American Journal of Medical Genetics Part C (Seminars in Medical Genetics) 160C:104110 (2012)

A R T I C L E

Working Up Autism:The Practical Role of Medical GeneticsFIORELLA GURRIERI1*

The autism spectrum disorders (ASD) comprise a group of neurobehavioral phenotypes of heterogeneousetiology. In spite of a worldwide extensive research effort to unravel the genetic mystery of autism, medicalgeneticists are still facing an embarrassing lack of knowledge in dealing with the diagnosis, and consequentlyprognosis, of a child with autism. However, some lessons can be learned from accumulating experience in theclinical and molecular genetic evaluation of children with this condition. Patient evaluation, indications formolecular testing and counseling are the three aspects that will be discussed in this review.

2012 Wiley Periodicals, Inc.

KEYWORDS: autism spectrum disorders; molecular tests; physical examination; genetic counseling; CGH microarray

How to cite this article: Gurrieri F. 2012. Working up autism: The practical role of medical genetics.Am J Med Genet Part C Semin Med Genet 160C:104110.

INTRODUCTION

Autism spectrum disorders (ASD)

include a group of neurobehavioral

conditions that have in common im-

pairment in socialization and communi-

cation, restriction and peculiarity of

interests and stereotypic behavior

[DiCicco-Bloom et al., 2006]. The

diagnosis is usually made no earlier

than 18 months without upper limits

(41 months on average) [Autism and

Developmental Disabilities Monitoring

Network Surveillance Year 2000 Prin-

cipal Investigators; Centers for Disease

Control and Prevention, 2007]. ASD

affects about 1:110 children, with a

4:1 male/female ratio. In about 70% of

cases the onset is gradual whereas in the

remaining 30% it is of regressive nature.

In spite of the more or less stringent

diagnostic criteria established by the

Diagnostic and Statistical Manual of

Mental Disorders, 4th edition (DSMIV)

[American Psychiatric Association

only because the clinical spectrum is

highly variable, but also because the phe-

notype may evolve and change over

time. For example, language, which is

usually delayed, can be developed at a

good level in 10% of ASD children, or

can remain mildly impaired in about

30%. On the other hand, 10% of

children develop no language at all and

40% have severely impaired language

[Howlin et al., 2004]. Along the same

line, about 20% of ASD children end up

in a regular school program, still with

some social impairment, whereas the

majority (up to 75%) remain within the

autistic spectrum phenotype; only a small

percentage (no more than 5%) may

completely recover [Zappella, this issue].

In addition, ASD is commonly as-

sociated with other medical issues,

such as epilepsy (about 30% of cases)

[Tuchman and Rapin, 2002], intel-

lectual disability (between 30% and

80%) [Fombonne, 1999; Chakrabarti

and Fombonne, 2005] and attention def-

icit hyperactivity disorders (ADHD),

which add complexity to the clinical

picture and make it difcult to reach a

causal diagnosis, without which there is

no possible clue to prognosis and family

counseling.

Therefore, it is crucial to fully com-

prehend the patients presentation both

at the neuropsychological, developmen-

tal, and behavioral picture. In addition to

that, laboratory testing, to detect possi-

ble genetic and metabolic alterations, is

also a relevant part of the diagnostic

work-up in ASD.

The purpose of thiswork is to brief-

ly describe the inuence of genetic and

environmental factors in ASD, and to

put more emphasis on aspects of the

clinical genetic evaluation, molecular

diagnosis, and counseling.

CLINICAL GENETICEVALUATION

Once the neuropsychological diagnosis

of an ASD disorder is established, it

is crucial to proceed with a medical

examination in order to detect con-

comitant issues that require treat-

ment. Among those, seizures, feeding

and gastrointestinal problems, sleep

Fiorella Gurrieri is associate professor of Medical Genetics at the Catholic University of Rome,School ofMedicine. She is involved in clinical andmolecular genetics.Her research is focusedon thegenetic aspects of autismand specically she has investigated quantitative andqualitative genomicalterations and their phenotypic consequences. A second research eld includes the application ofnew genomic technologies to identify the causes of congenital defects.

*Correspondence to: Fiorella Gurrieri, Istituto di Genetica Medica, Universita` Cattolica delS. Cuore, L.go F. Vito 1, 00168 Roma, Italy. E-mail: fgurrieri@rm.unicatt.it

DOI 10.1002/ajmg.c.31326Article rst published online in Wiley Online Library (wileyonlinelibrary.com): 12 April 2012

2012 Wiley Periodicals, Inc.

disturbances and dental abnormalities

[Olivie, 2012].

