Prognostication plays an essential role in decision making by offering patients and their families the necessary information to set realistic and achievable care goals. It helps determine eligibility for specific benefits and targets interventions to those who are most likely to benefit. Prognostication includes three main components: clinicians estimate the likelihood of a particular outcome over a certain period using their clinical judgment or other tools; this estimate is shared with the patient according to their preferred way of receiving information; and the patient or their surrogate uses this information to make informed clinical decisions (1).

Age, stroke severity, stoke mechanism, infarct location, comorbid conditions, clinical findings, and related complications influence stroke prognosis. Whether reperfusion therapies (fibrinolysis and/or thrombectomy) are conducted and the quality of stroke unit care given, including early start of physical rehabilitation and prompt introduction of secondary prevention measures, also influence the outcome of ischemic stroke (2).

Patient history, clinical examination and imaging technology have been the cornerstone in assessing stroke patients, guiding therapeutic decisions, and monitoring disease progression. Akin to the established gold standard in Cardiology, where biomarkers such as high sensitivity troponins have revolutionized the assessment of myocardial infarction (3), incorporating biomarkers into stroke management holds immense potential. These biomarkers offer a promising avenue to identify patients at heightened risk of severe disease, tailor treatment strategies, predict the response to reperfusion therapies and reasonably predict the overall prognosis and outcomes. While numerous proteins serve as markers of brain tissue damage, inflammation and coagulation/thrombosis, their utility is hindered by a lack of specificity to ischemic stroke, given that various other diseases processes can also damage the brain tissue (4).

For a biomarker to prove useful in stroke management, it should meet some basic criteria: be specific to the brain tissue, rise immediately within hours of a tissue insult, proportionally reflect the extent of brain damage, and ultimately serve as a reliable prognostic indicator for the event (5). Numerous biomarkers have been associated with short-and long-term clinical outcomes after stroke, but most of them have failed to improve the prediction capacities of conventional clinical variables. Biomarkers of ischemic brain injury include S100 calcium binding protein B (S-100B), neuron-specific enolase (NSE), myelin basic protein and glial fibrillary acid protein, among others (6).

The NSE is a dimeric intracellular neuronal glycolytic enzyme, primarily located in the cytoplasm of neurons, cells of the diffuse neuroendocrine system and erythrocytes. Elevated levels of NSE have been observed in response to sudden central nervous system events, such as cerebral infarction, subarachnoid hemorrhage, head injury, hypoxia, seizures, and cardiac arrest. These conditions are characterized by the disruption of the blood–brain barrier and subsequent damage of neuronal cells, leading to a leakage of NSE, which can be detected in cerebral spinal fluid but also in saliva or blood samples (7). Several studies propose NSE as a marker of brain damage following an ischemic event. There is also a correlation between NSE levels and acute ischemic stroke severity, infarction volume, extent of brain tissue damage [as clinically measured by the National Institutes of Health Stroke Scale (NIHSS) score], and poor functional outcomes (5, 8, 9).

NSE levels change dynamically following symptom onset, reflecting the necrosis of neuronal cells within the ischemic penumbra. Lower NSE levels are associated with clinical diffusion mismatch (7), serving as an alternative indicator for salvageable ischemic tissue that may be more responsive to interventions such as intravenous fibrinolysis and/or mechanical thrombectomy. However, high NSE levels are not exclusive to ischemic stroke and can also be found in other diseases such as neuroendocrine cell cancers like small cell lung cancers, neuroblastomas, melanomas and carcinoid tumors (10).

Since its approval by the European Stroke Organization in 2008, intravenous thrombolysis with alteplase has been a recognized systemic reperfusion treatment for patients with acute ischemic stroke (11). The pivotal MR CLEAN study published in 2015 further solidified the landscape by demonstrating the safety, efficacy and favorable impact on functional outcomes associated with intraarterial thrombectomy (12).

Building upon these advancements, our study aimed to evaluate the relationship between NSE levels at patient admission and 48 h post-reperfusion therapies, and functional outcome in ischemic stroke patients.

