TABLE 1.
Neurophysiological and neuroimaging biomarkers for stroke prognosis.
| Type of biomarker | Biomarker | Measurement time points (biomarker) | Analytical method | Type of stroke | Functional outcomes | Measurement time points (functional outcome) | Main findings | References |
|---|---|---|---|---|---|---|---|---|
| Neurophysiological biomarkers (EEG/MEG) | BSI-delta | Baseline (<3 weeks), Week 5, Week 12, Week 26 | Frequency domain analysis | Ischemic stroke | NIHSS, FMA-UE | Baseline (<3 weeks), Week 5, Week 12, Week 26 | (1) Directional BSI-delta was negatively correlated with FMA-UE, which was borderline significant after correction for time. | Saes et al. (2020) |
| (2) Directional BSI-delta was positively correlated with NIHSS, and the association became stronger after correction for time. | ||||||||
| BSI-theta | Baseline (<3 weeks), Week 26 | Frequency domain analysis | Ischemic stroke | FMA-UE | Baseline (<3 weeks), Week 26 | BSI-theta was the strongest predictor of FMA-UE at week 26, with higher BSI-theta value reflecting more impairment 6 months after stroke. | Saes et al. (2021) | |
| delta/alpha ratio | Baseline (<3 weeks), Week 5, Week 12, Week 26 | Frequency domain analysis | Ischemic stroke | NIHSS, FMA-UE | Baseline (<3 weeks), Week 5, Week 12, Week 26 | Delta/alpha ratio over the affected hemisphere showed a trend toward a negative association with FMA-UE scores. | Saes et al. (2020) | |
| SSEP | Baseline (<1 week) | Time-domain analysis | Ischemic and hemorrhagic, supratentorial stroke | mRS | Baseline, 6 months | Pathological alterations of the SSEP (absence of cortical N20 response, abnormal bilateral N20-P25 amplitude ratio, Cant grading 3, Judson grading 3, or Haupt grading 3 to 4) were predictive of unfavorable outcomes 6 months after stroke. | Su et al. (2010) | |
| Movement-related ERD | Baseline (7-14 days after stroke onset), 1 month, 2 months, 4 months | Time frequency domain analysis | Ischemic stroke | NIHSS, FMA-UE | Baseline (7–14 days after stroke onset), 1 month, 2 months, 4 months | (1) Recovered patients showed a significant increase of ipsilesional ERD and/or a decreasing trend of contralesional ERD. | Tangwiriyasakul et al. (2014) | |
| (2) The only non-recovery patient showed an increase in ERD over the contralesional hemisphere and ERD over the ipsilesional hemisphere remained absent. | ||||||||
| Movement-related beta ERD/Postmovement beta ERS | Baseline (2–4 weeks), Week 5, Week 12 | Time frequency domain analysis | Ischemic stroke | FMA-UE, ARAT | Baseline (2–4 weeks), Week 5, Week 12 | (1) Ipsilesional M1 beta ERD was significantly positively correlated with concurrent FMA-UE scores at all three time points. | Tang et al. (2020) | |
| (2) Ipsilesional beta ERS at baseline was positively correlated with FMA-UE scores at Week 12. | ||||||||
| FC-delta (M1-M1) | Admission (5–71 days), Discharge (10–116), Day 90 | Coherence analysis | Ischemic and hemorrhagic stroke | FMA-UE, FIM motor scores | Admission (5–71 days), Discharge (10–116), Day 90 | A decrease in M1-M1 coherence at delta band occurring correlated with FIM-motor scores and FMA-UE improvement during the inpatients’ stay. | Cassidy et al. (2020) | |
| Neurophysiological biomarkers (TMS) | MEP | Baseline (<1 week), 6 months | Time domain analysis | Ischemic stroke | MRC, BI | Baseline (<1 week), 6 months | The patients with present MEP evoked by the ipsilesional M1 stimulation showed better motor and functional recovery than those without. | Escudero et al. (1998) |
