OUP user menu

Efficacy of ivabradine, a new selective If inhibitor, compared with atenolol in patients with chronic stable angina

Jean-Claude Tardif, Ian Ford, Michal Tendera, Martial G. Bourassa, Kim Fox
DOI: http://dx.doi.org/10.1093/eurheartj/ehi586 2529-2536 First published online: 7 October 2005


Aims Ivabradine, a new If inhibitor which acts specifically on the pacemaker activity of the sinoatrial node, is a pure heart rate lowering agent. Ivabradine has shown anti-ischaemic and anti-anginal activity in a placebo-controlled trial. The objective of this study was to compare the anti-anginal and anti-ischaemic effects of ivabradine and the beta-blocker atenolol.

Methods and results In a double-blinded trial, 939 patients with stable angina were randomized to receive ivabradine 5 mg bid for 4 weeks and then either 7.5 or 10 mg bid for 12 weeks or atenolol 50 mg od for 4 weeks and then 100 mg od for 12 weeks. Patients underwent treadmill exercise tests at randomization (M0) and after 4 (M1) and 16 (M4) weeks of therapy. Increases in total exercise duration (TED) at trough at M4 were 86.8±129.0 and 91.7±118.8 s with ivabradine 7.5 and 10 mg, respectively and 78.8±133.4 s with atenolol 100 mg. Mean differences (SE) when compared with atenolol 100 mg were 10.3 (9.4) and 15.7 (9.5) s in favour of ivabradine 7.5 and 10 mg (P<0.001 for non-inferiority). TED at M1 improved by 64.2±104.0 s with ivabradine 5 mg and by 60.0±114.4 s with atenolol 50 mg (P<0.001 for non-inferiority). Non-inferiority of ivabradine was shown at all doses and for all criteria. The number of angina attacks was decreased by two-thirds with both ivabradine and atenolol.

Conclusion Ivabradine is as effective as atenolol in patients with stable angina.

  • Angina
  • Heart rate
  • Pharmacology

See page 2482 for the editorial comment on this article (doi:10.1093/eurheartj/ehi575)


Angina results when myocardial perfusion is insufficient to meet myocardial metabolic demand. A high heart rate (HR) induces myocardial ischaemia and subsequent angina as it both increases myocardial oxygen demand and decreases myocardial perfusion, the latter by shortening the duration of diastole. Beta-blockers are effective at reducing angina largely by decreasing HR1 and are usually preferred as initial therapy in the absence of contraindications.2 Despite the demonstrated safety and effectiveness of beta-blockers, physician use and patient compliance may be somewhat limited by the side effects of these agents, including fatigue, sexual dysfunction, depression, cold extremities, light-headedness, gastrointestinal disturbances, bronchospasm, and atrioventricular (AV) block.36

If, a mixed Na+–K+ inward current activated by hyperpolarization and modulated by the autonomic nervous system, is one of the most important ionic currents for regulating pacemaker activity in the sinoatrial (SA) node. Ivabradine (Procoralan®) is a novel specific HR lowering agent that acts in SA-node cells by selectively inhibiting the pacemaker If current in a dose-dependent manner.7,8 It slows the diastolic depolarization slope of SA-node cells9 and reduces HR at rest and during exercise in animals912 and human volunteers.13 Ivabradine has demonstrated anti-ischaemic and anti-anginal activity at doses of 5 and 10 mg bid in a placebo-controlled study involving 360 patients with stable angina.14 The primary objective of this trial was to compare the effects of ivabradine and the beta-blocker atenolol on exercise capacity in patients with stable angina.


Study population

A randomized double-blinded, parallel-group, trial involving 144 centres in 21 countries was performed to compare the effects of (1) 4 weeks of 5 mg ivabradine bid vs. 50 mg atenolol od and (2) 12 additional weeks of 7.5 or 10 mg ivabradine bid vs. 100 mg atenolol od (Figure 1). Eligible patients were aged ≥18 years with (1) a history of stable effort angina for ≥3 months prior to study entry; (2) evidence of CAD manifested by ≥1 of five criteria (myocardial infarction ≥3 months before study entry, coronary angioplasty ≥6 months or bypass surgery ≥3 months before entry, coronary angiogram showing ≥1 diameter stenosis ≥50%, or scintigraphic/echocardiographic evidence of exercise-induced reversible myocardial ischaemia); (3) two positive exercise tolerance tests (ETT) prior to randomization defined as occurrence of limiting angina (moderate/severe pain ordinarily causing the patient to stop exercise during normal daytime activity) and 1 mm horizontal or downsloping ST-segment depression (measured 0.08 s after the J-point on ≥3 consecutive complexes) between 3 and 12 min of initiation; and (4) time to 1 mm ST-segment depression (TST) of the two ETTs within ±20% or ±1 min of each other.

