Heart failure: a cardiovascular outcome in diabetes that can no longer be ignored

Personal View

Heart failure: a cardiovascular outcome in diabetes that can no longer be ignored John J V McMurray, Hertzel C Gerstein, Rury R Holman, Marc A Pfeffer

In patients with type 1 or type 2 diabetes, glycaemic exposure assessed as HbA1c correlates strongly with risk of future microvascular and macrovascular complications. Improved glucose control substantially reduces the risk of microvascular complications and, with extended follow-up, modestly reduces the risk of atherosclerotic events. The lowering of HbA1c concentrations by newly developed glucose-lowering drugs (alone or when added to other glucoselowering drugs) has been used, until recently, as a surrogate measure of their potential to lower cardiovascular risk. This assumption is no longer acceptable, and now demonstration of cardiovascular safety has been mandated by regulatory authorities. A major concern, however, is the universal absence in any large-scale trials of new glucoselowering drugs of hospital admission for heart failure as a prespecified component of the primary composite cardiovascular outcomes. This omission is important because hospital admission for heart failure is a common and prognostically important cardiovascular complication of diabetes. Moreover, it is the one cardiovascular outcome for which the risk has been shown unequivocally to be increased by some glucose-lowering therapies. As such, we believe that heart failure should be systematically evaluated in cardiovascular outcome trials of all new glucose-lowering drugs.

Introduction Regulatory approval of a new therapeutic drug is based on the achievement of a cumulative level of clinical trial evidence showing that the intervention is both effective and safe (as far as it is possible to establish safety within the limits of early discovery). The definitions of effective and safe are highly dependent on the disease condition and standards of international regulatory agencies. In patients with type 1 or type 2 diabetes, improved glycaemic control substantially reduces the risk of microvascular complications, and modestly reduces the risk of cardiovascular events with extended follow-up.1–4 Lowering of HbA1c concentration by glucose-lowering drugs has been used, until recently, as a surrogate measure of their potential to reduce cardiovascular risk.5–9 Registration trials generally included stable patients with a raised HbA1c and had follow-up measured in weeks or months.5–9 When sufficient evidence of glucose-lowering efficacy had been accumulated for regulatory submission, tabulations of serious and less serious adverse events were also provided, usually with a focus on treatmentrelated outcomes such as symptomatic hypoglycaemia. The prevailing assumption was that lowering of glucose would, in the long run, lead to beneficial patient outcomes, defined as fewer microvascular and hopefully macrovascular events.5–9 Many cardiovascular and diabetes researchers, however, strongly advocated the importance of explicit measurement of macrovascular events during the evaluation of new drugs for diabetes. After a high-profile meta-analysis suggesting that a particular drug that was effective at lowering glucose might also increase the risk of myocardial infarction, important and constructive changes in the regulatory process for assessment of new drugs to treat diabetes were finally announced.9 The US Food and Drug Administration (FDA) and the European Medicines Agency (EMA) both issued new guidance aimed at assuring more statistically robust

demonstration of cardiovascular safety for new treatments for diabetes before they can enter the market.10,11 In effect, sponsors must show that their drug will not result in an unacceptable increase in cardiovascular risk.5–11 This guidance has resulted in a huge increase in the number of randomised trials in patients with diabetes, done mainly to evaluate safety, with so-called major adverse cardiovascular events (MACE) as the primary outcome. MACE generally has been defined as the composite of cardiovascular death, myocardial infarction, or stroke, although the guidance states that other relevant cardiovascular events such as hospital admission for acute coronary syndrome, urgent revascularisation procedures, and other endpoints can be considered. On a positive note, the new guidance from the FDA and EMA has led to substantial increases in exposure of patients with cardiovascular disease to new compounds before they are approved. At present roughly 150 000 patients with either known cardiovascular disease or a high burden of cardiovascular risk factors are being followed up in randomised trials of new glucose-lowering drugs. This international effort will undoubtedly improve definition of the cardiovascular safety and efficacy profile of these drugs, and provide these data at an earlier marketing stage than hitherto. A major concern, however, is the universal absence in any of these trials of hospital admission for heart failure as a prespecified component of their primary composite cardiovascular outcomes. In our opinion, hospital admission for heart failure is one of the most common and prognostically important cardiovascular complications of diabetes, and the one cardiovascular outcome for which the risk has been shown unequivocally to be increased by certain glucose-lowering therapies. Thus, we believe that heart failure should be systematically evaluated in cardiovascular outcome trials of new glucose-lowering drugs, either as a component of the primary composite outcome or as a prespecified secondary endpoint.

www.thelancet.com/diabetes-endocrinology Published online March 13, 2014 http://dx.doi.org/10.1016/S2213-8587(14)70031-2

Lancet Diabetes Endocrinol 2014 Published Online March 13, 2014 http://dx.doi.org/10.1016/ S2213-8587(14)70031-2 BHF Cardiovascular Research Centre, University of Glasgow, Glasgow, UK (Prof J J V McMurray MD); Department of Medicine and Department of Clinical Epidemiology and Biostatistics, McMaster University and Population Health Research Institute, Hamilton, ON, Canada (Prof H C Gerstein MD); Diabetes Trials Unit, Oxford Centre for Diabetes, Endocrinology and Metabolism, University of Oxford, Oxford, UK (Prof R R Holman FRCP); and Department of Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA (Prof M A Pfeffer MD) Correspondence to: Prof John J V McMurray, Institute of Cardiovascular and Medical Sciences, BHF Glasgow Cardiovascular Research Centre, University of Glasgow, Glasgow G12 8TA, UK [email protected]

1

Personal View

What is heart failure? Heart failure is the term used to describe a common clinical syndrome arising as a consequence of impaired cardiac pump function.12,13 The most typical symptoms are breathlessness and fatigue and the most common sign is peripheral oedema. The syndrome of heart failure can arise as a result of almost any abnormality of the structure, mechanical function, or electrical activity of the heart. Many of the typical clinical symptoms and signs of heart failure do not arise directly as a result of the cardiac abnormality, but rather from secondary dysfunction of other organs and tissues—eg, the kidneys, bone marrow, and muscles.12,13 These secondary consequences of pump failure are myriad and are not explained solely by reduced perfusion; it is generally believed that other systemic processes such as Trials comparing intensity of glucose lowering

Trials in patients with diabetic nephropathy

ADVANCE (n=11 140)

ACCORD (n=10 251)

VADT (n=1791)

ADOPT (n=4351)

ORIGIN (n=12 537)

SAVOR-TIMI 53 (n=16 492)

DIABHYCAR31 (n=4912)

RENAAL32 (n=1513)

IDNT33,34 (n=1715)

ALTITUDE35 (n=8561)

