Numerous clinical situations remain as problematic issues which continue to vex neuroanesthesiologists and neurointensivists as they work to optimize outcomes for their patients. Nascent research ongoing at this time may provide a window to a future approach to these problems. In this chapter we review many of these new research areas and speculate as to how they may eventually translate to clinical care of neurosurgical patients in the operating room and neurointensive care unit. Areas to be reviewed include genomics, stem cells, neuroprotection, ICP management, technology, and pharmacology.
Genomics
Each cell has to produce proteins in order to function. The specific protein structure is determined by the sequence of base pairs on the organism’s DNA combined with post-translational modifications. One estimate is that there are some 35,000–40,000 genes with about 6,000 proteins active at any given time. However, actual protein composition varies according to the needs (and regulation) of the moment such that as many as 100,000 proteins are thought to be transcribed by a cell. Thus, the actual genetic structure of an individual has an important role in the development of and response to disease and the response to therapy.
The field of genomics is significantly complex. Notably the word “gene” has a multitude of definitions, reflecting the wide variety of approaches to the field. The field includes work with single nucleotide polymorphisms (SNPs), transcript mRNA studies, siRNA epigenetics and copy number variants. SNPs are naturally occurring variation in the specific DNA nucleotides that code for specific proteins resulting in naturally occurring variations in proteins with associated variations in function. After a gene has been transcribed into mRNA then further regulation with potential for variation arises such that there can be heritable changes in gene expression or translation without a change in DNA sequence, so-called epigenetics work. This concept is more recently and specifically described in siRNA (short interfering RNA) epigenetics studies wherein a transcribed mRNA is bound by a complementary siRNA, thus interfering (or regulating) the translation of the DNA nucleotide sequence into a protein. These observations have spawned an area of genetics research examining the possible uses of these interfering RNAs as future research or therapeutic tools.
To date, SNPs are the variant type of choice for association studies in common diseases and complex traits. This has resulted in many case control and association studies that have provided valuable information on age-related macular degeneration, diabetes, obesity, cardiovascular diseases, prostate cancer, and breast cancer. However, SNPs are not the only source of polymorphism of the human genome. Another abundant source of polymorphism, the so-called copy number variants, is one that involves deletions, insertions, duplications and complex rearrangements of genomic regions. Such polymorphisms can result in deletion of specific genes, as occurs with rhesus blood type, or with increased numbers of copies of genes, as has been demonstrated with alpha hemoglobin. Indeed, up to 10% of the variation in the human genome has been attributed to copy number variation (also called copy number polymorphism). Notably, the number of combinations and permutations of these causes of variation seem endless. Thus, biostatistical issues with sample size, repeated measures, sample independence, and association versus causation become essential considerations in experimental design and the ability to draw conclusions from the data. One practical result of these constraints with SNP and other studies is that the most robust conclusions can only be derived from variants which have a high enough frequency to reasonably allow for inferences to be drawn.
These and other current advances in causes of genomic variation foretell a day when an individual patient’s genome will be part of his/her history and physical examination. This information will be used to allow the anesthesiologist to optimize or individualize specific anesthetic effects, such as hyperemia or neuroexcitation in specific brain areas, define likely anesthetic tolerance and thus appropriate dosing, foretell susceptibility to ischemic and other types of brain damage, and adapt the application of general medical therapies to a person’s specific genomic signature. Moreover, this information will undoubtedly lead to efforts to alter an individual’s genotype, genomic regulation, or phenotype either with respect to the diseases anesthesiologists see in the operating room (OR) or with respect to specific responses to anesthetic interventions. A series of annual reviews have underscored these issues.
Specific Anesthetic Effects of Genetic Factors
Blood Flow and Metabolism
Probably the first report to detail a neurophysiologic effect of an anesthetic drug based on a single nucleotide polymorphism was published by Kofke et al. They determined the cerebral blood flow (CBF) response to increasing doses of remifentanil in human volunteers, determining an element of limbic system activation. However, subjects with an ApoE4 genotype had a different pattern of limbic system response from those who did not possess this SNP. If similar analogous effects of anesthetics are found with other SNPs we can expect an SNP-based selection of anesthetics based on the likelihood of producing some side effects such as cerebral hyperemia, ictal activation or suppression, and variable neuroprotection.
In a non-neurologic context early studies in rodents indicate that changes in chromosome composition affect cardiovascular responses to propofol. Countless other physiologic responses to anesthetics will also undoubtedly be determined with each patient’s genome studied for anticipated interactions between genome and reaction to an anesthetic or other perioperative drug.
MAC, Analgesia, and Other Anesthetic Side Effects
Redheads are known to require more anesthesia than the rest of the population. As such this represents probably the first observation of a heritable condition’s effect on MAC, although the specific SNP or set of SNPs or other causes of genetic variation producing red hair and linking this to anesthesia are not known. Early studies in nematodes have isolated specific genes which impact on sensitivity to anesthetics. , Moreover, specific genetic alterations in mice are reported to affect anesthetic sensitivity to disparate anesthetics pentobarbital, ketamine, and nitrous oxide. A clinical correlate of these observations can be found in Mullholland et al.’s report of the effect of various SNPs on several of the EEG responses to desflurane anesthesia in humans, though one might suggest that this is not that important as future management paradigms will most likely entail individual titration. Nonetheless this type of information will likely also relate to pain tolerance, anticipated needs for postoperative pain control regimens and, possibly, susceptibility to addiction. Such studies in humans have been published describing specific SNPs affecting pain tolerance and need for postoperative analgesia. In addition, neuropathic pain and more focused pain therapy is suggested in some preclinical studies evaluating effects of siRNA treatment on pain. Moreover, SNPs are now being described in humans regarding genetic predisposition to postoperative nausea and vomiting, , malignant hyperthermia, and cognitive dysfunction.
