ARTICLE IN PRESS

E U R O P E A N J O U R N A L O F PA E D I AT R I C N E U R O L O G Y 11 (2007) 310 – 317

Official Journal of the European Paediatric Neurology Society

Original article

Seizures in Rett syndrome: An overview from a one-year

calendar study

Le Jiana, Lakshmi Nagarajanb, Nicholas de Klerka, David Ravinec,

John Christodouloud,e, Helen Leonarda,
a
Centre for Child Health Research, The University of Western Australia, Telethon Institute for Child Health Research, PO Box 855, West Perth,
WA 6872, Australia
b
Department of Neurology, Princess Margaret Hospital for Children, Roberts Road, Subiaco, Perth, Western Australia, Australia
c
West Australian Institute of Medical Research and Centre for Medical Research, The University of Western Australia, Perth, Western
Australia, Australia
d
School of Paediatrics and Child Health, University of Sydney, New South Wales, Australia
e
Western Sydney Genetics Program, Children’s Hospital at Westmead, Sydney, Australia

ar t ic l e i n f o abs tra ct

Article history:
Received 18 November 2006 Background: Rett syndrome is a neurodevelopmental disorder mainly affecting females.
Received in revised form It is principally caused by mutations in the MECP2 gene. Seizures occur in about 80% of
13 February 2007 subjects but there has been little research into the factors contributing to their frequency.
Accepted 14 February 2007 Aims: To investigate seizure frequency in Rett syndrome and its relationship with other
factors, including genetic characteristics and the use of anti-epileptic drugs.

Keywords:
Methods: Information on daily seizure occurrence and health service utilization and
monthly anti-epileptic drug use was provided on 162 Rett syndrome cases for a calendar
Rett syndrome
year. Age at onset of seizures, developmental history and other clinical and genetic
Seizure rate
characteristics were obtained from a contemporaneously completed questionnaire and
Epilepsy
from the Australian Rett Syndrome Database. Negative binomial regression was used to
MECP2 mutations
investigate factors associated with seizure rates.
Early development
Results: Seizure rates were highest in the 7–12 year age group. They were lower in those
Clinical severity
with p.R294X, p.R255X mutations and C terminal mutations. Those who had early
Anti-epileptic drugs
developmental problems and poorer mobility had higher seizure rates as did those with
greater clinical severity and poorer functional ability. Many different combinations of
medications were being used with carbamazepine, sodium valproate and lamotrigine
either singly or in combination with another being the most common.
Conclusions: Seizure frequency in Rett syndrome is age-dependent, more common in
those with more severe early developmental problems and influenced by mutation type.
& 2007 European Paediatric Neurology Society. Published by Elsevier Ltd. All rights reserved.

Corresponding author. Tel.: +61 8 9489 7790; fax: +61 8 9489 7700.
E-mail address: [email protected] (H. Leonard).
Abbreviations: AED, anti-epileptic drug; ARSD, the Australian Rett Syndrome Database; CS2000, calendar study in 2000;
FUS2000, follow-up study in 2000; NLS, the nuclear localization signal region; TRD, the transcription repression domain; SD,
standard deviation; HC, head circumference; BMI, body mass index; NHMRC, Medical and Health Research Council; APSU, the
Australian Paediatric Surveillance Unit
1090-3798/$ - see front matter & 2007 European Paediatric Neurology Society. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.ejpn.2007.02.008
 ARTICLE IN PRESS
E U R O P E A N J O U R N A L O F PA E D I AT R I C N E U R O L O G Y 11 (200 7) 310 – 3 17 311