Figure 1 Study design summary.

Exclusion criteria included significant heart disease other than CAD; known high-grade left main CAD; congestive heart failure stage III/IV NYHA; symptomatic hypotension or uncontrolled hypertension [resting systolic blood pressure (SBP) >180 mmHg or diastolic blood pressure (DBP) >100 mmHg]; atrial fibrillation/flutter, pacemaker, or implanted defibrillator; second/third degree AV block, resting HR <50 bpm or sick sinus syndrome; any condition that interferes with ability to perform or interpret exercise tests (e.g. Wolff–Parkinson–White, left bundle branch block, left ventricular hypertrophy); contraindications to atenolol; recent treatment with amiodarone (<3 months) or bepridil (<7 days); ALT >3 times normal value; serum creatinine >180 µmol/L; electrolyte disorders; thyroid disorders unless controlled by thyroxine for ≥3 months; haemoglobin <100 g/L; or history of severe psychiatric disorders. Written informed consent was obtained from all patients. The study was approved by the institutional review board of each participating hospital.

Study protocol

The study design is shown in Figure 1. Placebo washout of anti-anginal medications first lasted 2–7 days depending on the previously administered treatment (≥5 half-lives). At the end of this period, a first ETT was performed (selection visit). A 7-day single-blind placebo run-in period followed during which the patient underwent a second ETT to meet qualifying and stability criteria. Patients were then randomized using permutation blocks into one of the three treatment groups: ivabradine 5 mg bid for 4 weeks increasing to 7.5 mg bid for 12 weeks, ivabradine 5 mg bid for 4 weeks increasing to 10 mg bid for 12 weeks, or atenolol 50 mg od for 4 weeks increasing to 100 mg for 12 weeks. Capsules of ivabradine (taken twice daily), atenolol (taken in the morning), and placebo (for evening administration in the atenolol group) were identical in appearance and taste. Patients, investigators, central readers of ETT data, and the sponsor were blinded to the treatment received by the patients. After 4 weeks (M1) and 16 weeks (M4) of therapy, treadmill ETTs were performed at trough (12±1 h after previous evening administration) and at peak of drug activity (4±1 h after morning administration). At each visit, tolerability and symptoms were evaluated through questioning and diary records (for frequency of angina attacks and short-acting nitrate consumption), and patients also underwent a physical examination and resting 12-lead electrocardiogram (ECG). Samples for blood biochemistry and haematology were taken prior to randomization, at M1 and M4, and analysed by a central laboratory. The M4 visit was followed by a 2-week double blind run-out period during which atenolol was progressively decreased, whereas patients in the ivabradine groups received placebo during that period.

Concomitant treatment with drugs that could interfere with the natural course of angina (long-acting nitrates, calcium antagonists, other beta-blockers, potassium-channel openers, molsidomine, or trimetazidine) or interpretation of the ST-segment changes (anti-arrhythmic agents, digitalis, MAO inhibitors) was not allowed during the trial.

Exercise testing

Symptom-limited ETTs (modified Bruce protocol15) included a minimum of three continuously monitored ECG leads and a 12-lead ECG recorded every 30 s, with smoking and nitroglycerin prohibited for ≥2 h prior to the test. The ST segment was measured at 0.08 s after the J-point in three consecutive QRS complexes with a flat baseline. If ST-segment depression was present at rest (maximum allowed ≤0.5 mm at rest), the change was calculated from the value at rest to the value during exercise. If ST-segment elevation was present at rest, ST depression during exercise was calculated from the ECG-isoelectric line. TST was calculated as the duration to 1 mm ST depression in the case of an isoelectric or elevated ST segment at rest or the duration to a further 1 mm depression compared with the negative basal value for those patients with ST depression at rest. When 1 mm ST depression did not occur during the randomized treatment phase, TED was used as the measure of TST. Reasons for terminating exercise were limiting angina, dyspnea, or extreme fatigue. For efficacy analyses, the original ECG printouts from the ETTs were re-analysed centrally by an independent physician blinded to treatment allocation.


