Blood Pressure Response to the First 36 Hours of Heart Failure Therapy with Perindopril Versus Captopril

An open randomized hospital study conducted in 169 centers in France compared the blood pressure response to the first 36 h of treatment with perindopril (PER), 2 mg once daily, with that to captopril (CAP), 6.25 mg t.i.d., in 725 patients (mean age, 70 years; men, 67%) with echocardiographic left ventricular systolic dysfunction (fractional shortening, ≤28%) due to ischemia (56.7%) or hypertension (34.5%) and a systolic blood pressure (SBP) ≥120 mm Hg. Each dose of CAP induced a sharp and rapid decrease in blood pressure (maximum, 1.5-2 h); with PER, the decrease was gradual (maximum, 6 h) and variation was less marked. However, at 36 h, the decrease in blood pressure versus baseline was similar on both treatments. Over the 36-h period, there were 22 (3%) dropouts due to marked orthostatic hypotension (SBP, <90 mm Hg), and they were fewer with PER than with CAP: 16 cases in the CAP group versus six in the PER group (p = 0.036). At 36 h, heart rate was lower with CAP than with PER: 75.2 versus 77.5 beats/min, respectively (p = 0.039). As initial therapy for stabilized left ventricular systolic dysfunction, the first dose of PER (2 mg) induced a significantly smaller decrease in blood pressure than the first dose of CAP (6.25 mg); dropouts due to orthostatic hypotension were also significantly fewer with PER than with CAP.

Angiotensin-converting enzyme inhibitors (ACEIs) are an effective treatment for heart failure ⁽¹⁾: they relieve symptoms, improve hemodynamics, delay disease progression, and reduce mortality. However, the initiation of ACEI therapy may be marked by severe hypotension ⁽²⁾, with adverse consequences ⁽³⁾. Not all ACEIs induce the same decrease in blood pressure (BP) with the first dose. In a study in heart failure patients who had stopped all diuretic therapy for 48 h, the decrease in BP with perindopril (an ACEI with a long half-life: the mean half-life of perindoprilate is 30 h) was slight (similar to that observed, contrary to all expectations, with placebo) and much smaller than that induced either by captopril (a reference ACEI with a short half-life of 2 h) or enalapril (another ACEI with a long half-life) ⁽⁴⁾.

In light of these differences in first-dose response to ACEIs, we compared the changes in BP observed in daily hospital practice after initiating heart-failure therapy with perindopril versus captopril in patients with echocardiographic left ventricular systolic dysfunction.

MATERIALS AND METHODS

Study design

After approval by Ethics Committee (Saint-Germain-en-Laye Hospital), this randomized study was conducted in 169 French general hospital cardiology departments on an open prospective multicenter basis over a 36-h period in 725 patients randomized into two parallel groups to receive, according to the approved prescribing information and standard cardiology department practice (Comité de Consultation de Protection des Personnes dans la Recherche Biomédicale), either a single dose of perindopril, 2 mg, at 0800 (n = 357), or captopril, 6.25 mg, 3 times daily at hours 0800, 1200, and 1800 (n = 368).

All concomitant treatments were permitted except other ACEIs. The dosage of any other vasodilator used had to remain unchanged for 48 h on either side of inclusion.

Inclusion criteria

Male and female patients were aged >18 years; hospitalized for an episode of heart failure or acute coronary event; in heart failure with stable left ventricular dysfunction (echocardiographic evidence within the previous 3 months of overall impairment of left ventricular kinetics and/or fractional shortening ≤28% and stabilized), meaning in practice that antihypertensive, vasodilator, and/or diuretic therapy had not been changed in the 72 h before inclusion; supine systolic BP (SBP) ≥120 mm Hg, without orthostatic hypotension (defined as an SBP <100 mm Hg after standing for 2 min); no indication for invasive investigations or exercise testing in the first 48 h of ACEI therapy; and informed written consent.

Exclusion criteria

Patients must not have aortic stenosis (mean continuous Doppler gradient, ≥20 mm Hg); mitral stenosis; known allergy to ACEIs; serum potassium >5.0 mM; serum creatinine >200 μM; pregnancy, breastfeeding, or absence of effective contraception in women of childbearing age; anticipated noncompliance on clinical or psychological grounds; any serious disease associated with left ventricular dysfunction; unstable angina, or myocardial infarction within the previous 72 h; treatment with an ACEI <5 half-lives of the ACEI in question before inclusion; and indication for an interventional procedure or surgery in the next 72 h.

Outcome measures

Changes in BP over a 36-h period were measured automatically (ambulatory BP monitoring) or semiautomatically (Dinamap) with the heart rate at the following intervals: every 30 min from 0800 to 1400; hourly from 1400 to 2000; and every 2 h from 2000 to 0800. End of study was defined as the last measure taken at 2000 on day 2. BP variation was evaluated from the curves, trough values, and peak - trough differences in systolic, diastolic, and mean BP. Tolerability was measured by dropouts due to disabling effect or severe orthostatic hypotension (SBP, <90 mm Hg).