Age 26–65 years; new diabetes; no previous drug therapy; current HF excluded

Age ≥55 years; macrovascular or microvascular disease; microvascular disease; or ≥1 cardiovascular risk factor

HbA1c ≥7·5%; age 40–79 years with CVD or 55–79 years with cardiovascular risk factors

Age >40 years; HbA1c ≥7·5%

Age 30–75 years; new diabetes; no previous antidiabetic drug therapy

Age ≥50 years and previous cardiovascular event, CVD, or cardiovascular risk factors and T2D, IFG, or IGT

Age ≥40 years and previous CVD or ≥55 years (≥60 years for women) and cardiovascular risk factors

Age ≥50 years; albuminuria ≥20 mg/L; serum creatinine ≤150 μmol/L

Age 31–70 years, UACR ≥300 mg/g; serum creatinine 115–265 μmol/L

Age 30–70 years; hypertension; urinary protein excretion ≥900 mg per day; serum creatinine 88–265 μmol/L

Age ≥35 years; UACR ≥200 mg/g and eGFR ≥30 mL/min per 1·73m² or ≥20 mg/g to <200 mg/g and eGFR ≥30 to <60 mL/min per 1·73m² or CVD and eGFR ≥30 to<60 mL/min per 1·73m²

NYHA class II–IV HF and LVEF <25%

NYHA class II–IV HF

NYHA NYHA Cardioclass I–IV HF class I–IV HF myopathy likely to cause death within 5 years

··

NYHA class III/IV

25

Exclusion criteria

··

··

Median/mean duration of diabetes (years)

··

8

10

5

Median/mean follow-up (years)

Trials evaluating glucose-lowering drugs

UKPDS (n=3867) 3

Participants

neurohumoral activation and inflammation are involved. In developed countries, left ventricular dysfunction is the commonest underlying problem and is caused, mainly, by myocardial infarction (leading to systolic ventricular dysfunction—ie, failure of normal contraction and emptying of the heart), hypertension (causing systolic dysfunction, diastolic dysfunction—ie, failure of normal relaxation and filling of the heart), or, often, both infarction and hypertension.12,13 Patients with cardiac dysfunction can remain asymptomatic for months, years, or even decades. However, once symptomatic heart failure has developed the syndrome is characterised by progressive worsening of the patient’s symptoms, of cardiac function, and of the function of other tissues (eg, skeletal muscle, bone marrow) and organs (eg, the kidneys). Quality of life is greatly diminished, hospital

26

10

3·5

27

28

29

30

11·5

<2·0

5·4*

10·3

5·6

4

6·2

2·1

··

1156 (9%)

529 (3%)

··

662 (5%)

543 (3%)

Insulin or ACE History inhibitor/ARB of HF treatment; congestive HF 9·8

>5·0†

14·9

>5·0‡

3·4

2·6

2·7

354 (7%)

169 (11%)

135 (8%)

461 (5%)

139 (3%)

118 (8%)

117 (7%)

289 (3%)

4

Outcomes Cardiovascular deaths, n (%)

··

542 (5%)

All MI, n (%)

573 (15%)

Fatal MI

297 (8%)

Non-fatal MI

298 (8%)

309 (3%)

All stroke, n (%)

203 (5%)

484 (4%)

Fatal stroke Non-fatal stroke All HF, n (%)

58 (1%) 158 (4%) 116 (3%)

Fatal HF

··

Non-fatal HF

··

679 (6%) ··

·· 423 (4%) ·· ·· 451 (4%)

229 (2%) ·· 32 (<1%) 421 (4%) ·· 20 (<1%) 128 (1%) 276 (3%) 39 (<1%) ··

67 (4%) 142 (8%) ·· ·· 64 (4%)

7 (<1%) 61 (1%) 52 (1%)

··

··

··

··

··

··

··

··

··

··

650 (5%)

298 (2%)§

234 (5%)

97 (6%)

64 (4%)

·· ·· 269 (3%)

··

··

··

··

··

··

··

··

··

··

··

··

··

··

··

··

··

··

··

··

158 (9%)

50 (1%)

··

··

··

··

653 (5%)

517 (3%)

187 (4%)

191 (13%)

225 (13%)

··

··

··

··

··

··

·· ·· 424 (5%)

Empty cells denote outcomes not reported for that trial. HF=heart failure. CVD=cardiovascular disease. T2D=type 2 diabetes. IFG=impaired fasting glucose. IGT=impaired glucose tolerance. UACR=urinary albumin–creatinine ratio. eGFR=estimated glomerular filtration rate. NYHA=New York Heart Association. LVEF=left ventricular ejection fraction. ACE=angiotensin-converting enzyme. ARB=angiotensin receptor blocker. MI=myocardial infarction. *In patients with established diabetes. †90% >5 years. ‡82% >5 years. §Ischaemic stroke.

Table 1: Prospective randomised treatment trials in patients with type 2 diabetes reporting heart failure as a cardiovascular outcome

2

www.thelancet.com/diabetes-endocrinology Published online March 13, 2014 http://dx.doi.org/10.1016/S2213-8587(14)70031-2

Personal View

admission for exacerbations are common, and mortality is high.12,13 In clinical trials, unplanned hospital admission for heart failure is typically defined as requiring all of presentation with typical symptoms, the presence of typical signs, and treatment with intravenous therapy.14

Diabetes as a causal factor for heart failure? Whether diabetes should be considered a causal factor or a comorbidity in heart failure is unclear.15,16 Its potential role in causation of systolic and diastolic dysfunction is uncertain. Individuals with diabetes have a higher prevalence of heart failure than those without diabetes. Diabetes accelerates the development of coronary atherosclerosis and is often associated with hypertension. Whether it directly causes a specific cardiomyopathy is, however, uncertain. Diabetes is also associated with an increased risk of developing heart failure in patients with other causes—eg, acute myocardial infarction. Diabetes is believed to promote the development of myocardial fibrosis and diastolic dysfunction. Diabetes is also associated with increased autonomic dysfunction and worsened renal and endothelial function, impaired gas exchange, as well as worsened functional status and a poor prognosis.15,16

Frequency of heart failure in diabetes Observational data illustrate the frequent occurrence of heart failure compared with other cardiovascular events.

