2023 Volume 87 Issue 10 Pages 1369-1379
Background: The effects of obstructive sleep apnea (OSA) on the prognosis of acute coronary syndrome (ACS) without revascularization remain unclear, so the aim of the present study was to elucidate the association of OSA with subsequent cardiovascular events in ACS patients with and without revascularization.
Methods and Results: We prospectively recruited hospitalized ACS patients undergoing sleep monitoring between June 2015 and January 2020. OSA was defined as an apnea-hypopnea index ≥15 events/h. The primary endpoint was a major adverse cardiovascular and cerebrovascular event (MACCE), including cardiovascular death, myocardial infarction, stroke, ischemia-driven revascularization, or hospitalization for unstable angina or heart failure. Among 1,927 patients, 52.6% had OSA and 69.4% underwent revascularization. During a 2.9-year follow-up (1.5–3.6 years), the risk of MACCE was similar in patients with or without revascularization. OSA was an independent predictor of MACCE in the non-revascularization group (22.6% vs. 14.6%; hazard ratio (HR) 1.861; 95% confidence interval (CI) 1.239–2.796; P=0.003) but not in revascularization group (22.3% vs. 19.3%; HR 1.135; 95% CI 0.882–1.460; P=0.324). The incremental risk in the non-revascularization group was attributable to more hospitalizations for unstable angina (14.2% vs. 8.6%; HR 1.896; 95% CI 1.124–3.199; P=0.016).
Conclusions: For patients with ACS, OSA was independently associated with higher risk of recurrent cardiovascular events among patients without revascularization but not among patients undergoing revascularization. The benefits of suitable OSA treatment for patients without revascularization need further investigation.
Obstructive sleep apnea (OSA) is characterized by frequent partial or complete closure of the upper airway during sleep.1 OSA is associated with increased risk of cardiovascular disease, but it remains uncertain whether OSA increases recurrent cardiovascular events.2 Several longitudinal observational studies reported that OSA was an independent predictor of poor outcomes in patients with known cardiovascular diseases.3,4 However, the Impact of Sleep Apnea Syndrome in the Evolution of Acute Coronary Syndrome/Effect of Intervention With Continuous Positive Airway Pressure (ISAACC) study reported that OSA does not increase the risk of cardiovascular events in patients with acute coronary syndrome (ACS).5 Interventional studies also failed to find a protective effect of continuous positive airway pressure (CPAP) treatment on the recurrence of coronary artery disease (CAD) or ACS,6,7 emphasizing the need to better identify high-risk subsets of ACS patients.
Coronary revascularization is the most effective treatment for CAD, but residual angina or recurrent cardiac event after revascularization is not uncommon.8–10 OSA was reported as a risk factor for recurrent events after percutaneous coronary intervention (PCI).11–13 However, the Sleep and Stent Study found no significant difference between OSA and non-OSA groups in the incidence of stent-related adverse events (target vessel revascularization and stent thrombosis).14 Another study showed that the incidence of recurrent events among OSA patients was related to non-culprit lesions rather than the culprit lesion.15 Moreover, >40% of patients who underwent elective coronary angiography had an indication for PCI.16 The effects of OSA on the prognosis of ACS patients without revascularization are still unclear. Therefore, we used data from a large prospective cohort study to evaluate the impact of OSA on the risk of cardiovascular events in ACS patients with or without revascularization.
The study dataset was extracted from the OSA-ACS project (NCT03362385), a prospective cohort study that has been reported previously.17,18 Briefly, patients aged 18–85 years and admitted for ACS in Beijing Anzhen Hospital, Capital Medical University between June 2015 and January 2020 were eligible for inclusion. ACS included ST-segment elevation myocardial infarction (STEMI), non-STEMI (NSTEMI) and unstable angina (UA). Acute MI required evidence of myocardial injury (defined as increased cardiac troponin or high-sensitivity cardiac troponin with ≥1 value above the 99th percentile of the upper reference limit) in a clinical setting consistent with myocardial ischemia and meeting other criteria required by guidelines.19,20 AMI patients with ST-segment elevation in at ≥2 contiguous leads were designated as STEMI. Patients without ST-segment elevation at presentation were designated as NSTEMI.19,20 UA manifested as angina occurring at rest and prolonged usually >20 min; new-onset angina of at least Canadian Class Score (CCS) class III severity; previously diagnosed angina that has become distinctly more frequent, longer in duration, or lower in threshold (i.e., increased by ≥1 CCS class to at least CCS class III severity).21 Some ACS may occur in the absence of anatomical coronary artery stenosis, including ischemia with non-obstructive CAD (INOCA) and MI with non-obstructive CAD (MINOCA). INOCA was defined as angina with non-obstructive CAD (<50% diameter stenosis). MINOCA was defined as MI with non-obstructive CAD (<50% diameter stenosis). For patients with INOCA or MINOCA, coronary microvascular dysfunction (CMD) was further evaluated by angiography-derived index of microcirculatory resistance (angio-IMR) (FlashAngio IMR, Rainmed Ltd., Suzhou, China), which has been proven to be reliable in previous studies.22,23 Exclusion criteria were cardiogenic shock, cardiac arrest, malignancy, and failed sleep monitoring (inadequate or unsatisfactory signal recording). Patients with predominantly central sleep apnea (≥50% central events and central apnea-hypopnea index [AHI] ≥10/h) and those receiving continuous positive airway pressure (CPAP) therapy after discharge were also excluded. The study was approved by the Ethics Committee of Beijing Anzhen Hospital, Capital Medical University (2013025). All participants provided written informed consent. Patients or the public were not involved in the design, or conduct, or reporting, or dissemination plans of our research.
