Elevated neuron-specific enolase level is associated with postoperative delirium and detection of phosphorylated neurofilament heavy subunit: A prospective observational study.

BACKGROUND: Delirium is the most common central nervous system complication after surgery. Detection of phosphorylated neurofilament heavy subunit in the serum reflects axonal damage within the central cervous system and is associated with the severity of postoperative delirium. Neuron-specific enolase and S100 calcium-binding protein β have been identified as possible serum biomarkers of postoperative delirium. This study examined the association of the levels of these markers with incidence of postoperative delirium and detection of phosphorylated neurofilament heavy subunit.
METHODS: This study represents a post hoc analysis of 117 patients who participated in a prospective observational study of postoperative delirium in patients undergoing cancer surgery. Patients were clinically assessed for development of postoperative delirium within the first five days of surgery. Serum levels of phosphorylated neurofilament heavy subunit, neuron-specific enolase, and S100 calcium-binding protein β levels were measured on postoperative day 3.
RESULTS: Forty-one patients (35%) were clinically diagnosed with postoperative delirium. Neuron-specific enolase level (P < 0.0001) and the proportion of patients positive for phosphorylated neurofilament heavy subunit (P < 0.0001) were significantly higher in the group of patients with postoperative delirium. Neuron-specific enolase level discriminated between patients with and without clinically diagnosed postoperative delirium with significantly high accuracy (area under the curve [AUC], 0.87; 95% confidence interval [CI], 0.79-0.95; P < 0.0001). Neuron-specific enolase level was associated with incidence of postoperative delirium independently of age (adjusted odds ratio, 8.291; 95% Cl, 3.506-33.286; P < 0.0001). The AUC for the serum neuron-specific enolase level in detecting phosphorylated neurofilament heavy subunit was significant (AUC, 0.78; 95% CI, 0.66-0.90; P < 0.0001).
CONCLUSION: Elevated serum neuron-specific enolase was associated with postoperative delirium independent of age as well as detection of phosphorylated neurofilament heavy subunit in serum. Serum neuron-specific enolase and phosphorylated neurofilament heavy subunit might be useful as biomarkers of postoperative delirium. TRIAL REGISTRATION: University Medical Information Network (UMIN) trial ID: UMIN000010329; https://clinicaltrials.gov/.

Introduction

Delirium is the most common complication of the central nervous system (CNS) after surgery. It usually occurs between postoperative days two and five [1]. The incidence of postoperative delirium (POD) ranges between 26 and 52% [2]. POD is associated with increased morbidity and mortality [3], prolonged intensive care unit (ICU) stay [4], and higher cost of hospital stay [5]; therefore, early detection of POD is important. However, a study of delirium assessment in the ICU showed that delirium is frequently not diagnosed when present [6]. Using delirium as diagnosed by a psychiatrist, neurologist, or geriatrician as the standard, the Confusion Assessment Method for the ICU (CAM-ICU) had a sensitivity of 64% and specificity of 82%; with sensitivity and specificity of the Intensive Care Delirium Screening Checklist (ICDSC) was 43% and specificity of 95%, respectively. The sensitivity of the ICU physician’s clinical impression was only 29%. POD can lead to long-term cognitive impairment and brain atrophy even in patients who recover [7,8]. Although the causal relationship between POD and anatomical or functional CNS alterations has not been clarified, mechanism-based diagnostics and measures of POD severity in the early phases are urgently needed.

Surgical and anaesthetic stresses trigger CNS inflammations via microglial activation [9]. CNS inflammation can induce the expression of neuron-specific enolase (NSE) in neurons and S100 calcium-binding protein β (S100β) in astrocytes, which results in loss of blood-brain barrier (BBB) integrity and neurotoxicity [10]. As a consequence, neurons are structurally damaged, which may release phosphorylated neurofilament heavy subunit (pNF-H) into the cerebrospinal fluid (CSF) and peripheral blood [11]. Because pNF-H is a cytoskeletal protein specifically localized within CNS axons, it is not normally detected in the blood. Therefore, unlike NSE, which is released during neuroinflammation regardless of neuronal damage, the detection of serum pNF-H directly reflects structural neuronal damage. Previously, we reported that serum pNF-H level is directly related to clinical severity of POD [12]. In that study, serum pNF-H was detected in more than 65% of patients with POD but in less than 10% of patients without POD. Although the utility of both NSE and S100β as diagnostic markers of POD in the ICU has been investigated, their role in POD pathogenesis remains controversial [13-15]. Furthermore, the association between serum NSE and S100β levels, POD development, and detection of serum pNF-H (i.e., axonal damage) has not been thoroughly investigated.

This study aimed to explore the levels of potential CNS-derived biomarkers of POD in the acute postoperative period. First, we investigated the accuracy of NSE and S100β in POD screening; then, we evaluated the association of the levels of these markers with detection of serum pNF-H.

