XVII. European Stroke Conference
Nice, France

Oral Session:

Recovery and rehabilitation
Date: Thursday 15 May 2008  
Time: 16:30 - 16:40 -  Room: Calliope
Chair: J. Bernhardt, Australia and V. Hömberg, Germany

01
SAFETY IN THE FIRST 170 PATIENTS OF A VERY EARLY REHABILITATION TRIAL (AVERT PHASE III)
J. Bernhardt    H. Dewey    J. Collier    A. Thrift    R. Lindley    M. Moodie    G. Donnan                     
on behalf of the AVERT Collaboration

National Stroke Research Institute

AUSTRALIA

Objectives: Very early mobilisation (VEM) may be an important component of effective stroke unit care. The Phase III AVERT trial will test the efficacy and cost-effectiveness of this simple intervention. We report the safety results from the first 170 patients completing 3 month assessments. Methods: Study design: Phase III randomized controlled trial of 2104 patients, with blinded outcome assessments. Inclusion criteria: Patients over 18 years, admitted within 24 hrs of confirmed cerebral infarct or haemorrhage, whose physiological parameters fall within set limits. Exclusion Criteria: Patients with severe premorbid disability, severe comorbidities or requiring palliative care. Treatment with tPA is not an exclusion. Randomisation: Web-based, blocked randomisation by site and stroke severity (NIHSS). Intervention: The VEM protocol, delivered by a nurse/physiotherapy team, begins within 24 hrs of stroke and continues for up to 14 days. Control patients receive standard care (SC). Safety outcomes: Deaths and serious adverse events (SAEs), including serious falls, at 3 months. Results: Phase III commenced in July 2006 in nine acute stroke units across Australia. To January 2008, 170 patients have completed 3 month assessments. Mean age 70.2 (SD 14.0) years, 80.6% with first ever stroke, 83.5% infarct, mean NIHSS score of 8.4 (SD 6.1), 40% with NIHSS > 7 (moderate and severe stroke). To date 19 patients (11.2%) have been treated with tPA, 63% of whom had experienced a TACI. Only 10 (5.8%) patients have died within 3 months of stroke, four having received tPA. Thirty eight SAEs have been reported including 8 stroke progressions and 7 recurrent strokes. There has been only one serious fall with fracture reported. Conclusion: VEM within a 24 hour time window appears safe with an acceptable rate of serious adverse events and only 5.8% deaths occurring in the first 170 patients.

Recovery and rehabilitation
Date: Thursday 15 May 2008  
Time: 16:40 - 16:50 -  Room: Calliope
Chair: J. Bernhardt, Australia and V. Hömberg, Germany

02
Motor network changes associated with successful motor skill relearning after acute ischemic stroke. A longitudinal functional MRI study
T. Askim    B. Indredavik    T.  Vangberg    A. Håberg                                          
 

St. Olavs Hospital, Trondheim University Hospital and Norwegian University of Science and Technology

NORWAY

Background: Brain plasticity after stroke has been studied extensively, but results are contradictory. Motor learning mechanisms may be operative in stroke recovery and possibly reinforced by rehabilitative training. In the present study we wanted to assess the association between motor skill relearning and early motor network changes after acute ischemic stroke. Methods: Twelve patients with first-ever stroke not encompassing the primary sensorimotor cortex, treated in an evidence based comprehensive stroke unit with emphasize on intensive rehabilitation, were assessed with functional MRI and clinical tests, such as the upper limb items of Motor Assessment Scale (UL-MAS), Nine Hole Peg Test (NHPT), transversal grip strength (TVGS) and key grip strength (KGS), within the first week after stroke and three months later. Statistical parametric mapping was used for statistical inferences of fMRI data. Results: All patients, except one, fulfilled the criteria for complete recovery of the affected arm and hand. Improvement in UL-MAS correlated positively with increased activation within the area of contralesional SII and SMA, plus ipsilesional cerebellum. Increased activation in contralesional MII, SI, SII, bilateral SMA and cerebellum correlated with increased speed on NHPT. The improvement in TVGS was mainly associated with increased activation in ipsilesional somatosensory association areas, contralesional SII and insula, bilateral thalami and cerebellum, while for KGS the activated regions were bilateral MI and SMA. Discussion: The recruitment of contralesional cortical structures is considered to represent cortical plasticity involved in successful relearning of dexterity while the bilateral activation seems important for recovery of grip strength after ischemic stroke.

