The future of newborn screening belongs to obstetricians

MI Evans1 and HL Levy2

Department of Obstetrics and Gynecology1, Hutzel Hospital, Wayne State University, Detroit, MI, USA;Children’s Hospital2, Harvard Medical School, Boston, MA, USA

Evans MI, Levy HL. The future of newborn screening belongs to obstetricians. Acta Pædiatr 1999;88 Suppl 432: 55–7. Stockholm. ISSN 0803–5326

Newborn screening has been a mainstay of genetic testing for several decades. Molecularbiological techniques, pervasive in genetics and in medicine, are shifting the focus of such testingfrom enzymology to molecular diagnosis. Concomitant developments in the isolation of fetal cellsfrom maternal blood enable molecular analysis to be performed not just postnatally, but in fact asearly as the late first trimester. Potential advantages of first trimester prenatal diagnosis of thosedisorders currently performed postnatally include more reproductive options for couples, and, inselected cases, the fact that certain genetic disorders can be treated much earlier in development,thereby potentially ameliorating the sequelae of these disorders. Thus, the future of newbornscreening will belong as much to the obstetrician as to the pediatrician, if not more so.& Geneticscreening, prenatal care, prenatal diagnosis

MI Evans, Department of Obstetrics and Gynecology, Hutzel Hospital, Wayne State University,4707 St. Antoine, Detroit MI 48201, USA

While neonatal and newborn screening has a richtradition extending back for decades, obstetricians/gynecologists likewise have a long history of perform-ing screening tests on women—pregnant, or not. ThePapanicolaou smear dates back nearly half a century,and the use of maternal serum alpha-fetoprotein (AFP)for the detection of neural tube defects is now more than25 years old (1). Use of AFP, and, later, multiple markerscreening for Down syndrome and other aneuploidieshas been going on for nearly 15 years. The majordifference between obstetrical screening for geneticdisorders and neonatal screening is that in mostjurisdictions, neonatal screening is not only sanctioned,but also mandated by law. Obstetrical screening isvoluntary; it requires informed consent discussions; andits utilization varies markedly among patient popula-tions. In the United States, of approximately fourmillion births per year, about two million women availthemselves of prenatal genetic screening, with markedvariances due to ethnicity, religion and geography.Other aspects of obstetrical screening include maternalserologic and infectious status. This may be “routine”,but can have significant implications for the fetus andnewborn.

Maternal screening issues, protocols, and directionshave been extensively explored in what is now anenormous body of literature, and there is not space torepeat here. There are, however, multiple chapters,articles, etc., available to the interested reader.

The main thesis of this paper concerns the directionof current newborn screening protocols. We look at itfrom the perspective of fetal therapists, believing that

much more could be done for the fetus if geneticscreening and diagnostics could be accomplished earlyin pregnancy rather than after birth. Postnatal testscorrespond to the American saying of “locking the barndoor after the horse has been stolen”.

Diagnosis in the first trimester would allow forcouples discovered to have a fetus with a seriousgenetic disorder to consider whether or not they wish tocontinue the pregnancy. In some selected cases, itwould be therapeutically possible to treat these dis-orders, either to ameliorate or even completely curecertain genetic diseases. Pharmacological therapy hasproven effective in the treatment of congenital adrenalhyperplasia, and the prevention of masculinization ofexternal female genitalia (2). Hematopoietic stem celltherapy has proven effective in the treatment of severecombined immunodeficiency disorder, and has beenattempted and will hopefully be successful in thetreatment of many inborn errors of metabolism,hematologic, and other immunodeficiencies (3).

In order to make neonatal screening disorderspossible in the first trimester of pregnancy, tests mustbe done on a molecular basis, as opposed to abiochemical/enzymatic one. In addition, fetal cell mustbe isolated from the maternal circulation to make fetalDNA readily available to the geneticist and laboratory.

Newborn screening tests become molecularTraditional newborn screening targets metabolites orenzymes. For instance, screening for phenylketonuria

Scandinavian University Press 1999. ISSN 0803-5326

Acta Pñdiatr Suppl 432: 55±7. 1999

(PKU) measuresthe concentrationof phenylalanineinnewborn blood specimens.Screeningfor congenitalhypothyroidismmeasuresthyroid stimulatinghormoneor thyroxine. Screening for galactosemiameasuresactivity of galactose-1-phosphateuridyltransferase(GALT) or galactose.All theseand other targetsofscreeningaredirector indirectproductsof genes.Thus,theyarenot theprecisesourcesof thedisorder.

