Comparative outcomes after unprotected left main stem percutaneous coronary intervention: a national linked cohort study of 5,065 acute and elective cases from the BCIS Registry (British Cardiovascular Intervention Society).

OBJECTIVES: The goal of this study was to report outcomes from percutaneous coronary intervention (PCI) to an unprotected left main stem (UPLMS) stenosis according to presenting syndrome, including ST-segment elevation myocardial infarction (STEMI), non-ST-segment elevation acute coronary syndrome (NSTEACS), and chronic stable angina (CSA). BACKGROUND: There are no published whole-country data concerning patient outcomes following PCI to UPLMS. METHODS: This study is a prospective national cohort study using data from the British Cardiovascular Intervention Society (BCIS) registry from January 1, 2005, through December 31, 2010. RESULTS: Of 5,065 patients having PCI to an UPLMS, 784 (15.5%) presented with STEMI, 2,381 (47.0%) with NSTEACS, and 1,900 (37.5%) with CSA. Crude 30-day and 1-year mortality rates were STEMI: 28.3% and 37.6%, NSTEACS: 8.9% and 19.5%, and CSA: 1.4% and 7.0%, respectively. Unadjusted in-hospital major adverse cardiovascular and cerebrovascular event rates were STEMI: 26.6%, NSTEACS: 6.6%, and CSA: 3.3%. Risk of 30-day mortality was much greater for STEMI and NSTEACS patients than CSA (STEMI adjusted odds ratio [aOR]: 29.45, 95% confidence interval [CI]: 19.37 to 44.80, NSTEACS aOR: 6.45, 95% CI: 4.27 to 9.76). More than 40% of patients presenting with STEMI had cardiogenic shock, in whom mortality was higher than in STEMI cases without shock (30 days: 52.0% vs. 11.7%, 1 year: 61.1% vs. 20.9%). Radial access, compared with the femoral approach, was associated with a lower risk of 30-day mortality (STEMI aOR: 0.37, 95% CI: 0.21 to 0.62; NSTEACS aOR: 0.66, 95% CI: 0.45 to 0.97). CONCLUSIONS: More than one-half of the patients who received UPLMS PCI were acute where outcomes were much worse than elective cases. Cardiogenic shock is common in STEMI patients, of whom more than one-half die at 30 days. The radial approach was associated with reduced early mortality in acute cases.

Fractional flow reserve versus angiography in guiding management to optimize outcomes in non-ST-elevation myocardial infarction (FAMOUS-NSTEMI): rationale and design of a randomized controlled clinical trial.

BACKGROUND: In patients with acute non-ST-elevation myocardial infarction (NSTEMI), coronary arteriography is usually recommended; but visual interpretation of the angiogram is subjective. We hypothesized that functional assessment of coronary stenosis severity with a pressure-sensitive guide wire (fractional flow reserve [FFR]) would have additive diagnostic, clinical, and health economic utility as compared with angiography-guided standard care. METHODS AND DESIGN: A prospective multicenter parallel-group 1:1 randomized controlled superiority trial in 350 NSTEMI patients with ≥1 coronary stenosis ≥30% severity (threshold for FFR measurement) will be conducted. Patients will be randomized immediately after coronary angiography to the FFR-guided group or angiography-guided group. All patients will then undergo FFR measurement in all vessels with a coronary stenosis ≥30% severity including culprit and nonculprit lesions. Fractional flow reserve will be disclosed to guide treatment in the FFR-guided group but not disclosed in the "angiography-guided" group. In the FFR-guided group, an FFR ≤0.80 will be an indication for revascularization by percutaneous coronary intervention or coronary artery bypass surgery, as appropriate. The primary outcome is the between-group difference in the proportion of patients allocated to medical management only compared with revascularization. Secondary outcomes include the occurrence of cardiac death or hospitalization for myocardial infarction or heart failure, quality of life, and health care costs. The minimum and average follow-up periods for the primary analysis are 6 and 18 months, respectively. CONCLUSIONS: Our developmental clinical trial will address the feasibility of FFR measurement in NSTEMI and the influence of FFR disclosure on treatment decisions and health and economic outcomes.