| CCT | Baseline (<1 week), 6 months | Time domain analysis | Ischemic stroke | MRC, BI | Baseline (<1 week), 6 months | The degree of recovery was better in patients with normal CCT than in those with delayed CCT. | Escudero et al. (1998) | |
| SICI | Day 5–8, Day 30, Day 90 | Time domain analysis | Ischemic and hemorrhagic stroke | NIHSS, Lindmark scale, BI | Day 5–8, Day 30, Day 90 | At 3 months post stroke, patients with lower BI scores at baseline (BI < 60) still showed a reduction of SICI in the unaffected M1 compared to the healthy controls, whereas the SICI in the unaffected M1 was increased to normal values in patients with higher BI scores (BI ≥ 60) at baseline. | Manganotti et al. (2008) | |
| IHI (during the premovement phase) | Week 1, Week 4, Week 12, Week 24, Week 52 | Time domain analysis | Ischemic stroke | FMA-UE, Strength of finger flexion, Finger individuation index | Week 1, Week 4, Week 12, Week 24, Week 52 | The premovement IHI was normal in stroke patients during the acute and subacute period, and became abnormal (i.e., the contralesional M1 remained inhibitory to the ipsilesional M1 prior to the movement onset) in patients at the chronic stage. The longitudinal emergence of abnormal premovement IHI was negatively correlated with the improvement of hand functions in the finger individuation task. | Xu et al. (2019) | |
| Neurophysiological biomarkers (TMS-EEG) | TMS-evoked potential | Baseline (<14 days) | Time domain analysis (LMPF, the number of deflections), time-frequency domain analysis (the natural frequency to TMS) | Ischemic stroke | Composite recovery score (MI, ARAT, and grip force), NIHSS | Baseline (≤14 days), 3–6 months | (1) A positive correlation between the numbers of deflections of the TMS-EEG response and better motor recovery. | Tscherpel et al. (2020) |
| (2) The general neurological recovery (changes of the NIHSS scores) was significantly linked to all TMS-EEG outcomes. | ||||||||
| Neuroimaging biomarkers (MRI) | Lesion (PLIC, CS, and CR) | Baseline (<12 h), Day 3, Day 30 | DTT | Ischemic stroke | NIHSS motor scores | Baseline (<12 h), Day 3, Day 30, Day 90 | (1) Damage to the PLIC in the first 12 h was the best predictor of severity of motor deficits at day 90. | Puig et al. (2011) |
| (2) Damage to the CS and CR at day 3 was associated with poor motor outcome at day 90. | ||||||||
| Lesion (CST) | Baseline, 3 months | DWI | Ischemic stroke | FMA-UE, NIHSS motor scores | Baseline, 3 months | A weighted CST lesion load measured at the acute phase, in addition to the initial motor impairment, predicted upper limb motor outcomes (assessed by FMA-UE) at 3 months after stroke. | Feng et al. (2015) | |
| Doughty et al. (2016) | ||||||||
| CST degeneration | Baseline (≤6 h), 12 ± 6 h, 24 ± 6 h, 7 ± 2 days, 30 ± 10 days | ADC sequence | Ischemic stroke | NIHSS motor score | Baseline (≤6 h), 12 ± 6 h, 24 ± 6 h, 7 ± 2 days, 30 ± 10 days | CST ADC decreases after acute stroke in patients with poor motor outcome | DeVetten et al. (2010) | |
| FA asymmetry at the pons | Baseline (≤12 h), Day 3, Day 30 | DTT | Ischemic stroke | NIHSS motor subscore, mRS BI MI | Baseline (≤12 h), Day 3, Day 30, Day 90, Year 2 | The FA at day 30 is a predictor of motor outcome 2 years after stroke. | Puig et al. (2013) | |
| Intra- and interhemispheric motor connections | Baseline, 1 month, 6 months | DSI | Ischemic stroke | NIHSS FIM mRS | Baseline, 1 month, 6 months | The contralesional structural motor connections at baseline, combined with patients’ age and initial NIHSS scores, predicted functional improvement at 6 months. | Granziera et al. (2012) | |
