Rett syndrome—a review and discussion of syndrome delineation and syndrome definition

SYNDROME DELINEATION AND NOSOLOGY

Rett Syndrome-A Review and Discussion of Syndrome Delineation and Syndrome Definition John M Opitz, MD and Susan 0 Lewin, MD

The current status of clinical genetic and pathogenetic knowledge in Rett syndrome is reviewed and updated. Some of the concepts which are highlighted include among others, causal homogeneity vs heterogeneity; the difficulty of dealing with potential variability while the possibility of heterogeneity still exists; progress in examining genetic hypotheses. We review our experience with 8 new patients, note the evidence for the presence of congenital hypotonia in all and several instances of minor anomalies and again raise the issue of whether girls with the Rett syndrome are in fact normal from the time of birth and in early infancy until the characteristic phenotype is recognizable. Previous recommendations for standardized international history, examination and investigation protocols are strongly reinforced, and the question as to who should study brains of Rett syndrome patients is raised by a parent. As is the case in most conditions now being studied by molecular geneticists, finding a marker will be a major etiologic breakthrough, but the work of delineating the pathogenesis will remain to be done. Key words: Rett syndrome, review, genetic hypotheses. Opitz JM, Lewin SO. Rett syndrome - a review and discussion of syndrome delineation and syndrome definition. Brain Dev 1987;9:445-50

At this occasion, Dr Lewin and I express our deep admi· ration to Professor Andreas Rett for his extraordinary accomplishments in pediatric neurology (of which the delineation of the Rett sy~drome is only one of many) and for his kindness in inviting us to present our views and data on the Rett "syndrome." We were specifically asked to address the concepts of syndrome identification and delineation; to the extent these topics were covered at the Baltimore conference [1] we shall not reiterate them here, except for certain elementary considerations pertaining to the concept of syndrome. The Concept of Syndrome with Respect to the Rett "Syndrome" The genetic definition of syndrome [2] is: "A group of specific characters (symptoms) characteristic of a certain (abnormal) genetic condition." An "inheritable" or "genetic" (syndrome) may be transmitted to the offspring as a unit." In this context the words genetic or inheritable refer to cause (or "etiology," a less accurate synonym

From Shodair Children's Hospital, Helena Montana (JMO, SOL) and the Universities of Washington/Seattle and Wisconsin/Madison and Montana State University, Bozeman, Montana (JMO). Correspondence address: John M. Opitz, Shodair Children's Specialty Hospital, Box 5539, Helena, MT 59604, USA.

since etiology means the study of cause rather than cause). Strictly speaking, the cause of the Rett syndrome is unknown; hence, it is defmed nosologically as a "syn· drome" from a medical, not from a genetic point of view. A medical dictionary, such as the 24th edition of the Stedman dictionary [3], defines a syndrome as: "The aggregate of signs and symptoms associated with any morbid process, and constituting together the picture of the disease." Clinical geneticists would be tempted to add non·disease producing combinations of structural anomalies as well. Non-causal definition of "syndromes" means at least a pathogenetic relationship between the various manifesta· tions. In the absence of evidence for a singIe basic (bio, chemical) mechanism these manifestations are presumed to exist together in 2 or more patients on the basis of the prevalence of the condition, and the small likelihood of chance concurrence of the component manifestations. Since cause is unknown it can not be excluded at this moment that the Rett syndrome is heterogeneous, i.e. that it is a causally nonspecific "phenotype" reo presenting a final pathogenetic pathway common to a number of different causes. An example is mucopolysaccharl'dosis III which is now known to be a hetero· geneous phenotype consisting of at least four causally delineated entities or Sanfilippo syndrome type A, B, C, D each representing a different, specific, autosomal recessive, non-allelic inborn error of metabolism [4] .