For example, Juhaeri and colleagues17 examined the incidence of hospital admission for myocardial infarction, stroke, and heart failure during a 6-year period among 65 619 patients with type 2 diabetes treated with insulin, in a large US claims database. The rate of myocardial infarction was 97/10 000 (1134 cases), stroke 151/10 000 (1746 cases), and heart failure 243/10 000 (2786 cases). Despite such epidemiological findings, a review of 19 trials in patients with type 2 diabetes (or a mix of type 1 and type 2 diabetes) showed that 12 did not report heart failure as an outcome; another review reached similar conclusions and two more recent trials also failed to report heart failure.18–21 Inspection of reports of the diabetic subgroup among patients enrolled in 12 trials of chronic arterial disease or hypertension revealed similar findings.18,22–24 Tables 1 and 2 summarise drug-therapy trials that did report heart failure as an outcome.3,25–42 Comparison across trials is difficult because of the differing entry criteria for each and because of varying definition of the endpoints. Some trials included so-called silent myocardial infarction with clinically recognised events and some trials included episodes of heart failure not leading to hospital admission in the definition of non-fatal heart failure. Despite these difficulties, it is clear that in the recent large-scale trials of glucose-lowering therapies, heart failure occurred at a frequency similar to that of stroke

Diabetic subgroup from trials in patients with chronic arterial disease, hypertension, or both HOPE36 (MICRO-HOPE; n=3577)

EUROPA37 LIFE38,39 (PERSUADE; (n=1195) n=1502)

Participants

Age ≥55 years with CVD; or ≥1 cardiovascular risk factor

Age >18 years; CHD

Exclusion criteria

Nephropathy, HF, or LVEF <40%

HF

Median/mean follow-up (years)

4·5

4·3

Diabetic subgroup from trials in patients with acute myocardial infarction

VALUE40 (n=5250)

ACCOMPLISH18 (n=6946)

SAVE41 (n=496)

VALIANT42 (n=3400)

Age 55–80 years; hypertension; LVH

Age ≥50 years; hypertension; CVD or cardiovascular risk factors

Age ≥50 years; hypertension; CVD or cardiovascular risk factors

Age ≥21–80 years; 3–16 days after acute MI; LVEF ≤40%

Age ≥18 years; 0·5–10 days after acute MI

HF or LVEF <40%

HF requiring an ACE inhibitor

HF or LVEF <40% requiring an ACE inhibitor

Symptomatic LV systolic HF requiring an dysfunction, HF, or both ACE inhibitor

4·7

4·2

2·5

3·5

1·0*

Outcomes Cardiovascular deaths, n (%)

284 (8%)

107 (7%)

99 (8%)

286† (5%)

136 (2%)

133 (27%)

All MI, n (%)

414 (12%)

134 (9%)

91 (8%)

299 (6%)

168 (2%)

94 (19%)

Fatal MI

··

··

··

··

··

··

Non-fatal MI All stroke

··

··

184 (5%)

41 (3%)

·· 116 (10%)

548 (16%) ·· ··

··

··

··

308 (10%)

234 (4%)

134 (2%)

··

97 (3%)

Fatal stroke

··

··

··

··

··

··

··

Non-fatal stroke

··

··

··

··

··

··

··

All HF

434 (12%)

39 (3%)

··

412 (8%)

141 (2%)

Fatal HF

··

··

··

··

··

156 (31%) ··

664 (21%) ··

Non-fatal HF

··

··

87 (7%)

··

··

··

··

Empty cells denote outcomes not reported for that trial. MI=myocardial infarction. CVD=cardiovascular disease. CHD=coronary heart disease. LVH=left ventricular hypertrophy. LVEF=left ventricular ejection fraction. HF=heart failure. ACE=angiotensin-converting enzyme. LV=left ventricular. *1 year follow-up was used for this analysis. †Cardiac death.

Table 2: Cardiovascular events in diabetic subgroups of patients in clinical trials of patients with chronic arterial disease, hypertension, or acute MI

www.thelancet.com/diabetes-endocrinology Published online March 13, 2014 http://dx.doi.org/10.1016/S2213-8587(14)70031-2

3

Cumulative percentage of patients admitted to hospital for heart failure (%)

Personal View

LIFE, atenolol group LIFE, losartan group RENAAL, losartan group RENAAL, placebo group

20

15

10

5

0 0

1

2

3

4

5

Years of follow-up

Figure 1: Hospital admission for heart failure in patients with diabetes in LIFE38,39 and RENAAL32 Cumulative proportion of patients admitted to hospital with heart failure in LIFE (losartan:atenolol; hazard ratio [HR] 0·57, 95% CI 0·36–0·91, p=0·019) and RENAAL (losartan:placebo, HR 0·74, 95% CI 0·55–0·98, p=0·037). Reproduced from Carr et al,44 with permission of Elsevier.

and myocardial infarction. It should be noted that this outcome was despite most trials excluding patients with more than mildly symptomatic heart failure (or heart failure altogether) and the exclusion by many trials of elderly patients, who are particularly likely to develop heart failure. In the 5145 low-risk participants in the Look AHEAD (Action for Health in Diabetes) trial, the rate of development of heart failure in the control group (0·42 per 100 patient-years) was greater than that of stroke (0·36) or cardiovascular death (0·22).43 The rate of myocardial infarction was 0·71 per 100 person-years. Only in the older UK Prospective Diabetes Study (UKPDS) was the risk of heart failure substantially less than that of myocardial infarction.3 Of the 3867 patients enrolled in UKPDS, 573 (15%) had a fatal or non-fatal myocardial infarction, 203 (5%) a stroke, and 116 (3%) developed heart failure over a median 10 years of followup. UKPDS, however, enrolled only patients with newly diagnosed diabetes and without known cardiovascular disease. Conversely, heart failure appears to be a more common complication of advanced diabetes, particularly when nephropathy develops (figure 1).44 Indeed, as shown in table 1, the occurrence of heart failure was actually more frequent than myocardial infarction in the Reduction in Endpoint with the Angiotensin Antagonist Losartan (RENAAL) and Irbesartan Diabetic Nephropathy Trial (IDNT). Not only was heart failure a common first cardiovascular event in these patients, but it recurred more frequently than other major cardiovascular outcomes. For example, in IDNT, among the 1715 patients randomly assigned to treatment groups, 225 (13%) had 336 episodes of heart failure compared 4

with 117 (7%) patients (128 episodes) with myocardial infarction and 69 (4%) patients (76 episodes) with stroke.34 A similar observation was made in patients with diabetes without nephropathy in the Prospective Pioglitazone Clinical Trial in Macrovascular Events trial (PROACTIVE).45–47 In that trial, 144 placebo-treated patients (5%) had 157 non-fatal myocardial infarctions (including silent infarctions), 107 (4%) had 119 strokes, and 108 patients (4%) had 153 hospital admissions for heart failure (90 [3%] patients also had 117 episodes of heart failure not needing hospital admission). Most recently, among the 8561 patients in the Aliskiren Trial in Type 2 Diabetes Using Cardio-Renal Endpoints (ALTITUDE),35 hospital admission for heart failure was a more frequent occurrence (424 patients) than either fatal or non-fatal myocardial infarction (289 patients) or stroke (269 patients). Several trials of interventions in patients with chronic arterial disease or hypertension have reported cardiovascular outcomes separately in a diabetic subgroup. These reports also reveal that heart failure was at least as common as was myocardial infarction (and more common than stroke) in most studies (table 2, figure 2).17,24,36,39–42,48 In the two trials that enrolled patients with left ventricular systolic dysfunction or heart failure, or both, after myocardial infarction, the risk of heart failure in the diabetic subgroup was particularly high and much greater than that of myocardial infarction (table 2).41,42

