Intravenous megadose sodium ascorbate normalises elevations in intracranial pressure and restores pressor responsiveness to norepinephrine in ovine Gram-negative sepsis Authors: Maya S Bishop, Ruslan V Pustovit, Alemayehu H Jufar, Darius J R Lane, Clive N May, Scott Ayton, Mark P Plummer, Connie P C Ow, Yugeesh R Lankadeva Published: 2026 ABSTRACT AbstractAcute brain dysfunction (delirium) is common in sepsis, and many survivors develop persistent cognitive impairment. Cerebral hypoperfusion and edema are implicated, yet it is unknown whether intracranial hypertension contributes to the neurological sequelae of sepsis, and no therapies specifically target cerebral hemodynamic dysfunction. Intravenous megadose sodium ascorbate has been reported to reverse cerebral tissue ischemia and hypoxia and reduce vasopressor requirements in preclinical and clinical studies. Here, we test for the first time the effects of sepsis and megadose sodium ascorbate on intracranial pressure (ICP), cerebral perfusion pressure (CPP) and pressor responsiveness to norepinephrine in a clinically relevant ovine model of Gram-negative sepsis. Adult Merino ewes were surgically instrumented with lateral cerebral ventricular catheters and carotid arterial and jugular venous catheters. Non-anesthetized sheep were infused with intravenous live Escherichia coli for 31-h. After 23-h of sepsis animals received bolus fluid resuscitation (30 mL/kg from 23-23.5-h; sodium lactate) and were randomized to receive intravenous 3 g/kg of sodium ascorbate (n = 7) or fluid-matched placebo (n = 6), as a bolus followed by a 7-h infusion. ICP and mean arterial pressure (MAP) were measured continuously. Changes in MAP to escalating norepinephrine doses were recorded at baseline and 23-h of sepsis, and after treatment at 27.5 and 31-h of sepsis. Sepsis reduced MAP and CPP and increased heart rate and ICP (all P < 0.01). By 31-h of sepsis, sodium ascorbate improved MAP compared with placebo (75 ± 5 to 87 ± 11 vs. 69 ± 11 to 66 ± 7 mmHg; P = 0.020). In addition, sodium ascorbate normalised ICP, whereas it remained elevated with placebo (17.0 ± 3.4 to 9.3 ± 1.9 vs. 20.6 ± 3.7 to 19.3 ± 5.0 mmHg; P = 0.024). Sodium ascorbate improved CPP versus no effect with placebo (61 ± 11 to 75 ± 5 mmHg vs. 48 ± 16 to 49 ± 12 mmHg; P = 0.046). MAP responsiveness to norepinephrine was impaired at 23-h of sepsis in both groups and was fully restored at 27.5- and 31-h of sepsis by sodium ascorbate but not by placebo. Norepinephrine did not increase ICP in either group. We demonstrate that sepsis elevates ICP to clinically relevant levels when measured directly via a lateral cerebral ventricular catheter, contributing to reduced CPP. Intravenous megadose sodium ascorbate ameliorates the intracranial hypertension, improves CPP, and reverses vasopressor hypo-responsiveness without increasing ICP in a clinically relevant ovine model of sepsis. These findings identify elevated ICP as a modifiable component of septic neuro-cardiovascular hemodynamic dysfunction and support sodium ascorbate as a potential therapy to mitigate both systemic and cerebral derangements in sepsis.Keywords: Sepsis, Sodium ascorbate, Intracranial pressure, Cerebral perfusion pressure, Mean arterial pressure, Vasopressor hypo-responsivenessSubject terms: Diseases, Medical research, Neurology, Neuroscience Introduction Sepsis remains a leading cause of death in critically ill patients worldwide and is characterized by a dysregulated host response to infection resulting in circulatory and multi-organ dysfunction1,2. Septic shock is defined by persistent hypotension requiring vasopressor therapy2, with norepinephrine recommended as the first-line agent to restore mean arterial pressure (MAP)3. However, septic shock is frequently complicated by vasopressor hypo-responsiveness, necessitating escalating catecholamine doses that are independently associated with adverse cardiovascular and neurological outcomes4,5. In parallel, sepsis-associated delirium and coma are common6,7 and cognitive dysfunction has emerged as a major determinant of morbidity and mortality yet remains poorly understood and largely unaddressed by current therapeutic interventions8–10. Cerebral tissue microcirculatory perfusion and oxygenation are compromised in ovine Gram-negative sepsis11–13. Experimental and clinical studies indicate that sepsis-associated neuroinflammation, impaired cerebral autoregulation, breakdown of the blood-brain barrier, and cerebral microvascular dysregulation may contribute to cerebral edema10,14. These pathophysiological mechanisms can increase intracranial pressure (ICP)12,15,16 and reduces cerebral perfusion pressure (CPP)17–19, thereby coupling systemic hypotension and intracranial hypertension into a detrimental pathophysiological interaction leading to cerebral tissue ischemia and hypoxia. Despite the risk of cerebral ischemia and hypoxia, current sepsis management focuses almost exclusively on systemic hemodynamics, with little consideration of intracranial pathophysiology or the cerebral consequences of vasopressor therapy. Intravenous megadose sodium ascorbate, a pH-balanced formulation of vitamin C, is emerging as a promising adjunctive therapy in sepsis10. Preclinical and early-phase clinical