J Clin Pathol 1994;47:199-204

Guidelines for the fetal diagnosis of globin genedisorders

Globin Gene Disorder Working Party of the BCSH General Haematology Task Force

Globin Gene DisorderWorking Party of theBCSH GeneralHaematology TaskForceWorking PartyMembers:M BhavnaniM Brozovic (chair)JM OldA D StephensTask Force Members:B J BainM Brozovic (chair)D W DawsonJM EnglandRM RowanA D StephensS M LewisN K ShintonM F MurphyJ K WoodCorrespondence to:Dr JM Old, Institute ofMolecular Medicine, JohnRadcliffe Hospital, Oxford,OX3 9DU.Accepted for publication7 September 1993

Phenotypes of thalassaemias,

Type

Homozygous statea' thalassaemiathalassaemia:f,' or severe + mutationmild fl + mutation

°fl' thalassaemiaHb LeporeHb SHb CHb DHb E

Compound heterozygous statefl'/severe fl +mild f, + /fl0 or severe +

5fll/fl' or severe fl +6,80mild ,t +

5fl'/Hb LeporeHb Lepore/fl' or severe fl +

Hb C/fl, or severe fi +Hb D/fl' or severe fl+Hb E/fl' or severe ,B+O Arab/fl'Hb S/fl, or severe fi +Hb S/mild fi +Hb S/Hb CHb S/Hb D PunjabHb S/Hb 0 Arab

IntroductionThe inherited disorders of haemoglobin com-prise a vast group that include the thalas-saemias and sickle cell disease and caninvolve complex interactions between severaldifferent mutant genes. Most of these muta-tions can now be detected directly by DNAanalysis which allows fetal diagnosis to becarried out for couples at risk of an affectedfetus without the necessity for linkage studieson any children or relatives. Each mutationis detected by its own specific probes orprimers. It is therefore vital to determineaccurately the parental genotypes, preferablybefore fetal diagnosis to avoid mistakes ifmutations are missed because of an incorrectdiagnosis of the carrier state. It also meansthat a large number of probes and primers arerequired, especially for the identification offi thalassaemia mutations, and it is recom-mended that fetal diagnosis by DNA analysisis only undertaken in referral centres.The aim of this paper is to: (a) summarise

the types of haemoglobin disorders for whichfetal diagnosis may be indicated; (b) reviewthe current procedures which are used forfetal diagnosis; (c) provide guidelines forclinicians requesting a fetal diagnosis test;and (d) provide guidelines for clinicians andscientists carrying out a fetal diagnosis test.

Indications for prenatal diagnosisThe important haemoglobin disorders inUnited Kingdom populations at risk are

thethe

sickle cell disease, and various thalassaemia interactions

Phenotype

Hb Bart's hydrops fetalis

Thalassaemia majorThalassaemia intermediaThalassaemia intermediaVariable: intermedia to majorSickle cell diseaseNo clinical problemsNo clinical problemsNo clinical problems

Thalassaemia majorVariable: intermedia to majorVariable: intermedia to majorMild thalassaemia intermediaThalassaemia intermediaThalassaemia majorfi thalassaemia trait -+ intermediaVariable: usually similar to f, thalassaemia traitVariable: intermedia to majorSevere thalassaemia intermediaSickle cell diseaseUsually mild sickle cell diseaseSickle cell disease, variable severitySickle cell diseaseSickle cell disease

thalassaemias, sickle cell disease, and theirinteractions. These are summarised in thetable. All the clinical details, etc, of the tha-lassaemias and their interactions can befound in the authorative book by Weatheralland Clegg.'

ao THALASSAEMIAThe a thalassaemias are a group of disorderscharacterised by a reduction in a globin syn-thesis. They can be divided into the severetypes (a° thalassaemias), which have two aglobin genes deleted per chromosome, andthe mild types (a+ thalassaemias), which haveone a globin gene per chromosome eitherdeleted or non-functional. Alpha thalassaemiais most frequently due to deletions, althougha+ thalassaemia can also be caused by pointmutations.2 The homozygous state for a0thalassaemia results in haemoglobin Bart'shydrops fetalis syndrome and is lethal, lead-ing to stillbirth or early neonatal death.Prenatal diagnosis for the homozygous state isnormally required as the mother of such aninfant may have toxaemia of pregnancy,obstructed labour, and ante- or post-partumhaemorrhage, as well as the psychologicalburden of carrying a non-viable fetus to term.