In addition, a clinical genetics eval-

uation should be considered in ASD

children in order to identify syndromic

forms of autism, identify familial cases,

and drive diagnostic testing.

This workup has been recom-

mended by the American Academy of

Pediatrics [Johnson and Myers, 2007]

and the American College of Medical

Genetics [Schaefer and Mendelsohn,

2008; Shen et al., 2010; Roesser, 2011].

The rst duty of the clinical geneti-

cist in evaluating a child with autism is to

dissect the etiologic heterogeneity of

ASD by distinguishing essential autism

from complex (syndromic) autism

[Miles, 2011].

Essential autism is usually present in

about 75% of cases and is characterized

by absence of dysmorphic features,

higher male-to-female ratio (6:1),

higher sibling recurrence risk (up to

35%) and positive family history (up to

20% of cases).

Essential autism is usually

present in about 75% of cases

and is characterized by absence

of dysmorphic features, higher

male-to-female ratio (6:1),

higher sibling recurrence risk

(up to 35%) and positive

family history

(up to 20% of cases).

Complex, syndromic autism is usually

characterized by recognizable dysmor-

phic features, lower male-to-female

ratio (3.51), lower sibling recurrence

risk (46%), less frequent positive family

history (up to 9%) [Miles, 2011]. In this

latter group the prognosis is usually

worse.

The distinction between essential

and complex autism is important

because it implies a different prognosis

and a different recurrence risk for other

family members. In spite of all these

recommendations not all ASD children

usually undergo a clinical genetic evalu-

ation, but mainly those with evident

dysmorphic features, positive family his-

tory and intellectual disability.

As it turns out, this is not the best

practice, because in selected cases of

nonsyndromic ASD the recurrence

risk might be actually higher and parents

should be informed that the phenotype

in a second child can be even more

severe than in their rst child.

An additional duty of the clinical

geneticist is to collect information on

the family history in order to identify

in other relatives the occurrence of phe-

notypes that can be related to ASD. For

instance, family history can be positive

for alcoholism, depression, manic-de-

pression, obsessivecompulsive disor-

ders, substance abuse, seizures, anxiety

disorders, Tourrette-like motor tics, an-

orexia. These ndings occur in up

to 35% of ASD families [Miles et al.,

2003]. To investigate on family history

is important because the identication

of additional relatives in the ASD spec-

trum is suggestive of a higher recurrence

risk for the siblings of the proband.

On the other hand, the nding of an

environmental exposure reduces the re-

currence risk for the family, provided

that the environmental risk factor is

removed.

The clinical genetic evaluation can

recognize phenotypes related to known

genetic conditions such as fragile X syn-

drome, Rett syndrome, tuberous sclero-

sis, Angelman syndrome, SmithLemli

Opitz syndrome and others. This recog-

nition is crucial to drive appropriate

molecular testing.

On the other hand, the general

practice of testing all ASD patients for

FMR1mutations without a proper clin-

ical evaluation only yields positive results

in less than 0.5% of cases [Roesser,

2011]. It should also be kept in

mind that in most monogenic forms of

essential ASD, such as those caused by

mutations in neuroligins, neurexines,

SHANK3, FOXP2 and many others

[Miles, 2011] there is no recognizable

phenotype that drives the testing to-

wards one gene or another.

GENETIC FACTORS

ASD is one of the most heritable neu-

ropsychiatric disorders, with an in-

creased recurrence risk (more than 20-

fold) in rst-degree relatives [Bayley

et al., 1995]. This observation points

to a major genetic contribution. How-

ever, despite signicant research, includ-

ing high throughput technique

applications, efforts have failed to iden-

tify genes of large-effect, whose identi-

cation could impact strongly the

diagnosis, prognosis, and counseling to

ASD families. The outstanding question

is: Where is the heritable component of

autism?

So far, more than 100 genes and 40

genomic loci have been reported in re-

lation to ASD [Betancur, 2011] and as-

sociated/overlapping phenotypes such

as intellectual disability, ADHD, epilep-

sy, and schizophrenia. None of these

genes, however, is responsible by itself

for a high percentage of cases of ASD.

Therefore, it is suggested that multiple

genes (of minor effect) in combination

with environmental factors, contribute

to this complex neurobehavioral pheno-

type. Because of this wide genetic het-

erogeneity, the diagnostic yield of single

gene testing strategies is quite low (less

than 1%).

