More than 30 years ago the late Sam Pruzansky wrote about the value of time as a fourth dimension in the analysis of genetic syndromes [Pruzansky, 1977]. While the focus of Pruzansky's article was longitudinal change in the phenotypes of craniofacial malformations, his approach to change underscores the logic for revisiting the classification of the ectodermal dysplasia syndromes.
Pruzansky wrote about syndromes, he called them conditions, that improved with time (e.g., Robin sequence), worsened with time (e.g., Apert syndrome), or remained unchanged with time (e.g., mandibulofacial dysostosis). After describing time-related changes in such syndromes, he commented that “knowing the natural history of a syndrome is essential for diagnosis” and important for personal and parental counseling. One could extrapolate this way of thinking to changes in the way syndromes have been classified over the years, and draw conclusions similar to Pruzansky's; that is, proper classification is essential for diagnosis and prognosis, elucidation of etiology, and genetic counseling.
Before discussing classification of a group of disorders as complex as the ectodermal dysplasia syndromes, it is important to be clear about the terms used in the process.
This is outermost of the three primordial layers of cells of the embryo. The surface ectoderm gives rise to the outer layer of skin, dermal appendages (e.g., hair, nails, and sweat glands), teeth, and parts of the eye and inner ear. The neuroectoderm gives rise to much of the nervous system. Most discussions and case reports on ectodermal dysplasia syndromes center on dermal appendages and teeth.
For purposes of discussing ectodermal dysplasia syndromes, this term is best defined as abnormal organization of cells into tissue and organs. The implication here is that an embryonic anlage is abnormal in some way and gives rise to mature structures that are deficient in number (if one is referring to a composite structure, such as the dentition), size (hypoplasia might be an acceptable synonym), character, or function. Dysplasias are not caused by environmental factors (carious lesions of teeth and weathering of hair cuticles for instance) or aging (physiological hair loss). Deformations (abnormalities of shape caused by external forces) are clearly not dysplasias.
The word syndrome refers to a constellation of physical (maybe even physiological) manifestations that co-exist and are presumed to have a common cause. Other terms for co-existing anomalies are association and sequence. An association, however, is more a statistical concept than anything else (physical features that occur together more often than predicted by chance) and a sequence is co-existence of features based on embryonic pathways (one feature leads to another and another, as in the Robin sequence where failure of extension of the fetal head leads to pressure on the mandible, which in turn leads to obstruction of palatal shelf fusion and later to glossoptosis). In many cases the shared cause is known (single gene mutation, chromosomal aberration); in others a common etiology is merely suspected on the basis of pedigree analysis or by similarity with other syndromes.
Simply put, this term means related to the genetic material, but in clinical practice it refers to syndromes and conditions caused by variation in gene structure. The variation may be at the level of single genes (monogenic), several genes acting together (polygenic), or larger segments of genetic material (chromosomal). It is wrong to consider genetic to be synonymous with inherited and even heritable since the syndrome or condition may not have been inherited from a parent in the case of mutation, or the variation in genetic material may be of such nature that inheritance is precluded (genetic lethals).
Generally speaking, the most commonly accepted definition of ectodermal dysplasias for the past several decades has been: genetic disorders that include abnormalities of two or more ectodermal structures. A nosologist would be justified in rejecting this definition because it truncates the array of manifestations that might be referenced by the rubric. That is, the notion that two or more abnormalities must be involved eliminates from consideration those dysplasias that affect only one ectodermal derivative, and there are many. For instance, the enamel of teeth is an ectodermal derivative that may be dysplastic in and of itself; such disorders are called the amelogenesis imperfectas. Since the embryonic anlage is defective and since such defects may have a genetic basis, are not the several amelogenesis imperfectas really ectodermal dysplasias? And, what about dysplastic nails or inherent abnormalities of hair structure (the onychodysplasias and trichodysplasias)? Are not they really ectodermal dysplasias? Of course they are, and they may be an isolated trait in an individual or family, or may co-exist with other abnormalities.
What the commonly accepted definition has been referring to it seems are ectodermal dysplasia syndromes. So, let us agree that there are ectodermal dysplasias (isolated defects of ectodermal derivatives) and ectodermal dysplasia syndromes (co-existing defects of two or more ectodermal derivatives). What then? How many ectodermal dysplasia syndromes are there?
