Highlights
-
Endothelial activation biomarkers vWF and syndecan-1 are linked to poor clinical outcomes in aSAH patients, highlighting their potential prognostic value.
-
No significant association was found between endothelial biomarkers and the occurrence of DCI.
-
This study underscores the need for further research to clarify the role of endothelial activation in aSAH pathophysiology.
Introduction
In patients with aneurysmal subarachnoid hemorrhage (aSAH) who survive the initial 24 hours, the most important contributor to poor outcome is delayed cerebral ischemia (DCI), which occurs in approximately one out of three to four aSAH patients. Reference Rosengart, Schultheiss, Tolentino and Macdonald1,Reference Nieuwkamp, Setz, Algra, Linn, de Rooij and Rinkel2 The pathophysiology of DCI is believed to be multifactorial, including pro-inflammatory pathways, pro-coagulant activity and endothelial activation. Reference Budohoski, Czosnyka, Kirkpatrick, Smielewski, Steiner and Pickard3 However, diagnosis and treatment are currently still largely hampered by its incomplete understanding. Reference Dodd, Laurent and Dumont4,Reference Hoh, Ko and Amin-Hanjani5 The diagnosis of DCI in patients with reduced consciousness or sedated patients is challenging since clinical deterioration is difficult to assess in these patients. Biomarkers could be useful for prognostication of patients at risk for complications following aSAH and to improve identification of patients in need of additional diagnostics. Especially biomarkers involved in known pathophysiological pathways might contribute to earlier and more accurate diagnosis of DCI. Identification of biomarkers is additionally useful for quantification and better understanding of harmful processes that contribute to DCI damage.
Endothelial activation is a process critical for vascular repair and remodeling during injury or inflammation. It results in a permeable, pro-inflammatory and pro-thrombotic endothelium, which is essential for leucocyte migration and the elimination of toxins following aSAH. This process, however, can become harmful when persistent, uncontrolled or spreading. Reference Liao6 Endothelial activation and subsequent dysfunction are the driving forces behind microcirculatory dysfunction, which is nowadays regarded as an important mediator in the pathophysiology of DCI. Reference Motwani, Dodd, Laurent, Lucke-Wold and Chalouhi7
Von Willebrand factor (vWF) is regarded as a potent biomarker of endothelial activation and dysfunction. Other biomarkers of endothelial activation are E-selectin, thrombomodulin, syndecan-1 and matrix metalloproteinase-9 (MMP-9). E-selectin plays a role in the activation and rolling of leucocytes over activated endothelium. Reference Nissen, Mantle, Gregson and Mendelow8 Syndecan-1 is a component of the endothelial glycocalyx, which is cleaved by MMP-9 (among others) in case of endothelial damage, thus shedding syndecan-1 into the circulation. Thrombomodulin activates coagulation pathways and is therefore also used as a marker of endothelial activation. Reference Ait-Oufella, Maury, Lehoux, Guidet and Offenstadt9 Some studies have suggested that elevated levels of these biomarkers are associated with the development of DCI, Reference McGirt, Lynch, Blessing, Warner, Friedman and Laskowitz10–Reference Frijns, Kasius, Algra, Fijnheer and Rinkel17 as well as with worse prognosis. Reference Li, Hui and Ju14,Reference Hirashima, Nakamura, Endo, Kuwayama, Naruse and Takaku16,Reference Frijns, Fijnheer, Algra, van Mourik, van Gijn and Rinkel18 However, results have been inconsistent, Reference Nissen, Mantle, Gregson and Mendelow8,Reference Frijns, Kasius, Algra, Fijnheer and Rinkel17,Reference Chamling, Gross and Stoffel-Wagner19,Reference Kubo, Ogasawara and Kakino20 with most sample sizes being relatively small and the methods being too heterogeneous to pool the results. Therefore, we reevaluated these associations in a larger, prospective cohort.
In this prospective study, the association between serum levels of vWF, E-selectin, thrombomodulin, syndecan-1 and MMP-9 and the occurrence of DCI and poor clinical outcome six months after aSAH was evaluated. We hypothesized that elevated levels of these markers might be associated with the occurrence of DCI and subsequent poor outcome after aSAH.
