Year in Review 2022: Noteworthy Literature in Cardiac Anesthesiology

Abstract

Last year researchers made substantial progress in work relevant to the practice of cardiac anesthesiology. We reviewed 389 articles published in 2022 focused on topics related to clinical practice to identify 16 that will impact the current and future practice of cardiac anesthesiology. We identified 4 broad themes including risk prediction, postoperative outcomes, clinical practice, and technological advances. These articles are representative of the best work in our field in 2022.

Introduction

Research relevant to the practice of cardiac anesthesiology in 2022 was both incremental and potentially transformative for the future. Highlights include testing fundamental practice, such as choice of fluid for resuscitation, and applications of novel technology, such as artificial intelligence and successful xenotransplantation to humans. To identify articles for this review, we completed a structured review using our previously reported methods.¹ Specifically, we used the search terms: “cardiac anesthesiology” (n = 6), “cardiac anesthesia” (n = 11), “cardiopulmonary bypass” (n = 106), “cardiothoracic surgery” (n = 240) filtered on clinical trials, meta-analysis, or randomized controlled trial and the year 2022 in PubMed. We also reviewed clinical trials relevant to cardiac anesthesiology from the most impactful clinical journals to identify additional studies for inclusion in this narrative review (n = 26). This resulted in a total of 389 publications which were screened for their clinical significance based on expert opinion. We then focused on the studies with the greatest potential to impact the current and future practice of cardiac anesthesiology Table 1.

Table 1.

Summary of the Most Noteworthy Work Relevant To Cardiac Anesthesiology in 2022.

Topic	Authors	Conclusions
Prediction	Demirjian et al.2	Predictive model using usual care laboratory values discriminated between those who would develop moderate to severe AKI and require dialysis.
	Devereaux et al.3	High sensitivity troponin values were associated with 30-day mortality after cardiac surgery. They may be useful for providing early identification of high-risk patients. The optimal cutoff for distinguishing high-risk patients was much higher than previous reports.
Outcomes	Fernando et al.13	ECMO was associated with more frequent mental health diagnoses compared to critical illness survivors without ECMO.
	Chen et al.14	Acute brain infarcts were common after aortic arch surgery, but the long-term clinical consequences remain unknown.
Clinical practice	Meersch et al.20	Handovers between anesthesiologists did not impact clinical outcomes in a large prospective randomized trial.
	Kapadia et al.24	Cerebral embolic protection during TAVR did not reduce the incidence of stroke in the first 72 hours after the procedure.
	Finfer et al.29	Use of balanced multielectrolyte solution compared to saline for critically ill adults did not impact clinical outcomes.
	Pesonen et al.32	Use of 4% albumin solution compared to ringer acetate in cardiac surgery patients did not impact the incidence of major morbidity or mortality.
	Zhao et al.37	Exposure to colchicine reduced the risk of post-operative atrial fibrillation after cardiac surgery.
	Schlapboch et al.42	Nitric oxide administered by the cardiopulmonary bypass circuit during pediatric cardiac surgery did not impact clinical outcomes.
	Smith et al.48	Prothrombin concentrate demonstrated slight benefits over plasma, and had a comparable safety profile for correcting coagulopathy during cardiac surgery in a pilot randomized controlled trial.
	Shi et al.53	Higher dosing of tranexamic acid was associated with a reduction in blood transfusion requirements without increasing perioperative risk during cardiac surgery.
Technology	Hu et al.60	Urine oximetry as a novel technique for measuring renal perfusion was responsive to physiological changes in perfusion.
	Noseworthy et al.57	Artificial intelligence screening increased prediction of atrial fibrillation compared to usual care.
	Montgomery et al.58	Xenotransplantation of porcine kidneys into brain dead humans demonstrated that the approach is possible without hyperacute rejection up to 54 hours.
	Griffith et al.59	Xenotransplantation of a porcine heart to a human was successful and provided sufficient cardiac function for 60 days.

Abbreviations: AKI, acute kidney injury; ECMO, extracorporeal membrane oxygenation; TAVR, transcatheter aortic valve replacement.

Risk Prediction

Risk prediction is a critical component of the practice of cardiac anesthesiology by providing a reference point to determine whether the risk to benefit ratio justifies proceeding with a given procedure. Risk prediction is also useful for identifying individuals who may benefit from a higher level of care or monitoring. Two articles in particular contributed to our understanding of risk associated with cardiac surgery and anesthesia, one of which focused on predicting acute kidney injury (AKI) from standard perioperative laboratory testing,² and the other defined the diagnostic cut-off point indicating who is at risk for worse outcomes by postoperative high sensitivity troponin values.³

Predictive Accuracy of a Perioperative Laboratory Test-Based Prediction Model for Moderate to Severe Acute Kidney Injury After Cardiac Surgery²

Unfortunately, AKI is a common complication after cardiac surgery and is associated with an increased risk of mortality.⁴ Diagnosis of AKI relies on detecting changes in urine output and increased plasma creatine levels which occur up to several days after the initial injury.⁵ AKI is typically asymptomatic, so the delay in diagnosis may prevent timely interventions that could limit the extent of the injury. Identifying high risk patients may allow targeting of preventative strategies before the development of irreversible injury.⁶ Researchers have identified many biomarkers for AKI, but none have been implemented into routine clinical practice. The main issues are a lack of sensitivity and specificity of the biomarkers and the expense of the diagnostic test compared to usual care.⁷ A more pragmatic approach to preventing AKI after cardiac surgery would be to leverage existing pre-operative laboratory values that are included in routine care to identify high-risk patients and implement preventative strategies before injury occurs.