studies demonstrate that intravenous megadose of sodium ascorbate reduced vasopressor requirements and reversed multi-organ dysfunction in sepsis and septic shock20,21. We have also reported a reversal of cerebral tissue ischemia and hypoxia with sodium ascorbate in ovine sepsis11. However, whether these cerebrovascular effects are associated with reductions in ICP and restoration of vascular responsiveness to vasopressors remains unknown. In particular, the effects of megadose sodium ascorbate on ICP, CPP, and pressor responsiveness to norepinephrine during established sepsis have not been investigated. We have recently performed a dose-finding study that validates 3 g/kg of intravenous sodium ascorbate as the minimum therapeutic dose required to reverse cerebrovascular, cardiovascular and renovascular dysfunction in ovine Gram-negative sepsis13. Accordingly, this study investigated the effects of 3 g/kg intravenous sodium ascorbate on ICP, CPP, and pressor responsiveness to norepinephrine in non-anesthetized sheep with established live Gram-negative sepsis. The primary outcome was the reduction in ICP, while secondary outcomes included improvements in CPP and restoration of pressor responsiveness to norepinephrine. We hypothesized that sodium ascorbate would reduce ICP, improve CPP, and restore pressor responsiveness to norepinephrine in septic sheep. Methods Animal preparation Outbred adult female sheep (1.5–2.0 years of age) were acquired from a commercial supplier (Beverly Merino Stud; Redesdale, Victoria, Australia) and transported to the Florey Institute of Neuroscience and Mental Health and allowed a week of acclimatization to the laboratory environment before entering the study. Nineteen adult Merino ewes (31–50 kg) were housed in individual metabolic cages throughout the experiment with free access to food and water. All experiments were approved by the Animal Ethics Committee of The Florey Institute of Neuroscience and Mental Health (ID: 20–041-FINMH, Melbourne, Australia) under the guidelines of the National Health and Medical Research Council, and in accordance with the ARRIVE 2.0 guidelines22. All sheep underwent an aseptic surgical procedure under general anesthesia with intravenous sodium thiopental (15 mg/kg; Jurox, Rutherford, NSW, Australia) followed by intubation and maintenance with oxygen/air/isoflurane (1.5–2.0% v/v; Isoflo; Zoetis, Rhodes, NSW, Australia). A guide tube was inserted through a craniotomy with the tip above the lateral ventricle of the brain. To measure ICP, a 20-gauge sterile needle of appropriate length was inserted through the guide-tube into the lateral ventricle, as previously described23. The probe was connected to a Tygon cannula (Cole-Parmer, Boronia, Australia) filled with sterile 0.9% sodium chloride (Baxter, Old Toongabbie, NSW, Australia), connected to a pressure transducer (ITL Healthcare, Chelsea Heights, VIC, Australia), and then tied to the wool on the back of the head at the level of the lateral ventricles23. The left carotid artery was cannulated with a Tygon catheter and connected to a pressure transducer for arterial blood sampling and for measurement of arterial pressure and heart rate. The left jugular vein was cannulated with three polyethylene catheters (Protex, Smiths Medical International, Hythe, Kent, United Kingdom) for the intravenous administration of Escherichia coli (E. coli), fluid resuscitation with sodium lactate, norepinephrine and study drug interventions. To maintain the patency of carotid arterial and jugular venous catheters, they were continuously infused with heparinised saline (10 IU Heparin/mL at 3 mL/hr, Pfizer, West Ryde, NSW, Australia). Sheep were given 2.2 mg/kg intramuscular flunixin meglumine (Norbrook, Tullamarine, Victoria, Australia) for analgesia and 15 mg/kg intramuscular procaine penicillin (Troy Laboratories, Glendinning, NSW, Australia) intra-operatively, repeated at 24- and 48-h postoperatively. Animals were allowed a 3–4-day recovery period after surgery. Analogue signals of arterial pressure and ICP were recorded continuously at 100 Hz using a data-acquisition system (CED Micro-1401 interface and Spike2 software (Cambridge Electronic Design, Cambridge, UK) throughout the experiment. MAP and heart rate were calculated from the arterial pressure trace using Spike 2 software (Cambridge Electronic Design, UK), and CPP was calculated as MAP - ICP. Study design All experiments were performed in conscious, spontaneously breathing sheep without sedation, anesthesia, mechanical ventilation, or supplemental oxygen administration. Animals breathed ambient room air throughout the study (FiO₂ ≈ 0.21), ensuring that changes in arterial oxygenation reflected pathophysiological consequences of sepsis rather than alterations in supplemental oxygen concentrations. Following a 20-h baseline measurement period, sepsis was induced in conscious sheep with an intravenous infusion of live E. coli as a bolus followed by a continuous infusion for 31-h (bolus: 2.8 × 10⁹ colony-forming units (CFU) over 0.5-h, infusion: 1.26 × 10⁹ CFU/h) as previously described11,13,21,24. Animals surviving to 23-h of untreated sepsis were block-randomized using a computer-generated randomisation sequence to receive either sodium ascorbate