Skip left side navigation and go to content
heart picture

Neurocognitive Changes Following Cardiac Surgery

April 4-5, 2002
Cloister Mary Woodward Lasker Center,
National Institutes of Health

brain picture

[Main][Abstracts][Agenda (PDF 30 K)
Information on PDF][Planning Committee][Participants]


David J. Madden, Ph.D.

Measurable decline in cognitive function occurs during normal human aging even in the absence of significant disease. Research has been particularly concerned with the identification of age-related change in the component processes of cognitive abilities. Comparisons between younger and older adults' performance on cognitive tasks have consistently revealed age-related deficits in episodic (context-dependent) memory, which appears to reflect older adults' difficulty in generating and maintaining successful encoding and retrieval strategies. An important contributing factor appears to be a generalized age-related slowing of virtually all stages of information processing. Exceptions to age-related cognitive decline, however, are also evident. Even when response speed is measured, the activation and retrieval of semantic (context-independent) information is often comparable for younger and older adults. Some aspects of visual selective attention are also resistant to age-related decline. This mosaic of age-related cognitive change interacts with health status variables in ways that are at present not well defined. Evidence indicates that mild hypertension may lead to subtle cognitive deficits that are more apparent for middle-aged adults than for older adults. To understand the neurocognitive effects of cardiac surgery it will be necessary to take into account the age of the patient, the type of cognitive ability being tested, and health status prior to surgery.

  • References
  • Craik, F. I. M., & Salthouse , T. A. (Eds.), The handbook of aging and cognition (2nd ed.). Mahwah, NJ: Erlbaum.
  • Kausler, D. H. (1991). Experimental psychology, cognition, and human aging (2nd ed.). New York: Springer-Verlag.

Ola Selnes, Ph.D.

Short-term cognitive changes after CABG, primarily involving memory functions, have long been recognized. Until recently, the long-term outcomes after CABG have received less attention. Newman and colleagues reported a high incidence of decline from pre-surgical performance to 5 years, with 42% of their patients performing below their baseline. Similarly, in a cohort of CABG patients followed prospectivel for 5 years, Selnes et al. found significant decline in multiple cognitive domains between 1 and 5 years. These observations raise the possibility that the late decline in cognitive performance may be causally linked to events during the cardiac surgery. We have previously hypothesized that the pathophysiology of cognitive changes after CABG may be multifactorial, including nonspecific effect of surgery, anesthesia, hypopefusion and microemboli. The aditional role of risk factors or cerebrovascular disease in poducing late cognitive decline s currently being investigated.

  • References
  • Selnes OA, Royall RM, Grega MA, Borowicz LM, Jr., Quaskey S, McKhann GM. Cognitive changes 5 years after coronary artery bypass grafting: is there evidence of late decline? Arch Neurol 2001; 58(4):598-604.
  • Selnes OA, Goldsborough MA, Borowicz LM, Enger C, Quaskey SA, McKhann GM. Determinants of cognitive change after coronary artery bypass surgery: a multifactorial problem. Ann Thorac Surg 1999; 67:1669-1676.

Challa VR, M.D.

Permanent mild-to-severe intellectual and/or cognitive dysfunction are found in 16-38% of patients following heart surgery with cardiopulmonary bypass (CPB). Our research efforts for the past 11 years have been directed towards explaining the pathologic basis of the neurologic damage. We believe that evidence incriminates minute fatty microemboli as the major source of brain injury. We have identified millions of these emboli in the brain arterioles at autopsy in 22 patients who died after CPB. These emboli were not found in 38 adult control brains with hypertension or leukoaraiosis or Alzheimer's disease and no history of open-heart surgery. These 10-70 um emboli appear as Small Capillary and Arteriolar Dilatations (SCADs) in alkaline phosphatase-stained celloidin sections. Neuropathologists using routine paraffin-embedded, hematoxylin-eosin stained, 5 um sections can not detect these emboli. More recently, we demonstrated a directly proportional relationship between the time length of CPB and the number of these emboli. They appear to be slowly cleared in about 2 weeks or more. They are fat-soluble, and frequently show birefringence of a particulate nature. Elemental analysis using laser microprobe mass spectrometry (n=8) has shown that there is increased concentration of aluminum (Al) and silicon (Si) in the SCADs and surrounding neuropil. We found 10 times the number of SCADs in brain arterioles (p=<.005) of dogs in which shed blood from the thoracic cavity was returned to the circulation after filtration through a cardiotomy reservoir. It appears that most lipid microemboli to the brain during CPB are due to cardiotomy suction. These emboli contain Al and Si which are toxic to neural parenchyma. We recommend that long-term multicenter studies be undertaken to determine the residual pathologic effects of microemboli in patients who undergo major open heart surgery. (Supported by NIH grants NS20618, NS27500 and 1-PO1-AG05119)

  • References
  • Challa VR, Lovell MA, Moody Dm, Brown WR, Reboussion DM, Markesbery WR. Laser microprove mass spectrometric study of Aluminum and Silicon in brain emboli related to cardiac surgery. J Neuropathol Exp Neurol 1998:57:140-7
  • Brown WR, Moody DM, Challa VR, Stump DA, and Hammon JW: Lon-g-er duration of CPB is associate with greater numbers of cerebral microemboli. Stroke; 2000; 31: 707-713.

Alison Baird, M.D., Ph.D.

Pre- and post-operative brain and vascular imaging has an important role in the understanding of the mechanisms of brain injury after cardiac surgery, as well as in the identification of high-risk patients, prediction of outcome and the monitoring of therapeutic interventions. In earlier studies of brain injury after cardiac surgery conventional magnetic resonance imaging (MRI) and MR angiography (MRA) sequences were used to evaluate the effect of vascular stenoses and prior ischemic lesions on tolerance to the procedure and to study brain edema and post-operative stroke. Other imaging approaches include: functional MRI and positron emission tomography (PET) activation studies; serial measurements of brain volumes; measurement of tissue perfusion, hypoxia, metabolism and novel molecules with PET and single photon emission computed tomography (SPECT); imaging of cerebral metabolites with MR spectroscopy. But to date, no brain imaging method has clearly emerged that could be used as a marker of brain injury post-cardiac surgery.