Demirjian et al. introduced a new predictive model consisting of laboratory values from a routine metabolic panel to identify patients at risk for AKI prior to clinical diagnosis. The model included measurements of blood urea nitrogen, potassium, bicarbonate, sodium, albumin, pre-operative serum creatinine, and the absolute perioperative serum creatinine change adjusted for the time between surgery and the blood draw. This study included over 58,000 subjects and a validation cohort of greater than 4000 patients. The cohort included subjects undergoing routine cardiac surgery, such as coronary artery bypass grafting (CABG), single valve repair or replacement, and aortic surgery. The new model had excellent predictive value for moderate to severe AKI within 72 hours of surgery (c-index, 0.86, 95% CI 0.84–0.88) and for predicting dialysis within 72 hours of surgery (c-index, 0.88, CI 0.84–0.92). The predictive accuracy was comparable for both predicting moderate to severe AKI and need for dialysis up to 14 days after surgery. The key innovation of this study is that the model variables are derived from standard of care laboratory measurements for all cardiac surgery patients and they are available without additional expense. Thus, this pragmatic approach to AKI prediction may be generalizable and scalable resulting in a potential benefit for a larger patient population compared to other biomarker strategies.

There has been significant research into early identification and prediction of kidney injury. For example, a proprietary test measuring urinary tissue inhibitor of metalloproteinase-2 and insulin-like growth factor binding protein-7 was approved for clinical use by the Food and Drug Administration in 2014,⁸ but the test has not been incorporated into routine clinical practice. Clinical laboratory leaders may be reluctant to incur such additional expenses, so the novel risk prediction model presented by Demirjian and colleagues based on routine laboratory values addressed this problem. Their large study used basic metabolic panel data to provide a modest improvement over prior models to increase its clinical utility. However, an important clinical limitation of all kidney injury predictive strategies is the absence of a targeted intervention that can be employed beyond optimization of hemodynamics and volume status. The development of more effective medical interventions for treating kidney injury is also necessary for predictive models of AKI to make a major impact on clinical outcomes.

High-Sensitivity Troponin I After Cardiac Surgery and 30-Day Mortality³

Myocardial injury is common after cardiac surgery and is associated with increased mortality.⁹ Early detection of myocardial injury with laboratory tests would provide time to intervene prior to further injury. High sensitivity troponins are the global standard for diagnosing myocardial injury,¹⁰ but the appropriate cutoff value after cardiac surgery is unknown because of the troponin release due to the surgery itself and other confounding factors. Devereaux and colleagues addressed this problem by conducting an international multisite prospective cohort study to establish the high sensitivity troponin I levels that are associated with myocardial injury sufficient for increasing the risk of death after cardiac surgery. The study included 13,862 patients undergoing cardiac surgery at 24 hospitals in 12 countries and the primary outcome was death up to 30 days after surgery. The exposure of interest was peak high-sensitivity troponin I levels measured serially after surgery and the hazard ratio was adjusted with a regression model using the 18 variables from the European System for Cardiac Operative Risk Evaluation II (EuroSCORE II) risk predication model for mortality after cardiac surgery.¹¹ The key finding of the study was that for an isolated CABG procedure or aortic valve replacement or repair, the high sensitivity troponin I levels indicating increased risk of mortality was 218 times the upper reference limit. This cutoff is much higher than the current guideline consensus of 70 times the upper reference limit.¹² Similarly, the cutoff was much higher for patients undergoing more complicated cardiac procedures at 499 times the upper reference limit.¹² These findings have the potential to make a major impact on post-operative care after cardiac surgery. With usual care, there is ambiguity regarding when an elevated troponin level is clinically significant after cardiac surgery because there is a high number of false positive tests based on current recommendations.¹² With their recent work, Devereaux and colleagues have established rigorous norms for high sensitivity troponin I levels that will enable the test to provide more meaningful clinical information after cardiac surgery.

Clinical Outcomes

Optimizing clinical outcomes for patients guides every decision made by cardiac anesthesiologists during the perioperative care. Ideally, the cause and consequences of adverse outcomes are well-understood and the risk can be minimized through effective interventions. Although this is not always possible, research is a key first step to identifying and defining problems with current therapeutic strategies and then targeting those risks with effective interventions. Recently, cognitive outcomes have become a major focus of research in cardiac anesthesiology, and the nuances of cognitive injury are being characterized and defined by research teams worldwide. A key finding is that the negative impacts of subclinical neurological injury may be under-recognized and should be targets for future therapeutic and preventative strategies. Two articles this year expanded what is known in this field. One article identified extracorporeal membrane oxygenation (ECMO) as a key risk factor for new mental health diagnoses, such as anxiety disorders, posttraumatic stress disorder, and schizophrenia,¹³ and another revealed that the majority of patients undergoing aortic arch repair have subclinical neurological lesions detectable postoperatively by magnetic resonance imaging (MRI) despite current neuroprotective strategies.¹⁴

Association of ECMO with New Mental Health Diagnoses in Adult Survivors of Critical Illness¹³

Although ECMO is associated with a high mortality, the majority of patients who survive hospitalization survive for several years after the illness.¹⁵ Survivors of critical illness have an increased risk of mental health disorders,¹⁶ but it is unknown if ECMO is associated with specific risks to cognition and mental health relative to critical illness. To address this issue, Fernando and colleagues studied the association between ECMO and new diagnoses of mental health or social problems in patients surviving until hospital discharge. The study included 1054 ECMO-supported patients from a ten-year population-based cohort study across Ontario, Canada. The cases were matched 1:6 against non-ECMO intensive care unit (ICU) survivor controls on the basis of age, institution, and Charleston Comorbidity Index (CCI). The overall survival rate to discharge was 60.9%, which demonstrated the high acuity of this patient population. Baseline demographics and admission criteria were similar across patient groups, with respect to pre-existing mental health disorders, but ECMO patients had an increased rate of chronic obstructive pulmonary disease, markers of more severe illness including increased multiple organ dysfunction scores (MODS), increased need for mechanical ventilation, renal replacement therapy, tracheostomy, overall increased hospital length of stay, and need for discharge to a location other than home without assistance. In addition, the average length of post-discharge follow up for the non-ECMO survivors was approximately twice as long at 1390 days vs 730 days. This likely reflected a difference in the years of data acquisition with relatively increasing ECMO rates from 2017–2020, whereas non-ECMO matched survivors were equally distributed across 2010–2020.

