SEARCH

SEARCH BY CITATION

Keywords:

  • beta thalassaemia carrier;
  • action values;
  • antenatal screening

Summary

  1. Top of page
  2. Summary
  3. Materials and methods
  4. Results
  5. Discussion
  6. Author contributions
  7. Conflict of interest disclosure
  8. References

National antenatal screening of all pregnant women in England is carried out using standards and guidelines produced by the National Health Service Sickle Cell and Thalassaemia Screening Programme. The algorithms for detection of beta thalassaemia carrier status rely on action criteria, which are set using the percentage Hb A2 and mean corpuscular haemoglobin (MCH) values. Three groups of samples: MCH <27 pg and Hb A2 3·5–3·9%, MCH ≥27 pg and Hb A2 4–4·3% and MCH ≥27 pg and Hb A2 3·5–3·9% were selected from a sample population of 59 500 to assess the validity and predictive value of the action criteria – 25 false positives (0·042% of total) and nine false negatives (0·015% of total) were detected. These findings support the continuation of the current action values.

In England, national antenatal screening of all pregnant women is carried out according to standards and guidelines produced by the National Health Service (NHS) Sickle Cell and Thalassaemia Screening Programme. Algorithms for laboratory screening are determined by the prevalence of sickle cell disease in the population. In low prevalence areas (defined as an area where the estimated fetal prevalence of sickle cell disease is less than 1·5 per 10 000 pregnancies) all women in the antenatal population are offered screening for thalassaemia using red cell indices [mean corpuscular haemoglobin (MCH) <27 pg]. A family origin questionnaire is used to assess the risk of the mother and/or the baby's father being a carrier of a haemoglobin variant. Women who are identified as high risk are offered laboratory testing (NHS Sickle Cell & Thalassaemia Screening Programme, 2012).

In high prevalence areas, all pregnant women are offered screening with a full blood count (FBC), haemoglobin variant screening and Hb A2 measurement, the latter two via high performance liquid chromatography (HPLC) or capillary zone electrophoresis. In both high and low prevalence areas where a carrier mother is identified, the baby's father is offered screening. The most recent data report (NHS Sickle Cell & Thalassaemia Screening Programme, 2013) shows that 6655 women had a positive screen for thalassaemia (α and β combined) from over 730 000 pregnancies. There were 173 high-risk pregnancies for thalassaemia, of which 75 had pre-natal diagnosis and 87·5% of those with an affected result and known pregnancy outcome had a termination.

Beta (β) thalassaemia carrier status is assessed via the red cell indices and Hb A2 measurement but there is limited evidence about the appropriate action values for Hb A2. Screening values of an MCH <27 pg and Hb A2 ≥4·0% are well established and considered to have a high positive predictive value for β thalassaemia carrier status. Testing of the baby's father is recommended for pregnant women with these findings. Additionally, testing of the baby's father is also recommended for those women with an MCH <27 pg and Hb A2 ≥3·5% but less than 4·0%, and for those with an MCH ≥27 pg and Hb A2 ≥4%. Both groups are classified as possible β thalassaemia carriers but the specificity is uncertain. Women with an MCH ≥27 pg and Hb A2 <4% are not considered to be β thalassaemia carriers and testing of the baby's father is not recommended (NHS Sickle Cell & Thalassaemia Screening Programme, 2012). The algorithm relies on the reliability and reproducibility of the Hb A2 result and the positive predictive value of the Hb A2 action value of 3·5% in conjunction with the MCH.

Hb A2 measurement is dependent on a number of factors. including different laboratory HPLC analysers and laboratory methods in use. A study evaluating the results produced by different analysers for a the UK National External Quality Control Service (UK NEQAS) demonstrated bias in Hb A2 values generated from different analysers, which would lead to differences in clinical assessment of β thalassaemia carrier status (Batterbee et al, 2010). There are additional difficulties associated with the separation and quantitation of small peaks or bands in any chromatographic or electrophoretic system. Some HPLC systems overestimate Hb A2 in the presence of Hb S, although this is not problematic if Hb A is greater than Hb S. A number of pre analytical factors are also known to affect the Hb A2 value. Hb A2 may be lowered by up to 0·5% in cases of severe iron deficiency anaemia. Elevated Hb A2 levels have been reported in hyperthyroidism, megaloblastic anaemia and human immunodeficiency virus (HIV) infected patients on antiretroviral therapy. Other acquired and inherited conditions can also affect the Hb A2 value. This includes the presence of haemoglobin variants e.g: Hb Yaounde and Hb La Desirade (Stephens et al, 2011).