How many permutations of ectodermal defects and genetic variations can one imagine? To illustrate the possibilities, take only two ectodermal derivatives—teeth and the hair. Teeth might be dysplastic in several ways: none (anodontia), too few (oligodontia or hypodontia, and there are several patterns of hypodontia), too many (supernumerary), enamel may be defective in at least three ways (hypoplasia, hypocalcification and hypomaturation), shape may be abnormal (pegged, globodontia, taurodontia), size may be abnormal (microdontia, macrodontia), and more. Hair shafts may be thin or thick; shafts may be twisted, flat, grooved, or beaded; there may be pigment variations, variations in the course of shafts (twisted, corkscrew, wooley), and more. Any two or more of these dysplasias may co-exist, and any combination may have a different genetic cause. In short, the number of potential ectodermal dysplasia syndromes is staggering.
The possibilities suggested above are the tip of the iceberg—the ectodermal dysplasia syndromes with co-existing defects of ectodermal derivatives only. What about the myriad of syndromes in which defects of two or more ectodermal derivatives co-exist with defects of structures not derived from the ectoderm? What about the orofaciodigital syndromes, the Ellis-vanCreveld syndrome, the trichodentoooseous syndromes, and others? So, let us extend our area of agreement. There are many ectodermal dysplasias, at least dozens of syndromes that involve ectodermal derivatives only (pure ectodermal dysplasia syndromes), and an untold number of syndromes that involve defects of ectodermal and non-ectodermal derivatives (complex ectodermal dysplasia syndromes).
Clearly, classification of ectodermal dysplasia is analogous to those syndromes that Pruzansky described as getting worse with time. The original definition has proven too restrictive in as much as it ignored isolated ectodermal dysplasias, and the number of disorders to be classified has grown exponentially. A fresh look at the classification of this interesting group of disorders is in order.
STATUS OF CLASSIFICATION
As early as 1977, Estabrook, following Opitz's lead [Opitz et al., 1969] suggested that objective classification of such things as syndromes pass through three developmental stages: descriptive, revision, and hypothesis testing [Estabrook, 1977].
Descriptive Stage of Classification
In his overview, Estabrook likened clinicians to biologists who compare their collections of specimens to those of other biologists, proclaiming a “new” type each time a specimen does not match one in a previous collection. This sort of thing went on with the ectodermal dysplasia syndromes for the last quarter of the 20th century. For years, one could not pick up a clinical genetics journal without seeing a title that listed a concurrence of several physical findings purported to be a “new ectodermal dysplasia.”
In some instances, the new ectodermal dysplasia syndrome was quickly found among “specimens in other collections,” as with the otodental dysplasia (concurrence of globodontia, hypodontia, taurondontia, and sensorineural deafness). First reported in 1975 [Levin et al., 1975], the syndrome had been around for generations, but not recognized. After the initial report, cases were cited from several places throughout the world [Witkop et al., 1976; Stewart and Kinirons, 1982; Chen et al., 1988].
In other instances, similar “specimens” were not found quickly or not at all. There are many examples of this in the classic catalogue of ectodermal dysplasia syndromes by Freire-Maia and Pinheiro 1984. Failure of corroboration does not mean the syndrome in question is not a distinct entity, merely that the syndrome is rare enough or its changes are so subtle that others have not found it in their “collections.”
In the aforementioned catalogue, for instance, is a syndrome called the Jorgenson syndrome (concurrence of sparse hair, nail dysplasia, and sparse bodily secretions). After the initial report of this syndrome in 1974 [Jorgenson, 1974] no other cases have been cited in the literature. Nonetheless, a family with Jorgenson syndrome attended a family conference sponsored by the National Foundation For Ectodermal Dysplasias in 2003. The proposita, a 45-year-old woman of Afro-American extraction, had pinched nares, long philtrum, sparse hair (scant body hair and scalp hair that was thick and coarse in childhood, but shed in her late teens), thin nails with longitudinal grooves, hypohidrosis, and by report had sparse salivary, lacrimal, and vaginal secretions. She had poorly developed dermal ridges of the fingers and palms, lost her teeth early in life and had supernumerary nipples. She had two children, a 17-year-old son who was similarly affected though he still had his teeth, and a 14-year-old daughter who had coarse hair, onychodyplasia, hypohidrosis, pinched nares, and poorly formed dermal ridges. This family corroborates the existence of the syndrome and expands knowledge about it (found in families of both Caucasian and African ancestry, and in males as well as females), but does not guarantee it will be recognized in the “collections” of other health professionals.
Revisionary Stage of Classification
After enough similar specimens have been collected, an analysis can be done to determine which characteristics are most useful in separating them into discrete sub-sets, and characteristics other than those used for the preliminary grouping can be incorporated into the equation that differentiates the specimens. In the process, errors in the original grouping may come to light. This is probably the stage in which classification of the ectodermal dysplasia syndromes is now mired, as the following examples make clear.