New functional sequences MRI (including diffusion-weighted imaging [DWI}, MR perfusion imaging and MRA) have given valuable new insights into the understanding of the pathophysiology of ischemic stroke. These sequences are now widely available and provide high-resolution images in a scanning time of around 20 to 30 minutes. On DWI lesions as small as 0.5mm in diameter are detectable. Lesion volume on DWI has been used as an outcome measure in several trials of neuroprotective agents in stroke, although this application is somewhat controversial because of variable correlations of lesion volumes with clinical severity. After vascular interventions such as carotid angioplasty silent DWI lesions have been detected in up to 20% of patients. Recently the first results using these newer functional MRI techniques have been reported in a small series of patients with stroke and encephalopathy after cardiac surgery. It is possible that this methodology may also detect brain injury that leads to cognitive decline after cardiac surgery. But this needs to be systematically studied with pre and post-operative imaging, clearly defined neuropsychological correlations of imaging changes and with controls. The attraction of the MRI methodology is that is widely available, repeatable and fast to obtain, relative to other imaging methods which are unlikely to be practical on a large scale. Imaging also has an important role in the pre-operative imaging of the heart and aorta for potential sources of embolism. The identification of the highest risk patients and subsequent modification of the surgical technique could significantly reduce the incidence of post-operative neurological complications.

  • References
  • Baird AE, Dambrosia J, Janket S-J, Eichbaum Q, Chaves C, Silver B, Barber PA, Darby D, Parsons M, Davis SM, Caplan LR, Edelman RR, Warach S (2001). A three-item scale for the early prediction of stroke recovery. Lancet 357: 2095-2099.
  • Staroselskaya I, Chaves C, Silver B, Linfante I, Edelman RR, Caplan L, Warach S, Baird AE (2001). Relationship between magnetic resonance arterial patency and perfusion-diffusion mismatch in acute ischemic stroke and its potential clinical use. Arch Neurol 58: 1069-74.

William A. Baumgartner, M.D.

Several thousand patients in the United States undergo cardiac surgery. This entails procedures for coronary artery disease, valvular heart disease, congenital heart disease, aortic disease, transplantation and a variety of other combined procedures. In evaluating coronary artery bypass surgery alone, this procedure has demonstrated improved long-term survival compared to medical therapy for patients with significant coronary artery disease. In addition to survival, there is a significant improvement in the quality of life with reduction of angina and recurrent cardiac procedures.
Neurologic injury is a significant risk for patients undergoing cardiac surgery. The spectrum of neurologic injury involves: stroke, (0.8%-5.2%) neurocognitive changes (30-79% at 2 weeks and 24-57% at 6 months), depression, encephalopathy, delirium, and confusion (approximately 10%).

Stroke is predominantly related to emboli from the heart, aortic wall, its major branches to the head, or due to decreased flow secondary to hypoperfusion of potentially compromised areas of the brain. The mechanism for neurocognitive deficits, encephalopathy and decreased level of consciousness however, are probably multifactorial. These can include hypoperfusion, micro-emboli, metabolic derangements due to acid-base imbalance, general anesthesia, and initiation of an inflammatory state. Our laboratory investigations have suggested the role of glutamate excitotoxicity as the mechanism of neurologic injury, especially when the technique of deep hypothermic circulatory arrest is used.

This review will summarize the current status of operative techniques involved in current-day cardiac surgery. These include off-pump coronary artery bypass grafting (OPCAB), influence of a single versus double cross clamp, use of echocardiographic imaging to detect aortic atherosclerosis and air (specific for valve, aortic, and transplant procedures), use of carbon dioxide in certain procedures and a brief description of techniques specific for patients undergoing deep hypothermic circulatory arrest for aortic disease.

Off -Pump Coronary Artery Bypass Grafting
With the introduction of OPCAB surgery, the scientific community has an opportunity to study the influence of cardiopulmonary bypass as it pertains to these neurologic sequelae. Several clinical studies have been reported and published. These studies report a variety of findings, many contradictory with one another. There are no consistent findings favoring one technique over the other.(1-10) To-date, there has not been a randomized study comparing these two operative techniques. Recently a randomized study (OPCAB vs. standard CABG) has been proposed, designed and initiated within the Veterans Administration system to address this question. What has been lacking in most clinical studies that have been previously reported is an adequate control group of patients which would control for the major operative procedure including general anesthesia with or without the use of cardiopulmonary bypass. Potential trade off in the OPCAB operation is a decrease in the number of coronary arteries that are bypassed, unknown long-term patency of bypass grafts, and isolated reports of ascending aortic dissection.(11) The incidence of neurologic injury between the two groups is also conflicting. One recent study from the Cleveland Clinic demonstrated a decrease in the number of bypass grafts performed in the OPCAB group with similar mortality and stroke seen in both groups.(12) There was however a significant reduction in encephalopathy in the OPCAB group compared to standard coronary artery bypass grafting.

Influence of the Aortic Cross Clamp
The techniques involved in aortic cross clamping vary within institutions and from surgeon to surgeon. The use of an aortic cross clamp is necessary for performance of standard coronary artery bypass grafting. A partial occlusion clamp is then applied to the ascending aorta for placement of the proximal vein graft anastomosis. There has been an increasing use of a single aortic cross clamp for both distal and proximal anastomoses. This adds a brief incremental period of ischemia to the operation, but potentially results in less neurologic injury. Like the OPCAB technique, there are few randomized studies, although the concept intuitively seems appropriate. We have looked at the influence of a single clamp technique at Johns Hopkins over the past couple of years. The data is predominantly that of a single surgeon whose operative technique has not changed during the period of study. Unfortunately, like most reported studies, this one was not randomized. Although there was a trend towards less stroke with the single clamp technique (1.1% versus 2.9%) this was not significant. However, if one compared neurologic injury (stroke, encephalopathy, seizure, delirium) there was a significant difference in incidence favoring the single clamp technique (3.2% versus 9.6%). There have been a variety studies reported in the literature favoring the use of a single clamp technique.

