Diamond–Blackfan Anaemia and Other Ribosomopathies

Diamond-Blackfan anaemia syndrome (DBA): congenital red cell aplasia due to haploinsufficiency of ribosomal proteins, with congenital anomalies, short stature, immunodeficiency and cancer predisposition; it includes Blackfan-Diamond syndrome and the former Aase syndrome, and belongs to the group of ribosomopathies (5q− syndrome, Treacher Collins syndrome, Shwachman-Diamond syndrome, cartilage-hair hypoplasia, isolated congenital asplenia, dyskeratosis congenita and RNA polymerase III leukodystrophies).

Literature and regulatory information updated: 17 October 2026.

0. Initial Reference Table

1. Introduction

1.1. Clinical description and first decades

The disease was described in 1938 by two American paediatricians, Diamond and Blackfan, as a congenital anaemia due to lack of red cell production; the original article was not located in PubMed or Crossref during data collection, so the 1938 description is presented by attribution through Smith's and a recent meta-analysis, which mentions even earlier descriptions from 1936 [3, 16]. Diamond preferred the designation "congenital hypoplastic anaemia" (congenital hypoplastic anemia), used for decades and taken up in a 1976 review by Diamond, Wang and Alter cited by the international consensus [1]. In 1969, Aase and Smith described two siblings with triphalangeal thumbs and congenital anaemia, a picture that became known as Aase syndrome or Aase-Smith syndrome II; the clinical and molecular overlap with DBA led to Aase syndrome no longer being considered a separate entity [2, 3, 17]. Corticosteroids have been used in DBA for more than 70 years and were for a long time the only treatment with an effect on erythropoiesis, alongside transfusions [1].

1.2. Molecular discovery

The first gene was identified in 1999 by Draptchinskaia and colleagues, who cloned the 19q13 breakpoint in a patient with a reciprocal X;19 translocation and found variants of the ribosomal protein S19 gene (RPS19) in 10 of 40 unrelated patients, including nonsense, frameshift, splicing and missense variants and two intragenic deletions [18]. The genes RPS24 (2006) and RPS17 (2007) followed, which the appendix to the 2024 consensus cites in its gene table, and RPL35A in 2008, identified in two patients with 3q deletions [1, 19]. In the same year, Gazda and colleagues described variants of RPL5, RPL11 and RPS7, showed that RPL5 was associated with craniofacial, thumb and cardiac anomalies and RPL11 mainly with isolated thumb malformations, and recorded that, at the time, RPS19, RPS24, RPS17 and RPL35A explained about 30% of cases [20]. RPS10 and RPS26 were described in 2010, RPL26 in 2012, RPS29 in 2014, RPL27 and RPS27 in 2015, RPS15A in 2017, and RPL18 and RPL35 in the 2017 genomic characterisation by the US National Cancer Institute (NCI) [1, 3, 21, 22].

The first non-ribosomal cause was described in 2012: exome sequencing of two siblings without a variant in ribosomal protein (RP) genes revealed a variant at a splicing site of GATA1 that impaired production of the full-length isoform of the transcription factor, and a third patient had another variant at the same site [23]. In 2014, Gripp and colleagues studied six families with DBA and Treacher Collins-type mandibulofacial dysostosis and found probable variants in five: two in RPS26, one X-linked in TSR2 (a direct binding partner of RPS26) and two de novo at the initiation codon of RPS28 [24]. In 2014, Ludwig and colleagues showed that GATA1 mRNA is poorly translated when there is a ribosomal protein deficiency, linking the two groups of genes in a common mechanism [25]. In 2018, exome sequencing of 472 individuals with a clinical diagnosis of DBA found relevant rare variants in 78%, mainly single loss-of-function variants in 19 known RP genes, 31 deletions detected by exon coverage, extended splicing variants, 7 candidate RP genes and phenocopies, including 9 patients with biallelic ADA2 variants (formerly CECR1); no new gene explained more than 5% of cases [8]. Biallelic variants of HEATR3, which impair nuclear import of the proteins uL18 (RPL5) and uL5 (RPL11), were described in 2022, and germline gain-of-function variants of TP53 in 2022 and 2026 [26–28]. A 2020 review summarises the pathophysiology and clinical features of the disease [29].

1.3. Nomenclature and reclassifications

The 2008 consensus distinguished classic and non-classic DBA and used major and minor diagnostic criteria; the 2024 international consensus, drawn up by 53 experts from 27 countries (the European EuroDBA consortium and DBAR) using a modified Delphi method and an agreement threshold above 85%, replaced it and introduced five changes of nomenclature and definition: the designation "DBA syndrome", which includes individuals without anaemia or with atypical presentation; the term "treatment independence" instead of "remission" when counts normalise; simplified diagnostic criteria (one of two criteria, genetic or phenotypic); genetic classification into ribosomopathy, other genes and phenocopies; and the designation "DBA syndrome currently without phenotype" (DBA syndrome currently without phenotype) for carriers of a pathogenic variant without manifestations, instead of "silent carrier", because these people may go on to develop anaemia or cancer [1, 30]. The consensus states that all available evidence is level C (expert opinion, retrospective series and registry data), as there are no randomised trials or large non-randomised studies in the disease [1].

The classification of patients with GATA1 variants differs between sources. The DBA GeneReviews includes GATA1 among the genes of X-linked DBA, and the 2024 consensus places it in the category "other DBA syndrome genes"; the authors of the GeneReviews on GATA1-related cytopenia prefer the designation GATA1-related cytopenia, because many of these DBA diagnoses were made before GATA1 cytopenia was recognised as an entity and because missense variants in the zinc fingers give a different picture (macrothrombocytopenia, dyserythropoiesis and haemolytic anaemia) [1, 2, 6]. The consensus excludes from the nosology the phenocopies with different pathophysiology that present as pure red cell aplasia (PRCA), namely ADA2 deficiency and the homozygous EPO variant [1].

1.4. Definition of the entities within scope

DBA syndrome is defined, according to the 2024 consensus, by the presence of a pathogenic or likely pathogenic variant in a gene of the syndrome, or by a compatible haematological picture (macrocytic anaemia with reticulocytopenia and marrow erythroblastopenia, without dysplasia, dyserythropoiesis or sideroblasts) with exclusion of the differential diagnoses; it covers classic DBA (anaemia in the first year of life, with or without congenital anomalies), mild or late-onset forms and individuals with congenital anomalies, cytopenias or cancer without anaemia [1]. Blackfan-Diamond syndrome and Aase syndrome correspond to the same entity [2, 3].

A ribosomopathy is a disease caused by an alteration of the structure or function of the ribosome or of its biogenesis. The chapter by Ferreira, Darling and Vockley organises the non-mitochondrial ribosomopathies by the step affected: transcription of pre-ribosomal RNA (rRNA) by RNA polymerase I (Treacher Collins syndrome and Cincinnati-type acrofacial dysostosis due to POLR1A); transcription of 5S rRNA and tRNAs by RNA polymerase III (leukodystrophies and syndromes due to POLR3A, POLR3B and POLR1C); processing and modification of pre-rRNA (dyskeratosis congenita due to DKC1 and NOP10, cartilage-hair hypoplasia due to RMRP, and POP1 and NEPRO deficiencies); maturation of the 40S and 60S subunits (DBA, the most frequent phenotype of ribosomal dysfunction); and assembly of the active 80S ribosome (Shwachman-Diamond syndrome) [12]. 5q− syndrome is the acquired, somatic form of ribosomal haploinsufficiency (loss of one copy of RPS14 in the myelodysplastic clone) [2, 31], and isolated congenital asplenia due to RPSA haploinsufficiency is the only known human developmental anomaly confined to a lymphoid organ [32]. These entities are summarised for context and differential diagnosis, as are Treacher Collins syndrome, Shwachman-Diamond syndrome, Fanconi anaemia, TAR syndrome and the telomere diseases.

2. Molecular Genetics and Pathophysiology

2.1. The ribosome and its biogenesis

The eukaryotic ribosome is formed by a small subunit (40S) and a large subunit (60S), which together comprise four RNA species and about 80 ribosomal proteins [2]. RNA polymerase I synthesises the precursor of the large rRNAs, RNA polymerase II the mRNAs and RNA polymerase III the tRNAs and 5S rRNA; each polymerase has at least 12 subunits [12]. Pre-rRNA is processed and modified by complexes that include RNase MRP (whose RNA component is encoded by RMRP) and the small nucleolar ribonucleoproteins of pseudouridylation (with dyskerin, encoded by DKC1, and NOP10); ribosomal proteins are imported into the nucleus, some with the help of chaperones such as HEATR3 (for uL18/RPL5 and uL5/RPL11) and TSR2 (for RPS26), and final maturation of the active 80S requires removal of the factor eIF6 from the pre-60S subunit, a step that fails in Shwachman-Diamond syndrome [1, 12, 26]. Mitochondrial ribosomal proteins are encoded in the nucleus and mitochondrial rRNAs in mitochondrial DNA; the corresponding diseases fall outside the scope [12]. Cytoplasmic ribosomal proteins now have a universal nomenclature (for example, RPS19 = eS19, RPL5 = uL18, RPL11 = uL5, RPL35A = eL33), used alongside the gene symbols [2].

2.2. Mechanism of erythroid failure

The most frequent mechanism is haploinsufficiency of a ribosomal protein; rarely, a missense variant acts with a dominant negative effect [2]. Lack of RPS19 impairs maturation of 18S rRNA in yeast and in human cells, and deficiencies of RPS24 and RPS7 (small subunit) and of RPL35A, RPL5, RPL11 and RPL26 (large subunit) alter processing of the pre-rRNA of the corresponding subunit [2, 20]. The excess of free ribosomal proteins and unused precursors triggers "ribosomal stress", which stabilises p53 and leads to cell-cycle arrest and apoptosis of erythroid progenitors; increased apoptosis has been demonstrated in haematopoietic cell lines and in marrow with RPS19 and RPL35A deficiency, and zebrafish with rps19 disruption show imbalance of the p53 family [2]. A second pathway explains erythroid specificity: GATA1 mRNA has a complex 5′ untranslated region and is translated inefficiently when the number of ribosomes is reduced, which decreases GATA1 protein in erythroid progenitors; this observation explains why germline GATA1 variants produce a DBA phenotype [2, 25]. Slow globin synthesis in ribosome-deficient erythroblasts leaves excess free haem, the rationale for restricting the uptake of glycine (a haem precursor) with bitopertin in preclinical models [33]. Recent models also describe activation of the NLRP1 inflammasome and a lipid metabolism signature (Scd1) in an Rpl5 mouse model, as yet without clinical translation [34, 35]. The marrow, initially normocellular, becomes hypocellular with age, and patients have slow haematological recovery and increased toxicity after chemotherapy or radiotherapy [1].

Gain-of-function TP53 variants directly activate the p53 pathway, unlike the loss-of-function variants of Li-Fraumeni syndrome, and reproduce the haematological phenotype [1, 27, 28]. Biallelic HEATR3 variants reduce nuclear import of uL18 and uL5; in this case, the proliferation defect demonstrated in vitro was independent of p53 [26].

2.3. Genes, proteins and relative frequency

The table brings together the genes recognised by the 2024 consensus (appendix, page 4) and by the 2025 GeneReviews. The GeneReviews frequencies derive from HGMD database data; those of the consensus come from registries. The two sets differ for some genes and are therefore presented side by side [1, 2].

The consensus appendix also lists seven candidate genes requiring functional validation (RPS11, RPL3, RPL10, RPL10A, RPL19, RPL34 and the gene for protein uL10) and two phenocopies (ADA2 and EPO) [1]. For RPL35A and RPL31, the appendix gives the chromosome arms 9q and 12q, which do not match the GeneReviews cytobands 3q29 and 2q11.2; the table follows GeneReviews, consistent with the 3q29 deletions described for RPL35A [2, 36]. The correspondence between numbered subtypes and genes follows the GeneReviews phenotype table and the list of OMIM numbers associated with ORPHA:124 [2, 5].

2.4. Variant types and normalisation of nomenclature

Causal variants are mainly loss-of-function: initiation codon changes, nonsense, small insertions or deletions and splicing variants leading to nonsense-mediated mRNA decay (NMD), and genomic deletions of exons, of the whole gene or of several contiguous genes [1]. Missense variants are less frequent, are concentrated in the RPS genes and may require functional validation (rRNA processing, animal models); a nonsense variant in an RP gene almost always causes DBA [1]. Many variants are private to each family, without hotspots, and there are no allele-by-allele genotype-phenotype correlations of confirmed clinical utility [2, 36]. Deletions carry variable weight by gene: according to GeneReviews, targeted deletion and duplication analysis detects more than 65% of RPS17 variants, about 50% of RPL35A variants, about 20% of RPL15 variants, less than 20% of RPS26 variants, about 10% of RPL11 variants, less than 5% of RPL5 variants and about 5% of RPS19 variants. In the table of laboratory considerations, GeneReviews, citing an Italian series studied by multiplex ligation-dependent probe amplification (MLPA), gives large deletions in about 20% for RPL5, RPL11, RPL35A, RPS17, RPS19 and RPS26, a value that does not match the proportions above gene by gene [2, 37]. A Japanese series described extensive deletions and, in the 2018 study, 31 RP gene deletions were validated by exome coverage [8, 38].

