Neuroprotection via optimizing cerebral blood flow after cardiac arrest (neuro-intact) study – A phase II proof-of-concept study protocol Authors: Shir Lynn Lim, Kai Lee Woo, Mihir Gandhi, Erta Beqiri, Peter Smielewski, Marcus E.H. Ong, Vijay Kumar Sharma Published: 2026 ABSTRACT IntroductionOptimizing cerebral blood flow after return of spontaneous circulation (ROSC) is key to neuroprotection in out-of-hospital cardiac arrest (OHCA) patients. Randomized trials comparing fixed mean arterial pressure (MAP) targets have been uniformly neutral, as between- and within-patient heterogeneity in cerebral autoregulation renders uniform targets inadequate. An individualized blood pressure strategy targeting patient-specific cerebral perfusion may reduce secondary brain injury.MethodsNEUROprotectIoN via optimizing cerebral blood flow afTer cArdiaC arresT (NEURO-INTACT) is a single-center, single-arm, phase II proof-of-concept trial enrolling 49 comatose OHCA survivors within 12 h post-ROSC. With constant ventilatory settings and continuous electrographic, pulse oximeter and end-tidal carbon dioxide monitoring, MAP is actively manipulated via vasopressor titration to generate slow MAP fluctuations. Changes in cerebral oxygenation (near-infrared spectroscopy) and blood flow velocities (transcranial Doppler) are continuously correlated with MAP by ICM+ software to characterize each patient’s cerebral autoregulation and derive an individualized optimal MAP (MAPopt). MAP is then actively titrated to MAPopt, with this assess-and-titrate cycle repeated at three timepoints across 72 h post-ROSC. Primary endpoint is the mean change in neuron-specific enolase from baseline to 72 h. Secondary endpoints include peak cardiac and renal injury biomarkers within 72 h, neurological outcome at discharge, and health-related quality of life at six months, compared against historical controls.ConclusionThis study will explore the effect of individualized, autoregulation-guided hemodynamic management on a validated surrogate marker of brain injury, demonstrate protocol feasibility and safety, and generate effect-size estimates to inform a definitive multicenter trial. Trial registration: ClinicalTrials.gov ID NCT05679739. Introduction Brain injury accounts for approximately two-thirds of mortality in resuscitated out-of-hospital cardiac arrest (OHCA) patients.1 Central to neuroprotection is optimizing cerebral blood flow, which is determined by mean arterial pressure (MAP) and local cerebrovascular resistance. The optimal blood pressure in resuscitated OHCA patients is unknown; current standard-of-care uses a MAP target of ≥65 mmHg, a practice adopted from septic shock guidelines.2 This target has been associated with decreased cerebral oxygenation during the critical first 6–12 h post-ROSC.3, 4 Additionally, randomized controlled trials comparing fixed MAP targets after OHCA have been uniformly neutral.5, 6, 7 The difficulty in defining optimal MAP targets stems from marked pathophysiological heterogeneity. Cerebral autoregulation may be totally absent in some patients and significantly impaired in others, with the preserved range varying substantially across individuals.8, 9 Furthermore, cerebral blood flow demonstrates a characteristic temporal pattern of early hyperemia, prolonged hypoperfusion, and later partial normalization, superimposed on regional microcirculatory heterogeneity.10 Consequently, the cerebrovascular response to any given MAP differs between patients and across post-resuscitation phases, rendering any fixed target liable to produce hypoperfusion at some timepoints and hyperperfusion at others. An individualized blood pressure strategy tailored to real-time cerebral blood flow is a biologically plausible alternative, but must be instituted early post-ROSC, while secondary injury is still modifiable, and dynamically adjusted as autoregulatory status evolves. Transcranial Doppler ultrasound (TCD) and near-infrared spectroscopy (NIRS) provide complementary, non-invasive, bedside measurements of cerebral blood flow velocity and regional cerebral oxygen saturation (rSO2), enabling autoregulation assessment by correlating slow signal changes with MAP fluctuations,11 to derive the autoregulatory limits and the individualized optimal MAP (MAPopt).12, 13 Prior observational studies have been constrained by single-modality NIRS assessment and non-interventional designs, limiting both the precision of autoregulatory characterization and ability to test causal hypotheses.14, 15 Our preliminary case-series of 16 patients using concurrent NIRS-TCD monitoring post-ROSC confirmed protocol feasibility, identified four distinct cerebral autoregulatory patterns, and demonstrated that preserved autoregulation was associated with lower neuro-injury biomarkers and better neurological recovery,16 underscoring both the heterogeneity that necessitates individualization and the feasibility of the multimodal approach that underpins this study. The continuous cerebral autoregulation research community recommends that proof-of-concept studies establishing MAPopt feasibility, signal yield, and