Alpha0 thalassaemia is particularly com-mon in South East Asia (China, Thailand,Vietnam, Malaysia, the Phillipines) and alsooccurs in the eastern Mediterranean countries(Greece, Cyprus). It is extremely rare inAfrica, the Caribbean, and India, where a+thalassaemia predominates. The compoundheterozygous condition of a°/a+ thalassaemiaresults in HbH disease in which three out ofthe four a globin genes are deleted or non-functional. Many patients with HbH disease,however, lead a relatively normal life and thusfetal diagnosis is not normally requested. Inthe United Kingdom there have been 28prenatal diagnoses carried out for homozy-gous a° thalassaemia up to 1992. These werefor couples mainly of Chinese, Vietnamese,and Cypriot extraction.3

f THALASSAEMIABeta thalassaemia is now known to be causedby more than 100 different mutations,4although in each at-risk population, only alimited number of molecular defects are usu-ally prevalent.5 The mutations differ greatly intheir phenotypic effect. At one end of thespectrum are a group of rare mutationsinvolving exon 3 of the I) globin gene. Theseare so severe that they behave in a dominantmanner and may produce thalassaemia inter-media in the heterozygous state.6 At the otherend are mild alleles which produce the

199

group.bmj.com on February 15, 2012 - Published by jcp.bmj.comDownloaded from

Globin Gene Disorder Working Party

phenotype of ft thalassaemia intermedia intheir homozygous state and, in just a fewcases, are so mild in the heterozygous statethat they are phenotypically silent with nor-mal mean corpuscular volume concentrationsand Hb A,.' However, most ft thalassaemiamutations lie inbetween these two extremes.These are the f,+ and f, 0 alleles which cause ftthalassaemia major either in the homozygousstate for one allele or in the compound het-erozygous state with two different alleles. Bydefining the mutations it is now possible, inmany cases, to predict whether a fetus is atrisk for transfusion dependent thalassaemiamajor or the milder non-transfusion depen-dent thalassaemia intermedia.The phenotype of thalassaemia intermedia

is itself heterogeneous with both mild andsevere forms, because there are many differ-ent genetic determinants responsible for pro-ducing this phenotype.8 These include: thehomozygous state for a mild f,+ allele; thecombination of a mild f,+ allele with a severeft+ or f,t in some cases (unfortunately the sit-uation is not very clear as this combinationsometimes results in ft thalassaemia major);two severe ft+ or ft0 alleles in combinationwith either a thalassaemia or a raised HbFdeterminant, such as the -158 Gy globin pro-moter mutation; one severe ft+ or ,B 0 allele incombination with a triplicated a gene allele;homozygous cfB 0 thalassaemia; cfl 0 thalas-saemia in combination with f,+ thalassaemiaor Hb Lepore; and a ft or severe ft+ allele incombination with Hb Lepore or Hb 0 Arab.

HB S AND ITS INTERACTIONSThe sickle cell gene is caused by an A -. Tsubstitution in codon 6 of the ft globin gene. ftglobin haplotype analysis has shown that thesickle cell gene has arisen at least three timesin Africa and once in Asia. In patients withsickle cell anaemia bearing homozygousforms of the four most common African hap-lotypes (the Senegal, Benin, Bantu andCameroon) the phenotypic expression of thisdisease varies considerably and it is notpossible to predict how severe the disease willbe in any particular family because there areseveral different factors (genetic, cellular, andphysiological) which can modify the course ofsickle cell disease. Recent evidence suggests,however, that the Bantu (also known asCentral African Republic, CAR) is oftenassociated with a clinically severe condition,9while homozygotes, for the Asian haplotypefrom Saudi Arabia and India have highconcentrations of Hb F coupled with only amodest level of anaemia and a relatively mildcourse of the disease, although bone patho-logy (painful crises, avascular necrosis of thefemoral head, and osteomyelitis) remainscommon.10 The latter effects seem to be mostsevere in patients with homozygous a+ tha-lassaemia.The sickle cell gene also interacts with ft

thalassaemia genes and several structuralhaemoglobins to produce sickling disorders ofvariable severity. The clinical course of sicklecell ft thalassaemia is very variable, depending

on the particular type of ft thalassaemia muta-tion that is co-inherited. It tends to be partic-ularly mild in African populations because ofthe likelihood of the co-inheritance of a mildft+ transcription mutation, although even hereoccasional severe interactions are encoun-tered. A severe course, similar to that foundin sickle cell anaemia, results from the inter-action of the sickle gene with either a ft ° or asevere ft+ thalassaemia allele. Haemoglobinsickle cell disease can be mild, though it maybe associated with ocular, central nervoussystem, and bone complications, and throm-botic problems in pregnancy. However, thecombination of Hb S/D Punjab, Hb S/0Arab, and Hb S/Lepore result in diseasessimilar in severity to homozygous sickle cellanaemia.

