The Effects of Hyper- and Hypoventilation on Cerebral Physiology in a Healthy and Compromised Brain State: An Experimental Pig Model with State-of-the-Art Neuromonitoring

Abstract

Arterial carbon dioxide (pCO₂) strongly affects cerebrovascular tone and cerebral physiology. While moderate hyperventilation is often used to reduce intracranial pressure (ICP) in acute brain injury, its broader physiological effects remain unclear. In this experimental study, 10 anesthetized pigs underwent multimodal neuromonitoring, including ICP, cerebral perfusion pressure, common autoregulatory indices (pressure reactivity index [PRx], cerebral blood flow index [CBFx], oxygen reactivity index), CBF, brain tissue oxygenation (pbtO₂), and microdialysis. Animals were exposed to four ventilatory intervals (normoventilation, moderate and severe hyperventilation, and hypoventilation), first in a healthy state and then following induction of intracranial hypertension (ICP 30–40 mmHg) via epidural balloon inflation. In the healthy brain, moderate and severe hyperventilation numerically, but non-significantly, reduced CBF without affecting pbtO₂ or cerebral energy metabolism, while hypoventilation increased CBF and pbtO₂. Under intracranial hypertension, moderate hyperventilation improved PRx and preserved CBF, pbtO₂, and metabolism, but severe hyperventilation reduced pbtO₂. Hypoventilation produced variable responses: Animals with higher baseline blood pressure showed improved perfusion and oxygenation, whereas those with lower pressure experienced reduced CBF, impaired oxygenation, and metabolic distress. These findings underscore the complex and context-dependent effects of pCO₂ on cerebral physiology, indicating that ventilatory strategies may both benefit and harm the injured brain depending on individual vulnerability and hemodynamic status.

Keywords

carbon dioxide cerebral blood flow hyperventilation multimodal monitoring neurointensive care

Introduction

Neurointensive care (NIC) aims to mitigate secondary brain injury by optimizing the cerebral environment following severe acute brain disorders.^1,2 This is achieved by controlling several key physiological parameters along the downstream cascade of cerebral hemodynamic and metabolic regulation, including arterial blood pressure (ABP), intracranial pressure (ICP), cerebral perfusion pressure (CPP), cerebral autoregulation (CA; the capacity to regulate the blood flow to the brain over a wide range of CPPs), cerebral blood flow (CBF), and the arterial content of oxygen and glucose.^1–7 These parameters are highly interdependent, and therapeutic interventions that aim to optimize one aspect of cerebral physiology may inadvertently impair another. One physiological variable that exemplifies this complexity is arterial carbon dioxide tension (pCO₂).⁸ pCO₂ modulates cerebrovascular tone primarily through pH-mediated mechanisms, though direct CO₂ effects may also contribute,^9,10 within a physiological range of approximately 2–10 kPa.¹¹ Hypocapnia induces cerebral vasoconstriction, whereas hypercapnia promotes vasodilation, both of which can influence multiple brain physiological variables such as ICP, CPP, CA, and CBF.^5,8,12–16 In the context of traumatic brain injury (TBI), early management has often prioritized treatment of elevated ICP, frequently employing hyperventilation-induced hypocapnia.^8,17–20 While this strategy lowers ICP and can prevent brain herniation,^8,21 it may do so at the cost of reduced CBF^19,20 and increased oxygen extraction,^19,20 while the effects on CA capacity remain elusive.^{12,13,22–24} Interestingly, the energy metabolic state does not appear to be significantly affected overall,^13,19,20 although substantial variability has been observed across different brain regions and between patients in TBI.^19,20 A landmark clinical trial by Muizelaar et al. demonstrated that severe hyperventilation (target pCO₂ ≈ 3.3 kPa) was associated with worse 6-month outcomes compared to more moderate levels (4.0–4.5 kPa).¹⁷ However, the lack of clear evidence regarding optimal pCO₂ intervals, other than avoiding extremely low values, contributes to substantial variability in clinical practice across NIC units¹⁸ Conversely, in other forms of acute brain injury, such as aneurysmal subarachnoid hemorrhage (aSAH), the primary challenge may be less related to elevated ICP and more to cerebral vasospasm causing critically reduced CBF.^1,25,26 Experimental strategies involving hypercapnia have shown some promise in improving CBF,^27,28 though elevated pCO₂ may also increase cerebral blood volume (CBV), raise ICP, and potentially reduce CPP.²⁹

Despite its clinical relevance, the net physiological impact of pCO₂ modulation in both healthy and injured brain states remains insufficiently understood and incompletely studied.^8,12 To address this knowledge gap, we conducted this experimental, reverse translational study on the effects of hypo- and hyperventilation on brain physiology across a wide pCO₂ spectrum using a large-animal model with state-of-the-art neuromonitoring techniques, mirroring those used in modern NIC units. The aim was to explore the effects of certain pCO₂ targets on the downstream cascade of brain physiology, including ICP, CPP, CA, pbtO₂, and cerebral energy metabolism using microdialysis (MD), in both a healthy and a compromised state in a pig model.

Materials and Methods

Animals

In this experimental study, ten pigs (Sus scrofa domesticus, three-breed cross; Norwegian Landrace [1/4], Yorkshire [1/4], and Hampshire [1/2]) aged 2–3 months with mixed gender were included (Supplementary Table S1).

Anesthesia and mechanical ventilation

Sedation was initiated with tiletamine/zolazepam (6 mg/kg, Zoletil Forte, Virbac, Denmark) and xylazine (2.2 mg/kg, Rompun, Elanco, Denmark), followed by a fentanyl bolus (5 µg/kg, Braun, Sweden) after intravenous access. Anesthesia was maintained throughout the experiment using ketamine (30 mg/kg/h; Abcur, Helsingborg, Sweden), fentanyl (4 µg/kg/h; Braun, Danderyd, Sweden), and midazolam (0.12 mg/kg/h; Accord-Healthcarre, Solna, Sweden). After confirming adequate anesthesia through lack of nociceptive reflexes, rocuronium (2.5 mg/kg/h, Braun, Sweden) was administered for muscle relaxation. Ringer acetate (Baxter, Kista, Sweden) was infused at 10 mL/kg/h for the first hour, then 5 mL/kg/h. After induction, the animals were intubated via tracheostomy and mechanically ventilated (Flow I, Maquet, Sweden). They were canulated in supine position, then repositioned to prone for intracranial catheters placement, and stayed prone for the duration of the session.

Study design

All animals underwent a standardized protocol for placement of multimodal monitoring, including ICP, CPP, CBF, pbtO₂, and brain energy metabolism, alongside systemic monitoring of ABP, arterial blood gas (ABG), and cardiac output (Supplementary Fig. S1). A balloon catheter was inserted epidurally in the contralateral frontotemporal region to allow induction of a compromised brain state with high ICP and low CPP.^30,31

The effects of different ventilatory pCO₂ targets on brain physiology were explored under both healthy and compromised cerebral conditions (Fig. 1) by adjusting ventilatory settings while aiming to maintain stable arterial oxygenation (pO₂). Four distinct ventilatory pCO₂ targets were studied: normoventilation (4.5–6.0 kPa), moderate hyperventilation (4.0–4.5 kPa), severe hyperventilation (2.5–3.5 kPa), and hypoventilation (7.0–8.0 kPa). Normoventilation served as the baseline and reference. Moderate hyperventilation (target pCO₂ 4.0–4.5 kPa) was included to reflect commonly employed ranges in clinical NIC settings for ICP reduction,^13,18 and is largely consistent with the control group in Muizelaar’s trial.¹⁷ Severe hyperventilation and hypoventilation represented extreme physiological perturbations for comparison.^10–12 Systemic and cerebral physiological variables were collected over 30 min for each pCO₂ interval. Normoxia (pO₂ 12–18 kPa) was targeted throughout the experiment. The respiratory settings for each pCO₂ target interval were standardized as described below, but they were individually adjusted as needed to achieve comparable pCO₂ and pO₂ targets across phases based on end-tidal capnography and ABG analyses.

FIG. 1.

Flowchart of the experimental design. The figure illustrates the experimental design of this pig study. Each phase included 30 min of monitoring. ICP, intracranial pressure.