Prognostic implications of heart failure in diabetes It is not only the frequent occurrence of heart failure that is important in patients with diabetes. The occurrence of this complication is associated with a poor subsequent prognosis, as shown in both observational studies and clinical trials. Bertoni and colleagues49 examined mortality rates among 151 738 beneficiaries with diabetes in a 5% sample of Medicare claims between 1994 and 1999. During these 60 months, patients with incident heart failure had a mortality rate of 32·7 per 100 person-years compared with 3·7 per 100 person-years in patients with diabetes who remained free of heart failure. The findings of clinical trials are consistent. For example, in DIABHYCAR,31,50 658 (13%) of 4912 patients died during follow-up. Among patients not admitted to hospital for heart failure, 590 (12%) of 4725 died during a mean followup of 47 (SD 16) months, equating to an annual mortality rate of 3·2%. Among patients admitted to hospital for heart failure, 68 (36%) of 187 died an average of 12 (SD 11) months after their first admission. Development of heart failure in RENAAL (mean 3·4 years) and LIFE (mean 4·7 years) was associated with a four-fold and six-fold increase, respectively, in risk of death (figure 3).44 Further insight into the importance of heart failure is provided by the RECORD trial, with 5·5 years mean follow-up.51,52 In RECORD, eight (28%) of 29 patients in the control group surviving a first admission for heart

www.thelancet.com/diabetes-endocrinology Published online March 13, 2014 http://dx.doi.org/10.1016/S2213-8587(14)70031-2

Personal View

0·10

Baseline diabetes New-onset diabetes Never diabetes

Proportion of patients with first event

0·09 0·08 0·07 0·06 0·05 0·04 0·03 0·02 0·10 0 0

500 1000 1500 Time to fatal or non-fatal myocardial infarction (days)

2000

0

500 1000 1500 Time to fatal or non-fatal heart failure (days)

2000

Figure 2: Cumulative risk of myocardial infarction and heart failure in VALUE40 overall (valsartan and amlodipine groups combined), according to diabetes status Reproduced from Akses et al,48 with permission of Wolters Kluwer Health. A 0·6

B Without heart failure With heart failure

Cumulative event rate

0·5

0·4

0·3

0·2

0·1

0 0 Number at risk Without heart failure 1147 With heart failure 0

500 1088 22

1000 Days

1500

1032 46

961 60

2000

0

300

600

900

1200

1500

1107 108

544 76

77 15

Days 195 1431 9 0

1348 46

1259 83

Figure 3: Mortality in patients with diabetes with and without heart failure in LIFE38,39 and RENAAL32 Kaplan-Meier curves illustrating cumulative mortality rate in patients who did and did not develop heart failure during (A) the LIFE trial and (B) RENAAL. The heart failure:no heart failure hazard ratio for mortality was 5·98 (95% CI 3·90–9·17, p<0·0001) in LIFE and 3·99, 95% CI 3·02–5·25, p<0·0001) in RENAAL. Reproduced from Carr et al,44 with permission of Elsevier.

failure subsequently died compared with eight (15%) of 52 patients surviving a first hospital admission for acute myocardial infarction (overall, 7% of patients in RECORD died). In other words, in RECORD, subsequent outcomes in patients admitted to hospital for heart failure were worse than for patients admitted with acute myocardial infarction.

Effects of diabetic therapy on incidence of heart failure Regulatory agencies have stressed the importance of ensuring that new glucose-lowering drugs are safe or beneficial with respect to myocardial infarction or stroke. No glucose-lowering drug has, however, been shown

clearly to cause either of these problems. Conversely, two such drugs (the thiazolidinediones pioglitazone and rosiglitazone) have each been shown to increase the risk of heart failure (table 3).45,51,52 The two major outcome trials with these drugs, RECORD and PROACTIVE, identified a substantial increase in risk of heart failure, although heart failure was not a prespecified, adjudicated endpoint in PROACTIVE.45–47 In PROACTIVE, 198 (8%) patients on placebo had 302 heart failure events compared with 281 (11%) patients (417 events) in the pioglitazone group (p<0·0001). Of patients in the placebo group, 108 needed hospital admission for heart failure (153 admissions) compared with 149 (209 admissions) in the pioglitazone group (hazard ratio [HR] 1·41, 95% CI 1·10–1·80, p=0·007).

www.thelancet.com/diabetes-endocrinology Published online March 13, 2014 http://dx.doi.org/10.1016/S2213-8587(14)70031-2

5

Personal View

RECORD51,52

PROACTIVE45,46

Control (n=2227)

Rosiglitazone (n=2220)

Placebo (n=2633)

Pioglitazone (n=2605)

Participants

Age 40–75 years; BMI >25 kg/m²; maximum dose of metformin or SU; HbA1c >7·0% to <9·0%

Age 40–75 years; BMI >25 kg/m²; maximum dose of metformin or SU; HbA1c >7·0% to <9·0%

Age 33–75 years; HbA1c >6·5% despite diet or oral hypoglycaemic drugs; macrovascular disease

Age 33–75 years; HbA1c >6·5% despite diet or oral hypoglycaemic drugs; macrovascular disease

Exclusion criteria

HF excluded

HF excluded

NYHA class II–IV HF excluded

NYHA class II–IV HF excluded

Median/mean duration of diabetes (years)

Roughly 7

Roughly 7

Median/mean follow-up (years)

5·5

5·5

8

8

2·9

2·9

136 (5%)

127 (5%)

Outcomes Cardiovascular deaths, n (%)

71 (3%)

60 (3%)

All MI, n (%)

56 (3%)

64 (3%)

Fatal MI

10 (<1%)

Non-fatal MI

52 (2%)

60 (3%)

144 (5%)*

119 (5%)*

All stroke, n (%)

63 (3%)

46 (2%)

107 (4%)

86 (3%)

7 (<1%)

·· ··

Fatal stroke

5 (<1%)

0

··

Non-fatal stroke

··

··

··

All HF Fatal HF Non-fatal HF

29 (1%) 2 (<1%) 29 (1%)

·· ··

·· ··

61 (3%)

198 (8%)

281 (11%)

10 (<1%)

22 (1%)

25 (1%)

57 (3%)

··

··

Empty cells denote outcomes not reported for that trial. SU=sulfonylurea. NYHA=New York Heart Association. HF=heart failure. MI=myocardial infarction. *Including silent MI.