Methods

Study population

This study represents a post hoc analysis of a prospective observational study conducted at the University of Tokyo Hospital, Saitama Red Cross Hospital, and Tsukuba University Hospital [16]. Details regarding study methodology and patient characteristics may be found in our previous report. Briefly, patients were enrolled and followed up between July 23, 2013 to February 28, 2015. Serum samples from patients who participated in the initial study were collected on postoperative day 3 and stored for later testing. Patients scheduled to undergo cancer surgery under general anasthesia, irrespective of the affected organ, were eligible for inclusion. Among these we enrolled patients with an American Society of Anesthesiologists physical classification score < 4. Exclusion criteria were as follows: (1) patients with pre-existing clinically relevant cognitive dysfunction or neurological disorder; (2) patients who were prescribed tranquillizers that could influence delirium [17]; (3) patients who required neurosurgery for lesions located in the brain and/or spinal cord; and (4) patients who required cardiothoracic surgery with cardiopulmonary bypass. These neurosurgical and cardiothoracic procedures can potentially cause CNS ischemia and a consequent increase in serum pNF-H level. Therefore, patients scheduled to undergo them were excluded.

Ethics

The local ethics committee of each institution approved the trial protocol. Written informed consent was obtained from all participants. The study was registered in the University Medical Information Network (UMIN trial ID: UMIN000010329).

Patient assessment

Patients were assessed for delirium-associated symptoms by the attending nurses at least three times a day during regular ward rounds for the first week after surgery, as described in our previous study [16]. Suspected POD was confirmed by nurses using the CAM-ICU [18] and by investigators using ICDSC [19]. Patients diagnosed with POD within the first five days after surgery were included in the POD group; those not were included in the no POD group.

Measurement of biomarkers

Blood samples were collected on postoperative day 3 and stored at −20°C. Detectio of pNF-H in serum was used as a proxy for CNS axonal damage. Enzyme-linked immunosorbent assay (ELISA) was performed to measure levels of serum pNF-H (Human Phosphorylated Neurofilament H ELISA; BioVendor, Modrice, Czech Republic). The threshold for detection of pNF-H (70.5 ng/mL) was determined according to manufacturer instructions. For the measurement of NSE and S100β, bead-based multiplex assays were performed (Procarta Immunoassay kit, human by request; Panomics Inc., Fremont, CA, USA) according to manufacturer protocol [20,21].

Statistical analysis

Patient characteristics and biomarker levels were compared between groups using the Wilcoxon or Pearson’s chi-square test. To eliminate the influence of age on serum biomarker levels, analysis of covariance (ANCOVA) was performed between groups. To identify independent biomarkers of POD, we conducted receiver operating characteristic (ROC) analyses of serum levels of NSE and S100β in participants with and without POD. NSE and S100β were analysed as dependent variables, with either dichotomous measures of postoperative delirium (i.e., presence or absence) or detection of serum pNF-H as the independent variable. Cut-off values for continuous variables were determined using the Youden index [22,23]. Subsequently,　multiple logistic regression was performed with direct entry of variance using a model based on the log-transformed concentrations of the potential candidate variables. Analyses were performed using JMP Pro 15 software (SAS Institute, Cary, NC, USA) and SPSS software version 22 (IBM Corp, Armonk, NY, USA). P ≤0.05 was considered significant.

Results

Patient characteristics

A total of 119 patients who underwent elective cancer surgery under general anesthesia were eligible for study inclusion, and 117 patients were analyzed (Fig 1). The baseline characteristics of the 117 patients with and without postoperative delirium are presented in Table 1. Age was significantly higher in the patients with POD (P < 0.0001).

Fig 1

Study flowchart.

Table 1

Characteristics of patients grouped according to development of postoperative delirium.

	PODa (n = 41)	no PODa (n = 76)	P value
Age; y	78 (73−81)	67 (58−74)	< 0.0001
Gender; male	24 (58.5%)	40 (52.6%)	0.54
BMIb; kg/m2	22.8 (20.2−24.1)	21.8 (19.8−24.1)	0.32
pNF-Hc positive (≥ 70.5 pg/mL)	23 (56.1%)	7 (9.2%)	< 0.0001

Values are presented as numbers (proportion) or medians (interquartile range).

aPOD, postoperative delirium;

bBMI, body mass index;

cpNF-H, phosphorylated neurofilament heavy subunit.

Association between POD and CNS-derived biomarkers

Forty-one patients (35.0%) were clinically diagnosed with POD (Table 1). The proportion of patients whowere positive for pNF-H (Table 1) and serum levels of pNF-H (coefficient of variation [CV]: no POD, 161.6; POD, 533.1), NSE (CV: no POD, 31.5; POD, 69.7), and S100β (CV: no POD, 94.2; POD, 119.0) were significantly higher in the POD group (Fig 2). After performing ANCOVA, using age as a covariate (S1 Table), the levels of pNF-H, NSE, and S100 β remained significantly higher in the POD group.

Fig 2

Comparison of serum biomarker levels in patients with and without postoperative delirium.