Use of Echocardiography
Use of epi-aortic echocardiography has been advocated by Dr. Kouchoukos and associates.(13,14) He has demonstrated the increased sensitivity of this technique over palpation alone. This technique is fairly easy to carry out in the operating room but does require a number of probes and machines if multiple procedures are occurring simultaneously. Use of transesophageal echocardiography has become very routine in the current era of cardiac surgery. Its use can detect aortic atherosclerosis particularly in the aortic arch and descending thoracic aorta. Unfortunately, its ability to detect atherosclerotic plaque in the ascending aorta is limited due to inadequate visualization of the aorta in this area.

Transesophageal echocardiography has been demonstrated to be helpful detecting air following a procedure in which the heart has been opened.(15) This has become a fairly reliable method to "de-air." For the past few years, we have been using a constant insufflation of carbon dioxide into the pericardial well. In our experience this has decreased the incidence of postoperative air as detected by transesophageal echo.

Deep Hypothermic Circulatory Arrest
This particular topic will be addressed in detail by Dr. Jonas in a subsequent session. In regards to operative technique, there are a few points to make in regards to use of deep hypothermic circulatory arrest in adult patients. The first is that our laboratory investigations suggest that the neurocognitive deficits may be the result of glutamate excitoxicity and thereby could potentially be prevented by pharmacologic intervention. To-date there has been no drug compatible with clinical use. Diazoxide a potassium dependent ATP channel opener, is being clinically investigated to determine its efficacy in reducing neurologic injury associated with cardiac surgery.

In a recent report to the American Surgical Association in April of 2001, Dr. Bavaria and his group reported on 104 consecutive patients who underwent repair of acute Type A dissection using hypothermic circulatory arrest.(16) The overall in hospital mortality was 9%. Only 5% developed a new CVA following repair. These outstanding results were attributed to a number of changes made in the operative management of these patients. These included intraoperative transesophageal echocardiography, retrograde cerebral perfusion, hypothermic circulatory arrest established after 5 minutes of EEG silence in neuromonitored patients, after 45 minutes of cooling in non-neuromonitored patients, and use of EEG for monitoring purposes during the operation.

The entire area of neurocognitive changes following cardiac surgery is influenced by the assessment of these outcomes. Experts in this area should develop a concensus in regards to study design, types of tests used and techniques of evaluations. There does appear to be some long-term decline in neurocognitive function following coronary artery bypass surgery. The clinical significance of this decline is unknown. Likewise the causal relationship between the operation and the decline is not validated. However, based upon these studies there is evidence suggesting that cardiopulmonary bypass contributes to this decline. A prospective study controlling for both general anesthesia and cardiopulmonary bypass is needed to better define the etiology of neurologic injury.

It appears that stroke will continue to be a major cause of morbidity and mortality in patients undergoing cardiac surgery, irregardless of technique employed. Further refinement of specific operative techniques such as OPCAB, anastomosis stapling techniques and the use of sophisticated imaging techniques to detect aortic atherosclerosis will be the thrust of investigational research in the future. Appropriate use of future phamacologic agents may also mitigate neurologic injury.

  • References
  • Hernandez F, et al. In-hospital outcomes of off-pump versus on-pump coronary artery bypass procedures; a multi-center experience. Presented at the Society of Thoracic Surgeons, January, 2001.
  • Magee MJ, et al. The elimination of cardiopulmonary bypass improves early survival in multivessel coronary artery bypass patients. Presented at the Society of Thoracic Surgeons, January, 2001.

John M Murkin, M.D.

Perioperative Assessment
Advanced age, peripheral vascular disease, previous stroke or carotid stenosis, and diabetes are important patient risk predictors, while emergency surgery, complex valve/coronary operations, and ascending aortic and/or aortic arch procedures are surgical factors, all of which are associated with increased risk of stroke and central nervous system (CNS) injury. Whether individual genetic susceptibility (eg. allele Apoe4) increases risk of injury remains an intriguing hypothesis. The use of a preoperative risk index score (eg. Perioperative Stroke Risk Index) can be an important tool to identify patients at greatest risk of perioperative stroke, potentially allowing interventional strategies (eg. epiaortic scanning, neurological monitoring, selective brain cooling) to be employed. Much work remains to be done in this area, however.

Direct epiaortic ultrasound scanning (EAS) would increasingly appear to be the 'gold standard' for intraoperative detection of aortic atheromatosis, and if associated with appropriate surgical modifications, is associated with improved CNS outcomes. Alternatively, screening with transesophageal echo (TEE) can reliably assess the descending aorta and detection of descending atherosclerosis should trigger direct EAS. Precisely what to do and how best to manage unexpected ascending aortic atheromatosis remains unclear, and is an area worthy of much more study. Similarly, optimal cannulation sites (eg. axillary vs femoral arteries vs ascending aorta vs distal aortic arch) and type of aortic cannula (eg. intraaortic filter, differential cerebral perfusion, flow diffusor) have not been well studied.

