Postoperative cognitive dysfunction (POCD) is a severe complication after surgery. Currently, a complicated battery of neuropsychological tests both before and after surgery with other characteristics-matched population as control are needed for the diagnosis of POCD. This diagnosis is also delayed, which could not be used to screen for high risk patients who may need intervention beforehand. The current trial targeted a surgical population of elderly patients undergoing cardiac surgery under cardiopulmonary bypass (CPB), which is a population of the highest incidence of POCD, to screen for possible predictive or diagnostic biomarkers in the serum for POCD. Myeloid differentiation factor 2 (MD2), also known as lymphocyte antigen 96, is a protein involved in biding lipopolysaccharide with Toll like receptor-4 (TLR4). Recently the investigators have found that increased MD2 expression in the hippocampus of the mice after surgery stimuli. On the other hand, the investigators have reported that cystatin C (CysC) as an endogenous neuroprotective factor for stroke. It may also be involved in endogenous neural protection against POCD. This trial is to investigate whether serum MD2, CysC can be used for prediction and diagnosis of POCD in surgical population. Serum based DNA methylation biomarkers will also be tested for prediction or diagnosis of POCD development. Also in our orevious research, SNPs cites at rs6739405、rs12467815、rs12472215、rs11126727、rs11126731、rs993607 were revealed as possible susceptibility variations for POCD (diagnosed with MMSE only, NCT02084030) in patients undergoing CPB. This study will also test the SNP variations in study populations to varify if one or conbination of morethan one of these varuations can be a risk factor for POCD when diagonosed with NPT.

Postoperative cognitive dysfunction (POCD) is a common complication of surgical patients, which manifested by reduced memory, attention and calculation abilities etc. Among different types of surgeries, the incidence of POCD in cardiac surgical patients are the highest, the prevalence of POCD on the day of discharge was as high as 53%, and 26% of patients still showed cognitive dysfunction three months after surgery. POCD can increase patients' hospitalization days, increase the consumption of medical resources and social security resources, seriously affect patients' postoperative quality of life, and even increase the mortality rate one year after surgery. There is no definite serum marker for the prediction and diagnosis of POCD. The diagnosis currently applied are a series of neuropsychological battery tests both before and after surgery. it's both time consuming and inapplicable to patients in severe conditions. The purpose of this trial is to screen out POCD serum markers as effective indicators for POCD prediction, early diagnosis of POCD, and assessment for therapeutic efficacy.

At present, the pathogenesis of POCD is not clear. Aging, general anesthesia, heart surgery, preoperative cognitive impairment and cardiovascular diseases are high risk factors for POCD. At present, preoperative and postoperative neuropsychological behavior assessments is the only way to diagnose POCD, but it is time consuming, resource occupying, and also has learning effect. However, studies on serum markers that can predict the risk of POCD or for early diagnosis of POCD are still at a very preliminary stage. Study has found that plasma inflammatory factors: interleukin (IL)-1β, IL-8 and tumor necrosis factor-α level raised in POCD patients [1-2]. In addition, in microglial cells, inflammatory reaction mediated by S100β modulated the receptor for advanced glycation end products (RAGE) signaling pathway [3], and raise pro-inflammatory factor expressions [4], so S100β expression may be related to POCD occurrence. However, inflammatory related factors such as IL-1, IL-8 and TNF-α lack neurological or disease specificity. S100B, previously thought to have diagnostic effects of nerve injury, has not been confirmed in small sample correlation studies [5]. Therefore, it is of great clinical significance to collect reliable and powerful clinical research data to explore serum biomarkers for early prediction and diagnosis of POCD.

The investigators have previously found that higher endogenous CystatinC level is an important protective mechanism against ischemic brain injury. Elevated serum Cystatin C, accompanied with brain cystatin C level elevation, can inhibit lysosome damage, promote autophagy [6] and induce ischemic tolerance in the brain. Myeloid differentiation protein (MD-2) is a secreted protein, composed of 160 amino acids and participate in TLR4 signaling pathway, mediated inflammatory response [7]. The investigator using an animal model of POCD also found that the MD2 expression is significantly increased after surgery.

In addition, the methylation is an important modification way of proteins and nucleic acids, it regulates the expression of genes and is closely related to many diseases such as alzheimer's. Studies have shown that COASY and SPINT1[17], NCAPH2/LMF2[18] promoter region DNA methylation has diagnostic value for alzheimer's disease and mild cognitive impairment. Therefore, we speculate that changes in the DNA methylation markers of the central nervous system may be of early diagnostic value for POCD after cardiac surgery. Therefore, the detection of the central nervous system methylation label in the serum of patients after cardiac surgery will provide a new research direction for the clinical detection of central nervous system injury.

Our previous clinical trial (NCT 02084030) indicates that, 6 SNP mutations on the CTNNA2 gene (SNPs cites at rs6739405、rs12467815、rs12472215、rs11126727、rs11126731、rs993607) has altered risk for POCD in elderly patients undergoing CPB. Since this gene functions as a linker between cadherin adhesion receptors and the cytoskeleton to regulate cell-cell adhesion and differentiation in the nervous system, it also regulates morphological plasticity of synapses and cerebellar and hippocampal lamination during development, we aim to verify if patients that has any of these mutations is more susceptible to POCD then patients that has none, or less of these mutations, using a more integrated neuropsychological test battery tests as compared to MMSE used in the previous trial.

In conclusion, the investigators believe that serum central nerveous system (CNS) specific methylation markers, MD2 level and CystatinC level may have predictive and diagnostic value for POCD. This trial is to to evaluate POCD in 200 patients with cardiac surgery under CPB, test their perioperative serum MD2, CystatinC, and DNA methylation markers from the central nervous system, and explore their role in prediction and diagnosis of POCD.

Inclusion Criteria:

• Patients age≥18 yrs scheduled for coronary artery bypass graft (CABG), valve replacement or CABG+valve replacement surgery under CPB and general anesthesia in Xijing Hospital, from Nov, 2018.

Exclusion Criteria:

• Has neurodegenerative disease: dementia, Alzheimer's or Parkinson's Disease
• Has received neuropsychological tests before
• Psychological disorder that needs medication
• Preoperative Mini-Mental State Examination (MMSE)< 24
• Didn't finish elementary school
• Has symptomatic cerebrovascular disease.
• Has received cardiac surgery or neurosurgery before
• Has cardiac arrest experience and received cardiopulmonary resuscitation
• Renal dysfunction (serum creatinine>2 mg/dL or 176.82 μmol/L)
• Hepatic pathology (AST, ALT exceeded 1.5 times of the upper limit of normal range)
• Unable to comply or non-cooperative
• Can't finish process under instruction
• Can't understand mandarin
• Has severe visual or auditorial impairment
• Has severe alcohol or drug dependence (has been drinking over 100 ml of ≥50° alcohol per day , for over 3 months. And other drug abuse problem)
• Has been enrolled in the same study before or is currently involved in other clinical trials.

Patients undergoing cardiac surgery under cardio-pulmonary bypass (CPB) and general anesthesia. Non-surgical controls are age and gender-matched community residents.