Effectiveness and safety of ivabradine in Chinese patients with chronic heart failure: an observational study

Abstract

Aims

A therapeutic strategy for chronic heart failure (HF) is to lower resting heart rate (HR). Ivabradine is a well‐known HR‐lowering agent, but limited prospective data exist regarding its use in Chinese patients. This study aimed to evaluate the effectiveness and safety of ivabradine in Chinese patients with chronic HF.

Methods and results

This multicentre, single‐arm, prospective, observational study enrolled Chinese patients with chronic HF. The primary outcome was change from baseline in HR at 1 and 6 months, measured by pulse counting. Effectiveness was also evaluated using laboratory tests, the Kansas City Cardiomyopathy Questionnaire (KCCQ) clinical summary score (CSS) and overall summary score (OSS), and New York Heart Association (NYHA) class. Treatment‐emergent adverse events (TEAEs) were assessed. A post hoc analysis examined the effectiveness and safety of ivabradine combined with an angiotensin receptor–neprilysin inhibitor (ARNI) or beta‐blocker. A total of 1003 patients were enrolled [mean age 54.4 ± 15.0 years, 773 male (77.1%), mean baseline HR 88.5 ± 11.3 b.p.m., mean blood pressure 115.7/74.4 ± 17.2/12.3 mmHg, mean left ventricular ejection fraction 30.9 ± 7.6%, NYHA Classes III and IV in 48.8% and 22.0% of patients, respectively]. HR decreased by a mean of 12.9 and 16.1 b.p.m. after 1 and 6 months, respectively (both P < 0.001). At Month 6, improvements in the KCCQ CSS and OSS of ≥5 points were observed in 72.1% and 74.1% of patients, respectively (both P < 0.001). Left ventricular ejection fraction increased by 12.1 ± 11.6 (P < 0.001), and 66.7% of patients showed improvement in NYHA class (P < 0.001). At Month 6, the overall proportion of patients in NYHA Classes III and IV was reduced to 13.5% and 2.1%, respectively. Serum brain natriuretic peptide (BNP) and N‐terminal pro‐BNP changed by −331.9 ng/L (−1238.6, −134.0) and −1113.8 ng/L (−2202.0, −297.2), respectively (P < 0.001). HR reductions and improvements in NYHA and KCCQ scores with ivabradine were similar with and without use of ARNIs or beta‐blockers. Of 498 TEAEs in 296 patients (29.5%), 73 TEAEs in 55 patients (5.5%) were considered related to ivabradine [most frequent sinus bradycardia (n = 7) and photopsia (n = 7)]. TEAEs were reported in a similar number of patients in ARNI and beta‐blocker subgroups (21.9–35.6%).

Conclusions

Ivabradine treatment reduced HR and improved cardiac function and health‐related quality of life in Chinese patients with chronic HF. Benefits were seen irrespective of whether or not patients were also taking ARNIs or beta‐blockers. Treatment was well tolerated with a similar profile to previous ivabradine studies.

Introduction

Heart rate (HR) is well established as an independent prognostic factor in chronic heart failure (HF). ¹ In the Systolic Heart failure treatment with the I _f inhibitor ivabradine Trial (SHIFT), patients with the highest HR (≥87 b.p.m.) were at more than two‐fold higher risk for cardiovascular death or hospital admission due to worsening HF than patients with the lowest HRs. ¹ Consequently, lowering resting HR has been targeted as a therapeutic strategy for chronic HF. ¹ , ²

Guideline‐recommended therapy for chronic HF treatment should include renin–angiotensin–aldosterone system (RAAS) blocking agents [angiotensin‐converting enzyme (ACE) inhibitors/angiotensin receptor blockers (ARBs) or angiotensin receptor–neprilysin inhibitors (ARNIs)], beta‐blockers, mineralocorticoid receptor antagonists (MRAs), and sodium–glucose cotransporter‐2 (SGLT2) inhibitors. ¹ Some challenges associated with the use of these treatments include drug intolerance or contraindications, which may contribute to suboptimal adherence, particularly in elderly populations. ¹ , ² The HR‐lowering drug ivabradine inhibits the hyperpolarization‐activated cyclic nucleotide‐gated channel, a principal regulator of the cardiac pacemaker current (I _f ) in the sinoatrial node. ³ The efficacy and safety of ivabradine in patients with chronic HF with left ventricular systolic dysfunction have been previously demonstrated in the randomized, double‐blind, placebo‐controlled SHIFT. ⁴

Although the results of SHIFT were further supported by a post hoc analysis in Chinese patients with chronic HF, ⁵ there are limited prospective studies examining the safety and efficacy of ivabradine in Chinese patients. Therefore, a prospective, multicentre, post‐marketing, real‐world surveillance study [POSt‐authorization drug Intensive surveillance monitoring sTudy of IVabradinE in patients with chronic heart failure (POSITIVE)] was conducted to investigate the effectiveness and safety of ivabradine in Chinese patients with chronic HF. Interim results for 655 patients have been published previously. ⁶ Here, we present the final results of the POSITIVE study.

Methods

Study design

The design of this multicentre, single‐arm, prospective, observational study has been described previously. ⁶ Briefly, the study enrolled patients with chronic HF from 72 sites in China between July 2016 and December 2019. This study was conducted in accordance with the ethical standards of the 1964 Declaration of Helsinki and its later amendments, and the study protocol and subsequent amendments were approved by the ethics committee of the leading site, Zhongshan Hospital, Fudan University, and all other 71 sites. The study was registered at the International Clinical Trials Registry Platform (ISRCTN 11703380). Written informed consent was collected from all participants before initiating the trial.

