Medical genetics: 2.

The diagnostic approach to the child

with dysmorphic signs
Review
Alasdair G.W. Hunter Synthèse

Abstract
Dr. Hunter is a consultant
DYSMORPHOLOGY IS THE BRANCH OF CLINICAL GENETICS in which clinicians and re- with the Eastern Ontario
searchers study and attempt to interpret the patterns of human growth and struc- Genetics Program, Children’s
tural defects. Reaching an accurate diagnosis for children with dysmorphic signs is Hospital of Eastern Ontario,
important to their families, because it makes available all the accumulated knowl- and Professor of Paediatrics,
edge about the relevant condition and may provide the family with the opportunity University of Ottawa, Ottawa,
for interaction with patient or parent support groups. I show in this review that Ont.
reaching a diagnosis in dysmorphology involves an approach that is not fundamen-
tally different from that of other medical disciplines. Cytogenetic and molecular This article has been peer reviewed.
techniques continue to improve our ability to make precise syndrome diagnoses;
CMAJ 2002;167(4):367-72
however, these tests are expensive and should be used selectively.
ß See related article page 373
Case 1
This boy was born at 38 weeks’ gestation weighing 2750 g (> 10th percentile). Initial Series editor: Dr. Judith Hall,
Professor of Pediatrics and Medical
feeding problems improved by 7 months, and he became quite a voracious eater by
Genetics, Department of Pediatrics,
the age of 6 years. His motor and language development were moderately delayed.
Children’s and Women’s Health
At the age of 10 years, a cranial CT scan showed mild bilateral frontal “atrophy.” A Centre of British Columbia,
G-banded (Giemsa-stained) chromosome study showed a lengthened short (i.e., “p”) Vancouver, BC
arm of chromosome 14 (14p+), which was interpreted as being the common normal
variant that is due to variable amounts of repetitive DNA in that region. Neither par-
ent had a 14p+ “variant” and the developmental delay was unexplained. At 12 years,
this boy was referred to a genetics clinic. He was at the 90th percentile for height and
had a head circumference and weight that were above the 97th percentile. Bifrontal
narrowing, a slight upslant to the palpebrae (eyelids), which were of normal length,
and some Brushfield spots (small white spots on the periphery of the iris) were noted.
His mouth was small, and his hands were at the 60th percentile for length.

Case 2
This girl was born at term weighing 2280 g (3rd–10th percentile). The family his-
tory was noncontributory. There were early feeding difficulties and marked
growth and developmental delay. Chromosome studies were done twice, be-
cause this girl’s appearance was reminiscent of that typical of Down’s syndrome.
At 4 years, she was diagnosed as having Williams1 syndrome, but at the age of
13 years this diagnosis was ruled out using a fluorescence in situ hybridization
(FISH) probe specific for the elastin gene on chromosome 7, which showed that
the gene was not missing from one of the chromosomes as is seen to occur in
Williams syndrome. The family moved and requested an assessment at the local
genetics clinic. At the age of 13 years, this girl’s history featured the recent onset
of seizures, long-standing behaviour problems including stubbornness, impul-
siveness, violent temper outbursts and self-abuse, and a disturbed sleep pattern.
She also had a history of putting objects in orifices. At 13 years, her height was
that of an 8 year old, and her head circumference was much lower than the third
percentile. Physical signs included mild brachycephaly, malar hypoplasia,
slightly upslanting palpebral fissures and scooped nasal shape, with the philtral
area running parallel to the angle of the nose. The upper lip was thin and the
mouth downturned. The lower lip was everted and the chin small. Her chest was
barrel-shaped with hypoplastic nipples. Her hands were small and blunt.