HB E/If THALASSAEMIAThis disorder is restricted mainly to people ofOriental, Indian, or Pakistani background. Itshows remarkable clinical variability, rangingfrom a symptomless disorder to a conditionas severe as homozygous ft thalassaemiawhich occurs when Hb E interacts with a f 0

thalassaemia allele. It seems likely thatinteraction with a severe ft+ thalassaemia typeof allele results in a slightly milder conditionbut is still a serious disorder in contrast tothe more rare interaction of Hb E with amild ft+ thalassaemia allele which results inan extremely mild, almost symptomless,disorder.

Globin chain synthesisThis technique may be useful in fetal diagno-sis when or if the putative father is not avail-able for tests. Guidelines to the analysis ofglobin chain biosynthesis in adult blood sam-ples for carrier status determination may befound in the review by Clegg (1983)"; thoserelating to the determination of globin chainsynthesis in fetal blood samples for fetal diag-nosis can be found in the review by Alter(1989).12

DNA analysisFetal DNA for prenatal diagnostic studies canbe isolated from aminiotic fluid cells orchroionic villus biopsy specimens.'3 SufficientDNA (10-70,ug) is usually obtained from achorionic villus sample for both polymerasechain reactions (PCR) and Southern blottechniques. However, 20 ml of amniotic fluidwill yield only about 7 ,ug DNA, and some-times much less than this, or occasionallyeven none at all. Thus it is an unreliablesource of DNA, and although usually suffi-cient DNA is obtained for PCR, the amountwill be borderline for Southern blot analysis;therefore the establishment of a backup cellculture is recommended for amniocyte DNAdiagnosis.

SOUTHERN BLOT ANALYSISThis technique is used for identifying a and ftglobin gene rearrangements caused by the

200

group.bmj.com on February 15, 2012 - Published by jcp.bmj.comDownloaded from

Guidelines for the fetal diagnosis ofglobin gene disorders

various a+ and a° thalassaemia deletion genes,Bfl0 thalassaemia deletions, and the HPFHdeletions. Prenatal diagnosis for a0 thalas-saemia is achieved by hybridising Southernblots of BamH I and BgI II digested DNA to32P-labelled a gene and 4 gene probes. Thehomozygous state for a° thalassaemia is diag-nosed by the presence of fragments contain-ing abnormal 4 gene in combination with theabsence of those containing normal a gene.14The procedure usually takes 10-14 days toachieve a result. Southern blot analysis givesan overall picture of the a globin genearrangement and is still very useful for carrierdiagnosis of a thalassaemia and for the con-firmation of prenatal diagnosis resultsobtained by PCR.

PCR TECHNIQUESTechniques based on PCR are now used forthe diagnosis of a thalassaemia, ,B thalas-saemia, some Bfl thalassaemias and HPFHdeletions, the analysis of fi globin restrictionfragment polymorphisms (RFLPs) and thedetection of the genes for Hb S, Hb C, Hb D,Hb E and Hb 0 Arab. A fetal diagnosis resultis usually obtained in one to two days if themutations requiring analysis are known. If themutations have not been identified before-hand, the parental DNA samples will have tobe analysed first before the fetal DNA and thefetal diagnosis will take two to five days.

Detection ofknown mutationsThere are three PCR based techniques ingeneral use in fetal diagnosis laboratories: (1)the hybridisation of 32P-labelled allele specificoligonucleotide probes, either to amplifiedDNA fixed to nylon filters by dot blotting'5 orby reverse dot blotting with the probes fixedto the filter'6; (2) the use of allele specificprimers 7; and (3) agarose or polyacrylamidegel electrophoresis. The first two techniquesdetect point mutations and require a batteryof specific probes-one for every mutationlikely to be encountered. The task of screen-ing parental DNA samples for ,B thalassaemiamutations is greatly simplified by the knowl-edge of the ethnic origin of each person. TheDNA is then screened first for the commonmutations to that particular ethnic group, andthen if the tests prove negative, the rarermutations are analysed for. The mutationscarried by both partners can therefore beidentified in more than 95% of couples. Theremaining few per cent of people with anunidentifiable mutation either have an unde-scribed one or a rare one from a different eth-nic group. In these cases fetal diagnosis canoften be carried out by the indirect methodsof RFLP linkage or denaturing gradient gelelectrophoresis.The mutations for Hb S, Hb E, Hb D and