or placebo. Treatment allocation was not blinded during administration due to logistical requirements of study drug preparation and delivery. However, investigators responsible for physiological data acquisition, signal processing, data analysis, and statistical evaluation were blinded to group allocation throughout the study. Following 23-h of sepsis, all animals received intravenous fluid resuscitation with sodium lactate (Hartmann’s solution 30 mL/kg over 0.5-h, Baxter Healthcare, Australia). Sheep then received either 3 g/kg of intravenous sodium ascorbate (0.5 g/kg as a 0.5-h bolus followed by 2.5 g/kg as an infusion, Biological Therapies, Victoria, Australia) or fluid-matched placebo treatment (5% w/v glucose; Baxter, Old Toongabbie, NSW) from 23.5–31-h of sepsis. Sodium ascorbate was prepared as a 1:1 dilution with 5% w/v d-glucose. Pressor responsiveness to norepinephrine was assessed using escalating infusion doses (0.2, 0.3, and 0.4 µg/kg/min; Hospira, Melbourne, VIC, Australia) at four defined experimental timepoints: premorbid baseline (Time 0), 23-h of untreated sepsis, 27.5-h of sepsis, and 31-h of sepsis. At each timepoint, animals underwent a structured, within-subject dose–response protocol. For the first norepinephrine dose at each timepoint, a 10-minute equilibration period was applied to allow catheter dead space clearance and attainment of hemodynamic steady state prior to data acquisition. MAP was then recorded over the subsequent 10-minute infusion period, with the pressor response quantified as the average MAP during the final 5 min of infusion. The second and third norepinephrine doses were administered sequentially within the same assessment period. A 5-minute equilibration period was allowed between dose increments to achieve hemodynamic stabilization prior to data collection. For each dose, the MAP response was calculated as the mean MAP during the final 5 min of infusion and referenced to the immediate pre-infusion hemodynamic state at that dose step. This sequential, within-timepoint design enabled construction of a dose–response relationship under stable measurement conditions while minimising hemodynamic disruption in a critically unstable septic model. Although doses were administered in ascending order without full washout periods, this approach is consistent with established large-animal sepsis methodologies, where rapid disease progression and cardiovascular lability preclude prolonged equilibration or washout between interventions. The highest norepinephrine dose was selected to achieve approximately a 20 mmHg increase in MAP, as previously described24. Arterial blood samples were collected at baseline (Time 0) and 23-h of sepsis, and then following treatment at 27.5-h, and 31-h of sepsis for measurement of blood gases, electrolytes and lactate (ABL System 625; Radiometer Medical, Copenhagen, Denmark). Neither analgesia nor antibiotics were given during the intervention period. Statistical analysis All physiological and biochemical variables were assessed for normality using the Shapiro–Wilk test and are presented as mean ± standard deviation (SD). Pre-morbid baseline and 23-h sepsis (pre-treatment) values for MAP, heart rate, ICP, CPP, arterial sodium and potassium levels were compared between treatment groups using unpaired Student’s t tests. To account for inter-animal variability at the point of randomization, treatment effects were assessed using delta change analyses relative to the 23-h untreated sepsis timepoint immediately prior to treatment randomization. Delta values were calculated between 23-h of sepsis and 27.5- and 31-h of sepsis, corresponding to 3.5- and 7-h of sodium ascorbate or placebo administration, respectively, to align with arterial blood sampling timepoints for biochemical measurements. Between-group comparisons of delta values were analyzed using unpaired Student’s t tests. Within-group changes in MAP, HR, ICP, CPP, and arterial sodium and potassium concentrations at 27- and 31-h of sepsis were compared with the 23-h untreated sepsis timepoint using paired Student’s t tests, with Bonferroni correction applied for multiple comparisons (k = 2). Absolute changes in MAP and ICP in response to escalating norepinephrine doses were assessed at baseline and compared with corresponding responses after sepsis induction and after 3.5 and 7-h of treatment, using repeated-measures ANOVA24. Factors in the analysis were time (PTime), dose (PDose), and their interaction (PInteraction). P values from within-subjects were conservatively adjusted using the Greenhouse-Geisser method25. Statistical analysis was performed using GraphPad Prism 10.6. (Graphpad Software, La Jolla, CA) for Windows. A P value ≤ 0.05 was considered statistically significant. Results The study design and methodology are illustrated in Fig. 1, which outlines the timeline of animal acclimatisation, surgical pre-instrumentation, sepsis induction, mortality during sepsis induction prior to randomization, and intervention phases. A total of 19 sheep were screened for health, with 6 excluded due to mortality during the 23-h of untreated sepsis (overall mortality 32%), prior to randomization. Following randomization at 23-h, the remaining animals were assigned to either the