Table 3: Occurrence of cardiovascular events in two large trials of thiazolidinediones

Retrospective adjudication of these events showed close concurrence with the investigator reports.47 In RECORD, 61 (3%) of 2220 rosiglitazone-treated patients had at least one admission to hospital for heart failure whereas 29 (1%) of 2227 patients in the placebo group had this outcome (HR 2·6, 95% CI 1·1–4·1, p=0·001). Despite previous speculation that thiazolidinedionesinduced heart failure does not carry the same prognostic importance as heart failure occurring in other patients, and might just reflect fluid retention caused by these drugs, recent evidence from RECORD suggests otherwise. In RECORD, the risk of hospital admission for heart failure was roughly doubled in the rosiglitazone group (table 3).52 Of the 61 rosiglitazone-treated cases of heart failure, the first event was fatal in four and 17 (30%) of the 57 surviving patients died during the remainder of the trial. In the control group, 29 patients had a first hospital admission for heart failure, none of which was fatal subsequently, and eight (28%) of these 29 patients died during follow-up (compared with 6·1% and 7·5% deaths overall in the rosiglitazone and control groups, respectively). In PROACTIVE, although mortality from what was categorised as serious heart failure was lower in pioglitazone-treated patients than in those treated with placebo, prognosis was poor in both groups: 40 (27%) of 149 pioglitazone-treated patients died after developing heart failure compared with 37 (35%) of 108 placebo-treated patients (compared with 6·8% and 7·1% deaths overall in the pioglitazone and placebo groups, respectively).46 The relative safety of pioglitazone compared with rosiglitazone is unlikely 6

ever to be known after the premature termination of the Thiazolidinedione Intervention with vitamin D Evaluation (TIDE).53 Recently, the dipeptidyl peptidase-4 (DPP-4) inhibitor saxagliptin has also been shown to increase the risk of hospital admission for heart failure, compared with placebo (3·5% vs 2·8%, p=0·007) in the Saxagliptin Assessment of Vascular Outcomes Recorded in patients with diabetes mellitus–Thrombolysis In Myocardial Infarction (SAVOR-TIMI) 53 trial in which 16 492 patients with type 2 diabetes and established cardiovascular disease, or risk factors for cardiovascular disease, were followed up for a median of 2·1 years.30 Whether this completely unexpected finding reflects the play of chance, is drug specific, a DPP-4 inhibitor class effect, or even an issue for other incretin-based therapies (eg, glucagon-like peptide-1 analogues) is unknown at this point, but once again emphasises the importance of heart failure as an outcome in diabetes. Not only was heart failure a much more frequent event than stroke, but also it was nearly as common as myocardial infarction and was the only major cardiovascular event (including death) affected by the study drug. The findings of SAVOR-TIMI 53 are clearly at odds with the observation that heart failure is characterised by high DPP-4 activity and that reducing this activity with DPP-4 inhibitors has beneficial effects in experimental models of heart failure.54–57 Another trial of a DPP-4 inhibitor, the EXamination of CArdiovascular OutcoMes: AlogliptIN vs. Standard of CarE in Patients with Type 2 Diabetes Mellitus and Acute Coronary Syndrome trial (EXAMINE) has also reported heart

www.thelancet.com/diabetes-endocrinology Published online March 13, 2014 http://dx.doi.org/10.1016/S2213-8587(14)70031-2

Personal View

failure as part of an exploratory composite outcome. Although not published in full at the time of writing, the EXAMINE investigators have reported a greater number of hospital admissions for heart failure in the alogliptin group compared with the placebo group, although this difference was not statistically significant.21 The effect of other antidiabetic treatments on the occurrence of heart failure is unknown, with the exception of insulin glargine, which was tested in the Outcome Reduction with an Initial Glargine Intervention trial (ORIGIN) in patients with impaired fasting glucose, impaired glucose tolerance, or early diabetes, and cardiovascular disease or risk factors (table 1).29 The rate of hospital admission for heart failure was 0·85 per 100 patient-years in the insulin group compared with 0·95 per 100 patient-years in the control group.29 It should be noted, however, that the 88% of patients in ORIGIN with diabetes had their disease for a mean duration of 5·4 years, 6% had a new diagnosis, and 23% were receiving no pharmacological therapy before randomisation. Patients with diabetes of longer duration are at greater risk of heart failure and might be at greater risk of drug-induced heart failure.

Why might treatments for diabetes increase the risk of heart failure? Because little attention has been paid to heart failure as an outcome in patients with diabetes, it is not surprising that even less is known about the possible mechanisms underlying the effects of thiazolidinediones and possibly DPP-4 inhibitors on this cardiovascular outcome. Although thiazolidinediones were recognised to cause fluid retention, possibly through a distal tubular action, this process cannot be the whole explanation for the findings of the RECORD and PROACTIVE trials because, in many cases, patients developed pulmonary oedema necessitating emergency admission to hospital. Not only that, but such events were associated with a dismal subsequent prognosis.52 One possibility is that fluid retention unmasks heart failure in susceptible individuals with undiagnosed underlying left ventricular dysfunction, a common problem in patients with diabetes.58 Thiazolidinediones are not known to have a direct toxic or negative inotropic action. The explanation for the findings of SAVOR-TIMI 53 (assuming they are not just the play of chance) is even less clear. At the time of writing, no further information was avalable about the nature of the heart failure identified or the outcomes in patients who developed heart failure in this trial. DPP-4 has many substrates and its inhibition might affect many biological pathways including, potentially, those as diverse as enzymes affecting cardiac collagen turnover and the sodium/hydrogen exchanger isoform 3 in the renal proximal tubule.55–57 A direct toxic or negative inotropic action of DPP-4 inhibitors has not been proposed to date, and most studies have suggested the opposite effect.59,60 Alternatively, a detrimental myocardial

energetic consequence of reduced blood and tissue glucose and insulin might provide a unifying explanation for the findings with both thiazolidinediones and DPP-4 inhibitors.61 This unsatisfactory state of affairs needs urgent attention and rectification with both improved studies on myocardial function in patients with concomitant diabetes and heart failure, as well as on the cardiac effects of new pharmacological treatments for diabetes. Fortunately, at least some of the ongoing trials in patients with type 2 diabetes are collecting information about possible cases of heart failure, either as part of an expanded secondary cardiovascular composite outcome62 or as a stand-alone endpoint (see also NCT01147250, NCT01144338).63

Limitations Our focus has been clinical trials and these have mainly reported cases of heart failure leading to hospital admission. Although patients admitted to hospital have the worst prognosis, even heart failure that does not lead to admission is an ominous development. Our focus on patients admitted to hospital will also have underestimated the incidence of heart failure in patients with diabetes. Another potential limitation of our critique is that some might argue that composite outcomes should reflect the same disease process—eg, atherosclerosis or atherothrombosis—as reflected in the MACE outcome of cardiovascular death, non-fatal myocardial infarction, and non-fatal stroke used in almost all recent and ongoing trials of antidiabetes drugs. A counter argument is that composites might also reflect all major and common events that represent the burden of the disease in question to patients and society; heart failure is highly relevant from this perspective. For any composite outcome, interpretation is difficult when a treatment has divergent effects on individual components and this problem might be more likely with components that do not reflect the same disease process.

