Skip to main content
Download PDF

Proteomic profile and predictive markers of outcome in patients with subarachnoid hemorrhage

Clinical Proteomics volume 21, Article number: 51 (2024) Cite this article

Abstract

Background

The molecular mechanisms underlying development of posthemorrhagic hydrocephalus (PHH) following subarachnoid hemorrhage (SAH) remain incompletely understood. Consequently, treatment strategies tailored towards the individual patient remain limited. This study aimed to identify proteomic cerebrospinal fluid (CSF) biomarkers capable of predicting shunt dependency and functional outcome in patients with SAH in order to improve informed clinical decision making.

Methods

Ventricular CSF samples were collected twice from 23 patients with SAH who required external ventricular drain (EVD) insertion (12 patients with successful EVD weaning, 11 patients in need of permanent CSF shunting due to development of PHH). The paired CSF samples were collected acutely after ictus and later upon EVD removal. Cisternal CSF samples were collected from 10 healthy control subjects undergoing vascular clipping of an unruptured aneurysm. All CSF samples were subjected to mass spectrometry-based proteomics analysis. Proteomic biomarkers were quantified using area under the curve (AUC) estimates from a receiver operating curve (ROC).

Results

CSF from patients with SAH displayed a distinct proteomic profile in comparison to that of healthy control subjects. The CSF collected acutely after ictus from patients with SAH was moreover distinct from that collected weeks later but appeared similar in the weaned and shunted patient groups. Sixteen unique proteins were identified as potential predictors of shunt dependency, while three proteins were identified as potential predictors of functional outcome assessed six months after ictus with the modified Rankin Scale.

Conclusions

We here identified several potential proteomic biomarkers in CSF from patients with SAH capable of predicting (i) shunt dependency and thus development of PHH and (ii) the functional outcome assessed six months after ictus. These proteomic biomarkers may have the potential to aid clinical decision making by predicting shunt dependency and functional outcome following SAH.

Background

Aneurysmal subarachnoid hemorrhage (SAH) is associated with high mortality and morbidity [1]. Despite neurosurgical intervention and specialized neuro-intensive care, more than one third of patients with SAH develop an unfavorable long-term functional outcome. An unfavorable outcome is characterized by an affected cognitive performance including impaired memory and executive functions as well as sleep disturbances and fatigue [2, 3] and it remains incompletely understood why some patients experience worse outcomes [4]. Immediately after the hemorrhage, extravasated blood enters the cerebrospinal fluid (CSF) spaces and the intracranial pressure (ICP) rises. Several pathological alterations commonly occur within minutes to days, i.e., cerebral vasospasms, acute and delayed brain ischemia, inflammation, seizures, and cortical spreading depression that all contribute to the risk of an unfavorable outcome [5,6,7,8]. The molecular mechanisms underlying the initial injury expansion following SAH and the factors contributing to an unfavorable outcome remain to be fully elucidated, which limits effective (pharmaceutical) treatment options [9]. Posthemorrhagic hydrocephalus (PHH) is a common and serious complication of SAH. Extravasated blood can obstruct the CSF drainage pathways either directly via formation of blood clots or indirectly by promoting meningeal inflammation and subsequent fibrosis [8, 10, 11]. Hypersecretion by the CSF-producing tissue, the choroid plexus, in response to the sudden presence of hemorrhage-derived molecules or inflammatory markers in the CSF may, moreover, contribute to the pathological CSF accumulation observed in PHH [11,12,13,14]. Acute PHH is often treated with external ventricular drainage (EVD) to alleviate the ICP and mitigate neurological damage [15, 16]. Although lifesaving, EVD placement comes with a significant risk of infection, which increases with the duration of the EVD [17,18,19,20]. ICP alleviation weighed against removal of the EVD as early as possible is often a difficult clinical decision in management of PHH in the acute and subacute phase. In some patients with SAH, CSF dynamics normalize after the initial hemorrhage, and they can successfully be weaned from the EVD. However, a considerable proportion of the patients fails EVD weaning and undergoes subsequent shunt insertion due to development of chronic PHH. A recent review reported the published risk of developing shunt-dependent PHH after SAH to be between 8 and 63% [21]. Such wide variation implies uncertainty regarding both the acute and subacute management of PHH and prediction of shunt-dependent chronic PHH development [15, 16, 22, 23]. Although several publications have investigated clinical, radiological, and treatment features as predictive indicators for PHH development following SAH, none have proven valid as predictors of shunt dependency [15, 16, 22,23,24,25,26,27,28,29]. Our group recently identified a subset of inflammatory proteins in ventricular CSF from patients with SAH as potential biomarkers of shunt dependency and functional outcome [30]. This suggests that neuroinflammation plays a pivotal role in the development of PHH following SAH. However, to delineate the precise pathological alterations following SAH and elucidate why some patients develop PHH while others do not, the molecular and cellular alterations occurring within the brain of patients with SAH require further investigation. Here, we assessed the proteomic profile in CSF from patients with SAH by comparison to CSF from healthy individuals with unruptured aneurysms. We quantified whether certain proteomic patterns could be used to distinguish patients with SAH who went on to develop PHH from those who did not, and we identified possible biomarkers of shunt dependency and functional outcome. Such biomarkers could potentially serve as clinical tools in management of SAH.

Methods

Patients and CSF collection

This study included paired CSF samples collected from 23 patients with SAH (median age: 61 years; range: 34–77 years; 18 F/5 M) who were diagnosed and treated at the Department of Neurosurgery, Rigshospitalet, Copenhagen, Denmark, between June 2019 and September 2021. The CSF samples were collected as part of a prospective trial on EVD weaning (clinical trial identifier: NCT03948256 [31]). The first CSF samples (“start samples”) were collected from the SAH patients through their EVD either acutely within 24 h of ictus (n = 19) or as soon as possible thereafter (n = 4). The time interval from ictus to first CSF collection was on average 23 h, range: 2 h – 7 days. Of the 23 patients with SAH, 12 patients could be successfully weaned off the EVD (“weaned”) and did not require further CSF diversion, while 11 patients underwent ventriculoperitoneal shunt surgery upon EVD weaning due to chronic PHH development (“shunted”). The last CSF samples (“end samples”) were collected prior to shunt insertion or EVD withdrawal (n = 23 in total). The time interval between the paired CSF samples (start and end) was on average 19 days, range: 5–30 days. The CSF samples from the patients with SAH were collected directly for analytical purposes and the CSF therefore did not reside in the EVDs prior to collection. The control group consisted of 10 patients with unruptured aneurysms undergoing preventive surgery (vascular clipping) (median age: 60 years, range: 39–71 years, 6 F/4 M) from whom the CSF was collected from the basal cisterns during surgery prior to clipping of the aneurysm. All CSF samples were centrifuged at 2000 × g for 10 min at 4 °C within 2 h from collection, and the supernatant aliquoted in polypropylene microtubes (Sarstedt, Nümbrecht, Germany) and subsequently stored at -80 °C [32].

