What this is
- This research investigates () in elderly patients after elective orthopedic surgery.
- It focuses on identifying serum biomarkers that can predict the occurrence of .
- The study finds that higher preoperative serum levels of () are significantly associated with .
Essence
- Higher preoperative serum levels predict in elderly patients undergoing elective orthopedic surgery. The study identifies as a potential biomarker for assessing delirium risk.
Key takeaways
- The incidence of () was 26.7% among the studied patients. This aligns with existing literature on the prevalence of in orthopedic surgery.
- Advanced age and lower preoperative cognitive scores were significant risk factors for developing . Specifically, older patients with lower Mini-Mental State Examination (MMSE) scores were more likely to experience delirium.
- Preoperative serum levels were significantly higher in patients who developed . This finding suggests that could serve as a reliable biomarker for identifying patients at risk for .
Caveats
- The study's single-center design and relatively small sample size may limit the generalizability of the findings. Larger, multi-center studies are needed to confirm the results.
- The lack of postoperative level measurements restricts understanding of its role in the development of . Future research should consider this aspect.
- Information on perioperative non-steroid anti-inflammatory drug use was not collected, which may influence the study's outcomes and should be addressed in further research.
Definitions
- Postoperative delirium (POD): An acute and fluctuating disturbance in awareness, cognition, and attention occurring after surgery.
- Prostaglandin E2 (PGE2): A lipid compound involved in inflammation that may influence cognitive function and is elevated in certain conditions.
Simplified
Introduction
Postoperative delirium (POD), characterized by an acute and fluctuating disturbance in awareness, cognition and attention, is one of the most common complications following major surgery [1, 2]. Although the incidence of POD in the general surgical population is 2.5–3%, it can afflict 12% to 51% of patients undergoing orthopedic surgery [3]. In high-risk patient groups, the incidence of POD is reported to be 50–70% [4]. Notably, the incidence of POD is significantly associated with increased morbidity and mortality, activities of daily living impairment, and global cognitive decline [5, 6].
The risk factors for POD are multiple and its pathophysiology remains poorly understood. Previous studies have demonstrated that advanced age, high-risk surgical procedures, cardiopulmonary comorbidities, poor functional status, preoperative fragility and cognitive impairment, pain and metabolic derangement might contribute to the occurrence of POD [7–9]. In animal studies, it has been proposed that neuroinflammation, neurohormone alterations, apoptosis, and oxidative stress may be the underlying molecular mechanism of POD [10–12]. As effective strategies exist to prevent the occurrence and development of POD [3, 4], especially in the high-risk population, there is a growing necessity to identify those patients who are vulnerable to POD.
Biomarkers not only indicate a certain pathological state, but also provide information about disease activity and progression. Although several biomarkers are reported to predict POD [10, 13], biomarkers with high sensitivity and specificity are still lacking. Therefore, we conducted this study aiming at identifying novel and reliable biomarkers for the prediction of POD. In the present study, a panel of 16 biochemical parameters in serum samples from elderly patients undergoing elective orthopedic surgery were assessed, including brain derived neurotrophic factor (BDNF), fibroblast growth factor-23 (FGF-23), irisin, C3a, C3, C5a, interleukin-17A (IL-17A), interleukin-33 (IL-33), prostaglandin E2 (PGE2), S100β, glial fibrillary acidic protein (GFAP), neurofilament light (NfL), vascular cell adhesion molecule 1 (VCAM1), E-selectin and matrix metalloprotein 9 (MMP9). Among them, BDNF, FGF-23 and irisin were selected because they are associated with neurocognitive regulation [14–16]. S100β, NfL and GFAP were implicated in neuronal injury [17, 18]. In addition, biochemical parameters associated with complement cascade (C3a, C3, C5a) and inflammatory response (IL-17A, IL-33, E-selectin, MMP9) were involved as well. These pathophysiological changes have been reported to play an essential part in the occurrence and development of POD [3, 4].