Hb O-Arab affect a restriction enzyme site inthe fi globin gene and are detected directly byamplification, followed by restriction enzymedigestion of the product and agarose gel elec-trophoresis.'4 Similarly, 13 fi thalassaemiamutations can be detected this way'8 and thismethod provides an alternative approach to

check prenatal diagnosis results involvingthese mutations. Gel electrophoresis ofamplified product is used for diagnosing the fithalassaemia deletion mutations,'9 a0 thalas-saemia, 0 Hb Lepore I and Bfl thalassaemia."Primers which are complementary tosequences flanking the deletion are used togenerate a specific product only in the pres-ence of the mutation (with the exception of ,Bthalassaemia, normal DNA is not amplifiedbecause the primers are too far apart).Normal DNA is amplified using a flankingprimer in combination to one complementaryto the deleted sequence.

Detection of unknown mutationsThere are now several PCR methods forcharacterising ,B thalassaemia mutations thatare undefined by direct detection methods.These are denaturing gradient gel elec-trophoresis (DGGE), chemical mismatchcleavage analysis, and single-stranded confor-mation polymorphism (SSCP) analysis, fol-lowed by direct sequencing of amplifiedsingle stranded DNA. The method mostwidely used is DGGE, a polyacrylamide gelsystem that separates DNA fragments as afunction of the melting temperature. It hasbeen found to be particularly useful for thediagnosis of,B thalassaemias in geographicalareas where a very heterogeneous spectrum ofmutations occurs.23

Indirect detection by RFLPSNearly all the /3 globin RFLPs discovered bySouthern blotting can now be analysed byPCR and restriction enzyme digestion.'4 Bycarrying out a family study with DNA fromchildren or grandparents, or both, the linkagephase of informative RFLPs can be estab-lished and ,B thalassaemia genes can be diag-nosed indirectly in fetal DNA withoutidentifying the mutations involved. The tech-nique has a slightly higher risk of error(03%) than a direct method because of thechance of recombination between the RFLPand the mutation locus. RFLP linkage analy-sis provides a useful alternative approach tothe diagnosis of /3 thalassaemia in the 80%of families that are found to be informative,and should be used for confirmation of adiagnosis obtained by direct mutationanalysis whenever possible.

PCR errorsMisdiagnosis may occur for several technicalreasons-that is, failure to amplify one ofthese two target DNA segments, maternalDNA contamination, contamination withpreviously amplified target DNA sequencesthrough aerosol contact, mispaternity, sampleexchange, incomplete digestion of amplifiedproduct by restriction endonucleases, andmispaternity and incorrect phenotype infor-mation resulting in a missed allele or incor-rect assignment of polymorphic markers.These errors can be encountered by care-ful laboratory practice in setting up PCRreactions and running appropriate controls,

201

group.bmj.com on February 15, 2012 - Published by jcp.bmj.comDownloaded from

Globin Gene Disorder Working Party

duplicate analysis of the fetal sample, the useof the minimum number of amplificationcycles, the use of a second different diagnos-tic approach and the amplification of a suit-able polymorphic marker, such as a variablenumber of tandem repeat (VNTR) sequence,to monitor the presence of maternal contami-nation.24

Requirements for prenatal diagnosisWhen a potentially at-risk couple is detectedthey will require counselling and if fetal diag-nosis is requested it will be necessary to: (a)confirm the parental phenotypes as deter-mined by the first centre; (b) send family orparental blood samples to a specialist DNAdiagnostic laboratory; and (c) arrange forfetal sampling and correct despatch of thefetal sample for fetal diagnosis. Ideally confir-mation of the parental phenotypes and identi-fication of the DNA mutations involvedshould be done before arrangements aremade for fetal sampling to define the exactnature of risk and to avoid unnecessary inva-sion of a pregnancy not at risk. To diagnosea+ f, thalassaemia or fi thalassaemia with nor-mal Hb A2 and low mean corpuscular haemo-globin may require a combination of carefulfamily studies, globin chain synthesis, andDNA analysis.