Late gadolinium enhancement cardiac magnetic resonance imaging for the prediction of ventricular tachyarrhythmic events: a meta-analysis.

AIMS: Approaches to the risk stratification for sudden cardiac death (SCD) remain unsatisfactory. Although late gadolinium enhancement cardiac magnetic resonance imaging (LGE-CMR) for SCD risk stratification has been evaluated in several studies, small sample size has limited their clinical validity. We performed this meta-analysis to better gauge the predictive accuracy of LGE-CMR for SCD risk stratification. METHODS AND RESULTS: Electronic databases and published bibliographies were systematically searched to identify studies evaluating the association between the extent of LV scar on LGE-CMR and ventricular arrhythmic events [SCD, resuscitated cardiac arrest, the occurrence of ventricular arrhythmias, or appropriate implantable cardioverter defibrillator (ICD) therapy]. Only studies enrolling patients with CAD or non-ischaemic cardiomyopathy were included. Summary estimates of the relative risk (RR) and likelihood ratios (LRs) were calculated using random effects models. Eleven studies comprising 1105 patients were identified. During a mean/median follow-up of 8.5-41 months 207 patients had ventricular arrhythmic events. Ventricular arrhythmic events were more common in patients with a greater extent of LV scar: RR 4.33 [95% confidence interval (CI) 2.98-6.29], positive LR 1.98 (95% CI 1.66-2.37), and negative LR 0.33 (95% CI 0.24-0.46). CONCLUSION: The extent of LGE on CMR is strongly associated with the occurrence of ventricular arrhythmias in patients with reduced LVEF and may be a valuable risk stratification tool for identifying patients who will benefit from ICD therapy. However, uncertainties regarding clinical application persist and need to be addressed prior to introduction into broad clinical practice.

Can QRS scoring predict left ventricular scar and clinical outcomes?

AIMS: Identifying patients with potential to benefit from implantable cardioverter defibrillator (ICD) therapy is challenging. Myocardial scar detected using cardiovascular myocardial resonance imaging with late gadolinium enhancement (CMR-LGE) is associated with ventricular arrhythmia. Its use is constrained due to limited availability, unlike electrocardiogram (ECG) which is widely available. Selvester QRS scoring detects scar, although the reported performance varies. The study aims were to determine whether QRS score (a) detects scar (b) varies with scar characteristics, and (c) can meaningfully predict sudden cardiac death. METHODS AND RESULTS: We investigated 64 consecutive ICD recipients (age 66 ± 11 years, 80% male, median left ventricular ejection fraction 30%) with coronary artery disease who had undergone CMR-LGE prior to device implantation, over 4 years in a single centre (2006-2009). A modified QRS score was measured on the ECG performed prior to ICD implantation. Clinical end points were (i) appropriate ICD therapy and (ii) all cause mortality. QRS score was associated with CMR scar (r = 0.42, P = 0.001) and scar surface area (r = 0.41, P = 0.001), but not subendocardial scar. Strongest correlation was seen in those patients with transmural scar only (r = 0.62, P = 0.01). During 42 ± 13 months follow-up, QRS score was not predictive of appropriate ICD therapy, but was significantly related to all cause mortality (hazard ratio = 1.16; confidence interval = 1.03-1.30; P = 0.01). CONCLUSION: QRS scoring performed best in quantifying transmural scar, and shows association with medium-term mortality risk, but not with risk of ventricular arrhythmia. It may be that the score is best suited as a risk stratifier of those with least potential to benefit from ICD.

The relationship between left ventricular scar and ventricular repolarization in patients with coronary artery disease: insights from late gadolinium enhancement magnetic resonance imaging.

