EMQN Best Practice Guidelines for Molecular and Haematology Methods for Carrier Identification

SUPPORTING INFORMATION

EMQN Best Practice Guidelines for molecular and haematology methods for carrier identification and prenatal diagnosis of the haemoglobinopathies

Joanne Traeger-Synodinos1, Cornelis L Harteveld2, John M. Old3, Mary Petrou4, Renzo Galanello5, Piero Giordano2, Michael Angastioniotis6, Barbara De la Salle7, Shirley Henderson3, Alison May8, on behalf of contributors to the EMQNhaemoglobinopathiesbest practice meeting.

Department of Medical Genetics, University of Athens, Choremeio Research Laboratory, St Sophia's Children's Hospital, Athens 11527, Greece.
Department of Human and Clinical Genetics, LeidenUniversityMedicalCenter, Einthovenweg 20,2333ZC Leiden, The Netherlands
NationalHaemoglobinopathyReferenceLaboratory, MolecularHaematology, L4, JohnRadcliffeHospital, Oxford, OX3 9DU, UK.
Haemoglobinopathy Genetics Centre, University College London Hospitals NHS Foundation Trust and Institute of Women’s Health, University College London, 86-96 Chenies Mews, London WC1 E6HX, UK.
Ospedale Regionale Microitemie, Via Jenner (sn), 09121 Cagliari, Italy
Thalassaemia International Federation, PO Box 28807, 2083 Nicosia, Cyprus.
UK NEQAS, General Haematology, PO Box 14, WD18 0FJ, UK.
Department of Haematology, Cardiff University Medical School, University Hospital of Wales, Heath Park, Cardiff, UK.

FIRST-LINE HAEMATOLOGICAL METHODS

Haemoglobin (Hb) pattern analysis

The methods for a presumptive identification of abnormal haemoglobins include

Haemoglobin electrophoresis at pH 8.6 using cellulose acetate membrane. This method will reliably detect the common haemoglobin variants, ie Hb’s S, C, D-Punjab, E, O-Araband the Lepore Hbs. Hb H and Hb Barts may also be detected if suitable run times are used. Many other variants are also detectable, eg Hb’s J’s, N’s, Q’s, other Ds and Gs and Hasharon.
Haemoglobin electrophoresis at pH 6.0 using acid agarose or citrate agar gel. This method is useful in selected cases for distinguishing Hb’s C, E, and O-Arab from each other, also Hb S and Hb D-Punjab from each other. (Of note is that the migration patterns are different for acid agarose or acid citrate agar gels).
Isoelectric focusing (IEF). IEF is a sensitive method, giving good separation of haemoglobin variants but requires considerably more expertise for interpretation than electrophoresis since adduct fractions also separate.
High Performance Liquid Chromatography (HPLC) is a recommended method for simultaneous automatic detection and quantitation of haemoglobin fractions. Since systems are automated, operation of analysers is simple. However, interpretation of the chromatograms requires expertise. Also, attention must be paid to quality control, especially for measurement of Hb A2. (16;17;53).
Capillary electrophoresis (CE) is a high through-put, automated alternative to alkaline electrophoresis, providing separation and quantitation of normal and abnormal Hb fractions.(16-18)

Recommendations for application and interpretation of haemoglobin pattern analysis

When a haemoglobin variant is detected further separation tests are recommended to confirm its identity. However, second or even third line separation procedures do not always ensure identification for all variants. In these cases molecular (DNA) analysis is the only method that ensures identification.

