Cerebrospinal fluid levels of neuron-specific enolase predict the severity of brain damage in newborns with neonatal hypoxic-ischemic encephalopathy treated with hypothermia.

OBJECTIVES: To investigate whether cerebrospinal fluid levels of neuron-specific enolase (CSF-NSE) during the first 72 hours correlate with other tools used to assess ongoing brain damage, including clinical grading of hypoxic-ischemic encephalopathy (HIE), abnormal patterns in amplitude integrated electroencephalography (aEEG), and magnetic resonance imaging (MRI), as well as with the neurodevelopmental outcomes at two years of age.
MATERIAL AND METHODS: Prospective observational study performed in two hospitals between 2009 and 2011. Forty-three infants diagnosed with HIE within 6 hours of life were included. HIE was severe in 20 infants, moderate in 12, and mild in 11. Infants with moderate-to-severe HIE received whole-body cooling. Both the HIE cohort and a control group of 59 infants with suspected infection underwent measurement of CSF-NSE concentrations at between 12 and 72 hours after birth. aEEG monitoring was started at admission and brain MRI was performed within the first 2 weeks. Neurodevelopment was assessed at 24 months.
RESULTS: The HIE group showed higher levels of CSF-NSE than the control group: median 70 ng/ml (29; 205) vs 10.6 ng/ml (7.7; 12.9); p <0.001. Median levels of CSF-NSE in infants with severe, moderate, and mild HIE were 220.5 ng/ml (120.5; 368.8), 45.5 ng/ml (26, 75.3), and 26 ng/ml (18, 33), respectively. CSF-NSE levels correlated were significantly higher in infants with seizures, abnormal aEEG, or abnormal MRI, compared to those without abnormalities. Infants with an adverse outcome showed higher CSF-NSE levels than those with normal findings (p<0.001), and the most accurate CSF-NSE cutoff level for predicting adverse outcome in the whole cohort was 108 ng/ml and 50ng/ml in surviving infants.
CONCLUSIONS: In the era of hypothermia, CSF-NSE concentrations provides valuable information as a clinical surrogate of the severity of hypoxic-ischemic brain damage, and this information may be predictive of abnormal outcome at two years of age.

Introduction

Despite the advent of therapeutic hypothermia (TH) as the sole specific intervention shown to improve neurodevelopmental outcomes, hypoxic-ischemic encephalopathy (HIE) continues to contribute to major worldwide perinatal mortality and to long-term disability in full-term and near-term newborns [1, 2].

Early and accurate assessment of the severity of brain damage after a perinatal hypoxic–ischemic event remains one of the most difficult challenges in neonatal care. A variety of clinical, neuroimaging, and neurophysiological tools, and combinations of such, are used to predict long-term outcome [3-7]. However, current methods of assessing the risk of brain injury in the newborn have inherent limitations during the first hours of life, and uncertainty regarding the severity of ongoing brain damage and eventual neurological outcome persists during this early period [8, 9].

A biochemical index of brain injury would be highly desirable to increase the reliability of predictions of neurological sequelae after hypoxic-ischaemic brain injury. Several central nervous system-specific molecules have been investigated in blood serum and in cerebrospinal fluid (CSF) as possible quantitative indexes of perinatal brain injury [10-14].

Neuron specific enolase (NSE) is a well-known biomarker of neuronal injury, both in adults and in pediatric patients [14-17]. In newborns with HIE, the concentrations of NSE have been measured in blood serum and in CSF [18-29]. Serum NSE levels in the setting of HIE after asphyxia show vast variability, which entails significant limitations for its use as a viable biomarker [25]. This variability is probably related to the inconsistent and unpredictable blood-brain-barrier permeability, the release of NSE from other extraneural sources in case of multiple organ injury, and confounding technical factors, such as the processing of the sample, the effect of hemolysis, the impact of the measuring method, and the of the reduction of concentrations as a function of storage time [30].

CSF levels of neuron-specific enolase (CSF-NSE) in infants with HIE have been shown to play a more consistent role as a surrogate marker of the extent of brain damage, and they provided good prediction of the outcome in infants with HIE in the pre-hypothermia era [20, 21, 26–28]. TH appears to lower CSF-NSE levels in infants with HIE; however, the predictive value of CSF-NSE for neurodevelopmental impairment at 12 months of age is not affected by cooling [29]. To our knowledge, no study has examined the correlation between CSF-NSE and other markers of ongoing brain damage during the acute phase of HIE, such us abnormal amplitude-integrated electroencephalogram (aEEG) and abnormal findings on cerebral MRI, nor its association with neurodevelopmental outcomes at 24 months in the hypothermia era.

The aim of this prospective cohort study of infants with HIE in the era of therapeutic hypothermia was to determine whether CSF-NSE correlates with the severity of HIE and other neonatal markers of ongoing brain damage, including aEEG and MRI, as well as with neurodevelopment at 2 years of age. This study shows that CSF-NSE within the first 72 hours may be a useful biomarker for estimating ongoing brain damage and that it provides a surrogate outcome measure.

Material and methods

The study population included infants with HIE consecutively born at ≥34 weeks gestational age with a birthweight ≥1800 grams, admitted to two universitary third-level hospitals in Barcelona, Spain, between April 2009 and July 2011. Infants were considered to have HIE if they met the following three criteria: 1) at least one of the following clinical surrogates of hypoxic-ischemic insult: altered fetal heart rate pattern, sentinel event, or labor dystocia; 2) Apgar score ≤5 at 5 and 10 minutes, need for resuscitation, including mask ventilation for more than 10 minutes after birth or endotracheal intubation, or acidosis (pH ≤7.0 and/or base deficit ≥16 mmol/L) in umbilical cord blood or within 60 minutes from birth; and 3) early neonatal encephalopathy defined as a syndrome of neurological dysfunction in the first 6 hours manifested by a subnormal level of consciousness with or without seizures (moderate or severe HIE) or palmary hyperexcitability, tremor, overactive myotatic reflexes, hypersensitivity to stimulation, or startle responses (mild HIE).

Newborns were excluded if a) they presented with congenital abnormalities, b) showed other identifiable etiologies of neurological dysfunction such as infection or genetic disease, or c) parental consent was not granted.

The severity of HIE was assessed immediately after admission, and always before starting TH, by three investigators (AGA, AA, GA); AGA was on call for the evaluation of infants of this project. Encephalopathy was classified as mild, moderate, or severe according to our semi-quantitative score which includes the aEEG findings [31, 32]. Newborns with moderate or severe HIE received whole-body cooling (Techotherm TSmed 200 N or Criticool, MTRE Ltd.) with rectal temperature maintained at 33–34°C for 72 hours, following which they were slowly rewarmed (≤0.5°C per hour). All cooled infants received sedation with fentanyl infusion throught the treatment.

Neuro-specific enolase in cerebrospinal fluid

A CSF sample was obtained by lumbar puncture at between 12 and 72 hours of life. The greater value for CSF-NSE was chosen if two lumbar punctures were performed. The CSF-NSE concentrations were measured blind to clinical data, including the grading of HIE as well as the aEEG traces and the neuroimaging findings.

CSF aliquots of 0.4 ml were distributed in plastic tubes, which were immediately frozen and stored at -80ºC until analysed within 6 months after sampling. NSE concentrations in CSF were determined with an automated chemiluminescence immunoassay NSE Kritpor (Thermo Scientific® NSE Kryptor, BRAHMS GmbH., Germany). The coefficients of variation for interassay and intraassay variability were 2.8 to 6.2% and 2.2 to 9.7%, respectively. The sensitivity of this method was 1 ng/ml.

To compare CSF-NSE levels between patients with HIE and a control group, we included 59 patients without any abnormal neurological signs or confirmed central nervous system infections in whom lumbar puncture was performed as part of an early onset infection workup within 80 hours of life and for whom sufficient CSF was available for NSE analysis as part of a project on biomarkers in CSF.

Amplitude-integrated electroencephalography (aEEG)

aEEG recordings were immediately started at admission in the NICU and continued for at least 24 hours in infants with mild HIE and throughout TH and rewarming in infants with moderate to severe HIE. The aEEG recordings were collected using CFM 6000 (Olympic Medical, Natus Inc, Seattle, WA, USA) or Nicolet One TM System (Viasys Healthcare, San Diego, California, USA) and digitally stored. The traces were blindly assessed by one researcher (GA) attending to the background pattern, cyclicity, and seizure activity. Classification of aEEG background patterns included the following categories: continuous (CNV), discontinous (DCN), low voltage (LV), burst-supression (BS), and inactive or flat (FT) [33]. aEEG background trace was considered severely abnormal if BS, LV, or FT traces were observed. Electrical seizures were also noted.

