Diagnostic performance of serum neuron-specific enolase for cranial ultrasound-detected brain injury in preterm neonates

Introduction

Preterm birth is a major concern of contemporary perinatology and public health. Annually, 12–15 million preterm infants worldwide are born, accounting for 11% of all births. Preterm birth contributes to up to 70% of all neonatal deaths. In addition to high mortality, preterm birth carries a high risk of morbidity, lasting health consequences, and high healthcare costs (1).

Despite advances in neonatology that have contributed to increased survival, neurological complications remain the leading cause of disability in preterm infants (2). Cranial ultrasound (CUS) is the recommended first-line and preferred diagnostic imaging tool for the detection of peri-/intraventricular hemorrhage (PIVH) and periventricular leukomalacia (PVL) in routine neonatal care. However, CUS is operator-dependent and has limited sensitivity for noncystic PVL and diffuse white matter injury, which are better detected by magnetic resonance imaging (MRI). Therefore, an objective blood-based biomarker may complement CUS by supporting early risk stratification (3,4).

The etiology of neonatal brain injury (NBI) in preterm infants is complex and involves a wide range of risk factors during prenatal, perinatal, and postnatal periods. Consequently, there is a growing interest in biochemical markers that allow early prediction and identification of cerebral injury and thus enable early intervention and the prevention of long-term effects (5,6).

Studies have reported elevated levels of neuron-specific enolase (NSE), a glycolytic enzyme located in neurons and neuroendocrine cells, in NBI, although the results have not always been consistent (7,8).

The aim of this study is to determine the correlation between serum NSE levels measured in the first 48 to 96 hours of life and CUS findings in preterm infants with regard to the occurrence and severity of PIVH and PVL. Our main research hypothesis is that there is an association between elevated NSE levels and pathological CUS findings in preterm infants (any form of PVL & PIVH). We present this article in accordance with the STARD reporting checklist (available at https://tp.amegroups.com/article/view/10.21037/tp-2025-aw-795/rc).

Methods

This prospective single-center cohort study included 63 preterm neonates (<37 weeks of gestational age) admitted consecutively to the Neonatal Intensive Care Unit (NICU) of the Clinical Center of Montenegro, the only tertiary institution for the care of preterm infants in Montenegro. After applying exclusion criteria, 55 infants were included in the final analysis.

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of The Faculty of Medicine, University of Montenegro (approval No. 409/3), and by the Ethics Committee of the Clinical Center of Montenegro (approval No. 03/01-12438/1). Written informed consent was obtained from the parents of all participants.

The results reported in this paper contribute to the overall objectives of the study, which investigates the associations between brain injury, NSE levels, nutritional status, and other biochemical markers.

Participants

The study included all preterm neonates born before 37 weeks of gestation, admitted to the Neonatology Department at the Clinical Center of Montenegro. The patients were divided into a primary study group with abnormal CUS results (PIVH and PVL) and a control group with normal ultrasound (US) results. Eligible participants were preterm neonates with complete clinical, biochemical, and CUS data obtained during the first week of life.

Participants were classified according to the level of prematurity into four different categories: late preterm [34–37 gestational weeks (gw)], moderate preterm (32–34 gw), very preterm (28–32 gw), and extremely preterm neonates (<28 gw).

Sample size: no formal a priori sample size calculation was performed. The sample size was based on feasibility and included consecutive recruitment of all eligible preterm neonates during the predefined study period at our single center. Because this study was designed as an exploratory diagnostic accuracy evaluation, diagnostic performance measures are presented with exact (Clopper-Pearson) 95% confidence intervals (CIs). Given the limited number of controls (n=20), the CI around specificity is wide and this estimate should therefore be interpreted with caution.

General exclusion criteria for the study, selected due to their potential influence on NSE concentrations, included major congenital malformations of the central nervous system, genetic or chromosomal disorders, signs of hypoxic-ischemic encephalopathy (HIE) at birth, serum hemolysis, patients with proven neuroendocrine tumors, severe renal impairment, recent neurosurgical procedures, and cases of maternal preeclampsia during pregnancy. Infants with confirmed or suspected sepsis or septic shock were excluded to avoid potential NSE elevation unrelated to brain injury.

Flow and timing: the index test and reference standard were performed within the first week of life; NSE sampling was at 48–96 hours (median day 3), and CUS was performed within 72 hours with follow-up scans as clinically indicated. No protocolized therapeutic interventions targeting brain injury were scheduled between the two assessments.