a composite score that increases to a maximum of 126 with

1. Introduction increasing independent functioning.7 The Kerr score was
developed from the system suggested as a guideline for
Rett syndrome is a neurodevelopmental disorder mainly reporting clinical severity in Rett syndrome,8 the Pineda score
affecting females and caused principally by mutations in the from the scaling system originally developed by Monros et al.9
MECP2 gene.1 Although the phenotype is generally severe, the and the Percy score from the score modified by Schanen and
clinical spectrum is variable with a number of associated Percy10 from Amir and Zoghbi.11 In contrast to the WeeFIM,
comorbidities including reduced somatic growth, gastro-in- these numerical scores increase with increasing severity.
testinal disease, osteopenia, autonomic dysfunction and sco-
liosis. Seizures, which have a considerable impact for those 2.3. Mutation testing
affected and their families occur in about 80% of subjects. The
Australian Rett Syndrome Database (ARSD) has been using MECP2 gene mutation testing has been completed in the
multiple sources to ascertain Rett syndrome cases in Australia majority of cases (154/162, 95.1%), with pathogenic mutations
among individuals born since 1976.2 Information on functional, identified in 118 (72.6%). Details of the methods employed for
medical, educational and other aspects of the cohort have been MECP2 mutation testing have been previously described.6 X
collected every two years since 2000 through questionnaires inactivation was categorized as skewed when one X-chromo-
completed by parents and carers. During 2000 families also some was active in 25% or less of all analyzed cells.12
completed a daily calendar in which they reported episodes of
seizures, medical and other health appointments and hospita- 2.4. Data management
lizations experienced by the subjects.3 In this report we use
these contemporaneously recorded data to determine seizure To examine the relationship between parent-reported seizure
rates and to investigate their relationship with demographic, activity and genetic findings, the cases were categorized into
clinical, genetic and other factors. We also describe the range the following mutation types: p.R168X, p.T158M, p.R294X,
and combinations of anti-epileptic drugs (AEDs) being used by p.R270X, p.R255X, p.R133C, p.R306C, p.R106W, large deletions
this Rett syndrome population. involving exon 3 and 4, C-terminal deletions in exon 4, early
truncating (truncations up to and including the nuclear
localization signal (NLS) region within the transcription
2. Material and methods repression domain (TRD), except for p.R270X, p.R255X and
p.R168X), and a final group that included all other pathogenic
2.1. Data sources mutations. Separate categories were included for those in
whom a MECP2 mutation was sought but not identified and
Families and caregivers of 162 verified female cases ascer- those who were not tested.
tained from the ARSD (www.ichr.uwa.edu.au/rett/aussierett/),4 The z-score for HC in 2000 was calculated based on the
participated in a year long calendar study in 2000 (CS2000).3 following formula: (HC of child – reference mean HC)/
These cases represented 81.4% of the cases who were in the reference standard deviation (SD) of HC. The reference mean
ARSD at that time. Over three quarters (78.4%) were classical and SD were obtained from a Dutch study which provided
according to the recently revised criteria5 compared with 67% population norms.13 Z-scores for body mass index (BMI,
in a more recent cohort.4 Families were asked to record, in a kg/m2) were calculated using the formula and reference
calendar format, and return on a monthly basis information, median values for females aged 2–20 years provided by the
detailing daily events in seven categories: medical, health and Centre for Disease Control.14 The reference value for age 20
therapy appointments; hospital stays; nursing care; seizure years was also applied to any individual over this age. HC z-
activity; and other health events. The reasons for each score in 2000 and BMI z-score in 2000 were divided into three
appointment and the type of doctor or health professional categories based on the distribution of the cases in this study:
involved were recorded. Details of all prescribed and non- a reference group with z-scores ranging from 25th percentile
prescribed medications were also recorded on a monthly (P25) to 75th percentile (P75), with the other two groups below
basis. As well as completing the calendar, families also P25 and above P75.
participated in a follow-up study in 2000 (FUS2000) the Information on the use of AEDs was collected from CS2000.
methodology for which has been previously described.2 Cases were classified as users of an AED if they used it for one
Information provided by families and caregivers, for example or more months in the year. Combination(s) of AEDs used the
regarding mobility and the presence of health problems such same ‘‘no/yes’’ model. The AEDs were categorized as ‘‘old’’
as abnormal breathing and sleep disturbance as well as drugs if they belonged to the first line AEDs according to the
weight, height and head circumference (HC) measurements, specifications of the Australian Pharmaceutical Benefits
was available for this analysis. Epilepsy diagnosis was based Scheme (PBS).15 The rest (lamotrigine, topiramate, clobazam,
on the age at seizure onset as previously6 or if AEDs were vigabatrin, gabapentin and tigabine) were categorized as
being used for seizure control (n ¼ 4). ‘‘new’’ drugs. Cases using any new medication were coded
as being on new medication irrespective of the continuation
2.2. Severity scores of the additional use of old medications. To investigate the
relationship between parent-reported seizure episodes and
Several severity scores derived from the FUS2000 data2 were the pattern of AED use, those on AEDs in 2000 were also
also included in the current analysis. The modified WeeFIM is divided into two groups according to whether they had
 ARTICLE IN PRESS
312 E U R O P E A N J O U R N A L O F PA E D I AT R I C N E U R O L O G Y 11 (2007) 310 – 317

changed their AEDs during that year. If the same medication Cases with abnormal early development and those
was withdrawn but subsequently reinstated, the case was whose mobility was impaired had increased rates of
considered in the ‘‘no medication change’’ group. seizures (Table 4). Those whose parents reported abnormal