With respect to the primary outcome, 36.8% of ECMO survivors vs 40.9% of non-ECMO survivors were diagnosed with a new mental health condition post-discharge. However, when adjusted for length of follow up, the true incidence rate was higher in the ECMO survivors when compared to non-ECMO patients 22.1% (95% CI 19.5–25.1) vs 14.5% (95% CI 13.8 to 15.2) per 100 person years with a hazard ratio of 1.24 (95% CI 1.01–1.52). With varying follow up intervals between patient groups and increasing ECMO utilization in the latter years of this dataset, there may have been fundamental differences between these groups that were incompletely captured with the current data analysis. Secondary outcome analyses did not identify a difference in future substance misuse or deliberate self-harm. Within ECMO supported patients, regression analysis identified other previous mental health diagnosis HR2 39 (95% CI, 1.78 to 3.20) and previous outpatient psychiatric visits in the past year HR 1.82 (95% CI 1.25 to 2.65) as factors associated with new mental health diagnoses.

With an increasing focus on post-intensive care syndrome (PICS) and the associated mental health burden as the use of ECMO continues to grow, this study serves to identify ECMO survivors as being a high-risk group that warrants further intensive outpatient follow up post-discharge. While an exact causal link between ECMO support and mental health diagnosis is not postulated and may just reflect a marker of severity of illness and post-discharge quality of life, further studies to identify a potential mechanistic link may allow for additional changes in ECMO management and patient selection.

Acute Infarcts on Brain MRI Following Aortic Arch Repair with Circulatory Arrest: Insights From the ACE CardioLink-3 Randomized Trial¹⁴

Approximately 5 to 10% of patients experience clinically apparent strokes after aortic arch surgery despite cerebral protection strategies.¹⁷ Imaging studies demonstrate that the percentage of patients experiencing subclinical neurological injuries may be much higher.¹⁸ Chen and colleagues studied this issue of occult neurological injury after cardiac surgery with a secondary analysis of the randomized controlled Aortic Surgery Cerebral Protection Evaluation (ACE) CardioLink-3 trial to explore the frequency and distribution of new ischemic lesions on MRI after proximal aortic arch surgery.

In the original ACE Cardiolink-3 trial, the safety and efficacy of innominate artery vs axillary artery cannulation was compared for antegrade cerebral protection during elective proximal aortic arch surgery.¹⁹ In this subgroup analysis of 102 patients with 53 patients undergoing innominate artery cannulation and 49 patients undergoing axillary artery cannulation, 71 patients (70%) had new ischemic lesions on diffusion weighted imaging (DWI) with approximately equal proportions within each group. Within these 71 patients, 391 new ischemic lesions were identified with 47% falling within the middle cerebral artery territory and an additional 24% located in the infra-tentorial territory. Of those patients with a new ischemic lesion there were 5.5 ± 4.9 new lesions identified with an even split between the left and right cerebral hemispheres, suggesting that most of these lesions were embolic in nature. Based on a multivariable logistic analysis of numerous clinical and procedural characteristics as well as pre-operative MRI features, only advanced age, lowest nadir nasopharyngeal temperature, and a higher baseline burden of white matter changes were found as potential risk factors for new post-operative ischemic lesions. Axillary vs innominate artery cannulation was not found to affect outcomes.

This study represents the most comprehensive analysis of the frequency and distribution of new acute infarcts after elective aortic arch surgery. MRI is very sensitive for new ischemic lesions based on DWI and apparent diffusion co-efficient (ADC) mapping allowing detection of new ischemic changes consistent with embolism. However, correlation with clinical symptoms was not possible based on study methodology. Routine serial imaging after cardiac surgery may be cost-prohibitive, but further research into peri-operative stroke prevention in high-risk patients may be beneficial. The elderly and patients with pre-existing white matter changes are particularly vulnerable to neuronal injury after aortic arch surgery and further research has the potential to identify novel therapeutic strategies. This study reveals that aortic surgery results in substantial sub-clinical neuronal injury despite modern cerebral protective strategies and there is a need to continue to improve strategies to prevent neuronal injury.

Clinical Practice

The clinical practice of cardiac anesthesiology is an amalgam of empiric knowledge from generations of clinical practice and formal evidence-based medicine derived from systematic studies of existing and novel practices. As cardiac surgery and perioperative care become more complex and patients present with higher-risk co-morbidities, it is imperative to continually re-evaluate clinical practice and formally study current standards. This year 8 articles addressed topics related to current and future clinical practices to assess and improve clinical outcomes after cardiac surgery and anesthesia.

Effect of Intraoperative Handovers of Anesthesia Care on Mortality, Readmission, or Postoperative Complications Among Adults The HandiCAP Randomized Clinical Trial²⁰

Providing safe anesthetic care requires providers who are vigilant and able to make appropriate real-time decisions to render clinical care effectively. Surgical case volume and complexity is increasing as the population ages, and improvements in perioperative care have resulted in a more favorable risk to benefit profile for more patients. The increased number of longer and more complicated surgical procedures presents new challenges to providing perioperative care. Longer duty hours increase fatigue that may compromise vigilance, but previous work demonstrated that intraoperative handovers may be associated with worse outcomes in a retrospective study.²¹ Complete handovers may result in a loss of information,^22,23 which may impact clinical decisions and lead to worse patient outcomes. The HandiCAP trial built on this preliminary data from retrospective studies to test the hypothesis that complete anesthesia care handover results in worse patient outcomes compared to patients who are not exposed to an anesthesia care handover.

