The Global Burden of Disease (GBD) study reported 55.9 million deaths in 2017, of which 834,700 were due to various forms of blood cancer and 104,600 due to haemoglobinopathies and haemolytic anaemias [7]. A separate report from 2016 reported a global prevalence of the major forms of blood cancer of almost 4 million people, a number dwarfed by the prevalence of all forms of haemoglobinopathy and haemolytic anaemia of around 2 billion people [8]. Thus, although the number of deaths due to blood cancer markedly exceeds that due to haemoglobinopathies and haemolytic anaemias by a factor of around 8, the prevalence of haemoglobinopathies and haemolytic anaemias exceeds that of blood cancer by some 500-fold.

These data, focussing on mortality, fails to address the morbid aspects of these conditions, which can be expressed as years lived with disability (YLD). The GBD report on this topic [8] told of some 460,000 YLDs in 2016 due to the major forms of blood cancer, data which is also dwarfed by the 6,572 million YLDs linked to haemoglobinopathies and haemolytic anaemias, an increase of some 14-fold, although that the lower YLD of blood cancer may be due to its higher mortality. Indeed, in 2011, the World Health Organisation estimated that around 7% of the world’s population to be carriers of a potentially pathological haemoglobin (Hb) gene, whilst more recently it has been estimated that 3 million people are living with sickle cell disease (SCD) alone, with over 300,000 affected births each year [9, 10]. But just as there are numerous variants of blood cancer and haemoglobinopathy, there are also many other haematological diseases with a genetic component, several with an impact into other forms of anaemia and haemostasis.

Of the hundreds of non-malignant haematological diseases, with the exception of a few, hard data on mortality and morbidity are scarce. The GBD study reported deaths in 2017 due to the most common of these: 7,200 to the thalassaemias, 38,400 to sickle cell disorders, 16,700 to glucose-6-phosphate dehydrogenase (G6PD) deficiency, and 42,200 to other haemoglobinopathies and haemolytic anaemias [6]. G6PD deficiency was recognised as a leading public health issue in India in 1966 [11]. The GBD also provides data regarding the prevalence and YLDs of these conditions (Table 1), which point to great variation in the extent to which each of the conditions bring disability. For example, crudely dividing the YLDs by the prevalence of each condition brings figure of between 0.01 to 0.08 for the haemoglobinopathies (smaller in each of the traits), but of the order or 1 × 10⁻⁶ for G6PD deficiency and 7.8 × 10⁻⁵ for its trait. This may be interpreted as a reduced impact of G6PD in terms of YLDs per case and so point to the haemoglobinopathies bringing most of the burden of disease in this setting.

Perhaps the first reports of molecular genetics to dissect the genome with a view to molecular pathology was in the haemoglobinopathies, with the development of techniques such as the generation of globin gene cDNA from reticulocyte mRNA, and its subsequent insertion into bacterial plasmids [11–16]. The haemoglobinopathies may be classified with relative ease as the thalassaemias, the sickling disorders, and their hybrids.

During the 1960’s and early 1970’s, the use of Hb electrophoresis was the gold standard techniques for the study of abnormal globin chains, with advances in cell biology showing aberrant mRNA activity in thalassaemia. By mid-1970s the use of restriction nucleases and Southern blotting was consolidating the precise aetiology of the haemoglobinopathies at the level of the DNA [14–16]. These advances included reports of undetectable α-globin mRNA and deletions in the gene for α-globin as the cause of α-thalassaemia, and a case report of β -thalssaemia [17–19]. The latter is of historical value as it charts the use of specific techniques, such as the purification of reticulocyte mRNA by an oligo-thymine column, the use of this mRNA to generate cDNA using an RNA-dependent DNA-polymerase from the avian myeloblastosis virus, hybridisation of cDNA to ‘native’ DNA, and use of an in vitro cell-free translation model where globins are generated from purified mRNAs.