Recovery and rehabilitation
Date: Thursday 15 May 2008  
Time: 16:50 - 17:00 -  Room: Calliope
Chair: J. Bernhardt, Australia and V. Hömberg, Germany

03
5000 finger grip movements with a new robotic hand rehabilitation device – effects on grip abilities and functional MRI
C. Enzinger    P. Grieshofer    G. Reiter    R. Scherer    R. Linderl-Madrutter    C. Pargfrieder    S. Pegritz    W. Wurm    C. Neuper    F. Fazekas
 

Medical University Graz

AUSTRIA

Background: Successful rehabilitation of highly impaired finger and hand movements after stroke is challenging and often remains unsatisfactory with conventional therapies. We therefore developed a new mechatronic device for rehabilitation of hand function which allows delivering well characterised, high frequent, repetitive movement sequences in an individualised manner. The goal of the ongoing exploratory study is to assess the consequences of such training both directly regarding gains in performance and indirectly using functional MRI. Methods: To date, 11 stroke patients with a moderate to high-grade paresis of the upper limb (pinch grip force grades 2 or 3 according to MRC; spasticity according to the Ashworth Scale <3) have been included (mean age 62, range 47-78, interval to their stroke 41 to 434 days). Before and after three weeks of standardised training using a hand robot (developed at the clinic Judendorf-Strassengel) with an average of 5000 grip movements during 15 therapy sessions in addition to conventional physiotherapy, patients were both tested behaviourally and a subset underwent repeated fMRI using an identical paradigm at 3T. The fMRI experiment consisted of active and passive flexion and extension of the digits II-V of both hands. The Motricity Index (M.I.) before and after training and force measurements on the robot during each session served to assess gains in functional strength. Results: Following the robot training, we observed significant improvements of the patients’ functional strength, both clinically and by measurements of force using the device (M.I. pinch grip pre 23.3+/-6.6 vs. post 26.3+/-4.6, p=0.03; finger flexion pre 7.75+/-4.5N vs. post 11.9+/-4.8N, p=0.04). At baseline, movement of the paretic hand vs. rest was associated with activation of a more bihemispheric network with additional recruitment of ipsilateral motor areas in the undamaged hemisphere compared to movement of the unaffected hand. Subsequent to robot therapy, significant increases in activation in the cerebellum and basal ganglia were noted with movement of the paretic, trained hand. fMRI activation patterns with movement of the healthy (not-trained) hand vs. rest did not change (robust activation of contralateral primary sensorimotor cortex, supplementary and cingulate motor areas, ipsilateral cerebellum in expected somatotopy). Conclusion: These preliminary results suggest distinct changes in sensorimotor networks associated with robotic-assisted rehabilitation of hand function after stroke. The causes and clinical significance of these changes will need further exploration.

Recovery and rehabilitation
Date: Thursday 15 May 2008  
Time: 17:00 - 17:10 -  Room: Calliope
Chair: J. Bernhardt, Australia and V. Hömberg, Germany

04
The effects of functional strength training on lower limb strength and function after stroke.
E.V.Cooke    R.C.Tallis    S. Miller    V.M.Pomeroy                                          
 

