Copyright: ©Author(s) 2026.
World J Hematol. Sep 10, 2026; 12(2): 121492
Published online Sep 10, 2026. doi: 10.5315/wjh.121492
Published online Sep 10, 2026. doi: 10.5315/wjh.121492
Table 1 Genomic and structural characteristics of the F8 and F9 genes of their variant databases
| Feature | F8 gene | F9 gene |
| Chromosomal location | Xq28 | Xq27 |
| Gene size | Approximately 186 kb[11] | Approximately 34 kb[11] |
| Number of exons | 26 | 8 |
| Protein encoded | Coagulation factor VIII | Coagulation factor IX |
| Protein structure/domains | A1-a1-A2-a2-B-a3-A3-C1-C2; encoded partly by large exon 14[11] | Signal peptide, propeptide, Gla domain, EGF1, EGF2, activation peptide, serine protease domain[11] |
| Key genomic features | Intron 22 region containing intronic gene copies (F8A and F8B), predisposing to Inv22 (40%-50% of severe HA)[11]; intron 1 low-copy repeats predisposing to Inv1 (2%-5% of severe HA); CpG hotspots accounting for approximately 40% of small variants despite comprising approximately 2% of sequence[11] | Compact gene with modular exon-domain correspondence; promoter contains androgen-responsive elements responsible for haemophilia B Leyden phenotype[11] |
| Predominant mutation classes | Inv22 (40%-50% of severe); Inv1 (2%-5% of severe); nonsense, frameshift, splice-site defects dominate in severe HA; missense variants common in mild/moderate HA | Missense (approximately 68%) dominate; nonsense/frameshift variants associated with severe phenotype; splice-site variants relatively frequent; large deletions rare (1%-3%) but high inhibitor risk |
| Mutation hotspots | Inv22 and Inv1 mediated by homologous recombination | Noncomparable to F8 inversions; mutations distributed across gene |
| Typical diagnostic methods | Long-range PCR for inversion detection; Sanger/NGS sequencing for point mutations, indels; MLPA for deletions/duplications | Direct sequencing (gene small enough for full coverage); MLPA when deletions suspected |
Table 2 Mutational spectrum of hemophilia A (F8 variants) and associated phenotypes
| Mutation type | Approximate distribution | Associated severity | Molecular mechanism |
| Inv22[15-20] | 40%-50% of severe HA globally; lower in some regions (e.g., 10.5% in Albania, approximately 30% in parts of India/Asia) | Severe | Homologous recombination between int22h-1 and extragenic int22h-2/int22h-3 repeats - disrupted F8 transcription |
| Inv1[15-17] | 2%-5% of severe HA | Severe | Homologous recombination within intron 1 |
| Nonsense mutations[14] | Common among severe, inversion-negative HA | Severe | Premature stop codons - truncated nonfunctional FVIII; often triggers NMD |
| Frameshift mutations (small insertions/deletions)[17-19] | Frequent in severe HA; multiple novel variants identified in several populations | Severe | Reading-frame disruption - premature truncation |
| Canonical splice-site mutations[14,17] | Common among severe phenotypes | Severe | Aberrant splicing - exon skipping or truncation |
| Missense mutations[16,17,22] | Predominant in mild and moderate A; smaller contribution to severe HA (especially at conserved residues) | Mild-moderate; occasionally severe | Residue substitution affects FVIII structure, stability, or cofactor function |
| Small deletions/insertions (non-frameshift)[22] | Less common but clinically significant | Mild-severe depending on domain affected | Disruption of local protein domains without full truncation |
| Large deletions/multiexon deletions[21] | Rare | Severe | Loss of entire domains - absent FVIII |
| Composite (double) mutations[18] | Rare | Severe | Combined effects of two pathogenic variants |
Table 3 Mutational spectrum of haemophilia B (F9 variants) and structural consequences
| Mutation type | Approximate distribution | Associated severity | Structural mechanism |
