Clin Neuroradiol (2010) 20:189–220 DOI 10.1007/s00062-010-0025-6

Abstracts

DGNR 2010 Jahrestagung der Deutschen Gesellschaft für Neuroradiologie 22. – 25. September 2010 im Rahmen der Neurowoche Mannheim | 21. – 25. September 2010

Tagungsleitung der Deutschen Gesellschaft für Neuroradiologie (DGNR) Prof. Dr. C. Groden

190

Abstracts

Intervention V1422 Endovascular treatment of blood blister-like aneurysms: initial experience with stent-assisted coil embolization S. Meckel, M. Cronqvist, T. Singh, C. Phatouros, W. McAuliffe (Freiburg; Lund, SE; Perth, AU) Purpose: Blood blister-like aneurysm (BBA) is a rare type of intracranial aneurysm which generally presents with severe SAH associated with high degree of morbidity and mortality. Due to their fragile walls and awkward geometry they are difficult to treat both surgically and endovascularly. The aim of this study was to present clinical and angiographic results of endovascular BBA treatment using predominately stent-assisted coil embolization. Methods:  13 patients, harbouring 13 ruptured BBA, from two different institutions (men/women, 6/7; mean age, 49.3 years) were included in this study from September 2006— May 2009. Clinical and angiographic findings, treatment strategies, anticoagulation routines, procedure-related complications, and outcome i mmediate post-coiling and midterm follow-up (mean, 12 months [6–27 months]) were retrospectively analyzed. Results:  11 BBAs were located in the supraclinoid ICA and two the basilar artery trunk. Most BBAs showed early growth (61.5%), 7 were ≤ 3 mm in largest diameter. Primary stent-assisted coiling was performed in 11 patients. Complementary treatment due to early/late aneurysm regrowth after incomplete occlusion was required in three patients including parent artery occlusion in two. Two patients had early re-hemorrhage, one presented with asymptomatic minor bleeding, and one with major fatal hemorrhage. Late angiographic follow-up showed complete occlusion/progressive occlusion in 92.3% (12/13). At mid-term clinical follow-up, 92.3% (12/13) showed good functional outcome (MRS 0–2). Reduced heparinazation (50%) and antiplatelet routines (30%) were used in stent-assisted procedures without evidence of early or late thromboemboli. Conclusion:  Stent-assisted coiling for treatment of ruptured BBAs is technically challenging but feasible. Reduced anticoagulation/antiplatelet regimes appear safe relative to symptomatic thrombembolic complications. Regrowth and early re-rupture remains a problem in incompletely occluded BBAs underlining the importance of early angiographic follow-up, and re-treatment including PAO may be necessary for definite occlusion. The mid-term aniographic and clinical outcome was excellent in our series.

V1423 Endovascular treatment of cerebral aneurysms with the pipeline embolization device M. Hartmann, S. Rohde, C. Braun, S. Hähnel, M. Bendszus (Heidelberg) Objectives:  The “flow diverters”, a family of high metal surface area coverage, self-expanding, braided, microstents, represent a novel technology which has recently become co mmercially available in Europe for the treatment of intracranial aneurysms. We present our initial experience with the Pipeline embolization device (PED, eV3, Irvine, CA) for a series of 8 patients. Methods:  From Oct/09 to March/10 eight patients (age range, 9–56 years; average age, 45.9 years; 4 female) harbouring 9 non-ruptured large and giant wide- necked aneurysms were treated. All patients were pretreated with dual antiplatelet medications for at least 72 hours or with a loading dose before treatment and continued taking both agents for at least 3 months and one agent for at least 6 months after treatm ent. Follow-up was performed with Dyna-CT and/or contastenhanced MR angiography. Mean followup (FU) time was 2 months (range, 0.2–5 months). Results:  A total of 22 PEDs were used (single PED in 4, 3 overlapping PEDs in 3, and 5 overlapping PEDs in 2 aneurysms). All cases were uneventfully completed and no procedural complications were encountered. All patients emerged from general anesthesia neurologically intact. 6 patients were without any new clinical deficit at discharge and FU. One patient experienced an ipsilaterlal parenchymal hemorrhage within 24 hours after treatment within the parenchyma supplied by the circulation ipsilateral to the treated lesion, but remote from the targeted aneurysm and implanted flow diverting device. One patient died due to mass effect and SAH from a treated giant basilar aneurysm after 72 hours. From the 6 patients with a follow-up of more than 2 months 3 aneurysm showed complete thrombosis and 3 progressive thrombosis.

Abstracts

191

Conclusion:  Our preliminary experience with the PED shows that the procedure is safe and a the use of flowdiverters can achieve curative reconstruction of the parent arteries giving rise to widenecked and fusiform intracranial aneurysms. Delayed spontaneous ipsilateral parenchymal hemorrhage, subarachnoid hemorrhage and increasing mass effect following aneurysm treatment may represent complications related to the application of flow diverting stent-like devices. V1424 Management von distal zerebellären Aneurysmen A. Harati, F. Lohmann, J. Stasierowski, R. Luckner, C. Loehr, M. J. A. Puchner, A. Mpotsaris, W. Weber (Recklinghausen) Einleitung:  Distal zerebelläre Aneurysmen sind eine seltene Entität, deren Behandlung aufgrund der operativ schwierigen anatomischen Lage und der häufigen Assoziation mit einer arteriovenösen Malformation (AVM) komplex ist. Die rein neurochirurgische Sanierung war über lange Zeit die einzige Therapieform, die häufig mit einer hohen Morbidität und Mortalität vergesellschaftet war. Im Rahmen der Fallserie stellen wir unser Konzept der kombinierten neurochirurgisch-endovaskulären Behandlung vor. Methoden:  Zwischen      Okt. 2007 und Dez. 2009 wurden ca. 200 Patienten mit intrakraniellen Aneurysmen in unserer Klinik behandelt. Neun Patienten hatten mind. ein distal zerebelläres Aneuryma mit einer klinischen und angiographischen Verlaufskontrolle nach einem Jahr. Ergebnisse:  Der Anteil der distal zerebellären Aneurysmen an allen intrakraniellen Aneurysmen war 4,8 %. Fünf der neun Patienten hatten eine AVM der hinteren Schädelgrube. Drei der fünf Patienten hatten multiple Aneurysmen. Ein Patient hatte 5 AVM-assozierte Aneurysmen der A. cerebelli superior (Abb. 1 vor/Abb. 2 nach der Behandlung). Abb. 1

Abb. 2

Alle Patienten mit einem AVM-assoziierten distal zerebellärem Aneurysma wurden durch ein Blutungsereignis symptomatisch. Blutungsquelle war dabei i mmer das Aneurysma und nicht die AVM. Im Patientenkollektiv der nicht AVM-assozierten distal zerebellären Aneurysmen trat ein Blutungsereignis hingegen nur bei einem Patienten auf. Im

Gesamtkollektiv wurden drei Patienten rein endovaskulär, ein Patient rein neurochirurgisch und zwei Patienten kombiniert endovaskulär-neurochirurgisch behandelt. Drei Patienten ohne ein vorangegangenes Blutungsereignis wurden nicht behandelt und im Verlauf kontrolliert. Sieben der neun Patienten hatten ein gutes bis exzellentes Outcome (GOS 4–5). Ein Patient hatte ein GOS von 3. Ein Patient verstarb 7 Tage nach dem Blutungsereignis. Konklusion:  Die AVM-assoziierten distal zerebellären Aneurysmen haben gegenüber den nicht AVM-assoziierten ein deutlich höheres Blutungsrisiko. Primär sollte die Ausschaltung des Aneurysmas im Vordergrund stehen. Die Behandlung der AVM kann im Intervall erfolgen. Bei den nicht AVM-assozierten distal zerebellären Aneurysmen, die einen deutlich günstigeren Krankheitsverlauf haben, sollte individuell entschieden werden, ob und welche Therapieform zur Ausschaltung des Aneurysmas angewendet werden soll. In jedem Fall sollte ein interdisziplinäres Vorgehen mit Berücksichtigung des klinischen Zustandes des Patienten und der anatomischen Lokalisation der Pathologie erfolgen. V1425 Selective disconnection of cortical and leptomeningeal venous drainage during transvenous occlusion of cavernous sinus fistulas G. Benndorf, A. Biondi, H. Morsi, S. Gopinath (Houston, US; Paris, FR) Background/purpose:  Direct and indirect cavernous sinus fistulas (CSFs) may present with cortical and leptomeningeal venous drainage that can lead to intracranial hemorrhage during the course of the disease or during transvenous coil occlusion (TVO). Selective disconnection of such venous exit during TVO can be difficult due to the complex anatomy of the sinus and its afferent and efferent venous co mmunications. Material & methods:  Six patients, 2 direct carotid cavernous sinus fistula (CCFs) and 4 dural cavernous sinus fistulas (DCSFs) with cortical or leptomeningeal venous drainage were treated with TVO. If leptomeningeal venous drainage was directed posteriorly, selective coil occlusion of the cavernous sinus-superior petrosal sinus junction was performed. If cortical drainage was directed anterolaterally, selective coil occlusion of the cavernous sinus-Sylvian vein junction was performed. Precise visualization of the co mmunication between cavernous sinus and Sylvian vein standard AP and lateral views was difficult, but overcome by using a caudal/ oblique working projection. In two patients, simultaneous placement of 2 microcatheters through the inferior petrosal sinus (IPS) was used to secure selective occlusion of the anterior drainage into the superior ophthalmic vein, and of the anterolateral drainage into the Sylvian vein during packing of the main fistulas compartment.

192

Results:  In all patients, selective disconnection of the cortical and leptomeningeal venous drainage was successfully performed prior to complete occlusion of the fistula. Dual catheter placement in the cavernous sinus facilitated selective occlusion of the different venous exits. All patients fully recovered from their neuroophthalmological symptoms and showed stable occlusions in follow-up exams. Conclusion: Selective disconnection of cortical and leptomeningeal venous drainage from the cavernous sinus is feasible and should be aimed for, preferably at the beginning of TVO of CSFs. It can help to prevent persisting or rerouting venous drainage into leptomeningeal and cortical veins that may lead to venous hypertension and intracranial hemorrhage. V1426 Protective stenting in skull base tumours F. Götz, A. Giesemann, O. Majdani, H. Lanfermann, T. Lenarz (Hannover) Purpose:  Protective stenting is a novel method to reinforce an artery against invasion by tumor or injury during surgery. We analysed the indications for protective stenting in our cohort and report on the follow up of treated patients. Method:  We reviewed our angiographic data base and extracted all stent-cases from 2005 to 2010. Patients with intracranial stenosis or aneurysms were excluded. Patients with stent implantation and concurrent skull-base tumors were analysed in detail. All available CT, MRI and DSA examinations were reviewed on a PACS workstation by two radiologists. Clinical charts, surgery and pathology reports as well as clinical notes during follow-up visits were analysed. Implanted stents were registered according to type, number, length and diameter. Results:  Thirteen self-expanding intracranial stents (Leo and Leo+, ab medica) were implanted in nine patients. We used stents with a nominal diameter from 3,5 to 5,5  mm and a length between 25–75  mm. It was indicated to cover arterial segments form the co mmon carotid to the C5-segment. The segment C2 of the internal carotid artery was covered most (in 8 patients). The V3 and V4 segments of the vertebral artery were stented once. Usually one stent per patient was used but three overlapping stents were implanted twice in order to cover a longer distance. Implantation of stents was completed without neurological complications in all patients. On follow-up angiography severe narrowing of one carotid (three stents with overlap) occurred, but patency of all stented arteries was demonstrated. Surgery was done not earlier than 6 weeks after the stenting procedure. One operation had to be discontinued due to

Abstracts

anaesthesiological problems. In all patients dissection was facilitated due to reinforcement of the arterial wall by the stent. Extended resection of tumor and even dissection of adventitia of the artery was possible. Discussion:  Using protective stenting, extended tumor resection and vessel preservation was achieved in all operated patients. In highly selected patients protective stenting is a promising method to keep cerebral circulation alive and preserve an artery in danger of tumor erosion or sacrifice during surgery. Using this novel method carotid occlusion testing is no longer required. V1427 Rate of recanalization of large intracranial arteries in acute stroke—a comparison of i. a. thrombolysis monotherapy, the penumbra system and solitaire stent M. Schick, W. Schlötzer, M. Müller, B. Schmitz (Ulm) Purpose:  The rate of recanalization in large intracranial vessel occlusion in acute stroke with intraarterial (i. a.) thrombolysis is limited. The purpose of the study was to evaluate additional tools like penumbra system or solitaire stent in combination with i. a. thrombolysis. Methods:  From January 2008 to March 2010 patients with occlusion of major vessels in acute stroke including carotidT, M1 of the middle cerebral artery, the basilar artery and P1 of the posterior cerebral artery received i. a. thrombolysis as monotherapy or in combination with penumbra system or solitaire stent. The end point was revascularisation of the target vessel to grade 2b or 3 on the Thrombolysis in cerebral infarction (TICI) scale. Results:  51 patients with 99 occluded major arteries were treated. 18 patients with 31 occluded arteries underwent i. a. thrombolysis as monotherapy, 7 patients with 11 occluded arteries received i. a. thrombolysis in combination with penumbra system and 18 patients with 38 occluded vessels in combination with solitaire stent. In 9 patients with 21 occluded vessels i. a. thrombolysis in combination with solitaire stent with mechanical clot retrieval while applying continuous aspiration was performed. A recanalization to TICI 2b or 3 was achieved in 23% of patients with i. a. thrombolysis monotherapy, 45% and 39% in combination with penumbra system and solitaire stent, respectively. The use of the solitaire stent with clot retrieval under aspiration resulted in 81% rate of recanalization to TICI 2b or 3.

Abstracts

193

Wirbelsäule

Fig. 1 Rate of recanalization

Therapy

i.a.thrombolysis

39%

penumbra systems and i.a. thrombolysis

39%

27%

soiliaire stent and i.a. thrombolysis

27%

21%

soiltaire stent with clot retrieval and i.a. thrombolysis

10%

0%

45%

39%

39%

10%

10%

V1430 Multi-level MR Neurography of traumatic plexus injuries: more than imaging of root avulsions M. Pham, P. Bäumer, T. Kästel, A. Bartsch, S. Heiland, M. Bendszus (Heidelberg)

23%

81%

20%

30%

40% TICI 0/1

50% TICI 2a

60%

70%

80%

90%

100%

TICI 2b/3

There were no device related complications like dissections or perforations. Conclusions: The use of a mechanical thrombectomy device improves the rate of recanalization of occluded major arteries in acute stroke. Especially the solitaire stent, if used as a clot retrieval tool achieved excellent rates of recanalization. V1428 Rekanalisation einer chronisch verschlossenen Arteria carotis interna – erste Erfahrungen an 3 Patienten J. Hattingen, R. Glombik, E. Gmelin, A. Schwartz (Hannover) Einleitung: Symptomatische chronische Verschlüße der Arteria carotis interna (ACI) werden zumeist mit einem Extra-Intracraniellem Bypass behandelt. Die interventionelle Rekanalisation von akuten Verschlüßen der ACI ist ein geläufiges Verfahren der Rekanalisation in der Akutsituation des Schlaganfalls. Das es technisch möglich ist, auch chronische Verschlüsse der ACI wieder zu eröffnen, möchten wir anhand dreier Fälle dokumentieren. Materal und Methode:  Im Zeitraum zwischen dem 1.1.2009 und 31.3.2010 behandelten wir drei Patienten (2 W, 1 M) mit symptomatischen Verschlüssen der ACI bei insgesamt komplexen Gefässbildern. Zur Vorbereitung der neuroradiologischen Interventionen gehörten neben Doppler- und Duplex Untersuchungen auch MRT Untersuchungen mit DWI, PWI und CE-MRA. Ergebnisse:  Die Interventionen wurden unter Flußumkehrbedingungen durchgeführt. Es gelang, das wahre Lumen der ACI zu erreichen und nach Freisetzen von mehreren Stents u. a. auch intracraniell einen antegraden Fluß zu erzielen. Postinterventionell waren alle Patienten neurologisch beschwerdefrei. IM MRT fanden sich bei zwei der drei Patienten DWI Läsionen. Schlussfolgerung:  Die neuroradiologisch interventionelle Rekanalisation eines symptomatischen Verschlusses der ACI ist technisch möglich. Sie kann eine alternative Methode zum Extra-Intracraniellen Bypass sein.

