Rationale and design of the randomised Treatment of sleep apnoea Early After Myocardial infarction with Adaptive Servo-Ventilation trial (TEAM-ASV I)

Polygraphy (SOMNOscreen™ plus RC, SOMNOmedics, Randersacker, Germany) will be performed within 3 days after PCI and repeated at the 12-week follow-up visit in both treatment groups (Fig. 1). Polygraphy evaluation includes breathing parameters (such as pressure cannula, snoring, thoracic and abdominal respiratory effort measures), oxygen saturation, ECG (including estimates of systolic and diastolic blood pressure based on the pulse-transit time method using validation references), recording of body position and light detection as well as a patient marker.

All polygraphy data are analysed by a central scoring lab to ensure consistency and quality. Each centre will send three respiratory recordings to the central lab for checking and validation as part of quality control. Apnoeas and hypopnoeas will be scored according to American Academy of Sleep Medicine (AASM) 2012 criteria by one experienced sleep technician blinded to the clinical data [22]. SDB is defined as an apnoea-hypopnoea index (AHI) ≥ 15/h (based on total recording time) and is categorised as predominantly OSA (≥ 50% of apnoeas obstructive) or predominantly CSA (< 50% of apnoeas obstructive).

Myocardial salvage, infarct size and left ventricular function will be examined in both groups using CMR at baseline and after 12 weeks of treatment. The baseline CMR is performed to visualise myocardial oedema (i.e. the area at risk) and baseline infarct size. The second CMR is performed to measure the final infarct size.

All CMR images will be acquired with a 1.5 or 3.0 Tesla clinical scanner, as previously described [13]. The scan protocol will be standardised at all sites. Patients will be examined in the supine position with care taken to ensure identical position in the scanner for both exams. All images will be acquired during breath hold and using an ECG gating technique. Cine images will be acquired with a steady-state free precession sequence in two-chamber, four-chamber and short-axis planes. Short-axis T2w-STIR will be used for myocardial oedema imaging. Infarct size will be assessed in delayed enhancement short-axis images (segmented inversion recovery SSFP technique) covering the whole ventricle, acquired 10–15 min after bolus injection of gadolinium contrast (0.2 mmol/kg body weight). Image analysis for the assessment of ejection fraction, area at risk and infarct size, total left ventricular mass and volumes using a semi-automatic approach will be performed in anonymised form by blinded operators with commercially available software. The salvaged myocardium is defined as area at risk minus final infarct size. MSI (primary endpoint) is defined as area at risk minus final infarct size as a percentage of the area at risk (Fig. 1).

Venous blood samples (10 mL + optional 40 mL) and urine (100 mL) will be collected from every patient participating in the study during the first 2 h after morning awakening at baseline and at 12 weeks (Fig. 1). Part of each sample (10 mL) will be used to measure levels of plasma amino terminal-pro B-type natriuretic peptide (NT-proBNP), total cholesterol, low-density lipoprotein (LDL) cholesterol, serum creatinine, high-sensitivity C-reactive protein (hsCRP), blood count and fibrinogen. The examination will be done in an assigned central laboratory.

Patients will be asked to complete the Seattle Angina Questionnaire [23] at baseline and at the final visit. Baseline and follow-up data will serve to document differences between these time points rather than evaluating changes over time. This approach accounts for the difficulties of quality of life assessment soon after a life-changing event (AMI). The Seattle Angina Questionnaire assesses subjective limitations of physical and daily activities by symptoms of coronary disease. Questionnaire scoring is undertaken as described previously [23].

Deaths, recurrent AMI, recurrent PCI, coronary bypass surgery, stent thrombosis, cerebral infarction, ventricular arrhythmia, rehospitalisation for heart failure or angina pectoris, cerebral infarction, syncope and other unexpected events will be documented at any visit and at first notice throughout the entire follow-up period. The investigator must report in detail all adverse signs and symptoms which are either volunteered by patients or observed during or following the course of investigational product administration. All AEs (including serious, unexpected/unanticipated device-related AEs (UADEs)) have to be documented in the CRF and source documents. Safety analyses will be performed in the safety population including participants who will have at least one post-baseline safety assessment. Patient informed consent includes long-term follow-up phone interviews after the 3-month trial period to assess mortality at 1, 3 and 5 years after randomization.

The maximum possible effect of ASV therapy on MSI would be a reduction of 25, as calculated from a previous prospective observational study (MSI in no SDB versus SDB group, 77 versus 52) [13]. We consider that the minimum clinically relevant effect would be an MSI improvement of 10 in the ASV group compared with the control group. According to Eitel et al. [2] this degree of improvement in the MSI is associated with a 0.6 times lower risk of developing a major cardiovascular event after AMI. Therefore, to detect a MSI difference of 10 with a standard deviation of 15 ¹³ with a power of 1 − β = 80%, at a two-sided alpha significance level of 0.05, the required sample size for a randomised trial is 37 patients per group (total 74 patients). With an estimated loss to follow-up rate of approximately 17% based on previous observational data, 90 randomised patients (45 per group) are required.

The point estimate for the primary endpoint (MSI) will be presented as mean and standard deviation (SD) with the corresponding 95% confidence interval (CI). MSI values will be compared between the ASV and control groups by an analysis of covariance (ANCOVA) with treatment (ASV therapy and control) as the fixed factor and infarct size at baseline, infarct location and study centre as an additional covariable. The results will be presented in terms of least-squares means and corresponding 95% CI. The primary analysis will be performed on an intention-to-treat basis (ITT analysis set, all randomised patients). In addition, we will perform a sensitivity analysis in a per-protocol analysis set (PP population) to assess the robustness of the results. The PP population consists of all patients in the ITT population who show no major protocol violations. Major and minor protocol deviations will be identified before database lock. In addition, an on-treatment analysis, including patients with positive airway pressure support use for ≥ 4 h/day and an on-treatment AHI of < 10/h, will be performed. All safety data will be analysed in the safety population (all patients who had at least one post-baseline safety assessment). Additional pre-specified subgroup analyses are: LAD versus non-LAD lesions; central versus obstructive sleep apnoea; and PCI < 24 h versus ≥ 24 h after symptom onset.

No imputation methods for missing values will be used for the analyses of the primary and secondary endpoints (complete case analysis). In case of 5 to 40% missing data regarding the primary endpoint, a multiple imputation method will be used for further sensitivity analyses. If more than 40% of the data are missing, the trial results may only be considered as hypothesis generating results [24].

All secondary endpoints will be analysed in an exploratory manner and summarised using descriptive statistics. Safety variables will be analysed descriptively. The interim analysis after randomisation of the 35th patient will include safety variables only. All efficacy analyses will be performed on the ITT population and will be two-sided at a significance level of 0.05. All statistical analyses will be performed using SAS software version 9.4 or higher.