Eligible patients included inpatients or outpatients at the designated clinical sites who were ≥18 years of age, had a resting HR ≥ 75 b.p.m. in sinus rhythm, as measured by pulse rate counting, and had New York Heart Association (NYHA) Class II, III, or IV chronic HF with systolic dysfunction [left ventricular ejection fraction (LVEF) <50% in the last 3 months]. ⁷

Patients were excluded if they had been treated with ivabradine within 2 weeks prior to enrolment, had experienced a myocardial infarction within 1 month prior to enrolment, had a history of acute HF or cardiogenic shock, suffered from sick sinus syndrome or sinoatrial block, or had a history of third‐degree atrioventricular block or severe hypotension (<90/50 mmHg). Key inclusion and exclusion criteria are shown in Supporting Information, Table S1 .

Data were collected at the initial clinical evaluation at enrolment (baseline) and at two follow‐up visits at 1 month (M1) and at 6 months (M6) after enrolment. During baseline evaluation, demographic characteristics, clinical history, and vital signs were recorded, and a complete physical examination was performed, including assessment of HF symptoms, evaluation of cardiac function, and 12‐lead electrocardiogram. A detailed outline of the data collected at baseline is provided in Supporting Information, Table S2 . In addition, patients completed the self‐administered 23‐item Kansas City Cardiomyopathy Questionnaire (KCCQ) to monitor changes in patients' health‐related quality of life (QoL) over time. This questionnaire is scaled to generate a score between 0 and 100, in which higher scores reflect better health‐related QoL status. ⁸ After clinical evaluation of cardiac function and HF symptoms, current cardiovascular medications were considered, the appropriate treatment regimen was determined, and the patient was scheduled for a follow‐up visit.

Patients were also assessed for clinical outcomes and adverse events (AEs) at baseline and at M1 and M6. Investigators recorded information on ivabradine‐related adverse drug reactions (i.e. AEs described by the investigator as having a definite or suspected causal relationship to ivabradine), consequences of AE (no impact, disease exacerbation, or hospitalization), and how each AE was managed. Any clinically significant changes in vital signs, physical examinations, or laboratory tests seen during follow‐up were recorded as AEs.

Treatment

Ivabradine treatment was administered (according to the Chinese government National Medical Products Administration recommendations) at an initial dose of 5 or 2.5 mg twice daily (b.i.d.) based on the patient's condition or age; the lower starting dose (2.5 mg b.i.d.) is recommended in patients aged ≥75 years. Investigators could adjust the ivabradine dosage based on the patient's condition at each visit and according to the prescribing information. The ivabradine dose was maintained in patients with a resting HR of 50–60 b.p.m., increased up to a maximum of 7.5 mg b.i.d. in patients with a resting HR persistently >60 b.p.m., and reduced (or ivabradine discontinued) in patients with a resting HR that was persistently <50 b.p.m. or who developed symptoms of bradycardia. ⁹

Endpoints

The primary effectiveness endpoint was the change from baseline in HR (determined by pulse counting) at M1 and M6. Secondary effectiveness endpoints included analysis of clinical symptoms, KCCQ clinical summary score (CSS), and overall summary score (OSS); these were summarized by individual subgroups and overall. Effectiveness endpoints were analysed in the following patient subgroups: (i) inpatients at study inclusion and (ii) outpatients at study inclusion [who used ivabradine for <3 months after discharge (i.e. during the ‘vulnerable period’ of the first 3 months after discharge from an HF hospitalization ¹⁰ ) and who used ivabradine for >3 months after discharge].

A post hoc analysis investigated the effectiveness and safety of ivabradine when combined with an ARNI or beta‐blocker by analysing outcomes according to the following subgroups: (i) in combination with ARNI overall (i.e. ARNI use at any visit), use of ARNI at all visits, use of ARNI at some visits, or no use of ARNI at any visit and (ii) in combination with beta‐blockers at all visits, use of beta‐blockers at some visits, or no use of beta‐blockers at any visit.

Statistical analysis

Details of the statistical methods used are provided in Supporting Information, Table S3 . Continuous variables were described as either the mean and standard deviation (SD) or the median and inter‐quartile range (IQR). Categorical variables were summarized as an absolute number (n) and percentage (%). No missing data were included in the percentage calculation unless otherwise specified.

AEs were coded using the Medical Dictionary for Regulatory Activities (MedDRA) Version 23.0. The analysis of AEs was based on ivabradine treatment‐emergent AEs (TEAEs).

Results

A total of 1023 patients were screened for study inclusion, and 1003 patients were included in the overall full analysis set (FAS) (Figure  1 ).

Figure 1.

Flow of patients through the study. FAS, full analysis set; M1, Month 1 visit; M6, Month 6 visit.

Overall, 849 patients received ivabradine within the vulnerable period, 95 patients used the drug outside the vulnerable period, and 59 patients could not be grouped in this way due to missing data. The effectiveness population included 832 patients, and the per‐protocol set included 799 patients.

Demographics and clinical characteristics

Patient demographic and clinical characteristics are shown in Table 1 . The mean age of the FAS population was 54.4 years, with 773 males (77.1%). Mean age, sex distribution, tobacco use, weight, height, and body mass index (BMI) in inpatients and outpatients were similar to those in the overall population.

Table 1.