Traumatic plexus injuries affect young individuals and often lead to severe functional impairment. It is of paramount importance to determine if injury is severe enough to prevent spontaneous recovery. This assessment, to date mainly by protracted clinical monitoring supplemented by electrophysiology, indicates either surgical nerve repair or conservative treatment. The significance of imaging so far, either by MRI or CT-myelography, is limited to identify intradural avulsions. However, nerve injury may involve distal elements, which so far elude evaluation. It is desirable to improve detection of distal level injuries to correctly indicate surgery and identify sites for nerve reconstruction. Obviously, undetected lesions either proximal or distal to the site of surgical reconstruction prevent successful nerve regeneration. Methods:  MR Neurography (MRN) was employed in brachial plexus (BPI, N = 10) and lumbar plexus injury (LPI, N = 5) with standard surface array coils combined with an adjustable multi-channel surface array coil (CPC, NORAS). Intradural rootlets were imaged by T2-w 3DTrueFISP (isotrop-vox.: 0,4 mm, TR/TE 8,1/3,6 ms), extradural distal plexus elements with PD-w 3DSPACE (isotrop-vox.: 0,6 mm, TR/TE 1170/33 ms) and sagittally oriented TIRM sequences (vox.: 3 × 0,8 × 0,8 mm, TR/TE/TI 7170, 53/180 ms). Structured reports of T2-w intraneural lesions in-continuity and neuroma formation or avulsion injury was made on anatomical drawings by consensus rating between MP and PB both blinded clinical data (e.g. Fig. 3). MRN findings were finally compared with follow-up and, when available, with intraoperative findings. Results:  Distal lesions were detected by MRN in 7/15 of all patients: 5/10 BPI (2 lesions in continuity, neuroma in 3); 2/5 LPI (neuroma in 2). Intradural root avulsions were identified in 6/10 BPI and 5/5 LPI patients. Combined intradural and distal injury was present in 4/15 patients (e.g. Figs. 1, 2). Prediction of sufficient spontaneous recovery based on clinical follow-up or, when available, of the intraoperative lesion pattern was correct in 14/15 patients. A single case eluding correct detection of the full lesion pattern by MRN was surgically confirmed as an axillary nerve neuroma in the far lateral axillary fossa not covered by MRN. Conclusion: The full pattern of BPI/LPI including distal lesions can be reliably identified by single session multilevel MRN. MRN is valuable to support early decision for either surgical or conservative treatment.

194 Fig. 1

Abstracts Fig. 3 Cords

Divisions

Trunks

Roots C5

Dorsal scapular nerve Suprascapular nerve

C6 Nerve to subclavius

C7

rd Co La te ra

Lateral pectoral nerve

rd

Musculocutaneous nerve Axillary nerve

Median nerve

C8

l

Po

r rio ste al

di Me

Co

T1

rd

Long thoracic nerve

Co

Upper subscapular nerve Medial pectoral nerve Thoraco dorsal nerve Lower subscapular nerve

Ulnar nerve

Radial nerve Medial cutaneous nerve of the arm Medial cutaneous nerve of the forearm

Fig. 2

V1431 MR Neurography reveals accumulating intraneural microlesions underlying peripheral neuropathy of Neurofibromatosis type II M. Pham, V.-F. Mautner, P. Bäumer, T. Kästel, M. Schuhmann, M. Bendszus (Heidelberg, Tübingen) Recent reports mainly based on clinical and electrophysiological findings have raised increasing awareness of peripheral neuropathy (PN) in Neurofibromatosis type 2 (NFII) as a prevalent condition and important cause of morbidity. However, there has been no pathomorphological in-vivo evidence of underlying lesions in clinically affected peripheral nerves. Methods:  High-resolution MR Neurography (MRN) was performed in 6 patients confirmed with NFII and 10 healthy control subjects employing dedicated surface array coils at 3 Tesla (Magnetom VERI O, Siemens AG). An extended imaging protocol covered, according to the main focus of clinical involvement, either both upper (N = 2) or lower extremities (N = 4). The nerval plexus and peripheral nerves along the extremities including the wrist or ankle region were investigated. High-resolution MRN sequences were of the T2-w TSE type with spectral fat saturation. An in-plane resolution between 0.8 (plexus) to 0.2 mm (wrist or ankle) was achieved. Lesion count was done by consensus rating between two neuroradiologists. Results:  Non-tumorous intraneural T2-w hyperintense microlesions (Figs. 1–2) were revealed by MRN in all patients with NFII associated PN but none of the control subjects (Fig. 3).

Abstracts Fig. 1

195 Fig. 3

Fig. 2

Lesion count along clinically affected peripheral nerves revealed a tendency towards higher lesion burden as compared to asymptomatic peripheral nerves in patients, which, however was not significant on statistical analysis (p = 0.1). In 5 of 6 patients, in addition, macroscopic nerve sheath tumors (Schwannomas of NFII) were found associated with the foraminal spinal nerves, however at levels not related to the clinical syndrome. Conclusion:  We present strong evidence that NF-2 associated peripheral neuropathy is caused by the cumulative effect of intraneural microlesions, which for the first time could be visualized in clinically affected peripheral nerves by MR Neurography. V1432 Cardiac-gated phase-contrast magnetic resonance imaging in the diagnosis of syringomyelia U. M. Mauer, A. Gottschalk, B. Danz, U. Kunz (Ulm) Introduction:  Syringomyelia is caused by a disturbance of cerebral spinal fluid (CSF) flow. Causal treatment requires that the exact location of this disturbance be identified by imaging. Cardiac-gated phase-contrast magnetic resonance imaging (MRI) is an effective tool for this purpose. Material and methods:  From 2001 to 2009, cardiac-gated phase-contrast MRI of the skull and the spine was used to detect the site of CSF obstruction in 693 patients with syringomyelia (417 women, 276 men; mean age: 38.5 years,

196

range: 1–79 years). CSF flow is demonstrated in the craniocaudal direction in a strictly median sagittal plane. Results:  Neither MRI nor a comprehensive medical history revealed the cause of syringomyelia in 188 patients. Chiari malformation was the most co mmon cause of syrinx formation (154 patients). Intradural adhesions in the region of the spine, which were detected and localised only by dynamic MRI, were found in 148 patients. For a variety of reasons, we did not obtain interpretable phase-contrast MRI scans in 11% of the cases. Conclusions:  In a number of patients, the cause of syringomyelia cannot be identified without cardiac-gated phasecontrast MRI. In these cases, causal treatment cannot be given if surgery is needed. V1433 Anatomy and pathology of the lumbar spine with 7 tesla magnet resonance imaging A. E. Grams, O. Kraff, L. Umutlu, S. Maderwald, P. Da mmann, M. E. Ladd, M. Forsting, E. R. Gizewski (Gießen, Essen) Question:  Aim of this study was to demonstrate feasibility of lumbar spinal imaging with 7 Tesla MRI by evaluating anatomy and complex spinal malformations. Different sequences, which are established at lower field strengths or assumed to be applicable, were evaluated. Methods:  Five healthy volunteers and a patient with spina bifida and meningocele received MRI examinations of the lumbar spine with a 7 Tesla whole body scanner (Magnetom 7T, Siemens Healthcare, Germany) and a custom built 8 channel transmit-receive spine coil. The imaging protocol included a T2-TSE (TR/TE = 3500/88, TA = 1:50, voxel size = 0.57 × 0.57), a 3D-DESS (TR/TE = 10.75/4.01, TA = 5:39, voxel size = 0.78 × 0.78), a 3D-CISS (TR/ TE = 5.74/2.78, TA = 10:45, voxel size = 1.04 × 1.04) and a 3DVIBE (TR/TE 20/3.06, TA = 12:59, voxel size = 0.57 × 0.57) sequence. Differentiation between the anatomical structures: intervertebral disc-vertebral body, vertebral body-cerebrospinal fluid (CSF) and CSF-spinal cord was evaluated by a measurement of contrast ratios (CR). Furthermore intraforaminal structures, spinal cord, facet joints and the pathologies (spina bifida and meningocele) were evaluated qualitatively with a three level scale (no diagnostic value, moderate visualisation, reliable visualisation). Results:  For the differentiation of intervertebral disc and vertebral body best CRs were received with the VIBE (0.52) and the CISS (0.51). In the discrimination between vertebral body and CSF best CRs were provided by the T2-TSE (0.91) and the CISS (0.86) (in all sequences CRs were  >  0.5). CSF and spinal cord could be distinguished best from each other with the T2-TSE (CR 0.56). Intraforaminal structures could be visualized moderately with the VIBE and the DESS, the other sequences had no diagnostic value. No sequence was

Abstracts

able to depict different structures within the spinal cord. Facet joints were seen reliably in VIBE, CISS and DESS, whereas the T2 had no diagnostic value due to decreased signal homogeneity in the lateral parts. Bony malformations could be visualized reliably with the VIBE and the DESS and only moderate with the CISS and the T2, the meningocele could be visualized moderately with the CISS and had no diagnostic value in the others. Conclusions:  These results indicate that the healthy lumbar spine, as well as complex spinal malformations can be visualized with 7 Tesla MRI. However, this study provides an first basis for further clinical applications of 7 Tesla MRI. Functional imaging/Varia V1435 Individual MRI segmentation reveals enlarged auditory cortex in Williams syndrome as a neural substrate of training-independent musicality M. Wengenroth, M. Blatow, M. Bendszus, P. Schneider (Heidelberg) Background:  Professional musicians exhibit up to twice the volume of the auditory cortex as compared to non-musicians. There is an ongoing controversy in the neuroscience co mmunity as to whether such volume differences are attributable to training-induced neuroplastic mechanisms or rather mirror an innate disposition (i.e. musicality). Methodology/principal findings: To elucidate the potential contribution of genetic factors to the auditory profile, we investigated individuals with the rare genetic disorder Williams-Beuren syndrome (WS) who are known for their strong affinity to music and sounds. Since musical education in this group was negligible due to psychomotor constraints and control subjects were carefully matched for this parameter, we hypothesized that any changes compared to the control group would reflect the contribution of genetic factors to auditory processing and musicality. High-resolution structural T1-weighted MR images were obtained at 3 Tesla. Images were corrected for inhomogeneity, transformed into anterior co mmissure-posterior co mmissure plane (ACPC) and subsequently normalized in Talairach space (TAL). To account for the high inter-individual variability of peripheral cortical structures, we employed an individual analysis method of the auditory cortex by semi-automated segmentation. In keeping with previous volumetric studies the WS group exhibited reduced total brain volume. On the other hand, the AC in WS subjects showed markedly enhanced gyrification and increased gray matter volume. After adjusting for total brain volume reduction, HG volume was 2.2-

Abstracts

fold increased in the left and 1.2-fold in the right hemisphere in WS as compared to control subjects. Moreover, despite their negligible musical education the absolute size of left auditory cortex in these subjects reaches that of professional musicians. Conclusions/significance: These results not only unravel the neural substrate for a particular auditory phenotype but in addition have direct implications for the neuroplasticity debate (“nature vs. nurture”). We therefore propose WS as a unique genetic model for training-independent auditory system properties. V1436 Novel fMRI paradigm for emotion processing in neuropsychiatric disorders M. Wengenroth, J. Reinhardt, C. Stippich, K. Budischewski, F. G. Brecht, M. Bendszus, M. Blatow (Heidelberg; Basel, CH) Background:  Functional magnetic resonance imaging (fMRI) is meanwhile a well-established tool within the clinical routine for patients with brain tumors or epilepsies prior to neurosurgery. However, due to methodological and interpretational constraints fMRI is not routinely applied in patients with neuropsychiatric or neurodegenerative disorders, such as schizophrenia, depression, autism or dementia. These patients suffer from a vide range of cognitive deficits; one of the most perturbing being emotional impairment. We therefore designed an fMRI paradigm for emotions applicable in neuropsychiatric patients within the clinical set-up. In contrast to previous studies it focuses on emotional experience rather than recognition of emotional content. Methods and results:  Visual stimulation material for adverse and positive emotions was selected and modified from the International Affective Picture System (IAPS, University of Florida). Neutral images were used for the baseline condition. Stimulation material was validated in 100 subjects by psychological assessment of elicited emotional responses. The block-designed fMRI paradigm consisting of 4 subsequent and independent runs of 4 min. comprising all tested emotional qualities was optimized in 25 subjects. Attention control and real-time stimulus rating was probed by touch pad response. System control conditions included visual and somatosensory stimulation. Particular emphasis of the experimental set-up was directed at working with cognitively impaired persons, i.e. training session, permanent interaction possibility and short scanning time. Further, the experiment was designed in a way that sudden termination would not automatically lead to a total loss of acquired data. BrainVoyager® (BrainInnovation, Maastricht, Netherlands) was used for standardized processing and analysis of MRI-data. The paradigm was applied in 30 healthy subjects and reliably led to differential activation of emotion processing networks, including amygdala, hippocampus,

197

anterior cingulate, fusiform gyrus and dorsolateral prefrontal cortex. Ongoing studies with schizophrenic patients show practicability of the protocol and robust activation of the above mentioned networks. Conclusions:  We propose a novel fMRI paradigm for implementation in clinical routine as a diagnostic and therapy monitoring tool in neuropsychiatric patients with emotional impairment V1437 Suszeptibilitätsgewichtete zerebrale MRT-Bildgebung bei Patienten mit M. Wilson D. Fritzsch, D. Weber, P. Günther, M. Philipp, K.-T. Hoffmann (Leipzig) Einleitung: Beim M. Wilson, einer autosomal rezessiven Erkrankung mit biliärer Kupferexkretionshe mmung, kann infolge verstärkter Kupferakkumulation eine vermehrte zerebrale Speicherung von Cu vor allem in den Basalganglien auftreten. Die Differenzierung zwischen normal-niedriger und pathologisch erniedrigter Signalintensität mittels T2-gewichteter Sequenzen ist oft problematisch. Ziel war, herauszufinden, ob mit suszeptibilitätsgewichteter Bildgebung (SWI), eine verbesserte Diskriminierung bei Patienten mit M. Wilson und einem Normalkollektiv möglich ist bzw. in welchen zerebralen Kerngebieten die Veränderungen sehr ausgeprägt sind. Methode:  Untersucht wurden Patienten (n = 9; 2w; mittleres Alter 38J.) mit M. Wilson und eine Kontrollgruppe (n = 9; 6w; mittleres Alter 40J.). 4 der Patienten leiden an einer Pseudoskleroseform, 2 an einem Pseudoparkinsonsyndrom, 2 mit milder Neurologie bei überwiegend hepatischer Verlaufsform. Eine Patientin mit hepatischer Verlaufsform ist neurologisch asymptomatisch. SWI-Sequenzen wurden bei 3 Tesla (TrioTim, Siemens) durchgeführt. Es erfolgte eine ROI-basierte Auswertung der Signalintensitäten der SWI-Bilder (OsiriX Foundation) jeweils bds. in der Substantia nigra, Ncl.ruber, Ncl. caudatus, Globus pallidus, Putamen, Ncl.dentatus. Referenzregion war Splenium des Balkens. Der Gruppenvergleich erfolgte mit dem Mann-Whitney-U-Test. Ergebnisse:  Eine Signalabsenkung ergab sich bei der Patientengruppe im Vergleich zur Kontrollgruppe in der Substantia nigra, im Putamen und Globus pallidus (jeweils beidseits, p < 0,01), eine Tendenz zur Signalabsenkung in den Corpora nuclei caudati (p = 0,01 links; p = 0,07 rechts) und den Nuclei ruber (p = 0,06 links; p = 0,01 rechts). Nicht signifikant waren die Unterschiede im Caput nuclei caudati und Ncl.dentatus (jeweils p > 0,2 beidseits). Die Werte der asymptomatischen Patientin lagen oberhalb der Standardabweichung für die Mittelwerte der symptomatischen Patienten, ausgeno mmen im Ncl.ruber rechts, hier fand sich eine Signalabsenkung und im Ncl.dentatus rechts bei hoher Standardabweichung der Wilsongruppe.