A total of 85 consecutive patients who fulfilled the established criteria for reperfusion treatment (either thrombolysis, thrombectomy or both) were included in the study. Of these patients six were excluded: two patients due to follow-up loss at 90 days post-stroke, three due to hemolyzed blood samples that invalidated the NSE level result, and one due to lung cancer history. Consequently, 79 patients (50.6% female; mean age 69.2 ± 14.4 years) were enrolled into the present study. The median time from symptom onset to hospital door was 113 min and the median time from symptom to reperfusion treatment as 165 min (when patients underwent both treatments, the time of rtPA bolus was chosen as the reperfusion treatment time, as it was the first treatment applied). The median NIHSS score was 10 points (range 0 to 30) before, and 8 points (range 0 to 32) 24 h after reperfusion therapies. The median NSE level at admission (NSE₀) was 16.3 ng/mL (4.7–45.1) and the median NSE level at 48 h post-reperfusion therapies (NSE_48h) was 17.5 ng/mL (4.7–117.0). At 90 days post-stroke, patients were assessed and distributed into two groups according to the mRS score at that time: 44 (55.7%) were placed in the favorable outcome group with a mRS score of 0–2 points, and 35 (44.3%) were in placed in the unfavorable outcome group with a mRS score of 3–6 points. Both groups were compared regarding baseline characteristics as shown in Table 1. There were no significant differences between the two groups in terms of sex, history of DM, hyperlipidemia, AF, previous stroke, smoking, CAD, time from symptom onset to hospital-door, time from symptom onset to treatment, fibrinolysis or thrombectomy and ICH. However, compared to favorable outcome group, patients in the unfavorable outcome group were more likely to be older (p = 0.006), have history of HTN (p = 0.012), HF (p = 0.019), and history of previous medical treatment (p = 0.029). They also presented a higher NIHSS score both at admission (p < 0.0001) and at 24 h (p < 0.0001) post-reperfusion therapies. Lastly, they were also more likely to have suffered a TACI type-stroke (p < 0.0001), post-stroke medical complications (p < 0.0001) and sICH (p = 0.006) (Table 1).

Of the 14 patients with ICH, 8 were asymptomatic, 6 had sICH and 1 died of sICH within 7 days after treatment. Among the 13 deceased patients, 4 died due to malignant infarct, 4 due to nosocomial pneumonia, 3 due to sepsis (urinary tract and endocarditis primary infections), 1 due to sICH and 1 due to hemorrhagic shock (gastric ulcer).

This study evaluated the predictive value of neuron-specific enolase (NSE) levels in determining stroke outcomes for patients undergoing reperfusion therapies. It was found that older age, history of hypertension (HTN), heart failure (HF), and previous medication use were significantly associated with unfavorable outcomes (mRS score of 3–6). Higher NIHSS scores at baseline and 24 h post-reperfusion were linked to worse outcomes and increased mortality. Notably, the study revealed that higher NSE levels at 48 h post-stroke were significantly associated with more severe outcomes and greater neurological disability, while NSE levels at admission showed a non-significant trend towards worse outcomes.

Clinicians are often expected to predict outcome after stroke, whether by patient, family, or other healthcare workers. There are a wide variety of factors that influence stroke prognosis, including age, stroke severity, stroke mechanism, infarct location, reperfusion treatments, comorbid conditions, neurological disability, and stroke complications (23, 24).

The modified Rankin Scale has been used as a measure of stroke-related handicap and is frequently used as a global measure of the functional impact of stroke. In addition, the mRS at 90 days after intravenous thrombolysis or endovascular interventions for acute ischemic stroke is the proposed “core metric” of most interventional trials and worldwide comprehensive stroke centers (25). The mRS score shows moderate correlation with the volume of cerebral infarction (26).

