Possibility Of Superficial Siderosis Of The Central Nervous System Formation After Single Episode Of Traumatic Subarachnoid Hemorrhage:A Study With MRI Enhanced Gradient Echo T2Star Weighted Angiography Technique

Objective:To study the possibility of superficial siderosis of the central nervous system (SS-CNS) formation after traumatic subarachnoid hemorrhage (tSAH) by enhanced gradient echo T2star weighted angiography (ESWAN) technique.Materials and Methods:From January2009to October2012,30consecutive patients with tSAH were recruited in this prospective study. All patients met the following criteria:(1) Patient had a history of head injury and was diagnosed as SAH with plain computed tomography (CT) or cerebrospinal fluid examination.(2) Patient was recove after conservative treatment. A follow-up plain CT examination over6weeks showed absorption of the SAH.(3) No history of craniotomy or central nervous system disease before head injury. All the patients suffer from road traffic injury. Informed consents were obtained from all of the recruited patients and contral volunteers. The first head CT scans were completed within6hours after head injury. A subsequently lumbar puncture was performed within24hours after head injury in patient with negative CT diagnosis but high clinically suspiction.The MR ESWAN sequence were performed in these patients6weeks later to the first CT scanning, whose followup CT scanning were normal.T2*-weighted gradient recalled echo(GRE T2*WI) were performed at the same time in22cases. Totally30patients fulfilled the criteria and constituted the tSAH group. The patients examined in the study consisted of20men and10women, aged from8to87years old (mean age,49.0±16.8years old). MRI was performed on a1.5-T (n=21) or3.0-T (n=9) scanner,6to167weeks after the tSAH. As a control, we also investigated25healthy volunteers in this study.Two experienced radiologists evaluated the MR images of the tSAH group (30cases) and control groups (25cases) individually,emphasizing on the presence or absence of SS-CNS on the magnitude image of ESWAN sequence. Kappa coefficients agreement was calculated to estimate repeatability of imaging evaluations.Two experienced radiologists evaluated the magnitude image of ESWAN sequence and GRE T2*WI sequence for the presence or absence of SS-CNS with a consensus agreement. The positive rate of two sequences were compared by using Chi-square test,the number of SS-CNS distributing regions detected by two sequences were compared with the Wilcoxon signed-rank test. The SS-CNS positive patiens on the magnitude image of ESWAN sequence were then divided into2subgroups (n≥5or n≤4) according the number of SS-CNS regions.The univariate analyses were done for the following variables:age (≥50years); sex; Fisher grade (≥3); subdural hematoma on initial CT; brain contusions and intracerebral hematomas on initial CT; and interval between initial CT and MRI examination (<26weeks), the Fisher’s exact test was used to find influence factors. The patients diagnosed with tSAH by plain CT and SS-CNS by ESWAN sequence, Pearson correlation coefficient was performed for comparing correlation coefficients between the number of SS-CNS regions on the magnitude image of ESWAN sequence and the number of SAH regions on the CT. The relationship between the tSAH Fisher grade and the number of SS-CNS regions on the magnitude image of ESWAN sequence were asessed by using Kendall’s tau rank correlation coefficient.Results:(1)Two radiologists evaluated of the presence or absence of SS-CNS in tSAH patients on the magnitude ESWAN images with a Kappa coefficient of0.651(P<0.01), which indicating a well agreement.(2) Twenty-nine of30(96.7%) patients demonstrated SS-CNS on the magnitude ESWAN images.22patients were identified by both ESWAN and GRE T2*WI.The positive rate was95.5%(21/22) on ESWAN and68.2%(15/22) on GRE T2*WI, with a significant differences (χ2=4.167, P<0.05).The number of SS-CNS regions on magnitude ESWAN images and GRE T2*WI MR images was127and33spectively, with a significant difference between the two different sequences (Z=4.034,P<0.01). No SS-CNS was seen in healthy volunteers on either of the two different sequences. Of the29patients with SS-CNS, the diposition proportion was found in the frontal (69.0%), temporal (34.5%), parietal (58.6%), occipital areas (17.2%), sylvian fissure (51.7%), cerebellum (31.0%) and Brainstem (13.8%).(3) The SS-CNS-positive patients was divided into2subgroups according to the number of SS-CNS regions (n≥5or n<4). We demonstrated on Fisher’s exact test that Fisher grade(≥3), brain contusions on initial CT were significantly associated with the extent of SS-CNS.(4) Twenty-seven patents diagnosed with tSAH by plain CT and SS-CNS by ESWAN sequence, the number of SS-CNS regions on the magnitude ESWAN images and the number of SAH regions on the CT are highly correlated (r=0.785,P<0.05). The tSAH Fisher grade was significantly positively correlated the number of SS-CNS regions on the magnitude ESWAN images (Kendall’s tau-b=0.610, P<0.01). Conclusion:(1) ESWAN is an effective MR sequence to detect SS-CNS with high sensitivity and good repeatability.(2) Patient has a high possibility to develop SS-CNS after a single episode of tSAH.(3) SS-CNS preferentially deposited at frontal, parietal, sylvian fissure, temporal, cerebellum, occipital and the brainstem after tSAH. There is a high correlation between the extention of tSAH and SS-CNS deposion regions.(4) The extention of SS-CNS was significantly associated with CT Fisher grade (≥3) combined with brain contusions. There is a positive correlation between Fisher scale for grading tSAH and range of SS-CNS deposition.