Neuroimaging studies

All MRI studies were performed in one of the two participating centers (HSJD). Brain magnetic resonance imaging (MRI) was performed using a 1.5 Tesla unit (General Electric) with a specific neonatal head coil. The imaging protocol has been published elsewhere [34]. Minimally, axial and sagittal T1-weighted, axial and coronal T2-weighted, and axial diffusion-weighted images were available for all patients. Some patients underwent an “early” study performed around the fourth day of life, and a “late” study during the second week of life. For infants with only one study, this was considered, while for infants with both an early and a late study, only the score of the late study was considered in establishing the correlation with CSF-NSE values.

Two researchers (TA, AGA) masked to clinical data and CSF-NSE levels reviewed MRIs. Images were scored according to the scheme previously reported by Rutherford et al [35, 36]. This score grades the severity of damage in four brain regions: the posterior limb of internal capsule (PLIC), the basal ganglia and thalami, the white matter, and the cortex. According to the findings, each site except for the PLIC is graded as normal, or as mildly, moderately, or severely affected. The PLIC is graded as normal, equivocal, or absent. Discrepancies in the scoring of the images were discussed and resolved by consensus. Moderate-to-severe injury was defined as moderate-to-severe damage in any of the regions examined. Infants were grouped into 4 patterns of injury: normal, basal ganglia/thalami injury (including basal ganglia/thalami ± cortical injury without white matter injury), watershed pattern (which included moderate to severe white matter injury without any other damage), and global injury (including minimally basal ganglia/thalami and white matter injury).

Neurodevelopmental outcome

Neurodevelopmental assessments were made in all surviving infants at 24 months of age (median 24.3 months, IQR 23.3, 26.7 months, min-max 13–35 months) using the Bayley Scales of Infant and Toddler Development, Third Edition (BSID-III), which contains 3 individual developmental scores: a cognitive composite score, a language composite score (with receptive and expressive scores), and a motor composite score (with gross and fine motor sub-scores). These 3 scales have a mean score of 100 ± 15. A developmental quotient score < 70 (>2 SD below the mean) indicates significant delay, and a score of < 85 (>1 SD below the mean) indicates at least mild-to-moderate delay [37].

Cerebral palsy was defined according to the Surveillance of CP in Europe [38]. Adverse outcome was defined as death, cerebral palsy, or Bayley-III cognitive, language, or motor scores <85.

Statistical analysis

Regarding perinatal, neurological, and outcome findings, qualitative variables were summarized with absolute and relative frequencies, and qualitative variables with median and interquartile range (IQR) or mean (standard deviation). Continuous variables were compared using Mann-Whitney’s U test or Kruskal-Wallis test, as appropriate. Categorical variables were compared using the Chi-squared test or Fisher’s exact test.

CSF-NSE values were log-transformed to approximate a normal distribution and to achieve homogeneity of variance. Then differences between groups were tested by analysis of variance, with Tukey post hoc pairwise analysis. Spearman correlation coefficients (rs) were used to assess the correlation between quantitative variables. Regression models were used to evaluate the relationship between biomarker levels and the presence and severity of HIE (controls, and mild, moderate, and severe HIE), including interaction for the age at which lumbar puncture was performed. In order to estimate the predictive capacity of the CSF-NSE regardless of the degree of HIE and the age at lumbar puncture, the CSF-NSE-outcome association was estimated by multiple regression, adjusting for these covariates. To determine the optimal cutoff value for CSF-NSE, defined as the highest value of sensitivity multiplied by specificity, receiver operating characteristic curve (ROC) analysis for CSF-NSE levels was made. The area under the curve, sensitivity (S), specificity (Sp), positive predictive value (PPV), and negative predictive value (NPV), including 95% confidence intervals (95% CIs) were expressed. All reported P values were 2-sided, and a P value of <0.05 was considered as indicating statistical significance. Statistical analyses were performed using SPSS verson 20 (IBM, Armonk, NY, USA).

We planned to recruit 50 subjects over a 2-year period. Accounting for 10–15% patient loss for analysis due to clinical instability of the patient, haemorrhage, or insufficient CSF, a priori power calculations were based on analysis of data from 40 subjects.This sample size provided at least 90% statistical power (two-sided alpha risk 5%) to detect a minimum expected effect size of 50 points (SD 48) of difference in biomarker levels between patients with adverse neurological outcome and those with non-adverse neurological outcome.

Ethical considerations

Written information was given to the parents and written consent was obtained at the bedside on admission of each infant after explanation of the study and before its onset. The research was conducted according to the Declaration of Helsinki principles and those of the human studies committee that approved the study (Clinic Hospital Protocol Number: HBC/2010/6049).

Results

Fifty-nine consecutive patients with HIE were recruited during the study period but 6 were excluded due to perinatal infection with Listeria monocytogenes (2 infants), multiple hemorrhagic strokes (1 infant), spinal cord injury (1 infant), late admission at 5 days of age (1 infant), or lack of parental informed consent (1 infant). Of the 53 candidate patients, CSF-NSE could not be performed in 10 infants due to clinical instability, haemorrhage, or insufficient CSF.

CSF-NSE was performed at a median age of 48 hours (15, 73) in the group of infants with HIE and 32 hours (24, 58) in the control group; p = 0.344. Compared to controls, patients with HIE had fewer spontaneous deliveries (7% vs 61%), lower cord pH values (6.94 ± 0.16 vs 7.22 ± 0.11), and lower Apgar scores at 5 minutes (5 (3, 6) vs 10 (8, 10)). Birthweight was lower in the HIE group compared to controls: 3004 ± 587 grams vs 3266 ± 489 grams (p = 0.034). No differences were found between the HIE and control groups regarding gestational age or gender.

The main perinatal characteristics of the 43 recruited infants with HIE, according to the severity of encephalopathy, are shown in Table 1. All infants with moderate or severe HIE were cooled, except for 3 infants with severe HIE, due to late transfer beyond 12 hours of life in 2 infants and to critically ill status in one case. Infants with mild HIE were not cooled. MRI was performed in 37 patients, as five patients died before an MR scan could be performed (4 severe and one moderate HIE). Both aEEG background pattern and MRI findings were associated with severity of HIE (Table 1). Brain MRI was performed at a median age of 10 days; IQR 4.4, 13.9.

Table 1

Perinatal, neurological and outcome findings of the 43 HIE infants enrolled in the study.

	Total HIE N = 43	Mild HIE N = 11	Moderate HIE N = 12	Severe HIE N = 20	P value
Gestational age, weeks	38 (37, 40)	38 (37, 40)	39 (37, 41)	38 (36, 40)	0.685
Weight, grams	3000 (2530, 3500)	3160 (2600, 3700)	3115 (2423, 3498)	2835 (2500, 3355)	0.346
Sex (female)	18/43 (42)	4/11 (36)	5/12 (42)	9/20 (45)	0.897
Intrauterine growth retardation	4/43 (9)	0/11	2/12 (17)	2/20 (10)	-
Sentinel event	16/43 (37)	4/11 (36)	4/12 (33)	8/20 (40)	0.929
Spontaneous delivery	3/43 (7)	1/11 (9)	1/12 (8)	1/20 (5)	-
Umbilical cord pHa	6.97 (6.81, 7.03)	7.00 (6.99, 7.05)	6.90 (6.79, 7.00)	6.94 (6.81, 7.06)	0.190
Apgar score at 5 minutes	5 (3, 6)	6 (4, 7)	5.5 (4, 7)	4 (2, 5)	0.051
Advanced resuscitationb	31/43 (72)	6/11 (55)	8/12 (67)	17/20 (85)	-
Electrical seizures during TH	24/43 (56)	0/11 (0)	6/12 (50)	18/20 (90)	<0.001
Abnormal aEEG BGP in first 6 hours of life	20/39 (51)	0/8 (0)	6/12 (50)	14/19 (74)	<0.001
Abnormal aEEG BGP in first 80 hours of lifec	24/43 (56)	0/11 (0)	6/12 (50)	18/20 (90)	<0.001
MRI Score	2.5 (0, 9.3)	0 (0, 1)	0 (0, 4)	9.5 (6.5, 11)	<0.001
Moderate-severe injury (MRI)	20/38 (51)	1/11 (9)	3/11 (27)	16/16 (100)	<0.001
Global injury pattern (MRI)	17/38 (45)	0/11 (0)	2/11 (18)	15/16 (94)	<0.001
Death	14/43 (33)	0/11	1/12 (8)	13/20 (65)	<0.001
Adverse outcomec	22/42 (52)	0/11	3/11 (27)	19/20 (95)	<0.001
Cerebral palsy	5/28 (18)	0/11	1/10 (10)	4/7 (57)	0.006
CP and/or any abnormal BSID-III score < 85d	8/28 (29)	0/11	2/10 (20)	6/7 (86)	<0.001
CP and/or any abnormal BSID-III score < 70d	5/28 (18)	0/11	0/10	3/7 (43)	0.006
BSID-III Score (Motor)e	94 (83, 99)	99 (94, 103)	96 (90, 98)	79 (45, 88)	0.009
BSID-III Score (Cognitive)e	98 (86, 105)	105 (95, 105)	105 (90, 110)	80 (45, 85)	0.006
BSID-III Score (Language)e	91 (83, 100)	97 (91, 100)	91 (88, 103)	77 (45, 83)	0.019

