The Harmonising Outcomes with Revascularisation and Stents in Acute Myocardial Infarction (HORIZONS-AMI) was a prospective open label, multi-centre controlled trial involving patients with ST-elevation myocardial infarction (STEMI) undergoing primary percutaneous coronary intervention (PPCI). It incorporated two factorial randomised phases to allow a comparison of the direct thrombin inhibitor bivalirudin alone with heparin plus glycoprotein IIb-IIIa inhibitor use and a comparison of paclitaxel eluting stents (PES) with bare metal stents (BMS). Two primary clinical end-points were pre-specified: ischaemia driven TLR (analysis powered for superiority) and a composite safety end-point of major adverse cardiovascular events consisting of death, reinfarction, stroke and stent thrombosis (powered for non-inferiority with a 3.0% margin). The major secondary end-point was angiographic evidence of restenosis at 13 months

3602 patients with STEMI underwent randomisation for the pharmacological phase of the trial and of these 3006 underwent randomisation in a 3:1 ratio to the stent phase. 2257 were assigned to PES and 749 received BMS. 1 year follow up data were available for 2186 and 715 patients respectively. Patients with PES, when compared to those receiving BMS, had significantly lower ischaemia driven TLR (4.5% vs 7.5%, HR 0.59, 95%CI 0.43-0.83, p=0.002) and target vessel revascularisation (TVR) (5.8% vs 8.7%, HR 0.65, 95% CI 0.48-0.89, p=0.006), with non-inferior rates of the composite safety end-point (8.1% vs 8.0%, HR 1.02, 95%CI 0.76-1.36; absolute difference 0.1%, 95%CI -2.1 to 2.4, p=0.01 for non-inferiority, p=0.92 for superiority). Both PES and BMS had similar 12 month rates of death (3.5% and 3.5%, p=0.98) and stent thrombosis (3.2% and 3.4%, p=0.77) respectively. The 13 month rate of binary restenosis was significantly lower with PES than with BMS (10% vs 22.9%, HR 0.44, 95%CI 0.33-0.57, p<0.001).

There are some limitations of this study. Firstly, logistic complexities meant that an open label design was necessary. High rates of protocol compliance and the use of blinded clinical event adjudication and core laboratory assessments were measures taken to mitigate potential bias. There was a slightly higher rate of thienopyridine use in the PES arm than the BMS arm in the period between 6 months and 1 year. Very few patients with cardiogenic shock were included in this study and patients with unprotected left main stem disease or bifurcation disease requiring a 2 stent strategy were not included – therefore the results should not be extended to these patient populations. Longer term follow up data will be necessary to characterise the late safety and efficacy profiles of PES in patients with evolving STEMI particularly as the use of dual anti-platelet therapy declines over time after stent implantation. This is particularly important when considering the increased risk of stent thrombosis with DES when compared to BMSwhich may only become apparent at more than 1 year after stent implantation.

Conclusion:

The HORIZONS-AMI trial provides data indicating that PES may be used in patients with evolving STEMI. The data from HORIZONS-AMI show that approximately 30 fewer/1000 patients with PES required TVR at 1 year when compared to BMS.

· Stone GW, Lansky AJ, Pocock SJ et al. Paclitaxel-eluting stents versus bare-metal stents in acute myocardial infarction. The New England Journal of Medicine (2009) vol. 360 (19) pp. 1946-59

Despite a strong association between elevated blood glucose levels and increased rates of mortality among patients hospitalised with acute myocardial infarction (AMI), the benefit of acutely lowering levels of blood glucose remains controversial. Several questions remain unanswered:

(1) Does normalisation of admission hyperglycaemia improved outcome?

(2) What level of blood glucose is associated with lowest mortality?

(3) Do outcomes differ between patients who spontaneously achieve normalisation of glucose levels as compared to those who require insulin?

In order to address these questions Kosiborod et al retrospectively analyzed data from 7820 hyperglycaemic patients hospitalised for biomarker-confirmed AMI, using a database which received patient data from 40 hospitals across the US over a 6 year period (Jan 2000-Dec 2005). Measurements of blood glucose were divided into 5 levels as shown in the table below. Admission blood glucose level was defined as the initial blood glucose level, and postadmission blood glucose level was defined as the mean of all glucose levels obtained after the initial measurement. Patients with admission glucose levels <7.8 were excluded from the analysis. Outcome was defined as all cause in hospital mortality.