Using a cut-off value of 201.2 ng/mL, the area under the curve (AUC) for serum NSE level in predicting delirium was 0.87 (sensitivity, 0.76; specificity, 1.00; 95% confidence interval [CI], 0.79–0.95; Fig 3). In contrast, serum S100β level exhibited lower correlation using a cut-off value of 305 pg/mL (AUC, 0.74; sensitivity, 0.78; specificity, 0.61; 95% CI, 0.64–0.83). Multivariate logistic regression analysis showed that age (adjusted odds ratio [OR], 1.080; 95% CI, 1.8–1.172; P  =  0.0269) and NSE (adjusted OR, 8.291; 95% Cl, 3.506–33.286; P < 0.0001; Table 2) were associated with POD.

Fig 3

Receiver operating characteristic curves analysis of serum biomarker levels for the diagnosis of postoperative delirium.

NSE, neuron-specific enolase; S100β, S100 calcium-binding protein β.

Table 2

Logistic regression analysis for prediction of postoperative delirium.

	Unadjusted ORa (95% CIb)	P value	Adjusted ORa (95% CIb)	P value
agec; y	1.139 (1.081−1.215)	< 0.0001	1.080 (1.008−1.172)	0.0269
log (pNF-Hd)	1.533 (1.292−1.819)	< 0.0001	1.105 (0.761−1.527)	0.567
log (NSEe)	9.972 (4.667−34.648)	< 0.0001	8.291 (3.506−33.286)	< 0.0001
log (S100βf)	2.042 (1.371−3.210)	0.0002	1.580 (0.842−2.997)	0.1466

aodds ratio;

bconfidential interval.

cOR for age per each additional year of age.

dphosphorylated neurofilament heavy subunit;

eneuron-specific enolase;

fS100 calcium-binding protein β.

Association between the detection of pNF-H and CNS-derived biomarkers

The association between the detection of pNF-H and serum levels of NSE and S100β is shown in Table 3. Serum NSE level was significantly higher in patients with detectable pNF-H. Although serum S100β level was also higher in the patients with detectable pNF-H, the difference was not significant.

Table 3

Association between detection of phosphorylated neurofilament heavy subunit and candidate biomarker levels.

	pNF-Ha positive (n = 30)	pNF-Ha negative (n = 87)	P value
NSEb; ng/mL	529.010 (758.925−763.164)	54.164 (45.144−63.656)	< 0.0001
S100 βc; pg/mL	528.8 (211.9−706.8)	304.7 (143.7−616.7)	0.0789

Values are presented as medians (interquartile range).

aphosphorylated neurofilament heavy subunit;

bNSE: Neuron-specific enolase;

cS100 calcium-binding protein β.

Using a cut-off value of 201.2 ng/mL, the AUC for serum NSE level in predicting detection of pNF-H was 0.78 (sensitivity, 0.73; specificity, 0.90; 95% CI, 0.66–0.90; Fig 4). In contrast, serum S100β level exhibited a lower association (AUC, 0.61; sensitivity, 0.53; specificity, 0.69; 95% CI, 0.49–0.72) using a cut-off value of 502.0 pg/mL/.

Fig 4

Receiver operating characteristic curves analysis of serum biomarker levels for detection of phosphorylated neurofilament heavy subunit.

NSE, neuron-specific enolase; S100β, S100 calcium-binding protein β.

Discussion

This study examined the clinical usefulness of serum CNS-derived biomarkers in diagnosing POD. NSE level was significantly higher in POD patients (Fig 2), and significantly associated with POD independent of age (Table 2). Furthermore, NSE level showed high accuracy for diagnosing POD (Fig 3). Because NSE level was associated with POD independent of age, this marker should be useful in elderly patients. Although both NSE and S100β in serum indicates a loss of BBB integrity, only NSE level was significantly associated with detection of serum pNF-H in our study (Table 3, Fig 4) [10]. S100β is mainly located in astrocytes, whose end-foot processes interacts with the BBB [10]. The reason that serum NSE increases without an accompanying increase in S100β is unclear. In our previous study [16]. we reported P-selection, which is expressed on endothelial cells and involved in recruitment of circulating leukocytes [24], is independently associated with detection of serum pNF-H. Taken together, compared with astrocytes, neurons might be structurally vulnerable against a loss of BBB integrity, However, anticancer agents induce oxidative stress, DNA damage, and dysregulation of neuronal repair processes [25] and serum pNF-H increases in a cumulative dose-dependent manner in breast cancer patients undergoing chemotherapy [26]. Considering that some patients in our study were undergoing chemotherapy at the time of blood sampling, the toxic effect of these drugs on neurons may have affected our results. To clarify the influence of chemotherapy-induced neurotoxicity in future studies, baseline preoperative biomarker levels should be further examined in addition to postoperative day 3 levels.

Serum NSE level was highly associated with POD and detection of pNF-H, which is proxy of for CNS axonal damage, suggesting that NSE can reflect both the onset and severity of POD. Our ROC analysis showed that the serum NSE cut-off values for the diagnosis of POD and detection of pNF-H were same (201.2 ng/mL for both; Figs 3 and 4). These results suggest that POD is associated with the development of inflammation-induced neuronal damage.