The primary efficacy criterion was the change in TED during ETT from inclusion (M0) to end of treatment (M4) performed at trough of drug activity. Secondary efficacy criteria included changes in time to limiting angina (TLA), time to angina onset (TAO), TST, HR, and rate pressure product (RPP), and TED and other ETT criteria at peak of drug activity; TED, TLA, TAO, and TST from M0 to M1 and angina attack frequency and short-acting nitrate consumption as recorded in patients' diaries.

Statistical analysis

Sample size was estimated to demonstrate the non-inferiority of ivabradine (for the 7.5 and 10 mg bid groups) in comparison to atenolol (100 mg od) on the changes in TED from M0 to M4 at trough of drug activity. Because no established equivalence limit existed in stable angina, it was agreed with the European Medicines Agency that the equivalence limit for TED should be set between 30 and 40 s. Committee for Proprietary Medicinal Products (CPMP)-approved independent expert committee subsequently recommended a clinically plausible equivalence standard of 35 s (i.e. change in TED with atenolol could exceed that with ivabradine by no more than 35 s), on the basis of the mean change from baseline in TED and its standard deviation (SD), calculated by independent statisticians in a blinded review of the first 203 patients. This equivalence limit represented less than one-third of the SD of the TED changes. With a one-sided test, a Type 1 error rate of 2.5% and a power of 95%, the analysis of non-inferiority was estimated to require 285 patients per group. The main analysis of outcomes was performed on the full analysis set (all randomized patients with documentation of coronary artery disease, having taken at least one dose of study medication and who had an evaluation of the primary efficacy criterion during the randomized therapy period).

To ensure the robustness of the main analysis conclusion, an additional analysis was performed using the per-protocol population, defined as randomized patients with documentation of the studied disease, with an evaluation of the main study criterion (TED) at baseline and 4 months, without deviation interfering with these evaluations and with sufficient exposure to study drug. Additional analyses were carried out on the randomized set. For patients without ETT post-baseline, a ‘maximal bias method’, designed to emphasize any difference between the two treatments, was used. This assumed no change in TED at M4 in the ivabradine groups (worst case imputation) and a clear increase in TED in the atenolol group (the third quartile of TED observed in this group at M4, a best case imputation).

The non-inferiority of ivabradine was tested on the primary efficacy criterion using a covariance analysis with adjustment for country factor and baseline as a covariate, taking into account the equivalence limit and according to a stepwise procedure (first comparing ivabradine 7.5 mg to atenolol 100 mg and then ivabradine 10 mg to atenolol 100 mg). The comparison of ivabradine 5 mg to atenolol 50 mg at M1 was performed independently of this stepwise procedure. The same analyses were performed on secondary efficacy criteria to which the equivalence limit applied. For other ETT criteria, number of angina attacks and short-acting nitrate consumption, the 95% confidence intervals (CI) of treatment differences were calculated. Safety analyses were performed on all patients having received ≥1 dose of study drug.


Patient characteristics

Patients (n=939) were randomized to treatment with either ivabradine 5/7.5 mg bid (315 patients) or ivabradine 5/10 mg bid (317 patients) or atenolol 50/100 mg od (307 patients), and 884 (94%) were included in the full analysis set (300, 298, and 286 patients, respectively). Baseline clinical and ETT characteristics for all randomized patients were similar in the three study groups (Table 1). A total of 121 patients withdrew early from study medications: 43 (13.7%) in the ivabradine 5/7.5 mg group, 43 (13.5%) in the ivabradine 5/10 mg group, and 35 patients (11.4%) in the atenolol group (Figure 2).

Figure 2 Patient disposition in study. The difference in number of withdrawals due to adverse events was not statistically significant.

View this table:
Table 1

Baseline characteristics of randomized population

Ivabradine 5/7.5 mgIvabradine 5/10 mgAtenolol 50/100 mg
Age, year60.8±8.561.1±8.461.6±6.6
Male, n (%)266 (84.4)275 (86.8)257 (83.7)
Angina class
 I64 (20.3)68 (21.5)62 (20.2)
 II225 (71.4)222 (70.0)215 (70.0)
 III26 (8.3)27 (8.5)30 (9.8)
Previous MI, n (%)168 (53.3)171 (53.9)167 (54.4)
Previous PCI, n (%)65 (20.6)73 (23.0)48 (15.6)
Previous CABG, n (%)60 (19.0)63 (19.9)52 (16.9)
Supine BP (mmHg) systolic135.6±16.5136.5±16.9136.3±17.3
Supine BP (mmHg) diastolic80.8±8.981.5±9.481.0±8.8
Total exercise duration (s)
 Min–max  232–1025   86–1054  163–929
Time to limiting angina (s)
 Min–max  232–980   86–1069  163–929
Time to angina onset (s)
 Min–max  122–780   75–920  120–780
Time to 1 mm ST depression (s)
 Min–max  120–840   60–840  120–900
Heart rate at rest (bpm)
 Min–max   53–140   53–137   46–120
Heart rate at peak exercise (bpm)
 Min–max   78–175   58–170   80–194
RPP at rest (bpm×mmHg)
 Mean±SD10 943±248210 683±252210 801±2418
 Min–max  6440–19 992  5850–22 040  6270–17 646
RPP at peak exercise (bpm×mmHg)
 Mean±SD21419±462121 127±462921 643±5195
 Min–max10 032–35 190 11 900–39 500 11 440–43 680

