Table 1.
Trials in which cerebrospinal fluid (CSF) protein biomarkers were associated with neurological outcome
| Author (ref) | Study population | CSF biomarker | Biological function of biomarker | Number of patients (ABI/control) | Source of protein (ABI/control) | Time point of first sampling* | Method of biomarker detection | Outcome measure | Relationship of biomarker with outcome | Newcastle–Ottawa risk of bias |
|---|---|---|---|---|---|---|---|---|---|---|
| Jiang et al. 2020 [10] | TBI | Caspase-3, cytochrome C, sFas and caspase-9 | Apoptosis | 45/25 | vCSF/lCSF | Day 1 after injury | ELISA | 6-month GOS | ICP and caspase-3 were significant predictors of outcome at 6 months | ★★★★★ |
| Mertens et al. 2018 [11] | Ischemic | Procarboxypeptidase U (proCPU, TAFI, proCPB2) | Inflammation, coagulation | AIS (n = 58) or TIA (n = 14)/32 | lCSF/lCSF | Day 1 after symptoms onset | ELISA | 3-month mRS | Increased proCPU levels were associated with stroke progression and worst mRS | ★★★★ |
| Kerr et al. 2018 [12] | TBI | Caspase-1, apoptosis-associated speck-like protein containing a caspase recruitment domain (ASC) | Apoptosis | 21#/30 | NS/Biobank | Day 1 after injury | ELISA | GOSE | Higher protein levels of ASC were consistent with poorer outcomes after TBI | ★★★★★ |
| Wąsik et al. 2017 [13] | SAH | Clusterin | Apoptosis | 27/25 | vCSF/lCSF | Day 1 after bleeding | ELISA | 3- month GOS | Higher levels of CSF clusterin were found 5–7 days after SAH in patients with good outcome | ★★★★ |
| Kellermann et al. 2016 [14] | Mixed | S-100β | Cytoskeleton | 45 SAH—57 TBI/no control | vCSF/no control | Day 1 after EVD placement | ELISA | GOS | In TBI and SAH patients, S-100β concentrations in CSF and serum were significantly higher in patients with unfavorable outcome (GOS 1–3) | ★★ |
| Failla et al. 2016 [15] | TBI | BDNF | Cytoskeleton | 203/10 | vCSF and serum/lCSF and serum | NR | ELISA | 1-year mortality | Higher CSF levels predicted mortality | ★★★★★★ |
| Wu et al. 2016 [16] | SAH | NLRP1, ASC and caspase 1 | Apoptosis | 24/10 | vCSF-lCSF/lCSF | Between 24 and 72 h after injury | SDS-PAGE | 3-month GOS |
Higher levels of inflammasome proteins were associated with severe SAH and poor outcome at 3 months |
★★★ |
| Papa et al. 2015 [17] | TBI | UCH-L1, MAP-2, SBDP150, SBDP145, SBDP120, MBP and S-100β | Apoptosis, cytoskeleton | 131/21 | vCSF/mixb | 6 h after injury | ELISA | 6-month mortality | MAP-2 in combination with clinical data provide enhanced prognostic capabilities for mortality at 6 months | ★★★ |
| Manevich et al. 2014 [18] | TBI | Peroxiredoxin (Prdx) VI | Redox | 21/10 | vCSF/lCSF | During EVD placement after injury | Western Blot | Scale of neurological deficits at discharge | Reduction of Prdx appeared to correlate with milder neurological deficits | ★★★ |
| Liu et al. 2014 [19] | TBI | Matrix metalloproteins (MMP-9) | Inflammation | 6/85 | vCSF/vCSF | During EVD placement | ELISA | ICP and GCS | MMP-9 was negatively correlated with the Glasgow Coma Scale | ★★★ |
| Gatson et al. 2013 [20] | TBI | NSE and Ab42 | Energy, neurodegeneration | 18/no control | vCSF/no control | Within 72 h after injury | ELISA | GOS-E and DRS | CSF oligomer levels correlated with GOS-E scores | ★★ |
| Mondello et al. 2013 [21] | TBI | Alpha-synuclein | Neurodegeneration | 12/22 | vCSF/lCSF | NR | ELISA |
6-month GOS-E Mortality |
Rising levels predicted mortality with 100% specificity and high sensitivity (83%) |
★★★ |
| Goyal et al. 2013 [22] | TBI | S100β | Cytoskeleton | 138/15 | vCSF /lCSF | First 6 days post-injury | ELISA |
GOS DRS Mortality |
Mean and peak levels were associated with mortality and GOS scores, but not with DRS | ★★★★★ |
| Zanier et al. 2013 [23] | SAH | H-FABP and tau protein | Cytoskeleton | 38/16 | vCSF/lCSF | Day 1 after injury | ELISA | GOS | Higher H-FABP and tau levels in patients with unfavorable outcome (death, vegetative state or severe disability) | ★★★★★ |
| Adamczak et al. 2012 [24] | TBI | ASC, caspase-1 and NALP-1 | Apoptosis | 23/9 | vCSF/vCSF | Within 12 h of injury and up to 72 h after injury | Western Blot | 5-month GOS |
Expression of each protein correlated significantly with the GOS at 5 months post-injury |
★★★ |
| Böhmer et al. 2011 [25] | TBI |
NSE, S-100β and glial fibrillary acidic protein |
Cytoskeleton | 20/20 | vCSF/lCSF | Between 2 and 4 h after hospitalization | ELISA | Survival | At admission, CSF NSE level predicted brain death more accurately than S-100β | ★★★★★★ |
| Stein et al. 2011 [26] | TBI | S100β, NSE | Cytoskeleton | 23/no control | vCSF/no control | Upon insertion of the EVD or as soon as possible after consent was obtained | ELISA |
ICH CH |
S-100β and NSE levels were associated with ICH and CH |