Anesthetic Agents
As a categorical statement, intravenous anesthetics are associated with proportionately decreased cerebral blood flow (CBF) and cerebral metabolic rate (CMR), while inhalational anesthetics increase or do not change CBF while decreasing CMR. While various laboratory studies have demonstrated a cerebroprotective potential for many anesthetic agents (eg, thiopental, propofol, isoflurane), none have reliably demonstrated clinically effective cerebroprotection despite several clinical trials. Overall it does not appear that anesthetic agents appreciably influence the incidence or extent of perioperative CNS injury. Similarly, several studies to date have investigated the cerebroprotective efficacy of calcium channel blockers (eg. nimodipine) with no beneficial CNS effects identified. Meta-analysis suggests that administration of serine protease inactivators is associated with a decrease in perioperative stroke in cardiac surgical patients.

Because cerebral emboli are felt to account for a significant proportion of neurological deficits following CPB, unnecessary elevations of CBF have the potential to directly increase delivery of microgaseous and microparticulate emboli into the cerebral circulation. During CPB principles of optimal management are thus aimed at preservation of appropriate CBF, and maintenance of cerebral flow/metabolism coupling. CO2 is a potent cerebral vasodilator and will directly increase CBF. This response is maintained during hypothermia, profoundly influencing CBF and disrupting cerebral autoregulation. This is the rationale for use of alpha-stat pH-management which avoids cerebral hyperemia and preserves cerebral flow/metabolism coupling rather than pH-stat management during moderate hypothermic CPB. In several prospective randomized trials alpha-stat pH-management has been shown to result in decreases in the incidence of cognitive and neurological dysfunction.

In a prospective study in infants undergoing surgical repair of congenital heart defects using deep hypothermic circulatory arrest (DHCA), use of pH-stat strategy was associated with lower postoperative morbidity, shorter time to first electroencephalographic activity, and numerically fewer seizures (though not statistically significantly less), compared with those infants managed using alpha-stat. It appears then, that in contrast to continuous perfusion during moderate hypothermic or normothermic CPB, reperfusion injury rather than cerebral embolization per se, may be a greater factor in postoperative injury following DHCA. There is no similar data in adults undergoing DHCA. In the presence of cerebral reperfusion (eg. following DHCA), pH-stat may thus be more advantageous especially in non-atherosclerotic patients (eg.infants), potentially because of more homogeneous brain cooling and mitigation of calcium-related reperfusion injury.

Perfusion Pressure
Cerebral hypoperfusion is also detrimental and can independently or synergistically exacerbate brain injury. To this end, preservation of cerebral perfusion pressure (CPP) should focus not only on elevating mean arterial pressure (MAP) but also on recognizing cerebral venous outflow obstruction (eg. atrial cannula-mediated) via monitoring proximal superior vena cava pressure (eg. via introducer port of pulmonary artery catheter). Specific patient factors (eg. hypertension, age, diabetes, carotid stenosis) may predicate a higher CPP and/or hematocrit requiring individualized neuromonitoring (eg. EEG, NIRS).

Hyperthermia, Hyperglycemia and DHCA
Ancillary factors such as rate of rewarming, cerebral hyperthermia, and hyperglycemia can exacerbate CNS injury possibly by increasing release of various excitotoxins (eg. glutamate), as well as producing an exponential increase in CMR and substrate requirements. Monitoring of brain temperatures (eg. jugular bulb and/or nasopharyngeal thermistry, as well as avoidance of CPB aortic inflow temperatures > 37oC, can avoid cerebral hyperthermia and exacerbation of CNS injury. Preliminary evidence suggests that maintaining lower brain temperatures in the postoperative period may also have salutary effects on CNS injury. During DHCA, effective brain cooling and avoidance of premature rewarming (eg. ice packs to head) appear to be the most important interventions. Use of electroencephalographic (EEG) monitoring, jugular bulb thermistry and jugular O2 saturation may all be useful endpoints for optimal brain cooling. Retrograde cerebral perfusion may have moderate additional CNS benefits, but this is debatable and more likely a result of improved brain cooling rather than because of metabolically effective cerebral perfusion. As discussed above, optimal pH-management in adults during DHCA remains unclear.

Hyperglycemia is associated with increased brain injury likely through development of intracellular lactic acidosis and impairment of energy-dependent neuronal ion pumps Surprisingly small amounts of intraoperative glucose, on the order of 100 g (eg. about 2 L D5W) -- particularly if administered during CPB -- have been shown to produce significant hyperglycemia with average peak serum glucose values in the range of 31 mmol/L (438 mg/dL). This effect is multifactorial and reflects acute glucose intolerance in the form of insulin suppression, stress hormone-induced gluconeogensis, and impaired glucose excretion as a consequence of enhanced renal tubular resorption. The result is that without tight intraoperative control of perioperative blood glucose, hyperglycemia can and does occur in virtually all CPB patients receiving glucose-containing solutions. Aggressive blood glucose management is associated with better CNS outcomes specifically, and with improved patient outcomes overall (eg. wound infection, renal failure). Avoidance of glucose in CPB prime and cardioplegic solutions, and frequent blood glucose monitoring and aggressive management of hyperglycemia during the perioperative period are all advocated.

Effective deairing to minimize microgaseous emboli in the presence of open-chamber procedures is currently not clinically achievable, though pre-closure CO2 insufflation into the ventricular chambers and image-directed ventricular aspiration would appear to be potentially effective measures.

  • References

David A. Stump, Ph.D.

The purpose of intraoperative cerebral monitoring is to provide timely information to the surgical team regarding the status of the brain. The definition of Atimely@ is the critical factor that determines the usefulness of the information. Unfortunately, most monitoring modalities are akin to fire alarms which tell you that the fire is in progress and the best that can be achieved is to limit the damage. Current cerebral monitoring tools provide information on a continuum from Adamage is about to happen@ to Athis is the extent of the injury@.