In some cases, ASD is part of the

phenotypic expression of a single-gene

disorder, while in others it results from a

combination of common genetic factors

that add up to overcome a threshold. In

the former situation, a clinical diagnosis

needs to be done rst, in order to rec-

ognize the basic disorder and determine

proper molecular testing. Even an oli-

gogenic inheritance of multiple hypo-

morphic mutations in genes whose

severe alterations cause known genetic

syndromes (TSC1 and 2, UBE3A,

PTEN, MECP2, and SHANK3) has

been observed in ASD [Schaaf et al.,

2011]. This observation suggests a new

genetic model for ASD.

In general genetic alterations re-

sponsible for ASD can be classied

into three subgroups: cytogenetic alter-

ations detectable on standard karyotype

(up to 5%), copy number variants

(CNVs), which can be found in a

ARTICLE AMERICAN JOURNAL OF MEDICAL GENETICS PART C (SEMINARS IN MEDICAL GENETICS) 105

variable percentage of cases (between

10% and 35%), and single gene muta-

tions (less than 5%) [Miles, 2011].

Abnormalities of the standard kar-

yotype are commonly found in ASD

patients with dysmorphic features and

intellectual disability [Reddy, 2005].

These abnormalities have been reported

in almost all chromosomes with the

same frequency, except for the duplica-

tion of the 15q11-q13 region in the

form of inv dup (15), which seem

more specically associated with ASD

[Dykens et al., 2004]. This duplication

occurs in about 13% of patients, but its

incidence might be higher because some

interstitial duplications of this same

region can only be detected with

array-CGH.

Other classical chromosomal syn-

dromes, such as Down syndrome (up

to 7% of cases) and Turner syndrome,

as well as other sex chromosome disor-

ders have been associated with autism

[Creswell and Skuse, 1999; Kent et al.,

1999].

Submicroscopic alterations (CNVs)

can be found in about 10% of patients

from simplex families and less than 2% of

patients from multiplex ones [Sebat

et al., 2007]. About 7% are de novo.

Again, one expects to identify such

CNVs mostly in syndromal autism.

One main distinction that needs to be

made is between pathogenic CNVs

(usually de novo, large and with signi-

cant gene-content), polymorphicCNVs

(usually inherited, small and with poor

gene-content). A third category

includes CNVs of unknown clinical sig-

nicance (medium sized, with signi-

cant gene-content, usually inherited

from a parent with a border-line pheno-

type). The most common ASD-related

CNVs are the 15q11-q13 duplication,

the 7q21 deletion, and the 16p11.2

microdeletion with its reciprocal micro-

duplication [Weiss et al., 2008], but oth-

er ones are being recurrently reported.

All CNVs specically associated with

ASD are annotated in a dedicated data-

base at projects.tcag.ca/autism_500k/:

with the highest stringency, a total of

276 CNVs result as being specic for

ASD in this database.

Because in some instances the same

CNVs have been associated with vari-

able phenotypes including autism,

atypical autism, schizophrenia, dyslexia,

intellectual disability, ADHD and

others, it is likely that these represent

predisposing genomic alterations lead-

ing to a variable nal phenotype corre-

lated with the genetic background and

the environment.

The last group of genetic alterations

in ASD includes single gene disorders.

Unless there is a recognizable clinical

diagnosis, such as fragile X, Angelman

or Rett syndrome, the likelihood of

identifying a single gene mutation in a

nonsyndromic ASD patient is extremely

low. A list of clinically recognizable sin-

gle gene disorders frequently associated

with ASD is reported in Table I.

With respect to fragile X syndrome,

both FMR1 full mutations and pre-

mutations can be found in children

with ASD: it is estimated that 13% of

ASD childrenmay have alterations in the

FMR1 gene. This is not surprising con-

sidering the overlapping of the neuro-

behavioral phenotypes in ASD and

fragile X syndrome.

Rett syndrome is also frequently

associated with autism: MECP2 muta-

tions are reported in approximately 1%

of children diagnosed with autism. On

the other hand, about 18% of girls end-

ing upwith a diagnosis ofRett syndrome

are initially considered autistic.

Among other single gene disorders,

tuberous sclerosis is frequently associated

with ASD with 25% of patients fullling

the diagnostic criteria for autism [Baker

et al., 1998]; the frequency of autistic

features is higher in younger children, up

to 60% [Jeste et al., 2008], and decreases

as the child gets older.

Mutations in the PTEN gene have

also been detected in 18% (according to

different studies) of children with ASD

TABLE I. Clinically Recognizable Single Gene Disorders in Which Autism Is Frequently Reported

Syndrome Gene locus % ASD among patients

Fragile X FMR1 Up to 30%

PTEN extreme macrocephaly PTEN n.a.