Of the several important proofs of concept were that, in contrast to normal mRNA, that from the affected patient generated α- and γ-globin but no β-globin, and their cDNA failed to anneal to normal β-globin DNA. A subsequent report identified three types of abnormal mRNA in beta°-thalassaemia–one with no detectable message, a second whose mRNA hybridised incompletely with cDNA, and a third that had intact mRNA that failed to translate, whilst another report pointed out that clinically severe and mild thalassaemias were linked to the deletion of both or a single α-globin gene, respectively [20, 21]. Orkin et al. used extensive restriction endonuclease digestion of DNA from patients with β^°-thalassaemia, with fragments hybridised to β-globin cDNA, to demonstrate various deletions of the β-globin gene, some partial, some complete [22]. Other studies followed, showing a functioning ζ (zeta) gene in α-thalassaemia, and further studies of α-globin gene deletion [23, 24]. To these were added demonstration of the deletions of three of the four α-globin genes in HbH disease, and of all four genes in Hb Bart’s foetal hydrops, [25–27]. The latter, and others [28, 29] also examined the role of mRNA in these conditions, a process made relatively easy by harvesting mRNA from reticulocytes and other cells [13–15, 20]. Baralle and colleagues [30] used both Sanger and Maxam/Gilbert methods to sequence a 2,000-nucleotide section upstream of the ε-globin gene, a region now known to include a regulatory promotor region. Work of this nature was fundamental in that it subsequently pointed to a further genetic lesion, i.e., defects in a region outside an apparently normal functional gene (its promotor) could lead to a haemoglobinopathy [31].

A key point as regards many of these reports is that gene analysis was supported by protein electrophoresis, high-performance liquid chromatography (HPLC) and, in some cases, studies of the synthesis of globin chains, thus unequivocally defining the molecular basis of these diseases. A good example of this is the elucidation of the genetic basis of Hb Lepore, whose abnormal protein was demonstrated by starch block electrophoresis in 1958 [32], and whose genetic basis was confirmed 20 years later with the demonstration of a fusion of the β-globin gene and the δ-globin gene [33]. By the end of the 1970s, the mapping of the various isogenes of Hb, and their pathology, was becoming established and their locations defined: chromosome 16 for the α-globin gene locus (two copies: HBA1 and HBA2) and chromosome 11 for the β-globin gene (HBB) and the γ-globin gene locus [12, 34–37], locations subsequently further defined as 16p13.3 and 11p15.4 respectively [38, 39].

A key methodological aspect of the work thus described was the used of section-specific restriction endonucleases to determine the presence or absence of particular sections of DNA and so produce a genetic map of a particular locus [40–42], but by the mid 1980s, nucleotide chain sequencing was becoming established. Sanger’s method was used to demonstrate the precise nucleotide sequence of a section of an intron in the β-globin gene of a G to T transversion [43], a T to G transversion [44] and an A to G transition [45], all causing a β-thalassemia. The Maxam and Gilbert method was used to report a case of α-thalassaemia with single base mutation in an α-globin gene at codon 116, that being GAG to UAG that formed a premature termination [46], whilst other part-sequenced the ε-globin gene [47]. We now describe these mutations as a single nucleotide polymorphism (SNP) [48].

A further initiative developed in the 1980s was the ability to use molecular genetics in the pre-natal diagnosis of hemoglobinopathies, initially using restriction endonucleases, then by lesion-specific oligonucleotide probes [49–54]. However, there remained a place for HPLC in determining the presence of abnormal Hbs in an antenatal setting [55–57]. At the end of the decade, Higgs and colleagues [58], and Cao and Murro [59], reviewed the genetics of the α-globin and β-globin gene loci respectively, summarising the various forms of abnormalities deletions, insertions, and SNPs causing different phenotypes of thalassaemia and other diseases of the α-gene and β-gene complex. Fuelled by work on transgenic mice, the 1990s saw molecular genetics used to identify the mechanism for the class switches of genes controlling the Hb isotypes from embryo to adult [60], the so-called “locus control region” (LCR). On the β-globin gene locus, this region is some 34 kB upstream (5′) of the ε gene [61], whilst others provided evidence of the involvement of the Erythroid Kruppel-like factor (subsequently re-named KLF-1, coded for by KLF at 19p13.13) in LRC regulation, binding a CACCC box in promotor regions, other important regions being a TATA box [62, 63]. A large deletion in a section of the LCR results in a form of gamma-delta-beta thalassaemia [64].