biomarker response should precede large outcome trials, given heterogeneity in monitoring yield and the absence of standardized intervention protocols.17 Accordingly, this proof-of-concept study uses neuron-specific enolase (NSE), a validated surrogate biomarker of neuronal injury post-OHCA, as the primary endpoint.18, 19 The primary objective is to determine the effect of individualized, autoregulation-guided MAP management on the change in NSE from baseline to 72 h (Fig. 1). Secondary objectives include estimating effects on cardiac and renal injury biomarkers, neurological outcome and health-related quality of life against historical controls, recovery trajectory through exploratory functional, cognitive, and neuropsychological assessments at one year, and protocol feasibility (spanning enrolment, monitoring signal yield, and protocol adherence) and safety.Fig. 1Conceptual figure.Conceptual figure comparing standard-of-care versus proposed strategy. The varying impairment in cerebral autoregulation and phasic variations in cerebral blood flow after return of spontaneous circulation suggest a potential neuroprotective benefit of actively tailoring systemic blood pressure targets to maintain optimal cerebral blood flow. This is guided by bedside transcranial Doppler ultrasound and near-infrared spectroscopy, by correlating changes in surrogates of cerebral blood flow with changes in mean arterial pressure. A tailored blood pressure strategy may also benefit extra-cerebral organ perfusion (heart and kidneys), the autoregulatory mechanisms of which are similarly impaired following successful resuscitation.Abbreviations: OHCA, out-of-hospital cardiac arrest; BP, blood pressure; TCD, transcranial Doppler ultrasound; NIRS, near-infrared spectroscopy; NSE, neuron-specific enolase; hsTNT, highly-sensitive Troponin T; NT-proBNP, N-terminal pro-brain natriuretic peptide; penKID, proenkephalin; bio-ADM, bioactive adrenomedullin; TIMP2, tissue Inhibitor of metalloproteinase 2; IGFBP7, insulin-like growth factor-binding protein 7. Conceptual figure. Conceptual figure comparing standard-of-care versus proposed strategy. The varying impairment in cerebral autoregulation and phasic variations in cerebral blood flow after return of spontaneous circulation suggest a potential neuroprotective benefit of actively tailoring systemic blood pressure targets to maintain optimal cerebral blood flow. This is guided by bedside transcranial Doppler ultrasound and near-infrared spectroscopy, by correlating changes in surrogates of cerebral blood flow with changes in mean arterial pressure. A tailored blood pressure strategy may also benefit extra-cerebral organ perfusion (heart and kidneys), the autoregulatory mechanisms of which are similarly impaired following successful resuscitation. Abbreviations: OHCA, out-of-hospital cardiac arrest; BP, blood pressure; TCD, transcranial Doppler ultrasound; NIRS, near-infrared spectroscopy; NSE, neuron-specific enolase; hsTNT, highly-sensitive Troponin T; NT-proBNP, N-terminal pro-brain natriuretic peptide; penKID, proenkephalin; bio-ADM, bioactive adrenomedullin; TIMP2, tissue Inhibitor of metalloproteinase 2; IGFBP7, insulin-like growth factor-binding protein 7. Methods This protocol is reported in accordance with the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) 2025 guidelines.20 The study has been approved by the National Healthcare Group Domain Specific Review Board (NHG DSRB Ref: 2020/01455). Design and setting NEURO-INTACT is a single-center, single-arm, phase II proof-of-concept trial conducted in the Coronary Care Unit at National University Hospital, Singapore. The active intervention spans 72 h post-ROSC, corresponding to the early and intermediate phases of post-cardiac arrest syndrome. All survivors are prospectively followed for one year from the cardiac arrest event. This design was chosen over a parallel-group randomized design because high early mortality, narrow enrolment windows, and the need for simultaneous neurosonographer and neuromonitoring equipment availability preclude feasible randomization; proof-of-concept objectives are achievable within a single-arm framework; and within-patient paired comparisons provide internally valid primary endpoint estimation. Historical control Historical controls are derived from ENFORCE (Early Neuroprognostic biomarkers For resuscitated Out-of-hospital Cardiac arREst), a prospective observational cohort at the same institution with consistent inclusion/exclusion criteria and NSE assayed at 0, 24, and 72 h post-ROSC.21, 22 Neurological outcome by Cerebral Performance Category (CPC) is assessed at hospital discharge and six months, and health-related quality of life is measured using EQ-5D-5L at six months. ENFORCE historical control data are used for comparison with the secondary endpoints of favorable neurological outcome (CPC 1–2) and EQ-5D-5L at six months, as well as for contextual comparison of the NSE trajectory from 0 to 72 h post-ROSC. Eligibility criteria The target enrolment is 49 patients. Inclusion criteria are age 21–79 years and comatose after sustained ROSC. Sustained