198

Schlussfolgerung:  SWI ermöglicht eine gute Diskrimination von symptomatischen Patienten mit M. Wilson im Vergleich zu normalen Kontrollpersonen v. a. anhand von Veränderungen in der Substantia nigra und im Ncl.lentiformis. Signifikant verminderte Signalintensitäten bei M. Wilson sind am wahrscheinlichsten durch paramagnetische Effekte von Cu erklärbar, die mit SWI besser als mit T2/ T2*-Sequenzen erfassbar sind.

Abstracts Fig. 1

V1438 Post-surgical assessment of thalamic deep-brain stimulation electrode position using Diffusion Tractography M. Beckmann, V. Kumar, L. Läer, R. Krüger, W. Grodd, A. Gharabaghi (Tübingen) Introduction:  Over the past years, thalamic ventral intermediate nucleus (VIM) deep-brain stimulation has successfully been used for patients with Essential Tremor. At the same time, it has been intricate to determine the exact location of intralamic sub-nuclei in surgical planning. Striving to maximize surgical precision, identification of these subregions on the basis of MR-diffusion images and the post-surgical depiction of electrode position would be desirable. Recent segmentation approaches identified individual subregions in the thalamus using probabilistic diffusion tractography and Bayesian inference (Behrens, 2003, Jbabdi, 2009). In this study we evaluate thalamic DBS electrode position in relation to the VIM as identified by Diffusion segmentation in patients treated with bilateral deep brain stimulation for essential tremor Methods: Study was performed using a 3T system (TimTrio, Siemens), 32 channel head coil in 4 essential tremor patients (Age: 50–80 years). Structural images were acquired with 1 mm isotropic resolution with MPRAGE (TR/TE = 2300/3.03 ms). Accordingly, isotropic DTI data were acquired using EPI sequence with 2.5 mm isotropic resolution, 64 directions, 2 averages, 56 slices, b = 1500  s/ mm2. DTI data were motion and eddy current corrected. Cortical regions of interest (ROIs) were delineated and thalamic probabilistic parcellation performed. DBS sugery used medtronic 3389 DBS electrode (length = 10–50 cm, diameter 1.27 mm, Number of electrodes = – 4, electrode length = 1.5 mm, electrode spacing edge to edge = 0.5 mm). Pre and post surgery CT and MR data were acquired. Electrode was delinated on post-CT images. Analysis used FSL and AMIRA. Results:  VIM was consistently identified in all of our four patients. The probabilistic thalamic subregions were fused with post-operative CT and MR. Electrodes were rendered on the parcellated thalamus and could confirm electrode position (see Patient A, Fig. 1).

Conclusion:  Fusion of post-surgical CT images with diffusion-based parcellation of the individual’s thalamus allows for determination of surgical accuracy in the implantation of VIM deep brain electrodes. In the future, a probability-based thalamic parcellation might be applied for surgical planning to determine different target regions within the thalamus. V1439 MR Neurography in ulnar neuropathy at the elbow: a prospective diagnostic study P. Bäumer, T. Dombert, F. Staub, T. Kaestel, M. Bendszus, M. Pham (Heidelberg, Dossenheim) Background: Compressive ulnar neuropathy at the elbow (UNE) is the second most frequent nerve entrapment syndrome after carpal tunnel syndrome (CTS). Unlike in CTS, lesion localization in UNE by clinical symptoms and electrophysiology is difficult, and unsatisfactory outcomes after conservative or surgical treatment are more frequently observed. High-resolution MR Neurography (MRN) is an emerging diagnostic tool that allows determination of severity and precise localization of neuropathic lesions. Novel diagnostic criteria of MRN have never been investigated in UNE. Methods:  We conducted a prospective diagnostic study of MRN in 20 patients with UNE based on standardized clinical and electrophysiological findings, and compared findings with 20 healthy controls. MRN of the ulnar nerve

Abstracts

199 Fig. 2 1.00 CNR 0.75 Sensitivity

from mid-upper arm to mid-forearm level was performed at 3 Tesla (Magnetom VERIO, Siemens) with a standard 8channel extremity array coil (Invivo GmbH) combined with a dedicated surface-array coil (NORAS GmbH). Regions of interest were precisely drawn at the epineurial border zone of the ulnar nerve to measure cross-sectional nerve area (Ø) and T2-w signal intensity (SI) on high-resolution TIRM (TR/TE/TI 6120/66/180 ms, voxel size 0.4 × 0.4 × 3.0 mm). For calculation of contrast-to-noise ratios (CNR) additional ROIs were read out in adjacent muscle and air. Coregistration of read-out parameters was by anatomical referencing to the center of the bony retroepicondylar groove.

area_nerve 0.50

0.25

0.00 0.00

0.50

cnr ROC area: 0.9792 Reference

100 ulnar neuropathy at the elbow (UNE)

90

healthy controls

80 70 60

0.75

1.00

1-Specificity

Fig. 1

CNR (of intraneural T2-w versus muscle)

0.25

area_nerve ROC area: 0.7667

Comparison between both criteria showed that CNR is significantly superior to Ø (p = 0.024). Fig. 3 shows a representative example of ulnar nerve lesion localization to the cubital tunnel (versus wrist). Fig. 3

50 40 30 20 10 0 –20

–10

0

10

20

30

Results:  Along the precise anatomical extension of the cubital tunnel (Fig. 1), the intraneural CNR of the ulnar nerve was found to be significantly higher (p < 0.001) in patients (50.3 ± 14) than in controls (19.8 ± 7.1), as was Ø as the second diagnostic criterion (p <  0.001; 11.9 ± 2.7 mm² versus 9.2 ± 1.7). In ROC curve analysis (Fig. 2) CNR proved to be highly significant, rated with an area-under-the-curve value (AUC) of 0.98 as “excellent”. Ø alone with AUC of 0.77 was likewise significant and rated as a “fair/good” criterion.

Conclusion:  Our findings demonstrate that an intraneural nerve lesion as revealed by T2-w contrast is a highly accurate diagnostic criterion of MRN for the diagnostic confirmation and precise lesion localization of UNE. It may in addition guide targeted decompression by precise lesion localization.

200

V1440 Comparative evaluation of Magnetic Resonance Imaging and Multislice Computed Tomography in the imaging of dental and periodontal structures: a subjective image quality in-vitro study C. Gaudino, R. Csernus, S. Heiland, M. Pham, R. Cosgarea, M. Bendszus, S. Rohde (Heidelberg) Purpose:  In clinical practice imaging of teeth and jaw is co mmonly performed with conventional X-ray or Computed Tomography (CT), providing good image quality of most dental structures. However, detection of pathological processes of the periodontal space is possible only in an advanced stage, when there is already resorption of the alveolar bone. Due to its high soft tissue contrast Magnetic Resonance Imaging (MRI) might represent an alternative in this situation. The aim of this experimental study was to compare the feasibility of high-resolution MRI and multislice CT in detection of dental and periodontal structures in an ex-vivo porcine mandible model. Methods:  Six preparations of porcine mandibles were examined with (1) 3T MRI (Tim-Trio, Siemens) and (2) multislice CT (Somatom Sensation 16, Siemens). MRI was performed with an 8-channel surface-coil (Multifunction-Coil-CPC, NORAS) using 3D-T2-TSE (TR/ TE750/123, SL0.6 mm, DF50%, FOV120 mm) and T1-SE (TR/TE680/16, SL2 mm, DF0%, FOV105 mm) sequences with highest space resolution and tissue contrast optimisation. CT scans were performed with high resolution helical scan protocol (mA80, kV120, SL0.75 mm/0.5  mm, FOV160 mm, WL H70 h). To compare image quality, two observers independently reviewed MR and CT images on multiplanar reconstructions and assessed image quality of different dental structures on a five-point scale (1 = excellent to 5 = no visible). Results:  Teeth roots, pulpa chamber and dentin were imaged accurately with both modalities (mean visibility scores for MR vs CT: 1.3 ± 0.4 vs 1.6 ± 0.6, 1.2 ± 0.4 vs 1.6 ± 0.5, 1.4 ± 0.6 vs 1.8 ± 0.6; n.s.). Periodontal space as well as cortical bone was better depicted by MRI (1.9 ± 0.6 vs 2.8 ± 0.7, 2.2 ± 0.4 vs 4.2 ± 0.5; p < 0.05). MRI could excellently display the lamina dura (1.5 ± 0.7), which separates periodontal space and alveolar bone; this structure could not be detected with CT (mean 5.0 ± 0; p < 0.05). Conclusions:  High-field MRI of the dental apparatus is technically feasible in an experimental setting with adapted surface coil and optimised high-resolution sequences. Compared to the CT it allows excellent depiction of the periodontal space and the lamina dura. Our results encourage further experimental and clinical MRI-studies of the periodontal space, e.g. in inflammatory

Abstracts

disease. In order to verify the feasibility of our MRI-setting in vivo, we are doing first examinations on healthy volunteers and on patients with periodontitis. V1441 Dual energy CT after peri-interventional subarachnoid haemorrhage: a feasibility study C. Brockmann, J. Scharf, I. Nölte, M. Seiz, C. Groden, M. A. Brockmann (Mannheim) Purpose:  Peri-interventional subarachnoid haemorrhage (SAH) due to vessel perforation frequently results in extravasation of contrast agent after intra-arterial injection during digital subtraction angiography. Extravasated contrast agent with high density values above 200 HU in the subarachnoid space outshines the subarachnoid bleeding which expresses considerably lower density values around 60 HU in postinterventional CT (Fig. 1A+B). The goal of our study was to evaluate the feasibility of dual-source dual-energy computed tomography (DE-CT) for detection and estimate of peri-interventional subarachnoid haemorrhage (SAH) by subtraction of the iodine-signal. Methods:  Three patients underwent DE-CT (Siemens Somatom Definition) after suspected peri-interventional SAH. Source data images, iodine maps and virtual non-contrast images were analysed regarding distribution, localization and extent of post-interventional SAH. Results:  Despite a strong hyperdense contrast on source data CT images (Fig. 2A) and a high signal in the iodine map along the subarachnoid spaces (Fig. 2B: red signal), virtual non-contrast images revealed only little post-interventional additional subarachnoid haemorrhage in all cases (Fig. 2C). Conclusion:  DE-CT has potential as a valuable tool in the acute assessment of peri-interventional SAH. DE-CT allows the detection and estimate of SAH in cases where contrast agent cannot be discriminated from blood. Fig. 1

Abstracts Fig. 2

201

Conclusion: Small arterial vessels, as demonstrated by representative ophthalmic artery branches, are clearly better visualized by IV-FP-CTA than MD-CTA. Pathological findings on IV-FP-CTA showed excellent agreement in those cases undergoing follow-up IA-DSA.

V1442 Intravenous flat-panel volume CT Angiography for the cerebral arterial vasculature M. Pham, S. Heiland, M. Hartmann, M. Bendszus (Heidelberg) Spatial resolution of latest multi-slice CT angiography (MDCTA) units at about 0,4–0,7 mm (X-Y plane) is still clearly inferior to intraarterial digital subtraction angiography (IA-DSA). High-resolution intravenous CTA (0,1–0,2 mm) has become possible with the advent of flat-panel biplane angiographic C-arm systems (IV-FP-CTA). Its employment in patients was described in few early case reports. Here, IVFP-CTA due to its superior spatial resolution was indicated in selected patients. Anatomical vessel resolution was analyzed for selected small arteries (branches of opht­halmic a.). Methods:  Indications for IV-FP-CTA (AXIOM artis, Dyna-CT, Siemens) were intracranial arterial stenosis (N = 5), aneurysm detection (N = 3) and follow-up after clipping (N = 5). Angular range of acquisition was 200° at increments of 0.4°/image and in-plane resolution (1 × 1 binning) of 0.1 × 0.1 mm. Acquisition was over 20s and started with a delay of 10s after initiation of contrast application (100ml of iodinated contrast agent, 20ml saline flush, 5ml/s via cubital vein). Quality of arterial visualization was compared to 15 age/sex matched patients undergoing otherwise indicated MD-CTA according a standard clinical protocol (Somatom Sensation 16, Siemens). Representative orbital vessel segments were rated as “clearly definable” (CD), “limited definition” (LD), or “not definable”(ND). Because of their relatively constant diameter at the limit of spatial resolution, according to the anatomical literature, the ophthalmic a. at crossing with optic n. (I), origin of lacrimal a. (II), ant. ethmoidal a. (III), supraorbital a. (IV) were selected. Results:  In IV-FP-CTA, vessel positions II-IV were more frequently rated as CD than in the MD-CTA group (p < 0.01). In all patients of the MD-CTA group at least one position between II–IV was rated as ND, and two positions between II–IV as ND or LD. At vessel position I diagnostic rating scores were similar between both groups (p = 0.38). In 5 patients with IV-FP-CTA, IA-DSA was indicated within 2 months. In this subgroup, diagnostic findings of previous IV-FP-CTA were confirmed with excellent agreement by IA-DSA: degree of stenosis (N = 2), aneurysm morphology/ size (N = 1), residual aneurysm after clipping (N = 2).

V1443 Neuroradiological findings in orbital infections F. Götz, A. Giesemann, T. Lichte, H. Lanfermann, T. Stöver (Hannover, Frankfurt/M.) Objective:  Orbital infection is a serious condition and ophthalmological diagnosis can be difficult. Imaging is a useful tool in delineating the spread of the infla mmatory process. The aim of our study was to correlate preoperative neuroradiological findings and post treatment results in orbital infection. Methods:  During an eight year period all patients admitted to the ENT department with suspected orbital infection were identified. Qualifying criteria for inclusion in the study were positive clinical findings, availability of a preoperative CT scan of the midface and complete documentation of the operative procedure. The original CT studies were retrospectively analysed by two neuroradiologists. Imaging findings and operative reports were documented on a standardised questionnaire and categorised. Sensitivity and specificity of neuroradiological diagnosis was calculated for acute sinusitis, preseptal cellulitis, subperiosteal abscess and intraconal infla mmation. Results:  41 patients with an overall average age of 21 years and 5 months were included in the study. Patient ages ranged from 1 to 84 years with a sex distribution of approximately 2:1 male to female. Moderate swelling of the eye-lids was present in 24 cases and profound swelling in 15 cases. Exophthalmus was found in 28 patients and acute sinusitis in 21 out of the 41 patients. Following the radiographic examination preseptal cellulitis was diagnosed in 1 patient, subperiosteal abscess in 15 and intraconal infla mmation in 22. The preoperative diagnosis of subperiosteal abscess and orbital phlegmone were corroborated after surgery in 7 of the 15 and 14 of the 22 patients giving a specificity of 0.7 in both analyses.

202

Abstracts

Conclusion:  Sinusitis is a co mmon but not obligatory prerequisite in orbital infection, even in a cohort of severely ill patients. In our cases, CT was only moderately helpful in differentiating between intra- and extraconal orbital infection. Studies using MRI for preoperative work-up in patients with orbital infection are pending.

