Enhancing Neuroprotection in Cardiac and Aortic Surgeries: A Narrative Review

Abstract

Background: Neurological injury poses a significant challenge in aortic surgery, encompassing spinal cord injury from thoraco-abdominal aorta intervention or stroke post-surgery on the arch and ascending aorta. Despite ample literature and proposals, a fully effective strategy for preventing or treating neurological injury remains elusive. This narrative review aims to analyze the most common neuroprotective strategies implemented for aortic arch surgery and aortic surgery. Results: Results from the reviewed studies showed that several strategies, including deep hypothermia cardiac induction (DHCA) and cerebral perfusion techniques (retrograde cerebral perfusion, RCP, and selective anterograde cerebral perfusion, SACP) aim to mitigate these risks. Monitoring methods such as electroencephalogram (EEG), somatosensory evoked potential (SEPs), and near-infrared spectroscopy (NIRS) offer valuable insights into cerebral function during surgery, aiding in the management of hypothermia and perfusion. Pharmacological agents and blood gas management (pH stat vs. alpha stat, hematocrit level, glycemic control) are crucial in preventing post-operative complications. Additionally meticulous management of atheromatous debris is essential to minimize embolic risks during surgery. Methods: For this narrative review, PubMed, Scopus, and Medline have been used to search articles about neuroprotection strategies in aortic and aortic arch surgeries. The search was narrowed to articles between 1975 and 2024. A total of 3418 articles were initially identified to be potentially relevant for this review. A total of 66 articles were included and were found to match the inclusion criteria. Conclusions: While an overabundance of neuroprotection strategies exists for cardiac surgery, particularly in procedures involving the aorta and the arch, their efficacy varies, with some well-documented and others still under scrutiny. Further research is imperative to advance our comprehension and refine prevention techniques for cardiac-surgery-related brain injury. This is crucial given its substantial contribution to both mortality and, notably, post-operative morbidity.

1. Introduction

Neurological complications arising from cardiac surgery pose significant challenges, leading to heightened morbidity and mortality rates. Prolonged hospital stays and increased reliance on rehabilitation programs incur substantial costs, ultimately diminishing post-operative quality of life. Over recent years, mitigating postoperative neurological complications has emerged as a key objective. The reported incidence of neurological side effects linked to cardiac surgery ranges from 1.4% to 6% [1], with higher rates observed in elderly patients (up to 9% in those over 75 years) [2,3] or in patients with significant preoperative risk factors (up to 16%), such as age > 75 years, cerebrovascular disease, peripheral vascular disease, carotid disease, smoke, severe left ventricular dysfunction, diabetes mellitus, renal failure, hypertension and COPD, alcohol abuse, EuroScore II, depression, low level of education, depression, sleep disturbance ad apnea, preoperative cognitive reserve dysfunction, or aortic arch variants [4]. Furthermore, neuropsychological dysfunction can affect up to half of all patients, manifesting as reversible cognitive impairment impacting various cognitive function such as memory, attention, and processing speed. In some cases, cognitive decline persists for years post-surgery [5]. The occurrence of post-operative neurological complications significantly amplifies mortality rates, with a three- to fourfold increase noted. Major neurological damage, such as stroke or seizures, escalates this risk even further, leading to a tenfold increase in both hospital mortality and within the first-year post-surgery. Additionally, intensive care unit and hospital stays are prolonged. Differentiating between two types of adverse cerebral outcomes, type 1 (including stroke, transient ischemic attack, and coma) and type 2 (comprising neuropsychological dysfunction), provides insights into varied causes and prognoses [5]. The cardiopulmonary bypass circuit (CBP), a core component of cardiac surgery, contributes to neurological damage through the exposure of blood to foreign surfaces, triggering inflammatory responses, coagulation abnormalities, and embolism [6]. Most neurologic complications arising from cardiac surgery are attributed to the use of CPB. The etiology is multifactorial, with neurological side effects stemming from embolic and inflammatory reactions associated with CPB. Various mechanisms contribute to neurologic complication post-CPB:

-

Embolism: Embolic particles, categorized as macro- or micro-emboli, are a primary cause of neurologic damage. Macro-emboli, often from surgical manipulation, lead to focal defects, while micro-emboli, originating from sources like air, lipids, cellular aggregates, or exogenous material, typically affect smaller vessels and are associated with cognitive decline [7,8,9].

-

Inflammatory response activation: CBP triggers an inflammatory response due to blood contact with non-endothelial surfaces, leading to platelet aggregation, protein denaturalization, and damage to the blood–brain barrier, resulting in increased permeability [10].

-

Disorders in neuronal metabolism: While hypothermia aids in lowering cerebral metabolic rate and preventing blood–brain barrier dysfunction, it can also reduce tissue oxygen transfer and increase the risk of cerebral hyperthermia during rewarming [11].

-

Cerebral hypoperfusion: Prolonged cerebral hypoperfusion during normothermia or decreased blood flow can lead to permanent neurologic damage. Cerebral perfusion is regulated by an autoregulation mechanism dependent on CO₂ arterial pressure (pCo₂) [12]. Risk factors for neurological complication after cardiac surgery encompass various preoperative and perioperative conditions that predispose patients to embolic events, cerebral hypoperfusion, inflammatory reactions, and metabolic disturbances in neuronal tissue [13,14,15].

Among the preoperative risk factors are included:

-

Age: older age, especially beyond 70 to 75 years increase the susceptibility to stroke, primarily attributed to severe aortic atherosclerosis and underlying cognitive deterioration.

-

Carotid disease: factors such as previous stroke, smoking, left main coronary disease, or age over 65 years predict carotid disease, increasing the risk of stroke during combined endarterectomy (CEA) and coronary bypass grafting procedures.

-

Cerebrovascular disease: Patient with history of previous stroke have a higher perioperative stroke incidence.

-

Peripheral vascular disease: aortic atherosclerosis significantly increases the risk of perioperative cerebrovascular disease and future stroke.