AIMS: The markers of ventricular repolarization corrected QT interval (QTc), QT dispersion (QTD) and Tpeak-to-Tend interval (Tpeak-end) have shown an association with sudden cardiac death (SCD) in the general population. However, their mechanistic relationship with SCD is unclear. The study aim was to evaluate the relationship between QTc, QTD, and Tpeak-end, and the extent and distribution of left ventricular (LV) scar in patients with coronary artery disease at high SCD risk. METHODS AND RESULTS: We included 64 consecutive implantable cardioverter defibrillator (ICD) recipients (66 ± 11 years, 80% male, median left ventricular ejection fraction 30%) who had undergone late gadolinium enhancement cardiac magnetic resonance (CMR) imaging prior to device implantation over 4 years. Scar was quantified using the CMR images and characterized in terms of percent LV scar and number of LV segments with subendocardial/transmural scar. Repolarization parameters were measured on an electrocardiogram performed prior to ICD implantation. After adjustment for potential confounders there was a strong association between the number of limited subendocardial (1-25% transmurality) scar segments and QTc (P = 0.003), QTD (P = 0.002), and Tpeak-end (P = 0.008). However, there was no association between the repolarization parameters and percent LV scar or the amount of transmural scar. During a mean follow-up of 19 ± 10 months 19 (30%) patients received appropriate ICD therapy, but none of the repolarization parameters were associated with its occurrence. CONCLUSION: In this pilot study there was a strong association between limited subendocardial LV scar and prolonged QTc, QTD, and Tpeak-end. However, there was no association between any of these repolarization markers and the delivery of appropriate ICD therapy.

Left ventricular scar burden specifies the potential for ventricular arrhythmogenesis: an LGE-CMR study.

INTRODUCTION: The extent of left ventricular (LV) scar, characterized by late gadolinium enhancement cardiac MRI (LGE-CMR), has been shown to predict the occurrence of ventricular arrhythmias in implantable cardioverter defibrillator (ICD) recipients. However, the specificity of LGE-CMR for sudden cardiac death (SCD) versus non-SCD is unclear. The aim of this retrospective, observational study was to evaluate this relationship in a cohort of ICD recipients. METHODS AND RESULTS: We included consecutive patients who had undergone LGE-CMR before ICD implantation over a 4-year period (2006-2009). Scar (defined as myocardium with a signal intensity ≥50% of the maximum in scar tissue) was characterized in terms of percent scar and number of transmural LV scar segments in a 17-segment model. The endpoints were appropriate ICD therapy and all-cause mortality. Sixty-four patients (average age 66 ± 11 years, 51 male, median LVEF 30%) were included. During 42 ± 13 months follow-up, appropriate ICD therapy occurred in 28 patients (44%), and 14 patients (22%) died. Number of transmural scar segments (P = 0.005) and percentage LV scar (P = 0.03) were both significantly associated with appropriate ICD therapy. However, neither number of transmural scar segments (P = 0.32) or percent LV scar (P = 0.59) was significantly associated with all-cause mortality. CONCLUSION: In this observational study, in medium-term follow-up, the extent of LV scar characterized by LGE-CMR was strongly associated with the occurrence of spontaneous ventricular arrhythmias but not all-cause mortality. We hypothesize that scar quantification by LGE-CMR may be more specific for SCD than non-SCD, and may prove a valuable tool for the selection of patients for ICD therapy.

Proteomic profiling to identify prognostic biomarkers in heart failure.