2.5. Statistical analysis Table 2 – Association of age with seizure rate in Rett
syndrome cases with epilepsy
If multiple seizures (n ¼ 31) but not the exact number (n ¼ 12)
were reported we used the average for those who had Age groups n Seizure rate 95% CI p
provided the number of multiple seizures. Annual (and (years) ratio
monthly) seizure rates were calculated by dividing the total
number of seizures recorded during the year by the number of o7 16 1.00
7–12 36 4.47 1.32–15.16 0.016
days of participation in the study. Because of the large
12–17 41 2.42 0.73–8.00 0.149
dispersion in seizure rates, negative binomial regression
417 45 1.17 0.36–3.81 0.791
was used to investigate the factors associated with seizure
rates. Seizure rate ratios (RR) were used to estimate the
associations of other variables with seizure rates. STATA
version 9.016 was used for analysis. Table 3 – Association of mutation with seizure rate in Rett
syndrome cases with epilepsy

3. Results MECP2 n Seizure rate 95% CI p

mutations ratio
One hundred and thirty eight of the 162 cases (85.2%) in No mutation 29 1.00
CS2000 had been diagnosed with epilepsy (Table 1). The mean Mutation 102 0.64 0.27–1.51 0.302
age of this group was 14.0 years (range 2.3–24.6), and of the identified
remainder 10.1 (range 1.9–24.5). Three cases died during 2000
No mutation 29 1.00
and a further nine have died since then. During the course of p.T158M 11 2.50 0.65–9.67 0.184
2000 parents/carers reported seizure episodes in 100/138 p.R168X 13 0.37 0.10–1.32 0.125
(72.5%) cases. Compared with the o7 years group, cases aged p.R294X 7 0.12 0.02–0.60 0.010
between 7 and 12 years had a higher rate of monthly seizures p.R270X 12 0.51 0.14–1.91 0.320
(Table 2). When comparing classical (n ¼ 112) with atypical p.R255X 6 0.11 0.02–0.68 0.017
p.R133C 7 0.80 0.16–4.03 0.792
(n ¼ 26) the seizure RR was 0.33 (0.13–0.79) indicating that the
p.R306C 6 0.35 0.06–1.96 0.234
seizure rate was significantly lower (p ¼ 0:013) in the classical
p.R106W 3 0.17 0.02–1.77 0.139
cases. When mutation groups were compared, there was no C terminal 13 0.15 0.04–0.54 0.004
difference in seizure rates overall between those with a Early truncatinga 8 0.62 0.14–2.87 0.545
MECP2 mutation and those in whom a mutation was not Large deletions 5 0.72 0.11–4.58 0.727
identified. However, those cases with p.R294X and p.R255X Others 11 0.53 0.14–2.05 0.358
mutations and those with mutations in the C-terminal Random XCI 54 1.00
region had significantly lower seizure rates than those Skewed XCI 43 0.55 0.23–1.33 0.185
without an identified MECP2 mutation (Table 3). The
seizure rate was lower but not significantly so in those with a
Not including p.R270X, p.R255X and p.R168X.
skewed X inactivation.

Table 1 – Proportion of Rett syndrome cases with epilepsy and monthly seizure rates in those with a diagnosis of epilepsy
by age group

Age groups Number of Rett syndrome Number of cases with an epilepsy Monthly seizure rates
(years)a cases (%) diagnosis (%)b
P10 P50 P90 mean
(95%CI)

o7 24 (14.8) 16 (66.7) 0 0.2 18.3 3.0 (2.8–3.3)

7–12 45 (27.8) 36 (80.0) 0 2.6 30.3 16.8
(16.4–17.2)
12–17 44 (27.2) 41 (93.2) 0 4.8 16.5 9.1 (8.8–9.4)
417 49 (30.3) 45 (91.8) 0 0.5 9.8 3.5 (3.4–3.7)