Rett syndrome MECP2 Up to 18%

Tuberous sclerosis TSC1 and TSC2 50%

Timothy syndrome CACNA1C and CACNA1F High

Phenylketonuria PAH 6%

Creatine biosynthesis and transport disorders SLC6A8 Up to 80%

L-arginine:glycine amidinotransferase

Guanidinoacetate methyltransferase

SmithLemliOpitz syndrome 7-Dehydrocholesterol reductase 5080%

Sotos syndrome NSD1 n.a.

Moebius syndrome Unknown 30%

Duchenne muscular dystrophy Dystrophin

PhelanMcDermid syndrome SHANK3 Up to 90%

106 AMERICAN JOURNAL OF MEDICAL GENETICS PART C (SEMINARS IN MEDICAL GENETICS) ARTICLE

and extreme macrocephaly [Buxbaum

et al., 2007; Varga et al., 2009].

Another gene which is likely

to cause a recognizable syndromic au-

tism, is SHANK3. Heterozygous muta-

tions and intragenic deletions have been

reported in about 15% of patients with

ASD, epilepsy, tendency to overgrowth,

hypotonia, and absent language [Her-

bert, 2011]. A complete deletion of

this gene is observed in the Phelan

McDermid syndrome, which is caused

by a microdeletion of several genes on

the 22q13 region.

Among the genetic factors, inborn

errors of metabolism should also be

included:

Mitochondrial disorders can be detected

in 45% of ASD children even with-

out additional neurological issues

[Hass, 2010].

Phenylketonuria, adenylosuccinase lyase

deciency, creatine deciency syn-

dromes usually show behavioral alter-

ations overlapping with autistic

symptoms in addition to severe intel-

lectual disability and seizures [Van den

Berghe et al., 1997; Baieli et al., 2003;

Newmeyer et al., 2007].

SmithLemliOpitz syndrome is proba-

bly the condition most frequently as-

sociated with ASD: variable features

of the spectrum can be detected in up

to 80% of these patients [Sikora et al.,

2006].

In addition, it has been repeatedly

suggested that individuals gifted with

mathematical minds might be

more likely to have a child with

ASD [Baron-Cohen, 2006; Buchen,

2011].

It should also be taken into account

that an increasing number of genes,

whose mutations are associated with

autism, are being annotated. In a recent

review on the genetics of ASD a list

of the most signicant genes involved

in this condition was reported [Miles,

2011]. This list has rapidly grown

as more and more genes have been

identied either by candidate gene

approach [Schaaf et al., 2011] or

through exome sequencing of trios

(proband parents) [ORoak et al.,

2011]; however, none of thesemutations

has led to a clinically distinguishable

phenotype.

ENVIRONMENTALFACTORS

It has been widely observed that there

has been an increase in incidence of ASD

over the past years. However, it is still

debated whether this increase is related

to a diagnostic improvement, raised

awareness towards ASD, or to emerging

environmental factors, not present in the

past, that have inuenced this epidemi-

ologic change [Duchan and Patel, 2012].

If this is the case, it is crucial to recognize

these factors because they are the ones

most amenable to elimination.

Environmental risk factors may be

related to in utero exposure: for instance,

children whose mothers consumed

antiepileptic drugs have a sevenfold in-

creased risk for ASD [Palac and

Meador, 2011]; maternal alcohol con-

sumption is also a risk factor [Eliasen

et al., 2010]. Emphasis has been placed

also on oxytocin levels at delivery:

lower levels seem to reduce the capabili-

ty to socialize and to increase the

risk for communication impairment

[Gurrieri and Neri, 2009; Gregory

et al., 2009].

Assisted reproductive technologies

or short interval between pregnancies

may represent additional risk factors

[Zachor and Ben Itzchak, 2011].

Other environmental modiers in-

clude advanced paternal age (risk in-

creased 2.2 times with paternal age

>50) [Hultman et al., 2011], oxidativestress and environmental pollutants (such

as air pollution, organophosphates, and

heavy metals).

Epilepsy, food intolerance, immune

and hormonal dysfunction, mitochon-

drial and metabolic unbalance (i.e., low

glutathione, low antioxidant and detox-

icant activity) epigenetic modications

[Grafodatskaya et al., 2010] and the

microbiome composition also play a

role in ASD, but it is difcult to establish

whether these issues are primarily

involved in its etiology or rather repre-

sent concomitant medical problems

[Herbert, 2010].

All these factors need to be consid-

ered when collecting anamnestic data in

ASD children.

