- Email: [email protected]
Rett variants: A suggested model for inclusion criteria
Review Articles
Rett Variants: A Suggested Model for Inclusion Criteria B e n g t A. H a g b e r g , M D * a n d O l a H . S k j e l d a l , MD*
A model for the clinical delineation of atypical cases of Rett syndrome is presented. It is based on the presence, at age >t10 years, of combined clusters of at least 3 of 6 primary criteria and at least 5 of 11 supportive manifestations appearing through childhood with advancing age. The model was applied to 16 mentally retarded females, aged 11--47 years (median: 23) who were considered to manifest atypical variants of the syndrome (8 formes frustes, 6 late regression, 2 congenital variants). Two of the 16 patients had an early seizure history as the initial abnormality. In parallel, the number of supportive manifestations in a series of 41 females over 10 years of age with classic Rett syndrome are given. The differentiating power was tested on 8 patients with a chromosome-verified Angelman syndrome. It is concluded that the model applied here has the capacity to identify and distinguish Rett syndrome variants of different types, to sift out other developmental disorders in routine clinical work, and to have potential as a useful research tool. Hagberg BA, Skjeldal OH. Rett variants: A suggested model for inclusion criteria. Pediatr Neurol 1994; 11:5-11.
Introduction Females with Rett syndrome (RS) appear to represent a more heterogeneous phenotype than was first realized. Increasing experience in recent years indicates that a number of clinical variants do exist and that the RS concept represents a broader and larger group than was previously imagined [1]. Transitions in clinical presentation between subtypes are numerous [2] but some main types emerge (Table 1). As the cause of RS is unknown and diagnostic biological markers are lacking, we must focus on clusters of characteristic clinical criteria in distinguishing the key group of classic RS [1,3,4]. To delineate RS variants, diagnostic support can be obtained from the presence of combinations of particular manifestations. Each of them may be seen in other conditions but rarely in distinct combinations.
From the Department of Pediatrics; Ostra Sjukhuset; G6teborg; Sweden; t Department of Pediatrics; Rikshospitalet; Oslo, Norway.
© 1994 by Elsevier Science Inc. • 0887-8994/94/$7.00
In a collaborative Swedish-Norwegian project, a model for RS variants has been constructed and applied to a pilot series of atypical RS cases. This model is based on combined clusters of primary and supportive criteria, the latter of which are particularly informative as evidenced by long clinical experience [1,2].
Methods Study Design. It is well known that a peculiar cluster of odd behavioral developmental and neurologic deviations appears with increasing age in the majority of classic RS schoolgirls and adolescents [ 1,3]. For patients with a high clinical suspicion of RS, but who do not fulfill the classic criteria, we constructed a list of criteria for the diagnosis of atypical RS. The model was simultaneously applied to 22 Swedish females and 19 Norwegian females with classic RS, aged 10-52 years (median: 20). The Swedish patients had been reevaluated by one of the authors during the past 3 years. The Norwegian patients had been evaluated by the other author during the past 8 months. As in most clinically derived series, it was not possible to obtain complete coverage of information on all the items in every single patient. Information on all 11 supportive items was available in 23 of the 41 patients with classic RS. The median number of supportive items was 9 (range: 7-11). The 6 primary and 11 supportive criteria suggesting atypical RS are listed in Table 2. The primary criteria (A) consist of 6 items. The first 5 items (A1-A5) have been modified from the corresponding ones for classic RS, now without age limitations and including atypical hand stereotypies. The well-known, long-term RS disease profile has been added as item A6. It consists of a somewhat improved communicative and perceptive ability with increasing age, contrasting with the slowly decreasing/deteriorating gross neuromotor function [1]. The supportive criteria (B) comprise: (B1) irregular breathing with episodic hyperventilation and/or breath-holding, only present while awake [1]; (B2) air swallowing of bloating type [1]; (B3) teeth grinding with the characteristic creaking sound [1]; (B4) obvious, dyspraxic gait problems in late childhood and adolescence; (B5) double-curved kypho-scoliosis of neurogenic type or high thoracic kyphosis in the case of preserved good walking ability [1]; (B6) successively developing distal lower limb neurologic abnormalities with dystonic plantar-flexed toes and/or malpositioned cavus feet [l]; (B7) small, bluish-red feet, cold and abiotrophic [1]; (B8) characteristic electroencephalographic (EEG) development [1,3]; (B9) episodic daytime and/or nighttime laughter, or alternatively, violent screaming spells [1,3]; (B10) indication of impaired/delayed reactions to pain stimuli (i.e., deranged nociceptive functions [1,3]); (B 1 l) intensive eye communication often developed to a primitive "language" of useful eye pointing [1]. Atypical RS patients were arbitrarily required to fulfill at least 3 of the 6 main (A) criteria and at least 5 of the 11 supportive ones (B). The exclusion criteria were those of the Rett Diagnostic Criteria Work Group
[4].
Communications should be addressed to: Dr. Hagberg, Department of Pediatrics; Ostra Sjukhuset; S-416 85 G6teborg, Sweden. Received December 17, 1993; accepted May 13, 1994.
Hagberg and Skjeldal: Rett Syndrome---Diagnosis of Variants
5
Table 1.
Rett syndrome phenotype expressions
I. Classic female RS ([1,3,4]) \ Subtype: Early seizure onset / H. Atypical female RS [2] 1. Formes frustes (FF) a) Original (/>13 years old) b) Tentative (10-13 years old) 2. Late childhood regression 3. Preserved speech 4. Congenital 5. Other 111. Male RS Abbreviation: RS = Rett syndrome
Subjects. The pilot series comprised 16 females, aged 11-47 years (median: 23), with clinical indications suggestive of phenotypic variations of RS, 14 from the updated Swedish RS series [5] and 2 from a recently initiated national Norwegian RS project. Six of the 16 patients ($30, $48, $90, S121, S127, S128) were of the age-defined traditional forme fruste (FF) type [6,7], the remaining 10 fulfilled neither the criteria of classic RS [8] nor those of traditional FF. The majority of patients included have been followed since the mid-1980s. The differentiating diagnostic power of the supportive (B) criteria was tested on 8 patients with verified Angelman syndrome (15q abnormality).
Results The clusters of primary (A) and supportive (B) criteria found in the pilot series of atypical RS are presented in Table 3. The presence or absence of classic RS criteria [4,8] is shown in Table 4. It is notable that, at the last follow-up, a microcephalic head circumference (<2 S.D. for age) was noted in only 2 of the 16 atypical patients, one FF ($48) and one congenital variant (N59). Table 5 shows the distribution of supportive (B) criteria in the pilot series as well as in the series of classic RS. None of the patients with Angelman syndrome fulfilled more than 3 of the supportive criteria. Detailed data have previously been given for the Swedish patients $30 [7], S 117 [2], and S 120 [2,9] included in the pilot series. Three additional illustrative RS variants with the model applied are presented in condensed form below. Patient $143. This patient is a 14-year-old Swedish girl with unrelated parents. She was born at term; the perinatal history revealed no abnormalities. Psychomotor development was convincingly normal during the first 6 months of life, followed by a discrete general stagnation during the next 6 months. Between 1 and 2V2 years of age, a slow loss of communicative ability and acquired words was observed and the girl was regarded as autistic. On the paternal side, a 15-year-old female second cousin has classic RS. RS was suspected at 8 years of age during initial exam-
6
PEDIATRIC NEUROLOGY
Vol. 11 No. 1
ination at Ostra Sjukhuset. However, she did not lulfill the classic criteria but conformed to diagnosis of ~sitnple mental retardation" and "infantile autism." At present (14 years of age) she is 161 cm long and has a normal build (head circumference 56 cm), is autistic yet communicative, and jumps around aggressively and agitatedly. She communicates, particularly with her mother, with the characteristic intense RS eye pointing. She has a pincer grasp but overall hand dyspraxia, marked breathing irregularities with hyperventilation broken by long breathholding periods, all very irregular and of the RS type; and bloating mannerisms. Her history reveals long, sudden, and completely unexpected screaming episodes of fear and/or rage. Her feet are normal in size, but are trophically impaired with intermittent profuse sweating. Additional abnormal neurology (mainly ataxic signs and tremor) developed during the 6 years after the first examination and she is now convincingly showing the characteristic RS disease profile (item A6). Comments. At age 14 years, this girl demonstrates a convincing cluster of RS manifestations and a characteristic disease profile. She is considered to represent an FF RS variant, certainly dyspraxic but with surprisingly wellpreserved motor abilities. According to the model, she Table 2.