First, in the healthy state, baseline data were collected during normoventilation (phase A), achieved with a respiratory rate of 20–25 breaths/min, tidal volume of 260–300 mL (8.9–10.0 mL/kg), and fraction of inspired oxygen (FiO₂) of 25%. Subsequently, moderate hyperventilation (phase B) was induced by increasing the respiratory rate by ∼10 breaths/min, while maintaining tidal volume unchanged and reducing FiO₂ by 1%. This was followed by severe hyperventilation (phase C), achieved by increasing the respiratory rate to 55 breaths/min and raising tidal volume by 20 mL (≈ 0.7 mL/kg), with a further 1% reduction in FiO₂. Thereafter, hypoventilation (phase D) was induced by lowering the respiratory rate to 10–15 breaths/min, decreasing tidal volume by 40 mL (≈ 1.4 mL/kg) relative to baseline, and increasing FiO₂ by approximately 10%. Second, to induce a compromised brain state, the epidurally placed balloon catheter was inflated with approximately 3–7 mL of water to induce intracranial hypertension (ICP ≈ 30–40 mmHg) and a CPP of ≈ 30 mmHg, thresholds that have previously been associated with cerebral hypoxia and cerebral energy metabolic decompensation using this model.³² Balloon adjustments were individualized based on downstream multimodal monitoring to ensure significant, but not critically irreversible ischemia, accounting for variations in baseline ABP and intracranial compliance across animals. After stabilization of ICP, the same ventilatory manipulations were repeated: baseline normoventilation (phase E), moderate hyperventilation (phase F), severe hyperventilation (phase G), and hypoventilation (phase H), following the same respiratory settings as mentioned above.

Monitoring, data acquisition, and analysis

Systemic (ABP, ABG, SpO₂, and CO) and cerebral variables (ICP, CPP, CBF, pbtO₂, and cerebral energy metabolism) were continuously or intermittently monitored. ABP was continuously measured via an arterial line (Careflow^TM, Merit Medical, Singapore Pte. Ltd., Singapore) in the right carotid artery. Blood samples were taken every 15 min through a separate femoral arterial line for ABG analysis (ABL800 Flex, Version 6.20 Build 1068; TRIOLAB AB, Mölndal, Sweden). The estimated plasma osmolality was calculated as “2[Na] + [Glucose]”,³³ i.e., using the traditional formula without urea [Urea] in blood since it was not available. Cardiac output was measured every 15 min using intermittent bolus thermodilution method via a pulmonary artery catheter (Merit Medical Systems, USA). Systemic vascular resistance (SVR) was approximated as mean arterial blood pressure (MAP)/cardiac output*80,³⁴ i.e., using the traditional formula but without central venous pressure which was not recorded. ICP and pbtO₂ were measured using a Neurovent PTO probe (Raumedic, Germany), inserted together with the MD catheter into the brain parenchyma in separate directions via a burr hole. Regional CBF and brain temperature were assessed with a thermal diffusion probe (Bowman Perfusion Monitor, Hemedex, USA) inserted through a separate burr hole using a bolt for stabilization of the probe. Signals were collected using the Moberg CNS Monitor (USA), with artifacts removed manually and automatically. CPP was calculated as the difference between MAP and ICP. Three CA metrics were also calculated. PRx, a global CA metric, was calculated as the Pearson correlation coefficient between 10-sec averages of ABP and ICP over 5 min.³⁵ High PRx indicates impaired CA, e.g., when an increase in ABP leads to increased ICP due to pressure passive vessels, whereas a low PRx indicates preserved CA.³⁵ The oxygen reactivity index (ORx), a focal pbtO₂-based CA-metric, was calculated as the Pearson correlation coefficient of pbtO₂ and CPP based on 30 sec values over 30 min, i.e., shorter than the traditional 60-min calculations to match the length of each study phase.³⁶ High ORx values, i.e., positive correlations between CPP and pbtO₂, suggest impaired CA, as changes in perfusion pressure directly influence oxygenation. In contrast, near-zero ORx values indicate preserved autoregulation, where oxygen delivery remains stable despite CPP fluctuations. The physiological significance of negative ORx values remains unclear.³⁷ We also calculated a similar index (CBFx) based on the Pearson correlation coefficient of CBF and CPP using 30 sec values over each 30-min phase.^31,38 Positive CBFx values indicate impaired CA, where CBF varies passively with changes in CPP. Near-zero values suggest intact CA, with stable CBF despite CPP fluctuations. The physiological interpretation of negative CBFx values remains uncertain.³¹ MD was performed using 71 High Cut-Off catheters (M Dialysis AB, Sweden; 100 kDa membrane, 10 mm length) perfused at 2 µL/min with artificial cerebrospinal fluid. Samples were collected every 15 min and stored at −20°C until analysis. Metabolites (glucose, lactate, pyruvate, and urea) were quantified using ISCUSflex (M Dialysis AB), and the lactate/pyruvate ratio (LPR) was calculated. Brain tissue urea was assumed to remain stable throughout the experiment and used as an internal quality control for the MD measurements.³⁹

Statistical analysis

Statistical analyses were performed in RStudio (version 2022.12.0). For each physiological variable, median values were calculated per 30-min phase. Comparisons between normoventilation and the subsequent hyper- and hypoventilation phases were performed separately for the healthy and compromised brain states, using the normoventilation period preceding each condition as the respective baseline. Statistical analysis was conducted using the Wilcoxon signed-rank test. Additionally, the Mann–Whitney U test was used to compare physiological variables within phases G and H between animals that showed improvement versus deterioration in CBF following hypoventilation in the compromised brain state. A p value <0.05 was considered statistically significant.

Ethics

The study was approved by the Animal Ethics Committee in Uppsala, Sweden (Dnr 5.8.18–21799/2022) and conducted in accordance with the European Communities Council Directive (2010/63/EU) at the Hedenstierna Laboratory, Uppsala University. All institutional and national guidelines for animal care and use were followed. The study complies with the ARRIVE 2.0 guidelines for the reporting of animal experiments.

Results

The effects of hyper- and hypoventilation on brain physiology in a healthy brain state

The pCO₂ levels were maintained within the targeted ranges during each phase (Table 1 and Fig. 2), with median values of 5.4 (IQR 5.3–5.6) kPa in phase A (normoventilation), 4.2 (IQR 4.1–4.3) kPa in phase B (moderate hyperventilation), 2.8 (IQR 2.7–2.9) kPa in phase C (severe hyperventilation), and 7.7 (IQR 7.7–8.1) kPa in phase D (hypoventilation). Moderate and severe hyperventilation (phases B and C) induced alkalosis, while hypoventilation (phase D) resulted in acidosis (all p < 0.01). Sodium levels and estimated plasma osmolality decreased significantly during the hyperventilation phases (p < 0.05). Arterial glucose levels decreased while lactate levels increased in response to hyperventilation, with partial normalization during hypoventilation. Severe hyperventilation also led to significant reductions in MAP (p < 0.05) and SVR (p < 0.01), accompanied by increases in heart rate and cardiac output (p < 0.05).

FIG. 2.

Dynamic changes in systemic and cerebral physiology in relation to pCO₂ in a healthy brain state—a multimodality monitoring analyses. Each phase corresponds to a certain pCO₂ ventilatory target; A (normoventilation; 4.5–6.0 kPa), B (moderate hyperventilation; 4.0–4.5 kPa), C (severe hyperventilation; 2.5–3.5 kPa), and D (hypoventilation; 7.0–8.0 kPa). The subfigures A to J represent each of the physiological variables of interest. ABG, Arterial blood gas; CBF, Cerebral blood flow; CBFx, CBF index; CPP, Cerebral perfusion pressure; ICP, Intracranial pressure; LPR, Lactate-pyruvate ratio; MD, Microdialysis; ORx, Oxygen reactivity index; PbtO₂, Partial pressure of brain tissue oxygenation; pCO₂, Partial pressure of carbon dioxide; PRx, Pressure reactivity index.

Table 1.