Methods
Patients
We used serum samples collected from patients with aSAH who were admitted to the Amsterdam University Medical Centre between October 2018 and November 2020. Ninety-eight consecutive aSAH patients were screened for eligibility. Inclusion criteria were (1) SAH on non-contrast head CT or confirmation by lumbar puncture, (2) aneurysmal origin of the SAH on angiographic imaging (CT angiography or digital subtraction angiography), (3) 18 years or older and (4) first blood withdrawal within 24 hours of ictus and before aneurysm treatment. Patients (n = 21) who were administered tranexamic acid as part of the ULTRA-trial Reference Post, Germans and Tjerkstra21 were excluded because of several studies suggesting an ameliorating effect of tranexamic acid on levels of endothelial biomarkers. Reference Diebel and Liberati22–Reference Kim, Lee, Lee, Kim, Jun and Choi24 Patients underwent a standard follow-up appointment six months after SAH. A control cohort of patients with unruptured intracranial aneurysms (referred to as controls) was included to obtain reference values for all of the investigated markers. These patients visited the outpatient clinic of the Amsterdam University Medical Center within the same time frame. The study was conducted in accordance with the Declaration of Helsinki. The study protocol was approved by our Institutional Review Board (MEC 2017_318). Deferred informed consent was obtained from all patients or their legal representatives, except from those who suffered imminent death.
Data collection
Data were prospectively collected in our institutional SAH registry by a trained research nurse, and queries were handled by a neurosurgeon (BAC) and clinical epidemiologist (DV). Data included gender, age, smoking status, cardiovascular comorbidities (cardiovascular diseases, diabetes mellitus, hypertension, hypercholesterolemia), prior anticoagulant use, clinical condition at admission as graded by the World Federation of Neurological Surgeons (WFNS) scale, Reference Teasdale, Drake, Hunt, Kassell, Sano, Pertuiset and De Villiers25 the extent of aSAH as graded by the modified Fisher scale, Reference Frontera, Claassen and Schmidt26 aneurysm location (anterior/posterior circulation), treatment modality (surgical, endovascular, none), the occurrence of DCI and other complications (rebleeding, hydrocephalus, meningitis, seizures, delirium, pneumonia, sepsis, deep venous thrombosis), mortality at six months after aSAH and clinical outcome at six months after aSAH as assessed by the modified Rankin Scale (mRS) score. Reference Stienen, Visser-Meily, Schweizer, Hänggi, Macdonald and Vergouwen27 During a follow-up appointment, a certified research nurse conducted an interview using a standardized and validated method to assess the mRS score. Reference Wilson, Hareendran, Hendry, Potter, Bone and Muir28,Reference Janssen, Visser, Dorhout Mees, Klijn, Algra and Rinkel29
Sample collection
Venous blood samples of aSAH patients were collected in citrated tubes as soon as possible at the emergency department (day 0, baseline sample) and on days 3–5 and days 9–11 after aSAH ictus. For the controls, only one blood sample was drawn upon inclusion. Blood would be extracted from the central artery line in the days after admission if one had been inserted after ICU admission. Samples were centrifuged within 2 hours of collection by a standardized protocol and subsequently frozen and stored at -80°C. All samples were thawed once to transfer to aliquots with study ID and timing stamps and refrozen. Quantitative evaluation of the markers was performed by a Luminex-200 platform (manufactured by Bio-Rad), with kits from Bio-Techne, as described by the manufacturer. In this study, we employed a cost-effective and time-efficient approach by avoiding the utilization of duplicate samples. The samples were 50-fold diluted with the Calibrator Diluent RD6-52. We used 25 µl of diluted sample per well. The plates were centrifuged for 2 hours and read out by the Bioplex-200. Plate-to-plate variance was not accounted for; instead, we conducted a comparative analysis exclusively on the positive samples that were assessed on both plates. Also, we applied a calibration curve derived from one plate to all plates. To maintain consistency in the experimental procedure, we derived a calibration curve from one of the plates and subsequently applied this curve to all other plates. Values that fell below the detection limit were imputed with the lower limit of quantification derived from the calibration curve for univariate comparisons.
