Noninvasive Ventilatory Correction in Patients With Acute Ischemic Stroke
A Systematic Review and Meta-Analysis
Background and Purpose—Even though current guidelines suggest that noninvasive ventilatory correction (NIVC) could be considered for acute ischemic stroke patients with obstructive sleep apnea, available evidence is conflicting, with no adequately powered randomized clinical trial being available to date.
Methods—We conducted a systematic review and meta-analysis of all available literature data evaluating the effect of NIVC on neurological improvement (based on decrease in National Institutes of Health Stroke Scale score), vascular events (recurrent stroke, transient ischemic attack, myocardial infarction and unstable angina), and mortality during the follow-up period.
Results—We identified 4 randomized clinical trials and 1 prospectively matched observational cohort, comprising a total of 389 patients (59.8% males, mean age: 64.4 years). The risk of both performance and detection bias was considered high in most of the included randomized clinical trials because of the lack of blinding in participants, personnel and/or outcome assessors. The mean decrease in National Institutes of Health Stroke Scale scores during the first (≤30) days of acute ischemic stroke was found to be greater in NIVC-treated patients in comparison to controls (standardized mean difference, 0.38; 95% confidence interval, 0.11–0.66; P=0.007). However, no significant differences were detected between NIVC-treated acute ischemic stroke patients and controls on both the risk of vascular events (risk ratio, 0.53; 95% confidence interval, 0.25–1.14; P=0.11) and mortality (risk ratio, 0.71; 95% confidence interval, 0.37–1.36; P=0.30). No evidence of heterogeneity (I2=0%; P for Cochran Q>0.50) or publication bias were detected in all analyses.
Conclusions—NIVC seems to be associated with greater short-term neurological improvement in acute ischemic stroke patients with obstructive sleep apnea. This finding deserves further investigation within the settings of an adequately powered, sham-control, randomized clinical trial.
- acute ischemic stroke
- bilevel positive airway pressure
- continuous positive airway pressure
- early neurological deterioration
- noninvasive ventilatory correction
Obstructive sleep apnea (OSA) is an independent risk factor for stroke and death from any cause.1 OSA is highly prevalent among acute ischemic stroke (AIS) patients and has been associated with worse functional outcomes and increased stroke recurrence.2 Even though current American Heart Association/American Stroke Association guidelines suggest that noninvasive ventilatory correction (NIVC) could be considered for AIS patients with OSA, available evidence is conflicting, with no adequately powered randomized controlled clinical trial (RCT) being available to date.3
In view of the aforementioned considerations, we conducted a systematic review and meta-analysis of all available literature data evaluating the effect of NIVC on different outcomes of AIS patients.
This meta-analysis has adopted the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines for systematic reviews and meta-analyses.4 We included all eligible RCTs and prospective observational studies of AIS patients receiving NIVC, with either continuous positive airway pressure or bilevel positive airway pressure. The outcome events of interest were changes in stroke severity during the first 30 days of ictus (assessed using National Institutes of Health Stroke Scale scores), vascular events (recurrent stroke, transient ischemic attack, myocardial infarction, and unstable angina), and deaths during the follow-up period between NIVC and control groups.
Unadjusted mean differences of reported National Institutes of Health Stroke Scale changes between NIVC and control groups were pooled as standardized mean differences, while random-effects model (DerSimonian Laird) was used to calculate the pooled standardized mean differences. We also performed sensitivity analyses for all reported outcomes of interest by including only data from included RCTs. Further details on the identification of included studies, the software used for statistical analyses, the assessment of heterogeneity, bias and funnel plot asymmetry are available in the online-only Data Supplement.
We identified 5 eligible studies (Figure 1),5–9 comprising a total of 389 patients (59.8% males, mean age 64.4 years). Details on literature search results, study characteristics, and bias assessments are available in the online-only Data Supplement.
The mean decrease in National Institutes of Health Stroke Scale scores during the first (≤30 days) of ictus was greater in NIVC-treated patients in comparison to controls (standardized mean difference, 0.38; 95% confidence interval [CI], 0.11–0.66; P=0.007; Figure 2A). No evidence of heterogeneity was found between estimates (I2=0%). In view of the small number (n=3) of included studies, inspection of funnel plot revealed no evidence of publication bias (Figure III in the online-only Data Supplement).
No significant differences were detected between NIVC-treated AIS patients and controls on both the risk of vascular events (risk ratio, 0.53; 95% CI, 0.25–1.14; P=0.11; Figure 2B) and mortality (risk ratio, 0.71; 95% CI, 0.37–1.36; P=0.30; Figure 2C) during the follow-up period, with no evidence of heterogeneity across included studies (I2=0%).
In the sensitivity analysis, after excluding non-RCT data from 1 study,9 the mean National Institutes of Health Stroke Scale decrease was greater in NIVC-treated patients in comparison to controls (standardized mean difference, 0.46; 95% CI, 0.03–0.89; P=0.04; Figure 3A) with no evidence of heterogeneity (I2=0%; P=0.89 by Cochran Q statistics). Similarly, no differences were detected between the 2 groups on the risk of vascular events (RR, 0.62; 95% CI, 0.26–1.48; P=0.28; Figure 3B) during the follow-up period, with no evidence of heterogeneity (I2=0).