In 1973, Harvey Levy wrote, “Since the directanalysis of genes, however desirable, is still notpossible, genetic screeningusually refers to testingrelatively large numbers of individuals for geneproducts or resulting metabolites with the aim ofidentifying disordersdueto mutantgenes”(4). Today,direct analysis of genes in newborn screening ispossible. In fact, it has been possible for over adecade,ever since Ed McCabe and his colleaguesshowedthatDNA couldbeextractedfrom thenewborndried blood specimen(5). Molecular screeningof thenewborncan be conductedfor sickle cell disease,themusculardystrophias,andmetabolicdisorderssuchasmedium-chain acyl-CoA dehydrogenasedeficiency(MCADD). Secondarymolecular screeningcould beincluded in newborn screening for virtually everygenetic disorder screened, including galactosemia,PKU, maple syrup urine disease, homocystinuria,biotinidasedeficiency,congenitaladrenalhyperplasia,cystic fibrosis, and a host of additional disordersforwhich expandedscreeningby tandemmassspectro-metry is conducted.In fact, almostall newbornscreen-ing for cystic fibrosis doesindeedinclude second-tiermoleculartestingfor the mostfrequentcystic fibrosis-associatedmutations, and some programs employsecondary molecular screening for the two mostfrequent galactosemia-associated GALT mutations,Q188R and N314D (6), and for confirmation ofmedium-chain acyl-CoA dehydrogenasedeficiency,long-chainhydroxyacyl-CoAdehydrogenasedeficiencyandsickle cell hemoglobinopathies.

Nevertheless,molecularscreeninghasnot replacedproduct or metabolitenewborn screening.Why not?The potentialadvantagesseemto be clear. Molecularscreeningoffers thepromiseof 100%detection(i.e. nofalse negatives),100% reliability (i.e. no false posi-tives),theidentificationof carriers,andtotalspecificity.Nevertheless,regardlessof theseassumedadvantages,thereareseveralmajor obstacles.First, for all geneticdisordersexceptsickle cell disease,therearemultiplemutations.This meansthatdetectionof all infantswithadisorderwouldrequiretestingnot for one,or eventwomutations, but for many mutations. In PKU, forinstance,over 300 mutationshavebeenidentified. Ingalactosemia,almost100 mutationshavebeenfound.Even in MCADD, in which the G985A mutationaccountsfor over 90% of affected alleles in somepopulations,thereareseveralothermutations.Testingfor hundredsor thousandsof mutations would beenormouslylaboriousand expensive;testing for only

thetwo or threepredominantmutationswould resultinundetectedinfants.

Secondly, molecular testing would result in theidentificationof hugenumbersof carriers,potentiallya seriousdisadvantagerather than an advantage.Forinstance,theheterozygoticratein PKU is 2%.Thus,theNew England Newborn Screening Program, whichscreensover500infantsadaywould,eachday,identifyat least10whocarryagenefor PKU.Addingto this thenumber of identified infants who carry a gene forgalactosemia,maplesyrupurine disease,andso forth,would result in perhapscloseto 100 infants eachdaylabeled as carriers. These infants and their familieswould requireconfirmatoryandfamily testing,aswellasgeneticcounseling.

An evengreaterconcernis that the sheermagnitudeof suchanenterprisewouldengenderenormousanxietyfor thesefamiliesfor afindingthat,for most,at leastforthe proband,is of no consequence.The alternative,disregardingcarriers,is ethicallyuntenable.In someofthesecarrierinfantsbothparentscarrythemutantgene.Approximately25% of thesecoupleswill produceanaffected infant in subsequentpregnancies.When thisoccurs,the lack of prior notification and concomitantgeneticcounselingwill havedeprivedthesefamiliesofthe family planning information that they otherwisemight havereceived.

