JOURNAL REVIEWS

EDITORS CHOICE

Myasthenia gravis: immunoglobulins or plasma exchange

are we any wiser?
The choice of immunomodulatory therapy for neuromuscular
respiratory failure following Guillain-Barr syndrome or
myaesthenia gravis (MG) crisis has always been between intravenous immunoglobulins (IVIG) and plasma exchange
(PLEX). Evidence from small institutional series has looked at
small numbers of patients, with no discernible difference
between the two. It is in this context that this 6-year retrospective population-based cohort study looking into over 1600 MG
patients by Mandawat et al, gains importance.The authors did
an inpatient database search from 1000 acute-care hospitals in
the United States from 2000 to 2005, identifying 6034 patients
with MG or MG crisis, based on the ICD-9 coding system. After
excluding patients who were transferred between hospitals
(thus avoiding duplicate data), and those with a prolonged
hospital stay for more than a year, 1606 patients (900 in MG
crisis) were identified who received either IVIG or PLEX.
Patients who received both treatments were not included in
the analysis. Although MG patients not in crisis were similar in
age, gender and comorbidities between the IVIG and PLEX
groups, patients with MG crisis receiving PLEX were more
likely to be male (45.37% vs 29.59%, p<0.001), likely to stay in
hospital longer (10 vs 5 days, p<0.001) and were more likely to
have been discharged to a rehabilitation facility (18.34% vs
11.24%, p<0.001). The authors ascertain that patients in MG
crisis who receive PLEX had a significantly higher unadjusted
mortality rate than those given IVIG (5.67% vs 0.59%, p<0.01).
However, when this was adjusted for covariates, there was no
difference in the mortality or complication rates. Acute respiratory failure, cardiac complications and acute renal failure
were associated with an increase in mortality. IVIG was better

Parkinsons disease: nature and neurturin

As the disease progresses, patients accumulate motor disability due to

nigrostriatal degeneration,typically developing dyskinesias due in part to
increasing levodopa requirements. In recent years, many researchers
have turned their attention to developing therapies capable of
minimising these long-term motor complications, either by surgically
implanting dopamine-producing neurones into the striatum or by
changing the genetics of the neurons involved in the damaged circuitry.
Gene delivery is exciting because it has the potential to deliver longlasting expression of a particular protein or growth factor to the brain
following a single surgical procedure. One such growth factor is neurturin, an analogue of glial cell line derived neurotrophic factor (GDNF),
that has been shown to protect dopaminergic neurones in animal
models of Parkinsons disease (PD). A vector capable of carrying the
gene for neurturin (adeno-associated vector type 2;AAV2) was studied in
a recent double-blind, randomised trial published in Lancet Neurology.
A North American collaboration,supported by the company Ceregene,
reported the safety and efficacy of stereotactic injection of AAV2-neurturin versus sham surgery in patients with advanced PD (disease duration
around 10 years,Hoehn andYahr stage 3).The vector genomes were delivered to the putamen bilaterally through frontal burr holes,with the expectation that some of the protein would be transported from the striatum to
nigral neurones. Patients undergoing sham surgery underwent an identical procedure, except the dura was not breached and the intracranial
injections were not done.
The study found that AAV2-neurturin treated patients (n=38) and
sham surgery patients (n=20) did not differ in terms of the primary study

44 > ACNR > VOLUME 10 NUMBER 6 > JANUARY/FEBRUARY 2011

tolerated in the elderly and in those with complex comorbid

conditions, including acute respiratory failure. The authors
conclude that patients receiving IVIG have a shorter hospital
stay and less total inpatient costs.
Essentially, this tells us what we already know IVIG and
PLEX are as effective as each other and the choice depends on
availability and comorbid diseases. Economically, IVIG has
always been thought to be more expensive at the point of
delivery, and although the prices have come down recently,
supply can be erratic. This study suggests that IVIG is cheaper
based on the overall hospital costs. But there is an inherent
bias in reaching this assumption - since there were almost 9
times more patients in the PLEX group who had acute respiratory failure than in the IVIG group (181 vs 21, p<0.0001). It is
possible that the most unwell patients were more likely to
receive PLEX than IVIG and would have had a longer length of
stay irrespective of the treatment given. This reinforces the
common perception among neurologists that PLEX may be
marginally superior to IVIG in MG crisis, when the patient is
severely unwell. This remains a gut feeling and robust Class I
evidence comparing the overall clinical effect and economical impact of these two therapies is unlikely to materialize in
the near future.
Saiju Jacob, Consultant Neurologist, Queen Elizabeth
Neurosciences Centre, University Hospitals of Birmingham
NHS Foundation Trust, Edgbaston, Birmingham, UK.
Mandawat A, et al. Comparative analysis of therapeutic
options used for myasthenia gravis.
ANNALS OF NEUROLOGY 2010;68:797-805.

