg/dL (6.6 and dence of renal or cognitive impairment hypoglycemia to allow use of alternative 7.3 mmol/L) and mean A1C is close to (12,13). Although a causal relationship treatment goals or methods. 7.0% (53 mmol/mol). Even when basal in- between hypoglycemia and poor medical sulin is expertly titrated, significant pran- outcomes has been difficult to verify in THE PRANDIAL PROBLEM PERSISTS dial hyperglycemia may persist. For type 2 diabetes, recurrent or severe hy- After longer duration of type 2 diabetes example, an experienced research group poglycemia calls for measures to mitigate (and for most patients with type 1 diabe- showed that for 61 patients whose mean risks (13,14). The simplest such measure tes), control of postprandial hyperglyce- baseline A1C was 9.5% (80 mmol/mol) on is relaxation of goals for A1C, typically by mia is the main unmet need, and it tends one or two oral agents, 36 weeks of titra- increasing the target range from ,7.0% to become more difficult over time. Be- tion of insulin glargine with continuation to between 7 and 8% for higher-risk pa- cause type 2 diabetes is now often diag- of metformin led to excellent control tients. Use of the new basal insulins (in- nosed before age 40, increasing numbers of fasting glucose (mean 104 mg/dL sulin degludec [Tresiba] and insulin of patients no longer maintain adequate [5.75 mmol/L]). However, the mean A1C glargine 300 units/mL [Toujeo]), both of glycemic control with basal therapies was 7.14% (54 mmol/mol), and half the which have longer and flatter glucose- alone at an age when they should have group still had A1C .7%, mainly because lowering profiles, may reduce the risk several decades of productive life ahead, of daytime hyperglycemia (8). of hypoglycemia while maintaining provided microvascular complications Attainment of A1C levels ,7.0% is desired levels of glycemic control. Di- can be prevented. more likely when basal insulin is started rect comparisons of degludec (15) and The usual recommendation for treat- before A1C is markedly elevated. Analysis glargine 300 units/mL (16) with glargine ment intensification in this setting has of data from .2,000 participants in 100 units/mL (Lantus) as ongoing basal been to start basal-bolus insulin therapy 12 studies showed that when baseline therapy in type 2 diabetes have shown or at least twice-daily injection of pre- A1C was between 7 and 8% (mean 7.6% ;30% lower risk of confirmed hypoglyce- mixed insulins. However, several lines of [60 mmol/mol]), a level ,7.0%wasreached mic events at night and smaller reduc- evidence suggest this may no longer be in 6 months by 75% of patients (9). The tions of hypoglycemia at any time of the best approach. First, population-based Outcome Reduction With Initial Glargine day. Presumably, a subset of patients studies in the U.S. show that although the Intervention (ORIGIN) trial further veri- would be especially likely to benefit proportions of patients who are able to fied the safety and efficacy of timely ini- from using one of the new basal insulins. maintain desired A1C levels by using tiation of basal therapies (10). More than The analysis of predictors of hypoglycemia lifestyle alone or with oral agents have care.diabetesjournals.org Riddle 293

increased in the last decade, insulin-treat- meals, among them GLP-1, which also control (34). Practice guidelines from the ed patients continue to have poor control suppresses glucagon, slows gastric empty- American Diabetes Association also (23). Second, in the Action to Control Car- ing, and provides a more sustained satiety suggest ,180 mg/dL as a target for con- diovascular Risk in Diabetes (ACCORD) signal. Neural factors may further regulate trol of glucose levels 1–2 h after a meal trial and the Veterans Affairs Diabetes Trial hepatic glucose production and peripheral (35). Selection of 180 mg/dL (10 mmol/L) (VADT), which enrolled patients with long- insulin sensitivity. In type 2 diabetes, peak by both groups as the upper end of a duration type 2 diabetes and high cardio- levels of insulin are lower than normal and nominally desirable range speaks to the vascular risk, seeking excellent glycemic occur 90–120 min after the meal begins difficulty of controlling postprandial hy- control with extensive use of basal-bolus rather than in the first 30 min. The rise of perglycemia with current methods. An insulin had mixed results. Microvascular amylin is also delayed and reduced and early report of CGM measurements for outcomes and some nonfatal cardiovascu- thus may fail to suppress glucagon, regu- a group of 15 patients with type 1 and lar events were reduced (24,25), but all- late gastric emptying, and limit food in- 15 with type 2 diabetes whose mean cause mortality increased 20% during take during the usual duration of the A1C was 7.5% (58 mmol/mol) found that intensive glycemic therapy in ACCORD meal. Secretion or action of GLP-1 may glucose was above 180 mg/dL 33% of the (25) and significant hypoglycemia and also be impaired. As for basal glucose, time (36). In addition to limiting attain- weight gain occurred in both studies. mechanisms for controlling postprandial ment of A1C targets, this much variability Third, studies comparing ways of intensi- glucose are more severely disturbed in beyond the desired range is suspected of fying insulin therapy for type 2 diabetes type 1 diabetes. When both endogenous contributing to both the microvascular have shown that basal-bolus treatment insulin and amylin are entirely lacking, and cardiovascular complications of dia- offers little further improvement of A1C prandial insulin treatment faces serious betes, independent of the mean level of when compared with simpler insulin regi- challenges. control reflected by A1C (37). mens (26,27). For all these reasons, it is In both type 1 diabetes and type 2 di- time to reconsider the prandial problem abetes requiring prandial therapy, the size TACTICS TO IMPROVE PRANDIAL in type 2 diabetes, which includes interre- and timing of a prandial