Sleep StudyAll recruited patients underwent an overnight sleep study during hospitalization after clinical stabilization with portable cardiorespiratory polygraphy (ApneaLink Air, Resmed, Australia). Nasal airflow (nasal pressure transducer), snoring episodes (integrated pressure transducer), thoraco-abdominal movements (respiratory inductance plethysmography belts) and arterial oxygen saturation (SaO2) (pulse oximetry) were recorded. We scored apneas and hypopneas according to the American Academy of Sleep Medicine criteria (2007). The AHI was calculated as the mean number of apneas and hypopneas per hour of total recording time. Most studies classified patients using a cutoff of AHI ≥15, for the reason that the incidence of long-term clinical outcomes was statistically different between AHI ≥15 and AHI <15 groups.24,25 Thus, OSA was defined as AHI ≥15 with absence of airflow despite respiratory movement or exertion. The sleep research staff was blinded to the demographic and clinical characteristics of the patients, and the records were imported into a specialized sleep apnea database. Decisions concerning referral of OSA patients to a sleep clinic for further evaluation and consideration of CPAP therapy were made in accordance with local clinical practice.
Procedure and ManagementDiseased vessel was defined as ≥1 lesion with ≥50% diameter stenosis. Revascularization was not considered in patients without significant coronary artery stenosis (≤70% diameter stenosis). Treatment of these patients was focused on lifestyle intervention and secondary prevention medications such as lipid-lowering, antiplatelet, and antianginal medications (β-blockers, nitrates, or calcium-channel blockers) to improve angina symptoms.
Revascularization strategies were determined according to the lesion and clinical status of patients (NSTE-ACS or STEMI), based on local protocols and current guidelines.26,27 Shared decision-making between the clinician and patient was fully considered. PCI was performed in a major coronary artery with severe stenosis (>70% stenosis or >50% stenosis for the left main artery) or fulfilling the physiological criteria for revascularization (i.e., fractional flow reserve, FFR <0.80). The extent of revascularization was free to local protocols and investigators’ decision. Staged procedures were allowed to achieve complete revascularization for multivessel disease. There was no restriction to treating complex lesions that required modification devices. For patients with multivessel CAD with surgical indications, the cardiac surgeon and interventional cardiologist reviewed and discussed the case to reach consensus. Coronary artery bypass grafting (CABG) procedures could be done with or without extracorporeal circulation. After the procedure, dual antiplatelet therapy was recommended to all patients, and duration of antiplatelet therapy followed contemporary guidelines.28,29 Other management followed the current standard of care.
Endpoints and Follow-upThe primary endpoint was major adverse cardiovascular and cerebrovascular events (MACCE), including cardiovascular death, hospitalization for ACS [including non-fatal MI and UA], stroke, ischemia-driven revascularization, and hospitalization for heart failure. The secondary endpoints were individual components of the primary endpoint. Definitions of events were in accordance with those proposed by the Standardized Data Collection for Cardiovascular Trials Initiative.30
Included patients were followed until December 2020. Follow-up contact was made at 1, 3, and 6 months after discharge, and every 6 months thereafter. Clinic visits, medical record reviews, or telephone calls were used to collect clinical event information. Clinical adverse events were collected and confirmed by source record and were independently evaluated by adjudicators blinded to the results of the sleep study.
Statistical AnalysisContinuous variables are expressed as mean±standard deviation or median (1st and 3rd quartiles) and compared between groups by Student’s t test or the Mann-Whitney U test. Categorical variables are expressed as number and percentage and compared between groups by χ2 statistics or Fisher’s exact test, as appropriate. Kaplan-Meier curves were generated for the OSA and non-OSA groups stratified by revascularization and compared by log-rank test. Furthermore, hazard ratios (HRs) and 95% confidence intervals (CIs) were tested with Cox hazard proportional regression model. The proportional hazards assumption was checked using log-log plots or the Schoenfeld residuals test. A multivariable marginal Cox model was used to identify independent predictors of MACCE and secondary endpoints. Model 1 covariates included age, sex, body mass index (BMI), current smoking, hypertension, diabetes, and dyslipidemia. Model 2 included Model 1 covariates and prior stroke, prior MI, history of revascularization, diagnosis (STEMI vs. non-STE-ACS), and optimal medical therapy (OMT; defined as the combination of aspirin, P2Y12 inhibitor, statin, angiotensin-converting enzyme inhibitor/angiotensin-receptor blocker, and β-blocker31) at discharge. The variables were carefully chosen based on clinical relevance or variables that showed a univariate relationship with outcome. The multiplicative interaction between revascularization and OSA was also tested in a Cox model. In the sensitivity analyses, we considered AHI ≥30, ODI ≥15 events/h and T90 (percentage of time with SaO2 <90%) ≥median as the alternative to AHI ≥15. Patients without revascularization were further stratified by the presence or absence of any lesion with ≥50% diameter stenosis. Patients undergoing revascularization were further stratified by PCI or CABG, and NSTE-ACS or STEMI. If a patient experienced ≥1 event, only the first event was included in the analysis. A two-sided P value <0.05 was considered statistically significant. Statistical analyses were performed using SPSS version 26 (IBM Inc., Armonk, NY, USA).