Rett syndrome: A variant delineation model
Inclusion criteria: A girl of at least 10 years of age with mental retardation of unexplained origin and with at least 3 of the 6 following primary criteria: A1 Loss of (partial or subtotal) acquired fine finger skill in late infancy/early childhood A2 Loss of acquired single words/phrases/nuanced babble A3 RS hand stereotypies, hands together or apart A4 Early deviant communicative ability A5 Deceleration of head growth of 2 S.D. (even when still within normal limits) A6 The RS disease profile: A regression period (stage II) followed by a certain recovery of contact and communication (stage III) in contrast to slow neuromotor regression through school age and adolescence and, in addition, at least 5 of the following 11 RS supportive manifestations: B1 Breathing irregularities (hyperventilation and/or breath-holding) B2 Bloating/marked air swallowing B3 Characteristic RS teeth grinding B4 Gait dyspraxia B5 Neurogenic scoliosis or high kyphosis (ambulant girls) B6 Development of lower limb neurologic abnormalities B7 Small blue/cold impaired feet, autonomic/trophic dysfunction B8 Characteristic RS electroencephalographic development B9 Unprompted sudden laughing/screaming spells B 10 Impaired/delayed nociception B11 Intensive eye communication--"eye pointing" Exclusion criteria according to the Diagnostic Criteria Work Group [4] Abbreviation: RS = Rett syndrome
Table
3.
Distribution
Case L a s t s e e n at a g e (yr) Atypical type
of variant
and
supportive
criteria
in a pilot study
of 16 females
with
atypical
Rett
syndrome
S30 19
$48 25
S90 30
Sl17
S120
S121
S122
S123
S126
S127
S128
S130
S142
S143
N21
N22
25
30
32
47
21
27
21
36
12
14
14
11
12
FF
FF
LR, ES
LR
LR
LR
LR
C
LR
FF
FF
FF
FF
FF
C
FF, ES
+
+
+
+
+
+
+
-
+
NA
+
+
+
+
-
+
+
+
+
+
+
+
+
-
+
+
NA
+
-
+
-
+
AI.3RS h a n d stereotypies AI.4 Loss of communicat i r e ability
(+)A
(+)A
(+)A
+
+
+
(+)A
+
(+)A
+
+
+
(+)A
(+)A
+
+
-
+
-
+
+
NA
+
-
+
NA
+
-
+
+
-
+
AI.5 Deceleration o f
-
+
NA
NA
NA
-
-
-
NA
-
NA
-
-
-
+
+
+
+
+
+
+
+
+
+
+
+
-t-
+
+
+
+
+
+
-
--
+
--
+
--
+
+
+
+
+
+
+
-
+ +
+
+
--
+
+
+
--
+
--
-t-
+
+
--
--
-
-
NA
+
+
-
-
-
+
+
-
+
+
( + )P
+
+
-
+
(+)P
+
N
+
+
+
+
+
-
N
-
+
+
+
+
+
+
+
+
+
+
+
+
+
-
+
+
+
+
+
+
+
+
--
+
.
+
--
+
+
+
+
+
+
+
--
-t-
+
+
+
--
+
+
+
+
+
NA
+
-
NA
+
NA
NA
NA
+
NA
NA
NA
NA
-
NA
pattem B9 Laugh-
-
+
+
+
+
+
+
+
+
+
+
-
+
+
-t-
+
ing/ screami n g spells B10 Noci-
+
+
-
+
-
+
--
+
-
NA
NA
+
+
+
+
NA
+
+
+
+
-
+
+
+
+
+
+
+
+
+
Variant criteria AI. 1 Loss of finger skill AI.2 Loss of babble/ speech
head growth AI.6 RS disease profile
Supporting criteria B 1 Breathing i r r e g ularities B2 Bloating B3 Teeth grinding B4 Gait dyspraxia B5 S c o l i o sis/high kyphosis B6 Lower l i m b abnormalities B7 Feet disturbances B8 RS
.
.
.
EEG
ception impaired B 11 R S e y e + + pointing Abbreviations: A = Atypical C = Congenital EEG = Electroencephalographic
ES FF LR
= Early seizure onset = Forme fruste = Late regression
N NA P
= N e v e r able to w a l k = Data incomplete/not available = P r e v i o u s l y a b l e to w a l k , not at last f o l l o w - u p
Hagberg and Skjeldal: Rett Syndrome--Diagnosis
of Variants
7
Table 4.