This study has several limitations. First, the biological half-lives of serum NSE and S100β (approximately 48 hours each) are shorter than that the pNF-H half-life (approximately 96 hours) [27]. Therefore, we cannot exclude the possibility that a shorter time window may be required to follow the severity of delirium by NSE than pNF-H. Second, this study was a post hoc analysis of a previous prospective observational study [16]. Therefore, the causal relationship between the onset and severity of POD and changes in biomarkers over time should be further investigated in a prospective manner. Third, we excluded patients with clinically relevant pre-existing cognitive dysfunction or neurological disorders because serum pNF-H level increases in certain CNS disorders, such as spinal cord injury, Alzheimer’s disease, and febrile convulsions [28-30], and such dysfunction or disorders may share symptoms and signs with POD. Finally, although NSE was significantly associated with POD independent of age, multiple logistic regression cannot completely control for each variable. Therefore, further investigation of the association between age and levels of CNS-derived biomarkers is warranted.

Conclusion

NSE is an accurate diagnostic biomarker for POD that is highly associated with detection of pNF-H, a proxy for CNS axonal damage. These findings indicate that the timing of neuronal structural changes might be consistent with both neuronal changes and the onset of POD.

Diagnostic accuracy of POD might be increased with early monitoring of serum NSE level in combination with clinical assessment.

Comparison of biomarkers using analysis of covariance with age as a covariate.

(XLSX)

Click here for additional data file.

30 Sep 2021

PONE-D-21-24158Elevation of neuron-specific enolase predicts postoperative delirium accompanying axonal damage: a prospective observational studyPLOS ONE

Dear Dr. Sumitani,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Nov 12 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Aleksandar R. Zivkovic

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Thank you for stating the following financial disclosure:

This study was supported by JSPS KAKENHI (to K. Mietani, Grant Number: 19H03749) and a Health Labour and Science Research Grant for research on chronic pain (to M. Sumitani, Grant Number: H26-Cancer-060).

Please state what role the funders took in the study.  If the funders had no role, please state: "The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript."

If this statement is not correct you must amend it as needed.

Please include this amended Role of Funder statement in your cover letter; we will change the online submission form on your behalf.

3. Thank you for stating the following in the Acknowledgments/ Funding Section of your manuscript:

This study was supported by JSPS KAKENHI (to K. Mietani, Grant Number: 19H03749) and a Health Labour and Science Research Grant for research on chronic pain (to M. Sumitani, Grant Number: H26-Cancer-060).

We note that you have provided additional information within the Acknowledgements Section that is not currently declared in your Funding Statement. Please note that funding information should not appear in the Acknowledgments section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form.

Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows:

This study was supported by JSPS KAKENHI (to K. Mietani, Grant Number: 19H03749) and a Health Labour and Science Research Grant for research on chronic pain (to M. Sumitani, Grant Number: H26-Cancer-060).

Please include your amended statements within your cover letter; we will change the online submission form on your behalf.

4. Thank you for stating the following in the Competing Interests section:

The Department of Pain and Palliative Medical Sciences, where M. Hasegawa-Moriyama works, is sponsored by Shionogi Co., Ltd. (Osaka, Japan) and Nippon Zoki Pharmaceutical Co., Ltd. (Osaka, Japan).

Please confirm that this does not alter your adherence to all PLOS ONE policies on sharing data and materials, by including the following statement: "This does not alter our adherence to  PLOS ONE policies on sharing data and materials.” (as detailed online in our guide for authors http://journals.plos.org/plosone/s/competing-interests).  If there are restrictions on sharing of data and/or materials, please state these. Please note that we cannot proceed with consideration of your article until this information has been declared.

Please include your updated Competing Interests statement in your cover letter; we will change the online submission form on your behalf.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer #1: This study evaluated plasma levels of (3) proteins in patients that underwent surgery for cancer-associated conditions. The proteins of interest (phospho-neurofilament, neuron-specific enolase, and S100) have all been previously associated with delirium, so this study does not particularly add to the literature since they used a very targeted approach, rather than an unbiased, untargeted method to look at different groups of proteins. Suggestions are made to improve the quality of the manuscript.

-Please report the median with the interquartiles and 95% CI for all proteins reported, rather than only the mean +/- stdev. Graphical representation of the distribution of plasma levels comparing control vs delirium groups (including individual data points for each patient sample) would be beneficial to the reader.

-The % coefficient of variation should be reported for all assays performed, distinguishing between technical replicates and biological replicates.

-Please clearly state in the methods section the timeframe for evaluation of biomarker associations with delirium (i.e., delirium development anytime within the week, before or after plasma collection).

-Since the patient population that developed postop delirium was significantly older than the non-delirium group, age should be considered within the analysis.

-Please include an analysis comparing biomarker levels (in plasma collected on day 3) to the delirium severity and/or incidence on that day.

-Some of the statements go too far in their conclusion given the very limited biochemical analysis performed in this study (ex- pg. 12, lines 6-8). Please revise throughout.