CMAJ • AUG. 20, 2002; 167 (4) 367

© 2002 Canadian Medical Association or its licensors

Hunter

T
he term dysmorphic is derived from the Greek of fetal movement. A general question about unusual or
words “dys” (disordered, abnormal, painful) and untoward events during the pregnancy will often uncover
“morph” (shape, form). Dysmorphology is a disci- parental beliefs about causation, and it is important to con-
pline of clinical genetics that studies and attempts to inter- sider these during counselling. Events surrounding the de-
pret the patterns of human growth and structural defects. livery, including evidence of fetal distress, hydramnios or
These include malformation (an intrinsic developmental oligohydramnios, birth weight, length and head circumfer-
anomaly, e.g., spina bifida), disruption (an event disrupting ence, and perinatal behaviour of the child may be impor-
intrinsically normal development, e.g., amniotic bands), tant and, if thought to be significant, original records
deformation (an external force altering the shape of devel- should be sought. A child with dysmorphic signs may be at
opment, e.g., face shape due to severe oligohydramnios) risk from the stress of birth, and later delay may be erro-
and dysplasia (abnormal growth and maturation of cells, neously attributed to birth injury.3 A careful developmen-
e.g., achondroplasia). However, the child with dysmorphic tal history with an emphasis on milestones, formal assess-
signs often does not have a major malformation, and he or ments and behaviour is also required. Medical records
she may simply have an appearance that is unusual com- should be sought to validate any diagnosis or treatment of
pared with the general population and out of keeping with a malformation.
that of unaffected close relatives. A syndrome is simply a
recognizable pattern of dysmorphic signs that have a com- The role of the dysmorphologist
mon cause.
A dysmorphologist will conduct a complete physical ex-
amination that includes all the body systems. Although ma-
Why strive for a diagnosis for a child with jor malformations will be noted, a great deal of the emphasis
dysmorphic signs? is devoted to the overall appearance (gestalt) and the pres-
ence of any minor anomalies (Table 1), defined as physical
Many children who are dysmorphic have significant in- variations that occur in less than 5% of the population but
ternal or external malformations or developmental delay, are of no clinical significance, in the face (e.g., upslanting
or some combination of these. The inability to provide a palpebrae, unusual ear helix, anteverted nares [nostrils]) and
definitive diagnosis and thus a presumptive cause, even other areas including the hands, genitalia and skin. Where
when having a diagnosis does not imply that there will be a possible, subjective observation should be supplemented by
therapeutic intervention, can greatly frustrate a family. A objective measurement and photographic records.
precise diagnosis makes available all the accumulated Most syndrome diagnoses are suggested by the gestalt
knowledge and experience of that condition and generally of the patient, in the same way that most people would
provides a better estimate of the risk of recurrence. It in- recognize a child with Down’s syndrome. However, the
forms prognosis and permits interventions that may pre- dysmorphologist has more interest in and experience of di-
vent, anticipate or more successfully treat complications.2 agnosis and has honed his or her skills in pattern recogni-
In many jurisdictions it facilitates getting support, such as tion. Uncommon single or combinations of anomalies can
financial and educational aid, and it allows families to inter- be used to direct a computer search. Notwithstanding an
act with specific support groups. Furthermore, an accurate initial impression, the dysmorphologist performs a de-
diagnosis is key to research into the identification of tailed examination and compares the observations and his-
causative genes, interventions and treatments. tory with those expected in the syndrome being consid-
ered or with the lead suggested by the database. It is
important not to rush too early to a diagnosis because,
How does one approach a diagnosis in once applied to a patient’s condition, labels are hard to re-
dysmorphology? move. Concerns about the subjectivity of some syndrome
diagnoses, especially at the mild end of a syndrome’s spec-
The approach in dysmorphology is not fundamentally trum, have led some dysmorphologists to explore more
different from that of other medical disciplines. An en- objective diagnostic approaches such as photogrammetry
quiry into the family history should extend beyond the and anthropometrics.4 Photogrammetry uses objective
findings in the immediate patient to look for partial signs measurements from standardized photographs, and an-
of a syndrome or other malformations that could result thropometry from standardized physical landmarks, to as-
from different unbalanced products of a familial chromo- sess patients objectively.
some rearrangement. The 3-generation pedigree should As is the case for the general population, the facial ap-
include a careful search for consanguinity. The pregnancy pearance of an individual with a syndrome is expected to
history should include details of previous losses, pregnancy change with growth and maturation. Sometimes the facial
planning, parental occupation and health at the time of appearance will become more characteristic; in others it be-
conception, exposure to drugs, medications and alcohol, comes less apparent with age (e.g., fetal alcohol syndrome,
and evidence of maternal fever, rash or illness, and details Beckwith-Wiedemann syndrome), in which case pho-