Initially, several distinct ectodermal dysplasia syndromes with associated orofacial clefting were described, including Rapp-Hodgkin syndrome (RHS), Hay-Wells (ankyloblepharon-ectodermal dysplasia-clefting—AEC) syndrome and ectrodactyly-ectodermal dysplasia-clefting (EEC) syndrome. As time went by and based in part on family history, some authors proposed that these syndromes were causally related [Gorlin et al., 1990]. More recently, the correct grouping of the clefting disorders and subsequent molecular studies, demonstrated that all three are manifestations of defects in the P63 gene complex, and therefore can be considered the same entity [Celli et al., 1999; Rath et al., 2001; Dianzani et al., 2003]. In short, a preliminary classification was revised when the molecular defects were added to the differentiating formula.
Revisionary classification hinges on the choice and integrity of characters. That is, the characters studied in specimens of the collection should be those most likely to reveal meaningful differences among the specimens, and the characters should be measured correctly; if not, the data are compromised and revision impossible. Confusion about the identity and inheritance of the Schöpf syndrome (sometimes called the Schöpf–Schultz–Passarge syndrome) illustrates this point. In 1971 two sisters were reported with a “new” syndrome comprising cystic eyelids, palmo-plantar keratosis, hypodontia and hypotrichosis [Schöpf et al., 1971]. The women's parents were consanguineous, no doubt prompting the authors to suggest autosomal recessive inheritance. The mode of inheritance would not only be a poor choice as a discriminating character since so many syndromes are autosomal recessive, but also it was most likely assigned incorrectly, regardless of the consanguinity. In fact, the mother of the women and two of her own sisters had hypotrichosis, and a daughter of one of the women had palmo-plantar keratosis. Later, a family was reported wherein four brothers had hypodontia, palmo-plantar hyperkeratosis and onychodysplasia; one had cysts of the eyelid [Font et al., 1986]. Since no antecedents were affected (and most likely influenced by the earlier report), autosomal recessive inheritance was postulated; by this time an autosomal recessive inheritance pattern was established as characteristic of the syndrome.
How wrong these reports were! There were individuals with variable expression of the pleiotropic features of the syndrome in three consecutive generations in the original report—an aspect of family history characteristic of autosomal dominant inheritance. The only valuable information about inheritance from the second report was that males also could be variably affected (all cases in the first report were female). Furthermore, in 1988 another family with probable autosomal dominant inheritance of the Schöpf syndrome was studied (Fig. 1). Note that there were affected individuals in consecutive generations, males and females were similarly affected, and there was probable male-to-male transmission. Clearly, the choice of a discriminating character (inheritance) and improper assessment of the physical phenotype (failure to account for variable expression and pleiotropy) has muddied the waters for years.
Hypothesis Testing Stage
Have we arrived yet at the point where hypotheses about the classification of ectodermal dysplasia syndromes can be proposed to explain their etiopathogeneses? Probably not. Many scientists might argue that the molecular evidence gives definitive answers about classification, but all it does is posit another classification profile; the same old syndromes shuffled around and grouped in another way on the basis of yet another character (gene mutation). There is fresh insight and unexpected groupings are coming to attention, but there is still work to be done.
Before closing a discussion classification, a few more words are in order. First, it should be noted that most medical treatises that purport to be classifications are nothing more than catalogues—compilations of disorders that share one or more features. Syndromes of the Head and Neck [Gorlin et al., 1990], Heritable Disorders of Connective Tissues [Beighton, 1993], and Genetic Skin Disorders [Sybert, 1997] are examples of catalogues, and thankfully the authors do not claim anything different. There is nothing wrong with catalogues; most are useful for clinicians seeking diagnoses, potential causes, and information for counseling. They are not, however, classifications regardless of what contrivance authors use to group the entities contained therein. Second, until now classifications of ectodermal dysplasia syndromes have been quite artificial. The seminal works of Freire-Maia and Pinheiro 1984 and Pinsky 1975 must be looked at as such and taken to the next step—hypothesis testing. That is to say, the syndromes (specimens) in the classification need to be validated: are they real, have the best descriptors been used, do they overlap too much with the collective descriptors of other syndromes? Then, newly discovered syndromes have to be plugged into the classification schemata to see if and where they fit. Only after this test of time will we know whether or not the classification holds water.
