Left ventricular end-diastolic pressure (LVEDP) is directly related to left ventricular (LV) compliance and intravascular volume and pressure1, and one of the main consequences of LV diastolic dysfunction is increased LV filling pressures2.

In the context of coronary artery disease, elevated filling pressures can be caused by acute ischemia, an infarction scar, or myocardial hibernation. Acute ischemia results in an upward and rightward shift of the diastolic pressure-volume curve, which may be due to various factors: calcium overload in cardiomyocytes resulting in late activation and incomplete relaxation, pericardial constriction due to increased atrial and ventricular volume, intraand interventricular asynchrony and papillary muscle dysfunction3.

Elevated filling pressures are known to have a short-term prognostic impact in patients with acute coronary syndromes (ACS), as demonstrated by Forrester4 at the end of the 1970s and reconfirmed many years later5. The importance of LVEDP in the development of heart failure (HF) within 30 days of ACS has also been shown in patients with ST-elevation myocardial infarction6. On the other hand, a substudy of the Survival and Ventricular Enlargement (SAVE) trial analyzed the importance of filling pressures in late development of HF in a population of patients with left ventricular ejection fraction (LVEF) of less than 40% and a history of ACS, and concluded that LVEDP was not an independent predictor of subsequent HF7.

In view of the uncertainty surrounding the importance of LVEDP in the long term, the present study aimed to assess this parameter during the acute phase of ACS in consecutive patients from a single center treated by an invasive strategy and with no significant left ventricular systolic dysfunction, in order to determine whether LVEDP has prognostic impact on in-hospital outcomes and during one-year follow-up.

We performed a prospective, longitudinal and observational study (prospective data collection with retrospective outcome analysis) of 1459 consecutive admissions for ACS between May 10, 2004 and December 31, 2006 to a coronary care unit of a single center. Of these, 1329 patients were identified, 130 readmissions being excluded. Of this total, 786 treated by an invasive strategy during the index admission were selected, for 675 of whom LVEDP data were available. Patients with LVEF of less than 40% were excluded, leaving a total of 489.

LVEDP was assessed in the hemodynamic laboratory immediately prior to left ventriculography, measured 50 ms after the beginning of the QRS complex, usually coinciding with the R wave. The decision to perform ventriculography and LVEDP assessment was the responsibility of the hemodynamicist on duty.

Within 24 hours of the invasive study, LV systolic function was assessed by echocardiography, LVEF being calculated by the Simpson method. All patients with LVEF of less than 40% were excluded from the analysis.

In accordance with the universal definition of myocardial infarction, diagnosis was made on the basis of markers of myocardial necrosis (troponin I) above the 99th percentile, together with evidence of myocardial ischemia8. ST-elevation myocardial infarction (STEMI) was defined as anginal chest pain, together with de novo ST elevation of >1 mm in two contiguous leads. Non-ST elevation myocardial infarction (NSTEMI) was defined as anginal chest pain lasting more than five minutes, with or without electrocardiographic (ECG) alterations (ST-segment depression or T-wave inversion). A diagnosis of unstable angina was based on the presence of new-onset anginal chest pain (at least CCS class III), worsening angina or rest angina, with or without associated ECG alterations.

For patients with NSTEMI or unstable angina, the decision on whether to adopt an invasive or a conservative strategy was based on clinical data and the results of ischemia testing, as previously described9.

Standard records from the hospital stay were analyzed, including demographic, clinical, electrocardiographic, echocardiographic and laboratory data, therapy prescribed during hospital stay, occurrence and type of in-hospital complications, treatment in the hemodynamic laboratory, duration of hospital stay and medication at discharge.

Patients were followed for a year after discharge, by telephone or by consulting hospital records. The primary endpoint in the present analysis was readmission for congestive heart failure (CHF). The secondary endpoint was the combined outcome of major adverse cardiac events (MACE) (death from cardiovascular cause, nonfatal infarction, readmission for unstable angina, or unscheduled percutaneous coronary intervention) and all-cause mortality in the first year.

The population was divided into three groups based on the distribution of percentiles (P33) of LVEDP values: group A – LVEDP ≤19 mmHg (n=186); group B – LVEDP >19 and ≤27 mmHg (n=172); and group C – LVEDP >27 mmHg (n=131).

LVEDP data were available for 489 patients. Mean LVEDP was 22.8±7.8 mmHg, and around 80% presented LVEDP of >17 mmHg. Mean age was 62.4±12.4 years, and 359 patients were male.

The demographic, clinical, laboratory and electrocardiographic characteristics of the three groups are shown in Tables 1 and 2.