-

Severe left ventricular dysfunction: preoperative cardiovascular symptoms like sustained hypotension and low cardiac output predispose patient to neurologic side effects.

-

Other factors: diabetes mellitus, renal failure, hypertension, and chronic obstructive pulmonary disease indirectly increase the risk of neurological complications.

Additionally, among the perioperative risk factors are included:

-

Type of surgical procedure: open cardiac chamber surgery and combined procedures are associated with a higher risk of cerebral injury due to increased debris and air emboli.

-

Duration of procedure: longer surgeries increase the risk of neurological side effect, with emboli number increasing proportionally to the duration of surgery.

-

Post-operative complications: any condition leading to low cardiac output or embolic situations, such as perioperative myocardial infarction or post-operative atrial fibrillation, significantly increase the risk of early and late neurological complications.

As previously mentioned, one of the key factors determining the risk of neurological damage is the type of surgery, and among these, aortic surgery, particularly involving the aortic arch is one of the procedures more at risk for the neurological damage. The practice of temporary cessation of circulation to achieve a bloodless surgical field remains pivotal in open aortic arch replacement procedures. This complex process involves systemic cooling, induction of circulatory arrest, and subsequent rewarming, representing one of the most intricate aspects of circulatory management and surgical procedures in contemporary medicine. While all tissues experience ischemic stress during circulatory arrest, nervous tissue is particularly vulnerable due to its heightened metabolic demands and reliance on aerobic energy production. Consequently, the significant and often irreversible morbidity associated with ischemic brain injury underscores the critical importance of prioritizing neuroprotection during circulatory management in aortic arch surgery [16]. Although the literature is abundant, it remains highly debated and discordant in its findings regarding the neuroprotective strategies to be employed in aortic arch surgery. However, the most reliable neuroprotective methods involve two modalities: inducing hypothermia before circulatory arrest to reduce metabolic demand, thereby enhancing ischemic tolerance [17]. Additionally, establishing a circuit for regional cerebral perfusion and implementing cerebral cooling after the onset of systemic circulatory arrest are essential strategies [18,19,20].

In addition to these approaches, pharmacological strategies and neurological monitoring are also employed to minimize neurological damage as much as possible. In fact, there is abundant literature on pharmacological strategies that could be used to reduce or prevent neurological damage, such as the use of propofol, thiopentone, magnesium, lidocaine, mannitol, and glucocorticoids, but their actual efficacy has never been demonstrated [21]. Despite ample literature and proposals, a fully effective strategy for preventing or treating neurological injury remains elusive. This narrative review aims to analyze the most common neuroprotective strategies implemented for aortic arch surgery and aortic surgery.

2. Materials and Methods

This review is intended to discuss the relevant literature that has studied neuroprotection strategy in cardiac surgery and aortic and aortic arch surgery. The literature was searched on PubMed and Scopus. The search was narrowed to articles between 1975 and 2024. The search terms were “neuroprotection cardiac surgery”, “neuroprotection aortic arch surgery”, “brain damage after aortic arch surgery”, “neuroprotection in descending aorta surgery”, which showed 1998 results for “neuroprotection cardiac surgery”, 241 results for “neuroprotection aortic arch surgery”, 288 results for “brain damage aortic arch surgery”, and 183 results for “neuroprotection in descending aorta surgery”. The review was aimed to find answers to the following questions: “what are the strategies for preventing brain damage associated with cardiac surgery? What are the neuroprotective strategies during aortic and arch surgery? Are they sufficient to prevent damage or is further study needed in this field?”. After adding these inclusion criteria, the total papers were 449. Exactly 253 papers were included after reading the titles and the abstract. Finally, after reading the full text, a total of 83 articles were included based on the quality of the studies (Figure 1).

3. Results

3.1. Deep Hypothermic Circulatory Arrest

Deep hypothermic circulatory arrest (DHCA) without adjunctive cerebral perfusion is a straightforward method of cerebral protection during aortic arch surgery, especially suitable for surgeons with limited experience in this procedure [22], and was used in Griepp’s initial successful series of arch replacements. It involves inducing systemic hypothermia through cardiopulmonary bypass to achieve a nasopharyngeal temperature of 14.1 °C to 20 °C. A temperature gradient (arterial inflow to venous return) that does not exceed 10 °C is maintained during the cooling period to prevent the formation of gaseous emboli.

This approach does not require complex perfusion strategies or extensive monitoring. Although DHCA alone has demonstrated excellent neuroprotection for arch replacement with circulatory arrest times under 40 min [23], a longer arrest time can increase the risk of adverse neurological outcomes and mortality [24]. Therefore, adjunctive cerebral perfusion has been introduced to mitigate these risks. DHCA alone is a straightforward and effective method for cerebral protection during aortic arch surgery, but adjunctive cerebral perfusion may be necessary for longer circulatory arrest times to minimize neurological complications.

3.2. Retrograde Cerebral Perfusion

Retrograde cerebral perfusion pioneered by Ueda et al. [25] involves cannulating and occluding the superior vena cava to infuse hypothermic arterial blood from the cardiopulmonary bypass circuit into the brain in a retrograde direction during circulatory arrest. The goal is to flush embolic material from the cerebral circulation, maintain cerebral hypothermia and support cerebral metabolism. The cardiopulmonary bypass technique typically involves bicaval venous cannulation, with the arterial line incorporating Y-connectors connected to the venous line limbs which are clamped during anterograde cerebral flow. Usually, upon initiating retrograde cerebral perfusion (RCP), the superior vena cava cannula is snared, anterograde flow is stopped, the arterial cannula is clamped, and the limb linking the arterial return line to superior vena cava canula is opened. Perfused blood is then redirected to the oxygenator via cardiotomic suction into the open thoracic aorta, to the surgical field, and through the inferior vena cava cannula. In some centers, the retrograde perfusion technique may encompass the entire venous system using a single atrial cannula. Retrograde cerebral perfusion can be administered continuously during the interruption of anterograde cerebral flow or intermittently. Some clinicians briefly apply RCP at the conclusion of the hypothermic circulatory arrest (HCA) period solely to flush embolic debries from the cerebral vasculature [26]. Clinical studies [27] have shown that adding RCP to deep hypothermic circulatory arrest (DHCA) significantly reduces mortality and stroke rates during arch reconstruction procedures. However, the incidence of temporary neurological dysfunction (TND) remained significant [28], suggesting that while RCP improves cerebral protection, further optimization may be required to minimize neurological complications during arch reconstruction.

