TO THE EDITOR:
In a recent issue of Pediatrics(1), Dr. Susan Astley reported on a
personally conducted comparison of her 4-digit diagnostic system for
children prenatally exposed to alcohol with our clarifications of the
Institute of Medicine (IOM) criteria for diagnosis of fetal alcohol
spectrum disorders (FASD)(2). A thoughtful reading of the entire
literature regarding the teratogenicity of ethanol and objective
assessment of our system for diagnosis in the FASD continuum reveal that
Dr. Astley’s conclusions are inaccurate, based on faulty and misleading
data analysis and a narrow understanding of the developmental biology
underlying the features of FASD and other dysmorphic syndromes associated
with developmental disabilities. We are surprised by the publication in
Pediatrics of this attempt to discredit years of clinical and population-
based research on the teratogenic effects of alcohol, much of it based on
the recommendations of the Institute of Medicine (IOM). In this letter we
will address many of the issues raised by Dr. Astley.
GENERAL COMMENTS:
To begin, Astley assumes that her 4-digit diagnostic system is the “gold
standard” for diagnosis in FASD. An assumption that any variance of our
criteria from hers leads to incorrect diagnoses is not demonstrated by our
data or her analysis. She has chosen to include in her comparison only
children diagnosed with full-blown FAS; this invalidates many of her
arguments, since the point of our article was to delineate the entire
clinical spectrum of FASD. In fact, a comparison of children diagnosed
with FAS and Partial FAS by all existing diagnostic systems would more
likely display more similarities than differences. As Dr. Astley points
out, many investigators have an ongoing interest in this area. We welcome
third party evaluation of all diagnostic systems, especially by busy
clinicians who routinely evaluate children exposed prenatally to alcohol
as part of their clinical practices. It is our opinion that refinement of
methods for the clinical characterization of FASD should focus on both
accuracy and practical clinical applicability. Fine-tuning the diagnosis
of FASD is an important and ongoing process, continually being re-
evaluated as new data are analyzed. Rather than the current published work
being the end point of such investigation, it should be the beginning.
This research should be conducted in a scientific and collegial
atmosphere, and we encourage Dr. Astley to add to this discourse in a
positive way.
Despite arguing the need for a multidisciplinary approach, Astley’s
system ultimately rests on the dysmorphology assessment of the face. She
contends that “the validity of an FASD diagnosis rests solely on the
specificity of the facial phenotype for the exposure (alcohol) and outcome
(FAS).” This statement places undue emphasis on features that are not
unique to alcohol exposure. Requiring 3 cardinal facial features rather
than 2 does not make “the face” any more specific for alcohol
teratogenicity. In fact, extensive animal as well as human data indicate
that the face of fetal alcohol syndrome is a non-specific part of the
holoprosencephaly spectrum, accompanying a midfacial deficiency directly
related to an insult to the developing median forebrain that can be either
genetic or teratogenic in origin(3-6). Holoprosencephaly can accompany
mutation in or downregulation of a number of early developmental genetic
pathways in the brain(7-10). Recent research indicates that one important
mechanism of alcohol teratogenicity is the downregulation of these genetic
pathways(11-13). The facial features observed in animal models are also
highly variable depending on timing of the insult(3). Thus, stating that
the three cardinal facial features are 100% specific for FAS is neither
borne out by the literature nor clinical experience. Despite Dr.
Astley’s assertion that “an astute clinician would not mistake FAS for
William’s [sic] syndrome,” many malformation syndromes are
phenotypically variable, including Williams syndrome(14,15) and FASD.
Astley’s comments about the absolute specificity of the face of FAS
reflect a general lack of appreciation of the entire field of
developmental disabilities and the causation and variability of
malformation syndromes. Although alcohol may contribute to the phenotype
of a child with a known genetically determined malformation syndrome, such
data are lacking in the literature. In fact, the non-specificity of the
facial phenotype implies the absolute necessity of a multidisciplinary
diagnostic approach, including input from a pediatrician knowledgeable in
the entire spectrum of developmental disabilities.
ERRORS IN FACT:
While there are many factual errors in Astley's paper, we will highlight
only a recurring few. First, she writes that “Hoyme, et al. refer to
their diagnostic guidelines as a clarification of the 1996 IOM criteria.