The HandiCAP trial was a 12-center randomized trial of 1772 cases. The investigators achieved good separation between the handover and no-handover groups based on randomization and they considered a thorough list of primary and secondary endpoints while blinding anesthesia providers to their participation in the study. The trial found no significant difference in all-cause mortality, readmission to any hospital, or serious postoperative complications within 30 days of surgery between the 2 groups. There was also no significant difference in the incidence of ICU admission or ICU length of stay. Notably, the study excluded off-hour emergency cases where handovers may have a larger impact given the additional clinician stress and fatigue with higher risk cases. The study also had a small number of cardiac surgery patients and included anesthesia providers with various levels of training. The trial was therefore unable to isolate the effect of intraoperative handovers from years in training. The study concluded that intraoperative anesthesia handovers were not associated with differences in patient outcomes within 30 days after a major surgery.

The results may not be generalizable to usual practice without trainees, but handovers may carry less risk than fatigue or attention lapses in providing anesthesia. The potential for loss of critical details during a handoff means there is potential for suboptimal care and patient harm. The upmost emphasis should be placed on being detailed and thorough while transferring care, especially as transfers of care are becoming increasingly inevitable with duty-hour limitations and increasing surgical volume and complexity. Individual vigilance is necessary, but it will not compensate for the systemic factors contributing to excessive fatigue and attention lapses.

Cerebral Embolic Protection During Transcatheter Aortic-Valve Replacement²⁴

Transcatheter techniques for aortic valve replacement avoid surgery requiring a sternotomy and cardiopulmonary bypass which offers more rapid recovery without compromising treatment outcomes.²⁵ Despite these benefits, a subset of patients may develop neurological injury secondary to ischemic lesions. As many as 98% of patients develop ischemic lesions detectable by MRI after transcatheter aortic valve replacement (TAVR),²⁶ with disabling neurological injury developing in about 2% of low-risk patients.²⁷ Cerebral embolic protection strategies employ a mechanical filter during TAVR to prevent debris from entering the cerebral circulation with the goal of reducing the risk of a disabling stroke.²⁸ The approach presumes that the key mechanism for cerebral ischemia associated with TAVR is embolic debris created through dislodgement of calcifications or athero-sclerotic tissue during the procedure. The persistent level of post-TAVR stroke despite clinical implementation of cerebral protection devices suggests that a rigorous clinical trial to determine the risk vs benefits of the approach is necessary.

Kapadia and colleagues addressed this question with a prospective, randomized, international clinical trial of TAVR with and without cerebral embolic protection. The study included 3000 patients and the primary outcome was the incidence of clinically apparent stroke within 72 hours after TAVR. Patients in the control group without cerebral embolic protection experienced a 2.9% rate of stroke compared to a 2.3% rate in patients with cerebral embolic protection. The rate of a disabling stroke was 1.3% in the control group compared with .5% in the cerebral protection group. The differences between the groups were not statistically significant, demonstrating that the impact of cerebral protection during TAVR may be subtle, but the study was underpowered to detect a difference between the groups if one exists. Importantly, the study demonstrated a similar safety profile with and without cerebral protection suggesting that cerebral protection may be useful since the sequalae of stroke are so debilitating for patients. Despite this rigorous study, clinical equipoise remains regarding the utility of cerebral protection. The results do suggest that there is likely another mechanistic explanation for the incidence of stroke, such as thrombus formation associated with TAVR, that supersedes the risk due to embolism during the procedure.

Balanced Multielectrolyte Solution vs Saline in Critically Ill Adults²⁹

Intravenous fluid administration for drug delivery and resuscitation is common and fundamental practice in modern medical care. Several types of intravenous fluid are approved for clinical use, and they vary in their composition and similarity to human plasma. Clinicians administer .9% sodium chloride most frequently, but its use may result in a supraphysiologic chloride concentration that has been associated with worse outcomes after surgery or during critical illness and intensive care.^30,31 Usual practice and experience suggest that both saline and balanced multielectrolyte solutions are safe on a population level, but it is unknown whether the physiological impact of exposure to hyperchloremic intravenous fluid results in differences in clinical outcomes in critically ill patients compared to a balanced solution with an electrolyte composition more similar to plasma.

Finfer and colleagues addressed this question by completing a prospective, randomized, double-blind, multinational trial including 5037 critically ill patients randomized to receive saline or a balanced multielectrolyte solution and assessed death within 90 days as the primary outcome. The groups had a similar mortality rate of 22% in the saline group and 21.8% in the balanced electrolyte group, which was not statistically significant. Patients typically received a median of 6 days of intravenous fluid therapy resulting in an exposure of 3.7 liters in the saline group and 3.9 liters in the balanced electrolyte solution group. Secondary clinical outcomes including the incidence of AKI or need for dialysis did not differ between the groups, but they did observe statistically significant differences in physiological outcomes with chloride levels being higher and arterial pH being lower in the saline group. These differences were not clinically significant given the comparable clinical outcomes. In summary, the results demonstrated that the choice of intravenous fluid therapy between saline and balanced electrolyte solutions did not impact clinical outcomes for critically ill patients only requiring modest amounts of intravenous fluid. Saline administration did result in a quantifiable increase in plasma chloride and decreased arterial pH. These physiological effects may have clinical significance with an exposure greater than the 4 L of intravenous fluid observed in the study groups, but the current study did not include enough patients requiring substantial resuscitation to determine a potential dose effect relationship.

Effect of 4% Albumin Solution vs Ringer Acetate on Major Adverse Events in Patients Undergoing Cardiac Surgery With Cardiopulmonary Bypass: A Randomized Clinical Trial³²

Cardiac surgery with cardiopulmonary bypass (CPB) often requires substantial resuscitation due to the need to prime the bypass circuit and to replace blood loss after the procedure. Crystalloid is the most common and least expensive option for intravenous fluid administration, but preliminary work suggests that colloids such as albumin-containing solutions may have notable benefits. Specifically, colloid remains intravascular longer than crystalloid and may have a beneficial effect on the endothelial glycocalyx and inflammatory cascade in addition to other physiological effects.^33,34 Albumin-based resuscitation for patients with severe sepsis and septic shock was found to provide a survival benefit and be cost-effective.³⁵ However, there is limited high-quality clinical data from prospective randomized controlled trials to guide practice during cardiac surgery despite over 70 years of cardiac surgery with CPB. Pesnonen and colleagues addressed this clinical equipoise to determine the impact of albumin only fluid resuscitation compared to crystalloid only resuscitation on patient outcomes after cardiac surgery.