3.3. Selective Anterograde Cerebral Perfusion (SACP)

Selective anterograde cerebral perfusion, introduced by De Bakey and colleagues, ensures sufficient cerebral blood flow during circulatory arrest [29]. In 1986, Frist and colleagues [30] reported a Stanford experiment with 10 patients using a combination of low-flow CPB and moderate (25–28 °C) hypothermic selective antegrade cerebral perfusion (SACP). In 1989, Kazui and colleagues introduced a branch graft technique for arch replacement using hypothermic SACP at 25 °C that remains in widespread use [31].

Various techniques for SACP delivery exist, such as unilateral SACP (uSACP) via right axillary artery cannulation or bilateral SACP (bSACP) via innominate and left carotid artery. SACP can be delivered bilaterally or unilaterally, wherein contralateral perfusion depends on collateral pathways, most prominently the circle of Willis.

In SACP, before sternotomy, an 8 mm graft is anastomosed end-to-side to the right axillary artery and utilized as the arterial inflow conduit to initiate cardiopulmonary bypass. During hypothermic circulatory arrest (HCA), blood flow is reduced to 8–10 mL/kg/min, and vascular clamps are applied to the base of the innominate and left common carotid arteries. Occlusion of the left carotid artery is performed to increase pressure in the extracranial collateral system and minimize steal phenomenon. This facilitates blood flow via the right common carotid and right vertebral arteries to perfuse the brain and spinal cord. Studies show comparable outcomes between unilateral and bilateral SACP [32].

The adoption of selective anterograde cerebral perfusion in conjunction with hypothermia for cerebral protection has led to a shift from deep hypothermia to moderate levels of hypothermia. This strategy is based on the premise that SACP modifies the total body circulatory arrest to lower body circulatory arrest. With cerebral perfusion maintained by cold blood during circulatory arrest, systemic hypothermia primarily aims to safeguard visceral organs, skeletal muscle, and the spinal cord through metabolic suppression. Since the metabolic rate of visceral organs and skeletal muscle is lower than of the brain, they require less hypothermia for optimal protection and are more tolerant to ischemia. Moderate hypothermic circulatory arrest combined with selective anterograde cerebral perfusion (MHCA/SACP) is now the preferred circulation management approach in many high-volume aortic centers worldwide.

3.4. Neurophysiological Intraoperative Monitoring

Intraoperative neuromonitoring is often used as an aid in managing hypothermia and monitoring for cerebral hypoperfusion [33,34].

3.4.1. EEG and Peripheral Somatosensory-Evoked Potentials (SEPs)

The primary modality of neurophysiologic intraoperative monitoring (NIOM) utilized during aortic arch surgery is electroencephalography (EEG) [16,34]. EEG monitoring is essential during the cooling process to assess electrocerebral activity in real time, serving as an indicator of the extent of hypothermia-induced metabolic suppression of the brain. Typically, EEG monitoring involves the placement of gold disk electrodes on the scalp according to the International System of Electrode Placement. Continuous EEG monitoring begins upon initiation of cardiopulmonary bypass with sensitivity adjustments made as cooling progresses to allow assessment of low-amplitude activity. The electrocerebral activity pattern during hypothermic circulatory arrest in aortic arch replacement has been extensively characterized, showing predictable changes as cooling progresses [35,36]. These changes include lateralized, generalized, or bilateral independent periodic discharges, transient increase in EEG wave amplitude, and eventually, a burst suppression pattern followed by complete electrocerebral inactivity. During rewarming, a reversed progression occurs, with electrocerebral activity returning to normal-amplitude continuous activity. Although changes in electrocerebral activity correlate with temperature, the exact temperature thresholds vary among patients and procedural factors, highlighting the importance of direct monitoring during aortic arch surgery to guide patient-specific management.

In addition to EEG, median-nerve-somatosensory-evoked potentials (SEPs) have been utilized during aortic arch surgery with hypothermic circulatory arrest to provide complementary information on brain activity [16,34]. Cortical (N20), subcortical (P13/14 and N18), and peripheral responses can be monitored, with cortical responses being lost first, followed by subcortical and peripheral responses. Similar to EEG, SEPs can demonstrate cerebral metabolic suppression before circulatory arrest initiation, enhancing the comprehensive neurophysiologic assessment during surgery.

Anesthetic agents exert significant effects on neurophysiology, thus necessitating careful management during aortic arch surgery to optimize neurophysiologic intraoperative monitoring interpretation.

If use of propofol or a barbiturate is planned as an adjunct for cerebral protection, administration is delayed until after the EEG endpoint or the target temperature for HCA (hypothermic circulatory arrest) has been reached.

During circulatory arrest, administration of medications, including anesthetics, neuromuscular blocking agents, and antibiotics, is not feasible. Hence, all necessary drugs should be administered before the initiation of hypothermic circulatory arrest (HCA). Intravenous drug infusion and volatile anesthetics agents should be stopped throughout the HCA period. Anesthetics needs decrease during deliberate hypothermia, and general anesthesia becomes unnecessary once EEG burst suppression begins and electrocortical silence ensues. However, anesthesia should be reinstated during rewarming to ensure adequate sedation once the nasopharyngeal temperature reaches around 30 °C or when consistent EEG activity resumes, typically about 30 min after rewarming initiation.