endpoint change in motor subscore of Unified Parkinsons Disease

Rating Scale (UPDRS III) from baseline to 12 months. However, the treatment group did show a modest, but significant, improvement compared
with the sham surgery group at 18 months in a subgroup analysis (least
squares mean change from baseline to 18 months of -11.96 in the AAV2neurturin group compared with -4.34 in the sham surgery group). After
looking at available autopsy specimens, the authors proposed that the
delayed improvement may be due to impaired or delayed transport of
the trophic factor to the substantia nigra. As well as longer-term followup,the authors plan to carry out further trials using direct injections into
the substantia nigra, perhaps combined with higher putaminal concentrations.
This trial is an important early step in establishing whether gene
therapy has a place in the long-term treatment of motor complications in
PD. Future results will need to show improved outcomes if this approach
is to compete with deep brain stimulation and cell transplantation, not
only in terms of effectiveness clinical and cost - but also safety.The risk
of tumour induction using trophic factors is a particular concern.Whilst
one patient treated with AAV2-neurturin in this study was found to have
a glioblastoma, this was later felt to have been present on brain imaging
prior to study entry and probably unrelated to the AAV2-neurturin
(although the potential for accelerated tumour growth due to the trophic
factor must be considered). Other serious adverse events related to the
surgical procedure in the AAV2-neurturin group all of which were selflimiting included confusion, seizure, haemorrhage, cerebral oedema,
and caudate nucleus infarct (all in one patient),transient mental change
(in one patient), and urinary retention (in two patients).

JOURNAL REVIEWS

Over the next few years, it will be fascinating to see how successful
each of the surgical interventions will be in demonstrating and communicating their efficacy and safety record. Deep brain stimulation has a
head-start but research funding and pharmaceutical backing will help
cell transplantation and gene therapy to stake their claim. This type of
competition will only benefit PD patients in the long run. One must
realise, however, that these treatments may never be able to treat the
extra-nigral manifestations of PD (e.g. memory and neuropsychiatric
problems) and we still need to better understand and treat these.
David P Breen, Clinical Research Fellow in Neurology,
Cambridge Centre for Brain Repair, University of Cambridge.
Marks WJ Jr, et al. Gene delivery of AAV2-neurturin for Parkinsons
disease: a double-blind, randomised, controlled trial.
LANCET NEUROLOGY 2010;9:1164-72.

Pathogenic antibodies to intracellular

antigens?

Over the last ten years,there has been an increase in the number of neurological syndromes associated with highly specific autoantibodies
directed against proteins expressed in the central nervous system, with
well-established antigenic targets such as ionotropic receptors (eg.NMDA
receptors) and components of the voltage-gated potassium channel
complex (Lgi1, Caspr2). The antibodies bind to cell-surface exposed
determinants and are presumed to be pathogenic. By contrast, current
dogma suggests that antibodies to intracellular targets, such as typical
onconeural antigens,are not pathogenic since the antigen will be unavailable to circulating antibodies. However, Geis et al studied antibodies to
amphiphysin, an intracellular antigen involved in synaptic vesicle recycling, in the pathogenesis of paraneoplastic stiff person syndrome (SPS).
Immunoglobulin fractions from two patients with SPS were intrathecally
infused for 15 days, thus circumventing the blood brain barrier. Rats
developed some of the clinical features of the disease including muscle
stiffness and spasms, and these features were not seen in appropriate
control animals. Electrophysiological investigations showed reduction of
the Hoffman reflex and dorsal root potential amplitudes.In vitro neuronal
preparations demonstrated highly specific uptake of amphiphysin antibodies into the nerve terminals, and a reduction in stimulated GABA
release, and to a lesser extent, glutamate release.These extensive findings
provide strong evidence for the pathogenicity of amphiphysin antibodies
in SPS, but interesting questions arise: amphiphysin antibodies were
derived from the patients' plasma,would it be possible to recapitulate the
pathophysiological features by peripheral administration of immunoglobulins? And what is the mechanism by which the antibodies are taken up
into the nerve terminals? This study sets a benchmark for further demonstrations of pathogenic effects of antibodies to both intracellular and
extracellular antigens in neurological disease.
Philippa Pettingil, Research Fellow, Department of Clinical
Neurology, Oxford University.
Geis C, et al. Stiff person syndrome-associated autoantibodies to
amphiphysin mediate reduced GABAergic inhibition.
BRAIN 2010;133:3166-80, with editorial by Angela Vincent,
BRAIN 2010;133:3164-5.

