HIGH ALTITUDE MEDICINE & BIOLOGY Volume 15, Number 2, 2014 ª Mary Ann Liebert, Inc. DOI: 10.1089/ham.2013.1151

Intracranial Pressure at Altitude Mark H. Wilson, MB, BChir, PhD, FRCS (SN),1,2,4 Alex Wright, MB, BChir, FRCP,2 and Christopher H.E. Imray, MB, BS, PhD, FRCS3

Abstract

Wilson, Mark H., Alex Wright, and Christopher H.E. Imray. Intracranial pressure at altitude. High Alt Med Biol. 15:123–132, 2014.— Rapid ascent to high altitude can result in high altitude headache, acute mountain sickness, and less commonly, high altitude cerebral or pulmonary edema. The exact mechanisms by which these clinical syndromes develop remain to be fully elucidated. Direct and indirect measures of intracranial pressure (ICP) usually demonstrate a rise in pressure when human subjects and animals are exposed to acute hypoxia. However, the correlation of ICP changes to symptoms and altitude-related illnesses has been difficult to establish. Headache, for example, may occur with vessel distension prior to a rise in ICP. This article reviews the literature both supporting and refuting an increase in ICP as the underlying mechanism of headaches and other related neurological sequelae experienced at high altitude.

Introduction

I

n 1783 the Edinburgh physician Alexander Monro (Monro, 1823) stated that as the substance of the brain, like that of the other solids of our body, is nearly incompressible the quantity of blood within the head must be the same, or very nearly the same, at all times. His student, George Kellie, subsequently demonstrated this at both animal and human autopsy (Kellie, 1824). Neither made any reference to cerebrospinal fluid (CSF), principally because, at this time, the scientific community did not believe in its existence. In 1842, the French physiologist Franc¸ois Magendie (1842) performed an experiment that involved puncturing the cisterna magna, and this convinced his peers of CSFs. Burrows, and subsequently Harvey Cushing, then incorporated CSF into the Monro-Kellie doctrine ( Reid, 1846; Burrows, 1848; Cushing, 1926) establishing the doctrine to how it is known today: that within a fixed skull, the total volume of brain, blood, and CSF is constant and if the volume of one constituent changes, there is a reciprocal change in one or both of the others. In reality, the dynamic mismatch of arterial supply and venous drainage, as originally described by Monro, is of greater clinical importance in the development of acute intracranial hypertension, although Burrows and Cushing’s amendment has detracted from this. The relationship between intracranial pressure and headache was originally demonstrated by Ray and Wolff

(1940). They mapped referred pain after stimulation of specific cerebral structures. Inflation of a balloon within the frontal horn of the lateral ventricle resulted in an ipsilateral frontal headache, whilst traction of the veins resulted in nausea and headache; whereas stimulation of the large arteries did not induce pain. Frontal headache appears to be a common feature of high altitude headache (Headache Classification Subcommittee of the International Headache Society, 2004). In 1985, Ross (1985) applied the Monro-Kellie doctrine to explain the random nature of mountain sickness. His ‘‘tight fit’’ hypothesis proposed that inter-individual variation in neuroaxis compliance (i.e., the inability of some to be able to accept brain swelling compared to others) accounted for a similar variability in susceptibility to acute mountain sickness (AMS) (Fig. 1). A number of factors influence such compliance: brain volume compared to skull volume (the atrophy of aging causes an increase in ventricular and sulcal size); and the volume of the spinal canal when compared to that of the spinal cord. A Pressure-Volume Index (PVI) can be calculated (the volume of fluid that needs to be added to CSF to raise the CSF pressure by a factor of 10) and this varies widely between individuals. For some 20 years, this ‘‘tight fit’’ hypothesis (in which it had been assumed that edematous brain increases intracranial pressure once the limits of compliance are reached) has been considered key to explaining differences in susceptibility to AMS, despite

1

The Brain Injury Centre—St Mary’s Hospital, Imperial College, London, United Kingdom. Birmingham Medical Research Expeditionary Society, Birmingham, United Kingdom. 3 University Hospital Coventry and Warwickshire NHS Trust and Warwick Medical School, Coventry, United Kingdom. 4 The Institute of Pre-Hospital Care, London’s Air Ambulance, Barts and the London Medical School, Queen Mary University of London, The Helipad, The Royal London Hospital, Whitechapel, United Kingdom. 2

123

124

FIG. 1. 2009.

WILSON ET AL.

Pressure/Volume curves representing compliant and noncompliant systems, from Wilson, Newman, and Imray,

paucity of supporting data. If it is a key factor, then increases in cerebral volume (and thus pressure) need also to be explained. Hackett and Roach (2001) summarized a physiological component to this baseline theory. Climbers who were more prone to hypoxemia (especially with exercise) would in turn have more cerebral swelling, hence a combination of this physiological predisposition with a lack of (anatomical predisposition) cerebral compliance could account for interindividual differences. However, the evidence to support these theories is not entirely consistent. This review article analyzes the different techniques that have been used to measure ICP in both hypobaric and normobaric hypoxia to build a profile of how much support there is. Human Studies Direct ICP monitoring

Invasive ICP monitoring is the gold standard, but is difficult to perform in the field at altitude, both on a practical and an ethical basis. A telemetric ICP study using implanted monitors in three climbers was undertaken on a high altitude expedition. (Wilson and Milledge, 2008). Only the youngest subject developed AMS. All of them had normal ICPs at rest at all altitudes. However the youngest subject suffered a dramatic rise in ICP at 4725 m during any form of mild exertion. Mild exertion and neck turning are both simple maneuvers that reduce venous outflow and hence could tip an individual from the ‘‘compensating’’ part of the cerebral compliance curve into the decompensated part, which results in a steep rise in ICP (Fig. 1). Unfortunately, this subject had a ventriculo-atrial shunt for mild hydrocephalus following a brain injury, and hence changes he exhibited may not reflect what happens ordinarily.