3.4.2. NIRS (Near Infrared Spectroscopy)

NIRS monitors cerebral regional saturation (rSO₂) in the frontal cortex. NIRS provides continuous monitoring unaffected by anesthetic agents or no pulsatile perfusion cardiopulmonary bypass. During HCA and SACP, rSO₂ either increases or remains at baseline. Recovery towards the baseline occurs during reperfusion. Any sudden unilateral decrease in rSO₂ indicates a regional cerebral perfusion decrease, warranting immediate communication with the surgical team [37,38,39,40].

During temporary clamping of the common carotid or the innominate artery, an ipsilateral decrease indicates decreased blood flow, while contralateral decrease suggests vascular steal via circle of Willis through the contralateral carotid artery that is open to the aortic arch. In this situation, clamping the contralateral carotid artery during SACP may improve cerebral perfusion to the contralateral hemisphere [37,38,39,40].

Unilateral SACP delivery may cause ipsilateral rSO₂ decrease due to cannula misplacement, while a severe contralateral decrease may prompt contralateral cannulation. Bilateral decrease may also result from aortic dissection or graft thrombosis [37,38,39,40].

While NIRS holds promise as a monitoring tool during aortic arch surgery, concerns persist, especially in adult patients with larger cerebral volumes and more interfering tissue. Questions arise regarding the reliability of NIRS in assessing deeper brain tissue oxygenation levels. Moreover, there is a lack of data defining thresholds of duration of NIRS-detected brain hypoxia without neurocognitive harm. Consequently, a decision based on NIRS-derived rSO₂ levels remains somewhat arbitrary [16,34].

3.4.3. Transcranial Doppler

The Transcranial Doppler (TCD) is a valuable tool for assessing blood flow in major intracranial arteries, such as the middle cerebral artery. It is effective in detecting both micro- and macro-emboli during aortic arch surgery. This technique enables the measurement of cerebral blood flow and its distribution, allowing optimization of anterograde cerebral perfusion during circulatory arrest. In cases where an incomplete circle of Willis is identified, TCD facilitates prompt adjustment from unilateral to bilateral cerebral perfusion [41,42]. Unlike NIRS, TCD provides real-time detection of cerebral blood flow, potentially enhancing perfusion strategy optimization at an earlier stage. However, the accuracy of TCD depends on the image quality obtained through the trans-temporal window, stable monitoring position, and the expertise of the sonographer. Further large-scale studies are needed to confirm the validity of TCD monitoring during selective anterograde cerebral perfusion, currently viewed as a complementary monitoring approach alongside NIRS [43].

3.5. Topical Cooling

Topical cooling involves placing ice bags around the patient’s head to enhance or sustain cerebral cooling. However, there is insufficient evidence to validate the efficacy of current techniques and devices for topical cooling [44]. The theoretical risks of thermal or pressure injury and retinal ischemia are noted [45].

3.6. Blood Gas Management

The debate between using pH-stat or alpha stat management continues [46]. The alpha stat approach aims to maintain pH in a temperature-uncorrected format, keeping the partial pressure of carbon dioxide (pCO₂) within the normal range at 37 °C, despite the patient’s lower body temperature during HCA. In contrast, the pH-stat approach involves adding CO₂ to maintain a pCO₂ of 40 mmHg at the patient’s true (low) temperature, resulting in increased brain perfusion due to a cerebral vasodilatory effect (this approach is widely used in infants undergoing HCA).

Blood gas management during cooling for HCA and rewarming in adult patients typically follows the alpha-stat method, as in routine CBP without hypothermia. This approach is supported by evidence indicating that it helps preserve cerebral blood flow autoregulation and reduces the risk of complications such as cerebral thromboembolism, cerebral edema, reperfusion injury, and unintended cerebral hyperthermia during rewarming [47].

3.7. Pharmacological Agents

Pharmacological agents for cerebral protection are frequently used, although their efficacy remains uncertain. Barbiturates or propofol are commonly employed to suppress cerebral activity and metabolism before HCA. However, concerns about adverse effects, such as myocardial depression and delayed emergence from anesthesia, exist, particularly with long-acting barbiturates [48,49]. Administration of these agents may also interfere with neurophysiological monitoring during surgery. If barbiturates or propofol are used, dosing should be guided by the goal of achieving electrocortical silence during HCA, rather than using a fixed dose.

Mannitol, often included in cardiopulmonary bypass pump solutions for its diuretic effects, lacks sufficient evidence supporting its use for cerebral protection.

Some institutions also use anti-inflammatory or anticonvulsant agents, lidocaine, magnesium, or calcium channel blockers, despite limited evidence for their efficacy. Despite minimal evidence supporting their efficacy, clinicians may justify their use based on theoretical physiological and biochemical rationales.

Lidocaine is thought to improve neurocognitive outcomes by reducing the cerebral metabolic rate of oxygen. Magnesium may offer protection by decreasing voltage-sensitive N-methyl-D-aspartate (NMDA)-activated calcium channels, potentially reducing vasospasm [50].

Glucocorticoids administered before CPB believed to reduce brain damage by suppressing the proinflammatory response and acting as an antioxidant for ischemic tissue. However, they are not recommended for cerebral protection due to concerns about adverse effects, such as abnormal glucose metabolism and immune system modulation [51].

3.8. Hematocrit

A higher hematocrit (25–30% vs. 10–20%) has been shown to improve functional outcomes in animal studies, reduce intracranial pressure, and improve perfusion pressures [52,53].

3.9. Rewarming Strategies

Rewarming strategies are crucial to prevent cerebral hyperthermia which can worsen brain damage and exacerbate ischemia–reperfusion injury post circulatory arrest. This is especially significant during open aortic surgery, where there is a risk of thromboembolic events due to air or particulate debris. The rewarming process involves several key considerations. Firstly, after completing the circulatory arrest phase, CBP is reintroduced with initial reperfusion of the brain at the original cold target temperature for about 10 min before starting rewarming. During rewarming, a temperature gradient of no more than 10 °C is maintained between the venous inflow and arterial outlet until the outlet temperature reaches 30 °C. Subsequently, this gradient should not exceed 4 °C. Rewarming should be gradual, limited to <0.5 °C/minute. Throughout, arterial blood outlet temperature should not exceed 36.5 °C to prevent cerebral hyperthermia [54,55].