Quantitative variables are expressed as median (interquartile range), and categorical values are expressed as n/N (%). Continuous variables were compared using Kruskal-Wallis test. Categorical variables were compared using the Chi-squared test. P value of <0.05 was considered as indicating statistical significance. aEEG: amplitude integrated electroencephalogram; BGP: background pattern; BSID-III: Bayley Scales of Infant and Toddler Development Third Edition; CP: cerebral palsy; MRI: magnetic resonance imaging; TH: therapeutic hypothermia.

aData were not available in 1, 2, and 3 infants with mild, moderate, and severe HIE, respectively.

bAdvanced resuscitation: tracheal intubation, chest compressions and/or adrenaline.

cAdverse outcome: death or cerebral palsy and/or BSID-III score <85.

dOutcome in 28/29 surviving infants.

eBSID-III test was performed in 24 out of 29 surviving infants

Fourteen infants (33%) did not survive the neonatal period, one with moderate HIE and 13 with severe HIE. In the only patient with moderate HIE who died, death was due to severe coagulopathy. In the 13 infants with severe HIE, death was preceded by an end-of-life decision (EoL). The EoL decision was considered for those patients with persistent severe encephalopathy (coma) in combination with persistent severe altered aEEG and severe neuroimaging findings (brain ultrasound scans and/or MRI). CSF-NSE levels were not considered in making an end-of-life decision.

Of the 29 neonates who survived, 24 were evaluated with BSID-III; additionally, 4 were contacted by phone, with parents reporting that they had a normal outcome. One infant were lost to follow-up. Of the overall cohort, 22 infants (54%) had adverse outcomes, and 8 infants out of the 29 neonates who survived had CP and/or BSID-III Score < 85. Table 1 shows the outcome results according to the severity of HIE.

CSF-NSE and neurological variables

The HIE group showed significantly higher levels of CSF-NSE than the control group: median 70 ng/ml (29; 205) vs 10.6 (7.7; 12.9); p <0.001. According to the severity of HIE, all subgroups presented higher CSF-NSE levels than the control population (p<0.001). Infants with severe HIE had higher levels of CSF-NSE than those with moderate HIE (p<0.001), and infants with moderate HIE had higher levels than those with mild HIE (p = 0.024). CSF-NSE levels in infants with severe, moderate, and mild HIE were: 220.5 ng/ml (120.5; 368.8), 45.5 ng/ml (26, 75.3) and 26 ng/ml (18, 33), respectively (Fig 1). All severe cases had CSF-NSE levels above 109 ng/ml except for two infants with levels of 50 ng/ml and 70 ng/ml, respectively. All mild cases had a level below 38 ng/ml.

Fig 1

Cerebrospinal fluid levels of neuron-specific enolase in 59 controls and 43 infants with HIE, according to the severity of the encephalopathy.

CSF-NSE levels in the control group were 10.6 (7.7; 12.9). CSF-NSE levels in 20, 12, and 11 infants with severe, moderate, and mild HIE, respectively, were 220.5 ng/ml (120.5; 368.8), 45.5 ng/ml (26, 75.3), and 26 ng/ml (18, 33). No statistically significant differences were found between any of the grades of HIE compared to controls or between any of the grades of HIE (analysis of variance after log-transformation of CSF-NSE values, with Tukey post hoc pairwise analysis); p<0.05. HIE: hypoxic-ischaemic encephalopathy; CSF-NSE: cerebrospinal fluid–neuron specific enolase.

Regression analysis showed that age at lumbar puncture did not influence the association between CSF-NSE levels and the presence or severity of HIE.

CSF-NSE concentrations were higher in infants with abnormal neurological findings including seizures, aEEG background activity, and MRI, compared to those without abnormalities (p<0.001) (Table 2).

Table 2

Cerebrospinal fluid levels of neuron-specific enolase of 43 infants with HIE according to their neurological findings and outcomes.

Variables	Yes		No		P value
	n	CSF-NSE (ng/ml)	n	CSF-NSE (ng/ml)
Moderate or severe encephalopathy	32	116.5 (56, 267.8)	11	26 (18, 33)	<0.001
Electrical seizures during TH	24	133 (71.8, 267.8)	19	28 (21, 37)	<0.001
Abnormal aEEG BGP in first 6 hours of life	20	133 (77.8, 368.8)	19	33 (25, 59)	<0.001
Abnormal aEEG BGP in first 80 hours of life	24	183.5 (87,5; 294.5)	19	32 (23; 50)	<0.001
Moderate-severe injury (MRI)	20	133 (72.5, 267.8)	18	28.5 (20.8, 36.3)	<0.001
Global injury pattern (MRI)	17	166 (96.5, 274.5)	21	29 (22, 46)	<0.001
Cerebral palsy	5	268 (130; 290)	23	33 (23; 59)	0.002
CP and/or BSID-III Score < 85a	8	157.5 (56, 277.8)	20	28.5 (21.5, 37)	0.001
CP and/or BSID-III Score < 70a	5	268 (130, 290)	23	33 (23, 59)	0.002
Adverse outcomeb	22	203 (112.3; 322.3)	20	29 (21.5; 36.8)	<0.001
Death	14	218.5 (121.5, 483)	29	34 (25, 73.5)	<0.001

CSF-NSE values are expressed as median (interquartile range). Mann-Whitney U test was used to analyze the differences in CSF-NSE levels between groups. P value of <0.05 was considered as indicating statistical significance. aEEG: amplitude integrated electroencephalogram; BGP: background pattern; BSID-III: Bayley Scales of Infant and Toddler Development Third Edition; CP: cerebral palsy; MRI: magnetic resonance imaging; TH: therapeutic hypothermia.

aOutcome in 28/29 surviving infants

bAdverse outcome: death or cerebral palsy and/or BSID-III score <85.

Regarding MRI injury, the 20/38 infants that had moderate-to-severe injury had higher CSF-NSE levels compared to the infants who did not have injury (15/38) or those with mild injury (3/38): 133 (73; 268) vs 28.5 (20.8; 36.3) (p<0.001). The pattern of damage on MRI was predominantly global (17/38), and only 2 and 3 infants out of the 38 infants with MRI had solely basal ganglia/thalamus and watershed injury, respectively. Infants with global injury had higher CSF-NSE levels: 166 (96.5, 275) vs 29 (22, 46) (p<0.001). CSF-NSE levels showed a positive correlation with the MRI score: rs 0.824, p<0.001.

CSF-NSE and outcome

Infants with an adverse outcome showed higher CSF-NSE levels than those with normal findings (p< 0.001) (Fig 2 and Table 2). ROC analysis revealed a CSF-NSE cutoff value of maximum predictive accuracy of 108 mg/dL (AUC 0.97, S 0.86, Sp 1.0, PPV 1.0, NPV 0.87) for prediction of adverse outcome. Among surviving infants, CSF-NSE optimal cutoff-point to predict CP and/or BSID-III score <85 was 50 mg/dL (AUC 0.93, S 1.0, Sp 0.8, PPV 0.67, NPV 1.0) (Table 3).

Fig 2

Cerebrospinal fluid levels of neuron-specific enolase of 43 patients according to their neurological findings and outcomes.

Each CSF-NSE value represents a square-shaped infant, trigulum, or circle according to the severity of HIE (mild, moderate, or severe). Filled shapes represent infants with adverse ouctomes (death or cerebral palsy and/or BSID-III score < 85). Empty dots represent those infants without adverse outcome. The figure above shows the entire cohort of 43 children, and the one below shows the 29 surviving infants. Two infants with very severe HIE with extreme CSF-NSE values (2855 ng/ml and 1490 ng/ml) and one infant lost in the follow-up are not represented in the figures. The dashed line represents the cutoff value of 108 mg/dL and 50 mg/dl for adverse outcome, according to the maximum predictive accuracy in the receiver operating curve analysis, for the total cohort (figure above) and only the surviving infants (figure below), respectively. HIE: hypoxic-ischaemic encephalopathy; CSF-NSE: cerebrospinal fluid–neuron specific enolase; aEEG: amplitude integrated electroencephalogram; BGP: background pattern; BSID-III: Bayley Scales of Infant and Toddler Development Third Edition; CP: cerebral palsy; MRI: magnetic resonance imaging.

Table 3

Receiver operating characteristic curve analysis of cerebrospinal fluid levels of neuron-specific enolase and neurological findings.