insulin dose THERAPY lated difficulties with hyperglycemia, hy- should match the needs posed by each Behavioral interventions for postpran- poglycemia, and weight control. meal. However, meals differ widely in dial hyperglycemia and weight control, composition, size, and timing. Faulty dos- although centrally important, are be- PRANDIAL PHYSIOLOGY AND ing decisions lead to a mismatch between yond the scope of this article. For the PATHOPHYSIOLOGY the meal and the profile of insulin deliv- present discussion, several other ap- Growing knowledge of the mechanisms ered, and either marked postprandial hy- proaches deserve comment. that normally regulate plasma glucose perglycemia or later hypoglycemia can Metabolic Surgery has clarified the problems posed by both occur. Even when prandial insulin dosing Gastrointestinal procedures intended to basal and postprandial hyperglycemia is accurate, the resulting smaller increase alter food intake and absorption and (28–30). Fasting glucose is tightly con- of glucose may lead to further suppression otherwise improve prandial physiology trolled through regulation of hepatic glu- of endogenous insulin and amylin, favor- are now entering the mainstream of cose production by variable release of ing reduced satiety and thus continued treatment options (38). The mechanisms insulin into the portal vein, with modula- calorie intake. These difficulties contribute underlying their effects are not well un- tion of insulin’shepaticeffectsbyglucagon to the postprandial hyperglycemia, later derstood (39), but a large body of infor- and free fatty acids (FFA). In type 2 diabe- hypoglycemia, and continuing weight mation on short-term clinical outcomes tes, basal insulin secretion is impaired and gain that are commonly seen during is available. Diabetes cannot be said to FFA and glucagon levels are high during basal-bolus insulin treatment. In addition, be cured by this approach, but very im- fasting. Injection of long-acting insulin sup- frequent testing of glucose (or CGM) to pressive weight loss and improvement of presses hepatic glucose production by act- guide prandial dosing and the multiple in- prandial and overall glycemic control are ing directly on the and indirectly by jections needed are burdens that hinder common (40). A recent joint statement reducing FFA release from adipose tissues. regimen . by international diabetes organizations In type 1 diabetes, abnormalities of basal has endorsed metabolic surgery as a glucose regulation are more marked and INFORMATION OBTAINED BY CGM treatment “recommended” for patients delivery of basal insulin that closely For clinical care, CGM facilitates individu- with type 2 diabetes and BMI 40 kg/m2 matches physiological needs is mandatory alized decisions on prandial dosing, and or higher and “considered” under certain for good control. Analogs of human insulin for research, it offers new quantitative conditions for individuals with less se- and improved delivery devices have end points for testing treatment regi- vere obesity (41). However, long-term greatly helped in meeting this need. mens. Various formulas have been pro- assessment of the risks versus benefits A more complex set of mechanisms posed to describe glycemic variability, of this approach is not yet available. comes into play after meals. Insulin secre- and there is active debate about how to tion rapidly and markedly increases, ac- use them (31–33). Among the most ap- a-Glucosidase Inhibitors companied by cosecretion of amylin (the pealing is the “time in range.” An expert a-Glucosidase inhibitors (AGIs) are not other b-cell ), which suppresses panel recently proposed the percentage widely used in the U.S. but have greater glucagon, slows gastric emptying, and of measurements in a day between acceptance in some other countries. Their acutely signals satiety. Various other gas- 70 and 180 mg/dL (3.9 and 10.0 mmol/L) leading drawbacks are gastrointestinal trointestinal are modulated by as a summary indicator of glycemic side effects, mainly flatulence and 294 The Prandial Problem Diabetes Care Volume 40, March 2017

. They can reduce postprandial sustained improvement of glycemic con- Amylin Receptor Agonists hyperglycemia in type 2 diabetes by trol in routine management of either type Even less attention has been directed to ;50% and are effective in combination 1 or type 2 diabetes remains unknown. pramlintide (Symlin), an amylin analog with therapies that target basal glucose (62). This agent is approved in the U.S. control (42). Secondary analyses of data Closed-Loop Insulin Delivery Systems as an adjunct to basal-bolus treatment Computerized algorithms linking CSII to from the Study to Prevent Non-Insulin- of type 1 or type 2 diabetes. It is given CGM are under intense scrutiny for Dependent-Diabetes (STOP-NIDDM) as fixed doses by type 1 diabetes (50). The hardware and suggested that might reduce before meals, has a short duration of ac- software for such systems continue to im- cardiovascular risk for patients with im- tion, and markedly blunts postprandial prove, and there are encouraging reports paired glucose tolerance or type 2 dia- hyperglycemia by suppressing glucagon of their use for ambulatory patients betes (43). A large study testing this and slowing gastric emptying. Like the (51,52). In addition to the pharmacoki- hypothesis is now underway (44). If car- GLP-1 agonists, it reduces food intake netics of currently used insulins, limita- diovascular benefit is confirmed, use of and favors weight loss. Pramlintide’sdis- tions of these systems include the cost AGIs for patients with type 2 diabetes advantages for its current indications in- of the devices themselves and of clinical and adequate basal control but persist- clude the need for several daily injections support for their use, the risk of system ing postprandial hyperglycemia should in addition to multiple injections of insu- failure, and observations that, to date, be reconsidered. lin, frequent dosing decisions, and in- substantial