NSE has predictive value for determining severity and early neurobehavioral outcomes after stroke (9). However, the role of NSE in predicting long-term outcomes is still an evolving area, and evidence regarding its role in ischemic stroke patients undergoing reperfusion therapies is still scarce.

In this study we demonstrated that risk factors such as older age, a history of HTN, HF and previous medication were associated with unfavorable outcomes (mRS score of 3–6 points). Previous studies (27–29) have shown a clear associated between age, history of HTN and HF with stroke prognosis. In our study, hyperlipidemia and CAD showed a non-significant trend to be more prevalent among patients with unfavorable outcomes. Recent study by Chang et al. (30) showed a poorer outcome in stroke patients with CAD. A study from Menet et al. (31), mention that hyperlipidemia exacerbate vascular damage and is importantly associated to an increased risk of mortality in stroke patients.

According to Muir et al. (32), the NIHSS score offers the best specificity, sensitivity, and accuracy for predicting prognosis, with high scores correlating proportionally with higher mortality rates (33). As was demonstrated by Chalos study et al. (34), our study showed that higher NIHSS₀ and NIHSS₂₄ post-reperfusion therapies are associated with unfavorable outcome and higher mortality rates.

In Yang’s study (35), TACI type-strokes exhibited worse 3-month clinical outcomes and higher mortality rates. Similarly, in our study, patients with ischemic stroke undergoing reperfusion therapies, we found that the TACI type-stroke was associated with stroke severity, unfavorable outcomes (85.7%) and mortality (100%) when compared to non-TACI type-strokes (PACI, LACI and POCI).

Bustamante et al. (36) reported that post-stroke complications significantly impact stroke outcomes. In our study, stroke complications were statistically significantly associated with unfavorable outcomes (p < 0.0001). Thirty-nine patients (49.4%) experienced post-stroke complications, with pneumonia and urinary tract infection being the most frequent. Although infections commonly occur after stroke and are strongly associated with an unfavorable outcome in these patients, effective management strategies for post-stroke infection remain scarce, presenting a significant challenge for stroke care (37).

Our study demonstrated that the median values of the biomarker CRP at admission were significantly higher among patients with unfavorable outcomes. Two studies (38, 39) indicate a strong association of poor outcomes with high sensitivity (hs)-CRP measurements at 24–48 h and 7 days post-stroke, reflecting impairment of the recovery process due to prolonged inflammation after ischemic stroke. Similar associations were also observed with CRP values at admission (40).

Since 2005 (41), we have known that NSE levels are higher in stroke patients than in controls. However, it was after Zaheer et al. (42) that a correlation was made between NSE levels on day 1, infarct volume and functional outcomes at 30 days post-stroke. Consistent with this, previous studies focusing on stroke patients not-submitted to reperfusion treatments suggested that the initial NSE level positively associated with the degree of neurological deficit (41). One single study (43) reported that NSE values did not differ between patients who underwent intravenous thrombolysis alone versus conservative medical treatment.

In our study, the NSE_48h levels were statistically significantly associated with an unfavorable outcome and the NSE₀ levels were non-significantly higher in the unfavorable outcome group. Wunderlich et al. (44) reported that the release pattern of NSE rises 2–3 h after stroke onset, then decreases until 12 h, followed by a secondary increase until the last measurement on day 5 after stroke onset, probably reflecting secondary mechanism of brain damage, ongoing neuronal cell death or persistent disturbance of the blood–brain barrier. NSE levels have been shown to change dynamically after an ischemic stroke. A pathological neurovascular status on admission resulted in a significantly higher release of NSE from 48 h onward, although NSE concentrations from 18 h onward were highly correlated with the severity of the corresponding neurological deficit as quantified by the NIHSS. Considering that reperfusion therapies may increase the odds of vessel recanalization rates, possibly interfering with NSE levels, we measured the NSE level at admission (immediately before reperfusion therapies), and at 48 h post-reperfusion therapies.

A previous study (45) showed that the NSE level at 24 h post-intravenous fibrinolysis highly correlated with the severity of the corresponding neurological deficit as quantified by the NIHSS score at 24 h post-fibrinolysis.