368 JAMC • 20 AOÛT 2002; 167 (4)

 The child with dysmorphic signs

tographs from infancy or early childhood of the patient and Referral to a dysmorphologist for assessment
parents may be helpful. Because of possible changes
through time, and the continued description of new syn- Most dysmorphologists are medical geneticists, increas-
dromes, follow-up of patients, especially newborns and ingly with certification from the Royal College of Physi-
young infants, may ultimately result in a diagnosis. cians and Surgeons of Canada, who are attached to genet-
Unfortunately, for many individuals with malforma- ics programs in major centres. The proportion of their
tions and/or an unusual appearance, with or without devel- time spent in dysmorphology will vary with the size of the
opmental delay, the history and examination do not sug- centre. It is important for the general practitioner to in-
gest an immediate diagnosis. If there are objective (e.g., form the patient and family, and the genetics clinic, why
absent thumb) and not overly common (e.g., developmen- the referral is being made (e.g., for diagnosis, for repro-
tal delay is common and nonspecific) findings, the dysmor- ductive counselling, a suspected syndrome). Many genetic
phologist can use computer software to elucidate possible centres will send out a family history questionnaire and
diagnoses from the over 3000 reported syndromes.5,6 Of- seek medical records before seeing a patient, and the
ten, photographs are available for comparison with the pa- process can be expedited if the relevant history and test re-
tient. However, these programs are systems for experts and sults are included with the referral. When and whether to
not expert systems.6 Several books describe many of the make a referral will depend upon the physician’s own com-
best-defined syndromes.1,7,8 Dysmorphologists will often fort in assessing a patient with developmental delay, dys-
circulate photographs of their undiagnosed patients to col- morphic signs and/or with a deteriorating medical condi-
leagues, and telemedicine will likely expand this approach. tion, and whether there is a question to be answered or a
A challenge for the dysmorphologist is the large number problem to be solved. If a person looks somewhat unusual,
of single-family reports of “syndromes” consisting of de- but is otherwise completely healthy and developmentally
velopmental delay and a “characteristic,” but in fact very normal, there is generally no reason to pursue matters fur-
subtle, difference in appearance. Some may simply repre- ther. A dysmorphologist is very unlikely to make a diagno-
sent nonspecific developmental delay combined with fam- sis9 in a developmentally delayed person unless there are
ily traits rather than a distinct syndrome. It should be some dysmorphic signs, but even in the absence of a diag-
noted that formerly used pejorative descriptors such as nosis the medical geneticist may be able to answer a fam-
gargoylism, FLK, happy puppet and elfin face are unac- ily’s concern regarding future pregnancies by using em-
ceptable today. piric recurrence data.

Table 1: A list of some minor anomalies that are used in syndrome delineation
Craniofacial Other body areas Skin
• Sagittal fontannelle • Bifid xiphoid • Hypopigmented patches
• Upslanting palpebrae • Unusual umbilical position • Aplasia cutis congenita
• Short palpebrae • 5th finger clinodactyly • Various nevi
• Anteverted nares • Excess nuchal skin • Hairy patch on lower spine
• Natal teeth • Supernumerary nipple • Sebaceous nevus
• Malar underdevelopment • Deep sacral dimple • Mild skin syndactyly
• Bifid uvula • Prominent heels • Café au lait spots
• Posteriorly rotated ears • Pectus excavatum • Pigment streaking
• Open metopic suture • Umbilical hernia
• Brushfield spots • Single palmar crease
• Ocular heterochromia • Single umbilical artery
• Flat philtrum • Shawl scrotum
• Hypoplasia anguli oris
• Single central inscisor
• Micrognathia
• Preauricular pits or tags
• Multiple hair whorls
• Epicanthic folds
• Wide/close-spaced eyes
• Low nasal bridge
• Hypodontia
• Missing lip frenulum
• Ear helix anomalies

