Complex Autism Spectrum Disorders and Cutting-Edge Molecular Diagnostic Tests.

Abstract

Autismspectrumdisorders (ASDs)areincompletelyunderstood neurodevelopmental disorders diagnosed solely on the basis of behavioral assessments of social, communicative, and repetitivesymptoms.1Although ASDisbehaviorallydistinctive andreliablyidentifiedbyexperienced clinicians, the disorder is clinically and genetically extremelyheterogeneous.Psychologists,whobegantodefineautismsubgroups inthe1990s, foundneitherbehavioralmeasures of core ASD symptoms nor cognitivemeasures reliably identified subgroups with similar outcomes or risk of recurrence in siblings.2Atthesametime,geneticistswereattemptingwithnegligible success to find “autism genes” usingmolecular linkage andassociationanalysisthathadsuccessfullyidentifiedthegenes for cystic fibrosis andHuntingtondisease.3Aboveall, their failures were attributed to an inability to assemble homogeneous ASD cohorts for analysis. This spurred the search for biomarkers, sometimescalledendophenotypes, thatmight sortout the etiologic heterogeneity associated with ASD.4 The goal was to identify features thatoccur consistently inaportionofpatients withASDandare relativelydiscrete,quantifiable, andmost importantly etiologically relevant. Physical dysmorphology appeared to fit the criteria. The premise, which grew out of the pioneering work of Smith in the 1960s,5 is that individuals for whom there is evidence of an insult to early morphogenesis would have developedtheirautismonadifferentbasis than individuals forwhom morphogenesis proceeded normally. Minor physical anomalies (MPAs)were recognized tobedevelopmental endpointsof geneticmutationsthathaddisruptedearlymorphogenesis.This work laid the foundation formedical geneticists todefinehundreds of specific syndromes caused by genetic mutations. Subsequent studies revealed that childrenwith ASDwith multipleminor anomalies constituted an important subset of individuals with a cohesive phenotype defined by lower IQ scores, abnormal brain magnetic resonance images (MRIs), poorer functional outcomes, a lower male-female ratio, and lower rate of sibling recurrence.6Miles and colleagues6 documenteddysmorphic features in260consecutive childrenwith ASD and analyzed etiologic differences between those with substantial dysmorphology (n = 46) and those with no evidence of disrupted morphogenesis (n = 187). Because ASD is usually not associatedwith physical anomalies, thiswas considered “essential” autism,whereas substantial dysmorphologydefined the“complex” subgroup.Only childrenwith complex ASD were found to have an abnormal karyotype (2.3%) or a clinically recognized syndrome (1.9%). ThedistinctionbetweencomplexandessentialASDwasconsidered tobe the first attempt at dissecting the etiologic heterogeneity within ASD. In their report in this issue of JAMA, Tammimies and colleagues7 used cutting-edge genome-widemolecular testing (chromosomalmicroarrayanalysis [CMA]andwhole-exomesequencing [WES]) to study 258 consecutively ascertained (from 2008to2013)unrelatedchildren(meanageatdiagnosis,4.5years) whowerediagnosedwithASDat2developmentalpediatricclinics in Newfoundland and Labrador. Chromosomal microarray analysiswasperformedforallprobandsand95proband-parent trioswere testedbyWEStodetermine theirmoleculardiagnosticyieldsinASD.WithCMAplusWEStesting,15.8%(15/95)ofchildrenwere identifiedwithageneticmutation.Chromosomalmicroarray analysis alone revealed amolecular diagnosis in 9.3% (24/258)andWESin8.45%(8/95).This isasubstantialenhancementover the2.3%of similarlyascertainedchildren (from1995 to 2001) diagnosed by karyotype alone. In addition, eachchildunderwent adetailedevaluationby an experienced medical geneticist/dysmorphologist that includedstandardizeddysmorphologyprotocol toenumerate the numberofMPAsandmajor congenital anomalies.Usingapublished schema,6 the geneticist categorized each child, based on his or her level of dysmorphology, as either essential ASD, complexASD,orequivocal.Dysmorphologyscores simplysum the number ofMPAs in the child that are not found in the parents, plus anymajor congenital anomalies.When a brainMRI wasavailable, thescorewasadjusted forabrainanomalyscore. Theprobandswere stratified into 3 groups ofmorphologic severity based on the number of MPAs: essential autism (0-3 MPAs), complex autism (≥6 MPAs), or equivocal (4-5 MPAs). Using this subgroupingschema, theauthors foundchildren with complex autism (basedon clinical evidenceof an insult to earlymorphogenesis)weremore likelytoreceiveamoleculardiagnosisthanthosewithessentialASD.ChildrenidentifiedashavingessentialASDreceivedamoleculardiagnosis4.2%(7/168)of thetimefromCMAand3.1%(2/64)fromWES,foracombinedrate of6.3%.ThosewithcomplexASDwerediagnosedwithagenetic mutationmuchmore frequently: 24.5% (13/53) by CMA, 16.7% (4/24) with WES, for combined rate of 37.5%. As might be expected, those with less evidence of dysmorphology (scored as equivocal)receivedamoleculardiagnosis28.6%(2/7)ofthetime. ThestudybyTammimiesandcolleagues7 illustrates the remarkablestrides inmoleculardiagnosisof the lastdecade.ChromosomalmicroarrayanalysisandWEStestinghavequadrupled themolecular diagnostic rate of karyotype alone, and thismay beanunderestimate.Twentypercentof childrenconsecutively diagnosedwithASDbetween1995and2001exhibitedsignificant Related article page 895 Opinion