Distribution o f classic criteria in a pilot study o f 16 females with atypical Rett s y n d r o m e
Case Last seen at age (yr) Atypical type
$30 19
$48 25
$90 30
Sl17 25
S120 30
S121 32
S122 47
S123 21
S126 27
S127 21
S128 36
S130 13
S142 14
S143 14
N2! 11
N22 12
FF
FF
LR, ES
LR
LR
LR
LR
C
LR
FF
FF
FF
FF
FF
C
FF, ES
+
+
+
+
+
NA
+
+
+
+
+
+
+
+
+
+
(+ )
+
-
+
+
+
--
__
+
+
+
--
+
+
-
+
+
+
NA
+
+
+
+
NA
NA
+
+
+
+
NA
+
+
-
+
NA
NA
NA
NA
-
-
NA
+
NA
-
+
-
+
-
-
(+ )
-
--
NA
NA
+
+
+
-
+
+
+
-
-
+
-
+
--
+
--
+
-
+
-
-
NA
NA
-
+
+
-
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
A
A
A
+
+
+
A
+
A
+
+
+
A
A
+
+
( + )P
+
+
-
+
(+)P
+
N
+
+
+
+
+
-
+
-
Classical criteria Pre- and perinatal period uneventful Early development normal Head circumference at birth normal Decelerating head growth Early loss of hand skills Early loss of babble/ speech Early loss of communicative ability Appearance of mental deficiency Hand stereotypies Gait abnormalities
Abbreviations: A = Atypical C = Congenital ES = Early seizure onset
FF = Forme fruste LR = Late regression N = Never able to walk
fulfills 4 of the 6 main criteria and 6 of the 11 supportive ones. Patient N21. This patient is an 11-year-old Norwegian girl born at term with a younger sister born showing a very similar phenotype. The parents are thought to be unrelated. Pre- and perinatal history revealed no abnormalities. At 1 month of age, she displayed weak sucking, hypotonia, and failure to thrive. At 4 months, microcephaly was observed and she was suspected of having some type of "cerebral palsy." When she was 2-3 years of age, a general developmental delay was diagnosed. She had no speech but good social contact through intense communication by eye pointing. Between 9 and 11 years of age, RS-like hand steremtypies appeared. In addition, finger dyspraxia was observed. At 11 she is nonambulatory and hyperventilates episodically with irregular breath-holding in between. There is no bloating but there is intense teeth
8
PEDIATRIC NEUROLOGY
Vol. 11 No. 1
N A = Data incomplete/not available P = Previously able to walk, not at last follow.up
grinding of the creaking RS type. Her feet are small, blue, and cold, and trophicaUy impaired. Distal lower limb neurologic abnormalities are present with plantar flexion dystonia, particularly in the toes. She has spastic tendon reflexes, scoliosis of the neurogenic type, and impaired/ delayed nociception. The intense communication by eye pointing is impressive. Comments. At 11 years of age, this girl presents a convincing cluster of RS-related symptoms, signs, and peculiarities. She has, however, never been considered to be completely normal. She is regarded to represent a congenital RS variant. Using the applied model she fulfiUs 3 of the 6 main RS criteria and 9 of the 1 t supportive criteria. Her younger sister at age 4 years fulfills 3 of the 6 main criteria and 8 of the 11 supportive ones. Patient 3122. This patient is a 14-year-old Norwegian girl with unrelated parents. She was born at term, birth
Table 5. Representation of supportive criteria in a pilot series of 16 atypical RS cases 11-47 years of age (median 23) and in a series of 41 classic RS cases 10-52 years of age (median 20)
B1 B2 B3 B4 B5 B6 B7 B8 B9 BI0 Bll
Breathing irregularities Bloating/marked air swallowing Characteristic RS teeth grinding Gait dyspraxia Neurogenic scoliosis/high kyphosis Development of additional lower limb abnormalities Small blue/cold impaired feet Characteristic EEG development Laughing/screaming spells Impaired/delayed nociception Intensive eye communication
Classic Series
Pilot Series
37/41 23/41
90% 56%
1 1 / 1 6 69% 9/16 56%
35/39
90%
34/37 36/40
92% 90%
1 1 / 1 5 73% 1 5 / 1 6 94%
38/41
93%
1 0 / 1 6 63%
36/41
88%
1 4 / 1 6 88%
19/27
70%
33/40 26/37 41/41
83% 70% 100%
8/15
3/6
53%
50%
1 4 / 1 6 88% 9/13 69% 1 5 / 1 6 94%
Abbreviation: EEG = Electroencephalographic
weight 3,400 gm and head circumference 35 cm. Preand perinatal history revealed no abnormalities. Normal development was seen during the first 7 months. At 8-9 months of age she began to have frequent mainly generalized tonic-clonic, epileptic seizures. When she was 12 months old, she spoke a few words and was able to walk a few steps without support. She had frequent epileptic seizures despite anticonvulsant treatment. Her EEG was of the generalized epileptic type. Through age 1-2 years, developmental regression occurred with the loss of acquired skills, decreasing communicability, and autistic traits. No deceleration of head growth was observed. At age 12 years, head circumference was still normal (55 cm). She demonstrated a convincing RS picture with typical hand stereotypies, marked hyperventilation with irregular breath-holding, and characteristic RS teeth grinding. She showed no bloating. There was partial hand dyspraxia. Her gross motor skills were surprisingly good. She was also able to jump on a trampoline, even to ski crosscountry, but not to use poles simultaneously with the lower limb movements. No spastic or dyskinetic signs were noted, but she had cavus feet. They were small, cold, blue-red in color, and trophically impaired. There was a slight S-shaped scoliosis of the neurogenic type. At age 14 years, intense communication by eye pointing and sudden unexpected and long-lasting laughing episodes are present. Comments. At age 14 years, this girl presents a convincing cluster of RS symptoms, signs, and peculiarities. When the model is applied, she fulfills 5 of the 6 primary criteria for atypical RS and 7 of the 11 supportive ones. She is considered to represent FF RS with an early seizure onset symptomatology.
Discussion The international criteria for defining classic RS were originally drawn up in Vienna in September 1984 [8] and were then slightly modified by the professional advisory board of the International Rett Syndrome Association and the Centers for Disease Control in the United States [4]. These criteria were intentionally restrictive so as not to become lost when searching for causation. It was, however, obvious from the outset that the biological concept might be considerably wider. The first atypical RS to be added was the FF type, defined with a lowest inclusion age of 13 years and presented at the International Rett Conference in Baltimore in November 1985 [5]. At the same conference a variety of other atypical RS case reports were discussed. Gouti~res and Aicardi reported 5 girls displaying an absence of normal development during the first months of life [10]. Back in 1984 in Vienna, Hanefeld had described 1 RS girl with late onset infantile spasms [11] and this was followed by a number of reports of RS patients with early seizure onset [2,12]. Over the years, an additional array of more atypical RS presentations has been reported [2]. Atypical RS thus constitutes a wide panorama of different phenotypes, some quite difficult to recognize and delineate in the clinical setting. A systematic analysis of such variants is needed. The construction of the presented model and its application in a pilot study is an attempt to address the diagnostic problem. The grouping of atypical RS now presented (Table 1) should be seen as provisional in the ongoing process of delineating RS phenotypes. It could be argued that early seizure onset RS does not constitute a specific subgroup but is due to constitutional variability in the convulsion threshold [12]. Nevertheless, infantile seizures are abnormal and the postnatal normality during the first 5-6 months of life, which is required for classic RS, is thus not displayed. Moreover, this subgroup might contain important information as to pathology, treatment, and long-term prognosis. As it occurs in both otherwise classic and atypical RS we have regarded it as a separate subtype, until it can be further elucidated. For this particular study, the FF subgroup has been further divided by age into one group above 13 years of age according to the original definition [6] and one group 10-13 years of age, in order to lower the age limit for the FF concept. With the broader diagnostic model that has now been applied, we consider the lower age limit of 10 years to be possible. Most of the FF females in the present series have been observed for years but it was not previously possible to accept them as RS because of an atypical presentation, which was not compatible with the original classic criteria. In particular, some have shown a surprisingly well-preserved, yet somewhat dyspraxic, hand function, as well as the absence of the classic handwringing RS stereotypies. Moreover, in the vast majority the head circumference curves have remained within normal limits
Hagberg and Skjeldal: Rett Syndrome----Diagnosis of Variants
9