Total 162 (100) 138 (85.2) 0 2.1 18.3 8.7 (8.5–8.8)

a
Age in 2000.
b
10th, 50th and 90th percentile.
 ARTICLE IN PRESS
E U R O P E A N J O U R N A L O F PA E D I AT R I C N E U R O L O G Y 11 (200 7) 310 – 3 17 313

Table 4 – Association of developmental factors and markers of clinical severity with seizure rate in Rett syndrome cases
with epilepsy

No. Seizure rate ratio 95% CI p

Ambulation
Walked o18 months 33 1.00
Walked o18 months then lost/walked X18 months 22 2.12 0.69–6.52 0.188
Walked 18–30 months then lost/walked430 months 20 2.19 0.69–6.95 0.183
Never walked 52 3.69 1.49–9.15 0.005

Mobility at 10 months
Walking/crawling 32 1.00
Bottom-shuffling 32 1.46 0.53–4.00 0.462
Rolling 43 3.35 1.31–8.57 0.012
Not moving around 27 5.60 1.95–16.02 0.001

Abnormal development within 0–10 months

No problems 35 1.00
0–6 months only 44 1.67 0.65–4.24 0.284
0–6 and 7–10 months 58 2.25 0.93–5.44 0.072

Respiratory dysfunction
No 22 1.00
Yes 101 1.85 0.69–4.94 0.219

Sleep disturbance
No 15 1.00
Sleep disturbance past only 27 1.30 0.35–4.90 0.698
Sleep disturbance past and present 84 1.99 0.63–6.30 0.244

Body mass index z-score in 2000

P25 to P75 (3.23–0.36) 47 1.00
oP25 (9.08–3.23) 25 3.07 1.14–8.30 0.027
4P75 (0.36–2.82) 20 0.95 0.33–2.78 0.928

Head circumference z-score in 2000

P25 to P75 (2.18–0.53) 45 1.00
oP25 (4.88–2.18) 25 0.29 0.11–0.77 0.014
4P75 (0.53–2.18) 21 0.52 0.18–1.48 0.219

WeeFIM score 127 0.96 0.93–0.99 0.007

Kerr score 127 1.08 0.99–1.17 0.081
Percy score 127 1.08 1.01–1.16 0.021
Pineda score 127 1.11 1.03–1.21 0.010

Hospital admission 138 1.03 0.95–1.12 0.415

Total medical appointment 138 1.07 1.03–1.11 0.001
Neurological appointment 138 1.33 1.12–1.58 0.001

development both in the period up to six months and in the Compared with those whose HC z-scores were in the inter
period from seven to 10 months had over twice the rate of quartile range those whose scores were in the lowest quarter
seizures of those who had normal development up to 10 of the group were least likely to have seizures (p ¼ 0:01) but
months (p ¼ 0:07). Infants who were not moving around or this effect did not remain after age adjustment.
only rolling at 10 months had significantly more seizures than Clinical severity as measured by the Kerr, Percy and Pineda
those who were walking or crawling at 10 months. Those who scores correlated positively with seizure rates as did poorer
had never walked had seizure rates nearly four times that of functional ability as measured by the WeeFIM. Although
those who walked normally (p ¼ 0:005). seizure rates were increased in the presence of parent-
In those with seizures, BMI z-scores were available on reported breathing abnormalities and sleep problems, the
92/138 (66.7%) and HC z-scores on 91/138 (65.9%). BMI z-scores relationship was not statistically significant. One hundred
were generally low, being more than 3 SD below normal for and thirty five subjects had medical appointments, 66
over a quarter of cases. Compared with those whose BMI neurological appointments and 51 hospitalizations during
z-scores were in the inter quartile (P25–P75) range, cases with 2000. The number of medical appointments ranged from 0 to
scores in the lower quarter (oP25) had an increased rate of 40 (median 8) and the number of neurological from 0 to 15
seizures (p ¼ 0:03). One quarter of the 91 cases had HC (median 0) per subject (i.e. more than half of the subjects had
z-scores more than 2 SD below the norm for their ages. no neurological appointments). The number of days spent in
 ARTICLE IN PRESS
314 E U R O P E A N J O U R N A L O F PA E D I AT R I C N E U R O L O G Y 11 (2007) 310 – 317