Outcome measurements
The outcome measurements in this study were DCI occurrence and poor outcome six months after aSAH. DCI was diagnosed according to the proposed definition of clinical DCI by Vergouwen et al: “the occurrence of a decrease in consciousness (two or more points on the total Glasgow Coma Scale or one point drop in the M-score) or focal neurological impairment, which lasts for at least one hour and cannot be attributed to any other cause by means of clinical assessment, CT or MRI scanning of the brain, and appropriate laboratory studies.” Reference Vergouwen, Vermeulen and van Gijn30 It is important to note that while neuroimaging is required to rule out other causes, radiological confirmation of DCI is not mandatory for this clinical diagnosis. We used the clinical definition of DCI because it aligns with current clinical practice, where management decisions are based on clinical signs and symptoms rather than imaging findings alone. In contrast, the radiological definition often identifies DCI too late for effective intervention, reducing its relevance in guiding treatment. Because we recognize the limitations that come with this clinical diagnosis, we performed a sensitivity analysis for radiological DCI. Radiological DCI is defined as “the presence of cerebral infarction on CT or MR imaging, which was not present on CT or MR imaging within 24–48 hours after SAH and cannot be attributed to the aneurysm treatment or other causes.” Clinical outcome at six months was assessed by the mRS and dichotomized into good (mRS 0–3) and poor (mRS 4–6) outcomes, as suggested by Stienen et al. Reference Stienen, Visser-Meily, Schweizer, Hänggi, Macdonald and Vergouwen27
Definitions
The location of the aneurysm was classified as either anterior or posterior circulation according to Rosner et al. Reference Rosner, Reddy and Neuroanatomy31 Rebleeding was defined as sudden neurological deterioration with increased subarachnoid blood on CT compared to earlier imaging or high clinical suspicion without any other explanation for the neurological deterioration. Hydrocephalus was defined as a gradual clinical deterioration based on the Glasgow Coma Scale, accompanied by either enlarged ventricles on non-contrast CT or elevated intracranial pressure. Meningitis was diagnosed through a positive liquor culture, performed after clinical suspicion (decreased consciousness, meningeal rigidity, fever or elevated inflammatory markers). Pneumonia was diagnosed upon clinical symptoms, with either consolidations on chest X-ray or a positive sputum culture. Further elaborations and definitions of the other complications can be found in the supplemental data.
Statistical analysis
Categorical data were described in proportions with percentages. Numerical data were described by mean and standard deviation (SD) or median and interquartile range (IQR) depending on normality. The cohort was dichotomized into patients with and without DCI and with good and poor clinical outcomes after aSAH. Baseline characteristics were compared between the DCI groups. For categorical variables, a chi-square test was used. For continuous variables, an independent T-test or a Mann–Whitney U test was used, depending on the normality of the data. The Shapiro–Wilk test was used to test for normality, with a test statistic (W) < 0.9 to reject the hypothesis of normality.
The samples were collected on three separate occasions after aSAH: T0 (at admission), T1 (days 3–5) and T2 (days 9–11). The biomarker concentrations for the separate time points were compared between aSAH patients and controls by independent T-test or Mann–Whitney U test. The median biomarker concentrations with IQRs of aSAH patients with and without DCI were plotted in a graph. The biomarker concentrations of patients with DCI and without DCI were compared for each separate time point by a T-test or Mann–Whitney U test, depending on the normality of the data. A complete case analysis was employed. For longitudinal analysis of biomarker concentrations in aSAH patients with and without DCI, a linear mixed effects model was used. Missing data were handled by available case analysis, where we retained the cases with partially incomplete data but excluded the cases with missing data at all time points. The sensitivity analysis for radiological DCI was performed similarly. For poor outcome, we dichotomized the group into good (mRS 0–3) and poor outcomes (mRS 4–6) and performed similar analyses as described for patients with and without DCI. SPSS version 28 was used for statistical analyses, and graphs were made by GraphPad Prism 9.