Physiological Changes in Relation to pCO₂ in the Healthy Brain State

Variable	Healthy state
Variable	Normoventilation (phase A)	Moderate hyperventilation (phase B)	Severe hyperventilation (phase C)	Hypoventilation (phase D)
ABG
pH (scale)	7.46 (7.43–7.48)	7.55 (7.53–7.57) ^b	7.69 (7.65–7.70) ^b	7.31 (7.30–7.33) ^b
pCO₂ (kPa)	5.4 (5.3–5.6)	4.2 (4.1–4.3) ^b	2.8 (2.7–2.9) ^b	7.7 (7.7–8.1) ^b
pO₂ (kPa)	15.2 (14.4–15.6)	15.1 (14.8–15.4)	15 (14.4–15.3)	15.6 (15.3–16.8) ^a
Hb (g/L)	89 (82–91)	89 (81–90)	90 (84–92)	86 (79–89) ^b
Sodium (mM)	134 (134–136)	134 (133–135) ^a	132 (132–134) ^b	134 (133–134) ^a
Glucose (mM)	6.6 (5.9–7.3)	6.2 (5.8–7.2) ^b	5.9 (5.3–6.6) ^b	6.0 (5.3–7.1) ^a
Lactate (mM)	1.2 (0.8–1.7)	1.5 (1.3–1.8) ^a	2.1 (1.8–2.3) ^b	1.7 (1.6–2.2) ^a
Estimated osmolality (mmol/kg)	276 (274–278)	276 (272–276) ^b	271 (270–274) ^b	273 (273–275) ^a
Systemic physiology
HR (bpm)	81 (76–85)	83 (69–86)	88 (81–92) ^b	77 (73–81)
MAP (mmHg)	81 (74–87)	80 (74–85)	73 (71–84) ^a	79 (66–89)
Cardiac output (L/min)	3.2 (3.1–3.5)	3.6 (3.1–3.7)	3.7 (3.4–4.3) ^a	3.0 (2.7–3.5)
SVR (dyn*s/cm⁵)	2132 (1899–2239)	1969 (1599–2162)	1639 (1255–1855) ^b	2044 (1863–2337)
Cerebral physiology
Brain temperature (°C)	38.6 (38.1–38.8)	38.4 (38.2–38.5)	38.3 (38.2–38.4)	38.6 (38.3–38.8)
ICP (mmHg)	9 (7–12)	10 (8–13) ^a	11 (8–15) ^a	17 (11–21) ^b
CPP (mmHg)	73 (66–80)	68 (66–80)	65 (57–76) ^a	58 (52–70) ^a
PRx (coefficient)	−0.14 (−0.31–0.07)	−0.24 (−0.37–0.13)	−0.14 (−0.48–0.06)	−0.18 (−0.39–0.04)
ORx (coefficient)	−0.19 (−0.38–0.25)	0.00 (−0.09–0.21)	0.16 (−0.19–0.35)	−0.33 (−0.66–0.31)
CBFx (coefficient)	−0.02 (−0.11–0.39)	0.23 (−0.06–0.47)	0.32 (0.07–0.66)	−0.17 (−0.50–0.49)
CBF (mL/100 g/minute)	30 (25–37)	25 (20–33)	28 (14–32)	66 (58–73) ^b
PbtO₂ (mmHg)	12 (10–15)	11 (9–15)	11 (8–14)	18 (13–22)
Relative pbtO₂ (%)	100 (100–100)	100 (94–108)	96 (88–105)	148 (143–170) ^a
MD-glucose (mM)	0.41 (0.29–0.51)	0.45 (0.28–0.66)	0.36 (0.24–0.62)	0.35 (0.26–0.72)
MD-pyruvate(µM)	26 (15–48)	27 (15–40)	31 (19–47)	25 (19–38)
MD-lactate (mM)	0.55 (0.44–0.89)	0.52 (0.44–0.67)	0.64 (0.59–0.87)	0.68 (0.59–0.72)
MD-LPR (ratio)	27 (20–34)	21 (15–30)	24 (16–35)	24 (17–37)

The table shows median values (interquartile ranges) and compares phase A (normoventilation) with phase B (moderate hyperventilation), phase C (severe hyperventilation), and phase D (hypoventilation), respectively, using the Wilcoxon signed-rank test. Bold and italics indicate statistical significance.

^ap value <0.05. ^bp value <0.01. Relative pbtO₂ (%) indicates the pbtO₂ value at each phase in relation to the baseline value at normoventilation (phase A).

ABG, arterial blood gas; CBF, cerebral blood flow; CBFx, CBF index; CPP, cerebral perfusion pressure; Hb, hemoglobin; HR, heart rate; ICP, intracranial pressure; MAP, mean arterial blood pressure; LPR, lactate-pyruvate ratio; MD, microdialysis; ORx, oxygen reactivity index; PbtO₂, partial pressure of brain tissue oxygenation. pCO₂, partial pressure of carbon dioxide; PRx, pressure reactivity index; SVR, systemic vascular resistance.

Regarding cerebral physiology, ICP increased modestly during hyperventilation from 9 (phase A) to 10 and 11 mmHg (in phases B and C, respectively), and more markedly during hypoventilation (phase D) to 17 mmHg. This was associated with a reduction in CPP during both severe hyperventilation and hypoventilation (p < 0.05). The autoregulatory metrics, PRx, ORx, and CBFx, showed variable changes, but no consistent or significant alterations across the phases. CBF showed a non-significant numerical decrease during both moderate and severe hyperventilation (p > 0.05), whereas it increased more than twofold during hypoventilation compared to normoventilation (p < 0.01). There was a significant increase in relative (%) pbtO₂ (the pbtO₂ in the corresponding phase divided by the baseline value during normoventilation) but not when assessed as absolute pbtO₂ values (mmHg), during hypoventilation. The MD variables of cerebral energy metabolism remained largely stable across all phases.

The effects of hyper- and hypoventilation on brain physiology in a compromised brain state

In the compromised brain state (elevated ICP), the targeted arterial pCO₂ levels were also achieved (Table 2 and Fig. 3), with median values of 5.1 (IQR 5.0–5.3) kPa in phase E (normoventilation), 4.0 (IQR 3.9–4.2) kPa during phase F (moderate hyperventilation), 2.7 (IQR 2.6–2.8) kPa during phase G (severe hyperventilation), and 7.7 (IQR 7.6–8.1) kPa during phase H (hypoventilation). Similar to the healthy state, hyperventilation led to alkalosis (p < 0.01) and hypoventilation to acidosis, although without reaching statistical significance (p = 0.103). Sodium, estimated osmolality, glucose, and lactate in blood also showed similar dynamics for ventilatory phases as in the healthy state. Regarding systemic physiology, MAP and SVR decreased during both severe hyperventilation and hypoventilation (p < 0.05), whereas heart rate and cardiac output remained relatively stable.

FIG. 3.

Dynamic changes in systemic and cerebral physiology in relation to pCO₂ in a compromised brain state—a multimodality monitoring analyses. Each phase corresponds to a certain pCO₂ ventilatory target; E (normoventilation; 4.5–6.0 kPa), F (moderate hyperventilation; 4.0–4.5 kPa), G (severe hyperventilation; 2.5–3.5 kPa), and H (hypoventilation; 7.0–8.0 kPa). Extreme LPR values from Pig 6 in phases G and H were excluded from the graph to preserve visual clarity. In pig 9 and 10, the pbtO₂ probe was manipulated after the completion of the healthy brain state phases and before the start of the compromised brain state phase, due to technical concerns (dislocation or extreme values during epidural balloon inflation). ABG, Arterial blood gas; CBF, Cerebral blood flow; CBFx, CBF index; CPP, Cerebral perfusion pressure; ICP, Intracranial pressure; LPR, Lactate-pyruvate ratio; MD, Microdialysis; ORx, Oxygen reactivity index; PbtO₂, Partial pressure of brain tissue oxygenation; pCO₂, Partial pressure of carbon dioxide; PRx, Pressure reactivity index.

Table 2.