hospital ranged from 0 to 94 with a median of 0 (i.e. more than 21 (19%) on three or more. Those who were on monotherapy
half of the subjects had no days in hospital). Rates of seizures with carbamazepine or sodium valproate or a combination of
were higher in those who had more medical (p ¼ 0:001) and the two had a median of one or less seizures per month. In
neurological appointments (p ¼ 0:001) but not in those with our group those on lamotrigine or combination therapy with
more hospitalizations (p ¼ 0:415). lamotrigine appeared to have higher seizure rates. Overall for
During 2000 AEDs were being prescribed to 119/138 (86.2%) the whole year the pattern of medication use varied by age
of those cases with a diagnosis of epilepsy including23 in group with younger children less likely to be using those
whom no seizures occurred during that year. The most medications classified as ‘‘new’’ such as lamotrigine, topir-
commonly used medications were sodium valproate used by amate, clobazam, vigabatrin, gabapentin and tigabine (Fig 1).
68 (57%) followed by lamotrigine, used by 52 (44%) and Those who had been prescribed ‘‘new’’ medications also had
carbamazepine used by 49 (41%) (Table 5). Diazepam pre- much higher seizure rates (po0.001) as did those whose
sumably as a ‘‘rescue’’ medication was used in 13 cases. The medication had been changed during the year (p ¼ 0:002)
109 cases on medication in the first month of the study were (Table 7).
on 39 different single or combinations of medications with
the commonest six being shown in Table 6. For 33 of the
groupings there were three or less cases on each. The 4. Discussion
commonest combinations were sodium valproate and lamo-
trigine (in 17 cases) followed by carbamazepine and sodium Although seizures are an important comorbidity in Rett
valproate (in six cases) and then carbamazepine and lamo- syndrome this is the first study to our knowledge to have
trigine (in five cases). In the first month of the study 46 (42%) defined parent-reported seizure rates in a totally ascertained
cases were on monotherapy, 42 (39%) on two medications and population of cases of Rett syndrome. This analysis combined

Table 5 – Range of anti-epileptic medications ever used during 2000 by age group

Medication(s) Rett syndrome cases ever on this medication/combinationa

0–7 years (n ¼ 12) 7–12 years (n ¼ 32) 12–17 years (n ¼ 36) 417 years (n ¼ 39) Total (n ¼ 119)

Sodium valproate 6 17 22 23 68
Lamotrigineb 1 16 21 14 52
Carbamazepine 3 13 13 20 49
Topiramateb 0 3 8 5 16
Clonazepam 1 4 4 5 14
Diazepam 1 5 1 6 13
Clobazamb 0 4 6 2 12
Vigabatrinb 0 3 3 1 7
Nitrazepam 1 3 1 0 5
Gabapentinb 1 1 1 1 4
Phenytoin sodium 1 0 1 2 4
Phenobarbital 0 1 1 0 2
Tiagabineb 0 1 0 1 2
Primidone 0 0 0 1 1

a
Including used in combination.
b
New medication.

Table 6 – Commonest AEDs (and combinations) used in January of 2000 and associated seizure rates and median age of
subjects

Medication(s) Cases on this Median age Monthly seizure rate

medication/
combination
P10 P50 P90 Mean (95% CI)

Carbamazepine 19 15.5 0.0 0.6 15.3 3.4 (3.1–3.6)

Sodium valproateþlamotrigine 17 13.8 0.1 3.8 30.3 26.2 (25.5–29.6)
Sodium valproate 16 10.5 0.0 0.2 13.7 3.4 (3.1–3.7)
Carbamazepineþsodium valproate 6 16.5 0.0 1.0 2.8 1.2 (0.9–2.0)
Lamotrigine 5 9.7 0.0 2.8 24.8 7.6 (6.9–8.3)
Carbamazepineþlamotrigine 5 14.1 0.5 5.4 30.3 9.0 (8.2–9.3)
 ARTICLE IN PRESS
E U R O P E A N J O U R N A L O F PA E D I AT R I C N E U R O L O G Y 11 (200 7) 310 – 3 17 315

30
Old AEDs New AEDs
25
25
22
21
20
17

15
11 11 11
10

5
1
0
0-7 7-12 12-17 >17
years

Fig 1 – Rett syndrome cases on new and old medications by age group.