Results
Data were collected from 75 aSAH patients who had at least one available blood sample. Baseline characteristics are shown in Table 1. Twelve (16%) aSAH patients developed DCI with a median onset on day 9 (IQR: 7–10) after ictus. Patients with DCI were more often female (Table 1; p = 0.04), had more treatment-related aneurysm ruptures (Table 1; p = 0.02) and were more likely to develop meningitis (Table 1; p = 0.01) compared to patients without DCI. Clinical outcome at six months follow-up was available for 72 patients (4% lost to follow-up). A total of 30 (42%) had poor outcomes after six months, of whom 19 patients (26%) had died. Ten control patients with unruptured intracranial aneurysms were included, with one blood sample available.
Table 1. Baseline characteristics of 75 aSAH and 10 control patients
Data are N (percentage). Percentages might not equal 100 due to rounding and missing data. * = statistically significant difference by p < 0.05. #n = missing data. aSAH = aneurysmal subarachnoid hemorrhage; DCI = delayed cerebral ischemia; mFisher = modified Fisher; SD = standard deviation; WFNS = World Federation of Neurosurgical Societies.
Comparison of aSAH patients and controls showed that serum levels of vWF at admission were significantly elevated compared to control patients (Table 2, p < 0.001), whereas E-selectin levels were lower on days 3–5 in aSAH patients than in controls (Table 2, p = 0.02). Syndecan-1 was lower in aSAH patients at admission (p = 0.02) than in controls. The concentrations of thrombomodulin and MMP-9 were not significantly different between aSAH patients and controls.
Table 2. Median and interquartile range (IQR) marker concentration in the control group (n = 10), compared to subarachnoid hemorrhage (SAH) patients at admission (T0), on days 3–5 (T1) and days 9–11 (T2)
* = statistically significant difference by p < 0.05. N does not equal the total number of included aSAH patients due to logistical difficulties leading to missing data. ASAH = aneurysmal subarachnoid hemorrhage; IQR = interquartile range; MMP-9 = matrix metalloproteinase-9; vWF = von Willebrand factor.
Analyses conducted at individual time points revealed no statistically significant differences between patients who developed DCI and patients who did not (Figure 1A–E, p-values and medians with IQR listed in Table S1 in the supplemental material). The biomarkers showed no significant longitudinal differences between patients with DCI and those without DCI (p-values listed in Table S1 in the supplemental material). In the sensitivity analysis for radiological DCI (n = 5), no statistically significant differences were found between radiological DCI patients and no radiological DCI patients (Table S2 in the supplemental material). For poor clinical outcomes, vWF concentrations at admission (T0) were significantly elevated, compared to good clinical outcomes at six months after aSAH (Figure 2A, p = 0.02). Syndecan-1 was significantly elevated on days 9–11 (T2) in patients with poor outcomes (Figure 2D, p < 0.05). For the other biomarkers, no significant differences could be detected when comparing patients with good outcomes to those with poor outcomes (Figure 2B, C and E; p-values and medians with IQR listed in Table S3 in the supplemental material). There were no significant longitudinal differences in any of the marker concentrations between patients with good and poor clinical outcomes (p-values listed in Table S3 in the supplemental material).
Figure 1. Median levels with interquartile range of (A) von Willebrand factor (vWF), (B) E-selectin, (C) thrombomodulin, (D) syndecan-1 and (E) matrix metalloproteinase-9 (MMP-9) in patients with and without delayed cerebral ischemia (DCI) on admission (T0), on days 3–5 (T1) and days 9–11 (T2).
Figure 2. Median levels with interquartile range of (A) von Willebrand factor (vWF), (B) E-selectin, (C) thrombomodulin, (D) syndecan-1 and (E) matrix metalloproteinase-9 (MMP-9) in patients with poor outcome and good outcome, on admission (T0), on days 3–5 (T1) and days 9–11 (T2).