CMAJ • AUG. 20, 2002; 167 (4) 369

Hunter

How is the laboratory helpful in rebirth in cytogenetics, and 3 approaches or techniques are
dysmorphology? commonly applied in dysmorphology. First, probes that are
specific to the locus, that is, the physical position of a gene
Cytogenetics on the chromosome, can bind to a segment of DNA, too
small to be visible by light microscopy, on a specific chro-
Cytogenetics is a mainstay of diagnosis in dysmorphol- mosome. If the segment of DNA is deleted, the FISH probe
ogy. However, chromosome studies are labour intensive and will not bind, thus demonstrating the deletion. Hence the
relatively expensive (> $400). Referral to a dysmorphologist term microdeletion syndrome that is applied to a growing list
may lead to the diagnosis of a single gene disorder or sug- of syndromes, including Prader-Willi, Angelman, Smith-
gest a syndrome with a known microdeletion, that is, a Magenis, Miller-Dieker and velo-cardio-facial/DiGeorge,
chromosome deletion not visible by standard G-banding, now often called del(22q11), several of which are quite com-
thus obviating the need for a standard karyotype. To be visi- mon.1 In some syndromes, cases may result from either
ble, a chromosome deletion or duplication probably in- FISH-detectable deletions or point mutations (e.g., Alagille,
volves at least 3–4 kilobases of DNA10 (perhaps 15–30 genes, Angelman). These tests should be requested because of clin-
depending upon the location and the chromosome). Given ical signs (dysmorphisms/behaviour) suggestive of the spe-
the high proportion of genes involved in the development cific syndrome, and when used in this way the diagnostic
and functioning of the brain, major malformations in a de- yield can be high.9
velopmentally normal child are not expected to have a cyto- Second, in whole chromosome painting (WCP), FISH
genetic cause. I have shown that a standard karyotype is very probes are specific to a complete individual chromosome,
unlikely to be informative in a developmentally delayed rather than a single locus, and will paint the entire chromo-
child who is found by a dysmorphologist to lack dysmorphic some. Different approaches include combining fluorescent
signs. Indeed, the yield of abnormal chromosome results is dyes to give each chromosome pair plus the X and Y chro-
proportional to the number of dysmorphic signs.9 mosomes a different colour on a single metaphase spread
Unless a nonchromosomal diagnosis is apparent, a and examining each chromosome separately, or a restricted
neonate or young infant with an unusual appearance or ma- group of chromosomes, in individual wells on a single slide.
jor malformations, or both, should have a standard chromo- WCP is very useful for identifying the origin of additional
some study. In the older infant or child, generally, one would chromosome material that is microscopically visible but not
also require some developmental delay. The earlier in the distinctive enough to be assigned to a specific chromosome.
metaphase that chromosomes are examined, the longer and It can also be used to search for light microscopically invisi-
less condensed they will be, revealing more bands and theo- ble (cryptic) translocations where suspicion of a chromo-
retically allowing detection of smaller abnormalities. A re- some abnormality remains, despite a normal standard kary-
peat karyotype may be justified in a patient who has signs otype. The exchange of similarly sized and banded material
suggestive of a chromosome abnormality, if a negative kary- between 2 chromosomes, which is not visible in a standard
otype was reported for fewer than 550–650 bands. black and white G-banded study, becomes visible because
Fluorescence in situ hybridization (FISH) has caused a of the exchange of different colours.

A B

Fig. 1: (A) Partial fluorescence in situ hybridization (FISH) metaphase from patient 1 showing the differentially coloured chromo-
somes 20 (turquoise) with a similar coloured segment attached to the chromosome 14 labelled 14p+. (B) Partial FISH metaphase
from patient 1 showing the chromosomes 20 with their short (p) arms labelled red, with a similar coloured segment attached to