Table 7 – Relationship between seizure rates and the use of ‘‘new’’ medications and requirement for medication change
during 2000

n Monthly seizure rate Seizure rate ratio 95% CI p

P10 P50 P90

On new medication(s)
No 55 0.0 0.5 10.0 1.00
Yes 64 0.1 4.9 30.1 3.74 1.93–7.22 0.000
a
Change medication(s)
No 92 0.0 2.1 15.3 1.00
Yes 27 0.2 4.9 35.1 3.41 1.55–7.52 0.002

a
Exclude cases using the same medication or medication combinations but off and on in various months.

information reported back by families in a calendar format to In a previous study we determined the factors influencing the
the study centre with data on other parameters ascertained age of onset of seizures in Rett syndrome.6 In this report we
from a comprehensive questionnaire completed in the same have focused on the factors influencing the frequency of
year. This combined approach yielded detailed information seizures. Seizures usually start in Rett syndrome around four
on seizure burden and associated factors in Rett syndrome, as years of age with 78.5% having seizures by 10 years.6 This is
well as AED usage, over a whole year. Using a calendar format later than in many other conditions with neurological
as previously described,3 we were also able to collect disability. We found seizure rates were highest in the 7–12
information about medical appointments and hospitaliza- age group and appeared to fall after that. This is consistent
tions which in this analysis helped to further develop the with previously reported data2 and implies that seizures in
morbidity picture of the group. Families also provided full Rett syndrome ‘‘wind down’’ in late adolescence and adult-
details of the medications being used by the subjects on a hood. Almost a quarter of Rett syndrome subjects are not
monthly basis. This allowed us to track changes being made expected to live beyond 25 years,4 with nine (6%) of cases in
to the medication regimen. Although some families did not this 2000 calendar cohort deceased by 2006. Our study is only
complete the calendar for every month we were able to adjust a cross-sectional analysis relating to one calendar year, which
for the missing time periods when calculating seizure means that it is not possible to rule out the possibility of
frequency. In this study we made the decision to accept and selection bias, particularly arising from deaths.
focus on those events families felt were seizures6 because, We found that seizure rates were reduced in subjects with
irrespective of their aetiology, these are the events that result p.R294X, p.R255X and C terminal deletions. Those with the
in a burden for them and their child. We acknowledge this most distal of these mutations (p.R294X and C terminal
may have resulted in under- or overreporting of true epileptic deletions) also had a significantly reduced seizure onset in the
events.17 first four years of life.6 Other genotype–phenotype relation-
Seizures, their unpredictability, frequency and severity, ships observed with these sorts of mutations also indicated
have a considerable impact on the quality of life for all people milder clinical presentations for other parameters.18–21
with epilepsy, but especially in conditions such as Rett Not surprisingly children with early developmental pro-
syndrome where there are also many other comorbidities. blems as well as those with greater motor disability had a
 ARTICLE IN PRESS
316 E U R O P E A N J O U R N A L O F PA E D I AT R I C N E U R O L O G Y 11 (2007) 310 – 317