Indirect invasive measures of ICP Lumbar CSF pressures. The evidence for a rise in CSF pressures understandably comprises a small number of studies with relatively few subjects. Schaltenbrand (1933) found a rise in lumbar CSF pressures in several subjects during simulated high altitude, however the conditions and distress may have been confounders. Singh et al. (1969) demonstrated elevated lumbar spinal fluid pressures in 34 Indian soldiers who were rapidly transported from sea level to 5867 m. Lumbar CSF pressures were 6–21cm H2O higher than pressures after recovery. It is unclear whether these soldiers had AMS or high altitude cerebral edema (HACE). While Hartig and Hackett demonstrated, in three subjects, that acute hypoxic gas inhalation resulted in a rise in lumbar CSF pressure, more gradual decompression to a simulated ascent of 5000 m was not associated with any change, even in those in whom AMS developed (Hartig and Hackett, 1992). Similarly, Bailey et al. found lumbar CSF pressures to be normal after exposure to 16 hours of 12% O2 and identified no difference in pressures between AMS and non-AMS groups (Ba¨rtsch et al. 2004). Indirect noninvasive measures of ICP

A number of indirect noninvasive techniques have been developed that infer changes in ICP. However, none have yet proved sufficiently reliable to enter routine clinical practice. Tympanic membrane displacement (TMD). Wright et al. used a TMD technique to study 24 subjects before and during ascent to 5200 m (Wright et al., 1995). The results suggested a rise in inferred ICP with acute hypoxic exposure (ascending rapidly to 3400 m), but there was no additional rise in inferred ICP in those who developed mild to moderate AMS. It was concluded that a rise in ICP, if it occurs, is a late phenomenon in AMS/HACE. However, more recent studies have

INTRACRANIAL PRESSURE

questioned the ability of TMD to truly reflect ICP (Shimbles et al., 2005a; Gwer et al., 2013). The predictive limits of linear regression are too wide though it may be of use in serial measurements in one patient (Shimbles et al., 2005b). Distortion product otoacoustic emmisions (DPOAE).

DPOAE is another noninvasive indirect technique used for monitoring ICP (Sakka et al., 2012). The technique has had limited verification with relatively small numbers (Bu¨ki et al. 1996; 2000; Frank et al., 2000). An increase in ICP results in a decrease in DPOAE and attempts to use this method to measure ICP changes at altitude in a relatively small number of subjects have been made. Climbers ascending to 7400 m were studied and were found to exhibit changes in DPOAE. However as with TMD, the response to hypoxia itself may not permit discrimination of subjects with rise in ICP due to AMS or altitude gain (Olzowy et al., 2008) Optic nerve sheath diameter (ONSD). ONSD has been used as a surrogate for the measurement of ICP at altitude and in those with AMS (Fagenholz et al., 2007; Sutherland et al., 2008). In a study of 13 mountaineers, ONSD was measured ultrasonically with regression analysis used to explore correlation with a number of variables. ONSD was positively associated with increasing altitude (0.1 mm increase per 1000 m 95% CI 0.05–0.14) and AMS score (0.12 mm per Lake Louise Score (CI 0.06–0.18). Associations were also found with resting heart rate and arterial oxygen saturation (0.2 mm increase per 10% Sao2 decrease) (Sutherland et al., 2008). A study of 287 subjects similarly demonstrated a larger mean optic nerve sheath diameter (ONSD: 5.34 mm, 95%CI 5.18–5.51 mm) in AMS sufferers compared to nonsufferers (4.46 mm, 95%CI 4.39-4.54 mm) (Fagenholz et al., 2009). However, a more recent study of 23 subjects has failed to show any difference in ONSD between AMS sufferers and nonsufferers (Lawley et al., 2012). The method of ONSD image acquisition and analysis is critical and any increases in ONSD, even if small, may reflect the mild interstitial edema that occurs with the hypoxia induced increased permeability in other tissues, for example, eye, subcutaneous tissues, lung, and kidney (proteinuria) independent of AMS. Another study has demonstrated that ONSD does increase with AMS; however wide inter-individual variation and because the degree of ONSD values were often below the clinical threshold for raised ICP, the authors concluded that their study did not support an increase in ICP in mild to moderate AMS (Keyes et al., 2013). Pulsatility Index (PI). Gosling’s Pulsatility Index (PI) is calculated from transcranial Doppler measurements of the middle cerebral artery (MCA): PI = systolic velocity diastolic velocity/mean velocity; a normal value being less than 1. In recent years, a good correlation between PI and ICP has been found in the context of nonspecific intracranial pathologies (Bellner et al., 2004), trauma (Moreno et al., 2000; Tan et al., 2001; Voulgaris et al., 2005; Bor-Seng-Shu et al., 2006), cerebral mass lesions (including hematomas) (Cardoso and Kupchak, 1992; Harada et al., 1993; Czosnyka et al., 1996), hydrocephalus (Norelle et al., 1989; Quinn and Pople, 1992; Nadvi et al., 1994; Goh and Minns, 1995; Hanlo et al., 1995; Iacopino et al., 1995; Vajda et al., 1999; Rainov et al., 2000) and subarachnoid hemorrhage (Soehle et al., 2007). In general, correlation is strongest when ICP is over