The variants cited were normalised in VariantValidator (GRCh38, MANE Select transcript) and, for EPO and RMRP, in ClinVar; conversions of historical numbering were not done by hand.

The variants RPL15 p.Tyr81* and RPL18 p.Leu51Ser appear in the sources only with protein notation. For RPL18, ClinVar has a single record, NM_000979.4:c.152T>C p.(Leu51Ser), included in the table. For RPL15 (NM_002948.5), p.Tyr81* corresponds in ClinVar to more than one nucleotide change (c.243C>G and c.242dup), so it is kept as the published designation. The GeneReviews notation for the GATA1 initiation codon variants (p.Met1Thr, p.Met1Ile) differs from the current HGVS recommendation, which describes the predicted effect as p.(Met1?), because the resulting protein depends on the reinitiation codon; in the case of GATA1, translation reinitiates at an internal methionine and produces only the short isoform GATA-1s, lacking the first 83 amino acids [6].

2.5. Genotype-phenotype correlations by gene

RPL5 is associated with craniofacial, cardiac and thumb anomalies that are more frequent and severe than with RPL11 and RPS19: cleft lip or palate in 45% of patients with RPL5 in the Gazda series and in 50% of an Italian group, and fetal growth restriction (small for gestational age) in 7 of 8 Czech patients with RPL5 (88%) versus 43% with RPS19 [2, 20, 44, 45]. RPL11 is associated mainly with thumb anomalies [20, 45]. The consensus summarises the established associations: abnormal thumbs with RPL5 and RPL11; cleft lip or palate with RPL5, RPL11 and RPS26; more than one anomaly more frequent with RPL5 and RPL11 than with RPS19; and anomalies in general more frequent in RPL genes than in RPS genes [1]. In the NCI cohort of 121 patients followed for 20 years, with a pathogenic variant in 71% (86 of 121, from 54 families), RPL genes were associated with more intellectual disability and gastrointestinal anomalies than RPS genes (p<3.5×10⁻⁴), RPS29 carriers were those who least needed treatment for anaemia, and there was no difference in overall survival or cancer incidence between genes of the two subunits [46]. In the Canadian registry, RPL5 was more frequent among corticosteroid non-responders (22.0% versus 0%, p=0.004) [47].

RPL35A is associated with severe neutropenia and immunodeficiency [1, 2]. In the multicentre series of 45 patients with RPL35A (plus 21 from the literature), 22 of 45 had large 3q29 deletions; these patients more frequently had steroid-resistant anaemia, neutropenia, craniofacial anomalies, chronic gastrointestinal problems and intellectual disability (p<0.01), interpreted as an effect of contiguous genes (LMLN and IQCG among the co-deleted genes of interest) [36]. The consensus generalises that large genomic deletions are associated with complex malformations and developmental delay [1].

RPL15 has a severe fetal presentation: in a European series of six individuals, three of four carriers of truncating variants had fetal hydrops requiring intrauterine transfusion, and the three carriers of p.Tyr81* became treatment-independent between 4 and 16 months, with normal blood counts at last follow-up and no demonstrated genetic reversion [4, 48]; another case of non-immune fetal hydrops due to RPL15 was published in 2023 [49]. RPL26 has a mainly malformative phenotype: the first proband, with a de novo 2 bp deletion, had multiple anomalies and typical haematological DBA, and a family published in 2025 had two siblings with limb malformations without anaemia and raised eADA in the proband [4, 50]. RPS28 and TSR2 are associated with mandibulofacial dysostosis resembling Treacher Collins syndrome (downslanting palpebral fissures, malar and mandibular hypoplasia, pinna anomalies, cleft palate), short stature and cardiac and renal anomalies; a third patient with RPS28 c.2T>C had Pierre Robin sequence, severe sensorineural hearing loss, developmental delay and severe anaemia from 2 years and 10 months [4, 24]. RPL18 was described in one family of the NCI cohort of 87 families and in a three-generation family with VACTERL association (vertebral, anorectal, cardiac, tracheo-oesophageal, renal and limb anomalies) [21, 42]. HEATR3 combines bone marrow failure, short stature, facial and acromelic dysmorphism and intellectual disability [26]. Gain-of-function TP53 gave, in the two French patients, normocytic anaemia, transient neutropenia, distinct neurological abnormalities and no response to corticosteroids; in the two Russian patients with de novo variants, both became transfusion-independent, which the authors attributed to L-leucine [27, 28]. Also published in 2026 were a neonatal case of DBA with persistent neutropenia due to an RPS15A variant and an Indian family with an RPS26 variant and variable phenotype, both consulted only by title [51, 52]. GATA1 is associated with thrombocytopenia, neutropenia and, in some patients, dyserythropoiesis and dysmegakaryopoiesis, a distinct subgroup within the syndrome [1].

2.6. Penetrance, expressivity and mosaicism

Penetrance is almost complete for loss-of-function variants of RPL5, RPS19 and RPS26 and high for loss-of-function variants of the other RP genes; missense variants (for example, of RPS19) appear less penetrant [2, 8]. Expressivity varies within the same family, from fetal hydrops to absence of anaemia with macrocytosis, raised eADA or HbF as the only signs [1, 2]. In the case of RPL11 c.157+1G>C, the carrier father had a milder picture than his son with macrocytic anaemia, cleft palate and heart disease [41].

Revertant mosaicism, for example an acquired uniparental disomy that eliminates a de novo RPS19 variant in the marrow, rarely explains a spontaneous remission and may render the test on leucocyte DNA of a parent or of the patient negative [2, 53–55]. Gonadal mosaicism has been documented in a Czech family with two affected siblings and parents without the variant in blood and in a family published in 2025 [56, 57]. Processed pseudogenes of RPS19 and RPS28 (the latter with several pseudogenes) may confound read alignment; in the patient with RPS28 c.2T>C, the variant was initially filtered out of the exome because of low depth (10 and 2 reads) and allelic imbalance [4, 56].

2.7. Allelic disorders, modifiers and phenocopies

GeneReviews recognises no other germline phenotype for most DBA genes [2]. The exceptions are GATA1, whose germline variants cause the spectrum of GATA1-related cytopenia (thrombocytopenia, anaemia of variable severity, platelet dysfunction, mild β-thalassaemia, neutropenia, congenital erythropoietic porphyria; cryptorchidism and hypospadias in some patients) and whose somatic variants, in newborns with trisomy 21, cause transient abnormal myelopoiesis, followed by acute megakaryoblastic leukaemia in 10–20% [2, 6]; and RPS20, which the consensus appendix also cites in connection with a study of germline predisposition to non-polyposis colorectal carcinoma without DNA mismatch repair deficiency [1]. A patient with one variant in RPL27 and another in RPS19, each inherited from an unaffected parent, was interpreted as digenic [4]. RPS14 has no established germline disease; its somatic haploinsufficiency causes 5q− syndrome [4, 31].

ADA2 deficiency accounted for 6% of patients without a genetic diagnosis in the German DBA registry and, in the 2018 exome series, 9 patients with a clinical diagnosis of DBA had biallelic ADA2 variants [1, 8]. The homozygous EPO variant NM_000799.4:c.530G>A p.(Arg177Gln) produces a PRCA that responds to recombinant erythropoietin, unlike DBA [2].

2.8. Other ribosomopathies: molecular basis

3. Clinical Spectrum and System-Specific Manifestations

3.1. Haematological manifestation and age at presentation

The central manifestation is a profound, normochromic and usually macrocytic anaemia, with reticulocytopenia, normal leucocytes and platelets or changes without clinical relevance, and a marrow with scarce or absent erythroid precursors [1, 2]. Ninety per cent of patients have symptoms in the first year of life; in the series cited by GeneReviews, the median age was 2 months at presentation and 3 months at diagnosis [2, 65]. The disease may manifest in utero with fetal anaemia and non-immune hydrops, or only in adulthood, up to the sixth decade, when it is easily confused with acquired PRCA or MDS; the consensus notes a trend towards later diagnoses and cites an unpublished British analysis in which only 77% (112 of 146) of patients had anaemia in early childhood [1, 2]. Presenting signs are pallor, weakness and faltering weight gain [2]. Haemolysis, hepatomegaly and splenomegaly are not part of the picture [1].

At presentation there may be thrombocytosis (only in infants), thrombocytopenia or neutropenia, generally without clinical relevance: in a study of 38 children cited by the consensus, 21 had thrombocytosis, 12 thrombocytopenia and 3 also leucopenia [1]. The marrow is normocellular at presentation, with absent or reduced erythroid activity, left shift rarely beyond the pronormoblast and normal erythroid morphology, without dyserythropoiesis (except in GATA1 forms), ring sideroblasts or vacuoles [1, 2]. With age, the marrow becomes hypocellular: among steroid-refractory patients, 75% developed moderate to severe hypoplasia and, of these, 43% neutropenia and 29% thrombocytopenia; pancytopenia of the severe aplasia type is not characteristic outside clonal evolution [1].

3.2. Spectrum of severity

The spectrum is wide, even within the same family. Alongside typical DBA, GeneReviews describes three atypical phenotypes: mild anaemia; absence of anaemia with subtle erythroid changes (macrocytosis, raised eADA and/or increased HbF); and malformations without anaemia. At the severe extreme are fetal anaemia and non-immune fetal hydrops [2]. Examples of this variability are the RP variant carriers with congenital heart disease and without anaemia, the siblings with RPL26 and limb malformations without anaemia, and the three-generation family with VACTERL association and RPL18 [1, 42, 50].

3.3. Congenital anomalies by system

More than half of patients have at least one congenital anomaly and about 25% have more than one [1, 2]. Frequencies differ according to source and period; the table compares the summary in the appendix of the 2024 consensus (median and range of several registries) with Table 2 of GeneReviews, based on the British, Italian and North American registries published between 1996 and 2006 [1, 2, 66, 67].

Smith's adds as occasional findings prematurity, downslanting palpebral fissures, choanal atresia, retinopathy, 11 pairs of ribs, bifid thoracic vertebra, agenesis of the clavicle, hip dysplasia, hypoplasia of the distal sacrum and coccygeal vertebrae, dysplastic middle phalanx of the fifth finger and multicystic kidneys, and describes the triphalangeal thumb and mild radial hypoplasia as the main findings of the form with skeletal anomalies [3]. The triphalangeal thumb also occurs in Holt-Oram syndrome, and cleft lip or palate in a patient with a radial defect points to DBA, Roberts syndrome or Cornelia de Lange syndrome [68].

3.4. Growth and endocrine system

Low birth weight occurs in 25% of affected infants and growth failure in about 30%; final height is influenced by chronic anaemia, corticosteroid therapy and iron overload, which makes it difficult to separate the intrinsic effect of the disease [1, 2, 69]. Smith's describes a mild growth deficit, around the 3rd centile [3]. According to the consensus appendix, more than 50% of patients have endocrine problems related to corticosteroids and iron overload: adrenal insufficiency, sex hormone insufficiency, growth hormone dysfunction, hypogonadism, thyroid and parathyroid abnormalities, pancreatic insufficiency and diabetes [1]. The French PNDS (National Diagnosis and Care Protocol) lists six causes of short adult stature in DBA: syndromic short stature with fetal growth restriction, rare growth hormone deficiencies (with or without an anomaly of the pituitary region on imaging), corticosteroids, iron overload, pubertal delay and chronic anaemia; it therefore recommends at least one paediatric endocrinology consultation for all children [9].

3.5. Immunity and infections

Up to 55% of patients have quantitative deficiencies of immunoglobulins and/or of T, natural killer (NK) and B lymphocytes, independent of corticosteroids and of the RP gene involved [1]. The consensus appendix cites a British series of 107 unselected cases with impaired cellular and humoral immunity [1]. In an Egyptian series of 34 children with DBA or a similar picture, 41% (14) had an immunodeficiency infection score of 6 or more, with lower immunoglobulins in this group (IgG 471 versus 1057 mg/dL; IgA 24 versus 98.5 mg/dL); mortality was 23.5% (8), with two deaths from infection, but in the molecular study of 8 probands 7 had ADA2 deficiency, which limits extrapolation to ribosomal DBA [70]. Severe persistent lymphopenia detected by newborn screening for inborn errors of immunity has been described in a child with DBA [71]. Treatment-related infections include catheter infections, transfusion-transmitted agents, viral infections on corticosteroids and Campylobacter, Yersinia and mucormycosis infections associated with iron overload and aggravated by deferasirox [1]. The French PNDS describes frequent lymphopenia, accentuated by corticosteroid therapy, and progression to a common variable immunodeficiency-like picture, which underpins annual measurement of IgG, IgA and IgM in all patients [9].

