We Care, Therefore We Are: Anesthesia-related Morbidity and Mortality

EMERY Andrew Rovenstine, M.D. (Professor, Department of Anesthesiology, New York University, New York; 1895–1960), was a giant in the founding of our specialty, and I feel greatly honored to have been selected to deliver this Rovenstine Memorial Lecture. The closest I came to Dr. Rovenstine was the privilege of occupying his former office at the old Bellevue Hospital in 1974. The textbooks he left behind, with underlined passages and notes written in the margins, were fascinating, and I have made them available through the Wood Library Museum of Anesthesiology.The message I want to bring to you today is the message that I have brought to this society since way back when I became the American Society of Anesthesiologists' Vice President of Scientific Affairs: To make the future of our specialty shine, we need to grab hold of the questions about anesthesiology that our patients and recent research have brought to the forefront. Whether you are a clinician, a clinician researcher, a nonclinician researcher, or a potential patient (as we all are), we need to show how much we care by supporting efforts to answer questions that address our patients' fears. When their fears become our research focus, public advocacy will convert our poor showing in National Institutes of Health (NIH) research funding into a scientific powerhouse.1Responses like “no time,”“no money,”“medical students with previous research experience choose other specialties,”“department chairs don't have sufficient research background,” or “we've solved all of our major problems” are not acceptable, and they are, in fact, not true. For example, take “we've solved all of our major problems.” What about mortality? We have made substantial progress in the sense that we are anesthetizing sicker patients who undergo more complex procedures without an increase in mortality, but mortality from anesthesia has not decreased significantly since the 1960s.2What about “medical students who do research don't go into anesthesiology”? Our residents have done research, and like residents who match into other highly desired specialties—such as dermatology, plastic and reconstructive surgery, orthopedics, and radiology—they use their qualifications to gain admission to a specialty that receives little NIH funding.1What about the qualifications of anesthesia chairs? I am confident that our chairpersons compare favorably to chairs in specialties that receive the lion's share of funding.Most of all, we need to reject the rejoinder that anesthesiology receives little research funding because it is a specialty without a disease. As my friend Jeff Balser, M.D., Ph.D. (Professor, Department of Anesthesiology, Vanderbilt University, Nashville, Tennessee), says, “Every disease in every patient who is going to be anesthetized is ‘our' disease.” In addition to diseases that are known to affect anesthetic management—such as coronary artery disease, chronic lung disease, and renal failure—there may be many that we are yet to discover. What about conditions that are exacerbated by anesthetics? These may include fetal apoptosis (brain cell suicide), Alzheimer disease (AD), Parkinson disease, Huntington disease, and asymptomatic cognitive dysfunction that becomes symptomatic as postoperative cognitive dysfunction (POCD). These are the diseases whose connection to anesthesia can connect us to grant proposals that—if properly formulated—NIH cannot refuse.So today I want to talk about some of the morbidities that we need to investigate most urgently. We will review characteristics of the young brain, discuss anesthetic effects on the young brain—effects that are being delineated today—and talk about characteristics of the old brain and anesthetic effects on the older brain. In doing so, we will consider the controversy over which is worse—deep anesthesia or light anesthesia—and talk about factors that probably aggravate cognitive dysfunction. Then we will consider the apparent lack of difference between regional versus general anesthesia, some of the confounding factors in that discussion, and some potential alleviating procedures, such as the use of continuous peripheral nerve catheters instead of deep sedation after regional anesthesia.We will also look at the neurodegenerative diseases. What happens when patients with AD have anesthesia and surgery? Do anesthetics hasten the progression of cognitive dysfunction? In light of that risk, I will review some of the agents and techniques that have shown protective potential in laboratories. Preconditioning is one such technique. Could we deliver a nondamaging stressor 24–48 h before surgery—a stressor that would generate a protective cascade, protective against an ischemic cascade? And what about neurogenesis? Can we stimulate the production of new brain cells that would repair damage not prevented by preconditioning? These prospects have left the arena of science fiction, but their clinical utility remains unknown.Last but not least, I want to add a lagniappe —“a little something extra” (for those of you who are not from Louisiana or Mississippi)—and tell you about work being done at my institution that involves memory, targeted prevention and erasure of memory, and the potential that research might have for treating phantom pain, preventing memory of intraoperative awareness, and treating posttraumatic distress syndrome. Then I will conclude with a vision of how we can make ourselves and our specialty more profoundly appreciated by the press, the