and it has not been possible to establish an early convincing deceleration. The late regression subgroup represents fairly recent experience [2] and appears to be an important subject for further evaluation in clinical RS research. The same applies to the more controversial preserved-speech subgroup which was previously emphasized by Zapella [13] and recently reviewed [2]. The congenital subgroup (group II:5 in Table 1) still raises particular diagnostic problems [21. In addition to the presence of primary criteria (A), the model constructed for inclusion as atypical RS depends on the clustering of supportive manifestations (B) known to be peculiar to the condition [1,3]. We consider the existence of a cluster of at least 5 supportive deviations to be essential for the model. A prominent complex profile of multiple symptoms, usually not fully evident until late childhood, provides the diagnostic power required to reveal potential RS cases even when combined with an incomplete array of the obligatory criteria otherwise required within the concept of classic RS [4,8]. Of necessity, the items which have been chosen have a variable power of specificity. Thus, for example, the weight of diagnostic importance differs considerably between breathing irregularities and decreased or delayed nociception (items B1 and B10 in Table 2). The unique disturbed breathing regulation in RS girls, noted only when awake, is a quite specific RS manifestation as thoroughly delineated by Kerr and collaborators [14], whereas the inadequate reaction to pain stimuli is more nonspecific, infrequent, and less studied, but is nevertheless peculiar when present. Apart from clusters of behavioral oddities (e.g., creaking bruxism, peculiar laughing and screaming episodes, bloating, and eye pointing), the subsequent appearance with age of additional distal lower limb neurologic abnormality is an important clinical finding when present. Specific abnormalities on EEG have not been observed in RS. However, a characteristic stepwise abnormal EEG development, starting with normality during the first years of life, has been found [15]. This pattern (item B8), which develops over time, has a supportive diagnostic power which we believe will be increasingly useful with the current increasing awareness of RS behind developmental regression in young girls. In this pilot series of atypical RS, however, it was only possible to evaluate the EEG development over time in a few patients because the data available for analysis were either too fragmentary or nonexistent. We therefore had not 11 but only 10 supportive items to use in most patients. The diagnostic model is considered to have the capacity to identify and distinguish relatively different types of atypical RS and to exclude other clinical conditions with developmental abnormalities. Its differentiating power was tested on a series of 8 children with Angelman syndrome who had a partly RS-like clinical presentation. All had a verified chromosome 15q abnormality. These children appeared to have at most 3 of the 11 supportive (B)
10 PEDIATRICNEUROLOGY Vol. 11 No. 1
criteria, usually gait abnormality (however, more ccrebellar ataxia than dyspraxia), and unexpected laughing episodes (items B4 and B9 in Table 2). Impaired nociception (item B10 in Table 2) was indicated in 3 of the 8 children. In addition, 1 of the girls with Angelman syndrome displayed RS-like handwringing (item A3 in Table 2). The model was found to be particularly suitable tor identifying mild, protracted, and initially indistinct RS clinical phenotypes, the most vague FF and the late regression variants. Congenital onset phenotypes, with early severe motor disabilities, remained more difficult to identify. For epidemiologic research, this model, added to the original one delineating classic RS [4,8] and with criteria defined since 1985 [8], was considered to be valuable in the present process of mapping geographic RS clusterings, like those revealed by previous Swedish experience [16]. Parallel to this, the model can certainly help to visualize additional familial RS cases of discrete and easily overlooked types [17]. Thus, attention can now be focused on unexplained and unspecific cases of mentally retarded females reported as existing in the same family. For diagnostic application, we would again emphasize the necessity of the presence of the characteristic cluster of a number of the supporting (B) criteria. The presence of single RS pecularities (e.g., isolated handwringing) should be regarded with great caution. Moreover, the sensitivity and specificity of the model is dependent on meticulous and initiated history-taking, detailed clinical observations, and repeated follow-up evaluations. Not least important is the age of the girl. Our experience indicates that some atypical RS patients, regarded as representing cases of unspecific mental deficiency in early childhood, have only convincingly demonstrated a full array of peculiar RS symptoms late in childhood (after age 9-10 years). Our preliminary message from the pilot series of atypical Swedish and Norwegian RS cases is as follows. The FF type appears to be the most prevalent atypical expression of RS. In these girls, the RS supportive characteristics often do not appear until late childhood. The late childhood regression type (i.e., with a regression period delayed until school age and usually with RS symptoms which develop fairly slowly) appears not to be extremely rare but is overlooked by the medical profession. In their early years, these patients are often considered to represent cases of so-called "simple mental retardation" of unexplained origin or to have an undifferentiated label of "infantile autism." The proportion of FF and late regression phenotypes taken together appears to approach onefifth of all RS, as judged from our Swedish series updated as of 1993 and now comprising 155 RS females aged 2-50 years (April 1994). An early seizure onset phenotype characterizes 7% of all RS in the Swedish series, the majority of whom represent classic RS with a few FF or late regression phenotypes (e.g., our case $90 [Table 3] with initial seizures and muted RS symptoms over an extended
period). The congenital variant, revealed in relatively few Swedish and Norwegian RS females, is difficult to distinguish and cannot usually be diagnosed accurately without many years of detailed observation and follow-up. This is particularly true when a later regression is indistinct or apparently lacking. It was not our intent to present a complete and fully tested methodology nor to show that the model has proven diagnostically reliable. Our intention has been to present a diagnostic model which has functioned well in a routine neuropediatric setting. Increasing the requirement to at least 6 or 7 supportive inclusion criteria could be an alternative for a future strict scientific design. In conclusion, with no diagnostic biologic RS marker yet available, our model was found to be useful as a complement to provide better insight into the RS concept. It has to be tested on independent series of both cases suggestive of atypical RS and those with other well-defined developmental disorders. We thank Alan K. Percy, MD, for a most valuable dialogue during the construction and application of the model. The study was supported by the W:6 Rett Research Programme and Gunnar and M/irta Bergendahl's Foundation, G6teborg, Sweden; and by the Co-ordination Council for the Mentally Retarded, Norway. References
[1] Hagberg B. Clinical criteria, stages and natural history. In: Hagberg B, ed. Rett syndrome~linical and biological aspects. Clinics in developmental medicine No. 127. London: Mac Keith Press, 1993;4-20. [2l Hagberg B, Gillberg C. Rett variants---Rettoid phenotypes. In: Hagberg B, ed. Rett syndrome-421inical and biological aspects. Clinics in developmental medicine No. 127. London: Mac Keith Press, 1993;40-60.
[3] Hagberg B. Rett syndrome: Clinical peculiarities, diagnostic approach and possible cause. Pediatr Neurol 1989;5:75-83. [4] Trevathan E, Moser HW. Diagnostic criteria for Rett syndrome. Ann Neurol 1988;23:425-8. [5] Hagberg B, Witt Engerstr6m I, The Swedish series of females with Rett syndrome 1960-92. In: Hagberg B, ed. Rett syndrome-Clinical and biological aspects. Clinics in developmental medicine No. 127. London: Mac Keith Press, 1993;21-5. [6] Hagberg B, Witt-EngerstrOm I. Rett syndrome: A suggested staging system for describing impairment profile with increasing age towards adolescence. Am J Med Genet 1986;24(suppl 1):47-59. 17] Hagberg B, Rasmussen P. "Forme fruste" of Rett syndrome--A case report. Am J Med Genet 1986;24:175-81. [8] Hagberg B, Gouti~res F, Hanefeld F, Rett A, Wilson J. Rett syndrome: Criteria for inclusion and exclusion. Brain Dev (Tokyo) 1985; 7:372-3. [9] Gillberg C. The borderland of autism and Rett syndrome: Five case histories to highlight diagnostic difficulties. J Autism Dev Disord 1989; 19:545-59. [10] Gouti~res F, Aicardi J. Atypical forms of Rett syndrome. Am J Med Genet 1986;24:184-94. [11] Hanefeid F. The clinical pattern of the Rett syndrome. Brain Dev 1985;7:320-5. [12] Hagberg B, Witt Engerstr6m I. Rett syndrome: Epidemiology and nosology--Progress in knowledge 1986---A conference communication. Brain Dev 1987;9:451-7. [13] Zapella M, The Rett girls with preserved speech. Brain Dev 1992;14:98-101. [14] Kerr AM. A review of the respiratory disorder in Rett syndrome, Brain Dev 1992;14 (suppl):S43-5. [15] Hagne I, Witt Engerstr6m I, Hagberg BA. EEG development in Rett syndrome. A study of 30 cases. Electroencephalogr Clin Neurophysiol 1989;72:1~5. [16] /~kesson HO, Hagberg B, Wahlstr6m J, Witt Engerstr6m I. Rett syndrome: A search for gene sources. Am J Med Genet 1992;42: 104-10. [17] Anvret M, Wahlstr6m J, Skogsberg P, Hagberg B. Segregation analysis of the X-chromosome in a family with Rett syndrome in two generations. Am J Med Genet 1990;37:31-5.