Heterogeneity in the Rett Syndrome? One of the synonyms for genetic heterogeneity is the term "genocopy" coined by Nachtsheim in 1957 [5] . Is there evidence for genetic heterogeneity or the existence of genocopies in the Rett syndrome (presuming for a moment that the Rett syndrome is indeed a genetic condition)? The first example that comes to mind is that published by Hyman and Batshaw [6]. This pertains to an 8-year-old girl with profound mental retardation, microcephaly after initially normal development in childhood, and presence of seizures, hyperventilation, ataxia, amimia and "hand wringing" stereotypies. But she had several symptoms which set the course of her condition clearly apart from that of the Rett syndrome, namely, recurrent episodes of vomiting and hyperammonemic coma. The diagnosis of ornithine hanscarbamylase deficiency was finally made in her. Hyperammomenia has been suspected as a metabolic component of the Rett syndrome and was mentioned in the first descriptions of the condition by Rett in 1966 and 1968 [7,8]. We now know that this is probably not a pathogenetic component of the Rett syndrome. This instance of ornithine transcarbamylase deficiency is Family 1 reported by Batshaw et al [9] . The patient's mother was found to be a heterozygote and the condition was extensively inherited in the family. Ornithine transcarbamylase deficiency is an X-linked disorder and an X-linked defect is suspected in the Rett syndrome. Through several ingenious cytogenetic and genetic studies ornithine transcarbamylase has been mapped to Xp21.1 in the following order: Xpter-DXS9DXS67 -(GK, CAH)-DMD-DXS84-DXI48-0TC-DXS7 -cen. Thus, we still have not given up the hope that the region Xp21.1, or it's vicinity, on the short arm of the X chromosome, might be an appropriate place to search with probes to detect a possible mutation in the Rett syndrome. Ornithine transcarbamylase deficiency in the Rett syndrome was studied by Hanefeld and coworkers [10] with negative results. A possible second instance of heterogeneity was cited in the addendum to our Baltimore paper of 1986 [1] . In this 3.5-year-old girl with initially relatively normal development and purportedly acquired microcephaly, autism, sleep disturbance, acrocyanosis, mouthing of hands, congenital hypotonia/lymphedema, mental retardation and vascular dysplasia, cytogenetic evaluation showed mosaicism with an isochromosome 13q form of trisomy 13: 46,XX,del(13p)/46,XX,i(13q), the del(13p) chromosome presumably being an unstable "telocentric" converting spontaneously into an isochromosome. Another possible example of heterogeneity was presented by Gordon and coworkers at the 1986 David Smith meeting [11] on malformation and morphogenesis in Burlington, Vermont. This was a 30-year-old woman;

446 Brain & Development, Vol 9, No 5, 1987

deceleration in psychomotor development began at 8 months. She had no speech and functioned at the mental age of a lO-month-old child. She had abnormal hand movements, ataxic gait, acquired microcephaly and shortness of stature, hyperventilation, bruxism, seizures with abnormal EEG and a karyotype indicating a mosaic 18q- syndrome: 46,XX/46,XX,del (18q). Another patient with Rett syndrome (or more properly Rett "phenotype") was found to have a metabolic defect of biotin dependency with reduced propionylCoA and 3-methy1crotonyICoA carboxylase activities [12]. Thus, we conclude at the moment that heterogeneity in the Rett syndrome remains unsubstantiated but is also not ruled out. However, the relative non-specificity of Rett syndrome manifestations singly or in combination make it tempting to see the Rett syndrome "under every stone," especially in institutions for the mentally retarded. Heterogeneity vs Variability Lack of knowledge of cause makes it impossible to study heterogeneity and may lead to artificial truncation of the phenotypic spectrum and underestimation of the true extent of variability of the Rett syndrome. However, it seems better at this stage in history to preserve the relative "purity" of samples of Rett syndrome patients by making the diagnostic criteria rather strict than to try and estimate the total incidence of the condition in the population including every "forme fruste" and possible case in males. We support Hagberg in postulating that variations of the Rett syndrome exist; however, the only way in which these can be conclusively accepted as the Rett syndrome proper is if they occur in sisters of patients with the classic syndrome; to our knowledge to date none of the forme fruste patients has had such a classically affected sister. The Rett Syndrome Phenotype What is the Rett syndrome phenotype? In most cases it seems to involve initially apparently relatively normal development until about the age of 6-18 months. The affected girls acquire microcephaly between 6 months and 4 years, however, cases with a nomlal head circumference with all other manifestation of the Rett syndrome are known. Regression of skills occurs in infancy and early childhood with an ultimate mental age of 4-6 months having at one time been higher in most cases. Most children develop seizures and in many one obtains a history of sleep disturbance, episodic hyperventilation/aerophagia, bruxism, constipation, stereotypic hand movements (in about 90% of cases), together with "autistic" behavior. A point to be made about the hand movement stereotypies is that every affected

child has some apraxia and abnormality of hand movement in a pattern which is characteristic of each child and which need not necessarily involve joint movement of hands in the midline.