★★ |
| Darwish et al. 2010 [27] | TBI | Cytochrome c and activated caspase-9 | Apoptosis | 9/5 | vCSF/lCSF | 2 to 6 h after injury | ELISA | GOS | Activated caspase-9 showed weak correlation with poor neurologic outcome | ★★★ |
| Mondello et al. 2010 [28] | TBI | SBDP145- SBDP120 | Apoptosis | 40/24 | vCSF/vCSF | First 24 h after injury | ELISA | 3-month survival |
CSF SBDP levels predicted injury severity and mortality after severe TBI |
★★★ |
| Papa et al. 2010 [29] | TBI | UCH-L1 | Neurodegeneration | 41/25 | vCSF/vCSF | 6 h after injury | ELISA | GOS, 6-week mortality | Higher levels in patients with lower GCS score at 24 h, in those with post-injury complications, in those with 6-wk mortality and in those with a poor 6-month dichotomized GOS | ★★★★★ |
| Brouns et al. 2010 [30] | Ischemic | MBP, GFAP, S100β, NSE | Cytoskeleton, energy | 89/35 | lCSF/ lCSF | NR | ELISA |
3-month mRS Infarct volume |
MBP was a marker for infarct location. GFAP and S-100β correlated with stroke severity and outcome | ★★★ |
| Fountas et al. 2009 [31] | SAH | CRP | Inflammation | 41/no control | vCSF | Admission | Nephelometry | GOS, mRS |
Increased CRP in CSF associated with increased risk of vasospasm and bad outcome |
★★★ |
| Pineda et al. 2007 [32] | TBI | SBDP | Apoptosis | 41/11 | vCSF/vCSF | 6 h after injury | SDS-PAGE | 6-month GOS, severity of injury, computed tomography (CT) scan findings | SBDP correlated with severity of injury, computed tomography (CT) scan findings and outcome at 6 months post-injury | ★★★ |
| Lewis et al. 2007 [33] | SAH | α-2 spectrin and SBDP | Apoptosis | 20/10 | vCSF /lCSF | NR | SDS-PAGE | 6-month GOS, vasospasm | SBDP levels were significantly increased in patients with vasospasm | ★★★★ |
| Ost et al. 2006 [34] | TBI | c-tau | Cytoskeleton | 39/20 | vCSF /lCSF | First 24 h after injury | ELISA | GOSE | vCSF total tau on days 2 to 3 post-trauma correlated to morbidity and mortality at 1 year | ★★★ |
| Selakovic et al. 2005 [35] | Ischemic | NSE | Energy | 55/16 | lCSF/ lCSF | 1–2 days [21 patients], 3–4 days [14 patients], and 5–7 days [20 patients] from the onset of symptoms | ELISA |
Infarct volume, Canadian neurological scale and Barthel index |
Significant correlation between NSE concentration and infarct volume and degree of neurological and functional deficit | ★★★ |
| Kay et al. 2003 [7] | SAH | Apo-E and S-100β | Inflammation, cytoskeleton | 19/28 | vCSF/lCSF | Within 72 h after injury | ELISA | 3-month GOS | SAH patients with more severe injury and less favorable outcome had lower CSF apo-E concentration | ★★★ |
| Zemlan et al. 2002 [36] | TBI | C-tau | Cytoskeleton | 28/154 | vCSF/ lCSF | NR |
ELISA Immunoblotting |
GOS | C-tau levels-independent predictor of clinical outcome | ★★★★ |
| Aurell et al. 1991 [37] | Ischemic |
S-100β and glial fibrillary acidic protein |
Cytoskeleton | 28/18 | lCSF/lCSF | 12–48 h after onset of symptoms |
ELISA (S-100β) Radioimmunoassay (GFAP) |
Clinical state: Simplified activities of daily living test Size of infarct: computed tomography |
Increment was significantly correlated with size of infarction and clinical state of patients | ★★★★★ |
| Strand et al. 1984 [38] | Ischemic | MBP, tau-fraction, albumin, IgG and transferrin | Cytoskeleton, inflammation | 40/37 | lCSF/lCSF | 24 h after symptoms onset | Radioimmunoassay (MBP); crossed immunoelectrophoretic method (tau-fraction); electroimmunoassay (albumin, IgG and transferrin) |
Disability groups, mortality |
MBP increased with extent of brain injury; high values indicated poor short-term prognosis for the patient. No clear patterns for other markers | ★★★★★ |
ABI acute brain injury, AIS acute ischemic stroke, Apo-E apolipoprotein E, ASC apoptosis-associated speck-like protein containing a caspase recruitment domain, BDNF brain-derived neurotrophic factor, CRP C-reactive protein, lCSF lumbar CSF, C-tau cleaved tau protein, DRS Disability Rating Scale, ELISA enzyme-linked immunosorbent assay, EVD external ventricular drainage, GCS Glasgow Coma Score, GOS Glasgow Outcome Scale, GOS-E extended Glasgow Outcome Scale, H-FABP heart-type fatty acid binding protein, ICP intracranial pressure, MAP-2 microtubule-associated protein, mRS modified Rankin Scale, MBP myelin basic protein, MMP matrix metalloproteinase, NALP1 nacht leucine-rich-repeat protein-1, NR not reported, NSE neuron-specific enolase, UCH-L1 ubiquitin C-terminal hydrolase, SDS-PAGE sodium dodecyl sulfate polyacrylamide gel electrophoresis, S-100β S-100 beta, SBDP spectrin breakdown products, TAFI thrombin-activatable fibrinolysis inhibitor, TBI traumatic brain injury, TIA transient ischemic attack, vCSF ventricular CSF
#18 CSF samples; *as reported by the author