Afaaf Shakir1, Didem Aksoy2, Michael Mlynash3, Odette A Harris4, Gregory W Albers5, Karen G Hirsch6. 1. Stanford University School of Medicine, Stanford, CA. 2. Stanford University School of Medicine, Department of Radiology, Stanford Stroke Center, Stanford, CA. 3. Stanford University School of Medicine, Department of Neurology, Stanford Stroke Center, Stanford, CA. 4. Stanford University School of Medicine, Department of Neurosurgery, Stanford, CA. 5. Stanford University School of Medicine, Department of Neurology, Director, Stanford Stroke Center, Stanford, CA. 6. Stanford University School of Medicine, Department of Neurology, Director, Neurocritical Care, Stanford, CA.
Abstract
BACKGROUND AND PURPOSE: Data about the predictive value of quantitative diffusion-weighted MRI in traumatic brain injury (TBI) patients is lacking. This study aimed to determine if specific apparent diffusion coefficient (ADC) thresholds could be determined that correlate with outcome in moderate-severe TBI. METHODS: This retrospective observational study investigated patients with moderate-severe TBI. MRIs obtained post-injury days 1-13 were analyzed. MRIs were obtained on a 1.5T scanner; 20-23 contiguous diffusion-weighted imaging (DWI) sections with a spin-echo echo planar imaging DWI 256×256 reconstructed matrix; field of view 24×24 cm; slice thickness/gap of 5/1.5 or 5/2.5 mm. The ADC value of each brain tissue voxel was determined. The percentage of voxels below different ADC thresholds was calculated and correlated with outcome. A good outcome was defined as discharge to home or a rehabilitation facility. RESULTS: Seventy-six patients were analyzed. Thirty-five patients (46%) had a good outcome. The timing of MRI scans did not differ between groups, but the mean age did (42±18 years vs. 56±19 years, p<.01, good vs. poor outcome). Patients with poor outcome had significantly higher percentage of brain volume with ADC < 400×10(-6) mm2 /second (.85±.67% vs. .60±.29%, poor vs. good outcome, p<.05). Using a ROC curve analysis and Youden's index, an ADC <400×10(-6) mm2 /second in ≥.49% of brain was 85% sensitive and 46% specific for poor outcome (p<.05). CONCLUSIONS: Quantitative MRI offers additional prognostic information in acute TBI. A whole brain tissue ADC threshold of <400×10(-6) mm2 /second in ≥.49% of brain may be a novel prognostic biomarker.
BACKGROUND AND PURPOSE: Data about the predictive value of quantitative diffusion-weighted MRI in traumatic brain injury (TBI) patients is lacking. This study aimed to determine if specific apparent diffusion coefficient (ADC) thresholds could be determined that correlate with outcome in moderate-severe TBI. METHODS: This retrospective observational study investigated patients with moderate-severe TBI. MRIs obtained post-injury days 1-13 were analyzed. MRIs were obtained on a 1.5T scanner; 20-23 contiguous diffusion-weighted imaging (DWI) sections with a spin-echo echo planar imaging DWI 256×256 reconstructed matrix; field of view 24×24 cm; slice thickness/gap of 5/1.5 or 5/2.5 mm. The ADC value of each brain tissue voxel was determined. The percentage of voxels below different ADC thresholds was calculated and correlated with outcome. A good outcome was defined as discharge to home or a rehabilitation facility. RESULTS: Seventy-six patients were analyzed. Thirty-five patients (46%) had a good outcome. The timing of MRI scans did not differ between groups, but the mean age did (42±18 years vs. 56±19 years, p<.01, good vs. poor outcome). Patients with poor outcome had significantly higher percentage of brain volume with ADC < 400×10(-6) mm2 /second (.85±.67% vs. .60±.29%, poor vs. good outcome, p<.05). Using a ROC curve analysis and Youden's index, an ADC <400×10(-6) mm2 /second in ≥.49% of brain was 85% sensitive and 46% specific for poor outcome (p<.05). CONCLUSIONS: Quantitative MRI offers additional prognostic information in acute TBI. A whole brain tissue ADC threshold of <400×10(-6) mm2 /second in ≥.49% of brain may be a novel prognostic biomarker.
Authors: Peter A Ferrazzano; Bedda L Rosario; Stephen R Wisniewski; Nadeem I Shafi; Heather M Siefkes; Darryl K Miles; Andrew L Alexander; Michael J Bell Journal: J Neurosurg Pediatr Date: 2019-02-08 Impact factor: 2.375