CARRIER STATE DETERMINATIONThis requires a blood count, including redcell indices, haemoglobin electrophoresis, asickle solubility test and an estimation of HbA2 and Hb F. Guidelines for these routinescreening tests are detailed in the associatedguidelines. These tests should be repeated atthe centre that has referred blood and fetalsamples for DNA analysis from an at-riskcouple. The referral form (see appendix)should be completed and accompany theblood samples to the DNA diagnostic labora-tory. Details of the ethnic origin (not place ofbirth) are particularly important for DNAanalysis.

a thalassaemiaThe typical haematological features for a tha-lassaemia carriers are shown in the table ofthe accompanying guidelines on the investiga-tion of a and JB thalassaemias. Some impor-tant points to note are that carriers for a+thalassaemia are haematologically normal orhave only slightly reduced indices (mean cor-puscular haemoglobin of 24-30 pg). Theindices for ao thalassaemia traitusually fall below the normal range (meancorpuscular haemoglobin of 28-32 pg) in theabsence of iron deficiency, but are indistin-guishable from those found in people withhomozygous a+ thalassaemia (mean corpus-cular volume of 19-24 pg).25 Therefore,definitive carrier diagnosis can only be doneby DNA studies. It is recommended that allpregnant women of Chinese, South EastAsian, or eastern Mediterranean originshould be screened for a° thalassaemia traitand offered prenatal diagnosis when indi-cated.

,B thalassaemiaThe haematological phenotype of heterozy-gous fi thalassaemia can be modified by inter-acting genetic or environmental factors. Theco-inheritance of a thalassaemia (usually thedeletion of two a globin genes) may raise themean corpuscular haemoglobin and volumesufficiently, in some cases, to values that liewithin the normal range. Alpha thalassaemia,however, does not affect the Hb A2 concen-tration, unlike iron deficiency which maydecrease the Hb A2 value, although it nor-mally remains within the fi thalassaemia car-rier range. Therefore in any population inwhich a and fi thalassaemia occur together,the best method of screening for fi thalas-saemia in pregnancy is to measure the Hb A2concentration.A subgroup of ,B thalassaemia alleles exists

in which heterozygotes have normal or mini-mally increased Hb A2 concentrations. Theones with reduced red cell indices may beconfused with heterozygous a° thalassaemia.Those with normal red cell indices are trulysilent (the only abnormality being an imbal-ance of the f,/a globin chain synthesis ratio)and are not detectable by haematologicalscreening techniques. Fortunately, the latterare both rare and very mild alleles, hithertofound only in one or both parents of patientsaffected by thalassaemia intermedia. Thesealleles are the Mediterranean - 101 mutationand the Indian CAP + 1 mutation. Thetriple a globin gene arrangement is anothertruly silent allele, which in combination witha fi thalassaemia mutation, can result in thephenotype of fi thalassaemia intermedia.Normal Hb A2 fi thalassaemia with reduced

red cell indices is usually caused by thedouble heterozygosity for 5 and fi thalas-saemia. The mutation causing defective 5gene expression may be in cis or trans to the fithalassaemia allele, which may be a mildallele such as codon 27 (Hb Knossos) orIVSI-5 (Corfu Bfl thalassaemia), or severesuch as IVSI-110, ISII-745, and the f,°codon 39 mutation. However, several fithalassaemia mutations by themselves havebeen observed in people with borderline tonormal Hb A2 values, the best knownexample being the Mediterranean mutationIVSI-6. Finally, adults heterozygous for therare condition of (y5ff)' thalassaemia havered cell indices typical of f, thalassaemiatrait but with normal Hb A2 and Hb F con-centrations.

Adults with normal or low mean cor-puscular volume of haemoglobin, normalto reduced Hb A2 concentration and highHb F are heterozygous for 5fl thalassaemiaor hereditary persistence of fetal haemo-globin. Although the distinction between thetwo conditions is not always clearcut, ingeneral heterozygotes for c5fl thalassaemiahave 5-15% Hb F, hypochromic microcyticred cells, and a heterogenous distribution ofHb F in peripheral red blood cells; those withHPFH have 15-25% Hb F, normal indices,and a homogenous distribution of Hb F. Adefinitive diagnosis can be made by globin

202

group.bmj.com on February 15, 2012 - Published by jcp.bmj.comDownloaded from

Guidelines for the fetal diagnosis ofglobin gene disorders

chain synthesis (normal in HPFH, unbal-anced in Bfl thalassaemia), and by DNAanalysis.

SAMPLE REQUIREMENTSAdult blood samples in heparin or EDTA forcarrier determination by globin chain synthe-sis have to be fresh-received by the labora-tory within a few hours-and are besttransported at 40C. It is essential to contactthe referral laboratory about transportarrangements for blood samples for such testsbefore arranging for the patient to attend.Fetal blood sampling is usually carried out atthe prenatal diagnosis centre.