Patients with higher postadmission glucose levels had a higher incidence of heart failure (p<0.001) and diabetes (p<0.001) as compared to those with lower postadmission glucose levels. Furthermore these patients were less like to receive percutaneous intervention, had worse renal function, higher leukocyte count, were less frequently treated with antiplatelet drugs (including aspirin) and experienced longer hospital stays. Lower postadmission glucose levels were associated with better in hospital survival both in unadjusted and multivariate adjusted analyses (see table).

Lowest mortality was observed among patients with a postadmission glucose level of 4.4 – 7.2mmol/L. Interestingly, following multivariate adjustment admission glucose level failed to predict in hospital mortality (p=0.89) whereas postadmission glucose level remained a robust predictor (p<0.001). Propensity matching failed to demonstrate a difference in outcome among patients treated with insulin and those who were not.

Conclusion:

Glycaemic control, rather than insulin use, appears to be the important factor following AMI. However these results should be interpreted with caution as normalisation of glycaemia may simply represent a surrogate marker of healthy physiology and a recovery from the stress of AMI rather than identify a causal mechanism.

· Kosiborod M, Inzucchi S, Krumholz H et al. Glucose normalization and outcomes in patients with acute myocardial infarction. Arch Intern Med 2009;169(5):438-46

The ACTIVE Trial (Atrial Fibrillation Clopidogrel Trial with Irbesartan for the Prevention of Vascular Events) was designed to assess the role of clopidogrel and aspirin in the prevention of stroke and other vascular events in patients with atrial fibrillation (AF). ACTIVE W has already reported and compared clopidogrel + aspirin with a vitamin K antagonist. The ACTIVE A trial, the results of which are discussed here, compared combination therapy with clopidogrel and aspirin versus aspirin alone in patients with AF who were considered unsuitable for a Vitamin K antagonist.

This was a randomised, double-blind multi-centre trial involving 7554 patients from 580 centres in 33 countries. Patients were eligible for ACTIVE (either A or W) if they had AF at the time of enrolment or had had 2 episodes of AF in the preceding 6 months. Additionally, patients were required to have at least one of the following risk factors: age > 75 years, systemic hypertension, previous stroke or TIA, LVEF <45%, peripheral vascular disease or age 55-74 years and diabetes mellitus or coronary artery disease. Patients considered to be candidates for vitamin K antagonists were enrolled in ACTIVE W and those not considered suitable in ACTIVE A. Patients in both trials could also be randomised to ACTIVE I which involved factorial randomisation to either irbesartan or placebo – this study is ongoing.

Patients in ACTIVE A were randomised to either aspirin and clopidogrel (n=3772) or aspirin (n=3782). Follow up visits to assess compliance and clinical outcome were held at 1 month and then 3 monthly for the first year and 6 monthly until the end of the trial. The primary outcome was any major vascular event (stroke, non central nervous system systemic embolism, myocardial infarction or death from vascular causes). The main secondary outcome measures were the other individual components of the primary outcome and the composite of the primary outcome and major haemorrhage.

At a median of 3.6 years follow up, major vascular events had occurred in 832 patients receiving clopidogrel (6.8%/year) and 924 patients receiving placebo (7.6%/year), (relative risk with clopidogrel 0.89, 95%CI 0.81-0.98, p = 0.01). The difference was predominantly due to a reduction in the rate of stroke with clopidogrel – stroke occurred in 296 patients in the dual anti-platelet therapy arm (2.4%/year) versus 408 placebo patients (3.3%/year), (RR 0.72, 95%CI 0.62-0.83, p < 0.001). Myocardial infarction occurred in 90 patients receiving clopidogrel (0.7%/year) and 115 receiving placebo (0.9%/year, RR 0.78, 95%CI 0.59-1.03, p = 0.08). Major bleeding occurred in 251 patients receiving clopidogrel (2%/year) and 162 patients receiving placebo (1.3%/year) (RR1.57, 95%CI 1.29-1.92, p<0.001). Rates of discontinuation of the study medication were 16.3% and 15.2% for clopidogrel and placebo respectively at 1 year, rising to 39.4% and 37.1% at 4years. All patients were receiving aspirin at the time of randomisation and this decreased to 92.9% at 1 year and 81.1% at 4 years in both groups. A total of 777 patients (10.3%) received a vitamin K antagonist following discontinuation of the trial medication.