glycemic variability related to creased risk of or hypoglycemia. Stepwise Addition of Prandial Insulin meals remains. Whether such systems will Because of these difficulties and limited Many patients with type 2 diabetes ad- be widely acceptable for type 2 diabetes is marketing efforts, it is not widely used. vised to use basal-bolus insulin therapy unknown, as is whether weight gain asso- do not consistently follow the treatment ciated with their use can be prevented. PROOF-OF-CONCEPT STUDIES plan. To address this problem, a simpler, Systems supplementing insulin infusion FOR SHORT-ACTING GLP-1 stepwise approach to adding prandial in- with timely delivery of subcutaneous glu- AGONISTS AND AN AMYLIN sulin to basal therapy has been proposed cagon to protect against hypoglycemia RECEPTOR AGONIST (45). Mean A1C reductions of 0.3% to have yielded promising early results (53). Recent studies of GLP-1 and amylin agonists 0.5% after adding a single prandial injec- added to basal insulin for type 2 diabetes tion before breakfast or another impor- GLP-1 Receptor Agonists Most studies of currently available GLP-1 have been very encouraging. Key findings tant meal are typical and sometimes agonists have focused on the longer- from these are summarized in Table 1. sufficient (46). At the time of initiation acting products. (Victoza), of basal insulin for type 2 diabetes or Versus Placebo extended-release exenatide (Bydureon), soon after, adding a single injection of In one study (63), 261 patients with (Trulicity), and prandial insulin may be as effective as type 2 diabetes with elevated A1C on oral (Tanzeum) all have relatively long dura- full basal-bolus treatment and also as ef- therapies plus insulin glargine (Lantus) tion of action and greater effects on basal fective as two injections of premixed in- were randomized to add twice-daily than on postprandial glucose (54–57). All sulin but with less risk of hypoglycemia exenatide (Byetta) or placebo injec- are very effective in reducing A1C in pa- (27,47). Of course, this approach is not tions before breakfast and dinner tients who start with high levels, and they appropriate for type 1 diabetes. (Fig. 1). Glargine was titrated during have desirable effects on weight. Their 30 weeks of treatment. Mean baseline More Rapidly Absorbed Prandial main side effects are nausea and other A1C was .8.0% in both groups, and Insulins gastrointestinal complaints, and up to the reduction from baseline was 0.69% Normal secretion of endogenous insulin 20% of patients discontinue treatment greater with exenatide. After treatment, increases plasma levels sooner after a in clinical studies. They are strongly mar- A1C #7.0% was attained by 60% of the meal than injection of the current gener- keted as alternatives to basal insulin. group assigned to exenatide and 35% of “ ” ation of rapid-acting insulin analogs. In contrast, the shorter-acting GLP-1 the placebo group. Body weight in- fi fi Hence, insuf ciently rapid onset of pran- agonists, unmodi ed exenatide (Byetta) creased with placebo but decreased dial insulin has been proposed as the and (Adlixin), have received with exenatide, with a between-treatment main limitation of basal-bolus therapy, limited attention (58,59). When given difference of 2.7 kg. Gastrointestinal com- whether by multiple injections or by con- just prior to a meal, both slow gastric plaints were more frequent with exena- tinuous subcutaneous insulin infusion emptying, prevent an inappropriate rise tide, but the frequency of hypoglycemia (CSII). Also, the subcutaneous depot of of glucagon, potentiate insulin secretion, did not differ. Most of the between-group insulins aspart, lispro, or glulisine may and thereby markedly blunt postprandial difference in A1C was due to blunting of continue to increase plasma levels longer hyperglycemia. They also favor weight the postprandial increments of glucose af- than necessary to match needs after loss. Dose-ranging studies have shown ter the exenatide doses at breakfast and fi small meals containing mainly re ned that these agents can reduce postprandial dinner. . Insulin formulations that hyperglycemia at low doses (e.g., 5 mg are more rapidly absorbed after subcuta- before a meal) that rarely cause gastroin- Exenatide Versus Rapid-Acting Insulin neous injection are under study (48), and testinal side effects (60,61). In principle, Another study directly compared basal an inhalable formulation with rapid early their best use should be for patients with insulin plus twice-daily exenatide with uptake (49) is now available as well. good control of basal glucose but persist- conventional basal-bolus therapy of Whether these modifications will lead to ing postprandial hyperglycemia. type 2 diabetes (64). All patients had care.diabetesjournals.org Riddle 295

Table 1—Proof-of-concept studies of short-acting GLP-1 and amylin receptor agonists added to basal insulin for prandial therapy Participants, Δ Weight N (duration Baseline Attained (kg) vs. Authors (ref.) diabetes, years) Prior therapy Comparison A1C (%) A1C (%) comparator Other comments Buse et al. (63) 261 (mean 12) Oral agents + Exenatide b.i.d. vs. 8.35, 8.53 6.6, 7.5 22.7 More nausea with basal insulin placebo b.i.d. exenatide Diamant et al. (64) 627 (median 11) Oral agents + Exenatide b.i.d. vs. 8.3, 8.2 7.2, 7.2 24.6 More nausea but less basal insulin rapid-acting insulin hypoglycemia with t.i.d. exenatide FLAT- Trial 102 (median 15) Oral agents + Exenatide b.i.d./t.i.d. 7.3, 7.4 7.1, 7.2 25.45 Less glycemic Investigators (65) basal-bolus vs. rapid-acting insulin variability with insulin t.i.d. exenatide Riddle et al. (66) 446 (mean 9) Oral agents + Lixisenatide q.d. vs. 7.6, 7.6 7.0, 7.3 20.9 More nausea and basal-bolus placebo q.d. hypoglycemia with insulin lixisenatide Rosenstock et al. (67) 298 (mean 12) Oral agents + Lixisenatide q.d. vs. 7.8, 7.7, 7.8 7.2, 7.2, 7.0 21.7, 22.0 More nausea but less basal-bolus rapid-acting insulin hypoglycemia with insulin q.d. vs. rapid-acting lixisenatide insulin t.i.d. Riddle et al. (68) 113 (mean 10) Oral agents 6 Pramlintide t.i.d. vs. 8.2, 8.3 7.2, 7.0 24.7 More nausea but less basal insulin rapid-acting insulin hypoglycemia with t.i.d. pramlintide