sodium ascorbate (n = 7) or placebo (n = 6) group. Measurements of MAP, HR, ICP, CPP, and arterial blood gases were conducted to assess the effects of the interventions. Notably, malfunctions of the ICP probes occurred in 2 animals in the sodium ascorbate group and 1 in the placebo group, leading to their exclusion from analyses. Schematic representation of the study design, depicting the timeline of sepsis induction, mortality events, randomization, and subsequent intervention phases. The schematic highlights the experimental timeline, surgical instrumentation, induction of sepsis and associated mortality and procedural issues related to malfunction of the lateral ventricular guide tube used for intracranial pressure (ICP) monitoring, which affected cerebral perfusion pressure (CPP) calculations (mean arterial pressure (MAP)/ICP) in a subset of animals. Additionally, measurements of MAP, heart rate (HR), and arterial blood gases (ABG) were assessed. MAP and HR responses to sepsis and sodium ascorbate (n = 7) or placebo (n = 6) treatment By 23-h post E. coli infusion, prior to randomization, both groups of septic sheep exhibited significant reductions in MAP, from 88 ± 7 mmHg to 75 ± 5 mmHg in the sodium ascorbate group, and from 85 ± 6 mmHg to 69 ± 11 mmHg in the placebo group (P = 0.002 and P = 0.018, respectively; Fig. 2a). HR increased significantly in both groups, from 72 ± 9 beats/min to 154 ± 15 beats/min in the sodium ascorbate group and from 74 ± 11 to 147 ± 19 beats/min in the placebo treatment group (both P < 0.001) (Fig. 2b). The reductions in MAP (P = 0.2) and the elevations in HR (P = 0.5) did not differ significantly between the two groups prior to randomization (Fig. 2a & d). Changes in mean arterial pressure and heart rate during live Gram-negative sepsis followed by treatment with sodium ascorbate or placebo. Data are presented as mean ± standard deviation. Panel (a) shows mean arterial pressure (MAP) at baseline and 23 h of sepsis, with significant reductions indicated by **P < 0.01, prior to being randomised to sodium ascorbate (n = 7). ##P < 0.01 depicts a significant reduction in MAP, in sheep prior to be randomised to placebo treatment (n = 6). P values were derived from a students paired t test. Delta values for MAP at 28 h (b) and 31 h (c) post-treatment reveal a significant increase in MAP in the sodium ascorbate group compared to placebo (*P < 0.05). P values were derived from a students unpaired t test. Panel (d) depicts heart rate (HR) at baseline and 23 h of sepsis, with significant elevations shown by **P < 0.01, prior to being randomised to sodium ascorbate (n = 7). ###P < 0.001 depicts increases in HR, in sheep prior to be randomised to placebo treatment (n = 6). P values were derived from a students paired t test. Delta values for HR at 28 h (e) and 31 h (f) post-treatment indicate a significant decrease in HR in the sodium ascorbate group compared to placebo (*P < 0.05 and ***P < 0.001, respectively). P values were derived from a students unpaired t test. Following treatment, sodium ascorbate significantly improved MAP at both 28-h (87 ± 11 mmHg vs. 69 ± 10 mmHg; P = 0.021) and 31-h (85 ± 7 mmHg vs. 66 ± 7 mmHg; P = 0.020) compared with placebo (Fig. 2b & c). Additionally, sodium ascorbate led to a reduction in HR from 23 to 28 h (115 ± 9 beats/min vs. 140 ± 22 beats/min; P = 0.029) and at 31 h of sepsis (106 ± 15 beats/min vs. 143 ± 15 beats/min; P < 0.001) compared with placebo treatment (Fig. 2e & f). ICP and CPP responses to sepsis and sodium ascorbate (n = 5) or placebo (n = 5) treatment By 23-h post E. coli infusion, prior to randomization, both groups of septic sheep exhibited significant elevations in ICP, with values increasing from 7.9 ± 4.8 to 17.0 ± 3.4 mmHg in the sodium ascorbate group, and from 12.3 ± 2.3 to 20.6 ± 3.7 mmHg in the placebo group (P = 0.01 and P = 0.008, respectively; Fig. 3a). CPP decreased in both groups, falling from 78 ± 10 to 61 ± 11 mmHg in the sodium ascorbate group (P = 0.008) and from 72 ± 3 to 47 ± 15 mmHg in the placebo treatment group (P = 0.02) (Fig. 3d). The elevations in ICP (P = 0.2) and the reductions in CPP (P = 0.2) did not differ significantly between the two groups prior to randomization (Fig. 3a & d). Changes in intracranial pressure and cerebral perfusion pressure during live Gram-negative sepsis followed by treatment with sodium ascorbate or placebo. Data are presented as mean ± standard deviation. Panel (a) shows intracranial pressure (ICP) at baseline and 23 h of sepsis, with significant increases indicated by **P < 0.01, prior to being randomised to sodium ascorbate (n = 5). ##P < 0.01 depicts a significant increase in ICP, in sheep prior to be randomised to placebo treatment (n = 5). P values were derived from a students paired t test. Delta values for ICP at 28 h (b) and 31 h (c) post-treatment reveal a significant reduction in ICP in the sodium ascorbate group compared to placebo at 31 h (*P < 0.05). P values were derived from a students unpaired t test. Panel (d) depicts cerebral perfusion pressure (CPP) at baseline and 23 h of sepsis, with significant reductions shown by **P < 0.01, prior to being randomised to sodium ascorbate (n = 5). ###P < 0.001 depicts decreases in CPP, in sheep prior to be