Functional outcome was assessed for each SAH patient six months after ictus by the modified Rankin Scale (mRS) [33]. Thirteen patients had a favorable functional outcome (mRS 0–2), while 10 patients had an unfavorable functional outcome (mRS 3–6), including two deceased patients (mRS 6). Written informed consent was obtained from all patients or their next of kin, depending on the capability of the patients, and the study was approved by the Danish National Committee on Health Research Ethics (approval no. H-19001474 and H-17011472/69197) and the Danish Data Protection Agency (VD-2019-210). Aliquots of the CSF samples have been analyzed for other components in unrelated studies [11, 30, 34,35,36].

Protein digestion and evotips loading

Human CSF sample preparation was performed on an Agilent Bravo Liquid Handling Platform (Agilent) according to an optimized version of previously published protocols [37, 38]. Briefly, CSF samples were aliquoted into a 96-well format plate and introduced to the Bravo Robot (Agilent). 20 µl CSF sample was mixed with 30 µl PreOmics Lysis buffer (P.O. 00001, PreOmics GmbH) and incubated at 95 °C for 10 min in order to denature proteins, reduce disulfide bridges, and alkylate cysteines [39]. After cooling the sample for 15 min at room temperature, trypsin and LysC (0.5 µg/µl, Promega) were added in a ratio of 1 µg enzyme to 100 µg proteins and the mixture incubated at 37 °C for 4 h. Similar protein quantity was observed in tested random samples (on average five samples per plate) using Nanodrop. The peptide mixtures were diluted in 100 µl 99% isopropanol, 1% Trifluoro-acetic acid (TFA) and desalted using two-gauge reversed-phase styrenedivinylbenzene (SDB-RPS) stage-tips. Afterwards, the stage-tips were washed using 200 µl 99% isopropanol, 1% TFA, followed by 200 µl 0.2% TFA. The purified peptides were eluted using 80% acetonitrile (VWR chemicals) containing 1% ammonia (Merck) and subsequently dried down. Peptides were resuspended in solvent A (0.1% formic acid (FA) in water) and loaded onto Evotips (Evosep Biosystem, Denmark) according to the manufacturer’s recommendations. The Evotips were wetted with isopropanol for 5 min, activated with 20 µl solvent B (99% CAN, 0.1% FA) and centrifuged at 700 x g for 1 min. 20 µl of solution A was then added to equilibrate the tips followed by sample loading. Finally, 20 µl buffer A was used to wash the Evotip and 100 µl was added to avoid drying.

Liquid chromatography and mass spectrometry (MS) analysis

The samples were injected in single shots into an Exploris 480 Thermo Fischer Scientific system using Evosep One (Evosep Biosystem). A preset chromatographic method was used corresponding to 60 samples per day. The peptides were separated on an 8 cm Pepsep column (150 μm, ID 1.5 μm bead size Reprosil-Pur C18 beads, Marslev, Denmark) at 1 µl/min flow rate with a 21 min gradient. The heated capillary temperature was set to 275 °C, the spray voltage to 2300 V and the funnel radiofrequency to 40 Hz. The mass spectrometer was operated in a data-independent mode (DIA) with a full MS range from 350 to 1650 m/z at a resolution of 60,000 at 200 m/z. The AGC target was set to 300% with an injection time of 50 ms. The AGC value of the targeted MS2 experiment was set to 1000%. Twenty-two windows of variable sizes were defined for target MS2 (tMS2) acquisition and subjected to high-energy collisional dissociation (HCD) fragmentation with a normalized collision energy at 30%. Each tMS2 scan was acquired at a resolution of 30,000 with a maximum ion injection time (IT) of 28 ms for a scan range of m/z 349.5 to 1650.5. Window sizes are provided in Supplementary file 3, Supplementary Table 7.

Data handling

The MS raw files were processed with Spectronaut version 15 (Biognosys, Switzerland). A CSF spectral library, generated in connection with an unpublished CSF study, was imported from MaxQuant software analyses. The library contained 2733 protein groups and 17,301 peptides. DIA files were searched against the library using default parameters except for the normalization, which was set to local, as it offered more stringent criteria for data analysis and thus helps avoid false hits. Dynamic mass and retention time tolerances (for both MS1 and MS2) were applied. Q-value cutoff was set to 1% both at precursor and protein level using a mutated decoy method [40]. The calibration was performed based on a local regression model [41]. The MS2-based quantitation was used for all further analysis. Protein data was exported from Spectronaut and further processed using the clinical knowledge graph (CKG) [42] together with their matching experimental and clinical data. Intensities were log-transformed before further statistical analysis.

Statistical analysis

Statistical analyses were carried out using R v. 4.1.0 (R Core Team, Vienna, Austria). Unpaired analyses were conducted for comparisons between healthy control subjects and SAH patients, while paired analyses were conducted for comparisons using CSF samples collected from the same SAH patients (start versus end). Continuous data were presented as mean and standard deviation or median and interquartile range depending on normality. Proteomics data were analyzed to assess which proteins were measured above the detection level. As cutoff, only proteins detectable in at least five CSF samples from healthy control subjects (corresponding to 50% of the control subjects) and 10 CSF samples from patients with SAH (corresponding to 43% of the SAH patients) were included in subsequent analyses. A principal component analysis (PCA) plot with complete case analysis was conducted to reveal systematic group differences [43]. A Volcano plot was used to investigate if protein abundance was higher in one group compared to another, similar between groups, or unclear. Statistical analysis was performed with Student’s t-test as indicated in figure and table legends, and P-values were corrected for multiple testing using a Bonferroni correction. Since data initially were skewed, but logarithmic transformed before analyses, the normality assumption was met. Proteins with adjusted P-values < 0.05 and a fold change in protein abundance > 2 were assigned as significantly different between groups to ensure that minor variation would not induce too many false positives. Conversely, proteins with similar abundance in groups were defined by uncorrected P-values > 0.05 and a fold change in protein abundance < 2. Lack of blood contamination of the CSF samples was verified by quantification of blood proteins (Supplementary file 1, Supplementary Fig. S1-S2). The sensitivity and specificity of CSF proteins for prediction of shunt dependency and/or functional outcome in SAH patients were assessed using receiver operating characteristic (ROC) curve analysis, and the corresponding area under the curve (AUC) values were computed as a measure of accuracy. CSF proteins with a lower confidence limit ≥ 0.7 using the AUC with a 95% confidence interval (CI) were considered potential predictors. The lower confidence limit of ≥ 0.7 was selected to ensure predictions of moderate to high accuracy [44].