The best Aneuroprotection@ is the prevention of brain injury which begs the question of exactly what we should be monitoring. Should we be looking for indications of acute brain insult or for the etiologic factors that either cause or exacerbate the damage? Brain injury associated with cardiac surgery is secondary to focal or global ischemia due to either hypoperfusion or embolization. The extent of the injury can be affected by the temperature of the tissue when the insult occurs, inflammatory processes, and the type and level of anesthetic and pharmacological interventions before, during and after the insult.
Direct intraoperative cerebral monitoring methodology can be subdivided into three categories based upon the timeliness and usefulness of the information: Immediately clinically relevant, technically relevant and research relevant. If we broaden the question to Amonitoring for neuroprotection@, then a very useful fourth category can be added and includes monitoring for the precursors of injury. Table 1 grossly catalogues current cerebral monitoring methods under these four headings. Clinically relevant information directs therapeutic interventions for the patient on the suegical the table (i.e. low SjVO2). Technically relevant information is feedback that will improve surgical methods that will benefit future patients (i.e. changing clamping method to reduce embolic load). Research tools provide data which will eventually improve overall outcome (i.e. optimizing CBF). Neuroprotection monitoring uses non-cerebral monitoring modalities to either prevent or ameliorate brain injury in the patient on the table (i.e. filter deterioration allowing more emboli to pass). There is considerable overlap in the categories.

Clinically Relevant

Technically Relevant

Research Tool
Embolus Detection Embolus Detection Embolus Detection Arterial line Embolus Detection
Jugular Venous Bulb O2 Sats CBF -Xe133 CBF-Xe133 TEE/Epiaortic* scanning
?TCD TCD TCD Arterial line temperature-rate change and final temp
?Evoked Potentials Processed EEG Processed EEG
Blood product utilization
?Near-Infrared Spectroscopy   Near-Infrared Spectroscopy  
    Retinal Fluorescein Angiography  
    Biochemical markers*
    Genetic Markers*
    New MRI methods*

*extra-operative method

To protect the brain from insult during cardiopulmonary bypass three major areas need to addressed:

1. Aortic manipulation. Macroemboli from the aorta are a major cause of neurologic injury. Reduced aortic manipulation and epiaortic scanning reduce macroemboli.
2. Temperature management. Emboli are delivered via the blood. The proportion of cardiac output delivered to an autoregulated brain is a function of temperature. Therefore the number of emboli delivered to the brain can be reduced by manipulating temperature.
2a. The size of a neurologic foci is related to temperature. The actual brain temperature at the time of the lesion affects the size of the penumbra. Hyperthermia is to be avoided. The rate of temperature change is also relevant. The temperature of the arterial inflow is the best indicator of brain temperature.
3. The return of shed and processed blood has a variety of negative effects. For example, the return of fatty emboli via cardiotomy suction and increased inflammatory processes.

Protecting the brain does not necessarily mean it has to be directly monitored. Monitoring for the conditions which may cause or exacerbate a cerebral lesion may be more neuroprotective if the information is used to maintain a favorable environment for the brain.

  • References
  • Hammon JW, Stump DA, Kon ND, Cordell AR, Hudspeth As, Oacks TE, Brooker RF, Rogers At, Hilbawi R, coker LH, Troost BT. Risk Factors and solutions for the development of neurobehavorial changes after coronary artery bypass granfting. Ann Thorac surgeon 1997; 63: 1613-1618.
  • Stump DA, Brown WR, Moody DM, Rorie KD, Manuel JC, Kon ND, Butterworth JF, Hammon JW. Microemboli and neurologic dysfunction after cardiovascular surgery. Semin Cardiothrorac Vasc Anesth 1999; 3 (1):47-55

L. Henry Edmunds Jr., M.D.

Cognitive defects associated with open cardiac surgery and/or the use of extracorporeal perfusion (including circulatory and respiratory assist applications) may be produced by ischemia, hypoxia, toxins and apoptosis, but a leading cause is the production and circulation of microemboli. Microemboli are defined as emboli less than 500 microns in diameter. The major types of microemboli are gas emboli, foreign material and emboli generated from blood elements. The cardiotomy suction system, retained air within the heart, stopcocks and bubble oxygenators are the most important sources of gas emboli. Inert particles in blood aspirated from the wound, particulate matter from the perfusion circuit and emboli composed of fat and/or denatured protein, which are largely produced in the wound and cardiotomy suction system, are major sources of foreign microemboli. Fibrin; platelet, leukocyte-leukocyte and leukocyte-platelet aggregates; and red cell debris account for most microemboli produced by blood elements.

Prebypass filtration, micropore filters within cardiotomy reservoirs, membrane oxygenators and use of arterial line filters reduce, but do not eliminate microemboli. Pore sizes for cardiotomy filters range from 20-40 microns and are 170 microns in arterial line filters; thus even if 100 percent efficient (which filters are not), countless microemboli enter and block systemic arterioles and capillaries. Pressure gradients pre-empt smaller pore sizes.

Contact with the wound and perfusion circuit activates five plasma protein systems and five different blood cells. As a result, thrombin is generated and fibrin is formed despite large doses of heparin. One consequence of platelet, neutrophil and monocyte activation is the formation of aggregates and microparticles. Hemolysis contributes to red cell debris.

Strategies to reduce the production of microemboli from blood elements include attempts to discover a non thrombogenic material that can be fabricated; tissue engineering; coating or formulating biomaterials to be less thrombogenic; and development of reversible inhibitors of selected blood elements. "Blood anesthesia" is a useful term to describe reversible, selective inhibition of blood elements during cardiac surgery and all applications of extracorporeal perfusion, including the artificial heart and temporary and destination left ventricular assist devices.

  • References
  • Edmunds LH Jr, Stenach N. Blood-Surface Interface. In Cardiopulmonary Bypass ed. By. GP Gravlee, RF Davis, M Kurusz, JR Utley. Lippincott, Williams and Wilkins, Philadelphia, 2000, p149-66.
  • Hessel EA II, Hill AG. Circuitry and Cannulation Techniques. In Cardiopulmonary Bypass ed. By. GP Gravlee, RF Davis, M Kurusz, JR Utley. Lippincott, Williams and Wilkins, Philadelphia, 2000, p69-104.