The flow diagram is presented in Figure 1. Of 2,160 patients with ACS recruited, 2,109 patients underwent an overnight sleep study. Patients with central sleep apnea and loss of follow-up were excluded, leaving 1,969 patients with OSA for follow-up. Among them, 42 (2.1%) patients received regular CPAP therapy (>4 h/day and >21 days/month). Finally, data for 1,927 patients were analyzed and of them, 1,337 (69.4%) patients underwent revascularization, and 1,014 (52.6%) had OSA.
Study flowchart. CPAP, continuous positive airway pressure; OSA, obstructive sleep apnea.
Compared with patients without revascularization, patients with revascularization were younger, had higher prevalence of male sex, smoking and history of revascularization, and a higher prevalence of STEMI and lower prevalence of dyslipidemia. The use of guideline-directed medical therapy other than statins was more frequent in the revascularization group than in the non-revascularization group. The prevalence of OSA was higher in the revascularization group, but the measurements of the sleep study were similar between groups (Supplementary Table 1).
Patients with OSA were more obese, had a higher proportion of male sex and hypertension than patients without OSA, regardless of revascularization status. For patients undergoing revascularization, the OSA group was characterized by a higher proportion of prior PCI. The prevalence of dyslipidemia, diabetes, and smoking did not differ significantly between the OSA and non-OSA group for patients with and without revascularization (Table 1). For lesion and procedural characteristics, the OSA group showed a higher proportion of any coronary artery with stenosis ≥70% and a lower proportion of MINOCA than the non-OSA group among patients without revascularization (Table 2).
| Variables | Non-revascularization (n=590) | Revascularization (n=1,337) | ||||
|---|---|---|---|---|---|---|
| Non-OSA (n=302) |
OSA (n=288) |
P value | Non-OSA (n=611) |
OSA (n=726) |
P value | |
| Demographics | ||||||
| Age, years | 56.8±10.0 | 59.0±9.8 | 0.163 | 55.9±10.6 | 56.0±10.8 | 0.959 |
| Male | 226 (74.8%) | 244 (84.7%) | 0.003 | 517 (84.6%) | 642 (88.4%) | 0.041 |
| BMI, kg/m2 | 26.12±3.51 | 28.07±3.58 | <0.001 | 25.93±3.34 | 28.03±3.55 | <0.001 |
| Neck circumference, cm | 40 (37, 42) | 41 (39, 43) | <0.001 | 40 (38, 42) | 42 (39, 44) | <0.001 |
| Waist-hip ratio | 0.96 (0.93, 1.00) | 0.99 (0.96, 1.03) | <0.001 | 0.97 (0.94, 1.01) | 0.99 (0.96, 1.03) | <0.001 |
| Systolic blood pressure, mmHg |
129 (120, 138) | 129 (120, 140) | 0.327 | 125 (116, 137) | 126 (116, 138) | 0.510 |
| Diastolic blood pressure, mmHg |
75 (70, 83) | 79 (70, 86) | 0.014 | 75 (69, 83) | 76 (70, 85) | 0.002 |
| Medical history | ||||||
| Hypertension | 183 (60.6%) | 197 (68.4%) | 0.048 | 373 (61.0%) | 494 (68.0%) | 0.008 |
| Diabetes | 94 (31.1%) | 96 (33.3%) | 0.566 | 196 (32.1%) | 223 (30.7%) | 0.593 |
| Dyslipidemia | 108 (35.8%) | 108 (37.5%) | 0.661 | 186 (30.4%) | 235 (32.4%) | 0.450 |
| Current smoking | 116 (38.4%) | 122 (42.4%) | 0.328 | 303 (49.6%) | 377 (51.9%) | 0.394 |