Results

CSF from patients with SAH displays a distinct proteomic profile

To determine whether the proteomic profile of CSF from patients with SAH differed from that of healthy control subjects, ventricular CSF collected upon EVD placement (start samples) from patients with SAH was compared to cisternal CSF collected from healthy control subjects undergoing vascular clipping of an unruptured aneurysm. Principal component analysis (PCA) of the CSF protein abundance revealed no clear overall clustering that could distinguish the SAH patient group from the group of healthy control subjects (Fig. 1a). Of the 1,205 unique proteins identified in the CSF samples using MS-based proteomics, 416 proteins met the requirements for comparison (detected in at least five control samples and 10 SAH start samples). Proteomic analysis employing a volcano plot for visualization (Fig. 1.b) revealed that three proteins were significantly increased in abundance in the CSF from patients with SAH when compared to that of healthy control subjects: immunoglobulin heavy variable 1–2 (IGHV1-2), protein S100-B (S100B), and serum amyloid A-1 protein (SAA1), Table 1. Conversely, 15 proteins were significantly decreased in abundance in the SAH patient CSF: amyloid beta precursor like protein 1 (APLP1), amyloid-beta precursor protein (APP), chromogranin-A (CHGA), secretogranin-1 (CHGB), neural cell adhesion molecule L1-like protein (CHL1), calsyntenin-1 (CLSTN1), beta-Ala-His dipeptidase (CNDP1), dickkopf WNT signaling pathway inhibitor 3 (DKK3), neural cell adhesion molecule 1 (NCAM1), neuronal cell adhesion molecule (NRCAM), proSAAS (PCSK1N), prostaglandin-H2 D-isomerase (PTGDS), secretogranin-3 (SCG3), SPARC-like protein 1 (SPARCL1), and thy-1 cell surface antigen (THY1), Table 2. Of the remaining 398 proteins, 213 proteins were detected in similar abundance in the SAH patient CSF and control CSF, while 185 proteins were categorized as unclear (Supplementary file 2, Supplementary Table 1).

Fig. 1
figure 1

Proteomic profile of CSF from healthy control subjects and patients with SAH over time. a PCA plot of the proteomic profiles of patients with SAH (n = 23) and healthy control subjects (n = 10). b Volcano plot of proteomic data from patients and control subjects from panel a. Ctrl: controls; SAH: subarachnoid hemorrhage. c PCA plot of the CSF samples collected upon EVD placement (start samples, light red) and just before EVD removal (end samples, dark red) from patients with SAH (n = 23). d Volcano plot of proteomic data from patients with SAH (n = 23) employing both the start and end samples. Significance threshold was defined as a two-fold change in protein abundance combined with statistical significance with a t-test followed by Bonferroni correction (P < 0.05). Grey dots represent proteins, which were categorized as unclear, either not reaching a significant Bonferroni-corrected P-value or a two-fold change in protein abundance. Black dots represent proteins found in similar abundance in the two groups

Full size image
Table 1 Proteins significantly more abundant in CSF from patients with SAH versus healthy control subjects
Full size table
Table 2 Proteins significantly less abundant in CSF from patients with SAH versus healthy control subjects
Full size table

The proteomic profile in CSF from patients with SAH changes over time

To elucidate whether the progression of SAH from ictus to EVD removal was associated with alterations in CSF protein composition, the proteomic profiles of paired CSF samples collected upon EVD placement (start samples) and EVD removal (end samples) from patients with SAH were compared. PCA revealed a tendency towards clustering of the start and end samples into distinct proteomic distributions (Fig. 1c), indicating that the proteomic profile of SAH patient CSF may change over time after the initial hemorrhagic event. Of the 344 unique proteins detected in the CSF samples that met requirements for comparison (detected in at least 10 start samples and 10 end samples), 13 proteins were significantly increased in abundance in the start samples: serum amyloid P-component (APCS), apolipoprotein B-100 (APOB), apolipoprotein C-I (APOC1), complement component C6 (C6), carboxypeptidase N subunit 2 (CPN2), coagulation factor X (F10), insulin-like growth factor-binding protein complex acid labile subunit (IGFALS), IGHV1-2, immunoglobulin heavy variable 3–49 (IGHV3-49), inter-alpha-trypsin inhibitor heavy chain H2 (ITIH2), plasma kallikrein (KLKB1), serum paraoxonase/arylesterase 1 (PON1), and S100B, Fig. 1d; Table 3. Conversely, three proteins were significantly increased in abundance in the end samples: apolipoprotein E (APOE), monocyte differentiation antigen CD14 (CD14), and lysozyme (LYZ), Fig. 1d; Table 4. Of the 328 remaining proteins, 140 proteins were detected in similar abundance in the start and end samples, while 188 proteins were categorized as unclear (Supplementary file 2, Supplementary Table 2).

Table 3 Proteins significantly more abundant in the CSF samples collected acutely after ictus (start samples)
Full size table
Table 4 Proteins significantly more abundant in the CSF samples collect upon EVD removal (end samples)
Full size table

The proteomic profile is similar in weaned and shunted patients with SAH

To determine whether shunted SAH patients displayed a distinct proteomic CSF profile in comparison to patients who could be successfully weaned from their EVD, we compared the proteomic profiles of the two patient groups (weaned: n = 12; shunted: n = 11, see Methods) using both the CSF samples collected upon EVD placement (start samples) and EVD removal (end samples). Regardless of the time of CSF collection (start samples and end samples), PCA revealed no clear clustering of the weaned and shunted SAH patient groups, albeit a tendency towards clustering of the end sample was apparent (Fig. 2a-b). In agreement, none of the proteins identified using MS-based proteomics were detected at significantly different levels in either the start samples or the end samples (Fig. 2c-d). All proteins identified in the start and end samples are listed in Supplementary file 2, Supplementary Tables 3–4.

Fig. 2
figure 2

Weaned and shunted SAH patients display a similar proteomic profile. a PCA plot of the start samples (light red) collected from weaned (open circles, n = 12) and shunted (filled circles, n = 11) patients with SAH. b PCA plot of the end samples (dark red) collected from weaned (open circles, n = 12) and shunted (filled circles, n = 11) patients with SAH. c Volcano plot of proteomic data (n = 380 proteins) using the start samples collected from patients with SAH (n = 23). Significance threshold was defined as a two-fold change in protein abundance combined with statistical significance with a t-test followed by Bonferroni correction. Grey dots represent proteins, which were categorized as unclear (n = 35 proteins), either not reaching a significant Bonferroni-corrected P-value or a two-fold change in protein abundance. Black dots represent proteins found in similar abundance in the start samples from weaned and shunted SAH patients (n = 345 proteins). d Volcano plot of proteomic data (n = 577 proteins) using the end samples collected from patients with SAH (n = 23 patients). Significance threshold was defined as a two-fold change in protein abundance and a significant Bonferroni corrected P-value. Grey dots represent proteins which were categorized as unclear (n = 91 proteins), either not reaching a significant Bonferroni-corrected P value or a two-fold change in protein abundance. Black dots represent proteins found in similar abundance in the end samples from weaned and shunted SAH patients (n = 486 proteins)