previously used insulin glargine (Lantus) exenatide). Glycemic variability, defined diabetes of 12 years who were previously with oral agents, and after 12 weeks of as the coefficient of variation of glucose taking basal insulin and oral agents. After titration of glargine dosage, 627 of them levels measured by CGM, was the primary optimization of insulin glargine (Lantus) were randomized to add either exena- end point. Variability was reduced more for 12 weeks, they were randomized to tide (Byetta) or . From with the exenatide-based regimen. Glu- addition of lixisenatide once daily, rapid- mean baseline levels of 8.2% and 8.3% cose values were in the 70–180 mg/dL acting insulin once daily, or rapid-acting (66 and 67 mmol/mol), both treatments range75%ofthetimewithexenatideand insulin three times daily for 26 more reduced A1C to 7.2% (55 mmol/mol). basal insulin and 71% of the time with weeks (67). From a mean A1C of 7.6% Weight increased with lispro but de- basal-bolus insulin. Weight was unchanged (60 mmol/mol) after glargine titration, creased with exenatide, with a be- with basal-bolus therapy but declined levels declined to 7.2, 7.2, and 7.0% (55, tween-treatment difference of 4.6 kg. with exenatide and basal insulin, with 55, 53 mmol/mol) with lixisenatide and Hypoglycemia was less frequent but a between-treatment difference of 5.5 kg. the two rapid-acting insulin regimens, re- gastrointestinal symptoms more com- spectively. The change of weight during Lixisenatide Versus Placebo randomized treatment favored the lixise- mon with exenatide. Reduced postpran- In patients with type 2 diabetes previ- natide arm versus the two insulin arms dial increments accounted for most of ously treated with basal insulin and oral by 21.7 and 22.0 kg. the glycemic effect of both regimens. agents, insulin glargine (Lantus) was A smaller but similarly designed study titrated for 12 weeks, and patients with Pramlintide Versus Rapid-Acting examined the effects of exenatide versus A1C .7.0% (n = 446) were randomized Insulin rapid-acting insulin on glucose profiles to add lixisenatide or placebo injected A randomized, open-label comparison of measured by CGM (65) (Fig. 2). After a once-daily before breakfast (66). After pramlintide with rapid-acting insulin, with run-in of 8 to 12 weeks on basal-bolus 24 weeks, mean A1C had declined from each of these prandial therapies added to insulin, 102 patients with long-duration 7.6 to 7.0% (60 to 53 mmol/mol) with titrated basal insulin, was performed in (median 15 years) type 2 diabetes and lixisenatide and to 7.3% (56 mmol/mol) 133 patients with type 2 diabetes (68) (Fig. high cardiovascular risk were randomized with placebo, with a difference of 3). After 24 weeks, A1C was similarly reduced to continue basal-bolus therapy or to 0.32%. Values of A1C #7.0% were at- from 8.2 and 8.3% (66 and 67 mmol/mol) to switch to insulin glargine (Lantus) plus tained by 56% of patients taking lixise- 7.2 and 7.0% (55 and 53 mmol/mol) with the two or three mealtime doses of exena- natide versus 39% of those assigned to pramlintide and rapid-acting insulin regi- tide (Byetta). Adjustment of placebo. Gastrointestinal symptoms mens, respectively. Reductions of fasting glu- during 26 weeks of randomized treat- and hypoglycemia were more frequent cose and postprandial glucose increments ment attained the intended goal of with lixisenatide. were not significantly different between equivalent A1C between 6.7 and 7.3% the regimens. Weight increased with rapid- (50 and 56 mmol/mol) with each regi- Lixisenatide Versus Rapid-Acting acting insulin but not pramlintide, with a dif- men (final mean values 7.2% [55 mmol/ Insulin ference of 4.7 kg. Pramlintide caused more mol] with basal-bolus insulin and 7.1% This head-to-head study examined 298 nausea but less hypoglycemia than rapid- [54 mmol/mol] with basal insulin and patients with a mean duration of type 2 acting insulin. 296 The Prandial Problem Diabetes Care Volume 40, March 2017

NEW PERSPECTIVES: REDIRECTING EFFORT FROM BASAL TO PRANDIAL CONTROL The preceding discussion suggests that basal hyperglycemia is no longer the main problem. Instead, the prandial problem deserves more attention, and there are promising but incompletely explored options for addressing it. From both the commercial and the pub- lic health points of view, the return on investment could be improved by shift- ing effort and resources from basal to prandial therapy.

Immediate Clinical Measures Together with earlier diagnosis and timely initiation of basal therapy for type 2 di- abetes, prandial treatment could be added earlier than in the past. The glyce- mic increment persisting after optimized basal therapy could be blunted by adding, with one or more meals, an AGI, a dose of rapid-acting or regular human insulin, or a short-acting GLP-1 agonist. Progressing stepwise in this fashion from a basal- only regimen to basal plus prandial therapy before attempting a full basal- prandial strategy is simpler and likely to be better tolerated. At present, continu- ation of basal therapies used alone, even with appropriate titration of dose, typi- cally allows A1C levels to rise gradually to levels well above 7.0% after 10 or more years while additional treatment to re- store A1C to the ,7.0% “target” is de- layed. A better approach might be to consider 7% A1C a “ceiling” and to seek values below this level by timely addition of prandial therapy. Overtitration of basal insulin leading to hypoglycemia would be less common if prandial ther- apy were regarded as easy and desirable rather than difficult and dangerous. Even without incorporating new methods, ear- lier prandial therapy might extend the time excellent glycemic control is main- tained, leading to reduction of long-term microvascular and cardiovascular risks.