the chromosome 14 labelled 14p+.

370 JAMC • 20 AOÛT 2002; 167 (4)

 The child with dysmorphic signs

Third, an exciting new development is that of FISH dysmorphic appearance. Examples include disorders of per-
probes specific to the subtelomeric region of the short (“p”) oxisomes (Zellweger syndrome, chondrodysplasia punctata),
and long (“q”) arm of each chromosome, the telomeres be- a number of abnormalities in the energy pathways that are
ing the ends of the chromosomes. The subtelomeric re- associated with brain cell migrational defects17 and disorders
gions of chromosomes are gene rich, subject to frequent of cholesterol metabolism (Smith-Lemli-Opitz syndrome).18
breaks and exchanges, and are notoriously difficult to see
on a standard G-banded karyotype (chromosome study). The cases revisited
The indications for the use of this technology are still
evolving, but it appears that up to 8% of children with dys- Case 1
morphic signs and moderate-to-severe delay will have a
subtelomeric abnormality that was not apparent on a stan- The eating history, bifrontal narrowness and small
dard karyotype.11 Subtelomeric probes are likely to be supe- mouth raise the possibility of Prader-Willi syndrome, but
rior to WCP techniques for uncovering cryptic transloca- the height, head circumference and hand size are against the
tions. 12 Syndromes originally described as single gene diagnosis. Although Brushfield spots are associated with
disorders have now been shown to be the result of sub- Down’s syndrome, they are common in the general popula-
telomeric chromosome changes.13,14 WCP and subtelomeric tion, there were no other findings of Down’s syndrome
FISH are expensive technologies that currently should be mentioned and the G-banded karyotype was normal. The
used at the discretion of dysmorphologists in consultation clue was the initial karyotype that was reported as being
with a cytogeneticist. normal, but with a 14p+ variant not found in the parents.
True variants tend to be stable and transmitted from a par-
Molecular (DNA) diagnostics ent. The geneticist ordered a repeat karyotype. The 14p+
was not typical of the usual normal variant; whole chromo-
Technologies and information from the human genome some painting was requested and showed that the large “p”
project have accelerated the discovery of genes responsible arm of chromosome 14 was caused by translocated material
for many genetic syndromes. Increasingly, mutation analy- from chromosome 20 (Fig. 1A). Specific subtelomeric
sis can confirm a suspected syndrome diagnosis. At present, probes showed its origin was the short arm of chromosome
this is most practical for small genes and/or those that show 20 (20p) (Fig. 1B), and the child was thus trisomic for part
recurrent mutations at a small number of specific codons (a of chromosome 20p, which accounted for his mild dysmor-
codon is one of 61 triplet combinations of DNA bases that phic signs and moderate developmental delay. The parents
specifies one of the 20 amino acids or one of the 3 non- were relieved to finally have an explanation and the ex-
sense [stop] codons). For larger genes, and those with mu- tended family was informed that other family members
tations that occur at many different sites, current technol- were not at risk of having similarly affected children.
ogy remains slow, labour intensive, expensive and unable to
find all mutations. This situation will improve with the fur-
ther development of computer-driven chips and microar-
rays.15 Availability is not a good reason to order mutation
analysis. An unambiguous clinical diagnosis does not re-
quire molecular confirmation. Molecular diagnosis may be
indicated if the parents are at risk of having a second af-
fected child and plan prenatal diagnosis, where there is
some doubt as to the diagnosis, or in cases where predictive
testing for a familial genetic disease may be of benefit. In all
cases, the practitioner must be aware of the limitations of
the testing, such as causal heterogeneity of a syndrome or
the frequent inability to find all mutations in a gene, and
the broader implications surrounding DNA diagnostics.16