Neuroradiological evaluation

All participants underwent an initial CUS within the first 72 hours of life, followed by longitudinal evaluation throughout hospitalization, including follow-up scans beyond 7 days after birth, and when clinically indicated, up to term equivalent age.

NSE results were not available at the time of CUS acquisition and reporting; therefore, CUS assessors were blinded to the index test.

CUS was chosen as the reference standard because it is the recommended first-line bedside modality for PVL/PIVH detection in preterm infants and standard of care in our NICU.

US examination was performed using a small footprint curvilinear probe with frequencies of 2–5 MHz or higher, via the anterior fontanelle. Four parasagittal planes per hemisphere and six standard coronal planes of examination were part of the protocol established.

US findings were classified as normal or abnormal. PIVH was graded using the Volpe classification (grades I–IV). Given the consistently low (sometimes absent) frequency of type IV PVL in prior de Vries-based studies, we used the more recent Agut et al. grading for PVL (grades 1–3: mild, moderate, severe) (9). Transient periventricular echogenicity that resolved within 7 days without pathological evolution was classified as normal (Figure 1).

Figure 1 Cranial ultrasound images. (A) Classic appearance of periventricular leukomalacia with increased echogenicity in the periventricular white matter. (B) Normal cranial ultrasound with symmetric ventricular contours and no abnormal echogenicity.

Biochemical analysis

Blood samples, obtained by venipuncture, were immediately sent to the Department of Laboratory Diagnostics, Institute for Children’s Diseases, Clinical Center of Montenegro for further processing. The index test was serum NSE measured within the first 48–96 hours of the child’s life (the exact time depended on clinical indications for routine biochemical control; the median sampling time was the third day of life), from the remaining portion of the serum, without the need for additional patient sampling. The analysis was performed on the Alinity immunochemical analyzer (Abbott, USA) using the chemiluminescent microparticle immunoassay (CMIA) method. Laboratory personnel were blinded to CUS results and clinical data.

Statistical analysis

Data were recorded in Microsoft Excel. Continuous data were reported as mean ± standard deviation and/or median (interquartile range), and categorical data as n (%). Normality was assessed by the Shapiro-Wilk test. Between-group comparisons (normal vs. abnormal CUS; secondary analyses PVL vs. normal and PIVH vs. normal) were performed using the Mann-Whitney U test; effect size was quantified as Cohen’s d. Subgroup analyses contrasting PVL vs. PIVH were pre-specified; all other subgroup evaluations were exploratory. Correlations between NSE level and lesion severity were analyzed with Spearman’s rank correlation (ρ) within PVL and PIVH subgroups.

Diagnostic performance was quantified by receiver operating characteristic (ROC) analysis, reporting area under the curve (AUC) with 95% CIs obtained by stratified bootstrap resampling (5,000 iterations). The diagnostic positivity threshold for NSE (43.2 µg/L) was derived post hoc from the ROC curve using Youden’s index and is therefore considered exploratory; at this threshold, we report sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) with exact (Clopper-Pearson) 95% CIs.

Besides non-parametric group comparisons, a multivariable quantile (median) regression was performed to assess the independent association of abnormal CUS findings with serum NSE, adjusted for gestational age, birth weight and sex. Furthermore, a 1:1 nearest-neighbor matching by gestational age was performed (±1 week, without replacement), and paired NSE values were compared using the Wilcoxon signed-rank test. Analyses were performed as complete-case (no imputation). Participants with missing index test or reference-standard data, or with indeterminate results, were excluded from the primary diagnostic-accuracy analyses.

Results

Study population characteristics

Among 63 consecutively assessed preterm neonates, 55 met inclusion criteria and were enrolled in this prospective single-center study. Eight neonates were ruled out because of several confounding factors that could have influenced values of NSE, namely maternal preeclampsia and specimen hemolysis. The included neonates had a mean gestational age of 32.1±3.5 weeks with a range of 23 to 36 weeks. Their mean birth weight was 1.75±0.57 kg, with a mean body length of 43.8±4.8 cm. The sex distribution was even: 28 male (50.9%) and 27 female (49.1%) infants. A flow diagram summarizing screening, exclusions with reasons, and final numbers included is provided in Figure 2.

