This prospective open-label, single-arm study was conducted at Quanta Diagnóstico por Imagem, a high-volume imaging center in Curitiba, Brazil, which has performed more than 6000 MPS procedures annually since 2003.

Patients had to meet the following eligibility criteria to be included in the study: (1) diagnosis of CA; (2) age of 18–70 years; (2) symptomatic even when on beta blockers (at maximum tolerated doses); (4) resting HR ≥ 70 beats per minute (bpm); (5) sinus rhythm; (6) left ventricular ejection fraction (LVEF) > 40%; and (7) presence of myocardial ischemia at the baseline MPS, performed with an ETT to evaluate ischemia induced under controlled medical conditions. An upper age limit of 70 years was selected to ensure that all patients in our study could start on the same dose of ivabradine 5 mg twice daily (b.i.d.) (the recommended starting dosage in patients aged ≥ 75 years is 2.5 mg b.i.d.).

Exclusion criteria were: (1) unstable IHD; (2) acute myocardial infarction (MI) < 6 months previously; (3) LVEF ≤ 40%; (4) HR < 70 bpm; (5) presence of left bundle branch block; (6) a pacemaker;(7) inability to perform the ETT due to physical limitations; (8) pregnancy or nursing status (women); (9) contraindications to ivabradine in accordance with the summary of product characteristics; and (10) a contraindication to MPS.

The schedule of study visits and observations are summarized in Fig. 1. Treatment with ivabradine 5 mg b.i.d. was initiated by study investigators at the baseline visit (visit 1). Four weeks after ivabradine initiation, the dose was titrated according to HR, persistence of symptoms, and tolerance of the drug. The dose was increased to 7.5 mg b.i.d. if the HR was > 60 bpm, but titrated down to 2.5 mg (half of a 5 mg tablet) b.i.d. if, during the treatment with 5 mg ivabradine b.i.d., the HR decreased to < 50 bpm at rest or the patient experienced symptoms related to bradycardia, such as dizziness, fatigue, or hypotension. Similarly, in patients who were receiving 7.5 mg ivabradine b.i.d., the dose was titrated down to 5 mg b.i.d. if the HR decreased to < 50 bpm at rest or patients developed symptoms of bradycardia. A second MPS was performed 3 months after the initiation of ivabradine (2–3 months after reaching the optimal dose), without interruption or change in the dose of the beta blocker and/or ivabradine.

Study design and procedures. Asterisk indicates that the MPS images were analyzed in a blinded manner (both patient identity and ivabradine administration unknown) by two independent investigators. b.i.d. twice daily, ETT exercise treadmill test, MPS myocardial perfusion scintigraphy, SAQ-7 seven-item Seattle Angina Questionnaire

The primary endpoint was the percentage of myocardial ischemia, which was evaluated in a blinded manner by the same two independent investigators at baseline and at 3 months. During the analysis of images, investigators were aware of neither patient identity nor ivabradine administration. Time to ischemia (angina, horizontal or down-sloping ST-segment depression of ≥ 1 mm), duration of exercise, and peak HR during the ETT were also evaluated. Patients were asked to answer items from the physical limitation, angina frequency, and QoL domains of the shortened seven-item Seattle Angina Questionnaire (SAQ-7) [26] at baseline (first visit), during the second visit (1 month after initiation of ivabradine), and during the last visit (3 months after initiation of ivabradine).

After providing written informed consent, patients entered the study. Patients received background therapy for CCS according to contemporary guidelines. The study was approved by the Research Ethics Committee of the Hospital de Clinicas da Universidade Federal do Paraná, in Curitiba, Paraná, Brazil (ethics approval number 3.182.920), and all procedures were followed in accordance with the Helsinki Declaration of 1964, as revised in 2013.