The groups were similar in age, gender, cardiovascular risk factors, cardiovascular history and medication prior to admission. With regard to admission diagnosis, there was a progressively stronger association between higher LVEDP and STEMI in the three groups (35.4 vs. 45.9 vs. 56.7%, p<0.01), the opposite being found with NSTEMI (44.6 vs. 38.3 vs. 28.2%, p=0.01) (Table 1).

There was a weak but significant negative correlation between LVEF and LVEDP (r=−0.22, p<0.001), although there was a significant progressive decrease in LVEF in the three groups (56.5±7.0 vs. 55.3±7.6 vs. 53.0±7.5%, p<0.01). There was also an increasing association with markers of myocardial necrosis, particularly peak troponin I (27.4±41.4 vs. 39.4±66.0 vs. 49.4±60.2 U/l, p<0.01) and CK-MB (Table 2).

With regard to hemodynamic laboratory data, an association was found between time to catheterization and LVEDP, especially in the group with non-ST elevation ACS. There was no statistically significant correlation between coronary disease severity, percutaneous intervention or surgical revascularization and LVEDP in the three groups (Table 3).

No statistically significant differences were observed between the groups in terms of pharmacological therapy within 24 hours of admission or at discharge (Table 3).

In-hospital outcome and follow-up

The percentage of individuals lost to follow-up was 4.7%, with data available for 465 patients.

In-hospital mortality was similar in the three groups, as were 30-day and one-year mortality, and the combined outcome of MACE (Table 4).

Table 4.

In-hospital outcomes and at one-year follow-up

	Group A	Group B	Group C	p
In-hospital mortality (%)	0/186 (0.0)	1/172 (0.6)	0/131 (0.0)	0.39
Hospital stay, days (mean, SD)	5.3±2.4	5.2±2.3	5.6±2.4	0.27
30-day mortality (%)	0/176 (0.0)	0/165 (0.0)	1/124 (0.8)	0.25
One-year all-cause mortality (%)	5/176 (2.8)	4/165 (2.4)	9/124 (7.3)	0.07
One-year CV mortality (%)	5/176 (2.8)	2/165 (1.2)	3/124 (2.4)	0.55
Non-fatal MI at one year (%)	7/176 (3.9)	8/165 (4.8)	5/124 (4.0)	0.91
Readmission for UA at one year (%)	6/176 (3.4)	2/165 (1.2)	2/124 (1.6)	0.35
Unscheduled PCI at one year (%)	8/176 (4.5)	11/165 (6.7)	5/124 (4.0)	0.49
MACE at one year (%)	19/176 (10.9)	21/165 (12.8)	16/124 (12.9)	0.82
Heart failure at one year (%)	1/176 (0.6)	4/165 (2.4)	7/124 (5.6)	0.02

CV: cardiovascular; MACE: major adverse cardiac event; MI: myocardial infarction; PCI: percutaneous coronary intervention; UA: unstable angina.

Regarding the primary endpoint of the analysis, survival free of readmission for CHF was higher in the group with lower LVEDP – 99.4 vs. 97.6 vs. 94.4%, log rank p=0.02 (Figure 1).

Figure 1.

Admission for congestive heart failure within a year of ACS. ACS: acute coronary syndrome.

(0.1MB).

The present study highlights the importance of LVEDP in predicting readmission for CHF within a year of ACS in a population of patients with LVEF ≥40%. Based on the Cox model proposed, a 1.97 times higher risk for the primary endpoint was found for each increase of 5 mmHg in LVEDP.

In this cohort there was a direct correlation between LVEDP values and higher levels of myocardial necrosis markers, and, in the case of non-ST elevation ACS, time to cardiac catheterization. This finding is supported by experimental data of pacing-induced ischemia, which demonstrated decreased myocardial distensibility in this context, and hence increased filling pressures10. Interestingly, as reported by other authors, LVEDP was not directly related to the anatomical extent of coronary disease7.

In our population, LVEDP had no prognostic impact on in-hospital mortality, which may be due to the small number of patients with LVEF ≥40% treated by an invasive strategy who died in hospital. In addition, of the patients referred for catheterization, 111 did not undergo LVEDP assessment. In-hospital mortality in this group was 9%. Mean LVEF in this group was 36.7±10.0 (versus 52.6±11.6%, p<0.01, in those treated by an invasive strategy without LVEDP measurement and who survived the index admission. These findings highlight the importance of LVEF in predicting in-hospital prognosis during the acute phase of ACS.

Elevated LVEDP was a common finding in the acute phase of ACS in this population with LVEF ≥40% treated by an invasive strategy. LVEDP was a significant long-term prognostic marker and an independent predictor of readmission for CHF one year after the index event.

The authors have no conflicts of interest to declare.