| Prior heart failure | 6 (2.0%) | 8 (2.8%) | 0.528 | 8 (1.3%) | 13 (1.8%) | 0.481 |
| Prior stroke | 24 (7.9%) | 35 (12.2%) | 0.089 | 62 (10.1%) | 86 (11.8%) | 0.324 |
| Prior MI | 50 (16.6%) | 58 (20.1%) | 0.261 | 89 (14.6%) | 119 (16.4%) | 0.359 |
| Prior PCI | 69 (22.8%) | 78 (27.1%) | 0.234 | 96 (15.7%) | 156 (21.5%) | 0.007 |
| Prior CABG | 5 (1.7%) | 9 (3.1%) | 0.241 | 6 (1.0%) | 9 (1.2%) | 0.656 |
| Baseline tests | ||||||
| Hemoglobin, g/L | 145 (134, 156) | 148 (137, 158) | 0.110 | 147 (137, 156) | 137 (147, 157) | 0.769 |
| Platelets, 109/L | 220 (185, 258) | 208 (179, 248) | 0.090 | 218 (186, 262) | 223 (187, 260) | 0.423 |
| LDL-C, mmol/L | 2.19 (1.71, 2.92) | 2.27 (1.78, 2.90) | 0.357 | 2.50 (1.94, 3.20) | 2.54 (1.98, 3.13) | 0.549 |
| hs-CRP, mg/L | 1.10 (0.51, 2.69) | 1.51 (0.65, 4.35) | 0.007 | 1.90 (0.65, 5.63) | 3.30 (1.19, 8.76) | <0.001 |
| Glycosylated hemoglobin, % | 6.1 (5.6, 6.7) | 6.1 (5.7, 6.9) | 0.190 | 6.0 (5.6, 7.0) | 6.1 (5.6, 7.2) | 0.065 |
| eGFR, mL/min/1.73 m2 | 107.2 (91.0, 125.2) |
105.7 (91.2, 121.3) |
0.556 | 106.6 (90.4, 122.3) |
101.9 (86.7, 118.2) |
0.002 |
| LVEF, % | 63 (60, 67) | 62 (59, 66) | 0.155 | 61 (55, 65) | 60 (55, 65) | 0.086 |
| Sleep study | ||||||
| AHI, event/h | 7.9 (4.8, 10.8) | 28.4 (20.8, 39.8) | <0.001 | 7.5 (3.9, 10.7) | 29.3 (20.8, 43.7) | <0.001 |
| ODI, event/h | 8.5 (5.2, 11.8) | 26.6 (20.7, 36.5) | <0.001 | 8.6 (4.6, 11.9) | 27.9 (20.1, 40.6) | <0.001 |
| T90, % | 0.4 (0.1, 2.9) | 7.0 (2.0, 17.0) | <0.001 | 0.6 (0.0, 3.0) | 6.0 (2.0, 15.0) | <0.001 |
| Mean SaO2, % | 94 (93, 95) | 93 (92, 94) | <0.001 | 94 (93, 95) | 93 (92, 94) | <0.001 |
| Minimum SaO2, % | 87 (85, 89) | 82 (77, 86) | <0.001 | 88 (84, 90) | 83 (77, 86) | <0.001 |
| Epworth Sleepiness Scale | 6.0 (3.0, 11.0) | 8.5 (5.0, 12.0) | 0.007 | 6.0 (3.0, 10.0) | 8.0 (4.0, 12.0) | <0.001 |
| Medications at discharge | ||||||
| Aspirin | 285 (94.4%) | 273 (94.8%) | 0.822 | 605 (99.0%) | 714 (98.3%) | 0.289 |
| P2Y12 inhibitor | 231 (76.5%) | 221 (76.1%) | 0.944 | 599 (98.0%) | 717 (98.8%) | 0.289 |
| β-blocker | 193 (63.9%) | 204 (70.8%) | 0.073 | 496 (81.2%) | 595 (82.0%) | 0.715 |
| ACEI/ARB | 155 (51.3%) | 182 (63.2%) | 0.004 | 375 (61.4%) | 483 (66.5%) | 0.050 |
| Statin | 294 (97.4%) | 284 (98.6%) | 0.278 | 606 (99.2%) | 713 (98.2%) | 0.124 |
| OMT | 103 (34.1%) | 136 (47.2%) | 0.001 | 312 (51.1%) | 402 (55.4%) | 0.116 |
| Calcium-channel blocker | 78 (25.8%) | 89 (30.9%) | 0.171 | 93 (15.2%) | 143 (19.7%) | 0.032 |
Data are presented as n (%), mean±standard deviation, or median (1st quartile, 3rd quartile). ACEI, angiotensin-converting enzymes inhibitor; AHI, apnea-hypopnea index; ARB, angiotensin-receptor blocker; BMI, body mass index; CABG, coronary artery bypass grafting; eGFR, estimated glomerular filtration rate; hs-CRP, high-sensitivity C-reactive protein; LDL-C, low-density lipoprotein cholesterol; LVEF, left ventricular ejection fraction; MI, myocardial infarction;ODI, oxygen desaturation index; OMT, optimal medical therapy; OSA, obstructive sleep apnea; PCI, percutaneous coronary intervention; SaO2, arterial oxygen saturation; T90, percentage of time with SaO2 <90%.