Figure 2 STARD participant flow diagram (screening, exclusions with reasons, and study population). CUS, cranial ultrasound; NICU, neonatal intensive care unit; NSE, neuron-specific enolase; STARD, Standards for Reporting Diagnostic Accuracy Studies.

Neonates were classified into four different groups based on gestational age:

• Late preterm: 34⁺⁰–36⁺⁶ gw;
• Moderate preterm: 32⁺⁰–33⁺⁶ gw;
• Very preterm: 28⁺⁰–31⁺⁶ gw;
• Extremely preterm: <28⁺⁰ gw.

Patient distribution across gestational age subgroups, with corresponding sex, CUS findings, and serum NSE concentrations, is summarized in Table 1. Across all categories, median NSE values were higher in infants with pathological CUS findings, and the difference reached statistical significance in late and very preterm groups.

US-based grouping and lesion distribution

In this study, patients were categorized into two main cohorts based on CUS findings: one with normal CUS results (control), and the other with any sort of abnormal results.

Out of 55 preterm neonates in total, 20 had normal CUS results, whereas 35 had abnormal US results. In total, PVL lesions—either alone or together with PIVH—were diagnosed in 27 preterm infants. Grade 1 PVL was present in 19 patients, and grade 2 PVL in 8. There were no instances of PVL grade 3.

PIVH lesions were present in 17 patients, with the full range of severity from mild (grade 1) to severe (grade 4).

When considering only patients with isolated lesions, the mean serum NSE concentration was 63.44 µg/L in the PVL group (n=18) and 84.76 µg/L in the PIVH group (n=8).

Levels were higher in the PIVH group, yet there was considerable overlap between individual measurements from the two conditions.

NSE differences between normal and abnormal US groups

Normality of the NSE value distribution in both groups was first tested using the Shapiro-Wilk test, which indicated that the values did not follow a normal distribution. According to this, the nonparametric equivalent of the t-test, i.e., Mann-Whitney U test was applied as the test of comparison of groups.

• Normal US group (n=20): mean 34.04±16.67 µg/L; median 39.1 µg/L;
• Abnormal US group (n=35): mean 82.54±39.89 µg/L; median 78.9 µg/L.

The resulting P value of 4.89×10⁻⁷ (P<0.001) is far below the traditional threshold (P<0.05), as well as stricter benchmarks (P<0.001), indicating a highly significant difference in NSE values between patients with normal vs. abnormal CUS findings.

Large effect size was seen with Cohen’s d =1.45, reflecting a clinically relevant difference.

This is presented in graphical format in Figure 3 (boxplot of NSE values per group).

Figure 3 This boxplot visually illustrates the difference in NSE distributions between the groups, consistent with the Mann-Whitney U test findings. CUS, cranial ultrasound; NSE, neuron-specific enolase.

Furthermore, Mann-Whitney U tests illustrated that the serum NSE levels were significantly higher in infants with any form of PVL (P=3.67×10⁻⁶) or PIVH (P=3.77×10⁻⁵) compared to those with normal CUS findings. These results are consistent with the observation that NSE levels are significantly elevated in the presence of both these types of brain injury. Together, the group comparison and correlation analyses support the potential clinical usefulness of NSE as an indicator of the severity of neonatal brain damage.

Adjusted and gestational age matched analyses

In multivariable quantile regression, adjusting for gestational age, birth weight and sex abnormal findings on CUS remained independently associated with higher serum NSE concentrations, with a median difference of +38.3 µg/L (95% CI: 15.3–61.3, P=0.002). Gestational age was inversely associated with NSE (−7.1 µg/L/week, P=0.006), i.e., lower NSE levels in more mature infants, while birth weight was weakly positively associated (P=0.043).

This relationship was confirmed by the gestational-age-matched sensitivity analysis, which showed that, of 20 pairs, the median paired difference was +34.8 µg/L (95% CI: 14.0–52.0; Wilcoxon P<0.001). Thus, these results confirm that an abnormal CUS is associated with high NSE values independently of gestational age.

ROC analysis—diagnostic performance of NSE

At the threshold of 43.2 µg/L, derived within this cohort using Youden’s index, sensitivity was 91.4% (95% CI: 76.9–98.2%) and specificity 85.0% (95% CI: 62.1–96.8%). Corresponding positive and NPVs were 91.4% (95% CI: 76.9–98.2%) and 85.0% (95% CI: 62.1–96.8%), respectively. The ROC analysis yielded an AUC of 0.911 (95% CI: 0.817–0.981).