Alzheimers Disease: evidence for a

neuroprotective effect of an
apolipoprotein E genetic polymorphism

The association between APOE genetic polymorphism and differing risks

of developing Alzheimers disease has been well documented. While
APOE4 allele,which is present in 25% of the general population,has been
associated with more rapid rates of memory decline, the reports
regarding the role of much less prevalent (5% of population) APOE2

allele have been inconsistent. One study has reported a protective role
for this allele due perhaps to greater entorhinal cortical thickness in the
adolescents studied,whereas a few others have associated APOE2 carrier
status with increased loss of mnemonic neural substrates. Armed with
the neuroimaging data and CSF analysis results of the Alzheimers
Disease Neuroimaging Initiative (ADNI), this study aimed to test the
hypothesis that compared to cognitively normal APOE3/3 carriers,
APOE2/2 and APOE2/3 carriers demonstrate reduced rates of
hippocampal atrophy and episodic memory decline and less immunobiochemical evidence of Alzheimers related pathological processes in
their CSF.Analysis of the longitudinal data obtained from 1.5 T MRI scans
using Freesurfer software showed evidence of hippocampal volume loss
in both groups with a statistically significant greater degree of atrophy in
APOE3/3 carriers.The MRI findings were not supported by the neuropsychological evidence as there was a longitudinal improvement in the
performance of both groups which was attributed to practice effect.CSF
findings however, were compatible with the assumed neuroprotective
role of APOE2 carrier status.
The authors discuss that the findings of their study support the
assumed neuroprotective role for APOE2.They postulate that decreased
Alzheimers neuropathological changes, evidenced by the CSF findings,
might be the mechanism behind the reduced hippocampal atrophy in
APOE2 carriers.Two major caveats to this hypothesis are also presented:
i) a recent post mortem study among a population aged greater than 90
found increased Alzheimers pathology but a reduced risk of clinical
dementia amongst APOE2 carriers which suggests a mechanism other
than decreased neuropathological changes for the reduced risk of
dementia in this group; and ii) APOE2 has been associated with an
increased risk of cerebral amyloid angiopathy leading to lobar haemorrhage, and the authors remind us that such patients would have been
excluded from the ADNI study due to abnormal screening MRI which in
turn could have confounded the inferences made. Regardless of mechanisms involved the study provides objective evidence of reduced
hippocampal atrophy rates amongst careers of APOE2 allele compared
to the prevalent APOE3/3 carrier status.
Seyed Sajjadi.Clinical Research Fellow and Honorary SpR in
Neurology. Herchel Smith Building for Brain and Mind Sciences and
Neurology Unit, Addenbrooke's Hospital.
Chiang et al. Hippocampal atrophy rates and CSF biomarkers in elderly
APOE2 normal subjects. NEUROLOGY 2010;75:1976-81.