| Non-revascularization (n=590) | Revascularization (n=1,337) | |||||
|---|---|---|---|---|---|---|
| Non-OSA (n=302) |
OSA (n=288) |
P value | Non-OSA (n=611) |
OSA (n=726) |
P value | |
| Diagnosis | 0.429 | 0.035 | ||||
| STEMI | 20 (6.6%) | 24 (8.3%) | 159 (26.0%) | 227 (31.3%) | ||
| NSTE-ACS | 282 (93.4%) | 264 (91.7%) | 452 (74.0%) | 499 (68.7%) | ||
| Non-obstructive CAD | ||||||
| INOCA | 43/223 (19.3%) | 38/201 (18.9%) | 0.921 | – | – | – |
| MINOCA | 22/54 (40.7%) | 15/64 (23.4%) | 0.044 | – | – | – |
| CMD# | 53/272 (19.5%) | 48/258 (18.6%) | 0.796 | – | – | – |
| Diseased vessel (≥50%) location | ||||||
| LM | 7/277* (2.5%) | 16/265* (6.0%) | 0.043 | 54 (8.8%) | 61 (8.4%) | 0.777 |
| LAD | 167/277 (60.3%) | 160/265 (60.4%) | 0.983 | 523 (85.6%) | 606 (83.5%) | 0.285 |
| LCX | 105/277 (37.9%) | 111/265 (41.9%) | 0.344 | 400 (65.5%) | 474 (65.3%) | 0.946 |
| RCA | 93/277 (33.6%) | 108/265 (40.8%) | 0.084 | 401 (65.6%) | 486 (66.9%) | 0.613 |
| No. of diseased vessels (≥50%) | ||||||
| 0 | 65/277 (23.5%) | 53/265 (20.0%) | 0.328 | 0 (0%) | 0 (0%) | – |
| 1 | 106/277 (38.3%) | 89/265 (33.6%) | 0.256 | 145 (23.7%) | 172 (23.7%) | 0.986 |
| 2 | 58/277 (20.9%) | 75/265 (28.3%) | 0.046 | 209 (34.2%) | 259 (35.7%) | 0.575 |
| 3 | 48/277 (17.3%) | 48/265 (18.1%) | 0.811 | 257 (42.1%) | 295 (40.6%) | 0.597 |
| ≥2 | 106/277 (38.3%) | 123/265 (46.4%) | 0.055 | 466 (76.3%) | 554 (76.3%) | 0.986 |
| Any severely diseased vessel (≥70%) |
132/277 (47.7%) | 151/265 (57.0%) | 0.030 | 611 (100%) | 726 (100%) | – |
| Procedure | ||||||
| PCI | – | – | – | 543 (88.9%) | 668 (92.0%) | 0.050 |
| Drug-eluting stent | – | – | – | 471 (77.1%) | 581 (80.0%) | 0.191 |
| Drug-coated balloon | – | – | – | 37 (6.1%) | 48 (6.6%) | 0.678 |
| Plain balloon | – | – | – | 33 (5.4%) | 31 (4.3%) | 0.335 |
| CABG | – | – | – | 71 (11.6%) | 60 (8.3%) | 0.040 |
| TIMI flow grade of target vessel | ||||||
| Baseline TIMI flow grade 0 or 1 |
– | – | – | 178 (29.1%) | 245 (33.7%) | 0.071 |
| Final TIMI flow grade 3 | – | – | – | 538 (88.2%) | 655 (90.4%) | 0.205 |
Data are presented as n (%), n/N (%), mean±standard deviation, or median (1st quartile, 3rd quartile). #CMD defined as abnormal angio-index of microvascular resistance (IMR) (target vessel angio-IMR ≥40 in STEMI, angio-IMR ≥25 in ≥1 vessel in NSTE-ACS) with symptoms of myocardial ischemia and absence of obstructive CAD (<50% diameter stenosis or FFR >0.80). *48 patients failed to undergo coronary angiography before discharge for: (1) rejection by the patient or their principals, (2) contraindications, or (3) patients with infectious diseases who can only undergo the procedure in infectious disease hospital. ACS, acute coronary syndrome; CAD, coronary artery disease; CMD, coronary microvascular dysfunction; INOCA, ischemia with non-obstructive coronary artery disease; LAD, left anterior descending; LCX, left circumflex; LM, left main; MINOCA, myocardial infarction with non-obstructive coronary artery disease; NSTE, non-ST-segment elevation; non-PCI, percutaneous coronary intervention; PTCA, percutaneous transluminal coronary angioplasty; RCA, right coronary artery; TIMI, Thrombolysis in Myocardial Infarction. Other abbreviations as in Table 1.
During a median follow-up of 2.9 (1.5, 3.6) years, 389 (20.2%) patients met the primary composite endpoint (MACCE), including 33 (1.7%) cardiac deaths, 50 (2.6%) non-fatal MI, 159 (8.3%) ischemia-driven revascularizations, 272 (14.1%) hospitalizations for UA, 21 (1.1%) hospitalizations for heart failure and 43 (2.2%) strokes.
The risk of MACCE was similar in patients with and without revascularization (adjusted HR 1.110; 95% CI 0.881–1.399; P=0.376). The incidence of hospitalization for UA (adjusted HR 1.376; 95% CI 1.033–1.833; P=0.029) and ischemia-driven revascularizations (adjusted HR 1.671; 95% CI 1.117–2.499; P=0.012) were higher in patients undergoing revascularization than in patients without revascularization (Supplementary Table 2).
Outcomes of OSA vs. Non-OSA Groups in the Overall Population and Those With and Without RevascularizationIn the overall population, the presence of OSA significantly predicted MACCE in the multivariable model excluding revascularization (adjusted HR 1.28; 95% CI 1.03–1.59; P=0.023) and in the model including revascularization (adjusted HR 1.28; 95% CI 1.03–1.58; P=0.026) (other variables were age, sex, BMI, current smoking, hypertension, diabetes, dyslipidemia, prior stroke, prior MI, history of revascularization, diagnosis, and OMT at discharge).