Search strategy and selection criteria To examine heart failure occurring in trials in diabetes, electronic databases including PubMed, Embase, CINAHL, and the Cochrane Library were searched up to Dec 1, 2013. We restricted the search to randomised controlled trials done in among human adults (age ≥19 years) and published in English; there were no date restrictions. Our search of the Medline database was done through PubMed using the medical subject headings (MeSH) terms for heart failure: “Heart Failure” or “Cardiomyopathy, Dilated”, which include the following definitions: cardiac failure; myocardial failure; left sided heart failure; right sided heart failure; congestive heart failure; heart insufficiency; systolic heart failure; diastolic heart failure; cardiac oedema; paroxysmal dyspnoea; dilated cardiomyopathy; congestive cardiomyopathy; familial idiopathic cardiomyopathy. Diabetes was identified using MeSH terms “diabetes mellitus”, “type 2”, “haemoglobin A1c”, “glycosylated”, “glycated”, “glucose”, “cardiovascular disease”, and “heart failure” as well as the non-MeSH terms “glycaemic control”, “glucose control”, and “aggressive/intensive/tight”. Bibliographies of reviews, guidelines, and all publications identified by the search strategies were systematically reviewed. International experts and pharmaceutical firms were also consulted to identify uncompleted, unpublished, or overlooked studies.

www.thelancet.com/diabetes-endocrinology Published online March 13, 2014 http://dx.doi.org/10.1016/S2213-8587(14)70031-2

7

Personal View

Conclusions

6

Clearly, the occurrence of heart failure in patients with diabetes, although poorly reported in existing clinical trials, is frequent and is of ominous prognostic importance. Admission to hospital for heart failure should have equal status to myocardial infarction and stroke as an important cardiovascular outcome in studies of new diabetic therapies, and even added as a component to the primary composite cardiovascular outcome of trials to more fully capture the potential cardiovascular risks and benefits. It is notable that while the argument is made that the lack of a demonstrable beneficial effect of glucose-lowering drugs on atherothrombotic events reflects the short duration of contemporary clinical trials (with the hypothesis that a much longer and sustained reduction in glucose is necessary to decrease macrovascular events), the timeframe of these trials has been more than sufficient to detect an adverse effect on heart failure.4 Contributors JJVMM wrote the first draft of the paper. All other authors critically reviewed and edited the content. Declarations of interest JJVMM reports that his employer, Glasgow University, has been paid for his time spent as a steering committee member, data safety monitoring board member, and endpoint committee member in diabetes trials by Bayer, GlaxoSmithKline, Lilly, Merck, Novartis, Sanofi-Aventis, and Oxford University. HCG reports consulting fees from Sanofi-Aventis, Novo Nordisk, Lilly, Bristol-Myers Squibb, Roche, AstraZeneca, Boehringer Ingelheim, and GlaxoSmithKline, lecture fees from Sanofi-Aventis and Bayer, and support for research or continuing education through his institution from Sanofi-Aventis, Lilly, Merck, Novo Nordisk, Boehringer Ingelheim, Bristol-Myers Squibb, Takeda, and AstraZeneca. RRH has received research funding from Bayer, BMS, and Merck and advisory board honoraria from Bayer, Elcelyx, Merck, Novartis, and Novo Nordisk. MAP has received research grant support from Amgen, Celladon, Novartis, Sanofi Aventis; has acted as a consultant for Aastrom, Abbott Vascular, Amgen, Cerenis, Concert, Fibrogen, GlaxoSmithKline, Hamilton Health Sciences, Medtronic, Merck, Roche, Servier, Teva, and University of Oxford; and The Brigham and Women’s Hospital has patents for the use of inhibitors of the renin-angiotensin system in selected survivors of myocardial infarction with Novartis, for which he is a co-inventor, but his share of the licensing agreement is irrevocably transferred to charity. References 1 The Diabetes Control and Complications Trial Research Group. The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. N Engl J Med 1993; 329: 977–86. 2 Nathan DM, Cleary PA, Backlund JY, et al, and the Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications (DCCT/EDIC) Study Research Group. Intensive diabetes treatment and cardiovascular disease in patients with type 1 diabetes. N Engl J Med 2005; 353: 2643–53. 3 Turner RC, Holman RR, Cull CA, et al, and the UK Prospective Diabetes Study (UKPDS) Group. Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). Lancet 1998; 352: 837–53. 4 Holman RR, Paul SK, Bethel MA, Matthews DR, Neil HA. 10-year follow-up of intensive glucose control in type 2 diabetes. N Engl J Med 2008; 359: 1577–89. 5 Hennekens CH, Hebert PR, Schneider WR, O’Brien P, Demets D, Borer JS. Academic perspectives on the United States Food and Drug Administration’s guidance for industry on diabetes mellitus. Contemp Clin Trials 2010; 31: 411–13.