Tourette Syndrome: habit reversal therapy

There is a new vogue for behavioural therapies in tic disorders which has
been driven by Doug Woods from Milwaukee and supported primarily by
the US Tourette Syndrome Association. A number of factors lie underneath this renaissance, the chief amongst which is the poor tolerability of
most anti-tic medications, even though these drugs can definitely
suppress tics (but then, so does a general anaesthetic). Extra-pyramidal
side-effects, obesity, impaired glucose tolerance and school refusal are
commonly encountered with dopamine blocking agents of all types. This
has made the search for alternative forms of therapy more urgent. A
further factor is a paradigm shift amongst Tourette doctors and lay organisations which has enabled the Tourette establishment to accept a nonbiological form of treatment for the first time: much of the early
campaigning work was based around using a biological model to prevent
blaming of children and parents for the occurrence of tics. It is interesting
to note theriderat the end of this paper which states....acknowledging
that behavioural and learning processes play a role in tic severity does
not imply that tics have a purely psychological etiology or that patients
can suppress tics by force of will. The authors are defending people with
Tourette Syndrome (TS) against the charge that this is just a bad habit
andthey could stop doing it if only they tried hard enough. This is ironic
in that habit reversal (the basis of the behavioural therapy) is pretty much
a technique for trying really hard not to tic.
The trial itself is excellently constructed and executed. The basic
finding of the RCT is that behavioural therapy (based on habit reversal) is

ACNR > VOLUME 10 NUMBER 6 > JANUARY/FEBRUARY 2011

> 45

JOURNAL REVIEWS

superior to supportive therapy (listening to patients talking about tics and

generally supporting them) in around 120 patients with mild to moderate
TS. Habit reversal involves performing a competing movement to the tic
and holding this until the urge to tic subsides. The therapy also involves
concentrating on the urge feelings and trying to develop a mastery over
them. Effect sizes are reasonable (slightly below those of the drug trials)
with a much better tolerability than the drugs in other trials.
There are some quibbles about the methodology, its difficult to see
how the trial was blinded. For instance it would be interesting to see
how the two arms of the trial were presented in the patient information
sheet.The behavioural treatment appears to be much more intense and
time-consuming than the control intervention and, therefore, vulnerable
to bias. Habit reversal is clearly not a treatment for the workers. Around
80% of all the trial participants came from a higher professional background (which may reflect access to US health insurance). The treatment and control groups did not appear to be well-matched for antipsychotic use (13.1% in the treatment group compared to 4.6% in the
control group).Perhaps the most valid outcome in the trial is theimpairment score on the Yale Global Tourette Syndrome Severity Scale.This is
only one question worth 50 marks covering the effect on school, occupation and social function.The treated group improved from a score of
25 to 12.2 over 10 weeks.The control group improved from 23.4 to 16.4.
The improvements in both groups suggest a strong placebo component
but the effect size of 0.57 is reasonable.
Behavioural therapy is clearly developing a place in the management
of tic disorders and this trial is a further building block in that process.
Clinics seeking to provide specialist services for people with tic disorders should be aiming to provide behavioural therapy as part of a range
of treatment options. However, this requires the training and supervision
of allied health professionals. At the moment I know of only 3 such practitioners in the UK.
Hugh Rickards. Hon. Reader in Neuropsychiatry, Birmingham
University and Consultant in Neuropsychiatry, Birmingham and
Solihull Mental Health Foundation Trust.
Piacentini J, et al. Behavioural Therapy for Children with Tourette
Disorder: A Randomised Controlled Trial.
JOURNAL OF THE AMERICAN MEDICAL ASSOCIATION 2010;303:1929-37.

Tourette Syndrome: greying on top

Brain imaging research in neuropsychiatric research is an esoteric and

problematic field. Adequate controls are difficult to find, biases are
common, particularly in relation to ascertainment, multiple comparisons,
medication and co-morbidity, and causative chains are hard to establish.
Finally, one can sometimes get the feeling that a new piece of kit is
routinely tried out on a series of complex conditions, which may not be a
bad thing. This paper is no exception. The authors used two different
methods for examining brain structure in Tourette Syndrome (TS); cortical
thickness estimation and voxel-based morphometry. They chose adults
with TS and healthy controls to compare with and found a reduction in
grey matter in the orbito-frontal, anterior cingulate and venterolateral prefrontal cortices.The cortex also appeared thinner in the mesial temporal
lobe.Cortical volume was increased in the primary somatosensory cortex
associated with higher premonitory sensation intensity.
There are some limitations to this paper. The sample are adults, who
represent the few children with TS who didnt grow out of it, and attend a
hyper-specialised clinic where previous research has indicated a high level
of personality disorder,also associated with structural brain abnormalities.
Over half of the sample had been medicated (probably chronically in most
cases).The authors have made some assumptions in relation to causation
which are not supported by their data. For instance, they have argued that
the reductions in grey matter are primary (i.e. reflecting developmental
abnormality) and the increases (for instance in the somatosensory cortex)
are a reflection of neural plasticity. This implies a causal relationship and
is not warranted. In the end, this paper does not tell us much that is new,
only that adult TS brains are somehow different to controls. Whether these
differences are the primary cause of the disorder, a reflection of the
disorder or an epiphenomenon remains to be seen. But, this is not to
dismiss the findings out of hand. The data represent a starting point for
46 > ACNR > VOLUME 10 NUMBER 6 > JANUARY/FEBRUARY 2011

