TY - JOUR
T1 - Correlations among brain and spinal MRI findings in spontaneous intracranial hypotension
AU - Wu, Jr Wei
AU - Wang, Yen Feng
AU - Fuh, Jong Ling
AU - Lirng, Jiing Feng
AU - Chen, Shih Pin
AU - Hseu, Shu Shya
AU - Wang, Shuu Jiun
N1 - Publisher Copyright:
© International Headache Society 2018.
PY - 2018/12/1
Y1 - 2018/12/1
N2 - Objectives: Several brain and spinal magnetic resonance imaging signs have been described in spontaneous intracranial hypotension. Their correlations are not fully studied. This study aimed to explore potential mechanisms underlying cerebral neuroimaging findings and to examine associations among spinal and brain magnetic resonance imaging signs. Methods: We conducted a retrospective review of magnetic resonance myelography and brain magnetic resonance imaging records of patients with spontaneous intracranial hypotension. Categorical principal component analysis was employed to cluster brain neuroimaging findings. Spearman correlation was employed to analyze associations among different brain neuroimaging findings and between brain and spinal neuroimaging findings. Results: In patients with spontaneous intracranial hypotension (n = 148), categorical principal component analysis of brain neuroimaging signs revealed two clusters: Cerebral venous dilation and brain descent. Among all brain magnetic resonance imaging signs examined, only midbrain-pons angle associated with anterior epidural cerebrospinal fluid collection length (surrogate spinal cerebrospinal fluid leak severity) (n = 148, Spearman’s ρ = −0.38, p <.001). Subgroup analyses showed that the association between midbrain-pons angle (within brain descent cluster) and spinal cerebrospinal fluid leak severity was presented in patients with convex margins of the transverse sinuses (n = 122, Spearman’s ρ = −0.43, p <.001), but not in patients without convex margins (n = 26, Spearman’s ρ = −0.19, p =.348). The association between severity of transverse sinus distension and spinal cerebrospinal fluid leak severity was only presented in patients without convex margins (n = 26, Spearman’s ρ = 0.52, p =.006). Conclusion: This study indicates that there are two factors behind the brain neuroimaging findings in spontaneous intracranial hypotension: Cerebral venous dilation and brain descent. Certain brain neuroimaging signs correlate with spinal cerebrospinal fluid leakage severity, depending on different circumstances.
AB - Objectives: Several brain and spinal magnetic resonance imaging signs have been described in spontaneous intracranial hypotension. Their correlations are not fully studied. This study aimed to explore potential mechanisms underlying cerebral neuroimaging findings and to examine associations among spinal and brain magnetic resonance imaging signs. Methods: We conducted a retrospective review of magnetic resonance myelography and brain magnetic resonance imaging records of patients with spontaneous intracranial hypotension. Categorical principal component analysis was employed to cluster brain neuroimaging findings. Spearman correlation was employed to analyze associations among different brain neuroimaging findings and between brain and spinal neuroimaging findings. Results: In patients with spontaneous intracranial hypotension (n = 148), categorical principal component analysis of brain neuroimaging signs revealed two clusters: Cerebral venous dilation and brain descent. Among all brain magnetic resonance imaging signs examined, only midbrain-pons angle associated with anterior epidural cerebrospinal fluid collection length (surrogate spinal cerebrospinal fluid leak severity) (n = 148, Spearman’s ρ = −0.38, p <.001). Subgroup analyses showed that the association between midbrain-pons angle (within brain descent cluster) and spinal cerebrospinal fluid leak severity was presented in patients with convex margins of the transverse sinuses (n = 122, Spearman’s ρ = −0.43, p <.001), but not in patients without convex margins (n = 26, Spearman’s ρ = −0.19, p =.348). The association between severity of transverse sinus distension and spinal cerebrospinal fluid leak severity was only presented in patients without convex margins (n = 26, Spearman’s ρ = 0.52, p =.006). Conclusion: This study indicates that there are two factors behind the brain neuroimaging findings in spontaneous intracranial hypotension: Cerebral venous dilation and brain descent. Certain brain neuroimaging signs correlate with spinal cerebrospinal fluid leakage severity, depending on different circumstances.
KW - Spontaneous intracranial hypotension
KW - brain MRI
KW - cerebrospinal fluid
KW - magnetic resonance myelography
UR - http://www.scopus.com/inward/record.url?scp=85057542587&partnerID=8YFLogxK
U2 - 10.1177/0333102418804161
DO - 10.1177/0333102418804161
M3 - Article
C2 - 30301380
AN - SCOPUS:85057542587
SN - 0333-1024
VL - 38
SP - 1998
EP - 2005
JO - Cephalalgia
JF - Cephalalgia
IS - 14
ER -