Baseline demographic and clinical characteristics

	Outpatient population (n = 233)	Inpatient population (n = 770)	Overall population (N = 1003)
Age (years), mean ± SD	53.3 ± 14.02	54.8 ± 15.29	54.4 ± 15.01
Male, n (%)	183 (78.5)	590 (76.6)	773 (77.1)
Height (cm)	167.5 ± 7.76	167.4 ± 8.15	167.4 ± 8.06
Weight (kg)	70.90 ± 16.084	70.65 ± 17.303	70.71 ± 17.020
BMI (kg/m2), mean ± SD	25.10 ± 4.59	25.04 ± 4.89	25.05 ± 4.83
Tobacco use, n (%)
Never	90 (38.6)	362 (47.0)	452 (45.1)
Previously	96 (41.2)	247 (32.1)	343 (34.2)
Current	47 (20.2)	161 (20.9)	208 (20.7)
Duration of HF (months), mean ± SD	21.1 ± 32.60	19.7 ± 36.35	20.0 ± 35.49
Primary cause of HF, n (%)
Dilated cardiomyopathy	108 (46.6)	362 (47.0)	470 (46.9)
Coronary heart disease	89 (38.4)	282 (36.6)	371 (37.0)
Hypertension	12 (5.2)	56 (7.3)	68 (6.8)
Valvular heart disease	7 (3.0)	19 (2.5)	26 (2.6)
Myocarditis	4 (1.7)	7 (0.9)	11 (1.1)
Other	12 (5.2)	44 (5.7)	56 (5.6)
Missing	1	0	1
Cardiovascular disease history a , n (%)
Hypertension	95 (40.8)	355 (46.1)	450 (44.9)
Diabetes	69 (29.6)	237 (30.8)	306 (30.5)
Dyslipidaemia	82 (35.2)	168 (21.8)	250 (24.9)
Myocardial infarction	64 (27.5)	154 (20.0)	218 (21.7)
Chronic kidney disease	18 (7.7)	78 (10.1)	96 (9.6)
PCI	59 (25.3)	136 (17.7)	195 (19.4)
CABG	3 (1.3)	15 (1.9)	18 (1.8)
PAD	4 (1.7)	27 (3.5)	31 (3.1)
Peripheral vascular procedures	0 (0)	1 (0.1)	1 (0.1)
Other significant comorbidities, n (%)	94 (40.3)	528 (68.6)	622 (62.0)
HR b (b.p.m.), mean ± SD	87.2 ± 10.7	88.9 ± 11.5	88.5 ± 11.3
BP (mmHg), mean ± SD
Systolic	115.4 ± 17.08	115.8 ± 17.28	115.7 ± 17.22
Diastolic	73.5 ± 11.60	74.6 ± 12.50	74.4 ± 12.30
HF symptoms, n (%)
Dyspnoea	146 (62.7)	656 (85.2)	802 (80.0)
Asthenia	153 (65.7)	618 (80.3)	771 (76.9)
NYHA functional class, n (%)
Class I	0 (0)	2 (0.3)	2 (0.2)
Class II	150 (64.4)	141 (18.3)	291 (29.0)
Class III	72 (30.9)	417 (54.2)	489 (48.8)
Class IV	11 (4.7)	210 (27.3)	221 (22.0)
LVEF (%), mean ± SD	33.26 ± 7.57	30.15 ± 7.42	30.87 ± 7.57
HF medications, n (%)
ACE inhibitors	87 (37.3)	250 (32.5)	337 (33.6)
ARBs	47 (20.2)	148 (19.2)	195 (19.4)
ARNIs	72 (30.9)	255 (33.1)	327 (32.6)
Beta‐blockers	211 (90.6)	648 (84.2)	859 (85.6)
MRAs	192 (82.4)	686 (89.1)	878 (87.5)
Loop diuretics	167 (71.7)	688 (89.4)	855 (85.2)
Digoxin	43 (18.5)	225 (29.2)	268 (26.7)
Implanted devices, n (%)
Artificial pacemaker, SC	1 (0.4)	1 (0.1)	2 (0.2)
Artificial pacemaker, DC	2 (0.9)	3 (0.4)	5 (0.5)
Cardioverter‐defibrillator	3 (1.3)	10 (1.3)	13 (1.3)
CRT‐D	5 (2.1)	17 (2.2)	22 (2.2)
CRT‐P	1 (0.4)	7 (0.9)	8 (0.8)

ACE, angiotensin‐converting enzyme; ARBs, angiotensin receptor blockers; ARNIs, angiotensin receptor–neprilysin inhibitors; BMI, body mass index; BP, blood pressure; CABG, coronary artery bypass graft; CRT‐D, cardiac resynchronization therapy defibrillator; CRT‐P, cardiac resynchronization therapy pacemaker; DC, double‐chamber; HF, heart failure; HR, heart rate; LVEF, left ventricular ejection fraction; MRAs, mineralocorticoid receptor antagonists; NYHA, New York Heart Association; PAD, peripheral artery disease; PCI, percutaneous coronary intervention; SC, single‐chamber; SD, standard deviation.

^a Patients could have more than one concurrent cardiovascular comorbidity or other cardiovascular conditions.

^b Measured by pulse counting.

The mean duration of HF (defined as the International Classification of Functioning, Disability and Health signature date − the initial HF diagnosis date + 1) was approximately 20 months in the overall population, inpatients, and outpatients. The major causes of HF were dilated cardiomyopathy and coronary heart disease and were generally similar among the groups. Concurrent cardiovascular diseases were common (Table  1 ).

In the overall FAS population, the mean HR at baseline was 88.5 b.p.m., mean systolic blood pressure (BP) was 115.7 mmHg, mean diastolic BP was 74.4 mmHg, mean LVEF was 30.9%, and most patients had NYHA Class III. However, the majority of outpatients were classified as NYHA Class II at baseline, while the majority of inpatients were classified as NYHA Class III (Table  1 ).