public, our patients, and funding agencies!The developing nervous system is highly susceptible to neurotoxic insults during rapid synaptogenesis, during the brain growth spurt, before neurons have migrated to their final destination and fully differentiated.3In human newborns, the brain has a full complement of 40–50 billion neurons.4However, in utero , for every neuron that survives, 1–2 neurons undergo apoptosis and die.5The signal for initiating those cell suicides is normally a lack of synaptic feedback due to failure to form synaptic connections, such that rates of cell death average approximately 8,000 neurons per second during the last 11 weeks of gestation. This suggests that neurons have a ready capacity to trigger cell death quickly in response to a lack of stimulation, or in response to a neurotoxic insult. With that response potential in mind, let's look at some of the evidence about adverse effects of anesthetics in children and newborns.6In 1953, James Eckenhoff, M.D. (Professor, Department of Anesthesiology, University of Pennsylvania, Philadelphia, Pennsylvania; 1915–1996), looked at personality changes in children after anesthesia and surgery.7They developed more temper tantrums, more phobias, and more bed-wetting. Those changes were most frequent (38%) in the youngest children (younger than 4 yr) and least frequent (8%) in the oldest children (older than 8 yr). Unsatisfactory anesthetic induction was a prominent risk factor. So way back in 1953, there was an indication that anesthesia might have an adverse effect on young children.One of the first animal models to test the effect of anesthesia on fetuses was developed while I was at New York University. Jack Chalon, M.D. (Professor, Department of Anesthesiology, New York University, New York), exposed pregnant mice to halothane and found that their offspring performed significantly more slowly than those of control mice.8Unfortunately, early reports indicating a potential problem did not receive the attention they deserved. That changed in 2003 with the publication of Jevtovic-Todorovic et al. 's “Early Exposure to Common Anesthetic Agents Causes Widespread Neurodegeneration in the Developing Rat Brain and Persistent Learning Deficits”—a title that says it all.9Several studies have augmented those findings,10–13including a recent investigation by Fredriksson et al. 14They injected mice with propofol, ketamine, thiopental, propofol with thiopental, ketamine with thiopental, or high-dose propofol during the critical stage of mouse development (10 days old) and found increases in brain cell death and significant reductions in functional performance. Now the laboratory evidence is sufficient to heighten concern about anesthetizing pregnant women during their third trimester, about anesthesia in our neonatal and pediatric patients, and about sedation in the neonatal intensive care unit. What are the proposed mechanisms for this neurotoxicity in laboratory animals? Do anesthetics trigger the same mechanisms in human fetuses and neonates?15,16And if they do, can we prevent or reduce that effect?Olney et al. have proposed that anesthetic drug effects on fetal and neonatal γ-aminobutyric acid and N -methyl-d-aspartic acid receptors cause translocation of a Bcl-2–associated protein to mitochondrial membranes, leading to an apoptotic cascade.15If we can interfere with the apoptotic cascade in abnormally inhibited neurons, we might be able to prevent anesthetic-induced neuronal apoptosis. Some ways to do that have been found in laboratory animals. Using the early postnatal rat model in which Jevtovic-Todorovic et al. found that anesthesia generates neural apoptosis and long-term learning deficits, Yon et al. 17found that melatonin reduced that damage in the most vulnerable brain regions: “Melatonin-induced neuroprotection was mediated, at least in part, via inhibition of the mitochondria-dependent apoptotic pathway since melatonin caused an up-regulation of the antiapoptotic protein, bcl-XL, reduction in anesthesia-induced cytochrome C release into the cytoplasm and a decrease in anesthesia-induced activation of caspase-3.” An analogous effect of melatonin has been found in fetal sheep.18Augmentation of another endogenously generated substance with neuroprotective potential, erythropoietin,19,20has also shown promise against N -methyl-d-aspartic acid receptor antagonist neurotoxicity in rat21and mouse22neonates and in hypoxic-ischemic injured neonatal rats.23Perhaps the problem can even be alleviated by anesthetic choice in pregnant females. Ma et al. 24have presented evidence that “xenon mitigates isoflurane-induced neuronal apoptosis in the developing rodent brain.” What about hypothermia? A human trial looking at whole body hypothermia (72 h at 33.5°C) in neonates with hypoxic ischemia encephalopathy found a decrease in death or moderate to severe disability from 62% down to 44%.25,26So what can we say about anesthesia and the young brain? What recommendations can we take home? We have been warned about this potential problem since 1953, and those warnings have been supported by substantial laboratory evidence published over the past I that and we that human fetuses and do not anesthetic we or anesthesia, in and early postnatal and ketamine because they to be the most we need to to look for ways to reduce fetal and anesthetic in the in We have made progress in the potential problem and but I this is that it our specialty with an and to older brain is but like the young brain, it is Some the older brain as over Some it as over I it as over and