Table 2.
Review of LOE of the role of MRI in the prediction of hemorrhagic transformation.
| Study | LOEa | No of Study Participants | Treatment Modality (%) | OR 95% CI, p | |
|---|---|---|---|---|---|
| Noncontrast MRI | |||||
| Leukoaraiosis (hyperintensity on T2 or FLAIR | Shi et al.45 | IV | 105 | IV thrombolysis and/or thrombolytic or mechanical IA therapy (100) | 3.4 (1.23–9.57), p = 0.019 |
| Contrast-enhanced MRI | |||||
| Contrast enhancement at T1-weighted | Yokogami et al.49 | IV | 35 | IA urokinase | p < 0.01 |
| Vo et al.50 | IV | 22 | IV thrombolysis (27.2) | p = 0.013 | |
| Kim et al.51 | IV | 55 | IV thrombolysis (27.2) | p = 0.003 | |
| Hjort et al.52 | IV | 33 | IV thrombolysis (100) | p = 0.043 | |
| Hyperintensity of CSF space termed hyperintense acute reperfusion marker on FLAIR | Latour et al.54 | IV | 144 | IV thrombolysis and/or thrombolytic or mechanical IA therapy (28.4) | 8.11 (2.85–23.1) p < 0.001 |
| Sulcal hyperintensity on FLAIR | Cho et al.55 | IV | 88 | IV thrombolysis and/or thrombolytic or mechanical IA therapy (100) | 13.64 (1.51–123.28) |
| Kim et al.56 | IV | 14 | IV thrombolysis and/or thrombolytic or mechanical IA therapy (100) | p = 0.031 | |
| DWI | |||||
| Large-sized lesion on DWI | Singer et al.58 | IV | 217 | IV thrombolysis and/or thrombolytic or mechanical IA therapy (100) | p = 0.004 |
| Selim et al.61 | IV | 29 | IV thrombolysis (100) | p = 0.032 | |
| Campbell et al.63 | IV | 49 | IV thrombolysis and/or thrombolytic or mechanical IA therapy | p = 0.003 | |
| ADC values ≤550 × 10−6 mm2/s | Tong et al.59 | IV | 17 | IV thrombolysis and/or thrombolytic or mechanical IA therapy (100) | p < 0.001 |
| Tong et al.60 | IV | 27 | IV thrombolysis and/or thrombolytic or mechanical IA therapy (59.2) | p = 0.02 | |
| Selim et al. 61 | IV | 29 | IV thrombolysis (100) | 1.176; p = 0.042 | |
| Very low CBV | Campbell et al. 62 | IV | 91 | IV thrombolysis (39.5) | 0.727 p = 0.0002 |
| Campbell et al.63 | IV | 49 | IV thrombolysis and/or thrombolytic or mechanical IA therapy | p = 0.002 | |
| PWI | |||||
| Prolonged perfusion deficit | Tong et al.60 | IV | 27 | IV thrombolysis and/or thrombolytic or mechanical IA therapy (59.2) | p = 0.03 |
| Decreased signal intensity at later time points in perfusion MRI acquisition | Bang et al.64 | IV | 32 | IV thrombolysis and/or thrombolytic or mechanical IA therapy | p < 0.001 |
| Increased relative recirculation of the contrast agent in T2* PWI | Thornhill et al.66 | IV | 18 | IV thrombolysis (44.4) | p = 0.006 |
| Large area of severe perfusion delay | Kim et al.67 | IV | 183 | IV thrombolysis and/or thrombolytic or mechanical IA therapy | 12.91 (3.69–45.17), p < 0.001 |
| T2*-weighted GRE sequences and SWI | |||||
| Cerebromicrobleeds: small, rounded, homogeneous, hypointense lesions | Nighoghossian et al.70 | IV | 100 | IV thrombolysis (27) | p < 0.001 |
| Kidwell et al.71 | IV | 41 | p < 0.05 | ||
| Abnormal visibility of transcerebral veins | Hermier et al.72 | IV | 49 | IV thrombolysis (100) | p = 0.001 |
ADC: apparent diffusion coefficient; CBV: cerebral blood volume; CSF: cerebrospinal fluid; DWI: diffusion-weighted MRI; FLAIR: fluid-attenuated inversion recovery; GRE: gradient-echo sequences; IA: intra-arterial; IV: intravenous; LOE: level of evidence; MRI: magnetic resonance imaging; PWI: perfusion-weighted MRI; SWI: susceptibility-weighted imaging.
LOE adopted from Ackley et al. as follows: level 1: systematic review or meta-analysis, level II: randomized controlled trial (RCT), level III: nonrandomized controlled trials, level IV: case control studies, cohort studies, level V: meta-synthesis, and level VI: single descriptive or qualitative study.20