3.6. Other intrinsic manifestations

The consensus describes severe colitis as a recent association [1]. Patients with 3q29 deletions of RPL35A have more chronic gastrointestinal problems [36]. The psychosocial burden of chronic disease is recognised as a manifestation requiring follow-up [1].

3.7. Cancer predisposition

The risk of MDS, AML and solid tumours is increased, as shown by the North American registry and the NCI cohort of inherited bone marrow failure syndromes with 15 years of follow-up [2, 72]. In the North American registry (DBAR), with 608 patients and 9,458 person-years of prospective follow-up since 1991, 15 solid tumours, 2 AML and 2 MDS occurred in non-transplanted patients, at a median age of 41 years; the observed/expected (O/E) ratio was 5.4 for all cancers, 287 for MDS, 28 for AML, 36 for colon carcinoma, 33 for osteosarcoma and 12 for female genital cancers; the cumulative incidence of solid tumour or leukaemia was about 20% at 46 years, lower than in Fanconi anaemia and dyskeratosis congenita [10]. The 2018 update of the same registry, cited by the consensus, gave O/E of about 5 for all cancers, 45 for colon cancer, 42 for osteosarcoma, 29 for AML and 350 for MDS [1, 73]. A DBAR analysis summarised in the preliminary colorectal screening recommendations gave a relative risk of cancer of 4.8 and a cumulative incidence of 13.7% at 45 years, with colorectal cancer and osteosarcoma as the most frequent solid tumours [74]. Colorectal cancer arises in young adults and osteosarcoma in children and adolescents, often diagnosed at an advanced stage and with reduced tolerance of full chemotherapy [75]. In a literature compilation by Shimamura and Alter, cited by the consensus, 15 cases of MDS/AML and 19 solid tumours (6 osteosarcomas) were recorded among 970 cases [1]. The French PNDS, drawing on the two North American registry studies, gives 36 years as the median age of the first solid tumour, identifies osteosarcoma and colon cancer as the two tumours with a clear excess and interprets early MDS as premature ageing of the marrow; it places them mainly from 50–60 years in the annex devoted to DBA, but from 30 years in the general table of constitutional bone marrow failures, and notes that there are no published descriptions of these MDS [9, 10, 73].

The consensus also describes smaller series: in 62 patients, 3 MDS between 25 and 29 years and single cases of colorectal cancer, breast cancer, lymphoma and multiple myeloma between 28 and 70 years; in another registry, an MDS at 4 years, gastric and thyroid cancer, lymphoma and osteosarcoma between 11 and 48 years; and, after HCT, two osteosarcomas and a cardiac Purkinje cell tumour at 19, 18 and 9 years, whose short latency suggests tumours intrinsic to DBA [1]. No association between gene and cancer is known, except between GATA1 and MDS with monosomy 7 [1]. In the GeneReviews on GATA1 cytopenia, carriers of variants that produce only GATA-1s (including heterozygous women) may develop myeloproliferative disease, MDS or myeloid leukaemia associated with acquired trisomy 21 and other chromosomal anomalies [6].

3.8. The disease in adults

In the French registry of 235 adults, at last observation 78% had anaemia, 31% neutropenia, 15% thrombocytopenia and 11% pancytopenia, without severe aplasia outside clonal evolution; among those not receiving specific treatment, 52% had anaemia, 35% neutropenia and 10% thrombocytopenia, which shows that treatment independence is not equivalent to haematological remission [76]. There were 4 MDS/AML, all with high-risk features and unfavourable course (O/E 155 for MDS and 55.4 for AML); with a median follow-up of 32.2 years (interquartile range 26–44), overall survival was 98.0% (95% confidence interval [95% CI] 95.8–100) at 30 years, 90.6% (84.2–97.5) at 40 and 70.7% (57.3–87.2) at 50, and solid and haematological cancers were the main cause of death [76]. In DBAR, the cumulative incidence of death in adults was 23% at 45 years, mainly from complications of HCT, iron overload, infections and leukaemia or solid tumours [1, 73].

3.9. Response to treatment and treatment independence as data on treated natural history

Corticosteroids produce an initial erythroid response in about 60–80% of patients and 30–40% maintain a lasting response [1, 2]. In the Canadian registry, 30 of 80 evaluable patients were responders and 50 non-responders (including loss of response); the median age at diagnosis was 20 months in responders and 3 months in non-responders (p=0.004), with a cut-off of 9.5 months on the ROC (receiver operating characteristic) curve, without difference in overall survival [47]. A 2026 paper argues that age at the start of treatment, and not at diagnosis, predicts response; only the title was consulted [77]. About 20% of treated patients become treatment-independent; in the French registry of 222 treated patients, the proportion was 21% (30% in those initially steroid-sensitive and 5% in non-responders), about 70% of them in the first decade; neither corticosteroid response nor genotype predicts independence, the anaemia may return and the cancer risk persists [1]. In an earlier analysis of the same registry, cited by the French PNDS, 42 of 125 corticosteroid responders were in treatment independence; these phases occur mainly before 2 years of age or in adolescence, and somatic reversion in a haematopoietic stem cell has been demonstrated in only 2 published cases, so it is considered a minority mechanism [9]. Untreated patients retain subtle abnormalities (haemoglobin at the lower limit, macrocytosis, raised HbF, tendency to leuconeutropenia), may have relapse of anaemia in adulthood and retain the same risk of solid tumours and hypogammaglobulinaemia [9]. Transfusional iron overload and cancer are the main causes of death in non-transplanted patients; DBA has higher non-transferrin-bound iron than thalassaemia or sickle cell disease, because iron is not used in erythropoiesis, with faster overload, frequent cardiac complications and early pancreatic loading [1].

Treatment-related complications are summarised in the consensus appendix: alloimmunisation, agranulocytosis (deferiprone), hepatic toxicity, renal toxicity with phosphate wasting and gastrointestinal toxicity (deferasirox), urinary calculi (deferasirox and desferrioxamine), ototoxicity (desferrioxamine), cataracts (corticosteroids and deferasirox), osteopenia and fractures (corticosteroids), cirrhosis and cardiomyopathy with arrhythmias (iron), infertility (iron and HCT) and increased risk of cancer after HCT [1].

3.10. Gene-specific forms with distinctive manifestations

RPL35A forms with a 3q29 deletion combine steroid-resistant anaemia, neutropenia, immunodeficiency, craniofacial anomalies, gastrointestinal problems and intellectual disability; RPL15 forms may present with fetal hydrops and progress to early treatment independence; RPL26 and RPL18 may give congenital anomalies with normal haematology; RPS28 and TSR2 combine DBA and mandibulofacial dysostosis; GATA1 combines thrombocytopenia, neutropenia and dyserythropoiesis; gain-of-function TP53 combines neurological abnormalities and no response to corticosteroids; and HEATR3 combines bone marrow failure, short stature, dysmorphism and intellectual disability (see 2.5) [1, 24, 26, 28, 36, 42, 48, 50].

3.11. Other ribosomopathies: summary of the clinical picture

In 5q− syndrome, somatic loss of RPS14 in an MDS clone produces refractory macrocytic anaemia in adults, making it the main acquired phenocopy of DBA at this age [1, 31]. In cartilage-hair hypoplasia, there is short-limbed dwarfism (birth length about −3 standard deviations; adult height 104–151 cm, median 131 cm in men and 122 cm in women), fine sparse hair, ligamentous laxity, cellular immunodeficiency, Hirschsprung disease in about 10% and severe varicella in up to about one third [7, 63]. The anaemia is macrocytic: mild in about 80%, with spontaneous resolution in childhood in most, and severe and persistent, resembling DBA, in about 6%, of whom 50–75% need lifelong transfusions or HCT [7]. In the Finnish cohort, 14 of 123 patients (about 11%) had a neoplasm over 39 years of follow-up, with a Kaplan-Meier probability of cancer (excluding basal cell carcinoma) of 41% at 65 years, mainly non-Hodgkin lymphoma and basal cell carcinoma [7, 78]. The Finnish studies describe impaired T-cell function in vitro in 88%, anaemia in 79%, lymphopenia in 62% and neutropenia in 27% [63].

In isolated congenital asplenia, absence of the spleen without heart disease or altered situs exposes the patient to fulminant bacterial infections: in the French national series of 20 patients from 10 families, the median age at first infection was 12 months and at diagnosis of asplenia 11 months, 15 of 20 had 18 episodes of invasive bacterial infection (61% due to Streptococcus pneumoniae) and 9 of 20 (45%) died of fulminant infection [4, 79]. Treacher Collins syndrome is characterised by downslanting palpebral fissures, lower eyelid coloboma, malar hypoplasia, micrognathia, microtia and conductive hearing loss (40–50%), with critical airway problems at birth in 12–18% and intellectual disability in less than 5% [59]. Shwachman-Diamond syndrome combines exocrine pancreatic insufficiency, neutropenia (intermittent or persistent, the most frequent cytopenia), metaphyseal changes, short stature and AML or other haematological neoplasms in more than 5% [12, 80]. In the table of the French PNDS (National Diagnosis and Care Protocol), SBDS explains about 90% of cases (SRP54 and EFL1 less than 10%); pancreatic insufficiency improves after 4 years in half of patients; there is neutropenia in 90% and anaemia in 80%; MDS is defined only with more than 5% blasts, because mild granulocytic dysplasia is frequent without MDS; the cumulative incidence of MDS/AML is 18–36% at 30 years, with male predominance; small clones with somatic TP53 variants are frequent and only those with a high allele fraction are significant, and iso(7q) and del(20q) have no prognostic value [9]. In dyskeratosis congenita, the classic triad is dysplastic nails, reticulated pigmentation of the chest and neck and oral leucoplakia, with progressive bone marrow failure, MDS/AML, solid tumours and pulmonary fibrosis [2]. The RNA polymerase III leukodystrophies combine neurological dysfunction, abnormal dentition, hypogonadotropic hypogonadism and progressive myopia [12].

4. Clinical Recognition and Warning Signs

4.1. When to suspect, by age and context

In the prenatal period, DBA enters the differential diagnosis of fetal anaemia and non-immune fetal hydrops, especially with a family history of congenital anaemia or unexplained macrocytosis; RPL15 forms are those most associated with hydrops requiring intrauterine transfusion [2, 48, 49]. Ultrasound identifies the radius and ulna from the beginning of the second trimester, but thumb hypoplasia is detected only when bones are absent [68].

In the infant, suspicion arises in the presence of pallor, weakness, faltering weight gain and macrocytic anaemia with reticulocytopenia without another significant cytopenia, especially in the first months of life, with or without thumb anomalies, cleft lip or palate, heart disease, urogenital anomalies or fetal growth restriction [1, 2]. In the older child, the main difficulty is distinguishing transient erythroblastopenia of childhood (TEC), which usually appears after the first year and has normal mean corpuscular volume (MCV), eADA and HbF [1, 2].

The consensus recommends considering DBA syndrome even without anaemia when there is a family history, macrocytosis, compatible congenital anomalies, associated cancers (MDS, osteosarcoma or early colorectal cancer) or excessive haematological toxicity from chemotherapy [1]. In adults, late-onset or unrecognised DBA presents as macrocytic anaemia, PRCA or suspected MDS; at this age, acquired del(5q) with RPS14 haploinsufficiency is the main differential [1]. In a family donor being considered for HCT, macrocytosis, raised eADA or raised HbF may reveal a form without anaemia [1, 2].

The French PNDS (National Diagnosis and Care Protocol) describes the typical infant picture as isolated pallor with dyspnoea that hinders feeding, without splenomegaly or haemolysis, and adds to the circumstances of diagnosis the investigation of a B-lymphocyte deficiency with hypogammaglobulinaemia, MDS before 60 years and a cancer unusual for the age or characteristics, with an unexpected transfusion requirement during chemotherapy [9]. In the investigation of bone marrow aplasia, DBA is an exceptional cause, rarely severe on the blood count; fetal growth restriction, malformations (thumb, face, heart, genital tract) and an abnormal blood count for more than 6 months point to a constitutional form, and the PNDS includes eADA among the investigations targeting constitutional forms, performed before 18 years or in the presence of a family history or suggestive signs [9].

4.2. Signs pointing to a gene or entity

4.3. Warning signs requiring urgent action

Severe anaemia in the infant, with very low haemoglobin at diagnosis (2.6 and 4.4 g/dL in the two 2025 cases described in section 10), requires transfusion, but blood for eADA, HbF and genetic studies should be collected beforehand, because transfusion invalidates these markers for weeks [1, 40]. Warning signs during follow-up are a rapid fall in red cells, leucocytes or platelets (possible MDS/AML), fever in a patient on deferiprone (agranulocytosis), fever or infection in a patient on high-dose corticosteroids or with lymphopenia, heart failure or arrhythmia with cardiac iron overload (cardiac T2* <10 ms indicates imminent risk of heart failure), unexplained bone or joint pain (osteosarcoma) and an atypically high transfusion requirement (alloimmunisation, haemorrhage) [1, 2]. In isolated congenital asplenia, any fever in the child is potentially a fulminant bacterial infection [79, 82]. In bone marrow failure, the French PNDS defines fever as a temperature >38.5 °C once or >38 °C twice 4 hours apart and considers that any fever with neutrophils <0.5 × 10⁹/L requires urgent hospital admission with intravenous antibiotics, a rule that covers agranulocytosis on deferiprone and the severe neutropenias due to RPL35A; in France, there is a DBA-specific Orphanet emergency sheet ("Bonnes pratiques en cas d'urgence — Anémie de Blackfan-Diamond") and the reference centre gives patients a bone marrow aplasia emergency card [9].