THE STAKE HOLDERS
One of the pundits in the early days of dysmorphology (M. Michael Cohen Jr.) said that it would be easier to get one dysmorphologist to use another's toothbrush than his nomenclature. This may be true of proposed classifications of ectodermal dysplasia syndromes as well, because the interests of stakeholders are so different. There are several groups with vested interests in such classifications: patients, clinicians, scientists, foundations, insurance carriers, and government.
The interests of patients and clinicians largely overlap; from any classification they want clear-cut diagnosis, prognosis, and etiology. Catalogues work well for these stakeholders; the problem is how to find the syndrome of interest among dozens or hundreds that might be listed in a given catalogue. Almost any polythetic classification that lists all known syndromes (even those reported only once in the literature) suffice, as long as someone is able to peruse the lists and find a match.
Scientists, including the ubiquitous clinician–scientists, want classifications that can be proposed, tested, and verified in the laboratory. A given classification is unlikely to be widely accepted in scientific circles, however, because scientists from various disciplines use such different characters to classify things. Some will want a classification based on gene mutation (genomic classification), others will want one based on variations in protein structure (proteinomic or biochemical classification), still others will want one based on homology of biological structures (taxonomic classification). And, the classifications are unlikely to overlap; that is, similar specimens in their collections will be grouped differently. The situation is analogous to asking several people to separate a collection of bottles into discrete sub-groups. Some might separate the bottles based on size (small, medium, large), some might separate them by composition (plastic, glass, metal), some might separate them by shape (long necks, short necks, flat bottom, concave bottom), and still others might use color as a differentiating character. Each system of classification is legitimate, but the same bottles will not be grouped together from one to another.
Foundations look at classifications differently. Foundations are constituted and developed around a constituency. They raise funds and expend their time and resources on behalf of that constituency, but with time are approached by supplicants whose claim to be part of the constituency are more or less tenuous. The NFED, for example, was founded in 1980 to address the needs of ectodermal dysplasia. At the time, the founders probably thought there was dysplasia only one type of ectodermal dysplasia (hypohidrotic ectodermal dysplasia or the Chirst–Siemens–Tourraine syndrome) and constituency was quite limited. In short order, the reality of the situation became clear, and the foundation expanded its mission to address the needs of people with ectodermal dysplasia(s), really ectodermal dysplasia syndromes. With its success, the foundation drew a wider constituency, but found its resources stretched by requests for funds (particularly for scholarship and treatment programs) from individuals with disorders that might not be considered ectodermal dysplasia syndromes, for instance individuals with isolated hypodontia or only trichodysplasia. Since hypodontia, as a case in point, may affect as many as 5% of the population [Jorgenson, 1980], the demand for the Foundation's limited resources were soon strained by applications for treatment funds to replace missing teeth.
It might seem that a change in mission or in name, from ectodermal dysplasias to ectodermal dysplasia syndromes, would ameliorate the dilemma of the Foundation's executives as they try to meet its fiduciary responsibilities. But, how would the foundation then meet address the question of legitimacy for applications from individuals with many syndromes with features that overlap those of ectodermal dysplasia syndromes? Is the Down syndrome an ectodermal dysplasia syndrome because it involves a higher than average frequency of hypodontia and microdontia [Cervenka, 1976]? Would the orofaciodigital syndromes be considered ectodermal dysplasia syndromes solely on the basis of hypodontia caused by disruption of the dental lamina by hypertrophied oral frenula (the embryonic anlage being inherently normal)? Do the superficial features of a disorder such as dyskeratosis congenita derive from an abnormal embryonic anlage, thereby qualifying it as an ectodermal dysplasia syndrome?
Insurance carriers are interested in a proper classification for much the same reason as foundations—how to expend limited resources. For planning purposes insurance carriers might also want to know how common ectodermal dysplasia syndromes are individually and as a group, especially since treatment for some, such as the AEC syndrome, is extensive. In projecting costs it will not do to rely on statements that this one or that one is more common than the others—a purely anecdotal assertion made about HED. Without proper classification, it is difficult to know which group of specimens is more common than others and impossible to estimate costs for treatment of the afflicted population.
Governmental agencies cannot go about the work of setting health care policies, determining entitlement, propagating demographic data, or establishing research priorities without a clear understanding of the disorders of interest.
Classification of ectodermal dysplasia syndromes has become complex over the years as the notions of dysplasias and syndromes have evolved, and the classifications that do exist are threatened with disarray by the onset of molecular biology and potential regrouping of disorders on the basis of genetic variation. Furthermore, the divergent interests of stakeholders make it unlikely that a single classification will be universally accepted. It is imperative, then, to address the classification of this fascinating group of disorders from more than one perspective and develop a scheme that will move each group of stakeholders forward to meet their respective charges.