Pharmacokinetics
Genetic influences on pharmacokinetics have become reasonably well known for a handful of antihypertensive drugs as they relate to genetic effects on metabolism (e.g., hydralazine) and perioperative drugs. Certainly, genetic factors are known for such entities as pseudocholinesterase deficiency, the interaction of thiopental with porphyria, or the genetics of malignant hyperthermia. , Cytochrome P450 is important in the metabolism of many drugs including anesthetics. These genomic variations in cytochrome P450 have been reported to impact on the metabolism of midazolam and opioids, , with clinical relevance made manifest in a report of genomic differences with cytochrome contributing to mortality from non-medical fentanyl ingestion. Multiple other effects of genetic variation on anesthetic metabolism and side effects have been reviewed. It is clear that increased information will become available with respect to SNPs, arrays of SNPs, and other causes of genetic variation and their impact on anesthetic metabolism.
Ischemic Tolerance
One group reports that up to 10% or more of genes undergo alteration in expression after brain ischemia. , In humans, alterations in the expression of genes responsible for the inflammatory response have similarly been reported after stroke. , Many others report increased risk of stroke associated with different genotypes. Thus, the effect of one’s genomic inheritance on one’s predisposition to stroke and susceptibility to its sequelae may be an important piece of information. However, the preponderance of such research in humans deals with genomic contribution to risk of stroke and not with direct genomic contribution to tolerance of, or vulnerability to, brain ischemia, factors which, arguably, are of greater interest to the neuroanesthesiologist.
These risk-oriented studies are nonetheless important areas of research. However, they are confounded by the natural heterogeneity of clinical stroke and provide no information on the possible genomic contributors in humans to congenitally determined ischemic tolerance or vulnerability. Three lines of genetic research have introduced the notion of gene-based ischemic tolerance which may be a therapeutic target. In the first, altering the genetic makeup of animals subsequent to cerebral ischemia can alter the animal’s ischemic tolerance. In the second line of work, researchers have demonstrated that environmental factors introduced in advance of a severe ischemic insult can induce genes to produce proteins that provide tolerance to subsequent ischemia. The prototypical paradigm is one of ischemia or other insult inducing upregulation of heat shock proteins, which are important contributors to subsequent tolerance to a greater ischemic stress. Other genes have also been suggested as contributing to ischemic preconditioning. , Indeed, some authors are now suggesting induction of ischemic tolerance as the basis for recovery in an ischemic penumbra. And, thirdly, Kofke et al. have observed genomic factors apparently contributing to a greater release of biomarkers of brain damage during cardiac surgery.
Cerebral ischemia clearly has a significant effect on transcription and translation of important genes, perhaps related to organismal survival and evolution. , , , The phenomenon of ischemic tolerance has been known for many years. Although incompletely understood, a basis in altered genetic regulation seems likely , , with the most attractive corollary that such information will lead to gene-based therapeutics. , Gidday et al. observed that the extent of cerebral infarction in a neonatal rat model was substantially attenuated if animals were pretreated with small doses of hypoxia first. This work was further developed and supported with similar observations in other ischemic models. Recent work has suggested convincingly that elaboration of heat shock proteins has an important role in ischemic tolerance. Other candidate contributing genes have also been suggested. HIF-1 (hypoxia inducible factor) is one such alternate inducible protective protein. Also described in this context are erythropoietin, glial cell-line-derived neurotrophic factor, and TGF beta-1. One report provides convincing evidence that a significant component of HIF is, in fact, induced expression of erythropoietin.
The preischemic stress does not necessarily have to be an hypoxic/ischemic stress, as fever and acidosis (among others) also induce subsequent ischemic tolerance mediated by induction of a protective protein. Similarly, ischemic tolerance can be pharmacologically induced by pretreatment with estrogen (Bcl-2), cobalt chloride (HIF-1), desferrioxamine (HIF-1), or isoflurane. In contrast, ischemia, despite the induction of a plethora of putative protective genes, also induces the BRCA-1 associated protein BARD-1 that mediates apoptosis, which is thought to be deleterious. Genetic manipulation of mice has also been used and also demonstrates how genetic factors can alter endogenous vulnerability to an ischemic insult. , , , Bernaudin et al. identified several genes that appear to be regulated by hypoxia, leading them to suggest that these could have a role in a genetic predisposition to ischemic tolerance. Genetic makeup and gene expression clearly contribute to one’s vulnerability to brain ischemia.
Nonetheless the notion of ischemic preconditioning is an increasingly attractive one. Recent research underscores the complexity of this phenomenon with numerous reports on only some of the factors thought to play a role. One important practical consideration in translating these findings to clinical management will be defining whether a given preconditioning therapy has, indeed, introduced increased ischemic tolerance. This consideration underlies work which is endeavoring to identify biomarkers, indicating, before ischemia, that tolerance has been enhanced.