The antenatal screening programme in England is designed to detect significant maternal haemoglobinopathies including Hb SS, SC, SDPunjab, SE, SOArab, S/Lepore, S/β thalassaemia, β thalassemia major and intermedia, Hb H disease; and significant carrier states in the mother, including Hb AS, AC, ADPunjab, AE, AOArab, A Lepore, β thalassaemia carriers, δβ thalassaemia carriers, α0 thalassaemia carriers and carriers of hereditary persistence of fetal haemoglobin (HPFH). The aim of the screening programme is to offer sickle cell and thalassaemia screening to all women and couples in a timely manner in pregnancy, to facilitate informed choices regarding participation in the screening programme and provide help for those couples identified by screening as being at higher risk. Screening programmes do not expect to have a diagnostic sensitivity of 100% and therefore it is likely that some cases will be missed. Also, it is recognized that over-identification of variants with no or little clinical significance will cause unnecessary parental concern with little clinical benefit. There is therefore a balance to be drawn between algorithms that allow identification of all haemoglobinopathies, many of which are not significant, resulting in a large number of false positives, and those that may generate fewer false positives, but may also miss significant haemoglobinopathies. These risks are documented in the risk assessments included with the screening algorithms (NHS Sickle Cell & Thalassaemia Screening Programme, 2012). For β thalassaemia these include:

  1. ‘silent’ or ‘near silent’ β thalassaemia carriers. Some β thalassaemia carrier genotypes are associated with borderline raised Hb A2 levels and an action value of 3·5% in conjunction with an MCH of <27 pg will miss some cases. Examples of such mutations include c.−50A>C, c.92+6T>C, c.−151C>T and the Poly A site mutations (c.*+110T>C, c.*+111A>G, c.*+112A>G, c.*+113A>G).
  2. Any β thalassaemia carriers where the MCH is ≥27 pg.
  3. possibly some β thalassaemias obscured by severe iron deficiency anaemia.
  4. β thalassaemia carriers with a co-existing δ chain mutation that is silent with the first line screening technique or who have co existing δ thalassaemia.
  5. β thalassaemia carriers with co-existing Hb H Disease, as some cases have normal Hb A2 values.

The aim of this study was to assess the validity and positive predictive value of current action values for MCH and Hb A2 in the detection of β thalassaemia carriers. This was achieved by reviewing the prevalence of β thalassaemia in samples that would currently fall into the ‘possible β thalassaemia carrier’ category and those with an Hb A2 ≥3·5% but less than 4·0% and an MCH that falls above the action value of <27 pg.

Materials and methods

  1. Top of page
  2. Summary
  3. Materials and methods
  4. Results
  5. Discussion
  6. Author contributions
  7. Conflict of interest disclosure
  8. References

All samples referred to the Guy's and St Thomas' Hospital laboratory for FBC and HPLC between 1 July 2008 and 31 May 2011 were included in the study. Samples originated from a variety of sources including antenatal screening samples, in-hospital and pre-operative screens and haemoglobinopathy clinic referrals. Peripheral venous whole blood samples collected in ethylenediaminetetraacetic acid (EDTA) anticoagulant were analysed for FBC using the automated cell counter Coulter® LH 750 System (Beckman Coulter, High Wycombe, UK). The haemoglobinopathy screen was performed by HPLC using a Variant II operating with Hb A2/Hb A1c Dual Program Kit (Bio-rad Laboratories Ltd, Hemel Hempstead, UK).

Molecular analysis

Genomic deoxyribonucleic acid (gDNA) was extracted from the nuclear material of leucocytes using the QIAcube whole blood kit by QIAGEN® (Manchester, UK). HBB (β-globin gene) sequencing and multiplex polymerase chain reaction for the common alpha thalassaemia deletions (α3·7 kb, α4·2 kb, α20·5 kb, MED, SEA, FIL, THAI) and triplicated HBA (α-globin gene) were carried out using methods adapted from previously described procedures (Fodor & Eng, 1999; Mai et al, 2004) and were performed on all samples where:

  1. MCH <27 pg and Hb A2 3·5–3·9% who would all be referred for testing of baby's father, to determine the positive predictive value of these action limits
  2. MCH ≥27 pg and Hb A2 4–4·3% who would all be referred for testing of baby's father, to determine the positive predictive value of these action limits
  3. MCH ≥27 pg and Hb A2 3·5–3·9% who would currently not be referred for testing of baby's father, to determine the number of false negatives in this category

Results

  1. Top of page
  2. Summary
  3. Materials and methods
  4. Results
  5. Discussion
  6. Author contributions
  7. Conflict of interest disclosure
  8. References

A total of 59 500 samples were processed for HPLC analysis over the 34 months; 242 samples (0·41%) fell within the investigation criteria and were referred for genetic analysis. The ethnicity of the majority of samples was not stated at the time of testing. 132/242 samples originated from antenatal clinics.