The crude number of MACCE between the OSA vs. non-OSA groups in patients with or without revascularization is shown in Supplementary Table 3. After dividing the population according to whether or not to receive revascularization, a higher incidence of MACCE (22.6% vs. 14.6%, log-rank P=0.004), hospitalization for ACS (17.4% vs. 10.9%, log-rank P=0.008) and hospitalization for UA (14.2% vs. 8.6%, log-rank P=0.012) was observed in the OSA group than in the non-OSA group among patients without revascularization (Figure 2). In the multivariable Cox-regression analysis, OSA was an independent predictor of MACCE after adjusting for potential confounders in patients without revascularization (HR 1.861; 95% CI 1.239–2.796; P=0.003) but not in patients with revascularization (HR 1.135; 95% CI 0.882–1.460; P=0.324). We also did a Cox-regression analysis using AHI as a continuous variable and did not find a significant association with MACCE in patients with revascularization (HR=1.00, 95% CI 0.99–1.01, P=0.82) or without revascularization (HR=1.01, 95% CI 1.00–1.02, P=0.056). The incremental risk associated with OSA in the non-revascularization group might be attributed to more hospitalizations for ACS (HR 1.831; 95% CI 1.149–2.917; P=0.011) or UA (HR 1.896; 95% CI 1.124–3.199; P=0.016). There was no significant difference in the incidence of cardiovascular death, MI, ischemic stroke, and ischemia-driven revascularization between the OSA and non-OSA groups in both the revascularization and non-revascularization groups (Table 3). The different risk of MACCE (P for interaction=0.112), hospitalization for ACS (P for interaction=0.078), hospitalization for UA (P for interaction=0.058) and other secondary endpoints between the OSA and non-OSA groups, according to categories of revascularization, showed no significant interaction.
Kaplan-Meier curves for (A) MACCE, (B) hospitalization for ACS, and (C) hospitalization for UA between the OSA and non-OSA groups in patients with or without revascularization. ACS, acute coronary syndrome; MACCE, major adverse cardiovascular and cerebrovascular event; OSA, obstructive sleep apnea; UA, unstable angina.
| Endpoints | Non-revascularization | |||||
|---|---|---|---|---|---|---|
| Univariable model | Multivariable model 1 | Multivariable model 2 | ||||
| HR (95% CI ) | P value | HR (95% CI) | P value | HR (95% CI) | P value | |
| MACCE | 1.750 (1.193, 2.568) | 0.004 | 1.755 (1.171, 2.630) | 0.006 | 1.861 (1.239, 2.796) | 0.003 |
| Cardiovascular death | 1.092 (0.383, 3.114) | 0.869 | 0.870 (0.290, 2.615) | 0.805 | 0.967 (0.322, 2.908) | 0.953 |
| Hospitalization for ACS | 1.795 (1.155, 2.787) | 0.009 | 1.790 (1.126, 2.846) | 0.014 | 1.831 (1.149, 2.917) | 0.011 |
| Non-fatal MI | 1.735 (0.708, 4.248) | 0.228 | 1.736 (0.676, 4.457) | 0.251 | 1.795 (0.697, 4.626) | 0.226 |
| UA | 1.859 (1.136, 3.040) | 0.014 | 1.842 (1.096, 3.098) | 0.021 | 1.896 (1.124, 3.199) | 0.016 |
| Stroke | 1.292 (0.468, 3.565) | 0.621 | 1.317 (0.448, 3.872) | 0.617 | 1.401 (0.474, 4.140) | 0.542 |
| Ischemia-driven revascularization |
1.679 (0.829, 3.402) | 0.150 | 1.610 (0.759, 3.418) | 0.215 | 1.636 (0.767, 3.488) | 0.202 |
| Hospitalization for heart failure |
0.897 (0.201, 4.013) | 0.887 | 0.709 (0.144, 3.485) | 0.672 | 0.559 (0.101, 3.099) | 0.506 |
| Endpoints | Revascularization | |||||
| Univariable model | Multivariable model 1 | Multivariable model 2 | ||||
| HR (95% CI) | P value | HR (95% CI) | P value | HR (95% CI) | P value | |
| MACCE | 1.213 (0.956, 1.537) | 0.111 | 1.157 (0.900, 1.486) | 0.255 | 1.135 (0.882, 1.460) | 0.324 |
| Cardiovascular death | 1.469 (0.578, 3.731) | 0.419 | 1.272 (0.480, 3.370) | 0.629 | 1.173 (0.431, 3.197) | 0.755 |
| Hospitalization for ACS | 1.140 (0.877, 1.482) | 0.326 | 1.089 (0.826, 1.436) | 0.545 | 1.058 (0.801, 1.398) | 0.689 |
| Non-fatal MI | 1.705 (0.798, 3.645) | 0.168 | 1.550 (0.679, 3.538) | 0.298 | 1.422 (0.620, 3.259) | 0.406 |
| UA | 1.097 (0.833, 1.445) | 0.509 | 1.065 (0.797, 1.424) | 0.669 | 1.039 (0.776, 1.392) | 0.796 |
| Stroke | 1.341 (0.628, 2.863) | 0.448 | 1.346 (0.609, 2.977) | 0.463 | 1.345 (0.601, 3.007) | 0.470 |
| Ischemia-driven revascularization |
1.255 (0.881, 1.787) | 0.209 | 1.208 (0.831, 1.756) | 0.322 | 1.153 (0.791, 1.680) | 0.460 |
| Hospitalization for heart failure |
1.117 (0.387, 3.222) | 0.838 | 1.009 (0.330, 3.083) | 0.988 | 0.911 (0.293, 2.828) | 0.872 |
Model 1: adjusted for age, sex, BMI, current smoking, history of hypertension, diabetes, dyslipidemia. Model 2: adjusted for Model 1 covariates and prior stroke, prior MI, history of revascularization, clinical presentation (ST-segment-elevation MI vs. non-ST-segment elevation ACS), and OMT at discharge. CI, confidence interval; HR, hazard ratio; MACCE, major adverse cardiovascular and cerebrovascular event; UA, unstable angina. Other abbreviations as in Tables 1,2.
