The Effect of Anesthesia and Surgery on Postoperative Changes in Plasma biomarkers of neuronal injury, Alzheimer's disease, and inammation in healthy subjects

Background: Anesthesia and surgery have been linked to neurological sequelae such as perioperative neurocognitive disorders (PND) and increased risk of Alzheimer's disease (AD). The exact mechanisms of PND remain ambiguous and controversial, which were deserved to explore further. Methods: Healthy subjects undergoing general anesthesia for orthognathic surgery were prospectively randomized to receive propofol or sevourane for anesthetic maintenance. Blood samples were taken preoperatively and at 3, 24, and 48 hours after surgery. Neurolament light (NFL), tau, Amyloid β (Aβ)40, Aβ42 and 21 inammatory mediators in plasma were measured using highly sensitive assays. Results: A total of 50 healthy subjects were enrolled. The mean (SD) age was 24.80(4.63) years. Plasma NFL increased at each measurement from a baseline mean (SD) of 22.3 (20.4) pg/mL to a maximal mean (SD) level of 35.1 (28.7) pg/mL, a maximum increase of 599%, at 3 hours postoperatively. NFL level began to decline at 24 hours, but remained higher at 48 hours. The levels of Aβ40 and Aβ42 decreased at 3 hours, and to minimum mean (SD) of 196.70 (38.61) pg/mL and 8.01(1.66) pg/mL at 24 hours postoperatively, respectively. There were no signicant differences in the concentrations of plasma tau after anesthesia and surgery. Plasma IL-6, IL-7, IL-8, IL-10, TNF-α, I-TAC and MIP-1β were signicantly increased at 3 hours postoperatively and then declined, which had a similar trajectory with a return to baseline. The peak levels of NFL, IL-6, IL-8, TNF-α and MIP-1β correlated with duration of surgery. The peak plasma NFL level signicantly correlated with the levels of IL-6 and IL-8. Conclusions: might be attributed to many of inammatory mediators. The data indicate systemic inammation after anesthesia and surgery may induce neuronal injury. These preliminary ndings in healthy subjects could help us to understand the effects of anesthesia/surgery on brain and the potential mechanisms of PND.

many of in ammatory mediators. The data indicate systemic in ammation after anesthesia and surgery may induce neuronal injury. These preliminary ndings in healthy subjects could help us to understand the effects of anesthesia/surgery on brain and the potential mechanisms of PND.

Background
It has become a common belief that general anesthesia acts reversibly on the central nervous system, which does not cause neuronal damage. However, impairments in cognitive ability are common complications experienced after anesthesia and surgery, particularly in the elderly [1]. The faith in the safety of anesthesia and surgery on the brain has been at odds with the clinical observations. These postoperative neurological complications include any form of acute event (postoperative delirium) and cognitive decline diagnosed up to 30 days after the procedure (delayed neurocognitive recovery) and up to 12 months (postoperative neurocognitive disorder, POCD) [1,2]. Now all forms of cognitive disorder above are referred to collectively as perioperative neurocognitive disorders (PND) [1]. The underlying pathogenesis of PND remains controversial. Growing evidence suggests that a systemic in ammatory reaction induced by surgical trauma may play important roles in PND [2][3][4]. In addition, data from animal and human studies have not been consistent in whether general anesthesia itself is a risk factor for PND [5][6][7]. There are other factors affect the prevalence of PND which include age, the depth of anesthesia, underlying disease, the types of surgery, and comorbidities [8].
Human biomarker studies hold promise for establishing causation and for risk strati cation and monitoring pathogenic processes or outcome of treatment [9,10]. Neuro lament light chain (NFL) has been a promising uid biomarker of brain damage in a wide variety of neurological disorders [11][12][13][14].
Amyloid β (Aβ) and tau are involved in Alzheimer's disease (AD) pathology or neuronal injury, which have been used as biomarkers for identifying the earliest stages of the disease [15]. Hence, in the present study, we limited confounders known to affect the prevalence of PND, and enrolled healthy adult subjects undergoing anesthesia and surgery. Using highly sensitive assays to measure NF-L, Aβ, tau and 21 in ammatory mediators in blood, the goals of our study were to (1) examine and compare levels of plasma biomarkers of neuronal injury, AD and in ammation, (2) examine the potential relationships between the markers of neuronal injury and in ammation.