On most HPLC systems, derivatives of Hb S, and other similarly-eluting haemoglobin variants, may co-elute with Hb A2 resulting in an overestimation of the Hb A2 level (3.5-4.5%). In such cases the peak value should not be reported as a percentage of Hb A2.
Measuring Hb A2 levels in the presence of Hb S and at least 50% Hb A (if no recent transfusion history) is of no diagnostic value, since the presence of more than 50% Hb A logically excludes co-existing β-thalassaemia. For the same reason Hb A2 estimation in the presence of  50% Hb A and Hb C, D, E, O, or any other -variant is of no diagnostic value.
When a high percentage of Hb A2 is found in the presence of Hb S, but in the absence of Hb A and in the presence of microcytosis, this may indicate either Hb S/βo-thalassaemia or homozygous Hb S and -thalassaemia. In such cases family and/or molecular analysis are recommended. ,
There are rare sickling variants (currently 11 known) with an additional amino acid substitution to that of Hb S that mostly do not migrate or elute like Hb S (Table 2supp) and are only detectable at the protein level if a sickle solubility test is used for all variant haemoglobins of any significant quantity (>20% total).
A sickle solubility test (see Supporting Information) can be used to confirm the presence of Hb S if suspected (54).
When Hb H disease or other unstable haemoglobin variants are suspected, always analyse fresh blood samples.
It is important to inspect carefully the chromatogram or electropherogram before the results are authorized to check for baseline, peak resolution and overlapping peaks.
An alkaline denaturation test may confirm the presence of significantly (>5%) increased Hb F
DCIP (2.6 dichlorophenolindophenol)) test (see Second-line haematology tests) can be used to screen for Hb E (28).
When the presence of a high affinity but electrophoretically silent Hb variant is suspected, oxygen dissociation analysis may be useful, but definitive identification is only achieved by molecular (DNA) analysis.

Quantification of Hb A2

Methods include (16;20;21):

Electrophoresis and elution, which is accurate but time-consuming.
Microchromatography, which is accurate but time-consuming.
HPLC which is accurate and high-throughput.
Capillary electrophoresis which is accurate and high throughput.
Hb electrophoresis with automatic densitometry, which is not recommended.

Quantification of Hb F

Methods for measuring Hb F levels include:

Alkali denaturation - The modified method of Betke (55) has excellent reproducibility, at low values of Hb F (<10%), giving worthwhile results in virtually all clinical situations. However, it is technically exacting and is not often used.
HPLC (17) and CE are much more accurate although some devices may overestimate Hb F due to overlap with the first glycated fraction.

NOTE: some variant haemoglobins, or adducts of chromatographically silent variants may co-elute with Hb F and cause false elevations, so confirmatory tests may be required. The use of standards and internal controls are recommended (16).

Iron (Fe) status

There are several parameters that can be measured to evaluate the the iron status of an individual, see below:

a) Zinc protoporphyrin (ZnPP). For this method the sample can be analysed from the same tube as the blood count. Analysis is simple and fast, although it requires a specific instrument. Zinc protoporphyrin is elevated in iron deficiency, but may be falsely high in lead intoxication or if the bilirubin levels are raised. Some studies have shown increased levels in thalassaemia carriers with normal iron levels and varying degrees of discrimination between uncomplicated heterozygous thalassaemia and iron deficiency have been reported.

b) Serum ferritin measurements are most commonly used for indicating iron deficiency, but may be falsely elevated during infection, inflammation, liver disease or neoplasia.

c) Transferrin saturation measures serum iron compared to the total iron binding capacity (Fe/TIBC). However, serum iron may be falsely low in infection and inflammation despite normal iron stores..

SECOND-LINE HAEMATOLOGICAL METHODS

There are a number of additional methods that may be useful for investigating unusual samples that do not have a definite diagnosis with conventional methods.

Red Cell Morphology (RCM)

Thalassaemia carriers will present with classic target cells, microcytes, fragmentocytes, basophilic stippling and anisopoikylocytosis.

Reticulocytes and F-cells can be counted using flow cytometry (25) or on a slide as a smear (14).

Single tube osmotic fragility test (OF).

This easy and cheap method is largely used in emerging countries doing population screening but can be very useful also in the modern laboratory. (25) .

Inclusion body test for alpha thalassaemia

Carriers of alpha thalassaemia with genotypes or combined genotypes causing loss of about half the usual alpha globin production have a very small number of red blood cells (often<1/5000) that contain sufficient HbH to precipitate on incubation with certain oxidative dyes and be visualized on smears under the microscope. This test has to be carried out on freshly-taken blood and 150 fields may have to be scrutinized. It has high specificity, giving a rapid confirmation of alpha thalassaemia for some making it a valued and useful test. It is not sensitive enough to be used either for all types of alpha thalassaemia or to exclude heterozygous alpha zero thalassaemia. Furthermore, it cannot distinguish between the various genotypes.