randomised to placebo treatment (n = 5). P values were derived from a students paired t test. Delta values for CPP at 28 h (e) and 31 h (f) post-treatment indicate a significant improvement in CPP in the sodium ascorbate group compared with placebo by 31 h hours (*P < 0.05). P values were derived from a students unpaired t test. Following treatment, sodium ascorbate reduced ICP at 31-h of sepsis (to 9.3 ± 1.9 mmHg). compared with placebo (to 19.3 ± 5.0 mmHg; P = 0.021) (Fig. 3c). Concurrently, sodium ascorbate led to an improvement in CPP from 23 to 31-h (75 ± 5 vs. 49 ± 12 mmHg; P < 0.001) compared with placebo (Fig. 3f). Arterial biochemical responses to sepsis and sodium ascorbate (n = 7) or placebo (n = 6) treatment By 23-h post E. coli infusion, prior to randomization, both groups of septic sheep exhibited reductions in arterial PO2, with values falling from 92 ± 6 to 84 ± 7 mmHg in the sodium ascorbate group, and from 95 ± 12 to 72 ± 16 mmHg in the placebo group (P = 0.029 and P = 0.018, respectively; Table 1). Arterial blood lactate concentrations increased in both groups, from 0.6 ± 0.2 to 1.6 ± 0.2 mmol/L in the sodium ascorbate group (P = 0.01) and from 0.6 ± 0.2 to 1.9 ± 0.8 mmol/L in the placebo treatment group (P = 0.045) (Table 1). The reductions in arterial PO2 (P = 0.1) and the increases in lactate concentration (P = 0.7) did not differ significantly between the two groups prior to randomization (Table 1). There were no significant differences in other arterial biochemical variables between groups. Table 1Arterial biochemistry in septic sheep treated with sodium ascorbate compared with placebo.VariablesGroupExperimental Time PointsTime of SepsisTime 0 (pre-morbid)23-h Sepsis28-h Sepsis31-h SepsispH Na Asc 7.53 ± 0.017.54 ± 0.017.50 ± 0.017.51 ± 0.01 Placebo 7.53 ± 0.017.53 ± 0.017.52 ± 0.017.52 ± 0.01PCO2 (mmHg) Na Asc 29.7 ± 1.5729.99 ± 1.7335.47 ± 1.7938.18 ± 1.23 Placebo 33.85 ± 1.0233.25 ± 1.0335.32 ± 0.8434.86 ± 0.62PO2 (mmHg) Na Asc 92.36 ± 2.3283.53 ± 2.80*83.53 ± 2.1984.25 ± 2.86 Placebo 94.5 ± 4.7472.15 ± 6.53#78.88 ± 7.6281.34 ± 9.34Haemoglobin (mmol/L) Na Asc 9.56 ± 0.2910.44 ± 0.409.76 ± 0.589.48 ± 0.52 Placebo 9.15 ± 0.4410.12 ± 0.568.82 ± 0.518.82 ± 0.41Oxygen Saturation (%) Na Asc 95.96 ± 0.3195.16 ± 0.5894.71 ± 0.4594.52 ± 0.81 Placebo 95.75 ± 0.8590.18 ± 3.7691.38 ± 3.8792.2 ± 4.25Base Excess (mmol/L) Na Asc 2.17 ± 0.414.27 ± 0.105.59 ± 0.716.93 ± 0.71 Placebo 5.27 ± 415.57 ± 1.555.71 ± 0.055.6 ± 0.93Lactate (mmol/L) Na Asc 0.53 ± 0.061.56 ± 0.25*1.36 ± 0.141.24 ± 0.17 Placebo 0.60 ± 0.071.75 ± 0.52#1.27 ± 0.291.62 ± 0.41 Values are presented as mean ± standard deviation. *P < 0.05 indicate significant differences between the pre-morbid baseline (Time 0) and 23-h of untreated sepsis in sheep prior to be randomised to the sodium ascorbate (Na Asc; N = 7). #P < 0.05 indicate significant differences between the pre-morbid baseline (Time 0) and 23-h of untreated sepsis in sheep prior to be randomised to the placebo (N = 6). P values were derived from a Student’s paired t test. Arterial biochemistry in septic sheep treated with sodium ascorbate compared with placebo. Values are presented as mean ± standard deviation. *P < 0.05 indicate significant differences between the pre-morbid baseline (Time 0) and 23-h of untreated sepsis in sheep prior to be randomised to the sodium ascorbate (Na Asc; N = 7). #P < 0.05 indicate significant differences between the pre-morbid baseline (Time 0) and 23-h of untreated sepsis in sheep prior to be randomised to the placebo (N = 6). P values were derived from a Student’s paired t test. Following treatment, sodium ascorbate significantly increased arterial blood sodium levels at both 28-h (143 ± 5 vs. 137 ± 4 mmol/L; P < 0.001) and 31-h of sepsis (150 ± 6 vs. 136 ± 3 mmol/L; P = 0.001) compared with placebo (Fig. 4b & c). Additionally, sodium ascorbate reduced arterial blood potassium from 23 to 28-h (3.6 ± 0.2 vs. 3.3 ± 0.4 mmol/L; P < 0.054), with a statistically significant reduction achieved at 31-h of sepsis (2.5 ± 0.2 vs. 3.6 ± 0.5 mmol/L; P = 0.015) when compared with placebo treatment (Fig. 4e & f). Changes in arterial blood sodium and potassium level during live Gram-negative sepsis followed by treatment with sodium ascorbate or placebo. Data are presented as mean ± standard deviation. Panel (a) shows arterial blood sodium (Na+) at baseline and 23 h of sepsis, prior to being randomised to sodium ascorbate; n = 5) or placebo. Delta values for arterial blood Na+ at 28 h (b) and 31 h (c) post-treatment reveal a significant increase in arterial blood Na+ in the sodium ascorbate group compared to placebo at 28-h (***P < 0.0001) and at 31 h (***P < 0.0001). P values were derived from a students unpaired t test. Panel (d) depicts arterial blood potassium (K+) at baseline and 23 h of sepsis, prior to being randomised to sodium ascorbate (n = 7) or placebo treatment (n = 6). Delta values for arterial blood K+ at 28 h (b) and 31 h (c) post-treatment reveal a significant reduction in blood K+ in the sodium ascorbate group compared to placebo at 31-hours (*P < 0.05). P values were derived from a students unpaired t test. Effects of sodium ascorbate on physiological responses to norepinephrine challenges Administration of norepinephrine