Rettsyndrome diagnosis The diagnosis of the Rett syndrome is made on the basis of female sex, with "male variants" not being really creditable cases unless they were a brother of a "classically" affected girl. The second criterion of the diagnosis is the classic history; the third is the finding of the combination of microcephaly, severe to profound mental retardation, seiZures with abnormal EEG, hypotonia, ataxia, spasticity with contractures and frequent later scoliosis. Most children have "typical" abnormal hand movements. They lack speech and most have episodes of hyperventilation. Acrocyanosis is universal. These children have a strong history of liking music and most of them ultimately show shortness of stature and underweight. Thus, the Rett syndrome emerges as a common disorder affecting only the female sex and apparently involving only the eNs. That is to say, this seems to be a primary "neuropathic" condition with many secondary changes reflecting impaired function of the CNS. A reading of the literature seems to suggest that the Rett syndrome is not a multiple congenital anomalies (MCA) syndrome. Recent Personal Observations Recently we had the opportunity of studying eight additional patients, five of them at Medical Lake, Washington at the gracious invitation of Dr Mary Harrison and the collaboration of her staff. These patients and three others seen in Coeur d'Alene, Idaho had also been studied previously in Seattle, Spokane and Baltimore, among others by Drs Vanja Holm and Andreas Rett. Their mean age was 14.4 years (range 5 1/2 to 28 years). In all eight patients the diagnosis of the Rett syndrome seemed firm on the basis of history, neurologic and psychologic fmdings. In all cases prenatal growth was normal. A few perinatal abnormalities were identified including: neonatal icterus in one case lasting 18 hours and rising to a maximum of 30 mg%; this child required two exchange transfusions. In another case there was a history of difficult labor with face presentation; this infant arched her back already in the neonatal period. In another case there was severe emotional and nutritional deprivation. With respect to family history all cases were sporadic an~ the family histories were unremarkable except in one where the mother had Parkinsonism. Past his't ory indicated early onset of puberty in two

cases, sleep disturbance in all, fits of screaming in most , failure to thrive or feeding difficulties, hyperthermia (i.e. an easy tendency to "run a high fever" without demonstrable infection) and a consistent history of very tou.chy/sensitive feet. All children took great pleasure in music. All were severely to profoundly retarded; one child never learned to sit. Seizures were present in 6 of 8 and an abnormal EEG in an additional patient. To variable degrees all patients showed evidence of congenital hypotonia with or without congenital lymphedema. In one child there was a clear history of failing to bear weight in infancy. Virtually all children had an increased number of white lines and there were several four-finger lines. Brachycephaly and/or plagiocephaly were seen in several children by us and by Nicola [13] . We also observed ptosis, crowding of toes, clinodactyly of toes and fifth fingers, striking and congenitally fixed camptodactyly of fmgers and toes in a couple of patients; micrognathia, and an extremely small muscle mass in all. We interpret a history of recurrent bronchitis and respiratory infections with otitis and severe recurrent constipation as additional manifestations of congenital hypotOnia. On physical examination we found striking hirsutism in all cases involving body hair, sideburns, synophris, moustache, hairy limbs; Nicola [13] also found hirsutism in the mothers of her patients. Hirsutism was also present in those patients not receiving diphenylhydantoin. We observed early pubarche and thelarche and prominent labia majora in several patients. Several girls showed evidence of a mild pigmentary dysplasia including patches of hypo- and hyperpigmented skin. Especially impressive was the finding in all girls of mUltiple pigmented nevi and the history and observation of pink nevi with a white halo. The minor anomalies that we observed include narrow forehead in virtually every case together with midface hypoplasia. Many children had macrostomia and several had short tips of toes and toenails; a couple of them had strikingly short metatarsals and metacarpals. One child had a cleft uvula as a mild malformation and another had two mild malformations including duplication of right ureter and mild aortic stenosis. These observations suggest the following three questions: 1) Is the Rett syndrome an MCA/MR syndrome after all? And does it involve the neural crest as suggested by the pigmentary dysplasias? Nicola [13] found a higher incidence of whorls and a greater-than-normal total ridge count in Rett syndrome propositae; if confirmed, that would suggest a first trimester onset of the pathogenesis! 2) Is there sufficient evidence to suggest presence of a neuroendocrine disorder in the Rett syndrome, and if