As technologies developed in the 2000’s, principally those of NGS analysis [65, 66], additional insights into the genetics of the thalassaemias were reported. These included α-gene duplication, defects in β-thalassaemia, α-gene quadruplication, enhanced prenatal diagnosis of β-thalassaemia, and population screening for both forms of thalassaemia [67–70]. A further advance was the ability to detect foetal nucleic acid in maternal blood, a technique used, alongside amplification refractory mutation system (ARMS) analysis and pre-implantation genetic diagnosis [71–74]. In Sardina, with a population of 1.7 million people, the carrier rate for β-thalassaemia is around 12%, such that 1 in 250 newborns is affected by this disease. Methods such as ARMS and other platforms found 25% of foetuses to be normal, 50% to be healthy carriers and 25% to be affected by the disease: over 98% of the latter pregnancies were terminated [75].

The ability to identify and quantify copy number variation (CNV) by quantitative RT-PCR can be improved by droplet digital PCR [76, 77], whilst whole-exome sequencing can be used to identify aberrant gene activity in thalassaemia [78, 79]. Described in 2010, third generation sequencing includes methods focussing on single-molecule sequencing technologies [80–82]. Examples of these methods include the Pac-Bio method, nanoball technology, and the Oxford Nanopore system, the latter used to detect rare and complex thalassaemia variants, and α-thalassaemia in preimplantation genetic testing [83–87]. These technologies have the ability to overcome the current shortcomings of short-read NGS for exon-level CNV of the globin genes, and provide a single test solution for combined analysis of single nucleotide variants and CNV.

Whilst bone marrow transplantation is an established option for treatment, molecular pathology may provide the next step [4, 5, 88, 89]. The discovery and development of molecular genetics that came to be known as clustered regularly interspaced short palindromic repeats (CRISPR) [90] that, linked with Cas-9 endonuclease [91], has promised much, winning the 2020 Nobel Prize for Doudna and Charpentier with hope for therapeutic genome engineering [92]. One of the first practical uses of this technology was to correct a mutated β-globin gene in an induced pluripotent stem cell line in vitro [93], a process repeated by many others, giving proof of concept, with Sanger sequencing showing a corrected HBB [94]. Originally reported in respect of lymphoid malignancies [95], a genome-wide association study of 4,305 Sardinians (a population at high risk of haemoglobinopathy) [96], showed that BCL11A (coded at 2p16.1) is strongly linked to levels of HbF [97], and that this was linked to SNPs in a 3 kB section of the second intron [98]. These and other data proposed the hypothesis that manipulation of this gene, normally minimally active in the healthy adult, may lead to an increase in red cell HbF, a desired treatment for many haemoglobinopathies [96–98].

The revolutionary paper supporting this hypothesis, reported a patient with transfusion-dependent β-thalassaemia, and another with sickle cell disease who were ‘transplanted’ with autologous CRISPR-Cas9–edited CD34⁺ stem cells that were engineered to reactivate the production of foetal Hb by targeting BCL11A [99]. Over a year later, Hb levels had risen from 90 to 141 g/L, and from 72 to 120 g/L respectively, with parallel improvements in quality of life. A similar clinical study reported editing the BCL11A enhancer to increase Hb levels from 82 to 150 g/L, and from 108 to 140 g/L in two patients with transfusion-dependent β-thalassaemia [100]. Although a successful strategy, a meta-analysis suggested that the promotor regions of gamma globin genes HBG1/2 may be a more efficient target than BCL11A [101]. Alternative uses of this technology include the silencing of genes for various red blood cell antigens, so creating red cells with considerable transfusion possibilities [102].

The precise lesions in the Hb molecules in the sickling disorders has long been known, and the electrophoresis technique developed in the 1950s was extended to further refine knowledge of the amino acid nature of the sickling condition during the 1960s [108–110]. In the 1970s this was extended to the different globin species and their roles in various haemoglobinopathies [111], enabling putative maps of the globin amino-acid chains [112]. As with the thalassaemias [17–20], the 1970s also saw the genesis of the molecular pathology of sickling disorders, primarily with restriction endonucleases [113–115]. Perhaps the first major use of molecular genetics was made by Marotta and colleagues [116] in the definition of the lesion that causes SCD. They purified mRNA from sickle cell reticulocytes using an oligo(dt)-column, and from this used avian myeloblastosis virus RNA-dependent DNA polymerase to generate cDNA that was subsequently phosphorylated. Restriction endonucleases digested this cDNA, and fragments were sequenced by the method of Maxam and Gilbert, showing the HbS to be due to the result of a single nucleotide base mutation in the β-globin gene that converts the glutamic acid codon (GAG) at amino acid position 6 to one for valine (GTG) [117]. A further advance being the detection of β-globin fragments in cells from amniotic fluid that were consistent with sickle-cell trait [118, 119].