The associated 2×2 contingency table for this threshold was as follows: true positive (TP) =32, false negative (FN) =3, false positive (FP) =3, true negative (TN) =17. Because the threshold was data-driven and derived from the same dataset in which diagnostic performance was assessed, it should be considered exploratory and requires external validation in independent cohorts. For transparency, the complete 2×2 cross-tabulation is provided in Table S1.

The resulting ROC curve is presented in Figure 4. These results illustrate NSE’s potential as a diagnostic biomarker of neurological injuries.

Figure 4 ROC curve demonstrating the diagnostic performance of NSE for predicting pathological ultrasound findings. AUC, area under the curve; discrim., line of no discrimination; NSE, neuron-specific enolase; Opt. thresh., optimal threshold; ROC, receiver operating characteristic.

Correlation of NSE with lesion severity

Spearman correlation was used to calculate the correlation between serum NSE levels and the severity of brain injury on CUS. Patients with both PVL and PIVH were excluded to allow independent analysis of each condition separately, reducing the sample size per group notably.

In the isolated-lesion analysis, the mean serum NSE concentration was 63.44 µg/L in the PVL group (n=18) and 84.76 µg/L in the PIVH group (n=8). Although higher values were observed in the PIVH group, considerable variability and overlap between the two distributions were noted.

In PVL group, there existed a significant statistical correlation between PVL grade and NSE concentration (Spearman’s ρ=0.685, P=0.002), which translates to higher NSE levels with more severe PVL.

Within the PIVH group, there was an upward trend (ρ=0.320), though not statistically significant (P=0.40) and most likely to be the result of the small number of patients in this group. Larger cohorts are needed to refine the correlation in the PIVH subgroup (Figure 5).

Figure 5 Scatter plot showing the relationship between lesion severity and NSE values for PVL and PIVH cases. NSE, neuron-specific enolase; PIVH, peri-/intraventricular hemorrhage; PVL, periventricular leukomalacia; ROC, receiver operating characteristic.

Adverse events

No adverse events were observed related to phlebotomy or US examinations.

Summary of key findings

• NSE levels were significantly increased in neonates with abnormal CUS findings (P<0.001) defined as the presence of any grade of PVL or PIVH;
• At 43.2 µg/L, sensitivity and specificity were 91.4% and 85.0%, respectively;
• NSE levels positively correlated with PVL severity.

Discussion

NSE is a cytosolic glycolytic enzyme highly expressed in neurons and neuroendocrine cells, with minimal expression elsewhere. When neuronal injury occurs and the blood-brain barrier becomes permeable, NSE diffuses from the cytosol of neurons into cerebrospinal fluid and subsequently into the bloodstream. Serum concentrations rise within hours and decline with an approximate one-day half-life, making NSE a time-sensitive indicator of acute brain injury.

In preterm infants, the severe vulnerability of germinal matrix and periventricular white matter to hemorrhage as well as ischemia explains the mechanisms of increased NSE release by both PIVH and PVL (10).

Most of the earlier studies on NSE in preterm infants have relied largely on clinical presentation of neurologic dysfunction, such as seizures, hypotonia, or altered state of consciousness. Celtik et al. (11), for example, described a significant correlation between serum levels of NSE and clinically staged HIE. Similarly, previous research studies on HIE in term or near-term infants typically have taken HIE to be indicated by clinical criteria, e.g., Sarnat stage, Apgar score/pH, and acid-base status, with CUS being used primarily to track lesion development rather than as an inclusion criterion (Celtik et al.; Chaparro-Huerta et al.; Giuseppe et al.) (11-15).

This heterogeneity in outcome definition (clinical vs. imaging criteria) likely contributes to between-study variability in effect sizes and differences in numeric cut-offs.

Perrone et al.’s comprehensive systematic review on NBI explicitly calls for biomarker studies stratified by injury type and severity—a principle we follow by grading PVL and PIVH on CUS and analyzing both pooled and lesion-specific subgroups (8).

Following this suggestion, our design was focused solely on preterm infants, with cohorts established based on ultrasonographically verified brain injury and a direct comparison of their serum NSE levels.