More Skillful Use of GLP-1 Agonists for Type 2 Diabetes Better awareness of the differences be- Figure 1—Comparison of treatment with basal insulin glargine plus either twice-daily injections tween long-acting and short-acting drugs of exenatide (black lines) or placebo (green lines), for type 2 diabetes previously treated with in this class is needed. Demonstration of a oral agents and basal insulin (adapted with permission from Buse et al. [63]). Dosage of glargine was titrated during 30 weeks of randomized treatment. The figure shows responses of A1C (A), 22% reduction of cardiovascular mortality weight change from baseline (B), and 7-point self-measured glucose profiles before (dashed with liraglutide (in the Liraglutide Effect lines) and after (solid lines) treatment (C). Error bars display 95% CI. *P , 0.001 and †P , 0.01 and Action in Diabetes: Evaluation of car- for between-treatment differences. PP, postprandial. diovascular outcome Results [LEADER] trial [69]) has increased the attractiveness of this agent, but this GLP-1 agonist does care.diabetesjournals.org Riddle 297

patients whose postprandial hyperglyce- mia seriously impairs attainment of A1C goals to limit microvascular risk. Exena- tide or lixisenatide should, for many pa- tients with type 2 diabetes, be as effective as prandial insulin without causing weight gain and hypoglycemia. Gastrointestinal symptoms could be limited by minimizing doses. Delivery devices allowing selection of doses between 2.5 and 10 mgwould permit the greatest flexibility, but they have not been made available. The best results are likely to derive from stepwise addition of up to three prandial doses (65). Pre- or postprandial glucose tests would not routinely be needed to guide dosing, allowing greater convenience and perhaps lower additional cost than pran- dial insulin. Further well-designed studies of this approach are needed.

Coadministration of an Amylin Analog With Insulin for Type 1 Diabetes Perhaps the most remarkable opportunity is for development of a fixed-ratio formu- lation of an amylin analog with insulin. Under normal conditions, endogenous in- sulin and amylin are cosecreted by b-cells in the same patterns during fasting and with meals. Use of pramlintide (off-label) as the only prandial therapy supplement- ing newly initiated basal insulin in the study described above (66) is a significant proof of concept, but not likely to be ef- fective for longer-duration type 2 diabetes or type 1 diabetes. Especially in type 1 diabetes, where little or no b-cell func- tion remains, coordinated replacement therapy for this bihormonal deficiency deserves further study. Preliminary re- ports show no obvious barriers to a com- bined formulation of pramlintide with regular human insulin for clinical use (71,72), and developing one could have Figure 2—Comparison of treatment with either basal insulin glargine plus exenatide injected both clinical and commercial appeal. Con- two or three times daily or basal-bolus insulin plus metformin, for long-duration type 2 diabetes tinuous delivery of a coformulation might previously stabilized on basal-bolus insulin (adapted with permission from FLAT-SUGAR Trial have several advantages. Titration deci- Investigators [65]). During randomized treatment for 26 weeks the aim was to keep A1C be- sions would be simpler than with separate tween 6.7 and 7.3%. The figure shows CGM measurements at baseline (black line) and 26 weeks (red line) for patients using basal-bolus insulin (A), CGM at baseline (black line) and 26 weeks dosing. Postprandial hypoglycemia might (blue line) for those using basal insulin with exenatide (B), and weight changes during random- be reduced, and weight control would ized treatment (C). Mean weights and SE are shown. surely be improved compared with insulin therapy alone. Nausea might be less fre- quent than with intermittent use of pram- not reliably control postprandial hyper- benefit was seen with once-daily lixise- lintide. Combined with a closed-loop glycemia. Whether long-term use of natide in the very high-risk population control system, this approach might attain short-acting GLP-1 agonists, which have studied in the Evaluation of Lixisenatide near-normal glycemic control more reli- better prandial effects (20,21), can also in Acute Coronary Syndrome (ELIXA) trial ably than an insulin-only or insulin-with- reduce cardiovascular risk in a similar (70). However, the short-acting GLP-1 glucagon system. These are testable population is unknown, and no such agents might be an excellent option for hypotheses. 298 The Prandial Problem Diabetes Care Volume 40, March 2017

Duality of Interest. M.C.R. has received re- search grant support through Oregon Health & Science University from AstraZeneca, Eli Lilly, and and honoraria for consulting or speaking from AstraZeneca, Biodel, Elcelyx, Eli Lilly, GlaxoSmithKline, Sanofi, Theracos, and Valeritas. These dualities of interest have been reviewed and managed by Oregon Health & Sci- ence University. No other potential conflicts of interest relevant to this article were reported.