| Neuroimaging biomarkers (fMRI) | Motor-related brain activation | Baseline (6 ± 3 days), 12 ± 4 days, 4–6 months | fMRI BOLD signals | Ischemic stroke | A composite score using ARAT and grip strength | Baseline (6 ± 3 days), 12 ± 4 days, 4-6 months | (1) Patients with higher levels of activity in the ipsilesional M1, SMA, dPMC, and contralesional cerebellum at the acute stage predicted a good motor outcome in the chronic phase. | Rehme et al. (2015) |
| (2) Patients with poor outcome showed enhanced activity in the contralesional mid-posterior insula and the ipsilesional cerebellum at the acute stage. | ||||||||
| Motor-related brain activation | Baseline (10 ± 14 days), 1-4 weeks (weekly), 3 months, 6 months, 12 months | fMRI BOLD signals | Ischemic stroke | Rankin disability scale, BI, OPSS, MI, NHPT, grip strength, ARAT, timed 10 m walk. | Baseline (10 ± 14 days), 1-4 weeks (weekly), 3 months, 6 months, 12 months | A negative correlation between task-related brain activation levels and functional recovery across time was found in a number of motor-related regions. | Ward et al. (2003) | |
| Combined approaches | SSEP+MEP | Baseline (2-5 weeks), 2 months | Time domain analysis | Ischemic and hemorrhagic stroke | FMA-UE, BI, Ashworth scale | Baseline (2-5 weeks), 2 months, 6 months, 12 months | (1) SSEPs and MEPs, in addition to clinical assessments, increased the possibility of predicting arm recovery in the long term. | Feys et al. (2000) |
| (2) The combination of the motor scores and SSEPs collected at the acute phase best predicted long-term recovery outcomes. | ||||||||
| (3) The combination of the three clinical assessments and MEPsat 2 months best predicted the long-term recovery outcomes. | ||||||||
| The PREP algorithm (SAFE scores in MRC, MEP, and the FA asymmetry of the PLIC) | Baseline (<3 days) | DWI, MEP | Ischemic stroke | ARAT | Week 2, Week 6, Week 12, Week 26 | The algorithm had positive predictive power of 88%, negative predictive power of 83%, specificity of 88%, and sensitivity of 73%. | Stinear et al. (2012) | |
| Stinear et al. (2017) |
Abbreviations: ADC, Apparent diffusion coefficient; ARAT, Action research arm test; BI, Barthel index; BOLD, Blood-oxygen-level-dependent; BSI, Brain symmetry index; CCT, Central conduction time; CR, Corona radiate; CS, Centrum semiovale; CST, Corticospinal tract; DSI, Diffusion spectrum imaging; DTT, Diffusion tensor tractography; DWI, Diffusion-weighted imaging; EEG, Electroencephalography; ERD, Event-related desynchronization; ERS, Event-related synchronization; FA, Fractional anisotropy; FC, Functional connectivity; FIM, Functional Independence Measure; FMA-UE, Fugl-Meyer Assessment-Upper Extremity; fMRI, Functional magnetic resonance imaging; IHI, Interhemispheric inhibition; LMPF, Local mean power field; M1, Primary motor cortex; MEG, Magnetoencephalography; MEP, Motor evoked potential; MI, Motricity index; MRC, Medical research council; mRS, Modified Rankin scale; NHPT, Nine hole peg test; NIHSS, National Institutes of Health Stroke Scale; OPSS, Orpington prognostic stroke scale; PLIC, Posterior limb of the internal capsule; PREP, Predict recovery potential; SAFE, Shoulder abduction and finger extension; SICI, Short-interval intracortical inhibition; SSEP, Somatosensory evoked potential; TMS, Transcranial magnetic stimulation.