ArticlePlus
Click on the links below to access all the ArticlePlus for this article.
Please note that ArticlePlus files may launch a viewer application outside of your web browser.
ALTHOUGH described more than 50 yr ago,1postoperative cognitive dysfunction (POCD) remains an enigma. The subject of a Rovenstine Lecture in 2003 and numerous panels at international meetings, POCD was again featured prominently at the American Society of Anesthesiologists Annual Meeting in October 2006 as the topic for the Journal's sponsored symposium. Each of the keynote speakers graciously agreed to be recorded and to provide their slides, and we are pleased to offer their presentations as a Web Enhancement to this month's issue of Anesthesiology. Also in this issue of the Journal are POCD-related review articles,2,3as well as articles that have evolved from posters presented at the symposium.4–6
Opening up proceedings at the Journal's sponsored symposium was Lars S. Rasmussen, M.D., Ph.D., Associate Professor, Department of Anesthesia, Centre of Head and Orthopaedics, Copenhagen University Hospital, Copenhagen, Denmark. Also the highly regarded leader of the International Study of Postoperative Cognitive Dysfunction (ISPOCD) group, he was tasked with addressing four questions: (1) Does POCD exist after noncardiac surgery? (2) If so, what is the incidence? (3) Which risk factors have been identified? and (4) Is the incidence affected by the choice of the anesthetic technique (general vs. regional)? The existence of POCD has been challenged; the hypothesis that the care associated with a surgical intervention produces a greater decline in cognitive performance than could have been expected to occur had these surgically scheduled patients not undergone surgery has never been formally tested. Absent a control group of surgical patients randomized to nonsurgical care, investigators have sought a variety of less appropriate alternatives, including age-matched healthy controls, or disease-matched patients who do not undergo the surgical procedure because of the choice of the patient or the surgeon. At the poster session that accompanied the Journal-sponsored symposium, Saager et al. 7reported their preliminary findings obtained from the Alzheimer's Disease Research Centre's database. The annual rate of cognitive decline was accelerated in an aged population after a major noncardiac surgical intervention when compared with a similar cohort, prospectively followed for at least 3 yr, who did not require a surgical intervention. Epidemiologic studies of this type are likely to shed light on the question of whether cognitive decline in surgical settings differs from the apparent decline experienced by aged patients receiving care for major medical illnesses.
Stanton Newman, D.Phil., Professor and Director, Center for Behavioral and Social Science in Medicine, Royal Free and University College of London Medical School, London, England, addressed methodologic aspects in the diagnosis of POCD. Coming from the doyen of behavioral psychologists in this field, Dr. Newman identified problems in the realm of testing and analysis of the neurocognitive data sets. These are further elaborated in a review article by Newman et al. 2in this issue of the Journal.
Mark Newman, M.D., Merel H. Harmel Professor of Anesthesiology, Chief, Division of Anesthesiology Cardiac Service, Chair, Department of Anesthesiology, Duke University Medical Center, Durham, North Carolina, is a pioneer in the field of POCD after cardiac surgery. He suggested that the distinction between cardiac and noncardiac POCD may be more apparent than real and raised the likelihood that these two artificially separated conditions may be governed by similar etiologic factors and pathogenic mechanisms. Previously, Newman et al. 8had provided data that demonstrate the independent adverse effect that POCD has on long-term morbidity and mortality. More recently, Newman's group has reported on the deterioration in quality-of-life indices that occurs in patients with POCD.9
The panelists agreed that POCD does indeed exist as an often subtle persistent deterioration in cognitive performance (including one or more of memory, attention, and speed of information processing) after a surgical intervention; its assessment requires neuropsychological testing, and it is diagnosed by applying firm criteria to yield a dichotomous result. Thereafter, there was lively debate regarding when the change in cognitive performance between two time points separated by a surgical intervention constitutes POCD. Specific issues addressing the degree of change, together with the timing, type, and analysis of the neurocognitive battery of tests, were explored in depth without reaching a consensus view.
In a fascinating mechanistic study, Wan et al. 4subjected anesthetized rats to splenectomies and demonstrated that there was a brief period of cognitive dysfunction, as assessed with a Y-maze, and that this was associated with biochemical markers of glial activation and inflammation (interleukin-1β and tumor necrosis factor-α messenger RNA) in the hippocampus. Of particular note was that anesthetized animals that had not undergone surgery showed none of these changes, reinforcing the concept that the observed changes were not related to anesthesia per se .
Interestingly, in most of the earlier clinical trials addressing the prevalence and risk factors associated with POCD, patients with existing cognitive impairment were excluded (e.g. , see ISPOCD1).10In two articles resulting from this Journal-sponsored symposium, both prospective5and retrospective6analyses were used to redress this deficiency.