St. George's University of London

UNITED KINGDOM

Background Stroke survivors often have permanent residual motor impairment. This may be due to a sub-optimal dose of conventional physical therapy (CPT) and the discouragement of strength training. However, experimental evidence suggests that strength training might be beneficial. Hypothesis Adding functional strength training (FST) to CPT improves muscle function and gait more than either CPT alone or increased intensity of CPT. Methods Multi-centred randomised controlled observer-blind trial. Subjects were within 3 months of stroke with ability to voluntarily move their paretic lower limb. A power calculation estimated sample size as 102. Subjects underwent baseline measurements before being allocated randomly to: CPT; or CPT+CPT; or CPT+FST for 6 weeks. All additional therapy was provided up to 1 hour/day, 4 times/week. Outcome measures were made at 6 weeks after baseline. Measurement battery included: muscle strength (maximum torque during isokinetic knee flexion); walking speed; and functional mobility (Rivermead). Analysis followed the intention-to-treat principle. Data was tested for differences between groups using the Kruskal-Wallis test. Results 109 subjects were recruited. Mean age was 68.3 (SD12.03) years. The attrition rate was 8.3%. For muscle strength the median (IQR) change following intervention was 4.17(9.7) Nm for CPT, 6.33 (12.7) Nm for CPT+CPT and 10.57 (17.5) Nm for CPT+FST (p=0.11). Median (IQR) change in walking speed following intervention was 0.03 (0.13) m/sec for CPT, 0.19 (0.44) m/sec for CPT+CPT and 0.09 (0.17) m/sec for CPT+FST (p=0.10). Rivermead median (IQR) score following intervention was 39.5 (20.5) for CPT, 41.0 (12.0) for CPT+CPT and 41.0 (8.50) for CPT+FST (p=0.255). Discussion Immediately after intervention no statistically significant differences were found between groups for muscle strength, walking speed or functional mobility. There was, however, a tendency for higher change scores in CPT+CPT and CPT+FST groups.

Recovery and rehabilitation
Date: Thursday 15 May 2008  
Time: 17:10 - 17:20 -  Room: Calliope
Chair: J. Bernhardt, Australia and V. Hömberg, Germany

05
Studied on Language Networks Damaged of Wernicke's Aphasia by Imaging
Z. Yumei    Z. Xingquan    W. Chunxue    W. Yongjun                                          
 

Beijing tiantan hospital,capital medical university, Beijing

CHINA

Background and purposes: There are a variety of differing types of aphasia classified by the patient’s deficit in language ability. Among them, Wernicke’s aphasia is known as a fluent aphasia, the major impairment is semantic. As the Geschwind-Lichtheim model, Wernicke’s aphasia damaged Brodmann 22 and 39 areas, namely Wernicke’s area. We used MR diffusion tensor imaging (DTI) and diffusion tensor fiber tractography (DT) to study Wernicke’s area on volunteers and Wernicke’s aphasia cases in order to offer functional anatomy results for language disorders. Methods: DTI and DTFT were implemented on Simens 3.0T MR system. Data were acquired in twenty volunteers and twenty Wernicke’s aphasia cases. Brodmann 22, Brodmann 39 areas and right hemisphere corresponding areas were performed as regions of interest ( ROIs), All fibers passing through these ROIs were reconstructed in three dimensions and visualized streamtubes. And we observed Wernicke’s area fibers distributing both volunteers and Wernicke’s aphasia cases, measured fractional anisotropy (FA) value. Results: From the results of volunteers, we found that Brodmann 22 area and 39 area fibers/ FA in left hemisphere had no significant different compared with that of the mirror side (p>0.05). On the other hand, we also found that Brodmann 22 and 39 areas in left hemisphere of Wernicke’s aphasia had lesser fibers than that of mirror side (p<0.05). Conclusions: The language networks of Wernicke’s aphasia were damaged , and DTI, DTFT can be used to elucidate and investigate major nerve pathways in the brain, and provide future information on the function of the normal and abnormal human brain.