Full size image

Predictors of shunt dependency and functional outcome

To determine whether certain proteins detected in the CSF from patients with SAH could serve as potential predictors of shunt dependency, the proteomic data were analyzed by a receiver operating curve (ROC). Proteins with a lower confidence limit of ≥ 0.7 using the area under the curve (AUC) with a 95% confidence interval were considered potential predictors of shunt dependency. A total of 16 unique proteins had a lower confidence limit of ≥ 0.7 and could thus be considered potential predictors of shunt dependency (Fig. 3; Table 5). For the start samples, three proteins were identified as possible predictors of shunt dependency: creatine kinase B-type (CKB), S100B, and tubulin alpha-1B chain (TUBA1B). For the end samples, 13 proteins were identified as possible predictors of shunt dependency: agrin (AGRN), lysosomal protective protein (CTSA), procathepsin L (CTSL), plasma alpha-L-fucosidase, (FUCA2), polypeptide N-acetylgalactosaminyltransferase 2 (GALNT2), hexosaminidase subunit beta (HEXB), hypoxanthine-guanine phosphoribosyltransferase (HPRT1), peptidyl-prolyl cis-trans isomerase B (PPIB), vitamin K-dependent protein Z (PROZ), protein S100-A8 (S100A8), transmembrane protein 132 A (TMEM132A), tripeptidyl-peptidase 1 (TPP1), and WAP, follistatin/kazal, immunoglobulin, kunitz and netrin domain containing 2 (WFIKKN2). Considering the average change in protein abundance per day, only one protein, CKB, was identified as a possible predictor of shunt dependency. The remaining proteins that did not reach a lower confidence limit of ≥ 0.7 are listed in Supplementary file 2, Supplementary Table 5.

Fig. 3
figure 3

Potential proteomic predictors of shunt dependency. The figure illustrates the calculated AUC and 95% CI for prediction of shunt dependency in patients with SAH (n = 23), employing the proteins detected in the start samples, the end samples, and the calculated average change in protein abundance per day. Proteins that reached a lower confidence limit of ≥ 0.7 were accepted as possible predictors (black lines). AUC: area under the curve; CI: confidence interval

Full size image
Table 5 Possible predictors of shunt dependency in SAH patients
Full size table

Predictors of functional outcome

To determine whether certain proteins detected in the CSF from patients with SAH could serve as potential predictors of functional outcome assessed six months after ictus (favorable: mRS 0–2, n = 13; unfavorable: mRS 3–6, n = 10, see Methods), we employed ROC analysis and AUC values (Fig. 4; Table 6). For the start samples, one protein, extracellular matrix protein 1 (ECM1), was identified as a possible predictor of functional outcome. For the end samples, two proteins, ADAM metallopeptidase domain 22 (ADAM22) and cell growth regulator with EF hand domain protein 1 (CGREF1), were identified as possible predictors of functional outcome. No proteins were identified as possible predictors of functional outcome when considering the average change in protein abundance per day. The proteins that did not reach a lower confidence limit of ≥ 0.7 are listed in Supplementary file 2, Supplementary Table 6.

Fig. 4
figure 4

Potential proteomic predictors of functional outcome. The figure illustrates the calculated AUC and 95% CI for prediction of functional outcome assessed after six months in patients with SAH (n = 23), employing the proteins detected in the start samples, the end samples, and the calculated average change in protein abundance per day. Proteins that reached a lower confidence limit of ≥ 0.7 were accepted as possible predictors (black lines). AUC: area under the curve; CI: confidence interval

Full size image
Table 6 Possible predictors of functional outcome in SAH patients
Full size table

Discussion

Proteomic profiles

SAH vs. controls

We here revealed that the proteomic profile of CSF from patients with SAH differs from that of healthy control subjects as evident by an altered abundance of select CSF proteins. Substantiating the notion that SAH promotes neuroinflammation [45], CSF collected acutely after ictus from patients with SAH contained an increased abundance of SAA1 and S100B, proteins which are both linked to inflammation. While SAA1 serves as an acute-phase protein produced in large quantities in response to injury, infection, and inflammation [46], S100B, a Ca2+ binding protein predominantly derived from astrocytes, has been widely recognized as a putative biomarker of brain injury [47] and, only more recently, as a possible neuroinflammatory mediator [48]. Elevated CSF S100B levels have previously been reported in patients with SAH [49,50,51,52] suggesting that the protein may serve as a possible biomarker of SAH in general. However, whether some of the neuroinflammatory actions of S100B, such as its ability to promote microglial activation [48], contribute to the subsequent disease progression following SAH awaits determination. Among the proteins detected in lower abundance in the SAH patient CSF, when compared to that of control CSF, was APLP1 and APP, two proteins belonging to the family of amyloid precursor proteins [53]. Corroborating the present findings, a recent study revealed significantly decreased CSF levels of APLP1 and a similar trend for APP in CSF from patients with non-traumatic acute brain injury (the majority of which had SAH) when compared to that of healthy controls undergoing elective clipping of an unruptured aneurysm [54]. However, another proteomics study revealed significantly elevated APLP1 levels in SAH patient CSF when compared to control CSF [55]. The study employed lumbar CSF for their proteomics analysis, the composition of which differs from that of ventricular CSF [56], as employed here. Hence, methodological differences may explain some of these discrepancies.

Time course

In the present study, we furthermore demonstrated that CSF collected acutely after ictus displays a distinct proteomic profile in comparison to CSF collected weeks later (on average 19 days, range: 5–30 days), thus substantiating that the CSF proteome in SAH patients is dynamic. Noticeably, the majority of proteins associated with an altered CSF abundance were elevated in the CSF samples collected acutely after ictus, while only a few proteins appeared elevated in the CSF samples collected weeks later. Among the proteins elevated acutely after ictus were different inflammatory mediators, components of the coagulation system, and different apolipoproteins. Proteins linked to the innate immune system were likewise elevated in the CSF samples collected just prior to EVD removal.

Shunt dependency

Regardless of the time of CSF collection, we found no evidence of distinct proteomic CSF profiles when comparing patients with SAH who went on to develop shunt-dependent PHH and patients who could be successfully weaned off their EVD. This observation may possibly be explained by the statistical requirement of correction for multiple comparisons that accompanied our unbiased quantification of the CSF protein content, which may have led to the neglection of potentially relevant proteins capable of predicting shunt dependency and/or functional outcome. We therefore sought to identify potential biomarkers through use of an ROC analysis.