Statistical analysis

Student's two-sided t test for unpaired samples was used for quantitative variables and the χ² test for qualitative variables, together with analysis of variance.

RESULTS

Population

The 725 patients (mean age, 69.9 years; 66.9% men) were a representative cross section of the patients usually seen by the 169 participating cardiology departments (Table 1): left ventricular dysfunction was the most frequent reason for hospitalization (50.1%), followed by congestive heart failure (31.9%), arrhythmia (30%), and postmyocardial infarction (17.4%). Heart failure was ischemic in origin in 56.7% of cases, and hypertensive in 34.5%; it was secondary to cardiomyopathy in 26.3%, valvular disease in 22.9%, and other causes in 15.8%.

Table 1

The perindopril and captopril groups were similar in age, sex, and etiology and severity of heart failure (New York Heart Association criteria), degree of left ventricular systolic dysfunction (Table 2) (mean fractional shortening on echocardiography was 20.3 and 20.1%, respectively), and particularly in baseline supine and standing systolic and diastolic BP (mercury sphygmomanometer) and heart rate.

Table 2

The perindopril group differed from the captopril group in the following (Table 2): lower rate of hospitalization for left ventricular heart failure (46.3 vs. 53.8%, respectively; p = 0.045); lower rate of history of myocardial infarction (53.3 vs. 65.1%, respectively; p = 0.036); higher rate of repolarization disturbances (69.5 vs. 59.6%, respectively; p = 0.008); and a higher rate of altered segmental kinetics on echocardiography (56 vs. 46.3%, respectively; p = 0.016).

Primary outcome: Blood pressure changes over 36 h

The profiles of the systolic, diastolic, and mean BP curves over the 36-h period were identical (Fig. 1). Diurnal BP in both groups showed three increases (at 0800, 1200, and 1800) immediately before dosing with either drug. These increases in pressure were reproducible in that they recurred identically on day 2, most probably expressing the alarm felt by resting patients at the conjunction of three events: BP measurement with the sphygmomanometer, testing for orthostatic hypotension, and mealtimes.

Fig. 1

With captopril, intake of each dose of 6.25 mg induced a sharp and rapid decrease in BP with a maximal reduction at 1.5-2 h. With perindopril, on the other hand, BP reduction was gradual, with a maximal effect at 6 h; the alarm-induced surges in BP were consistently lower than those with captopril.

Although for the 4 h after the first dose, captopril induced a significantly greater reduction in BP than did perindopril (Fig. 2), at 36 h, the reduction in BP versus baseline was similar in both groups (i.e., the two ACEIs had equivalent effects on afterload).

Fig. 2

Secondary outcomes: Dropouts due to orthostatic hypotension

Although the study population was theoretically at low risk of orthostatic hypotension, in that patients were stable with an SBP ≥120 mm Hg, and the study lasted only 36 h, there were 22 dropouts (3% of the total population) due to marked orthostatic hypotension (SBP <90 mm Hg). Dropouts for this reason were fewer with perindopril than with captopril (n = 6 vs. n = 16; p = 0.04), whereas dropouts for other reasons did not differ between the two groups (Table 3). The groups also differed in the change in mean BP over the 36-h period, which was lower with perindopril [41.4 vs. 44 mm Hg with captopril (p = 0.019)].

Table 3

Heart rate

Basal minimal and maximal heart rates did not differ between the two groups. The only difference was in the end-of-study heart rate: 75.2 beats/min with perindopril versus 77.5 beats/min with captopril (p = 0.039; Table 4).

Table 4

DISCUSSION

ACEI therapy for heart failure can raise problems in daily practice because of the very marked decrease in BP that may follow the initiation of treatment, irrespective of dose; Cleland et al. ⁽²⁾ described severe hypotension after the first dose of enalapril, 5 or 10 mg. The features of this first-dose hypotension, although not formally defined, are decreases in systolic and diastolic BP of >20 and >10 mm Hg, respectively. Its incidence, whether symptomatic or nonsymptomatic, is variously reported as 2-30%, and it is difficult to be more precise. Thus in the SOLVD study ⁽⁵⁾, which included patients with symptomatic and nonsymptomatic left ventricular systolic dysfunction, patients were randomized only if the acceptability of enalapril, 2.5 mg daily for 1 week, was considered satisfactory; the results thus underestimated the true incidence of the adverse events to be found in a nonselected randomized population; even so, persistent hypotension was still the most frequent adverse event (14.8 vs. 7% with placebo), leading to dropout in 5.4% of cases (vs. 2.8% with placebo) ⁽⁶⁾. The two problems raised by first-dose hypotension are its prognostic significance and its mechanism.

Prognostic significance of first-dose hypotension

Theoretically, hypotension is harmful in that it decreases the perfusion of essential organs, in particular the brain. Orthostatic hypotension can cause syncope and falls, which are particularly dangerous in the elderly. Persistent hypotension is also potentially harmful, as it can lead to noncompliance, depriving patients of the long-term benefits of treatment.