The most frequently used cardiovascular drug classes, with ≥50% use at baseline, were RAAS inhibitors (RAASis) (ACE inhibitor, ARB, or ARNI), aldosterone receptor antagonists, beta‐blockers, and loop diuretics (Table  1 ). Frequently used beta‐blockers and their relative target doses are shown in Supporting Information, Table S4 .

Patient demographic and clinical characteristics by subgroup are shown in Supporting Information, Table S5 . The demographics of patients in the combined ARNI subgroup were generally similar to those of patients who did not receive ARNI therapy. In the FAS, the mean BMI of patients in the subgroups with no beta‐blocker use was statistically lower than in the subgroups with beta‐blocker use at some or all visits (P < 0.001 for both); other demographic and baseline characteristics were generally similar between patients with vs. without beta‐blocker use and to those of the combined ARNI group.

Treatment regimen and cardiovascular medications

Most patients were administered with ivabradine at 2.5 mg b.i.d. or 5 mg b.i.d. at baseline (Table  2 ). At the M1 and M6 follow‐up visits, the proportion of patients receiving ivabradine 5 and 7.5 mg increased slightly, while the proportion receiving 2.5 mg decreased. Overall, the majority of patients had no change in ivabradine dose over the course of the study (Table  2 ).

Table 2.

Ivabradine dose and dose changes

	Baseline (n = 1033)	Month 1 (n = 818)	Month 6 (n = 574)
Ivabradine dose
2.5 mg	535 (53.3)	333 (40.7)	214 (37.3)
5.0 mg	462 (46.1)	431 (52.7)	312 (54.4)
7.5 mg	5 (0.5)	48 (5.9)	43 (7.5)
Other	1 (0.1) a	6 (0.7) b	5 (0.9) b
Change in ivabradine use c
No change	—	604 (73.8)	501 (87.3)
Dose increase	—	182 (22.2)	41 (7.1)
Dose decrease	—	33 (4.0)	32 (5.6)
Discontinuation	—	40 (4.9)	147 d (25.6)

Data are n (%).

^a 1.25 mg (n = 1).

^b 1.25, 3.75, or 10 mg.

^c Data are missing for one patient at Month 1.

^d Data include patients at both Month 1 and Month 6.

Heart rate

In the effectiveness population, the mean change from baseline in HR was −12.9 [95% confidence interval (CI) –13.83, −11.89] b.p.m. at M1 and −16.1 (95% CI –17.38, −14.89) b.p.m. at M6 (P < 0.001). The changes in HR from baseline at M1 and M6 were also significant for the inpatient and outpatient subgroups (Table  3 ).

Table 3.

Heart rate at baseline and during treatment with ivabradine

	Baseline		Month 1				Month 6
	N	Mean ± SD	N	Mean ± SD	LSM ± SE change from baseline	P value a	N	Mean ± SD	LSM ± SE change from baseline	P value a
Outpatient population b	201	87.2 ± 10.74	182	74.1 ± 10.28	−12.8 ± 12.33	<0.001	146	72.2 ± 10.91	−14.5 ± 14.29	<0.001
Inpatient population b	629	88.9 ± 11.45	560	76.1 ± 11.06	−12.9 ± 13.84	<0.001	400	72.0 ± 10.45	−16.7 ± 14.94	<0.001
Overall population b	830	88.5 ± 11.30	742	75.6 ± 10.90	−12.9 ± 13.48	<0.001	546	72.0 ± 10.57	−16.1 ± 14.79	<0.001

LSM, least squares mean; SD, standard deviation; SE, standard error.

^a Paired t‐test.

^b Effectiveness population (see Supporting Information, Table S3 ).

At M1 and M6, significant reductions in HR from baseline were observed in subgroups treated with ivabradine and an ARNI or beta‐blocker (P < 0.001). No significant differences were seen in HR reduction between subgroups using beta‐blockers at all visits compared with no beta‐blocker use (Supporting Information, Table S6 ).

Symptoms and quality of life

The mean KCCQ CSS and OSS increased significantly from baseline to M1 and M6 (Table 4 ; P < 0.001). Overall, 72.1% of patients had CSS improvements and 74.1% had OSS improvements of ≥5 points compared with baseline at M6 (Table  4 ).

Table 4.

Kansas City Cardiomyopathy Questionnaire (KCCQ) scores at baseline and during treatment with ivabradine

	Baseline	Month 1				Month 6
	Mean ± SD	Mean ± SD	LSM ± SE change from baseline	P value a	Increase of ≥5 points from baseline, n (%)	Mean ± SD	LSM ± SE change from baseline	P value a	Increase of ≥5 points from baseline, n (%)
Outpatient population b (n)	228	202	—	—	—	186	—	—	—
CSS	82.44 ± 17.48	88.75 ± 14.50	5.18 ± 14.30	<0.001	86 (42.6)	92.72 ± 10.71	8.76 ± 14.08	<0.001	100 (53.8)
OSS	77.49 ± 18.97	84.96 ± 15.93	6.47 ± 14.22	<0.001	88 (43.6)	89.76 ± 11.64	10.76 ± 15.28	<0.001	113 (60.8)
Inpatient population b (n)	614	202	—	—	—	490	—	—	—
CSS	66.05 ± 20.65	84.57 ± 16.57	17.31 ± 19.21	<0.001	454 (74.4)	88.08 ± 14.76	19.70 ± 19.27	<0.001	387 (79.1)
OSS	61.25 ± 21.12	79.87 ± 17.76	17.33 ± 19.82	<0.001	433 (71.0)	84.61 ± 16.03	21.20 ± 20.09	<0.001	387 (79.1)
Overall population b (n)	962	816	—	—	—	676	—	—	—
CSS	69.93 ± 21.12	85.60 ± 16.18	14.29 ± 18.85	<0.001	540 (66.5)	89.35 ± 13.92	16.68 ± 18.63	<0.001	487 (72.1)
OSS	65.10 ± 21.75	81.13 ± 17.46	14.62 ± 19.16	<0.001	521 (64.2)	86.03 ± 15.12	18.32 ± 19.45	<0.001	500 (74.1)

CSS, clinical symptom score; LSM, least squares mean; OSS, overall symptom score; SD, standard deviation; SE, standard error.