Hagberg and Skjeldal: Rett Syndrome--Diagnosis of Variants
11
Rett Variants: A Suggested Model for Inclusion Criteria B e n g t A. H a g b e r g , M D * a n d O l a H . S k j e l d a l , MD*
A model for the clinical delineation of atypical cases of Rett syndrome is presented. It is based on the presence, at age >t10 years, of combined clusters of at least 3 of 6 primary criteria and at least 5 of 11 supportive manifestations appearing through childhood with advancing age. The model was applied to 16 mentally retarded females, aged 11--47 years (median: 23) who were considered to manifest atypical variants of the syndrome (8 formes frustes, 6 late regression, 2 congenital variants). Two of the 16 patients had an early seizure history as the initial abnormality. In parallel, the number of supportive manifestations in a series of 41 females over 10 years of age with classic Rett syndrome are given. The differentiating power was tested on 8 patients with a chromosome-verified Angelman syndrome. It is concluded that the model applied here has the capacity to identify and distinguish Rett syndrome variants of different types, to sift out other developmental disorders in routine clinical work, and to have potential as a useful research tool. Hagberg BA, Skjeldal OH. Rett variants: A suggested model for inclusion criteria. Pediatr Neurol 1994; 11:5-11.
Introduction Females with Rett syndrome (RS) appear to represent a more heterogeneous phenotype than was first realized. Increasing experience in recent years indicates that a number of clinical variants do exist and that the RS concept represents a broader and larger group than was previously imagined [1]. Transitions in clinical presentation between subtypes are numerous [2] but some main types emerge (Table 1). As the cause of RS is unknown and diagnostic biological markers are lacking, we must focus on clusters of characteristic clinical criteria in distinguishing the key group of classic RS [1,3,4]. To delineate RS variants, diagnostic support can be obtained from the presence of combinations of particular manifestations. Each of them may be seen in other conditions but rarely in distinct combinations.
From the Department of Pediatrics; Ostra Sjukhuset; G6teborg; Sweden; t Department of Pediatrics; Rikshospitalet; Oslo, Norway.
© 1994 by Elsevier Science Inc. • 0887-8994/94/$7.00
In a collaborative Swedish-Norwegian project, a model for RS variants has been constructed and applied to a pilot series of atypical RS cases. This model is based on combined clusters of primary and supportive criteria, the latter of which are particularly informative as evidenced by long clinical experience [1,2].
Methods Study Design. It is well known that a peculiar cluster of odd behavioral developmental and neurologic deviations appears with increasing age in the majority of classic RS schoolgirls and adolescents [ 1,3]. For patients with a high clinical suspicion of RS, but who do not fulfill the classic criteria, we constructed a list of criteria for the diagnosis of atypical RS. The model was simultaneously applied to 22 Swedish females and 19 Norwegian females with classic RS, aged 10-52 years (median: 20). The Swedish patients had been reevaluated by one of the authors during the past 3 years. The Norwegian patients had been evaluated by the other author during the past 8 months. As in most clinically derived series, it was not possible to obtain complete coverage of information on all the items in every single patient. Information on all 11 supportive items was available in 23 of the 41 patients with classic RS. The median number of supportive items was 9 (range: 7-11). The 6 primary and 11 supportive criteria suggesting atypical RS are listed in Table 2. The primary criteria (A) consist of 6 items. The first 5 items (A1-A5) have been modified from the corresponding ones for classic RS, now without age limitations and including atypical hand stereotypies. The well-known, long-term RS disease profile has been added as item A6. It consists of a somewhat improved communicative and perceptive ability with increasing age, contrasting with the slowly decreasing/deteriorating gross neuromotor function [1]. The supportive criteria (B) comprise: (B1) irregular breathing with episodic hyperventilation and/or breath-holding, only present while awake [1]; (B2) air swallowing of bloating type [1]; (B3) teeth grinding with the characteristic creaking sound [1]; (B4) obvious, dyspraxic gait problems in late childhood and adolescence; (B5) double-curved kypho-scoliosis of neurogenic type or high thoracic kyphosis in the case of preserved good walking ability [1]; (B6) successively developing distal lower limb neurologic abnormalities with dystonic plantar-flexed toes and/or malpositioned cavus feet [l]; (B7) small, bluish-red feet, cold and abiotrophic [1]; (B8) characteristic electroencephalographic (EEG) development [1,3]; (B9) episodic daytime and/or nighttime laughter, or alternatively, violent screaming spells [1,3]; (B10) indication of impaired/delayed reactions to pain stimuli (i.e., deranged nociceptive functions [1,3]); (B 1 l) intensive eye communication often developed to a primitive "language" of useful eye pointing [1]. Atypical RS patients were arbitrarily required to fulfill at least 3 of the 6 main (A) criteria and at least 5 of the 11 supportive ones (B). The exclusion criteria were those of the Rett Diagnostic Criteria Work Group
[4].
Communications should be addressed to: Dr. Hagberg, Department of Pediatrics; Ostra Sjukhuset; S-416 85 G6teborg, Sweden. Received December 17, 1993; accepted May 13, 1994.
Hagberg and Skjeldal: Rett Syndrome---Diagnosis of Variants
5
Table 1.
Rett syndrome phenotype expressions
I. Classic female RS ([1,3,4]) \ Subtype: Early seizure onset / H. Atypical female RS [2] 1. Formes frustes (FF) a) Original (/>13 years old) b) Tentative (10-13 years old) 2. Late childhood regression 3. Preserved speech 4. Congenital 5. Other 111. Male RS Abbreviation: RS = Rett syndrome
Subjects. The pilot series comprised 16 females, aged 11-47 years (median: 23), with clinical indications suggestive of phenotypic variations of RS, 14 from the updated Swedish RS series [5] and 2 from a recently initiated national Norwegian RS project. Six of the 16 patients ($30, $48, $90, S121, S127, S128) were of the age-defined traditional forme fruste (FF) type [6,7], the remaining 10 fulfilled neither the criteria of classic RS [8] nor those of traditional FF. The majority of patients included have been followed since the mid-1980s. The differentiating diagnostic power of the supportive (B) criteria was tested on 8 patients with verified Angelman syndrome (15q abnormality).
Results The clusters of primary (A) and supportive (B) criteria found in the pilot series of atypical RS are presented in Table 3. The presence or absence of classic RS criteria [4,8] is shown in Table 4. It is notable that, at the last follow-up, a microcephalic head circumference (<2 S.D. for age) was noted in only 2 of the 16 atypical patients, one FF ($48) and one congenital variant (N59). Table 5 shows the distribution of supportive (B) criteria in the pilot series as well as in the series of classic RS. None of the patients with Angelman syndrome fulfilled more than 3 of the supportive criteria. Detailed data have previously been given for the Swedish patients $30 [7], S 117 [2], and S 120 [2,9] included in the pilot series. Three additional illustrative RS variants with the model applied are presented in condensed form below. Patient $143. This patient is a 14-year-old Swedish girl with unrelated parents. She was born at term; the perinatal history revealed no abnormalities. Psychomotor development was convincingly normal during the first 6 months of life, followed by a discrete general stagnation during the next 6 months. Between 1 and 2V2 years of age, a slow loss of communicative ability and acquired words was observed and the girl was regarded as autistic. On the paternal side, a 15-year-old female second cousin has classic RS. RS was suspected at 8 years of age during initial exam-
6
PEDIATRIC NEUROLOGY
Vol. 11 No. 1
ination at Ostra Sjukhuset. However, she did not lulfill the classic criteria but conformed to diagnosis of ~sitnple mental retardation" and "infantile autism." At present (14 years of age) she is 161 cm long and has a normal build (head circumference 56 cm), is autistic yet communicative, and jumps around aggressively and agitatedly. She communicates, particularly with her mother, with the characteristic intense RS eye pointing. She has a pincer grasp but overall hand dyspraxia, marked breathing irregularities with hyperventilation broken by long breathholding periods, all very irregular and of the RS type; and bloating mannerisms. Her history reveals long, sudden, and completely unexpected screaming episodes of fear and/or rage. Her feet are normal in size, but are trophically impaired with intermittent profuse sweating. Additional abnormal neurology (mainly ataxic signs and tremor) developed during the 6 years after the first examination and she is now convincingly showing the characteristic RS disease profile (item A6). Comments. At age 14 years, this girl demonstrates a convincing cluster of RS manifestations and a characteristic disease profile. She is considered to represent an FF RS variant, certainly dyspraxic but with surprisingly wellpreserved motor abilities. According to the model, she Table 2.