As the 1980s proceeded, the work with the use of restriction endonucleases regarding achieving gene maps of the α-globin and β-globin loci [33, 58, 120, 121] was being confirmed and extended by direct genetic analysis, with developments such as isotopic and non-isotopic probes, and their use (alongside PCR) in prenatal diagnosis [122–125]. Subsequent reports described in situ dot hybridisation, amplification by PCR of a 725 bp section of the β-globin gene with the sickle-cell anaemia point mutation in formalin-fixed bone marrow samples [126, 127], and allele-specific PCR, used to analyse the lesion in HbC [128, 129]. A 1992 review summarised the potential for using molecular genetics to insert functioning α-globin and β-globin genes into haemopoietic stem cells via a viral vector [130]. This period also saw the development of PCR heteroduplex analysis, used to screen for HbS and HbC, and the use of amplification refractory mutation system to detect HbA/HbS chimerism after bone marrow transplantation for sickle cell anaemia [131–133].

Developments in cell biology and transfection were used in numerous studies of the basic biology of the sickle phenotype and genotype, with prototype gene therapy. Tang et al. [134] showed that the promoter sequences of the δ-globin gene differs from the β-globin gene by the absence of an erythroid Krüppel-like factor (EKLF) binding site and alteration of the CCAAT box to CCACC, and that restoration of this site can increase δ-globin gene expression, a strategy that has potential future clinical benefit. Knowledge of the sequence of the promotor region of the γ-globin gene enabled Graslund and colleagues [135] to insert an artificial transcription factor for this gene into a cell line, which resulted in an increased expression of α₂γ₂ HbF, and so a strategy for increasing levels of this Hb variant in SCD, a leading therapeutic option [136]. Ho and colleagues [137] generated recombinant mutants of sickle Hb in E coli to show that the formation of HbS polymerisation requires interaction between α-globin 114Pro and β-globin 87THr, and that a mutant 114Arg inhibits the polymerisation, whilst Cole-Strauss et al. [138] inserted a chimeric DNA/RNA to lymphoblastoid cells in a direct correction of the mutation in the Hb betaS allele.

A step closer to clinical application was taken by Pawlik and colleagues [139], who corrected SCD in a mouse model by inserting a lentivirus vector carrying a beta globin gene variant that prevents HbS polymerisation into haemopoietic stems cells, which was then passed into the mouse. They subsequently transduced human cord blood cells with the vector, and transplanted them into mice, using PCR to demonstrate integration of the human gene, with increased levels of the modified β-globin [140, 141]. Small-interfering RNA represents a new opportunity to treat various diseases [142], an example being use of a small interfering antisense RNA to silence a β^s-globin gene when transfected into HeLa cells and an erythroleukaemic cell line [143]. Vasavda and colleagues [144] used RTqPCT to measure circulating HBB DNA in the plasma of patients with SCD, HbS/β^o thalassaemia, HbSC and HbAA controls, finding no difference in steady state disease, but increased levels (alongside CRP and the white blood cell count) when in crisis.

A long-addressed strategy in SCD is to increase levels of HbF, and so ameliorate the clinical effects of the mutation, as does hydroxyurea, a drug that reduces levels of plasma HBB [145]. Using a variety of techniques that included mass spectrometry, RT- PCR and a custom-made Illumina platform, Sebastiani and colleagues [146] reported links between HbF and SNPs in TOX at 8q21.1, in GPM6B at Xp22.2, and at 15q22 (a region that includes AQP9, MAP2K1, SMAD3, and SMAD6), suggesting manipulation of these gene may influence levels of HbF. As in thalassaemia [99–101], BCL11A may be important in SCD. Zhou et al. [147] showed that knockdown of KLF1 in adult erythroid stem cells markedly reduces BLC11A levels and increases the ratio of γ-globin to β-globin. Uda et al. [96] used genome-wide association studies to show that BCL11A is associated with HbF levels in sickle cell patients (as it is in β-thalassaemia), whilst Ghedira and colleagues [148] used quantitative multiplex PCR to estimate that the loss of the BCL11A binding domain leads to an increase in HbF of 27 g/L. Chen and colleagues [149] used an electrophoretic mobility shift assay and other methods to show links between a SNP in a GATA-1 binding motif and high levels of HbF. Other genes linked to HbF include OR51B5 and OR51B6, both coded at 11p15.4, HBS1L-MYB at 6q23, and SAR1A at 10q22.1 [150–152].

