Blood Flow of the Middle Cerebral Artery With Sleep-Disordered Breathing: Correlation With Obstructive Hypopneas
To the Editor:
Netzer et al1 recently reported that obstructive sleep hypopnea and apnea are associated with reductions in middle cerebral artery (MCA) blood flow compared with central sleep apnea. This study may explain the increased incidence of ischemic stroke in obstructive sleep disorders,2 3 4 although it is important to remember that other mechanisms independent of cerebral hemodynamics have also been proposed.5 6
Transcranial Doppler blood flow measurements of the MCA are assumed to reflect total brain flow, but it should be noted that this technique does not evaluate regional cerebral blood flow (CBF). Several lines of evidence suggest that localized disturbances in CBF may be important in patients who sustain cerebrovascular events in association with sleep disorders. Non-REM sleep in normal subjects is associated with a global reduction in brain blood flow (measured by stable xenon computed tomography) and global glucose hypometabolism (measured by positron emission tomography). Against this global fall in CBF are important regional variations, with some brain regions (frontal cortex, basal ganglia, thalamus, pons, cerebellum) affected to a greater degree while others (temporal cortex) are relatively unaffected.7 8 During REM sleep there is increased CBF to the associated visual area, presumably reflecting the activation involved in processing visual dream experiences, with a reduction in inferior frontal cortical flow.9 Preliminary observations in sleep apnea syndromes indicate the presence of focal cortical perfusion abnormalities in the awake or apneic state that are reversed with therapeutic transnasal continuous positive airway pressure (CPAP).10 11
We recently assessed regional CBF in 7 patients with severe obstructive sleep apnea (OSA) using single-photon emission computed tomography (SPECT) imaging and the cerebral flow tracer [99mTc]ethyl cysteinate dimer (ECD; Neurolite, DuPont Pharma). The diagnosis of OSA was made from overnight polysomnography (apnea-hypopnea index of >40). Subjects had no other sleep disorders, no past history of cerebral vascular or cardiovascular disease, and normal findings on neurological examination, and they were not taking any medications. All patients underwent 2 morning sleep studies of 3 hours’ duration. For the first examination, [99mTc]ECD was injected intravenously at the onset of an apneic episode, and the patient was scanned 1 hour later. The exact time of injection, sleep stage, subsequent arousal, oxygen saturation, and total and postinjection duration of the desaturation episode (oxygen saturation of <85%) were carefully noted. The same procedure was repeated with the second sleep study, except that nasal CPAP (mean, 14 cm water; range, 10–20 cm) was applied throughout the study. [99mTc]ECD was injected in the sleep study at the same time as with the apneic scan. Patients underwent awake brain scanning on a third occasion. Six normal control subjects were recruited from among hospital employees, and polysomnography confirmed a normal sleep pattern. [99mTc]ECD injection was performed after 2 hours of normal sleep, with SPECT scanning 1 hour later. An awake SPECT scan was performed on a second occasion.
Subjects were imaged on a dual-head camera (Helix, Elscint Ltd) fitted with high-resolution, low-energy parallel-hole collimators and a custom-built SPECT headholder. This imaging system has a reconstructed resolution of 9 mm in air of the center of a 15-cm radius of rotation. The scans were reconstructed and reoriented parallel to the orbitomeatal plane with use of external fiduciary markers. A set of region-of-interest (ROI) templates defined from cross-sectional neuroanotomical atlas were superimposed on the patients’ scans with a technique that requires minimal operator intervention. Six transaxial slices (slice thickness interpolated to 1 cm) were used to extract average ROI activity for temporal cortex, parietal cortex, frontal cortex, occipital cortex, anterior cingulate, striatal nuclei, thalamus, and cerebellum. Regional flow measures were normalized to cerebellar activity and to whole cortex activity at the midstriatal level.
The following preplanned comparisons were performed: sleep (control) versus sleep-apneic (OSA), sleep-apneic (OSA) versus sleep-nonapneic (OSA with CPAP), and awake (control) versus awake (OSA). After Bonferroni adjustment for multiple comparisons, no significant differences were identified for any of the brain ROIs studied. Visual assessment of the scans confirmed a normal symmetrical pattern of cortical, subcortical, and cerebellar perfusion.