MOLECULAR DIAGNOSISAND TESTING STRATEGIES

More than half (between 50% and 70%)

of the parents have the perception that

the cause of ASD in their children might

be of genetic nature [Harrington et al.,

2006; Selkirk et al., 2009]. It is crucial for

the clinical geneticist to assess the

parents expectations of genetic testing

and to inform them of the limited utility

of the genetic testing in providing

answers or suggesting treatment plans.

In order to determine appropriate

biochemical and molecular testing, a

clinical genetic evaluation is crucial or

unnecessary effort will be put into the

search of a genetic or organic (metabol-

ic) cause in each ASD patient. Evenwith

an extensive clinical workup, physicians

can expect to identify a genetic cause in

less than 25% of ASD patients.

After clinical genetic evaluation one

can expect three possible scenarios: the

patient has nonsyndromic autism, a spe-

cic genetic syndrome is suspected,

or the patient has morphological alter-

ations on physical exam, but a specic

genetic condition cannot be identied.

After clinical genetic

evaluation one can expect

three possible scenarios: the

patient has nonsyndromic

autism, a specic genetic

syndrome is suspected, or the

patient has morphological

alterations on physical exam,

but a specic genetic condition

cannot be identied.

In the case of essential autism, al-

though a genetic basis is possible, no

specic test for monogenic ASD should

be recommended because the possibility

ARTICLE AMERICAN JOURNAL OF MEDICAL GENETICS PART C (SEMINARS IN MEDICAL GENETICS) 107

of identifying a pathogenic mutation is

negligible. Only array-CGH testing is

worthwhile because there is a 10%

chance of identifying a possibly patho-

genic CNV [Sebat et al., 2007]. Standard

karyotype is the rst step when array-

CGH is not available, even in essential

autism, otherwise the latter should be

the rst choice. The diagnostic yield

of standard cytogenetic testing is about

2% [Roesser, 2011].

It has been shown that the highest

diagnostic yield among the different

laboratory tests is reached by array-

CGH. There have been variable

reports of different authors who have

identied CNVs in a percentage of

ASD cases varying from 10% [Sebat

et al., 2007] to 18% [Shen et al., 2010].

It has been shown that the

highest diagnostic yield among

the different laboratory tests is

reached by array-CGH. There

have been variable reports of

different authors who have

identied CNVs in a

percentage of ASD cases

varying from 10% to 18%.

It should bementioned that the results of

array-CGH are not always easy to inter-

pret: for instance, the sameCNVs can be

detected in an ASD child and his/her

healthy or border-line parent or a de

novo CNV can have a nonsignicant

gene content so that it is difcult to

consider it pathogenic. Also, potentially

detrimental CNVs detected in ASD can

be also be found, although at a lower

frequency, in the normal population.

Interpreting array-CGH can be a very

difcult task that should be left to expe-

rienced medical geneticists.

Not infrequently, array-CGH in

ASD patients detects CNVs commonly

associated with specic microdele-

tion or microduplication syndromes:

among those, the 22q11 deletion (asso-

ciated with DiGeorge velo-cardio-facial

syndrome), the 17p11 deletion (associ-

ated with SmithMagenis syndrome),

the 22q13 deletion (associatedwith Phe-

lanMcDermid syndrome) or even the

MECP2 duplication (for which there is

no specic phenotype). In these cases the

phenotypic expression of the known

syndrome is atypical and therefore not

easily recognizable.

For children with normal results on

array-CGH I would not recommend

further testing but follow-up and even-

tually propose autism-specic-gene se-

quencing, when such diagnostic tools

will become available and affordable.

Fragile X testing is frequently rec-

ommended as a rst step molecular test

in ASD. However, in cohorts not

screened by clinical evaluation the diag-

nostic yield is quite low: less than 0.5%

[Reddy, 2005].

Figure 1 shows a possible diagnostic

itinerary in ASD patients.

COUNSELING

If a genetic cause of clear pathogenic

signicance is identied, the recurrence

risk for sibs is relatively easy to establish

according to the etiologic diagnosis. If

no genetic alteration is found, and this

happens in the majority of patients,

an empirical 1020% risk for sibs

should be given [Ozonoff et al., 2011].

If a genetic cause of clear

pathogenic signicance is

identied, the recurrence risk

for sibs is relatively easy to

establish according to the

etiologic diagnosis. If no

genetic alteration is found,

and this happens in the

majority of patients, an

empirical 1020% risk for

sibs should be given.