evoked dose-dependent increases in MAP (Fig. 5). By 23-h of sepsis, the mean arterial pressor responses to norepinephrine were significantly less than the responses in the premorbid baseline period in both groups (Fig. 5a & b). For example, the highest dose of norepinephrine (0.4 µg/kg/min) increased MAP by 20 ± 5 mmHg at baseline but only by 12 ± 7 mmHg at 23-h of sepsis in placebo-treated animals (Fig. 5a). In placebo-treated sheep, the pressor responses to the highest dose of norepinephrine remained diminished at 27.5-h (14 ± 6 mmHg) and 31-h (13 ± 4 mmHg) of sepsis (Fig. 5c & e). In contrast, in sodium ascorbate-treated sheep, the pressor responses to norepinephrine at 27.5-h (15 ± 5 mmHg) (Fig. 5d) and 31-h (19 ± 10 mmHg) (Fig. 5f) of sepsis were restored towards baseline levels (20 ± 3 mmHg). Administration of norepinephrine had no effect on ICP at any doses in either treatment group (all PTime and PDose > 0.05; Fig. 6). Mean arterial pressure pressor responsiveness to norepinephrine during sodium ascorbate or placebo treatment in ovine Gram-negative sepsis. Data are expressed as mean ± standard deviation. Responses in pre-morbid sheep in the placebo group (open red boxes, n = 6) and sodium ascorbate group (open orange circles; n = 7)) and at 23-h of sepsis in the placebo group (closed red boxes, a) and sodium ascorbate group (closed orange circles, b). Pressor responsiveness after 3.5-h infusion of placebo (closed boxes, c) or sodium ascorbate (closed circles, d) and after 7-h of placebo (closed boxes, e) or sodium ascorbate (closed circles, f). Dose-response curves were compared using two-way repeated-measures analysis of variance, and P values represent significant changes in mean arterial pressure pressor responsiveness to norepinephrine with time of sepsis. Intracranial pressure responses to norepinephrine during sodium ascorbate and placebo treatment in ovine Gram-negative sepsis. Data are expressed as mean ± standard deviation. Responses in pre-morbid sheep in the placebo group (open red boxes; n = 5) and sodium ascorbate group (open orange circles; n = 5) and at 23-h of sepsis in the placebo group (closed red boxes, a) and sodium ascorbate group (closed orange circles, b). Intracranial pressure changes after 3.5-h infusion of placebo (closed boxes, c) or sodium ascorbate (closed circles, d) and after 7-h of placebo (closed boxes, e) or sodium ascorbate (closed circles, f). Dose-response curves were compared using two-way repeated-measures analysis of variance, and P values represent significant changes in intracranial pressure to norepinephrine with time of sepsis. Discussion This study yielded three clinically relevant novel findings. First, that sepsis causes a physiologically significant increase in ICP (measured directly from a lateral cerebral ventricle) and reduced CPP. Second, intravenous megadose sodium ascorbate reduced the sepsis-induced elevation in ICP and restored CPP. Norepinephrine challenges had no effect on ICP, suggesting that the reduction in ICP observed with sodium ascorbate is independent of direct catecholamine mediated cardiovascular effects. Third, megadose sodium ascorbate fully restored the pressor responsiveness to norepinephrine infused within a clinically relevant dose range during sepsis. Vitamin C, as l-ascorbic acid (pH ~ 5.4)26, has been evaluated in two large randomized controlled trials using lower dosing regimens (200 mg/kg/day, intravenously) and found to be associated with higher rates of 28-day mortality or persistent organ dysfunction in patients with sepsis and COVID-1927,28. In contrast, intravenous sodium ascorbate (pH ~ 7.2) is emerging as a highly effective and potentially safer formulation of ascorbate for the treatment of sepsis26. In ovine Gram-negative sepsis, megadose sodium ascorbate administered at 150 g (3.75 g in a 40 kg sheep) improved vasopressor responsiveness, reversed multi-organ function and improved the clinical state from malaise, lethargy, and somnolence to a normal, alert, mobile state11,21. Similar physiological signals of benefit were observed in a Phase Ia trial in 30 patients with septic shock using 60 g of intravenous sodium ascorbate (0.8 g/kg in an 80 kg human)20. In this study, we used 3 g/kg of sodium ascorbate as we have shown this is the minimum therapeutic dose required to achieve optimal plasma ascorbate concentrations (~ 10 mmol/L) and reverse sepsis-induced vasoplegia and multi-organ dysfunction in sheep13. Sepsis-associated cerebral ischemia, edema, impaired cerebral autoregulation, and elevations in ICP have been reported in patients with sepsis17–19,29, using indirect and non-invasive techniques. Ovine Gram-negative sepsis is characterised by cerebral tissue ischemia and hypoxia12, which can be effectively reversed with intravenous megadose sodium ascorbate11,13. Importantly, pre-morbid ICP values in our sheep model (approximately 8–12 mmHg) were physiologically comparable to normal ICP values reported in healthy adult humans (typically 7–11 mmHg in the supine position)30,31. This similarity supports the translational validity of our ovine model and provides a rationale for interpreting sepsis-induced ICP elevations relative to clinically recognised human thresholds. Current clinical guidelines define