Blood samples in heparin or EDTA forDNA analysis can be sent by overnight deliv-ery or first class post without refrigeration.The maximum delay before deterioration isabout three days so blood taken on a Fridayis best stored in a fridge until Mondayand then posted to arrive Tuesday.Anticoagulated blood (EDTA or heparin) (10ml) is normally sufficient from each parent. IfRFLP linkage analysis is to be performed inaddition to direct detection of /3 thalassaemiamutations then additional blood samples arerequired from either: (i) a homozygous nor-mal child or an affected child with /B thalas-saemia major; (ii) a child with /3 thalassaemiatrait plus one set of grandparents; (iii) if nochildren are available then both sets of futuregrandparents would be required. The samplesshould be clearly labelled and be accompa-nied with a request form detailing the follow-ing relevant particulars.

(1) Haematological details: Hb, RBC,MCV, MCH, sickle test, Hb electrophoresis,Hb A2, Hb F, H bodies, ferritin, serumiron/total iron binding capacity.

(2) Patient particulars: surname, firstname, date of birth, address, ethnic origin,referring consultant, gestational age.

(3) Details of family history and previouschildren.

(4) Confirmation the family has beencounselled as to the nature of the disease andobstetric risks and error rates.

(5) Name and address of contact forarrangement of follow up studies.

Chorionic villus samples must be dissectedfree of any contaminating maternal deciduaby microscopic dissection before sending tothe fetal diagnosis referral centre. Thesesamples can be sent by overnight deliveryeither in tissue culture medium, or prefer-ably, in a special lysis buffer obtainablefrom the DNA diagnosis laboratory.Amniotic fluid samples (15-20 ml) are bestsent to the DNA laboratory as quickly as pos-sible and must be received within 24 hours. Ifa longer transit time is anticipated the amnio-cytes should be sent resuspended in tissueculture medium. In many cases it is now pos-sible to arrange for the fetal sample to go to aregional molecular genetics laboratory forDNA extraction. The fetal DNA can then beposted to the haemoglobinopathy diagnosislaboratory without risk of deterioration ifdelayed in transit.

RISKS AND MISDIAGNOSIS RATESBecause both chorionic villus sampling andfetal DNA analysis are still in the stages ofdevelopment, it is vital that parents are givenaccurate advice about the current rates offetal loss and also the likely misdiagnosis rate.Current figures suggest that the fetal loss ratefollowing chorionic villus sampling in the firsttrimester is in the range of 2-3%. The errorin DNA analysis due to such factors asgenetic recombination, technical problems inthe laboratory, and maternal contaminationof chorionic villus sampling specimens hasbeen determined to be 1% for fetal diagnosescarried out using Southern blotting. Thereare not yet sufficient data to give anabsolutely accurate assessment of the misdi-agnosis rate for the recently introduced PCRbased techniques but experience to date sug-gests it may be in a similar range of 1-2%.

It is essential that follow up data areobtained on all cases that have undergonefetal diagnosis by chorionic villus samplingDNA analysis; ideally this should includeboth haematological and developmentalassessment, as well as any fetal abnormalities.These data should be available to both theDNA laboratory and the referral centre:

(a) Information on data of birth andbirthweight and any information observed inthe neonatal period.

(b) Neonatal study of cord blood (elec-trophoresis or globin biosynthesis), or stan-dard carrier test at six months or later.

(c) Confirmation of genotype by DNAanalysis.

COUNSELLINGNo couple should be offered fetal diagnosiswithout proper counselling involving properlytrained counsellors. Counselling requires anaccurate determination of the parental geno-types and a good understanding of clinicaloutcomes of the haemoglobin disorders andtheir various interactions. It must include thenature and prognosis of the disorder involvedand the treatment available; the genetic risksand possible methods of fetal diagnosis; theobstetric risks and risk of misdiagnosis.