^a Paired t‐test.

^b Full analysis set population (see Supporting Information, Table S3 ).

The incidence of dyspnoea and fatigue significantly decreased at M1 and M6 compared with baseline (Supporting Information, Tables S7 and S8 ; P < 0.001).

An improvement of ≥1 NYHA class was experienced by 497 patients (58.0%) and 478 patients (66.7%) by the M1 and M6 visits, respectively (P < 0.001; Supporting Information, Table S9 ). The improvement in NYHA class was similar in the inpatient and outpatient subgroups at each follow‐up visit.

In all three subgroups of patients using ARNIs and in those not using ARNIs, mean CSS at M1 and M6 was significantly higher than those at baseline (P < 0.001; Supporting Information, Table S10 ). Mean OSS at M1 and M6 was also significantly higher than baseline in all groups (P < 0.001).

Mean CSS and OSS at M1 and M6 were significantly higher than baseline in the subgroups using beta‐blockers at all visits, some visits, and no visits (P < 0.001; Supporting Information, Table S10 ). There was no significant difference in the CSS and OSS between the subgroups with beta‐blockers at some or all visits and the subgroup with no beta‐blocker use.

Change in NYHA class showed a trend towards improvement at M1 and M6 in the ARNI use subgroups (data not shown). There was no significant difference in NYHA class improvement from baseline between the subgroups with beta‐blockers at some or all visits and the subgroup with no beta‐blocker use.

A patient's perspective is presented in Box 1.

Box 1. Patient perspective.

This patient has been diagnosed with chronic HF for 3 years.

Underlying disease: hypertension for 20 years.

Enrolled in POSITIVE study in 2018.

Main treatment for chronic HF: ARNI, bisoprolol, spironolactone, and ivabradine.

‘I have been diagnosed with chronic heart failure for 3 years. I was feeling unwell for so long and even have trouble in my daily routine. At the worst of my disease, I was reluctant to take a walk outside, even a few steps, because I could hardly breathe when I did that. Moreover, I suffered from chest tightness and heart racing frequently. These got even worse under swelter or severe cold. In addition, I was unable to sleep well and needed sleeping pills to solve this problem.

During admission, my treating physician optimized the anti‐HF therapy based on my condition. Then I started to adhere to take medications and follow up regularly after I was discharged home.

On these treatments, my blood pressure is around 130/80 mmHg and heart rate is basically between 60 and 70 bpm, measured by my electronic blood pressure monitor. I do feel markedly better and hardly appear to be breathlessness and do not suffer any heart racing. I am able to take a long walk, like more than ten thousand of steps and also sleep well. I think I have been brought back to a normal life.’

Biomarkers and left ventricular ejection fraction

In the overall population, N‐terminal pro‐brain natriuretic peptide (NT‐proBNP) levels decreased at M1 (P < 0.001) and M6 (Table 5 ; P < 0.001). Changes in NT‐proBNP in the inpatient subgroup closely resembled trends in the overall population, while those in the outpatient group decreased only slightly at M1 (P = 0.014), before ultimately reaching levels similar to those in the overall population by M6 (P = 0.012). BNP levels decreased significantly in the overall population at M1 (P = 0.002) and at M6 (P < 0.001); similar trends were seen in the inpatient subgroup. However, no significant changes in BNP were seen in the outpatient subgroup (Table  5 ).

Table 5.

Biomarker levels and LVEF at baseline and during treatment with ivabradine (full analysis set)

	Baseline		Month 1				Month 6
	N	Median (Q1, Q3)	N	Median (Q1, Q3)	Median change from baseline	P value a	N	Median (Q1, Q3)	Median change from baseline	P value a
NT‐proBNP (ng/L)
Outpatient population	73	1143.00 (410.70, 2984.00)	14	1084.00 (230.50, 2439.00)	−494.15 (−2161.00, −110.00)	0.014	10	347.60 (183.00, 751.60)	−772.25 (−2476.20, −303.80)	0.012
Inpatient population	367	2508.00 (1031.00, 5393.00)	94	1297.50 (593.80, 2901.00)	−1157.25 (−3132.00, −13.00)	<0.001	43	545.00 (111.00, 1775.00)	−1136.50 (−2194.98, −228.60)	<0.001
Overall population	440	2233.50 (891.90, 4905.50)	108	1206.50 (575.10, 2899.50)	−976.50 (−3057.50, −30.95)	<0.001	53	452.00 (135.20, 1515.00)	−1113.75 (−2202.00, −297.20)	<0.001
BNP (ng/L)
Outpatient population	57	476.99 (211.00, 1307.00)	19	510.30 (63.08, 785.00)	40.50 (−90.96, 113.06)	0.715	6	218.50 (10.40, 312.00)	−65.95 (−354.89, 34.00)	0.123
Inpatient population	240	1080.45 (452.00, 2191.50)	43	273.60 (75.00, 602.00)	−415.00 (−1301.40, −101.00)	<0.001	26	100.84 (38.00, 190.00)	−431.00 (−1362.60, −204.81)	<0.001
Overall population	297	941.80 (365.29, 2014.20)	62	318.00 (75.00, 749.10)	−159.00 (−1023.30, 58.00)	0.002 a	32	113.19 (37.00, 223.90)	−331.90 (−1238.60, −134.00)	<0.001 a