Proteomic biomarkers

Shunt dependency

Evaluation of CSF biomarkers is gaining significance as potential clinical tools to guide treatment and predict the functional outcome of various neurological conditions. At present, it remains unresolved why some patients with SAH can be successfully weaned off their EVD, while other patients require permanent shunt insertion due to development of chronic PHH. Identification of predictive biomarkers for PHH is thus greatly needed. In the present study, we identified 16 unique proteins as possible predictors of shunt dependency, the majority of which were derived from the CSF samples collected weeks after ictus, just prior to EVD removal (end samples). Biomarkers of shunt dependency should ideally be detectable in the CSF as early as possible to minimize the risk of infection, which increases with the duration of the EVD [17,18,19,20]. However, the pathophysiological events culminating in development of PHH may manifest in the weeks following SAH and thus not be adequately reflected in the CSF samples collected acutely after ictus. Of all the proteins identified as possible biomarkers of shunt dependency, only one protein, S100B, was also found significantly more abundant in the SAH patient CSF when compared to the control CSF. Noticeably, employing the CSF samples collected acutely after ictus, S100B appeared less abundant in patients who required permanent CSF shunting compared to patients who could be successfully weaned from their EVD. The present observation thus suggests that lower CSF levels of S100B indicate a higher probability of permanent shunt requirement. However, this finding contradicts an earlier observation of higher CSF S100B levels in SAH patients requiring permanent shunt placement [29]. Hence, whether CSF S100B levels may assist clinical decision making with regards to permanent shunt placement after SAH requires further elucidation.

Functional outcome

Employing the mRS for evaluation of functional outcome six months after ictus in patients with SAH, we identified three possible predictors, ADAM22, CGREF1, and ECM1, the levels of which were all elevated in patients with a favorable outcome (mRS 0–2) and lower in patients with an unfavorable outcome (mRS 3–6). To the best of our knowledge, none of these proteins have previously been identified as possible predictors of functional outcome after SAH and additional studies are thus required to elucidate their biomarker potential. Potential biomarkers of functional outcome previously identified in patients with SAH comprise, among others, inflammatory markers [30] and various proteins, including S100B [57,58,59,60], the protein here identified as a possible biomarker of shunt dependency. Hence, although the present study did not identify S100B as a possible biomarker of functional outcome, the protein appears intimately linked to the SAH pathophysiology.

Strengths and limitations

To the best of our knowledge, previous proteomic studies of CSF from patients with SAH have generally relied on control CSF obtained from the lumbar compartment [55, 61], the protein composition of which differs from that of ventricular CSF [56]. The present study sought to limit this potentially confounding factor by comparing ventricular CSF collected from patients with SAH to cisternal CSF collected from healthy control subjects undergoing vascular clipping for an unruptured aneurysm. The different CSF sampling sites (ventricular versus cisternal) were dictated by ethical limitations but could potentially influence our results if the protein content and concentration of individual proteins differ between these CSF compartments. Although a recent study revealed no overall difference in cisternal and ventricular CSF protein content [62], further exploration is required to elucidate the extent to which the ventricular and cisternal CSF proteomes are comparable. Moreover, we cannot exclude that the surgical EVD insertion itself may have led to release of cellular debris or other proteins into the CSF, which could potentially have altered the protein content in the SAH patient samples and thus our present results, as others have previously demonstrated altered CSF protein levels upon EVD insertion [63]. Furthermore, we cannot exclude that alterations in the protein content may have occurred within the limited time frame from CSF collection to storage (maximum of 2 h), albeit the CSF samples were kept on ice to minimize such alterations. Although the total number of unique CSF proteins identified in the present study (1205 proteins) is similar to that identified in one of our recent studies (1251 proteins) [64], others have previously reported greater numbers (~ 2000–3000 CSF proteins) utilizing different separation methodologies and MS-based proteomics [65,66,67,68]. Hence, it is possible that some CSF proteins of potential relevance to the SAH pathophysiology were undetected. Lastly, the present study is limited by the relatively limited number of patient samples (n = 23) and control samples (n = 10), which may preclude detection of all potentially statistically relevant markers.

Conclusions

Biomarkers capable of distinguishing SAH patients with only a temporary need for CSF diversion from those in need of permanent CSF shunting due to chronic PHH development are in great demand. Here we identified several possible biomarkers of shunt dependency and functional outcome in patients with SAH using an unbiased MS-based proteomics approach. The applicability of these biomarkers as potential clinical tools should be validated in larger patient cohorts, preferentially with additional targeted CSF analysis techniques to validate the promising biomarkers, such as benchtop immunoassays (e.g. ELISAs).

Data availability

No datasets were generated or analysed during the current study.

Abbreviations

AUC:

Area under the curve

CSF:

Cerebrospinal fluid

EVD:

External ventricular drainage

ICP:

Intracranial pressure

mRS:

Modified Rankin Scale

MS:

Mass spectrometry

PCA:

Principal component analysis

PHH:

Posthemorrhagic hydrocephalus

ROC:

Receiver operating curve

SAH:

Subarachnoid hemorrhage

References

  1. Olsen MH, Orre M, Leisner ACW, Rasmussen R, Bache S, Welling K-L, et al. Delayed cerebral ischaemia in patients with aneurysmal subarachnoid haemorrhage: functional outcome and long-term mortality. Acta Anaesthesiol Scand. 2019;63:1191–9.

    Article  PubMed  Google Scholar 

  2. Western E, Sorteberg A, Brunborg C, Nordenmark TH. Prevalence and predictors of fatigue after aneurysmal subarachnoid hemorrhage. Acta Neurochir (Wien). 2020;162:3107–16.

    Article  PubMed  Google Scholar 

  3. Al-Khindi T, Macdonald RL, Schweizer TA. Cognitive and functional outcome after aneurysmal subarachnoid hemorrhage. Stroke. 2010;41:e519–36.

    Article  PubMed  Google Scholar 

  4. Claassen J, Park S. Spontaneous subarachnoid haemorrhage. Lancet. 2022;400:846–62.

    Article  PubMed  PubMed Central  Google Scholar 

  5. Cahill J, Cahill WJ, Calvert JW, Calvert JH, Zhang JH. Mechanisms of early brain injury after subarachnoid hemorrhage. J Cereb Blood Flow Metab. 2006;26:1341–53.

    Article  CAS  PubMed  Google Scholar 

  6. Grote E, Hassler W. The critical first minutes after subarachnoid hemorrhage. Neurosurgery. 1988;22:654–61.

    Article  CAS  PubMed  Google Scholar 

  7. Chaudhry SR, Hafez A, Rezai Jahromi B, Kinfe TM, Lamprecht A, Niemelä M, et al. Role of damage associated molecular pattern molecules (DAMPs) in aneurysmal subarachnoid Hemorrhage (aSAH). Int J Mol Sci. 2018;19:2035.

    Article  PubMed  PubMed Central  Google Scholar 

  8. Papaioannou V, Czosnyka Z, Czosnyka M. Hydrocephalus and the neuro-intensivist: CSF hydrodynamics at the bedside. Intensive Care Med Exp. 2022;10:20.

    Article  PubMed  PubMed Central  Google Scholar 

  9. Al-Mufti F, Amuluru K, Smith B, Damodara N, El-Ghanem M, Singh IP, et al. Emerging markers of early brain injury and delayed cerebral ischemia in aneurysmal subarachnoid hemorrhage. World Neurosurg. 2017;107:148–59.