Summary of guidelines for centressending samples for fetal diagnosis

(1) A well run diagnostic service requiresclose liaison between haematologists, theirobstetric colleagues, and the DNA referrallaboratory.(2) Haematological findings determinedpreviously elsewhere should be recheckedfor accuracy by the referring hospital.(3) Counselling of families requires theaccurate determination of parental geno-types, a good understanding of the interac-tions producing the different haemoglobindisorders and the clinical picture of eachdisorder, and accurate advice on the cur-rent rates of fetal loss and the likely fetalDNA analysis error.(4) All haematology results and patientdetails should be sent to the DNA referral

203

group.bmj.com on February 15, 2012 - Published by jcp.bmj.comDownloaded from

Globin Gene Disorder Working Party

laboratory. A referral form to accompanyblood samples is provided.(5) It is essential to contact the DNAreferral laboratory to discuss transportarrangements for samples.(6) Family or parental blood samples forDNA analysis are required for fetal diag-nosis of both thalassaemia and sickle celldisease.Blood samples for DNA analysis ideallyshould be sent at least one week before afetal sample to allow time for mutationidentification. However, current tech-niques now allow parental and fetal sam-ples to be sent at the same time, on theunderstanding that the fetal diagnosis maytake a few days longer and there is a smallchance of diagnostic failure from not beingable to identify one of the ,B thalassaemiamutations.(7) It is essential to obtain follow up datafor chorionic villus sampling DNA diag-noses. These should always be returned tothe DNA laboratory and the referringcentre.

Summary of guidelines for thosecarrying out fetal diagnosis

(1) Ensure that adequate haematologicaldata have been provided.(2) Ensure that the mother (or couple)have been adequately counselled.(3) Ensure that chorionic villus sampleshave undergone careful microscopic dis-section to remove any contaminatingmaternal decidua.(4) Always analyse parental DNA samplesand appropriate control DNAs alongsidethe fetal DNA sample.(5) Always repeat the fetal DNA analysistest to double check the result.(6) Use a second alternative DNA analysistest whenever possible to confirm a result.(7) Use alimited number of amplificationcycles to minimise co-amplification ofmaternal DNA.(8) Check for maternal DNA contamina-tion by amplification of polymorphicmarkers.(9) The fetal diagnosis report should detailthe types of DNA analysis used and clearlystate the risk of misdiagnosis.

Usefil addresses(1) Characterisation of mutations and fetaldiagnosis for thalassaemia and sickle celldisease is available as a national servicefrom:Dr John Old, National HaemoglobinopathyReference Service, Institute of MolecularMedicine, John Radcliffe Hospital, OxfordOX3 9DU Tel: 0865-222 449 Fax: 0865-222 500(2) Fetal diagnosis for thalassaemia andsickle cell disease is also provided by thefollowing regional centres:Dr M Petrou, Dr B Modell, PerinatalCentre Department of Obstetrics and

Gynaecology, 88-96 Chenies Mews,London WC1E 6HX Tel: 071-387 9300extn. 5230 Fax: 071-380 9864Dr M Layton, Department of Haemato-logical Medicine, King's College Hospital,Denmark Hill, London SE5 9RS Tel:071-326 3239 Fax: 071-326 3514For sickle cell disease only:Dr M Patton, SW Thames RegionalGenetics Service, St Georges HospitalMedical School, Cranmer Terrace,London SW17 ORE Tel: 081-672 9944.

1 Weatherall DJ, Clegg JB. The thalassemia syndromes. 3rdedn. Oxford: Blackwell Scientific Publications, 1981.

2 Higgs DR. The molecular genetics of the a globin genefamily. EurJ Clin Invest 1990;20:340-7.

3 Petrou M, Brugiatelli M, Old J, et al. Alpha thalassaemiahydrops fetalis in the UK: the importance of screeningpregnant women of Chinese, other South East Asianand Mediterranean extraction for alpha thalassaemiatrait. BrJ ObstetGynaecol 1992;99:985-9.

4 Kazazian HH Jr. The thalassemia syndromes: molecularbasis and prenatal diagnosis in 1990. Semin Hematol1990;27:209-28.

5 Huisman THJ. Frequencies of common fl-thalassaemiaalleles among different populations: variability in clinicalseverity. BrJHaematol 1990;75:454-7.

6 Thein SL, Hesketh C, Taylor P, etal. Molecular basis fordominantly inherited inclusion body fl-thalassemia.ProcNadAcad Sci USA 1990;87:3924-8.

7 Gonzalez-Redondo JM, Stoming TA, Kutlar A, et al.AC -a T substitution at nt-101 in a conserved DNAsequence of the promoter region of the /5 globin gene inassociation with 'silent',B thalassemia. Blood 1989;73:1705-11.