References 1. Reaven GM, Hollenbeck C, Jeng CY, Wu MS, Chen Y-DI. Measurement of plasma glucose, free fatty acid, lactate, and insulin for 24 h in patients with NIDDM. Diabetes 1988;37:1020–1024 2. O’Meara NM, Shapiro ET, Van Cauter E, Polonsky KS. Effect of glyburide on re- sponsiveness to glucose in non-insulin-dependent diabetes mellitus. Am J Med 1990;89(2A):11S– 16S; discussion 51S–53S 3. Riddle MC. Evening insulin strategy. Diabe- tes Care 1990;13:676–686 4. Monnier L, Colette C, Dejager S, Owens DR. Figure 3—Comparison of treatment with basal insulin glargine plus either prandial injections of Near normal HbA1c with stable glucose homeo- pramlintide (solid squares) or a rapid-acting (open circles), for type 2 diabetes stasis: the ultimate target/aim of diabetes ther- previously treated with oral agents 6 basal insulin (adapted with permission from Riddle et al. apy. Rev Endocr Metab Disord 2016;17:91–101 [68]). Dosage of glargine was titrated during 24 weeks of randomized treatment. The figure 5. Monnier L, Lapinski H, Colette C. Contributions shows dosage of glargine with both regimens (solid lines) and rapid-acting insulin plus glargine in of fasting and postprandial glucose increments to the basal-bolus arm (dashed line) (A), responses of A1C (B), change of weight from baseline (C), overall diurnal hyperglycemia of type 2 diabetic and responses of fasting glucose (D). Means and SE are shown. ***P , 0.001 and **P , 0.01 for patients. Diabetes Care 2003;26:881–885 between-treatment differences. 6. Monnier L, Colette C, Dejager S, Owens D. Re- sidual dysglycemia when at target HbA(1c) of 7% (53 mmol/mol) in persons with type 2 diabetes. Development of Other Gut-Peptide be studied, along with coformulation of Diabetes Res Clin Pract 2014;104:370–375 Related Therapies an amylin agonist with insulin. In addition, 7. Riddle M, Umpierrez G, DiGenio A, Zhou R, This topic deserves a separate review, but investment in closed-loop delivery sys- Rosenstock J. Contributions of basal and post- several points are clear. There are many tems and surgical modification of intesti- prandial hyperglycemia over a wide range of fi gastrointestinal peptide , some A1C levels before and after treatment intensi - nal function should proceed but not be cation in type 2 diabetes. Diabetes Care 2011; with known structures and metabolic unlimited. They are what Lewis Thomas 34:2508–2514 functions (73,74). Research on their roles called “half-way technologies” (75). While 8. Yki-Jarvinen¨ H, Kauppinen-Makelin¨ R, in health and disease will be central to amazingdwe are lucky to have them Tiikkainen M, et al. Insulin glargine or NPH com- understanding the genesis of obesity d bined with metformin in type 2 diabetes: the today they will become obsolete (like – and diabetes and perhaps the links be- LANMET study. Diabetologia 2006;49:442 451 tuberculosis sanatoria and iron lungs) 9. Riddle MC, Vlajnic A, Zhou R, Rosenstock J. tween these and when research uncovers the root causes Baseline HbA1c predicts attainment of 7.0% and . Formulations of these pep- of the problems they address. The pre- HbA1c target with structured titration of insulin tides, or synthetic analogs, will become sent challenge is to focus use of closed- glargine in type 2 diabetes: a patient-level anal- available for research and clinical use. ysis of 12 studies. Diabetes Obes Metab 2013; loop CSII and metabolic surgery on – fl 15:819 825 Their dosing may be more exible than subgroups of patients for whom they of- 10. Gerstein HC, Bosch J, Dagenais GR, et al.; that of insulin, allowing alternative (non- fer the best balance of benefittocostand ORIGIN Trial Investigators. Basal insulin and car- injection) routes of delivery. risk. Other resources should be reserved diovascular and other outcomes in dysglycemia. to explore new insights into islet, gut, and N Engl J Med 2012;367:319–328 ALLOCATION OF RESOURCES 11. ORIGIN Trial Investigators. Characteristics as- brain physiology (76) to produce the de- , Where can we find the resources for sociated with maintenance of mean A1C 6.5% in finitive therapies of the future. Regulatory people with dysglycemia in the ORIGIN trial. Di- these new efforts? Product development and financial incentives toward this end abetes Care 2013;36:2915–2922 and basic research are expensive. How- are needed. The prandial problemd 12. ORIGIN Trial Investigators. Predictors of ever, some of the resources now used including postprandial hyperglycemia, nonsevere and severe hypoglycemia during glu- cose-lowering treatment with insulin glargine or for fine-tuning the treatment of basal hy- weight gain, and hypoglycemia caused by perglycemia may now be better used for standard drugs in the ORIGIN trial. Diabetes overreliance on injected insulindis an en- Care 2015;38:22–28 the prandial problem. We have multiple docrine and neurologic puzzle that calls for 13. Zoungas S, Patel A, Chalmers J, et al.; classes of oral and injected agents with further basic and clinical research. ADVANCE Collaborative Group. Severe hypoglyce- powerful effects on basal glucose, as mia and risks of vascular events and death. N Engl well as many choices in each class. Rather J Med 2010;363:1410–1418 14. ORIGIN Trial Investigators. Does hypoglycae- than more of these redundant therapies, Funding. Support for this work was from the mia increase the risk of cardiovascular events? A better ways to deliver prandial insulins Rose Hastings and Russell Standley Memorial report from the ORIGIN trial. Eur Heart J 2013;34: and short-acting GLP-1 agonists should Trusts. 3137–3144 care.diabetesjournals.org Riddle 299