Both these strategies came to clinical fruition in the decade that followed, fuelled by the slow and steady increase in knowledge of practicalities of gene therapy, such as inserting genes into cells [153–160] (Table 3). A comparative study concluded that BCL11A is the most clinically relevant approach for CRISR/Cas9 insertions into stem cells [161]. In 2017, Ribeil and colleagues [162] published a case report of a thirteen-year-old boy with homozygous SCD who was transplanted with his own CD34⁺ stem cells that had been engineered with a Lentivirus vector to carry the HBB variant β^A-T87Q, a molecule that inhibits the sickling process. He was discharged on day 50, and after 15 months, the patient’s total Hb increased from 101 to 118 g/L with a fall in reticulocytes from 238,000 to 141,000 per mm³, with no recurrence of pathological sickle crises. In 2021, Esrick and colleagues [163] reported six patients with SCD who were transplanted with autologous CD34⁺ stem cells transduced with a lentivirus vector carrying a short-hairpin RNA targeting BCL11A. HbF induction was robust and stable, with all indices of HbF improved (e.g., HbF rising from 9.0 to 18.6 pg/cell) with a reduction or absence of clinical manifestation of the disease.

As descrived previously, the same transplant strategy was used by Frangoul et al. [99] to treat a 19-year-old female with the β^o/β+ thalassaemia but also a 33-year-old female with SCD (βS/βS and a single α-globin deletion). This resulted in increases in total Hb from 90 g/L to 142 g/L after 15 months, and from 72 to 120 g/L, whilst the % of red cell expressing HbF increased from 10.1 to 100, and from 33.9% to 98.1%, respectively. Vaso-occlusive events were eliminated in the patient with SCD. The following year, Kanter and colleagues [164] reported the use of a lentivirus vector to insert a gene coding an antisickling engineered variant of HbA into autologous stem cells, subsequently transplanted back into the patient with SCD. The largest study so far (n = 35) of this engineered Hb saw total Hb rise from a median of 85 g/L to over 120 g/L (53% being the engineered Hb) after 36 months, with no severe vaso-occlusive events after transplantation. Levels of serum lactate dehydrogenase and serum indirect bilirubin, and the reticulocyte count, all fell markedly. Magrin and colleagues [165] reported a 36, 42, and 60 months follow up of three SCD patients with the antisickling Hb variant β^A-T87Q, noting marked variation in the % of total Hb being composed of HbF and HbA^A-T87Q of 43.7%, 14.0% and 51.1% respectively. Sharma et al. infused stem cells bearing a CRISPR-Cas9 product engineered to disrupt HBG1 and HBG2 into three patients with severe SCD [166]. This resulted in sustained induction of red cell foetal Hb, that being 19%–26.8% of total Hb, with 69.7%–87.8% of erythrocytes classified as F cells, and a decrease in the manifestations of SCD. Extending their previous work [163], Esrick and colleagues now [167] compared their patients transplanted with BCL11A-silenced stem cells with transplanted patients being treated with standard care hydroxyurea. The % of red cells from transplanted patients containing ≥10 pg of HbF was comparable to that of hydroxyurea high-responders, but superior to that of hydroxyurea low-responders, but the % of cells carrying HbS was lower than both hydroxyurea groups.

Whilst gene therapy represents an alternative to bone marrow transplantation, which although well-established, has its drawbacks [168], the number of treated patients is small and follow-up short, such that full assurance is as yet unforthcoming. Indeed, despite the apparent success of some cases where all three patients in the OTq923 study (targeting BCL11A) [166] showed a clinical improvement, they continued to have mild haemolysis and some symptoms of sickle-cell disease, a finding which may have contributed to the decision by the sponsor to discontinue the development of this product [169]. The treatment cost of $2 million may be balanced by reduced cost of, for example, vaso-occlusive events, acute chest pathology etc. over the hoped-for decades of life remaining and the inestimable cost of improved quality of life [170]. Chapman et al. have pointed out the danger of vector insertion related leukaemias in other diseases from 2003 to 2014 and found an increased frequency of potential driver mutations associated with myeloid neoplasms or clonal haematopoiesis in both genetically modified and unmodified stems cells after SCD transplantation [171]. It remains to be seen if these mutations precipitate other disease and prompts the possible need to screen for these mutations before gene therapy is commenced.