Variable	N	Cutoff-point	Area under the ROC (95%CI)	Sensitivity (95%CI)	Specificity (95%CI)	PPV (95%CI)	NPV (95%CI)
Moderate-to-severe HIE	32	50	0.91 (0.83,0.99)	0.81 (0.64,0.93)	1.00 (0.72,-)	1.00 (0.86,1.00)	0.65 (0.42,-)
Severe HIE	20	110	0.98 (0.95,1.00)	0.90 (0.68,0.98)	1.00 (0.85,-)	1.0 (0.82,1.00)	0.92 (0.73,-)
Electrical seizures during TH	24	50	0.85 (0.71,0.99)	0.92 (0.73,0.99)	0.79 (0.54,0.94)	0.85 (0.64,0.98)	0.88 (0.65,0.97)
Abnormal aEEG BGP in first 6 hours of life	20	77	0.87 (0.75,0.99)	0.80 (0.56,0.94)	0.89 (0.67,0.99)	0.89 (0.65,0.97)	0.80 (0.58,0.97)
Abnormal aEEG BGP in first 80 hours of life	24	77	0.92 (0.83,1)	0.83 (0.63,0.95)	0.95 (0.74,0.99)	0.95 (0.76,0.99)	0.82 (0.60,0.99)
Moderate-severe injury (MRI)	20	80	0.93 (0.85,1.00)	0.75 (0.51,0.91)	1.00 (0.81,-)	1.00 (0.79,1.00)	0.78 (0.55,-)
Global injury pattern (MRI)	17	80	0.95 (0.88,1.00)	0.82 (0.57,0.96)	0.95 (0.76,0.99)	0.93 (0.69,0.99)	0.87 (0.65,0.99)
Cerebral palsy	5	205	0.95 (0.84,1.00)	0.80 (0.28,0.99)	1.00 (0.85;-)	1.00 (0.50,1.00)	0.96 (0.69,-)
CP and/or BSID-III Score < 85a	8	50	0.93 (0.83,1.00)	1.00 (0.63,-)	0.80 (0.56,0.94)	0.67 (0.39,-)	1.00 (0.77,1.00)
CP and/or BSID-III Score < 70a	5	205	0.95 (0.84.1.00)	0.80 (0.28,0.99)	1.00 (0.85,-)	1.00 (0.49,1.00)	0.96 (0.69,-)
Adverse outcomeb	22	108	0.97 (0.93,1.00)	0.86 (0.65,0.97)	1.0 (0.83,-)	1.0 (0.82,1.0)	0.86 (0.66,-)
Death	14	108	0.91 (0.83,0.99)	1.00 (0.77,-)	0.83 (0.64,0.94)	0.74 (0.51,-)	1.00 (0.85,1.00)

aEEG: amplitude integrated electroencephalogram; BGP: background pattern; BSID-III: Bayley Scales of Infant and Toddler Development Third Edition; CI: confidence interval; CP: cerebral palsy; MRI: magnetic resonance imaging; NPV: negative predictive value; PPV: positive predictive value; ROC: Receiver operating curve; TH: therapeutic hypothermia.

aOutcome in 28/29 surviving infants.

bAdverse outcome: death or cerebral palsy and/or BSID-III score <85. CSF-NSE values are expressed in ng/ml

CSF-NSE levels showed a moderate negative correlation with the three BSDI-III domains (motor, cognitive, and language composite scores): rs -0.650, p = 0.001; rs -0.641, p = 0.001; and rs -0.604, p = 0.002 respectively.

Regression analysis showed that age at lumbar puncture nor the severity of HIE did not interact with the association between CSF-NSE level and outcome variables.

The predictive values for the neurological variables included in the study (severity of HIE, aEEG background pattern, MRI findings, and the CSF-NSE cutoff-point of 108 ng/ml) are shown in Table 4.

Table 4

Predictive values of the different neurological tools related to the outcome.

Adverse outcome	Sensitivity	Specificity	PPV	NPV	Odds Ratio (95%CI)
Severe HIE	0.86	0.95	0.95	0.86	120.3 (11.5;1262.8)
Moderate-to-severe	1.00	0.55	0.71	1.00	54.5 (2.9;1021.5)
Abnormal aEEG BGP in first 80 hours of life	0.86	0.75	0.79	0.83	19 (3.9;92.6)
Moderate-severe injury (MRI)	0.94	0.85	0.84	0.94	90.7 (8.5;964.0)
Global injury pattern (MRI)	0.82	0.9	0.88	0.86	42 (6.2,286.7)
CSF-NSE > 108 ng/ml	0.82	1.00	1.00	0.83	168.6 (8.5,3347.3)

aEEG: amplitude integrated electroencephalogram; BGP: background pattern; CI: confidence interval; MRI: magnetic resonance imaging; NPV: negative predictive value; PPV: positive predictive value; ROC: Receiver operating curve; TH: therapeutic hypothermia.

aAdverse outcome: death or cerebral palsy and/or BSID-III score <85.

Discussion

The study shows that CSF-NSE within the first 72 hours may be a useful biomarker for estimating ongoing brain damage, based on its correlation with the clinical grading of encephalopathy, aEEG tracings, and MRI findings. More importantly, NSE concentrations in CSF can provide a surrogate outcome measure.

Several methods have been used for early estimation of the cerebral damage after hypoxic-ischemic injury and prediction of outcome. The most readily available tools include clinical grading of the encephalopathy, electrophysiological information such us assessment of electrical activity by means of an amplitude-integrated EEG and evoked responses, neuroimaging, and cerebral blood flow studies [7, 9, 39–41].

However, all these evaluations have some limitations during the first 72 hours. During this time frame, the level of alertness is often confounded by the use of sedatives during cooling and/or antiepileptic drugs if clinical or electrical seizures occur, rendering clinical evaluation less reliable [42, 43]. In addition, qualitative evaluation of significant brain injury on both diffusion-weighted and conventional MRI during the first 48h may be underestimated [2, 43, 44]. Other quantitative techniques such as quantitative diffusion parameters and proton magnetic resonance spectroscopy have been used to overcome these limitations [45]. However, in clinical practice their use in infants with perinatal HIE during this narrow short time frame is limited. Sick infants may not tolerate transfer to radiology rooms, and early MRI is challenging as it requires specialized interpretive expertise.

Amplitude-integrated EEG background patterns in neonatal encephalopathy correlate closely with neurological outcome [46]. However, their predictive value depends on how long the background patterns are altered in relation to the hypoxic-ischemic insult, and the greater predictive value for adverse outcome peaks at 48–72 hours of life in cooled infants [41].

Despite the limitations in accurate assessment of brain damage during the first 72 hours after the hypoxic-ischemic insult, this challenge has to be addressed in order to provide early neuroevolutive prognostic information to parents [9].

The concept of “window of opportunity” is used in severe brain injury to indicate as the period of time in which a critically ill neonate might die after removal from the ventilator. After this window, the infant would no longer be dependent on such treatment and would probably survive with subjectively intolerable disability. This window of opportunity is a determining factor in the urgency of having neurological outcome prognostic indicators during the first 72 hours [47].

To improve early identification of infants at risk of neuroevolutive sequelae, reliable early biomarkers of brain damage are a valuable clinical tool to improve accurate identification of infants at risk after a hypoxic-ischemic insult.

Neuron specific enolase (NSE) is a well-known biochemical marker of neuronal injury in adults [16, 17], and pediatric patients [14, 15]. NSE is stable in biological fluids, it originates predominantly in the cytoplasm of neuroendocrine cells and neurons, and its measurement does not elicit immunologic cross-reactivity with non-neuronal enolase.

For clinical purposes, measurement of this potentially sensitive biomarker for brain damage in peripheral blood may have advantages compared to CSF. Blood sampling is less invasive, and can be performed more often, even when the patient is unstable. However, serum concentration both in the pre-hypothermia era and in cooled infants has shown inconsistent results in relation to the severity of brain injury assessed by clinical grading of HIE, neuroimaging, and outcome beyond 12 months of age [18-25]. There are several possible explanations for these controversial results. First, brain injury does not necessarily result in consistent, predictable blood-brain-barrier disruption. Second, increased serum NSE is not necessarily of CNS origin. Several tissues, including muscle, kidney, heart, adipose tissue, adrenal glands, lymphocytes, red blood cells, and platelets contain significant amounts of NSE [48-52]. A hypoxic-ischemic event can result in multiple organ injury, including all those tissues and disseminated intravascular coagulation. Finally, technical confounders must be taken into account. These include the processing of the sample, the effects of the time on the sampling stored since after 6 months at -80ºC there might be a significant decrease in concentrations in CSF, and, particularly, the effect of hemolysis [53]. Moreover, the choice of measuring methods may have influence on the levels of NSE [54], so care should be taken in using our specified cut-off level of NSE for determination of prognosis if another NSE determination method was used.

The main advantages of CSF sampling is the fact that the brain is in direct contact with the CSF and the damaged brain tissue releases high quantities of proteins into this fluid. In addition, there are no sources of NSE other than neurons [30]. Hence, the consistent value of CSF-NSE levels in the first 96 hours of life seen in the prehypothemia era are not surprising, and this biomarker reflects brain damage in accordance with the clinical severity of HIE, as do MRI and aEEG findings [14, 20, 26, 28].

