J Neurol (2009) 256:285–286 DOI 10.1007/s00415-009-0136-7 LETTER TO THE EDITORS JON 3136 An 80-year old woman was admitted with sudden right-sided hemiparesis, dysphasia and altered mental status. There was no pre- ceding infection or vaccination. She had an epileptic seizure after which we started phenytoin. Vital signs were BP 130/70, pulse 87, tempera- ture 37.1 °C. MRI revealed white matter abnormalities, exerting a slight mass effect (Fig. 1 A). No gadolinium enhancement was ob- served. Serum and cerebrospinal fluid studies were within normal limits. After exclusion of infection we started dexamethasone 4 mg/ day and observed a clear recovery. The initial follow-up MRI was comparable with the previous MRI. Diffusion weighted images sug- gested that the white matter ab- normalities represented vasogenic edema (Fig. 1 D). A T2*-weighted gradient echo sequence revealed multiple petechial hemorrhages or cerebral microbleeds (CMB, Fig. 1 B). In the absence of the clas- sical vascular risk factors, we diag- nosed CAA and continued dexame- thasone in a gradually tapering scheme. The patient recovered, but relapsed after two months. She responded well to a temporary increase of dexamethasone. Four months later we could stop the dexamethasone. She had a mild residual dysphasia with a full recovery of cognition. Repeated MRI showed a clear reduction in brain edema (Fig. 1 C). Following the Boston criteria [3], we classified our patient as ‘probable CAA’. Post-mortem neuropathological examination is still considered the most reliable tool for diagnosing CAA. CAA can be identified in vivo by histological analyses upon evacuating a hema- toma or by cortical biopsy. Because our patient responded well to ste- roid treatment, we did not perform a brain biopsy. The need for biopsy in the diagnosis of CAA is likely to decrease in the future, since Knud- Richard Daniëls Jeroen J. G. Geurts Joost C. Bot Wouter J. Schonewille Bob W. van Oosten Steroid-responsive edema in CAA-related inflammation Received: 11 June 2008 Received in revised form: 7 October 2008 Accepted: 6 November 2008 Published online: 23 January 2009 Sirs: Cerebral amyloid angiopathy (CAA) forms a heterogeneous group within the cerebral small vessel diseases, characterized by congophilic amyloid vessel wall deposits [1]. Several studies have emphasized the importance of CAA in cerebral brain hemorrhage of non-traumatic and non-hyper- tensive origin, and to a lesser extent in ischemic stroke and leukoencephalopathy with cogni- tive impairment and cortical dys- function [2]. CAA classically pres- ents as a spontaneous intracerebral hemorrhage. We present a less common manifestation of CAA. sen et al. reported that a reliable diagnosis of CAA can be reached from MR imaging and clinical find- ings alone [3]. In recent years, inflammatory changes associated with CAA have been increasingly recognized. A study with 42 patients diagnosed with CAA suggested that in these patients accumulation of cerebral vascular fibrillar Aβ promotes perivascular inflammation, accord- ing to localization and close rela- tion of immune cells to Aβ- containing plaque deposits. The clinical presentation of these pa- tients differs from most CAA pa- tients. As in other CAA patients, multiple CMB and small infarc- tions in the cerebral cortex are seen, resulting in subacute cogni- tive decline and seizures. However, lobar hemorrhagic stroke is rare. These differences might have a ge- netic background, since the APOE ε4/ε4 genotype is much more fre- quently seen in patients who pres- ent with CAA-related inflammation [4]. Furthermore, these patients have a considerable chance of recovery with steroids [5, 6]. In our case, radiographic imag- ing showed extensive edema, which was also reported in similar CAA cases in the literature [7]. The cor- responding reversible, asymmetric T2-hyperintense MRI changes are fairly specific and enable differenti- ation from CAA without associated inflammation and reversible poste- rior leukoencephalopathy [6]. CMB will further support a diagnosis of CAA and should be looked for us- ing T2*-weighted (FLASH2D gradi- ent-echo) MR sequences. This leads Kinnecom et al. to suggest that in the appropriate clinical setting CAA-related inflammation can be diagnosed without brain biopsy [6]. Most of these patients respond well to anti-inflammatory treat- ment [6]. With this case study, we wish to stress the importance of recognizing this less well-known R. Daniëls · J. J. G. Geurts · J. C. Bot Dept. of Radiology VU University Medical Center P.O. Box 7057 1007 MB Amsterdam, The Netherlands J. J. G. Geurts Dept. of Pathology VU University Medical Center P.O. Box 7057 1007 MB Amsterdam, The Netherlands W. J. Schonewille Dept. of Neurology St Antonius Ziekenhuis P.O. Box 2500 3430 EM Nieuwegein, The Netherlands B. W. van Oosten () VU University Medical Center Dept. of Neurology, Room 2F 29 P.O. Box 7057 1007 MB Amsterdam, The Netherlands Tel.: +31-20/444-1981 Fax: +31-20/444-2800 E-Mail: bw.vanoosten@vumc.nl