Within the categories we examined:

  1. 27 samples (0·05% of total samples) with MCH<27 pg and Hb A2 3·5–3·9%
  2. 13 samples (0·02% of total samples) with MCH >27 pg and Hb A2 4–4·3%
  3. 202 samples (0·34% of total samples) with MCH >27 pg and Hb A2 3·5–3·9%

Of the category A samples, 12/27 had mutations associated with β thalassaemia, representing 44% of the samples tested in this category and 0·02% of the total samples analysed during the study period (Table 1). Thus, 15/27 samples (56%) in this category had no β thalassaemia mutations detected; 2 had β chain variants detected, Hb Yaounde and Hb La Desirade (these Hb variants have no reported clinical significance), both had concomitant α thalassaemia, 6 had α+ thalassaemia and 7 had no mutation detected.

Table 1. Category A: red cell indices, Hb A2% and beta gene mutations for positive cases.
RBC (×1012/l)Hb (g/l)MCV (fl)MCH (pg)RDWHb A2 (%)HBB mutation detected (all heterozygous)
  1. RBC, red blood cells; Hb, haemoglobin; MCV, mean corpuscular volume; MCH, mean corpuscular haemoglobin; RDW, red cell distribution width; HBB, β-globin gene.

4·42996922·318·53·7c.93−21G>A Beta plus
5·841166519·917·33·8c.92+5G>C Beta plus (severe)
5·991246620·615·93·8c.92+5G>C Beta plus (severe)
5·671186620·816·93·8c.92+5G>C Beta plus (severe)
5·221116621·314·73·8c.92+6T>C Portuguese Beta plus
4·34956721·815·73·6c.92+6T>C Portuguese Beta plus
4·911096822·216·93·9c.92+6T>C Portuguese Beta plus
4·491067223·7153·9c.92+6T>C Portuguese Beta plus
4·461187926·315·73·9c.92+6T>C Portuguese Beta plus
4·641147424·614·33·5c.*+112A>G Beta Plus
5·21167022·416·43·6c.*+110T>C Beta Plus
4·831217825163·7c.*+110T>C Beta Plus

For the category B samples, 3/13 (23%) had mutations associated with β thalassaemia detected (Table 2). Of the remaining; 1 was found to have the β chain variant Hb La Desirade and the remainder of samples (9/13) had no mutations detected.

Table 2. Category B: red cell indices, Hb A2% and beta gene mutations for positive cases.
RBC (×1012/l)Hb (g/l)MCV (fl)MCH (pg)RDWHb A2 (%)HBB mutation detected (all heterozygous)
  1. RBC, red blood cells; Hb, haemoglobin; MCV, mean corpuscular volume; MCH, mean corpuscular haemoglobin; RDW, red cell distribution width; HBB, β-globin gene.

5·411568328·913·64·3c.−151C>T Beta plus (silent)
4·471248127·712·14c.−151C>T Beta plus (silent)
51438628·3134c.−151C>T Beta plus (silent)

Two of 202 Category C samples had mutations associated with β thalassaemia detected (1%), 2/202 (1%) had unreported mutations that occurred within the IVS-II splice site, 3/202 (1·5%) were found to have previously unreported mutations of uncertain significance and 1/202 (0·5%) had a mutation that is reported as probably non pathogenic (Table 3). A further 6 (3%) were shown to have the 3·7 kb alpha triplication, (Table 4) and the remaining 192 had no mutations detected.

Table 3. Category C: red cell indices, Hb A2% and beta gene mutations for positive cases.
RBC (×1012/l)Hb (g/l)MCV (fl)MCH (pg)RDWHb A2 (%)HBB mutation detected (all heterozygous)
  1. RBC, red blood cells; Hb, haemoglobin; MCV, mean corpuscular volume; MCH, mean corpuscular haemoglobin; RDW, red cell distribution width; HBB, β-globin gene.

  2. a
5·211448427·713·23·5c.*−151C>T Beta plus (silent)
5·031438628·814·83·9c.*+112A>G Beta Plus
4·361218427·715·13·6c.316−12T>C (unreported splice site mutation)
4·461288628·7143·6c.316−15A>G (unreported splice site mutation)
3·761088428·814·93·7c.*+74A>G (unreported, uncertain significance)
4·011198829·615·13·6c.*+74A>G (unreported, uncertain significance)
3·821159030·113·43·6c.93−23T>C (unreported, uncertain significance)
3·721189131·612·93·8c.*+96T>C (reported as probably non pathogenica)
Table 4. Category C red cell indices and Hb A2% for triplicated HBA cases.
RBC (×1012/l)Hb (g/l)MCV (fl)MCH (pg)RDWHb A2 (%)Mutation detected (all heterozygous)
  1. RBC, red blood cells; Hb, haemoglobin; MCV, mean corpuscular volume; MCH, mean corpuscular haemoglobin; RDW, red cell distribution width.