Recovery and rehabilitation
Date: Thursday 15 May 2008  
Time: 17:20 - 17:30 -  Room: Calliope
Chair: J. Bernhardt, Australia and V. Hömberg, Germany

06
Upper plexus anaesthesia in rehabilitation of hand function in subacute stroke
K.M.Stephan    M.  Stollorz    J. Bohland    J. Specht    H. Krause    T. Steinforth    V. Hömberg                     
 

St. Mauritius Therapieklinik Meerbusch, Augusta Krankenhaus Düsseldorf, NTC Universität Düsseldorf

GERMANY

Muellbacher et al. (2003) demonstrated improved hand motor function after plexus anaesthesia in chronic stroke patients. We were interested to see, whether this procedure would also help patients in the subacute phase after stroke (4 to 8 weeks after stroke). We investigated a group of 12 patients after subacute stroke, mean age 64 years (5f; 7m). All patients had a moderate paresis after their first clinical stroke. Upper plexus anaesthesia was performed on the affected side by an experienced anaesthetist (T.S.) before training started. A special daily training course was implemented for all participants lasting for two weeks. Motor testing was performed by occupational therapists and neurophysiological parameters (motor evoked potentials and somatosensory evoked potentials) were collected before and after therapy. All patients gave informed consent. A second group of subacute stroke patients with similar demographic and clinical data served as a control group. They received the same special training, but had no upper plexus anaesthesia. Both, the patients of the experimental group and of the control group showed a clear improvement of arm function during their training. There were no statistical significant differences of overall recovery between the two groups within the two weeks. However, while patients without plexus anaesthesia showed most recovery during the first week, in patients with plexus anaesthesia the speed of recovery was greater in the second week. Furthermore, proximal and distal function showed different time courses of recovery. The present data suggest, that upper plexus anaesthesia influences the sequence and possibly also the mode of recovery. Our period of observations (2 weeks) was however not long enough to show whether the intervention led to a lasting clinical improvement over and above the effect of intensive sensorimotor training.

Recovery and rehabilitation
Date: Thursday 15 May 2008  
Time: 17:30 - 17:40 -  Room: Calliope
Chair: J. Bernhardt, Australia and V. Hömberg, Germany

07
The nature of fatigue following stroke and TIA.
M. Martin    S.M.Walsh    D. Martin    J.A.Harbison                                          
 

Department of Medical Gerontology, Trinity College Dublin

IRELAND

Depending on the subgroup studied and assessment tool used fatigue reportedly affects up to 70% of patients following stroke, Fatigue may be sub-classified as either physical or mental and the term fatigue may include symptoms such as excess daytime sleepiness or somnolence. We sought to define what subjects with stroke and TIA mean when they report ‘fatigue’. Methods: Subjects were recruited from a secondary prevention clinic. All subjects recruited were independent (modified Rankin Score (mRS) <3) to reduce the confounding effect of physical disability on fatigue scores. Assessment scores were completed for each subject including, the fatigue Severity Score (FSS), Epworth Sleepiness Score (ESS) Abbreviated Mental Test Score (AMTS) and a simple 5 point Lickert scales of frequency of physical and mental fatigue. Fatigue occurring ‘often’ or ‘always’ was recorded as significant. Results: 40 subjects were recruited in the pilot phase (35% Male, median age 74years). 9 (23%) subjects reported significant physical fatigue, and 4 (10%) mental fatigue. All those reporting mental fatigue also had physical fatigue. Median value for FSS was (3.6, range 1.0-6.6), for AMTS (9, range 6-10) for ESS (4.5 range 0-12). FSS correlated significantly with Likert scores for physical fatigue (r=0.447, p=0.005) and mental fatigue (r=0.351, p=0.03). FSS correlated with mRS (r=0.388, p=0.021. Spearman’s rho) but Likert fatigue scores did not. No measure of fatigue correlated significantly with ESS or AMTS. ESS however correlated with AMTS (r=-0.534, p<0.001). Conclusion: One quarter of independent subjects with stroke or TIA were found to have significant fatigue which was predominantly physical, rather than mental fatigue. Degree of somnolence was not related to fatigue severity.