The key finding of our study is that, among preterm neonates sampled 48–96 hours after birth (median day 3), serum NSE were significantly higher in the presence of any CUS-detected brain injury (PVL or PIVH) than in infants with normal CUS findings (P<0.05), with discrimination at an assay-specific cut-off of 43.2 µg/L (sensitivity 91.4%, specificity 85.0%). Importantly, this difference also remained significant in isolated subgroup analyses: NSE was higher in infants with PVL alone and in those with PIVH alone (both P<0.05) compared with neonates with normal CUS findings. After adjustment for gestational age, birth weight, and sex, abnormal CUS findings remained independently associated with higher NSE concentrations, confirming that this relationship is not explained by prematurity or growth differences. Collectively, these results indicate that NSE rises in both injury types and—together with correlation analyses—support its potential clinical utility as an indicator of neonatal brain-injury severity.

A particularly notable finding in our study was the inverse relationship between gestational age and baseline NSE, with mean concentrations ranging from 49.4±26.8 µg/L in late preterm infants to 101.2±57.2 µg/L in extremely preterm infants. This maturational gradient—likely reflecting progressive integrity of the blood-brain barrier, greater metabolic stability, and reduced susceptibility to perinatal stress—is consistent with published data showing higher baseline NSE in preterm than term infants and a decline with advancing maturity, including the gestational age-dependent pattern reported by Abbasoglu et al. These observations suggest that adult reference ranges (4.7–18 µg/L) are not applicable to neonates and that neonate-specific reference intervals, which are dependent on gestational age, are needed. It should be noted that there are currently no officially approved reference limits for preterm infants (16).

Given the very limited number of directly comparable studies of NSE in early brain injury in preterm infants, our study provides important additional data, which we compare primarily against two recent pivotal studies—Metallinou et al. [2024] and Efstathiou et al. [2023]. These studies serve as our main external reference parameters (4,17).

Both comparative studies demonstrated an association between NBI and increased serum NSE, supporting its potential utility as a biomarker. Notably, both studies also found an increase in NSE on day 3, especially in severe intraventricular hemorrhage (IVH). Metallinou et al. reported higher NSE on day 3 in IVH compared to controls with PVL, with significantly higher NSE values in grade II–IV hemorrhage (cutoff 5.43 µg/L). In parallel, Efstathiou et al. showed that subgroups of IVH grades III–IV already had elevated NSE on day 3, while differences between controls and cases with encephalopathy became more pronounced later. These findings support the view that sampling on day 3 preferentially captures the kinetics of NSE caused by hemorrhage, while phenotypic-agnostic separation emerges as subacute processes develop (4,17).

Temporally, our data follow the expected kinetics. Within 48–96 hours of life, a single CLIA-based NSE measurement distinguished abnormal from normal CUS, with a cutoff value on day 3 of 43.2 µg/L. Although this threshold is numerically higher than the cutoff value on day 3 reported by Metallinou, the absolute values are not directly comparable across studies using different analytical platforms and calibrations.

Consistent with Metallinou, our phenotype-specific assays also show stronger early discrimination for PIVH than for PVL: the mean NSE on day 3 was higher in isolated PIVH than in PVL (84.8 vs. 63.4 µg/L), while still tracking the severity of PVL.

In addition to the previously observed differences in absolute NSE values—which likely reflect variations in assays and methodology—the main difference in results between our study and the two comparative studies is the broader early discrimination observed in our study: a single NSE value on day 3 discriminated between normal from any abnormal CUS (PVL and/or PIVH) in this cohort.

The different kinetics of NSE observed in PIVH compared with PVL reflect different pathophysiology. In PIVH, the rapid early increase in NSE arises from convergent mechanisms—direct neuronal and glial injury due to mechanical disruption, heme iron-mediated oxidative toxicity, acute inflammatory activation with cytokine release, secondary ischemia associated with pressure in periventricular tissues, and compromise of the blood-brain barrier—each of which promotes rapid membrane damage and rapid efflux of NSE into the circulation. In contrast, PVL is predominantly a white matter injury of premyelinating oligodendrocytes and developing axons, characterized by more gradual cycles of ischemia-reperfusion, excitotoxicity, microglial activation, and apoptosis, leading to slower kinetics of NSE. Consistent with this pathophysiology, in our study (48–96 hours), NSE on day 3 was lower in PVL than in isolated PIVH, but remained significantly higher than in neonates with normal CUS (P<0.05), indicating that NSE is not only a biomarker of hemorrhage, but also of white matter damage (18,19).