ROSC requires ≥20 min of sustained circulation; comatose is defined as Glasgow Coma Scale ≤8 with inability to obey verbal commands. Patients must be enrolled and the first autoregulation assessment initiated within 12 h of sustained ROSC. Exclusion criteria are total downtime (from OHCA to sustained ROSC) ≥60 min prior to admission, early limitations of care or withdrawal of life-sustaining treatment within 24 h, inability to complete study activities through Day 4 (72 h post-ROSC), inability to obtain TCD signals due to poor acoustic windows, and pregnancy. General management All patients receive standard post-cardiac arrest care per institutional protocol, including targeted temperature management at 34°C for 24 h with controlled rewarming at 0.3–0.5°C per hour, emergent coronary angiography and percutaneous coronary intervention as clinically indicated, and organ support as required.23, 24 Continuous blood pressure monitoring via radial or femoral arterial catheter is standard. The use of intra-aortic balloon pump and choice of vasopressors and/or inotropes is at the discretion of the managing cardiologist. Intervention (Fig. 2) Cerebral autoregulation is assessed at three timepoints: on admission to the Coronary Care Unit (within 12 h of ROSC), and at 24 h and 48 h post-ROSC. The assessment frequency in the first 72 h reflects the evolving phase of brain injury and hemodynamic adjustments. Additional assessments are performed if significant clinical changes occur, including new neurological signs, seizures, or arrhythmias. Each assessment spans 30–45 min.Fig. 2Schematic representation of the NEURO-INTACT study design. Comatose OHCA survivors undergo serial cerebral autoregulation assessment using combined TCD and NIRS at three timepoints over 48 h, with MAP titrated to the individualized MAPopt for 72 h. Biomarker sampling and clinical outcome assessments are performed at specified intervals through one year.*Severity of illness on admission will be assessed using the Pittsburgh Cardiac Arrest Category and complemented by serum levels of lactate and neuron-specific enolase at baseline. These will be accounted for in the post-hoc adjustment.Abbreviations: CCU, Coronary Care Unit; OHCA, out-of-hospital cardiac arrest; MAP, mean arterial pressure; H, hour; CPC, Cerebral Performance Category; HRQoL, health-related quality of life; IV, intravenous; CVP, central venous pressure; TCD, transcranial Doppler ultrasound; NIRS, near-infrared spectroscopy; BP, blood pressure; MAP, mean arterial pressure; PaCO2, arterial partial pressure of carbon dioxide; FiO2, fraction of inspired oxygen; COx, cerebral oxygenation index; Mxa, mean velocity index. Schematic representation of the NEURO-INTACT study design. Comatose OHCA survivors undergo serial cerebral autoregulation assessment using combined TCD and NIRS at three timepoints over 48 h, with MAP titrated to the individualized MAPopt for 72 h. Biomarker sampling and clinical outcome assessments are performed at specified intervals through one year. *Severity of illness on admission will be assessed using the Pittsburgh Cardiac Arrest Category and complemented by serum levels of lactate and neuron-specific enolase at baseline. These will be accounted for in the post-hoc adjustment. Abbreviations: CCU, Coronary Care Unit; OHCA, out-of-hospital cardiac arrest; MAP, mean arterial pressure; H, hour; CPC, Cerebral Performance Category; HRQoL, health-related quality of life; IV, intravenous; CVP, central venous pressure; TCD, transcranial Doppler ultrasound; NIRS, near-infrared spectroscopy; BP, blood pressure; MAP, mean arterial pressure; PaCO2, arterial partial pressure of carbon dioxide; FiO2, fraction of inspired oxygen; COx, cerebral oxygenation index; Mxa, mean velocity index. Cerebral rSO2 is measured using the INVOS cerebral oximeter (Medtronic, Dublin, Ireland) with bilateral frontotemporal sensors. Mechanical ventilatory settings are kept constant during assessments, and continuous pulse oximetry and end-tidal carbon dioxide monitoring ensures stable partial pressures of oxygen and carbon dioxide, minimizing confounding of the rSO2 signal and ensuring stationarity of the autoregulatory mechanism. Sedation and analgesia infusion rates are kept stable for ≥30 min before and throughout each assessment. Cerebral blood flow velocities are measured using Dolphin 4D (Viasonix, Yokneam, Israel) with a 2-MHz monitoring probe mounted on a Spencer's headframe. The headframe ensures continuous insonation at a fixed angle after an optimal Doppler signal has been obtained from the proximal middle cerebral artery at 50–60 mm depth. Intravenous ultrasound contrast (SonoVue, Braco, Italy) is used in the ∼10–15% of patients lacking adequate temporal acoustic windows.25 Each assessment comprises a 5-min baseline recording epoch followed by a supervised 30- to 40-min MAP manipulation phase. MAP is titrated upward then returned to baseline via vasopressor (noradrenaline as first choice) titration, targeting a total MAP excursion of ≥20 mmHg to generate sufficient pressure variability for valid autoregulatory curve construction. Absolute