Rett syndrome: A variant delineation model
Inclusion criteria: A girl of at least 10 years of age with mental retardation of unexplained origin and with at least 3 of the 6 following primary criteria: A1 Loss of (partial or subtotal) acquired fine finger skill in late infancy/early childhood A2 Loss of acquired single words/phrases/nuanced babble A3 RS hand stereotypies, hands together or apart A4 Early deviant communicative ability A5 Deceleration of head growth of 2 S.D. (even when still within normal limits) A6 The RS disease profile: A regression period (stage II) followed by a certain recovery of contact and communication (stage III) in contrast to slow neuromotor regression through school age and adolescence and, in addition, at least 5 of the following 11 RS supportive manifestations: B1 Breathing irregularities (hyperventilation and/or breath-holding) B2 Bloating/marked air swallowing B3 Characteristic RS teeth grinding B4 Gait dyspraxia B5 Neurogenic scoliosis or high kyphosis (ambulant girls) B6 Development of lower limb neurologic abnormalities B7 Small blue/cold impaired feet, autonomic/trophic dysfunction B8 Characteristic RS electroencephalographic development B9 Unprompted sudden laughing/screaming spells B 10 Impaired/delayed nociception B11 Intensive eye communication--"eye pointing" Exclusion criteria according to the Diagnostic Criteria Work Group [4] Abbreviation: RS = Rett syndrome
Table
3.
Distribution
Case L a s t s e e n at a g e (yr) Atypical type
of variant
and
supportive
criteria
in a pilot study
of 16 females
with
atypical
Rett
syndrome
S30 19
$48 25
S90 30
Sl17
S120
S121
S122
S123
S126
S127
S128
S130
S142
S143
N21
N22
25
30
32
47
21
27
21
36
12
14
14
11
12
FF
FF
LR, ES
LR
LR
LR
LR
C
LR
FF
FF
FF
FF
FF
C
FF, ES
+
+
+
+
+
+
+
-
+
NA
+
+
+
+
-
+
+
+
+
+
+
+
+
-
+
+
NA
+
-
+
-
+
AI.3RS h a n d stereotypies AI.4 Loss of communicat i r e ability
(+)A
(+)A
(+)A
+
+
+
(+)A
+
(+)A
+
+
+
(+)A
(+)A
+
+
-
+
-
+
+
NA
+
-
+
NA
+
-
+
+
-
+
AI.5 Deceleration o f
-
+
NA
NA
NA
-
-
-
NA
-
NA
-
-
-
+
+
+
+
+
+
+
+
+
+
+
+
-t-
+
+
+
+
+
+
-
--
+
--
+
--
+
+
+
+
+
+
+
-
+ +
+
+
--
+
+
+
--
+
--
-t-
+
+
--
--
-
-
NA
+
+
-
-
-
+
+
-
+
+
( + )P
+
+
-
+
(+)P
+
N
+
+
+
+
+
-
N
-
+
+
+
+
+
+
+
+
+
+
+
+
+
-
+
+
+
+
+
+
+
+
--
+
.
+
--
+
+
+
+
+
+
+
--
-t-
+
+
+
--
+
+
+
+
+
NA
+
-
NA
+
NA
NA
NA
+
NA
NA
NA
NA
-
NA
pattem B9 Laugh-
-
+
+
+
+
+
+
+
+
+
+
-
+
+
-t-
+
ing/ screami n g spells B10 Noci-
+
+
-
+
-
+
--
+
-
NA
NA
+
+
+
+
NA
+
+
+
+
-
+
+
+
+
+
+
+
+
+
Variant criteria AI. 1 Loss of finger skill AI.2 Loss of babble/ speech
head growth AI.6 RS disease profile
Supporting criteria B 1 Breathing i r r e g ularities B2 Bloating B3 Teeth grinding B4 Gait dyspraxia B5 S c o l i o sis/high kyphosis B6 Lower l i m b abnormalities B7 Feet disturbances B8 RS
.
.
.
EEG
ception impaired B 11 R S e y e + + pointing Abbreviations: A = Atypical C = Congenital EEG = Electroencephalographic
ES FF LR
= Early seizure onset = Forme fruste = Late regression
N NA P
= N e v e r able to w a l k = Data incomplete/not available = P r e v i o u s l y a b l e to w a l k , not at last f o l l o w - u p
Hagberg and Skjeldal: Rett Syndrome--Diagnosis
of Variants
7
Table 4.
Distribution o f classic criteria in a pilot study o f 16 females with atypical Rett s y n d r o m e
Case Last seen at age (yr) Atypical type
$30 19
$48 25
$90 30
Sl17 25
S120 30
S121 32
S122 47
S123 21
S126 27
S127 21
S128 36
S130 13
S142 14
S143 14
N2! 11
N22 12
FF
FF
LR, ES
LR
LR
LR
LR
C
LR
FF
FF
FF
FF
FF
C
FF, ES
+
+
+
+
+
NA
+
+
+
+
+
+
+
+
+
+
(+ )
+
-
+
+
+
--
__
+
+
+
--
+
+
-
+
+
+
NA
+
+
+
+
NA
NA
+
+
+
+
NA
+
+
-
+
NA
NA
NA
NA
-
-
NA
+
NA
-
+
-
+
-
-
(+ )
-
--
NA
NA
+
+
+
-
+
+
+
-
-
+
-
+
--
+
--
+
-
+
-
-
NA
NA
-
+
+
-
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
A
A
A
+
+
+
A
+
A
+
+
+
A
A
+
+
( + )P
+
+
-
+
(+)P
+
N
+
+
+
+
+
-
+
-
Classical criteria Pre- and perinatal period uneventful Early development normal Head circumference at birth normal Decelerating head growth Early loss of hand skills Early loss of babble/ speech Early loss of communicative ability Appearance of mental deficiency Hand stereotypies Gait abnormalities
Abbreviations: A = Atypical C = Congenital ES = Early seizure onset
FF = Forme fruste LR = Late regression N = Never able to walk
fulfills 4 of the 6 main criteria and 6 of the 11 supportive ones. Patient N21. This patient is an 11-year-old Norwegian girl born at term with a younger sister born showing a very similar phenotype. The parents are thought to be unrelated. Pre- and perinatal history revealed no abnormalities. At 1 month of age, she displayed weak sucking, hypotonia, and failure to thrive. At 4 months, microcephaly was observed and she was suspected of having some type of "cerebral palsy." When she was 2-3 years of age, a general developmental delay was diagnosed. She had no speech but good social contact through intense communication by eye pointing. Between 9 and 11 years of age, RS-like hand steremtypies appeared. In addition, finger dyspraxia was observed. At 11 she is nonambulatory and hyperventilates episodically with irregular breath-holding in between. There is no bloating but there is intense teeth
8
PEDIATRIC NEUROLOGY
Vol. 11 No. 1
N A = Data incomplete/not available P = Previously able to walk, not at last follow.up
grinding of the creaking RS type. Her feet are small, blue, and cold, and trophicaUy impaired. Distal lower limb neurologic abnormalities are present with plantar flexion dystonia, particularly in the toes. She has spastic tendon reflexes, scoliosis of the neurogenic type, and impaired/ delayed nociception. The intense communication by eye pointing is impressive. Comments. At 11 years of age, this girl presents a convincing cluster of RS-related symptoms, signs, and peculiarities. She has, however, never been considered to be completely normal. She is regarded to represent a congenital RS variant. Using the applied model she fulfiUs 3 of the 6 main RS criteria and 9 of the 1 t supportive criteria. Her younger sister at age 4 years fulfills 3 of the 6 main criteria and 8 of the 11 supportive ones. Patient 3122. This patient is a 14-year-old Norwegian girl with unrelated parents. She was born at term, birth
Table 5. Representation of supportive criteria in a pilot series of 16 atypical RS cases 11-47 years of age (median 23) and in a series of 41 classic RS cases 10-52 years of age (median 20)
B1 B2 B3 B4 B5 B6 B7 B8 B9 BI0 Bll
Breathing irregularities Bloating/marked air swallowing Characteristic RS teeth grinding Gait dyspraxia Neurogenic scoliosis/high kyphosis Development of additional lower limb abnormalities Small blue/cold impaired feet Characteristic EEG development Laughing/screaming spells Impaired/delayed nociception Intensive eye communication
Classic Series
Pilot Series
37/41 23/41
90% 56%
1 1 / 1 6 69% 9/16 56%
35/39
90%
34/37 36/40
92% 90%
1 1 / 1 5 73% 1 5 / 1 6 94%
38/41
93%
1 0 / 1 6 63%
36/41
88%
1 4 / 1 6 88%
19/27
70%
33/40 26/37 41/41
83% 70% 100%
8/15
3/6
53%
50%
1 4 / 1 6 88% 9/13 69% 1 5 / 1 6 94%
Abbreviation: EEG = Electroencephalographic