Hundreds of nucleotide polymorphisms are under active investigation of their contribution to stroke and other cardiovascular diseases. Such studies in humans generally deal with the role of a gene in risk of stroke. Many genes have been suggested, based on animal work such as that described above, to have important roles in determination of tolerance to ischemia. However, of the apparent myriad of potentially important genes only a relative handful of genes have been studied. Nonetheless, one can surmise, based on knowledge about the pathophysiology of cerebral ischemia, that evaluation of pathogenetically grouped genes would be a rational approach to screening for SNPs and other causes of genetic variation with potential therapeutic value. From such an approach one could develop “ischemic axes” with corresponding variant structure and regulation of genes, gene number, transcription, and translation, as depicted in Fig. 29.1 . This knowledge could be used to develop new neuroprotective therapies or, if known for an individual patient, be used to tailor optimal neuroprotective therapy. Finding such clinically relevant effects for a given patient’s or patient group’s genetic variables will then lead to more intense scrutiny of the protein eventually translated that is transcribed by a given gene. Such information will then lead to rational selection of genes for possible gene delivery or perhaps development of drugs which mimic the effect of favorable proteins or inhibit the effects of unfavorable proteins. Examples of the validity of this approach are now appearing.
Several years of work from many investigators has shown the relevance of ApoE polymorphisms to neurologic outcomes. Thus, it should be clear that unlike most gene association studies used to predict a given patient’s risk based on his/her genomic profile, knowledge of SNP associations with ischemic tolerance will likely result in new genome-based gene, proteomic, or other categories of neuroprotective therapies. This is simply one example of potential translational outcomes from such genomic association studies. In the case of ApoE, work has followed on the neuroprotective potential of so-called apo-mimetic peptides. Preclinical reports indicate neuroprotective efficacy in hypoxia-ischemia, subarachnoid hemorrhage, peripheral nerve injury, traumatic brain injury, intracerebral hemorrhage, autoimmune encephalitis, and spinal cord injury. It thus seems likely that identified genomic contributors to tolerance or vulnerability to acute brain insults will lead to gene-based therapies and also associated proteomic-based therapies.
Risk of Comorbidities
Gene association studies with stroke are just one example of the role of genomics in defining a given medical risk. Such studies are burgeoning regarding a host of medical conditions. There have been some studies showing statistical associations between specific SNPs and postoperative complications. Specific examples include renal failure, , post-transplant kidney infection, total hip arthroplasty failure, , allograft dysfunction, pain intensity and analgesic requirements, vasopressor requirements, myocardial infarction, inflammatory response, thromboembolism, , stroke, and vascular graft patency. , These studies will undoubtedly contribute to risk assessment in individual patients and lead to new therapeutic strategies designed to minimize perioperative complications.
Application of Therapies to a Specific Genomic Signature
At the time of writing, characterization of the entire genomes of many humans has been accomplished. This, combined with the abovementioned material, indicates a future that will encompass patients arriving for surgery with their entire genomes on record. One expectation is personal genome sequencing in many healthy people who later present for neurosurgery. This will likely also include other personal “omics” profiling. Applied across a population, including neurosurgical patients, this suggests characterization of the genome, epigenome, transcriptome, proteome, cytokine-ome, metabolome, auto antibody-ome, and microbiome (gut, urine, nose, tongue, skin) for possibly billions of individuals. With this we will have a large amount of impressively big data (e.g., a million genes and other -omic information in millions to billions of people). One hoped for consequence of this will be an ability to predict and monitor diseases, resulting in personalized therapies identified from big data analyses that will increase tolerance to various brain insults in individual patients. In addition, -omic profiling may allow us to tailor anesthetics to individuals or at least predict the consequences of anesthesia and surgery. For example, metabolomic profiling using proton magnetic resonance spectroscopy in children under anesthesia has demonstrated that sevoflurane anesthesia results in higher brain lactate and glucose levels in children compared with propofol. Analysis showed a positive correlation between lactate and glucose levels with agitation and delirium.
Stem cells
Stem cells seem certain to be an important future therapeutic tool. , In fact, recent work has shown that transplantation of mesenchymal stem cells into patients following ischemic stroke not only is safe, but also may improve survival and functional outcome. In addition, recent work has shown that injection of mesenchymal stem cells into the intervertebral disc in patients with discogenic back pain may improve pain and function for up to 2 years post-procedure. Larger trials to investigate the utility of this therapy are currently underway.
This will have important implications for neuroanesthesiologists who may encounter patients who are having neural stem cells placed as part of an operative procedure or who have had them placed earlier with a variable extent of post-implantation differentiation. Given the recent data suggesting specific neurotoxicity of some anesthetics in developing brain, it becomes reasonable to suggest that anesthetics used in the context of stem cell placement could significantly affect the success of the graft. Moreover, the physiologic milieu of the implantation may also be important. One well-known example of this is the effect of hyperoxia on immature retinal cells. Blood pressure, pCO2, temperature, and other aspects of anesthetic administration may have a similarly important impact. Future research will be needed to resolve these issues.
Stem cells
Stem cells seem certain to be an important future therapeutic tool. , In fact, recent work has shown that transplantation of mesenchymal stem cells into patients following ischemic stroke not only is safe, but also may improve survival and functional outcome. In addition, recent work has shown that injection of mesenchymal stem cells into the intervertebral disc in patients with discogenic back pain may improve pain and function for up to 2 years post-procedure. Larger trials to investigate the utility of this therapy are currently underway.