4.4. Newborn screening

There is no newborn screening targeting DBA in the sources consulted. Newborn screening for inborn errors of immunity by quantification of T-cell receptor excision circles and kappa-deleting recombination excision circles (TREC/KREC) identified severe persistent lymphopenia in a child with DBA syndrome; the report was consulted only by title, which indicates that the lymphopenia of the syndrome can be captured by this screening [71].

5. Diagnosis: Criteria, Tests and Algorithms

5.1. Diagnostic criteria and their evolution

The 2024 consensus simplified the diagnosis: one of two criteria suffices [1]. The 2025 GeneReviews retains a clinical diagnosis with six cumulative criteria and a molecular diagnosis, and the 2008 consensus used major and minor criteria; the table compares the three versions [1, 2, 30].

Diagnosis by genetics alone, for example in a typical infant or in a relative with known segregation, should be accompanied by marrow confirmation, which in infants is usually deferred until just before corticosteroids are started [1]. eADA and the rRNA processing test are not required [1]. Corticosteroids should not be started without an established diagnosis of DBA [1].

5.2. Recommended investigations

GeneReviews adds, in the post-diagnosis evaluation, extended red cell phenotyping for future unit selection, renal and urinary tract ultrasound, echocardiography, orthopaedic assessment if Klippel-Feil, Sprengel or thumb anomalies are suspected, assessment of growth and development, and genetic counselling, and notes that macrocytosis is not always present [2].

5.3. Biomarkers

eADA activity, sampled before transfusion, has a sensitivity of 84% and a specificity of 95% for distinguishing DBA from other anaemias and inherited bone marrow failure syndromes; a normal eADA does not exclude the diagnosis, especially in GATA1 forms [1]. GeneReviews describes raised eADA in 80–85% of patients; in TEC only 10% have raised eADA [2]. In the 2025 genome series, a patient with RPS7 had an eADA of 2,138 mU/g haemoglobin (laboratory reference 400–900), whereas the patient with intronic RPS19 had normal eADA (620 mU/g) [40]. Raised eADA is not exclusive to DBA: the GATA1 variants p.(Arg307Cys) and p.(Arg307His) give haemolytic anaemia with raised eADA and erythroid hyperplasia, and very high erythrocyte ADA1 hyperactivity (about 50-fold) has been described as a cause of non-spherocytic haemolytic anaemia, a distinct entity [6, 81]. Raised HbF is reliable after 6 months of age [1]. The French PNDS (National Diagnosis and Care Protocol) requests eADA in a non-transfused patient or more than 3 months after the last transfusion, a longer interval than the minimum 6 weeks of the consensus, and, in the investigation of bone marrow aplasia, considers HbF >5% before any transfusion to favour a constitutional form [1, 9]. Serum erythropoietin is raised in DBA, unlike the phenocopy due to an EPO variant [1, 2].

The rRNA processing test, available in specialised research laboratories, is useful when the genetics are ambiguous [1]. The DNA methylation episignature, studied in 80 patients (RPS19 30, RPS17 6, RPL35A 8, RPL5 15, RPL11 13, RPS26 8), distinguished DBA from Fanconi anaemia and other diseases with an episignature and included an RPL5 versus RPL11 classifier; in 6 cases without a molecular diagnosis, 3 had the pattern and genome sequencing found the cause in 2 (RPL5 NM_000969.5:c.325-380A>G and RPL26 NM_000987.5:c.-6+3_-6+25del), and two cases with somatic reversion retained the episignature; a correction to the article was published; the tool remains a research tool, without guideline validation [39, 84].

5.4. Parvovirus B19 and the marrow

Parvovirus B19 has tropism for erythroid colony-forming units (CFU-E) and, in individuals with normal red cell turnover, interrupts erythropoiesis without causing anaemia; in infants and in patients with chronic haemolysis or immunodeficiency it may cause severe anaemia [1]. Diagnosis combines serology (IgG and IgM) and PCR for viral DNA in blood or marrow; the marrow in acute infection lacks mature erythroid precursors and shows characteristic giant pronormoblasts [1]. Viral genome positivity was much higher in the marrow (22%) than in peripheral blood cells (0.8%) in infected individuals, and B19 DNA is found in the marrow of 2% of healthy seropositive individuals, so persistence of DNA may correspond to infectious virus or to residual DNA from a past infection [1].

5.5. Genetic testing

The consensus recommends genetic testing for all patients with suspected disease and family studies after a pathogenic variant has been identified; the strategy may be stepwise or "all in one test" (panel, microarray, exome, genome), provided it covers single-nucleotide variants, exon and whole-gene deletions and contiguous-gene microdeletions; some experts prefer to test simultaneously the DBA genes, the phenocopies and the other inherited bone marrow failure syndromes [1]. GeneReviews proposes as option 1 a multigene panel with the DBA and differential genes, with sequencing and deletion/duplication analysis, and as option 2, in atypical phenotypes, exome or genome; exome and genome detect deletions and duplications by breakpoints or read depth, with lower sensitivity than targeted analysis (quantitative PCR, long-range PCR, MLPA, targeted microarray) [2]. Interpretation follows ACMG/AMP; there are as yet no ClinGen DBA-specific specification rules [1]. ClinGen assigned DBA (RPS19, RPS10, RPS24; document AC1045) strong actionability in the paediatric context, with no assertion for adults, and classifies RPS19 haploinsufficiency as having sufficient evidence (score 3) [4].

The 2023 French PNDS identifies a variant in about 80% of patients, with RPS19 in 25% and six genes (RPS19, RPL5, RPL11, RPS26, RPS24 and RPL35A) in about 90% of patients with a known genotype, and mentions 7 other RP genes under validation [9]. The first-line test is targeted next-generation sequencing; if negative, large deletions, which the PNDS estimates at about 20% of cases, should be sought by custom comparative genomic hybridisation microarray or by MLPA (multiplex ligation-dependent probe amplification); without a variant, exome or genome sequencing under the France Genomic Medicine Plan (SeqOIA and AURAGEN platforms), on fibroblast DNA, is performed in a research context or after discussion at a national meeting [9]. When the presentation is bone marrow failure, the PNDS recommends the panel preferably on a non-haematopoietic tissue (fibroblasts from a skin biopsy), because reversion and clonal evolution can give false negatives in blood, and in France samples with suspected DBA are sent preferentially to the reference laboratory at Hôpital Robert-Debré, Paris [9]. The gene table in annex 2 of the PNDS announces 23 RP genes but lists 20 distinct ones (RPS20 appears twice) and does not include RPS29, RPL4, RPL9 or RPL17, which the 2024 consensus incorporates; it also estimates familial forms at 20–45% [1, 9].

Table 1 of GeneReviews gives, for each gene, the proportion of variants detectable by each method according to the HGMD database; these values are not clinical sensitivities [2].

5.6. Limitations of the tests

A negative result does not exclude DBA: 20–30% of patients have no identified genetic cause [1]. The known limitations are as follows. Exon or whole-gene deletions escape sequencing alone and weigh mainly in RPS17, RPL35A, RPL15, RPS26 and RPL11 [2]. Long indels of RPL5, RPS10 and RPS24 may escape the exome [2, 8]. In RPS26, six extended or cryptic splicing variants and one 3′ untranslated region variant have been described [2, 8]. Exome and panel captures lack probes for non-coding exons, and variants such as RPS7 c.-19G>C (last nucleotide of a non-coding exon) and RPS19 c.172+350C>T (deep intronic) were found only by reanalysis of an initially negative genome and confirmed by RNA sequencing, after clinical laboratories had classified them as VUS [40]; a second case of a deep intronic RPS19 variant was published in 2026 [85]. The RPS19 pseudogenes and the several processed pseudogenes of RPS28 may cause alignment errors, and low read depth with allelic imbalance led to a true RPS28 variant being filtered out [4, 56]. Revertant mosaicism may reduce or eliminate the variant in blood, and parental mosaicism may give false negatives in the parents' leucocyte DNA [2, 55]. Missense variants may require functional studies to move out of the VUS category [1]. In minute somatic 5q deletions, karyotype and FISH may be normal and diagnosis requires SNP microarray [58].

5.7. Report content, VUS and segregation

The report should state the gene, the reference transcript with version (for example, NM_001022.4 for RPS19), the HGVS c. and p. description, the zygosity (heterozygous in RP genes, hemizygous in GATA1 and TSR2 in males, biallelic in HEATR3), the ACMG/AMP class, the methods used and the regions not covered, including the ability to detect deletions [1, 2, 40]. Pathogenic and likely pathogenic variants have the same clinical value; a VUS neither confirms nor excludes the diagnosis and therefore cannot be the basis for predictive testing or prenatal diagnosis [2]. Family segregation, the erythroid markers of relatives (MCV, eADA, HbF), RNA sequencing and the rRNA processing test may help to reclassify a VUS, as happened in the two 2025 cases [1, 40]. When no variant is identified, storing the proband's DNA for future analyses is recommended [2].

5.8. Practical algorithm

In the presence of macrocytic anaemia with reticulocytopenia in an infant or child, the first step is to collect, before any transfusion, a full blood count with reticulocytes, eADA, HbF, a sample for genetics and parvovirus B19 serology and PCR, and then to transfuse if necessary [1]. The second step is the bone marrow aspirate, which confirms isolated erythroblastopenia and excludes dysplasia, dyserythropoiesis, sideroblasts and other causes, before any corticosteroid [1]. The third is panel genetic testing with copy-number analysis, completed by tests targeting the other bone marrow failure syndromes when the picture suggests it (chromosome breakage, telomere length, faecal elastase, mitochondrial DNA, ADA2) and, if negative, by exome or genome with reanalysis and RNA sequencing in the centres that offer it [1, 2, 40]. Malformation assessment (echocardiography, abdominal and renal ultrasound, hand radiograph, ophthalmological and auditory assessment when indicated), immunological assessment and HLA typing of the patient and siblings complete the initial work-up [1, 2].

5.9. Prenatal and preimplantation diagnosis

Once the familial variant has been identified, prenatal diagnosis and preimplantation genetic testing are possible, according to each country's legislation; the consensus recommends discussing medically assisted reproduction with people who request these options [1, 2]. A VUS does not allow prenatal diagnosis [2]. Ultrasound does not replace genetic testing, because thumb hypoplasia is detected only when bones are missing and many forms have no malformations [1, 68]. In the pregnancy of a woman with DBA or of a couple with an affected child, screening for fetal anaemia is recommended [1].

5.10. Diagnosis in the other ribosomopathies

In cartilage-hair hypoplasia, suspicion is clinical and radiographic (metaphyseal changes, usually visible after 2 years) and molecular confirmation starts with the founder variant NR_003051.4:n.72A>G in patients of Amish or Finnish ancestry, in whom this variant is present in 100% and 92% of cases, respectively [7, 63]. In isolated congenital asplenia, Howell-Jolly bodies on the blood film and abdominal ultrasound guide the diagnosis, and the causal RPSA variants lie in translated exons and in the 5′UTR region; these RPSA variants explained 55% of patients (40 of 73) and 41% of families (23 of 56), with 92% (12 of 13) in families with several affected members and 26% (11 of 43) in sporadic ones [32, 83]. In 5q− syndrome, diagnosis rests on marrow morphology, karyotype, FISH and somatic MDS mutations, with normal eADA [1]. Diagnosis of Treacher Collins and Shwachman-Diamond syndromes has its own criteria; recognition of the SBDSP pseudogene is necessary in the interpretation of SBDS [61]. In the presence of bone marrow failure, the French PNDS directs the search for Shwachman-Diamond syndrome through fat-soluble vitamins, lipase, faecal elastase and, if in doubt, pancreatic MRI [9].

6. Differential Diagnosis

6.1. Comparative table

6.2. Main differentials in prose

TEC remains the main differential of DBA in children, although it is rare: it appears after the first year in more than 80% of cases, in a healthy child, without congenital anomalies, with normal MCV, eADA and HbF and with spontaneous recovery; transfusion is needed only in severe cases [1, 2]. The consensus advises against corticosteroids before a diagnosis of DBA has been established, because they mask the course and make the distinction difficult [1]. Parvovirus B19 infection is excluded by PCR and serology; seroprevalence reaches 50% by 15 years and 90% in the elderly [2].