The Albumin in Cardiac Surgery (ALBICS) trial was a single center, double-blind, randomized trial, with 1386 patients for completion. Patients were adequately randomized into an albumin 4% only group and a Ringer acetate only group, and fluid choice for each group was consistent for both circuit priming and resuscitation. Investigators included a 90 day follow up that was sufficient to capture likely adverse events related to fluid choice. The primary outcome for the study was a composite of major adverse events with varying severity, and the trial found no significant difference in the composite endpoints between the 2 groups. Interestingly, however, the group with albumin only fluid exposure had a significantly lower incidence of myocardial injury (defined by postoperative plasma CK-MB) and fluid requirements at the expense of a significantly increased risk of bleeding, resternotomy, and infection. The albumin only group also had significantly higher chest tube drainage, red blood cell (RBC) transfusion, and platelet transfusion.

The discordance between individual and composite primary outcomes indicates that optimal fluid choice in cardiac surgery patients remains controversial. Furthermore, the individual adverse outcomes suggest that there are risks with the use of albumin in cardiac surgery patients. This is supported by an earlier trial investigating the use of albumin in on-pump cardiac surgical patients, which showed a significant increase in blood products received by the albumin group and impaired thromboelastrography.³⁶ It is still currently unknown whether albumin dose-response effects are present or whether the benefits of albumin may somehow be preserved while reducing the risks. Future studies including multiple study sites and patients with pre-existing unstable coronary syndrome may improve generalizability of the findings and may support the use of albumin in cardiac surgery patients.

A Meta-Analysis of Colchicine in Prevention of Atrial Fibrillation Following Cardiothoracic Surgery or Cardiac Intervention³⁷

Post-operative atrial fibrillation frequently complicates cardiac surgery and predisposes patients to worse outcomes. Persistent atrial fibrillation after surgery may require additional procedures or life-long anticoagulation to prevent stroke.³⁸ Inflammation and oxidative stress after surgery increase the risk of developing atrial fibrillation and therefore represent potential pathways for intervention.³⁹ Colchicine is a potent anti-inflammatory medication that prevents the release of pro-inflammatory cytokines by disrupting tubulin formation and inhibiting macrophages,⁴⁰ which has potential as a rational therapeutic approach based on the pathophysiology of post-operative atrial fibrillation. Preliminary work demonstrated that colchicine may provide benefit after cardiac surgery by modulating inflammatory pathways⁴¹ and this has been tested in prospective randomized controlled trials. This preliminary work has not led to clinical adoption since the results of smaller trials have not been tested in a large multi-center format.

Zhao and colleagues conducted a meta-analysis of 9 randomized controlled trials including 2031 patients to determine if the existing data supports the efficacy of colchicine for the prevention of post-operative atrial fibrillation. There was no apparent heterogeneity between the studies, but there was an unclear risk of bias for 8 of the studies and one study had a high risk of bias due to incomplete non-critical data as determined by the authors. In the primary analysis, colchicine reduced the incidence of atrial fibrillation compared to placebo after cardiac surgery with a relative risk of .62 (95% CI 0.52–.74, P < .001). This effect remained consistent in the subgroup analyses of different surgical procedures and durations of colchicine exposure. Colchicine use increased the risk of adverse events, but they were relatively mild compared to the sequalae of post-operative atrial fibrillation. The most common adverse events were nausea and gastrointestinal symptoms. These results may support the use of colchicine for the prevention of post-operative atrial fibrillation after cardiac surgery because there is some indication that its use is relatively safe and effective. The results of the prior studies and meta-analysis should be confirmed in larger clinical trials.

Effect of Nitric Oxide via Cardiopulmonary Bypass on Ventilator-Free Days in Young Children Undergoing Congenital Heart Disease Surgery: The NITRIC Randomized Clinical Trial⁴²

Nitric oxide is a key signaling molecule that regulates microvascular blood flow and contributes to endothelial cell function. Microvascular blood flow is essential for maintaining cellular gas exchange and supporting metabolic function through the delivery of oxygen and removal of carbon dioxide.⁴³ Cardiac surgery with CPB disrupts endothelial function and results in derangements in physiology resulting in myocardial injury and dysfunction.⁴⁴ Preliminary work demonstrated that repleting nitric oxide through the CPB circuit during cardiac surgery may prevent myocardial injury and dysfunction after cardiac surgery.⁴⁵ Infants undergoing surgery to correct congenital cardiac disease may be especially vulnerable to the stress of CPB and represent a unique patient population without chronic co-morbidities due to their early life stage. Given the promising preliminary work supporting nitric oxide supplementation during cardiac surgery, Schlapbach and colleagues tested the hypothesis that nitric oxide exposure during CPB would result in fewer ventilator-free days after congenital heart disease surgery compared to standard care.⁴²

The study was a multi-center, international prospective randomized controlled trial to evaluate the potential benefit from nitric oxide administered through the CPB circuit. A total of 1371 patients who met inclusion criteria were randomized to routine CPB or CPB with an infusion of 20 parts per million (ppm) nitric oxide. The primary outcome of the study was ventilator free days. Secondary outcomes included low cardiac output syndrome, death, ICU and overall length of stay, and post-operative troponin levels. Post-hoc analysis evaluated organ dysfunction as described by the PELOD-2 score⁴⁶ and AKI. There was no significant effect or trend toward significant effect of the intervention for either the primary or secondary outcomes. Post-hoc analysis also did not show any difference between the two study arms.