Conclusion:

For patients unable to take vitamin K antagonists, the addition of clopidogrel to aspirin therapy significantly reduced the rate of major vascular events, primarily driven by a reduction in strokes. However, one extra major bleeding episode would occur for ever 143 patients treated for one year with this combination.

· ACTIVE Investigators et al. Effect of clopidogrel added to aspirin in patients with atrial fibrillation. The New England Journal of Medicine (2009) vol. 360 (20) pp. 2066-78

Animal studies have suggested that bone marrow cell injection can improve myocardial perfusion and improve ventricular function in chronic ischaemia, but the two randomized trials performed to date in humans have yielded conflicting results.

To investigate the effect of intramyocardial bone marrow cell injection on myocardial perfusion and left ventricular (LV) function further, Ramshorst et al. randomised 50 patients with chronic myocardial ischaemia to either intramyocardial injection of autologous bone marrow cells or placebo solution (25 patients per group). The primary outcome measure was myocardial perfusion as assessed by the summed stress score, a 17-segment score for stress myocardial perfusion assessed by Tc-99m tetrofosmin single-photon emission computed tomography (SPECT), while secondary measures included LV ejection fraction (LVEF) as assessed by MRI, Canadian Cardiovascular Society (CCS) angina class, and the Seattle Angina Questionnaire quality-of-life score.

After 3-months of follow-up, the summed stress score improved from 23.5 (SD, 4.7) to 20.1 (SD, 4.6; p<0.001) in the bone marrow cell group, compared with a decrease from 24.8 (SD, 5.5) to 23.7 (SD, 5.4; p=.004) in the placebo group. In those patients who underwent MRI, a 3% increase in LVEF was seen in the bone-marrow cell treated patients at 3 months (95% CI, 0.5% to 4.7%), but the placebo group showed no improvement. Improvements in CCS class an quality-of-life score were significantly greater in bone marrow cell-treated patients that in placebo-treated patients (p=.03 and p=.04, respectively).

Conclusion:

In this small study of patients with angina refractory to medical treatment, intramyocardial bone marrow injection resulted in a statistically significant improvement in myocardial perfusion, ejection fraction, and quality-of-life at 3 months. Whether this translates into an improvement in morbidity and mortality outcomes requires verification in further long-term studies.

· Van Ramshorst J, Bax JJ, Beeres SLMA et al. Intramyocardial bone marrow cell injection for chronic myocardial ischemia: a randomized controlled trial. JAMA (2009) vol. 301 (19) pp. 1997-2004

The recurrence of atrial fibrillation (AF) after cardioversion may be partially related to a process known as remodeling - the electrical, mechanical and structural properties of the atrial tissue are altered in a progressive and irreversible manner resulting in a more favourable substrate for AF. From animal models, blockade of the rennin-angiotensin-aldosterone system (RAAS) has been shown to alter the remodeling process in a beneficial manner.

The GISSI-AF (Gruppo Italiano per lo Studio della Sopravvivenza nell’Infarto Miocardico-Atrial Fibrillation) study was a prospective, multi-centre, randomised, double-blind, placebo controlled trial. It assessed whether the addition of the AIIRB valsartan to established treatments could reduce the recurrence of AF in patients with a history of arrhythmia. Patients were eligible if they were at least 40 years of age and had had 2 or more episodes of symptomatic atrial fibrillation in the previous 6 months or a successful cardioversion (electrical or pharmacological) 48 hours-14 days before randomisation. Patients had to have been in sinus rhythm for a minimum of 48 hours prior to enrolment. Participants were also required to have been on a stable regimen for treatment for atrial fibrillation or other cardiac conditions for 1 month.

1442 patients were randomly assigned to either valsartan (n=722) or placebo (n=720). Study visits were scheduled at 2, 4, 8, 24 and 52 weeks with follow up concluding at 1 year. An ECG was performed at each visit and to increase the likelihood of AF being detected patients were provided with a transtelephonic device with instructions to activate this if symptoms occurred. Additionally they were to activate this device on a weekly basis - regardless of symptoms - and a 30 second ECG was sent to both the co-ordinating centre and the responsible physician. The study was designed with 2 primary end-points: time to first recurrence of AF and the proportion of patients who had more than 1 episode of AF in the follow up period. Secondary end-points included: total number of episodes of AF/patient, hospitalisation for any reason, hospitalisation for a cardiovascular event, the composite of death and thromboembolic events, number of patients in sinus rhythm at each study visit, duration of and ventricular rate at the first recurrence of AF and a safety profile.