3·361169432·713·63·6Triplicated alpha 3·7 kb
4·09123903012·43·5Triplicated alpha 3·7 kb
3·741168931·112·63·7Triplicated alpha 3·7 kb
3·991249131·114·73·7Triplicated alpha 3·7 kb
4·611449131·212·93·5Triplicated alpha 3·7 kb
4·27115842715·23·7Triplicated alpha 3·7 kb

Discussion

  1. Top of page
  2. Summary
  3. Materials and methods
  4. Results
  5. Discussion
  6. Author contributions
  7. Conflict of interest disclosure
  8. References

This study reviewed the validity of current action limits for laboratory antenatal screening algorithms for β thalassaemia carriers in England. We looked at the positive predictive value of two groups of patients: MCH <27 pg and Hb A2 3·5–3·9% and MCH >27 pg and Hb A2 4–4·3%. These categories made up a very small proportion of our total samples at 27 (0·05%) and 13 (0·02%), respectively. In the former category, 12/27 were correctly identified with β thalassaemia mutations detected and the remaining 15 (0·025% of total samples) can be considered false positives. In the latter category 3/13 had β thalassaemia mutations detected, the remaining 10 (0·017% of total samples) can be considered false positives. Haemoglobin variants known to affect the level of Hb A2 were detected in both of these categories. If we continue with the current screening algorithm there are small numbers of false positive samples but if these were excluded from the screening algorithm there would be similar numbers of patients who were β thalassaemia carriers who would be missed. In view of the very small numbers of patients falling into this category and risk of missing patients with significant β thalassaemia mutations the current screening algorithm is valid and should continue to be used.

This study also reviewed the risks of not referring patients with MCH >27 pg and Hb A2 3·5–3·9% for further investigation to determine if this strategy led to patients with β thalassaemia mutations being missed. There were more samples within this category although it still only made up 0·34% of the total samples. Of these, only 2/202 (0·003% of total samples) had known β thalassaemia mutations detected; when the triplicated alpha samples are included this rises to 8/202 (0·014% of total). Triplicated alpha genes are common and although moderately increased Hb A2 levels have been reported by some authors, this is not consistent and there is little effect on the β thalassaemia carrier phenotype (Giordano et al, 2009; Stephens et al, 2011). These mutations are covered within the screening algorithm risk assessment and these findings support continuation with the current algorithm.

We have not investigated samples with Hb A2 <3·5% and an MCH <27 pg: investigation of samples with an Hb A2 value between 3·1 and 3·5% would be of interest to ascertain the magnitude of ‘missed’ samples, when the algorithm is applied. Likewise, as our extended testing did not include Hb A2 <3·5%, we cannot comment on the section of algorithm pertaining to screening for α thalassaemia (i.e., MCH <25 pg). We have also not investigated samples that are considered to have a high positive predictive value (MCH<27 pg and Hb A2 ≥4·0%), which occur at an incidence of approximately 1·5% in our population.

Further study limitations include a lack of detailed history or clinical details as the samples were routine referrals into the department, therefore no account has been made of previous blood transfusions or factors which may have influenced the Hb A2 level, such as HIV or vitamin B12/folate status. The service supports an inner London Hospital with a diverse ethnic mix that is predominantly African and African-Caribbean and it is likely that different results would be obtained were the patients largely of Asian or Mediterranean origin. The African-Caribbean ethnicity of this population had most impact on Category A samples, where 8/27of the false positive results had an α thalassaemia mutation. This may be less of an issue in a more ethnically mixed population. Finally, we have not considered investigation for other causes of a raised Hb A2 such as KLF1 mutations (Perseu et al, 2011) and, in view of the variation of Hb A2 results with different HPLC instruments, accept that a similar study using a different instrument may have had a slightly different outcome (Batterbee et al, 2010).

In conclusion, with the HPLC instrument and reagents reported, this study validates the programme algorithm for use in such populations but also demonstrates the fact that false positives and negatives occur, with 25 (0·042%) and 8 (0·014%), respectively, being detected. In comparison with other programmes, e.g., breast or cervical screening (Karnon et al, 2004; Hofvind et al, 2012), these rates are relatively low, confirming the high performance of this screening programme.