Richard Jonas, M.D.

This presentation will review the challenges and the advantages of studying the neurological morbidity of cardiopulmonary bypass in a pediatric population.

Problems in Studying Neurological Morbidity of Cardiopulmonary Bypass in Children

The wide variability and heterogeneity of congenital heart disease complicates the assessment in children of the neurological morbidity of cardiopulmonary bypass. There is a wide spectrum of congenital anomalies and within each anomaly a wide spectrum of severity. Many different procedures are employed for correction of these anomalies.

Another important problem is the lack of uniform testing modalities in children who are of different ages both at the time of surgery as well as at the time of assessment. The young age at which surgery is presently undertaken prevents subjects from acting as their own control through pre and postoperative studies.

Recent advances in understanding of genetic associations with congenital heart anomalies have demonstrated frequent association of microdeletions such as 22 q 11 with conotruncal anomalies as well as with developmental delay. Until all such genetic associations with developmental delay are defined studies will need to include a large number of subjects to eliminate outliers with genetically determined developmental delay.

A wide range of equipment is applied for congenital heart surgery as it is only a recent development that oxygenators designed specifically for neonates and young infants have been available. Although total priming volumes have decreased tremendously over the last 10 years there is still a marked disparity between the blood volume of the neonate and total circuit volume. The major changes that have occurred in equipment and techniques over the last 10 years reduce the relevance of previous studies of developmental outcome in pediatric patients.

Advantages of Studying Neurological Morbidity of Cardiopulmonary Bypass in Children

Despite the problems listed above, the absence of atherosclerotic cerebrovascular disease and micro and macroembolization of debris mean that the pediatric patient undergoing cardiopulmonary bypass is a useful subject to improve understanding of non embolic bypass related factors in causation of neurological morbidity of bypass. At Children's Hospital Boston extensive laboratory studies and randomized prospective clinical trials have been undertaken over the last 14 years. These studies strongly suggest that the combination of hypothermia, alkaline pH and hemodilute perfusate as well as marginal flow rates interact in an additive fashion to importantly limit oxygen delivery during bypass. Cognitive studies of high altitude mountaineers who climb without oxygen have previously documented the occurrence of permanent cognitive dysfunction secondary to hypoxia but without obvious manifestations of brain injury.

Deep hypothermic circulatory arrest is an extreme example of flow rate and temperature manipulation that continues to be widely used in pediatric cardiac surgery and is also being used with increasing frequently for adult aortic surgery. Laboratory studies strongly suggest that use of a higher hematocrit and the pH stat strategy increase the safe duration of hypothermic circulatory arrest and should be applied both in adults as well as in children.

Although there are a number of proponents of normothermic cardiopulmonary bypass there are few data available to support this approach. At least one prospective randomized trial in adults demonstrated a worse outcome with normothermic bypass because of a higher stroke incidence


There are many aspects of neuroprotection of the child undergoing cardiopulmonary bypass that need study. Both catastrophic brain injury such as choreoathetosis and stroke as well as more subtle degrees of cognitive decline are of enormous importance when they occur in the young child.

  • References
  • Belling DC, Wypij D, du plessis AJ, Rappaport LA, Riviello J, Jonas RA, Newburger JW. Development and neurologic effects of alpha stat versus pII stat strategies for deep hypothermic cardiopulmonary bypass in infants. J Thorac cardiovasc surg 2001; 121:374-383.
  • Bellinger DC, Wypij D, Kuban KC, Rappaport LA, Hickey PR, Wernovsky G, Jonas RA, Newburger JW. Developmental and neurological status of children at 4 years of age after heart surgery with hypothermic circulatory arrest or low flow. Circulation1999;100:526-532.

Donald S. Prough, M.D.

Because of the high incidence of neurocognitive disorders after cardiac surgery, the integrity of cerebral blood flow and cerebral metabolism during cardiopulmonary bypass is important. During cardiopulmonary bypass, multiple physiologic factors are modified, including pulsatile flow, body temperature, hematocrit, and arterial blood gases. During the past twenty years, investigators have repeatedly studied the cerebral circulation during cardiopulmonary bypass in experimental animals and in humans undergoing cardiac surgery. Techniques used have included Xenon133 clearance, the Kety-Schmidt technique, transcranial Doppler ultrasonography and jugular venous bulb oxyhemoglobin saturation. Accumulated data support the following generalizations: During cardiopulmonary bypass, cerebral blood flow responds as expected to reductions in body temperature and hematocrit. During hypothermic cardiopulmonary bypass, the cerebral metabolic rate for oxygen is reduced as expected.
During hypothermic cardiopulmonary bypass with alpha-stat management of PaCO2, cerebral pressure autoregulation is well maintained. During hypothermic cardiopulmonary bypass with pH-stat management of PaCO2, cerebral pressure autoregulation is impaired, consistent with relative hypercarbia. Frank global cerebral ischemia is rare during cardiac surgery in the absence of severe hypotension or cardiac arrest. Major focal episodes of cerebral ischemia are rare during cardiac surgery. During cardiopulmonary bypass in patients with severe cerebral vascular disease, mean arterial pressure should be maintained at a higher level to limit the likelihood of focal cerebral ischemia. Conditions that produce relative luxury perfusion increase the delivery of experimental microemboli to the cerebral circulation. Rapid rewarming after hypothermic cardiopulmonary bypass is associated with jugular venous desaturation under certain circumstances.

  • References
  • O'Dwyer C, Prough DS, Johnston WE. Determinants of Cerebral Perfusion During Cardiopulmonary Bypass. J Cardiothoracic and Vascular Anesthesia
  • Schell RM, Kern FH, Greeley WJ et al. Cerebral Blood Flow and Metabolism
    During Cardiopulmonary Bypass. Anesth Analg 1993;76:849-865.