Physiological Changes during Hypo/Hypercapnia in the Compromised Brain State

Variable	Compromised brain state
Variable	Normoventilation (phase E)	Moderate hyperventilation (phase F)	Severe hyperventilation (phase G)	Hypoventilation (phase H)
Arterial blood gas
pH (scale)	7.49 (7.46–7.49)	7.57 (7.55–7.58) ^b	7.71 (7.70–7.71) ^b	7.33 (7.30–7.35)
pCO₂ (kPa)	5.1 (5.0–5.3)	4.0 (3.9–4.2) ^b	2.7 (2.6–2.8) ^b	7.7 (7.6–8.1) ^b
pO₂ (kPa)	15.0 (14.7–15.4)	15.1 (14.4–15.4)	14.8 (13.9–15.1)	15.6 (14.8–16.8)
Hb (g/L)	85 (77–86)	84 (76–88)	85 (77–91)	84 (76–87)
Sodium (mM)	132 (130–134)	132 (131–133) ^a	131 (130–133) ^a	132 (131–134)
Glucose (mM)	6.3 (5.3–6.7)	6.0 (5.0–6.3) ^a	5.7 (4.8–5.8) ^b	5.7 (4.9–6.1) ^b
Lactate (mM)	1.3 (1.2–1.7)	1.5 (1.2–1.8)	1.9 (1.6–2.6) ^b	1.5 (1.2–2.1)
Estimated osmolality (mmol/kg)	270 (268–274)	268 (267–272) ^a	268 (265–271) ^b	269 (267–273)
Systemic physiology
HR (bpm)	79 (71–83)	79 (73–82)	80 (77–84)	75 (69–81)
MAP (mmHg)	75 (62–94)	77 (62–94)	72 (59–77) ^a	66 (58–73) ^a
Cardiac output (L/min)	3.3 (2.9–4.0)	3.4 (2.8–3.8)	3.5 (2.9–3.9)	3.2 (3.1–3.8)
SVR (dyn*s/cm⁵)	1965 (1532–2274)	1967 (1602–2177)	1683 (1271–1914) ^b	1501 (1324–1884) ^a
Cerebral physiology
Temperature (°C)	38.4 (38.2–38.8)	38.4 (38.2–38.7)	38.3 (38.2–38.5)	38.3 (38.1–38.7)
ICP (mmHg)	31 (28–42)	30 (28–37)	32 (29–37)	42 (40–48) ^b
CPP (mmHg)	48 (28–54)	50 (32–60)	40 (24–48)	19 (16–24) ^b
PRx (coefficient)	−0.19 (−0.40–0.05)	−0.49 (−0.55–−0.29) ^a	−0.60 (−0.65–−0.52)	0.57 (0.35–0.81) ^b
ORx (coefficient)	0.53 (0.41–0.76)	0.15 (0.08–0.70)	0.54 (0.14–0.77)	0.80 (0.69–0.84)
CBFx (coefficient)	0.66 (0.31–0.73)	0.39 (0.20–0.75)	0.62 (0.40–0.78)	0.73 (−0.24–0.89)
CBF (mL/100 g/minute)	27 (22–29)	25 (22–30)	20 (12–25)	12 (2–34)
PbtO₂ (mmHg)	8 (6–13)	9 (7–12)	6 (5–9) ^a	4 (3–5)
Relative pbtO₂ (%)	100 (100–100)	100 (94–106)	76 (60–92) ^a	63 (40–70) ^a
MD-glucose (mM)	0.14 (0.09–0.29)	0.13 (0.06–0.19) ^a	0.08 (0.05–0.14) ^b	0.06 (0.03–0.16) ^a
MD-pyruvate(µM)	23 (11–39)	24 (15–42)	26 (15–47)	18 (7–30)
MD-lactate (mM)	1.02 (0.86–1.23)	1.11 (0.91–1.32)	1.27 (1.05–1.65)	1.17 (0.97–1.79)
MD-LPR (ratio)	53 (23–85)	44 (22–76)	58 (22–81)	111 (28–188) ^a

The table shows median values (interquartile ranges) and compares phase E (normoventilation) with phase F (moderate hyperventilation, phase G (severe hyperventilation), and phase H (hypoventilation), respectively, using the Wilcoxon signed-rank test.

p value <0.05. ^bp value <0.01. Relative pbtO₂ (%) indicates the pbtO₂ value at each phase in relation to the baseline value at normoventilation (phase E).

ABG, arterial blood gas; CBF, cerebral blood flow; CBFx, CBF index; CPP, cerebral perfusion pressure; Hb, hemoglobin; HR, heart rate; ICP, intracranial pressure; MAP, mean arterial blood pressure; LPR, lactate-pyruvate ratio; MD, microdialysis; ORx, oxygen reactivity index; PbtO₂, partial pressure of brain tissue oxygenation; pCO₂, partial pressure of carbon dioxide; PRx, pressure reactivity index; SVR, systemic vascular resistance.

For the cerebral physiological variables, median ICP was 31 (IQR 28–42) mmHg during phase E (normoventilation), did not change significantly during phases F and G (hyperventilation), but increased to 42 (IQR 40–48) mmHg following hypoventilation in phase H (p < 0.01). The latter phase was also accompanied by a significant drop in CPP to a median value of 19 (IQR 16–24) mmHg (p < 0.01). PRx improved during phase F (moderate hyperventilation, p < 0.05) and deteriorated during phase H (hypoventilation, p < 0.01), while the other CA metrics did not change significantly across the phases. Notably, CBF decreased progressively, but non-significantly, from normo-, to hyper-, and hypoventilation. PbtO₂ decreased during both phase G (both in mmHg and % in relation to baseline following severe hyperventilation, p < 0.05) and during phase H (only in % in relation to baseline, p < 0.05). MD-glucose levels dropped during both (moderate and severe) hyperventilation and hypoventilation (p < 0.05), while MD-LPR increased significantly during hypoventilation (p < 0.05). The other MD-metabolites remained largely unchanged. Notably, the response to hypoventilation was heterogeneous: Four pigs exhibited improved CBF, whereas the remaining animals showed deterioration (Fig. 4). The former group had slightly higher CPP and maintained MAP during the transition from phase G to H, in contrast to the subgroup with reduced CBF (Supplementary Table S2 and Supplementary Fig. S2).

FIG. 4.

Heterogeneous response in relation to hypoventilation in a compromised brain state. The figure illustrates the individual physiological responses during the transition from phase G (severe hyperventilation) to phase H (hypoventilation). In all pigs, ICP increased and CPP decreased during hypoventilation. However, for pig 1 CPP remained above 30 mmHg in phase H and both CBF and pbtO₂ improved during this phase. In contrast, pig 8 started with a lower CPP, which dropped below 30 mmHg during phase H, resulting in a marked, but partially reversible, decline in CBF and pbtO₂. Pig 10 showed dynamic changes in ICP and CPP, with CPP probably just around the lower limit of autoregulation, accompanied by highly variable changes in CBF and pbtO₂. CPP, Cerebral perfusion pressure; ICP, Intracranial pressure; PbtO₂, Partial pressure of brain tissue oxygenation.

Discussion

This experimental pig study with state-of-the-art neuromonitoring demonstrates that arterial pCO₂ has a strong and complex impact on systemic and cerebral physiology. In the healthy brain, hyperventilation caused modest reductions in CBF, whereas hypoventilation elicited marked increases; however, CA, pbtO₂, and cerebral energy metabolism remained stable. In the compromised brain with elevated ICP, moderate hyperventilation improved CA, with a slight decline in cerebral glucose levels, but preserved pbtO₂ and LPR, while severe hyperventilation lowered pbtO₂. Hypoventilation in the compromised brain state led to diverging physiological responses depending on the balance between ICP and CPP. These findings highlight the importance of individualized ventilation strategies in NIC and the role of multimodal monitoring to assess the net cerebral impact of pCO₂ modulation on the different aspects of brain physiology.

pCO₂ levels in relation to systemic physiology

Before discussing the cerebral effects, the systemic consequences of altered pCO₂ must be acknowledged. Hyperventilation resulted in decreased MAP and SVR, while heart rate and cardiac output typically increased. Mechanical hyperventilation is expected to decrease MAP due to a reduction in right and left ventricle preloads due to increased intrathoracic pressure.⁴⁰ Interestingly, cardiac output increased by 10–15% during hyperventilation despite a decrease in MAP in the healthy state. This may be due to measurement inaccuracies related to the high ventilation settings affecting thermodilution via the pulmonary artery catheter,⁴¹ or a complex hemodynamic response involving changes in pre-/afterload (with peripheral vasodilation) and autonomic regulation.^42,43 Notably, in contrast to the healthy condition, while SVR still decreased, there was no increase in cardiac output observed during severe hyperventilation in the compromised brain state. Although the exact mechanisms remain unclear, we speculate that impaired autonomic compensation may contribute to this blunted cardiac response. Also, in the compromised brain state, hypoventilation was associated with a significant decrease in SVR, in contrast to the healthy state when SVR remained unchanged. Again, while we lack definitive explanations, it is possible that brain injury induced a stress response and impacted on autonomic function, which modulated the complex systemic vascular response to hypercapnia.⁴⁴