Methods
Study overview
Participant flow diagram
Delirium and cognition assessment
Mini-mental state examination (MMSE) was used to assess the preoperative cognitive function of participants on the day before surgery. The Confusion Assessment Method (CAM) was utilized for the evaluation of delirium, which comprises four clinical features: (a) acute onset of cognitive changes with a fluctuating course, (b) inattention, (c) disorganized thinking, and (d) altered level of consciousness. The diagnosis of delirium was based on the presence of (a) and (b) plus either (c) or (d). For baseline delirium assessment, participants enrolled were evaluated by a well-trained anesthesiologist on the day before surgery. For POD diagnosis, participants enrolled were evaluated by the same anesthesiologist twice daily on postoperative days 1–3.
Biochemical measurements
Patients had blood draws in the morning on the day before surgery. The collected serum samples were stored at -80℃ for the later detection. The serum levels of 16 biochemical parameters were measured by enzyme linked immunosorbent assay (ELISA, CUSABIO, Wuhan, China), including BDNF, FGF-23, irisin, C3a, C3, C5a, IL-17A, IL-33, PGE2, S100β, GFAP, NfL, VCAM1, E-selectin and MMP9.
Statistical analysis
This was a secondary analysis of data from a prospective observational study. Statistical analysis was performed using the SPSS version 25.0 software (SPSS, Chicago, IL). Normally distributed quantitative data were presented as mean and SD and were analyzed with independent-sample t-test, while non-normally distributed quantitative data were presented as median and interquartile range (IQR) and were analyzed with Mann–Whitney U test. Categorical data were analyzed with χ2 test or Fisher exact test. The effect size of all significant group differences was calculated with the formula Cohen’s d = (MA- MB) / σ or r = Z /\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\sqrt{N}$$\end{document}N depending on the distribution of the data [19]. The risk factors for POD were evaluated by multivariate logistic regression analysis, in which variables with P values less than 0.05 were included. The prediction of serum biomarkers for POD was analyzed by using receiver operating characteristic (ROC) curve. P < 0.05 was regarded as statistically significant.
Results
Patient characteristics at baseline
| Characteristics | POD group (= 35)n | Non-POD group (= 96)n | valueaP | effect size |
|---|---|---|---|---|
| Sex (n, male/female) | 15/20 | 39/57 | 0.818 | / |
| Age (year) | 77(70 ~ 82) | 72(67 ~ 76) | 0.001 | 0.3 |
| ASA status (I/II/III) | 2/22/11 | 8/73/15 | 0.169 | / |
| BMI (kg/m)2 | 21.88(19.58 ~ 25.40) | 23.50(22.04 ~ 26.38) | 0.058 | / |
| Preoperative MMSE | 22(18 ~ 27) | 26(23 ~ 29) | 0.001 | 0.3 |
| Medical comorbidities (n) | ||||
| Hypertension | 29(82.86%) | 78(81.25%) | 0.833 | / |
| Diabetes mellitus | 13(37.14%) | 30(31.25%) | 0.525 | / |
| Coronary artery disease | 11(31.43%) | 27(28.13%) | 0.712 | / |
| History of stroke | 10(28.57%) | 19(19.79%) | 0.284 | / |
| Chronic pulmonary disease | 6(17.14%) | 16(16.67%) | 0.949 | / |
| Duration of surgery (min) | 85(75 ~ 95) | 80(70 ~ 90) | 0.454 | / |
| Blood loss(ml) | 250(200 ~ 300) | 250(200 ~ 350) | 0.815 | / |
Differences in the levels of preoperative biochemical parameters between patients with and without POD
| Parameters | POD group (= 35)n | Non-POD group (= 96)n | valueaP | effect size |
|---|---|---|---|---|
| BDNF (ng/ml) | 0.43(0.38 ~ 0.54) | 0.48(0.41 ~ 0.63) | 0.109 | / |