Discussion
The current prospective study in a large, consecutive cohort of aSAH patients did not detect significant differences between patients with and without DCI in serum levels of vWF, E-selectin, thrombomodulin, syndecan-1 or MMP-9. Following aSAH, the most prominent change compared to controls is a prompt increase in serum levels of vWF. In addition, vWF and syndecan-1 were significantly elevated, respectively, on admission and in the late phase, in patients who had poor clinical outcomes at six months after aSAH.
Levels of vWF were elevated directly after hospital admission when compared to control patients, suggesting prompt endothelial activation after hemorrhage. While several previous studies demonstrated associations between elevated vWF and the development of DCI Reference McGirt, Lynch, Blessing, Warner, Friedman and Laskowitz10,Reference Li, Hui and Ju14,Reference Hirashima, Nakamura, Endo, Kuwayama, Naruse and Takaku16,Reference Frijns, Kasius, Algra, Fijnheer and Rinkel17 or poor clinical outcomes, Reference Li, Hui and Ju14,Reference Frijns, Fijnheer, Algra, van Mourik, van Gijn and Rinkel18 our findings did not reveal a direct link to DCI development. However, we did find that elevated vWF levels at hospital admission are associated with a six-month poor outcome. vWF levels showed a decrease in the 3–5 days after admission in both groups, resulting in the attenuation of the initial disparity between good- and poor-outcome patients. A possible explanation for the initial greater rise of vWF in poor-outcome patients could be that there is more initial neurological and vascular damage in this group, leading to more endothelial activation, more systemic complications and worse outcomes.
Soluble syndecan-1, an indicator of endothelial glycocalyx degradation, has previously been shown to increase at DCI onset in a small case series including three patients. Reference Bell, Rhind, Di Battista, Macdonald and Baker12 Our prospective cohort study showed that syndecan-1 levels were higher in controls than in aSAH patients at admission. This unexpected finding could suggest an altered syndecan-1 metabolism or clearance in the acute phase after hemorrhage. In all aSAH patients, syndecan-1 seemed to increase over time, suggesting that syndecan-1 is a late-phase marker that represents a delayed response. In the development of a diagnostic test, it is customary to first compare test results of patients with healthy control subjects. Reference Bouter and MCJMv32 We chose patients with unruptured intracranial aneurysms since they represent a population with a similar baseline regarding age, sex and vascular condition, without the acute critical illness associated with hemorrhage. The subsequent step in identifying new biomarkers would be investigating if the combined elevation of vWF and depression of syndecan-1 is aSAH specific or is also present in other critically ill patients with, for example, sepsis or traumatic brain injury. Reference Bouter and MCJMv32
Intriguingly, late-phase (days 9–11) syndecan-1 elevation was associated with poor outcome. We found a nonsignificant trend of lower syndecan-1 in DCI patients compared to patients without DCI, suggesting that the poor prognosis associated with late-phase elevated syndecan-1 is not provoked by DCI development. Secondary brain injury can occur after aSAH, caused by hydrocephalus or cerebral edema due to inflammatory processes or a positive fluid balance. Reference Wiegers, Lingsma and Huijben33 This is thought to lead to downstream cascades, including the disruption of the blood-brain barrier (BBB), cerebral edema, neuroinflammation and oxidative stress. Reference Lauzier, Jayaraman and Yuan34 Specifically, BBB disruption causes endothelial dysfunction and leads to further vasogenic edema. In addition, infused fluids are shown to contribute to glycocalyx shedding in critically ill patients. Reference Hippensteel, Uchimido and Tyler35 Collectively, these processes are associated with poor outcomes. Reference Cahill, Calvert and Zhang36 The temporal alignment of this biomarker elevation with the later phase of aSAH supports its relevance in terms of disturbed recovery and long-term worse prognosis. Given that metalloproteinases cleave