in another I to have the to it as over I to for in which it will be over let's look at the older brain has cognitive to also has anesthetic are many potential mechanisms that might in cognitive and anesthetic may not work as We have some that can are also that were when we were but as we And there is with reduced of such as and do we look for after anesthesia and surgery? We to in memory or not weeks after anesthesia, with a of weeks to A of is if we have evidence from a of first major of cognitive dysfunction after anesthesia was published in by M.D. patients older than who general in of patients to be or that with recent The that cognitive was to anesthetic agents and that on be to and that not be a The major to to first of patients older than the of cognitive dysfunction after was than in and after for with of patients of to at a postoperative of anesthesia, a second postoperative and were risk factors for early a of significantly reduced due to a to at and use before significant risk factors for and were not significant early or risk consider the most recent and of a of patients after the of also found a substantial between and death of risk factors for the at and a of without with many more may cognitive as previous may reduction of cognitive this most recent did not of anesthesia to be a risk factor. of patients at the was even with patients, than in the of with patients, about in coronary artery of us have or say something like he not the he as not as The New York brought attention to this problem with an the Can the the of and for memory on a patient who back to work and found that he with his patient who not that he performed for many That a of the stage for an published a in the New of by et al. found in of coronary artery patients at and in of patients weeks That down to at after surgery, but came back to at after surgery—a of early by that was by at factors that cause in cognitive capacity patients also affect some of those risk such as of to may be by damage done to patients by to ischemia during and the of the versus make a A found a of more than patients found a of in patients exposed to with of as a significant risk patients were not for a between and not be that patients may be at risk for factors aggravate the development of in the caused by is an early We about the up-regulation of and this in can affect the anesthetic cascade of the anesthetic γ-aminobutyric acid and N -methyl-d-aspartic acid receptors and increases production of and we that even in patients, can cause cognitive if there is of are another factor. For example, et al. shown the of and protein in to cognitive caused by after anesthetics et al. that memory is for weeks after h of and in What about In a more recent et al. that exposed to for 4 h more to a test and made before their first with control over the In a of they found that the same but reduced the of whose a that is in and may be to of acid normally to an by from and that deep is with more a recent by et al. at anesthesia versus anesthesia and found that of anesthesia were with cognitive weeks that As et al. have anesthetic not Using as a of we can that anesthetics affect brain to a or most showing a reduction in some showing in and a in which increases during if synaptic feedback is to preventing the apoptotic cascade, neurons of that do not receive sufficient from neurons to generate that are to trigger might be at risk than neurons that are in a of and the of to anesthetized neurons be in anesthetized this brain a neuron whose are by anesthetized neurons is in it may The fetal neuron and the neuron may undergo apoptosis for the same in the of the fetal that is because it has not made functional connections, for the many of anesthetized have become Whether a neuron in because it to make or because it is in a from which most of to have left neuron in an anesthetized brain the effect may be the is at and there is much to As by et al. the need for studies to the of to this is also to the lack of difference in or the of the difference in between patients who receive general anesthesia and patients who receive regional anesthesia with sedation with regional anesthesia generates of neurons by anesthetized neurons, a that is analogous to the effect of general et al. that sedation to anesthesia to a in and et al. the found that postoperative sedation with continuous peripheral nerve catheters was with a in the of postoperative there are to a between and and there is evidence that patients with the experience postoperative at more than the of patients without about the effects of anesthetics on the neurodegenerative from of anesthetics can those and reduce the of to protein of protein to into and if those are the can to that are to into neurons and to be to to the by Alzheimer in the early are to have AD by the of the of them will need to be and many of them will have been anesthetized before they became of anesthetics in the has been in by a of work from et al. , supported in by a mouse addition to the et al. human cell to add anesthesia-induced apoptosis as a to and Jevtovic-Todorovic and that old rat are or more to anesthetic neurotoxicity than young rat the rodent and cell to from studies are of et al. the risk of developing AD of during anesthesia versus the risk of developing AD of coronary the general for of and of procedures, the patients developed AD at the of coronary patients patients more factors than coronary patients, but in evidence supporting mechanisms for a between anesthesia and it would be to anesthesia as an to the of et al. et al. a of AD