Early brain swelling in acute hypoxia

Abstract

Acute mountain sickness (AMS) and high-altitude cerebral edema share common clinical characteristics, suggesting cerebral swelling may be an important factor in the pathophysiology of AMS. Hypoxia and hypocapnia associated with high altitude are known to exert strong effects on the control of the cerebral circulation, yet how these effects interact during acute hypoxia, and whether AMS-susceptible subjects may have a unique response, is still unclear. To test if self-identified AMS-susceptible individuals show altered brain swelling in response to acute hypoxia, we used quantitative arterial spin-labeling and volumetric MRI to measure cerebral blood flow and cerebrospinal fluid (CSF) volume changes during 40 min of acute hypoxia. We estimated changes in cerebral blood volume (CBV) (from changes in cerebral blood flow) and brain parenchyma swelling (from changes in CBV and CSF). Subjects with extensive high-altitude experience in two groups participated: self-identified AMS-susceptible (n = 6), who invariably experienced AMS at altitude, and self-identified AMS-resistant (n = 6), who almost never experienced symptoms. During 40-min hypoxia, intracranial CSF volume decreased significantly [-10.5 ml (SD 6.9), P < 0.001]. There were significant increases in CBV [+2.3 ml (SD 2.5), P < 0.005] and brain parenchyma volume [+8.2 ml (SD 6.4), P < 0.001]. However, there was no significant difference between self-identified AMS-susceptible and AMS-resistant groups for these acute-phase changes. In acute hypoxia, brain swelling occurs earlier than previously described, with significant shifts in intracranial CSF occurring as early as 40 min after exposure. These acute-phase changes are present in all individuals, irrespective of susceptibility to AMS.

Fig. 1.

T1- and T2-weighted image pair used to quantify cerebrospinal fluid (CSF) volume. A: whole brain, high-resolution, three-dimensional (3D) T1-weighed image is used to define a mask of intracranial CSF (bounded inferiorly at the foramen magnum). B: high-resolution 3D heavily T2-weighted image is dominated by signal from CSF (note: almost no residual signal from blood vessels, gray or white matter). The intracranial mask is applied to this image to quantify the intracranial volume of CSF.

Fig. 2.

Reproducibility of CSF measurement. Linear regression of CSF measurements was made at 2 sessions 2–13 days apart in 5 subjects. Regression line has slope 0.97 and intercept 5.8 (r = 0.97).

Fig. 3.

Change in CSF volume during 40-min normoxia following baseline measurements (n = 5) (measurements made at the same time intervals as the main study, but without hypoxia; see Fig. 4, top). There is an initial decrease in mean CSF volume in the first 20 min after the baseline measurement that is independent of inspired O₂. During the subsequent 20 min, there is no further CSF volume change. Error bars indicate ±1 SD.

Fig. 4.

Decrease in intracranial CSF volume and corresponding increase in intracranial space occupancy by blood and brain parenchyma during 20- and 40-min acute hypoxia. Top: both self-identified acute mountain sickness-susceptible (AMS-S) and -resistant (AMS-R) groups show a significant decrease in intracranial CSF volume following 20 min of hypoxia and a further decrease after 40 min. The decrease in CSF is indicative of increased space occupancy by other intracranial tissues (i.e., brain swelling and/or increased cerebral blood volume). Middle: calculated increase in steady-state cerebral blood volume from the steady-state cerebral blood flow change. Both AMS-S and AMS-R groups show increased blood volume following 20 min of hypoxia and further increases after 40 min. Bottom: calculated increase in brain volume (brain parenchyma swelling). Both groups show increased brain volume following 20 min of hypoxia and further increases after 40 min. The increased blood volume and brain swelling were not significantly different between the AMS-R and AMS-S groups. Error bars indicate ±1 SD.