Regarding PVL specifically, neither the study by Metallinou et al. nor the study by Efstathiou et al. (4) showed a significant difference in NSE concentrations when PVL was analyzed in isolation. A direct comparison with Metallinou et al. (17) is not appropriate because PVL was categorized and reported differently (de Vries classification and primarily compared with IVH in a matched case-control setting), while our study uses Agut et al. (9) grading and lesion-specific analyses. For Efstathiou et al. (4), the very small subgroup of PVL (n=6) severely limits statistical power and does not allow a robust interpretation of the early dynamics of NSE in PVL.

The aforementioned differences in absolute NSE values between studies may be explained by methodological heterogeneity, in particular by test characteristics and analytical platforms.

Our CLIA-based approach (Abbott Alinity CMIA) differs significantly from published enzyme-linked immunosorbent assay (ELISA) methods and immunochemiluminometric assay (ICMA) techniques used in comparative studies. These different calibration standards and dynamic ranges make numerical thresholds non-replaceable across platforms, emphasizing the need for platform-specific validation.

The study populations also differed considerably in size. Our study included all preterm infants born <37 weeks of gestation, whereas the comparison cohorts used narrower gestational age time frames (28–33 and <34 weeks, respectively). Outcome definitions also differed: Efstathiou et al. (4) stratified the severity of encephalopathy by MRI-defined injury and subsequent developmental outcomes, while Metallinou et al. (17) used a matched case-control design with the degree of brain injury at discharge.

While our study yielded significant results, it is important to acknowledge several limitations.

The limited sample size, particularly the relatively small number of neonates with normal CUS findings, reduces the precision of specificity estimates and results in wide 95% CIs; therefore, diagnostic performance metrics—especially specificity—should be interpreted with caution. This constraint should be viewed in the context of a highly specific and vulnerable population of very preterm neonates, in whom recruitment of large single-center cohorts is inherently challenging.

In addition, the diagnostic cut-off was derived within the same dataset and should be considered exploratory, requiring validation in independent, preferably multicenter, cohorts.

Since it was a single-center trial, generalizability is currently limited until multicenter trials reproduce the findings. The higher rate of pathological CUS findings likely indicates our tertiary-care setting, in which only neonates with severe clinical indications are admitted.

The use of CUS as the sole reference standard introduces an imperfect gold standard bias, as CUS has limited sensitivity for noncystic PVL and diffuse white matter injury; consequently, reliance on CUS-detected lesions may underrepresent the true burden of brain injury. Some neonates classified as controls may therefore, in the absence of MRI, have harbored subtle white matter injury below the detection threshold of US, potentially leading to an underestimation of specificity. Accordingly, this study validates serum NSE only against US-detected brain injury, and further studies incorporating MRI are needed to define its incremental diagnostic value.

Future studies should examine serum NSE on day 3 of life in multicenter cohorts of preterm infants—explicitly including infants with PVL and PIVH—and establish reference interval values for NSE according to gestational age. Cross-platform comparisons (CLIA/ICMA vs. ELISA) are needed to obtain test-specific thresholds or conversion factors, prioritizing automated methods with high sensitivity. The development of NSE assays that could be applied in intensive care units could facilitate clinical application of this biomarker, allowing rapid decision-making.

Conclusions

Serum NSE levels measured on the third day of life were associated with CUS-detected brain injury in preterm infants, including PIVH and PVL. In this single-center cohort, NSE demonstrated promising diagnostic accuracy for identifying US-detectable brain injury. Because the proposed threshold was derived within the same dataset, these findings should be considered exploratory and require validation in larger, preferably multicenter, cohorts before clinical application.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STARD reporting checklist. Available at https://tp.amegroups.com/article/view/10.21037/tp-2025-aw-795/rc

Data Sharing Statement: Available at https://tp.amegroups.com/article/view/10.21037/tp-2025-aw-795/dss