safety MAP limits of 60 mmHg (floor) and 110 mmHg (ceiling) apply throughout. Quantitative autoregulatory indices, the cerebral oximetry index (COx, derived from NIRS rSO2) and the mean velocity index (Mxa, derived from TCD cerebral blood flow velocity), are derived using the ICM+ brain monitoring software (Cambridge Enterprise, Cambridge, UK).26 COx and Mxa are moving Pearson correlation coefficients between rSO2 and mean middle cerebral artery flow velocity, respectively, with MAP; values ≥0.3 indicate impaired autoregulation.13, 27 Raw NIRS, TCD, and arterial pressure signals are streamed continuously, averaged into 10-s epochs to isolate slow vasogenic wave frequencies (0.005–0.05 Hz), then binned within 5 mmHg MAP intervals using the ICM+ error bar chart. Plotting Fisher-transformed COx and Mxa against the patient's MAP range produces a U-shaped curve; the nadir corresponds to the optimal MAP (MAPopt) at which autoregulation is most effective (Fig. 3).27 Indices are per modality per side, yielding up to 4 MAPopt values. When COx-derived and Mxa-derived MAPopt values are concordant, the consensus value is used; when discordant, the TCD-derived (Mxa) MAPopt is prioritized given its direct measurement of macrovascular flow.Fig. 3(i) Conceptual illustration of the cerebral oximetry index (COx) plotted against MAP, demonstrating derivation of MAPopt from the curve nadir. When autoregulation mechanism is intact, the COx–MAP curve is U-shaped; MAPopt represents the pressure at which autoregulation is most effective.(ii) A flat curve at uniformly low Mxa and/or COx across the full measured MAP range: autoregulation is fully preserved throughout — all assessed pressures are equally safe and the current MAP is maintained without adjustment.(iii) A monotonically decreasing Mxa and/or COx with increasing MAP (across the measured range) with no visible nadir: most commonly indicating that the measured MAP range falls below the patient's lower limit of autoregulation (LLA) – a trial of higher MAP is undertaken to seek the nadir before applying the default target.(iv) A flat curve at uniformly high Mxa and/or COx (>0.3): total vasoparalysis, representing complete loss of myogenic tone with fully pressure-passive CBF — the protocol reverts to the default MAP ≥75 mmHg with rSO2 ≥50%.Abbreviations: COx, cerebral oximetry index; MAP, mean arterial pressure; NIRS, near-infrared spectroscopy; MAPopt, optimal mean arterial pressure. (i) Conceptual illustration of the cerebral oximetry index (COx) plotted against MAP, demonstrating derivation of MAPopt from the curve nadir. When autoregulation mechanism is intact, the COx–MAP curve is U-shaped; MAPopt represents the pressure at which autoregulation is most effective. (ii) A flat curve at uniformly low Mxa and/or COx across the full measured MAP range: autoregulation is fully preserved throughout — all assessed pressures are equally safe and the current MAP is maintained without adjustment. (iii) A monotonically decreasing Mxa and/or COx with increasing MAP (across the measured range) with no visible nadir: most commonly indicating that the measured MAP range falls below the patient's lower limit of autoregulation (LLA) – a trial of higher MAP is undertaken to seek the nadir before applying the default target. (iv) A flat curve at uniformly high Mxa and/or COx (>0.3): total vasoparalysis, representing complete loss of myogenic tone with fully pressure-passive CBF — the protocol reverts to the default MAP ≥75 mmHg with rSO2 ≥50%. Abbreviations: COx, cerebral oximetry index; MAP, mean arterial pressure; NIRS, near-infrared spectroscopy; MAPopt, optimal mean arterial pressure. During MAP manipulation, TCD Doppler spectra, mean cerebral blood flow velocity and pulsatility index are visually inspected in real time as a qualitative cross-check: intact autoregulation produces stable cerebral blood flow velocity despite increasing MAP, with pulsatility index increasing as vessels vasoconstrict, whereas pressure-passive autoregulation shows cerebral blood flow velocity tracking MAP proportionally with no pulsatility index compensation. Four error bar curve patterns are recognized (Fig. 3): (i) U-shaped curve with an identifiable nadir: autoregulation is operative within measured MAP range — MAP is titrated to MAPopt, (ii) flat curve at uniformly low Mxa and/or COx across full measured MAP range indicating autoregulation is fully preserved – current MAP is maintained, (iii) monotonically decreasing Mxa and/or COx with increasing MAP without a nadir, indicating that the measured MAP range falls below the patient's lower limit of autoregulation and a trial of higher MAP is undertaken to seek the nadir before applying the default target, and (iv) flat curve at uniformly high Mxa and/or COx (>0.3), representing total vasoparalysis with fully pressure-passive cerebral blood flow —protocol reverts to MAP ≥75 mmHg with rSO2 ≥50%. Patterns (iii) and (iv) are distinguished: by whether elevated COx occurs uniformly (vasoparalysis) or only at the lower MAP range (below lower limit of autoregulation). The target of ≥75 mmHg exceeds the