BP, blood pressure; CABG, coronary artery bypass graft; MI, myocardial infarction; PCI, percutaneous coronary intervention.


In the full analysis set, TED at trough increased from M0 to M4 by 86.8±129.0 s with ivabradine 7.5 mg bid, 91.7±118.8 s with ivabradine 10 mg bid, and 78.8±133.4 s with atenolol 100 mg od (Table 2). The estimated differences (SE) when compared with atenolol 100 mg and adjusted for baseline value and country were 10.3 (9.4) s and 15.7 (9.5) s in favour of ivabradine 7.5 and 10 mg, respectively (P<0.001 for non-inferiority, Figure 3). The adjusted estimated difference (SE) in TED (Table 3) between the treatment groups at M1 was 6.7 (7.2) s in favour of ivabradine 5 mg bid (P<0.001 for non-inferiority, Figure 3). The analyses in the per-protocol population yielded similar results. When all randomized patients were analyzed and a maximal bias hypothesis was used, the non-inferiority of ivabradine nevertheless persisted. Non-inferiority of both ivabradine groups compared with atenolol was also demonstrated for TST, TLA, and TAO (Tables 2 and 3, Figure 4). HR was consistently reduced in all treatment groups at rest and peak exercise (Table 4).

Figure 3 Effects on total exercise duration at trough of drug activity.

Figure 4 Effects on time to 1 mm ST segment depression (TST) at trough of drug activity.

View this table:
Table 2

Changes in exercise tests from M0 to M4 at trough of drug activity in the full analysis set

Ivabradine 7.5 mg bid (n=300)Ivabradine 10 mg bid (n=298)Atenolol 100 mg od (n=286)
Total exercise duration (s)
 Differencea (SE)10.3 (9.4)15.7 (9.5)
 95% CIs[−8.3; 28.8][−2.9; 34.3]
P-value, non-inferiorityP<0.001P<0.001
Time to limiting angina (s)
 Differencea (SE)9.3 (9.7)15.1 (9.7)
 95% CIs[−9.6; 28.3][−3.9; 34.0]
P-value, non-inferiorityP<0.001P<0.001
Time to angina onset (s)
 Differencea (SE)12.1 (11.5)10.1 (11.6)
 95% CIs[−10.5; 34.7][−12.5; 32.8]
P-value, non-inferiorityP<0.001P<0.001
Time to 1 mm ST depression (s)
 Differencea (SE)4.3 (10.7)−3.3 (10.8)
 95% CIs[−16.8; 25.3][−24.4; 17.8]
P-value, non-inferiorityP<0.001P=0.002

SE, standard error.

aIvabradine–atenolol, estimate adjusted for baseline and country factors.

View this table:
Table 3

Changes in exercise tests from M0 to M1 at trough of drug activity in the full analysis set

Ivabradine 5 mg bid (n=595)Atenolol 50 mg od (n=286)
Total exercise duration (s)
 Difference (SE)6.7 (7.2)
 95% CIs[−7.4; 20.8]
P-value, non-inferiorityP<0.001
Time to limiting angina (s)
 Difference (SE)5.0 (7.3)
 95% CIs[−9.3; 19.3]
P-value, non‐inferiorityP<0.001
Time to angina onset (s)
 Difference (SE)0.7 (9.4)
 95% CIs[−17.7; 19.2]
P-value, non‐inferiorityP<0.001
Time to 1 mm ST depression (s)
 Difference (SE)4.9 (8.5)
 95% CIs[−11.9; 21.7]
P-value, non‐inferiorityP<0.001
View this table:
Table 4