Reviewer #2: General comments:

My biggest criticism of this study is that the authors claim, "to explore the changes in potential CNS-derived biomarkers…" yet I did not see any evidence that change has been studied. The best I can tell, this is a cross-sectional study of associations between POD and a handful of biomarkers. The discussion is more measured, but causality still creeps into the discussion.

The authors place a high level of faith in regression models to mitigate the difference in age between the POD positive and negative groups. While it is true that type III effects are calculated after account for the variability accounted for from other covariates, regression may not entirely remove the effect of covariates. For instance, if the effect of age is nonlinear or interacts with the variable of interest, controlling for the main effect will not fully remove the differences between groups.

Otherwise, this seems to be a fairly standard regression analysis. I have questions about how this study fits into the greater body of research since it sounds like associations between POD and biomarkers have been studied, I don't think this addresses pathogenesis. Hopefully other reviewers can address this.

I have other, smaller criticisms that are included in my specific comments below.

Specific comments (my page numbers are based on the PDF build from PLOS since I couldn't find page numbers on every page; it's helpful for reviewers to have page numbers and, better yet, line numbers that are continuous throughout the paper):

1. (abstract, lines 9-10) I think the abstract should at least mention how patients were sampled and, preferably, something about the analyses performed.

2. (p.12, line 13) Something like "Mietani et al" is needed here.

3. (p.12) Where were patients ascertained?

4. (p.13, lines 18-19) Please provide a citation for ROC analysis.

5. (p.14, line 3) Please provide a citation for the Youden index.

6. (p.14, lines 4-7) It's unclear why both bivariate screening (the p<0.1 part) and stepwise selection (lines 166-7) are used. Though, the larger problem is, with the sample sizes you have, stepwise variable selection procedures usually do a poor job of finding the most appropriate model (e.g., https://doi.org/10.1002/sim.3943). Stepwise procedures and any p-value based selection have quite a bit of evidence suggesting that they are suboptimal at selecting the appropriate variables. For a decent summary, see the link above. Generally, it's better to select based on more robust criteria, especially measures which assess the fit of the model, such as BIC, or, better yet, a shrinkage-based estimator such as lasso or lars.

7. (p.16, lines 5-7; p.17 lines 7-11) The sensitivity and specificity estimates should have 95% confidence intervals.

8. (Table 2) What are the units for the OR for age? For instance, is the OR for age per each additional year of age?

9. (Table 2) I suggest including the results for all variables considered instead of only those that are included in the final model.

10. (p.16 line 16; Table 3; p.17, line 10; p.18, line 4, …) In this context, I believe the term "association" would be preferred over "correlation".

Reviewer #3: Assistant professor Mietami and colleagues present results from a prospective observational trial examining the ability of neuromarkers to predict the occurrence of postoperative delirium, as well as the ability of the neuroinflammatory markers S100b and neuron-specific enolase to predict axonal damage assessed by neurofilament heavy subunit.

The article is well written, the methods are sound, and the statistical analyses appear correct - I would like to congratulate the authors for their work.

I have some suggestions for the authors:

1) I would suggest writing that the results originate from a 'prospective observational trial', which I believe is the case. It would facilitate reading, if this was defined in the methods section without the need to search for the reference of the preciously published paper.

2) It is not clear to me, whether the analyses were pre-planned, or whether the present analyses were conducted post hoc. This has implications for the interpretation and applicability of the results, and I think it should be written in the methods section. Further, I believe that it should be mentioned in the limitations section that the results should be considered hypothesis generating, and that the suggested NSE cut-offs should be validated externally in future prospective cohorts.

3) The study included 117 patients. It is unclear, whether the included patients were consecutive or not. If possible, a consort diagram showing in- and exclusions should be provided. As a minimum it should be defined more clearly, how patients were selected i.e. is there a risk of selection bias.

4) You present biomarker levels as picograms per milliliter. In most studies the presented unit of NSE is nanograms per milliliter. Can you provide a reference for the precision of your NSE assay? If the assay has a high enough precision to assess picograms per milliliter it is fine, but if the precision is lower, I would suggest presenting biomarkers as nanograms per milliliter in stead.

5) The NSE values you present seem very high compared to other studies. In resuscitated cardiac arrest, patients usually have NSE values from 20 - 200 nanograms per milliliter, and any value above 100 ng/mL is a strong predictor of very poor neurologic outcome or death (Stammet et al, JACC 2015). In patients undergoing cardiac surgery with cardiopulmonary bypass, NSE values usually range from 4 - 20 ng/mL. Can you comment on, whether the high values are due to the application of a different assay, or any other explanation? Or is there an error regarding the units you present?

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: Yes: Sebastian Wiberg

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

13 Oct 2021

Dr. Aleksandar R. Zivkovic

Academic Editor

PLOS ONE

PONE-D-21-24158

Elevated neuron-specific enolase level is associated with postoperative delirium and detection of phosphorylated neurofilament heavy subunit: A prospective observational study

Dear Dr. Zivkovic

We would like to thank you and the reviewers for assessing our manuscript and providing comments, which have helped us to improve the quality and clarity of our manuscript. Below, we have addressed each of the reviewer’s comments in blue and have highlighted all changes made to the manuscript in red.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

We rewrote our revised manuscript to meet the PLOS ONE style formatting.