Globin chain synthesis

This was one of the first methods used to support differential diagnosis of globin gene disorders. However it is quite a cumbersome, time consuming method involving the use of radio-isotopes. Nevertheless, it may provide useful information for diagnosing in selected atypical cases. Once radioactively labelled the haem-free denatured and solubilised globin chains can be separated for quantification by a number of methods:

a)CMC chromatography method. - Very accurate but time consuming method for evaluating relative rate of globin chains synthesised in reticulocytes.

b)HPLC – Potentially a less time consuming method, but it needs careful standardization to be accurate and reliable.

c)IEF- rapid and convenient, with potential to process multiple samples simultaneously (27) .

Globin chain separation

Can be undertaken either by HPLC or IEF, and is useful for indicating which globin chain is affected, thus giving evidence for the nature and potential significance of the variant(s) present.

Functional tests for Hb S

If there is an abnormal fraction that runs in the position of Hb S, then the variant can be confirmed without molecular analysis, either using one of the commercially available solubility tests or the sickling test.

Note: the solubility test may produce false positive results in the presence of elevated Hb F or with unstable haemoglobin variants that usually do not migrate or elute on the same position as Hb S. On the other hand there are 11 rare sickling variants with an additional amino acid substitution to that of Hb S that mostly do not migrate or elute like Hb S (Table 2Supp).

Mass spectrometry

Specialized method based on analysing tryptic digests of whole blood. Although there may be only a small shift in mass for some common variants, it is very effective for the characterisation of Hb variants especially if used in conjunction with other methods (29). However this method may identify the specific amino-acid change but does not necessarily confirm the DNA sequence variation or, in the case of duplicated genes such as - or -, the gene involved

DCIP (2.6 dichlorophenolindophenol) test

The DCIP test is also, available in a kit form, can be used for Hb E carrier screening whenever more complex and expensive tests are not readily available (28).

Heinz body formation

This is not a very specific, but useful for detecting the presence of some unstable variants

Oxygen dissociation curve

The measurement of the oxygen dissociation curve may be useful for confirming the presence of Hb variants with altered oxygen affinity. However, the abnormal fraction should ideally be purified first, the measurement is complex and few laboratories are equipped to do it. Pragmatic alternatives are the p50 saturation measurement in the presence of erythrocytosis and elevated PCV for variants with increased O2 affinity or some degree of cyanosis in carriers of variants with reduced O2 affinity.

Table 1supp. Traditional and HGVS nomenclature of globin gene variantsreferred to in these guidelines

Traditional nomenclature / HGVS nomenclature
Variantsinvolving HBB gene 1
Hb LeporeHollandia / NG_000007.3:g.63290_70702del
Hb Lepore-Baltimore / NG_000007.3:g.63564_70978del
Hb Lepore-Boston-Washington / NG_000007.3:g.63632_71046del
Hb C / c.19G>A
Hb C-Harlem / c.[20A>T;220G>A]
Hb C-Ndjamena / c.[20A>T;112T>G]
Hb C-Ziguinchor / c.[20A>T;176C>G]
Hb DPunjab / c.364G>C
Hb E / c.79G>A
Hb Hope / c.410G>A
Hb I-Toulouse / c.199A>G
Hb Jamaica Plain / c.[20A>T;205C>T]
Hb NorthShore. / c.404T>A
Hb OArab / c.364G>A
Hb O-Tibesti / c.[34G>A;364G>A]
Hb Quebec-Chori / c.263C>T
Hb S / c.20A>T
Hb S-Antilles / c.[20A>T;70G>A]
Hb S-Clichy / c.[20A>T ;26A>C]
Hb S-Cameroon / c.[20A>T;271G>A]
Hb S-Oman / c.[20A>T;364G>A]
Hb S-Providence / c.[20A>T;249G>T or 249G>C]
Hb Shelby / c.394C>A
Hb S-South End / c.[20A>T;399A>C]
Hb S-Travis / c.[20A>T;428C>T]
 -101 (CT) / c.-151C>T
 -92 (CT) / c.-142C>T
 +33 (CG) / c.-18C>G
IVS2-844 (CG) / c.316-7C>G
 +1480 (CG) / c.*6C>G
Cap+1 (AC) / c.-50A>C
 IVS1-6 (TC) / c.92+6T>C
Variaantsinvolving HBA1 or HBA2 genes 2
--FIL / NG_000006.1:g.11684_43534del31851
--MEDI / NG_000006.1:g.24664_41064del16401
--SEA / NG_000006.1:g.26264_45564del19301
--THAI / NG_000006.1:g.10664_44164del33501
-()20.5 / NG_000006.1:g.15164_37864del22701
-3.7 (Type I) / NG_000006.1:g.34164_37967del3804
-3.7 (Type II) / Not defined
-3.7 (Type III) / Not defined
-4.2 / Not defined
Hb Adana / c.179G>A (HBA2 or HBA1)
Hb Agrinio / c.89T>C
Hb Constant Spring / c.427T>C
Hb Hasharon / c.142G>C
Hb Taybee / c.118_120delACC (HBA1)
2 gene variantat the initiation codon (e.g. ATGACG / c.2T>C
IVS1 donor site (-GAGGT- 5 base pair deletion) / c.95+2_95+6delTGAGG
2 gene polyadenylation signal AATAAA->AATA- - / c.*93_*94delAA
2 gene polyadenylation signal AATAAA->AATGAA / c.*92A>G
2 gene polyadenylation signal AATAAA->AATAAG / c.*94A>G