Opitz and Lewin: Rett syndrome 447

so, should a standard protocol for the evaluation of such patients include recommendations for study of neuroendocrine functions? 3) Ii any Rett syndrome girl neurologically normal at birth? Is it possible to demonstrate evidence of congenital hypotonia in all? If the answer to the last question is yes then we will have demonstrated unequivocal evidence that the Rett syndrome is not an acquired but rather a congenital, developmental disorder. In his comments on this manuscript Hanefeld also expresses the view that Rett syndrome girls probably are not normal at birth, and may then atready have a "dyskifletic" disturbance of hand movemeq,t. Thus, we would suggest that a standardized protocol for the evaluation of Rett syndrome patients include the following: 1. A standardized approach to the physical examination. 2. Standardized cephalometry and anthropometry, including several skull and body segment measurements. 3. We think that a Poznanski metacarpophalangeal profile analysis of these patients would be a particularly productive investigation. 4. We urge continuing studies of dermatoglyphics of Rett syndrome girls compared to their sisters and brothers and to th~ general population. 5. Neuroendocrine studies may be indicated. 6. We recommend that nuclear magnetic resonance (NMR) studies be done in all patients upon first diagnosis and, if available at few University Hospital Centers, positron emission (PET) tomography on a few specific metabolic pathways such as the Dl • dopamine receptors. Genetic Data We are aware of at least three cases of parental consanguinity including the patient studied by Coleman et al [14] where the parents were 1 1/2 first cousins (or 1/2 second cousins); the patient mentioned by Killian where the parents were first cousins, and the possible Turkish twin pair studied at Heidelberg with first cousin parents. Killian found one instance of consanguinity in 215 families (15]. To our knowledge, the familial cases include the (maternal) half-sisters reported by Anvret et al (16] ; the two sets of sisters observed by Hanefeld [10] and the 1/2 second cousins studied by Philip art and Naidu with classical Rett syndrome (mentioned in AI-Mateen et al [17]; also published by Riccardi in 1986 [18] as cases that had been studied by Percy in 1985). Three or four sets of concordantly affected monozygotic twin girls are known [19] including the two black children studied by Coleman in Philadelphia and confirmed as virtually proven monozygotic twins by HLA typing. There exists controversy as to whether the "Heidelberg

448 Brain & Development, Vol 9, No 5, 1987

twins" have the Rett syndrome. Three sets of discordantly affected dizygotic twins have been studied [20] including a sister pair in Milwaukee studied by Dr Naidu, a brother/sister pair studied by Dr Naidu in St Louis, and another brother/sister pair studied by Dr Murphy in Maryland. Recently we were privileged to study a Rett syndrome girl referred from Denver whose (dizygotic) twin brother was unaffected by history. Thus, dizygotic twins seem to be always discordantly affected and monozygotic twins always concordantly affected; however, true familial cases are rare and to date constitute a fraction of one percent of over one thousand cases studied at major centers. Killian (15] has made an analysis of sex ratio, prenatal loss, and parental age to which the data of Coleman can be added. The 255 propositae had 219 normal brothers and 190 unaffected sisters with a male to female sex ratio of 1.15. In 1986 Killian reported (6 out of 601) spontaneous abortions and 8 stillbirths; this impresses as an incidence of fetal death less than that of the average population. Parental age studies by Killian (15] did not uncover any abnormalities; he found the mean maternal age of 27.3 years and a mean paternal age of 29.6 years in a total of 142 cases. Genetic Hypotheses A genetic hypothesis is needed for a condition with a dozen salient criteria: 1. This is a common disorder with a prevalence of about lout of 10,000. A high prevalence of course may mean heterogeneity, although this remains unproven. 2. Apparently sole involvement of the female sex. 3. Overwhelmingly sporadic occurrence with only a halfdozen familial occurrences known out of some 1 ,000 or so families studied. 4. Normal parents with unremarkable family history for mental retardation or other CNS defects. 5. Normal parental reproductive age. 6. Normal Sib/sex ratio. 7. Normal consanguinity rate. 8. Apparently no increased fetal wastage. 9. Many population groups involved. 10. Apparently normal chromosomes, by present methods, in almost all patients. II. No consistent metabolic defect has been demonstrated. 12. Concordance of MZ, discordance of DZ twins. Killian's excellent paper of 1986 [15] reviewed a number of genetic hypotheses several of which could be dismissed as rather unlikely. These would include an autosomal dominant mutation with female sex limitation; an autosomal dominant mutation prenatally lethal in males; an autosomal recessive trait with female sex