BACKGROUND: The ability to predict mode, as well as risk, of death in left ventricular systolic dysfunction (LVSD) is important, as the clinical and cost-effectiveness of implantable cardioverter defibrillators (ICD) therapy depends on its use in appropriately selected patient populations. The value of a proteomic approach in identifying prognostic biomarkers in LVSD is unknown. The aims of this pilot study were to use proteomic techniques to identify serum biomarkers associated with LVSD and to prospectively explore their association with prognosis. PATIENTS AND METHODS: Serum was analysed by surface-enhanced laser desorption ionisation time-of-flight mass spectrometry (SELDI-TOF MS) in patients with (n=78) and without (n=45) systolic heart failure (SHF). Spectra were compared to identify differentially expressed signal peaks as potential biomarker indicators. The ability of these peaks to predict all-cause mortality and survival with appropriate ICD therapy was then tested prospectively in patients with ICDs, on the background of LVSD (n=141). RESULTS: For the identification stage spectra (2-200 kDa) from SHF and control patients were randomly separated into two equally sized discovery and validation sets. Six protein peaks were identified that were differentially expressed in SHF in both sets. In the prospective phase, during a mean follow-up of 15±3 months, 11 patients died and 39 survived with appropriate ICD therapy. Five out of the six proteomic biomarkers predicted all-cause mortality but none predicted appropriate ICD therapy. CONCLUSION: These results provide proof-of-principle and are supportive of the SELDI proteomic approach as a high-throughput screening tool in identifying potentially prognostic protein peaks in patients with LVSD.

Multiobjective design optimisation of coronary stents.

We present here a multi-objective and multi-disciplinary coronary stent design optimization paradigm. Coronary stents are tubular, often mesh-like, structures which are deployed in diseased (stenosed) artery segments to provide a scaffolding feature that compresses atheromatus plaque, hence restoring luminal area and maintaining vessel patency. A three variable geometry parameterisation of a CYPHER (Cordis Corporation, Johnson & Johnson co.) type stent is proposed to explore the functionality of a sequence of circumferential rings connected by 'n' shaped links. The performance of each design is measured by six figures of merit (objectives/metrics) representing (i) acute recoil, (ii) tissue stresses, (iii) haemodynamic disturbance, (iv) drug delivery, (v) uniformity of drug distribution, and (vi) flexibility. These metrics are obtained from computational simulations of (i) structural deformation through balloon inflated expansion of a stent into contact with a stenosed vessel, (ii) pulsatile flow over the deformed stent embedded in the vessel wall, (iii) steady-state drug distribution into the tissue, and (iv) flexibility of a stent in response to an applied moment. Design improvement is obtained by a multi-objective surrogate modelling approach using a non-dominated sorting genetic algorithm (NSGA-II) to search for an optimal family of designs. A number of trade-offs between the different objectives are identified. In particular a conflict between pairs of the following objectives are shown -- (a) volume average stress vs. recoil, (b) volume average drug vs. volume average stress, (c) flexibility vs. volume average stress, (d) flexibility vs. haemodynamic disturbance, (e) volume average drug vs. haemodynamic disturbance, and (f) uniformity of drug vs. volume average stress. Different paradigms to choose the optimal designs from the obtained Pareto fronts are presented and under each such paradigm, the optimal designs and there relative positions with respect to a representative CYPHER stent are shown. The methodology and the results of this work could potentially be useful in further optimisation studies and development of a family of stents with increased resistance to in-stent restenosis and thrombosis.

Defining potential to benefit from implantable cardioverter defibrillator therapy: the role of biomarkers.