syndecan-1 and may mediate shedding and hence endothelial dysfunction, we measured levels of MMP-9. We found a trend toward an increase in MMP-9 levels on admission compared to controls, but in contrast to earlier studies, Reference McGirt, Lynch, Blessing, Warner, Friedman and Laskowitz10,Reference Fischer, Dietmann and Beer13 no differences for DCI or poor outcome. Possibly, other MMP proteins play a role in the observed syndecan-1 shedding following aSAH. Reference Endo, Takino and Miyamori37 Our results regarding E-selectin align with most preceding studies reporting no significant difference in plasma concentrations regarding patients with and without DCI. Reference Nissen, Mantle, Gregson and Mendelow8,Reference Frijns, Kasius, Algra, Fijnheer and Rinkel17,Reference Chamling, Gross and Stoffel-Wagner19,Reference Kubo, Ogasawara and Kakino20 Thrombomodulin has been proposed as a marker of endothelial activation and dysfunction in several studies investigating other vascular diseases. Reference Drożdż, Łątka, Drożdż, Sztefko and Kwinta38,Reference Nadar, Al Yemeni, Blann and Lip39 We hypothesized that this association could possibly be extrapolated to DCI after aSAH because of the partially overlapping endothelial activation processes that occur in DCI. Nonetheless, our findings in the main analysis did neither reveal any association between thrombomodulin concentrations and DCI development nor with poor outcome. In our sensitivity analysis for radiological DCI, no statistical differences were found, which is very probable due to the low numbers of radiological DCI in the analysis (n = 5, with additionally missing data for days 3–5 and days 9–11). A nonsignificant trend of elevated thrombomodulin was found in the patients with radiological DCI compared to no radiological DCI. This, however, may also indicate that the biomarkers for radiological DCI are interchangeable with those without radiological DCI. In groups this small, the likelihood of finding a statistically significant difference is low, but that does not imply that the difference exists.
A potential cause for the disagreement between previous studies and ours might be that there is great methodological variation, including the definition of DCI. Over the past decades, DCI has been addressed using a variety of terms, including cerebral vasospasm and delayed ischemic neurological deficit, corresponding with slightly different clinical and radiological definitions, leading to variety in patient selection. In 2010, an international panel of experts proposed the currently used definition of DCI for use in clinical trials and observational studies. Reference Vergouwen, Vermeulen and van Gijn30 Additionally, the timing of blood sampling varied greatly between the different studies, which further complicates comparison. Regarding the great variety in analysis methods, some studies compared mean peak serum concentrations Reference McGirt, Lynch, Blessing, Warner, Friedman and Laskowitz10,Reference Chamling, Gross and Stoffel-Wagner19 or only compared the delta concentrations before and after the onset of DCI. Reference Wang, Lin and Yang11,Reference Frijns, Kasius, Algra, Fijnheer and Rinkel17 Additionally, there were studies that applied distinct criteria for their patient selection by only including WFNS 1 and 2 patients Reference Nissen, Mantle, Gregson and Mendelow8 and aSAH patients who had neurosurgical clipping procedures Reference Kubo, Ogasawara and Kakino20 or excluding all patients with comorbidities or those using anticoagulants. Reference Li, Hui and Ju14,Reference Hirashima, Nakamura, Endo, Kuwayama, Naruse and Takaku16 The heterogeneous nature of patient selection criteria further complicates comparisons and warrants exploration of potential interactions of endothelial biomarkers with DCI in better-specified subgroups. However, the statistical power in this study was deemed insufficient for further stratification based on patient characteristics. Heterogeneity in the used definitions of DCI and great methodological variety may lead to inaccurate comparison of the available data. The field would benefit from more standardized definitions and methodologies to enhance comparability and facilitate accurate data interpretation.