	Baseline		Month 1				Month 6
	N	Mean ± SD	N	Mean ± SD	LSM ± SE change from baseline	P value b	N	Mean ± SD	LSM ± SE change from baseline	P value b
LVEF (%)
Outpatient population	223	33.26 ± 7.57	22	35.58 ± 10.24	1.79 ± 4.42	0.079	34	43.98 ± 9.54	9.34 ± 10.07	<0.001 b
Inpatient population	738	30.15 ± 7.42	126	35.99 ± 9.58	7.53 ± 9.32	<0.001	100	43.46 ± 12.79	13.05 ± 12.01	<0.001 b
Overall population	961	30.87 ± 7.57	148	35.93 ± 9.65	6.70 ± 9.00	<0.001	134	43.59 ± 12.02	12.11 ± 11.62	<0.001 b

BNP, brain natriuretic peptide; LSM, least squares mean; LVEF, left ventricular ejection fraction; NT‐proBNP, N‐terminal pro‐brain natriuretic peptide; Q1, first quartile; Q3, third quartile; SD, standard deviation; SE, standard error.

^a Paired t‐test (log‐transformed value).

^b Paired t‐test.

LVEF increased from baseline at each follow‐up visit in the overall population and the inpatient subgroup (both P < 0.001) and was higher than baseline at M6 (P < 0.001), but not at M1, in the outpatient subgroup (Table  5 ). The magnitude of the mean increase in LVEF between baseline and M6 was ~12%.

In the ARNI use analysis, the mean (95% CI) changes from baseline in NT‐proBNP at M1 and M6 were similar across the three subgroups using ARNIs at any time (Supporting Information, Table S11 ). The change in NT‐proBNP from baseline at each visit was significant for all three subgroups (M1: some visits P = 0.009, all visits P < 0.001, and combined P < 0.001; M6: some visits P = 0.031, all visits P = 0.015, and combined P < 0.001). In the beta‐blocker use analysis, there were significant changes in NT‐proBNP from baseline to each visit in the subgroup with beta‐blocker use at all visits (P < 0.001 at all visits for M1 and M6, respectively) and significant changes in NT‐proBNP from baseline only at M1 in the subgroup with beta‐blocker use at some visits (P = 0.002; Supporting Information, Table S11 ).

A summary of the efficacy outcomes is shown in Figure 2 .

Figure 2.

Summary of efficacy outcomes. BNP, brain natriuretic peptide; CSS, clinical symptom score; HR, heart rate; KCCQ, Kansas City Cardiomyopathy Questionnaire; LVEF, left ventricular ejection fraction; NT‐proBNP, N‐terminal pro‐brain natriuretic peptide; NYHA, New York Heart Association; OSS, overall symptom score.

Cardiovascular death and re‐hospitalization for worsening HF

In the FAS, cardiovascular (CV) death occurred in 12 out of 1003 patients (1.2%). In the ARNI use subgroups, CV death occurred in 7 out of 600 patients (1.2%) in the subgroup with no ARNI use and 5 out of 202 patients (2.5%) in the subgroup with ARNI use at some visits. By beta‐blocker use, CV death occurred in 2 out of 116 patients (1.7%) in the subgroup with no beta‐blocker use and 10 out of 312 patients (3.2%) in the subgroup with beta‐blocker use at some visits. No CV deaths occurred in the beta‐blocker use at all‐visit subgroups; the incidence of CV death in this group was significantly lower than in the subgroup with no beta‐blocker use (P < 0.05).

In the FAS, 71 out of 1003 patients were re‐hospitalized for worsening HF (7.10%). In the ARNI use subgroups, re‐hospitalization due to worsening HF occurred in 41 out of 600 patients (6.8%) in the subgroup with no ARNI use, 18 out of 202 patients (8.9%) in the subgroup with ARNI use at some visits, and 12 out of 201 patients (6.0%) in the subgroup with ARNI use at all visits. In the beta‐blocker use subgroups, re‐hospitalization occurred in 12 out of 116 patients (10.3%) in the subgroup with no beta‐blocker use, 16 out of 312 patients (5.1%) in the subgroup with beta‐blocker use at some visits, and 43 out of 575 patients (7.5%) in the subgroup with beta‐blocker use at all visits. The incidence of re‐hospitalization due to worsening HF in the beta‐blocker use at all‐visit subgroups was significantly lower than in the no beta‐blocker use subgroup (P < 0.05). Due to the low event rates for CV death or re‐hospitalization due to worsening HF (<15%) in all subgroups during the observation period, the survival time or time to re‐hospitalization was not obtained (Supporting Information, Table S12 ).