15. Ratner RE, Gough SCL, Mathieu C, et al. Hy- basal-bolus up to three prandial injections. Diabe- 45. Raccah D, Bretzel RG, Owens D, Riddle M. poglycaemia risk with compared tes Obes Metab 2014;16:396–402 When basal insulin therapy is not enough: what with insulin glargine in type 2 and type 1 diabetes: 28. Riddle MC, Drucker DJ. Emerging therapies next? Diabetes Metab Res Rev 2007;23:257–264 a pre-planned meta-analysis of phase 3 trials. Di- mimicking the effects of amylin and glucagon- 46. Owens DR, Luzio SD, Sert-Langeron C, Riddle abetes Obes Metab 2013;15:175–184 like peptide 1. Diabetes Care 2006;29:435–449 MC. Effects of initiation and titration of a single 16. Ritzel R, Roussel R, Bolli GB, et al. Patient- 29. Defronzo RA. From the triumvirate to the pre-prandial dose of while con- levelmeta-analysisof the EDITION 1, 2, and 3 stud- ominous octet: a new paradigm for the treat- tinuing titrated insulin glargine in type 2 diabetes: ies: glycaemic control and hypoglycemia with ment of type 2 diabetes mellitus. Diabetes 2009; a6-month‘proof-of-concept’ study. Diabetes new insulin glargine 300 U/ml versus glargine 58:773–795 Obes Metab 2011;13:1020–1027 100 U/ml in people with type 2 diabetes. Diabetes 30. Nolan CJ, Damm P, Prentki M. Type 2 diabe- 47. Rodbard HW, Visco VE, Andersen H, Hiort LC, Obes Metab 2015;17:859–567 tes across generations: from pathophysiology to Shu DHW. Treatment intensification with step- 17. Rosenstock J, Fonseca V, Schinzel S, Dain M-P, prevention and management. Lancet 2011;378: wise addition of prandial boluses Mullins P, Riddle M. Reduced risk of hypoglycemia 169–181 compared with full basal-bolus therapy (FullSTEP with once-daily glargine versus twice-daily NPH 31. Service FJ. Glucose variability. Diabetes Study): a randomised, treat-to-target . and number needed to harm with NPH to dem- 2013;62:1398–1404 Lancet Diabetes Endrocinol 2014;2:30–37 onstrate the risk of one additional hypoglycemic 32. Rodbard D. Evaluating quality of glycemic 48. Heinemann L, Muchmore DB. Ultrafast-act- event in type 2 diabetes: evidence from a long- control: graphical displays of hypo- and hyper- ing insulins: state of the art. J Diabetes Sci Tech- term controlled trial. J Diabetes Complications glycemia, time in target range, and mean glu- nol 2012;6:728–742 2014;28:742–749 cose. J Diabetes Sci Technol 2015;9:56–62 49. Boss AH, Petrucci R, Lorber D. Coverage of 18. Home PD, Bolli GB, Mathieu C, et al. Modu- 33. Kovatchev B, Cobelli C. Glucose variability: prandial insulin requirements by means of an lation of insulin dose titration using a hypoglycae- timing, risk analysis, and relationship to hypo- ultra-rapid-acting inhaled insulin. J Diabetes mia-sensitive algorithm: insulin glargine versus glycemia in diabetes. Diabetes Care 2016;39: Sci Technol 2012;6:773–779 neutral protamine Hagedorn insulin in insulin-na- 502–510 50. Russell SJ, Beck RW. Design considerations ive people with type 2 diabetes. Diabetes Obes 34. Bergenstal RM, Ahmann AJ, Bailey T, et al. Rec- for artificial pivotal studies. Diabetes Metab 2015;17:15–22 ommendations for standardizing glucose reporting Care 2016;39:1161–1167 19. Gough SC, Bode B, Woo V, et al.; NN9068- and analysis to optimize clinical decision making in 51. Anderson SM, Raghinaru D, Pinsker JE, et al.; 3697 (DUAL-I) trial investigators. Efficacy and diabetes: the Ambulatory Glucose Profile (AGP). Control to Range Study Group. Multinational safety of a fixed-ratio combination of insulin de- Diabetes Technol Ther 2013;15:198–211 home use of closed-loop control is safe and effec- gludec and liraglutide (IDegLira) compared with 35. American Diabetes Association. Glycemic tive. Diabetes Care 2016;39:1143–1150 its components given alone: results of a phase 3, targets. Sec. 5. In Standards of Medical Care in 52. Tauschmann M, Allen JM, Wilinska ME, open-label, randomised, 26-week, treat-to-target Diabetesd2016. Diabetes Care 2016;39(Suppl. 1): et al. Day-and-night hybrid closed-loop insulin trial in insulin-naive patients with type 2 diabetes. S39–S46 delivery in adolescents with type 1 diabetes: a Lancet Diabetes Endocrinol 2014;2:885–893 36. Mazze RS, Strock E, Wesley D, et al. Charac- free-living, randomized clinical trial. Diabetes 20. Meier JJ, Rosenstock J, Hincelin-Mery´ A, terizing glucose exposure for individuals with nor- Care 2016;39:1168–1174 et al. Contrasting effects of lixisenatide and lir- mal glucose tolerance using continuous glucose 53. Bakhtiani PA, Zhao LM, El Youssef J, Castle JR, aglutide on postprandial glycemic control, gas- monitoring and ambulatory glucose profile anal- Ward WK. A review of artificial pancreas technol- tric emptying, and safety parameters in patients ysis. Diabetes Technol Ther 2008;10:149–159 ogies with an emphasis on bi-hormonal therapy. with type 2 diabetes on optimized insulin glar- 37. Hirsch IB. Glycemic variability and diabetes Diabetes Obes Metab 2013;15:1065–1070 gine with or without metformin: a randomized, complications: Does it matter? Of course it 54. Scott LJ. Liraglutide: a review of its use in open-label trial. Diabetes Care 2015;38:1263– does! Diabetes Care 2015;38:1610–1614 adult patients with type 2 diabetes mellitus. 1273 38. Cefalu WT, Rubino F, Cummings DE. Meta- Drugs 2014;74:2161–2174 21. Drucker DJ, Buse JB, Taylor K, et al.; bolic surgery for type 2 diabetes: changing the 55. Syed YY, McCormack PL. Exenatide ex- DURATION-1 Study Group. Exenatide once weekly landscape of diabetes care. Diabetes Care 2016; tended-release: an updated review of its use versus twice daily for the treatment of type 2 di- 39:857–860 in type 2 diabetes. Drugs 2015;75:1141–1152 abetes: a randomised, open-label, non-inferiority 39. Batterham RL, Cummings DE. Mechanisms 56. Burness CB, Scott LJ. Dulaglutide: a review study. Lancet 2008;372:1240–1250 of diabetes improvement following bariatric/ in type 2 diabetes. BioDrugs 2015;29:407–418 22. Khunti K, Nikolajsen A, Thorsted BL, Andersen metabolic surgery. Diabetes Care 2016;39: 57. Blair HA, Keating GM. Albiglutide: a review M, Davies MJ, Paul SK. Clinical inertia with regard 893–901 of its use in patients with type 2 diabetes melli- to intensifying therapy in people with type 2 di- 40. Schauer PR, Mingrone G, Ikramuddin S, tus. Drugs 2015;75:651–663 abetes treated with basal insulin. Diabetes Obes Wolfe B. Clinical outcomes of metabolic sur- 58. McCormack PL. Exenatide twice daily: a review Metab 2016;18:401–409 gery: efficacy of glycemic control, weight loss, of its use in the management of patients with 23. Stark Casagrande S, Fradkin JE, Saydah SH, and remission of diabetes. Diabetes Care 2016; type 2 diabetes mellitus. Drugs 2014;74:325–351 Rust KF, Cowie CC. The prevalence of meeting 39:902–911 59. Scott LJ. Lixisenatide: a review of its use in A1C, , and LDL goals among peo- 41. Rubino F, Nathan DM, Eckel RH, et al.; Del- patients with type 2 diabetes mellitus. BioDrugs ple with diabetes, 1988–2010. Diabetes Care egates of the 2nd Diabetes Surgery Summit. 