| FGF-23 (pg/ml) | 504.07(456.31 ~ 577.63) | 513.83(466.68 ~ 588.38) | 0.894 | / |
| Irisin (ng/ml) | 47.64(18.94 ~ 74.45) | 63.01(26.67 ~ 115.62) | 0.07 | / |
| IL-17A (pg/ml) | 6.00(5.71 ~ 6.61) | 6.02(5.58 ~ 6.69) | 0.32 | / |
| IL-33 (pg/ml) | 11.50(5.35 ~ 33.07) | 16.21(5.13 ~ 185.00) | 0.353 | / |
| C3a (ng/ml) | 504.07(456.31 ~ 577.63) | 513.83(466.68 ~ 588.38) | 0.894 | / |
| C3 (μg/ml) | 264.87(136.56 ~ 297.21) | 207.81(103.38 ~ 271.51) | 0.065 | / |
| C5a (ng/ml) | 102.57(52.33 ~ 165.90) | 85.77(52.71 ~ 141.67) | 0.297 | / |
| iNOS (IU/ml) | 8.81(6.21 ~ 13.6) | 8.54(6.07 ~ 12.16) | 0.274 | / |
| PGE2 (pg/ml) | 181.23(126.08 ~ 235.65) | 61.63(39.96 ~ 105.65) | < 0.001 | 0.61 |
| S100β (pg/ml) | 199.22(183.52 ~ 283.19) | 181.36(170.91 ~ 207.67) | < 0.001 | 0.33 |
| GFAP (ng/ml) | 11.76(7.77 ~ 14.61) | 7.38(5.84 ~ 10.29) | < 0.001 | 0.36 |
| NfL (ng/ml) | 1.47 ± 0.2960 | 1.29 ± 0.2961 | 0.002 | 0.6 |
| VCAM1(ng/ml) | 691.11 ± 263.50 | 652.18 ± 196.83 | 0.364 | / |
| E-selectin (ng/ml) | 2.95(2.14 ~ 3.99) | 2.79(1.98 ~ 4.26) | 0.718 | / |
| MMP9 (ng/ml) | 4.00(1.37 ~ 5.32) | 3.36(1.71 ~ 9.68) | 0.649 | / |
Preoperative biomarker associations with POD
| Variables | OR | 95% CI | valueP |
|---|---|---|---|
| Age | 1.144 | 1.008 ~ 1.298 | 0.037 |
| Preoperative MMSE | 0.922 | 0.810 ~ 1.049 | 0.217 |
| PGE2 | 1.031 | 1.018 ~ 1.044 | < 0.001 |
| S100β | 1.002 | 0.997 ~ 1.008 | 0.456 |
| NfL | 1.003 | 1.000 ~ 1.005 | 0.036 |
| GFAP | 1.021 | 0.967 ~ 1.078 | 0.452 |
The predicative performance of age, NfL and PGE2 for POD
ROC curve analysis of age, preoperative serum NfL and PEC2 levels for distinguishing POD. (= 131). ROC: receiver operating characteristic; NfL: neurofilament light; PGE2: prostaglandin E2; POD: postoperative delirium n
Discussion
Our study showed that preoperative serum level of PGE2 was closely associated with the development of POD in elderly patients undergoing elective orthopedic surgery. Thus, preoperative PGE2 could serve as a new serum biomarker to identify patients at risk for developing POD. To the best of our knowledge, this is the first study to link preoperative PGE2 level with POD following elective orthopedic surgery.
The incidence of POD was 26.7% in our study, which was similar to 27% in orthopedic surgery reported by a systematic review [20], and was also comparable with 20.4% in the aged patients undergoing similar type of surgery in a recent study [21]. It is widely accepted that the occurrence of POD was resulted from the complicated interactions of multiple predisposing factors. Among the known risk factors for POD, advanced age and pre-existing dementia or cognitive impairment have been considered as the major risk factors [3]. Consistently, our study showed advanced age was an independent risk factor of POD, and patients with POD were prone to have poor preoperative cognition as reflected by MMSE. Although delirium is significantly associated with adverse outcomes, fortunately, it is preventable and treatable [3]. Currently, non-pharmacological multicomponent intervention is thought to be the most effective prevention strategy, which mainly includes the Hospital Elder Life Program and proactive old age medicine consultation [22–24]. In addition, the fundamental aim of all preoperative screening instruments, whether scale- or biomarker-based, is to identify those vulnerable patients, who then can benefit from optimized perioperative evaluation, monitoring and management. Thus, identifying high-risk patients is an important step in successfully reducing the risk for POD.