ADA2 deficiency is the most relevant genetic phenocopy, because it is treated differently: 6% of patients without a genetic diagnosis in the German DBA registry had DADA2, and in an Indian series 9 children with PRCA due to DADA2 (mainly homozygous for ADA2 NM_001282225.2:c.278T>C p.(Ile93Thr)), initially taken to have DBA, did not respond to corticosteroids [1, 89]. The PRCA and bone marrow failure of DADA2 are associated with variants with minimal residual activity, nonsense variants and indels, are refractory to tumour necrosis factor inhibitors and probably benefit from allogeneic HCT; eADA is raised in DBA and ADA2 activity is low in DADA2, which discriminates between the two [81]. In the Egyptian series, 7 of 8 probands studied with "DBA or DBA-like" disease had DADA2 [70].

The inherited bone marrow failure syndromes are distinguished by multilineage cytopenia and characteristic multisystem involvement; MCV and HbF may also be raised in these diseases, but eADA is normal [1]. In a patient with a radial or thumb defect, the Oxford Desk Reference approach proposes distinguishing DBA (anaemia, cleft lip or palate, short stature), Fanconi anaemia (bilateral anomalies, consanguinity, café-au-lait spots over time, renal ultrasound, chromosome breakage test), TAR syndrome (thumb present with bilateral absence of the radius and thrombocytopenia) and Holt-Oram, Okihiro and Nager syndromes [68]. The metabolic macrocytic anaemias of childhood (orotic aciduria, cobalamin and folate defects, Pearson syndrome and other respiratory chain diseases) enter the differential when there is neurological, metabolic or pancreatic involvement [88]. In adults, 5q− syndrome and acquired PRCA (thymoma, autoimmunity, lymphoproliferation) are the main differentials, and normal eADA, karyotype and immunology guide the diagnosis [1].

GATA1 missense variants in the zinc fingers give macrothrombocytopenia and dyserythropoiesis, and the variants NM_002049.4:c.919C>T p.(Arg307Cys) and c.920G>A p.(Arg307His) give haemolytic anaemia with raised eADA and erythroid hyperplasia, unlike the erythroid hypoplasia of DBA; raised eADA is therefore not exclusive to DBA [6].

7. Treatments and Risk Management Strategies

7.1. Goals and options

The goal of treating the anaemia is to maintain haemoglobin at 9–10 g/dL (90–100 g/L) or above, ensuring adequate growth and development in the child and quality of life in the adult, on an individual basis [1]. The three established options are transfusions with iron chelation, corticosteroids and HCT; supplements of folate, other vitamins or trace elements are not indicated, and malformations are corrected as far as possible [1]. All recommendations of the 2024 consensus have evidence level C [1]. GeneReviews refers to the 2024 consensus as the clinical practice guideline but retains its own values on some points (transfusion target, corticosteroid maintenance dose, chelation thresholds), presented below as divergences [2].

7.2. Red cell transfusions

The consensus considers restrictive transfusion strategies with a threshold of 6–7 g/dL harmful in DBA (panel unanimity) and recommends a pre-transfusion haemoglobin of at least 9–10 g/dL throughout life, regardless of age, or a higher value at which the patient is asymptomatic; the 2008 consensus accepted 8 g/dL and GeneReviews gives a target of 80–100 g/L [1, 2, 30].

Before each transfusion, a full blood count with reticulocytes and, if possible, red cell antibody screening are recommended, and ferritin and transferrin saturation are monitored every 1–3 months [1].

The French PNDS (National Diagnosis and Care Protocol) recommends a transfusion threshold of 9.0–9.5 g/dL, with an interval of 3–4 weeks between transfusions (longer in patients who retain significant reticulocytosis), and notes that tolerance of chronic anaemia decreases with age and that some adults need more intensive support; transfusion support should allow normal growth and psychomotor development in the child and quality of life in the adult [9]. It retains at least annual serology (HIV, hepatitis B and C and human T-lymphotropic virus [HTLV]) and hepatitis B vaccination in non-immune patients and, in bone marrow failure in general, phenotyped red cell concentrates (Rh and Kell) to prevent alloimmunisation [9].

7.3. Corticosteroids

Drugs and mechanism. Oral prednisone or prednisolone, of equivalent potency, is used [1]. The mechanism of the erythroid response in DBA is not established in the sources consulted. In the USA, the prednisone (ANDA080352, 2023 revision) and prednisolone (ANDA215673, 2024 revision) labels include among the haematological indications "congenital (erythroid) hypoplastic anaemia", but do not define a regimen for DBA; the doses come from the consensus [90]. National European summaries of product characteristics (SmPCs), including the Portuguese ones, were not consulted.

Starting. Any transfusion-dependent patient from 12 months; possible at 15–18 months in children with growth delay; earlier (about 9 months) if there is no venous access or no safe transfusions [1]. Ideally, after the first live vaccines (first dose of the measles, mumps, rubella and varicella vaccine) given at least 3 weeks before and after the diagnostic marrow assessment [1]. GeneReviews gives a start at 12 months or more and 10–14 days after the last transfusion; the 2024 consensus did not reach agreement between starting one day after transfusion or about 10–14 days later, and no longer makes the start conditional on the interval since the last transfusion [1, 2].

GeneReviews is internally inconsistent on the maintenance dose (0.3 mg/kg/day in Table 5 and 0.25 mg/kg/day in the text); this article follows the 2024 consensus, which sets the maximum maintenance at 0.3 mg/kg/day or 0.6 mg/kg on alternate days, lower than in 2008 [1, 2]. GeneReviews notes, from the author's personal observation, that in a 10 kg child the dose of 2.5 mg/day is obtained by breaking a 5 mg tablet and that 5 mg on alternate days is equivalent to 2.5 mg/day, reducing adrenal suppression [2].

Loss of response and stopping. The consensus did not reach agreement between "active" and "passive" tapering (letting the child grow on the same absolute dose). If the response is lost during tapering, the dose should immediately return to the previous level that maintained haemoglobin ≥9 g/dL; otherwise, the whole response may be lost and it may be necessary to restart at 2 mg/kg/day. If haemoglobin ≥9 g/dL is not maintained with ≤0.3 mg/kg/day, corticosteroids should be stopped and the patient moved to transfusions [1]. Corticosteroid "holidays" of 1–3 years before or during puberty, with transfusions, serve to optimise growth and most patients regain the response afterwards (expert experience, without definitive evidence) [1, 2]. A second corticosteroid trial in non-responders may be made 1–2 years later, especially before HCT; further trials are not recommended [1]. The French PNDS sets the pre-pubertal stop at 2–3 years, because even doses below 0.15 mg/kg/day prevent a normal pubertal growth spurt, a decision to be taken with paediatric endocrinology; it also indicates stopping during growth hormone treatment and in pregnancy with relatively high doses, and recommends returning to transfusions in patients who accumulate toxicity and insufficient haemoglobin, a frequent situation in adults owing to the relative loss of steroid sensitivity [9]. DBA is, according to the PNDS, the only case in which corticosteroids are prescribed from 1 year of age for an indefinite period, which requires systematic vitamin D supplementation and densitometry, because bone loss occurs even with very low doses [9].

Results. The initial response occurs in about 60–80% and 30–40% maintain a lasting response on corticosteroids [1, 2]. Older age at diagnosis was associated with response in the Canadian registry and RPL5 was more frequent in non-responders [47]. In DBA due to gain-of-function TP53, neither of the two French patients responded [28].

Particular scenarios (Table 4 of the consensus). Acute fall in haemoglobin: single transfusion. Persistent fall: increase the dose or declare non-response. Oral contraception with oestrogens: may limit the response. Pregnancy or systemic disease, including cancer: stop corticosteroids and move to transfusions. Immunosuppression or lymphopenia with relevant infection: taper. Live vaccines: avoid during the initial high dose; subsequent doses are given after the corticosteroids have been reduced [1].

Adverse effects and their management. The effects described are osteoporosis and pathological fractures, weight gain, cushingoid appearance, hypertension, diabetes, growth restriction, gastric ulcer, cataracts, glaucoma, adrenal insufficiency, increased susceptibility to infection and avascular necrosis, which should be considered in the presence of bone or joint pain alongside osteosarcoma [1, 2]. The consensus recommends follow-up by endocrinology (adrenal insufficiency during tapering and maintenance, hypertension, diabetes), annual ophthalmological examination (cataracts), annual densitometry (osteopenia), vitamin D and calcium in all patients on prolonged corticosteroids, bisphosphonates for significant osteoporosis, a proton pump inhibitor or H2 antagonist during the initial high dose or if symptomatic, and prophylaxis against Pneumocystis jirovecii pneumonia at 2 mg/kg/day according to local policy (no consensus); the 2008 consensus recommended this prophylaxis at doses above 2 mg/kg/day [1, 30]. If growth is severely compromised, especially at puberty, corticosteroids should be replaced by a temporary transfusion regimen [2].

7.4. Assessment of iron overload

In DBA, ferritin has low accuracy: it rises with inflammation, may be raised with little iron loading or falsely low with severe overload; a high ferritin with raised transferrin saturation confirms early overload [1]. Liver iron concentration (LIC) by magnetic resonance imaging (MRI), using T2* or R2 (FerriScan), is the standard for guiding chelation, with a target below 3 mg iron per gram dry weight (mg/g × 18 = µmol/g) [1]. Cardiac T2* is normal above 40 ms in volunteers; below 20 ms it is associated with dysfunction and reduced survival, below 10 ms with imminent risk of heart failure, and above 35 ms indicates no relevant loading; the panel's target is 20–35 ms, and many experts prefer more than 25 ms [1]. Pancreatic and pituitary haemosiderosis predicts endocrinopathy but is difficult to measure [1]. Liver and cardiac MRI should be performed as early as possible and at the latest at 5 years (earlier if loading is high or if HCT is planned), and then annually or more frequently in all patients on chronic transfusions; before 5 years sedation may be required [1]. GeneReviews also lists magnetic biosusceptometry (SQUID), non-invasive and little available, and liver biopsy, accurate but impractical for follow-up [2].

7.5. Iron chelation

When to start. The 2024 consensus recommends starting chelation after about 10 transfusions or when there is evidence of overload (serial ferritin >500 ng/mL, transferrin saturation >60% or raised LIC), which in children typically coincides with the first failed corticosteroid trial; in infants, one should wait for the first failed trial and start at a low dose with close monitoring [1]. The 2008 consensus and GeneReviews used higher thresholds: 10–12 transfusions (170–200 mL/kg of red cell concentrate), ferritin 1,000–1,500 µg/L or LIC 6–7 mg/g [2, 30]. The deferasirox labels set the start after transfusion of about 100 mL/kg of red cell concentrate (about 20 units for 40 kg) and ferritin persistently >1,000 µg/L [91]. Chelation is continuous and doses should be reduced as soon as iron falls, because of the toxicity of over-chelation; overload often persists in corticosteroid responders, in treatment-independent patients and after HCT, who remain under surveillance and on chelation or, only after HCT, phlebotomy [1].

Drugs (Table 5 of the 2024 consensus).

Common combinations are desferrioxamine 5 nights per week (about 12 h) with deferasirox on the remaining 2 days, or desferrioxamine 7 nights per week with deferasirox during the day [1]. In DBA, desferrioxamine at a dose ≤30 mg/kg below 3 years follows the analogy with the growth delay and bone changes described in thalassaemia; 50–60 mg/kg/day controls overload in older children and adults; the continuous 24 h intravenous route is used for cardiac loading, often with another chelator [1]. Agranulocytosis with deferiprone occurred in 10% (3 of 23) of the French series, non-fatal and resolved on withdrawal, and a fatal agranulocytosis has been described in a child with DBA [1, 94, 95]. Renal Fanconi syndrome secondary to deferasirox has been described in a case series with DBA, with recommendations for early diagnosis [96]. Deferiprone is prescribed by an experienced clinician, with an emergency plan for fever and agranulocytosis (stop the drug, antibiotics and granulocyte colony-stimulating factor [G-CSF] if necessary) [1].

Adjustment and targets. Doses are adjusted every 3–6 months, with targets of LIC <3 mg/g, cardiac T2* >20–35 ms and ferritin <500 ng/mL; without MRI (a non-standard situation), ferritin of 500–1,000 ng/mL prompts consideration of a reduction, 300–500 ng/mL a reduction or pause and <300 ng/mL a pause; if ferritin is <500 ng/mL with LIC >5 mg/g, low-dose chelation is maintained with intensified monitoring [1]. In rapid or severe overload, the consensus suggests considering genetic testing for haemochromatosis [1].

Toxicity monitoring. With desferrioxamine and deferasirox, annual audiometry; with deferasirox, annual ophthalmological examination, renal function (creatinine, phosphate wasting and proteinuria as signs of renal Fanconi syndrome), liver function and gastrointestinal symptoms; in pancreatic overload, fasting glucose, oral glucose tolerance test and fructosamine (instead of haemoglobin A1c in transfused patients); periodic renal ultrasound with desferrioxamine and deferasirox (renal stones); densitometry every 1–3 years with desferrioxamine; with deferiprone, weekly full blood count at the start and in any febrile episode, with withdrawal in the presence of progressive or unusual neutropenia, and a medicine "passport" with an emergency plan [1]. Toxicity is greater in patients with low ferritin [1, 97].