To our knowledge only one study has been published on CSF-NSE that included cooled infants with HIE, but this study did not examine the relation between CSF-NSE concentrations and aEEG and MRI findings as surrogates for the severity of hypoxic-ischemic event [29]. We examined the relationship between CSF-NSE and relevant clinical variables (seizures), in addition to aEEG, MRI, and neurodevelopmental outcomes. Our results are consistent with the previously reported value of the CSF-NSE to identify severe brain damage after sudden and unexpected postnatal collapse [55], and with the relationship between CSF-NSE levels and infarction volume and topology, as well as neurodevelopment at 2 years of age in infants with neonatal arterial ischemic stroke [56].

One strength of our study is that it offers cutoff points that are relevant in clinical practice. CSF-NSE levels ≥ 108 ng/ml showed the greatest specificity and positive predictive value for identifying infants at increased risk of adverse outcome. Considering only those who survived, optimal cutoff point to predict abnormal outcome was 50 ng/ml. CSF-NSE levels may be particularly useful when additional information is needed about brain injury and prognosis, especially if there are inconsistencies between prognostic tools, such as neurological examination and neurophysiological studies, and also if an end-of-life decision is considered within the first 72 hours of life [8].

Biochemical markers of brain injury, such as NSE, are released in a time sequence related to brain damage. In animal models of focal and global ischemia, an early peak of serum NSE levels reflects the ischemia-induced cytoplasmic loss of NSE in neurons that is detectable before irreversible neuronal damage occurs [57, 58]. This has also been corroborated in human adults with stroke, where, after an initial rise within 3 hours, NSE concentrations in blood serum decrease followed by a secondary increase until day 5, with concentrations after 24 hours reflecting the volume of infarcted brain areas [59, 60].

Several studies have determined serum NSE at different times in the first days of life. These studies have shown that serum NSE concentrations are relatively stable and higher during the first 24 hours after birth, decreasing thereafter [24, 61]. It is not known how soon NSE can be detected in the CSF after brain injury, nor are the kinetics of its release and clearance in this biological fluid well understood. Therefore, the time point of maximum diagnostic efficacy is unknown [49]. The scheduling of CSF sampling is generally decided arbitrarily, but in most studies lumbar puncture has been performed within the first 72 hours of life [21, 26, 28, 62]. Despite a range of 12–80 hours in our study, we did not find that the age at lumbar puncture was an interaction factor between the CSF-NSE levels and the variables analyzed. Whether serial measurements might provide indirect information regarding ongoing neuronal injury and might therefore help increase accuracy in predicting patient outcome is unknown. However, lumbar puncture is an interventional procedure, and it is not free from risks. It should not be performed when the infant is unstable or suffers from significant coagulopathy, which makes it difficult to obtain serial samples.

Our determination of CSF-NSE concentrations at a median age of 48 hours after birth may mirror irreversible neuronal damage as it is correlated with surrogate biomarkers of brain damage and the outcomes in this study. The reliability of CSF-NSE to predict subsequent adverse neurological outcomes was similar to the reliability of aEEG and neuroimaging. Therefore, CSF-NSE testing is suggested as both a valuable diagnostic tool in clinical management of the infant with HIE and a prognostic parameter during the course of the encephalopathy. In addition to helping to predict neurodevelopmental outcomes, biochemical brain damage biomarkers such as CSF-NSE should gain momentum in clinical research on HIE. We feel that they can be particularly useful to identify newborns at 48 hours who are at risk of adverse prognosis and potential candidates to receive adjuvant neuroprotective therapies to promote neuronal regeneration and structural recovery from injury during the tertiary phase of HIE (as is the case with stem cell transplantation and anti-inflammatory agents and gangliosides). Furthermore, this biochemical marker could be useful for rigorous monitoring of therapeutic interventions aimed at ameliorating brain damage.

Our study has some limitations. Like many longitudinal studies on neurobiological markers that may reflect the severity and progression of brain injury, the inability to obtain or analyze some CSF samples was a concern. Given the low incidence and prevalence of HIE, one for every 1000 births in our setting [63], the number of recruited infants was not huge. Larger validation studies would be advisable. Our study was not powered to evaluate death or disability separately, and since most infants with severe damage die, larger studies are needed to assess the relationship between biomarkers and outcome, independent of mortality. This is especially important for the group of infants with moderate HIE—the group with the greatest prognostic uncertainty. In our series 3 of the 32 infants with moderate-to-severe HIE were not cooled; all three had severe HIE. When analyses were performed eliminating these three patients, no difference in results emerged, even when including optimal cut-off values for CSF-NSE (ROC analysis) (S1 and S2 Tables). Finally, Since it is not ethical to perform lumbar puncture in healthy babies, we decided to use babies with suspected infection as control subjects. This approach has already been used in previous studies with the same purpose as ours, and we believe that CSF-NSE values are highly unlikely to be influenced by the consideration of infectious risk since their neurological status and CSF findings were normal [21, 26, 27].

Conclusion

HIE is a rare but potentially devastating disease associated with high morbidity and mortality. Our study suggests that NSE in CSF measured in infants with HIE at a median age of 48 hours of life might be a useful tool to identify those newborns with suspicion of severe brain injury based on clinical assessment, aEEG, and MRI. Furthermore, in the era of hypothermia, this biomarker appears to be able to distinguish infants will fully recover from those who are likely to present an adverse outcome.

Cerebrospinal fluid levels of neuron-specific enolase of the 40 cooled infants with HIE according to their neurological findings and outcomes.

The three infants that did not undergo hypothermia treatment had severe HIE and were not included in this analysis (NSE values were 166, 281, and 2855 ng/ml). CSF-NSE values are expressed as median (interquartile range). Mann-Whitney U test was used to analyze the differences in CSF-NSE levels between groups. P value of <0.05 was considered as indicating statistical significance. aEEG: amplitude integrated electroencephalogram; BGP: background pattern; BSID-III: Bayley Scales of Infant and Toddler Development Third Edition; CP: cerebral palsy; MRI: magnetic resonance imaging; TH: therapeutic hypothermia. aOutcome in 28/29 surviving infants. bAdverse outcome: death or cerebral palsy and/or BSID-III score <85.

(DOCX)

Click here for additional data file.

Receiver operating characteristic curve analysis of cerebrospinal fluid levels of neuron-specific enolase and neurological findings of the 40 cooled infants.

The three infants that did not undergo hypothermia treatment had severe HIE and were not included in this analysis (NSE values were 166, 281 and 2855 ng/ml). aEEG: amplitude integrated electroencephalogram; BGP: background pattern; BSID-III: Bayley Scales of Infant and Toddler Development Third Edition; CP: cerebral palsy; MRI: magnetic resonance imaging; PPV: positive predictive value; NPV: negative predictive value; ROC: Receiver operating curve; TH: therapeutic hypothermia. aOutcome in 28/29 surviving infants. bAdverse outcome: death or cerebral palsy and/or BSID-III score <85. CSF-NSE values are expressed in ng/ml.

(DOCX)

Click here for additional data file.

(XLSX)

Click here for additional data file.

3 Mar 2020

PONE-D-19-29027

Cerebrospinal fluid levels of neuron-specific enolase predict the severity of brain damage in newborns with neonatal hypoxic-ischemic encephalopathy treated with hypothermia

PLOS ONE

Dear Dr Garcia-Alix,

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 address the comments raised by both Reviewers, especially the major comments in order to further improve your manuscript.

We would appreciate receiving your revised manuscript by Apr 17 2020 11:59PM. When you are 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.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

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). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Pierre Gressens

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

http://www.journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and http://www.journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2) Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information.

3) We noticed you have some minor occurrence of overlapping text with the following previous publication(s), which needs to be addressed:

https://www.pedneur.com/article/S0887-8994(18)31326-2/fulltext

The text that needs to be addressed is in the Discussion section.

In your revision ensure you cite all your sources (including your own works), and quote or rephrase any duplicated text outside the methods section. Further consideration is dependent on these concerns being addressed.

4) Please change your reference to "p=0.000" to "p<0.001" or as similarly appropriate, as p values cannot equal zero.

5) Thank you for your ethics statement : "Written information was given to the parents and written consent was obtained at the bedside on admission of each infant after explanation of the study and before its onset. The Human Studies Committee of each institution approved the study"

Please amend your current ethics statement to include the full name of the ethics committee/institutional review board(s) that approved your specific study.

Once you have amended this/these statement(s) in the Methods section of the manuscript, please add the same text to the “Ethics Statement” field of the submission form (via “Edit Submission”).

For additional information about PLOS ONE ethical requirements for human subjects research, please refer to http://journals.plos.org/plosone/s/submission-guidelines#loc-human-subjects-research.

6) We note you have included a table to which you do not refer in the text of your manuscript. Please ensure that you refer to Table 3 in your text; if accepted, production will need this reference to link the reader to the Table.

7) Thank you for stating the following financial disclosure:

[The funders had no role in study design, data collection and analysis, decision to

publish, or preparation of the manuscript].