Despite various mechanisms have been implicated in the development of POD [12, 25], neuroinflammation and altered neurotransmitters are thought to be the two leading hypotheses [26]. For neuroinflammation, both preoperative and postoperative elevation of peripheral C-reactive protein and IL-6 were reported to be associated with higher risks of POD [27]. The underlying mechanism may be that peripheral inflammation causes the breakdown of the blood–brain barrier, leading to neuroinflammation, which in turn results in synaptic loss, neuronal death, and neurogenesis damage [28]. On the other hand, studies have showed alteration of neurotransmitters in the process of POD, including acetylcholine, glutamate, gamma-aminobutyric acid, dopamine, and serotonin systems, of which deficient choline or excessive dopamine, or both, are the most frequently linked to delirium [3, 4, 26].
In the present study, we assessed a panel of 16 parameters implicated in inflammatory response, neurotransmitter disturbance, complement cascade, and neuronal injury in serum samples from elderly patients undergoing elective orthopedic surgery. Importantly, we showed that serum level of preoperative PGE2 was significantly increased and appeared to be a potential biomarker of POD after elective orthopedic surgery. Indeed, PGE2 is reported to be an important modulator of inflammation, which is of particular interest due to its relatively high concentrations in the brain compared with other prostaglandins [29]. PGE2 works through four distinct G protein-coupled E-prostanoid receptors, EP1 to EP4, and a growing number of studies suggest that PGE2 exerts neurotoxic effects by binding to EP1, EP2 or EP3 receptors in models of chronic neurodegenerative disease like Alzheimer’s disease (AD) and Parkinson’s disease [29, 30]. In aged mice, blockade of peripheral PGE2 signaling is sufficient to reverse cognitive decline associated with aging [31]. Moreover, the cerebrospinal fluid concentration of PGE2 is elevated in both probable AD patients and in AD patients with memory decline [32, 33]. Strikingly, subsequent analysis of several studies indicated that chronic non-steroid anti-inflammatory drug use was associated with lower risk of AD in normal aging people [29, 34]. However, there are a number of studies suggesting that PGE2 had a role in synaptic plasticity and neuroprotection by binding EP4 receptors [35, 36]. Therefore, the role of PGE2 is complex and whether PGE2 inhibitor can be used as a prevention or treatment target for POD in elderly patients needs further research.
Our study also showed that NfL was also an independent risk factor for POD. As one of the neuroaxonal injury markers, NfL has been reported to be elevated in multiple conditions including stroke, traumatic brain injury, multiple sclerosis, AD and frontotemporal dementia [37]. Recently, a clinical study by Tamara and colleagues showed that patients with the highest preoperative NfL levels were more likely to develop POD, which is consistent with our finding [38]. Additionally, the increased serum levels of S100β and GFAP were observed in POD group. S100β, a calcium binding protein, has different effects during different stages of human development. In the first three months of life, S100β exerts neurotrophic effects on the central nervous system (CNS) [39]. However, elevated S100β levels at other points in life mostly correlate with blood–brain barrier injury, which has been observed in cerebrovascular and neurodegenerative diseases [40, 41]. S100β has been reported as a potential biomarker for delirium, but its particular relationship with POD remains controversial [42, 43]. GFAP, an intermediate filament protein, is mainly expressed in astrocytes and its plasma level has been proposed as a biomarker for nervous system injury in traumatic brain injury and stroke [44, 45]. Overall, these increased preoperative serum levels of S100β and GFAP reflect pre-existing brain function impairment in patients with POD, which makes them more susceptible to brain injury in response to surgical trauma.
A few limitations should be noted in our study. Firstly, despite suitable for identifying biomarker, this clinical study involved a single-center, single-surgery procedure and was conducted in a relatively small sample, which may produce biases unintentionally, thus further studies in larger sample size are required. Moreover, detecting PGE2 levels postoperatively would be more informative to determine the relationship between PGE2 and POD, which merits further investigation. Finally, the information about the perioperative use of non-steroid anti-inflammatory drug was not collected, which might have an effect on our study results and should be considered in future study.
Conclusions
Our study showed that preoperative serum PGE2 levels might be a biomarker to predict the occurrence of POD in elderly patients undergoing elective orthopedic surgery. Meanwhile, this finding may provide a target for possible interventions to prevent POD. However, further studies in different surgical populations are required to confirm our results.