The two sets of guidelines are authored by separate groups, however.” In
fact, the second author of our paper (May), who is also
the principal investigator on the NIAAA grants that have produced much of
the data leading to development of the diagnostic system in question, was
a member of the IOM committee on FAS for its entire existence, 1994-1996,
and authored several sections of the report16. Second, in numerous places
in her article, Dr. Astley fails to mention that our diagnostic scheme is
a more clinically-specific version of the exact four-category
classification system that was recommended by the IOM. Third, Astley
repeatedly calls our diagnostic methods a “gestalt” diagnostic
technique, which is meant to imply minimal organization and quantification
of FASD diagnoses. In fact our approach is well structured and quantified,
based on three decades of clinical experience of each of our senior
dysmorphologists working on research projects in populations throughout
the world. Fourth, she writes that one validation of the 4-digit facial
classification system is that it has produced high correlations with the
“most disabling feature” of FAS, significant cognitive and behavioral
impairments. Our IOM-based system provides for evaluation of not only
facial features, but also of other clinical features associated with FASD
through a structured, total dysmorphology score. This score has produced
higher, more significant correlations with low intelligence and behavioral
traits characteristic of FASD(17,18) than 4-digit studies of the face
alone(19).Total IQ, Verbal IQ, Performance IQ and behavioral problems have
also been significantly linked to specific aspects of our dysmorphology
assessments through highly significant correlations with a series of
measures of quantity, frequency, and timing of drinking in populations
where complete and accurate reporting has been achieved (20,21). Yet
Astley argues that it is nearly impossible to link alcohol quantity and
frequency to specific prenatal alcohol damage and that our system requires
“excessive” documentation of maternal drinking. Is the fact that the 4
-digit system requires a lower extent and lower quality of evidence for
quantity, frequency and timing of prenatal alcohol consumption a good
thing?
CONTRADICTIONS:
In addition to inaccuracies, Dr. Astley’s arguments are circular and
contradictory. First, at one point in the paper she writes that “since
the [Hoyme et al.] diagnosis is
based solely on the physical features of growth, facial anomalies and
structural brain abnormalities…an interdisciplinary clinical
team…would have no role in the derivation of an FAS diagnosis.”
Second, she states that we use the “gestalt” method of diagnosis,
stating that we have not standardized our clinical procedures and
observations in studies of the various populations in which we work.
Third, Astley writes in multiple places that the data used to illustrate
our diagnostic system in our paper are from “a non-representative
population base” (South Africans and Native Americans). None of these
assertions are accurate, and later she contradicts all three of them by
writing that “key strengths” of our study design and methods used to
formulate the diagnostic guidelines “include the use of skilled
multidisciplinary teams led by experts in the field of FASD diagnosis; use
of standardized objective measurement tools to enhance reliability; and
access to a reasonably large, population-based study sample.” Regarding
the populations we have studied, these populations are non-representative
to whom? These two populations exhibit rates of FASD that are among the
highest in the world and have proven to provide high degrees of
candor/honesty in reporting alcohol use and other maternal risk factors
(22,20). Although there are others, the fourth and last contradiction is
criticism of our adoption and use of Astley’s lip-philtrum guide to
quantify midfacial morphology. Speculating, Dr. Astley states that we must
have used the Caucasian version in South Africa, implying that the African
American guide would have been more appropriate. We did use the Caucasian
version in South Africa and the Plains, feeling it was the best guide
available, since both populations have experienced substantial degrees of
racial admixture with Caucasians. She also contradicts herself by writing
that the “direction of error [using the Caucasian scale] would be to
underestimate the prevalence of the FAS facial phenotype.” But later she
writes: “The extraordinarily high FAS prevalence rates (40.5 - 46.4
cases per 1000 subjects) reported by May et al. for a South African
community were based on FAS diagnoses that Hoyme et al. reported were
inaccurate and overestimated.” What Dr. Astley fails to recognize is
that all of our studies of other populations utilize normal control groups
of children and mothers randomly selected from those same populations.
These control groups provide the standardization that she said should be
done through adaptation of tools for the specific population. It is
equally legitimate and accurate, and preferable due to time, sampling, and
cost efficiency, to compare both subpopulations (FASD and random controls)
to a common standard whether or not the standard was developed
specifically from that exact population(23).