2. Thank you for stating the following financial disclosure:

This study was supported by JSPS KAKENHI (to K. Mietani, Grant Number: 19H03749) and a Health Labour and Science Research Grant for research on chronic pain (to M. Sumitani, Grant Number: H26-Cancer-060).

Please state what role the funders took in the study. If the funders had no role, please state: "The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript." If this statement is not correct you must amend it as needed.

Please include this amended Role of Funder statement in your cover letter; we will change the online submission form on your behalf.

We added the statement of financial disclosure and its role to the revised cover letter.

3. Thank you for stating the following in the Acknowledgments/ Funding Section of your manuscript:

This study was supported by JSPS KAKENHI (to K. Mietani, Grant Number: 19H03749) and a Health Labour and Science Research Grant for research on chronic pain (to M. Sumitani, Grant Number: H26-Cancer-060).

We note that you have provided additional information within the Acknowledgements Section that is not currently declared in your Funding Statement. Please note that funding information should not appear in the Acknowledgments section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form.

Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows:

This study was supported by JSPS KAKENHI (to K. Mietani, Grant Number: 19H03749) and a Health Labour and Science Research Grant for research on chronic pain (to M. Sumitani, Grant Number: H26-Cancer-060).

We removed the above funding-related text from the manuscript.

Please include your amended statements within your cover letter; we will change the online submission form on your behalf.

We have no funding status updates to provide.

4. Thank you for stating the following in the Competing Interests section:

The Department of Pain and Palliative Medical Sciences, where M. Hasegawa-Moriyama works, is sponsored by Shionogi Co., Ltd. (Osaka, Japan) and Nippon Zoki Pharmaceutical Co., Ltd. (Osaka, Japan).

Please confirm that this does not alter your adherence to all PLOS ONE policies on sharing data and materials, by including the following statement: "This does not alter our adherence to PLOS ONE policies on sharing data and materials.” (as detailed online in our guide for authors http://journals.plos.org/plosone/s/competing-interests). If there are restrictions on sharing of data and/or materials, please state these. Please note that we cannot proceed with consideration of your article until this information has been declared.

Please include your updated Competing Interests statement in your cover letter; we will change the online submission form on your behalf.

We removed the competing interests statement from our manuscript and added it to the revised cover letter. We declared that this does not alter our adherence to PLOS ONE policies on sharing data and materials.

Reviewers' comments:

Reviewer #1: This study evaluated plasma levels of (3) proteins in patients that underwent surgery for cancer-associated conditions. The proteins of interest (phospho-neurofilament, neuron-specific enolase, and S100) have all been previously associated with delirium, so this study does not particularly add to the literature since they used a very targeted approach, rather than an unbiased, untargeted method to look at different groups of proteins. Suggestions are made to improve the quality of the manuscript.

-Please report the median with the interquartiles and 95% CI for all proteins reported, rather than only the mean +/- stdev. Graphical representation of the distribution of plasma levels comparing control vs delirium groups (including individual data points for each patient sample) would be beneficial to the reader.

Values are presented as medians (interquartile range) in our new Table 1. In addition, we added a graph indicating the serum levels of CNS-derived proteins of pNF-H, NSE, and S100� in the POD and no POD groups to our new Figure 2. Individual data points for each patient sample have been added as suggested.

-The % coefficient of variation should be reported for all assays performed, distinguishing between technical replicates and biological replicates.

We added the % coefficient of variation of pNF-H, NSE, and S100� on Page 10, line 18.

The proportion of patients whowere positive for pNF-H (Table 1) and serum levels of pNF-H (coefficient of variation [CV]: no POD, 161.6; POD, 533.1), NSE (CV: no POD, 31.5; POD, 69.7), and S100� (CV: no POD, 94.2; POD, 119.0) were significantly higher in the POD group (Fig 2).

-Please clearly state in the methods section the timeframe for evaluation of biomarker associations with delirium (i.e., delirium development anytime within the week, before or after plasma collection).

As suggested, we added the following sentence to Page 8, line 5:

Patients diagnosed with POD within the first 5 days after surgery were included in the POD group; those not were included in the no POD group.

-Since the patient population that developed postop delirium was significantly older than the non-delirium group, age should be considered within the analysis.

We included age as a covariate for the analysis in our new Supplementary Table 1.

-Please include an analysis comparing biomarker levels (in plasma collected on day 3) to the delirium severity and/or incidence on that day.

Patients who developed delirium within 5 days after surgery were included in the POD group. Because patient records are preserved only for 5 years, we could not re-evaluate the relationship between onset and day of sumple collection Therefore, we added the following sentences to the limitations section of the discussion on Page 15, line 4

Second, this study was a post hoc analysis of a previous prospective observational study [16]. Therefore, the causal relationship between the onset and severity of POD and changes in biomarkers over time should be further investigated in a prospective manner.