125

20, meaning that PI may be a poor index of ICP in the normal or only slightly elevated range. The correlation between ICP and PI across such a range of pathologies implies that the technique may be of use in noninvasively assessing ICP at altitude. Ter Minassian et al. (2001) measured PI during a simulated ascent of Mount Everest (Operation Everest III). Eight subjects were studied in a hypobaric chamber decompressed to altitudes of 5000, 6000, 7000, and 8000 m. All measurements were done on day 3 after ‘‘arrival’’ at the altitude except those measured at the 8000 m altitude, which were done after 4 hours. They calculated both PI and Resistivity Index (RI): RI = (systolic velocity – diastolic velocity)/systolic velocity). Their demonstration of a very clear reduction in PI at each altitude gain seems to conflict with the proposal that PI should rise with ICP (if one assumes that ICP is rising). Of note however, the Paco2 values also fell dramatically, and this hyperventilatory response to hypobaric hypoxia may alter the usefulness of PI since the fall in Paco2 causes cerebral vasoconstriction. Variations in Paco2 have previously been shown to alter PI independently (Homburg et al., 1993; Czosnyka et al., 1996). Palma et al. (2006) studied 9 individuals ascending to 4300 m. They assessed PI and Dynamic Flow Index (DFI, Mean Flow Velocity/PI). They found that DFI was increased in subjects with AMS at 4300 m when compared to asymptomatic subjects. Whether this relates to changes in cerebral hemodynamics or intracranial pressure cannot be evaluated. There has been the belief that exercise at altitude increases the likelihood of an individual developing AMS and HACE. Imray et al. (2005) assessed cerebral perfusion during exercise in 9 individuals ascending to 5260 m. They demonstrated an increase in resting MCAv with increasing altitude and a further increase during exercise up to 50% of VO2Max, beyond which MCAv declined. Marked rises in blood pressure and an elevated MCAv could stress the blood brain barrier, possibly initiating vasogenic edema. Subudi et al. (2008) have similar results, and both groups suggest that cerebral blood flow and hypoxia may limit exercise performance. Recently however, moderate exercise in normobaric hypoxia has been shown not to exacerbate AMS (Schommer et al., 2012; Mairer et al., 2013). Retinal imaging. Many researchers have studied the optic disc and retinal vessels in hypobaric hypoxia. Bosch et al. (2008) demonstrated both arterial and venous retinal distension in hypoxia that correlated with acute mountain sickness. Similarly, we have demonstrated venous distension with high altitude headache (Fig. 2) (Wilson et al., 2013), using the sum of headache scores reported by 24 subjects over 8 mornings after an ascent over a 13 day period to 5340 m. Although a correlation between retinal venous distension and headache was demonstrated, such a correlation cannot be taken as an indicator of a causal relationship between the two variables because anything that changes with altitude will correlate with AMS scores, since these increase with altitude. Using a more sophisticated semiautomatic computerized retinal vessel analysis technique, Willmann et al. also recently confirmed vessel distension in 18 subjects ascending to 4559 m. His group recorded headache (measured on a 5-point scale or with LL and AMS-C scores with concurrently measured vessel diameter) and did not demonstrate a correlation. This may reflect the headache reporting technique or it may be

126

WILSON ET AL.

FIG. 2. Retinal venous distension: Retinal vasculature (A) at sea level and (B) after 24 hours at 5300 m. There is marked distension (venous greater than arterial) and the venous distension correlates with headache (Wilson, Davagnanam, and Holland, 2013). that, in reality, headache does not correlate with venous vessel distension (Willmann et al., 2013). While retinal vessel (and in particular venous) distension does not necessarily reflect ICP changes, it may reflect an increase in cerebral venous pressure. Papilledema (optic disc swelling) is the classic pathoneumoic sign of raised ICP. Fifty nine percent of climbers at 6865 m have been found to exhibit a degree of optic disc swelling (Bosch et al., 2008), although in other studies the incidence is lower (Singh et al., 1969). In a similar manner to retinal vessel distension, some groups have demonstrated a correlation with headache (Bosch et al., 2008), while others have not (Willmann et al., 2011). Again, this probably reflects the technique to measure headache burden. Papilledema has also been seen with high altitude cerebral edema (Dickinson et al., 1983). While this may reflect a rise in ICP, it does not aid in understanding the underlying physiological process. Animal and Direct Studies

Krasney and co-workers designed the first conscious animal model of HACE in 1990 (Krasney et al., 1990; CurranEverett et al., 1991). In addition to measuring ICP invasively in the lateral ventricle of sheep (n = 20), they also calculated cerebral oxygen extraction and investigated changes in wetto-dry brain weight (n = 9) with 72 hours of normobaric hypoxia (arterial oxygen tension of 40 mmHg giving an arterial oxygen saturation of 50%). Whilst ICP did not change with the 72 hours of hypoxia, wet-dry brain weight ratios increased in all regions, but especially in the white matter, caudate nuclei, and thalamus. The authors hypothesized that the lack of a rise in ICP with the increase in brain volume was due to reciprocal loss of volume from the ventricles. They noted that, in normoxia, it is possible to withdraw 3–4 mL of CSF from the ventricular catheter with ease. However, after 7 hours of hypoxia, it was often impossible to withdraw more than 0.5–1.0 mL.