guideline-recommended minimum of 65 mmHg because: (i) MAP of 65 mmHg has been associated with decrease in cerebral oxygenation during the critical first 6–12 h post-ROSC28; (ii) in a pressure-passive brain, MAP directly determines cerebral blood flow, making a higher floor physiologically prudent; and (iii) published MAPopt estimates in post-arrest cohorts consistently exceed 65 mmHg.27, 28 The rSO2 ≥50% co-target provides real-time verification of adequate cerebral oxygen delivery. Targets are achieved within one hour of each assessment and maintained until the next scheduled assessment or end of the 72-h intervention. Feasibility assessment Feasibility is evaluated across six pre-specified metrics. Adherence-related metrics include: (i) protocol completion rate (proportion completing all three assessments); (ii) biomarker completeness (proportion evaluable NSE at baseline and 72 h); and (iii) protocol deviation rate. Additional metrics include: (i) enrolment rate (proportion of eligible patients enrolled within the 12-h post-ROSC, screen-to-enrolment ratio); (ii) monitoring signal yield (proportion of assessments yielding a valid autoregulatory curve with MAPopt from at least one modality, and proportion with concordant bilateral signals; and (iii) MAP target achievement (proportion of time within MAPopt ± 5 mmHg, and median time to target MAP). Deviations from protocol are documented on standardized case report forms with reasons. Endpoints The single-arm design permits within-patient paired comparisons for serial biomarkers and comparison against historical controls for selected outcomes but precludes formal testing for measures lacking comparators. The primary endpoint is the mean change in NSE concentration from baseline (<12 h post-ROSC) to 72 h. Secondary endpoints include peak concentrations within 72 h of cardiac biomarkers (high-sensitivity troponin T and N-terminal pro-B-type natriuretic peptide) and renal biomarkers (proenkephalin, bioactive adrenomedullin, and the urinary [tissue inhibitor of metalloproteinases-2]·[insulin-like growth factor-binding protein 7] product) (Supplemental Table S1); favorable neurological outcome (CPC 1–2) at hospital discharge and 6 months29; and health-related quality of life using EQ-5D-5L at 3, 6 and 12 months (the 6-month timepoint serves as the prespecified secondary endpoint for comparison against historical controls). Safety endpoints include arrhythmias, pulmonary edema, limb ischemia, and skin reactions to NIRS sensors, systematically assessed at each study visit. Serious adverse events are reported to the institutional review board within 7 calendar days. Exploratory endpoints include means scores based on Duke Activity Status Index,30 Montreal Cognitive Assessment,31 Repeatable Battery for the Assessment of Neuropsychological Status,32 Depression Anxiety and Stress Scale,33 and digital neurocognitive assessments, collected at hospital discharge and 3, 6, and 12 months; they are reported descriptively to characterize the recovery trajectory and inform outcome selection for a future multicenter trial. Autoregulation parameters (MAPopt values, proportion of time spent with intact versus absent autoregulation, temporal evolution) and protocol feasibility metrics (proportion of time within MAP target ± 5 mmHg, time to achieve target, protocol deviations) are also exploratory and summarized descriptively. Participant timeline Table 1 summarizes the schedule of enrolment, interventions, and assessments.Table 1Schedule of enrolment, interventions, and assessments.Study activityScreeningDay 1≤12 h ROSCDay 112 ± 4 h(O)Day 224 ± 4 hDay 348 ± 4 hDay 472 ± 4 hDay 7±1 dayDischarge±3 days3 mo±2 wk6 mo±2 wk12 mo±2 wkENROLLMENTEligibility screenXDeferred consent (LAR)aXPatient re-consentbX INTERVENTIONAutoregulation assessment(NIRS + TCD)X(X)XXMAPopt determination (ICM+)X(X)XXMAP titration to MAPoptc←———→ECG monitoring (aiTriage)X(X)XXX ASSESSMENTSBiomarkersNSE (primary endpoint)XXXXhsTnT, NT-proBNPXXXXpenKid, bio-ADMXXXXUrinary TIMP2 × IGFBP7XXXXBlood/urine biobankXXXXX(X)(X)(X)Clinical/monitoringBaseline & CCU dataXXXXXOutcomesCPCXXXXDASIXXXXNeurocognition (digital)(X)(X)(X)(X)MoCAXXXXRBANSXXXXDASS-21XXXXEQ-5D-5LXXXXSafetyAdverse event assessmentXXXXXXXXXXX = required; (X) = optional; (O) = optional timepoint; ← → = continuous throughout 72 h intervention period.Abbreviations: bio-ADM, bioactive adrenomedullin; CCU, coronary care unit; COx, cerebral oximetry index; CPC, Cerebral Performance Category; DASI, Duke Activity Status Index; DASS-21, Depression Anxiety and Stress Scale; EQ-5D-5L, EuroQol 5-Dimension 5-Level; hsTnT, high-sensitivity troponin T; IGFBP7, insulin-like growth factor binding protein 7; LAR, legally acceptable representative; MAP, mean arterial pressure; MAPopt, optimal mean arterial pressure; MoCA, Montreal Cognitive Assessment; NIRS, near-infrared spectroscopy; NSE, neuron-specific enolase; NT-proBNP, N-terminal pro-brain natriuretic peptide; penKid, proenkephalin; RBANS, Repeatable Battery for the Assessment of Neuropsychological Status; ROSC, return of spontaneous circulation; rSO2, regional