    Article  PubMed  Google Scholar 

  10. Chen S, Luo J, Reis C, Manaenko A, Zhang J. Hydrocephalus after subarachnoid hemorrhage: pathophysiology, diagnosis, and treatment. Biomed Res Int. 2017;2017:1–8.

    Google Scholar 

  11. Lolansen SD, Rostgaard N, Barbuskaite D, Capion T, Olsen MH, Norager NH, et al. Posthemorrhagic hydrocephalus associates with elevated inflammation and CSF hypersecretion via activation of choroidal transporters. Fluids Barriers CNS. 2022;19:62.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  12. Karimy JK, Zhang J, Kurland DB, Theriault BC, Duran D, Stokum JA, et al. Inflammation-dependent cerebrospinal fluid hypersecretion by the choroid plexus epithelium in posthemorrhagic hydrocephalus. Nat Med. 2017;23:997–1003.

    Article  CAS  PubMed  Google Scholar 

  13. Robert SM, Reeves BC, Kiziltug E, Duy PQ, Karimy JK, Mansuri MS, et al. The choroid plexus links innate immunity to CSF dysregulation in hydrocephalus. Cell. 2023;186:764–e78521.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  14. Ben-Shoshan SD, Lolansen SD, Mathiesen TI, MacAulay N. CSF hypersecretion versus impaired CSF absorption in posthemorrhagic hydrocephalus: a systematic review. Acta Neurochir (Wien). 2023;165:3271–87.

  15. Capion T, Lilja-Cyron A, Juhler M, Mathiesen TI, Wetterslev J. Prompt closure versus gradual weaning of external ventricular drainage for hydrocephalus in adult patients with aneurysmal subarachnoid haemorrhage: a systematic review. BMJ Open. 2020;10:e040722.

    Article  PubMed  PubMed Central  Google Scholar 

  16. Capion T, Lilja-Cyron A, Bartek J, Forsse A, Logallo N, Juhler M, et al. Discontinuation of external ventricular drainage in patients with hydrocephalus following aneurysmal subarachnoid hemorrhage - a Scandinavian multi-institutional survey. Acta Neurochir (Wien). 2020;162:1363–70.

    Article  PubMed  Google Scholar 

  17. Zhu Y, Wen L, You W, Wang Y, Wang H, Li G, et al. Influence of Ward environments on external ventricular drain infections: a retrospective risk factor analysis. Surg Infect (Larchmt). 2021;22:211–6.

    Article  PubMed  Google Scholar 

  18. Champey J, Mourey C, Francony G, Pavese P, Gay E, Gergele L, et al. Strategies to reduce external ventricular drain–related infections: a multicenter retrospective study. J Neurosurg. 2019;130:2034–9.

    Article  Google Scholar 

  19. Walek KW, Leary OP, Sastry R, Asaad WF, Walsh JM, Horoho J, et al. Risk factors and outcomes associated with external ventricular drain infections. Infect Control Hosp Epidemiol. 2022;43:1859–66.

    Article  PubMed  PubMed Central  Google Scholar 

  20. Camacho EF, Boszczowski I, Basso M, Jeng BCP, Freire MP, Guimarães T, et al. Infection rate and risk factors associated with infections related to external ventricular drain. Infection. 2011;39:47–51.

    Article  CAS  PubMed  Google Scholar 

  21. Yang Y-C, Yin C-H, Chen K-T, Lin P-C, Lee C-C, Liao W-C, et al. Prognostic nomogram of predictors for shunt-dependent hydrocephalus in patients with aneurysmal subarachnoid hemorrhage receiving external ventricular drain insertion: a single-center experience and narrative review. World Neurosurg. 2021;150:e12–22.

    Article  PubMed  Google Scholar 

  22. Palasz J, D’Antona L, Farrell S, Elborady MA, Watkins LD, Toma AK. External ventricular drain management in subarachnoid haemorrhage: a systematic review and meta-analysis. Neurosurg Rev. 2022;45:365–73.

    Article  PubMed  Google Scholar 

  23. Rao SS, Chung DY, Wolcott Z, Sheriff F, Khawaja AM, Lee H, et al. Intermittent CSF drainage and rapid EVD weaning approach after subarachnoid hemorrhage: association with fewer VP shunts and shorter length of stay. J Neurosurg. 2019;132:1583–8.

    Article  PubMed  Google Scholar 

  24. Dorai Z, Hynan LS, Kopitnik TA, Samson D. Factors related to hydrocephalus after aneurysmal subarachnoid hemorrhage. Neurosurgery. 2003;52:763–9. 

    Article  PubMed  Google Scholar 

  25. Brisman JL, Berenstein A. Factors related to hydrocephalus after aneurysmal subarachnoid hemorrhage. Neurosurgery. 2004;54:1031.

    Article  PubMed  Google Scholar 

  26. Chan M, Alaraj A, Calderon M, Herrera SR, Gao W, Ruland S, et al. Prediction of ventriculoperitoneal shunt dependency in patients with aneurysmal subarachnoid hemorrhage. J Neurosurg. 2009;110:44–9.

    Article  PubMed  Google Scholar 

  27. Rincon F, Gordon E, Starke RM, Buitrago MM, Fernandez A, Schmidt JM, et al. Predictors of long-term shunt-dependent hydrocephalus after aneurysmal subarachnoid hemorrhage. Clinical article. J Neurosurg. 2010;113:774–80.

    Article  PubMed  Google Scholar 

  28. Lai L, Morgan MK. Predictors of in-hospital shunt-dependent hydrocephalus following rupture of cerebral aneurysms. J Clin Neurosci. 2013;20:1134–8.

    Article  PubMed  Google Scholar 

  29. Brandner S, Xu Y, Schmidt C, Emtmann I, Buchfelder M, Kleindienst A. Shunt-dependent hydrocephalus following subarachnoid hemorrhage correlates with increased S100B levels in cerebrospinal fluid and serum. Acta Neurochir Suppl. 2012;114:217–20.

    Article  CAS  PubMed  Google Scholar 

  30. Rostgaard N, Olsen MH, Capion T, MacAulay N, Juhler M. Inflammatory markers as predictors of shunt dependency and functional outcome in patients with aneurysmal subarachnoid hemorrhage. Biomedicines. 2023;11:997.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  31. Capion T, Lilja-Cyron A, Olsen MH, Juhler M, Møller K, Sorteberg A, et al. Prompt closure versus gradual weaning of external ventricular drainage for hydrocephalus following aneurysmal subarachnoid haemorrhage: protocol for the DRAIN randomised clinical trial. Acta Anaesthesiol Scand. 2023;67:1121–7.