Yang et al. (1993) studied ICP in goats exposed to a Pao2 of 40 Torr (equivalent to an altitude of 4000 m). Although there were methodological issues, ICP and cerebral blood flow increased, and intracranial compliance decreased with 2 hours of hypoxia. Exposed to the same Pao2, ICP rose (and cerebral edema occurred) in sheep which exhibited AMS behaviour (off food/water), but not in those behaving normally (Yang et al., 1994). In New Zealand white rabbits, ICP did not rise with exposure to simulated hypobaric hypoxia of 5000 m for 6 hours, whether or not steroids were administered (Pendon and King, 2003). It must be remembered that both the anatomy (especially biped nature of humans) and physiology may limit translation of animal work into humans. Brain Imaging

CT scans performed on climbers with HAPE and neurological dysfunction have demonstrated the presence of small ventricles and cisterns, and the disappearance of cerebral sulci (Koyama et al., 1988). Magnetic resonance imaging (MRI) allows improved assessment of edema and, while not measuring ICP, can infer changes in ICP from changes in brain volume (e.g., loss of ventricular space, sulci effacement). It has been used in a number of studies where subjects have developed AMS, and there are clinical reports of MRIs of patients who have suffered with HACE (Hackett et al., 1998b). However, while edema may occur, many of these subjects have atrophic brains, questioning whether an ICP rise actually occurs (Wilson et al., 2009). Cytotoxic (intracellular) (Houston and Dickinson, 1975) and vasogenic (extracellular water accumulation due to increased blood brain barrier permeability) (Hackett et al., 1998a) edema have both been postulated to be core mechanisms in HACE pathogenesis. A particularly useful MRI technique, diffusion weighting, differentiates between these two forms of edema. Susceptibility weighting, whilst not

INTRACRANIAL PRESSURE

demonstrating edema per se, is a useful MRI modality for demonstrating microhemorhages that are the pathognomonic hallmark of previous high altitude cerebral edema (Kallenberg et al., 2008, Schommer et al., 2013). Amongst 9 subjects exposed to a simulated sudden ascent to 4572 m, a 2.77% (36.2 mL) increase in brain matter volume was found at 32 hours (Morocz et al., 2001). These volume changes only occurred in gray (not white) matter, regardless of whether AMS symptoms were present. Meanwhile, in 10 subjects who had MRIs after 10 hours of exposure to a simulated altitude of 4500 m (eight of whom suffered AMS), none had cerebral edema (Fischer et al., 2004). They demonstrated that ventricular CSF volume decreased in all subjects, more so in those with severe AMS. This again implies that that the brain parenchyma or other intracerebral components expand in hypoxia. Meanwhile, in 22 subjects (half of whom suffered AMS, of which seven received metoclopramide and paracetamol), cerebral swelling of the order of 7 – 4.8 mL (approximately 0.5% of total brain volume) was observed after 16 hours of exposure to 12% O2 (equivalent to 4500 m) (Kallenberg et al., 2007). In addition, they studied T2 Relaxation Time (T2rT is a technique that reflects changes in parenchymal water content) and the Apparent Diffusion Coefficient (ADC, which reflects

127

changes in the diffusibility of water molecules). ADC helps differentiate vasogenic edema (as water moves intracellularly, its diffusibility, and hence ADC, falls) from cytotoxic edema (as water moves extracellularly its diffusibility and ADC increases). Specific regions of interest within the brain (white matter, basal ganglia, genu and splenium of corpus callosum, and cerebellar white matter) were studied. Hypoxia resulted in a general increase in T2rT representing an increase in parenchymal edema. ADC values were consistently lower in those who developed AMS symptoms. This suggested that hypoxia causes a generalized vasogenic edema, but that with AMS may be associated with an additional cytotoxic (intracellular) component. This may occur through a reduction in the Na + / K + ATPase pump (see below). This study also demonstrated that those developing AMS had a greater brain:intracranial volume ratio, supporting the ‘‘tight fit’’ hypothesis. It should be noted that the volume changes demonstrated in these studies are relatively small and hence are unlikely to significantly alter ICP in most subjects. In addition, other MRI studies have not demonstrated that ‘‘tighter’’ brains are more prone to AMS (Dubowitz et al., 2009). Schoonman et al. (2008) studied 9 students exposed to isobaric hypoxia (N2 enriched air) to obtain arterial oxygen saturations of 75%–80% for 6 hours. Seven of the 9 devel-

FIG. 3. Arterial (red lines), venous (light blue line), arterial-venous difference (dark blue lines) blood flow, and cerebral spinal fluid flow rate wave forms (green line) demonstrating normal intracranial compliance (low ICP) in normobaric (upper two panels) and lower intracranial compliance (higher ICP) in the hypoxic state (lower two panels) (Lawley, 2013).

128

oped AMS. Visual inspection of the MRIs failed to show any edema. However, there was a general increase in ADC with hypoxia while ADC values negatively correlated with severity of cerebral symptoms. Hence, similarly to Kallenberg, the authors concluded that vasogenic edema occurs in isobaric hypoxia irrespective of AMS, while severe AMS is associated with an additional mild cytotoxic component. Fischer et al. demonstrated that, whilst exposure to a simulated altitude of 4500 m did not induce demonstrable edema, a mean 10% reduction in intracranial CSF volume occurred at 10–12 hours of exposure. This may well reflect cerebral swelling caused by increased cerebral blood flow, even if edema does not form (Fischer et al., 2004). Although studies using ADC thus suggest the development of vasogenic edema in response to hypoxia, and possibly of cytotoxic edema in those developing severe AMS, a more recent hypoxic MRI study (Dubowitz et al., 2009) demonstrates that hypoxia causes brain swelling without edema formation. This study of 12 subjects found cerebral swelling and compression of ventricular CSF spaces after only 40 minutes of hypoxia. Without edema, this would suggest a change in intracerebral blood volume. Matsuzawa et al. (1992) reported slightly increased T2 signal intensity in 4 of 7 subjects with AMS in a 24 hour simulated altitude experiment. A recent study has used a complex multi (MRI)-imaging approach to infer ICP (MR-ICP) (Fig. 3). Thirteen subjects were studied in 12% normobaric hypoxia. The technique used a 3 tesla magnet and a 16 channel head and neck coil. Using a cardiac gated velocity encoding cine phase contrast imaging (VENC), a measurement of arterial inflow (internal carotid and vertebral arteries), venous outflow (internal jugular veins), and pulsatile CSF were obtained. Having obtained velocities and diameters for the arteries and veins, it was then possible to calculate time-dependent volumetric flow rate waveforms. The transcranial volumetric flow rate was calculated for each point in the cardiac cycle by subtracting the venous and the cerebrospinal fluid outflow rates from the arterial inflow rate. From this an intracranial volume change was derived, and from this