There are several possible explanations for this finding, but the explanation that garners the most consensus is related with the penumbra area involved in the ischemic stroke. Cells in the ischemic penumbra, as they suffer necrosis, release NSE that passes from the cerebral spinal fluid to the peripheral blood through an impaired blood–brain barrier (46). Patients submitted to thrombolysis show decreased NIHSS scores as the occluded artery recanalizes and the ischemic penumbra area reperfuses, leading to a decrease in NSE levels. However, no studies have been conducted with ischemic stroke patients submitted to thrombectomy. On the other hand, patients who poorly respond to thrombolysis (considered as no or subpar recanalization) have larger core and smaller penumbra areas, leading to a higher NSE release from disrupted neuronal cells and both higher serum NSE levels and NIHSS scores.

In our study, higher NSE₀ levels showed a non-significant trend toward an unfavorable outcome (p = 0.059), reflecting the above correlation between NSE and the penumbra area (42). By the other hand, higher concentration of NSE_48h were statistically significantly associated with an unfavorable outcome (p = 0.008), reflecting a larger area of necrosis and infarct core and consequently higher neurological disability.

Two previous studies (42–45) showed a similar NSE level threshold for poor functional outcomes. As mentioned before, low NSE levels might be associated with clinical-diffusion mismatch, which has been a surrogate for brain tissue at risk of infarct.

By comparing the two ROC curves of NSE₀ and NSE_48h through the DeLong test, we found no significant differences between the two curves, leading us to conclude that the NSE₀ and NSE_48h have identical predictive value for unfavorable outcome.

Our study shown that NSE_48h and NIHSS_24h were independently and significantly associated with unfavorable outcome, after multivariate analysis adjusted for other risk factors for unfavorable outcome.

A previous study (9) showed that the NSE levels had high predictive value for the degree of short-term disability (measured with NIHSS at day 7). In the current study the NSE_48h levels positively correlated with the severity of neurological long-term disability (mRS at 90d), reflecting the NSE utility as a marker for destructive process in the central nervous system, a marker of neuron loss and consequently associated with higher disability.

This study has some limitations that must be considered. First, our data result from hyperacute ischemic stroke treated with reperfusion therapies, without a control group. Whether reperfusion treatments interfere with the blood–brain barrier, increasing the serum NSE concentration is still unknown. Thus, the association between NSE, especially at 48 h, and stroke severity estimated by the NIHSS score may have been overestimated. Second, a small sample size might have confounded the results of factors predicting the severity and outcome of stroke and resulted in wider confidence intervals. The authors are still recruiting more patients to increase the sample size and further ascertain this hypothesis. We also note that the titration of NSE levels is time consuming and vulnerable to the collecting methods, resulting in hemolysis and making the interpretation of hemolytic samples impossible. Eventually, in the future, a saliva NSE sample test could be more feasible, safe, and economic (47).

In conclusion, our results show that NSE levels at 48 h are associated with patient’s deficits assessed by the mRS. Like NIHSS score, NSE can be used, on a large scale in clinical practice, as a valuable tool in clinical management and prognostic estimation in ischemic stroke patients undergoing reperfusion therapies.

In the future, NSE levels might also be used in combination with CT perfusion of brain to study the core and penumbra area, maximizing the sensitivity of both.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving humans were approved by Hospital Ethics Committee, Hospital Dr. Nélio Mendonça. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

TiF: Writing – original draft, Writing – review & editing. AC: Writing – review & editing. LN: Writing – review & editing. CB: Writing – review & editing. MM: Writing – review & editing. PeF: Writing – review & editing. DN: Writing – review & editing. PaF: Visualization, Writing – review & editing. TeF: Writing – review & editing, Visualization. SB: Writing – review & editing. SF: Writing – review & editing, Data curation, Methodology. EH: Writing – review & editing. AS: Writing – review & editing, Writing – original draft.