This will have important implications for neuroanesthesiologists who may encounter patients who are having neural stem cells placed as part of an operative procedure or who have had them placed earlier with a variable extent of post-implantation differentiation. Given the recent data suggesting specific neurotoxicity of some anesthetics in developing brain, it becomes reasonable to suggest that anesthetics used in the context of stem cell placement could significantly affect the success of the graft. Moreover, the physiologic milieu of the implantation may also be important. One well-known example of this is the effect of hyperoxia on immature retinal cells. Blood pressure, pCO2, temperature, and other aspects of anesthetic administration may have a similarly important impact. Future research will be needed to resolve these issues.
Neuroprotection
Neuroprotection has long been and should continue to be a central focus of neuroanesthesia and neurocritical care. At the time of writing, notwithstanding ample promising preclinical work, there are very few neuroprotectant strategies that have sustained scrutiny in clinical trials. There is significant laboratory work which forms much of the basis of neuroanesthesia practice. Much of this is based on the notion that a physiologic or anesthetic decision needs to be made and in the absence of satisfactory human data, the preclinical laboratory data are used to justify a decision. This current situation should eventually be rectified. Some encouraging approaches are currently being evaluated. These probably represent only a small fraction of what will eventually become a part of an evidence-based approach to neuroprotection during neurosurgical procedures.
Hypothermia
Definition of Subsets Suitable for Hypothermia
Hypothermia presently has evidence-based support for its use after global brain ischemia. , Clifton et al.’s multi-institutional study in TBI indicates that a TBI patient arriving hypothermic should not be rewarmed, although a protective effect of de novo induction of hypothermia could not be demonstrated. The authors postulated that earlier induction of hypothermia may improve outcome; however, a second trial did not demonstrate any benefit. Meanwhile, others have shown that induction of long-term (5 days) vs. short-term (2 days) mild hypothermia may be associated with decreased incidence of rebound intracranial hypertension on rewarming and improved functional outcome. The IHAST study in over 1000 patients undergoing aneurysm surgery showed no efficacy from the global practice of inducing moderate hypothermia in all patients undergoing cerebral aneurysm surgery. It still leaves open the question of the potential efficacy for a patient with active severe intraoperative focal temporary brain ischemia, without dealing statistically with all of the patients who did not need the hypothermia. The overwhelming evidence of efficacy in animal studies and humans with deep hypothermic circulatory arrest suggest a need to better focus the clinical studies to better identify the clinical subsets in which it will be found efficacious.
New Methods to Induce Hypothermia
As with most neuroprotective therapies, the efficiency and efficacy of induction of hypothermia seems an important factor. Thus some research is ongoing that suggests important advances in the means by which hypothermia will be induced and maintained.
Ice Crystals
Cold fluids have long been used to decrease temperature. This may have problems with the volume of fluid needed to affect the desired endpoints. Reports on using an ice slurry given intravenously may result in an improved method to induce hypothermia quickly, safely, and in a controlled manner. Work is needed to ensure that the infused ice crystals have no sharp edges and thus pose no risk potential for endothelial trauma.
Heat Exchange Techniques
Currently new devices available include surface cooling devices and intravascular cooling devices. Future such devices will certainly have servo control systems that will enable the physicians to program or prescribe the desired central temperature and the device with its software will ensure maintenance of an unchanging level of hypothermia. A current conundrum in temperature reduction therapy is shivering, which can produce adverse systemic effects and prevent achievement of the normo- or hypothermic therapeutic endpoints. , It is likely that better understanding of this process will yield a reliable and safe pharmacologic approach to prevent shivering.
Hibernation
Hibernating animals routinely tolerate physiologic insults that are known to be deleterious in nonhibernating animals. These animals are able to induce profound decreases in temperature, adjust immune function, and have enhanced antioxidant defenses. They are capable, in a regulated manner, to decrease metabolic rate to as low as 2% of normal. This has led to the notion that whatever neuroprotective mechanisms are operant in hibernating animals may have potential for translation to neuroprotection in humans. A variety of neurally controlled biochemical processes that underlie hibernation have been reported. Neural regulatory mechanisms include hibernation protein complex, and a variety of neurotransmitter systems, hydrogen sulfide , , and neural circuits. , Molecular mechanisms of the organismic tolerance of this condition include a switch to lipid-based biofuel, inactivation of pyruvate dehydrogenase, altered structure and concentration of thyroid hormones, altered neuronal morphology, delta opioids , and altered pH regulation. Moreover, since the heart ordinarily develops lethal arrhythmias at hibernating temperatures a variety of gene-regulated processes have been postulated to allow continued cardiac contraction. Evaluation of the genomes of hibernators is ongoing with the hope that identification of similarities with the human genome will lead to therapies that may activate these otherwise latent genes to better allow humans to tolerate ischemia or hypoxia without neural or other organ injury. These notions form the basis for ongoing research to develop suspended animation methods, with overt intent to copy hibernation biochemistry and physiology , , , translated to attenuate problems in human neuroprotection.
Infrared Light Lasers
Several reports suggest that infrared light can be used to penetrate the skull to improve tissue energetics, notwithstanding an anaerobic condition. Data are very preliminary at this time, but laboratory studies are encouraging. Shining low level infrared light at a specific wavelength improves ATP production in neuronal culture, and improves neurological outcome in animal stroke models. Although preliminary studies in humans with stroke appeared promising, further study failed to demonstrate any benefit (NEST-2, NEST-3). However, there may yet be a role for laser therapy in the neurosurgical setting.