In discrepant cases, the stereotactic-guided biopsy of both MET- and Cho/NAA “hot-spots” led to a 100% sensitivity and specificity. Conclusion:  In patients with suspected anaplastic gliomas, simultaneous acquired PET/MR imaging is feasible and markedly improves the diagnostic efficacy of the targeted biopsies.

Molecular Imaging

V1447 Vergleich von kontrastverstärkter Suszeptibilitätsgewichteter und T1-gewichteter MR-Bildgebung bei Patienten mit neu diagnostiziertem Glioblastom A. Radbruch, P. Bäumer, B. Wiestler, T. Boppel, N. Dörner, A. Bartsch, A. Wick, W. Wick, S. Heiland, M. Bendszus (Heidelberg)

V1446 Metabolic mapping of human gliomas: assessment with simultaneous PET/MR imaging for preoperative “hotspot” imaging of suspected anaplastic gliomas S. Bisdas, T. Nägele, R. Ritz, A. Kolb, B. Pichler, U. Ernemann (Tübingen) Purpose:  To evaluate the simultaneous PET/MR imaging using a newly launched integrated PET/MR system for metabolic mapping in patients with suspected anaplastic gliomas. Materials and methods:  Simultaneous acquisition of 11CMethionine(MET)-PET,  T1-weighted-MR-perfusion and 1H-MR-spectroscopy (3-dimensional chemical shift imaging technique, echo time: 135 ms) exams were performed in 18 consecutive patients with newly diagnosed intracerebral lesions indistinctive of being low-grade or anaplastic gliomas. The integrated PET/MR system consisted of a MR-compatible PET scanner (1.25  mm axial spatial resolution) mounted in a 3T MR unit. Lesion to brain ratios of MET uptake greater than 2 were considered positive (post-processed by VINCI software). The perfusion images were analysed using co mmercially-available software (Tissue 4D, Siemens Medical Systems), based on a modified Tofts-Kermode model with arterial input function, for the calculation of the transfer constant (Ktrans), which was supposed to be positive (Ktrans > 0) in anaplastic gliomas. Metabolic maps of Choline/N-acetyle-aspartate (Cho/NAA) ratios  > 3 were considered suggestive of anaplastic gliomas. The tumor “hot-spots” of each modality were correlated with the histological analysis of the subsequent stereotactic guided biopsies. Results:  All patients underwent the integrated PET/MR imaging without complications. All acquired images were of diagnostic quality. The MET-PET images demonstrated 24 positive lesions, the Ktrans maps indicated 14 tumor sites with abnormal permeability, and the Cho/NAA maps showed 26 localizations with anaplastic characteristics. A spatial discrepancy between MET and Cho/NAA “hotspots” was evident in 10 patients. There was no spatial discrepancy regarding positive findings between MET and perfusion maps. Sensitivity/specificity (%) of MET, perfusion and spectroscopy were 75/100, 70/100, and 85/84.

Fragestellung: Susceptibility-Weighted-Imaging (SWI) ist eine MR-Sequenz, die in den letzten Jahren vermehrt Eingang in die klinische Routine gefunden hat. In der Tumorbildgebung wird die SWI hauptsächlich zur Darstellung von Mikroblutungen (punktförmige Signalabsenkungen, pSAs) verwendet. Ziel der Studie ist, a) die Visualisierung einer Kontrastmittelaufnahme in der SWI, b) die Darstellung der pSAs bei Glioblastomen und c) der Vergleich beider Parameter mit einer standardmäßig verwendeten kontrastmittelverstärkten T1-Sequenz (T1-KM). Methode:  Wir untersuchten 12 Patienten mit neu diagnostiziertem, histologisch gesichertem Glioblastom präoperativ mit einer kontrastverstärkten SWI-Sequenz (SWI-KM) (TR: 26; TE: 19,2; Flipwinkel: 15 Grad) und einer 3D-T1-Sequenz (MPRAGE TI 1100; TR 1710; TE 4,0; TR 1710; Flipwinkel 15 Grad) an einem 3-Tesla Gerät (TIM Trio). Das maximale Ausmaß der Anreicherung in der SWI-KM und in der T1KM wurde jeweils auf der Schicht der größten Ausdehnung gemessen und die pSAs numerisch evaluiert. Ferner wurde qualitativ bewertet, ob die in der SWI-KM sichtbaren pSAs Korrelate in der T1-KM haben. Die Auswertungen erfolgten geblindet von zwei unabhängigen Neuroradiologen. Ergebnisse:  Als bislang noch nicht beschriebenes Phänomen zeigten alle 12 Patienten in der SWI-KM ein deutliches Enhancement, welches in der räumlichen Ausdehnung zu 91  ±  4 % dem Enhancement der T1-KM entsprach. pSAs in der SWI-KM konnten bei 7 Patienten nachgewiesen werden. Es konnte eine signifikante Korrelation zwischen Tumorgröße und Anzahl der pSAs gezeigt werden (R2 = 0,8). Sofern pSAs in der SWI-KM nachgewiesen wurden, konnte auf den T1-KM Bildern in 77  ±  10 % kein Korrelat gefunden werden, wobei als Korrelate Enhancement und Signalabsenkungen gewertet wurden. Schlussfolgerung:  Hier konnte erstmals gezeigt werden, dass mit der SWI-KM die Kontrastmittelaufnahme als auch pSAs als Kriterien für Malignität kombiniert visuali-

Abstracts

perf(max/norm)

perf(tu/norm)

perf(tu/norm)

ch0(tu/norm)

.820(**) .000

rho sig

ch0(max/norm)

ch0(perf/norm)

.272 .108

.223 .190

.228 .181

.236 .165

.196 .252

.182 .288

Group –.396(*) .017

ch0(perf/norm)

ch0(max/norm)

ch0(tu/norm)

perf(max/norm)

3 rho sig

.820(**) .000

.272 .108

rho sig

.223 .190

.196 .252

.941(**) .000

rho sig

.228 .181

.182 .288

.768(**) .000

5

rho sig

.236 .165

–.396(*) .017

4

5 –.321 .056

–.161 .350

.941(**) .768(**) .000 .000

rho sig

Group

Aim:  First-pass contrast-enhanced dynamic perfusion imaging (DSC MRI) and MR spectroscopic imaging (MRSI) are highly accurate to classify the malignancy even in heterogeneous gliomas. Parameter maps of the regional cerebral blood volume (rCBV) and of choline (Cho) provide the maximal rCBV value and Cho concentration of glioma tissue (rCBV_max, Cho_max) which presumably represent the most malignant tumor area. This study evaluates whether the area with the rCBV_max matches that with Cho_max and whether rCBV positively correlates with Cho (rCBV_max vs. Cho_max and Cho_tu vs. rCBV_tu). Methods:  46 patients with brain tumours highly suspect for primary or recurrent WHO grade III or IV gliomas were examined with 2D MRSI and DSC MRI at 3 Tesla. Of these, 36 remained in the study following histological confirmation (n = 30) or MRI and clinical findings suggestive of primary or recurrent GBM (n = 6). Voxel size of rCBV maps was adapted to the size of spectroscopic voxels and rCBV images were anatomically coregistered with the spectroscopic Cho maps. Within the tumor we analyzed: rCBV_tu, rCBV_max, Cho_tu, Cho_max and Cho_perf (where Cho_perf is the Cho concentration within the rCBV_max voxel); all values were normalized to the values of the contralateral hemisphere. The correlation between these parameters was analyzed with Spearman’s rho test while a binomial test was performed to check whether Cho_max = Cho_perf. Patients with other histological diagnoses (metastases, lymphomas, low-grade gliomas, therapy-induced changes) and patients without a focal rCBV increase were excluded from the analysis. Results:  In 32 of the 36 patients the area with the rCBVmax did not match the area with Cho_max (Cho_perf < Cho_max, p < 0.001). However, Cho_perf was highly correlated with with Cho_max (r = 0.78, p < 0.001; Fig.1 and Fig.2). The Cho_tu and rCBV_tu were only weakly correlated (r = 0.27, p < 0.11; Fig. 1), whereas Cho_max and Cho_perf were

Fig. 1

–.054 .755

.779(**) .000

.038 .825

.779(**) .000

–.321 .056

–.054 .755

–.161 .350

.038 .825

4 3

Fig. 2 GBM

Gliom III

recurrent GBM

10 cho(max/norm)

V1448 The area of maximal regional blood volume in gliomas WHO grade III and IV, does it fit with highest choline? E. Hattingen, M. Wagner, R. Nafe, A. Jurcoane, F. Zanella (Frankfurt/M.)

not correlated with rCBV_max (r = 0.19; p = 0.25; r = 0.18, p = 0.28; Fig. 1 and Fig.3). Cho_max was highly correlated with Cho_tu (r00.94, p = 0.001) and rCBV_max with rCBV_ tu (r = 0.82, p < 0.001). Conclusion:  In gliomas WHO grade III and IV the voxel with the maximal rCBV differs from the voxel with the maximal Cho. Further, different values of rCBV and Cho are not significantly correlated within the glioma tissue. Therefore it remains to evaluate, which parameter provides the most accurate spatial information about the most malignant tumor area.

8 6 4 2 0 0

1

2

3

4

5

0

1

2

3

4

cho(perf/norm)

cho(perf/norm)

Gliom III

GBM

5 0

1

2

3

4

5

cho(perf/norm)

Fig. 3 recurrent GBM

3 cho(tu/norm)

siert werden können. Die hohe Anzahl der in der SWI-KM sichtbaren pSAs, welche sich in der T1-KM nicht darstellten, lässt eine Anwendung der SWI-KM z. B. zur Biopsieplanung sinnvoll erscheinen. Aufgrund der weitgehenden räumlichen Übereinsti mmung des Enhancements in SWIKM und T1-KM wären hier keine Einbußen in der Identifikation der Tumorgrenzen zu erwarten. Hinsichtlich der unterschiedlichen Intensität des Enhancements erscheint ine detailliertere quantitative Untersuchung mittels Phantomstudie und Verdünnungsreihe erforderlich.

203

2 2 2 1 0 0 0

1

2

3

4

perf(tu/norm)

5

0

1

2

3

4

perf(tu/norm)

5 0

1

2

3

4

perf(tu/norm)

5

204

Abstracts

V1449 Gadofluorine-M versus Gd-DTPA im Tumormodell der Maus L. Jestaedt, M. Weiler, D. Lemke, S. Heiland, W. Wick, M. Bendszus (Heidelberg)

V1450 Darstellung von mit Eisennanopartikeln markierten Stammzellen im Gelphantom mit suszeptibilitätsgewichteten und T2*-gewichteten Sequenzen im 3Tesla-MRT D. Lobsien, A. Dreyer, J. Boltze, K.-T. Hoffmann (Leipzig)

Einleitung:  Gadofluorine-M (GfM) ist ein neues Kontrastmittel (KM) in der Magnetresonanztomographie (MRT), das aus einem fluorierten Gadolinium(Gd)-Komplex besteht und durch amphiphile Eigenschaften Nanomizellen in wässrigen Lösungen bildet. In früheren Studien wurde eine hohe Bindungsaffinität an die pathologisch veränderte Extrazellularmatrix gezeigt. Ziel dieser Studie war der Vergleich von Gd-DTPA und GfM im Tumormodell. Methoden:  In n = 21 weiblichen VM/dK-Mäusen wurden Maus-Astrocytome durch stereotaktische Injektion von je 5,000 SMA-560 Zellen in die rechte Großhirnhemisphäre induziert. Das Monitoring des Tumorwachtums erfolgte jeweils mit T1-w und T2-w Sequenzen an einem 2,3 Tesla Bruker Scanner. In Gruppe I (n = 9 Tiere) erfolgte am 8. postoperativen Tag (POD) die erste MRT-Untersuchung (5 min. nach Gd-DTPA i. v., 0,2  mmol/kg Körpergewicht (KG)). Direkt nach der Untersuchung wurde GfM i. v. injiziert (0,1  mmol/kg KG) und 24 h später (9. POD) folgte eine weitere MRT-Untersuchung. Weitere MRT-Kontrollen wurden ohne erneute KM-Injektion am 15. und 20. POD durchgeführt. In Gruppe II (n = 9 Tiere) wurden MRT-Untersuchungen am 8. und 15. POD durchgeführt (5 min. nach Gd-DTPA i. v., 0,2  mmol/kg KG). Direkt nach der Untersuchung am 15. POD erfolgte die i. v. Injektion von GfM (0,1  mmol/kg KG) und 24 h später (16. POD) die MRTKontrolle. Eine weitere MRT-Kontrolle wurde am 20. POD ohne weitere KM-Injektion durchgeführt. Gruppe III (n = 3 Tiere) diente als Kontrollgruppe. Aus allen 3 Gruppen wurde je ein Tier am 9., 16. und 20. POD zur Histologie aus dem Versuch geno mmen. Ergebnisse:  In Gruppe I zeigte GfM eine weit größere Tumorausdehnung als das T2-w Bild bzw. Gd-DTPA. Die GfM-Aufnahme im Tumor persistierte auch zu späteren Zeitpunkten im Tumor, obwohl das Kontrastmittel nicht mehr zirkulierte. Auch in Gruppe II zeigte sich eine weit größere Tumorausdehnung mit GfM als mit T2-w Sequenzen bzw. Gd-DTPA. Histologisch entsprach die Tumoranfärbung mit GfM (Fluoreszenz-gelabelt) vitalem Tumorgewebe. Schlussfolgerung:  GfM bildet die reale Gliominfiiltration präzisier ab als T2-w Sequenzen bzw. Gd-DTPA. Hierbei scheint GfM dauerhaft an die Extrazellularmatrix von Tumorzellen zu binden und verteilt sich mit dem Tumorwachstum vorwiegend an die Tumorperipherie. Das neue KM könnte im klinischen Einsatz eine weitaus präzisere Darstellung der Tumorausdehnung ermöglichen. Durch die Persistenz an Tumorzellen wäre es auch ein geeignetes KM für die intraoperative MR-Bildgebung.