Perhaps the oldest known genetic disease, as described in ancient Hebrew texts, haemophilia is the likely reason for the dispensation of the surgical removal of the foreskin (circumcision) in a neonate if older brothers died following prolonged bleeding after their own circumcision. In the modern era, the disease was clearly described over 200 years ago [207], although not named, and perhaps the oldest detailed clinical description of haemophilia is over 150 years ago [208], with a treatment of the infusion of serum at the beginning of the last century [209].

The gene for Factor VIII (F8) was cloned, sequenced (reporting 186 kB, 25 introns, and 26 exons) and its 2,351 amino acids (giving a mass of 267-kDa) determined in 1984 [210–213]. Shortly afterwards, FVIII-specific probes were used for the prenatal diagnosis of haemophilia, and its chromosomal location determined [214–217]. By the end of the decade, deletions, SNPs, inversions, and insertions in the gene had all been shown to be the cause of haemophilia [218–222], reviewed in [223]. Interestingly, a cluster of mutations, many of which are inversions, in intron 22 found in 40% of one series of haemophiliacs can cause defective joining of exons 22 and 23 in the mRNA and so result in the disease [224].

The closely-related Christmas disease was described in 1952 [225], the identity of the restorative agent (anti-haemophilic factor) was determined in the late 1950s as Factor VIII. Becoming known as haemophilia B to distinguish it from ‘standard’ haemophilia, subsequently known as haemophilia A, the gene coding Factor IX (i.e., FIX) was cloned in 1982, defects shown in 1985 to be the cause of deficiency of the protein [226–228]. Kurachi and Davie [229] used a baboon liver mRNA to synthesise a radiolabelled cDNA 14-mer primer to identify and clone sections of a human liver cDNA library that were inserted into plasmids. Restriction endonuclease digestion of positive clones provided fragments that were sequenced by the method of Maxam and Gilbert. The gene has eight exons transcribing a 2.8 kB RNA that translates to a pre- and pro-peptide and mature protein of 415 amino acids, subsequently cleaved to FIXa with a relative molecular mass of 57 kDa. Standard sequencing reported 55.8% of 1692 abnormalities to be point mutations, 20.8% to be polymorphisms,16.6% to be deletions, and 83.9% to be in exons, the most common being in the eighth (37%) and second (12%) [230]. The power of NGS was demonstrated by the analysis of DNA from 2401 patients with haemophilia A and 599 with haemophilia B, finding 924 unique variants, confirming the inversion in intron 22 to be the leading mutation in severe cases of haemophilia A (42% of cases), followed by frameshift and missense (both 17.4%), whilst missense mutation were the most common in mild and moderate disease (79.5%) [231]. In haemophilia B, the leading mutation was missense in both severe (47.2%) and mild/moderate (87.1%) disease.

The relatively small size of F9 (∼34 kb), normally coding a 2.8 kb mRNA, permits its insertion into a viral vector that can be delivered to hepatocytes that then produce the active zymogen product of some 461 amino acid residues [232]. Furthermore, molecular genetics has provided a further advance in identifying a mutated form (FIX-Padua) that generates a F9 with enhanced enzyme activity compared to normal F9, and which is effective in a clinical setting [233]. This part-contrast with F8, of 186 kb, coding an mRNA of 9 kb and ultimately a 2,300 amino acid product, that together provided a challenge for genetic engineers [234]. However, these have been overcome, and two therapeutics (Valoctocogene Roxaparvovec and Efanesoctocog Alfa) are now available for the treatment of haemophilia A [235].

By 1971 the hereditary nature of dysfibrinogenaemia (as autosomal dominant), afibrinogenemia, hypoprothrombinaemia, dysprothrombinaemia, defects of Stuart-Prower factor (Factor X) (all autosomal recessive) had been described, all pointing to a genetic lesion [236]. The molecular pathology of all other factor deficiencies was subsequently defined, locations of leading coagulation genes, and their population frequencies, are shown in Table 4.