Spinal Cord Injury Neuroprotection
Many avenues are being explored with many indications that this injury may be amenable to protection acutely and restoration chronically. Current research suggests the following as approaches likely to be translated eventually to clinical care.
Acute Protection
Promising studies in this area include pharmacological therapies, surgical decompression, , and hypothermia. Early laboratory studies demonstrated the therapeutic potential of surgically applied hypothermia for several hours in the acute management of spinal cord injury with dramatic attenuation of paraplegia demonstrated by Albin et al. In 2007, an NFL football player suffered a complete cervical spine injury and was immediately treated on the field with moderate hypothermia using cold saline IV. This player had a functional outcome that was better than expected and renewed interest in hypothermia as a treatment for SCI.
Maintenance of spinal cord blood flow is likely to be important in the acute management of spinal cord injury. Mesquita et al. are presently developing a near-infrared spectroscopy (NIRS)-based device that could be placed in a manner similar to an epidural catheter to yield continuous real-time spinal cord blood flow information. Preclinical studies in a spinal cord ischemia model are encouraging. This could have a dramatic impact on the management of patients presenting with spinal cord injury or spinal cord ischemia.
Stem Cell Therapy
This is based on the use of embryonic stem cells or progenitor cells from other sources such as bone marrow to promote regeneration of cells and remyelination. Animal studies support the feasibility of this approach, showing the capability of embryonic stem cells to develop into glial cells and into neural cells which make synaptic contacts. In addition to these sources the nervous system itself may also be a source of endogenous stem cells.
Genomic Therapy
Preclinical studies suggest a role for siRNAs in potentiating repair after SCI. After spinal cord injury, astrocyte activation promotes glial scar formation. Telomerase reverse transcriptase (TERT) is part of the telomerase enzyme complex that aids in prolonging cell life by protecting the telomere. It is also associated with multiple nontelomere actions including cell activation, proliferation, and apoptosis. Evidence supports the hypothesis that TERT may be involved in astrocyte activation that results in glial scar formation. However, an antisense nucleotide targeting astrocyte TERT mRNA failed to significantly reduce glial scar formation following SCI in rats. Astrocyte activation and glial scar formation is a poorly understood process and the search for other potential molecular targets continues.
Neurotrophin Therapy
A variety of approaches are currently being explored including cytokine blockade, implantation of trophin producing cells, , viral gene transfer, and implantation of stem cell progenitors of glial cells.
Activity-based Restorative Therapy
Animal studies indicate that the paradigm of Brus-Ramer et al. and others , offers promising opportunities to restore motor function using activity-dependent processes to strengthen the connectivity between “top-down” and “bottom-up” pathways. This notion has been further elaborated with early studies evaluating nanotechnology to produce focused electrical stimulation while monitoring consequent neurochemical processes deriving from such stimulation. Clinical implementation of this approach is in early stages of study and implementation , using functional electrical stimulation (FES) has produced encouraging results, indicating the feasibility of this approach, perhaps in combination with stem cell therapy, to promote neural proliferation in areas of the spinal cord adjacent to the injury. Notably, baclofen therapy seems to inhibit such regeneration. A functional translation of this approach is the FES bicycle, used for years to promote return of function. The best known example of this sustained, determined approach to achieve some recovery after a severe spinal cord injury is the case of actor Christopher Reeve who improved by two ASIA grades over 5 to 8 years post injury.
Multimodality Approach to Neuroprotection
Establishing efficacy of new neuroprotective therapies in neurocritical care and stroke has proven to be an exercise in futility. Over 2000 completed clinical trials are listed on the Internet Stroke Trials Registry with few apparent reproducible results of any demonstrable efficacy in the acute context. However, these negative studies belie the supportive basic laboratory studies that justified the time and enormous expense for such attempted translational clinical trials. We provide a rationale to suggest that such results, in retrospect, are altogether predictable and suggest an explanatory model for such reproducible futility in a complex biological system. Many of the points discussed are supported in an editorial by Grotta. It is to be expected that these issues will eventually lead to a commonplace approach to use multimodal therapies for neuroprotection. Donnan, in the 2007 Feinberg lecture, suggests: “We have reached a stage at which research in this area should stop altogether or radical new approaches adopted.” The new approaches may include a reevaluation of prospective randomized studies as the only path to new knowledge. A rationale for this approach as advocated by Kofke follows.
Imagine a factory that makes widgets. A number of processes are important for the quality of the final widget as it proceeds: conveyor speed(x 1 ), presence of raw materials and power(x 2 ), quality of bolts(x 3 ), quality of steel(x 4 ), and type of metal used for circuits(x 5 ). A weighting factor can be applied to each variable w i leading to the following general equation describing the widget quality:
Each variable x can be precisely known with very small variation so any change in any of the variables will produce a reproducible and predictable change in the widget quality Q.
In a biological system characterized by severity of a pathophysiologically complex injury, S, a similar equation can be derived with important pathophysiologic factors, x i , and weighting factors, w i :
Notably different from the widget, however, is that there are a large number of disparate and potentially interacting factors known to contribute to with also an unknown number of as yet unknown factors with correspondingly unknown weighting factors and variability, partially represented in Fig. 29.1 . Moreover, each pathophysiologic factor x i has to be described over a biologically diverse population such that each factor has an associated central tendency and large normal or non-normal distribution about that mean. Moreover, the weighting factors and pathophysiologic factors also vary as a function of time after the onset of the insult. Thus, the importance of restoring cerebral flow is very high early on but late in the course it becomes a less important contributor to final outcome (e.g., thrombolysis works well if done promptly, but is fruitless if done a day later as the infarction process has completed).