1 NM_000518.4< HBB;

2 HBA2; NM_000558.3 HBA1. All variantsare on the HBA2 gene unless stated otherwise between brackets.

Table 2supp. The sickling variants with two amino acid substitutions

Haemoglobin / Amino acid substitution / HPLC
Hb S-South End / beta 6(A3) GluVal & beta 132(H10) LysAsn / Runs with Hb A
Hb S-Antilles / beta 6(A3) GluVal & beta 23(B5) ValIle / Separate peak
Hb C-Ziguinchor / beta 6(A3) GluVal & beta 58(E2) ProArg / Unknown
Hb C-Harlem / beta 6(A3) GluVal & beta 73(E17) AspAsn / Runs with Hb A2
Hb S-Providence / beta 6(A3) GluVal & beta 82(EF6) LysAsn / Separate peak
Hb S-Travis / beta 6(A3) GluVal & beta 142(H20) AlaVal / Separate peak
Hb C-Ndjamena / beta 6(A3) GluVal & beta 37(C3) TrpGly / Separate peak
Hb S-Oman / beta 6(A3) GluVal & beta 121(GH4) GluLys / Separate peak
Hb S-Clichy / Beta6(A3) GluVal & beta 8(A5)LysThr] / Separate peak
Hb S-Cameroon / beta 6(A3) GluVal & beta 90(F6) GluLys / Separate peak
Hb Jamaica Plain / beta 6(A3) GluVal & beta 68(E12) LeuPhe / Runs with Hb S

Table 3supp. Known causes underlying Hb A2 levels outside the normal range

Increase (excluding -thalssaemia variants) / Reduction
Genetic / Genetic
KLF1 variants / -thalassaemia
Triplicated  gene / -chain variants
Some unstable variants / -chain variants
Hb variants eluting with or close Hb A2 / Hb Lepore1
-thalassaemia2
 and  thalassaemia; some mild  thalassaemia variants
Acquired / Acquired
Hyperthyroidism / Severe iron deficiency anaemia
Megaloblastic anaemia / Sideroblastic anaemia
Aplastic crisis in HS / Lead poisoning
Antiretroviral drugs / Leukaemia, aplastic anaemia
Pseudoanthoma elasticum

Note 1: a) In Hb Lepore carriers, the expected parameters are: 2.0-2.5% Hb A2, 1-3% Hb F, 3-5% Hb Lepore-Boston-Washington. However, falsely high Hb A2 levels may be observed (up to 10-15%) when using HPLC, since Hb Lepore co-elutes with Hb A2.

Note 2: Co-inheritance of -thalassaemia, particular Hb H disease may reduce Hb A2 levels.