higher frequency of seizures that was additional to the earlier on the comparative efficacy of the different AEDs in Rett
age of onset that we have already reported.6 It is known from syndrome.
studies of autonomic function in Rett syndrome22 that In summary, we have observed that seizure frequency in
breathing abnormalities occur almost universally, although children and women with Rett syndrome and epilepsy is age-
they may not always be noticed by parents or clinicians. What dependent. The maximal seizure frequency occurs between 7
we reported were parent-observed breathing problems which and 12 years of age with some reduction after this and this is
are probably an indication of the more severe hyperventila- information which will be useful to families and carers.
tion or apnoea. We did not find any statistically significant Seizure frequency appears to be higher in those with early
relationship between this parameter or sleep disturbance onset and more severe developmental disabilities, which may
(again a fairly gross measure) and seizure frequency. reflect a high general level of cerebral dysfunction in these
Although the BMI in Rett syndrome is generally low, those cases.
in the group with a higher BMI appeared to do better, perhaps On the basis of reports from parents and carers, seizures in
suggesting that it is important to maximize nutrition. The Rett syndrome appear to be well controlled in approximately
inverse relationship observed between head growth and a third of cases (38%), with a further 11% with an earlier
seizure rate is surprising, because it appears to conflict with diagnosis of epilepsy not using AEDs and having no seizures
the observations associated with BMI. However, the observa- during the 12 month period of the survey. The spectrum of
tion may have been confounded by age and possibly affected AEDs used in Australia in 2000 was similar to that reported
by the small sample size. The clinical severity scores from other studies.26 Whether the introduction of new AEDs
described elsewhere2 provide a composite picture of clinical such as levetiracetam and pregabalin and the increased use of
severity in Rett syndrome, with some placing more emphasis established options for epilepsy such as the vagal nerve
on the developmental and regression period and less on the stimulator27 and the ketogenic diet28 will have a significant
current status. They do, however, include reference to role in the management of seizures in Rett syndrome is yet to
epilepsy. Nevertheless, there was a significant correlation be determined. A future study using longitudinal data and
with these and seizure frequency suggesting that seizures including clinical information on seizure classification, EEG
tend to be worst in those subjects who are already generally findings and treatment approaches would aid further defini-
severe for other parameters. tion of the age-related profile that we have identified and also
The pattern of anti-convulsant use that we recorded in 2000 assist the task of clarifying the efficacy of the different AEDs.
is similar to the pattern in the report of Steffenburg et al.23
with some specific differences. In our study sodium valproate
was the most commonly used AED, while it was the
third most frequently prescribed in the Swedish study,
Acknowledgements
which reported a higher level of lamotrigine/sodium valpro-
The authors would like to acknowledge the funding of
ate combination usage than we observed. In our study
Australian Rett Syndrome research by the National Institutes
seizure rates were lowest in those on monotherapy with
of Health (1 R01 HD43100-01A1) and the National Medical and
carbamazepine or sodium valproate or on a combination of
Health Research Council (NHMRC) under project Grant
carbamazepine and sodium valproate. This might suggest
303189. HL is funded by NHMRC program Grant 353514 and
that carbamazepine and sodium valproate could have a
JC by NHMRC project Grants 185202 and 346603. Special
particular place in the treatment of seizures in Rett syn-
thanks to Alison Anderson who assisted with data manage-
drome. Caution with this interpretation is warranted because
ment and to Linda Weaving, Sarah Williamson and Mark
reduction in seizures may not be a response to the medica-
Davis for molecular work. We would also like to express our
tions but it may be that these medications are being used
gratitude to all the families who have contributed to the
more frequently in those with fewer seizures (such as those in
study; the Australian Paediatric Surveillance Unit (APSU) and
the youngest age group). Not surprisingly, perhaps, children
the Rett Syndrome Association of Australia who facilitated
with more difficult epilepsies were more likely to be on the
case ascertainment in Australia. The APSU is a Unit of the
‘‘new’’ AEDs or on polytherapy.
Division of Paediatrics, Royal Australasian College of Physi-
In total, 42% of cases, who were on AEDs, were on
cians and is funded by the Department of Health and Ageing
monotherapy, 39% were on two AED and 19% on three or
and the Faculty of Medicine of the University of Sydney.
more AED in the first month of the study. A second AED would
Ethical approval for the study has been provided by the Ethics
have been included if seizures were not well controlled on
Committee of the Women’s and Children’s Health Services in
one, so it is not surprising those on two or more AEDs had
Western Australia.
more seizures than those on one. Current practice in
Australia is to try the first line drugs before adding a ‘‘new’’
R E F E R E N C E S
agent and hence younger children were more likely to be on
the older drugs. Only 6/119 were on phenobarbitone or
phenytoin which may reflect reluctance in current practice
1. Amir RE, Zoghbi HY. Rett syndrome: methyl-CpG-binding
on the part of paediatric neurologists/paediatricians to use
protein 2 mutations and phenotype–genotype correlations.
these drugs on a long term basis in intellectually disabled
Am J Med Genet 2000;97(2):147–52.
children because of the potential cognitive and cosmetic 2. Colvin L, Fyfe S, Leonard S, et al. Describing the phenotype in
effects. Other than some small studies suggesting benefit Rett syndrome using a population database. Arch Dis Child
with lamotrigine24 and topimarate25 there are no good data 2003;88(1):38–43.
 ARTICLE IN PRESS
E U R O P E A N J O U R N A L O F PA E D I AT R I C N E U R O L O G Y 11 (200 7) 310 – 3 17 317