Additionally, in the context of clinical medicine there are also associated system factors, H i , such as nursing ratio, nursing experience, availability of drugs and technology, efficiency of rapid response teams, and so on, which are also important to the severity of injury such that the equation can be written as:
Given the above characterization of the multiple highly variable biological and system factors that enter into a given outcome, it should come as no surprise that clinical studies directed at improving only one of the abovementioned numerous complex factors tend to show no effect, especially if multi-institutional in design (increasing variation in H factors), unless it is truly a breakthrough phenomenon (large W factor such as early thrombolysis in ischemic stroke) or the therapy exerts a multifaceted effect (e.g., hypothermia). This then leads to the notion that the current widely accepted methods of advancing clinical knowledge for complex problems is generally futile, which subsequently produces innovation paralysis on the part of institutions, third party payers, clinicians, pharmaceutical companies, and investigators; an alternate method that is based on a multifactorial approach is needed. Rogalewski et al. have recently reviewed and endorsed this concept; however, they fail to suggest a rational means for building the multimodal approach other than trying everything at once. A rational method is needed. Kofke’s proposal entails a hybridization of QI methods and standard research techniques to create a novel model of incremental addition of therapies with ongoing evaluation of effects.
Intracranial pressure management
Traditional methods of managing intracranial hypertension are primarily phenomenological. That is, the intracranial pressure (ICP) is elevated and therapy is implemented to decrease it. If warranted, investigation may be implemented to identify the etiology of intracranial hypertension (e.g., hydrocephalus or masses). However, vascular factors contributing to intracranial hypertension typically receive scant attention as to their contribution to an ICP problem. These vascular factors include systemic hypertension, cerebral venous hypertension, and hyperemia. Grande and coworkers have emphasized the role of systemic hypertension exacerbating hydrostatic brain edema, leading to venous outflow obstruction, which, in turn, further exacerbates hydrostatic edema in a positive feedback cycle. Piechnik et al. provide a mathematical model underscoring this issue, which, although triggered by systemic hypertension, is propagated by cerebral venous hypertension. Early experimental support is provided by Hayreh and Nemoto.
Another little appreciated factor is the role of normotensive hyperemia in the genesis of hydrostatic edema. Observations after arteriovenous malformation resection and carotid endarterectomy suggest a role for hyperemia contributing to a normal perfusion pressure breakthrough syndrome. Moreover, static observations of multiple patients at different points on the path from normal to hepatic encephalopathy provides strong circumstantial support for the notion that hyperemia precedes brain edema and intracranial hypertension in patients with liver failure. ,
It seems likely that once appropriate monitors have been developed to provide reliable information regarding cerebral venous pressure/volume and cerebral blood flow, manipulation of these physiologic variables will become part of the standard paradigm in the treatment of intracranial hypertension.
Technology
Monitoring
Neuromonitoring has long been an area of active research and clinical contributions in anesthesia. Such a tradition is to be expected into the future. However, this area is full of challenges regarding validation of new monitors as accurately reflecting the desired parameter and, moreover, showing that making such measurements matters in a way that affects outcome. As noted in a quote of Pickering in an editorial on ICP measurement by Crosby and Todd : “Not everything that counts can be counted, and not everything that can be counted, counts.”
Continuous regional blood flow and metabolic rate monitoring, long a holy grail of neuromonitoring, should eventually be realized as a clinical reality. Currently attractive noninvasive methods tend to be based on NIRS with other invasive techniques that may also be useful.
Near-Infrared Spectroscopy-based Techniques
Since being first described by Jobsis in 1977 and hailed as the “Monitor of the Future,” NIRS has been developed as a noninvasive monitor of cerebral chromophores, oxyhemoglobin, deoxyhemoglobin, and cytochrome aa3. This is made possible by the property of infrared light being able to penetrate into tissue much deeper than visible light. , More recently researchers are reporting use of NIRS to facilitate continuous bedside rCBF monitoring. NIRS-based diffuse correlation spectroscopy (DCS), diffuse reflectance spectroscopy (DRS), and NIRS-based ICG blood flow index (BFI) and absolute CBF hold significant promise as bedside monitors of rCBF and metabolic rate (CMRO 2 ). DCS combined with ICG offers the potential for a truly continuous rCBF and CMRO 2 monitoring.
Diffuse Correlation Spectroscopy and Diffuse Reflectance Spectroscopy
DCS for continuous evaluation of rCBF and DRS for continuous evaluation of CMRO 2 and oxygen extraction fraction (OEF) are recently reported advances using infrared light to provide real-time continuous information on changes in rCBF and rCMRO 2 , with quantitative information on OEF, with promising studies reported in subprimate animal models and preliminary studies in humans.
Diffuse Correlation Spectroscopy
Near-infrared photons diffuse through thick living tissues. When diffusing photons scatter from moving blood cells, they experience phase shifts that cause the intensity of detected light on the tissue surface to fluctuate in time. These fluctuations are more rapid for faster moving blood cells. Therefore, one can derive information about tissue blood flow far below the tissue surface from measurements of temporal fluctuations impressed on diffusing light. Further details of the DCS method can be found elsewhere. , Changes in DCS CBF were evaluated by Kim et al. in subarachnoid hemorrhage (SAH) patients undergoing blood pressure manipulation with Xe-CT-CBF assessment. The DCS measure correlated well with the appropriate region of interest in the Xe-CT-CBF ( Fig. 29.2 ).