An alternative explanation for the lack of association between the biomarkers and DCI may be that the biomarkers determined in our study do not comprehensively represent the endothelial activation process, considering the existence of numerous other established biomarkers for endothelial activation. It has been proposed that the endothelial function might be best assessed by a combination of circulating biomarkers and assessment of vasomotor response. Reference Deanfield, Halcox and Rabelink40,Reference Leite, Borges-Canha, Cardoso, Neves, Castro-Ferreira and Leite-Moreira41 Alternatively, it is plausible that the contribution of endothelial activation to DCI’s pathogenesis might be overestimated since DCI is a complex and multifactorial disease where different pathways play different roles to varying degrees. Considering this, the development of an (AI-driven) multi-biomarker model may yield superior predictive value over a single marker approach. This model should potentially include clinical parameters such as transcranial Doppler and electroencephalogram data, Reference Carrera, Schmidt and Oddo42 mFisher score Reference van der Steen, Leemans and van den Berg43 and promising markers of different pathophysiological pathways of DCI, such as haptoglobin polymorphisms 2–1 and 2–2, ADAMTS13, neutrophil/leucocyte ratio or P-selectin. Reference Tjerkstra, Labib and Coert44
The strengths of our study are its prospective controlled study design and the utilization of multiple markers readily accessible in peripheral blood. We pioneered the investigation of syndecan-1 and thrombomodulin, being the first to explore these markers in the context of DCI occurrence. Additionally, our investigation involved a larger cohort compared to the majority of prior studies, and we applied a linear mixed model for the longitudinal analysis of biomarker data.
Even though this study has a larger sample size than previous cohorts, the power to demonstrate significant results is still limited due to the relatively small proportion of aSAH patients who developed DCI. Nevertheless, the lack of elevated biomarker levels in DCI patients does not seem to be driven by low power, as the point estimate aims in the opposite direction. Short-term fluctuations of the biomarkers were not accounted for. As a result, we are unable to determine how variable the markers were over the course of an hour or a few hours. Great temporal variability can however contribute to the observed large interquartile ranges. Logistical difficulties such as hospital transfers, withdrawn consent for further sampling, death or overdue processing of the samples resulted in missing data. From a statistical perspective, limitations include the use of multiple comparisons, which can increase the risk of false-positive findings. However, to interpret our results, we depended on the biomarkers’ biological plausibility, consistency over time and alignment with the clinical setting.
Conclusion
In this prospective study, we found no association between levels of vWF, E-selectin, thrombomodulin, syndecan-1 and MMP-9 and the occurrence of DCI after aSAH. Therefore, these biomarkers seem to have limited prognostic and diagnostic value in relation to DCI. It is possible that the studied biomarkers are not representative of the endothelial activation process during DCI. However, we did observe that patients with poor outcomes exhibited higher vWF levels on admission and higher syndecan-1 levels on days 9–11. Our findings suggest that enhanced endothelial activation may contribute to poor prognosis in aSAH patients. To improve prognostic and diagnostic approaches for aSAH patients and DCI, it may be beneficial to explore alternative pathways and develop a multimodal model that incorporates various biomarkers, baseline characteristics and clinical parameters.
Supplementary material
The supplementary material for this article can be found at http://doi.org/10.1017/cjn.2025.32.
Data availability statement
Data will be shared on request. Requests for noncommercial purposes can be submitted to the corresponding author.
Acknowledgment
Special thanks to Anita Tuip-de Boer from the research department of intensive care for the quantitative analysis of our samples.
Author contributions
MFS contributed to the study design, data analysis, data interpretation, writing of the article and drafting of the figures. MAT contributed to the study design, data interpretation and writing and revising the article critically. DV and NPJ contributed to the study design and data interpretation and revised the article critically. WPV contributed to the study design and revised the article critically. BAC and RP revised the article critically.
Funding statement
No funding was received to support this project.
Competing interests
The author and all coauthors declare no conflict of interest, as stated in the submitted ICMJE forms.
Ethical statement
The study was conducted in accordance with the Declaration of Helsinki. The study protocol was approved by our Institutional Review Board (MEC 2017_318). Deferred informed consent was obtained from all patients or their legal representatives, with the exceptionof those who suffered imminent death.
Instructions to authors
The submitted manuscript complies with the instructions to authors. The authorship requirements have been met, and the final manuscript was approved by all authors. This article has not been previously published, and it is not under consideration by any other journal.
Reporting checklist
The STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) checklist was employed to assess and ensure adherence to the quality requirements for a cohort study in this article.
Open access