| Missense variants[11,23-25] | Approximately 68% of all F9 variants; the dominant mutation class[11] | Mild-moderate most commonly; sometimes severe when essential residues affected | Alter FIX folding, γ-carboxylation, calcium binding, catalytic triad stability, or activation peptide processing |
| Nonsense mutations[25,28] | Less common than missense; significant proportion in severe HB | Severe | Premature truncation - absent or unstable FIX protein; NMD frequently triggered |
| Frameshift variants (insertions/deletions)[25] | Relatively uncommon but clinically important | Severe | Frameshift - truncated nonfunctional protein |
| Canonical splice-site mutations[25,28] | Common among severe HB due to compact exon-domain architecture | Severe | Aberrant splicing - exon skipping, truncated proteins, or defective post-translational processing |
| Promoter mutations (including HB Leyden)[11] | Rare overall | Childhood severe - spontaneous improvement after puberty | Androgen-responsive elements regulate transcription; puberty FIX expression |
| Synonymous pathogenic variants[27] | Rare but increasingly recognized | Mild-severe depending on impact on mRNA | Alter mRNA structure, translation rate, and co-translational folding (e.g., p.Val107Val) |
| Large deletions (partial or whole-gene)[29] | 1%-3% of F9 mutations | Severe | Loss of entire exons or full gene; often destabilizes neighboring genomic regions |
| Contiguous gene deletion syndromes[29] | Extremely rare | Severe haemophilia + syndromic features | Deletion of F9 plus adjacent genes - multi-system phenotype |
Table 4 Comparison of approved gene therapies for haemophilia A and B
| Feature | Valoctocogene roxaparvovec (roctavian)[46,50,52,53] | Etranacogene dezaparvovec (hemgenix)[47-50] |
| Indication | HA | HB |
| Vector capsid | AAV5 | AAV5 |
| Transgene | B-domain-deleted FVIII, codon-optimized FVIII-SQ | FIX-Padua variant (R338 L), approximately 5-10 × higher specific activity |
| Mechanism of action | Hepatic expression of FVIII-SQ leads to endogenous FVIII production sufficient to convert severe HA to mild/normal range | Hepatic expression of FIX-Padua generates supraphysiologic FIX activity at low vector doses |
| Peak factor levels | Median FVIII 11.9%-62.3% at weeks 49-52 in phase 3 | Approximately 30 IU/mL at 12 months across trials |
| Long-term factor expression | Decline over time typical: Approximately 50%-60% reduction from peak by 24 months | FIX expression more stable than FVIII; long-term persistence observed |
| Reduction in ABR | Meta-analysis: -7.58 treated bleeds/year; > 90% reduction in factor use | Meta-analysis: 5.64-fold ABR reduction; near-universal cessation of prophylaxis |
| Proportion of patients stopping prophylaxis | Majority (> 90%) discontinue FVIII prophylaxis | Majority discontinue FIX prophylaxis; FIX activity sufficient for stable haemostasis |
| Durability and challenges | Expression decline over several years; FVIII synthesis is hepatocyte-stressful (UPR/ER stress) | Durable expression; lower dose requirement improves safety margin |
| Key safety issues | Transaminase elevations common; very high vector dose raises hepatotoxicity concerns potential genomic integration events | ALT elevations possible but generally manageable lower vector dose reduces risk of hepatotoxicity |
| Advantages | First approved gene therapy for HA; major reductions in bleeding and factor use; high initial FVIII expression | High efficiency due to Padua transgene; stable FIX expression; very low dosing requirements |
- Citation: Bolou K, Kapsimali Z, Dettoraki A, Michalopoulou K, Triantafyllou G, Karangeli N, Piagkou M, Pergantou H. Molecular pathogenesis and therapeutic advances in haemophilia, an update of the current evidence. World J Hematol 2026; 12(2): 121492
- URL: https://www.wjgnet.com/2218-6204/full/v12/i2/121492.htm
- DOI: https://dx.doi.org/10.5315/wjh.121492