WILSON ET AL.

the CSF pressure gradient waveform was calculated. Finally, intracranial compliance was derived from the ratio of the maximal intracranial volume change and the maximal CSF pressure gradient pressure change. The major findings in this study were that, although there was no change in intracranial pressure or cerebral perfusion pressure despite an increase in brain volume, there was a statistically significant relationship between change in intracranial pressure and acute mountain sickness severity (R2 = 0.71, B = 2.3, p < 0.01) after 10 hours. Further studies are required to confirm these findings (Lawley, 2013). Overall, MRI studies of simulated ascents to ‘‘very high altitude’’ in hypobaric chambers demonstrate that acute hypoxia may cause brain parenchymal enlargement. This may be a combination of vasogenic oedema with hypoxia and cytotoxic oedema when in the context of AMS, but this is by no means conclusive and another ‘‘volume’’ could be contributing to this apparent parenchymal enlargement. Venous Hypothesis

Acute hypoxic exposure results in an increase in cerebral blood inflow and as a consequence, venous outflow must increase. However, venous outflow can be relatively compromised, either through limited intracranial venous drainage (e.g., transverse sinus stenosis) or from elevated central venous pressures. The resultant change in hydrostatic (Starling) pressures across the capillary/venous bed, in association with a more permeable endothelium could account for edema, a rise in ICP, and finally HACE (Fig. 4). Cerebral venous distension causes headache (Ray and Wolff, 1940) and this, prior to an ICP rise, could be the initiator mechanism of high altitude headache. Recently our group proposed, and then demonstrated, that hypobaric hypoxia results in both retinal and cerebral venodistension and have suggested that this reflects a relative inadequacy in venous drainage (Wilson and Imray, 2009; Wilson et al. 2011; 2013). The importance of the venous system in ICP is often forgotten. Interestingly at the turn of the last century, the venous role was much more appreciated (Hill, 1896).

FIG. 4. Diagram demonstrating the potential role of relative venous outflow compromise and venous hypertension (Wilson, Imray, and Hargens, 2011). Hypoxia results in increased cerebral blood flow. If the venous system cannot drain this adequately (e.g., because of anatomical limitations that are not normally clinically significant, or because of increased distal pressures), it will distend (causing headache) while CSF buffers the increased venous volume preventing an ICP rise initially. Increased hydrostatic pressures could then contribute to vasogenic edema.

INTRACRANIAL PRESSURE

It should be noted that once again there is independent evidence to both support and refute the hypothesis. Like us, Bosch et al. (2009) found a correlation between ‘venous retinal vessel diameter and AMS-C’ (OR 1.042, 95%CI (1.001–1.084, p = 0.046). However, a recent study by Willmann et al. (2013), whilst demonstrating arterial and venous retinal distension on rapid ascent to altitude (4559 m), failed to demonstrate a correlation with an isolated headache score; hence further studies are needed. Therapeutic significance of intracranial pressure

The importance of raised ICP as a possible mechanism for altitude-related cerebral symptoms and illnesses is also implied by the successful use of acetazolamide and dexa-

129

methasone. Classically both of these drugs lower ICP, but it is not clear if this is their critical mode of action in treating high altitude illness (Imray et al., 2011). Acetazolamide leads to bicarbonate diuresis through inhibition of renal carbonic anydrase. The resulting metabolic acidosis stimulates ventilation (Holl et al., 1990). It is also thought to reduce the production of cerebrospinal fluid, although this itself is unlikely to be responsible for its benefits. Interestingly, acetazolamide is also the primary medical treatment for idiopathic intracranial hypertension, a condition that bears many similarities to high altitude headache (Wilson et al., 2011). The exact mechanism of action of glucocorticoids such as dexamethasone is unknown, but some authors have speculated that they reduce capillary permeability and cytokine release. Dexamethasone’s clinical effects however are very

FIG. 5. The mechanisms that are thought to underlie AMS and HACE and contribute to a rise in ICP (Wilson and Imray, 2009). Progression from artery to vein (left to right). Mechanical factors increase intravascular pressure and hence can cause vasogenic edema and vessel wall damage. This pressure can act in the artery (increased hydrostatic pressure associated with increased flow) or vein (if there is venous outflow obstruction). The partial pressures of oxygen and carbon dioxide are thought to have direct vasoactive properties, with hypoxemia causing vasodilatation and hypocarbia causing vasoconstriction. A balance between these is mediated by the hypoxic ventilatory response. Cytotoxic edema might result from direct hypoxia-induced Na + /K + ATPase failure. Many chemical mediators have been implicated. Free radical formation could directly damage vessel basement membranes, causing vasogenic edema. Accumulation of HIF-1a and subsequent upregulation of VEGF could contribute to further basement membrane damage and edema. Local hyperkalaemia could trigger calcium-mediated nitric oxide release, which in turn can act on vessel smooth muscle to cause vasodilatation. Neuronally-mediated adenosine release could also cause vasodilatation. Vessel dilatation has been implicated in activating the trigeminovascular system, causing headache. An important element of HACE is microhemorrhage formation, which might be caused by vessel damage from chemical mediators or cytokines or by damage through increased hydrostatic pressure. AMS, acute mountain sickness; HACE, high-altitude cerebral edema; ICP, intracranial pressure.