intracranial hypertension as an ICP greater than 22 mmHg in patents with traumatic brain injury32, although the biological harm of raised ICP is likely determined by both the magnitude and duration of exposure. It is conceivable that the sodium ascorbate-induced reduction in ICP and the concomitant restoration of CPP contributes to improvements in perfusion and oxygenation to the brain parenchyma11,13. Our findings highlight that restoration of MAP alone with fluids and vasopressors does not reduce the sepsis-induced increase in ICP or decrease in CPP. Prior studies of fluid and vasopressor therapy in sepsis have largely focused on systemic endpoints and have not examined the interactions between increases in MAP and ICP and decreases in CPP3,33,34. Our study demonstrates that increases in ICP during sepsis, measured directly from a lateral cerebral ventricle, contribute to the reduced CPP during sepsis. We further demonstrate that megadose sodium ascorbate lowers ICP, which was associated with concomitant improvement in CPP. Norepinephrine had no effect on ICP, which suggests that sodium ascorbate improves intracranial hemodynamics through mechanisms distinct from direct vasoconstrictor effects. Sepsis-associated cerebral dysfunction is complex and multifactorial, involving disturbances in cerebral autoregulation, endothelial and microcirculatory dysfunction, neuroinflammation, metabolic dysregulation, and altered intracranial compliance8–10. Accordingly, the reductions in ICP observed following sodium ascorbate administration over a 7-h infusion period should not be interpreted as definitive evidence of reversal of sepsis-associated encephalopathy, as neurological function, delirium, neuropathology, cerebral autoregulation, and long-term neurological outcomes were not assessed. Rather, these findings demonstrate that in sepsis sodium ascorbate improves cerebral hemodynamics, which may contribute to its ability to improve cerebral perfusion and oxygenation and reverse sickness behaviour11,21. Recent clinical trials in sepsis have highlighted substantial heterogeneity in cerebral autoregulatory function and suggest that autoregulation-guided hemodynamic management may influence neurological outcomes, including delirium17–19,29. Future studies incorporating direct assessments of cerebral autoregulation, intracranial compliance, neurovascular coupling, and neurological outcomes will be required to determine the translational significance of these findings. The increase in arterial blood sodium following administration of sodium ascorbate is likely to have contributed to the reduction in ICP through osmotic mechanisms, analogous to the effects of established osmotherapy (i.e. hypertonic saline) used in neurocritical care35–37. In addition to its osmotic actions, sodium ascorbate may also exert neuro-protective effects through restoration of endothelial function, attenuation of inflammation and improvement in microcirculatory perfusion10. The ability of sodium ascorbate to reverse sepsis-induced frontal cortical tissue ischemia, hypoxia and hyperthermia11 do not depend on the sodium load as we demonstrated that these effects were not reproduced by infusion of equimolar sodium bicarbonate11. In sheep with established septic acute kidney injury, we have also found sodium ascorbate to mitigate renal tissue inflammation and improve nitric oxide bioavailability13. Future studies are warranted to delineate the osmotic effects of sodium from the biological effects of ascorbate on ICP and CPP in ovine sepsis. Using our recently validated optimal dose of sodium ascorbate (3 g/kg),13 the present study is an important advance demonstrating complete restoration of pressor responsiveness to norepinephrine in ovine sepsis. Sepsis induced vasodilation and vasopressor hypo-responsiveness have been proposed to result from excessive inflammation and dysregulated nitric oxide signalling10,38,39. Ascorbate can attenuate activation of nicotinamide adenine dinucleotide phosphate oxidase, suppress inflammation and inducible nitric oxide synthase expression40, which can limit excessive nitric oxide production and pathological vasodilation in sepsis. In accord, we and others have demonstrated that sodium ascorbate reduces nuclear factor kappa B activation, mitigating cerebral and renal tissue inflammation in preclinical sepsis11,13,41,42. Ascorbate can also improve endothelial function by stabilizing tetrahydrobiopterin and positively modulating the phosphorylation of endothelial nitric oxide synthase43, thereby preventing uncoupling of eNOS in ovine sepsis13. Ascorbate is also an essential cofactor for dopamine β-hydroxylase, supporting endogenous norepinephrine biosynthesis44,45, which can enhance adrenergic tone during sepsis. These synergistic beneficial effects of sodium ascorbate on inflammatory signalling, endothelial function, and catecholamine biology are likely to contribute to the restoration of norepinephrine pressor responsiveness observed in this study. It will also be interesting to determine if sodium ascorbate restores vasopressor responsiveness to angiotensin II in sepsis, which we demonstrated with clonidine and dexmedetomidine24,46. This study has several strengths and limitations. To our knowledge, it