In the sensitivity analyses, AHI ≥30, ODI ≥15 events/h, and T90 ≥median were not predictors of MACCE in either the non-revascularization or revascularization group (Figure 3A).
Kaplan-Meier curves for (A) MACCE based on other OSA-related characteristics in patients with or without revascularization, and (B) MACCE between the OSA and non-OSA groups stratified by different subgroups. CI, confidence interval; HR, hazard ratio; MACCE, major adverse cardiovascular and cerebrovascular event; ODI, oxygen desaturation index; OSA, obstructive sleep apnea; T90, percentage of time with SaO2 <90%.
When considering the non-revascularization group according to the presence or absence of any lesion with ≥50% diameter stenosis, the association between OSA and the primary endpoint reached statistical significance in the lesion subgroup (HR 1.576; 95% CI 1.028–2.416; P=0.037), but not in the non-lesion subgroup (HR 2.189; 95% CI 0.640–7.480; P=0.212) (Figure 3B).
When considering the revascularization group according to PCI or CABG, OSA was not a predictor of the primary endpoint in the PCI and CABG subgroups. When considering the revascularization group according to STEMI or non-STE-ACS, OSA was not a predictor of the primary endpoint in either group (Figure 3B).
The main findings of the present study are as follows. More than 30% of patients in this ACS cohort did not undergo revascularization. At a median follow-up of 2.9 years, OSA was associated with an increased risk of MACCE in the non-revascularization group, but was not evident in the revascularization group, although no significant revascularization by OSA interaction was noted. The incremental risk associated with OSA in patients without revascularization might be explained by more rehospitalizations for UA. When further dividing the non-revascularization group into the subgroups with and without any lesion with ≥50% diameter stenosis, the association between OSA and the primary endpoint reached statistical significance in the subgroup with any lesion with ≥50% diameter stenosis.
Although revascularization is the most effective treatment for ACS, a proportion of patients did not undergo it. In the present study, 52% of patients with ≥70% stenosis in the non-revascularization group did not undergo revascularization for the following reasons. First, according to the Chinese PCI guideline,26 only those with significant anatomic (>50% left main or >70% non-left main CAD) or physiological (FFR <0.80) stenosis are candidates for revascularization. Thus, most patients (>50%) in this subgroup only had 70% stenosis in ≥1 vessel. Second, for patients with >70% stenosis, most had significant stenosis in branch vessels (<2 mm) or the distal segment of main epicardial vessels. Third, some patients with significant stenosis did not undergo revascularization due to rejection by the patient or their relatives or because of contraindications (e.g., severe bleeding or thrombocytopenia).
In patients undergoing revascularization, recurrent angina after PCI is well recognized and may affect 20–40% of patients during the short- to medium-term follow-up, even when PCI is optimized using physiological or imaging-guided approaches.10 The mechanisms include structural (residual disease or disease progression, in-stent restenosis, diffuse atherosclerosis, etc.) or functional changes (epicardial coronary spasm or microvascular dysfunction) of the coronary circulation.10 Prevention of the progression of non-culprit lesions may be warranted for both patients with and without revascularization. However, the characteristics of patients without revascularization differ significantly from those of patients with revascularization and may relate to the different proportions of events. The effects of OSA on the prognosis of these 2 groups may be also inconsistent.
Effect of OSA on Patients With RevascularizationOur data suggested that OSA was not an independent risk factor of recurrent cardiovascular events in ACS patients undergoing revascularization. Several observational studies reported inconsistent results regarding the association of OSA with cardiovascular events in patients after PCI.12,13,32,33 However, either small sample size (<300) or short follow-up period (<1 year) may have resulted in inadequate statistical power. Also, in the early period, the usage rate of bare metal stents was high, with a high risk of restenosis. The effect of OSA on stent-related events may be more obvious.
The Sleep and Stent Study was a large observational study that suggested OSA is independently associated with subsequent MACCE in patients undergoing PCI.14 In that study, >30% of patients were diagnosed with stable angina, 13% of patients were implanted with bioresorbable vascular scaffold or bare metal stents, and the secondary events did not include hospitalization for UA.14 In our study with >1,900 patients enrolled between 2015 and 2020, all patients were admitted for ACS and all implanted stents were drug-eluting stents. Hospitalization for UA was one of the secondary endpoints with a high incidence. The differences in the baseline clinical and procedural characteristics could be another reason for the different results. Moreover, the Sleep and Stent Study did not show significant differences between the OSA and non-OSA groups in the incidence of any individual secondary events, including stent-related adverse events (target vessel revascularization and stent thrombosis),14 which was consistent with our study population with revascularization. The impact of OSA on stent-related events may be hidden in the context of contemporary revascularization strategy and OMT.