The hyperventilation phases also led to decreased arterial glucose and increased lactate levels, likely due to alkalosis-induced stimulation of phosphofructokinase activity leading to increased glycolysis.⁴⁵ Furthermore, both sodium and the estimated plasma osmolality decreased during hypocapnia, which may be explained by an alkalosis-induced intracellular sodium shift and antidiuretic hormone-mediated water retention.^46,47 Furthermore, pCO₂ critically regulates acid–base balance, which in turn alters hemoglobin—oxygen affinity via the hemoglobin—oxygen dissociation curve. Hypocapnia with alkalosis shifts the curve leftward, impairing oxygen release, while hypercapnia shifts it rightward, enhancing tissue oxygen delivery.⁴⁸ Taken together, these systemic changes provide essential context for interpreting the cerebral effects of different ventilatory settings and pCO₂ levels.

pCO₂ levels and ICP dynamics

Hyperventilation is expected to induce cerebral vasoconstriction,¹¹ leading to reduced CBV⁴⁹ and lower ICP.^8,21 However, in the healthy brain state, both moderate and severe hyperventilation were paradoxically associated with modest, but significant, ICP elevations. One plausible explanation is that the pigs maintained intact CA, as indicated by low PRx, CBFx, and ORx, while experiencing a slight drop in MAP. In such a scenario, the vasoconstrictive effects of hyperventilation-induced hypocapnia may have been offset by autoregulatory vasodilation triggered by reduced perfusion pressure, leading to a net increase or preservation rather than reduction of CBV. In the compromised brain state, ICP remained unchanged during moderate and severe hyperventilation. In this setting, the smaller MAP reduction compared to the healthy state, combined with impaired autoregulation, likely reduced the MAP-driven impact on ICP, thereby allowing the vasoconstrictive effects of hypocapnia to play a more dominant role. However, hypoventilation consistently resulted in marked ICP elevations in both the healthy and compromised brain, with the most pronounced increase observed in the injured state, i.e., when the intracranial compliance was low. Notably, despite high ICP and low CPP indicative of conditions near the lower limit of autoregulation, hypoventilation still induced an ICP rise, suggesting that some vasodilatory reserve remained despite approaching exhaustion. Altogether, these findings underscore that while hyperventilation is often used at the bedside,¹⁸ its efficacy to reduce ICP may be limited or even counteracted under certain physiological conditions, either due to insufficient vasoconstrictive effect or autoregulatory compensation in response to concurrent reduction in MAP. However, hypoventilation triggered a detrimental vasodilatory cascade that often resulted in critical cerebral hypoperfusion in the compromised brain state, underscoring the potential danger of hypercapnia in patients with impaired intracranial compliance, elevated ICP, and low MAP.

pCO₂ levels in relation to cerebral autoregulation

Neither hyper- nor hypoventilation significantly altered any of the autoregulatory indices, PRx, CBFx, and ORx, in the healthy brain state, indicating preserved vascular reactivity across a wide pCO₂ range. Although pCO₂ can shift both the limits and position of the autoregulatory plateau,^12,22,50 it is not surprising that CA remained intact under these optimized, healthy conditions with adequate MAP and CPP. However, in the injured state with compromised ICP and CPP, PRx decreased during moderate hyperventilation, suggesting improved CA. A similar trend was observed during severe hyperventilation, although statistical significance was not reached, likely due to hyperventilation-induced systemic hypotension in a subset of the pigs. These findings are consistent with clinical TBI studies reporting improved PRx during hyperventilation.^13,24 Notably, CBFx and ORx remained unchanged across all ventilatory phases. To understand these discrepancies, it is important to note that these three indices are surrogates of dynamic CA and are based on different signal inputs of cerebral physiology.^7,51 PRx, which reflects the correlation between MAP and ICP, is closely influenced by changes in cerebrovascular tone and CBV, which probably makes it particularly sensitive to the direct effects of pCO₂ on the vessels of the brain. Hyperventilation likely induced vasoconstriction in previously dilated vessels close to the lower limit of autoregulation in the compromised brain state, potentially restoring vasodilatory reserve and improving dynamic pressure reactivity. However, while this led to lower PRx values, CBF decreased slightly numerically although not significantly, likely due to a down-shift of the autoregulatory plateau.^12,22 As such, downstream variables like CBF may still decline despite apparent autoregulatory improvement. Otherwise, CBFx, reflecting the correlation between CPP and CBF, may be a less reliable surrogate of dynamic CA function during these acute pCO₂ changes, as CBF during ventilatory manipulation was likely more influenced by pCO₂ mediated changes in vascular tone rather than changes in perfusion pressure. Similarly, ORx, reflecting the correlation between CPP and pbtO₂, was likely only minimally affected, as pbtO₂ changed relatively little during these phases. Such pbtO₂ changes were probably also more driven by pCO₂ mediated effects on CBF. Lastly, hypoventilation significantly worsened PRx in the injured brain, most likely due to elevated ICP and critically low CPP, leading to pressure-passive cerebral perfusion. This deterioration highlights the vulnerability of cerebrovascular control in states of compromised intracranial compliance.

pCO₂ in relation to CBF, pbtO₂, and cerebral energy metabolism

Moderate hyperventilation resulted in a slight numerical, but non-significant, reduction in CBF in the healthy brain, while pbtO₂ and cerebral energy metabolism remained largely stable. In the injured brain, a similar, non-significant trend with lower CBF was seen during moderate hyperventilation, while downstream physiological variables were generally preserved, except for a slight, but significant decrease in cerebral glucose. Severe hyperventilation led to further, but non-significant, reductions in CBF in both states, but only in the injured brain did this translate into significantly lower pbtO₂ and lower cerebral glucose. Despite these changes, cerebral energy metabolism (LPR levels) remained largely unchanged. These findings suggest that the CBF reductions were modest in magnitude and did not reach a threshold of ischemic compromise, at least not at the group level, consistent with several clinical TBI studies.^13,19–21 It is worth noting that technical limitations, such as transient instability in CBF monitoring, may have influenced measurement accuracy and reduced statistical power in some comparisons. Nevertheless, the stability of pbtO₂ and LPR across most conditions supports the interpretation that cerebral oxygen delivery and metabolic function were not critically impaired during short-term hyperventilation, except for severe hyperventilation in the compromised brain.

The effects of hypoventilation were more complex. In the healthy brain, it consistently increased CBF and tended to improve pbtO₂ (in %change, but not mmHg, in relation to baseline), while cerebral metabolism remained unchanged. This suggests intact CO₂-reactivity, with pbtO₂ (to some extent) and cerebral glucose stability possibly reflecting already saturated oxygen and glucose gradients, a balanced increase in metabolic demand, or an early stage of “luxury perfusion”. In contrast, the compromised brain exhibited highly heterogeneous responses to hypoventilation. On group level, hypercapnia was associated with reductions in CBF (numerical, but non-significant) and pbtO₂ (p < 0.05), as well as decreased cerebral glucose and elevated LPR, suggesting impaired oxygen and substrate delivery due to low CPP. However, subgroup analysis revealed two distinct response patterns. In pigs with lower baseline MAP and CPP or where MAP dropped markedly due to hypoventilation, CBF, pbtO₂, and metabolic markers deteriorated substantially. Conversely, in animals with elevated ICP but more preserved MAP and CPP, hypercapnia was associated with increased CBF, improved oxygenation, and more favorable metabolic profiles. These findings suggest that hypoventilation in the compromised brain can exert diametrically opposing physiological effects depending on the ICP dynamics and hemodynamic conditions.