Diffuse Reflectance Spectroscopy
It is well known that the near-infrared photon fluence rate obeys a diffusion equation in highly scattering media such as tissue. In DRS light measurements employ intensity modulated light sources (i.e., the frequency domain technique). The amplitude of the input source is sinusoidally modulated, producing a diffusive wave within the medium. These disturbances are called diffuse photon density waves or simply diffusive waves . When everything works, a best estimate of the absorption and scattering coefficients at one or more optical wavelengths is obtained, from which can be derived contributions from different tissue chromophores. Oxy- and deoxyhemoglobin concentrations along with water concentration are the most significant tissue absorbers in the NIR. Their combination gives total hemoglobin concentration, which can be referred to as blood volume and blood oxygen saturation or StO2, both of which are useful physiological parameters. Combined with DCS calculation of CBF these measurements lead to a real time bedside assessment of CMRO 2 and OEF. , , A detailed theoretical treatment of these concepts can be found at http://www.lrsm.upenn.edu/pmi/nonflash-ver/publicationNF.html .
In early experiments, investigators at the University of Pennsylvania compared DCS measurements of flow variation to other standards. Direct comparisons were made to Doppler ultrasound, , to laser Doppler flowmetry, to arterial spin labeled MRI, and to reports in the literature. , , Moreover, this group carried out extensive studies in phantoms , , wherein the medium’s viscosity and the flow speed of scatterers are varied. Overall, these validation studies have shown that DCS measurements of blood flow variations are in good agreement with theoretical expectations and with other measurement techniques. In rodents this method has been demonstrated to detect hyperemia due to hypercapnia , and ischemia due to middle cerebral artery occlusion and cardiac arrest with appropriate changes in OEF and CMRO 2 .
Near-Infrared Spectorscopy-based Indocyanine Green, Blood Flow Index and Cerebral Blood Flow
NIRS also allows detection of other IR chromophores such as indocyanine green ( ICG ). ICG has several attributes that make it an ideal tracer to monitor rCBF. ICG has an IR absorption peak at 805 nm and after intravenous injection is limited to the intravascular compartment. These properties then make the use of ICG as a marker of tissue blood flow feasible. It is nontoxic and serious adverse reactions are rare. Notably it is rapidly removed from the circulation by the liver through biliary excretion with a circulation half time of 3.3 minutes. , These kinetic properties make ICG suitable for repetitive measurements, even with short between-study intervals, without accumulation of dye. With a maximal daily dosage of 5 mg/kg and a study dose of 0.1 mg/kg, up to 50 rCBF determinations can be made daily. Thus, although not a true moment-to-moment monitor, being able to evaluate BFI as a proportionate indicator of rCBF every 30 minutes as one manipulates and optimizes physiology or pharmacology at the bedside based on rCBF is, nonetheless, a very attractive notion.
The BFI is based on fluorescein flowmetry for measurement of relative blood flow changes in the intestine using intravital fluorescence microscopy. This method entails use of the ratio of the maximum fluorescence and the rise time of the fluorescence curve. Keubler et al. adapted the mathematical basis for fluorescein flowmetry to the first circulatory passage of an ICG bolus through the brain ( Fig. 29.3 ).
Several approaches to the use of ICG NIRS for rCBF determination or estimation have also been reported. Early studies involved using ICG to determine relative changes in CBF have been described by Roberts et al., Gora et al., and Kuebler et al., with subsequent validation studies reported in animals , , and humans. , Diop et al. report encouraging preclinical studies demonstrating the potential use of ICG combined with DCS to yield continuous absolute CBF, CMRO 2 , and OEF. ,
Other Near-Infrared Spectroscopy Cerebral Blood Flow Methods
Smith et al. have recently reported an alternate approach, also using NIRS, to measure OEF at bedside. Edwards and Elwell, using the Fick principle, suggested measuring CBF by monitoring changes in oxygenated hemoglobin in response to changes in FiO2 as a tracer. , This technique was found to correlate with 133-Xe rCBF measurements in neonates. , However, subsequent studies in animals and in adults showed an unacceptably high coefficient of variation and it did not correlate with other established methods. , Moreover, there were concerns that altering FiO2 could be deleterious in brain-injured patients. Acousto-optic techniques are also available which can be used to measure relative changes in CBF. Validation studies are ongoing and the technology is FDA approved
Reliable Measurement of Depth of Hypnosis and Analgesia and Potential for Automated Closed Loop Anesthetic Administration
BIS and patient state index (PSI) monitors have undergone extensive evaluation as monitors of depth of hypnosis with reasonable evidence for a role in some neurosurgical settings. However, this depth of anesthesia monitoring approach provides little insight into adequacy of analgesia. Recent early work in this area suggests that depth of analgesia monitoring will also eventually be developed and validated. Studies have been done on facial EMG, palmar conductance, , and pupillometry with reasonable results, suggesting that such monitoring should be clinically feasible for future applications. The Analgoscore is another means of monitoring intraoperative pain through the application of an algorithm to blood pressure and heart rate data. This score has been coupled with the closed-loop administration of remifentanil to control intraoperative pain and combined with BIS and neuromuscular blockade monitoring to create a prototype total anesthesia robot, , which is described in more detail later in this chapter.