AIMS: Implantable cardioverter defibrillator (ICD) therapy improves survival in patients at high sudden cardiac death (SCD) risk. However, some patient groups fulfilling indications for ICD therapy may not gain significant benefit: patients whose absolute risk of SCD is low and patients whose risk of death even with an ICD is high. The value of biomarkers in identifying patients' potential for survival benefit from ICD therapy is unknown. We performed a pilot study to investigate this. METHODS AND RESULTS: Five established cardiovascular biomarkers were measured in patients with ICDs on the background of left ventricular dysfunction: N-terminal pro-brain natriuretic peptide [NT-proBNP], soluble ST2 [sST2], growth differentiation factor-15, C-reactive protein, and interleukin-6. The endpoints were all-cause mortality and survival with appropriate ICD therapy. One hundred and fifty-six patients were enrolled (age 69 years [Q1-Q3 62-77], 85% male, 76% ischaemic aetiology). During a follow-up of 15 ± 3 months, 12 patients died and 43 survived with appropriate ICD therapy. In a Cox proportional hazards model, the strongest predictors of death were Log sST2 (P< 0.001), serum creatinine (P< 0.001), and Log NT-proBNP (P= 0.002). The strongest predictor of survival with appropriate ICD therapy was Log NT-proBNP (P= 0.01). CONCLUSION: The biomarkers NT-proBNP and sST2 are promising biomarkers for identifying patients with little potential to gain significant survival benefit from ICD therapy. However, their incremental benefit, in addition to currently available clinical risk prediction models, remains unclear. These results demand a confirmatory prospective cohort study, designed and powered to derive and validate prediction algorithms incorporating these markers.

The extent of left ventricular scar quantified by late gadolinium enhancement MRI is associated with spontaneous ventricular arrhythmias in patients with coronary artery disease and implantable cardioverter-defibrillators.

BACKGROUND: Characterization of sudden cardiac death (SCD) risk remains a challenge in the application of implantable cardioverter-defibrillator (ICD) therapy. Late gadolinium enhancement cardiac MRI (LGE-CMR) can accurately identify myocardial scar. We performed a retrospective, single-center observational study to evaluate the association between the extent and distribution of left ventricular scar, quantified using LGE-CMR, and the burden of ventricular arrhythmias in patients with coronary artery disease and ICDs. METHODS AND RESULTS: All patients included (2006 to 2009) had undergone LGE-CMR before ICD implantation. Scar (defined as myocardium with a signal intensity ≥50% of the maximum in scar tissue) was characterized in terms of percent scar, scar surface area, and number of transmural left ventricular scar segments. The end point was appropriate ICD therapy. Sixty-four patients (mean age, 66±11 years; male sex, 51) were included. During 19±10 months follow-up, appropriate ICD therapy occurred in 19 (30%) patients. In Cox regression analyses, both percent scar (hazard ratio per 10%, 1.75; 95% CI, 1.09 to 2.81; P=0.02) and number of transmural scar segments (hazard ratio per segment, 1.40; 95% CI, 1.15 to 1.70; P=0.001) were significantly associated with the occurrence of appropriate ICD therapy. CONCLUSIONS: In this pilot study, the extent of myocardial scar characterized by LGE-CMR was significantly associated with the occurrence of spontaneous ventricular arrhythmias. We hypothesize that scar quantification by LGE-CMR may prove a valuable risk stratification tool for the occurrence of ventricular arrhythmias, which may have implications for patient selection for ICD therapy.

Combining carotid intima-media thickness with carotid plaque on screening for coronary heart disease.

BACKGROUND: Ultrasound-detected carotid artery intima-media thickness (IMT) and carotid plaque are possible screening tests for coronary heart disease (CHD) among asymptomatic individuals. OBJECTIVE: To assess the increase in screening performance of combining carotid IMT and plaque compared with each measurement alone in the identification of individuals with CHD. METHODS: Ultrasound examination of left and right carotid arteries was performed on 100 individuals (median age 57), 55 with a history of CHD (unstable angina or myocardial infarction) and 45 without. IMT measurements were taken from the common carotid artery and plaque was identified above, at and below the carotid bifurcation. Associations between IMT and plaque were determined using logistic regression, and screening performance was assessed from the distributions of IMT and plaque among cases and controls. RESULTS: At a false-positive rate of 5%, IMT (cut-off >0.75 mm) identified 30% (95% CI 14-58) of affected individuals. There was an increase in the detection rate of 8 percentage points (1-33%) using IMT and plaque combined compared with IMT alone. As the false-positive increased, the difference in the detection rate increased, up to a maximum of 20 percentage points (5-38%) at a false-positive rate of 20%. The comparison of IMT and plaque combined with plaque alone could only be estimated for the false-positive rate observed using plaque alone (18%); at this point the detection rate was 72% for plaque and 75% for plaque and IMT combined, an increase of 3 percentage points (0-4%). CONCLUSION: In screening for CHD, combining carotid IMT measurement with plaque assessment is better than using either measurement alone, but the improvement in discrimination is not sufficient to make carotid ultrasound screening for CHD worthwhile.