Based on an amino acid sequence of a factor Va light chain, Jenny and colleagues [237] synthesised a 39-mer probe and used it to screen an oligo(dT)-primed human foetal liver library, with clones expressed in phage vectors. Overlapping fragments were sequenced by Sanger technology, reporting a 6672 bp coding region ultimately producing a mature protein of 330 kDa consisting of 2196 amino acids. Subsequent analysis showed the presence of 25 exons, and that removal of a domain from the protein structure generates FVa, formed from a 105 kDa heavy chain and a 71 or 74 kDa light chain [238]. The combined deficiency of FV and FVIII may follow loss-of-function mutations (mostly insertion/deletions and splice site mutations) in LMAN1 at 18q21.32, and MCFD2 at 2p21 that together code for membrane proteins of the endoplasmic reticulum and Golgi apparatus pathway for transporting the coagulation factors [239].

The gene for FVII is located 2.8 kB telomerically to that for FX, spanning 12 kB and with nine exons coding for a mature protein of 406 amino acids of 50 kDa, the leading cause of deficiency being a 10 bp insertion polymorphism, with other abnormalities linked to a variable reduction in biological activity [240]. Fibrinogen is formed from three proteins, coded for by FGA (7.6 kB, 6 exons), FGB (8 kB, 8 exons), both at 4q31.3, and FBG (8.5 kB, 10 exons) at 4q32.1, giving a total mass of 340 kDa: the estimated prevalence of deficiency (generally a deletion, most often in FGA) is around 1/million. Vitamin K dependent coagulation factor deficiency, producing low levels of several of the products of these genes, is linked to GGXC at 2p11.2, coding for gamma-glutamyl carboxylase, and VKORC1 at 16p11.2, coding for a subunit of the vitamin K epoxide reductase complex [241–243].

The most common genetic haemorrhagic disease is von Willebrand disease (VWD), with an estimated Caucasian population frequency of 0.57%–1.15%, translating to 5,700 to 11,500 per million [244], far exceeding that of haemophilia A at 200 per million males. However, clinical VWD runs a spectrum from asymptomatic to life-threatening, where the phenotype resembles haemophilia, such that the referral rate for the most severe cases has been estimated at 23 to 113 per million [245]. The variable nature of the disease, characterised by low levels of von Willebrand factor (vWf), can be accounted for by various mutations in VWF at 12p13.31, where it spans 176 kB, consisting of 52 exons between 52 base pairs and 1.3 kB generating an mRNA around 9 kB long [246, 247].

In contrast to low levels of plasma vWf, increased levels are common in cardiovascular disease, cancer, and in connective tissue diseases, reflecting an endothelial cell pathology, although the driver(s) are unlikely to be genetic [248]. Nevertheless, a leading feature of this molecule is as a chaperone for FVIII, increased levels of which are a risk for a venous thrombosis, with both molecules considered by some to be part of a thrombophilia screening panel [249], although this view is not universal [250]. Genes cited as leading genetic causes of thrombophilia are shown in Table 5, many of which may be part of a multi-target NGS thrombophilia panel for investigating those with, or at risk of, a venous thrombosis [250–255], as summarised in a 2022 guideline [256].

Nurden and Nurden [257] summarised the genetics of the ontogeny of thrombopoiesis in four stages:

THPO at 3q27.1, coding for thrombopoietin, and MPL at 1p34.2 coding the thrombopoietin receptor (loss of function mutations in either causing congenital amegakaryoctic thrombocytopenia), at the level of the haemopoietic stem cell.

The leading forms of genetic disease of platelets involve surface receptors. In Bernard Soulier syndrome (with a prevalence of 1 per million), there is a thrombocytopenia and a qualitative change, the latter caused by failure of any of GP1BA (the most frequently mutated gene), GP1BB, GP9, or GP5 (at 3q29) to correctly code molecules forming the GpIb-IX-V complex, the receptor for vWf. Failure of this receptor/ligand reaction results in platelets unable to form a thrombus, and so haemorrhage [258, 259]. Glanzmann thrombasthenia also has an incidence of 1/million, but in areas of high consanguinity this may increase five-fold. It is caused by mutations in ITGA2B or ITGB3, that together code for integrins GPIIb/IIIa, also known as the fibrinogen receptor [260]. The molecular pathology of the platelet extends other forms of haemorrhagic disease (Table 6) [257, 261–265]. Gebetsberger and colleagues [266] summarised Sanger, NGS panels, WES and WGS approaches to platelet disorders.