3.10. Glycemic Control

Hyperglycemia is frequently observed during cardiopulmonary bypass (CPB), attributable to various factors, including exogenous glucose administration, stress response to surgery and CBP, and hypothermia-induced insulin resistance [56,57]. These conditions contribute to peripheral insulin resistance and significant increases in glycemic levels, affecting up to 75% of patients, with a higher incidence among those with pre-existing diabetes mellitus. Research has explored the adverse effects of hyperglycemia during cardiac surgery, including its association with perioperative infection and immunomodulatory effects, potentially impairing white blood cell function and renal function in the long term. Additionally, hyperglycemia has been linked to poorer neurological outcomes post cerebral injury, attributed to factors such as increased lactate production, intracellular acidosis, release of excitotoxic amino acids, and enhanced inflammatory response, particularly relevant during CBP [58,59,60].

3.11. Atheroma Management

The aorta can release harmful embolic material, mainly atheromatous debris during cardiac surgery, potentially causing cerebral complications. Techniques to minimize this include precise cannula placement, specialized cannula designs, and imaging guidance (transesophageal echocardiography and epiaortic scanning to strategically avoid the atheromatous areas during cannulation and clamping). Avoiding aortic manipulation and using advanced cannulas with filtering technologies are also effective [61,62,63]. Further development in managing the atheromatous aorta is expected.

3.12. Carbon Dioxide Field Flooding

A structured review in cardiothoracic surgery examined whether carbon dioxide (CO₂) field flooding in heart valve surgery reduces post-operative neurological complications. Only some studies found significantly fewer intracardiac bubbles in the CO₂ group. Other studies found no neurocognitive benefits. Post-operative cerebrovascular complications were slightly lower in the CO₂ group (1.2% vs. 2.5%). Although CO₂ flooding reduces intracardiac air bubbles and improved survival in some small studies, no sustained reduction in cerebrovascular complications has been demonstrated [64].

3.13. Neuroprotection in Open and Endovascular Aortic Surgery in Thoraco-Abdominal Aortic Aneurysm

Spinal cord injury (SCI) is a serious complication associated with both open and endovascular thoraco-abdominal aneurysm repair with an incidence rate ranging from 2% to 15%. This variance depends on the aneurysm’s extent and etiology, the patient’s comorbidities, the urgency of the procedure, and the experience of the surgeon and the center [65]. The pathogenesis of SCI is multifactorial and not fully elucidated. Two primary theories include insufficient remodeling of the collateral blood supply network and atheroembolism of aortic plaque material into the spinal cord’s segmental arteries [65]. The spinal cord’s blood supply is highly intricate, more so than any other vital organ. It primarily receives blood from one anterior spinal artery and two posterior spinal arteries, which originate from the vertebral arteries and are supplemented by segmental arteries branching from the aorta [66]. These segmental arteries, especially prominent in the thoracic region, further divide into muscular and spinal branches, ultimately contributing to the anterior and posterior spinal arteries through radicular branches. The artery of Adamkiewicz, a major radiculomedullary artery in the thoracolumbar region, was historically considered crucial for spinal cord perfusion, but recent evidence suggests that the collateral network of spinal vessels plays a more vital role in preventing spinal cord injury (SCI) [66]. The pathogenesis of SCI involves the complex interplay of multiple vascular systems. The central arterial system, supplied by sulcal arteries, and the peripheral pial network form a robust collateral network capable of compensating for segmental artery occlusion. This network includes intraspinous and paraspinous vessels, which ensure alternative blood supply routes [66]. During open thoracoabdominal aortic aneurysm (TAAA) repair, aortic cross clamping can induce proximal hypertension, increased central venous pressure (CVP), increased intracranial pressure, and distal hypotension, the latter causing ischemia in the kidneys, intestines, and lower limbs, and reducing spinal cord perfusion pressure. Reperfusion after un-clamping can further damage the spinal cord due to inflammatory responses [67]. Thoracic endovascular aortic repair (TEVAR) avoids the physiological disruptions caused by aortic cross-clamping, but permanently occludes segmental arteries covered by the thoracic stent, potentially impairing spinal cord blood flow over time. Delayed paralysis is more common after TEVAR, attributed to gradual ischemia from stent deployment and the slow cessation of endoleaks. The remodeling of the collateral network following TEVAR plays a crucial role in maintaining spinal cord perfusion. This process involves the enlargement and reorientation of intraspinous and paraspinous vessels to sustain blood flow despite segmental artery exclusion. However, if this compensatory mechanism fails due to thrombus formation or embolism, spinal cord ischemia can ensue, particularly affecting the watershed areas of the spinal cord’s gray matter. Atheroembolism, resulting from dislodged aortic plaque during TEVAR, also contributes to SCI [67].

Preventive and definitive treatment measures for SCI following open aortic surgery and TEVAR are extensively debated. The main strategies focus on increasing mean arterial blood pressure and draining cerebrospinal fluid (CSF) to optimize spinal cord perfusion pressure. The 2022 American Heart Association Guidelines (AHA) [68] recommend CSF drainage for spinal cord protection in both open TAAA repair (thoraco-abdominal aneurysm) and TEVAR for high-risk patients (e.g., those with prior abdominal aortic surgery, extensive endovascular repairs of the descending thoracic aorta, or previous infrarenal aneurysm repair). In open TAAA repair patients, delayed paraplegia constitutes nearly 60% of spinal cord deficits. Despite initially intact neurological exams, patients may develop delayed deficit within the first two weeks postoperatively. The incidence of delayed SCI is approximately 5%, nearly double that of immediate postoperative deficits [68]. These delayed deficits often occur following hemodynamic insults, such as atrial fibrillation, hypovolemia, hemorrhage, or infection, and may respond to aggressive measures to optimize spinal cord perfusion (increase mean arterial pressure >100 mmHg, transfuse to a hemoglobin >10 g/dL, volume resuscitation, cardioversion for tachyarrhythmias, decrease intracranial pressure). CSF drainage rapidly reduces the intrathecal pressure and increases the spinal cord perfusion pressure (calculated as mean arterial pressure minus spinal cord fluid pressure). A significant proportion (57%) of patients with delayed deficits show neurological improvement, with 17% achieving complete resolution [68].

