Methods

The ACTION trial was a multicenter, randomized, placebo-controlled, double-blind trial comparing the effects of long-acting nifedipine versus placebo in patients with stable angina. A detailed description of the trial has been previously published. In brief, 7,665 patients who were ≥35 years, had angina pectoris that had been stable for ≥1 month, and required pharmacotherapy to treat or prevent angina were enrolled in the study. The patients had a history of myocardial infarction, angiographic coronary disease, a positive exercise test, or a myocardial perfusion defect on nuclear scintigraphy. The EF at baseline had to be ≥40% because of previous concerns of the safety of calcium channel blockers in patients with heart failure (HF). The exclusion criteria were overt HF, a major cardiovascular event or intervention in the previous 3 months, planned coronary angiography or intervention, clinically significant valvular or pulmonary disease, unstable insulin-dependent diabetes, systolic blood pressure of ≥200 mm Hg, diastolic blood pressure of ≥105 mm Hg, and renal dysfunction, as defined by creatinine more than twice the upper limit of normal.

The patients were seen in an outpatient clinic every 6 months and followed up for a median of 5.0 years. The outcomes assessed included death from any cause, cardiac death, and new overt HF. HF was diagnosed when new or worsening symptoms necessitated treatment and admission to the hospital without a noncardiac cause identified.

Baseline transthoracic 2-dimensional echocardiograms were obtained with simultaneous electrocardiographic lead II for all patients. The recorded images were sent to a central echocardiographic core laboratory for quantitative analysis at SOCAR (Nyon, Switzerland). A single ultrasonographer, who was unaware of the patient characteristics and treatment allocation, analyzed all echocardiograms using commercially available software. The LV endocardial boundaries were traced manually and then digitized at end-diastole and end-systole from paired biplane images (apical 4-chamber and apical 2-chamber views). The core laboratory supervisor independently analyzed and validated a subset of these echocardiographic studies, demonstrating low interobserver variability of <10%. The intraobserver variability rates were <7%.

The end-systolic volume (ESV) and end-diastolic volume (EDV) were calculated using the methods of disks, as recommended by the American Society of Echocardiography. The ESV and EDV indexed to the body surface area were then calculated, according to the body surface area as determined using Mosteller’s formula [√(height in meters × weight in kilograms)/3,600 U m ² ). All measurements were obtained as the median value of 3 different cardiac cycles. The LVEF was calculated as 100% × (EDV − ESV)/EDV.

The baseline characteristics were assessed according to gender. The patients were stratified by gender, EF, ESV, and EDV. The echocardiographic parameters were divided into the following strata for EF: EF <45%, EF ≥45% and <55%, and EF ≥55%. These strata were selected before any analyses according to (1) clinically accepted cutpoints for defining normal, mild, and moderate decreases in LV function, and (2) our objective of defining the change in hazard with EF decrease relative to normal cardiac function. Post hoc sensitivity analyses were also performed in which EF was dichotomized by ≥45% and <45%, and a similar trend in the data were observed. The indexed EDV was stratified by ≤83 and >83 ml/m ² and ESV by ≤43 and >43 ml/m ² . These values were selected because they represented the threshold at which the mortality rate curves diverged as a function of volume.

The following clinical outcomes were assessed in individual models: all-cause death, cardiac death, and incident HF. The event rates were calculated by dividing the number of patients with the event by the total patient-years until the event or the end of follow-up. Cox models were fit to assess the effect of the echocardiographic predictors and gender on the interval to the incident adverse outcome with inclusion of the following terms: gender, EF, and an interaction term between gender and EF. Similar models were constructed for the dichotomized variables EDV and ESV indexed to the body surface area. The echocardiographic parameters were treated as ordinal variables with one degree of freedom in each of these models. In the multivariable models, adjustments were made for age, diabetes, hypertension, tobacco use, creatinine (per 0.1 mg/dl >1.15 mg/dl), daily angina, history of stroke, and treatment group (nifedipine vs placebo). These covariates were chosen because they were predictive of an increased risk of mortality, and they were likely to be relevant to the relation between the echocardiographic parameters of function and size and gender. Additional iterations of the multivariate models were performed, adding the following variables to the presented model: previous myocardial infarction, lipid-lowering therapy, number of antianginal drugs (eg, nitrates, calcium channel blockers, β blockers), >3-vessel disease, β blockers, and angiotensin-converting enzyme-inhibitor therapy. These covariates had no significant effect on the risk estimates or interaction p values; thus, a more parsimonious model was presented. The interaction terms was assessed using likelihood ratio tests, with p <0.10 as the threshold for significance. All data analyses were performed using Statistical Analysis Systems, version 9.1 (SAS Institute, Cary, North Carolina).