CONCLUSIONS:
Rather than setting a standard for FAS diagnosis, the 4-digit system tends
to eliminate the value of clinical judgment beyond that which is applied
to a series of
Likert-scaled items. The bar is set so high in terms of facial morphology
(three cardinal features) and head circumference (< 2 1/2 percentile)
that it is no surprise that Astley and colleagues claim a specificity of
100% when using these cut points. But one might ask how many true positive
FAS cases are missed, particularly among offspring of binge drinkers who
produce exposed children with facial features that are not always as
consistent as required by the 4-digit system. Seventy percent of FAS cases
that we diagnose in most populations have all three key facial features
along with meeting other physical, behavioral, intelligence and maternal
exposure criteria. And we wonder how many false negative FAS cases the 4-
digit system would produce in individuals from populations with a normally
large head size (e.g. American Indians)(24-26). The 4-digit system is
hardly without flaws in spite of the statistical measures employed by
Astley. Therefore to hold up the 4-digit system as superior to our system
is unjustified. Our criteria for diagnosis of FASD, which Dr. Astley
refers to as the “Hoyme” methodology, were indeed coauthored by a
truly interdisciplinary research team, including representatives from the
disciplines of dysmorphology, human genetics, epidemiology, psychology,
neuropsychology, educational diagnostics, special education, and social
work. All have worked extensively with diverse populations on three
continents. Not only does our critic need to review more carefully the
total body of published literature in this area, but she needs to spend
more time with an interdisciplinary team of clinicians, practice more
“shoe leather” epidemiology in the field with diverse populations, and
be more understanding of the many possible faces and variations of FAS and
other FASD. This might help her appreciate the need for more sensitivity
in accommodating ethnic and racial variation, unique patterns from binge
drinking, and minor differentiation arising from other real world
variables. The key is to understand and accommodate substantial human
phenotypic diversity due to prenatal and postnatal factors without losing
accuracy of diagnosis in FASD.
REFERENCES:
1. Astley SJ. Comparison of the 4-digit diagnostic code and the Hoyme
diagnostic guidelines for fetal alcohol spectrum disorders. Pediatrics.
2006;118:1532-1545
2. Hoyme HE, May PA, Kalberg WO, Kodituwakku P, Gossage JP, Trujillo
PM, Buckley DG, Miller JH, Aragon, AS, Khaole N, Viljoen DL, Jones KL,
Robinson LK. A practical clinical approach to diagnosis of fetal alcohol
spectrum disorders:
clarification of the 1996 institute of medicine criteria. Pediatrics
2005;115:39-47
3. Sulik KK. Genesis of alcohol-induced craniofacial dysmorphism. Exp
Biol Med. 2005;230:366-375
4. Sulik KK, Johnston MC. Embryonic origin of holoprosencephaly:
interrelationship of the developing brain and face. Scan Electron Microsc.
1982; 309-322
5. Webster TA, Gibson BW, Keng T, Biemann K, Schimmel P. Primary
structures of both subunits of Escherichia coli glycyl-tRNA synthetase. J
Biol Chem. 1983;258:10637-10641
6. Edison R, Muenke M. The interplay of genetic and environmental
factors in craniofacial morphogenesis: holoprosencephaly and the role of
cholesterol. Congenit Anom (Kyoto). 2003;43:1-21
7. Ribeiro V, Reis H, Teixeira JF, Xavier JA, Cruz R. Middle
interhemispheric variant of holoprosencephaly: a case report. Rev Neurol.
2006;42:245-246
8. El Jaick KB, Powers SE, Bartholin L, Myers, KR, Hahn J, Orioli IM,
Ouspenskaia M, Lacbawan F, Roessler E, Wotton D, Muenke M. Functional
analysis of mutations in TGIF associated with holoprosencephaly. Mol Genet
Metab. 2006
9. Ming Z, Griffith BL, Breese GR, Mueller RA, Criswell HE. Changes
in the effect of isoflurane on N-methyl-D-aspartic acid-gated currents in
cultured cerebral cortical neurons with time in culture: evidence for
subunit specificity.