Finally, thespecificityof molecularscreeningis alsoa disadvantage,rather than an advantage.With thelimited specificityof metabolitescreening,a singlenetis cast wide with severaldisordersidentified, to besortedout by confirmatorytesting.With the absolutespecificity of molecular testing, screeningrequiresasingle net for eachdisorder,the very antithesisof theprinciple of screening.

Doesthis meanthat molecular(i.e. gene)screeninghas no role in newbornscreening?On the contrary,secondarymoleculartestingis very valuablein screen-ing for cystic fibrosis and galactosemia.It could alsoimproveotherscreening(e.g.PKU, maplesyrupurinedisease,etc.)by reducingfalsepositiveresults.It is notinconceivablethat there might be a role for primarymolecularnewbornscreeningfor a limited numberofdisorders.However,beforeuseof moleculartechnologyin newborn screeningis considered,there must becareful considerationnot only of the logistics of itsapplication,but alsoof its implications.

Application of molecularscreeninginpregnancyInvasiveprenataldiagnosisusingmoleculartechniqueshasdevelopedrapidly in thepastdecade,usingamnioticfluid cells,chorionicvilli, or fetal blood.Obviously,allof these proceduresare invasive, and can only beperformedon a small percentageof the population.Recentadvancesin isolatingfetalcellsthatleakthrough

56 MI EvansandHL Levy ACTA PÆDIATRSUPPL432(1999)

into the maternalcirculationsuggestthat this will be ascreeningtechniqueof very high specificity for aneu-ploidy. Although data are still very preliminary, itappearsas if, for Mendeliandisorders,the technologymay prove diagnostic (7). With the developmentofarray technologythat will allow literally hundredsofthousandsof reactionsto be performedon the samespecimen,the concernsraisedin the previoussectionshould be overcome.Array technologywill not onlyradically changeinvasiveprenataldiagnosis,but, oncethe logistics, costs, and necessity for appropriatecounselinghave beenworked out, should also allowfor thebenefitsof thesemoleculardiagnosesto bemadeavailableto theobstetricalpopulationat large.

ConclusionOverthecourseof thenextdecade,it is likely thatmanyof the testscurrentlyperformedin the newbornperiodwill be accomplishedin the early or mid-gestationalperiod.Therewill be no reasonto wait becauseof theadvantageousimplicationsfor reproductivechoiceandfetal therapy.Theshift from postnatalto obstetricalcarewill clearly be fraughtwith technicalchallenges,but ifboth physiciansandpatientshavethe correctmindset,the transitionwill besuccessful.

References1. EvansMI, O’Brien JE,Dvorin E, etal. Biochemicalscreening.In:

GleisherN, ButtinoJrL, ElkayamU, etal.,editors.Principlesandpracticesof medicaltherapyin pregnancy.3rd ed.Norwalk, CN:AppletonandLange,1998:199–207

2. YaronY, EvansMI. Fetalmetabolicandgenetherapy.In: ReeceEA, HobbinsJ,editors.Medicineof thefetusandmother.2nded.Philadelphia:Lippincott/Raven,1999:873–90

3. FlakeAW, RoncaroloMG, PuckJM, Almeida-PoradaG, EvansMI, JohnsonMP, et al. Successfulin uterocorrectionof x-linkedrecessiveseverecombinedimmuno-deficiency(X-SCID): fetalintraperitonealtransplantationof CD34 enrichedpaternalbonemarrowcells (EPPBMC).N Engl J Med 1996;335:1806–10

4. Levy HL. Geneticscreening.In: HarrisH, HirschhornK, editors.Advancesin humangenetics.Vol. 1. New York: PlenumPress,1973:1–104

5. McCabe ERB, Huang S-Z, Seltzer WK, Law ML. DNAmicroextractionfrom dried blood spotson filter paperblotters:potentialapplicationsto newbornscreening.HumGenet1987;75:213–6

6. Larson C, GerstelJ, ComeauA-M. Classificationof suspectedgalactosemicsby presenceof Duarte (N314D) and classicgalactosemia(Q188R) mutations. Book of abstracts.Fourthmeeting of the International Society for Neonatal Screening,Stockholm,13–16June1999:106

7. BianchiDW. Fetalcells in thematernalcirculation:feasibility forprenataldiagnosis.Br J Haematol1999;105:1–10

ACTA PÆDIATRSUPPL432(1999) Futureof newbornscreening 57