Biochemical laboratory testing

The traditional association between the biochemical lab-

oratory and dysmorphology has occurred with the storage
disorders such as the mucopolysaccharidoses (e.g., Hurler
syndrome). However, there is a growing list of inherited Fig. 2: Partial lateral photograph of patient 2 showing the un-
biochemical diseases that are associated with dysmorphic derdeveloped malar region, the philtrum running parallel to
signs. Thus, biochemical genetic disease should not be dis- the nose and an everted lower lip, which are all characteristic
missed simply because of the presence of malformations or a of Smith-Magenis syndrome.

CMAJ • AUG. 20, 2002; 167 (4) 371

Hunter

8. Winter RB, Donnai D, editors. Congenital malformation syndromes. London:

Chapman & Hall Medical; 1995.
9. Hunter AGW. Outcome of the routine assessment of patients with mental re-
tardation in a genetics clinic. Am J Med Genet 2000;90:60-8.
10. Gardner RJM, Sutherland GR, editors. Chromosome abnormalities in genetic
counselling. 2nd ed. New York: Oxford University Press; 1996. p. 269.
11. Knight SJL, Regan R, Nicod A, Horsley SW, Kearney L, Homfray T, et al.
Subtle chromosomal rearangements in children with unexplained mental re-
tardation. Lancet 1999;354:1676-81.
12. Schaffer LG, Kashork CD, Bejjani BA. What FISH methodology should I or-
der? [lecture]. Symposium on Cytogenetic Diagnostic Dilemmas, annual clin-
ical genetics meeting of the American Board of Medical Genetics; 2001 Mar
1-4; Miami.
13. Herens C, Jamar M, Alvarez-Gonzales ML, Lesenfants S, Lombet J, Bon-
nivert J, et al. Private multiple congenital anomaly syndromes may result from
unbalanced subtle translocations: t(2q;4p) explains the Lambotte syndrome.
Am J Med Genet 1997;73:127-31.
14. Verloes A, Lesenfants S, Jamar M, Dideberg V, Herens C. GOMBO syn-
drome: another “pseudorecessive” disorder due to a cryptic translocation. Am
J Med Genet 2000;95:185-6.
15. Sinclair A. Genetics 101: detecting mutations in human genes. CMAJ 2002;
167(3):275-9.
16. Baird PA. Identifying people’s genes: ethical aspects of DNA sampling in
populations. Perspect Biol Med 1995;38:159-66.
Fig. 3: Partial FISH metaphase from patient 2 showing the 2 17. Hunter AGW. The brain. In: Stevenson RE, Hall JG, Goodman RM, editors.
Human malformations and related anomalies. Vol 2. New York: Oxford Univer-
chromosome 17s highlighted by a pink fluorescent stain. The sity Press; 1993. p. 38-51.
chromosome marked with the white arrow is missing the stain 18. Nowaczyk MJM, Whelan DT, Heshka T, Hill RE. Smith-Lemli-Opitz syn-
for the Smith-Magenis locus and has only the control fluores- drome: a treatable inherited error of metabolism causing mental retardation.
CMAJ 1999;161(2):165-70.
cence. The normal chromosome has 2 fluorescent points. 19. Smith ACM, Gropman A. Smith-Magenis syndrome. In: Cassidy SB, Allan-
son JE, editors. Management of genetic syndromes. New York: John Wiley &
Sons; 2001. p. 363-87.
20. De Leersnyder H, de Blois MC, Vekemans M, Sidi D, Villain E, Kindermans
Case 2 C, et al. β1-adrenergic antagonists improve sleep and behavioural distur-
bances in a circadian disorder, Smith-Magenis syndrome. J Med Genet 2001;
38:586-90.
The geneticist had not seen a patient with Smith-
Magenis syndrome, but the behavioural pattern and subtle
facial features (Fig. 2) in this 13-year-old girl suggested the Additional resources
diagnosis. This chromosome microdeletion syndrome was
The following resources, which are included in the reference list, are excellent
confirmed by FISH studies with a probe for 17p11.2 (Fig. 3) general texts to aid in the diagnosis of syndromes and to provide some basic infor-
that showed this region to be missing from one chromo- mation on each.
Gorlin RJ, Cohen MM Jr, Hennekam RCM, editors. Syndromes of the head and
some 17. The diagnosis allows the possibility of providing a neck. 4th ed. New York: Oxford University Press; 2001.
more specific intervention. In the younger child, this in- Winter RB, Donnai D, editors. Congenital malformation syndromes. London: Chap-
man & Hall Medical; 1995.
cludes an intensive program of speech and signing, early in- Hunter AGW. Outcome of the routine assessment of patients with mental retarda-
tervention to manage maladaptive behaviours, and help with tion in a genetics clinic. Am J Med Genet 2000;90:60-8.
feeding and swallowing difficulties.19 Later education can fo- The following reference places a useful emphasis upon the management of some of
the more common syndromes.
cus on relative strengths such as receptive language, long- Cassidy SB, Allanson JE, editors. Management of genetic syndromes. New York: John
term memory and interest in computers, while working Wiley & Sons; 2001.
around major deficits in short-term memory and sequential Web site resources
processing. Furthermore, there is growing evidence that • Online Mendelian Inheritance in Man (OMIM) provides a catalogue of hu-
man single gene disorders: www3.ncbi.nlm.gov/Omim/
melatonin and β1-adrenergic antagonists may aid sleep by • GeneTests•GeneClinics provides excellent reviews of a broad range of ge-
altering the abnormal circadian melatonin secretion.19,20 netic diseases and syndromes: www.geneclinics.org
• National Organization for Rare Diseases (NORD) provides information and
Competing interests: None declared. links to support groups: www.rarediseases.org

References Correspondence to: Dr. Alasdair Hunter, Eastern Ontario

Genetics Program, Children’s Hospital of Eastern Ontario,
1. Jones KL, editor. Smith’s recognizable patterns of human malformation. 5th ed. 401 Smyth Rd., Ottawa ON K1H 8L1; fax 613 738-4822;
Philadelphia: WB Saunders; 1997. heddalasdair@aol.com
2. Cassidy SB, Allanson JE, editors. Management of genetic syndromes. New York:
John Wiley & Sons; 2001.
3. Hall DMB. Birth asphyxia and cerebral palsy: birth asphyxia is hard to define
but is rarely the cause of cerebral palsy. BMJ 1989;299:279-82.
4. Allanson JE. Objective techniques for craniofacial assessment: What are the Articles to date in this series
choices? Am J Med Genet 1997;70:1-5.
5. POSSUM. Version 5.2. Melbourne (Australia): Murdoch Institute; 2000.
6. Winter RM, Baraitser M. Dysmorphology photo library on CD-ROM. Version
Polifka JE, Friedman JM. Medical genetics: 1. Clinical teratol-
2.2. Oxford Medical Databases. Oxford: Oxford University Press; 2000. ogy in the age of genomics. CMAJ 2002;167(3):265-73.
7. Gorlin RJ, Cohen MM Jr, Hennekam RCM, editors. Syndromes of the head and
neck. 4th ed. New York: Oxford University Press; 2001.

372 JAMC • 20 AOÛT 2002; 167 (4)