Please provide an amended Funding Statement that declares *all* the funding or sources of support received during this specific study (whether external or internal to your organization) as detailed online in our guide for authors at http://journals.plos.org/plosone/s/submit-now.

Please state what role the funders took in the study.  If any authors received a salary from any of your funders, please state which authors and which funder. 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."

Please include your amended statements within 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's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: No

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: In this study the authors report elevated levels of neuron-specific enolase (NSE) in CSF following birth asphyxia, and correlate the levels to grade of encephalopathy, aEEG and MRI findings, as well as cognitive and motor outcome. NSE as a marker for brain injury and adverse outcome following birth asphyxia has previously been widely studied in serum as well as CSF, including studies that report elevated NSE in CSF in association with adverse outcome in cooled infants. The strength of this study, is that NSE is studied in a relatively large group of infants treated according to current protocols for therapeutic hypothermia, that a wider variety of clinical variables indicating significant brain injury are studied, and that the authors present cut-off values that are highly predictive for brain injury and adverse outcome. Introduction and Discussion are succinct and to the point. Patient material is clearly defined and clinical examinations and definitions widely accepted.

There are, however, issues mainly regarding presentation and analysis of data that need to be addressed

Major comments:

Methods:

One of the main motivations for this study is to investigate NSE as a marker in the age of hypothermia. Three infants with severe HIE were not cooled according to clinical protocol but still included. Reasons for inclusion?

Statistics and presentation:

- Only Mann-Whitney is stated as method for comparisons between groups (numerical values for NSE between groups). In Table 1, control infants are compared with asphyxiated infants with three different grades of encephalopathy. Both numerical and nominal variables are compared and only one p-value per variable is presented. Statistics? Kruskall-Wallis/Chi-Square – or controls compared to all HIE groups together? Add to statistics under Methods.

- Statistical method needs to be added to all Figure and Table legends to make them self-explanatory.

- Comparing NSE levels between controls and asphyxiated infants with different grades of HIE (line 265-271) seem to be performed with multiple Mann-Whitney tests without correction. Please clarify/motivate or use other method including corrections

- Use of regression analyses for confounding factors – last part of Results. It is highly relevant and important to correct for age at lumbar puncture. Is the correction performed within the asphyxia group only, or in all infants including controls?

- In the same section of results: What is the rational for using regression analysis to correct for severity of HIE as it is strongly correlated with NSE levels as well as outcome? If I misunderstood this part, please clarify.

Pictures and legends:

- Legends for Pictures 2 and 3 have obviously been mixed up!

- Add number of infants (n) in the different groups into figures or legends

Discussion:

The authors report a remarkably high sensitivity and specificity for CSF NSE as a marker for neurological signs and brain injury as well as outcome. This a major finding that warrants further emphasis in the Discussion.

Minor comments:

- Please use consistent terminology. PCI is introduced instead of CP without explanation.

- Timing and comparisons of MRI. As I understand infants were either imaged twice (early around day 4 and late during second week) or only early. If two images were obtained only the later was used. Why - when comparing with children with only early images? Motivate or clarify.

Reviewer #2: In this prospective observational cohort study Leon-Lozanzo et al examined NSE in the CSF of a cohort of neonates with HIE. A control group of term neonates without CNS infection who has an LP taken for an infection screen was used for comparison. Within the cohort of HIE neonates, the level of NSE was analysed for short and long term outcome measures. The results of this study suggest that CSF NSE taken in the first few days after birth in neonates with HIE is associated with HIE severity and outcome and might be used as a predictive tool. CSF NSE as a biomarker for HIE outcome has been reported previously, as noted in the manuscript. The novelty of this study, according to the authors, is the association of CSF NSE with other surrogate outcome biomarkers (EEG and MRI).

The study is appropriately designed, and the results support the conclusions drawn. The results are of interest to clinicians and scientists in the field. There are some major and minor issues with the manuscript which I have highlighted below.

1.0 Major issues:

1.1 I suggest the authors consider hiring a copy editor or similar. The English language requires improvement, there are a number of grammatical/typing errors and the manuscript could be less verbose.

1.2 The main question a reader might want answered by this manuscript is ‘does CSF NSE, in isolation or in combination with other biomarkers, provide greater predictive valve for long term outcome than the more readily available tools, i.e. clinical assessment of severity, EEG and MRI? i.e. is it worth performed an LP in a neonate with HIE?’. The discussion highlights the pitfalls in the predicting outcome using EEG and MRI but misses the opportunity to tell us if CSF NSE is better or additive. A further analysis of the available data could shed light on this important question

2.0 Minor issues:

2.1 Some terms are used in the manuscript that are not in common use i.e. eutocic (line 226 and table 1) or may have translated poorly, i.e. ‘the aggression’ (line 370, 374, 413). The authors should substitute for the more commonly used terms

2.2 Replace phentanyl with fentanyl

2.3 In the introduction (lines 84 – 86) and discussion (lines 401 – 403) mention is made of the technical factors which make serum NSE unreliable, including processing of the sample and the effect of temperature. It would be worthwhile including whether it is known if these same factors influence CSF samples.

2.4 In limitations, I suggest reference to the use of infants with suspected infection as control subjects

2.5 Clinical and electrographic seizures are grouped as an outcome measure. I make a note of this due to the poor correlation between ‘clinical’ and electrographic seizures, and the uncertainty around the interpretation of clinical seizures without electrographic correlate. Consider including only electrographic seizures in analysis, or if the combined group is used, then the method of diagnosing clinical seizures needs to be added to the methods

2.6 Report IQR for age at ND assessment (line 180)

2.7 Page line 246, within description for Table 1. Footnote for ‘c’ incorrect (repeated from ‘b’)

2.8 Define PCI in full at least once (page line 261 and 321)

2.9 Figure 1. 2 outliers not shown on the figure (2855 ng/ml and 1490 ng/ml for 2 neonates with severe HIE referred to elsewhere in manuscript). Suggest either include in the figure as outlier dots or mention omission in the figure legend.

2.10 Descriptions for figure 2 and figure 3 need to be swapped around

2.11 Figure 2 and Table 2 include identical data. Both are not necessary

2.12 Consider showing significant differences between groups on figure 1 and figure 2

2.13 Line 54 and 318 refer to ‘minor findings’. Define this or replace with clearer language

2.14 Figure 3 – this is a very worthwhile figure, but the legend requires more detail. Explain the different shapes, the meaning of empty versus filled shapes more clearly, and the reason for the dashed line

2.15 The paragraph about ‘window of opportunity’, line 373 to 380, would benefit from more delicate wording or a clearer explanation (i.e. line 378 add… survive with subjectively intolerable disability)

**********

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

[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 to be viewed.]

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 us at figures@plos.org. Please note that Supporting Information files do not need this step.

17 Apr 2020

PONE-D-19-29027

Cerebrospinal fluid levels of neuron-specific enolase predict the severity of brain damage in newborns with neonatal hypoxic-ischemic encephalopathy treated with hypothermia

PLOS ONE

Dear Dr Garcia-Alix,

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 address the comments raised by both Reviewers, especially the major comments in order to further improve your manuscript.

We would appreciate receiving your revised manuscript by Apr 17 2020 11:59PM. When you are 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.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see:http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

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). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

• A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

• An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Pierre Gressens

Academic Editor

PLOS ONE

Journal Requirements:

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

Thank you. Please note that all our answers are referred to line numbers in the marked-up-copy that highlights changes made to the original version

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

http://www.journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdfandhttp://www.journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

Response. Thank you. We have checked our manuscript to insure that we meet PLOS ONE´s requirements.

2) Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information.

Response. Thank you. We added captions for our supporting information files at the end of our manuscript and updated text citations to match accordingly. Please see line 839 at the end of the manuscript, and line 562 for in-text citations

3) We noticed you have some minor occurrence of overlapping text with the following previous publication(s), which needs to be addressed:

https://www.pedneur.com/article/S0887-8994(18)31326-2/fulltext

The text that needs to be addressed is in the Discussion section.

In your revision ensure you cite all your sources (including your own works), and quote or rephrase any duplicated text outside the methods section. Further consideration is dependent on these concerns being addressed.

Response. Thank you. We have reviewed the overlapping text and have rephrased any duplicated text outside the methods section. Please see text from line 418 to line 467 in the discussion section.

4) Please change your reference to "p=0.000" to "p<0.001" or as similarly appropriate, as p values cannot equal zero.

Response. Thank you. We have changed p=0.000 to p<0.001 in Tables 1 and 2

5) Thank you for your ethics statement: "Written information was given to the parents and written consent was obtained at the bedside on admission of each infant after explanation of the study and before its onset. The Human Studies Committee of each institution approved the study"

Please amend your current ethics statement to include the full name of the ethics committee/institutional review board(s) that approved your specific study.

Once you have amended this/these statement(s) in the Methods section of the manuscript, please add the same text to the “Ethics Statement” field of the submission form (via “Edit Submission”).

For additional information about PLOS ONE ethical requirements for human subjects research, please refer tohttp://journals.plos.org/plosone/s/submission-guidelines#loc-human-subjects-research.