This risk might be higher for having a

second child with milder symptoms, in-

cluding language, social, or other psy-

chiatric disorders. If the propositus has

essential autism and if there is already an

affected sib or a positive family history

the recurrence risk increases consistently

(up to 2530%) [Miles et al., 2005]. On

the other hand complex autism of un-

known etiology has a lower recurrence

risk (between 1% and 2%) [Miles, 2011].

Figure 1. Proposed genetic diagnostic itinerary for autism spectrum disorders.

108 AMERICAN JOURNAL OF MEDICAL GENETICS PART C (SEMINARS IN MEDICAL GENETICS) ARTICLE

The medical geneticist is often re-

quired to make predictions about the

clinical outcome in case a genetic alter-

ation is or is not detected: in general

it has been observed that the prognosis

is slightly better in patients without

positive testing and with essential

autism.

FUTURE DIRECTIONS

New light has been shed recently on the

general thinking about autism: the spec-

trum is highly variable to the point

that people with autism may be particu-

larly talented in many professional set-

tings, including scientic laboratory

[Mottron, 2011]. However early diag-

nosis of this disorder is crucial as it allows

for more effective intervention so that

any talent might be more easily involved

in our social world.

It is expected that high throughput

molecular screenings, such as high reso-

lution array-CGH, exome and full ge-

nome sequencing [ORoak et al., 2011],

as well as transcriptomic analysis

[Voineagu et al., 2011] will increase

our understanding of the genetic causes

of ASD.

It will be possible in the near future

to obtain diagnostic tools to screen hun-

dreds of autism-genes in a single shot so

the genetic prole of ASD patients will

be more easily outlined. However, if we

do not correlate these ndings with a

critical evaluation of the different autistic

phenotypes, there is no way that they

will be of any help in making diagnosis,

prognosis and counseling in ASD

families.

REFERENCES

Autism and Developmental Disabilities Monitor-ing Network Surveillance Year 2000Principal Investigators, Centers for DiseaseControl and Prevention. 2007. Prevalenceof autism spectrum disordersAutism anddevelopmental disabilities monitoring net-work, six sites, United States, 2000.MMWR Surveill Summ 56:111.

American Psychiatric Association. 1994. Diagnos-tic and statistical manual of mental disorders,4th edition. Washington, DC: AmericanPsychiatric Association.

Baieli S, Pavone L, Meli C, Fiumara A, ColemanM. 2003. Autism and phenylketonuria.J Autism Dev Disord 33:201204.

Baker P, Piven J, SatoY. 1998. Autism and tuberoussclerosis complex: Prevalence and clinicalfeatures J Aut Dev Disord 28:279285.

Baron-Cohen S. 2006. The hyper-systemizing,assortative mating theory of autism. ProgNeuro-Psychopharmacol Biol Psychiatry30:865872.

Bayley A, Le Couteur A, Gottesman I, Bolton P,Simonoff E, Yuzda E, Rutter M. 1995. Au-tism as a strong genetic disorder: Evidencefrom a British twin study. Psychol Med25:6377.

Bayley A. 2011. Etiological heterogeneity inautism spectrum disorders: More than 100genetic and genomic disorders and stillcounting. Brain Res 1380:4277.

Buchen L. 2011. When geeks meet. Nature479:2527.

Buxbaum JD, Cai G, Chaste P, Nygren G,Goldsmith J, Reichert J, Anckarsater H,Rastam M, Smith CJ, Silverman JM, Hol-lander E, Leboyer M, Gillberg C, Verloes A,Betancur C. 2007.Mutation screening of thePTENgene in patientswith autism spectrumdisorders and macrocephaly. Am J MedGenet Part B 144B:484491.

Chakrabarti S, Fombonne E. 2005. Pervasivedevelopmental disorders in preschoolchildren: Conrmation of high prevalence.Am J Psychiatry 162:11331141.

Creswell C, Skuse D. 1999. Autism in associationwith Turner syndrome: Genetic implicationsfor male vulnerability to pervasive develop-mental disorders. Neurocase 5:101108.

DiCicco-Bloom E, Lord C, Zwaigenbaum L,Courchesne E, Dager SR, Schmitz C,Schultz RT, Crawley J, Young LJ. 2006.The developmental neurobiology of autismspectrum disorder. J Neurosci 26:68976906.

Duchan E, Patel DR. 2012. Epidemiology ofautism spectrum disorders. Pediatr ClinNorth Am 59:2743.

Dykens EM, Sutcliffe JS, Levitt P. 2004. Autismand 15q11-q13 disorders: Behavioral,genetic, and pathophysiological issues.Ment Retard Dev Disabil Res Rev 10:284291.