Peer Review File: Available at https://tp.amegroups.com/article/view/10.21037/tp-2025-aw-795/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tp.amegroups.com/article/view/10.21037/tp-2025-aw-795/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.This study was approved by The Ethics Committee of the Faculty of Medicine, University of Montenegro (approval No. 409/3) and by The Ethics Committee of the Clinical Center of Montenegro (approval No. 03/01-12438/1). The study was conducted in accordance with the institutional ethical standards and with the Helsinki Declaration and its subsequent amendments.Written informed consent was obtained from the parents of all participants prior to inclusion in the study.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Walani SR. Global burden of preterm birth. Int J Gynaecol Obstet 2020;150:31-3. [Crossref] [PubMed]
2. Barfield WD. Public Health Implications of Very Preterm Birth. Clin Perinatol 2018;45:565-77. [Crossref] [PubMed]
3. Yao M, Mao SS. Research advances in the biomarkers of brain damage in preterm infants. Zhongguo Dang Dai Er Ke Za Zhi 2019;21:1138-43. [Crossref] [PubMed]
4. Efstathiou N, Soubasi V, Koliakos G, et al. Beyond brain injury biomarkers: chemoattractants and circulating progenitor cells as biomarkers of endogenous rehabilitation effort in preterm neonates with encephalopathy. Front Pediatr 2023;11:1151787. [Crossref] [PubMed]
5. Bersani I, Pluchinotta F, Dotta A, et al. Early predictors of perinatal brain damage: the role of neurobiomarkers. Clin Chem Lab Med 2020;58:471-86. [Crossref] [PubMed]
6. Tenovuo O, Janigro D, Sanchez JC, et al. Editorial: Biomarkers of Brain Damage - A Complex Challenge With Great Potential. Front Neurol 2021;12:664445. [Crossref] [PubMed]
7. Caramelo I, Coelho M, Rosado M, et al. Biomarkers of hypoxic-ischemic encephalopathy: a systematic review. World J Pediatr 2023;19:505-48. [Crossref] [PubMed]
8. Perrone S, Grassi F, Caporilli C, et al. Brain Damage in Preterm and Full-Term Neonates: Serum Biomarkers for the Early Diagnosis and Intervention. Antioxidants (Basel) 2023;12:309. [Crossref] [PubMed]
9. Agut T, Alarcon A, Cabañas F, et al. Preterm white matter injury: ultrasound diagnosis and classification. Pediatr Res 2020;87:37-49. [Crossref] [PubMed]
10. Vizin T, Kos J. Gamma-enolase: a well-known tumour marker, with a less-known role in cancer. Radiol Oncol 2015;49:217-26. [Crossref] [PubMed]
11. Celtik C, Acunas B, Oner N, et al. Neuron-specific enolase as a marker of the severity and outcome of hypoxic ischemic encephalopathy. Brain Dev 2004;26:398-402. [Crossref] [PubMed]
12. Efstathiou N. Biomarkers in neonatal brain injury: interpreting research into clinical practice. In: Rajendram R, Preedy VR, Patel VB, editors. Biomarkers in Trauma, Injury and Critical Care. Biomarkers in Disease: Methods, Discoveries and Applications. Cham: Springer; 2022:1-17.
13. Chaparro-Huerta V, Flores-Soto ME, Merin Sigala ME, et al. Proinflammatory Cytokines, Enolase and S-100 as Early Biochemical Indicators of Hypoxic-Ischemic Encephalopathy Following Perinatal Asphyxia in Newborns. Pediatr Neonatol 2017;58:70-6. [Crossref] [PubMed]
14. Surkov D. Prognostic values of serum neuronal biomarkers NSE and protein S-100 in acute period of severe hypoxic-ischemic encephalopathy in term infants. ScienceRise: Medical Science 2018;6:32-41. [Crossref]
15. Giuseppe D, Sergio C, Pasqua B, et al. Perinatal asphyxia in preterm neonates leads to serum changes in protein S-100 and neuron-specific enolase. Curr Neurovasc Res 2009;6:110-6. [Crossref] [PubMed]
16. Abbasoglu A, Sarialioglu F, Yazici N, et al. Serum neuron-specific enolase levels in preterm and term newborns and in infants 1-3 months of age. Pediatr Neonatol 2015;56:114-9. [Crossref] [PubMed]
17. Metallinou D, Karampas G, Pavlou ML, et al. Serum neuron-specific enolase as a biomarker of neonatal brain injury: new perspectives for the identification of preterm neonates at high risk for severe intraventricular hemorrhage. Biomolecules 2024;14:434. [Crossref] [PubMed]
18. Volpe JJ. Neurology of the Newborn. 6th ed. Philadelphia: Elsevier; 2018.
19. Back SA. White matter injury in the preterm infant. Brain Pathol 2017;27:245-52.