Author contributions

  1. Top of page
  2. Summary
  3. Materials and methods
  4. Results
  5. Discussion
  6. Author contributions
  7. Conflict of interest disclosure
  8. References

YD and JH contributed to conception and design, data analysis and interpretation, manuscript writing and final approval of the manuscript. RC contributed to collection and assembly of data, data analysis and interpretation, manuscript writing and final approval of the manuscript, KR contributed to data analysis and interpretation and final approval of the manuscript, AS contributed to conception and design, manuscript writing and final approval of the manuscript.

Conflict of interest disclosure

  1. Top of page
  2. Summary
  3. Materials and methods
  4. Results
  5. Discussion
  6. Author contributions
  7. Conflict of interest disclosure
  8. References

JH has done consultative work for Sangart. All other authors have no potential conflicts of interest to disclose.

References

  1. Top of page
  2. Summary
  3. Materials and methods
  4. Results
  5. Discussion
  6. Author contributions
  7. Conflict of interest disclosure
  8. References
  • Batterbee, H., De la Salle, B., McTaggart, P., Dove, C., Wild, B. & Hyde, K. (2010) Evaluation of UK NEQAS (H) HbA2 and Related Performance Data. UK NEQAS, Sheffield, UK.
  • Fodor, F.H. & Eng, C.M. (1999) Molecular Exclusion of haemoglobin SD disease by prenatal diagnosis. Prenatal Diagnosis, 19, 5860.
  • Giordano, P.C., Bakker-Verwij, M. & Harteveld, C. (2009) Frequency of a-globin gene triplications and their interactions with B-thalassaemia mutations. Hemoglobin, 33, 124131.
  • Hofvind, S., Ponti, A., Patnick, J., Ascunce, N., Njor, S., Broeders, M., Giordano, L., Frigerio, A. & Törnberg, S.; EUNICE Project and Euroscreen Working Groups., Van Hal, G., Martens, P., Májek, O., Danes, J., von Euler-Chelpin, M., Aasmaa, A., Anttila, A., Becker, N., Péntek, Z., Budai, A., Mádai, S., Fitzpatrick, P., Mooney, T., Zappa, M., Ventura, L., Scharpantgen, A., Hofvind, S., Seroczynski, P., Morais, A., Rodrigues, V., Bento, M.J., Gomes de Carvalho, J., Natal, C., Prieto, M., Sánchez-Contador Escudero, C., Zubizarreta Alberti, R., Fernández Llanes, S.B., Ascunce, N., Ederra Sanza, M., Sarriugarte Irigoien, G., Salas Trejo, D., Ibáñez Cabanell, J., Wiege, M., Ohlsson, G., Törnberg, S., Korzeniewska, M., de Wolf, C., Fracheboud, J., Patnick, J., Lancucki, L., Ducarroz, S. & Suonio, E. (2012) False-positive results in mammographic screening for breast cancer in Europe: a literature review and survey of service screening programmes. Journal of Medical Screening, 19, 5766.
  • Karnon, J., Peters, J., Platt, J., Chilcott, J., McGoogan, E. & Brewer, N. (2004) Liquid-based cytology in cervical screening: an updated rapid and systematic review and economic analysis. Health Technology Assessment, 8, 20.
  • Mai, M., Hoyer, J.D. & McClure, R.F. (2004) Use of multiple displacement amplification to amplify genomic DNA before sequencing of the δ and β haemoglobin genes. Journal of Clinical Pathology, 57, 637640.
  • NHS Sickle Cell & Thalassaemia Screening Programme (2012) Handbook for Laboratories, 3rd edn. Available at: http://sct.screening.nhs.uk/getdata.php?id=10756 (accessed November 19, 2013).
  • NHS Sickle Cell & Thalassaemia Screening Programme (2013) Data Report 2011/12, Trends and perfomance analysis. Available at: http://sct.screening.nhs.uk/getdata.php?id=12725 (accessed March 11, 2014).
  • Perseu, L., Satta, S., Moi, P., Demartis, F.R., Manunza, L., Sollaino, M.C., Barella, S., Cao, A. & Galanello, R. (2011) KLF1 gene mutations cause borderline Hb A2. Blood, 118, 44544458.
  • Stephens, A.D., Angastiniotis, M., Baysal, E., Chan, V., Fucharoen, S., Giordano, P.C., Hoyer, J.D., Mosca, A.D., Wild, B. & on behalf of the International Council for the Standardisation of Haematology (ICSH). (2011) ICSH recommendations for the measurement of Haemoglobin A2. International Journal of Laboratory Hematology, 34, 113.