limitation; and an autosomal recessive disorder prenatally lethal in homozygous males. A two-step mutational process was discussed by Anvret et al [16] and Hillig [21]; also cytoplasmic determination and an X-linked recessive disorder with non-pen trance in males and unusuallyonization in affected and in unaffected but transmitting females [1]. The latter hypothesis was based mostly on the model of the Martin-Bell syndrome; however, on reconsideration it would seem unlikely since some carrier fathers would then have to transmit the condition to one-half of their daughters. The most likely hypothesis still would seem to be that of an X-linked dominant disorder lethal in hemizygotes. This was first proposed by Rett and Killian in 1983 [21] and was also discussed by Comings [22] and Riccardi [18], with the latter also postulating a possible disturbance of heterochromatinization of the inactive X chromosome. At the Berlin congress Dr Francke and at this conference Dr De Gennaro discussed the model of an X-linked mutation involving a protein produced only in the CNS and necessary for the neuron to function and to survive . In a female heterozygous for such a mutation failure to produce the protein or production of a mutant protein might lead to death in a random manner of approximately half of such neurons; this might simulate the picture of a "temporary" progressive central nervous system degenerative disease which would leave in place only normal but a reduced number of neurons for study at the time of rare autopsies. The existence of such a protein was also postulated by Comings [22]. This seems to be a plausible hypothesis and we suggest that investigators collaborate with Dr Uta Francke at Yale University in performing appropriate tests of the hypothesis. Genetic Studies: Recommendations 1) We would concur with the recommendation that in all Rett syndrome cases high resolution prometaphase chromosome studies be done with the greatest of care. 2) In all cases accurate three generation pedigrees should be obtained. After study of the propositae all families should be invited to keep in contact with the center for a prospective assessment of the subsequent reproductive performance of the parents. 3) We strongly recommend continuing search of X chromosomal DNA of Rett syndrome patients and their parents for structural changes with many probes. 4) We still recommend that Rett syndrome propositae be studied carefully for any abnormalities of lyonization of the X chromosomes.

Final Question The father of a Rett syndrome patient recently wrote

to us asking for guidance in selecting centers best qualified to study the brains of children with the Rett syndrome. We hope that this conference will settle on such guidelines and recommendations applicable to individuals in North America, Europe and other parts of the world. It has also been suggested that the Birth Defects and Developmental Disabilities branch of the Center for Environmental Health of the CDC, Atlanta, Georgia undertake a North American registry of Rett syndrome patients. We strongly support such a venture but would also recommend that this be done in concert with similar efforts in Europe and elsewhere (given guarantees to safeguard the confidentiality of personal data) and that everywhere the patients be evaluated and recorded according to an agreed-upon standardized protocol. Colophon The methods of molecular biology have suddenly thrust medical and clinical genetics onto center stage of the most dramatic developments ever in human biology. We would venture to predict that an application of these methods to the study of the Rett syndrome will sooner, rather than later uncover the specific genetic defect in the syndrome. This will almost immediately provide an elegant explanation of the 12 (genetic) characteristics of the syndrome mentioned above and an intellectually most gratifying insight in to the causal genesis of the Rett syndrome. As a byproduct it will probably also yield a test for prenatal diagnosis and for the rare silent carrier state. But it will, as in so many other previous instances, bypass or sidestep the issue of the formal genesis of the Rett syndrome. It will be the task of biochemists, pathophYSiologists and neuropediatricians to elucidate in painstaking work the pathogenetic chain of events which begins after the initial action of the mutant gene and which brings about the catastrophic consequences now recognized as the Rett syndrome. And while this may be cause for sober reflection it ought to be accepted almost as a moral challenge by the rest of us engaged in the practice of clinical genetics to sharpen our focus and to deepen our insights into the signs and symptoms and natural history of the Rett syndrome in our patients; and, for having pointed us in the right direction toward this end, we remain forever grateful to Andreas Rett.