Maintaining adequate spinal cord oxygenation is crucial to prevent ischemia during impaired blood flow. Near-infrared spectroscopy (NIRS) measures tissue oxygenation by analyzing hemoglobin absorption spectra. It is theorized that paraspinous vasculature perfusion and oxygenation reflect the spinal cord tissue status. NIRS offers a noninvasive, real-time monitoring tool for spinal cord oxygenation. Lumbar NIRS values 30% below the baseline correlate with permanent paraplegia; comparative studies with motor-evoked potentials (MEPs) show a strong association between spinal cord function and paraspinous muscle oxygenation [67].

Left subclavian artery (LSA) coverage is required in up to 40% of TEVAR cases for descending thoracic aortic aneurysm. In addition to SCI, stroke can also be a serious complication associated with TEVAR. Addressing modifiable risk factors can improve outcomes. Preventing vertebrobasilar insufficiency, preserving a left internal mammary artery coronary bypass graft, and maintaining left upper extremity dialysis access are crucial considerations [68]. Preoperative LSA revascularization decreases stroke and SCI rates and prevents vertebrobasilar insufficiency and left arm ischemia. It is crucial for patients that depend on LSA flow, such as those with coronary bypass grafts or dialysis access. In 2023, Schachner et al. [69] formulated 18 Delphi consensus questions, based on the result of the first round of the aortic association survey. Of the 18 Delphi consensus questions, 13 achieved a high importance and high consensus:

• CSF drainage should be used in all patients undergoing OPEN types I, II, III, and V TAAA repair and should be considered in patients undergoing type IV repair if additional risk factors for symptomatic spinal cord injury are present (occlusion of 1 or more vascular territories feeding the collateral network): high importance, high consensus.
• CSF drainage should be used in all patients undergoing ENDOVASCULAR types I, II, III, and V TAAA repair and should be considered in patients undergoing type IV repair if additional risk factors for symptomatic spinal cord injury are present (occlusion of 1 or more vascular territories feeding the collateral network): high importance, no consensus.
• In OPEN types I, II, III, and V TAAA repair, despite CSF drainage, at least 1 additional method (MEPs, SEPs, or paravertebral NIRS) should be routinely used (high importance, moderate consensus) to monitor spinal cord perfusion.
• In ENDOVASCULAR types I, II, III, and V TAAA repair, despite CSF drainage, at least 1 additional method (MEPs, SEPs or paravertebral NIRS) should be considered (low importance, moderate consensus) to monitor spinal cord perfusion.
• Cerebral NIRS should be used in all patients undergoing OPEN type I or II TAAA repair (high importance, high consensus).
• Cerebral NIRS should be considered in patients undergoing an OPEN type III or V TAAAA repair (high importance, high consensus).
• Staged OPEN or hybrid repair (TEVAR+open repair of remaining downstream aortic segments) or preoperative minimally invasive segmental artery coil embolization (MISACE protocol) should be considered if feasible (high importance, high consensus).
• Staged ENDOVASCULAR TAAA repair or preinterventional minimally invasive segmental artery coil embolization (minimally invasive segmental artery coil embolization, MISACE protocol) should be considered, if appropriate, to minimize the risk of symptomatic spinal cord injury (high importance, high consensus) [68].
• In ENDOVASCULAR TAAA repair, an “intentional endoleak” (branch that remains initially open) may be a useful option to prevent symptomatic spinal cord injury (high importance, moderate consensus).
• In case of bloody puncture, the placement of the CSF drain should be discontinued, and the operation should be rescheduled (high importance, high consensus).
• In case of bloody puncture, delay of rescheduling the procedure and re-puncturing should be at least 24 h (high importance, high consensus).
• When puncturing for CSF drainage, the initial puncture pressure should be monitored (high importance, high consensus).
• Intraoperatively, the CSF pressure should not exceed 10–15 mmHg. However, initial pressures should be used as reference, and higher values might be accepted if the preoperative CSF pressure was higher (high importance, moderate consensus).
• Postoperatively, in the absence of symptomatic spinal cord injury, the CSF pressure should be kept to preoperative levels but should not exceed 10–15 mmHg. However, the initial pressure should be used as reference, and higher values might be accepted if preoperative CSF was higher (high importance, high consensus).
• If spinal cord injury is suspected intraoperatively, the CSF pressure should be kept below the preoperative CSF pressure (high importance, high consensus).
• In the absence of symptomatic spinal cord injury, the CSF drain can be removed 48–72 h after OPEN TAAAA repair (high importance, high consensus).
• In the absence of symptomatic spinal cord injury, the CSF drain can be removed 24–72 h after ENDOVASCULAR TAAAA repair (high importance, high consensus).
• In case of symptomatic spinal cord injury, CSF drainage should be kept at least 2 days beyond when the diagnosis is established, even if the CSF pressure has already returned to preoperative levels (high importance, high consensus).

4. Discussion

Despite significant advancements, cerebral injury remains a significant concern for cardiac surgical patients, particularly among older and sicker individuals (Figure 1). The risk factors for cardiac-surgery-related adverse neurological outcomes are classified into preoperative (age > 75 years, cerebrovascular disease, peripheral vascular disease, carotid disease, smoke, severe left ventricular dysfunction, diabetes mellitus, renal failure, hypertension and COPD, alcohol abuse, EuroScore II, depression, low level of education, depression, sleep disturbance ad apnea, preoperative cognitive reserve dysfunction), perioperative (longer surgeries, type of surgical procedure, anesthetic dose), and postoperative (atrial fibrillation, myocardial infarction).

