31 July 2009. Anesthesia is meant to induce a temporary fog, lifting as soon as the drugs wear off. But for some, the effects linger, with delirium or confusion for days or weeks following surgery. Over the last decade, a small cadre of scientists has begun investigating whether anesthesia might occasionally have permanent effects, pushing some people closer to Alzheimer disease (reviewed by Xie and Tanzi, 2006). Although cell culture and animal experiments show common anesthetics can cause amyloid-β production and apoptosis, the possible connection in people is hazier. “We have a smoking gun, but no victim at this point,” said Roderic Eckenhoff, an anesthesiologist who researches the topic at the University of Pennsylvania in Philadelphia.
Scattered case reports of mental fuzziness after surgery have been part of the scientific literature for years, and doctors occasionally hear of a patient who was never quite the same, cognitively, after a spell on the operating table. “That story is so common,” Eckenhoff said. “We give these drugs to tens of millions of patients every year, and blithely ignore that they could have long-term effects.” At a recent workshop in Philadelphia, Eckenhoff and others concluded that the current data on anesthesia and Alzheimer’s are worrisome enough to warrant further studies (for a consensus statement see Baranov et al., 2009). Researchers will also convene this fall to plan multi-site clinical studies.
With clinical data scarce, researchers say it is too soon to alter operating room practice. But a few are already modifying their treatment of elderly patients. “I’m telling my own family and friends to avoid isoflurane,” said Rudy Tanzi of the Massachusetts General Hospital in Boston, referring to a common inhalation anesthetic. Tanzi’s mother had several surgeries including hip and knee replacements when she was in her early seventies. She always experienced a few days of mild delirium following an operation—until the most recent surgery, when Tanzi asked the anesthesiologist to swap isoflurane for desflurane. Following that procedure, his mother awoke clearheaded, “like nothing happened,” Tanzi recalled. When she heard it was 7:30, she demanded he turn on the TV because the Red Sox were playing and she wanted to see David “Big Papi” Ortiz. “It’s an n of one, but it’s my mother, so it counts for a lot,” Tanzi said.
In vitro and animal studies link common anesthetics to both the Aβ plaques and tau tangles found in Alzheimer’s. Tanzi and Zhongcong Xie of Massachusetts General Hospital in Boston have shown that inhaled compounds such as isoflurane increase production of Aβ in cultured cells (Xie et al., 2006; Dong et al., 2009), and Eckenhoff and colleagues found that the same chemicals lower the concentration of Aβ needed to oligomerize in a test tube (Eckenhoff et al., 2004 and Carnini et al., 2007). Emmanuel Planel of Laval University in Québec City has found that anesthetics and the accompanying hypothermia can increase phosphorylation of tau, too (see ARF related news story and Planel et al., 2007).
Anesthesia also causes apoptosis of neurons, and the apoptotic pathways and Aβ can feed off of each other in a vicious cycle. Caspases, activated in apoptosis, destroy the chaperone GGA3, which normally restricts BACE activity. Unchecked, BACE cleaves more APP, setting it on the path to become Aβ (see ARF related news story and Tesco et al., 2007). At the same time, the addition of Aβ to cells increases caspase activity further (see ARF related news story and Xie et al., 2007).
Based on the cell and animal data, scientists hypothesize that similar damage might occur in the human brain. For a cognitively normal person, the effects of anesthesia may be minimal. But for a person already at risk for AD—maybe because of an ApoE4 genotype or mild cognitive impairment—anesthesia may be the final insult that launches the brain into Alzheimer’s mode. “It’s like a stepping stone,” said Walter Schulz-Schaeffer of the University of Göttingen in Germany. In animals, too, those most prone to AD-like pathology may be most sensitive to the effects of anesthesia. For example, in Planel’s tau experiments, wild-type mice quickly return to normal levels of tau phosphorylation after anesthesia, but animals carrying a tau mutation associated with frontotemporal dementia do not (Planel et al., 2009). However, researchers emphasize that without human studies, their ideas remain unproven.
Part of the challenge is that despite more than a century of experience with anesthetics, doctors still don’t know exactly how the drugs work (reviewed in Eckenhoff, 2001). Inhaled anesthetics interact with enzymes, receptors, transporters, channels, and structural proteins. At least some of the damaging effects of anesthesia may be due to the hypothermia that comes with it. Anesthesia throws off the body’s thermostat: “Patients become, for lack of a better term, cold-blooded during surgery,” Eckenhoff said. In Planel’s studies of tau under anesthesia, maintaining normal temperature abolished the increased tau phosphorylation that anesthesia usually produced. Keeping patients warm is an easy precaution that doctors can take now, Planel said: “It wouldn’t hurt.”
And temperature is just one factor that doctors should track so that they can analyze the effects of anesthesia. A surgical stay in the hospital will also involve pain, stress, inflammation, perhaps infection, all of which could confound studies of cognition. A person may have cognitive impairment before an operation, but friends and family might not notice until afterward, and mistakenly attribute the problem to the surgery. To truly sort out the relationship between AD and anesthesia will require a large, expensive study.
Clinical research has already confirmed the immediate cognitive effects of surgery, which fit two categories. Acute post-operative delirium (POD), in which patients may not know who or where they are, may last a few days. Post-operative cognitive dysfunction (POCD), causing people to struggle with basic tasks such as memory and word-finding, may not dissipate for weeks or months. Some scientists suspect these conditions are short-term variants of Alzheimer-like pathology, perhaps caused by a temporary Aβ increase that the brain eventually overcomes. Even a single episode of delirium is associated with quickened cognitive decline in people who already have Alzheimer’s (see ARF related news story and Fong et al., 2009). But the relationship is tricky to study, Xie said, because data are missing on both sides. Doctors know POD and POCD are real, but have no information on the underlying molecular mechanisms. Researchers have used mazes to test rodents for the effects of POCD on memory and learning (for examples, see Wan et al., 2007; Culley et al., 2004; Culley et al., 2003), but delirium is harder to assess in animal models. “How do you know if a mouse has lost touch with reality?” Eckenhoff asks. Conversely, scientists know plenty about the neuropathology of AD, but lack positive proof that it is associated with anesthesia.
Xie has commenced a couple of studies to look for anesthesia effects in people. In one project, he is examining brain tissue removed during surgery to treat epilepsy, looking for evidence of apoptosis markers, enzymes involved in Aβ processing, and phosphorylated tau. Xie and Eckenhoff are collaborating in a study of cerebrospinal fluid samples, collected when people undergo spinal anesthesia. They hope to correlate CSF Aβ levels with the severity of post-operative cognitive symptoms.
Should scientists conclusively link anesthesia to Alzheimer’s or Aβ, they could then look for ways to circumvent the problem. Anesthesiologists might screen for vulnerable people who need special care or should avoid certain anesthetics. For instance, one study linked the ApoE4 allele to a predisposition for POD (Ely et al., 2007). Scientists could also look for co-treatments that might diminish anesthesia’s impact on cognition. Tanzi and Xie found that the Alzheimer’s drug memantine reduced the caspase activation caused by isoflurane in a cell culture study (Zhang et al., 2008). Planel is also testing memantine to see if it attenuates the pathology he sees in anesthetized mice.
At this point, doctors and patients have precious little information to go on. Xie regularly fields e-mails from concerned people facing surgery, and he tells them to follow their anesthesiologist’s recommendations, and not to base medical decisions on limited laboratory results. “If you try another drug, how do you know the other drug is not also dangerous?” he asked. “Simply, we do not know the answers.”—Amber Dance.