AF recurred in 371/722 patients (51.4%) in the valsartan group compared to375/720 (52.1%) in the placebo group (adjusted hazard ratio (HR) 0.97, 96%CI 0.83-1.14, p = 0.73). 194/722 (26.9%) in the valsartan arm and 201/720 (27.9%) in the placebo arm (adjusted odds ratio 0.89, 99% CI 0.64-1.23, p = 0.34 had more than one episode of AF. The results were similar in all the pre-defined sub-groups including those who were not receiving ACE inhibitors.

These results provide demonstrate that valsartan did not prevent the recurrence of AF in this population. However there are some important caveats. The majority of the study population had reasonably well controlled hypertension (85%), 57% were already taking ACE inhibitors and >70% were taking a Class I or III anti-arrhythmic agent for the prevention of atrial fibrillation. In addition, 88% had undergone cardioversion (either electrical or pharmacological) in the 2 week period preceding randomisation. These factors would suggest that the substrate for AF was well established in the population studied, and that it may have been too late for medication to make a noticeable difference. Previous studies have suggested that patients with LV dysfunction or hypertrophy benefit the most from use of ACE inhibitors or AIIRBs for prevention of AF. However in this study population only 8% of patients had either heart failure or ventricular systolic dysfunction and data on presence of left ventricular hypertrophy were not provided. Finally, this was a study of the secondary prevention of AF and it is possible that the greatest benefit from ACE inhibitors and A II RBs will be in the prevention of new onset AF (‘primary’ prevention) since it may be that these drugs can prevent but not reverse the structural remodelling that acts as a substrate for AF.

The GISSI-AF Investigators. Valsartan for Prevention of Recurrent Atrial Fibrillation. N Engl J Med 2009;360:1606-17

Gillis AM (ed). Angiotensin Receptor Blockers for Prevention of Atrial Fibrillation – A Matter of Timing or Target? N Engl J Med 2009;360:1669-71

Should cardiologists who are not electrophysiologists be inserting implantable cardioverter-defibrillators (ICDs)? Curtis et al. performed a retrospective cohort study of patients who had an ICD fitted between January 2006 and June 2007. Patients were grouped according to whom had implanted their device: electrophysiologists, nonelectrophysiologist cardiologists, thoracic surgeons, and other specialists. The main outcome measures examined were the proportion of patients meeting the criteria for a defibrillator with cardiac resynchronization therapy (CRT-D) who received that device, and the in-hospital procedural complication rates.

111293 ICD implantations were included in the analysis. Overall, 70.9% were performed by electrophysiologists, 21.9% by nonelectrophysiologist cardiologists, 1.7% by thoracic surgeons, and 5.5% by other specialists. Patients who had an ICD inserted by anyone other than an electrophysiologist were found to be at an increased risk for complications both in unadjusted (p<0.001) and adjusted analyses (relative risk for nonelectrophysiologist cardiologists, 1.11; relative risk for thoracic surgeons, 1.44). Although 35841 patients met criteria for CRT-D, when a nonelectrophysiologist cardiologist or thoracic surgeon performed the procedure, patients were significantly less likely to receive the correct device (relative risk for nonelectrophysiologist cardiologists, 0.93; relative risk for thoracic surgeons, 0.81). The study also demonstrated that electrophysiologists were available in the majority of hospitals where thoracic surgeons and general cardiologist were performing their ICD insertions.

This study only looked at in-hospital outcomes, and it may be that long-term complications are equally as skewed; the same group is planning longer-term follow-up. Currently, no criteria exist for determining what procedures a physician may or may not perform; this paper demonstrates that this problem will become increasingly important as sub-specialisation in cardiology continues.

Curtis JP, Luebbert JJ, Wang Y, et al. Association of physician certification and outcomes among patients receiving an implantable cardioverter defibrillator. JAMA 2009; 301:1661-1670.