In addition to the aforementioned risk factors for neurological events, anatomical variants of the aortic arch must also be considered [70,71,72]. As suggested by Patrick Popielusko et al. in “A systematic metanalysis of variations in branching patterns of the adults aortic arch”, 7 anatomic variants of the aortic arch exist [73]. While most anatomical variations of the aortic arch do not result in physiological consequences, they can elevate the risk of complications during surgical interventions. An illustrative case of this is observed in endovascular graft replacement procedures, where inadequate knowledge of the aortic vasculature can precipitate leakage and ischemic injuries to the brain (and also extremities). Moreover, the type 2 AA (bovine arch) configuration has been documented to complicate and increase the risk of carotid stenting, contingent on the chosen approach [71,72]. This complication is typically due to a sharp curve involving the brachiocephalic trunk and the left common carotid arteries when accessing via femoral approach. Consequently, patients with a bovine arch might benefit from alternative approaches, such as brachial, radial, or innovative trans carotid artery revascularization procedures [71,72,73].

In a study by Swapnil Samadhiya et al. involving 200 ischemic stroke patients, 85.5% of these had the standard arch. Bovine arch has a prevalence of 6% to 31% and 33% in ischemic stroke patients [71]. Bovine arch type A and B are associated with earlier stroke onset and increased risk of embolic stroke. Type A correlates with anterior circulation stroke and type B with bilateral anterior circulation stroke. Abnormal aortic arches are linked to higher regional shear stress, endothelial injury, and thrombus formation, increasing stroke risk. Aortic arch evaluation with contrast-enhanced CT is recommended before endovascular interventions to mitigate stroke risk [71] (Figure 2).

These complications often lead to prolonged hospital stays, increased reliance on rehabilitation programs, and diminished post-operative quality of life. The incidence of cerebral injury varies, ranging from cognitive impairment to stroke, with stroke rates averaging around 2% [74,75] (Figure 3).

Cognitive dysfunction is more common than stroke, with incidence rates ranging from 80 to 90% in the early post-operative period to around 25% at one year after surgery [76]. Post-operative neurological complications significantly amplify mortality rates, particularly major neurological damage, such as stroke or seizures:

• Acute ischemic stroke is a frequent and serious complication after cardiac surgery, posing significant risks of morbidity and mortality. The incidence varies from 0% to 18%, influenced by the type of surgery (1–5% for aortocoronaric bypass, 1.5–17% for valve replacement, 1–10% for aortic surgery, 17% for aortic arch surgery) and patient’s health prior to surgery. Many ischemic strokes detected on imaging are asymptomatic. They typically result from embolization originating from the heart or major arteries, with less common causes being reduced cerebral flow due to arterial hypotension and vessels narrowing. Treatment options include intravenous tissue plasminogen activator (IV-tPA) and mechanical thrombectomy for large vessels occlusion in the brain’s front circulation. IV-tPA is generally avoided after major open-heart surgery. The timing of stroke onset relative to surgery is crucial for treatment decisions, often determined after the anesthesia effect wears off. Mechanical thrombectomy is a suitable option especially within 6 h of symptoms onset, confirmed by imaging showing large vessels occlusion and suitable brain conditions [76].
• Retinal artery occlusion can occur following heart surgery (<0.5% aortocoronary bypass), presenting as a sudden, painless vision loss. Central retinal artery occlusion leads to complete loss of central peripheral vision, often identified by a cherry-red spot on fundoscopy. The mechanism is typically embolic rather than hypoperfusion, with risk factors including aortic insufficiency, diabetic retinopathy, and hypercoagulability. Treatments like intra-arterial alteplase showed no benefit, but methods like ocular massage and intravenous acetazolamide may be attempted to improve blood flow [77].
• Ischemic optic neuropathy, occurring in <0.5% post-cardiac surgery, manifests as painless vision loss with an afferent pupillary defect [78].
• Spinal cord ischemia, reported in 0.5–10% following aortic aneurysm surgery, results from arterial hypotension, surgical injury, or emboli. Symptoms vary by location, commonly causing bilateral lower-limb weakness and sensory loss. MRI with diffusion-weighted imaging aids diagnosis [76].
• Intracerebral hemorrhage (ICH) is a significant concern due to anticoagulation use, necessitating prompt differentiation from ischemic stroke with a cerebral CT scan. ICH management involves stabilizing the patient, cerebral CT scan for diagnosis, and maintaining systolic blood pressure below 140 mmHg [76].
• Seizures post cardiac surgery are rare (<1%), stemming from a multifactorial cause, and may recur, warranting comprehensive evaluation. Cognitive decline risks remain debated, linked to cerebral microembolization and ischemia [76].
• Psychological complications, notably depression and anxiety, affect 14% of cardiac surgery patients. Interventions like cognitive–behavioral therapy mitigate these challenges effectively [76,79].

The exact causes of cerebral injury in cardiac surgery are not fully understood but may include factors such as microembolization, hypoperfusion, inflammation, cerebral edema, blood–brain barrier dysfunction, hyperthermia, and genetic predisposition. While microembolization is a well-studied cause, other factors like global cerebral hypoperfusion during cardiopulmonary bypass and inflammatory effects also contribute. Genetic factors play a role in both susceptibility to and recovery from cerebral injury after cardiac surgery. Variants such as the apolipoprotein genotype (APOE) have been linked to cognitive decline, but their individual impact is debated [80].

Aortic surgery particularly involving the aortic arch, poses a heightened risk of neurological damage due to the necessity of a temporary cessation of circulation to achieve a bloodless surgical field and due to supra-aortic vessels manipulation during the procedure. Various strategies, both non-pharmacological and pharmacological, aim to reduce cerebral injury in cardiac surgery. The non-pharmacological approach focuses on minimizing emboli generation and optimizing perioperative temperature, while the pharmacological approach targets specific pathways involved in the ischemic cascade. However, no pharmacological agents have shown sufficient efficacy for widespread clinical use yet. Neuroprotective strategies during aortic arch surgery include the following: (a) deep hypothermic circulatory arrest (DHCA) with or without adjunctive cerebral perfusion; (b) retrograde cerebral perfusion (RCP); (c) selective anterograde cerebral perfusion (SACP); and (d) neurophysiological intraoperative monitoring [81,82,83,84].