In practice, to judge from the major clinical trials with ACEIs in postinfarction with or without left ventricular systolic dysfunction, an initial decrease in BP did not prevent the majority of patients from deriving subsequent benefit from ACE blockade ⁽⁷⁾. It should be emphasized in this regard that ISIS IV (captopril) ⁽⁸⁾, GISSI III (lisinopril) ⁽⁹⁾, the Chinese study (captopril) ⁽¹⁰⁾, and HEART (ramipril) ⁽¹¹⁾ all screened closely for first-dose hypotension: the specific purpose of intensive BP monitoring was to cut short any hypotensive reaction immediately, whether by reducing the dose, discontinuing ACEI and/or vasodilator and diuretic treatment, or even administering plasma expanders. It is thus likely that if patients with persistent hypotension did not clinically deteriorate, they were being closely monitored, and the investigators considered their BP levels acceptable. However, this was not the case in CONSENSUS II ⁽¹²⁾, in which first-dose hypotension, which was more frequent with intravenous enalapril than with placebo (25.3 vs. 9.6%), was associated with a significant increase in mortality (17% in patients experiencing first-dose hypotension vs. 9.3% in patients not experiencing first-dose hypotension) despite the ACEI dosage having been reduced in the protocol precisely because of the excessive incidence of this adverse event.

Mechanism of the first-dose effect

The precise mechanism of the first-dose effect remains unelucidated, but the factors known to be associated with first-dose hypotension in response to ACEIs in heart failure are

* Nonspecific factors (i.e., advanced age, low basal SBP, diuretic-induced hyponatremia, and hypovolemia). In reviewing the results of ACEI studies in post-myocardial infarction and/or heart failure, Latini et al. ⁽⁷⁾ found that the patients at high risk of hypotension were those with a basal SBP ≤100 mm Hg, whereas values of 100-110 mm Hg were an indication for intensive surveillance over the first 24-h period of treatment. In GISSI III ⁽⁴⁾ and ISIS IV, ⁽³⁾, advanced age was not an added risk. Our own study showed no differences with respect to any of these parameters, probably because our population was too small for such subgroup analysis.

* Specific factors to the ACEI in question, in terms of dosage but more especially of pharmacokinetics. In a study of captopril versus enalapril in 42 patients with severe (NYHA III-IV) heart failure, Packer et al. ⁽³⁾ suggested that the first-dose effects were dependent on half-life, with the longer half-life of enalapril being responsible for a more sustained effect: thus the first dose of captopril, 25 mg, was followed by a transient decrease in mean BP, compared with the delayed, and above all, more prolonged decrease after the first dose of enalapril, 10 mg. For this reason, Packer et al. recommended initiating treatment with a short-half-life ACEI, by using the lowest possible dose and after reducing the diuretic therapy. If these conclusions are applied to perindopril, a long-half-life ACEI (mean half-life of perindoprilate of 30 h), the first dose should cause a delayed and prolonged decrease in BP. As demonstrated in our study, this is not the case. In a double-blind controlled study in 48 patients with symptomatic NYHA II-IV heart failure in whom diuretics had been discontinued ≥48 h previously, MacFadyen et al. ⁽¹³⁾ compared the diurnal BP response to first-dose perindopril (2 mg), captopril (6.25 mg), and enalapril (2.5 mg) versus placebo; at similar degrees of ACE inhibition, the decrease in BP induced by perindopril was slight and, in any event, less than that induced by captopril or enalapril. In other words, despite an identical pharmacologic effect, perindopril was better tolerated. These results were confirmed in another study in which diuretic therapy was discontinued for 48 h before the study. In this study, patients were randomized to either placebo, 2 mg oral perindopril, or 2.5 mg oral enalapril. Oral enalapril produced a greater and statistically significant reduction in BP. The BP profile after a first dose was specifically studied in several other trials. Dickstein et al. ⁽¹⁴⁾ showed that lisinopril (10 mg) induced a BP decrease similar to that obtained after 10 mg of enalapril. A comparison between lisinopril and captopril ⁽¹⁵⁾ showed that both induced hypotension of similar amplitude, but more prolonged under lisinopril. Finally, Squire et al. ⁽¹⁶⁾ showed a decrease in BP significantly higher after a dose of quinapril (2.5 mg) as compared with placebo.

Together with the findings of MacFadyen et al. ⁽⁴⁾, our study provides clear evidence that on initiating ACEI therapy in heart failure, there are distinct qualitative differences between the individual drugs. The mechanism involved remains unexplained and is not related to differences in half-life.

Limitations of the study

The nonblinded design of the randomized study could be a potential limitation, but this design was chosen to study a "true" daily practice with ACEI. The differences in initial BP response to ACEIs in heart failure already were assessed in a small double-blind randomized study ⁽⁴⁾. Our goal was then not only to confirm the results of that study on a much larger population but mainly also to demonstrate that these differences were of clinical importance in the daily practice.

Despite a randomized design, it should be noted that significant differences between the two groups were observed (higher percentages of left ventricular failure and of ECG sequelae of myocardial infarction in the captopril group), which can limit the interpretation of the results. The difference in altered segmental kinetics in echocardiography should be interpreted with caution, as there was no centralized analysis of the echocardiographic data.