Einleitung:  Wir untersuchten die Grenze der Detektierbarkeit von mit Eisennanopartikeln (VSOP) markierten Sta mmzellen in vitro in einem klinischen 3T MRT mit suszeptibilitätsgewichteten Sequenzen (SWI) und T2*-Sequenzen. Material und Methoden:  Die Messungen wurden an 4 Agarose-Gelphantomen durchgeführt. Die Phantome hatten jeweils 3 Schichten (Schichtfläche 5,46 cm2) VSOP markierter oviner Sta mmzellen mit 0, 100, 500 Zellen bzw. 1000, 10000, 100000 Zellen. Untersucht wurde bei 3Tesla mit 6 SWI- bzw. T2*-gewichteten Sequenzen in 3 verschiedenen Spulen. 8 Kanal Spule: T2 FLASH 3D SWI, TR 60, TE 20, 3 Voxelgrößen (0,37 × 0,31 × 0,15 bzw. 0,6 bzw. 1,2 mm), Messzeiten von ca. 4 h bzw. 50 min. bzw. 21 min. 12 Kanal Spule: T2 FLASH 3D SWI, TR: 40, TE: 20, Voxelgröße 0,39 × 0,31 × 0,7 mm, Messzeit 32 min. T2* FLASH 3D, TR: 6,2, TE: 20, Voxelgröße 0,39 × 0,31 × 0,70 mm, Messzeit ca. 9h. Ringspule (4 cm Spulendurchmesser): T2* FLASH 3D, TR 620, TE 20 und T2 FLASH 3D SWI, TR 60, TE 20 mit Voxelgröße jeweils 0,24 x 0,2 x 0,4  mm, Messdauer ca 8,5h bzw. 4,8 h. Die Bilder wurden von 4 Neuroradiologen geblindet analysiert und in einer vierstufigen Skala bewertet: 1 sicher nachweisbar, 2 unsicher nachweisbar, 3 unsicher nicht nachweisbar, 4 sicher nicht nachweisbar. Bei einem sich daraus ergebenden Mittelwert der Bewerter (MB) von 1 bzw. 1,25 wurde die Zellmenge als sicher nachweisbar bzw. bei einem MB von 3,75 bzw. 4 als sicher nicht nachweisbar gewertet. Ergebnisse:  Die gruppierte Interraterreliabilität lag bei Κ 0,4–0,7. 100000 Zellen konnten mit allen Sequenzen sicher nachgewiesen werden (MB = 1). 10000 Zellen konnten in der 8 Kanalspule mit SWI Sequenzen bei 0,6 mm und 0,15 mm Schichtdicke, in der 12 Kanalspule mit T2* sicher nachgewiesen werden. Bei 1000 Zellen konnte eine sichere Nachweisbarkeit in der 8 Kanalspule mit SWI bei 0,15 mm Schichtdicke, mit den T2* Sequenzen in der Ringspule sowie in der 12 Kanal Spule erreicht werden. 500 Zellen konnten lediglich in der 8 Kanalspule mit SWI und 0,15 mm Schichtdicke nachgewiesen werden (MB = 1,25), allerdings zeigte sich in dieser Sequenz falsch positiv eine sichere Nachweisbarkeit bei 0 Zellen. 100 Zellen ließen sich in keiner Sequenz sicher nachweisen. Null Zellen konnten in keiner Sequenz sicher ausgeschlossen werden. Fazit:  Mit SWI- als auch T2*-Sequenzen können 1000 VSOP markierte Sta mmzellen bei 3T sicher nachgewiesen werden, in einem Untersuchungsaufbau, der sich auch für größere Tiermodelle eignet. Ab 500 Zellen scheint die

Abstracts

Fragestellung:  Eine Erhöhung der inspiratorischen Sauerstofffraktion führt zu einer Erhöhung des Partialdruckes des gelösten Sauerstoffs (pO2) im Plasma. Ein potentieller Anstieg des Verhältnisses von Oxyhämoglobin (OxyHb) zu Deoxyhämoglobin (DeoxyHb) im Hirnparenchym ist bei Gesunden aufgrund der Sauerstoffbindungskurve v. a. im venösen Schenkel zu erwarten und wird möglicherweise durch eine Absenkung des zerebralen Blutflusses (CBF) verhindert. Ziel unserer Untersuchung war die Messung dieser Effekte mittels gepulstem arteriellen Spin Labeling (PASL) und Quantifizierung von T2’ (T2-unabhängiger Anteil der effektiven transversalen Relaxationszeit T2*). Methoden: T2’- und PASL-Messungen wurden an 12 gesunden Probanden (20–34 Jahre, kein kardiovaskuläres Risikoprofil) bei 3.0 T unter Ruhebedingungen durchgeführt, zunächst unter Atmung von Raumluft, 8 min. später während Maskenbeatmung mit 10l 100%-igem O2. Die T2’- Daten wurden auf makroskopische Magnetfeldinhomogenitäten und mit einem neu entwickelten Algorithmus auf Bewegung korrigiert. Aus der PASL- Messung wurden CBF- Karten berechnet und auf eine O2-bedingte Verkürzung der T1- Zeit korrigiert. Die Auswertung der koregistrierten und bewegungskorrigierten T2’- Daten erfolgte wie die der CBF- Karten mittels manuell gezeichneter regions of interest (ROI) im paramedianen frontoparietalen Kortex (T2*, T2’, PASL) und im Thalamus (T2*, T2’). Die statistische Auswertung erfolgte mittels Wilcoxon-Rangsu mmentest. Ergebnisse:  Unter Maskenbeatmung zeigten sich im intraindividuellen Vergleich zu den Werten bei Raumluftatmung im paramedianen frontoparietalen Kortex sowohl ein signifikanter Abfall der T2*- und der T2’-Werte (Mittelwert  ±  Standardabweichung: T2’ ohne O2 199.6 ± 72.6 ms; T2’ mit O2 156.0 ± 29.1 ms; p = 0.04) als auch eine signifikante Reduktion des CBF (CBF ohne O2 61.2 ± 9.6 ml/100 g/min; CBF mit O2 56.2 ± 8.5 ml/100 g/min; p = 0.01) (siehe jeweils Abb. 1). Schlussfolgerung:  Die Kombination von PASL als nichtinvasiver Methode zur Absolutquantifizierung des CBF und der quantitativen T2’- Messung ist vielversprechend zur Evaluation hämodynamischer Effekte auf die zerebrale Sauerstoffsättigung. Unsere Ergebnisse zeigen, dass bei Gesunden unter Ruhebedingungen eine O2- Maskenbeatmung keinen Nettoanstieg des Verhältnisses von OxyHb zu DeoxyHb im zerebra-

Abb. 1 ROI Thalamus

ROI Cortex

ROI PASL

ms 225

p = 0.04

200 T2' (mean +/– SE)

V1451 PASL- und T2’-Bildgebung bei 3 Tesla: Einfluss der O2- Beatmung auf die zerebrale Autoregulation M. Wagner, J. Magerkurth, S. Volz, A. Jurcoane, O. C. Singer, T. Neumann-Haefelin, R. Deichmann, E. Hattingen (Frankfurt/M., Fulda)

len Kortex bewirkt, sondern sogar ein gegenläufiger Trend zu sehen ist. Dieser Effekt ist am ehesten auf eine reaktive Erniedrigung des CBF zurückzuführen.

175

150

125

100 no O2

O2

Thalamus

no O2

O2 Cortex

ml/100g/min p = 0.01

65

PASL – CBF (mean +/– SE)

sichere Nachweisbarkeit durch zunehmenden Einfluss von Artefakten beeinträchtigt zu werden.

205

60

55

50

45 no O2

O2 (corrected)

O2 (uncorrected)

V1452 Suszeptibilitätsgewichtete Bildgebung zerebraler Kerngebiete im 3Tesla-MRT bei Patienten mit idiopathischem Parkinson Syndrom D. Fritzsch, M. Oberbeck, D. Lobsien, J. Schwarz, K.-T. Hoffmann (Leipzig) Ziel:  Motivation der Untersuchung war es herauszufinden, inwieweit mittels suszeptibilitätsgewichteter Bildgebung (SWI) Signalunterschiede in Basalganglien, Thalamus und Mittelhirn bei idiopathischem Parkinson-Syndrom (IPS) im Vergleich zu einer Kontrollgruppe nachweisbar sind.

206

Abstracts

Stroke/Pediatric-Imaging

Results: Striatal changes were present in all pts. with AEC and insidious-onset of dystonia, but the pattern of putaminal change differed: In 4/4 pts. with insidious-onset only the dorsolateral putamen was affected in contrast to 1/13 pts. after AEC. With AEC after insidious-onset, diffuse pallidal changes were superimposed on initial dorsolateral changes. Putaminal changes persisted on follow-up MRI in pts. with AEC or insidious-onset, indicating a lesional pattern consistent with irreversible clinical deficits. Conversely, the putamen was normal in all asymptomatic and late-onset pts. Pallidal changes were present in 20/43 pts., after AEC in 13 and “isolated” in 7 pts. In contrast to a lesional pattern after AEC, normalization was observed in a subgroup of prospectively followed asymptomatic patients. As shown previously, brain maturation was co mmonly delayed, but ultimately complete. Frontotemporal hypoplasia was present at birth and could resolve with therapy. Frequency and extent of white matter changes (26/43) increased with age. Conclusions:  In GA1 different (time) patterns exist: Lesional changes occur with AEC and with insidious-onset dystonia, the latter preferentially affecting the dorsolateral putamen. Pallidal changes in asymptomatic patients appear to be transient. Frontotemporal hypoplasia, i mmaturity of gyration and white matter in neonates indicate prenatal onset and these changes may completely resolve. White matter abnormalities progress with age. Lesional, transient, and progressive MR changes coexist, highlighting that changes are not strictly limited to a brief period of vulnerability for AEC and that these changes are not restricted to the striatum.

V1454 Neuroimaging in glutaric aciduria type I—lesional, transient and progressive patterns of MRI changes I. Harting, A. Seitz, E. Neumaier-Probst, C. Mühlhausen, B. Assmann, E. Maier, S. Gabarde, S. Kölker (Heidelberg, Mannheim, Hamburg, Düsseldorf, München)

V1455 Can malignant progression of paediatric solid gliomas be suggested with MRS? A comparison grading between children and adults L. Porto, M. Kieslich, K. Franz, T. Lehrnbecher, F. Zanella, U. Pilatus, E. Hattingen (Frankfurt/M.)

Objective:  Glutaric aciduria type I (GA1) is an inherited neurometabolic disorder. Bilateral striatal lesions develop in the majority of untreated patients following an acute encephalopathic crisis (AEC). We have previously shown that MR changes in GA1 are complex and dynamic and that extrastriatal changes are co mmon in GA1, both with and without AEC (Harting et al. Brain 2010). Here we analyse patterns and age-related MR changes, in particular of basal ganglia, with respect to different clinical presentations. Methods:  82 cranial MRIs of 43 pts. (9 d–66 yrs) were assessed using a standardized protocol. At the time of MRI 13 pts. were symptomatic after AEC, 3 had insidious-onset dystonia, 1 had insidious-onset dystonia and later AEC, and 4 had late-onset disease without motor symptoms.

Objective:  To investigate whether pathologically similar glial tumours in adults and children may also show metabolic similarities in proton magnetic resonance spectroscopy (MRS) and if the ratio of NAA/tCho could differentiate between low and high grade gliomas for the different groups. Material and methods:  49 patients with astrocytomas were evaluated retrospectively using normalized measures of total choline (tCho), N-acetyl-aspartate (NAA) and total creatine (tCr). These metabolites were used to differentiate between WHO II and WHO III astrocytoma in children and adults. Neuropathological grading was performed using WHO criteria. Twelve children (5 WHO II astrocytomas, 7 WHO III astrocytomas) and 37 adults (21 WHO II astrocytomas, 16 WHO III astrocytomas) were included in this study. MRS was performed before treatment. Metabolite concen-

Methodik:  Bei 34 konsekutiv und prospektiv untersuchten Patienten mit IPS (Hoehn und Yahr Stadien I–IV, Seitendominanz re. n = 14, li. n = 10, nicht eindeutig n = 10) und 34 alters- und geschlechtsgepaarten Kontrollen wurden im 3 Tesla-SWI-Bild relative Signalintensitäten (SI) von Putamen, Pallidum, Nc. caudatus, Thalamus, Nc. ruber und Substantia nigra ermittelt. Referenzregionen waren der Liquor im Seitenventrikel-Vorderhorn und das Splenium des Balkens. Die regionalen Mittelwerte der relativen SI der untersuchten Gruppen wurden verglichen. Ergebnisse:  Bei Patienten mit IPS waren die Mittelwerte der relativen SI in fast allen betrachteten Kerngebieten höher als in der Kontrollgruppe (Referenz Liquor und Splenium), wobei ein signifikanter Unterschied nur in der Substation nigra (links deutlicher als rechts). Schlussfolgerung:  Mittels SWI bei 3  Tesla wurden bei Patienten mit IPS Unterschiede in der relativen SI der untersuchten Kerngebiete, insbesondere in der Substania nigra im Vergleich zur Kontrollgruppe nachgewiesen, wobei eine Korrelation mit der Seitendominanz nachweisbar war. Die relative SI war bei Erkrankten höher als bei Gesunden. Als Ursachen werden, ähnlich wie bei anderen Basalganglienerkankungen, einerseits neurodegenerative Veränderungen mit glialen Umbaureaktionen diskutiert. Daneben können pathologische Veränderungen der regionalen Vaskularisation, die mit SWI sehr sensitiv erfasst werden können, ursächlich für die erhobenen Befunde sein.

Abstracts

trations of tCho, NAA and tCr were normalized to contralateral brain tissue. A Mann-Whitney U-Test was performed to evaluate differences within the respective groups. Results:  In both groups, loss of NAA and increase of tCho were more pronounced in WHO III than in WHO II astrocytoma. The best discriminator to differentiate between low and high grade gliomas was found to be the ratio of NAA/ tCho (p < 0.01) for the different groups. Conclusion:  Metabolite ratios were more informative than metabolite changes in differentiating low and high grade gliomas in children as well as in adults. V1456 Frühe zeitliche Dynamik von seriellen T2*-gewichteten MRT-Sequenzen als Prädiktor für Infarktgröße und Penumbra in einem Mausmodell des ischämischen Schlaganfalls A. Eisenbeis, F. Leypoldt, A. Gersing, N. Forkert, T. Magnus, J. Fiehler (Hamburg) Einleitung:  Die MRT-Bildgebung des akuten ischämischen Schlaganfalles hat einen festen Platz in der Schlaganfallbehandlung. Trotz umfangreicher Forschungsarbeiten existieren nur wenige Untersuchungen zu der Wachstumsdynamik eines ischämischen Schlaganfalles in der Frühphase während der ersten Minuten. In Vorarbeiten konnte die hohe Aussagekraft der T2*-gewichteten MRT- Bildgebung gezeigt werden. Diese Arbeit untersucht die zeitliche Dynamik und die räumliche Ausbreitung der akuten T2*Signalveränderungen nach Gefäßverschluss korrelierend zu der Diffusionsläsion nach 30 und 60 Minuten Ischämie. Zudem wird die Größe der endgültigen T2*-Signalveränderung mit der histopathologischen Infarktläsion verglichen. Methoden:  Wir führten bei n = 24 i. p. anästhesierten Mäusen und n = 6 Kontrollen eine temporäre Ischämie der A. cerebri media mittels Fadenmodell für 30 oder 60 Minuten durch. Unmittelbar vor Platzierung des Filaments wurden diese in einem 3T -MRT (Intera Philips, Hamburg) mit folgenden Baseline-Sequenzen untersucht: T2*w, qT2w, DWI inkl. ADC. Zwei Minuten nach Gefäßokklusion wurde die erste dynamische T2*-gewichtete Sequenz akquiriert. Nach 30 oder 60 Minuten T2*w-Messung erfolgte eine Diffusionswichtung. Zwischen 60 bis 90 Minuten nach Reperfusion wurde mittels Vitalfärbung (TTC) 1 mm messender coronarer Hirnschnitte eine Infarktgrößenquantifizierung vorgeno mmen. Ergebnisse:  Von n = 24 Mäusen waren n = 20 auswertbar. Bei diesen zeigten sich folgende Ergebnisse. In der akuten Ischämiephase (2–52 Minuten) nach einem Verschluss der A. cerebri media zeigte sich eine kontinuierliche Zunahme des T2*-Volumens, welches maßgeblich durch eine Zunahme