    Article  PubMed  Google Scholar 

  32. del Campo M, Mollenhauer B, Bertolotto A, Engelborghs S, Hampel H, Simonsen AH, et al. Recommendations to standardize preanalytical confounding factors in Alzheimer’s and Parkinson’s disease cerebrospinal fluid biomarkers: an update. Biomark Med. 2012;6:419–30.

    Article  PubMed  Google Scholar 

  33. Broderick JP, Adeoye O, Elm J. Evolution of the Modified Rankin Scale and its use in future stroke trials. Stroke. 2017;48:2007–12.

    Article  PubMed  PubMed Central  Google Scholar 

  34. Toft-Bertelsen TL, Andreassen SN, Rostgaard N, Olsen MH, Norager NH, Capion T et al. Distinct cerebrospinal fluid lipid signature in patients with subarachnoid hemorrhage-induced hydrocephalus. Biomedicines 2023;11:2360.

  35. Lolansen SD, Rostgaard N, Capion T, Norager NH, Olsen MH, Juhler M et al. Posthemorrhagic hydrocephalus in patients with subarachnoid hemorrhage occurs independently of CSF osmolality. Int J Mol Sci. 2023;24.

  36. Toft-Bertelsen TL, Barbuskaite D, Heerfordt EK, Lolansen SD, Andreassen SN, Rostgaard N et al. Lysophosphatidic acid as a CSF lipid in posthemorrhagic hydrocephalus that drives CSF accumulation via TRPV4-induced hyperactivation of NKCC1. Fluids Barriers CNS. 2022;19.

  37. Bader JM, Geyer PE, Müller JB, Strauss MT, Koch M, Leypoldt F, et al. Proteome profiling in cerebrospinal fluid reveals novel biomarkers of Alzheimer’s disease. Mol Syst Biol. 2020;16:e9356.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  38. Geyer PE, Wewer Albrechtsen NJ, Tyanova S, Grassl N, Iepsen EW, Lundgren J, et al. Proteomics reveals the effects of sustained weight loss on the human plasma proteome. Mol Syst Biol. 2016;12:901.

    Article  PubMed  PubMed Central  Google Scholar 

  39. Kulak NA, Pichler G, Paron I, Nagaraj N, Mann M. Minimal, encapsulated proteomic-sample processing applied to copy-number estimation in eukaryotic cells. Nat Methods. 2014;11:319–24.

    Article  CAS  PubMed  Google Scholar 

  40. Bruderer R, Bernhardt OM, Gandhi T, Xuan Y, Sondermann J, Schmidt M, et al. Optimization of experimental parameters in data-independent mass spectrometry significantly increases depth and reproducibility of results. Mol Cell Proteom. 2017;16:2296–309.

    Article  CAS  Google Scholar 

  41. Callister SJ, Barry RC, Adkins JN, Johnson ET, Qian W-J, Webb-Robertson B-JM, et al. Normalization approaches for removing systematic biases associated with mass spectrometry and label-free proteomics. J Proteome Res. 2006;5:277–86.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  42. Santos A, Colaço AR, Nielsen AB, Niu L, Strauss M, Geyer PE, et al. A knowledge graph to interpret clinical proteomics data. Nat Biotechnol. 2022;40:692–702.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  43. Josse J, Husson F, missMDA:. A Package for handling missing values in multivariate data analysis. J Stat Softw. 2016;70.

  44. Swets JA. Measuring the accuracy of diagnostic systems. Science. 1988;240:1285–93.

    Article  CAS  PubMed  Google Scholar 

  45. Lucke-Wold BP, Logsdon AF, Manoranjan B, Turner RC, McConnell E, Vates GE, et al. Aneurysmal subarachnoid hemorrhage and neuroinflammation: a comprehensive review. Int J Mol Sci. 2016;17:497.

    Article  PubMed  PubMed Central  Google Scholar 

  46. Ye RD, Sun L. Emerging functions of serum amyloid A in inflammation. J Leukoc Biol. 2015;98:923–9.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  47. Thelin EP, Nelson DW, Bellander B-M. A review of the clinical utility of serum S100B protein levels in the assessment of traumatic brain injury. Acta Neurochir (Wien). 2017;159:209–25.

    Article  PubMed  Google Scholar 

  48. Michetti F, Clementi ME, Di Liddo R, Valeriani F, Ria F, Rende M et al. The S100B protein: a multifaceted pathogenic factor more than a Biomarker. Int J Mol Sci. 2023;24.

  49. Kleindienst A, Meissner S, Eyupoglu IY, Parsch H, Schmidt C, Buchfelder M. Dynamics of S100B release into serum and cerebrospinal fluid following acute brain injury. Acta Neurochir Suppl. 2010;106:247–50.

    Article  CAS  PubMed  Google Scholar 

  50. Kleindienst A, Schmidt C, Parsch H, Emtmann I, Xu Y, Buchfelder M. The passage of S100B from brain to blood is not specifically related to the blood-brain Barrier Integrity. Cardiovasc Psychiatry Neurol. 2010;2010:801295.

    Article  PubMed  PubMed Central  Google Scholar 

  51. Petzold A, Keir G, Lim D, Smith M, Thompson EJ. Cerebrospinal fluid (CSF) and serum S100B: release and wash-out pattern. Brain Res Bull. 2003;61:281–5.

    Article  CAS  PubMed  Google Scholar 

  52. Kay A, Petzold A, Kerr M, Keir G, Thompson E, Nicoll J. Decreased cerebrospinal fluid apolipoprotein E after subarachnoid hemorrhage: correlation with injury severity and clinical outcome. Stroke. 2003;34:637–42.

    Article  CAS  PubMed  Google Scholar 

  53. Jacobsen KT, Iverfeldt K. Amyloid precursor protein and its homologues: a family of proteolysis-dependent receptors. Cell Mol Life Sci. 2009;66:2299–318.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  54. Santacruz CA, Vincent J-L, Imbault V, Bruneau M, Creteur J, Brimioulle S, et al. Cerebral apolipoprotein E and amyloid precursor-like protein 1 as risk factors for chronic neurodegeneration after non-traumatic acute brain injury (ABI). Crit Care. 2023;27:249.

    Article  PubMed  PubMed Central  Google Scholar 

  55. Sokół B, Urbaniak B, Zaremba B, Wąsik N, Kokot ZJ, Jankowski R. CSF proteomics of patients with hydrocephalus and subarachnoid haemorrhage. Transl Neurosci. 2019;10:244–53.

    Article  PubMed  PubMed Central  Google Scholar 

  56. Rostgaard N, Olsen MH, Ottenheijm M, Drici L, Simonsen AH, Plomgaard P, et al. Differential proteomic profile of lumbar and ventricular cerebrospinal fluid. Fluids Barriers CNS. 2023;20:6.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  57. Abboud T, Rohde V, Mielke D. Mini review: current status and perspective of S100B protein as a biomarker in daily clinical practice for diagnosis and prognosticating of clinical outcome in patients with neurological diseases with focus on acute brain injury. BMC Neurosci. 2023;24:38.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  58. Lai PMR, Du R. Association between S100B levels and long-term outcome after aneurysmal subarachnoid hemorrhage: systematic review and pooled analysis. PLoS ONE. 2016;11:e0151853.