-Some of the statements go too far in their conclusion given the very limited biochemical analysis performed in this study (ex- pg. 12, lines 6-8). Please revise throughout.

The relevant descriptions in the discussion and conclusion were rewritten throughout the manuscript.

Reviewer #2: General comments:

My biggest criticism of this study is that the authors claim, "to explore the changes in potential CNS-derived biomarkers…" yet I did not see any evidence that change has been studied. The best I can tell, this is a cross-sectional study of associations between POD and a handful of biomarkers. The discussion is more measured, but causality still creeps into the discussion.

The authors place a high level of faith in regression models to mitigate the difference in age between the POD positive and negative groups. While it is true that type III effects are calculated after account for the variability accounted for from other covariates, regression may not entirely remove the effect of covariates. For instance, if the effect of age is nonlinear or interacts with the variable of interest, controlling for the main effect will not fully remove the differences between groups.

Page 11, line 1

We performed analysis of covariance with regarding age as a covariate in Supplementary table 1.

After performing ANCOVA using age as a covariate (Supplementary table 1), the level of pNF-H, NSE, and S100� remained significantly higher in the POD group.

In addition, the following sentence was added to the limitations:

Page 15, line 11

Finally, although NSE was statistically associated with POD independent of age, multiple logistic regression cannot completely control for each variable. Therefore, further investigation of the association between age and levels of CNS-derived biomarkers is warranted.

Otherwise, this seems to be a fairly standard regression analysis. I have questions about how this study fits into the greater body of research since it sounds like associations between POD and biomarkers have been studied, I don't think this addresses pathogenesis. Hopefully other reviewers can address this.

As suggested by both reviewers 1 and 2, the possible overstatements concerning biological changes and pathogenesis were rewritten.

I have other, smaller criticisms that are included in my specific comments below.

Specific comments (my page numbers are based on the PDF build from PLOS since I couldn't find page numbers on every page; it's helpful for reviewers to have page numbers and, better yet, line numbers that are continuous throughout the paper):

We added page numbers to the revised manuscript.

1. (abstract, lines 9-10) I think the abstract should at least mention how patients were sampled and, preferably, something about the analyses performed.

We added the following sentences to the Methods of abstract:

Page 3, line 9

This study represents a post hoc analysis of 117 patients who participated in a prospective observational study of postoperative delirium in patients undergoing cancer surgery. Patients were clinically assessed for development of postoperative delirium within the first five days of surgery.

2. (p.12, line 13) Something like "Mietani et al" is needed here.

As suggested, we added “in our previous study” in the following sentence:

Page 13, line 16

In our previous study [16], we reported that P-selection, which is expressed on endothelial cells is involved in recruitment of circulating leukocytes [24], is independently associated with detection of serum pNF-H.

3. (p.12) Where were patients ascertained?

Patients were assessed postoperatively by the nursing staff during the regular ward rounds in the first week after surgery as indicated in the method section. Therefore, the patients were assessed in the ICU initially and then the surgical floor after transfer from the ICU.

4. (p.13, lines 18-19) Please provide a citation for ROC analysis.

The following literature was cited on page 9, line 2.

22. Lin X, Tang J, Liu C, Li X, Cao X, Wang B et al. Cerebrospinal fluid cholinergic biomarkers are associated with postoperative delirium in elderly patients undergoing Total hip/knee replacement: a prospective cohort study. BMC Anesthesiol 2020; 20: 246.

5. (p.14, line 3) Please provide a citation for the Youden index.

The following literature was cited on page 9, line 2.

23. Bantis LE, Nakas CT, Reiser B. Construction of confidence regions in the ROC space after the estimation of the optimal Youden index-based cut-off point. Biom J 2005; 47: 458–472.

6. (p.14, lines 4-7) It's unclear why both bivariate screening (the p<0.1 part) and stepwise selection (lines 166-7) are used. Though, the larger problem is, with the sample sizes you have, stepwise variable selection procedures usually do a poor job of finding the most appropriate model (e.g., https://doi.org/10.1002/sim.3943). Stepwise procedures and any p-value based selection have quite a bit of evidence suggesting that they are suboptimal at selecting the appropriate variables. For a decent summary, see the link above. Generally, it's better to select based on more robust criteria, especially measures which assess the fit of the model, such as BIC, or, better yet, a shrinkage-based estimator such as lasso or lars.

Instead of stepwise variable selection procedures, we used direct entry in our new Table 3.

Page 9, line 2

Subsequently, multiple logistic regression was performed with direct entry of variance using a model based on the log-transformed concentrations of the potential candidate variables.

7. (p.16, lines 5-7; p.17 lines 7-11) The sensitivity and specificity estimates should have 95% confidence intervals.

We have added 95% confidence intervals to the data in the place you have mentioned.

8. (Table 2) What are the units for the OR for age? For instance, is the OR for age per each additional year of age?