Safety

Overall, 498 TEAEs occurred in 296 patients (29.5%) (Supporting Information, Table S13 ). Of these, 73 TEAEs [reported by 55 patients (5.5%)] were considered to be related to ivabradine; these were experienced by 20 (8.6%) outpatients and 35 (4.5%) inpatients. The most frequent ivabradine‐related TEAEs, using MedDRA preferred terms, were sinus bradycardia and photopsia, both in seven cases (0.7%). The most common drug‐related AEs were cardiac disorders, which occurred in 23 (9.9%) outpatients and 115 (14.9%) inpatients. A total of 18 patients developed 21 serious study drug‐related AEs, 4 events in 4 (1.7%) outpatients and 17 events in 14 (1.8%) inpatients, including sinus bradycardia (n = 4, 0.4%), atrial fibrillation (n = 4, 0.4%), cardiac failure (n = 3, 0.3%), hypotension (n = 2, 0.2%), and atrioventricular block second degree, cardiac failure chronic, ventricular extrasystoles, bradycardia, myocardial infarction, palpitations, electrocardiogram QT prolonged, and dermatitis allergic (1 patient, 0.1% each) (Supporting Information, Table S13 ).

TEAEs were reported by a similar number of patients in all ARNI and beta‐blocker subgroups (21.9–35.6%). Further, a similar number of patients in all subgroups experienced serious TEAEs, although the incidence tended to be lower in subgroups using beta‐blockers or ARNIs at all visits (Supporting Information, Table S14 ).

Discussion

This study demonstrated the HR‐lowering effectiveness of ivabradine in combination with HF medications in Chinese patients with chronic HF, including inpatients and outpatients. It provides real‐world evidence to complement the findings from previous randomized placebo‐controlled trials, including the SHIFT study, ⁴ , ¹¹ a placebo‐controlled trial in Japanese patients with similar study and baseline characteristics as the SHIFT study (J‐SHIFT), ¹² and the ETHIC‐AHF trial. ¹³

Our real‐world data from Chinese individuals with chronic HF are very similar to those of a subgroup analysis of Chinese patients participating in the SHIFT study, ⁵ which showed a decrease in HR of 9.3 b.p.m. after a mean follow‐up of 15.6 months ⁵ ; our study demonstrated an HR reduction of 16.1 b.p.m. sustained for up to 6 months with ivabradine. The greater HR reduction in our cohort is likely due to the higher baseline HR compared with the SHIFT study (mean 89 vs. 80 b.p.m.), ⁴ as a higher HR at the time of ivabradine initiation is associated with better treatment outcomes. ¹⁴ Our findings of significant improvements in LVEF and NYHA class were also consistent with the Chinese subgroup analysis from SHIFT. ⁵

Interestingly, the proportion of patients with HF of non‐ischaemic aetiology was higher in the present study (63%) than in the SHIFT study (33%). ⁴ This is consistent with data from the Swedish Heart Failure Registry, which showed that patients who met the criteria for SHIFT study inclusion were more likely to have ischaemic heart disease than the general HF population, ¹⁵ and highlights the importance of supplementing and complementing the results of randomized trials with those of observational research in a more heterogeneous patient population. ¹⁶

In our analysis, 37% of the overall population had ischaemic aetiology. Some individuals underwent revascularization; however, their ischaemic severity and chronology of revascularization and HF diagnosis were not specifically recorded, so their HF aetiology may not have been primarily ischaemic. Although a more conservative approach in patients with chronic coronary syndrome (CCS) is encouraged by guideline recommendations and implications of clinical trials, ¹⁷ the benefits of revascularization by percutaneous coronary intervention (PCI) or coronary artery bypass graft (CABG) in ischaemic HF treatment remain uncertain. Compared with optimal medical therapy (OMT) alone, revascularization does not lead to a better prognosis for death from any cause or hospitalization for HF among patients with ischaemic left ventricular systolic dysfunction. ¹⁸ , ¹⁹ However, QoL improvement has been reported as an underlying benefit of revascularization in patients with HF, partly reflecting improved cardiac function. The overall improvement in cardiac function in our analysis suggests that revascularization may bring value to medical treatment of HF in real‐world practice. This improvement in cardiac function may further contribute to better HR reduction. The observed improvements in QoL in our study are in line with findings from previous clinical trials ²⁰ , ²¹ and meta‐analyses. ²² , ²³

Other benefits of ivabradine in our study included an improvement in NYHA class, which provides a measure of the severity of functional limitations attributed to HF. In the SHIFT study, 28% of patients in the ivabradine group experienced an improvement in NYHA class of ≥1 grade after 12 months, ⁴ whereas in the current study, 67% of patients receiving ivabradine showed this level of improvement in NYHA class at 6 months. While the proportion of patients with baseline NYHA Class II in the current study was very similar to that of the overall SHIFT study population, our study included more patients with NYHA Class IV (22.0% vs. 2.0% in SHIFT). ⁴ Interestingly, in the SHIFT subgroup analysis of Chinese patients, the proportion of patients with a NYHA class improvement with ivabradine was much higher than in the overall SHIFT population (54% vs. 28%) ⁵ and closer to the proportion observed in the present study.

OMT, including RAASi, beta‐blockers, and MRAs, has been well established to contribute to overall improvement of cardiac function. ²⁴ , ²⁵ , ²⁶ , ²⁷ In our study, the majority of patients were taking these fundamental therapies. As a result, the improvement in NYHA class might be perceived as being influenced by the overall effectiveness of OMT.

Capacity to perform physical activities is a key factor in evaluation of NYHA functional classification. Lower physical activity in elderly populations or symptoms of peripheral artery diseases (PADs), such as claudication and pain at rest, make NYHA evaluations less reliable in these situations. ²⁸ In our study, patients aged >70 years were also included, and 3.2% of the overall population were diagnosed with PAD. These factors, combined with the use of OMT, may have led to a small overestimation of the effectiveness of ivabradine in improving cardiac function graded by NYHA.