8

7 8

9

10

11

12 13 14

15

16 17

18

19

20

21

22

23

24

25

Woodcock J, Sharfstein JM, Hamburg M. Regulatory action on rosiglitazone by the U.S. Food and Drug Administration. N Engl J Med 2010; 363: 1489–91. Drucker DJ, Goldfine AB. Cardiovascular safety and diabetes drug development. Lancet 2011; 377: 977–79. Hiatt WR, Kaul S, Smith RJ. The cardiovascular safety of diabetes drugs--insights from the rosiglitazone experience. N Engl J Med 2013; 369: 1285–87. Nissen SE, Wolski K. Effect of rosiglitazone on the risk of myocardial infarction and death from cardiovascular causes. N Engl J Med 2007; 356: 2457–71. US Department of Health and Human Services. Guidance for industry. Diabetes mellitus—evaluating cardiovascular risk in new antidiabetic therapies to treat type 2 diabetes (December, 2008). http://www.fda.gov/downloads/Drugs/GuidanceCompliance RegulatoryInformation/Guidances/ucm071627.pdf (accessed March 3, 2014). European Medicines Agency. Guideline on clinical investigation of medicinal products in the treatment or prevention of diabetes mellitus (May, 2012). http://www.ema.europa.eu/docs/en_GB/ document_library/Scientific_guideline/2012/06/WC500129256.pdf (accessed March 3, 2014). McMurray JJ, Pfeffer MA. Heart failure. Lancet 2005; 365: 1877–89. McMurray JJ. Clinical practice. Systolic heart failure. N Engl J Med 2010; 362: 228–38. Hicks KA, Hung HMJ, Mahaffey KW, et al, on behalf of the Standardized Data Collection for Cardiovascular Trials Initiative. Standardized definitions for cardiovascular and stroke end point events in clinical trials. https://www.cardiac-safety.org/thinktanks/ecrf-forms-for-posting/Draft%20Definitions%20for%20 Testing%20November%209-%202012%20with%20MI%20 Preamble%20CLEAN.pdf/view (accessed March 3, 2014). MacDonald MR, Petrie MC, Hawkins NM, et al. Diabetes, left ventricular systolic dysfunction, and chronic heart failure. Eur Heart J 2008; 29: 1224–40. Voors AA, van der Horst IC. Diabetes: a driver for heart failure. Heart 2011; 97: 774–80. Juhaeri J, Gao S, Dai WS. Incidence rates of heart failure, stroke, and acute myocardial infarction among Type 2 diabetic patients using insulin glargine and other insulin. Pharmacoepidemiol Drug Saf 2009; 18: 497–503. Preiss D, Sattar N, McMurray JJ. A systematic review of event rates in clinical trials in diabetes mellitus: the importance of quantifying baseline cardiovascular disease history and proteinuria and implications for clinical trial design. Am Heart J 2011; 161: 210–19, e1. Barkoudah E, Skali H, Uno H, Solomon SD, Pfeffer MA. Mortality rates in trials of subjects with type 2 diabetes. J Am Heart Assoc 2012; 1: 8–15. Farkouh ME, Domanski M, Sleeper LA, et al, and the FREEDOM Trial Investigators. Strategies for multivessel revascularization in patients with diabetes. N Engl J Med 2012; 367: 2375–84. White WB, Cannon CP, Heller SR, et al, and the EXAMINE Investigators. Alogliptin after acute coronary syndrome in patients with type 2 diabetes. N Engl J Med 2013; 369: 1327–35. Whelton PK, Barzilay J, Cushman WC, et al, and the ALLHAT Collaborative Research Group. Clinical outcomes in antihypertensive treatment of type 2 diabetes, impaired fasting glucose concentration, and normoglycemia: Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT). Arch Intern Med 2005; 165: 1401–09. Barzilay JI, Davis BR, Pressel SL, et al, and the ALLHAT Collaborative Research Group. Long-term effects of incident diabetes mellitus on cardiovascular outcomes in people treated for hypertension: the ALLHAT Diabetes Extension Study. Circ Cardiovasc Qual Outcomes 2012; 5: 153–62. Weber MA, Bakris GL, Jamerson K, et al, and the ACCOMPLISH Investigators. Cardiovascular events during differing hypertension therapies in patients with diabetes. J Am Coll Cardiol 2010; 56: 77–85. Patel A, MacMahon S, Chalmers J, et al, and the ADVANCE Collaborative Group. Intensive blood glucose control and vascular outcomes in patients with type 2 diabetes. N Engl J Med 2008; 358: 2560–72.

www.thelancet.com/diabetes-endocrinology Published online March 13, 2014 http://dx.doi.org/10.1016/S2213-8587(14)70031-2

Personal View

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

Gerstein HC, Miller ME, Byington RP, et al, and the Action to Control Cardiovascular Risk in Diabetes Study Group. Effects of intensive glucose lowering in type 2 diabetes. N Engl J Med 2008; 358: 2545–59. Duckworth W, Abraira C, Moritz T, et al, and the VADT Investigators. Glucose control and vascular complications in veterans with type 2 diabetes. N Engl J Med 2009; 360: 129–39. Kahn SE, Haffner SM, Heise MA, et al, and the ADOPT Study Group. Glycemic durability of rosiglitazone, metformin, or glyburide monotherapy. N Engl J Med 2006; 355: 2427–43. Gerstein HC, Bosch J, Dagenais GR, et al, and the ORIGIN Trial Investigators. Basal insulin and cardiovascular and other outcomes in dysglycemia. N Engl J Med 2012; 367: 319–28. Scirica BM, Bhatt DL, Braunwald E, et al, and the SAVOR-TIMI 53 Steering Committee and Investigators. Saxagliptin and cardiovascular outcomes in patients with type 2 diabetes mellitus. N Engl J Med 2013; 369: 1317–26. Marre M, Lievre M, Chatellier G, Mann JF, Passa P, Ménard J, and the DIABHYCAR Study Investigators. Effects of low dose ramipril on cardiovascular and renal outcomes in patients with type 2 diabetes and raised excretion of urinary albumin: randomised, double blind, placebo controlled trial (the DIABHYCAR study). BMJ 2004; 328: 495–99. Brenner BM, Cooper ME, de Zeeuw D, et al, and the RENAAL Study Investigators. Effects of losartan on renal and cardiovascular outcomes in patients with type 2 diabetes and nephropathy. N Engl J Med 2001; 345: 861–69. Lewis EJ, Hunsicker LG, Clarke WR, et al, and the Collaborative Study Group. Renoprotective effect of the angiotensin-receptor antagonist irbesartan in patients with nephropathy due to type 2 diabetes. N Engl J Med 2001; 345: 851–60. Berl T, Hunsicker LG, Lewis JB, et al, and the Irbesartan Diabetic Nephropathy Trial. Collaborative Study Group. Cardiovascular outcomes in the Irbesartan Diabetic Nephropathy Trial of patients with type 2 diabetes and overt nephropathy. Ann Intern Med 2003; 138: 542–49. Parving HH, Brenner BM, McMurray JJ, et al, and the ALTITUDE Investigators. Cardiorenal end points in a trial of aliskiren for type 2 diabetes. N Engl J Med 2012; 367: 2204–13. Gerstein HC, Yusuf S, Mann JFE, et al, and the Heart Outcomes Prevention Evaluation Study Investigators. Effects of ramipril on cardiovascular and microvascular outcomes in people with diabetes mellitus: results of the HOPE study and MICRO-HOPE substudy. Lancet 2000; 355: 253–59. Daly CA, Fox KM, Remme WJ, Bertrand ME, Ferrari R, Simoons ML, and the EUROPA Investigators. The effect of perindopril on cardiovascular morbidity and mortality in patients with diabetes in the EUROPA study: results from the PERSUADE substudy. Eur Heart J 2005; 26: 1369–78. Lindholm LH, Ibsen H, Dahlöf B, et al, and the LIFE Study Group. Cardiovascular morbidity and mortality in patients with diabetes in the Losartan Intervention For Endpoint reduction in hypertension study (LIFE): a randomised trial against atenolol. Lancet 2002; 359: 1004–10. Okin PM, Devereux RB, Harris KE, et al, and the LIFE Study Investigators. In-treatment resolution or absence of electrocardiographic left ventricular hypertrophy is associated with decreased incidence of new-onset diabetes mellitus in hypertensive patients: the Losartan Intervention for Endpoint Reduction in Hypertension (LIFE) Study. Hypertension 2007; 50: 984–90. Julius S, Weber MA, Kjeldsen SE, et al. The Valsartan Antihypertensive Long-Term Use Evaluation (VALUE) trial: outcomes in patients receiving monotherapy. Hypertension 2006; 48: 385–91. Murcia AM, Hennekens CH, Lamas GA, et al. Impact of diabetes on mortality in patients with myocardial infarction and left ventricular dysfunction. Arch Intern Med 2004; 164: 2273–79. Aguilar D, Solomon SD, Køber L, et al. Newly diagnosed and previously known diabetes mellitus and 1-year outcomes of acute myocardial infarction: the VALsartan In Acute myocardial iNfarcTion (VALIANT) trial. Circulation 2004; 110: 1572–78. Wing RR, Bolin P, Brancati FL, et al, and the Look AHEAD Research Group. Cardiovascular effects of intensive lifestyle intervention in type 2 diabetes. N Engl J Med 2013; 369: 145–54. Carr AA, Kowey PR, Devereux RB, et al. Hospitalizations for new heart failure among subjects with diabetes mellitus in the RENAAL and LIFE studies. Am J Cardiol 2005; 96: 1530–36.