207

der randständigen hypoxischen T2*-hypointensen Areale bedingt war. Die zentralen, stark oligämischen Infarktareale blieben dahingegen über diesen Zeitraum größenkonstant. Das endgültige T2*-Volumen korreliert mit der mittels Vitalfärbung dargestellten Infarktgröße. Diskussion:  Unsere Ergebnisse zeigen, dass die T2*gewichtete Bildgebung in der akuten Phase einer Ischämie einen guten Prädiktor für die endgültige Infarktgröße darstellt. Diese Methode erlaubt außerdem während einer akuten Ischämie eine Unterscheidung von Infarktkern und Penumbra. Im murinen Schlaganfallmodell ist primär ein kontinuierliches Wachstum der Penumbra zu verzeichnen, während der Infarktkern bereits unmittelbar nach Ischämie determiniert erscheint. Die weitere Erforschung der T2*-Bildgebung erscheint uns auch beim humanen Schlaganfall vielversprechend. V1457 Standardization of dynamic whole brain perfusion CT: a comprehensive database of regional perfusion parameters L. Feyen, H. Seifarth, T. Niederstadt, W. Heindel, A. Ke mmling (Münster) Introduction:  In acute stroke imaging, there is a need for harmonization of quantitative methods to assess brain perfusion. Significantly different regional brain perfusion has been demonstrated using PET and SPECT but not in with dynamic CT perfusion which is widely used for assessing ischemic stroke. The purpose of this study was to present a comprehensive database of quantitative human regional blood flow parameters in healthy elderly using whole brain dynamic perfusion CT. Methods:  In a prospective fashion, CT-perfusion maps from brains without evidence for ischemia were acquired: during a 24 months period, 930 patients received whole brain dynamic CT-perfusion imaging for evaluation of acute stroke. 107 cases were consecutively selected without any perfusion abnormality, infarction or symptoms on follow up or vascular abnormality. CT perfusion maps (cerebral bloodflow [CBF],-volume [CBV], time to peak [TTP] and mean transit time) were calculated and linearly co-registered to standard MNI-152 space (FLIRT5.5). A weighted average perfusion value in a priori brain regions defined by the anatomical probability distribution in standardized brain atlas (Harvard Oxford cortical/subcortical atlas) was calculated for each map. Perfusion differences of brain regions were tested by paired t-test (p < 0.05) with Bonferroni correction. Results: Regional perfusion parameters in a total of 68 defined regions were examined in three normalized perfusion maps (CBV, CBF and TTP) derived from 107 healthy brains (Fig 1). Empirical normal values for perfusion parameters where significantly different for cortex, subcortical

208

grey matter and white matter. The highest cortical grey matter perfusion values were measured in the calcarine gyrus, Heschl’s gyrus and opercular cortex, the lowest in the temporal and frontal cortex. The highest subcortical grey matter perfusion values were found in the Hippocampus, Thalamus and Putamen. Fig. 1

Discussion:  Significantly different CTP derived regional blood flow parameters in cortical and subcortical grey- and white matter were demonstrated in a large series of patients. Our findings are in accord with regional blood flow parameters measured by PET and SPECT. There is evidence that regional rather than whole brain blood flow thresholds may be a more precise measure of predicting the likelihood of infarction. The presented standardized CT-perfusion maps can be used for automated stroke evaluation to improve assessment of regional stroke risk. Fig. 2

V1459 Arterial spin labeling in subacute ischemia using a 32channel head coil at 3.0 Tesla I. Nölte, S. Huck, C. Groden (Mannheim) Purpose:  Arterial spin labeling (ASL) is a promising method to evaluate cerebral perfusion without the need of intravenous contrast agent. However, co mmercially available ASL sequences have not been tested in a clinical setting in comparison to conventional perfusion measurements using intravenous Gd-DTPA. Therefore, the purpose of this study was to compare perfusion measurements with ASL to con-

Abstracts

ventional Gd-DTPA perfusion measurements (PWI) in patients with subacute ischemia using a co mmercially available 32-channel head-coil at 3-Tesla. Materials and methods:  Twelve patients with subacute ischemia were included in the study. ASL (cerebral blood flow) and PWI (time to peak) were performed and evaluation was done by two raters. Image quality and quality of delineation of areas with disturbed perfusion were graded from 0 (not possible) to 4 (very good). The severity of maximal perfusion restriction was graded from 0 (no restriction) to 4 (no perfusion). The area of restricted perfusion was measured in ASL and PWI. For comparison of ordinal data the Wilcoxon signed rank test was used, for scale data the student's t-test was used. Results:  Mean image quality (ASL/PWI) was 1.4/3.0 (p < 0.05), mean delineation quality was 1.5/2.25 (n. s.) and mean perfusion restriction was 2.25/2.25 (n. s.). Evaluation of perfusion restriction in ASL and PWI differed more than 1 grade in more than 25%. The volume of perfusion restriction was higher in PWI than in ASL (p < 0.05). Conclusion:  ASL using a co mmercially available sequence and 32-channel head coil at 3.0 Tesla is feasible but substantial improvement of signal to noise ratio has to be achieved in order to improve image quality and perfusion delineation before ASL can be used in clinical practice. V1460 Die Wertigkeit der Dual-Energy-CT-Angiographie für die Stenosenquantifizierung und Beurteilung der Plaquemorphologie von Carotisabgangsstenosen im Vergleich zur konventionellen Angiographie A. Korn, C. Thomas, B. Bender, S. Danz, M. Fenchel, T.-K. Hauser, T. Nägele, M. Heuschmid, U. Ernemann (Tuebingen) Einleitung:  Die Dual-Energy-CT-Technik (DECTA) erlaubt die Subtraktion von Knochen und Kalzifikationen (KPS). Kalzifikationen erschweren oft die Beurteilung von Carotisabgangsstenosen in der konventionellen CTA im 3D-MIP Modus. Die zeitaufwendige Durchsicht axialer MPR-Dünnschichten gilt demnach noch als Standard zur Beurteilung der supraaortalen Gefäße. Ziel dieser Arbeit war die Beantwortung der Frage, ob KPS mittels DECTA eine bessere Beurteilung von Carotisabgangsstenosen in 3D-MIP Darstellung ermöglicht. Material und Methoden:  36 Patienten mit insgesamt 46 Stenosen an der Karotisbifurkation wurden in diese Studie eingeschlossen. Nach Durchführung einer DECTA wurde bei allen Patienten eine konventionelle Angiographie durchgeführt. MPR und 3D MIP Darstellungen wurden vor und nach KPS berechnet und anschließend die Stenosegrade nach NASCET Kriterien beurteilt. Die Plaquemorphologie wurde eingeteilt in: nicht beurteilbar, glatt, irregulär und tief ulzeriert.

Abstracts

Ergebnisse:  Nur in 21 Fällen konnte der Stenosegrad vor KPS in den 3D-MIP Darstellungen beurteilt werden. Nach KPS und in den axialen MPR-Dünnschichten war eine Beurteilung des Stenosegrades in allen Fällen möglich. Die lineare Korrelation zwischen MPR, 3D-MIP vor und nach KPS mit der konventionellen Angiographie nach y = a * x + c ergab in allen Fällen eine signifikante Korrelation (a = 0.94/0.93/0.92, b = 6.7/7.5/10.5, Pearson r = 0.971/0.986/0.958). Die im Vergleich zur konventionellen Angiographie im Durchschnitt zunehmende Überschätzung des Stenosegrades im MPR- und 3D-MIP-Modus (b: 6.7  <  10.5) spiegelte sich auch in der höheren Rate an Pseudoverschlüssen (MPR = 9, 3DMIP nach KPS = 16) wieder. Die Plaquemorphologie war vor KPS in 14 Fällen nicht beurteilbar, nach KPS in 4 Fällen. Zusammenfassung:  DECTA mit KPS erlaubte die Beurteilung des Stenosegrades in 3D-MIP Darstellung in allen 46 Fällen, während in der konventionellen 3D-MIP eine Beurteilung nur in 21/46 Fällen möglich war. Im Vergleich zur axialen MPR und zur konventionellen Angiographie ko mmt es zu einer zunehmenden Überschätzung des Stenosegrades. Die 3D-MIP Darstellung ermöglicht jedoch eine bessere Beurteilung des Plaquetyps und insbesondere nach KPS eine bessere Beurteilung der Plaquemorphologie. Für die Beurteilung von Carotisabgangsstenosen ist das insofern von Bedeutung, als daß ulzerierende Plaques einen unabhängigen Risikofaktor für das Auftreten eines Schlaganfalls darstellen [1]. Literatur:  [1] Stroke 2000;31:615–21. V1461 Dual energy bone subtraction in computed tomography angiography of extracranial- intracranial bypass: comparison with digital subtraction angiography A. Ke mmling, I. Nölte, C. Groden, S. Diehl (Münster, Mannheim) Objectives:  Computed tomography angiography (CTA) has become an increasingly accepted non-invasive method for the evaluation of extracranial and intracranial vessels including extracranial-intracranial bypass (EC/IC-bypass). For optimized evaluation and presentation of reconstructed images automated dual energy based bone removal techniques have been proposed as other bone removal techniques are time comsuming and require user interaction (threshold method) or entail substantially increased radiation exposure (subtraction of non-enhanced computed tomography images). Therefore the aim of this study was to assess the accuracy of dual energy (DE) CTA images with fully automated DE bone removal (DEBR) for the evaluation of EC/IC bypass in comparison with conventional CTA images and digital subtraction angiography (DSA).

209

Materials and methods:  24 patients underwent cranial DE-CTA (80 kV and 140 kV) and conventional DSA for the evaluation of EC/IC-bypass function postoperatively. From the DE images virtual 120 kV single energy images were reconstructed. Using 5-point scales (0 = poor to 4 = excellent) two examiners rated overall image quality, the quality of bone removal, and the visualization and vessel integrity of bypass for three segments (extracranial, trepanation, intracranial) in CTA images with and without DEBR in comparison to DSA. Additionally, the time required for bone removal was noted. Friedmann's test and post hoc Wilcoxon rank test (p < 0.05) were used for statistical testing. Results:  Image quality was high in CTA with and without DEBR and DSA (3.78 +/ – 0.36, 3.78 +/ – 0.36, 3.27 +/ – 0.46, n. s.). No significant bone remnants were present on DE-CTA images, DEBR was fast (< 42 s). Mean scores of EC/IC bypass visualization were not significantly different in the extracranial and intracranial bypass segment for CTA, DEBR-CTA, and DSA. Fig. 1

However, in the trepanation segment pseudo-lesions of the bypass were present in DEBR-CTA (6 out 24 cases) significantly affecting rating of bypass integrity. Fig. 2

210

Conclusion:  CTA with automated DEBR of the extracranial and intracranial vessels is feasible and provides fast and complete removal of the cranial bones. Whereas visualization of EC/IC-bypass in the extracranial and intracranial segments is not different between DEBR-CTA and DSA, DEBR may result in artificial stenosis in the trepanation segment. V1462 Carotid artery stenting in acute stroke P. Papanagiotou, C. Roth, S. Walter, A. Haass, M. Politi, K. Fassbender, W. Reith (Homburg) Background and purpose:  Stroke caused by acute occlusion of the origin of the internal carotid artery (ICA) is associated with a significant level of morbidity and mortality. The purpose of this study is to demonstrate the technical success of recanalization using stent implantation for an acute ICA occlusion as well as the clinical results. Methods:  We retrospectively reviewed 17 patients that were treated for an extracranial ICA occlusion within 6 hours of symptom onset. In 14 patients there was an additional occlusion; at the level of the terminal segment of the ICA (2 patients) and at the level of the middle cerebral artery (MCA) in 12 patients. To treat the intracranial occlusion intraarterial thrombolysis or mechanical recanalization was performed. Recanalization result was assessed by follow-up angiography i mmediately after the procedure. The neurological status was evaluated before and after the treatment with a follow up of up to 30 days using the NIHSS (National Institutes of Health stroke scale) and mRS (modified ranking scale). Results:  Successful revascularization using acute stent implantation was achieved in 16 patients (94,1%). There was no acute stent thrombosis. There was one procedure related complication. After successful recanalization of the origin of the ICA the intracranial recanalization with TIMI 2/3 flow was achieved in 8 of the 14 patients (57.1%). Mean NIHSS on admission was 17 with a standard deviation of 3. 9 of the patients (52.9%) showed an improvement of more than 4 points on the NIHSS at the day of discharge. 8 patients (50%) showed a mRS score of  ≤ 3 at 90 days (43.7% mRS  ≤ 2). Mortality was 18.5%. Conclusion:  Recanalisation and stent implantation for an acute extracranial ICA occlusion in selected patients is a relatively safe procedure that can lead to a significant improvement of the clinical symptoms.

Abstracts

V1463 Stent-assisted mechanical recanalization for treatment of acute intracerebral artery occlusion P. Papanagiotou, C. Roth, S. Walter, S. Behnke, C. Becker, W. Reith (Homburg) Background and purpose:  The purpose of this study is to demonstrate a new approach to the use of a self-expanding stent in the treatment of acute ischemic stroke. Methods:  12 consecutive patients suffering from acute intracerebral occlusion were treated using a self-expandable intracranial stent, which was withdrawn, in its unfolded state. For this new technique we used the Solitaire AB (ev3, Calif., USA) that is the only intracranial stent that is fully recoverable. 4 patients had an occlusion of the basilar artery, 8 patients had a middle cerebral artery occlusion, 3 of these had to be treated for acute carotid occlusion first. Recanalization result was assessed by follow-up angiography i mmediately after the procedure. The neurological status was evaluated before and after the treatment with a follow up of up to 30 days using the NIHSS (National Institutes of Health stroke scale) and mRS (modified ranking scale). Results: Successful revascularization could be achieved in 91,6% (TIMI 2 and 3), a TIMI 3 state could be reached in 7 patients, partial recanalization or slow distal branch filling could be achieved in 4 patients (TIMI 2). In 11/12 (91.6%) cases i mmediate flow restoration was achieved after deployment of the device. The stent was removed in its unfolded state in all patients. The mean time from stroke system onset to recanalization was 265 min. with a standard deviation of 126 min. Mean NIHSS on admission was 17 with a standard deviation of 3. 66.7% of the patients improved more then 10 points on the NIHSS at discharge. 58.3% showed a mRS score of  < = 3 at 30 days (50% mRS  < = 1). Mortality was 16.6%. In one of the cases (8.3%) an asymptomatic ICH was detected on control CT. Conclusion:  The withdrawal of an unfolded fully recoverable intracranial stent shows very promising angiographic and clinical results. It combines the advantages of prompt flow restoration and mechanical thrombectomy. V1464 Cerebral venous thrombosis: diagnostic accuracy of combined, dynamic and static, contrast-enhanced 4D MR venography S. Meckel, C. Reisinger, J. Bremerich, M. Wolbers, S. Engelter, K. Scheffler, S. Wetzel (Freiburg; Basel, Zürich, CH) Background and purpose:  MR including MRV is an established method to diagnose CVT. However, it remains unsettled which MR imaging modalities offer the highest diagnostic accuracy. We evaluated the accuracy of a combined, dynamic (1.5 seconds per dataset) and static (voxel size, 1.1 x 0.9 x 1.5  mm), contrast-enhanced MRV method (combo-4D MRV) relative to other established MR/MRV modalities.