Thank you for your comments. The OR for age is per each additional year of age, which has been added to our new Table 2.

9. (Table 2) I suggest including the results for all variables considered instead of only those that are included in the final model.

We have added pNF-H and S100� data to our new Table 2.

10. (p.16 line 16; Table 3; p.17, line 10; p.18, line 4, …) In this context, I believe the term "association" would be preferred over "correlation".

We changed the term “correlation” to “association” throughout the manuscript.

Reviewer #3: Assistant professor Mietami and colleagues present results from a prospective observational trial examining the ability of neuromarkers to predict the occurrence of postoperative delirium, as well as the ability of the neuroinflammatory markers S100b and neuron-specific enolase to predict axonal damage assessed by neurofilament heavy subunit.

The article is well written, the methods are sound, and the statistical analyses appear correct - I would like to congratulate the authors for their work.

I have some suggestions for the authors:

1) I would suggest writing that the results originate from a 'prospective observational trial', which I believe is the case. It would facilitate reading, if this was defined in the methods section without the need to search for the reference of the preciously published paper.

The following sentence was rewritten as suggested.

Page 7, line 3

This study represents a post hoc analysis of a prospective observational study conducted at the University of Tokyo Hospital, Saitama Red Cross Hospital, and Tsukuba University Hospital [16].

2) It is not clear to me, whether the analyses were pre-planned, or whether the present analyses were conducted post hoc. This has implications for the interpretation and applicability of the results, and I think it should be written in the methods section. Further, I believe that it should be mentioned in the limitations section that the results should be considered hypothesis generating, and that the suggested NSE cut-offs should be validated externally in future prospective cohorts.

As noted above in our response to your first suggestion, this was a post hoc analysis of previously obtained data, which is now specified early in the methods section.

Further, I believe that it should be mentioned in the limitations section that the results should be considered hypothesis generating, and that the suggested NSE cut-offs should be validated externally in future prospective cohorts.

Page 15, line 4

Second, this study was a post hoc analysis of a previous prospective observational study [16]. Therefore, the causal relationship between the onset and severity of POD and changes in biomarkers over time should be further investigated in a prospective manner.

Page 15, line 11

Finally, although NSE was statistically associated with POD independent of age, multiple logistic regression cannot completely control for each variable. Therefore, further investigation of the association between age and levels of CNS-derived biomarkers is warranted.

3) The study included 117 patients. It is unclear, whether the included patients were consecutive or not. If possible, a consort diagram showing in- and exclusions should be provided. As a minimum it should be defined more clearly, how patients were selected i.e. is there a risk of selection bias.

We show a study flow chart in our new Figure 1.

4) You present biomarker levels as picograms per milliliter. In most studies the presented unit of NSE is nanograms per milliliter. Can you provide a reference for the precision of your NSE assay? If the assay has a high enough precision to assess picograms per milliliter it is fine, but if the precision is lower, I would suggest presenting biomarkers as nanograms per milliliter in stead.

NSE levels are as nanogram/mL as suggested. A multiplex immunoassay system was used to measure biomarker levels, This system detects protein with higher sensitivity than conventional ELISA.

We added the following references 20 and 21 to the Methods.

20. Menzenbach J, Frede S, Petras J, Guttenthaler V, Kirfel A, Neumann C et al. Perioperative Vascular Biomarker Profiling in Elective Surgery Patients Developing Postoperative Delirium: A Prospective Cohort Study. Biomedicines. 202; 9: 553.

21. Zhang Y., Birru R., Di Y.P. Analysis of Clinical and Biological Samples Using Microsphere-Based Multiplexing Luminex System. Mol. Toxicol. Protoc. 2014; 1105: 43–57.

5) The NSE values you present seem very high compared to other studies. In resuscitated cardiac arrest, patients usually have NSE values from 20 - 200 nanograms per milliliter, and any value above 100 ng/mL is a strong predictor of very poor neurologic outcome or death (Stammet et al, JACC 2015). In patients undergoing cardiac surgery with cardiopulmonary bypass, NSE values usually range from 4 - 20 ng/mL. Can you comment on, whether the high values are due to the application of a different assay, or any other explanation? Or is there an error regarding the units you present?

Please refer to our comments above.

Menzenbach J et al. addresses the limitations of the multiplex system in their report as follows: the use of multiplex arrays that allow for the measurement of a substantial number of biomarkers at once, while using only small serum sample sizes, could also have resulted in less accurate detection of some proteins over others.

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

15 Oct 2021

Elevated neuron-specific enolase  level is associated with  postoperative delirium and detection of phosphorylated neurofilament heavy subunit: A prospective observational study

PONE-D-21-24158R1

Dear Dr. Sumitani,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Aleksandar R. Zivkovic

Academic Editor

PLOS ONE

10 Nov 2021

PONE-D-21-24158R1

Elevated neuron-specific enolase level is associated with postoperative delirium and detection of phosphorylated neurofilament heavy subunit: A prospective observational study

Dear Dr. Sumitani:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Aleksandar R. Zivkovic

Academic Editor

PLOS ONE