Study participants
The study was approved by Ethics Committee of Hospital of Stomatology, Sun Yet-sen University, and registered with http://www.chictr.org.cn/index.aspx (ChiCTR1900021289). Fifty adult subjects with the American Society of Anesthesiologists (ASA) status I were enrolled who were undergoing orthognathic surgery.
All participants provided written informed consent, and underwent an evaluation that consisted of medical history, physical and neurological examinations, laboratory tests and neuropsychological assessments. Inclusion and exclusion criteria are provided in Additional le 1. The participants were assigned to receive general anesthesia that consisted of either a volatile agent (sevo urane) or intravenous anesthesia (propofol). The general anesthesia type was chosen according to the preference of the anesthesiologist. For orthognathic surgery, general anesthesia was usually administered in combination with local nerve blocks (Articaine).
Blood samples were taken consecutively before surgical anesthesia (baseline), after anesthesia induction (0 hour), at 3, 24, and 48 hours after surgery, and stored on ice in vacutainer tubes (Becton Dickinson) containing the anticoagulant ethylenediaminetetraacetic acid (EDTA). Within 1 hour, these tubes were centrifuged at 4°C for 10 minutes at 3000 revolutions per minute, and then aliquoted and stored at −80°C for subsequent batch analysis.

Neurology Assays
Plasma NFL, t-tau, Aβ40 and Aβ42 levels were measured on the Quanterix Simoa-HD1 Platform (Quanterix, MA, USA) at GBIO (Hangzhou, China). The multiplex assays Neurology 3-Plex A Kits (Cat No.101995) and NFL Kit (Cat No.102258) were purchased from Quanterix and used according to the manufacture's instruction. Samples were diluted at 1:4 ratio and were performed on single run using 1 batch of reagents. All plasma analyses were performed by skilled laboratory technicians blinded to clinical information.

Statistical analysis
Data were presented as mean (±standard deviation, SD). A repeated measure one-way analysis of variance (ANOVA) was applied to statistically compare longitudinal changes after surgery. The Bonferroni post-hoc test was used to test statistical differences only when the overall test was signi cant. Correlations between variables were performed by Pearson correlation coe cient. A P value of less than 0.05 was taken to indicate signi cance. Tests were performed using SPSS 20.0 version (IBM Corporation, Chicago, IL, USA).

Demographics
A total of 71 subjects were screened. There were 50 participants enrolled in this study. Enrollment is summarized in a summary and ow diagram in Additional le 2 ( Figure S). There were 24 males and 26 females; The mean age of the subjects was 24.80 (4.63) years. Perioperative plasma samples were available for a total of 50 consecutive participants. Baseline and intraoperative characteristics of the study subjects were summarized in Table 1. Patients were euthermic throughout, and none experienced unusual changes in physiology. Procedures were without complications from either the anesthesia or surgery.

Anesthesia and surgery induce neurological damage
The mean plasma NF-L, t-tau, Aβ40 and Aβ42 concentrations were summarized in Table 2. Plasma NF-L showed a signi cant increase at each measurement from a baseline mean (SD) of 3.73 (1.7) pg/mL to a maximal mean (SD) of 22.74 (18.55) pg/mL (P<0.001) at 3 hours postoperatively. (Table 2). Mean changes from baseline were 599% at 3 hours; 562% at 24 hours; and 540% at 48 hours postoperatively. NF-L levels began to decline after 24 hours, but remained higher levels after 48 hours. (Figure 1).
The concentrations of plasma tau were statistically unchanged after anesthesia and surgery ( Table 2).