45

46

47 48

49

50

51

52

53

54

55

56

57

58

59

60 61

62

63

Dormandy JA, Charbonnel B, Eckland DJA, et al, and the PROactive investigators. Secondary prevention of macrovascular events in patients with type 2 diabetes in the PROactive Study (PROspective pioglitAzone Clinical Trial In macroVascular Events): a randomised controlled trial. Lancet 2005; 366: 1279–89. Erdmann E, Charbonnel B, Wilcox RG, et al, and the PROactive investigators. Pioglitazone use and heart failure in patients with type 2 diabetes and preexisting cardiovascular disease: data from the PROactive study (PROactive 08). Diabetes Care 2007; 30: 2773–78. Rydén L, Thráinsdóttir I, Swedberg K. Adjudication of serious heart failure in patients from PROactive. Lancet 2007; 369: 189–90. Aksnes TA, Kjeldsen SE, Rostrup M, Omvik P, Hua TA, Julius S. Impact of new-onset diabetes mellitus on cardiac outcomes in the Valsartan Antihypertensive Long-term Use Evaluation (VALUE) trial population. Hypertension 2007; 50: 467–73. Bertoni AG, Hundley WG, Massing MW, Bonds DE, Burke GL, Goff DC Jr. Heart failure prevalence, incidence, and mortality in the elderly with diabetes. Diabetes Care 2004; 27: 699–703. Vaur L, Gueret P, Lievre M, Chabaud S, Passa P, and the DIABHYCAR Study Group (type 2 DIABetes, Hypertension, CARdiovascular Events and Ramipril) study. Development of congestive heart failure in type 2 diabetic patients with microalbuminuria or proteinuria: observations from the DIABHYCAR (type 2 DIABetes, Hypertension, CArdiovascular Events and Ramipril) study. Diabetes Care 2003; 26: 855–60. Home PD, Pocock SJ, Beck-Nielsen H, et al, and the RECORD Study Team. Rosiglitazone evaluated for cardiovascular outcomes in oral agent combination therapy for type 2 diabetes (RECORD): a multicentre, randomised, open-label trial. Lancet 2009; 373: 2125–35. Komajda M, McMurray JJ, Beck-Nielsen H, et al. Heart failure events with rosiglitazone in type 2 diabetes: data from the RECORD clinical trial. Eur Heart J 2010; 31: 824–31. Punthakee Z, Bosch J, Dagenais G, et al, and the TIDE Trial Investigators. Design, history and results of the Thiazolidinedione Intervention with vitamin D Evaluation (TIDE) randomised controlled trial. Diabetologia 2012; 55: 36–45. dos Santos L, Salles TA, Arruda-Junior DF, et al. Circulating dipeptidyl peptidase IV activity correlates with cardiac dysfunction in human and experimental heart failure. Circ Heart Fail 2013; 6: 1029–38. Lourenço P, Friões F, Silva N, Guimarães JT, Bettencourt P. Dipeptidyl peptidase IV and mortality after an acute heart failure episode. J Cardiovasc Pharmacol 2013; 62: 138–42. Shigeta T, Aoyama M, Bando YK, et al. Dipeptidyl peptidase-4 modulates left ventricular dysfunction in chronic heart failure via angiogenesis-dependent and -independent actions. Circulation 2012; 126: 1838–51. Gomez N, Touihri K, Matheeussen V, et al. Dipeptidyl peptidase IV inhibition improves cardiorenal function in overpacing-induced heart failure. Eur J Heart Fail 2012; 14: 14–21. Boonman-de Winter LJ, Rutten FH, Cramer MJ, et al. High prevalence of previously unknown heart failure and left ventricular dysfunction in patients with type 2 diabetes. Diabetologia 2012; 55: 2154–62. Takahashi A, Asakura M, Ito S, et al. Dipeptidyl-peptidase IV inhibition improves pathophysiology of heart failure and increases survival rate in pressure-overloaded mice. Am J Physiol Heart Circ Physiol 2013; 304: H1361–69. Scheen AJ. Cardiovascular effects of gliptins. Nat Rev Cardiol 2013; 10: 73–84. Stanley WC, Recchia FA, Lopaschuk GD. Myocardial substrate metabolism in the normal and failing heart. Physiol Rev 2005; 85: 1093–129. Marso SP, Poulter NR, Nissen SE, et al. Design of the liraglutide effect and action in diabetes: evaluation of cardiovascular outcome results (LEADER) trial. Am Heart J 2013; 166: 823–30, e5. Green JB, Bethel MA, Paul SK, et al. Rationale, design, and organization of a randomized, controlled Trial Evaluating Cardiovascular Outcomes with Sitagliptin (TECOS) in patients with type 2 diabetes and established cardiovascular disease. Am Heart J 2013; 166: 983–89 e7.

www.thelancet.com/diabetes-endocrinology Published online March 13, 2014 http://dx.doi.org/10.1016/S2213-8587(14)70031-2

9