Effects of clopidogrel on "aspirin specific" pathways of platelet inhibition.

The most widely accepted methods of assessing response to clopidogrel involve isolated ADP-induced platelet aggregation. Whilst poor response determined by these assays correlates with adverse clinical events, the number of "poor responders" is far higher than the number of events attributed to treatment failure. Clopidogrel may have effects that cannot be assessed using isolated ADP-induced aggregation. We have investigated the effect of clopidogrel on Arachidonic Acid (AA) induced platelet activation-an "aspirin specific" pathway using a novel near patient assay. Thirty four volunteers on no medication and 36 patients, on maintenance therapy with aspirin 75 mg daily, were recruited. Blood tests for Thrombelastogram PlateletMapping were taken immediately prior to and 6 hours after administration of a 600 mg clopidogrel loading dose. Changes in the area under the response curve at 15 minutes (AUC15) with both ADP- and AA-stimulation were calculated as were the corresponding percentage platelet and percentage clotting inhibition (%PIn and %CIn). There were predictable and significant changes in the AUC15 of the ADP channel in response to clopidogrel and the corresponding %PIn and %CIn in both volunteers and patients. There were also significant reductions in the AUC15 of the AA channel (presented as Mean +/- 95%CI), by 27.2 +/- 11.8%, p = 0.005 in volunteers and 35.0 +/- 8.2%, p < 0.001 in patients) and increases in the %PIn and %CIn calculated using the AA channel in volunteers (by 20.0 +/- 11.4%, p + 0.02 and 32.3 +/- 12.8%, p < 0.001 respectively) and patients (by 24.2 +/- 8.6%, p < 0.001 and by 18.0 +/- 8.6, p < 0.001 respectively). Clopidogrel has both independent and aspirin-synergistic effects on AA-induced platelet activation suggesting potentiation of the antiplatelet activity of aspirin. This effect may be clinically important and is not detected by current "gold standard" methods of assessing response to clopidogrel.

Atorvastatin restores endothelial function in offspring of protein-restricted rats in a cholesterol-independent manner.

Maternal protein restriction in rats leads to endothelial dysfunction and decreased NO bioavailability in the offspring. Statins (3-hydroxy-3-methylglutaryl-coenzyme A reductase inhibitors) are recognized to have pleiotropic actions including increasing NO bioavailability and reducing inflammation and oxidative damage. This study assessed statin treatment on vascular function in a model of endothelial dysfunction, which is independent of dyslipidemia. Wistar rats were fed a control (18% casein) or protein-restricted (9% casein) diet throughout pregnancy. At weaning, a subset of the protein-restricted group was given atorvastatin (10 mg/kg per day) in the drinking water. At 145 days of age, offspring were euthanized by CO(2) inhalation. Plasma samples were collected for markers of inflammation, vascular reactivity of the thoracic aorta, and small mesenteric arteries were assessed on the wire myograph, and tissues were snap frozen for molecular biology analysis. Thoracic aorta endothelial-dependent vasodilatation was attenuated in the male offspring from both protein-restricted groups compared with controls (P<0.05) but was similar in females (P value not significant). Endothelial-dependent dilatation of mesenteric arteries was attenuated in male and female protein-restricted offspring (P<0.05) and was corrected by atorvastatin. Maternal protein restriction increased plasma inflammatory markers granulocyte chemotactic protein, lipocalin-2, and beta(2)-microglobulin in male and C-reactive protein in female offspring (P<0.05). Atorvastatin had no effect on inflammatory markers in the males but restored C-reactive protein to control levels in the females (P<0.05). Aortic and mesenteric artery mRNA levels of endothelial NO synthase, superoxide dismutase 1, and tumor necrosis factor-alpha were unchanged. These data suggest that atorvastatin can restore endothelial function in this model, but its effects are gender specific and dependent on the vascular bed.