understanding the shape of the brain in TS.They need careful replication.

The whole field of TS needs longitudinal cohort studies to tease out the
causal relationships. Ideally, we should explore the longitudinal study of
those people at risk of developing the disorder (i.e.those with affected first
degree relatives). Until that happens, progress will probably be slow.
Hugh Rickards. Hon. Reader in Neuropsychiatry, Birmingham
University and Consultant in Neuropsychiatry, Birmingham and
Solihull Mental Health Foundation Trust.
Draganski B, et al. Multispectral brain morphometry in Tourette
syndrome persisting into adulthood. BRAIN 2010;133:3661-75.

BDNF or not BDNF that is the

question in Huntingtons Disease

The cause of Huntington's disease (HD) is known to be due to expression

of mutant huntingtin, but how this causes the actual cell loss is less clear
especially given that whilst all cells express the mutant protein, only a few
develop pathology. In the early part of this century, Elena Cattaneo and
colleagues showed that the mutant gene may be exerting its effects
through BDNF expression in the corticostriatal pathway. The evidence
supporting this hypothesis has slowly accumulated over the years, and
part of the attraction of the theory is that abnormalities in BDNF expression are known to occur in patients with HD and that abnormalities in its
expression could help explain some of the regional pathology seen in the
brains of patients with this disorder. In this new paper, Xie et al show that
they can rescue the phenotype on the YAC128 mouse model of HD by
overexpressing BDNF in the forebrain using a double transgenic
approach, in which BDNF is put in under the promotor for the alpha
subunit of Ca2+/calmodulin dependent protein kinase II. They show that
this overexpression not only rescues the behavioural phenotype but also
rescues the mouse at the histological and biochemical level. This study
therefore adds further weight to the theory that abnormalities in BDNF
signalling are critical to the expression of some features of HD.Whether it
accounts for all of them is less clear as is the route by which we can
exploit this fact for therapeutic gain in patients. However, knowing this
means there is hope for patients with HD. The challenge now is knowing
how we can do this effectively as has been tried to some extent using exercise and diet both of which are known to effect levels of this
neurotrophic factor in the CNS.
Roger Barker.
Xie Y, Hayden MR, Xu B. BDNF overexpression in the forebrain rescues
Huntington's Disease phenotytpes in YAC128 mice. J NEUROSCIENCE
2010;30:14708-18.

The Lynx1 effect

These authors from Harvard have identified a novel restrictor of CNS plasticity (akin to the axonal growth inhibiting chondroitin sulfate proteoglycans) relevant to visual cortex, which could be a therapeutic target for
amblyopia, with potential for the method to be taken further afield. The
molecule, Lynx1, is structurally similar to -bungarotoxin, and binds
directly to nicotinic acetylcholine receptors. Importantly, the protein was
identified through analysis of a transcriptome of mouse V1 visual cortex
(16,209 full-length cDNA clones: for light reading turn to Plessy et al.PLoS
ONE 2008), with special attention to the binocular zone of V1 and for
molecules expressed later than the early critical period of visual plasticity. Lynx1 knockout mice deprived of monocular vision for a short or
long time in maturity (>postnatal day 60) demonstrated electrophysiological evidence of reoorganisation and plasticity that wildtype mice did
not,and return of normal visual acuity measured by visual evoked potentials. The plasticity was not at the structural level of perineural nets or
increased myelination, and was achieved also purely by cholinergic
agonism in wildtypes, leading the authors to hypothesise that the mechanism is at the level of attentional control.
Mike Zandi.
Morishita, H. et al. Lynx1, a Cholinergic Brake, Limits Plasticity in Adult
Visual Cortex. SCIENCE 2010;330:1238-40.