Changes in heart rate and rate pressure product (RPP) at trough of drug activity

Ivabradine 5 mg bida (n=594)Atenolol 50 mg oda (n=286)Ivabradine 7.5 mg bidb (n=299)Ivabradine 10 mg bidb (n=298)Atenolol 100 mg odb (n=286)
Heart rate at rest (bpm)
 Difference (SE)2.7 (0.6)2.1 (0.8)1.1 (0.8)
 95% CIs[1.4; 4.0][0.6; 3.7][−0.4; 2.7]
Heart rate at peak exercise (bpm)
 Difference (SE)3.6 (0.8)5.6 (1.0)3.6 (1.0)
 95% CIs[2.0; 5.3][3.5; 7.6][1.6; 5.6]
RPP at rest (bpm×mmHg)
 Baseline10 809±249910 758±240710 919±249410 721±249910 759±2400
 Difference (SE)643 (115)682 (135)555 (136)
 95% CIs[417; 869][417; 948][288; 821]
RPP at peak exercise (bpm×mmHg)
 Baseline21 338±462921 587±525421 435±465821 063±465321 599±5214
 Difference (SE)1534 (255)1980 (302)1466 (300)
 95% CIs[1035; 2037][1387; 2573][878; 2054]



At peak of drug activity at M4, the non-inferiority of both ivabradine groups vs. atenolol was shown for TED and all secondary criteria except TST. TED increased by 93.8±150.9 s with ivabradine 7.5 mg bid, 99.8±132.5 s with ivabradine 10 mg bid, and 108.9±33.0 s with atenolol 100 mg od (P≤0.01 for non-inferiority). TST increased by 108.4±165.8 s with ivabradine 7.5 mg bid, 95.4±133.6 s with ivabradine 10 mg bid, and 140.5±141.9 s with atenolol 100 mg od. At peak of drug activity at M1, the non-inferiority of ivabradine 5 mg compared with atenolol 50 mg was demonstrated for all ETT criteria.

The number of angina attacks per week and consumption of short-acting nitrates were reduced in all treatment groups (Table 5). There were small changes in supine blood pressure from baseline to last value on treatment (+2.3±16.5 [95% CI: 0.41; 4.12] and +0.9±17.3 [−1.00; 2.73] mmHg in SBP and −1.6±9.2 [−2.65; −0.57] and −1.8±9.5 [−2.89; −0.80] mmHg in DBP with ivabradine 7.5 and 10 mg bid, respectively, and −4.9±15.9 [−6.81; −3.02] mmHg in SBP and −3.5±9.1 [−4.54; −2.42] mmHg in DPB with atenolol 100 mg od).

View this table:
Table 5

Treatment effects on angina attacks and nitrate consumption

Ivabradine 5a (n=606)Atenolol 50a (n=293)Ivabradine 7.5b (n=307)Ivabradine 10b (n=303)Atenolol 100b (n=294)
Weekly number of angina attacks (mean±SD)
 Last post M01.6±3.41.5±5.91.0±2.31.0±2.51.0±3.3
Short-acting nitrate consumption (units/week) (mean±SD)
 Last post M01.1±3.30.7±2.10.6±2.10.8±2.60.6±2.2




Ivabradine was well-tolerated, with the most frequent adverse drug reaction being visual symptoms. Most of these were mainly phosphenes (luminous phenomena) described as increases in brightness in limited areas of the visual field. These symptoms were transient, rated as non-serious, appeared on average after 40 days of treatment, occurred under well-defined conditions, such as light variation, and did not disturb patients' activities. Only five patients withdrew because of visual symptoms (two and three in the ivabradine 7.5 and 10 mg bid groups, respectively). Sinus bradycardia was reported as an adverse event in 2.2 and 5.4% in the ivabradine 7.5 and 10 mg groups and in 4.3% with atenolol. Headache was reported in 2.6, 4.8, and 1.6% in the three groups, respectively.

The corrected QT interval (QTc) decreased by 10±40 and 9±35 ms in the ivabradine 7.5 and 10 mg groups, respectively, and by 12±38 ms with atenolol from baseline to end of treatment. QTc increased by >60 ms in one, two, and two patients in the ivabradine 7.5 and 10 mg bid and in the atenolol groups, respectively. A total of six deaths (all cardiac) occurred during the treatment or run-out periods: two in the ivabradine 7.5 mg bid group, three in the ivabradine 10 mg bid group, and one in the atenolol 100 mg od group. No rebound phenomenon was observed after ivabradine discontinuation.