cerebral oxygen saturation; TCD, transcranial Doppler ultrasound; TIMP2, tissue inhibitor of metalloproteinase 2.aDeferred consent obtained from legally acceptable representative (LAR) within 3 working days of enrolment, following priority hierarchy: spouse → adult children → parents → siblings.bWritten informed consent from patient upon regaining decision-making capacity. Options: full consent, partial consent (retain data/samples, withdraw from further participation), or complete withdrawal.cMAP titrated to individualized MAPopt if autoregulation intact (identifiable U-shaped COx curve with nadir); default MAP ≥75 mmHg with rSO2 ≥50% if autoregulation absent. Safety limits: MAP 50–110 mmHg at all times. Targets achieved within 1 h and maintained until next assessment. Table 1Schedule of enrolment, interventions, and assessments.Study activityScreeningDay 1≤12 h ROSCDay 112 ± 4 h(O)Day 224 ± 4 hDay 348 ± 4 hDay 472 ± 4 hDay 7±1 dayDischarge±3 days3 mo±2 wk6 mo±2 wk12 mo±2 wkENROLLMENTEligibility screenXDeferred consent (LAR)aXPatient re-consentbX INTERVENTIONAutoregulation assessment(NIRS + TCD)X(X)XXMAPopt determination (ICM+)X(X)XXMAP titration to MAPoptc←———→ECG monitoring (aiTriage)X(X)XXX ASSESSMENTSBiomarkersNSE (primary endpoint)XXXXhsTnT, NT-proBNPXXXXpenKid, bio-ADMXXXXUrinary TIMP2 × IGFBP7XXXXBlood/urine biobankXXXXX(X)(X)(X)Clinical/monitoringBaseline & CCU dataXXXXXOutcomesCPCXXXXDASIXXXXNeurocognition (digital)(X)(X)(X)(X)MoCAXXXXRBANSXXXXDASS-21XXXXEQ-5D-5LXXXXSafetyAdverse event assessmentXXXXXXXXXXX = required; (X) = optional; (O) = optional timepoint; ← → = continuous throughout 72 h intervention period.Abbreviations: bio-ADM, bioactive adrenomedullin; CCU, coronary care unit; COx, cerebral oximetry index; CPC, Cerebral Performance Category; DASI, Duke Activity Status Index; DASS-21, Depression Anxiety and Stress Scale; EQ-5D-5L, EuroQol 5-Dimension 5-Level; hsTnT, high-sensitivity troponin T; IGFBP7, insulin-like growth factor binding protein 7; LAR, legally acceptable representative; MAP, mean arterial pressure; MAPopt, optimal mean arterial pressure; MoCA, Montreal Cognitive Assessment; NIRS, near-infrared spectroscopy; NSE, neuron-specific enolase; NT-proBNP, N-terminal pro-brain natriuretic peptide; penKid, proenkephalin; RBANS, Repeatable Battery for the Assessment of Neuropsychological Status; ROSC, return of spontaneous circulation; rSO2, regional cerebral oxygen saturation; TCD, transcranial Doppler ultrasound; TIMP2, tissue inhibitor of metalloproteinase 2.aDeferred consent obtained from legally acceptable representative (LAR) within 3 working days of enrolment, following priority hierarchy: spouse → adult children → parents → siblings.bWritten informed consent from patient upon regaining decision-making capacity. Options: full consent, partial consent (retain data/samples, withdraw from further participation), or complete withdrawal.cMAP titrated to individualized MAPopt if autoregulation intact (identifiable U-shaped COx curve with nadir); default MAP ≥75 mmHg with rSO2 ≥50% if autoregulation absent. Safety limits: MAP 50–110 mmHg at all times. Targets achieved within 1 h and maintained until next assessment. Schedule of enrolment, interventions, and assessments. X = required; (X) = optional; (O) = optional timepoint; ← → = continuous throughout 72 h intervention period. Abbreviations: bio-ADM, bioactive adrenomedullin; CCU, coronary care unit; COx, cerebral oximetry index; CPC, Cerebral Performance Category; DASI, Duke Activity Status Index; DASS-21, Depression Anxiety and Stress Scale; EQ-5D-5L, EuroQol 5-Dimension 5-Level; hsTnT, high-sensitivity troponin T; IGFBP7, insulin-like growth factor binding protein 7; LAR, legally acceptable representative; MAP, mean arterial pressure; MAPopt, optimal mean arterial pressure; MoCA, Montreal Cognitive Assessment; NIRS, near-infrared spectroscopy; NSE, neuron-specific enolase; NT-proBNP, N-terminal pro-brain natriuretic peptide; penKid, proenkephalin; RBANS, Repeatable Battery for the Assessment of Neuropsychological Status; ROSC, return of spontaneous circulation; rSO2, regional cerebral oxygen saturation; TCD, transcranial Doppler ultrasound; TIMP2, tissue inhibitor of metalloproteinase 2. Deferred consent obtained from legally acceptable representative (LAR) within 3 working days of enrolment, following priority hierarchy: spouse → adult children → parents → siblings. Written informed consent from patient upon regaining decision-making capacity. Options: full consent, partial consent (retain data/samples, withdraw from further participation), or complete withdrawal. MAP titrated to individualized MAPopt if autoregulation intact (identifiable U-shaped COx curve with nadir); default MAP ≥75 mmHg with rSO2 ≥50% if autoregulation absent. Safety limits: MAP 50–110 mmHg at all times. Targets achieved within 1 h and maintained until next assessment. Sample size To detect a 0.5 effect size (mean paired difference divided by standard deviation of paired differences) with 80% power at two-sided type-I error rate of 0.05, 34 patients are required to complete the study (calculated in Stata