This study did not demonstrate a benefit with infused nitric oxide in contrast to other studies that did demonstrate a benefit for adults and children undergoing cardiac surgery.^45,47 Prior studies were significantly smaller and underpowered for major clinical endpoints such mortality, while the NITRIC trial included a higher proportion of patients with low cardiac output syndrome, need for ECMO, and mortality. The nitric oxide dose of 20 ppm may be too low to have a physiologic effect and nitric oxide may be scavenged by plasma free hemoglobin that is produced with hemolysis during CPB. Approximately 12% of the patients in both arms received inhaled nitric oxide during their post-operative course which resulted in some patients in the control group being exposed to the intervention. Despite these methodological issues, the key study conclusion is that nitric oxide administered through the CPB circuit is unlikely to be beneficial.

Prothrombin Complex Concentrate vs Plasma for Post-Cardiopulmonary Bypass Coagulopathy and Bleeding: A Randomized Clinical Trial⁴⁸

Coagulopathy after cardiac surgery results in significant blood loss, which is associated with an increase in blood transfusion requirements and mortality.⁴⁹ Tissue damage from the surgical procedure and exposure to CPB results in activation of the inflammatory cascade and production of reactive oxygen species which damage coagulation proteins.⁵⁰ Damaged coagulation proteins are less effective for initiating the coagulation cascade resulting in increased blood loss after surgery. Plasma proteins are also lost through the wash step of red blood cell salvage systems. Managing the coagulopathy after cardiac surgery typically includes resuscitation with fresh frozen plasma derived from human donors. Transfusion of fresh frozen plasma exposes patients to the risks of transfusion related acute lung injury and other transfusion reactions.⁵¹ Prothrombin concentrate (PCC) is an Food and Drug Administration approved treatment for the urgent reversal of acquired coagulation factor deficiency due to vitamin K antagonists, but preliminary work has also studied PCC as a possible alternative treatment for post-cardiac surgery coagulopathy.⁵²

Smith and colleagues hypothesized that PCC administration in cardiac surgery would reduce transfusion requirements and bleeding without increasing the risk of thrombosis. To test this hypothesis, they performed a prospective randomized clinical trial in patients undergoing complex cardiac surgery, defined as multiple cardiac procedures, repeat sternotomy, and aortic replacement surgery. After protamine administration, patients with laboratory results of an International Normalized Ratio (INR) greater than 1.6/prothrombin time greater than 16.6 sec were randomly assigned to receive either plasma (10–15 mL/kg) or PCC (15 IU/kg) for microvascular post cardiopulmonary bypass bleeding based on direct observation and assessment of the surgical team. A total of 100 patients were included in the analysis.

The primary outcome was chest tube output from ICU admission through midnight of post-operative day one. There was no significant difference in the primary endpoint, with an average of 1022 mL of chest tube output in the plasma group and 937 mL in the PCC group (P = .84). Secondary outcomes were 30-day mortality, ICU and hospital length of stay, re-operation for bleeding within 48 hours, prothrombin time and INR, estimated blood loss, and intraoperative and postoperative transfusions. Though total intraoperative transfusion showed no statistical difference between treatment arms, the group who received PCC received less intraoperative RBC transfusion post intervention compared to the group treated with plasma (13.7% vs 30.6%; P = .04). Additionally, the PCC group had a higher postoperative hemoglobin level, possibly due to the higher cell saver volume, and more robust normalization of the INR. There were no differences in the pre-specified adverse outcomes such as mortality, thrombo-embolic events, AKI, new kidney replacement therapy, acute respiratory distress syndrome, and transfusion-related complications.

Smith and colleagues performed a well-designed prospective randomized trial of the potential differences between plasma and PCC administration in cardiac surgical patients at a single center with a relatively small number of subjects. The standardized bleeding intervention algorithm was well-defined and unambiguous and the authors focused primarily on high-risk surgeries, which may prevent the application of results to lower risk cardiac surgeries. Overall, the findings suggest that PCC administration may offer benefits without increased risk in thrombosis, kidney injury, renal replacement therapy, morbidity, or mortality when compared to plasma administration in cardiac surgery. These finding warrant a larger definitive trial to determine if the results are generalizable to other settings and patient populations. Based on these promising early results, PCC has the potential to become a new standard of care for managing post-CPB coagulopathy by avoiding the complications associated with plasma transfusion.

Effect of High- vs Low-Dose Tranexamic Acid Infusion on Need for Red Blood Cell Transfusion and Adverse Events in Patients Undergoing Cardiac Surgery: The OPTIMAL Randomized Clinical Trial⁵³

Anti-fibrinolytic therapy is one of the most effective preventative strategies for reducing blood loss after cardiac surgery.⁵⁴ Anti-fibrinolytic medications function by inhibiting the enzymatic degradation of hemostatic clots by plasmin. Tranexamic acid (TXA) is a highly effective anti-fibrinolytic medication, but it is also associated with a dose-dependent risk of seizure due to interactions with neuronal glycine receptors.⁵⁵ Early clinical studies demonstrated that TXA has a relatively wide therapeutic window, but clinicians may limit the dose due to concerns for seizure risk.⁵⁶

The authors of the OPTIMAL Study Group sought to further investigate the dose-dependent effects of TXA in cardiac surgery. In this trial, they compared “high-dosed TXA,” defined as a bolus of 30 mg/kg after induction of anesthesia, with a maintenance of 16 mg/kg/h, and a bypass pump prime dose of 2 mg/kg, to “low-dose TXA,” defined as a bolus of 10 mg/kg after induction of anesthesia, with a maintenance of 2 mg/kg/h, and a bypass pump prime dose of 1 mg/kg. This was performed in a multicenter, randomized, double-blind control trial across four academic Chinese hospitals. Patients 18 to 70 years-old undergoing elective cardiac surgery with CPB were included with one year follow-up. Patients were excluded if they had active intravascular coagulation, anti-thrombin III deficiency, breast-feeding or pregnancy, allergy, or contraindication to TXA, history of thrombophilia, previous convulsion or seizure, terminal illness with a life expectancy of less than 3 months, acquired defective color vision, mental or legal disability, or current enrollment in other perioperative interventional studies. The primary efficacy end point was the proportion of patients who received red blood cell (RBC) transfusion in the time between incision and discharge. The primary safety end point was the 30-day rate of postoperative seizure, thrombotic events, kidney dysfunction, and all-cause mortality. In total, 3031 patients were included in the primary efficacy analysis and 2983 patients were included in the primary safety analysis due to 48 patients being lost to follow-up at 30 days.