Discussion
In the present study, we limited confounders such as the ages, underlying disease, the type of surgery and comorbidities known to affect the prevalence of PND, and enrolled healthy adult subjects undergoing anesthesia and surgery. The present results demonstrated that plasma NFL was signi cantly increased, but Aβ42 and Aβ40 were decreased after anesthesia and surgery. The changes of biomarkers of neuronal injury might be attributed to the changes of multiple in ammatory mediators after surgery.
NFL is the major constituents of the neuronal cytoskeleton [16], which is detected in the cerebrospinal uid (CSF) and that increased CSF NFL levels are associated with neuronal injury in some neurodegenerative disorders, such as multiple sclerosis, HIV infection, and Alzheimer disease [17]. Plasma NFL levels are associated with the severity of injured and/or degenerating neurons and correlated highly with CSF levels [18,19]. In the present study, we found the NFL levels in plasma increased rapidly in response to anesthesia and surgery. In general, our results of NFL were in agreement with those reported by Evered et al [20], who measured in elderly surgical patients. In addition, our results con rmed a strong positive correlation between NFL levels and duration of anesthesia/surgery. Since the measurement of serum NFL may be useful to assess the severity of neuronal injury following traumatic brain injury [21] and ischemic stroke [22], PND is believed to be multifactorial and involves age and healthy state, and NFL is sensitive for neuro injury, our results from healthy adult subjects suggest that anesthesia and surgery are responsible for the neurotoxicity, although DiMeglio's study showed that serum NFL concentration did not change signi cantly after cardiac surgery [23].
Neuritic plaques containing Aβ and neuro brillary tangles consisting of tau protein are the primary neuropathologic criteria for AD diagnosis [24]. Both Aβ42 and Aβ40 isoforms in the CSF and/or blood have been used as biomarkers for the identi cation of the earliest stages of the certain forms of AD [9,[25][26][27]]. Evered's study showed that plasma levels of Aβ42 and Aβ40 were signi cantly lower in patients with POCD at 3 months than those without POCD after cardiac surgery [28]. The lower plasma levels of Aβ may indicate premorbid neurodegenerative disease as Aβ deposited selectively in the brain [29]. In the present study, the levels of Aβ42 and Aβ40 decreased over time after surgery, which remain lower levels at 48 hours postoperatively. There are studies showing that the level of plasma Aβ is associated with age [15,30]. Moreover, the rst signs of AD pathology and cognitive decline may occur from around 50-60 years of age [31,32]. The decreased blood Aβ of Evered's results may be involved in age, since the mean age of the patients enrolled was 68.0 years [28]. Our ndings showed lower level of plasma Aβ obtained from young healthy adults, which are unlikely attributed to Aβ accumulation. A recent study showed that exposure to surgery with general anesthesia during adult life did not induce increased Aβ deposition in brain [33]. In addition, Pikwer's study showed that there were no signi cant effects on Aβ in CSF after surgery and anesthesia (5 hours after induction) [34]. These data suggest that surgery and anesthesia may be involved in the complex mechanisms of Aβ metabolism, which remain to be explored.
Tau protein is primarily localized in CNS neurons and contributes to axonal integrity, which is considered as an important biomarker for neurodegenerative disease and brain injury [35]. Preclinical and clinical studies suggest that anesthesia and/or surgery have effect on this biomarker of AD, the changes of which might be associated with PND [36]. Previous studies have showed that blood tau increased rapidly from baseline postoperatively [20], and even remained elevated at 7 days and three months [23] in the elderly. In the present study, we did not detect a difference in levels of tau over time. A similar nding has been reported by Pikwer et al. [34]. They demonstrated that anesthesia and surgery have no effects on tau in adults. These results suggest that tau metabolism may be related to age in that aging brains are vulnerable to anesthesia and surgery.
Surgery and anesthesia can trigger a systemic in ammatory response which is coordinated by the immune system and mediated by endogenous mediators such as in ammatory cytokines [37]. For example, Hirsch's study showed that statistically signi cant changes compared to baseline were present in IL-5, IL-6, IL-8, IL-10, monocyte chemotactic protein (MCP)-1, MIP-1α in plasma of patients undergoing major knee surgery who received spinal anesthesia with intravenous sedation (propofol) [38]. There are other studies showing that the serum levels of cytokines, such as IL-1β, IL-2, IL-6 and IL-8, increased after anesthesia and surgery [39][40][41][42]. In the present study, we observed dynamic changes over time after anesthesia and surgery in speci c plasma in ammatory biomarkers, including IL-6, IL-7, IL-8, IL-10, TNF-α, I-TAC (CXCL11) and MIP-1β. The in ammatory mediators were increased and peaked at 3 hours, and many of the cytokines were not restored baseline at 48 hours postoperatively. IL-10 trajectory matched other proin ammatory biomarkers in plasma, which is an anti-in ammatory cytokine that maintains the balance of the immune response [43]. The data suggests a substantial activation of key pathways of the immune response following surgery and anesthesia. Moreover, we further observed that there were correlations between plasma in ammatory mediators (IL-6, IL-7, IL-8, IL-10) and duration of surgery and anesthesia. These results suggest that plasma levels of IL-6, IL-7, IL-8, IL-10 may be useful markers of the magnitude of surgical stress response to trauma and injury. In general, our results were consistent with previous studies, although there were contradictory results obtained in the pattern and extent of in ammatory response [44], which may be due to different types of surgery, underlying diseases, research methods and age of subjects.
Neuroin ammation has become a key hallmark of neurological complications including PND [4]. Surgery is known to induce a systemic in ammatory response causing blood-brain barrier breakdown and then triggering neuroin ammation [38,45,46]. In the present study, we demonstrated that there was a positive correlation between plasma concentrations of NFL and in ammatory cytokines (IL-6 and IL-8).
The changes of biomarkers of neuronal injury might be attributed to increase of in ammatory cytokines IL-6 and IL-8. These ndings lead to possibility that systemic in ammation might have profound impact on the brain after anesthesia and surgery. Experimental and clinical studies have suggested that different anesthetics may modulate immune signaling pathways, which can directly cause immune suppression by in uencing the functions of immunocompetent cells and in ammatory mediator gene expression and secretion [47,48]. Volatile anesthetics have been thought to have antiin ammatory properties [49,50]. In addition, a latest research showed that general anesthesia (sevo urane) without surgery in healthy volunteers did not provoke an in ammatory state or neuronal injury in the early hours after exposure [10]. Another study in healthy subjects without any surgical intervention or other nociceptive stimuli demonstrated that propofol exerted a partly pro-in ammatory but also slightly anti-in ammatory effect on the immune system [51]. Since anesthetics may have anti-in ammatory effects on the immune system, our results obtained in healthy subjects suggest that systemic in ammation induced by operative trauma might be main culprit causing neuronal injury.