(a)

Deep hypothermic circulatory arrest (DHCA) is a simple method for cerebral protection during aortic arch surgery, involving systemic hypothermia (14.1° to 20°) to minimize neurological risks during short arrest times (<40 min). Longer times may necessitate adjunctive cerebral perfusion.

(b)

Retrograde cerebral perfusion (RCP) infuses hypothermic arterial blood into the brain retrogradely to clear emboli and maintain cerebral hypothermia. It is beneficial but does not eliminate temporary neurological dysfunction.

(c)

Selective anterograde cerebral perfusion (SACP) ensures adequate cerebral blood flow during circulatory arrest using either unilateral or bilateral perfusion techniques. It supports cerebral and systemic protection with moderate hypothermia and has become the preferred method in many centers.

(d)

Neurophysiological intra-operative monitoring techniques such as EEG and peripheral somatosensory-evoked potentials (SEPs) provide valuable information on brain activity and can guide management during surgery. The EEG monitors brain activity during hypothermia showing predictable changes with temperature. The SEPs offer additional brain activity insights. The near-infrared spectroscopy (NIRS) measures cerebral regional saturation (rSO₂), useful in detecting perfusion changes. The Transcranial Doppler (TCD) assesses intracranial blood flow, identifying emboli and aiding perfusion strategy adjustments. Instead, limited evidence supports the efficacy of ice bags around the head for cerebral cooling, with risks of thermal injury. Pharmacological agents, including barbiturates, propofol, mannitol, lidocaine, magnesium, calcium channels blockers, and glucocorticoids, are frequently used for cerebral protection, although their efficacy remains uncertain. Barbiturates and propofol suppress cerebral activity before HCA. Other agents like mannitol, lidocaine, and magnesium lack strong evidence but are used based on theoretical benefits. As regards blood management, an alpha-stat management during hypothermia is preferred in adults, preserving cerebral blood flow autoregulation and minimizing complications. Higher hematocrit levels (25–30%) improve functional outcomes and perfusion pressures. Managing hyperglycemia during CBP is crucial due to its association with poorer neurological outcomes and other complications. A gradual rewarming strategy (<0.5°/min) with controlled temperature gradients prevents cerebral hyperthermia and ischemia–reperfusion injury. Techniques like precise cannula placement and imaging-guided aortic clamp can minimize cerebral embolic risks from aortic atheroma. Finally, spinal cord injury (SCI) is a risk mitigated by strategies like cerebrospinal fluid drainage (CSF) and maintaining spinal cord perfusion pressure. Preventive measures include managing blood pressure and revascularizing the left subclavian artery, and using NIRS for monitoring oxygenation can help to reduce this risk.

Summarizing, there are multiple strategies to prevent and reduce neurological damage in cardiac surgery. It is essential to differentiate between cardiac surgery and surgery of the aortic arch, descending aorta, and thoracoabdominal aorta. Among the strategies that help reduce the incidence of neurological damage in cardiac surgery (including aortic valve, mitral valve and tricuspid valve replacement or repair, aortocoronary bypass, ascending aorta replacement, Bentall, Wheat Tiron–David, Yacoub procedure) are use of NIRS, TCD, use of pharmacological agents (although strong evidence is lacking), echo-guided aortic clamping to prevent emboli dislodgement from atheroma, echo-guided cannula placement, and glycemic control hematocrit control.

Regarding aortic arch surgery, useful strategies to reduce the incidence of neurological damage include DHCA, RCP, SACP (today most centers agree on using SACP under moderate hypothermia to minimize post operative bleeding), EEG, SEPs, NIRS, TCD, use of pharmacological agents, echo-guided aortic clamping to prevent emboli dislodgement from atheroma, echo-guided cannula placement, and glycemic control hematocrit control.

Finally, regarding open or endovascular surgery of the thoracic or thoracoabdominal aorta, the most commonly used methods to prevent neurological damage are: SACP, EEG, SEPs, NIRS, TCD, use of pharmacological agents, echo-guided aortic clamping to prevent emboli dislodgement from atheroma, echo-guided cannula placement, glycemic control hematocrit control, paravertebral NIRS, potential CSF drain placement, and potential left subclavian revascularization [85,86,87]. Brain perfusion has proven to be the most validated neuroprotection method along with neurological monitoring to prevent cerebral damage. Selective brain perfusion techniques such as antegrade cerebral perfusion (ACP) and retrograde cerebral perfusion (RCP) are employed during induced hypothermia, and several studies have compared these perfusion methods to temperature management strategies. Clinical evidence generally support the superiority of ACP over RCP. However, the use of RCP remains controversial due to potential associated risks. Leshonower et al. [87] conducted a randomized controlled trial comparing deep hypothermic circulatory arrest with RCP versus moderate hypothermic circulatory arrest (MHCA) with ACP and found no significant difference in neurological outcomes between the groups. Nevertheless, magnetic resonance imaging (MRI) with diffusion-weighted imaging (DWI) revealed a higher incidence of acute cerebral infarction lesions (100%) 7 days post-cardiac-surgery in the MHCA with ACP group compared to 45% in the DHCA with RCP group. It is important to note that this RCT had a small sample size highlighting the need for further RCT [87]. Moreover, bilateral ACP requires a longer arrest time compared to unilateral ACP, yet both approaches demonstrate comparable rates of mortality, transient neurological dysfunction (TND), and permanent neurological dysfunction (PND) [88].

5. Conclusions

The management of neurological complications in cardiac surgery requires a multifaceted approach involving meticulous surgical techniques, advanced neuroprotective strategies, and comprehensive intraoperative monitoring to optimize patient outcomes and minimize morbidity and mortality. Further research is needed to refine existing strategies and develop novel interventions for neuroprotection in cardiac surgical patients.