Coromilas J. Physician credentials and ICD implantation. JAMA 2009; 301:1713-1174.

Should patients with type 2 diabetes be screened for coronary artery disease if asymptomatic? In the DIAD study (Detection of Ischemia in Asymptomatic Diabetics), 1123 patients with type 2 diabetes and no symptoms of coronary artery disease (CAD) were randomly assigned to be screened with adenosine-stress myocardial perfusion imaging (MPI) or to no screening. The main outcome measure was the incidence of cardiac death or nonfatal myocardial infarction (MI) during long-term follow-up.

The cumulative cardiac event rate was 2.9% over a mean follow-up of 4.8 years for an average of 0.6% per year. Seven nonfatal MIs and eight cardiac deaths (2.7%) occurred among the screened group, and ten nonfatal MIs and seven cardiac deaths (3.0%) among the group that received no screening (p=.73). However, of the patients who underwent screening, those with normal results (n=409) and those with small MPI defects (n=50) had lower event rates than the 33 with moderate or large MPI defects (0.4% vs 2.4% per year, p=.001).

Therefore screening using SPECT appeared to make no difference to the rate of coronary events in this study, and the event rate overall was low, which translates into a high cost of screening per event prevented. Indeed a low event rate was also seen in the recent ACCORD (Action to Control Cardiovascular Risk in Diabetes), leading some to question whether diabetic patients should truly be labelled as having the same risk as patients who have already had coronary events. For example, the ARIC (Atherosclerosis Risk in the Community) study found a lower mortality rate amongst diabetics than amongst patients with known coronary disease.

· Young LH, Wackers FJ, Chyun DA, et al. Cardiac outcomes after screening for asymptomatic coronary artery disease in patients with type 2 diabetes. JAMA 2009; 301: 1547-1555.

· Wackers FJ, Young LH, Inzucchi SE, et al. Detection of silent myocardial ischemia in asymptomatic diabetic subjects: the DIAD study. Diabetes Care 2004; 27: 1954-1961

It has been calculated that a polypill containing asprin, a beta-blocker, a statin and an angotensin-converting enzyme inhibitor could reduce cardiovascular events in people with cardiovascular disease by about 75%. It has further been suggested that the addition of folic acid, and the use of three separate agents to lower blood pressure (each in low doses), could enable a polypill preparation to reduce cardiovascular disease by more than 80%. The Indian Polycap Study (TIPS) was designed to examine the efficacy of just such a formulation, named the Polycap pill.

2053 patients without cardiovascular disease, all aged 45-80 years but with one cardiovascular risk factor, were randomised to the Polycap (n=412) or to one of eight other groups - each with about 200 patients - of aspirin alone, simvastatin alone, hydrochlorthiazide alone, three different combinations of two blood-pressure lowering drugs, three blood-pressure lowering drugs alone, or three blood-pressure-lowering drugs with aspirin.

Compared with groups not receiving blood pressure drugs, the Polycap reduced systolic blood pressure by 7.4mmHg and diastolic blood pressure by 5.6mmHg. Although the Polycap reduced LDL cholesterol by 0.70mmol/L, this was less than the effect of simvastatin alone (0.83mmol/L; p=0.04). Tolerability of the Polycap was similar to that of other treatments, with no evidence of higher intolerability as the number of active components increased.

This phase II study confirms that a polypill formulation can be tolerated and used to reduced multiple risk factors. Importantly, the study also demonstrates that the effects of a polypill may not necessarily be totally predicted by its components – the weakness of the cholesterol effect demonstrates this. Of course, the trial does not tell us about morbidity and mortality effects, although the authors predict these values and compare them to those projected by another study by Wald and Law (see table). Nonetheless, this study marks an important step in the ongoing development of the polypill.

· The Indian Polycap Study (TIPS). Effects of a polypill (Polycap) on risk factors in middle-aged individuals without cardiovascular disease (TIPS): a phase II, double-blind, randomized trial. Lancet 2009; DOI:10.1016/S0140-6736(09)60611-5.

· Cannon CP. Can the polypill save the world from heart disease? Lancet 2009; DOI:10.1016/S0140-6736(09)60652-8.

Surgical ventricular reconstruction is a technique which has been developed to reconstruct dysfunctional myocardial segments; it is usually performed in conjunction with coronary artery bypass grafting (CABG). Previous studies of this technique have not been randomised and have been limited to observational studies.