Implications

In acute brain injuries, normocapnia remains the standard target for ventilatory management.⁵² However, hyperventilation is commonly employed as a temporizing measure to lower ICP.^13,18,21,52 While this therapy usually reduces ICP in a clinical setting,²¹ our data also highlight that systemic hypotension may counteract any ICP-lowering benefit. This may be particularly relevant in pre- or interhospital settings, where continuous ABP monitoring and management may be more limited. Notably, moderate hyperventilation appeared to improve CA function, consistent with previous clinical findings in TBI^13,23,24 and may therefore offer physiological benefits beyond ICP control. The preservation of CBF, pbtO₂, and LPR further suggests that this strategy is likely safe, particularly when accompanied by careful MAP management. However, severe hyperventilation led to more severe downstream effects on CBF, pbtO₂, and energy metabolism, consistent with previous clinical studies^19,53 and is likely detrimental.

Hypoventilation, on the other hand, exhibited highly variable effects that were strongly dependent on the underlying systemic and cerebral physiological state. In the compromised brain state, characterized by reduced intracranial compliance and impaired autoregulation, elevated pCO₂ exacerbated ICP, reduced CPP, and was associated with impaired CA function, lower pbtO₂, and metabolic distress. These results reinforce the clinical caution against hypoventilation in conditions such as TBI, especially in the early phases of care when the risk of ICP compromise is greatest. Early airway protection and avoidance of hypoventilation are therefore crucial to prevent secondary cerebral injury in these patients. Conversely, in acute brain injuries not primarily dominated by raised ICP, such as aSAH, the situation may be different. In these cases, relatively preserved intracranial compliance may allow for a safe increase in CBV without critical ICP elevations. Thus, the net effect of hypoventilation may shift toward improved CBF and oxygenation, as previously suggested by small, preliminary clinical studies.^5,27,28 While speculative, our findings lend support to further investigation of targeted hypercapnia as a therapeutic strategy in selected patient populations characterized by hypoperfusion without major ICP problems.

Lastly, controlled hyperventilation is frequently used during elective neurosurgical procedures to reduce CBV, relax the brain, and facilitate safer craniotomy, dural opening, and brain dissection.⁵⁴ In this study, brief periods of moderate hypocapnia in the healthy brain were not associated with deleterious effects on CA, CBF, pbtO₂, or cerebral energy metabolism, suggesting that this strategy is physiologically well-tolerated. These findings support the safety of short-term moderate hyperventilation during elective neurosurgery under controlled conditions.

Methodological considerations

This study has several methodological strengths. The controlled experimental design enabled clearer causal inference compared with observational clinical studies. The use of juvenile pigs enhances the translational relevance of our findings, as their gyrencephalic brains and cerebrovascular anatomy more closely resemble those of humans compared to small, lissencephalic species.^55,56 Moreover, pigs share key features of human cardiorespiratory physiology, brain growth, and developmental trajectories, and their size allows for advanced multimodal monitoring, making them well suited for NIC research.^55,57 Furthermore, the use of a state-of-the-art multimodal neuromonitoring platform, encompassing ICP, CPP, CBF, pbtO₂, and cerebral MD, allowed for comprehensive evaluation of the cerebral physiological cascade under different pCO₂ conditions. This mirrors the setup of modern NIC units and strengthens the applicability of our results to clinical practice. Nevertheless, the study has some limitations. The number of pigs was limited, increasing the risk of both type I and type II errors. However, the high degree of physiological control in the experimental setup likely minimized variability and reduced noise. Additionally, we used median values and non-parametric statistics to mitigate the influence of outliers, thereby enhancing the robustness of observed significant differences. Furthermore, limited reliability of intracranial monitors represented a potential source of measurement error. In particular, the thermal diffusion probe used for CBF monitoring required intermittent re-calibration, which occasionally introduced baseline shifts. These fluctuations may have influenced some of the CBF comparisons. Moreover, while all monitoring devices were allowed to stabilize for approximately one hour prior to the start of the protocol, certain probes, such as the Neurovent PTO, may require up to 8 h to reach full equilibration, potentially affecting the reliability of absolute pbtO₂ readings. Furthermore, in a few cases (pig 9 and 10), minor adjustments were made to the pbtO₂ probe between phase D and E, i.e., after completion of the healthy brain state and before the start of the compromised brain state phases, due to technical concerns about the probe (dislocation or extreme values). This may have influenced absolute pbtO₂ values at the start of the compromised brain phase. However, the same probe was consistently used within each experimental period (healthy and compromised brain state, respectively), and all analyses were performed relative to the respective normoventilation baseline. Additionally, in the compromised brain state, a fully standardized level of compromise could not be achieved due to interindividual differences among the pigs, particularly in baseline MAP. Small changes in balloon volume caused large, variable shifts in ICP, reflecting reduced intracranial compliance and making it difficult to match ICP dynamics and CPP across pigs. This variability was most evident during hypoventilation, where physiological responses differed markedly with small changes in pressure. Although this reduced the overall homogeneity of results, it closely reflects the clinical reality and highlights how subtle physiological differences can lead to divergent responses to the same intervention. While we considered actively adjusting MAP to reduce variability, doing so would have masked the full systemic effects of the ventilatory manipulations and introduced confounding through the use of vasopressors, making it more difficult to isolate the specific impact of pCO₂ on cerebral physiology. Instead, our study design prioritized capturing the integrated physiological response to pCO₂ modulation, encompassing both systemic and cerebral effects. Moreover, each ventilatory phase was maintained for 30 min, allowing assessment of acute physiological responses. This time frame reflects clinically relevant decision-making windows but may not fully capture the effects of sustained pCO₂ alterations, particularly regarding cerebral metabolism. On one hand, longer durations might be necessary to elicit measurable changes in energy metabolic markers. On the other, the physiological impact of altered ventilation may diminish over time, as the acid-base effects of pCO₂ typically attenuate within a few hours due to compensatory mechanisms. Also, the analgosedation protocol used in this study, including ketamine, midazolam, and fentanyl, was optimized for experimental conditions and hemodynamic stability. However, we acknowledge that other sedatives such as propofol are more commonly used in clinical NIC. The choice of sedative may therefore influence the generalizability of the findings. Finally, while both male and female pigs were included, the animals were prepubescent (2–3 months old), and no systematic sex differences were observed. Nonetheless, we cannot exclude the possibility of subtle sex-dependent physiological differences that might become more apparent in sexually mature subjects.

Conclusions

This experimental pig study with state-of-the-art neuromonitoring illustrates that arterial pCO₂ exerts a strong and complex influence on both systemic and cerebral physiology. In the healthy brain, pCO₂ alterations induced predictable changes in CBF, with preserved CA, oxygenation, and energy metabolism. In contrast, the injured brain showed a more complex pattern; moderate hyperventilation improved CA while maintaining stable CBF, pbtO₂, and LPR, whereas severe hyperventilation and hypoventilation were associated with impaired oxygenation and signs of metabolic stress, particularly under conditions of reduced CPP. However, hypoventilation elicited divergent responses depending on the balance between ICP and MAP in the compromised brain state. These findings emphasize the need for individualized, physiology-based ventilation strategies in NIC and support the use of multimodal monitoring to guide safe and effective pCO₂ management. Future research should explore whether individual multimodal physiological characteristics, such as ABP, ICP, intracranial compliance, and CA function, could help guide pCO₂-targeted interventions.

Transparency, Rigor, and Reproducibility Statement

This experimental study was conducted in accordance with international guidelines for animal research and approved by the appropriate institutional ethics committee. To ensure transparency, detailed descriptions of animal preparation, anesthesia protocols, ventilatory settings, and multimodal neuromonitoring methods are provided in the Materials and Methods section. All data were collected prospectively according to a pre-defined protocol. Cerebral physiological parameters were continuously recorded using validated equipment with synchronized time-stamping, enabling precise temporal alignment with changes in arterial pCO₂. Rigor was maintained by standardizing experimental procedures across all animals. Data preprocessing and statistical methods are clearly outlined to facilitate reproducibility. Raw data are available upon reasonable request.

Authors’ Contributions

Conceptualization: T.S.W. Methodology: All authors. Formal analysis: M.A. Resources: T.S.W. Data curation: All authors. Writing—original draft preparation: T.S.W. Writing—review and editing: All authors.