In this 4 month randomized, double-blind, multicenter study, the non-inferiority of ivabradine 7.5 and 10 mg bid compared with atenolol 100 mg od was demonstrated for all exercise parameters in both the full analysis set and per-protocol population. The increase in TST by ∼1.5 min indicates that the improvement in total exercise capacity is associated with a relevant anti-ischaemic effect. When compared with baseline, HR and RPP were reduced at end of treatment, at rest, and peak of exercise in all study groups. The decrease in HR at peak exercise was greater with atenolol (14.0 bpm) than with ivabradine (8.6–10.3 bpm with 7.5–10 mg), showing that ivabradine induced a similar or greater improvement in exercise capacity than atenolol for a comparatively smaller reduction in RPP and HR. The improvement in exercise test parameters was associated with marked decreases in the number of angina attacks and short-acting nitrate consumption. At M1, the non-inferiority of ivabradine 5 mg bid vs. atenolol 50 mg od was also demonstrated. At peak of drug activity at M4, the non-inferiority of ivabradine was shown for all criteria except TST.

The doses of ivabradine 5 and 10 mg bid were known to be active in the treatment of angina in a placebo-controlled dose-ranging study.14 In the current study, the dose of 7.5 mg bid was studied because it was expected to retain most of the efficacy of the 10 mg bid dose while minimizing side effects. The efficacy of ivabradine was assessed against a positive control in agreement with CPMP guidelines, which recognize that it would be unethical to conduct a 4 months study using only placebo in some patients. The reference medication selected was atenolol, a selective beta-blocker well-established in the treatment of stable angina with a long duration of action which allows a once-daily regimen.1620 Maximal efficacy of atenolol is reached after 4–6 weeks and preserved over 1 year.21

Regulatory guidelines specify that total exercise capacity measured as TED at trough of drug activity, should be the primary efficacy criterion in anti-anginal trials such as this one (CPMP, 1996). The guidelines also specify how studies assessing the non-inferiority of a new drug against an active comparator should be conducted, and these recommendations were followed in the INITIATIVE trial. The anti-anginal and anti-ischaemic effects of ivabradine were initially established in a placebo-controlled trial involving 360 patients.14 The 5 and 10 mg bid dosages of ivabradine demonstrated superiority to placebo in the prevention of angina and ischaemia and showed the absence of rebound phenomena or tolerance. In contrast to ivabradine, the other HR reducing agent zatebradine did not provide anti-ischaemic benefit.22 One important reason for this difference is likely that ivabradine is a better If channel inhibitor,23 and its half-block concentration being 1.5 µmol/L whereas that of zatebradine is 80 µmol/L.24

The design of this study does not allow to strictly compare the safety of ivabradine and atenolol, because about two-thirds of the patients had previously received beta-blockers and were known to tolerate these drugs; patients with known intolerance or contraindications to atenolol were specifically excluded. Ivabradine was well tolerated, with transient visual symptoms being the main drug-related adverse event. These symptoms may be linked to the presence in the retina of ion channels similar to cardiac If channels and did not adversely affect the tolerability of the drug for most patients. The incidence of sinus bradycardia with ivabradine was similar to or lower than that usually reported with beta-blockers. There was a slightly higher number of deaths in the ivabradine groups (2 [0.6%] and 3 [1.0%], respectively) than in the atenolol group (1 [0.3%]) that was not statistically significant and falls within the expected mortality range for patients with chronic stable angina. Although this is likely a chance finding, a larger study in CAD patients will assess the effect of ivabradine on cardiovascular mortality.

Study limitations

One study limitation was the lack of an established placebo-controlled equivalence limit for the effect of atenolol. A clinically plausible equivalence standard of 35 s during ETT was estimated by an independent expert committee following a rule approved by the regulatory agency. At the doses employed, ivabradine tended to provide a nominally greater improvement in exercise tolerance than atenolol, and the equivalence limits were therefore not approached. No effort was made to titrate either drug to maximally tolerable doses; therefore, no inferences can be made about relative superiority of one to the other. Rather, these data indicate that, at clinically plausible doses, ivabradine is not inferior to atenolol. Although no important safety concerns were raised by this trial and, specifically, no severe or irreversible ophthalmologic problems were observed, the limited duration of observation suggests the need for longer follow-up before firm conclusions about safety in chronic use can be drawn.

In conclusion, the If inhibitor ivabradine is as effective as atenolol, a well-established reference drug for the treatment of stable angina.


This study was supported by Servier, France.

Conflict of interest: none declared.


View Abstract