130

rapid, implying that another mechanism, such as alteration in cardiac function, is also involved. Conclusion

Both direct and indirect measures of ICP have demonstrated that intracranial pressure may rise when animal or human subjects are exposed to hypoxia. However, the period of inter-individual compliance prior to such a rise and the relationship with the subjective symptom of headache are poorly understood and as such, increased ICP may not be the cause of AMS. Figure 5 demonstrates the various mechanisms (from vascular pressures through to edema formation) that may underlie fluid shifts and result in raised ICP in hypoxia. A greater understanding of the interrelationship of arterial, capillary, and venous blood, together with brain and CSF volumes, and how these buffer each other prior to a rise in ICP is required. It may well be within this compliant period that milder forms of altitude illness (perhaps relating to venous distension) such as headache are occurring. To date, studies have tended to focus on individual volume components, making interpretation complex. Newer technologies are giving better insights into the pathophysiology and may facilitate better preventative and therapeutic strategies. Author Disclosure Statement

No competing financial interests exist. References

Ba¨rtsch P, Bailey DM, Berger MM, et al. (2004). Acute mountain sickness: Controversies and advances. High Alt Med Biol 5:110–124. Bellner J, Romner B, Reinstrup P, et al. (2004). Transcranial Doppler sonography pulsatility index (PI) reflects intracranial pressure (ICP). Surg Neurol 62:45–51. Bor-Seng-Shu E, Hirsch R, Teixeira MJ, et al. (2006). Cerebral hemodynamic changes gauged by transcranial Doppler ultrasonography in patients with posttraumatic brain swelling treated by surgical decompression. J Neurosurg 104:93–100. Bosch MM, Merz TM, Barthelmes D, et al. (2009). New insights into ocular blood flow at very high altitudes. J Appl Physiol 106:454–460. Bosch MM, Barthelmes D, Merz TM, et al. (2008). High incidence of optic disc swelling at very high altitudes. Arch Ophthalmol 126:644–650. Burrows G. (1848). On Disorders of the Cerebral Circulation and on the Connection between Affections of the Brain and Diseases of the Heart. Philadelphia: Lea and Blanchard. Bu¨ki B, Chomicki A, Dordain M, et al. (2000). Middle-ear influence on otoacoustic emissions. II: Contributions of posture and intracranial pressure. Hearing Res 140:202–211. Bu¨ki B, Avan P, Lemaire J, et al. (1996). Otoacoustic emissions: A new tool for monitoring intracranial pressure changes through stapes displacements. Hearing Res 94:125–139. Cardoso ER, and Kupchak JA. (1992). Evaluation of intracranial pressure gradients by means of transcranial Doppler sonography. Acta Neurochir Suppl (Wien), 55:1–5. Curran-Everett DC, Iwamoto J, Meredith MP, and Krasney JA. (1991). Intracranial pressures and O2 extraction in conscious sheep during 72 h of hypoxia. Am J Physiol 261:H103–109. Cushing H. (1926). The Third Circulation in Studies in Intracranial Physiology and Surgery, London: Oxford University Press.

WILSON ET AL.

Czosnyka M, Richards HK, Whitehouse HE, and Pickard JD. (1996). Relationship between transcranial Doppler-determined pulsatility index and cerebrovascular resistance: An experimental study. J Neurosurg 84:79–84. Dickinson J, Heath D, Gosney J, and Williams D. (1983). Altitude-related deaths in seven trekkers in the Himalayas. Thorax 38:646–656. Dubowitz DJ, Dyer EA, THeilmann RJ, Buxton RB, and Hopkins SR. (2009). Early brain swelling in acute hypoxia. J Appl Physiol 107:244–252. Fagenholz PJ, Gutman HA, and Murrary AF. (2007). Evidence for increased intracranial pressure in high altitude pulmonary edema. High Alt Med Biol 8:331–336. Fagenholz PJ, Gutman HA, and Murrary AF. (2009). Optic nerve sheath diameter correlates with the presence and severity of acute mountain sickness: Evidence for increased intracranial pressure. J Appl Physiol 106:1207–1211. Fischer R, Vollmar C, Thiere M, et al. (2004). No evidence of cerebral oedema in severe acute mountain sickness. Cephalalgia 24:66–71. Frank AM, Alexiou C, Hulin P, Janssen T, Arnold W, and Trappe AE. (2000). Non-invasive measurement of intracranial pressure changes by otoacoustic emissions (OAEs)—A report of preliminary data. Zentralblatt Neurochirurg 61:177–180. Goh D, and Minns RA. (1995). Intracranial pressure and cerebral arterial flow velocity indices in childhood hydrocephalus: Current review. Childs Nerv Syst 11:392–396. Gwer S, Steward V, and Birch A. (2013). The tympanic membrane displacement analyser for monitoring intracranial pressure in children. Childs Nerv Syst 29:927–933. Hackett PH, Yarnell PR, Hill R, Reynard K, Heit J, and McCormick J. (1998a). High-altitude cerebral edema evaluated with magnetic resonance imaging: Clinical correlation and pathophysiology. JAMA 280:1920–1925. Hackett PH, Yarnell PR, Hill R, Reynard K, Heit J, and McCormick J. (1998b). High-altitude cerebral edema evaluated with magnetic resonance imaging: Clinical correlation and pathophysiology. JAMA 280:1920–1925. Hanlo PW, Gooskens RH, and Nijhuis IJ. (1995). Value of transcranial Doppler indices in predicting raised ICP in infantile hydrocephalus. A study with review of the literature. Childs Nerv Syst 11:595–603. Harada K, Hayashi T, Anegawa S, et al. (1993). [Transcranial Doppler sonography in acute intracranial hypertension model— Usefulness of pulsatility index]. No To Shinkei 45:851–856. Hartig GS, and Hackett P. (1992). Cerebral spinal fluid pressure and cerebral blood velocity in acute mountain sickness. In: JR Sutton, G Coates, CS Houston, eds. Hypoxia and Mountain Medicine. Oxford: Pergamon Press, pp. 260–265. Headache Classification Subcommittee of the International Headache Society (2004). The International Classification of Headache Disorders: 2nd edition. Cephalalgia 24:9–160. Hill L. (1896). The Physiology and Pathology of the Cerebral Circulation, J&A Churchill. Available at: https://archive.org/ details/physiologypatholhill. Last accessed 16 May 2014. Holl K, Hassler HE, Nemati N, Gaab MR, Haubitz B, and Becker H. (1990). [The effect of acetazolamide-induced, endogenous volume stress of the cerebrospinal system on intracranial pressure. Diamox—Effect on intracranial pressure]. Neurochirurgia 33:29–36. Homburg AM, Jakobsen M, and Enevoldsen E. (1993). Transcranial Doppler recordings in raised intracranial pressure. Acta Neurol Scand 87:488–493. Houston CS, and Dickinson J. (1975). Cerebral form of highaltitude illness. Lancet 2:758–761.