Response. Thank you. We have completed the information in the ethics statement. Please see line 230.

6) We note you have included a table to which you do not refer in the text of your manuscript. Please ensure that you refer to Table 3 in your text; if accepted, production will need this reference to link the reader to the Table.

Response. Thank you. We have added the reference to Table 3 in the text (at the end of the paragraph about CSF-NSE and outcome in the results section).

7) Thank you for stating the following financial disclosure:

[The funders had no role in study design, data collection and analysis, decision to

publish, or preparation of the manuscript].

a. Please provide an amended Funding Statement that declares *all* the funding or sources of support received during this specific study (whether external or internal to your organization) as detailed online in our guide for authors athttp://journals.plos.org/plosone/s/submit-now.

b. Please state what role the funders took in the study. If any authors received a salary from any of your funders, please state which authors and which funder. 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." Please include your amended statements within your cover letter; we will change the online submission form on your behalf.

Statement received by email (April 16):

Please clarify the sources of funding (financial or material support) for your study. List the grants or organizations that supported your study, including funding received from your institution.

b) State what role the funders took in the study. If the funders had no role in your study, please state: “The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.”

c) If any authors received a salary from any of your funders, please state which authors and which funders.

d) If you did not receive any funding for this study, please state: “The authors received no specific funding for this work.”

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

Response. Thank you. We have added the following statement to the cover letter: “The authors received no specific funding for this work”.

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: No

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: In this study the authors report elevated levels of neuron-specific enolase (NSE) in CSF following birth asphyxia, and correlate the levels to grade of encephalopathy, aEEG and MRI findings, as well as cognitive and motor outcome. NSE as a marker for brain injury and adverse outcome following birth asphyxia has previously been widely studied in serum as well as CSF, including studies that report elevated NSE in CSF in association with adverse outcome in cooled infants. The strength of this study, is that NSE is studied in a relatively large group of infants treated according to current protocols for therapeutic hypothermia, that a wider variety of clinical variables indicating significant brain injury are studied, and that the authors present cut-off values that are highly predictive for brain injury and adverse outcome. Introduction and Discussion are succinct and to the point. Patient material is clearly defined and clinical examinations and definitions widely accepted.

There are, however, issues mainly regarding presentation and analysis of data that need to be addressed

Major comments:

Methods:

One of the main motivations for this study is to investigate NSE as a marker in the age of hypothermia. Three infants with severe HIE were not cooled according to clinical protocol but still included. Reasons for inclusion?

Response. As the reviewer says, our intention was to describe the usefulness of NSE at the present time when therapeutic hypothermia is a standardized treatment in most countries. However, it is known that some patients do not receive this treatment for different reasons even though they were candidates. That is why we believe it is relevant to provide real data concerning why it is that some infants do not receive hypothermia. As stated in the text, in our series 3 newborns with severe HIE were not cooled: 166, 281, and 2855 ng/ml of CSF-NSE values.

However, as the reviewer notes, including these three children could influence the results if we wanted to draw conclusions regarding NSE in cooled infants. Thus, we have repeated the analyses without including these 3 children. The results are very similar. We decided to add a comment in the discussion section (line 559).

We added the analysis eliminating the three uncooled infants as supplementary material, in case the reader is interested in having a look at them (S1 and S2 Tables). If the editor and/or reviewers consider that these tables do not need to be added as supplementary material, we agree to their removal.

Statistics and presentation:

- Only Mann-Whitney is stated as method for comparisons between groups (numerical values for NSE between groups). In Table 1, control infants are compared with asphyxiated infants with three different grades of encephalopathy. Both numerical and nominal variables are compared and only one p-value per variable is presented. Statistics? Kruskall-Wallis/Chi-Square – or controls compared to all HIE groups together? Add to statistics under Methods.

Response. Thank you. In Table 1, comparisons are made between different grades of encephalopathy, but controls are not included. That is why the paragraph in the text (lines 247 to 254) does not refer to Table 1. Table 1 is referred to in the text only when addressing the characteristics of the infants with HIE. Thus, p values in Table 1 only refer to the comparison of the different grades of encephalopathy. We have clarified this in the legend of Table 1 (lines 258 to 260) and in the methods section (lines 194 to 198).

- Statistical method needs to be added to all Figure and Table legends to make them self-explanatory.

Response. Thank you. We have added the appropriate information to all figures and tables.

- Comparing NSE levels between controls and asphyxiated infants with different grades of HIE (line 265-271) seem to be performed with multiple Mann-Whitney tests without correction. Please clarify/motivate or use other method including corrections

Response. Thank you. This is a very appropriate suggestion. Before using the method with corrections we explored data trends and we noticed that an ANOVA post hoc analysis could better explore the data than using p-adjustment by Bonferroni. So, we transformed the CSF-NSE variable to normalize it (logarithm transformation) and performed a parametric test, which allowed analysis between groups without losing power. As sample size of the moderate and severe group was not large, significance may be affected using a non-parametric test.

The analysis showed that there were statistically significant differences in all comparisons between groups. Same results were found when we excluded the three infants that were not cooled. We added the analysis as supplementary information in case it wants to be checked (but we did not included it in the manuscript as it is too long) and a paragraph referring to this correction in the methods section (line 199).

- Use of regression analyses for confounding factors – last part of Results. It is highly relevant and important to correct for age at lumbar puncture. Is the correction performed within the asphyxia group only, or in all infants including controls?

Response. As the reviewer says, regression analysis to correct CSF-NSE levels for age at lumbar puncture was performed in all infants including controls. The analysis was performed by including time as a numerical variable, and also categorizing it into two periods <48 and > 48 hours. In both cases, time was not a confounding factor.

To better display the information we have changed the sentence on line 394 and moved the information regarding the regression analysis focused on the degree of HIE to line 294. We have left the regression analysis with respect to the outcome more clearly reported in lines 394 to 396 .

We also changed the paragraph in the methods section in order to make it more comprehensible (line 206)

- In the same section of results: What is the rational for using regression analysis to correct for severity of HIE as it is strongly correlated with NSE levels as well as outcome? If I misunderstood this part, please clarify.

Response. Thank you. We checked this issue with our statistics department and they agree that in order to estimate the predictive capacity of the NSE regardless of the degree of HIE and the time of determination, the NSE-outcome association should be estimated by multiple regression, adjusting for these covariates. This adjusted estimation helps to distinguish the outcome differences associated with NSE from those associated with the degree of HIE (the effect of NSE levels may be different depending on the degree of HIE).

In any case, as we noted above, we changed the paragraph at the end of the results section to explain the information more clearly (lines 206 to 211).

Pictures and legends:

- Legends for Pictures 2 and 3 have obviously been mixed up!

Response. Yes, thank you. We have corrected the legends.

- Add number of infants (n) in the different groups into figures or legends

Response. We completed legends for Figures. Numbers were already added in the tables.

Discussion:

The authors report a remarkably high sensitivity and specificity for CSF NSE as a marker for neurological signs and brain injury as well as outcome. This a major finding that warrants further emphasis in the Discussion.

Response. Thank you. To give more emphasis we moved the last paragraph at the end of the discussion section (lines 569-574) and rephrased the information in lines 542 to 550.

Minor comments:

- Please use consistent terminology. PCI is introduced instead of CP without explanation.

Response. Thank you. We changed PCI to CP (line 358 and 280).

- Timing and comparisons of MRI. As I understand infants were either imaged twice (early around day 4 and late during second week) or only early. If two images were obtained only the later was used. Why - when comparing with children with only early images? Motivate or clarify.

Response. Thank you very much for your observation. We have eliminated the word “early” when referring to the infant who underwent only one study in order not to be misunderstood. For infants in whom only one MRI study was performed, the time ranged from 38 to 584 days of life. That is why we preferred to use the “late MRI” in those with two images, in order to indicate as clearly as we could the severity of the damage. We know this could be a controversial aspect as an early MRI could be comparable to a late MRI and may even offer some advantage, especially if sequences such as diffusion and spectroscopy are incorporated.

In any case, we also explored the analyses using the early image rather than late MRI, and the results were very similar (without differences when excluding the three non-cooled infants) with slightly better area under the curve in the ROC analysis when using late MRI.