Results

The median follow-up period for the 7,665 ACTION patients was 5.0 years or 37,867 patient-years and was 97.3% complete. Of the 7,665 patients, 649 were excluded from the echocardiography analysis for the following reasons: 116 because the EF was determined by an alternative imaging modality (nuclear cardiology); 491 because of technically limited echocardiograms that precluded analysis by the core laboratory; and 42 because of baseline echocardiogram. Of the 7,016 patients included in the present analysis, 1,424 (21%) were women and 5,592 (79%) were men, with a mean age of 63.5 years; 98% were white. The demographics, clinical history, and baseline echocardiographic parameters by gender are listed in Table 1 .

Of the 7,016 patients, 534 died, of whom 313 died from cardiovascular events. Of the 534 patients who died, 434 were men and 100 were women. An EF <45% was associated with an increased risk of death (hazard ratio [HR] 1.9, 95% confidence interval [CI] 1.6 to 2.2) and HF (HR 2.4, 95% CI 1.8 to 3.3). An ESV >43 ml/m ² was associated with an increased risk of death (HR 1.7, 95% CI 1.4 to 2.0) and HF (HR 2.0, 95% CI 1.5 to 2.7). Also, an EDV >83 ml/m ² was associated with an increased risk of death (HR 1.5, 95% CI 1.3 to 1.8) and HF (HR 1.6, 95% CI 1.2 to 2.2).

The mean quantitative EF was 48 ± 6.4% (SD) in men and 49 ± 6.4% in women. A total of 2,082 subjects (1,705 men and 377 women) had an EF <45%. The increased mortality risk associated with systolic dysfunction differed significantly by gender (interaction p = 0.03), such that women with an EF <45% had an increased risk of death compared to men, as demonstrated in the Kaplan-Meier plots ( Figure 1 ). In the Cox models, women with an EF <45% had a HR of 4.78 (95% CI 2.16 to 10.57) for all-cause death and 22.44 (95% CI 3.07 to 164.27) for cardiovascular death compared to women with an EF of ≥55% ( Table 2 ). In contrast, men with an EF <45% had a HR of 1.99 (95% CI 1.42 to 2.79) for all-cause death and 2.77 (95% CI 1.73 to 4.43) for cardiovascular death compared to men with normal cardiac function. Similarly, women with mildly depressed ventricular function, as defined by an EF of 45% to 55%, had a HR of 2.09 (95% CI 0.94 to 4.63) and 7.84 (95% CI 1.06 to 57.98) for all-cause death and cardiovascular death, respectively. Men with an EF of 45% to 55% had a comparatively lower relative risk. In the multivariate models, after adjustment for age, diabetes, hypertension, tobacco use, creatinine, daily angina, and a history of stroke, the interaction p value retained significance at our threshold p <0.10 (adjusted p = 0.07 for all-cause death and adjusted p = 0.06 for cardiac death). Thus, women with chronic stable angina and moderate cardiac dysfunction were at a significantly greater risk of adverse events. No statistically significant evidence was found for effect modification by gender on the relation between EF and HF.

Kaplan-Meier estimates of all-cause mortality by ejection fraction according to female (A) and male (B) gender.

Table 2

Event rates and hazard ratios (HRs) for clinical outcomes by gender and ejection fraction (EF)

Outcome	Event Rate (per 100 person-years)			HR, EF <45% vs EF ≥55% (95% CI)
	EF <45%	EF ≥45% and <55%	EF ≥55%
All-cause death
Men	2.21	1.32	1.10	1.99 (1.42–2.79) ⁎
Women	2.63	1.15	0.54	4.78 (2.16–10.57)
Cardiac death
Men	1.50	0.70	0.54	2.77 (1.73–4.43) ⁎
Women	1.75	0.61	0.08	22.44 (3.07–164.27)
Heart failure
Men	0.89	0.41	0.35	2.52 (1.39–4.54)
Women	1.18	0.39	0.16	7.62 (1.79–32.51)

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

1. Left Ventricular Function and C-Reactive Protein Levels in Acute Myocardial Infarction
2. Is Accurate Intravascular Ultrasound Evaluation of the Left Circumflex Ostium from a Left Anterior Descending to Left Main Pullback Possible?
3. Risk Factors Predictive of Right Ventricular Failure After Left Ventricular Assist Device Implantation
4. Value of Color Doppler Three-Dimensional Transesophageal Echocardiography in the Percutaneous Closure of Mitral Prosthesis Paravalvular Leak

Stay updated, free articles. Join our Telegram channel

Join