The inclusion of patients in hospital settings in our study could explain why the baseline NT‐proBNP and BNP levels were four‐fold to five‐fold higher than in other those randomized studies, such as the J‐SHIFT, ¹² , ²⁹ while being closer to the findings of a large Chinese cohort study. ³⁰

The increase in the mean LVEF in our study is consistent with previous findings from randomized clinical studies. ¹² , ²⁹

The discharge time and the period immediately following discharge are recognized as vulnerable phases for HF patients, being associated with a high risk of poor outcomes; thus, optimization of therapy at this stage is critical. ¹⁰ The event rate for both CV death and HF re‐hospitalization in our analysis is comparable with that in the SHIFT study ⁴ at 6 months and is lower compared with other 6 month follow‐up study. ³¹ The greater loss to follow‐up seen in our study—and often seen in real‐world studies—may have influenced the reported CV death and hospitalization outcomes. The higher loss to follow‐up may have led to an overestimation of outcomes. Our study suggests that early initiation of ivabradine can be beneficial for patients with chronic HF, which is consistent with previous studies. ¹³ , ³² Subgroup analysis suggested that adding ivabradine to core therapies such as an ARNI or beta‐blockers provided clinical benefits and that HR was reduced during ivabradine treatment, irrespective of whether or not patients were using an ARNI or beta‐blocker. The majority of patients on combined beta‐blocker and ivabradine in our study were on <25% of target dosages of beta‐blockers. A retrospective real‐world study by Lee et al. demonstrated that simultaneous, rather than sequential, treatment with sacubitril/valsartan and ivabradine was a better strategy to improve outcomes and reverse left ventricular remodelling. ³³ In addition, in a post hoc analysis from the SHIFT database, it was reported that the magnitude of HR reduction achieved with a beta‐blocker plus ivabradine, rather than the background beta‐blocker dose, primarily determines subsequent effects on outcomes. ³⁴

Study strengths and limitations

Some of the strengths of this study include its prospective, observational design and the large number of enrolled patients (i.e. >1000). Additionally, the study population was regionally diverse because patients were enrolled at 72 clinical sites around China. However, because it was a real‐world observational study, there was a high rate of study withdrawal due to factors including loss to follow‐up, physician‐directed discontinuation, non‐compliance, and AEs (both ivabradine‐related and non‐related). Further, this study had no control group, some outcomes were subjective in nature (e.g. KCCQ), and no outcomes were evaluated after 6 months of treatment. In addition, major adverse cardiovascular events (MACEs) were not included as a primary endpoint. Finally, no clinical analysis was carried out after discontinuation of treatment in patients who withdrew or stopped treatment, which may have provided further insights into ivabradine‐specific outcomes.

Conclusions

In conclusion, treatment with ivabradine effectively reduced HR and improved QoL, symptoms, and cardiac function in Chinese patients with chronic HF. Addition of ivabradine to core therapies, such as beta‐blockers or ARNIs, resulted in clinical benefits, but the use of ARNIs or beta‐blockers did not change the response to ivabradine. This study confirms the effectiveness and safety of ivabradine in patients with chronic HF, some of which may be unique to the Chinese population.

Conflict of interest

J.Z., Y.X., Z.Z., S.Z., J.Y., Y.Z., B.T., H.H., Q.Z., F.L., W.D., C.Q., G.S., X.L., Y. Shen, B.S., X.K., Z.G., P.Z., X.G., H.Z., Y. Sun, Y.D., G.F., and J.G. have received fees, research grants, or both from Servier. L.F. is an employee of Servier.

Funding

This study was funded by Servier, China.

Supporting information

Table S1. Inclusion and exclusion criteria.

Table S2. Data collected at baseline.

Table S3. Details of statistical analysis.

Table S4. Patients who used β‐blockers: dosage and drug type.

Table S5. Baseline demographic and clinical characteristics in the angiotensin receptor neprilysin inhibitor and β‐blocker use subgroups.

Table S6. Heart rate during ivabradine treatment in the angiotensin receptor neprilysin inhibitor and β‐blocker use subgroups.

Table S7. Change in dyspnoea during ivabradine treatment.

Table S8. Change in asthenia during ivabradine treatment.

Table S9. Change in New York Heart Association cardiac function classification during ivabradine treatment.

Table S10. Analysis of covariance of change from baseline in quality of life according to Kansas City Cardiomyopathy Questionnaire clinical summary and total scores in the angiotensin receptor neprilysin inhibitor and β‐blocker use subgroups (full analysis set).

Table S11. Biomarker N‐terminal pro‐BNP levels during ivabradine treatment in the angiotensin receptor neprilysin inhibitor and β‐blocker use subgroups (full analysis set).

Table S12. Kaplan–Meier analysis for time to re‐hospitalization for worsening HF in the ARNI and β‐blocker use subgroups (FAS).

Table S13. Adverse events and treatment‐emergent adverse events (full analysis set).

Table S14. Summary of treatment‐emergent adverse events in the angiotensin receptor neprilysin inhibitor and β‐blocker use subgroups (full analysis set).

Zhou, J. , Xu, Y. , Zheng, Z. , Zhang, S. , Yang, J. , Zhang, Y. , Tang, B. , Han, H. , Zhang, Q. , Liu, F. , Ding, W. , Qian, C. , Su, G. , Liu, X. , Shen, Y. , Shi, B. , Kong, X. , Ge, Z. , Zhang, P. , Guo, X. , Zhang, H. , Sun, Y. , Dong, Y. , Fu, G. , Feng, L. , Ge, J. , and the POSITIVE investigators (2024) Effectiveness and safety of ivabradine in Chinese patients with chronic heart failure: an observational study. ESC Heart Failure, 11: 846–858. 10.1002/ehf2.14581.