of the previous to general anesthesia, which greatly their to evidence an effect of anesthesia to AD et al. found significant effects in the of a between AD and general anesthesia on of to anesthesia, to or more of general anesthesia and to or more of general anesthesia et al. performed a of AD In their the were Parkinson disease patients and patients with other but a between general anesthesia and other diseases has been that lack of a difference in anesthetic between AD patients and patients with Parkinson disease or other diseases not a lack of a effect of anesthetic a connection between anesthetics and AD has more attention than a between anesthesia and Parkinson or Huntington disease, that research is et al. found evidence that an risk of Parkinson disease as with and et al. found in laboratory evidence that may Huntington there is evidence that anesthetics are a problem for older patients, but before we make recommendations more than from that on be to need to more about to which older patients are most at today is to stimulate our research to more questions and do more research more might we reduce the risk of apoptosis in the young and those with or in older there evidence that some anesthetics are more than et al. at the American Society of indicating that propofol anesthesia is of the memory effects with other general anesthetic agents in this it that general anesthesia-induced memory may be a of the than the anesthetic In a rat with has been shown to prevent the in caused by during the protective potential of is being in and are laboratory has the effects of during ischemia in We found that neuron death in the is reduced in that have of was also after ischemia in that at patients, et al. that by a during and 4 in the of reduced days after at a of patients, et al. found reduced in patients who from to at days from to at weeks and from to at about melatonin and in the A review by et al. the effects of melatonin in models of AD is but the has looked for evidence that or with little to because there are other for the to take we can to need to many AD patients who are the evidence from patients not be without indication for fetuses and the are to and you might make you to clinical of patients with For example, ischemic during ischemia by the that be before and evidence of due to This was in a model of induction of of repair and the that for them are of has much of the from such that in the rat brain with memory laboratory as a potential the in of et al. neuroprotection by anesthetics is not that in anesthesia may to the induction of a of anesthesia the same protective mechanisms as a of how much anesthesia can we before what would have been a protective effect becomes a of are being In addition to laboratory promise for with the the most human trial of a published to et al. their to ischemic patients whose 8 h of the of et al. that of for days to to increases of in the nervous reduced of the of reduced and hold in the trial that is there is to that would be even more as a we effects by initiating 24–48 h before surgery, deliver it during surgery, and in the intensive care has the of effect in the of ischemic and in fetuses more of such as have been developed and are showing as old that is for the young was shown to be in in also to be for the that effects of and anesthesia on the as as the can be by that the in the that in the brain a response that in the production of new of growth factors can this response we may be able to to of also et al. In addition to ischemic and to be such has shown protective promise against N receptor antagonist neurotoxicity in rat21and in neonatal against neurotoxicity in an in AD deep brain has also been shown to in has shown promise in a human after brain all of you have that I have been about that have been in the since the and of anesthetics on the in the and not because when I The are but our of them has a over the past of that and the research it has my is that we will have anesthetic and to their effects on the young and the and that and of of repair and will deliver brain to the the and in the us are to before the and at my institution in an in model to to the that an protein in may be and sufficient for the of long-term also et al. That the of to or prevent memory and even to or for such as preventing of intraoperative and treating posttraumatic phantom pain, and even has been this has a to a long-term memory in a of will take Could NIH say to such even to another at my institution brain from AD patients and found of in of brain in memory was not found in of the memory and was not found in of the of to have effect on long-term memory through of acid synaptic Do we need to the effects of anesthetics on in but not least in this of research, I that in anesthesiology Dr. in last Scientific American the In addition to the presented in an about the that the receptors for the effects of anesthesia, it may become to patients at risk for intraoperative because they lack those drug to prevent or at least also be a little more something I would like to by you a in I was my office while students for an to our I one of them do The answer came to and them to the on a by a in at of them and the over with a down to the Then he through a and you he can grab you to a he a of and in a from a Then he you back with and and many of you through the message that I have been for a are who patients while do that would can show the public how much we care about what we do by our specialty through research that our patients' fears. When we show them how much we they will who we