Regarding the primary efficacy endpoint, 21.8% of patients in the high-dose group and 26% of patients in the low-dose group received a least one RBC transfusion (risk difference of −4.1% relative risk of .84, 1-sided 97.55% CI, −∞ to −1.1%) between the start of their operation and hospital discharge. Regarding the primary safety endpoint, high-dose TXA was demonstrated to be non-inferior to low-dose TXA; 17.6% of patients in the high-dose group and 16.8% of patients in the low-dose group qualified for the primary safety end (point risk difference = .8%; 1-sided 97.55% CI, −∞ to 3.9%; P = .003 for noninferiority). These results expand on previous data supporting TXA use by demonstrating that high dose TXA was more effective at reducing RBC transfusion, without an increase in adverse events.

The OPTIMAL study group performed a well powered, randomized, prospective, multi-site control trial to extend the safety and efficacy of TXA administration in cardiac surgery. The study demonstrated excellent adherence to the assigned protocol, effective follow-up, and included a diverse selection of cardiac surgical procedures. Additionally, they implemented a standardized transfusion algorithm for the primary outcome of RBC transfusion. Weaknesses to the study included a low 30-day mortality, possibly suggestive of low-risk subjects, a low number of urgent or emergent cases, and that the study was performed in only 4 academic centers. In summary, high-dose TXA appears safe and reduces the amount of RBC transfusion after cardiac surgery when compared to low-dose TXA.

Technology

Technological advances have the potential to transform what is possible with clinical medicine. The typical progression involves slow incremental improvements punctuated by a rapid expansion in unexpected directions once innovations are applied to human health. Improvements in computing capacity revolutionized other fields prior to becoming ubiquitous in healthcare documentation and record storage. The availability of large amounts of data coupled with advances in artificial intelligence are just now being applied to diagnostic support as demonstrated by work this year predicting atrial fibrillation from routine electrocardiograms.⁵⁷ Also, the revolutionary gene editing technology called Crispr-Cas9 has enabled an astonishing revolution in the viability of xenotransplantation using organs sourced from genetically modified pigs.^58,59 Using this technology, researchers systematically edited out immunogenic proteins in pigs and removed porcine viruses to enable xenotransplantation. With two reports of successful xenotransplantation, we have entered into a new phase of medicine where increasing the supply of organs for transplantation could now become an engineering challenge. Also this year, a novel device measuring urine oxygenation was able to detect physiological changes such as blood flow and mean arterial pressure,⁶⁰ which may make real-time renal perfusion monitoring in the operating room a possibility.

The Effects of Targeted Changes in Systemic Blood Flow and Mean Arterial Pressure on Urine Oximetry During Cardiopulmonary Bypass⁶⁰

Tissue and organ oxygen delivery is essential for maintaining homeostasis, and the disruption of this process is a common pathway for most adverse outcomes after exposure to surgery and anesthesia. Safe administration of anesthetics requires monitoring patients’ physiological processes such as blood pressure, heart rate, and cardiac output as surrogate measures for adequate organ perfusion. This is adequate for the vast majority of patients and procedures, but patients with significant co-morbidities undergoing cardiac surgery are especially vulnerable to undetected malperfusion. The kidneys are especially sensitive to malperfusion during cardiac surgery because of the low partial pressure of oxygen in the renal medulla even in patients without significant co-morbidities.⁶¹ This is a problem because AKI is a frequent complication of cardiac surgery and increases the risk of chronic kidney disease and increased mortality after surgery.⁶² AKI is typically diagnosed as an increase in plasma creatinine levels compared to baseline and a reduction in urine output.⁵ These clinical signs do not appear until several days after the injury, which does not allow for targeted interventions to prevent or limit the severity of the injury in real-time.

To evaluate urine oximetry as a continuous marker of end-organ perfusion, a single-site, crossover study with 20 volunteer participants was performed recruiting patients greater than 18 years old undergoing cardiac surgery requiring CPB for at least 60 minutes. Patients were randomized in a crossover fashion to receive either high pump flow (3.0 L/min/m²) and pressure (80 mmHg) or normal flow and pressure (2.4 L/min/m² and mean arterial pressure of 65 mmHg) for 20 minutes after cross clamp removal followed by 20 minutes of the alternate.

Those receiving the intervention (high CPB flow and elevated mean arterial pressure goal) had detectable differences in urine oximetry values as quickly as 17 minutes after the hemodynamic changes (22.6 vs 19.9 mmHg PuO2, P < .001) compared to those with standard perfusion goals.

Cardiac surgery patients have significant changes in hemodynamics with associated operative and post-operative complications. Preventing complications from malperfusion in these patients is challenging with limited options for real-time assessment of peripheral end-organ perfusion. This study is an early phase proof-of-principle for using urine oximetry to assess renal oxygen delivery. This technology provides a metric to test interventions to prevent AKI associated with cardiac surgery or other settings.

Artificial Intelligence-Guided Screening for Atrial Fibrillation Using Electrocardiogram During Sinus Rhythm: A Prospective Non-Randomized Interventional Trial⁵⁷

Artificial intelligence (AI) has the potential to revolutionize medicine in unexpected and transformative ways.⁶³ Health data was recently digitized with the widespread availability of electronic health records, but the full potential for discovery has not been realized due to insufficient access to the data and the lack of analytical methods capable of leveraging the information. AI may detect subtle patterns in data which may have applications for supporting diagnosis and management of disease.⁶⁴ Noseworthy and colleagues developed a practical application of AI to identify obscure patterns in routinely collected electrocardiograms of patients in sinus rhythm to predict who is at elevated risk of developing atrial fibrillation.