Effects of OSA on Patients Without RevascularizationIn non-revascularization group, OSA was associated with an increased risk of MACCE mainly due to a higher risk of hospitalization for UA. To our knowledge, this is the first study focusing on the effect of OSA in ACS patients without revascularization. The primary reason for not undergoing revascularization was the absence of indication for PCI or CABG, which does not mean the absence of lesions. Our data showed that atherosclerotic lesions (diameter stenosis ≥50% of any coronary artery) were observed in 78% of patients without revascularization, and severe lesions (diameter stenosis ≥70% of any coronary artery) were in 52% of patients. In the non-revascularization group, OSA patients were found to have a higher incidence of severely diseased vessels (≥70% stenosis) and numerically high incidence of diseased vessels (≥50% stenosis), thus contributing to the lower incidence of MINOCA in the OSA vs. non-OSA groups. Also, the sample size of STEMI patients was small in the non-revascularization group, which may have resulted in the statistically significant result. OSA is reported to facilitate the progression of coronary atherosclerosis, and increase plaque burden and plaque instability,34,35 which can be attributed to sympathetic activation, increased oxidative stress, proinflammatory responses, endothelial dysfunction, and platelet activation.36–38 Meanwhile, ≈3 years of follow-up may be not long enough for plaque progression to the point of requiring revascularization, which may explain why OSA increased the incidence of hospitalization for UA but not ischemia-driven revascularization.
OSA may also be a predisposing factor for ischemia without obstructive CAD (<50% diameter stenosis) including coronary artery spasm (CAS) and CMD, which could cause vasospastic angina and microvascular angina, respectively.39–41 Many patients with angina symptoms do not have significant stenosis. The prevalence of coronary vasomotor disorders is ≈50% in patients with angina, and the frequency of multiple coronary spasm (≥2 spastic arteries) by provocative testing in the Asian population is 19–24%.42 CMD affects up to 50% of subjects with chronic coronary syndromes, and up to 20% of those with ACS.43 Both CAS and CMD are less likely to be correctly diagnosed by invasive coronary angiography and patients thus unable to receive revascularization therapy.43,44 Systemic inflammation, platelet activation, and autonomic dysfunction attributed to OSA are important factors contributing to CAS and CMD.39,45 It might also explain why OSA increased the risk of hospitalization for UA but not ischemia-driven revascularization in patients without severe obstructive lesions.
Although coronary artery lesions were more severe in patients who required revascularization than those who did not, it does not mean postoperative residual stenosis in the revascularization group was still more severe than the untreated stenosis in non-revascularization group. In addition, the usage rate of OMT was significantly higher in the revascularization group than in the non-revascularization group. The detrimental effect of OSA on cardiovascular outcomes may be attenuated in patients with revascularization. For ACS patients with OSA, complete functional or anatomic revascularization may be needed to prevent future events.46 For patients with intermediate stenosis (e.g., 50–70%), either physiological or intravascular imaging evaluation, or even a combination of these modalities, is needed to identify ischemic or high-risk lesions that may benefit from revascularization. In the case of incomplete revascularization, close follow-up and intensive medical therapy after discharge are necessary. Furthermore, OSA treatment may be needed for those without revascularization, although more studies are warranted.
Study LimitationsFirst, OSA was diagnosed by portable polygraphy, which may underestimate AHI as a result of overestimating the actual sleeping time. Second, although OSA severity may be overestimated during the acute setting of ACS, this is true for OSA assessment in the setting of acute disease including heart failure. Also, the sleep study was performed after clinical stabilization during hospitalization. Third, portable polygraphy cannot capture the hypoxic burden and heart rate and related heart rate response, which are also related to cardiovascular events. Fourth, the proportion of complete revascularization and drug compliance was inaccessible in this database, which may have effects on the outcomes. Fourth, this study could not analyze the detailed information (e.g., CAS or CMD) in patients without obstructive lesions. The hypotheses that explain the association of OSA with outcomes in this subgroup remain to be verified.
For ACS patients in the present study, OSA was associated with a higher risk of subsequent cardiovascular events in patients without revascularization, but not in those undergoing revascularization. These findings highlight the importance of identifying OSA in ACS patients without revascularization, and the possible benefits of suitable OSA treatment for this high-risk subset need further investigation.
The authors thank Drs. Guanqi Zhao, Zexuan Li, Xin Huang, Siyi Li, Ge Wang (Center for Coronary Artery Disease, Division of Cardiology, Beijing Anzhen Hospital, Capital Medical University, Beijing, China) for study data collection.
S.N.: research grants to the institution from Boston Scientific, Abbott, Jiangsu Hengrui Pharmaceuticals, China Resources Sanjiu Medical & Pharmaceuticals, East China Pharmaceuticals. The remaining authors have no relevant relationships to disclose.
Not applicable.
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
This work was supported by grants from National Natural Science Foundation of China (82200495); Natural Science Foundation of Beijing, China (7222046); Beijing Nova Program (Z201100006820087); and National Key R&D Program of China (2022YFC2505600, 2020YFC2004800).
Study concept and design: J.D., S.N. Acquisition, analysis, or interpretation of data: Y.Z., W.H., J.F., R.G., X.W., J.D., S.N. Drafting of the manuscript: Y.Z. Critical revision of the manuscript for important intellectual content: All authors. Obtained funding: X.W., S.N. Administrative, technical, or material support: H.A., B.Q., S.N. S.N. had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. All authors read and approved the final manuscript.
The study was approved by the Ethics Committee of Beijing Anzhen Hospital, Capital Medical University (2013025).
Please find supplementary file(s);
https://doi.org/10.1253/circj.CJ-23-0164