3. Moore H, Leonard H, de Klerk N, et al. Health service use in 15. Australian Government, DoHa Ageing [cited 2006 5th Octo-
Rett syndrome. J Child Neurol 2005;20(1):42–50. ber]; available from /http://www.health.gov.au/pbsS.
4. Laurvick CL, de Klerk N, Bower C, et al. Rett syndrome in 16. StataCorp. Stata Statistical Software. In: 9.0 R, editor. 9.0 ed.
Australia: a review of the epidemiology. J Pediatr College Station, Texas: Stata Corporation; 2005.
2006;148(3):347–52. 17. Rugg-Gunn FJ, Harrison NA, Duncan JS. Evaluation of the
5. Hagberg B, Hanefeld F, Percy A, Skjeldal O. An update on accuracy of seizure descriptions by the relatives of patients
clinically applicable diagnostic criteria in Rett syndrome. with epilepsy. Epilepsy Res 2001;43(3):193–9.
Comments to Rett Syndrome Clinical Criteria Consensus 18. Colvin L, Leonard H, de Klerk N, et al. Refining the phenotype
Panel Satellite to European Paediatric Neurology Society of common mutations in Rett syndrome. J Med Genet
Meeting, Baden Baden, Germany, 11 September 2001. Eur J 2004;41(1):25–30.
Pediatr Neurol 2002;6(5):293–7. 19. Leonard H, Moore H, Carey M, et al. Genotype and early
6. Jian L, Nagarajan L, de Klerk N, et al. Predictors of seizure development in Rett syndrome: the value of international
onset in Rett syndrome. J Pediatr 2006;149(4):542–7. data. Brain Dev 2005;27(Suppl 1):S59–68.
7. Leonard S, Msall M, Bower C, Tremont M, Leonard 20. Smeets E, Terhal P, Casaer P, et al. Rett syndrome in females
H. Functional status of school-aged children with Down with CTS hot spot deletions: a disorder profile. Am J Med Genet
syndrome. J Paediatr Child Health 2002;38(2):160–5. Part A 2005;132(2):117–20.
8. Kerr AM, Nomura Y, Armstrong D, et al. Guidelines for 21. Ager S, Fyfe S, Christodoulou J, Jacoby P, Schmitt L, Leonard H.
reporting clinical features in cases with MECP2 mutations. Predictors of scoliosis in Rett syndrome. J Child Neurol
Brain Dev 2001;23(4):208–11. 2006;21(9):809–13.
9. Monros E, Armstrong J, Aibar E, Poo P, Canos I, Pineda M. Rett 22. Julu PO, Kerr AM, Apartopoulos F, et al. Characterisation of
syndrome in Spain: mutation analysis and clinical correla- breathing and associated central autonomic dysfunction in
tions. Brain Dev 2001;23(Suppl 1):S251–3. the Rett disorder. Arch Dis Child 2001;85(1):29–37.
10. Schanen C, Houwink EJ, Dorrani N, et al. Phenotypic 23. Steffenburg U, Hagberg G, Hagberg B. Epilepsy in a represen-
manifestations of MECP2 mutations in classical and atypical tative series of Rett syndrome. Acta Paediatr 2001;90(1):
Rett syndrome. Am J Med Genet 2004;126A(2):129–40. 34–9.
11. Amir RE, Van den Veyver IB, Schultz R, et al. Influence of 24. Stenbom Y, Tonnby B, Hagberg B. Lamotrigine in Rett
mutation type and X chromosome inactivation on Rett syndrome—treatment experience from a pilot study. Eur
syndrome phenotypes. Ann Neurol 2000;47(5):670–9. Child Adolesc Psychiatry 1998;7(1):49–52.
12. Weaving LS, Williamson SL, Bennetts B, et al. Effects of MECP2 25. Goyal M, O’Riordan MA, Wiznitzer M. Effect of topiramate on
mutation type, location and X-inactivation in modulating Rett seizures and respiratory dysrhythmia in Rett syndrome.
syndrome phenotype. Am J Med Genet Part A J Child Neurol 2004;19(8):588–91.
2003;118(2):103–14. 26. Guerrini R. Epilepsy in children. Lancet 2006;367(9509):
13. Fredriks AM, van Buuren S, Burgmeijer RJ, et al. Continuing 499–524.
positive secular growth change in The Netherlands 27. Wilfong AA, Schultz RJ. Vagus nerve stimulation for treatment
1955–1997. Pediatr Res 2000;47(3):316–23. of epilepsy in Rett syndrome. Dev Med Child Neurol
14. National Center for Health Statistics 22nd June 2006 [cited 2006;48(8):683–6.
2006 30th June]; available from /http://www.cdc.gov/nchs/ 28. Liebhaber GM, Riemann E, Baumeister FA. Ketogenic diet in
about/major/nhanes/growthcharts/datafiles.htmS. Rett syndrome. J Child Neurol 2003;18(1):74–5.