Anesthesiology, 2002;97:856-867
10. Wallis DE, Muenke M. Mutations in holoprosencephaly. Hum Mutat.
2000;16:99-108
11. Ahlgren SC, Thakur V, Bronner-Fraser M. Sonic hedgehog rescues
cranial neural crest from cell death induced by ethanol exposure. Proc
Natl Acad Sc USA. 2002;99;10476-10481
12. Eagleson GW, Dempewolf RD. The role of the anterior neural ridge
and Fgf-8 in early forebrain patterning and regionalization in Xenopus
laevis. Comp Biochem Physiol B Biochem Mol Bio. 2002;132:179-189
13. Tyas DA, Pearson H, Rashbass P, Price DJ. Pax6 regulates cell
adhesion during cortical development. Cereb Cortex, 2003;13:612-619
14. von Beust G, Laccone FA, del Pilar AM, Wessel A. Clinical aspects
and genetics of Williams-Beuren syndrome. Clinical and molecular genetic
study of 44 patients
with suspected Williams-Beuren syndrome. Klin Padiatr. 2000;212:299-307
15. Joyce CA, Zorich B, Pike SJ, Barber JC, Dennis NR. Williams-
Beuren syndrome: phenotypic variability and deletions of chromosomes 7,
11, and 22 in a series of 52 patients. J Med Genet. 1996;33:986-992
16. Stratton KR, Howe CJ, Battaglia FC. Institute of Medicine,
Division of Biobehavioral Sciences and Mental Disorders, Committee to
Study Fetal Alcohol Syndrome, & National Institute on Alcohol Abuse
and Alcoholism. Fetal alcohol syndrome diagnosis, epidemiology,
prevention, and treatment. Washington, D.C: National Academy Press; 1996
17. Kodituwakku P, Coriale G, Fiorentino D, Aragon AS, Kalberg WO,
Buckley D, Gossage JP, Ceccanti M, May PA. Neurobehavioral characteristics
of children with fetal alcohol spectrum disorders in communities from
Italy: Preliminary results. Alcohol Clin Exp Res. 2006;30:1551-1561
18. Hoyme HE, Ervalahti N, Fagerlund Å, Autti-Rämö I, Loimu L and
Korkman M. Relationship between dysmorphic features and general cognitive
functioning in children with fetal alcohol spectrum disorders. David W.
Smith Morphogenesis
and Malformations Workshop, Lake Arrowhead, CA, September, 2006.
Proceedings of the Greenwood Genetic Center (in press).
19. Astley SJ, Clarren SK. Measuring the facial phenotype of
individuals with prenatal alcohol exposure: correlations with brain
dysfunction. Alcohol Alcohol. 2001;36:147-159
20. May PA, Gossage JP, Brooke LE, Snell CL, Marais AS, Hendricks LS,
Croxford JA, Viljoen DL. Maternal risk factors for fetal alcohol syndrome
in the Western Cape Province of South Africa: a population-based study. Am
J Public Health
2005;95:1190-1199
21. May PA, Gossage JP, Marais AS, Adnams CM, Hoyme HE, Jones KL,
Robinson LK, Khaole NCO, Snell C, Kalberg WO, Hendricks L, Brooke L,
Stellavato C, Viljoen DL. The epidemiology of fetal alcohol syndrome (FAS)
and partial FAS in a South African Community: A third study of a very high
prevalence area. Drug Alcohol Dep 2006; (in press)
22. Viljoen D, Croxford J, Gossage JP, Kodituwakku PW, May PA.
Characteristics of mothers of children with fetal alcohol syndrome in the
Western Cape Province of South Africa: a case control study. J Stud
Alcohol. 2002;63:6-17
23. May PA, Brooke L, Gossage JP, Croxford J, Adnams C, Jones KL,
Robinson L, Viljoen D. Epidemiology of fetal alcohol syndrome in a South
African community in the Western Cape Province. Am J Public Health.
2000;90:1905-1912
24. Adams MS, Niswander JD. Birth weight of North American Indians.
Hum Biol 1968;40:226-234.
25. Adams MS, Niswander JD. Birth weight of North American Indians: A
correction and amplification. Hum Biol 1973;45:351-357
26. Maynard JE, Hammes LM. A study of the growth, morbidity and
mortality among Eskimo infants of western Alaska. Bull World Health Org.
1970;42:613-622.
Conflict of Interest:
None declared
eLetters: ![[Read eLetters]](/icons/misc/sectionDOWN.gif){kind=icon}