2013;27:509–523 2013;36:2271–2279 Metabolic surgery in the treatment algorithm 60. Kolterman OG, Kim DD, Shen L, et al. Phar- 24. Hayward RA, Reaven PD, Wiitala WL, et al.; for type 2 diabetes: a joint statement by inter- macokinetics, , and safety of VADT Investigators. Follow-up of glycemic con- national diabetes organizations. Diabetes Care exenatide in patients with type 2 diabetes mel- trol and cardiovascular outcomes in type 2 di- 2016;39:861–877 litus. Am J Health Syst Pharm 2005;62:173–181 abetes. N Engl J Med 2015;372:2197–2206 42. Chiasson J-L, Josse RG, Hunt JA, et al. The 61. Ratner RE, Rosenstock J, Boka G; DRI6012 25. ACCORD Study Group. Nine-year effects of 3.7 efficacy of acarbose in the treatment of patients Study Investigators. Dose-dependent effects years of intensive glycemic control on cardiovas- with non-insulin-dependent diabetes mellitus. of the once-daily GLP-1 receptor agonist lixisena- cular outcomes. Diabetes Care 2016;39:701–708 A multicenter controlled clinical trial. Ann Intern tide in patients with type 2 diabetes inadequately 26. Holman RR, Farmer AJ, Davies MJ, et al.; 4-T Med 1994;121:928–935 controlled with metformin: a randomized, dou- Study Group. Three-year efficacy of complex in- 43. Delorme S, Chiasson J-L. Acarbose in the pre- ble-blind, placebo-controlled trial. Diabet Med sulin regimens in type 2 diabetes. N Engl J Med vention of cardiovascular disease in subjects with 2010;27:1024–1032 2009;361:1736–1747 impaired glucose tolerance and type 2 diabetes 62. Ryan G, Briscoe TA, Jobe L. Review of pram- 27. Riddle MC, Rosenstock J, Vlajnic A, Gao L. mellitus. Curr Opin Pharmacol 2005;5:184–189 lintide as adjunctive therapy in treatment of Randomized, 1-year comparison of three ways 44. Holman RR, Bethel MA, Chan JCN, et al.; type 1 and type 2 diabetes. Drug Des Devel to initiate and advance insulin for type 2 diabetes: ACE Study Group. Rationale for and design of Ther 2009;2:203–214 twice-daily premixed insulin versus basal insulin the Acarbose Cardiovascular Evaluation (ACE) 63. Buse JB, Bergenstal RM, Glass LC, et al. Use of with either basal-plus one prandial insulin or trial. Am Heart J 2014;168:23–29.e2 twice-daily exenatide in basal insulin-treated 300 The Prandial Problem Diabetes Care Volume 40, March 2017

patients with type 2 diabetes: a randomized, con- 67. Rosenstock J, Guerci B, Hanefeld M, et al.; 71. Weyer C, Fineman MS, Strobel S, et al. Prop- trolled trial. Ann Intern Med 2011;154:103–112 GetGoal Duo-2 Trial Investigators. Prandial op- erties of pramlintide and insulin upon mixing. 64. Diamant M, Nauck MA, Shaginian R, et al.; tions to advance basal insulin glargine therapy: Am J Health Syst Pharm 2005;62:816–822 4B Study Group. Glucagon-like peptide 1 recep- testing lixisenatide plus basal insulin versus in- 72. Riddle MC, Yuen KCJ, de Bruin TW, et al. tor agonist or bolus insulin with optimized basal sulin glulisine either as basal-plus or basal-bolus Fixed ratio dosing of pramlintide with in type 2 diabetes. Diabetes Care 2014; in type 2 diabetes: the GetGoal Duo-2 Trial. Di- insulin before a standard meal in patients with 37:2763–2773 abetes Care 2016;39:1318–1328 type 1 diabetes. Diabetes Obes Metab 2015;17: 65. FLAT-SUGAR Trial Investigators. Glucose 68. Riddle M, Pencek R, Charenkavanich S, Lutz K, 904–907 variability in a 26-week randomized comparison Wilhelm K, Porter L. Randomized comparison of 73. Chaudhri O, Small C, Bloom S. Gastrointes- of mealtime treatment with rapid-acting insulin pramlintide or mealtime insulin added to basal tinal hormones regulating appetite. Philos Trans versus GLP-1 agonist in participants with type 2 insulin treatment for patients with type 2 diabe- R Soc Lond B Biol Sci 2006;361:1187–1209 diabetes at high cardiovascular risk. Diabetes tes. Diabetes Care 2009;32:1577–1582 74. Drucker DJ. Evolving concepts and transla- Care 2016;39:973–981 69. Marso SP, Daniels GH, Brown-Frandsen K, et al.; tional relevance of enteroendocrine cell biology. 66. Riddle MC, Forst T, Aronson R, et al. Adding LEADER Steering Committee; LEADER Trial Investi- J Clin Endocrinol Metab 2016;101:778–786 once-daily lixisenatide for type 2 diabetes in- gators. Liraglutide and cardiovascular outcomes in 75. Thomas L. The Technology of Medicine. In adequately controlled with newly initiated and type 2 diabetes. N Engl J Med 2016;375:311–322 The Lives of a Cell. Notes of a Biology Watcher. continuously titrated basal insulin glargine: a 70. Pfeffer MA, Claggett B, Diaz R, et al.; ELIXA New York, The Viking Press, Inc., 1974, p. 31– 24-week, randomized, placebo-controlled study Investigators. Lixisenatide for patients with 36 (GetGoal-Duo 1). Diabetes Care 2013;36:2497– type 2 diabetes and acute coronary syndrome. 76. Haring¨ H-U. Novel phenotypes of prediabe- 2503 N Engl J Med 2015;373:2247–2257 tes? Diabetologia 2016;59:1806–1818