Recovery and rehabilitation
Date: Thursday 15 May 2008  
Time: 17:40 - 17:50 -  Room: Calliope
Chair: J. Bernhardt, Australia and V. Hömberg, Germany

08
Mirror movements after stroke: Quantification with accelerometry and correlation with brain activation patterns
C. Calautti    S.P.Jones    M. Naccarato    N. Sharma    D.J.Day    T.A.Carpenter    E.A.Warburton    J.C.Baron              
 

Dept of Clinical Neurosciences, University of Cambridge

UNITED KINGDOM

Mirror movements (MM) of the unaffected hand (UH) when moving the affected hand (AH) are well described clinically after stroke, but their neural substrates remain the matter of controversy, with several conflicting hypotheses involving the contralesional M1, ipsilesional M1, both or neither being proposed. In addition, MM during non-effortful movement have not been quantitated using objective means so far. Aims of the present study were to quantify MM during simple index-thumb tapping (IT-taps) using tri-axial accelerometry (TAA), and to study their neural substrates using fMRI. 25 prospectively selected R-handed stroke patients (mean age 63yrs) partially recovered from L or R hemiparesis underwent TAA and fMRI during auditory-cued IT-taps at 1.25Hz. Using TAA over both index fingers we calculated a cross correlation coefficient (CCC) which reflects the occurrence of MM in the opposite hand during IT-taps of the AH (CCC-AH) and UH (CCC-UH) hand. To investigate their neural substrates, we correlated the CCC-AH with the fMRI maps using SPM. Clinical deficit was scored using the European Stroke Scale (ESS), maximum IT-taps in 15s (IT-Max) and tapping regularity index (RI). 28 normal right-handed subjects were also studied. The CCC in controls was 0.065 +/- 0.07 and 0.075 +/- 0.086 for the R and L hand, respectively (NS), without significant effects of age. The CCC-AH was significantly (p≤0.05) higher than both the CCC-UH (0.145+/-0.173 and 0.114+/-0.09, respectively) and the CCC of controls. There was no significant correlation between the CCC-AH and the ESS, time from stroke onset, IT-max and RI. No significant correlation between CCC-AH and the activation maps were found in either the contralesional or ipsilesional hemispheres. In this study, objective indices of MM during IT-taps were dissociated from both motor deficit and brain activation maps. We argue that MM after stroke could be the result of complex interactions between the individual’s built-in attentional system and transcallosal M1 connection strength, the type of task used and the effect of damage from the stroke on the motor system.

Recovery and rehabilitation
Date: Thursday 15 May 2008  
Time: 17:50 - 18:00 -  Room: Calliope
Chair: J. Bernhardt, Australia and V. Hömberg, Germany

09
Discrepancy between inner and overt speech: a potential for misdiagnosis of language impairments in post-stroke aphasia
S. Geva    E.A.Warburton                                                         
 

University of Cambridge

UNITED KINGDOM

Background: Patients with aphasia often complain that there is a poor correlation between the words they think (inner speech) and the words they say (overt speech). A few studies show that patients often attend to their intact inner speech and not their erroneous overt speech, which could explain the frequent observation that many patients do not notice their overt errors. This study tries to characterise the relation between inner and overt speech in post-stroke aphasia and its relation to error monitoring. Methods: 10 patients (8m, age: 72+/-8) with left MCA stroke and problems with speech production were studied at 32+/-21 months post-stroke. Following a structural MRI or CT scan, their language abilities were tested. Next, performance on inner speech tasks, such as homophone and rhyming judgement and recitation, was tested using words, pictures and automatic speech, and compared to overt speech errors. Results: For two patients inner speech was better preserved than overt speech, one patient had no inner speech, and the rest had similar levels of overt and inner speech. Patients relied on inner speech more than on overt speech for monitoring their speech errors. Discussion: It is generally assumed that when patients make overt speech errors this is due to their impaired language abilities and hence their subsequent therapy is based on a potential misdiagnosis. However, this study shows that inner speech can remain intact while there is a marked deficit in overt speech. Also, it is suggested that complex tasks using inner speech can be used as an improved therapy technique for patients who cannot speak, potentially leading to improved language abilities. This may have implications for diagnosis, prognosis and therapy of certain patients with post-stroke aphasia.