    Article  PubMed  PubMed Central  Google Scholar 

  59. Kellermann I, Kleindienst A, Hore N, Buchfelder M, Brandner S. Early CSF and serum S100B concentrations for Outcome Prediction in Traumatic Brain Injury and Subarachnoid Hemorrhage. Clin Neurol Neurosurg. 2016;145:79–83.

    Article  PubMed  Google Scholar 

  60. Aineskog H, Johansson C, Nilsson R, Koskinen L-OD, Lindvall P. Serum S100B correlates with health-related quality of life and functional outcome in patients at 1 year after aneurysmal subarachnoid haemorrhage. Acta Neurochir (Wien). 2022;164:2209–18.

    Article  PubMed  Google Scholar 

  61. Lad SP, Hegen H, Gupta G, Deisenhammer F, Steinberg GK. Proteomic biomarker discovery in cerebrospinal fluid for cerebral vasospasm following subarachnoid hemorrhage. J Stroke Cerebrovasc Dis. 2012;21:30–41.

    Article  PubMed  Google Scholar 

  62. Hopkins K, Price B, Ziogas J, Adamides A, Mangum J. Comparative proteomic analysis of ventricular and cisternal cerebrospinal fluid in haemorrhagic stroke patients. J Clin Neurosci. 2023;107:84–90.

    Article  CAS  PubMed  Google Scholar 

  63. Brandner S, Thaler C, Buchfelder M, Kleindienst A. Brain-derived protein concentrations in the cerebrospinal Fluid: contribution of trauma resulting from ventricular drain insertion. J Neurotrauma. 2013;30:1205–10.

    Article  PubMed  Google Scholar 

  64. Rostgaard N, Olsen MH, Lolansen SD, Nørager NH, Plomgaard P, MacAulay N, et al. Ventricular CSF proteomic profiles and predictors of surgical treatment outcome in chronic hydrocephalus. Acta Neurochir (Wien). 2023;165:4059–70.

    Article  PubMed  Google Scholar 

  65. Schutzer SE, Liu T, Natelson BH, Angel TE, Schepmoes AA, Purvine SO, et al. Establishing the proteome of normal human cerebrospinal fluid. PLoS ONE. 2010;5:e10980.

    Article  PubMed  PubMed Central  Google Scholar 

  66. Guldbrandsen A, Vethe H, Farag Y, Oveland E, Garberg H, Berle M, et al. In-depth characterization of the cerebrospinal fluid (CSF) proteome displayed through the CSF proteome resource (CSF-PR). Mol Cell Proteom. 2014;13:3152–63.

    Article  CAS  Google Scholar 

  67. Begcevic I, Brinc D, Drabovich AP, Batruch I, Diamandis EP. Identification of brain-enriched proteins in the cerebrospinal fluid proteome by LC-MS/MS profiling and mining of the human protein atlas. Clin Proteom. 2016;13:1–13.

    Article  Google Scholar 

  68. Zhang Y, Guo Z, Zou L, Yang Y, Zhang L, Ji N, et al. A comprehensive map and functional annotation of the normal human cerebrospinal fluid proteome. J Proteom. 2015;119:90–9.

    Article  CAS  Google Scholar 

Download references

Acknowledgements

We are grateful for the contribution from all the clinical staff at the Department of Neurosurgery and the Neurointensive Unit, Copenhagen University Hospital – Rigshospitalet, Copenhagen, Denmark, who participated in the collection of the CSF samples, and the support from Christina Christoffersen and Peter Plomgaard, Department of Clinical Biochemistry, Copenhagen University Hospital – Rigshospitalet, Copenhagen, Denmark, for scientific input to the proteomics strategy.

Funding

This research was funded by the Novo Nordisk Foundation (Tandem grant NNF17OC0024718 to NM and MJ).

Open access funding provided by Copenhagen University

Author information

Author notes

  1. Sara Diana Lolansen, Nina Rostgaard and Markus Harboe Olsen contributed equally to this work.

Authors and Affiliations

  1. Department of Neurosurgery, the Neuroscience Centre, Copenhagen University Hospital – Rigshospitalet, Copenhagen, Denmark

    Sara Diana Lolansen, Nina Rostgaard, Tenna Capion, Nicolas Hernandez Nørager & Marianne Juhler

  2. Department of Neuroscience, University of Copenhagen, Copenhagen, Denmark

    Sara Diana Lolansen & Nanna MacAulay

  3. Department of Neuroanaesthesiology, the Neuroscience Centre, Copenhagen University Hospital – Rigshospitalet, Copenhagen, Denmark

    Markus Harboe Olsen

  4. Department of Anaesthesiology, Zealand University Hospital, Køge, Denmark

    Markus Harboe Olsen

  5. NNF Center for Protein Research, University of Copenhagen, Copenhagen, Denmark

    Maud Eline Ottenheijm & Lylia Drici

  6. Department of Clinical Biochemistry, Copenhagen University Hospital – Bispebjerg and Frederiksberg Hospital, Copenhagen, Denmark

    Maud Eline Ottenheijm & Lylia Drici

  7. Department of Clinical Medicine, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark

    Marianne Juhler

Authors
  1. Sara Diana Lolansen
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Nina Rostgaard
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Markus Harboe Olsen
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Maud Eline Ottenheijm
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Lylia Drici
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Tenna Capion
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Nicolas Hernandez Nørager
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Nanna MacAulay
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Marianne Juhler
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

Conceptualization: M.J., N.M., N.R., M.H.O.; Patient samples: M.H.O., N.R., N.H.N., T.C.; Data curation: M.E.O., L.D.; Formal analysis: M.H.O., S.D.L.; Funding acquisition: N.M., M.J.; Methodology: M.E.O., L.D.; Writing – original draft: N.R., S.D.L., M.J.; Writing – review and editing: N.R., S.D.L., N.M., M.H.O., M.E.O. All authors have read and agreed to the published version of the manuscript.

Corresponding authors

Correspondence to Nanna MacAulay or Marianne Juhler.

Ethics declarations

Ethics approval and consent to participate

Written informed consent was obtained from all patients or their next of kin, depending on the capacity of the patients. The study was approved by the Danish National Committee on Health Research Ethics (H-19001474 and H-17011472/69197) and the Danish Data Protection Agency (VD-2019-210).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Lolansen, S.D., Rostgaard, N., Olsen, M.H. et al. Proteomic profile and predictive markers of outcome in patients with subarachnoid hemorrhage. Clin Proteom 21, 51 (2024). https://doi.org/10.1186/s12014-024-09493-6

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12014-024-09493-6

Keywords

Clinical Proteomics

ISSN: 1559-0275