Abstracts

Materials and methods:  A total of 39 patients with CVT (n  = 20) and control subjects (n = 19) underwent combo-4D MRV, 2D TOF MRV, GRE imaging, and T2W imaging. For these modalities, diagnostic accuracy using receiver operating characteristics (ROC) for CVT affecting 53 out of 234 predefined venous segments was determined. Sensitivity and specificity were separately calculated for different stages of CVT (acute/subacute/chronic). Results:  Combo-4D MRV showed the highest accuracy (AUC = area under the respective reader-specific ROC curve, 0.99 [95% CI, 0.97–1.0]; sensitivity, 97% [84%– 100%]) for thrombosed dural sinuses. For all thrombosed segments including cortical veins, its sensitivity was best (76% [64%–84%]; AUC, 0.92 [0.88–0.96]), followed by TOF MRV (72% [59%–81%]; AUC, 0.93 [0.88–0.97]). Even for chronic CVT, it showed a relatively high sensitivity of 67% (30%–90%). For thrombosed cortical veins alone, GRE images achieved the highest sensitivity (66% [46%–81%]; AUC, 0.88 [0.78–0.97]). Specificities of all modalities ranged from 96% to 99%. Conclusion:  Combo-4D MRV showed an excellent accuracy for the diagnosis of dural sinus thrombosis. The analysis of dynamic patterns of contrast enhancement in dural sinuses appeared useful to identify chronic thrombosis. To diagnose thrombosed cortical veins, GRE images should primarily be analyzed. Fig. 1

211 Fig. 2

POSTER Intervention/Varia P1465 In vitro comparison of different carotid artery stents: a pixel-by-pixel analysis using CT angiography and contrast-enhanced MR angiography at 1.5T and 3T M. Lettau, A. Sauer, S. Heiland, S. Rohde, M. Hartmann, J. Reinhardt, M. Bendszus, S. Hähnel (Freiburg, Heidelberg) Background and purpose:  CT angiography (CTA) and MR angiography (MRA) are increasingly used methods for evaluation of stented vessel segments. The purpose of this study was to compare CTA, contrast-enhanced MRA (CEMRA) at 1.5T, and CEMRA at 3T for the visualization of carotid artery stents and to define the best noninvasive imaging technique for each stent. Materials and methods:  CTA and CEMRA appearances of 18 carotid artery stents of different designs and sizes (4.0 to 10.0  mm) were investigated in vitro. The profile of the contrast-to-noise ratio (CNR) of the lumen of each stent was calculated semiautomatically by a pixel-by-pixel analysis using the medical imaging software OSIRIS®. For each stent, artificial lumen narrowing (ALN) was calculated. Results:  In all but one stents, ALN was lower on CEMRA at 3T than at 1.5T. With CEMRA at 3T and at 1.5T, ALN in most nitinol stents was lower than in the groups of stainless steel and cobalt alloy stents. In most nitinol stents, ALN on CEMRA at 3T was lower than on CTA. In all stainless steel stents and cobalt alloy stents, ALN was lower on CTA than on CEMRA. With CTA and CEMRA, in most stents ALN decreased with increasing stent diameter. Conclusion:  CTA and CEMRA evaluation of vessel patency after stent placement is possible, but considerably

212

impaired by ALN. Investigators should be informed about the method of choice for every stent. P1466 Optimized algorithm for polytrauma management and its implication on the patients’ prognosis B.A. Prümer, S. Terwey, C. Reckels, U. Haverkamp, A.R. Fischedick (Münster) Introduction:  A swift and standardized investigation procedure of polytrauma patients is of vital importance for an optimized diagnostic workup, therapy stratification and implication on the patients´ prognosis. Neurosurgery is one of the major columns, because evacuating an intracranial haemorrhage strongly influences the final overall outcome. Material & method: This study covers the investigation of 296 polytraumatized patients administered to our hospital during a four year time period. The patients were grouped according to their health- and consciousness state and to the transportation device and analysed according to the time of diagnostic work up, time to operation suite (if necessary) and their final outcome (survival, rehabilitation & daily life) Results:  203 patienets were delivered by ambulance, while 93 by helicopter. The average GCS read 8,2 (2,0 – 14) and 6,5 (4 – 11) respectively. The diagnostic workup was performed on a MDCT, (Aquilion 64, Toshiba) in a three step trauma triage concept. (A: head&neck, B: thorax&abdomen and C: lower limbs –if necessary). In 38% of cases only step A, in 33% steps A and B and in 3% all steps (A + B + C) were done. MDCT-Imaging included 0.5 mm acquisitions, 5 mm axial-, 8 mm coronal MIP(torso)- and 3 mm sagittal(spine) reconstructions. The average time to perform the diagnostic procedures – inclusive initial ER-treatment – took 54 minutes (20 – 128 minutes). There was an inverse correlation between the GCS and the duration of reanimation, while the time for diagnostic procedures varied only little but paralleled the GCS down to a score of 5, then for a GCS below 5 necessitating a moderate increase of imaging work up. The prognosis was favourable for patients with an initial high GCS above 10, while we saw a poor prognosis with a GCS below 6. Another independent predictor of a favourable outcome was the duration of the preoperational stabilizing and diagnostic work up. If this preparation took less than 42 minutes we saw an improved outcome with one third of patients in daily life within 10 weeks. Conclusion:  The average time for life saving and diagnostic procedure should take less than 1 hour (best results with  <  42 minutes). The prognosis is favourable for patients with an initial GCS  >  10, while a poor outcome becomes more likely with a GCS  <  6. The duration of the preoperational stabilization and diagnostic workup is an independent factor for predicting the posttraumatic outcome and evaluating the patients prognosis.

Abstracts

P1467 Optimized diagnostic work up of cryptic prepontine SAHs with a special focus on the role of MD-CT-Angiograohy B.A. Prümer, S. Terwey, C. Reckels, U. Haverkamp, A.R. Fischedick (Münster) Introduction:  A circumscribed small amount of subarachnoid haemorrhage in a prepontine localisation is a finding, trigger questions concerning the diagnostic work up as well as the therapeutic consequences. Literature says that one hardly detects a precise cause for the actual bleeding rendering therapeutic approaches rather questionable. Material & method:  This study covers the investigation of 135 consecutive patients, administered to our hospital with an acute SAH during a five year time period. Those patients presenting with a limited prepontine SAH received a DS-angiography (Integris-5000, Philips) as well as a MDCT-angio (Aquilion-64, Toshiba) to determine the bleedings cause. Both investigation techniques were compared to one another in terms of detecting an aneurysm, an AV-malformation, a vascular spasm or other associated pathologies. Results:  Out of 135 patients presenting with an acute SAH in 27 (20%) patients we discovered a limited prepontine SAH. The patients presented with a mean GCS of 10,4 (14 – 6) and a Hunt&Hess state of 1,8 (1 – 4). In all 27 patients a DSA and a MSCTA was technically feasible. A control study was performed after 12 days up to 2 weeks. The average time to perform the diagnostic procedures took 55 minutes for the DSA, while the MSCTA took 12 minutes. The DSA procedure caused one complication with a temporary hemiparesis resolving after two weeks, MSCTA showed none. With both modalities we discovered neither an aneurysm, nor an AVM responsible for the SAH. Also the control studies after 2 weeks and 6 months were negative. DSA demonstrated vasospasm in 6 patients (22%), while CTA only revealed spasm in 3 patients (11%). On the other hand MSCTA is superior in detecting associate pathologies like intraparenchymal – 1 patient (4%) and intraventricular bleeding – two patients (7%) or an evolving hydrocephalus – three patients (11%). Conclusion:  In all 27 cases of a cryptic prepontine SAH neither DSA, nor MSCTA could reveal a definitive bleeding cause. Neither an aneurysm. Nor an AVM was found. DSA has a higher sensitivity in finding vasospasm, while MSCT is favourable for detecting associate pathologies like intraparenchymal and intraventricular bleeding or an evolving hydrocephalus. We suggest a combined diagnostic workup of a cryptic prepontine SAH: 1.) Initially MSCTA after diagnosing a SAH, 2.) DSA after 12-14 days and 3.) MSCTA-control study after 4-6 months, given that the preceeding angio is negative.

Abstracts

P1468 Traumatic trigeminal cavernous fistula and treatment of persistent trigeminal artery variant pseudo-aneurysm S. Meckel, W. McAuliffe (Freiburg; Perth, AU) We report a case of a 53-year-old male patient presenting with a post-traumatic trigeminal-cavernous fistula that underwent spontaneous occlusion. A pseudo-aneurysm located on a rare persistent trigeminal artery (PTA) variant, an anomalous anterior inferior cerebellar artery (AICA) originating from the internal carotid artery (ICA), was unmasked as the cause of the fistula. This pseudo-aneurysm was successfully treated with coil embolization. The anatomical variants of the trigeminal artery and potential implications for endovascular treatment of trigeminal artery aneurysms are discussed. A review of the literature on aneurysms and trigeminal cavernous fistulas originating from the PTA or its variants is presented.

213

of small T1w hyperintense lesions and the necessitiy of CRu (Complete response unconfirmed defined as small persistent enhancing abnormality) as a response category. Material and methods:  MRIs of 29 patients with PCNSL of the “Freiburger ZNS NHL-Study” who were treated between January 2007 and December 2009 were evaluated retrospectively. Patients were included from whom MRI prior to therapy, one or more MRI during therapy and MRI after therapy were available. T1w-Sequences with and without contrast enhancement were analysed. Results:  23 patients (80%) showed T1w (without contrast) hyperintense lesions under therapy. In 4 cases hyperintense areas could be seen prior to therapy, in 12 cases they occurred after tumor resection or biopsy and in 7 cases they appeared spontaneously. In the majority of the cases the hyperintense signal was faint. 87% of the patients with T1w hyperintense lesions were categorized as CRu during therapy but only 17% of the patients without hyperintense lesions. CRu changed to CR (complete response) in 86% of the cases and remained CRu in 14%. Conclusion:  T1w hyperintense lesions occur frequently during high-dose chemotherapy of PCNSL and classification in remission state can be difficult. Minor hemorrhages in the affected area could be the reason for these changes. CRu is a very helpful remission category to handle small lesion remnants and prevent patients from unnecessary extension of therapy. POSTER Pediatric Imaging/Stroke P1470 A previously undescribed malformation in genetically confirmed Joubert syndrome: Interpeduncular heterotopia as a manifestation of abnormal migration due to ciliopathy? U. Kotzaeridou, J. Pietz, U. Moog, M.F. Bauer, A. Seitz, M. Bendszus, E. Boltshauser, I. Harting (Heidelberg, Ludwigshafen; Zürich, CH)

P1469 MR follow-up in Patients with Primary CNS Lymphoma during high-dose chemotherapy C. Hader, U. Feige, B. Kasenda, G. Illerhaus (Freiburg) Introduction:  Primary CNS Lymphoma (PCNSL) is a rare but aggressive Non-Hodgkin-Lymphoma exclusively invading the brain, leptomeniges, spinal cord or eyes. MRI is standard to evaluate the tumor response during therapy. Complete remission (CR) is defined as complete disappearance of all enhancing abnormalities on gadolinium-enhanced MRI but often small remnants occur and interpretation of the images can be difficult. We evaluated the frequency

Objective:  Joubert syndrome (JS) is a rare, most often autosomal-recessive disorder with a characteristic complex malformation of the hindbrain, seen as the so called molar tooth sign (MTS) on axial imaging. This results from vermis hypoplasia, a deep interpeduncular fossa and thickened, elongated superior cerebellar peduncles. Clinical diagnostic criteria are developmental delay and hypotonia, inconsistent findings are irregular breathing pattern in infancy and ocular apraxia. Joubert Syndrome and Related Disorders (JSRD) comprise disorders with additional CNS, ocular, renal, hepatic and/or skeletal manifestations. The molecular basis is a ciliopathy, currently 10 causative genes have been identified.

214

Case report:  We present a 16 months old girl with clinical and neuroimaging signs of JS and a previously undescribed cortexisointense heterotopia in the interpeduncular fossa. Additional MR findings comprised a vermal cleft, volume deficit of the right cerebral peduncle and pons, a dysplastic tectum, and myelination delay. Clinical manifestations were an abnormal breathing pattern as a neonate followed by global developmental delay, hypotonia, horizontal nystagmus, and diffuse retinal dystrophy. Renal and hepatic findings were normal on ultrasound. Molecular genetics revealed a homozygous mutation of CEP290. Discussion:  The patient's clinical and radiological findings suggested the diagnosis of JS which was confirmed by molecular analysis. CEP290 encodes a structural protein at the base of primary cilia. Primary cilia seem to have a role in signaling pathways controlling cell division, differentiation, and migration. CNS malformations are postulated to result from defective dorsal-ventral patterning and cranio-caudal segment specification in the mid-hindbrain boundary due to the ciliopathy. Our finding of an interpeduncular lesion with a signal intensity of either heterotopia or hamartoma has not yet been described. Hypothalamic hamartoma in combination with MTS is known and specific for oro-facio-digital syndrome VI. In our patient there are however neither a connection with the hypothalamus nor typical clinical findings (polydactyly, lingual hamartoma, frenula). A finding similar to ours is shown but not discussed by Alorainy et al. (2006; Fig. 1c) and heterotopic gray matter within the mesencephalon has been described (Poretti et al. 2007). It seems likely that these mesencephalic heterotopias also result from abnormal migration due to the ciliopathy. P1471 Atretische Cephalocele – eine Fallbeschreibung G. Mühlenbruch, S. Leyka mm, B. Wesseling, M. Mull, M. Wiesmann (Aachen) Einführung:  Atretische Cephalocelen sind seltene Formen des Cranium bifidum, die gehäuft mit Fehlbildungen des ZNS vergesellschaftet sind. Wir zeigen den Fall eines Neugeborenen mit parietal lokalisierter atretischer Cephalocele, die mit einer subependymalen Heterotopie und einem kongenitalen Hydrocephalus internus assoziiert ist. Fallbeschreibung:  Bei dem 5 Monate alten männlichen Säugling wurden bereits pränatal sonographisch die Verdachtsdiagnosen eines Hydrocephalus sowie einer Cephalozele gestellt. Klinisch zeigte sich eine durch intakte Kopfhaut bedeckte, palpatorisch weiche, nicht fluktuierende, rundlichumschriebenen Schwellung, die in der parietalen Mittellinie cranial der kleinen Fontanelle lag. Sonografisch zeigte sich eine echoinhomogene Struktur mit Verbindung nach intrakraniell. Zusätzlich war das Ventrikel-

Abstracts

system erweitert, insbesondere die Seitenventrikel auf 11  mm, gemessen in Höhe der Foramina Monroi. In der anschließend durchgeführten craniellen Magnetresonanztomographie (MRT) stellte sich die extrakranielle Läsion hyperintens im T2-gewichteten Bild und hypointens im T1-gewichteten Bild dar. Eine Kontrastmittelaufnahme lag nicht vor. Eine auch sonografisch dokumentierte Konturunregelmäßigkeit der linken Seitenventrikelbegrenzung entsprach einer isolierten subependymalen Heterotopie. Die Erweiterung der Seitenventrikel bestätigte sich, es fanden sich weder eine Aquäduktstenose noch ein intrakranieller Tumor. Assoziiert zu der extrakraniellen Läsion bestand ein vertikaler Verlauf des Sinus rectus. Dieser zog auf die Knochenlücke zu und mündete in den Sinus sagittalis superior. Im Verlauf war das Kind neurologisch unauffällig. Die Weite der Seitenventrikel nahm in den folgenden Ultraschallkontrollen zu und der Kopfumfang wies einen deutlichen Perzentilensprung auf, sodass in der 5. Lebenswoche die komplikationslose Implantation eines ventrikuloperitonealen Shunts erfolgte. Dies führte konsekutiv zur Abnahme des Ventrikeldurchmessers. Die Exzision der Cephalozele wurde im Alter von 4 Monaten als elektiver Eingriff durchgeführt und verlief komplikationslos. Der Befund wurde histologisch bestätigt. Diskussion:  Die extra- und intracranielle Sonographie als nichtinvasive Erstdiagnostik sowie die kranielle MRT sind die Methoden der Wahl zur Diagnostik einer atretischen Cephalocele. Die atretische Cephalocele kann chirurgisch extirpiert werden, hierbei geben Ultraschall und MRT wichtige OPPlanungs Informationen. Abb. 1

Abb. 3

Abb. 2

Abstracts

P1472 Multicompartment analysis of the FID in acute stroke B. Bender, T. Nägele, U. Ernemann, U. Klose (Tübingen) Introduction:  To characterize the FID evolution in the human brain He et al. [1] proposed a signal model which takes into account intracellular and extracellular signal components. Recently a simplified version of this model was used with EPI images to quantify interstitial fluid in the human brain [2]. In stroke early cytotoxic edema with cell swelling and later apoptosis as well as vasogenic edema should contribute differently to the signal components. Subjects and method:  2 patients with an acute and a subacute stroke (~19 h/~96 h after onset of symptoms) where included in this study. An enlarged stroke region was found in the first patient 3 days after the measurement in a control MR. The measurement procedure of [2] was used on a 3T scanner (Trio, Fa. Siemens) to acquire the FID (37 GREEPI images with varying TE between 21 – 301ms). After movement correction the FID signal was smoothed with a [4 4 5]  mm kernel and fitted to the simplified signal model on a voxel-by-voxel basis with a non-linear least square algorithm. The ADC map and the DWI images were used to select a ROI within the ischemic area. All fitting parameters of the stroke lesion were compared with a contralateral control ROI. The ADC map and the DWI image of the increased ischemic area was used to select a second ROI in the first patient. Only voxels which appeared normal in the initial DWI image were included. For significance testing Wilcoxon rank sum test was used. Results and discussion:  In figure 1 ROI selection from the DWI-image and ADC-map is shown. The enlarged region of the stroke 3 days after the initial measurement is clearly visible on the slice depicted. Table 1 su mmarizes the differences found in the three stadiums. Cell swelling leads to an increase in T2* at all times and to a decrease in the volume fraction lambda of the extracellular water in the acute stadium. Cell death becomes prominent after 24-72h [3], which is in good agreement with the enlarged extracellular water compartment in the subacute stroke. Cell death and consecutive electrolyte changes account probably also for the change in the frequency shift. Conclusion:  Analysis of T2* in stroke patients with a two compartment model yields results consistent with theory. Significant changes in T2* can be found in ischemic areas prior to DWI changes. References: [1] MRM 2007;57:115–26, [2] MRM 2009;61:834–41, [3] Trends Neurosci 199;22:391–7.