The study included a cohort of 1003 patients from across 40 states (average age of 74), who had risk factors for stroke but did not have a history of atrial fibrillation The patients had an electrocardiogram performed and underwent continuous ambulatory heart rhythm monitoring for up to 30 days. The AI system was then employed to separate patients into low-risk or high-risk groups. The AI-based risk stratification was successful, with atrial fibrillation being found in six (1.6%) of patients predicted to be low-risk vs 48 (7.6%) of patients thought to be high-risk (OR 4.98, 95% CI 2.11–11.75, P = .0002) In addition, AI-based screening was associated with a higher detection rate of atrial fibrillation in the high risk group compared to standard care over a median follow up of 9.9 months.

The strengths of this trial include its pragmatic design with results likely to generalize well if deployed outside a research setting. The trial also used propensity score matching with eligible but unenrolled controls to reduce selection bias. With follow-up at 9.9 months, this is the most appropriate window for assessing short-term outcomes. However, these results have some notable limitations. The dataset used for an AI-based prediction is relatively small in comparison to published results from other comparable fields. Historically, AI models trained and evaluated on small datasets tend to overfit, learning to recognize quirks of the dataset itself rather than the goal of generalizable insights. The trial was also underpowered for low-risk subjects, limiting its generalizability to this key population who likely stand to benefit the most. Early detection of disease is the most difficult clinically for low-risk subjects due to initially subtle evidence, and AI-based methods often excel in this domain compared to humans. We look forward to seeing the results when applied on a larger sample size. Finally, with a low consent rate of approximately 15% of eligible patients participating, it is challenging to assess if the results are representative of what we would see if this technology were broadly applied.

Medicine at large is undergoing a paradigm shift to leverage AI methods to supplement traditional clinical diagnoses. This trial is one of many highlighting a glimpse of how the future of AI-augmented medicine will present. Overall, the trial shows promising results for using AI-based prediction to improve cardiovascular disease (CVD) prevention but has several notable limitations that will hopefully be abrogated by follow-up work. Evaluation with larger sample sizes adequately powered for low-risk subjects, and higher participation rates will be needed to fully evaluate its effectiveness for widespread use. The pragmatic and scalable nature of the approach makes it a promising strategy for reducing the burden of CVD in the future. Scaling up this technology for use in clinical practice will still be a challenge requiring integrating data infrastructure and ensuring compliance with privacy laws and security.

Xenotransplantation

The definitive treatment of end-stage organ failure is human-to-human transplantation. The first documented successful human transplantation took place in 1906 with transplantation of cornea, followed by kidney, liver, heart, lung and bone marrow in the 1960s.⁶⁵ Further developments in the field increased the safety, reliability, and success rate of transplantation procedures which increased demand for organs. The limited number of donors and suitable organs led to long waitlists because demand far exceeded supply. Surgeons first attempted cross-species transplantation into humans (xenotransplantation) starting in the 19th century with experimental solid organ transplantation from non-human primate in the 1960s.⁶⁶ These experiments were not successful due to organ immune rejection by the patient due to incompatibility. Extensive research into the mechanism of immune rejection identified many of the factors promoting immune rejection, but a feasible way of applying this knowledge clinically remained elusive. A major breakthrough in gene editing technology using the Crispr Cas9 system enabled the production of non-immunogenic pigs, which are a possible source of donor organs.⁶⁷ In 2022, researchers reported the results of the first cases of successful cases of pig to human solid organ transplantation with both kidneys and a heart.

Results of Two Cases of Pig-to-Human Kidney Xenotransplantation⁵⁸

Montgomery and colleagues reported on 2 cases where genetically modified pig kidneys were transplanted into brain-dead human recipients. The study continued for 54 hours and demonstrated that the transplanted kidneys produced urine immediately on reperfusion and resulted in improved creatinine clearance and increased glomerular filtration rates in both subjects. The kidneys remained well-perfused throughout the duration of the study and serial biopsies confirmed that the organs were well-tolerated without evidence of rejection. These promising results support expansion of the technique and further evaluation to determine the efficacy of porcine-derived kidneys for transplant and the risks associated with the procedure.

Genetically Modified Porcine-to-Human Cardiac Xenotransplantation⁵⁹

Griffith and colleagues reported the results of cardiac transplantation with a heart from a genetically modified pig to a patient with nonischemic cardiomyopathy. The patient was not a candidate for standard heart transplantation or durable mechanical circulatory support and was dependent on venoarterial ECMO at the time of transplantation. The xenograft functioned well upon transplantation and ECMO was discontinued on day 4. Porcine cytomegalovirus DNA was detected on day 20 with subsequent failure of the graft requiring resumption of ECMO on day 49. Xenograft biopsy and histology demonstrated damage to the myocardial capillaries, but an absence of typical antibody-mediated or cellular rejection on day 50. On day 56, endomyocardial biopsy revealed antibody-mediated rejection and 40% myocardial necrosis. The cause of graft failure is not fully explained yet, but this proof of principle case demonstrates that a genetically modified porcine heart is capable of functioning in a human without hyperacute rejection. Endogenous porcine viruses appear to be the main barrier to durable cardiac xenotransplantation and further genetic modification of the donor pigs has the potential to overcome this barrier.

Conclusion

Perioperative research in 2022 was still somewhat limited due to the impact of the worldwide COVID-19 pandemic hindering clinical research, but potentially transformative discoveries were still made with direct relevance to the current and future practice of cardiac anesthesiology. Key opportunities for future research include integrating novel technologies such as artificial intelligence to support clinical decision making and diagnosis, while also leveraging the massive of amounts of health data available with the ubiquity of electronic health records. If the long-held aspiration of nearly unlimited availability of suitable organs for transplantation is realized through the practice of xenotransplantation, the main limitation for treatment of end-stage organ failure may be clinicians trained in cardiac surgery, anesthesiology, and critical care.