Transcript Slides - Clinical Trial Results

The Sleep Apnea cardioVascular
Endpoints study (SAVE)
R. Doug McEvoy
Professor of Medicine, Adelaide Institute for Sleep Health
Flinders University, AUSTRALIA
[email protected]
For the SAVE Investigators and Coordinators, on behalf of the
SAVE Executive , Operations, and Advisory Committees
Disclosures
• Personal disclosure: Research funding - Philips Respironics,
AirLiquide, ResMed and National Health and Medical
Research Council (NHMRC) of Australia
• Main funding for SAVE - Philips Respironics and NHMRC
• Additional support - ResMed, Fisher&Paykel, Australasian Sleep Trials
Network, Spanish Respiratory Society, and Fondo de Investigaciones
Sanitarias
Rationale – obstructive sleep apnea (OSA)
• OSA causes repetitive night-time hypoxic episodes, sympathetic
activation, BP surges, and cardiac stress
• OSA associated with elevated BP, insulin resistance and endothelial
dysfunction - improved by continuous positive airway pressure (CPAP)
• OSA associated with an increased CV risk (stroke, coronary artery
disease, heart failure) and sudden death (most at night) – reduced in
observational studies with CPAP treatment
• OSA affects 40-60% of CV patients
• RCT clinical endpoint data lacking
Objective
• To determine if CPAP treatment of moderate
to severe OSA in patients with CV disease
would reduce the incidence of CV events
Design
• International, multicentre, prospective, open-label with
blinded adjudication of endpoints, randomised
controlled trial
– CPAP + usual care
– Usual care alone
Patients
Inclusion criteria
Exclusion criteria
• Age 45-75 years
• Severe sleepiness or risk of fall-asleep accident
• Coronary or cerebrovascular disease
• Moderate-severe OSA
– ApneaLink™, ResMed; 4% oxygen desaturation
index, ODI, ≥ 12 events/hr1
• Able to use CPAP mask
– Av. >3hrs/night during 1-week sham-CPAP run-in
– Epworth sleepiness score >15; fall-asleep or near
miss accident last 12 months; or commercial driver
• Severe oxygen desaturation
– >10% recording time with SaO2 <80%
• Heart Failure NYHA Class III-IV
• Cheyne Stokes respiration
• Able and willing to give informed consent
• Prior CPAP use
1. Ganter D et al. Respirology 2010;15:952-60
• Other condition which in opinion of
investigator made patient unsuitable
Outcomes
• Primary outcome - composite of MI, stroke, hospitalisation
for unstable angina, TIA or heart failure, and CV death
• Secondary CV outcomes
– Other composites: ischaemic CV events; major CV
events; cardiac events; cerebral events
– Individual components of primary outcome
– New-onset atrial fibrillation, diabetes
– All cause mortality
Other outcomes
• OSA symptoms - snoring, sleepiness (ESS), mood (HADS)
• Health related quality of life (SF36)
• Days off work due to sickness
• Safety outcomes
– Serious adverse events
– Accidents causing personal injury, fall-asleep accidents
or near miss accidents, traffic accidents
Statistics
Sample size - revised from 5,000 to 2,500 in 2012
– Challenges with recruitment - meta-regression showed strong relationship of AHI and
CV events; blinded data of high event rates; good CPAP adherence
– Power - 25% RR reduction for primary endpoint (α 0.05, 1-β 0.90) for av. 4.5 years FU;
6.86% annual event rate with usual care, av. CPAP 3 hr/night - for total of 533 events
Statistics
– ITT, P<0.047 for primary outcome (1 formal interim analysis)
– Unadjusted survival analysis for time to first (adjudicated) event in Cox model
– Sensitivity analyses - Covariate adjustments - stratification factors, region, severity of
OSA; multiple events; events as clinician-reported
– CPAP complier (>4 hrs/night) PP-matched
Recruitment and Management
Dec 2008 - Dec 2015
 2717 patients from 89
centres in 7 countries

International Coordinating
Centre

Core sleep laboratory
Patient Flow
15,325 Screened
9481 Ineligible or declined
5844 ApneaLink sleep study
2598 Excluded
3246 Sham CPAP run-in
529 Excluded
2717 Randomised
1359 CPAP + Usual Care
1358 Usual Care
13 Withdrew consent or
were excluded
17 Withdrew consent or
were excluded
1346 included in ITT
1341 included in ITT
Baseline Characteristics
CPAP + usual care
N=1346
Usual care
N=1341
Age, mean, years
61
61
Male %
81
81
Ethnicity -Asian %
64
63
BMI, kg/m2
29
29
Snoring almost every day %
84
81
ODI 4% SaO2 dips/hr
28
28
AHI events/hr
29
30
Epworth sleepiness Score mean (0-24)
7.3
7.5
Baseline characteristics
Coronary artery disease, %
TIA/Stroke, %
Systolic/ diastolic BP, mmHg
Medical history, %
Past or Current Smoking
Hypertension
Diabetes mellitus
Medications, %
BP lowering drugs,
Lipid lowering drugs
Antiplatelet/antithrombotic
Oral anti-diabetic
Insulin
CPAP + Usual Care
N=1346
Usual Care
N=1341
51
49
132/ 80
51
49
131/ 79
57
79
30
55
78
29
78
60
75
22
6
78
60
75
22
6
CPAP adherence and OSA control
Primary endpoint – neutral
Hazard ratio (95% CI)
1.10 (0.91 - 1.32)
Sensitivity and secondary CV analyses
• No significant differences in primary outcome sensitivity analyses
• No significant differences in analyses of:
– Individual components of primary endpoint
• Except rate of TIA hospital admissions higher in CPAP group, P=0.04
– Other CV composites
– New-onset diabetes or newly diagnosed AF
– All cause mortality
Primary endpoint subgroups
Sleepiness, mood, QoL, work and safety
CPAP+Usual Care versus Usual Care
• Improved
–
–
–
–
Epworth sleepiness score (p<0.001)
HADS anxiety (p=0.002) and depression (p<0.001) scores
SF36 physical (p=0.002 ) and mental (p<0.001) component scores
Work days lost because of ill-health (p<0.001)
• No significant difference
– Serious adverse events
– Accidents
Per protocol analysis
CPAP good compliers
(i.e. >4 hrs/night;
n=561) matched 1:1
with usual care
patients by propensity
scores
Composite primary outcome, no. (%)
Components of primary endpoint
CV Death
Myocardial infarction
Stroke
Hospitalisation for heart failure
Hospitalisation for unstable angina
Hospitalisation for TIA
Other vascular endpoints
Composite of ischaemic CV events
Composite of major CV events
Composite for cerebral events
Composite for cardiac events
Hazard Ratio
(95% CI)
0.80 (0.60 to 1.07)
0.13
0.90 (0.41 to 2.01)
1.19 (0.59 to 2.39)
0.56 (0.32 to 1.00)
0.81
0.63
0.05
0.82 (0.34 to 2.03)
0.99 (0.64 to 1.51)
0.22 (0.03 to 2.01)
0.67
0.95
0.18
0.81 (0.59 to 1.10)
0.69 (0.46 to 1.04)
0.52 (0.30 to 0.90)
1.01 (0.74 to 1.39)
0.17
0.08
0.02
0.93
P
Conclusions
• Addition of CPAP to usual care did not reduce CV events in
patients with OSA and coronary or cerebrovascular disease
– Risk of stroke may be reduced in patients with high levels
of CPAP adherence
• CPAP significantly improved daytime alertness, mood, quality
of life, and reduced work-days lost because of ill-health
McEvoy RD et al. NEJM 2016, 28 August [Epub ahead of print].