Myron D. Ginsberg, M.D.

Hypothermia and rewarming represent major modulators of brain metabolism in the context of cardiac surgery. Issues of particular relevance relate to the rate and ultimate extent of rewarming, whether (inadvertent) cerebral hyperthermia is induced during rewarming, how these factors contribute to the production of neurocognitive or other abnormalities, and how to arrive at optimal management parameters.

In the settings of both cerebral ischemia and traumatic brain injury, the extent of injury is strikingly susceptible to alterations of brain temperature, as demonstrated convincingly in experimental as well as clinical investigations. Hypothermia is neuroprotective while hyperthermia significantly increases the extent of damage. For example, in patients with acute ischemic stroke, a 1oC rise in temperature doubles the odds of poor outcome. Moreover, the susceptibility of the injured brain to temperature-induced exacerbation of damage may persist for even days. Brain temperature, when directly measured (or estimated, for example, by jugular venous bulb temperature), is typically higher than core body temperature; thus, reliance upon core temperature measurements alone may underestimate the extent of cerebral hyperthermia. Experimentally, mild hyperthermia has been shown to exacerbate both functional and structural neurologic injury after deep hypothermic circulatory arrest. In astrocyte culture, mild hyperthermia after deep hypothermia leads to increased extracellular levels of the potentially excitotoxic neurotransmitter glutamate.

Neurocognitive changes after cardiac surgery may involve a number of metabolic mediators. Deserving of consideration are: oxygen-related mechanisms (e.g., primary tissue hypoxia vs. oxygen radical production and "reperfusion injury"), the role of neutrophils in radical-mediated injury, the release of multiple neurotransmitters and -modulators (e.g., glutamate, glycine, GABA, adenosine), altered brain water and ionic homeostasis, blood-brain barrier dysfunction, ischemic depolarizations, and secondary energy failure. Each of these represents a potential target for prophylactic neuroprotection.

  • References
  • Ginsberg MD, Busto R: Combating hyperthermia in acute stroke. A significant clinical concern. Stroke 29: 529-534, 1998.
  • Shum-Tim D, Nagashima M, Shinoka T, Bucerious J, Nollert G, Lidov HG, duPlessis A, Laussen PC, Jonas RA: Postischemic hyperthermia exacerbates neurologic injury after deep hypothermic circulatory arrest. J Thorac Cardiovasc Surg 116: 780-792, 1998.

John Detre, M.D.

Cognitive decline following cardiopulmonary bypass may have various etiologies including focal infarction due to macroemboli, diffuse ischemic or inflammatory injury due to microemboli, or conceivably reperfusion effects. Risk factors for cognitive decline are similar to stroke risk factors, suggesting that cardiovascular status contributes mechanistically to the outcome. The composition of microemboli includes microbubbles from the bypass circuit, lipid from scavenged blood from the surgical cavity, and fibrin-platelet complexes. Several studies have shown that neurocognitive decline is correlated with embolic load as determined by ultrasound monitoring, duration on bypass, and interventions that increase overall cerebral blood flow. But while reducing perfusion pressure or CBF can reduce emboli, it has also been hypothesized that the impact of microemboli may be adversely affected by reduced clearance at low flow. Preliminary perfusion MRI data obtained in our laboratory demonstrate that despite any embolization, CBF is generally increased postoperatively. Some of this increase may be in response to anemia, since current practice limits transfusion and crystalloid in the bypass circuit produces significant hemodilution. Thus, increased numbers of microemboli may continue to be delivered to the brain, and the reduced oxygen carrying capacity of the blood may produce hypoxia in regions at risk. The effects of an abrupt reduction in hemoglobin concentration are also unknown. Postoperative MRI has additionally shown brain swelling and increased water diffusion suggesting vasogenic edema, perhaps with blood brain barrier breakdown, findings that are atypical for acute ischemia. Suggested areas for further exploration include the development of animal models of cognitive impairment following CPB for the purposes of identifying the critical pathophysiological mechanisms involved, monitoring for emboli during the postoperative period, and further elucidation of the effects of profound anemia on oxygen delivery and cognitive function in CPB patients.

  • References
  • Caplan, L.R., Hennerici, M., Impaired clearance of emboli (washout) is an important link between hypoperfusion, embolism, and ischemic stroke. Arch Neurol,1998, 55(Nov): p. 1475-1482.
  • Hindmand, B.J. and Todd, M.M., Improving Neurologic Outcome after Cardiac Surgery. Anesthesiology, 1999 90(May): 1243-7

David J. Pinsky, M.D.

There is increasing recognition that both short and long-term cognitive changes occur following cardiac surgery (1). Although the causes are likely to be multifactorial, certain prototypical features of the cerebral microvasculature may predispose to neuronal injury in the setting of cardiopulmonary bypass procedures. The brain is critically dependent on continual nutritive flow of blood, and the quiescent microvasculature exhibits a number of features which maintain blood fluidity. Procoagulant mediators, such as tissue factor, are sequestered abluminally in the subendothelium, and others such as vonWillebrand factor are sequestered in discrete endothelial intracytoplasmic storage organelles. Critical endothelial transmembrane proteins such as CD39 actively catabolize vasoactive and prothrombotic mediators. There is scant expression of glycoprotein adhesion receptors, ensuring smooth and nonadherent passage of circulating leukocytes. When blood flow is interrupted, even for relatively brief periods, such as may occur with low flow secondary to bypass or macro- or micro-emboli, the microvasculature undergoes a major phenotypic shift, becoming prothrombotic (2) and inflamed (3). Proinflammatory cytokines synthesized locally or within the bypass circuitry can contribute to upregulation of adhesion receptors such as ICAM-1 and E-selectin in the cerebral microvasculature, leading to autoamplification of inflammatory cascades, leukocyte accumulation, microvascular obstruction, and neuronal injury. Upregulation of inflammatory cascades secondary to low flow may be potentiated by the brisk cytokine response to cardiopulmonary bypass (4). This is dominated by release of TNF and interleukins -6, -8, and 10, the latter of which may exert an anti-inflammatory function. Complement activation, which is a well documented consequence of cardiopulmonary bypass, may be particularly injurious to neurons, particularly those subjected to low flow conditions (5). Rapid accretion of intravascular, platelet-rich thrombi may occur in situ secondary to endothelial retraction and exposure of blood to procoagulant subendothelial matrix, loss of critical anticoagulant molecules, induced expression of procoagulant mediators, and suppression of endogenous fibrinolytic cascades. The shift in the vascular phenotype at locations downstream from the site of primary vascular occlusion likely represents a critical feature contributing to ischemic cerebral injury. Restoration of blood flow may not only restore nutritive flow to the region, but bring in activated leukocytes, elicit complement activation, and promote the formation of reactive oxygen intermediates, which can be directly injurious or which can trigger de novo transcription of inflammatory mediators. Because the vascular phenotype is modulated in predictable ways following cessation of intraluminal flow and the onset of low flow represents a discrete, anticipated event in the setting of cardiac surgery, this opens up significant possibilities for preoperative therapeutic intervention.