weight 3,400 gm and head circumference 35 cm. Preand perinatal history revealed no abnormalities. Normal development was seen during the first 7 months. At 8-9 months of age she began to have frequent mainly generalized tonic-clonic, epileptic seizures. When she was 12 months old, she spoke a few words and was able to walk a few steps without support. She had frequent epileptic seizures despite anticonvulsant treatment. Her EEG was of the generalized epileptic type. Through age 1-2 years, developmental regression occurred with the loss of acquired skills, decreasing communicability, and autistic traits. No deceleration of head growth was observed. At age 12 years, head circumference was still normal (55 cm). She demonstrated a convincing RS picture with typical hand stereotypies, marked hyperventilation with irregular breath-holding, and characteristic RS teeth grinding. She showed no bloating. There was partial hand dyspraxia. Her gross motor skills were surprisingly good. She was also able to jump on a trampoline, even to ski crosscountry, but not to use poles simultaneously with the lower limb movements. No spastic or dyskinetic signs were noted, but she had cavus feet. They were small, cold, blue-red in color, and trophically impaired. There was a slight S-shaped scoliosis of the neurogenic type. At age 14 years, intense communication by eye pointing and sudden unexpected and long-lasting laughing episodes are present. Comments. At age 14 years, this girl presents a convincing cluster of RS symptoms, signs, and peculiarities. When the model is applied, she fulfills 5 of the 6 primary criteria for atypical RS and 7 of the 11 supportive ones. She is considered to represent FF RS with an early seizure onset symptomatology.
Discussion The international criteria for defining classic RS were originally drawn up in Vienna in September 1984 [8] and were then slightly modified by the professional advisory board of the International Rett Syndrome Association and the Centers for Disease Control in the United States [4]. These criteria were intentionally restrictive so as not to become lost when searching for causation. It was, however, obvious from the outset that the biological concept might be considerably wider. The first atypical RS to be added was the FF type, defined with a lowest inclusion age of 13 years and presented at the International Rett Conference in Baltimore in November 1985 [5]. At the same conference a variety of other atypical RS case reports were discussed. Gouti~res and Aicardi reported 5 girls displaying an absence of normal development during the first months of life [10]. Back in 1984 in Vienna, Hanefeld had described 1 RS girl with late onset infantile spasms [11] and this was followed by a number of reports of RS patients with early seizure onset [2,12]. Over the years, an additional array of more atypical RS presentations has been reported [2]. Atypical RS thus constitutes a wide panorama of different phenotypes, some quite difficult to recognize and delineate in the clinical setting. A systematic analysis of such variants is needed. The construction of the presented model and its application in a pilot study is an attempt to address the diagnostic problem. The grouping of atypical RS now presented (Table 1) should be seen as provisional in the ongoing process of delineating RS phenotypes. It could be argued that early seizure onset RS does not constitute a specific subgroup but is due to constitutional variability in the convulsion threshold [12]. Nevertheless, infantile seizures are abnormal and the postnatal normality during the first 5-6 months of life, which is required for classic RS, is thus not displayed. Moreover, this subgroup might contain important information as to pathology, treatment, and long-term prognosis. As it occurs in both otherwise classic and atypical RS we have regarded it as a separate subtype, until it can be further elucidated. For this particular study, the FF subgroup has been further divided by age into one group above 13 years of age according to the original definition [6] and one group 10-13 years of age, in order to lower the age limit for the FF concept. With the broader diagnostic model that has now been applied, we consider the lower age limit of 10 years to be possible. Most of the FF females in the present series have been observed for years but it was not previously possible to accept them as RS because of an atypical presentation, which was not compatible with the original classic criteria. In particular, some have shown a surprisingly well-preserved, yet somewhat dyspraxic, hand function, as well as the absence of the classic handwringing RS stereotypies. Moreover, in the vast majority the head circumference curves have remained within normal limits
Hagberg and Skjeldal: Rett Syndrome----Diagnosis of Variants
9
and it has not been possible to establish an early convincing deceleration. The late regression subgroup represents fairly recent experience [2] and appears to be an important subject for further evaluation in clinical RS research. The same applies to the more controversial preserved-speech subgroup which was previously emphasized by Zapella [13] and recently reviewed [2]. The congenital subgroup (group II:5 in Table 1) still raises particular diagnostic problems [21. In addition to the presence of primary criteria (A), the model constructed for inclusion as atypical RS depends on the clustering of supportive manifestations (B) known to be peculiar to the condition [1,3]. We consider the existence of a cluster of at least 5 supportive deviations to be essential for the model. A prominent complex profile of multiple symptoms, usually not fully evident until late childhood, provides the diagnostic power required to reveal potential RS cases even when combined with an incomplete array of the obligatory criteria otherwise required within the concept of classic RS [4,8]. Of necessity, the items which have been chosen have a variable power of specificity. Thus, for example, the weight of diagnostic importance differs considerably between breathing irregularities and decreased or delayed nociception (items B1 and B10 in Table 2). The unique disturbed breathing regulation in RS girls, noted only when awake, is a quite specific RS manifestation as thoroughly delineated by Kerr and collaborators [14], whereas the inadequate reaction to pain stimuli is more nonspecific, infrequent, and less studied, but is nevertheless peculiar when present. Apart from clusters of behavioral oddities (e.g., creaking bruxism, peculiar laughing and screaming episodes, bloating, and eye pointing), the subsequent appearance with age of additional distal lower limb neurologic abnormality is an important clinical finding when present. Specific abnormalities on EEG have not been observed in RS. However, a characteristic stepwise abnormal EEG development, starting with normality during the first years of life, has been found [15]. This pattern (item B8), which develops over time, has a supportive diagnostic power which we believe will be increasingly useful with the current increasing awareness of RS behind developmental regression in young girls. In this pilot series of atypical RS, however, it was only possible to evaluate the EEG development over time in a few patients because the data available for analysis were either too fragmentary or nonexistent. We therefore had not 11 but only 10 supportive items to use in most patients. The diagnostic model is considered to have the capacity to identify and distinguish relatively different types of atypical RS and to exclude other clinical conditions with developmental abnormalities. Its differentiating power was tested on a series of 8 children with Angelman syndrome who had a partly RS-like clinical presentation. All had a verified chromosome 15q abnormality. These children appeared to have at most 3 of the 11 supportive (B)
