Abstract Archives of the RSNA, 2014
Sven Haller MD, MSc, Presenter: Nothing to Disclose
Panteleimon Giannakopoulos, Abstract Co-Author: Nothing to Disclose
Constantin Bouras, Abstract Co-Author: Nothing to Disclose
Eniko Kovari, Abstract Co-Author: Nothing to Disclose
Karl-Olof Lovblad, Abstract Co-Author: Nothing to Disclose
Microbleeds are radiologically defined as small, punctiform hypointense brain lesions of signal loss on gradient-echo T2* or susceptibility weighted imaging (SWI). Due to the increasing application of susceptibility imaging and increasing magnetic field strength, the detection rate or microbleeds increases notably in the elderly. We assessed the pre-mortem MRI radiologic-histopathologic correlation of cerebral microbleeds in the elderly
This retrospective study identified all cases of brain autopsy during a 10-year period from 1.1.2000 and 31.12.2010. 1064 cases were autopsied as part of a systemic procedure in the local academic geriatric hospital. 134 of those cases had pre-mortem MRI. Histopathologic slices were obtained in all cases with cerebral T2* lesions corresponding to the anatomic findings on MRI and evaluated independently by 2 experienced neuropathologists.
13 cases had 15 hypointense lesions on gradient-echo T2*w imaging on pre-mortem MRI. In 8 / 13 patients, there was a concordance between MRI and histopathology (true positive rate of 61.5%). In 3 / 13 patients, the T2* lesions did not correspond to histopathologic lesion (microcalcification as “microbleed mimic” in 1 / 13, absent lesion in 2 / 13), corresponding to a false positive rate of 23.1%. In 2/13 patients there were additional multiple microbleeds on histopathology indicating a false negative rate of 15.4%
The majority (8/13 cases) of hypointense lesions on T2* imaging correspond to histopathologically confirmed mircrobleeds. MRI was false positive in 3/13 cases and one microbleed mimics was a microcalcification. Conversely, in 2/13 cases MRI was false negative.
Additional post-mortem MRI is currently repeated in selected cases to assess the number of new microbleeds appearing in the delay between pre-mortem MRI and death.
Better understanding of the radiologic-histopathologic correlation of hypointense brain lesions on T2* and SWI will enable a more accurate discrimination of cerebral microbleeds versus microbleed mimics.
Haller, S,
Giannakopoulos, P,
Bouras, C,
Kovari, E,
Lovblad, K,
Radiologic-histopathologic Correlation of Pre-mortem and Post-mortem Cerebral Mircrobleeds in the Elderly. Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL.
http://archive.rsna.org/2014/14005743.html