Limitations
Although anesthesia and surgery have been proposed to increase the incidence of PND [7,52], we did not observe a corresponding behavioral phenotype to the changes we measured in a short period. Whether the changes have a prolonged effect on brain and cause cognitive dysfunction remained to be evaluated. Because of anesthesia and the accompanying surgical interventions, we cannot differentiate any adverse effects of the surgical trauma from those speci cally associated with the use of anesthetic agents. Moreover, due to invasiveness and low availability of the testing CSF biomarkers, the biomarkers of systemic in ammation might not be able to evaluate neuroin ammation, although systemic in ammatory processes have been linked to brain homeostasis and brain injury [53]. The causal relationship between the changes of plasma cytokines and neuronal damage will be required to determine. Clinical studies designed properly could elucidate the effects of in ammation on the pathogenesis of PND. The use of animals in research is also essential to further elucidate the underlying cellular mechanisms.

Conclusions
In summary, we limited some confounders which might affect the prevalence of PND in the present study. This is the rst study to report that a neuronal injury has taken place after anesthesia and surgery in the healthy subjects, although the neurotoxicity might sustain in a short term. Moreover, the neuronal damage might be largely driven by many of cytokines levels, which was in uenced by duration of anesthesia and extent of surgical trauma. The data provide highly valuable information in an understanding of neurological damage by anesthesia/surgery and the potential mechanisms of PND. Declarations WF developed and designed the study and wrote the manuscript. WF, LM and ZW had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. WF, LM and ZW contributed equally to this work. WF and ZW performed analyses. XZ and FH prepared the datasets. GS and YH performed necessary experiments. HH and XY provided supervision and direction for the whole study. All authors read and approved the nal manuscript.

Funding
This work was supported partly by the National Natural Science Foundation of China (No. 81771098 and 81541153).

Availability of data and materials
The data supporting the conclusions of this article are included within the article and additional le.

Ethics approval and consent to participate
This study was approved by Ethics Committee of Hospital of Stomatology, Sun Yet-sen University (GHKQ-201812-K2-01) and we followed all appropriate protocols.

Consent for publication
All the co-authors and participants have given their consent for publication.  Values are presented as arithmetic mean (±SD) concentrations (pg/mL) Compare to baseline*P <0.05, **P<0.01. Values are presented as arithmetic mean (±SD) concentrations (pg/mL) Compare to baseline*P <0.05, **P<0.01. Figure 1 The changes in plasma level of neuro lament light, T-tau, Aβ40 and Aβ42 over time after surgery and anesthesia. The line indicates the mean, and the error bars indicate the standard deviation.