Point-of-care platelet function assays demonstrate reduced responsiveness to clopidogrel, but not aspirin, in patients with Drug-Eluting Stent Thrombosis whilst on dual antiplatelet therapy.

BACKGROUND: To test the hypothesis that point-of-care assays of platelet reactivity would demonstrate reduced response to antiplatelet therapy in patients who experienced Drug Eluting Stent (DES) ST whilst on dual antiplatelet therapy compared to matched DES controls. Whilst the aetiology of stent thrombosis (ST) is multifactorial there is increasing evidence from laboratory-based assays that hyporesponsiveness to antiplatelet therapy is a factor in some cases. METHODS: From 3004 PCI patients, seven survivors of DES ST whilst on dual antiplatelet therapy were identified and each matched with two patients without ST. Analysis was performed using (a) short Thrombelastogram PlateletMappingtrade mark (TEG) and (b) VerifyNow Aspirin and P2Y12 assays. TEG analysis was performed using the Area Under the Curve at 15 minutes (AUC15) as previously described. RESULTS: There were no differences in responses to aspirin. There was significantly greater platelet reactivity on clopidogrel in the ST group using the Accumetrics P2Y12 assay (183 +/- 51 vs. 108 +/- 31, p = 0.02) and a trend towards greater reactivity using TEG AUC15 (910 +/- 328 vs. 618 +/- 129, p = 0.07). 57% of the ST group by TEG and 43% of the ST cases by Accumetrics PRU had results > two standard deviations above the expected mean in the control group. CONCLUSION: This study demonstrates reduced platelet response to clopidogrel in some patients with DES ST compared to matched controls. The availability of point-of-care assays that can detect these responses raises the possibility of prospectively identifying DES patients at risk of ST and manipulating their subsequent risk.

Long-term continuation on cardiovascular drug treatment in patients with coronary heart disease.

BACKGROUND: Combination therapy to reduce risk factors is effective in preventing recurrent cardiovascular disease events in patients with coronary heart disease (CHD), but medications need to be continued indefinitely to maximize the benefits. OBJECTIVE: To evaluate the extent of long-term continuation with cardiovascular drug therapy and its expected impact on the prevention of CHD. METHODS: We studied 242 patients with CHD who underwent percutaneous coronary intervention following an acute coronary syndrome over a 6 month period in 2004. We prospectively examined the extent to which specific drugs and drug combinations were continued over time by reviewing medication use at the time of hospital discharge and after 2 years. The results were used to estimate the expected loss in preventive efficacy due to discontinuation of therapy. RESULTS: The changes over a 2 year period in the proportions of patients taking each drug class were as follows: 15% reduction for aspirin (95% CI, -21 to -9), 10% reduction for statins (95% CI, -16 to -5), 19% reduction for angiotensin-converting enzyme inhibitors (95% CI, -26 to -12), 12% reduction for beta-blockers (95% CI, -18 to -6), 0% increase for calcium-channel blockers (95% CI, -5 to 6), 2% increase for thiazides (95% CI, -2 to 6), and 12% increase for angiotensin-II receptor blockers (95% CI, 6 to 18). The combination of aspirin, statin, and at least 2 blood pressure lowering drugs was prescribed to 81% of patients, three-quarters of whom remained on this combination after 2 years. The overall expected preventive effect on CHD of the combined medication taken during hospitalization and after 2 years was 80% and 74%, respectively. CONCLUSIONS: In patients with CHD, long-term continuation of combination cardiovascular drug therapy is considerably greater than generally perceived.