Footnotes

Acknowledgment

The authors would like to thank Hedenstierna lab for excellent support during the experiments, Henrik Nytén at PO Medica for helping them set up the neuromonitoring tools, and Katarina Uisk for support with performing the analyses of the MD vials. The study was funded by “Erik, Karin, and Gösta Selanders stiftelse” and “Jeanssons stiftelse.”

Author Disclosure Statement

No competing financial interests exist.

Funding Information

The study was funded by “Erik, Karin, and Gösta Selanders stiftelse” and “Jeanssons stiftelse.”

Supplemental Material

References

Svedung Wettervik

, Lewén

, Enblad

. Fine tuning of neurointensive care in aneurysmal subarachnoid hemorrhage: From one-size-fits-all towards individualized care. World Neurosurg X, 2023; 18:100160; doi: 10.1016/j.wnsx.2023.100160

Svedung Wettervik

, Lewén

, Enblad

. Fine tuning of traumatic brain injury management in neurointensive care-indicative observations and future perspectives. Front Neurol, 2021; 12:638132; doi: 10.3389/fneur.2021.638132

Carney

, Totten

, O’Reilly

, et al. Guidelines for the management of severe traumatic brain injury, fourth edition. Neurosurgery, 2017; 80(1):6–15; doi: 10.1227/neu.0000000000001432

Treggiari

, Rabinstein

, Busl

, et al. Guidelines for the neurocritical care management of aneurysmal subarachnoid hemorrhage. Neurocrit Care, 2023; 39(1):1–28; doi: 10.1007/s12028-023-01713-5

Svedung Wettervik

, Engquist

, Hånell

, et al. Cerebral blood flow and oxygen delivery in aneurysmal subarachnoid hemorrhage: Relation to neurointensive care targets. Neurocrit Care, 2022; 37(1):281–292; doi: 10.1007/s12028-022-01496-1

Lazaridis

. Brain shock-toward pathophysiologic phenotyping in traumatic brain injury. Crit Care Explor, 2022; 4(7):e0724; doi: 10.1097/cce.0000000000000724

Svedung Wettervik

, Fahlström

, Enblad

, et al. Cerebral pressure autoregulation in brain injury and disorders-a review on monitoring, management, and future directions. World Neurosurg, 2022; 158:118–131; doi: 10.1016/j.wneu.2021.11.027

Godoy

, Rovegno

, Lazaridis

, et al. The effects of arterial CO(2) on the injured brain: Two faces of the same coin. J Crit Care, 2021; 61:207–215; doi: 10.1016/j.jcrc.2020.10.028

Yoon

, Zuccarello

, Rapoport

. pCO(2) and pH regulation of cerebral blood flow. Front Physiol, 2012; 3:365; doi: 10.3389/fphys.2012.00365

10.

Kontos

, Raper

, Patterson

. Analysis of vasoactivity of local pH, PCO2 and bicarbonate on pial vessels. Stroke, 1977; 8(3):358–360; doi: 10.1161/01.str.8.3.358

11.

Harper

, Glass

. Effect of alterations in the arterial carbon dioxide tension on the blood flow through the cerebral cortex at normal and low arterial blood pressures. J Neurol Neurosurg Psychiatry, 1965; 28(5):449–452; doi: 10.1136/jnnp.28.5.449

12.

Meng

, Gelb

. Regulation of cerebral autoregulation by carbon dioxide. Anesthesiology, 2015; 122(1):196–205; doi: 10.1097/aln.0000000000000506

13.

Svedung Wettervik

, Howells

, Hillered

, et al. Mild hyperventilation in traumatic brain injury-relation to cerebral energy metabolism, pressure autoregulation, and clinical outcome. World Neurosurg, 2020; 133:e567–e575; doi: 10.1016/j.wneu.2019.09.099

14.

Hauerberg

, Ma

, Bay-Hansen

, et al. Effects of alterations in arterial CO2 tension on cerebral blood flow during acute intracranial hypertension in rats. J Neurosurg Anesthesiol, 2001; 13(3):213–221; doi: 10.1097/00008506-200107000-00006

15.

Rudenberg

, McGraw

, Tindall

. Effects of hyperventilation, CO2, and CSF pressure on internal carotid blood flow in the baboon. J Neurosurg, 1976; 44(3):347–352; doi: 10.3171/jns.1976.44.3.0347

16.

Forbes

, Hendrich

, Kochanek

, et al. Assessment of cerebral blood flow and CO2 reactivity after controlled cortical impact by perfusion magnetic resonance imaging using arterial spin-labeling in rats. J Cereb Blood Flow Metab, 1997; 17(8):865–874; doi: 10.1097/00004647-199708000-00005

17.

Muizelaar

, Marmarou

, Ward

, et al. Adverse effects of prolonged hyperventilation in patients with severe head injury: A randomized clinical trial. J Neurosurg, 1991; 75(5):731–739; doi: 10.3171/jns.1991.75.5.0731

18.

Neumann

, Chambers

, Citerio

, et al.; BrainIT Group. The use of hyperventilation therapy after traumatic brain injury in Europe: An analysis of the BrainIT database. Intensive Care Med, 2008; 34(9):1676–1682; doi: 10.1007/s00134-008-1123-7

19.

Coles

, Fryer

, Coleman

, et al. Hyperventilation following head injury: Effect on ischemic burden and cerebral oxidative metabolism. Crit Care Med, 2007; 35(2):568–578; doi: 10.1097/01.Ccm.0000254066.37187.88

20.

Diringer

, Videen

, Yundt

, et al. Regional cerebrovascular and metabolic effects of hyperventilation after severe traumatic brain injury. J Neurosurg, 2002; 96(1):103–108; doi: 10.3171/jns.2002.96.1.0103

21.

Brandi

, Stocchetti

, Pagnamenta

, et al. Cerebral metabolism is not affected by moderate hyperventilation in patients with traumatic brain injury. Crit Care, 2019; 23(1):45; doi: 10.1186/s13054-018-2304-6

22.

Dietvorst

, Kempen

, De Sloovere

, et al. The Influence of different arterial carbon dioxide levels on the cerebrovascular autoregulation curve in a porcine cranial window Model. Neurocrit Care, 2025; doi: 10.1007/s12028-025-02250-z

23.

Haubrich

, Steiner

, Kim

, et al. How does moderate hypocapnia affect cerebral autoregulation in response to changes in perfusion pressure in TBI patients? Acta Neurochir (Suppl 2012); 114:153–156; doi: 10.1007/978-3-7091-0956-4_28

24.

Steiner

, Balestreri

, Johnston

, et al. Effects of moderate hyperventilation on cerebrovascular pressure-reactivity after head injury. Acta Neurochir (Suppl 2005); 95:17–20; doi: 10.1007/3-211-32318-x_4

25.

Claassen

, Park

. Spontaneous subarachnoid haemorrhage. Lancet, 2022; 400(10355):846–862; doi: 10.1016/s0140-6736(22)00938-2

26.

Hoh

, Ko

, Amin-Hanjani

, et al. 2023 guideline for the management of patients with aneurysmal subarachnoid hemorrhage: A guideline from the american heart association/american stroke association. Stroke, 2023; 54(7):e314–e370; doi: 10.1161/str.0000000000000436

27.

Stetter

, Weidner

, Lilla

, et al. Therapeutic hypercapnia for prevention of secondary ischemia after severe subarachnoid hemorrhage: Physiological responses to continuous hypercapnia. Sci Rep, 2021; 11(1):11715; doi: 10.1038/s41598-021-91007-7

28.

Westermaier

, Stetter

, Kunze

, et al. Controlled hypercapnia enhances cerebral blood flow and brain tissue oxygenation after aneurysmal subarachnoid hemorrhage: Results of a phase 1 study. Neurocrit Care, 2016; 25(2):205–214; doi: 10.1007/s12028-016-0246-x

29.

Darkwah Oppong

, Wrede

, Müller

, et al. PaCO2-management in the neuro-critical care of patients with subarachnoid hemorrhage. Sci Rep, 2021; 11(1):19191; doi: 10.1038/s41598-021-98462-2

30.