INTRACRANIAL PRESSURE

Iacopino DG, Zaccone C, Molina D, Todaro C, Tomasillo F, and Cardia E. (1995). Intraoperative monitoring of cerebral blood flow during ventricular shunting in hydrocephalic pediatric patients. Childs Nerv Syst 11:483–486. Imray C, Booth A, Wright A, and Bradwell A. (2011). Acute altitude illnesses. BMJ 343:4943. Imray CH, Myers SD, and Pattinson KT. (2005). Effect of exercise on cerebral perfusion in humans at high altitude. J Appl Physiol 99:699–706. Kallenberg K, Dehnert C, and Dorfler A. (2008). Microhemorrhages in nonfatal high-altitude cerebral edema. J Cereb Blood Flow Metab 28:1635–1642. Kellie G. (1824). An account of the appearances observed in the dissection of two of the three individuals presumed to have perished in the storm of the 3rd, and whose bodie were discovered in the vicinity of Leith on the morning of the 4th November 1821 with some reflections on the pathology of the brain. Transact Medico-Chirurgical Society of Edinburgh, 84–169. Keyes LE, Paterson R, Boatright D, et al. (2013). Optic nerve sheath diameter and acute mountain sickness. Wildern Environ Med 24:105–111. Koyama S, Kobayashi T, Kubo K, et al. (1988). The increased sympathoadrenal activity in patients with high altitude pulmonary edema is centrally mediated. Jpn J Med 27:10–16. Krasney JA, Curran-Everett DC, and Iwamoto J. (1990). High altitude cerebral edema: An animal model. In: JR Sutton, G Coates, JE Remmers, eds. Hypoxia: the Adaptations. Ontario: Decker, pp. 200–205. Lawley JS. (2013). Acute mountain sickness. In: The Elusive Phenotype - PhD Thesis Chapter. Bangor University, p. 139. Lawley JS, Oliver SH, and Mullins P. (2012). Optic nerve sheath diameter is not related to high altitude headache: A randomized controlled trial. High Alt Med Biol 13:193–199. Magendie F. (1842). Recherches Anatomique et Physiologique sur le Liquide Ce´phalo-rachidien ou Ce´rebro-spinal. Paris. Mairer K, Wille M, Grander W, and Burtscher M. (2013). Effects of exercise and hypoxia on heart rate variability and acute mountain sickness. Intl J Sports Med 34:700–706. Matsuzawa Y. 1992. Cerebral edema in acute mountain sickness. In: G Ueda, JT Reeves, M. Sekiguchi, eds. High Altitude Medicine. Matsumoto, Japan: Sjinshu University Press, pp. 300–304. Minassian, ter A, Beydon L, Ursino M, Gardette B, Gortan C, and Richalet JP. (2001). Doppler study of middle cerebral artery blood flow velocity and cerebral autoregulation during a simulated ascent of Mount Everest. Wildern Environ Med 12:175–183. Monro A. (1783). Observations on the structure and function of the nervous system, Edinburgh: Creech and Johnson. Moreno JA. (2000). Evaluating the outcome of severe head injury with transcranial Doppler ultrasonography. Neurosurg Focus 8:e8. Morocz IA, Zientara GP, Gudbjartsson H, et al. (2001). Volumetric quantification of brain swelling after hypobaric hypoxia exposure. Exp Neurol 168:96–104. Nadvi SS, Du Trevou MD, Van Dellen JR, and Gouws E. (1994). The use of transcranial Doppler ultrasonography as a method of assessing intracranial pressure in hydrocephalic children. Br J Neurosurg 8:573–577. Norelle A, Fischer AQ, and Flannery AM. (1989). Transcranial Doppler: A noninvasive method to monitor hydrocephalus. J Child Neurol 4:S87–90. Olzowy B, von Gleichenstein G, Canis M, and Mees K. (2008). Distortion product otoacoustic emissions for assessment of