V17, College Station, TX). Accounting for an anticipated 30% dropout rate—reflecting the high early mortality (within 72 h post-ROSC) in this critically ill population—49 patients will be enrolled. Statistical analysis The primary analysis uses a paired t-test comparing NSE at 72 h to baseline in the modified intention-to-treat population (all enrolled patients surviving to 72 h), reported as mean change with 95% confidence interval. NSE trajectory is additionally compared against historical controls using linear regression adjusting for age, initial rhythm, and arrest duration. Wilcoxon rank-sum test and quantile regression are pre-specified non-parametric alternatives if data are highly skewed. The proportion achieving CPC 1–2 is compared against historical controls using logistic regression, and EQ-5D-5L scores using linear regression, adjusting for potential confounders. Exploratory endpoints lacking institutional historical comparators are summarized descriptively with no formal hypothesis testing. Autoregulation parameters (MAPopt values, proportion with intact versus absent autoregulation, temporal evolution across serial assessments) are summarized descriptively. Exploratory analyses examine the dose–response relationship between proportion of time within MAPopt ± 5 mmHg and clinical outcomes. Missing data, expected primarily due to early death, are addressed using multiple imputation under the missing-at-random assumption. Data monitoring and safety An independent data and safety monitoring committee, independent of the study team, will review primary safety endpoints, serious adverse events, and aggregate outcome data at prespecified intervals. No formal interim efficacy analysis is planned; the data and safety monitoring committee may recommend protocol modification or early termination if safety concerns arise. The study team meets quarterly to review adverse events and protocol adherence. Dissemination of findings Trial results will be submitted for publication in a peer-reviewed journal, deposited in the trial registry, and presented at relevant scientific conferences. Patient and public involvement There is no patient or public involvement in the study design, conduct or dissemination plans. Ethics and consent Deferred consent is obtained from a legally acceptable representative within three working days of enrolment. When patients regain decision-making capacity, an additional consent discussion is held and written informed consent obtained for ongoing participation. All data are de-identified prior to analysis and entered into the institutional REDCap electronic data management system. Discussion This study addresses a critical gap between observational evidence supporting individualized MAP targeting after cardiac arrest and the need for prospective interventional data. The scientific rationale rests on three converging lines of evidence. First, cerebral autoregulation is impaired in 35–50% of OHCA patients,8, 9, 34 with MAPopt values ranging from 55 to over 100 mmHg—heterogeneity that no single fixed target can accommodate.13, 14 Excess pressure above MAPopt risks cerebral hyperperfusion and vasogenic oedema, while pressure below perpetuates ischemia. Targeting the individualized MAPopt range is therefore expected to improve cerebral outcomes in a way that any fixed threshold cannot. Second, time spent outside the autoregulatory range follows a dose–response relationship with worse outcomes. Under standard care, MAP is outside the optimal range for half the monitoring period, with sub-MAPopt deviations associated with increased mortality.13, 34 Third, neutral fixed-target RCTs including Carbon dioxide, Oxygen and Mean arterial pressure After Cardiac Arrest and REsuscitation (COMACARE) and Blood Pressure and Oxygenation Targets in Post Resuscitation Care (BOX) paradoxically support individualization. With only ∼10–15 mmHg between-group separation and no stratification by autoregulatory status, bidirectional misalignment between assigned targets and individual MAPopt dilutes treatment effects.7, 18 NEURO-INTACT builds on foundational feasibility work demonstrating NIRS-derived MAPopt could be generated, albeit not in all patients.13, 14, 35 Our protocol addresses this limitation through two design innovations: active MAP manipulation, since spontaneous MAP fluctuations may lack sufficient range or variability for valid autoregulatory curve construction and do not ensure the causal directionality required by the correlation methodology; concurrent bilateral NIRS and TCD monitoring generates up to four independent MAPopt estimates per assessment, maximizing the probability of obtaining a valid computation. Together, these approaches, combined with real-time qualitative TCD assessment, enable autoregulatory classification across the full physiological spectrum. Our preliminary case-series of 16 patients confirmed feasibility and provided the rationale for serial reassessment and prioritizing Mxa-derived MAPopt when indices are discordant.16 NEURO-INTACT’s key innovative elements – active MAP-guided autoregulation assessment, multimodal concurrent