Svedung Wettervik

, Hånell

, Ahlgren

, et al. Preliminary Observations of the Loke Microdialysis in an Experimental Pig Model: Are We Ready for Continuous Monitoring of Brain Energy Metabolism? Neurocrit Care, 2025; 42(1):222–231; doi: 10.1007/s12028-024-02080-5

31.

Svedung Wettervik

, Hånell

, Ahlgren

, et al. The effects of hyperglycemia on brain physiology in a healthy and injured state: An experimental pig study with state-of-the-art multimodal neuromonitoring. J Cereb Blood Flow Metab, 2025:271678x251337633; doi: 10.1177/0271678x251337633

32.

Purins

, Enblad

, Wiklund

, et al. Brain tissue oxygenation and cerebral perfusion pressure thresholds of ischemia in a standardized pig brain death model. Neurocrit Care, 2012; 16(3):462–469; doi: 10.1007/s12028-012-9675-3

33.

Worthley

, Guerin

, Pain

. For calculating osmolality, the simplest formula is the best. Anaesth Intensive Care, 1987; 15(2):199–202.

34.

, Noh

, Hong

, et al. Is measurement of central venous pressure required to estimate systemic vascular resistance? A retrospective cohort study. BMC Anesthesiol, 2021; 21(1):310; doi: 10.1186/s12871-021-01522-3

35.

Czosnyka

, Smielewski

, Kirkpatrick

, et al. Continuous assessment of the cerebral vasomotor reactivity in head injury. Neurosurgery, 1997; 41(1):11–17; doi: 10.1097/00006123-199707000-00005

36.

Jaeger

, Schuhmann

, Soehle

, et al. Continuous assessment of cerebrovascular autoregulation after traumatic brain injury using brain tissue oxygen pressure reactivity. Crit Care Med, 2006; 34(6):1783–1788; doi: 10.1097/01.Ccm.0000218413.51546.9e

37.

Svedung Wettervik

, Beqiri

, Hånell

, et al. Revisiting the oxygen reactivity index in traumatic brain injury: The complementary value of combined focal and global autoregulation monitoring. Crit Care, 2025; 29(1):20; doi: 10.1186/s13054-025-05261-6

38.

Tackla

, Hinzman

, Foreman

, et al. Assessment of Cerebrovascular Autoregulation Using Regional Cerebral Blood Flow in Surgically Managed Brain Trauma Patients. Neurocrit Care, 2015; 23(3):339–346; doi: 10.1007/s12028-015-0146-5

39.

Ronne-Engström

, Cesarini

, Enblad

, et al. Intracerebral microdialysis in neurointensive care: The use of urea as an endogenous reference compound. J Neurosurg, 2001; 94(3):397–402; doi: 10.3171/jns.2001.94.3.0397

40.

Duke

. Cardiovascular effects of mechanical ventilation. Crit Care Resusc, 1999; 1(4):388–399.

41.

Stevens

, Raffin

, Mihm

, et al. Thermodilution cardiac output measurement. Effects of the respiratory cycle on its reproducibility. JAMA, 1985; 253(15):2240–2242; doi: 10.1001/jama.253.15.2240

42.

Richardson

, Wasserman

, Patterson

JL.

Jr ., General and regional circulatory responses to change in blood pH and carbon dioxide tension. J Clin Invest, 1961; 40(1):31–43; doi: 10.1172/jci104234

43.

Burnum

, Hickam

, Mc

. The effect of hypocapnia on arterial blood pressure. Circulation, 1954; 9(1):89–95; doi: 10.1161/01.cir.9.1.89

44.

Carr

, Ainslie

, Howe

, et al. Brachial artery responses to acute hypercapnia: The roles of shear stress and adrenergic tone. Exp Physiol, 2022; 107(12):1440–1453; doi: 10.1113/ep090690

45.

Kraut

, Madias

. Lactic acidosis. N Engl J Med, 2014; 371(24):2309–2319; doi: 10.1056/NEJMra1309483

46.

Elkinton

, Singer

, Barker

, et al. Effects in man of acute experimental respiratory alkalosis and acidosis on ionic transfers in the total body fluids. J Clin Invest, 1955; 34(11):1671–1690; doi: 10.1172/jci103221

47.

Stark

, Weinberger

, Ben-Bassat

. Persistent hyponatremia and inappropriate antidiuretic hormone secretion in children with extensive burns. J Pediatr Surg, 1979; 14(2):149–153; doi: 10.1016/0022-3468(79)90007-1

48.

Dash

, Korman

, Bassingthwaighte

. Simple accurate mathematical models of blood HbO2 and HbCO2 dissociation curves at varied physiological conditions: Evaluation and comparison with other models. Eur J Appl Physiol, 2016; 116(1):97–113; doi: 10.1007/s00421-015-3228-3

49.

Ito

, Ibaraki

, Kanno

, et al. Changes in the arterial fraction of human cerebral blood volume during hypercapnia and hypocapnia measured by positron emission tomography. J Cereb Blood Flow Metab, 2005; 25(7):852–857; doi: 10.1038/sj.jcbfm.9600076

50.

Nusbaum

, Brady

, Kibler

, et al. Acute hypercarbia increases the lower limit of cerebral blood flow autoregulation in a porcine model. Neurol Res, 2016; 38(3):196–204; doi: 10.1179/1743132815y.0000000094

51.

Zeiler

, Donnelly

, Menon

, et al. Continuous autoregulatory indices derived from multi-modal monitoring: Each one is not like the other. J Neurotrauma, 2017; 34(22):3070–3080; doi: 10.1089/neu.2017.5129

52.

Hawryluk

GWJ

, Aguilera

, Buki

, et al. A management algorithm for patients with intracranial pressure monitoring: The Seattle International Severe Traumatic Brain Injury Consensus Conference (SIBICC). Intensive Care Med, 2019; 45(12):1783–1794; doi: 10.1007/s00134-019-05805-9

53.

Marion

, Puccio

, Wisniewski

, et al. Effect of hyperventilation on extracellular concentrations of glutamate, lactate, pyruvate, and local cerebral blood flow in patients with severe traumatic brain injury. Crit Care Med, 2002; 30(12):2619–2625; doi: 10.1097/00003246-200212000-00001

54.

, Gelb

, Flexman

, et al. Definition, evaluation, and management of brain relaxation during craniotomy. Br J Anaesth, 2016; 116(6):759–769; doi: 10.1093/bja/aew096

55.

Duhaime

. Large animal models of traumatic injury to the immature brain. Dev Neurosci, 2006; 28(4–5):380–387; doi: 10.1159/000094164

56.

Melià-Sorolla

, Castaño

, DeGregorio-Rocasolano

, et al. Relevance of porcine stroke models to bridge the gap from pre-clinical findings to clinical implementation. Int J Mol Sci, 2020; 21(18):6568; doi: 10.3390/ijms21186568

57.

Klein

, De Sloovere

, Meyfroidt

, et al. Differential hemodynamic response of pial arterioles contributes to a quadriphasic cerebral autoregulation physiology. J Am Heart Assoc, 2022; 11(1):e022943; doi: 10.1161/jaha.121.022943

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.63 MB

The Effects of Hyper- and Hypoventilation on Cerebral Physiology in a Healthy and Compromised Brain State: An Experimental Pig Model with State-of-the-Art Neuromonitoring

Abstract

Keywords

Introduction

Materials and Methods

Animals

Anesthesia and mechanical ventilation

Study design

Monitoring, data acquisition, and analysis

Statistical analysis

Ethics

Results

The effects of hyper- and hypoventilation on brain physiology in a healthy brain state

The effects of hyper- and hypoventilation on brain physiology in a compromised brain state

Discussion

pCO2 levels in relation to systemic physiology

pCO2 levels and ICP dynamics

pCO2 levels in relation to cerebral autoregulation

pCO2 in relation to CBF, pbtO2, and cerebral energy metabolism

Implications

Methodological considerations

Conclusions

Transparency, Rigor, and Reproducibility Statement

Authors’ Contributions

Footnotes

Acknowledgment

Author Disclosure Statement

Funding Information

Supplemental Material

References

Supplementary Material

pCO₂ levels in relation to systemic physiology

pCO₂ levels and ICP dynamics

pCO₂ levels in relation to cerebral autoregulation

pCO₂ in relation to CBF, pbtO₂, and cerebral energy metabolism