10 PEDIATRICNEUROLOGY Vol. 11 No. 1
criteria, usually gait abnormality (however, more ccrebellar ataxia than dyspraxia), and unexpected laughing episodes (items B4 and B9 in Table 2). Impaired nociception (item B10 in Table 2) was indicated in 3 of the 8 children. In addition, 1 of the girls with Angelman syndrome displayed RS-like handwringing (item A3 in Table 2). The model was found to be particularly suitable tor identifying mild, protracted, and initially indistinct RS clinical phenotypes, the most vague FF and the late regression variants. Congenital onset phenotypes, with early severe motor disabilities, remained more difficult to identify. For epidemiologic research, this model, added to the original one delineating classic RS [4,8] and with criteria defined since 1985 [8], was considered to be valuable in the present process of mapping geographic RS clusterings, like those revealed by previous Swedish experience [16]. Parallel to this, the model can certainly help to visualize additional familial RS cases of discrete and easily overlooked types [17]. Thus, attention can now be focused on unexplained and unspecific cases of mentally retarded females reported as existing in the same family. For diagnostic application, we would again emphasize the necessity of the presence of the characteristic cluster of a number of the supporting (B) criteria. The presence of single RS pecularities (e.g., isolated handwringing) should be regarded with great caution. Moreover, the sensitivity and specificity of the model is dependent on meticulous and initiated history-taking, detailed clinical observations, and repeated follow-up evaluations. Not least important is the age of the girl. Our experience indicates that some atypical RS patients, regarded as representing cases of unspecific mental deficiency in early childhood, have only convincingly demonstrated a full array of peculiar RS symptoms late in childhood (after age 9-10 years). Our preliminary message from the pilot series of atypical Swedish and Norwegian RS cases is as follows. The FF type appears to be the most prevalent atypical expression of RS. In these girls, the RS supportive characteristics often do not appear until late childhood. The late childhood regression type (i.e., with a regression period delayed until school age and usually with RS symptoms which develop fairly slowly) appears not to be extremely rare but is overlooked by the medical profession. In their early years, these patients are often considered to represent cases of so-called "simple mental retardation" of unexplained origin or to have an undifferentiated label of "infantile autism." The proportion of FF and late regression phenotypes taken together appears to approach onefifth of all RS, as judged from our Swedish series updated as of 1993 and now comprising 155 RS females aged 2-50 years (April 1994). An early seizure onset phenotype characterizes 7% of all RS in the Swedish series, the majority of whom represent classic RS with a few FF or late regression phenotypes (e.g., our case $90 [Table 3] with initial seizures and muted RS symptoms over an extended
period). The congenital variant, revealed in relatively few Swedish and Norwegian RS females, is difficult to distinguish and cannot usually be diagnosed accurately without many years of detailed observation and follow-up. This is particularly true when a later regression is indistinct or apparently lacking. It was not our intent to present a complete and fully tested methodology nor to show that the model has proven diagnostically reliable. Our intention has been to present a diagnostic model which has functioned well in a routine neuropediatric setting. Increasing the requirement to at least 6 or 7 supportive inclusion criteria could be an alternative for a future strict scientific design. In conclusion, with no diagnostic biologic RS marker yet available, our model was found to be useful as a complement to provide better insight into the RS concept. It has to be tested on independent series of both cases suggestive of atypical RS and those with other well-defined developmental disorders. We thank Alan K. Percy, MD, for a most valuable dialogue during the construction and application of the model. The study was supported by the W:6 Rett Research Programme and Gunnar and M/irta Bergendahl's Foundation, G6teborg, Sweden; and by the Co-ordination Council for the Mentally Retarded, Norway. References
[1] Hagberg B. Clinical criteria, stages and natural history. In: Hagberg B, ed. Rett syndrome~linical and biological aspects. Clinics in developmental medicine No. 127. London: Mac Keith Press, 1993;4-20. [2l Hagberg B, Gillberg C. Rett variants---Rettoid phenotypes. In: Hagberg B, ed. Rett syndrome-421inical and biological aspects. Clinics in developmental medicine No. 127. London: Mac Keith Press, 1993;40-60.
[3] Hagberg B. Rett syndrome: Clinical peculiarities, diagnostic approach and possible cause. Pediatr Neurol 1989;5:75-83. [4] Trevathan E, Moser HW. Diagnostic criteria for Rett syndrome. Ann Neurol 1988;23:425-8. [5] Hagberg B, Witt Engerstr6m I, The Swedish series of females with Rett syndrome 1960-92. In: Hagberg B, ed. Rett syndrome-Clinical and biological aspects. Clinics in developmental medicine No. 127. London: Mac Keith Press, 1993;21-5. [6] Hagberg B, Witt-EngerstrOm I. Rett syndrome: A suggested staging system for describing impairment profile with increasing age towards adolescence. Am J Med Genet 1986;24(suppl 1):47-59. 17] Hagberg B, Rasmussen P. "Forme fruste" of Rett syndrome--A case report. Am J Med Genet 1986;24:175-81. [8] Hagberg B, Gouti~res F, Hanefeld F, Rett A, Wilson J. Rett syndrome: Criteria for inclusion and exclusion. Brain Dev (Tokyo) 1985; 7:372-3. [9] Gillberg C. The borderland of autism and Rett syndrome: Five case histories to highlight diagnostic difficulties. J Autism Dev Disord 1989; 19:545-59. [10] Gouti~res F, Aicardi J. Atypical forms of Rett syndrome. Am J Med Genet 1986;24:184-94. [11] Hanefeid F. The clinical pattern of the Rett syndrome. Brain Dev 1985;7:320-5. [12] Hagberg B, Witt Engerstr6m I. Rett syndrome: Epidemiology and nosology--Progress in knowledge 1986---A conference communication. Brain Dev 1987;9:451-7. [13] Zapella M, The Rett girls with preserved speech. Brain Dev 1992;14:98-101. [14] Kerr AM. A review of the respiratory disorder in Rett syndrome, Brain Dev 1992;14 (suppl):S43-5. [15] Hagne I, Witt Engerstr6m I, Hagberg BA. EEG development in Rett syndrome. A study of 30 cases. Electroencephalogr Clin Neurophysiol 1989;72:1~5. [16] /~kesson HO, Hagberg B, Wahlstr6m J, Witt Engerstr6m I. Rett syndrome: A search for gene sources. Am J Med Genet 1992;42: 104-10. [17] Anvret M, Wahlstr6m J, Skogsberg P, Hagberg B. Segregation analysis of the X-chromosome in a family with Rett syndrome in two generations. Am J Med Genet 1990;37:31-5.
Hagberg and Skjeldal: Rett Syndrome--Diagnosis of Variants
11