monitoring, and serial assessments at three timepoints – directly addresses the principal limitations of prior observational work: reliance on spontaneous MAP variability, single-modality measurement, and static single-timepoint design. Serial assessments capture the dynamic evolution of autoregulation during the post-resuscitation phases which no single-assessment study could characterize, while the cardiac and renal biomarkers will provide mechanistic insight into whether cerebral perfusion optimization confers benefits beyond the brain. Several limitations warrant acknowledgment. Conversion from a parallel-group randomized controlled trial to a single-arm design was driven by recruitment challenges inherent to emergency critical care trials, with narrow enrolment windows, deferred consent requirements, and high early attrition. The single-arm design introduces potential confounding from secular trends in clinical practice and case mix, precluding definitive causal inference Single-center conduct may limit generalizability. Individual values of NSE, the surrogate biomarker, may be confounded by hemolysis. Functional outcomes beyond CPC and EQ-5D-5L lack institutional historical comparators and are reported descriptively; they nonetheless provide valuable cohort characterization data to inform outcome selection for a future definitive trial. MAPopt may be unobtainable in a proportion of patients, and NIRS signal quality may be affected by frontal sinus anatomy, skin pigmentation, or cerebral edema. If this study demonstrates feasibility, a signal of benefit, and acceptable safety, the findings will inform a multicenter randomized trial; the concurrent use of two modalities will additionally clarify whether NIRS alone suffices, potentially simplifying the protocol for multicenter deployment. Statement of ethics This study protocol was reviewed and approved by the National Healthcare Group Domain Specific Review Board, Singapore, approval number [NHG DSRB Ref: 2020/01455]. This study is registered on ClinicalTrials.gov (ID NCT05679739). Given that prospective subjects are comatose, deferred consent is obtained as described in the Methods. This study is conducted in accordance with the Declaration of Helsinki and ICH Good Clinical Practice guidelines. Important protocol amendments will be communicated in writing to all investigators, and will be reflected in updated versions of the protocol. Funding sources This study is supported by the National Medical Research Council Transitional Award (MOH-001146). The funder had no role in the design, data collection, data analysis, and reporting of this study. Author contributions SL Lim, MEH Ong, M Gandhi and VK Sharma conceptualized and designed the study. KL Woo, SL Lim and VK Sharma contributed to study logistics, study execution as well as data curation. P. Smielewski and E. Beqiri provided technical expertise in cerebral autoregulation monitoring. SL Lim and KL Woo drafted the manuscript. SL Lim obtained study funding. All authors read and approved the final manuscript. Data availability statement As this is a study protocol, no datasets have been generated. Individual de-identified participant data generated during the trial will be made available upon reasonable request to the corresponding author, subject to institutional data sharing agreements and applicable regulations. CRediT authorship contribution statement Shir Lynn Lim: Writing – original draft, Funding acquisition, Data curation, Conceptualization. Kai Lee Woo: Writing – original draft, Project administration, Data curation. Mihir Gandhi: Writing – review & editing, Formal analysis. Erta Beqiri: Writing – review & editing, Resources. Peter Smielewski: Writing – review & editing, Visualization, Resources. Marcus E.H. Ong: Writing – review & editing, Supervision, Conceptualization. Vijay Kumar Sharma: Writing – review & editing, Supervision, Data curation, Conceptualization. Declaration of competing interest SL Lim is supported by the National Medical Research Council Transitional Award; she has received research grants from National University Health System, National Kidney Foundation of Singapore and Singapore Heart Foundation. She is an Associate Editor for Resuscitation and on the Editorial Board of Resuscitation Plus. MEHO reports grants from the Laerdal Foundation, Laerdal Medical, and Ramsey Social Justice Foundation for funding of the Pan-Asian Resuscitation Outcomes Study; an advisory relationship with Global Healthcare SG, a commercial entity that manufactures cooling devices; and funding from Laerdal Medical on an observation program to their Community CPR Training Center Research Program in Norway. MEH Ong is a Scientific Advisor to TIIM Healthcare SG and Global Healthcare SG. He is a member of the Editorial Board of Resuscitation. These organizations have no role in conducting this research. P Smielewski receives royalties from Cambridge Enterprise Ltd for licensing of the ICM+ software used in this study for data collection and analysis. All other authors reported no competing interests. All other authors reported no competing interests.