8 Wainscoat JS, Thein SL, Weatherall DJ. Thalassaemiaintermedia. Blood Rev 1987;1:273-9.

9 Powars DR.,tP-gene cluster haplotypes in sickle cellanemia. Hematol Oncol Clin North Am 1991;5:475-93.

10 Padmos MA, Roberts GT, Sackey K, et al. Two differentforms of homozygous sickle cell disease occur in SaudiArabia. BrJHaematol 1991;79:93-8.

11 Clegg JB. Haemoglobin synthesis. In: Weatherall DJ,eds. Methods in haematology. Edinburgh: ChurchillLivingstone, 1983:54-73.

12 Alter BP. Antenatal diagnosis using prenatal blood. In:Alter BP, eds. Methods in haematology. Edinburgh:Churchill Livingstone, 1989:114-33.

13 Old JM. Fetal DNA analysis. In: Davies KE, eds. Humangenetic disease analysis: A practical approach. Oxford:Oxford University Press, 1993:1-20.

14 Old JM, Ludlam CA. Antenatal diagnosis. In: Hann IM,Gibson BES, eds. Bailltire's dinical haematology-Paediatric haematology. London: Bailliere Tindall, 1991:391-428.

15 Saikl RK, Bugawan TL, Horn GT, etal. Analysis of enzy-matically amplified /5 globin and HLA-DQ a DNA withallele-specific oligonucleotide probes. Nature 1986;324:163-6.

16 Saikl RK, Walsh PS, Levenson CH, Erlich HA. Geneticanalysis of amplified DNA with immobilized sequence-specificoligonucleotide probes. Proc Nad Acad Sci USA1989;86:6230-4.

17 Old JM, Varawalla NY, Weatherall DJ. The rapid detec-tion and prenatal diagnosis of /5 thalassaemia in theAsian Indian and Cypriot populations in the UK. Lancet1990;ii:834-7.

18 Kazazian HH Jr. Use of PCR in the diagnosis of mono-genic disease. In: Erlich H, eds. PCR technology. NewYork: Stockton Press, 1989:153.

19 Faa V, Rosatelli MC, Sardu R, Meloni A, Toffoli C, CaoA. A simple electrophoretic procedure for fetal diagnosisof fl-thalassaemia due to short deletions.Prenat Diagn1992;12:903-8.

20 Bowden DK, Vickers MA, Higgs DR. A PCR-basedstrategy to detect the common severe determinants ofa-thalassaemia. BrJ Haematol 1992;81:104-8.

21 Camaschella C, Alfarano A, Gottardi E, etal. Prenataldiagnosis of fetal hemoglobin Lepore-Boston disease onmaternal peripheral blood. Blood 1990;75:2102-6.

22 Ghanem N, Vidaud M, Plassa F, Martin J, Goossens M.Direct carrier detection and prenatal diagnosis ofSicilian and Spanish (t5fl)'-thalassaemias. Mol CellProbes1993;7:167-8.

23 Losekoot M, Fodde R, Harteveld CL, Van Heeren H,Giordano PC, Bemini LF. Denaturing gradient gel elec-trophoresis and direct sequencing of PCR amplifiedgenomic DNA: a rapid and reliable diagnostic approachto beta thalassaemia. BrJHaematol 1991;76:269-74.

24 Decorte R, Cuppens H, Marynen P, Cassiman J-J. Rapiddetection of hypervariable regions by the polymerasechain reaction technique. DNA Cell Biol 1990;9:461-9.

25 Higgs DR a-thalassaemia. In: Higgs DR, Weatherall DJ,eds. Bailliere's dinical haematology. International practiceand research: The haemoglobinopathies. London: BailliereTindall, 1983:117-50.

204

group.bmj.com on February 15, 2012 - Published by jcp.bmj.comDownloaded from

doi: 10.1136/jcp.47.3.199 1994 47: 199-204J Clin Pathol

  Haematology Task Force.Working Party of the BCSH Generalgene disorders. Globin Gene Disorder Guidelines for the fetal diagnosis of globin

http://jcp.bmj.com/content/47/3/199.citationUpdated information and services can be found at:

These include:

References http://jcp.bmj.com/content/47/3/199.citation#related-urls

Article cited in:

serviceEmail alerting

box at the top right corner of the online article.Receive free email alerts when new articles cite this article. Sign up in the

Notes

http://group.bmj.com/group/rights-licensing/permissionsTo request permissions go to:

http://journals.bmj.com/cgi/reprintformTo order reprints go to:

http://group.bmj.com/subscribe/To subscribe to BMJ go to:

group.bmj.com on February 15, 2012 - Published by jcp.bmj.comDownloaded from