ACKNOWLEDGMENTS We are most grateful to Gail Aguilar and LaVelle M. Spano for expert secretarial collaboration, to the State of Montana for support under HB 430, to the Alberta Heritage Foundation for Medical Research for a Fellowship grant to Susan O. Lewin, and to Drs. Andreas Rett, Hugo Moser, Bengt Hagberg, F. Hanefeld, Mary Coleman and W. Killian for constructive and most helpful comments on this manuscript.

Opitz and Lewin: Rett syndrome 449

REFERENCES 1. Opitz JM. Editorial comments: Rett syndrome: some comments of terminology and diagnosis. Am J Med Genet 1986; 24 (Suppll):27-37. 2. Rieger R, Michaelis A, Green MM. A glossary of genetics and cytogenetics. 3rd ed. New York: Springer Verlag, 1968. 3. Stedman'S illustrated medical dictionary. 24th ed. Baltimore· London: Williams and Wilkins, 1982. 4. McKusick VA. Mendelian inheritance in man. Catalogs of autosomal dominant, autosomal recessive and X-linked phenotypes. 7th ed. Baltimore: Johns Hopkins Univ. Press, 1986. 5. Nachtsheim H. Mutation und Phiinokopie bei Siiugetier und Mensch. Ihre theoretische und praktische Bedeutung fUr Genetik und Eugenik. Experientia 1957; 13:57. 6. Hyman SL, Batshaw ML. A case of ornithine transcarbamylase deficiency with Rett syndrome manifestations. Am J Med Genet 1986;24 (Suppll):339-43. 7. Rett A. Uber ein zerebral-atrophisches Syndrom bei Hyperammoniimie. _Wien: Briider Hollinek, 1966. 8. Rett A, Stockl W. Untersuchungen tiber den Ammoniakgehalt im Blutserum hirngeschiidigter Kinder. Wien Med Wochenschr 1968;118:311-4. 9. Batshaw ML, Msall M, Beaudet AL, Trojak J. Risk of serious illness in heterozygotes for ornithine transcarbamylase deficiency. J Pediatr 1986;108:236-41. 10. Hanefeld F, Hanefeld U, Wilichowski E, Schmidtke J. Rett syndrome - Search for genetic markers. Am J Med Genet 1986;24 (Suppl 1):377-82. 11. Gordon K, Siu VM, Sergovich F, Jung J. 18q-associated with Rett syndrome. Program and abstracts. DW Smith

450 Brain & Development, Vol 9, No 5, 1987

Conference on Morphogenesis. Burlington: VT, 1986. 12. Bachmann C, Schaub I, Colombo IP, Burri BI, Sweetman L, Wolf B. Rett syndrome revisited: a patient with biotin dependency. EurJ Pediatr 1986;144:563-6. 13. Nicola M. Morphological aspects of Rett syndrome. Infans Cerebropathicus 1976; 1:81-107. 14. Coleman M, Brubaker J, Hunter K, Smith G. Rett syndrome: a survey of North American patients (in preparation). 15. Killian W. On the genetics of the Rett syndrome: analysis of family and pedigree data. Am J Med Genet 1986;24 (Suppl 1):369-76. 16. Anvret M, Johanssen I-M, Wahlstrom J, Hagberg B. Linkage analysis of the Rett syndrome using human chromosomal specific probes. Brain Dev (Tokyo) 1985; 7:361-4. 17. Al-Mateen M, Philipp art M, Shields D. Rett syndrome. A commonly overlooked progressive encephalopathy in girls. Am J Dis Child 1986;140:761-5. 18. Riccardi VM. The Rett syndrome: genetics and the future. Am J Med Genet 1986; 24 (Suppl 1):389-402. 19. Hagberg B. Data presented at the 4th Rett Conference, Vienna, 1986. 20. Moser H. Data presented at the 4th Rett Conference, Vienna, 1986. 21. Hillig U. On the genetics of the Rett syndrome. Brain Dev (Tokyo) 1985;7:368-71. 22. Comings DE. The genetics of the Rett syndrome: the consequences of a disorder where every case is a new mutation. Am J Med Genet 1986;24 (Suppll): 383-8. 23. Morrissey P. Personal communication, 1986. 24. Sever LE. Personal Communication, 1986.