STICH (Surgical Treatment for Ischaemic Heart Failure Trial) was a multi-centre, non-blinded, randomised trial conducted at 127 sites in 26 countries. There were 2 major hypotheses. The Hypothesis 1 substudy compared medical therapy and CABG with medical therapy alone and has not yet reported. The Hypothesis 2 substudy compared CABG alone with the combined procedure of CABG and ventricular reconstruction. 1000 patients were recruited from 96 medical centres in 23 countries. Patients were required to have coronary disease amenable to CABG, LVEF <35% and a dominant area of anterior myocardial akinesia or dyskinesia that was suitable for treatment with surgical ventricular reconstruction. All patients received standard medical and device therapy for heart failure.

Major peri-operative events and specified end-points were recorded at discharge or 30 days for patients still in hospital. The primary outcome was time to death from any causes or hospitalisation for cardiac causes. Secondary outcomes included death from any cause at 30 days, hospitalisation for any cause and for cardiovascular causes, myocardial infarction and stroke.

499 patients underwent CABG and 501 underwent the combined procedure of CABG and ventricular reconstruction. The median follow up was for 48 months. Surgical ventricular reconstruction reduced the end systolic volume index by 19% as compared with a reduction of 6% with CABG alone. Cardiac symptoms and exercise tolerance improved form baseline to a similar degree in the two groups. No significant difference was seen in the primary outcome which occurred in 292 (59%) patients undergoing CABG and 289 (58%) undergoing CABG with surgical ventricular reconstruction (hazard ratio for the combined approach 0.99, 95%CI 0.84-1.17, p = 0.90).

This large trial had some key strengths including the use of up-to-date surgical techniques and an attempt to standardise heart failure therapies according to guidelines throughout the entire study population. However there are some important limitations. Whilst the authors state that use of evidence based pharmacological and device therapies was monitored throughout the study, no information is given about rates of use of these over the time of the study. There was a greater reduction in the end systolic volume index with the combined procedure (16ml/m² body surface area) compared to CABG alone (5ml/m² body surface area) but these data were only obtained from 373 patients at baseline and at 4 months. More complete and longitudinal information on the end systolic volume index might have provided supporting evidence as to whether ongoing ventricular modelling did occur. Potential explanations for the lack of added efficacy of the combined procedure include the fact that the available heart failure therapies are effective at limiting adverse remodelling. Revascularisation with CABG may enhance this process resulting in a lack of any additional benefit being seen with surgical ventricular reconstruction. When the Hypothesis 1 substudy reports it is hoped that this will provide further insights into the benefits of medical therapies.

• Jones RH, Velazquez EJ, Michler RE et al. Coronary Bypass Surgery with or without Surgical Ventricular Reconstruction. N Engl J Med 2009;360:1705-17
• Eisen HJ (ed). Surgical Ventricular Reconstruction for Heart Failure. N Engl J Med 2009;360:1781-4

Depressive symptoms are prominent in elderly hospitalized patients with heart failure and have been associated with decreased functional status, increased hospital readmissions and greater mortality. However, little is known about the characteristics of heart failure patients prone to depression and their treatment.

In this analysis of the OPTIMIZE-HF (Organized Program to Initiate Lifesaving Treatment in Hospitalized Patients with Heart Failure) study, a multi-centre registry based in the US of some 48,612 patients at 219 hospitals, a history of depression was present in 10.6% of individuals. Depression was significantly more common in females, whites and those with concomitant diseases such as COPD and insulin-dependent diabetes. Patients with depression were much less likely to receive coronary interventions or cardiac devices or to be referred to outpatient disease management programs.

Due to the retrospective nature of the study, the authors are unable to comment on the reasons for these seeming disparities but these factors may go some way towards explaining the poorer outcomes of some patient groups. The randomized controlled trial MOOD-HF investigating the use of SSRIs in depressed heart failure patients will provide evidence based recommendations for management, but in the meantime an increased awareness of the impact of depression may provide an opportunity to improve outcomes in heart failure patients.

Nancy M. Albert, Gregg C. Fonarow, William T. Abraham et al. Depression and Clinical Outcomes in Heart Failure: An OPTIMIZE-HF Analysis. The American Journal of Medicine (2009) 122, 366-373