Reversal of the neurological deficit in acute stroke with the signal of efficacy trial of auto-BPAP to limit damage from suspected sleep apnea (Reverse-STEAL): study protocol for a randomized controlled trial

Stroke still constitutes a leading cause of death and ranks number one among all causes of serious long-term adult disability in Europe and the United States [1, 2], underscoring the need for adjuvant acute and preventive stroke therapies that reduce both the devastating human and economic burden of this disease.

More than two-thirds of acute ischemic stroke (AIS) patients suffer from sleep apnea of variable degree compared to 5 to 15% in the general population [3‐6]. Moreover, in a prospective cohort study, patients with moderate to severe obstructive sleep apnea (OSA) had an almost two-fold increased risk for stroke and death after adjustment for other traditional vascular risk factors, thus establishing OSA as an independent risk factor [7, 8]. Following stroke, sleep apnea negatively affects the length of hospitalization and short-term as well as long-term outcomes, likely through its association with an increased risk of neurological deterioration during acute and subacute phases of stroke [4, 9‐12]. Neurological deterioration during the early phase of AIS can occur between one and two out of five patients and it is independently predicted by sleep apnea and associated with unfavorable outcome [9, 13].

Numerous studies including serial transcranial Doppler (TCD) in patients with AIS provided evidence that persisting large-artery steno-occlusive disease and corresponding collateral failure, leading to depletion of oxygen and nutrients to penumbral tissues, appear to be the most relevant pathogenic mechanisms for neurological deterioration [14‐17]. Thus, it seems intuitive that further hemodynamic and perfusion derangements may do harm by facilitating hypoperfusion and ultimately lead to expansion of infarction to penumbral areas (or even oligemic tissues surrounding the ischemic penumbra) with neurological deterioration as the clinical correlate [18].

Several pathophysiological conditions associated with sleep apnea are believed to account for acute harmful effects on ischemic brain [19]. Recurrent nocturnal apneic episodes promptly entail intermittent hypoxemia, intrathoracic pressure changes and activation of the sympathetic nervous system resulting in cardiac arrhythmias and marked blood pressure swings [20, 21]. The cerebral autoregulatory system is essential to compensate for changes in systemic hemodynamics and to avoid harmful hypo- or hyperperfusion of brain tissue. The fate of ischemic brain tissues particularly in the setting of large steno-occlusive disease already depend on dilated arteries with no further or only minimal residual vasomotor capacity [22‐24] to counteract apnea-associated blood pressure changes and maintain blood flow according to local metabolic needs [19]. In fact, patients with sleep apnea generally have impaired cerebral vasoreactivity, even during wakefulness, potentially explaining their particular vulnerability to stroke-associated hemodynamic derangements [25‐28]. Repetitive oxygen desaturations in cerebral ischemic tissue during apneic periods may further compromise functional and structural integrity of neuronal cells [29]. Additionally, hypercapnia may lead to depletion of the collateral blood flow when vessels in the non-ischemic area dilate more leading to blood flow diversion from the ischemic area to the non-ischemic areas, possibly further worsening hypoperfusion in the ischemic brain area [23, 30, 31]. This ‘cerebral blood flow steal’ phenomenon was demonstrated by TCD in real time in AIS patients during voluntary breath holding and led to early neurological deterioration in those with acute proximal arterial obstructions and excessive daytime sleepiness [10]. When found during the initial hospitalization for AIS, this so called ‘reversed Robin Hood syndrome’ led to a four-fold increase in stroke recurrence within the same arterial territory [11].

Maintenance of stable systemic and cerebral hemodynamics, as well as adequate and constant brain tissue oxygenation, plays a crucial role in the management of AIS [32]. The ischemic penumbra may be present for many hours and be amenable to reperfusion. Consequently, any treatment that potentially maintains or augments cerebral perfusion may be justified [33]. Non-invasive ventilation with either continuous (CPAP) or bilevel positive airway pressure (BPAP) is the treatment of choice for sleep apnea. Although some studies have shown the safety and beneficial effects of CPAP and BPAP by improvement of quality of life and reduction of cardio- and cerebrovascular morbidity and mortality in the general sleep apnea population [34, 35], their role in AIS patients remains to be elucidated. Most previous investigations were conducted beyond the (hyper-) acute phase of ischemic stroke and pursued secondary stroke prevention goals rather than acute effects. For example, two studies confirmed safety and tolerability, and pointed to a potential short-term treatment effect of prompt initiation of non-invasive ventilation in the acute phase of stroke when hemodynamically compromised tissue can potentially be salvaged [36, 37]. In the first study, an observational study, AIS patients with proximal persisting arterial occlusions and suspected or known OSA had a tendency towards early neurological improvement (median National Institutes of Health Stroke Scale (NIHSS) score decrease by 2 points) during the hospital course when BPAP was initiated within the first 24 hours of admission, as compared with those who did not receive BPAP (median NIHSS score decrease by 1 point, P = 0.078) [36]. The other study randomly assigned fifty patients in the first night after AIS to either CPAP for three nights or control [37]. The investigators found that CPAP treatment was feasible, safe and associated with a greater neurological improvement until day eight in patients who tolerated CPAP well when compared with control patients (NIHSS 2.3 versus 1.4, P = 0.022).

As data on the use of non-invasive ventilation in the acute phase of ischemic stroke are lacking and its potential benefit on stroke outcomes is controversial [38], a randomized controlled efficacy trial appears justified as being timely and needed. The choice of using an auto-titrating BPAP ventilation mode is based on the fact that AIS patients will not have the necessary time or access to facilities to undergo comprehensive sleep laboratory testing prior to start of treatment. Auto-BPAP adapts pressure settings automatically according to apnea frequency recorded in real-time from breath-to-breath, thus sparing manual titration of therapeutic pressures. Also, BPAP mode delivers pressure levels for inspiration and expiration separately, easing inspiration while ensuring alveolar ventilation and splinting the airway, which has several advantages over standard CPAP that applies a fixed pressure setting during the entire respiratory cycle [39]. Collectively, auto-BPAP may be superior particularly in patients who require higher pressure levels [39, 40] and produces improved patient adherence, a well-known issue in recent studies that may be even more important for stroke patients who may not have appropriate understanding and motivation to allow themselves to be exposed to mask and airflow pressures. Finally, this approach could be utilized in stroke centers without experience in sleep medicine.

We expect that this study will advance our understanding of how early treatment with non-invasive ventilatory treatment can counterbalance, or possibly reverse, the deleterious effects of sleep apnea in the acute phase of ischemic stroke. Our assumptions for sample size were based on a relatively small effect size as even slight improvements on the NIHSS score appear to be a strong predictor of good functional outcome after stroke [42]. The study will provide preliminary data to power a phase III study of auto-BPAP in the acute phase of cerebral ischemia.

At the time of manuscript submission, recruitment of study patients is ongoing. Initial recruitment was started in May 2013.

Ulf Bodechtel and Andrei V Alexandrov are the joined senior authorship.