Chelation in infants and young children. In the French and Italian registries, among 167 transfused patients screened, 64 (38%) started chelation before 3 years (median 18 months), with ferritin ≥500 ng/mL (median 1,340) and more than 10 transfusions; 63% received deferasirox and 35% desferrioxamine; ferritin fell by 11% per year (p<0.001); at 5–6 years (n=28), 43% had ferritin <500 ng/mL and none >2,000; LIC was assessed in 31 of 64 (48%) at a median age of 3.2 years, with severe overload at first assessment in 45%, and none of the 22 patients with cardiac MRI had myocardial overload [98]. The study supports the feasibility, tolerability and efficacy of chelation before 3 years, below the age limits of the labels and without a comparator [98].

Recommendations of the French PNDS. The PNDS emphasises that, for the same transfusion support, patients with DBA accumulate more iron than other transfused patients, especially young children, who may have severe tissue overload, including cardiac, within a short time; chelation is discussed as soon as ferritin exceeds 500 µg/L, often before 2 years [9]. The targets are ferritin between 300 and 500 µg/L, liver iron concentration <36 µmol/g (about 2 mg/g with the consensus conversion factor, below the consensus target of 3 mg/g) and normal cardiac T2* [1, 9]. Desferrioxamine (Desferal®) is given at 40–60 mg/kg/day by subcutaneous infusion over at least 8 hours, ideally 12, often 5 days in 7 to favour adherence, with a pump supplied by a home care provider; it is the only chelator usable before 2 years, with a maximum of 30 mg/kg/day up to 3 years [9]. Deferasirox (Exjade®) is first choice above 2 years and in adults, at 14–28 mg/kg/day as a single daily dose, a range that in the SmPC corresponds to the film-coated tablets (with the dispersible tablet, 20–40 mg/kg/day) [99]; the general annex on chelation in aplasias, written for all bone marrow failures, places desferrioxamine as first choice and refers DBA to its own annex [9]. Deferiprone (Ferriprox®) is not recommended in DBA because of agranulocytosis (estimated at 10%); it is accepted only as third choice, in uncontrolled overload, after discussion at a multidisciplinary meeting and education of the patient and family about the infectious risk, with written instructions for any fever, in 3 doses and with a maximum of 75 mg/kg/day, below the label maxima (99 mg/kg/day in the USA; >100 mg/kg/day not recommended in the EU) [9, 92–94]. Ferritin is measured immediately before each transfusion, preferably every month; transferrin saturation is useful mainly at the start of the transfusion programme and to fine-tune chelation in a well-chelated patient, although it rarely falls below 60%; liver MRI is performed at the latest at 5–6 years and then every 12–24 months according to overload, and cardiac MRI at the first assessment and then according to the results [9]. An annual assessment looks for the endocrine, hepatic and cardiac complications of haemochromatosis [9]. In patients in remission or after HCT, phlebotomies replace chelators: 10–20 mL/kg per session (maximum 400–500 mL) every 15 days, with a blood count at each session and haemoglobin kept above 10 g/dL, adjusting volume and frequency to tolerance, and with the total number defined by ferritin and by liver and cardiac MRI [9].

7.6. Label doses and divergences from the consensus

In study 2 of the deferasirox label, 184 patients (85 with β-thalassaemia and 99 with other congenital or acquired anaemias, of whom 30 had DBA, 47 MDS and 22 others) received 5, 10, 20 or 30 mg/kg/day of the dispersible tablet according to baseline LIC, with an absolute reduction in LIC of −4.2 mg iron per gram dry weight at the end of the study; the label summary does not present the DBA subgroup result [91, 100]. The European SmPC for Exjade® mentions 411 adults and 292 children (52 aged 2–5 years) studied, including patients with DBA, and an iron excretion/intake ratio of 1.02 with 20 mg/kg and 1.67 with 30 mg/kg in β-thalassaemia [99].

7.7. Comparative regulatory status

No medicinal product has a specific authorisation for DBA. For chelators, DBA falls under the generic indication of transfusional iron overload or of "other anaemias", depending on the label [91, 99].

The search of the FDA orphan designation database, performed with the terms "Diamond" and "Blackfan", returned only these two designations; designations worded only as "congenital pure red cell aplasia" would not be captured [101]. National access (INFARMED, Brazilian Health Regulatory Agency [ANVISA]) was not verified.

7.8. Haematopoietic cell transplantation

HCT is the only curative therapy for erythroid failure, but it does not correct the malformations nor has it been shown to abolish the risk of solid tumours [1]. The 2024 consensus recommends HCT from an HLA-identical sibling (MSD, matched sibling donor) or from a 10/10 molecularly matched unrelated donor (MUD, matched unrelated donor) in transfusion-dependent children without adequate haemoglobin on a tolerable corticosteroid dose, generally before 10 years and ideally at preschool age (about 2–5 years) [1]. The 2008 consensus accepted only HLA-matched family donors; the inclusion of MUD as an equivalent alternative is one of the 2024 changes [1].

GeneReviews adds that HLA typing of the patient, siblings and parents should be performed at diagnosis and that a relative with the pathogenic variant is not a suitable donor, regardless of clinical status, because the marrow may fail or not engraft [2]. The 2021 recommendations of the European transplant group (EBMT) cover indications, pre-HCT assessment, choice of donor and source, conditioning, prophylaxis of rejection and GVHD, and follow-up; only the abstract was consulted, and the 2024 consensus incorporates them and refers to the EBMT for post-HCT follow-up [1, 105].

Series with reduced-toxicity or reduced-intensity conditioning (a US institutional series with distal alemtuzumab, fludarabine, melphalan and thiotepa, with median neutrophil engraftment at day +15, maximum grade II acute GVHD and all transfusion-free with donor chimerism at a median of 52 months; a St. Jude series) and the report of haploidentical HCT with post-transplant cyclophosphamide in a 21-year-old woman (with RPS19, transfusion-independent 5 years later, and a review of 4 published adult cases, all alive without severe GVHD) diverge from the consensus, which reserves these regimens and donors for trials or for the absence of an alternative [1, 111–113]. These data are small series or cases, without controlled comparison.

Recommendations of the French PNDS. The PNDS indicates HCT in any steroid-resistant transfusion-dependent child or one dependent on high corticosteroid doses (above 0.3 mg/kg/day), and in any patient with clonal evolution, whatever the age and type of donor; rarely, also in severe neutropenia due to RPL35A or in uncontrollable iron overload [9]. Results are better before 10 years: HCT should preferably be performed between 2 and 10 years, ideally before 5 (between 3 and 5 in the algorithm), and not before 2 years, except for early MDS, because of increased toxicity; above 10 years, disease-free and GVHD-free survival is lower, with excess GVHD perhaps linked to transfusion burden and alloimmunisation, and HCT is generally considered too risky in adults [9]. The donors accepted are the HLA-identical sibling in whom a silent form has been excluded (no macrocytosis, normal eADA and without the familial variant; in case of any doubt, an unrelated donor is preferred) and the 10/10 unrelated donor [9, 107]. The recommended conditioning is myeloablative, because the marrow of these patients is rich: busulfan adjusted to weight and to the SmPC, with pharmacokinetics and a target area under the curve of 900–1,350 µmol·min, from day −5 to day −2, fludarabine 40 mg/m²/day for 4 days (days −6 to −3) and rabbit antithymocyte globulin (Thymoglobuline®) 10 mg/kg total dose; alternatively, a reduced-toxicity regimen with treosulfan and fludarabine, with or without thiotepa, and the same antithymocyte globulin; total body irradiation has no indication [9]. GVHD prophylaxis combines a calcineurin inhibitor with a short course of methotrexate, or with mycophenolate mofetil when related cord blood is used; marrow is the preferred source, to reduce chronic GVHD; alternative donors (unrelated <10/10 or haploidentical) are discussed only in MDS or acute leukaemia, and unrelated cord blood gave unsatisfactory results in older studies [9]. In bone marrow failure, red cell concentrates are irradiated from the start of conditioning [9].

The PNDS text on treosulfan contains a unit error (g/kg instead of g/m²) and overlapping body surface area bands, and states that the drug had no paediatric authorisation outside malignant diseases; the annex of off-label medicines therefore includes it among the products used in conditioning [9]. The current European SmPC for Trecondi® (treosulfan; initial authorisation on 2019-06-20) indicates it, with fludarabine, for conditioning before allogeneic HCT in adults and in children older than 1 month with malignant and non-malignant diseases; in children, 10 g/m²/day if body surface area is <0.4 m², 12 g/m²/day between 0.4 and <0.9 m² and 14 g/m²/day from 0.9 m², as a 2-hour infusion on days −6, −5 and −4 (total 30–42 g/m²), with fludarabine 30 mg/m²/day on days −7 to −3 and, in the intensified regimen, thiotepa 5 mg/kg twice on day −2 [114]. The SmPC doses are the reference for treosulfan; the part of the PNDS on the indication is outdated.

Fertility before HCT. The PNDS recommends addressing fertility before any treatment and offering a reproductive medicine consultation to all HCT candidates, under the 2004 French bioethics law, even with conditioning of low gonadotoxicity, weighing the haemorrhagic and infectious risk and the delay to HCT [9]. Before puberty (before menarche, or before 12–13 years and Tanner stage 3 in boys), only cryopreservation of gonadal tissue is possible: ovarian cortex, used for about 20 years, with pregnancies described after reimplantation in women transplanted for non-malignant diseases; testicular tissue, experimental and without practical application. After puberty, sperm cryopreservation is offered to boys and ovarian stimulation for oocyte vitrification (or of embryos, in a couple) to girls; profound thrombocytopenia and neutropenia may contraindicate the collections, and preservation remains possible after HCT, which justifies specialised follow-up [9].

7.9. Other treatments and clinical trials

Leucine was proposed on the basis of rps19/rps14 zebrafish models and deficient human CD34+ cells, in which it improved the anaemia and developmental defects through activation of the mTOR pathway, an effect abolished by rapamycin [124]. The consensus considers that L-leucine may induce an erythroid response and linear growth in some transfusion-dependent patients, that some experts use it routinely, that it is not standard and that studies in steroid-sensitive patients are under way [1]. The trials are small, without control groups, with different populations (age, steroid resistance, genotype) and non-equivalent outcomes; they do not allow indirect comparison between interventions. The response rates are crude proportions without a comparator, and the negative or interrupted trials (bitopertin, sotatercept, trifluoperazine, lenalidomide, eltrombopag in most patients) are part of the evidence picture. The French PNDS considers corticosteroids the only drug treatment for the anaemia, declares erythropoietin, sotatercept and ciclosporin inactive, reserves eltrombopag for clinical trials and describes, with leucine, very rare haematological responses but a relatively frequent effect on height and weight gain, which may justify its use in children [9, 115]. Growth hormone treats the growth deficit associated with DBA; there are no data on increased cancer risk with growth hormone in DBA and studies in survivors of childhood cancer are reassuring; the panel considers it reasonable during corticosteroid "holidays" or before, preferably on transfusions, because corticosteroids may reduce its efficacy [1]. The French PNDS adds that growth hormone accelerated growth in the short term without data on adult height, that it is offered to children with the most severe delays within the usual indications, decided with paediatric endocrinology according to predicted adult height and expected gain, and that starting it is accompanied by stopping or reducing corticosteroids and by transfusion support [9].

7.10. Developments in the last 12 months and preclinical research

The developments in the window from October 2025 to October 2026 are as follows, by level of evidence. Clinical trials with published results: bitopertin, with no clinical responses in 15 adults [117]. Registered ongoing trials: APR-2020, the first RPS19 gene therapy trial in humans, recruiting since April 2026, and the DBAR CD34+ mobilisation study [11, 123]. Observational studies: chelation before 3 years in the French and Italian registries, haematological picture of 235 French adults, predictors of corticosteroid response in the Canadian registry, cord blood HCT (Eurocord) and systematic review of the role of iron in HCT [47, 76, 98, 109, 110]. Diagnostic research: methylation episignature and non-coding variants by genome and RNA sequencing [39, 40, 85]. Preclinical: "universal" gene therapy by erythroid-restricted expression of GATA1, which increased erythropoiesis in models of several genotypes without affecting stem cells or favouring premalignant clonal expansion; the St. Jude SJEFS-S19 lentiviral vector, which corrected rRNA processing, erythropoiesis and repopulation with polyclonal integration; spironolactone as an NLRP1 inflammasome inhibitor; SCD1 inhibition in an Rpl5 model [34, 35, 125, 126]. Regulatory: orphan designations for RPS19 gene therapy in the USA (2020) and the EU (2021), without marketing authorisation [101–103]. No clinical trial registration of GATA1 gene therapy was found up to October 2026, which does not prove that none exists.