Symptom-limited ETTs, using the standard Bruce protocol, were performed for the MPS stress test. Treatment for CCS, including beta blockers and ivabradine, was not interrupted when patients undertook the ETT. Time to onset of angina, time to limiting angina during the ETT, total exercise duration, and time to 1-mm horizontal or down-sloping ST-segment depression were determined by the investigator. HR was obtained from the electrocardiogram (ECG) recording, and the rate-pressure product was determined using BP measurements at rest and during exercise. Interruption criteria were limiting angina, dyspnea, and extreme fatigue. The depression of the ST-segment was measured at the J-point in three consecutive QRS complexes with a flat baseline.

The MPS studies were performed according to institutional protocols following current guidelines. Images were acquired on the Ventri Cardiac Imaging System (2006 model; GE HealthCare, Chicago, IL, USA) or the NM 530c Discovery Nuclear Medicine System (2012 model; GE HealthCare). Resting scans were performed 30–60 min after intravenous injection of the radiopharmaceutical 99mTc-sestamibi. Stress tests were conducted on a separate date with another administration of 99mTc-sestamibi, and images were acquired after 15–60 min of physical stress. Radiopharmaceutical activity was determined based on body mass index, and activity optimization was allowed by increasing acquisition time, as recommended by the guidelines [21, 27, 28]. The radiation dose applied during MPS in the clinic is 8.4 msV, consistent with American Society of Nuclear Cardiology recommendations of a dose < 9 msV [27]. Images were processed using quantified SPECT gated software (QGS; Cedars-Sinai, Los Angeles, CA, USA) and analyzed by standard semi-quantitative visual quantification. Using the 17-segment polar map, each segment was scored from 0 to 4 according to radiopharmaceutical uptake (0 = normal, 1 = discrete reduction, 2 = moderate reduction, 3 = significant reduction, 4 = absence of radiopharmaceutical uptake).

From the sum of values attributed to each segment in the stress phase, the summed stress score (SSS) and summed rest score (SRS) were obtained. The summed difference score (SDS) was calculated as SSS − SRS. The primary outcome of interest was the total ischemic burden, reflected by the percentage of ischemic myocardium. Normalized values were calculated by dividing the measured SSS, SRS, and SDS score by 68 (derived from a deficit of 4 points in each of the 17 segments of the MPS polar map) and then multiplying the result by 100% to convert the value to a percentage [21]. The degree of ischemia was also categorized as: (1) absent or minimal, if SDS < 2; (2) mild, if SDS 2–4; (3) moderate, if SDS 5–8, and (4) significant, if SDS > 8 [29, 30]. All MPS readings were performed independently by two trained nuclear cardiologists who were blinded to whether ivabradine was administered or not. Any disagreement was resolved by a consensus achieved with the assistance of a third trained nuclear cardiologist.

Since there are no published studies evaluating MPS in ivabradine-treated patients, we examined the sample sizes used in other published studies of ivabradine (between 12 and 59 patients) [15, 18, 19]. In addition, data from the nuclear sub-study of the COURAGE trial [23] showed a difference in the ischemic myocardium of –2.7% with percutaneous coronary intervention and optimal medical therapy and –0.5% with optimal medical therapy alone.

On this basis, we estimated that 20 patients would be necessary to detect a 4 ± 6 percentage change in myocardial ischemia after treatment with ivabradine with a power of 80% and a two-sided alpha of 5%, assuming a baseline myocardial ischemia of 8%.

Continuous variables were evaluated for the Gaussian distribution of the data and presented as mean ± standard deviation (SD) if normal, or median (25th percentile, 75th percentile) if non-normal. Categorical variables were presented as proportions. Continuous variables were compared before and after initiation of ivabradine treatment using the paired t-test or the Wilcoxon signed rank test if the delta between the two measurements was normal or non-normal, respectively. Categorical variables before and after the intervention were compared using the McNemar exact test. The level of significance was set as 0.05; p values were not adjusted for multiple comparisons. Analyses were performed using Stata version 15 (Stata Corp, College Station, TX, USA).