Eliasen M, Tolstrup JS, Nybo Andersen AM,Grnbaek M, Olsen J, Strandberg-LarsenK. 2010. Prenatal alcohol exposure and autis-tic spectrum disorders. A population-basedprospective study of 80,552 children andtheir mothers. Int J Epidemiol 39:10741081.

Fombonne E. 1999. The epidemiology of autism:A review. Psychol Med 29:769786.

Grafodatskaya D, Chung B, Szatmari P, WeksbergR. 2010. Autism spectrum disorders andepigenetics. J Am Acad Child AdolescPsychiatry 49:794809.

Gregory SG, Connelly JJ, Towers AJ, Johnson J,Biscocho D, Markunas CA, Lintas C,Abramson RK, Wright HH, Ellis P,Langford CF, Worley G, Delong GR,Murphy SK, Cuccaro ML, Persico A,Pericak-Vance MA. 2009. Genomic andepigenetic evidence for oxytocin receptordeciency in autism. BMC Med 7:62.

Gurrieri F,Neri G. 2009.Defective oxytocin func-tion: A clue to understanding the cause ofautism? BMC Med 7:62.

Harrington JW, Patrick PA, Brand DA. 2006.Parental beliefs about autism: Implications

for the treating physician. Autism 10:452462.

HassRH. 2010.Autism andmitochondrial disease.Dev Disabil Res Rew 16:144153.

Herbert MR. 2010. Contributions of the environ-ment and environmentally vulnerable physi-ology to autism spectrum disorders. CurrOpin Neurol 23:103110.

Herbert MR. 2011. SHANK3, the synapse andautism. N Engl J Med 365:173175.

Howlin P, Goode S, Hutton J, Rutter M. 2004.Adult outcome for children with autism.J Child Psychol Psychiatry 45:212229.

HultmanCM, Sandin S, Levine SZ, Lichtenstein P,Reichenberg A. 2011. Advancing paternalage and risk of autism: New evidence from apopulation-based study and a meta-analysisof epidemiological studies. Mol Psychiatry16:12031212.

Jeste SS, Sahin M, Bolton P, Ploubidis GB,Humphrey A. 2008. Characterization ofautism in young children with tuberous scle-rosis complex. J Child Neurol 23:520525.

Johnson CP, Myers SM, American Academy ofPediatrics Council on Children With Dis-abilities. 2007. Identication and evaluationof children with autism spectrum disorders.Pediatrics 120:11831215.

Kent L, Evans J, Sharp M. 1999. Comorbidity ofautistic spectrum disorders in children withDown syndrome. Dev Med Child Neurol41:153158.

Miles JH. 2011. Autism spectrum disordersAgenetics review. Genet Med 13:278294.

Miles JH, Takahashi TN, Haber A, Hadden L.2003. Autism families with a high incidenceof alcoholism. J Autism Dev Disord 33:403415.

Miles JH, Takahashi TN, Bagby S, Sahota PK,Vaslow DF, Wang CH, Hillman RE, FarmerJE. 2005. Essential versus complex autism:Denition of fundamental prognostic sub-types. Am J Med Genet Part A 135A:171180.

Mottron L. 2011. Changing perceptions: Thepower of autism. Nature 479:3335.

Newmeyer A, de Grauw T, Clark J, Chuck G,Salomons G. 2007. Screening of malepatients with autism spectrum disorder forcreatine transporter deciency. Neuropedi-atrics 38:310312.

Olivie H. 2012. Clinical practice: The medicalcare of children with autism. Eur J PediatrDOI: 10.1007/s00431-011-1669-1.

ORoak BJ, Deriziotis P, Lee C, Vives L, SchwartzJJ, Girirajan S, Karakoc E,Mackenzie AP,NgSB, Baker C, Rieder MJ, Nickerson DA,Bernier R, Fisher SE, Shendure J, EichlerEE. 2011. Exome sequencing in sporadicautism spectrum disorders identies severede novo mutations. Nat Genet 43:585589.

Ozonoff S, Young GS, Carter A, Messinger D,Yirmiya N, Zwaigenbaum L, Bryson S,Carver LJ, Constantino JN,DobkinsK,Hut-man T, Iverson JM, Landa R, Rogers SJ,Sigman M, Stone WL. 2011. Recurrencerisk for autism spectrum disorders: A babysiblings research consortium study. Pediatrics128:488495.

Palac S, Meador KJ. 2011. Antiepileptic drugs andneurodevelopment: An update. Curr NeurolNeurosci Rep 11:423427. Review.