Although our findings are based on a small number of subjects, we failed to find any significant regional disturbance in CBF during apnea or in the awake state in patients with OSA. Our observations complement the recent study by Netzer et al1 and suggest that global rather than regional changes in brain hemodynamics are likely to be more important in patients with obstructive sleep disorders.
This work was supported in part through clinical grant 95031 from DuPont Pharma.
- Copyright © 1999 by American Heart Association
Netzer N, Werner P, Jochums I, Lehmann M, Strohl KP. Blood flow of the middle cerebral artery with sleep disordered breathing: correlation with obstructive hypopneas. Stroke.. 1998;29:87–93.
Koskenvuo M, Kapnio J, Talakivi T, Partinen M, Heikkila K, Sarna S. Snoring as a risk factor for stroke in men. Br Med J. 1987;294:16–19.
Dyken ME, Somers VK, Yamada T, Ren Z-Y, Zimmerman B. Investigating the relationship between stroke and obstructive sleep apnea. Stroke.. 1996;27:401–407.
Dannenberg C, Bosse-Henck A, Barthel H, Bettin S, Sattler B, Knapp WH. Baseline and activation studies using Tc-99m-ECD-SPECT in patients with severe sleep apnea syndrome. Eur J Nucl Med.. 1996;23:1201.
Feistel H, Merkl M, Siegfied W, Möller C, Dertinger S, Ficker JH, Platsch G, Hahn EG, Wolf F. Brain perfusion during sleep apnea: a study with Tc-99m-HMPAO in sleep laboratory. Eur J Nucl Med. 1994;21:770.
Drs Leslie, Wali, and Kryger identify key elements and critical issues about our studyR1 and its relevance to understanding the relationships between sleep apnea and stroke. First, we agree that the technique of transcranial Doppler flow measurements of the MCA provides no information on regional blood flow, nor can its use in our study predict brain dysfunction or stroke in sleep apnea. We were careful to state that the method provided information concerning dynamic changes in vascular pressures and flows occurring in the MCA with repetitive obstructive apneas and obstructive hypopneas (heavy snoring). Our conclusions were limited to differences in the physiological events among apnea types.
Second, the relevance of studies of cerebral function in sleep apnea are to the cerebrovascular and, possibly, neurocognitive sequelae of sleep apnea.R2 Here data are scant. There are epidemiological studies showing association, and studies like oursR1 and that reported by Leslie et al that perform physiological studies in patients. There is a need for outcome and observational epidemiological studies of cerebrovascular control that might permit insight into the pathophysiology of stroke in the setting of sleep apnea. Of interest, animal studies suggest that exposure to hypoxia will increase brain microvascularity, and as a result, reduce the neuroanatomical consequences of experimental ischemic stroke.R3
Third, the instruments to assess the brain and its vascularity are currently focused on anatomic regions or large vascular structures, and interpretation of results is made without much regard for microvascular events, such as the coordination of localized blood flow to neuronal activity. The clinical vectors to produce cerebral dysfunction in sleep apnea are complex, given the state-related nature of the illness, the attendant intermittent and often profound hypoxemia and alterations in cardiovascular function, and the global consequences of repetitive arousal on daytime performance. Neurocognitive testing in sleep-apnea patients reveals deficits that correlate only moderately well with nighttime events.R2 The study reported by Leslie et al nicely illustrates a similar phenomenon, namely, the SPECT imaging shows no regional or global differences between patients with severe degrees of apnea (>40 per hour of sleep) and healthy control subjects or in the same patients before and with definitive treatment, after controlling for state. Other factors to consider that potentially confound measures of brain function and affect the power of any study of sleep apnea are markers of age, comorbidity, and length of illness. As our study suggests, the equivalent effects of heavy snoring and obstructive apnea on MCA flow profiles and the observation that central apneas show less dynamic impact may be important factors to consider in determining the effect size of sleep-disordered breathing on cerebrovascular function and illness.
Netzer N, Werner P, Jochums I, Lehmann M, Strohl KP. Blood flow of the middle cerebral artery with sleep disordered breathing: correlation with obstructive apneas. Stroke. 1989;29:87–93.
American Thoracic Society/American Sleep Disorders Association. Statement on health outcomes research in sleep apnea. Am J Respir Crit Care Med.. 1998;157:335–341.