  • References
  • Newman MF, JL Kirchner, B Phillips-Bute, V Gaver, H Grocott, RH Jones, DB Mark, JG Reves, & JA Blumenthal: Longitudinal assessment of neurocognitive function after coronary artery bypass surgery. N. Engl. J. Med. 344: 395-402, 2001.
  • Connolly ES Jr, CJ Winfree, TA Springer, Y Naka, H Liao, SD Yan, DM Stern, RA Solomon, J-C Gutierrez-Ramos, & DJ Pinsky: Cerebral protection in homozygous null ICAM-1 mice after middle cerebral artery occlusion: role of neutrophil adhesion in the pathogenesis of stroke. J. Clin. Invest. 97:209-216, 1996.

Giora Z Feuerstein, M.D., Ms.C., F.A.H.A.

Thromboembolic events leading to acute and chronic cerebrovascular consequences are well documented after cardiac interventions. In addition to acute brain damage (strokes), long-term cognitive deficits may lead to dementia with chronic disability at substantial cost burden to the individual and the health system. Strategies for prevention and treatments of cerebrovascular events associated with cardiac interventions have not been formulated as yet. In this review, three layers of prophylactic and therapeutic options will be proposed. I. Effective anti-thrombotic pharmacology; II. Acute treatment of cerebrovascular events peri-operative; III. Long term therapeutic and reconstructive strategies to prevent and arrest vascular dementia and neurodegenerative processes.

The first pharmacological strategy, anti-thrombotic agents, must include combination of anti-coagulants, ant platelet agents and possibly fibrinolytics agents. In this respect, effective anti-coagulation with UFH/LMWH , anti-platelet agents (ASA/clopidogrel) and tPA are currently available. New anti-thrombotic agents such as selective thrombin (IIa) or FXa inhibitors could become available within 3-5 years. In addition, the role of GPIIb/IIIa antagonists as adjunct anti-thrombotic therapy during vascular interventions supports the possibility that such agents (ReoPro=abciximab, integrelin=eptifibatide or aggrastat-tirofiban) may be useful agents to enhance the limited efficacy of ASA/clopidogrel anti-platelet combination.

The second pharmacological strategy, protecting the brain tissue under distress due to acute ischemia, may include molecular targets that are rapidly presenting toxic effects, which in the context of neuronal distress (hypoxia/hypoglycemia) may induce death signals. Three categories of targets carry the potential to be rapidly tried as neuroprotective agents in the context of cardiac interventions:  1. Excitotoxic neurotransmitters (glutamate; glycine) modulators; 2. Ion channels (Ca+2, Na+) antagonists, and 3. oxygen radicals scavengers (ROR might be associated with hypoxia and reoxygenation of ischemic brain tissue). These molecular targets present immediate opportunities in the form of pharmacological agents that have been introduced into advanced clinical development (e.g., NMDA or glycine antagonists; ion channels blockers) for treatment of ischemic brain injury. While such agents have failed in clinical trials for acute ischemic stroke when treatment commenced > 6 hrs post ictus, the elective nature of most of the cardiac interventions may allow for treatment to start at the peri-operative time frame where the respective antagonists can be delivered in ''real time' relative to the cerebrovascular event.

The third pharmacological strategy that needs to be considered in combating cerebrovascular consequences to thromboembolic events associated with cardiac intervention are aimed to mitigate secondary pathophysiological consequences of brain injury as well as enhancing regenerative and reconstructive capacities. Inflammation (1) is a major reaction that presents itself shortly after brain ischemia or trauma. Ant-inflammatory agents might therefore be of use to diminish secondary damage. In this respect, highly specific and inhibitors of TNT, such as TACE inhibitors need to be tried. Regenerative agents such as growth factors that enhance neuronal sprouting, or factors that may induce tolerance to I jury could well be applicable.

In summary, there are multitude opportunities for pharmacological interventions aimed at preventing or circumventing cerebrovascular consequences to thromboembolic events peri-cardiac procedures. Several pharmacological agents are already available for clinical trials as monotherapy or combination therapy.

  • References
  • Wang X., Feuerstein GZ. Role of immune and inflammatory mediators in CNS Injury. Drug News and Perspectives. 13:133-146,2000.
  • King SB, Montalescot G., reducing the Risk of Vascular Disease: Anti-Platelet Therapy in the New Millennium. Clin Cardiol 23 (Suppl VI) VI-1-VI-2, 2000.

Skip footer links and go to content
Twitter iconTwitterExternal link Disclaimer         Facebook iconFacebookimage of external link icon         YouTube iconYouTubeimage of external link icon         Google+ iconGoogle+image of external link icon