is the first study to simultaneously characterize the effects of megadose sodium ascorbate on MAP, ICP, CPP and pressor responsiveness to a clinically relevant dose range of norepinephrine in a large mammalian model of live Gram-negative sepsis. Importantly, ICP was continuously measured directly from a cannula in the lateral cerebral ventricle in the brain and the preclinical study was performed on non-anesthetized septic sheep without the confounding effects of general anesthesia. The study also has several limitations. First, the number of animals available for ICP and CPP analyses was relatively small, and a few ICP recordings were unavailable due issues with lateral ventricular catheter patency. Consequently, these analyses should be interpreted with caution given the limited statistical power and the influence of the inter-animal biological variability inherent to small sample sizes. Second, as randomization occurred after establishment of sepsis, there is a potential for survivor selection bias in the study cohort that may have excluded animals with more severe early disease trajectories, potentially influencing generalisability of the findings. Third, the study did not include an equimolar sodium group, so it is not possible to distinguish whether the observed effects are entirely attributable to the sodium load or also additional ascorbate mediated effects. Fourth, measures of cerebral edema, blood–brain barrier integrity, cerebral blood flow, and molecular markers of neuroinflammation and oxidative/nitrosative stress markers were not evaluated. Finally, formal neurological, behavioural, and cognitive outcome assessments were not undertaken. In addition, patients with severe sepsis commonly receive antibiotics, sedative agents, and opioid-based analgesia, which may modify systemic hemodynamics, inflammatory signalling, and microcirculatory function, and therefore may influence brain pathophysiology in ways not captured in the sheep model used in this study. Collectively, these limitations indicate that observed physiological changes in ICP and CPP should be interpreted as providing mechanistic and physiological insights requiring validation in larger, controlled studies to establish the translational significance of these observations. With this in mind, a multicentre randomized controlled Phase Ib trial is currently evaluating the safety and feasibility of intravenous megadose sodium ascorbate (180 g) in patients with septic shock, with the trial powered to detect a reduction in norepinephrine requirements with reductions in sequential organ failure assessment scores that includes neurological functional assessments as a secondary outcome [ANZCTRB12625001401448]. Conclusion In a non-anaesthetized ovine model of established Gram-negative sepsis, intravenous sodium ascorbate (3 g/kg) was associated with reductions in the elevated ICP, improvements in CPP, and restoration of MAP pressor responsiveness to norepinephrine compared with placebo treatment. The mechanisms underlying these effects, their generalisability across different sepsis phenotypes, and their impact on neurological functional outcomes remain to be determined in future preclinical mechanistic and ongoing multi-centre clinical trials [ANZCTRB12625001401448]. Acknowledgements The authors acknowledge the contribution of the late Professor Rinaldo Bellomo to preclinical and clinical investigations on sodium ascorbate for sepsis. We thank Tom Vale, Quan Nguyen and Miranda Marcon for their excellent technical assistance. Author contributions MSB, RVP, AHJ, DJRL, CPCO and YRL were involved in data acquisition, analysis and interpretation prior to submission. MSB, RVP, AHJ and CPCO were involved in the surgical instrumentation and experiments. CNM, SA, MPP, CPCO and YRL were involved in conception, study design, interpretation of data and substantial revision prior to submission. MSB and YRL wrote the manuscript. All authors read and approved the final manuscript prior to submission. Funding This study was supported by a Medical Research Future Fund grant [MRF2022388] from the National Health and Medical Research Council of Australia. MSB was supported by an Australian Government Research Training Program (RTP) Scholarship. YRL was supported by an Emerging Leader Investigator Grant from the National Health and Medical Research Council of Australia (GNT2025266). MPP was supported by an Emerging Leader Investigator Grant from the National Health and Medical Research Council of Australia (GNT2017388). SA was supported by a Leadership Investigator Grant from the National Health and Medical Research Council of Australia (GNT2008359). Data availability All data generated or analysed during this study are included in this article. Datasets are available from the corresponding author upon a request. Declarations Competing interests YRL and CNM are co-inventors on the use of sodium ascorbate for sepsis. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. These authors contributed equally: Connie P. C. Ow and Yugeesh R. Lankadeva. References Associated Data Data Availability Statement All data generated or analysed during this study are included in this article. Datasets are available from the corresponding author upon a request.