The results of the present meta-analysis suggest that NIVC seems to be associated with greater improvement of neurological deficit during the first days after stroke onset. Even though a trend of lower risk for both recurrent vascular events and all-cause mortality was detected in NIVC-treated patients, these differences did not reach statistical significance.
Early neurological deterioration is common in AIS and may be attributed to hemodynamic steal phenomena, which in turn are more prevalent in patients with excessive sleepiness and OSA.10 This hemodynamic impairment in AIS patients with OSA may be improved using NVIC, and this pathophysiological mechanism may account for the greater neurological improvement during the first days of acute cerebral ischemia that was documented in patients randomized to NIVC compared with controls.
The lack of significant differences in vascular events and mortality during the follow-up period of available RCTs5–8 may be partially attributed to short follow-up, the small sample sizes, and the resulting power limitations because of difficulties in patient recruitment, adherence, and follow-up. The SATS trial (Sleep Apnea Treatment after Stroke) reported both adherence and success of sham continuous positive airway pressure on allocation masking in the AIS setting, and the methodology of SATS can be, therefore, used in future RCTs.11 Thus, improvements in continuous positive airway pressure compliance and recruitment of multidisciplinary teams with expertise in behavioral interventions are methodological approaches that could ensure patients’ compliance and drastically reduce losses to follow-up.
Certain limitations of the present report need to be acknowledged. First and most important, a small number of studies with little overlap in reported outcomes was included in this meta-analysis. This heterogeneity of reported outcomes between included studies could be attributed to the different study designs (RCTs versus matched cohorts), definitions of OSA, durations of follow-up for outcome assessments (ranging from 1 to 60 months), and primary outcomes (short-term5,8,9 versus long-term6,7 end points) in each study protocol (Table II in the online-only Data Supplement). Given the findings of the present report and the available literature data on hemodynamic steal phenomena in OSA patients during the acute phase,10 we consider that future trials should use early neurological improvement as their primary end point. Notably, early neurological improvement was the only end point that was significantly associated with NIVC both in overall and sensitivity analyses without any evidence of heterogeneity across included studies.
Another methodological shortcoming of the present study is related to the fact that poor NIVC compliance or tolerability are common in stroke patients and may adversely affect the successful initiation and maintenance of NIVC therapy.1 It should be noted that high withdrawal rates (22%) during follow-up were reported in 2 of the study protocols,5,6 while no data on withdrawals/losses to follow-up were provided in the other studies (Table III in the online-only Data Supplement). Because reported study data are unadjusted for losses to follow-up, withdrawals, and poor NIVC compliance, the results of the present meta-analysis need to be interpreted with caution, while these potential confounders are of paramount importance for the successful design of future study protocols. Moreover, the NIVC treatment duration and the time frame of early neurological improvement varied between included studies (Table III in the online-only Data Supplement). Finally, one RCT included patients with ischemic (67%) and hemorrhagic (33%) stroke, and none of the included studies focused on AIS patients with intracranial occlusions who might be at highest risk of OSA-related neurological deterioration and possibly benefit most from early NIVC.
In conclusion, given the low power of available studies and the reported poor compliance rates, the benefit of NIVC in short-term neurological improvement of patients with AIS and OSA deserves further investigation within the settings of an adequately powered sham-control RCT.
Sources of Funding
Dr Mikulik has been supported by the project no. LQ1605 from the National Program of Sustainability II (MEYS CR) and by the project FNUSA-ICRC no. CZ.1.05/1.1.00/02.0123 (OP VaVpI).
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.117.017661/-/DC1.
- Received January 8, 2017.
- Revision received April 8, 2017.
- Accepted May 2, 2017.
- © 2017 American Heart Association, Inc.
- Lyons OD,
- Ryan CM
- Kernan WN,
- Ovbiagele B,
- Black HR,
- Bravata DM,
- Chimowitz MI,
- Ezekowitz MD,
- et al
- Liberati A,
- Altman DG,
- Tetzlaff J,
- Mulrow C,
- Gøtzsche PC,
- Ioannidis JP,
- et al
- Aaronson JA,
- Hofman WF,
- van Bennekom CA,
- van Bezeij T,
- van den Aardweg JG,
- Groet E,
- et al
- Parra O,
- Sánchez-Armengol Á,
- Capote F,
- Bonnin M,
- Arboix A,
- Campos-Rodríguez F,
- et al
- Minnerup J,
- Ritter MA,
- Wersching H,
- Kemmling A,
- Okegwo A,
- Schmidt A,
- et al
- Tsivgoulis G,
- Zhang Y,
- Alexandrov AW,
- Harrigan MR,
- Sisson A,
- Zhao L,
- et al
- Alexandrov AV,
- Nguyen HT,
- Rubiera M,
- Alexandrov AW,
- Zhao L,
- Heliopoulos I,
- et al