Hudetz et al. 5prospectively examined the independent effect of self-reported alcohol abuse on the development of POCD and demonstrated what others had intuitively suspected,11that with less “cognitive reserve” (lower baseline scores), patients are more likely to develop POCD. There are several caveats that need to be considered in this relatively small study (i.e. , only 14 patients in each of four groups). The surgical alcoholic group included two patients who developed postoperative complications, each of which has been independently shown to increase the risk of POCD10; therefore, the influence of alcohol abuse on the subsequent development of POCD is brought into question. Furthermore, the alcoholic surgical patients were more likely to require postoperative admission to the intensive care unit, where sleep deprivation is likely to occur12and can result in postoperative confusional states including delirium.13The fact that patients were assessed relatively soon after the surgical procedure (within 2 weeks) makes this a distinct possibility.
Silverstein et al. from ISPOCD have retrospectively reviewed their data from the first trial10to determine whether particular cognitive domains are more severely affected by a surgical intervention in patients in whom preoperative cognitive impairment can already be detected. As can be expected from their original inclusion criteria (> 23 on the Mini-Mental State Examination), only 74 patients (of a total of 1,185 patients) met their definition of preoperative impairment (visual verbal learning test score that is 1.5 SD below that of healthy age-corresponding controls; this definition was selected to resemble the amnestic minimally cognitively impaired cohort). Of the four cognitive domains (memory, information processing, planning, and attention) that the ISPOCD neurocognitive battery tests, Silverstein et al. noted that postoperative deterioration in memory was less likely in the preoperative cognitive impairment group.10However, this may be due to the phenomenon of “regression to the mean” in which a low score is not capable of dropping as much in absolute terms as a higher score. This becomes particularly problematic when use is made of the “change in score” (referred to as a “floor effect” by Silverstein et al. ). Residualized change scores is a method of measuring change that could have avoided the problem that Silverstein et al. 3seem to have encountered. In this calculation, the postoperative score is regressed at a “normal” rate of decline to yield a preoperative score, and the difference from this predicted value and the observed value provides the “unexpected change.”
While Hudetz et al. 5concentrate on the cognitive effects of alcohol abuse and Silverstein et al. purport to address the intermediate state between normal aging and Alzheimer's Disease,14both groups have ventured into what should become fertile territory. It will be necessary to comprehensively examine the effects of a surgical intervention on the slope of cognitive decline in those already cognitively impaired in order for these patients to be fully apprised of the possible cognitive consequences of a surgical intervention before providing consent for a non-life-preserving surgical intervention.
Although the original four Koch's postulates referred to the causal relationship between an infective agent and a disease (e.g. , tuberculosis and anthrax),15at least two of the four can be used to establish the existence and to understand the etiology of other “new” diseases such as POCD in a preclinical laboratory setting. In this vein, one may wish to investigate what are the minimal features are that are exhibited by all patients with POCD and whether these can be transferred to a healthy organism to recreate the disease. Hypotheses testing, as well as the development of treatment strategies, can be explored in validated animal models; in particular, the role of possible causative (as well as protective) interventions that occur in the perioperative period, and the contribution of genetic polymorphisms can be investigated.
With the aging of their populations, it is predicted that cognitive decline will constitute an increasing healthcare problem in the developed countries of the world, causing despair to patients and their providers alike while consuming a growing fraction of available healthcare resources. The latter should provide the “business case” to fund the type of studies that are required to address the genesis of this health problem and, in particular, the contribution of what role a surgical intervention plays in cognitive dysfunction. Were postoperative cognitive decline shown to be no different from that which occurs in the elderly confronted by any cause of failing health, it still requires the perioperative practitioner to understand this condition more fully because, uniquely, the elective surgical setting provides an opportunity to test preemptive therapies. Falling between several stools, a clinical discipline needs to take “ownership” of this putative perioperative problem, and in our view, this needs to reside within our specialty. The major large-scale clinical trials defining some of the issues involved have been led by the eminent keynote speakers at the Journal's sponsored symposium.8,10Although the existing data seem to rule out a putative anesthetic causative factor,16we can be at the forefront, hunting for and implementing solutions with the same success that has seen our discipline pioneer strategies to prevent and treat ischemic-reperfusion organ injury both in and outside the operating room.
* Department of Anaesthetics, Pain Medicine and Intensive Care, Imperial College London, London, England. m.maze@imperial.ac.uk. † Department of Anesthesia, Carver College of Medicine, The University of Iowa, Iowa City, Iowa.