ARTICLE AMERICAN JOURNAL OF MEDICAL GENETICS PART C (SEMINARS IN MEDICAL GENETICS) 109

Reddy KS. 2005. Cytogenetic abnormalities andfragile-X syndrome in Autism SpectrumDisorder. BMC Med Genet 6:3.

Roesser J. 2011. Diagnostic yield of genetic testingin children diagnosed with autism spectrumdisorders at a regional referral center. ClinPediatr 50:834843.

Schaaf CP, Sabo A, Sakai Y, Crosby J, Muzny D,Hawes A, Lewis L, Akbar H, Varghese R,Boerwinkle E, Gibbs RA, Zoghbi HY.2011. Oligogenic heterozygosity in individ-uals with high-functioning autism spectrumdisorders. Hum Mol Genet 20:33663375.

Schaefer GB, Mendelsohn NJ. 2008. Clinical ge-netics evaluation in identifying the etiologyof autism spectrum disorders. Genet Med10:301305.

Sebat J, Lakshmi B, Malhotra D, Troge J, Lese-Martin C, Walsh T, Yamrom B, Yoon S,Krasnitz A, Kendall J, Leotta A, Pai D, ZhangR, Lee YH, Hicks J, Spence SJ, Lee AT,Puura K, Lehtimaki T, Ledbetter D, Gre-gersen PK, Bregman J, Sutcliffe JS, Jobanpu-tra V, Chung W, Warburton D, King MC,Skuse D, Geschwind DH, GilliamTC, Ye K,Wigler M. 2007. Strong association of denovo copy number mutations with autism.Science 316:445449.

Selkirk CG, McCarthy Veach P, Lian F, Schim-menti L, LeRoy BS. 2009. Parents percep-tions of autism spectrum disorder etiologyand recurrence risk and effects of their per-ceptions on family planning: Recommenda-tions for genetic counselors. J Genet Couns18:507519.

Shen Y, Dies KA, Holm IA, Bridgemohan C,Sobeih MM, Caronna EB, Miller KJ, FrazierJA, Silverstein I, Picker J, Weissman L,Raffalli P, Jeste S, Demmer LA, Peters HK,Brewster SJ, Kowalczyk SJ, Rosen-SheidleyB, McGowan C, Duda AW III, Lincoln SA,Lowe KR, Schonwald A, Robbins M,Hisama F, Wolff R, Becker R, Nasir R,Urion DK, Milunsky JM, Rappaport L,Gusella JF, Walsh CA, Wu BL, Miller DT,Autism Consortium Clinical Genetics/DNA Diagnostics Collaboration. 2010.Clinical genetic testing for patients with au-tism spectrum disorders. Pediatrics 125:727735.

Sikora DM, Pettit-Kekel K, Peneld J, MerkensLS, Steiner RD. 2006. The near universalpresence of autism spectrum disorders inchildrenwith Smith-Lemli-Opitz syndrome.Am J Med Genet Part A 140A:15111518.

Tuchman R, Rapin L. 2002. Epilepsy in autism.Lancet Neurol 1:352358.

Van den Berghe G, Vincent MF, Jaeken J. 1997.Inborn errors of the purine nucleotide cycle:Adenylosuccinase deciency. J InheritMetab Dis 20:193202.

Varga EA, Pastore M, Prior T, Herman GE,McBride KL. 2009. The prevalence ofPTEN mutations in a clinical pediatriccohort with autism spectrum disorders, de-velopmental delay, and macrocephaly. GenetMed 11:111117.

Voineagu I,Wang X, Johnston P, Lowe JK, Tian Y,Horvath S,Mill J, CantorRM,Blencowe BJ,Geschwind DH. 2011. Transcriptomic anal-ysis of autistic brain reveals convergent mo-lecular pathology. Nature 474:380384.

Weiss LA, Shen Y, Korn JM, Arking DE, MillerDT, FossdalR, Saemundsen E, StefanssonH,Ferreira MA, Green T, Platt OS, RuderferDM,Walsh CA, Altshuler D, Chakravarti A,Tanzi RE, Stefansson K, Santangelo SL,Gusella JF, Sklar P, Wu BL, Daly MJ, AutismConsortium. 2008. Association betweenmicrodeletion and microduplication at16p11.2 and autism. N Engl J Med 358:667675.

Zachor DA, Ben Itzchak E. 2011. Assisted repro-ductive technology and risk for autism spec-trum disorder. Res Dev Disabil 32:29502956.

110 AMERICAN JOURNAL OF MEDICAL GENETICS PART C (SEMINARS IN MEDICAL GENETICS) ARTICLE