7.11. Treatment of associated manifestations

Cleft lip or palate is treated by a craniofacial team; glaucoma and cataract by ophthalmology; thumb anomalies by orthopaedic surgery with occupational and physical therapy; urogenital anomalies by nephrology or urology; heart defects by cardiology [2]. Cancer is treated in oncology with cautious chemotherapy, because of the prolonged cytopenia and toxicities described in DBA [1, 2]. Immunodeficiency requires measurement of IgG, IgA and IgM and lymphocyte subsets when indicated, assessment of vaccine responses, discussion of immunoglobulin and, in severe T lymphopenia, prophylaxis against Pneumocystis jirovecii pneumonia [1]. Vaccines have no restrictions: hepatitis B vaccine is a priority in transfused patients; the first dose of live vaccines should ideally precede the start of prednisone and subsequent doses are given after corticosteroid reduction; with significant hypogammaglobulinaemia, vaccine antibody titres are measured; human papillomavirus vaccine is recommended [1]. For constitutional bone marrow failures, the French PNDS recommends pneumococcal, meningococcal, Haemophilus influenzae and influenza vaccines, if not already given in the national schedule; antifungal prophylaxis is decided case by case, according to the severity of neutropenia; G-CSF is not routine and is reserved for short courses in severe infections; erythropoietin is useless without chronic kidney disease [9].

7.12. Agents and situations to avoid

Deferiprone should be avoided, because of the risk of neutropenia and agranulocytosis, except in severe cardiac overload or as third line [1, 2]. With deferiprone, medicines known to cause neutropenia or agranulocytosis should not be combined [93]. Transfusions with a restrictive threshold (6–7 g/dL) are considered harmful [1]. Live vaccines should be avoided during the initial high dose of corticosteroids [1]. Corticosteroids should not be used without an established diagnosis, nor at doses above 2 mg/kg/day, nor prolonged beyond 4 weeks at the initial dose without response [1]. Deferasirox may render hormonal contraceptives ineffective, so non-hormonal methods are recommended [91, 99]. Irradiation in conditioning and unmanipulated mobilised peripheral blood as a cell source should be avoided [1]. Whole-body MRI is not recommended for cancer screening [1].

7.13. Pregnancy, contraception and breastfeeding

Pregnancy in a woman with DBA is high risk and should be followed by high-risk obstetrics and haematology with experience in bone marrow failure [2, 127]. In the survey of 64 pregnancies from the French and German registries, there were vasculoplacental complications in 66% of pregnant women, with intrauterine growth restriction, pre-eclampsia, placental abruption, intrauterine fetal death and preterm delivery [1, 128]. The 2008 consensus also describes worsening of anaemia with loss of steroid sensitivity or increased transfusion requirement, accelerated haemochromatosis during the chelation break, miscarriage, caesarean section and recurrence of DBA in the fetus, including severe forms; the hormonal stress of pregnancy appears to contribute to a "relapse" that may be transient, and there are no prospective studies [30]. The 2024 consensus recommends keeping nadir haemoglobin above 10.0–10.5 g/dL, as in thalassaemia, so many patients, including corticosteroid responders, need transfusions in pregnancy; corticosteroids are stopped and treatment becomes transfusional; chelation should be intensified before conception, because chelators are contraindicated in pregnancy (thalassaemia data suggest that desferrioxamine may be considered in the last trimester in women with severe overload); aspirin prophylaxis may be considered because of placental vasculopathy; and screening for fetal anaemia is recommended [1]. GeneReviews mentions low-dose aspirin up to 37 weeks in women with a previous complicated pregnancy; the dose is not specified in the sources consulted [2, 128].

The French PNDS recommends follow-up in a level 3 maternity unit and haemoglobin always above 10.5 g/dL, by analogy with thalassaemia, which often requires starting transfusions (including in women in treatment independence with low haemoglobin, and when the corticosteroid dose needed would be excessive for the fetus) or increasing support in those already transfused [9]. Pregnancy should be anticipated with a period of intensified chelation, because chelators are contraindicated, and overload is reassessed at the end of pregnancy to decide on resuming chelation [9]. Ultrasound surveillance looks for severe fetal malformations and fetal anaemia with hydrops, rare but perhaps underestimated, which may require in utero transfusions; a blood count with reticulocytes and genetic testing on cord blood, or on a fetal blood sample before any transfusion, are justified; and maternal surveillance focuses on vasculoplacental complications [9].

The chelator labels advise against breastfeeding: desferrioxamine during treatment and for up to one week after the last dose, deferiprone during treatment and for up to 2 weeks afterwards, and deferasirox during treatment [91, 92, 97, 99]. In the EU, deferiprone is contraindicated in pregnancy and breastfeeding; the US label recommends a pregnancy test before starting and effective contraception during treatment and for 6 months afterwards in women and 3 months afterwards in men with a partner of reproductive potential [92, 93]. Oral contraception with oestrogens may reduce the response to corticosteroids [1]. No reduction in fertility has been described in men with DBA; infertility may result from iron overload and HCT [1].

7.14. Perioperative context and chemotherapy

The sources consulted do not describe a DBA-specific perioperative protocol. The applicable elements are optimisation of haemoglobin by transfusion, endocrinological assessment for adrenal insufficiency in patients on prolonged corticosteroids or tapering, and increased haematological toxicity, with slow recovery, after chemotherapy or radiotherapy, which warrants caution in oncological treatment [1, 2].

7.15. Transition to adult care

In adults, the goals come to include adjustment of anaemia therapy (many adults need higher haemoglobin for daily life and work), monitoring of toxicity and comorbidities, cancer screening, family planning and pregnancy surveillance, with specific tests for adrenal insufficiency and hypogonadism, psychosocial support and baseline marrow assessment before transition [1].

The French PNDS treats transition as a period of high risk of disruption of care and proposes a three-phase programme, with identified contacts in the adult team [9]. In the preparatory phase, in paediatrics and well before the move, the adolescent's autonomy is promoted, with part of the consultation without the parents, the adult physician is designated early (taking into account the patient's gender preference) and a summary of the record is prepared for the patient and the new physician. In the transfer phase, between 16 and 20 years according to maturity and preferably during a stable period of the disease, a first joint consultation is held, consultations alternate during the first year and a liaison nurse is involved, with a visit to the adult service. In the reception phase, the first adult consultation is extended, introduces the organ specialists involved, assesses knowledge of the disease and lifestyle habits and avoids non-essential investigations [9].

Therapeutic education, according to the 2014 HAS framework cited by the PNDS, is a continuous process integrated into care, which should ensure that the patient and family understand the disease and make known the patient organisations and the patient days organised by the associations linked to the reference centre and by the Francophone Society of Bone Marrow Transplantation and Cell Therapy (SFGM-TC) [9]. The reference centre distributes patient booklets (updated in October 2022), the SFGM-TC a national post-HCT follow-up booklet, and nursing and clinical pharmacist consultations are part of the programme [9].

7.16. Guideline recommendations and divergences

On most points, the 2023 French PNDS, published one year before the international consensus, follows the same direction, with its own thresholds and intervals; where they diverge, this article follows the 2024 international consensus as the reference and presents the PNDS values as French practice [1, 9]. For treosulfan, the reference dose is that of the European SmPC, because the PNDS contains a unit error [9, 114].

7.17. Management of the other ribosomopathies

In cartilage-hair hypoplasia, live vaccines are contraindicated when cellular immunity in vitro is impaired, varicella and zoster are treated with aciclovir, deformities with osteotomies, and HCT corrects severe immunodeficiency but not the bone; chelation was effective and well tolerated in transfused patients, and corticosteroids were effective in some patients with anaemia, without sufficient data to recommend them (immunosuppression, growth deficit) [7, 63]. In isolated congenital asplenia, antibiotic prophylaxis and pneumococcal vaccination are the basis of prevention [82]. In 5q− syndrome, treatment follows the standards for adult MDS, outside the scope. The treatments of Treacher Collins and Shwachman-Diamond syndromes, Fanconi anaemia, TAR syndrome and the telomere diseases are specific to each entity. In the RNA polymerase III leukodystrophies, treatment is symptomatic [12].

8. Reproductive Options and Family Counselling

8.1. Recurrence risks by inheritance pattern

In autosomal dominant DBA, about 40–45% of probands inherited the variant from a parent, with or without manifestations, and 55–60% have a de novo variant; the 2024 consensus estimates two thirds of cases to be de novo or sporadic and one third familial [1, 2, 129]. Each child of a person with the variant has a 50% probability of inheriting it; penetrance is high but variable according to the variant, and severity in the child is unpredictable, from atypical forms to fetal hydrops [1, 2]. The siblings of a proband have a 50% risk if a parent is affected or a carrier; if the variant is not detected in the parents' leucocyte DNA, the risk is slightly higher than that of the population, because of gonadal mosaicism or a false negative due to preferential loss, by somatic reversion, of the chromosome carrying the variant; if the parents are clinically unaffected but untested, the risk is increased [2]. An apparently isolated case may result from failure to recognise the disease in a parent, reduced penetrance, early death or late onset [2]. The French PNDS (National Diagnosis and Care Protocol) estimates 55% of cases to be sporadic, describes no anticipation and emphasises that severity in a child cannot be predicted from the gene, the variant or the phenotype of the parent or of an affected sibling; a parent with the variant or with raised eADA has a 50% risk in each pregnancy, without it being possible to exclude a severely affected child, and genetic counselling should be offered to all parents and adult patients [9].

The 2008 consensus guides cases without an identified variant: in a family with classic DBA in a parent and a child, or in two or more siblings, the recurrence risk reaches 50%; in a sporadic case, the risk depends on the presence of a variant or of manifestations in first-degree relatives; if asymptomatic first-degree relatives have raised eADA, HbF or MCV, a 50% risk should be considered; if the proband's variant is absent in both parents, it is probably new, with a residual risk linked to gonadal mosaicism or non-paternity [30].

In the autosomal recessive form due to HEATR3, the parents are presumed heterozygous and should both be tested; if only one is a carrier, the second variant may be de novo or post-zygotic in a mosaic parent, and apparent homozygosity may correspond to an exon deletion not detected by sequencing or to uniparental isodisomy. Siblings have a 25% probability of being affected, 50% of being asymptomatic carriers and 25% of not being carriers; the children of a patient are obligate carriers, unless the partner is a carrier; the parents' siblings have a 50% probability of being carriers [2].

In the X-linked forms (GATA1, TSR2), the father of an affected boy has neither the disease nor the variant; a mother with more than one affected son is an obligate heterozygote and, if the variant is not detected in her leucocyte DNA, gonadal mosaicism is likely; in an isolated case, the mother may be heterozygous, the variant may be de novo or there may be maternal mosaicism. A heterozygous mother transmits the variant in 50% of pregnancies: boys who inherit it are affected and girls are heterozygous, generally unaffected; an affected man transmits it to all his daughters and to none of his sons; maternal aunts and female cousins may be at risk [2]. In GATA1 cytopenia, heterozygous women may have menorrhagia or easy bruising and two platelet populations owing to random X inactivation [6].

8.2. Cascade testing and assessment of relatives

After identification of the pathogenic variant, family studies are recommended, because penetrance is variable [1]. Asymptomatic relatives, older and younger, undergo molecular testing when the familial variant is known; if it is not known, MCV, eADA and/or HbF are assessed, especially in candidate marrow donors; carriers have a blood count and, if anaemic, a full assessment [2, 30]. A normal blood count does not exclude the disease in a relative, as shown by RPL26 carriers without haematological involvement [4, 50]. Carriers without manifestations are classified as "DBA syndrome currently without phenotype" and remain under surveillance, because anaemia or cancer may develop [1]. In the RPS7 family of the 2025 series, cascade testing identified maternal carriers with variable anaemia and raised eADA, co-segregating with the variant, and allowed an HLA-identical sibling without the variant to be chosen as donor [40].

8.3. Germline and revertant mosaicism

Gonadal mosaicism, or somatic and gonadal mosaicism, in a parent has been demonstrated in a family with two affected siblings and parents without the variant in blood and, more recently, in a family published in 2025 (consulted by title) [2, 56, 57]. Revertant mosaicism in a parent may give a false negative in leucocyte DNA and delay molecular confirmation [2, 55].

8.4. Prenatal diagnosis, preimplantation testing and pregnancy in at-risk couples

Once the familial variant has been identified, prenatal diagnosis and preimplantation genetic testing are possible; the consensus recommends discussing medically assisted reproduction with those who request them, according to national legislation, and bearing in mind that the severity of the disease in the child cannot be predicted [1, 2]. In an at-risk pregnancy, screening for fetal anaemia is part of follow-up [1]. In France, preimplantation diagnosis requires the variant to be identified and classified as pathogenic and to be analysable in the DNA of a single embryo cell, and dual selection of an unaffected embryo that is HLA-identical to an affected sibling is no longer practised [9]. The 2008 consensus recommended collecting cord blood from subsequent pregnancies in families with a proband, as a source of stem cells; the 2024 consensus considers cord blood from a healthy matched sibling a good cell source, provided DBA has been excluded in the donor [1, 30, 109].

8.5. Pre-pregnancy counselling and counselling of young adults

Family planning before pregnancy and counselling of affected or at-risk young adults are part of follow-up, including optimisation of chelation before conception and the obstetric risks (section 7.13) [1, 2]. In probands without a molecular diagnosis, storing DNA is recommended [2].