215 Fig. 1

Tab. 1 T2*1 [ms]

∆ T2*1

∆f [Hz]

λ [%]

∆ ∆f

∆λ

S

C

Hyperacute

53

48

5 **

2.8

2.8

0.0

31.8

30.1

–0.1

Acute

62

48

14 **

2.6

2.5

0.1

14.9

28.2

–1.6 *

Subacute

65

51

14 **

2.6

2.9

–0.3 **

25.4

22.5

1.9 **

S–C

S

C

S–C

S

C

S–C

P1473 Endovaskuläre mechanische Rekanalisation akuter zerebraler Gefäßverschlüsse: initiale Ergebnisse mit dem Solitaire Stent S. Stampfl, M. Hartmann, S. Haehnel, P. Ringleb, M. Bendszus, S. Rohde (Heidelberg) Hintergrund:  In verschiedenen Fallserien wurde über die Akutbehandlung von intrakraniellen Gefäßverschlüssen mit ablösbaren Stents berichtet. Mit dem Solitaire Stent (ev3, Plymouth, USA) steht ein System zur Verfügung, das über einen 2.3F Mikrokatheter eingeführt und im verschlosse-

216

nen Gefäßsegement passager freigesetzt werden kann. Wir berichten über erste Ergebnisse mit dem Solitaire Stent bei der Rekanalisation akuter cerebraler Gefäßverschlüsse. Material und Methoden:  Seit 03/2009 wurde der Solitaire Stent bei 10 Patienten (mittleres Alter 63 Jahre; 33–82) mit akutem thrombembolischen Verschluß eines intrakraniellen Gefäßes (MCA n = 7; AB n = 3) eingesetzt. Die durchschnittliche Symptomdauer bis zum Beginn des Eingriffs betrug 255,4 min; bei 9 Pat. wurde der Stent in Kombination mit einer i. a. Lyse, bei 1 Pat. ohne weitere pharmakologische Maßnahmen angewendet. Anhand der angiographischen Befunde wurden der Gefäßverschluss und die Perfusion des abhängigen Territoriums vor und nach Rekanalisation mit Hilfe des Thrombolysis-in-myocardial-infarction Scores (TIMI 0–3) und des Thrombolysis-in-cerebral-infarction Scores (TICI 0–3) bewertet. Ergebnisse:  Bei allen Pat. bestand vor der Intervention ein vollständiger Gefäßverschluss (TIMI/TICI 0) mit fehlender bzw. unzureichender Kollateralisierung. Der Stent konnte bei allen Patienten im verschlossenen Gefäßsegment freigesetzt werden. Die Rekanalisation der Gefäßsegmente war bei 8 von 10 Pat. erfolgreich (TIMI 3, TICI 2,75). Bei jeweils einem Patienten mit einer verkalkten Mediastenose und einem Pat. mit einem kardialen Embolus konnte der Verschluss nicht wiedereröffnet werden. Durchschnittlich wurde der Stent während der Intervention 2x geöffnet (1–4), in 2 Fällen wurde mit dem geöffneten Stent eine Thrombektomie durchgeführt. Bei 2 Pat. wurde der Stent vollständig freigesetzt und abgelöst, um die erzielte Reperfusion sicherzustellen. Prozedurale Komplikationen wurden nicht beobachtet; bei 1 Patienten kam es 1 Stunde nach kombinierter mechanischer Rekanalisation und i. a. Lyse der A. basilaris zu einer intrakraniellen Blutung mit letalem Ausgang. Schlussfolgerung:  Der Solitaire Stent konnte bei allen Pat. im verschlossenen Gefäßsegement komplikationslos freigesetzt werden; eine Rekanalisation gelang in 80%. Der Stent stellt somit ein vielversprechendes Instrument zur mechanischen Gefäßeröffnung bei akuten zerebralen Gefäßverschlüssen dar und sollte in prospektiven Studien weiter untersucht werden. P1474 Basilar artery dissection with subarachnoid haemorrhage in a patient with Systemic Lupus Erythematosus (SLE) G. Benndorf, H. Morsi, S. Gopinat (Houston, US) Background/purpose:  Systemic Lupus Erythematosus (SLE) may be associated with vasculopathies in the cerebral circulation, including aneurysms, thrombosis and vasculitis-like changes. Subarachnoid hemorrhage (SAH) caused by saccular cerebral aneurysms is a relatively rare compli-

Abstracts

cation of SLE (3.9%). The association of SLE with acute intracranial dissection causing SAH has not been described in the literature. We report the unusual case of a patient that presented with SAH caused by acute basilar artery dissection that was successfully managed by endovascular means. Material & methods:  A 35-year-old female with known SLE presented with SAH H&H 3. Cerebral angiography revealed several stenotic and dysplastic arterial segments in both anterior and the posterior circulations. The bilateral middle cerebral arteries showed focal fusiform aneurysmal dilations. In addition, the posterior circulation revealed a narrowed proximal basilar artery segment caused by a dissecting aneurysm that was considered the source of the bleeding. Under full heparinization and premedication with dual antiplatelets, endovascular management was performed the next day by placement of two overlapping Enterprise Stents (4.5 × 28 mm, 4.5 × 30 mm). Results:  Immediate post stenting angiogram showed increased diameter of the basilar artery lumen and a slight reduction of the size of the pseudoaneurysm. A second, 4 day-angiogram showed a further minimal reduction of the aneurismal size. The patient fully recovered under medical treatment within 3 weeks and was discharged. A third control angiogram was performed after 6 weeks and showed subtotal occlusion of the aneurysm. A final 6-month FU arteriogram is pending. Conclusion:  Patients with SLE may not only present with various vasculopathic changes causing ischemic infracts, aneurysmal or non-aneurismal SAH. Intracranial dissection causing SAH in SLE patients appears to be extremely rare, but may occur and should be considered in the differential diagnosis. Endovascular management using flow-diverting implants can be a feasible and effective treatment option even in an caute setting. P1475 MRI abnormalities detected by proton density measurements in normal appearing brain tissue of patients with amyotrophic lateral sclerosis X.-Q. Ding, K. Kollewe, K.S. Blum, S. Schmalbach, S. Kehbel, R. Dengler, H. Lanfermann, S. Petri (Hannover) Purpose:  We performed this prospective magnetic resonance imaging (MRI) study including use of quantitative MR parameter mapping methods, with the aim to detect possible invisible pathological alterations in the brain of patients suffering from amyotrophic lateral sclerosis (ALS). Materials and methods:  Seven ALS patients diagnosed according to the revised El Escorial criteria and seven ageand sex-matched healthy controls underwent MRI. Brain maps of T2 relaxation time (T2) and relative proton density (PD), as well as apparent diffusion coefficient (ADC) were derived for each subject. Numeric values of T2, PD, and ADC were measured in 22 selected regions of interest

Abstracts

of gray matter and white matter including pyramidal tracts. The values of the patients were compared with those of the matched controls. Results:  By conventional MRI no morphological alterations were found except slight brain atrophy in two patients. However, significant differences were detected by quantitative MR measurements: higher PD was found in all of the ALS patients (6% – 12%, p < 0.02), with a maximum at the posterior limb of capsula interna bilaterally (~ 12 %); The T2 as well as the ADC showed also a tendency to an increase in patients, which was only significant at the genu

217

of corpus callosum (T2: 4%, p < 0.05; ADC: 10%, p < 0.05), and the posterior limb of capsular interna bilaterally (T2: 7%, p < 0.05; ADC: 11%, p < 0.05). Conclusion: Albeit a small sample size, we found that a general significant increase of proton density in ALS patients was evident, with partially significant increased T2 and ADC values. The findings indicate that cerebral microstructural changes in ALS involve not only the motor system but the whole central nervous system, and that the degrees of these changes are region-specific.

218

Abstracts

A Assmann, B.

V1454

B Bartsch, A. Bauer, M. F. Bäumer, P. Becker, C. Beckmann, M. Behnke, S. Bender, B. Bendszus, M. Benndorf, G. Biondi, A. Bisdas, S. Blatow, M. Blum, K. S. Boltshauser, E. Boltze, J. Boppel, T. Braun, C. Brecht, F. G. Bremerich, J. Brockmann, C. Brockmann, M. A. Budischewski, K.

V1430, V1447 P1470 V1430, V1431, V1439, V1447 V1463 V1438 V1463 V1460, P1472 V1423, V1430, V1431, V1435, V1436, V1439, V1440, V1442, V1447, V1449, P1465, P1470, P1473 V1425, P1474 V1425 V1446 V1435, V1436 P1475 P1470 V1450 V1447 V1423 V1436 V1464 V1441 V1441 V1436

C Cosgarea, R. Cronqvist, M. Csernus, R.

V1440 V1422 V1440

D Da mmann, P. Danz, B. Danz, S. Deichmann, R. Dengler, R. Diehl, S. Ding, X.-Q. Dombert, T. Dörner, N. Dreyer, A.

V1433 V1432 V1460 V1451 P1475 V1461 P1475 V1439 V1447 V1450

E Eisenbeis, A. Engelter, S. Ernemann, U.

V1457 V1456 P1466, P1467 V1456 V1433 V1420 V1455 V1437, V1452

G Gabarde, S. Gaudino, C. Gersing, A. Gharabaghi, A. Giesemann, A. Gizewski, E. R. Glombik, R. Gmelin, E. Gopinat, S. Gopinath, S. Gottschalk, A. Götz, F. Grams, A.E. Grodd, W. Groden, C. Günther, P.

V1454 V1440 V1456 V1438 V1426, V1443 V1433 V1428 V1428 P1474 V1425 V1432 V1426, V1443 V1433 V1438 V1441, V1459, V1461 V1437

H V1462 Haass, A. Hader, C. P1469 Haehnel, S. P1473 Hähnel, S. V1423, P1465 Harati, A. V1424 Harting, I. V1454, P1470 Hartmann, M. V1423, V1442, P1465, P1473 Hattingen, E. V1448, V1451, V1455 Hattingen, J. V1428 Hauser, T.-K. V1460 Haverkamp, U. P1466, P1467 Heiland, S. V1430, V1440, V1442, V1447, V1449, P1465 Heindel, W. V1457 Heinemann, T. S1418 Heuschmid, M. V1460 Hoffmann, K.-T. V1437, V1450, V1452 Huck, S. V1459 I

V1456 V1464 V1446, V1460, P1472

F Fassbender, K. Feige, U. Fenchel, M.

Feyen, L. Fiehler, J. Fischedick, A. R. Forkert, N. Forsting, M. Frankenberg, P. Franz, K. Fritzsch, D.

V1462 P1469 V1460

Illerhaus, G.

P1469

J Jestaedt, L. Jurcoane, A.

V1449 V1448, V1451

Abstracts

219 O

K Kaestel, T. Kasenda, B. Kästel, T. Kehbel, S. Kemmling, A. Kieslich, M. Klose, U. Kolb, A. Kölker, S. Kollewe, K. Korn, A. Kotzaeridou, U. Kraff, O. Krüger, R. Kumar, V. Kunz, U.

V1439 P1469 V1430, V1431 P1475 V1457, V1461 V1455 P1472 V1446 V1454 P1475 V1460 P1470 V1433 V1438 V1438 V1432

L Ladd, M. E. Läer, L. Lanfermann, H. Lehrnbecher, T. Lemke, D. Lenarz, T. Lettau, M. Leykamm, S. Leypoldt, F. Lichte, T. Lobsien, D. Loehr, C. Lohmann, F. Luckner, R.

V1433 V1438 V1426, V1443, P1475 V1455 V1449 V1426 P1465 P1471 V1456 V1443 V1450, V1452 V1424 V1424 V1424

M Maderwald, S. Magerkurth, J. Magnus, T. Maier, E. Majdani, O. Mauer, U. M. Mautner, V.-F. McAuliffe, W. Meckel, S. Moog, U. Morsi, H. Mpotsaris, A. Mühlenbruch, G. Mühlhausen, C. Mull, M. Müller, M.

V1433 V1451 V1456 V1454 V1426 V1432 V1431 V1422, P1468 V1422, V1464, P1468 P1470 V1425, P1474 V1424 P1471 V1454 P1471 V1427

N Nafe, R. Nägele, T. Neumaier-Probst, E. Neumann-Haefelin, T. Niederstadt, T. Nölte, I.

V1448 V1446, V1460, P1472 V1454 V1451 V1457 V1441, V1459, V1461

Oberbeck, M.

V1452

P Papanagiotou, P. Petri, S. Pham, M. Phatouros, C. Philipp, M. Pichler, B. Pietz, J. Pilatus, U. Politi, M. Porto, L. Prümer, B. A. Puchner, M. J. A.

V1462, V1463 P1475 V1430, V1431, V1439, V1440, V1442 V1422 V1437 V1446 P1470 V1455 V1462 V1455 P1466, P1467 V1424

R Radbruch, A. Reckels, C. Reinhardt, J. Reisinger, C. Reith, W. Ringleb, P. Ritz, R. Rohde, S. Roth, C.

V1447 P1466, P1467 V1436, P1465 V1464 V1462, V1463 P1473 V1446 V1423, V1440, P1465, P1473 V1462, V1463

S Sauer, A. Scharf, J. Scheffler, K. Schick, M. Schlötzer, W. Schmalbach, S. Schmitz, B. Schneider, P. Schuhmann, M. Schwartz, A. Schwarz, J. Seifarth, H. Seitz, A. Seiz, M. Singer, O. C. Singh, T. Stampfl, S. Stasierowski, J. Staub, F. Stippich, C. Stöver, T.

P1465 V1441 V1464 V1427 V1427 P1475 V1427 V1435 V1431 V1428 V1452 V1457 V1454, P1470 V1441 V1451 V1422 P1473 V1424 V1439 V1436 V1443

T Terwey, S. Thomas, C.

P1466, P1467 V1460

220

Abstracts

U Umutlu, L.

V1433

V V1451 Volz, S. von Kummer, R. S1417, V1419 W Wagner, M. Walter, S. Weber, D.

V1448, V1451 V1462, V1463 V1437

Weber, W. Weiler, M. Wengenroth, M. Wesseling, B. Wetzel, S. Wick, A. Wick, W. Wiesmann, M. Wiestler, B. Wolbers, M.

V1424 V1449 V1435, V1436 P1471 V1464 V1447 V1447, V1449 P1471 V1447 V1464

Z Zanella, F.

V1448, V1455