131

intracranial hypertension at extreme altitude? Eur J Appl Physiol 103:19–23. Palma J, Macedonia C, and Deuster P. (2006). Cerebrovascular dynamics and vascular endothelial growth factor in acute mountain sickness. Wildern Environ Med 17:1–7. Quinn MW, and Pople IK. (1992). Middle cerebral artery pulsatility in children with blocked cerebrospinal fluid shunts. J Neurol Neurosurg Psychiatry 55:325–327. Rainov NG, Weise JB, and Burkert W. (2000). Transcranial Doppler sonography in adult hydrocephalic patients. Neurosurg Rev 23:34–38. Ray BS, and Wolff HG. (1940). Experimental studies on headache. Arch Surg (Chicago, Ill. : 1960), 41:813–856. Reid J. (1846). Burrows on cerebral circulation. Edinburgh Monthly J Med Sci pp.111–123. Available at: http://hdl.handle .net/2027/hvd.32. Last accessed 16 May 2014. Roach RC, and Hackett PH. (2001). Frontiers of hypoxia research: Acute mountain sickness. J Expl Biol 204:3161–3170. Ross RT. (1985). The random nature of cerebral mountain sickness. Lancet 1 (8435):990–991. Sakka L Thalamy A, GIraudet F, Hassoun T, Avan P, and Chazal J. (2012). Electrophysiological monitoring of cochlear function as a non-invasive method to assess intracranial pressure variations. Acta Neurochirurg 114:131–134. Schaltenbrand G. 1933. Atmospheric pressure, circulation, respiration and cerebrospinal fluid pressure. Acta Aerophysiol 65–78. Schommer K, Hammer M, Hotz L, Menold E, Bartsch P, and Berger MM. (2012). Exercise intensity typical of mountain climbing does not exacerbate acute mountain sickness in normobaric hypoxia. J Appl Physiol 113:1068–1074. Schommer K, Kallenberg K, Lutz K, Bartsch P, and Knauth M. (2013). Hemosiderin deposition in the brain as footprint of high-altitude cerebral edema. Neurology 81:1776–1779. Schoonman GG, Sandor PS, Nirkko AC, et al. (2008). Hypoxiainduced acute mountain sickness is associated with intracellular cerebral edema: A 3 T magnetic resonance imaging study. J Cereb Blood Flow Metab 28:198–206. Shimbles S, Dodd C, Banister K, Mendelow AD, and Chambers IR. (2005a). Clinical comparison of tympanic membrane displacement with invasive ICP measurements. Acta Neurochirurg Supplement 95:197–199. Shimbles S, Dodd C, Banister K, Mendelow AD, and Chambers IR. (2005b). Clinical comparison of tympanic membrane displacement with invasive intracranial pressure measurements. Physiologl Measure 26:1085–1092. Singh I, Khanna PK, Srivastava MC, et al. (1969). Acute mountain sickness. New Engl J Med 280:175–184. Soehle M, Chatfield DA, Czosnyka M, and Kirkpatrick PJ. (2007). Predictive value of initial clinical status, intracranial pressure and transcranial Doppler pulsatility after subarachnoid haemorrhage. Acta Neurochirurg 149:575–583. Subudhi AW, Lorenz MC, Fulco CS, and Roach RC. (2008). Cerebrovascular responses to incremental exercise during hypobaric hypoxia: Effect of oxygenation on maximal performance. Am J Physiol Heart Circ Physiol 294:H164–171. Sutherland AI, Morris DS, Owen CG, Bron AJ, and Roach RC. (2008). Optic nerve sheath diameter, intracranial pressure and acute mountain sickness on Mount Everest: A longitudinal cohort study. Br J Sports Med 42:183–188. Tan H, Feng H, Gao L, Huang G, and Liao X. (2001). Outcome prediction in severe traumatic brain injury with transcranial Doppler ultrasonography. Chin J Traumatol 4:156–160. Vajda Z, Buki A, Veto F, Horvath Z, Sandor J, and Doczi T. (1999). Transcranial Doppler-determined pulsatility index in

132

the evaluation of endoscopic third ventriculostomy (preliminary data). Acta Neurochirurg 141:247–250. Voulgaris SG, Partheni M, Kaliora H, et al. (2005). Early cerebral monitoring using the transcranial Doppler pulsatility index in patients with severe brain trauma. Med Sci Monit 11:CR49–52. Willmann G, Fischer MD, Schatz A, et al. (2011). Quantification of optic disc edema during exposure to high altitude shows no correlation to acute mountain sickness. PloS One 6:e27022. Willmann G, Fischer M, Schommer K, Bartsch P, Gekeler, F, Schatz A. (2014). Missing correlation of retinal vessel diameter with high-altitude headache. Ann Clin Translat Neurol 1:59–63. Wilson M, and Imray C. (2009). High-altitude cerebral effects: Risks and mechanisms—Authors’ reply. Lancet Neurol 8: 605–605. Wilson MH, and Milledge J. (2008). Direct measurement of intracranial pressure at high altitude and correlation of ventricular size with acute mountain sickness: Brian Cummins’ results from the 1985 Kishtwar expedition. Neurosurgery 63:970–974; discussion 974–975. Wilson MH, Davagnanam I, Holland G, et al. (2013). Cerebral venous system and anatomical predisposition to high-altitude headache. Ann Neurol 73:381–389. Wilson MH, Imray CHE, and Hargens AR. (2011). The headache of high altitude and microgravity—Similarities with clinical syndromes of cerebral venous hypertension. High Alt Med Biol 12:379–386.

WILSON ET AL.

Wilson MH, Newman S, and Imray CH. (2009). The cerebral effects of ascent to high altitudes. Lancet Neurol 8: 175–191. Wright AD, Imray CH, Morrissey MS, Marchbanks RJ, and Bradwell AR. (1995). Intracranial pressure at high altitude and acute mountain sickness. Clin Sci 89:201–204. Yang SP, Bergo GW, Krasney E, and Krasney JA. (1994). Cerebral pressure-flow and metabolic responses to sustained hypoxia: effect of CO2. J ApplPhysiol 76:303–313. Yang YB, Sun B, Yang Z, Wang J, and Pong Y. (1993). Effects of acute hypoxia on intracranial dynamics in unanesthetized goats. J Appl Physiol 74:2067–2071.

Address correspondence to: Mark H. Wilson, MD The Brain Injury Centre St Mary’s Hospital Imperial College Praed Street London W2 1NY United Kingdom E-mail: [email protected] Received December 29, 2013; accepted in final form March 13, 2014.