The availability of depth of hypnosis, analgesia, and neuromuscular blockade, along with cardiovascular monitoring will then enable the neuroanesthesiologist to precisely titrate the anesthetic drug as a “magic bullet” directed to that element of the anesthetic state most in need of attention.
Continuous Blood Levels of Intravenous Anesthetics
Presently no clinically available technology provides continuous direct measurement of blood levels of anesthetics. One approach has been to use a pharmacodynamic measure such as BIS to derive the blood level that works, even if the exact level is not known, and then titrate anesthetic versus effect rather than a blood level. Pharmacokinetic models combined with target-controlled infusion technology have also been used to predict anesthetic levels and titrate accordingly.
Nonetheless, recent reports indicate that sensitive exhaled gas monitoring can provide end tidal concentrations of propofol. Indications are that such a monitor provides a reliable measure of blood concentration. No comparable technology has been reported for other intravenous anesthetic drugs. One might speculate that a hypnotic, paralytic, or analgesic drug of the future might be designed with a chromophore in its structure that would allow detection and thus continuous monitoring by transcutaneous spectroscopy.
Multimodality Brain Monitoring
Brain tissue pO2 (pbO2), microdialysis, and blood flow methods are receiving significant attention presently. There is ample retrospective evidence to associate a low pbO2, low CBF, , , and elevated lactate pyruvate ratio with poor outcome and certainly it makes physiologic sense. It seems reasonable to expect that these monitors will eventually find a place in titration of the physiologic contributors’ secondary processes in brain injury. Multimodality monitoring has recently been evaluated in a consensus statement.
Exhaled Gas Monitoring
Previously the province of nonportable GC-mass spectrometry types of equipment, improved sensor and computing technology has resulted in the capability to place a highly sensitive array of chemical sensors in exhalation tubing of a patient to make significant physiologic inferences. Currently published examples suggest an ability to detect bacterial pneumonia and sinusitis, asthma, aerodigestive tract tumor cells, lung cancer, and tuberculosis. The technology has been reported as being able to detect lipid peroxidation in food. Given that lipid peroxidation or perhaps release of other volatile organic compounds occurs with ischemia, it becomes plausible to suggest that this technology will have a place in screening for immediate evidence of ongoing cerebral (or other organ) ischemia in neuroanesthesia and neurocritical care.
Feedback Loops
Anesthesia Robot
Multiple physiologic parameters can be measured with digital output that is amenable to feedback loops. Technology is now available or in development that is demonstrating the feasibility of this concept for parameters that include neuromuscular blockade, depth of hypnosis monitors, facial EMG (analgesia), and blood pressure. Given that the administration of an anesthetic entails provision of hypnosis, analgesia, immobility, and sympathetic reflex control, this then presents the notion of the anesthesiologist explicitly prescribing an effect of a drug rather than a dose as the primary goal in the administration of anesthesia. Moreover, for the neuroanesthesiologist, brain oriented parameters such a brain pO2, tissue lactate or glutamate, CBF, or other measures may also be amenable to such an approach. In addition, the notion of providing , and measuring amnesia is also an attractive possibility, although amnesia monitors at this time remain undescribed (other than after the fact inquiry).
Putting all of these concepts together produces the concept of an anesthesia robot. Researchers at McGill University have described such a system that they have named McSleepy. , This innovation is both pharmacologic robot and anesthesia information management system. An automated, closed-loop system controls the three variables of general anesthesia: hypnosis, analgesia, and neuromuscular blockade. Bispectral index is used as the control variable for hypnosis, analgoscore for analgesia, and phonomyography for neuromuscular blockade. Based on these data, the computer titrates the appropriate medication (propofol, remifentanil, rocuronium) using three infusion pumps. The system can be used to control induction, maintenance, and emergence from general anesthesia. The system also alerts the anesthesiologist when certain actions should be performed, such as mask ventilation, intubation, and waking the patient. Several safety features have been incorporated into the software. For example, the system will not administer rocuronium until a BIS < 60 is reached and maximum and minimum limits can be set for drug administration. In addition, McSleepy can be controlled remotely from any PC, smartphone, or tablet. In fact, researchers used this capability to perform the first transcontinental anesthetic between Montreal, Canada, and Pisa, Italy. Additional features such as voice command and fluid management are on the horizon. A large scale trial is underway to compare the performance of the system to manually controlled anesthesia. This group is also exploring the performance a robotic intubation system called the Kepler Intubation System. ,
Magnetic Resonance in the Operating Room
MR imaging is undergoing evaluation as an aid during neurosurgery. The primary impetus for it is to aid in the provision of more complete tumor resection and to reevaluate stereotactic coordinates intraoperatively as brain anatomy may be altered during the course of a procedure. Another emerging use is for laser tumor ablation under MRI guidance. This introduces a variety of logistical challenges that can substantially increase the time in the operating room or create issues with transport of an anesthetized patient, mid-surgery, to an MR unit. Nonetheless, the introduction of this may translate into a technology providing information of value to both neurosurgeon and anesthesiologist, including biochemical information, such as high energy phosphates, lactate, n-acetylphosphate, and other relevant chemicals. In addition, rCBF and metabolic rate information may also become available during neurosurgical procedures through MRI methodology.
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