Variable N Cutoff-point Area under the ROC (95%CI) Sensitivity (95%CI) Specificity (95%CI) PPV (95%CI) NPV (95%CI)

Using the late MRI if two images were obtained

Moderate-severe injury (MRI) 20 80 0.93 (0.85,1.00) 0.75 (0.51,0.91) 1.00 (0.81,-) 1.00 (0.79,1.00) 0.78 (0.55,-)

Global injury pattern (MRI) 17 80 0.95 (0.88,1.00) 0.82 (0.57,0.96) 0.95 (0.76,0.99) 0.93 (0.69,0.99) 0.87 (0.65,0.99)

Using the early MRI if two images were obtained

Moderate-severe injury (MRI) 21 80 0.89 (0.80,0.99) 0.71 (0.48,0.89) 1.00 (0.80,-) 1.00 (0.79,1.00) 0.74 (0.51,-)

Global injury pattern (MRI) 23 80 0.87 (0.77,0.98) 0.65 (0.43,0.84) 1.00 (0.78,-) 1.00 (0.78,1.00) 0.65 (0.43,-)

Reviewer #2: In this prospective observational cohort study Leon-Lozanzo et al examined NSE in the CSF of a cohort of neonates with HIE. A control group of term neonates without CNS infection who has an LP taken for an infection screen was used for comparison. Within the cohort of HIE neonates, the level of NSE was analysed for short and long term outcome measures. The results of this study suggest that CSF NSE taken in the first few days after birth in neonates with HIE is associated with HIE severity and outcome and might be used as a predictive tool. CSF NSE as a biomarker for HIE outcome has been reported previously, as noted in the manuscript. The novelty of this study, according to the authors, is the association of CSF NSE with other surrogate outcome biomarkers (EEG and MRI).

The study is appropriately designed, and the results support the conclusions drawn. The results are of interest to clinicians and scientists in the field. There are some major and minor issues with the manuscript which I have highlighted below.

1.0 Major issues:

1.1 I suggest the authors consider hiring a copy editor or similar. The English language requires improvement, there are a number of grammatical/typing errors and the manuscript could be less verbose.

Response. Thank you. We have reviewed the language and changed it where appropriate.

1.2 The main question a reader might want answered by this manuscript is ‘does CSF NSE, in isolation or in combination with other biomarkers, provide greater predictive valve for long term outcome than the more readily available tools, i.e. clinical assessment of severity, EEG and MRI? i.e. is it worth performed an LP in a neonate with HIE?’. The discussion highlights the pitfalls in the predicting outcome using EEG and MRI but misses the opportunity to tell us if CSF NSE is better or additive. A further analysis of the available data could shed light on this important question

Response. We thank the reviewer for this comment as it demonstrates his/her clinical interest is in line with ours. In our opinion CSF-NSE is one more pillar when it comes to establishing the severity of the damage and relating it to the prognosis. As our research points out, it reflects the damage revealed by other tools such as clinical exam, aEEG, and neuroimaging. In our opinion, CSF-NSE should be considered as a key tool when it comes to obtaining as much information as possible regarding each infant. However, although we understand what the reviewer points out and its relevance, performing a combined analysis with all the tools would be advisable with a greater number of patients included for better consistency of statistical models. We are working on acquiring this data set shortly.

But to help the readers compare the probability values for each tool, we added a paragraph (line 397) in the results section and also Table 4 (line 400) in which it may be seen that the CSF-NSE compares favorably with the other classically used tools. We also added an appropriate sentence in the discussion section (line 538).

2.0 Minor issues:

2.1 Some terms are used in the manuscript that are not in common use i.e. eutocic (line 226 and table 1) or may have translated poorly, i.e. ‘the aggression’ (line 370, 374, 413). The authors should substitute for the more commonly used terms

Response. Thank you. We changed these two terms: we replaced ‘eutocic’ delivery with ‘spontaneous’ and hypoxic-ischemic ‘aggression’ with hypoxic-ischemic ‘insult’.

2.2 Replace phentanyl with fentanyl

Response. Thank you. We have corrected this.

2.3 In the introduction (lines 84 – 86) and discussion (lines 401 – 403) mention is made of the technical factors which make serum NSE unreliable, including processing of the sample and the effect of temperature. It would be worthwhile including whether it is known if these same factors influence CSF samples.

Response. Thank you very much for pointing out our imprecision on this issue. We have completed the phrase in the introduction (lines 8e to 88), also in the methods section (line 136), and we have tried to better clarify in the discussion (line 484) with a new reference (number 53). We also added a comment regarding the measuring method that may also be relevant in addressing the confounding technical factors (lines 486 to 488).

2.4 In limitations, I suggest reference to the use of infants with suspected infection as control subjects

Response. Thank you. We have completed the information and added references already included in the manuscript (numbers 21,26 and 27) of other studies that used the same approach as ours (lines 563 to 568).

2.5 Clinical and electrographic seizures are grouped as an outcome measure. I make a note of this due to the poor correlation between ‘clinical’ and electrographic seizures, and the uncertainty around the interpretation of clinical seizures without electrographic correlate. Consider including only electrographic seizures in analysis, or if the combined group is used, then the method of diagnosing clinical seizures needs to be added to the methods

Response. Thank you. We agree with the reviewer that it may be better to consider only electrographic seizures in the analysis. Thus, we re-analyzed this variable and modified Tables 1, 2, and 3. At any rate, the results were very similar to those of previous analyses.

2.6 Report IQR for age at ND assessment (line 180)

Response. Thank you. We have added the information.

2.7 Page line 246, within description for Table 1. Footnote for ‘c’ incorrect (repeated from ‘b’)

Response. Thank you. We have corrected it.

2.8 Define PCI in full at least once (page line 261 and 321)

Response. Thank you. This is a mistake. PCI is the abbreviation for cerebral palsy in Spanish. We have changed it to the correct abbreviation in English: CP.

2.9 Figure 1. 2 outliers not shown on the figure (2855 ng/ml and 1490 ng/ml for 2 neonates with severe HIE referred to elsewhere in manuscript). Suggest either include in the figure as outlier dots or mention omission in the figure legend.

Response. Since Figures 1 and 2 represent the enolase values in medians and interquartile range, those values outside P25 and P75 are not shown. If they were to be displayed, it would not have to be just the two values that the reviewer refers to, but all of them. In our opinion, this makes the graph ‘dirtier’; it is common practice to show only the median and the interquartile range when drawing the data in box-plots. Figure 3 does point out that these two patients are not represented as it is a figure in which all patients in the cohort are represented and is therefore not a half-IQR as Figures 1 and 2.

2.10 Descriptions for figure 2 and figure 3 need to be swapped around

Response. Thank you. We have corrected it.

2.11 Figure 2 and Table 2 include identical data. Both are not necessary

Response. We have eliminated Figure 2 and renamed the Fig numbers: Fig 3 has been changed to Fig 2

2.12 Consider showing significant differences between groups on figure 1 and figure 2

Response. Relevant information was added to the legend of Figure 1. Figure 2 was eliminated.

2.13 Line 54 and 318 refer to ‘minor findings’. Define this or replace with clearer language

Response. The reason for pointing out normal or minor findings was to refer to the outcome that did not correspond to "adverse outcome". But as the reviewer says, this could be misunderstood, so we've removed it from lines 55 and 355.

2.14 Figure 3 – this is a very worthwhile figure, but the legend requires more detail. Explain the different shapes, the meaning of empty versus filled shapes more clearly, and the reason for the dashed line

Response. Thank you. We have completed and clarified the information. Please note that Figure 3 changed the name to Figure 2.

2.15 The paragraph about ‘window of opportunity’, line 373 to 380, would benefit from more delicate wording or a clearer explanation (i.e. line 378 add… survive with subjectively intolerable disability)

Response. Thank you. We have added “with subjectively intolerable disability”.

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

[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 to be viewed.]

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 us at figures@plos.org. Please note that Supporting Information files do not need this step.

In compliance with data protection regulations, you may request that we remove your personal registration details at any time. (Remove my information/details). Please contact the publication office if you have any questions.

Submitted filename: 20200415_Response to the Reviewers.docx

Click here for additional data file.

19 May 2020

Cerebrospinal fluid levels of neuron-specific enolase predict the severity of brain damage in newborns with neonatal hypoxic-ischemic encephalopathy treated with hypothermia

PONE-D-19-29027R1

Dear Dr. Garcia-Alix,

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

Within one week, you will receive an e-mail containing information on the amendments required prior to publication. When all required modifications have been addressed, you will receive a formal acceptance letter and your manuscript will proceed to our production department and be scheduled for publication.

Shortly after the formal acceptance letter is sent, an invoice for payment will follow. To ensure an efficient production and billing process, please log into Editorial Manager at https://www.editorialmanager.com/pone/, click the "Update My Information" link at the top of the page, and update your user information. 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 enable them to help maximize its impact. If they will be preparing press materials for this manuscript, you must inform our press team as soon as possible and 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.

With kind regards,

Pierre Gressens

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #2: The authors have satisfactorily addressed the comments and suggestions made. The manuscript is improved by the changes made. I have no further comments.

**********

7. 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 #2: No

22 May 2020

PONE-D-19-29027R1

Cerebrospinal fluid levels of neuron-specific enolase predict the severity of brain damage in newborns with neonatal hypoxic-ischemic encephalopathy treated with hypothermia

Dear Dr. Garcia-Alix:

I am 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 notify them about your upcoming paper at this point, to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, 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.

For any other questions or concerns, please email plosone@plos.org.

Thank you for submitting your work to PLOS ONE.

With kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Pierre Gressens

Academic Editor

PLOS ONE