8.6. Testing of minors

Molecular testing of asymptomatic minor siblings has direct utility in the choice of an HCT donor and in haematological and oncological surveillance; ClinGen classifies the actionability of DBA as strong in the paediatric context [1, 2, 4]. The sources consulted do not describe DBA-specific rules for consent in minors beyond these indications.

8.7. Other ribosomopathies

Cartilage-hair hypoplasia is autosomal recessive: the siblings of a patient have a 25% risk, and in founder populations the carrier frequency is 1 in 10 among the Old Order Amish and 1 in 76 in Finland [7]. In isolated congenital asplenia (autosomal dominant), penetrance was complete in the 8 families of 2013 but incomplete for 6 of 18 coding variants (33%) and for the two 5′UTR variants described in 2018, and the variant was de novo in 7 of 23 probands tested; therefore, the relatives of any proband should have abdominal ultrasound and a blood film (Howell-Jolly bodies), even without an identified variant [4, 32, 64]. In Treacher Collins syndrome, expressivity is highly variable within the same family [59].

9. Special Considerations: Population Diversity and Epidemiology

9.1. Incidence and prevalence: what each figure measures

Incidence estimates refer mainly to classic DBA diagnosed through anaemia; individuals with "DBA syndrome currently without phenotype", undiagnosed adults and forms without anaemia are not included, so the frequency of the syndrome in the 2024 sense is unknown [1]. The unpublished British analysis cited by the consensus, in which only 77% of patients had anaemia in early childhood, illustrates this underestimation [1].

9.2. Data by region

The North American cohorts are DBAR (a prospective registry since 1991, 608 patients in the 2012 neoplasia analysis; registration NCT00106015, since 2004, with a target of 900 participants) and the NCI inherited bone marrow failure syndromes cohort (121 patients with DBA in the 2024 genotype-phenotype analysis; study NCT00027274) [10, 46, 72]. In Canada, the Canadian Inherited Marrow Failure Registry provided 80 evaluable patients for corticosteroid response [47].

In Europe, data come from national registries. In the United Kingdom, Ball and colleagues analysed 80 cases from a 20-year birth cohort and Orfali and colleagues described the clinical and genetic heterogeneity; a British series of 107 unselected cases characterised the immunodeficiency [1, 66, 129]. In France, the French registry provided data on growth (2005), agranulocytosis with deferiprone (2022), treatment independence (222 treated patients) and the haematological picture of 235 adults (2026), and there is a national bone marrow failure registry (NCT04781790) [1, 69, 76, 94]. The German and French registries provided the HCT data (70 children) and pregnancy data (64 pregnancies), and the German registry the data on DADA2 among patients without a genetic diagnosis (6%) [1, 107, 128]. In Italy, the series of Ramenghi, Quarello and colleagues described the RPL5 and RPL11 correlations and the MLPA deletions, and the French and Italian registries chelation before 3 years [37, 44, 98]. In the Czech Republic, Cmejla and colleagues studied RPL5, RPL11 and the RPS19 pseudogenes; in the Netherlands, there is a national paediatric cohort cited by Smith's [3, 45, 56]. The European RPL15 consortium and the EBMT (106 patients) complete the European data [48, 108].

Outside these regions, Japan contributed data on prednisolone and HCT and on extensive deletions; Egypt and India series of DBA and DADA2; Russia two patients with TP53; and Korea an RPL11 family [27, 38, 41, 65, 70, 89]. The 2024 consensus panel included a representative from Brazil (Ribeirão Preto, University of São Paulo) [1]. The sources consulted contain no epidemiological data on DBA in Portugal or Brazil, which does not prove that none exist.

9.3. Founder populations in the other ribosomopathies

In cartilage-hair hypoplasia, prevalence is 1–2 per 1,000 among the Old Order Amish (carriers 1 in 10) and 1 in 23,000 in Finland (carriers 1 in 76); the variant NR_003051.4:n.72A>G is present in 100% of Amish, 92% of Finnish and 48% of non-Finnish patients, and has an ancient founder origin common to both populations [7, 43, 130]. The original description of the disease, by McKusick in 1965, was made in the Amish, and the Finnish cohort of 108 patients (1992) and the extended follow-up of 2008 are the main series [63, 78]. In isolated congenital asplenia, three RPSA variants were each found in two independent families, suggesting a hotspot or founder effect [32]. In Shwachman-Diamond syndrome, gene conversion with the SBDSP pseudogene explains the recurrence of the same variants in different populations [61].

9.4. Special populations

Pregnant women with DBA have vasculoplacental complications in 66% of pregnancies and need nadir haemoglobin >10.0–10.5 g/dL (section 7.13) [1, 128]. Infants have their own thresholds for corticosteroids (from 12 months) and chelation (after the first failed corticosteroid trial, at a reduced dose) [1, 98]. Adults have a haematological picture with additional cytopenias, a cancer risk that increases with age and a late diagnosis that can be confused with MDS [1, 76]. Transplanted patients retain the risk of cancer and of residual iron overload [1].

9.5. Reference networks and patient organisations

The European reference network for rare haematological diseases is ERN-EuroBloodNet (https://eurobloodnet.eu/); no DBA-specific sheet or recommendation from this network was located during data collection. The EuroDBA consortium and DBAR (https://www.dbar.org/) coordinated the 2024 consensus [1]. The consensus appendix lists the patient organisations that took part: DBA Foundation (USA, dbafoundation.org), Daniella Maria Arturi Foundation (USA), DBAS UK (United Kingdom, diamondblackfan.org.uk, registered charity no. 1083179, with a helpline, a national patient registry being set up and a standard of care document in preparation), Diamond-Blackfan-Anämie-Selbsthilfegruppe (Germany, diamond-blackfan.de), Association Francophone de la Maladie de Blackfan-Diamond (France, afmbd.org), DBA Italia ODV (Italy, fondazionedba.org), Asociación Afectados Blackfan Diamond España (Spain, anaemiablackfandiamond.org.es), patient groups from Sweden, Austria, China, the Czech Republic, Finland and Denmark, DBA Canada (dbacanada.com) and Maddie's Vision (Australia, mrv.org.au) [1, 131]. GeneReviews also lists the National Organization for Rare Disorders (NORD) and the NCI inherited bone marrow failure syndromes study (marrowfailure.cancer.gov) [2]. The sources consulted did not identify organisations specific to Portugal or Brazil.

In France, DBA is managed by the reference centre for acquired and constitutional bone marrow aplasias (coordinated at Hôpital Saint-Louis, Paris, with Hôpital Robert-Debré as the constituent paediatric site) and by the network of competence centres, within the MaRIH health filière for rare immuno-haematological diseases (https://www.marih.fr/); the centre organises two national multidisciplinary meetings per month and has maintained, since 2017, the national observatory of bone marrow failures (RIME), with a biobank of blood, marrow and, in constitutional forms, fibroblasts [9]. Molecular diagnosis of DBA is concentrated in the reference laboratory at Hôpital Robert-Debré and eADA in the biochemistry laboratory at Hôpital Necker; genomes are performed under the France Genomic Medicine Plan [9]. Bone marrow aplasia may be recognised as an off-list long-term condition, with exemption from patient co-payment, through a care protocol drawn up by the attending physician with the specialist; the PNDS also identifies the medicines used off-label [9]. The Association Francophone de la Maladie de Blackfan-Diamond (AFMBD) took part in drafting the PNDS, the reference centre publishes information at https://www.aplasiemedullaire.com/ and Orphanet has a DBA-specific emergency sheet [9].

10. Illustrative Clinical Case

10.1. DBA due to a non-coding RPS7 variant diagnosed by genome reanalysis

Wen and colleagues described, in 2025, a boy assessed at 16 months, born small for gestational age (weight below the 3rd centile at 2 weeks of life) [40]. At 2 months, before any transfusion, he had haemoglobin of 4.4 g/dL, MCV of 106.7 fL and almost absent reticulocytes; he started red cell transfusions every 3 weeks. eADA was 2,138 mU/g haemoglobin (reference 400–900) and the marrow had 30% cellularity, with markedly reduced erythropoiesis and left shift. He had no congenital anomalies (normal renal ultrasound and echocardiogram) and development was normal. A maternal cousin had non-transfusion-dependent anaemia and reticulocytopenia since 2 years of age, and several maternal relatives had raised eADA [40].

The corticosteroid trial, at 12 months, did not achieve transfusion independence. An inherited bone marrow failure syndrome panel, exome and a clinical genome were negative. Reanalysis of the genome identified the maternal variant RPS7 NM_001011.4:c.-19G>C, at the last nucleotide of the first non-coding exon, absent from gnomAD; RNA sequencing showed intron retention up to c.-19+98 in about 20% of transcripts in the proband and the mother, absent in the father and in 11 controls, with 26 of 27 reads carrying the variant showing retention, and the variant, classified as a VUS by the clinical laboratory, was reclassified as pathogenic [40]. Cascade testing showed that the maternal relatives with variable anaemia and raised eADA carried the variant, and a 12/12 HLA-matched sibling without the variant was chosen as donor. After HCT, the patient had 100% donor chimerism at day +100 and, more than 14 months later, was asymptomatic, transfusion-independent and had a normal blood count [40].

Figure 1 of the article (https://pmc.ncbi.nlm.nih.gov/articles/PMC12144513/figure/fig1/), consulted, shows a photograph of the proband at 16 months, the genome read depth, the reads in the IGV viewer and the three-generation maternal pedigree with eADA values [40]. The management coincides with the 2024 consensus: corticosteroid trial from 12 months, transfusions with an adequate target haemoglobin and HCT from a matched sibling after the variant had been excluded in the donor [1].

10.2. Second case from the same series: deep intronic RPS19 variant with normal eADA

In the same series, a 20-month-old girl had severe anaemia from 6 weeks (haemoglobin 2.6 g/dL and MCV 110.3 fL before transfusion), growth restriction below the 3rd centile up to 7 months, pure red cell aplasia in the marrow and a fenestrated atrial septal defect, a muscular ventricular septal defect and a patent foramen ovale; she received transfusions every 3 weeks for nadir haemoglobin >9.0 g/dL [40]. eADA was normal (620 mU/g haemoglobin). The panel, chromosome breakage, karyotype and marrow microarray were negative; genome sequencing identified the de novo variant RPS19 NM_001022.4:c.172+350C>T, deep intronic, which creates a splicing donor and includes an 81 bp pseudo-exon with two stop codons in about half of the transcripts, absent in 13 controls; the patient was being prepared for HCT from a 12/12 sibling without the variant [40]. Figure 3 (https://pmc.ncbi.nlm.nih.gov/articles/PMC12144513/figure/fig3/), consulted, shows the pedigree with eADA, a photograph of the patient at 1 year, the position of the variant and the predicted donor gain [40]. The case illustrates that a normal eADA does not exclude DBA and that exomes and panels do not detect deep intronic variants [1, 40].

11. Prognosis and Long-Term Follow-up

11.1. Prognosis in figures

In DBAR, median survival was 56 years (2012 analysis) and the cumulative incidence of death in adults 23% at 45 years, mainly from complications of HCT, iron overload, infections and leukaemia or solid tumours [1, 10, 73]. In the French registry of 235 adults, overall survival was 98.0% at 30 years, 90.6% at 40 and 70.7% at 50, with cancer as the main cause of death [76]. The O/E ratio for cancer was 5.4 (DBAR 2012), with a cumulative incidence of solid tumour or leukaemia of about 20% at 46 years, and about 5 in the 2018 update [1, 10, 73]. About 20% of treated patients become treatment-independent (21% in the French registry), but the anaemia may return and the cancer risk persists [1]. After HCT, OS reaches 84–92% in recent European series, with better results before 10 years and after 2000 [1, 107, 108]. Iron overload and cancer are the main causes of death in non-transplanted patients [1]. These data come from registries of patients treated according to the practices of different periods and do not describe the natural history without treatment.

11.2. Surveillance

Whole-body MRI is not recommended [1]. The differences in interval between the French PNDS and the consensus are summarised in section 7.16; the PNDS also asks that patients and their physicians be informed of the cancer risk and of the symptoms that warrant consultation, and that they follow at least the general population screening programmes [9]. The consensus also recommends education on healthy lifestyle habits (avoiding tobacco, alcohol, toxic substances and unprotected sun exposure) and vaccination against human papillomavirus [1].

11.3. Follow-up after HCT and in patients without phenotype

After HCT, the EBMT follow-up recommendations apply, with attention to cancer risk, residual iron overload (chelation or phlebotomy) and infertility [1, 105]. Treatment-independent patients and carriers with "DBA syndrome currently without phenotype" continue with an annual blood count, colorectal cancer screening and genetic counselling, because the anaemia may return and the cancer risk persists [1].

12. Summary and Final Considerations

Diamond-Blackfan anaemia syndrome is the most frequent ribosomopathy, with an incidence of 5–10 cases per million live births. In most patients it results from haploinsufficiency of one of 24 ribosomal proteins or of the chaperones TSR2 and HEATR3, and more rarely from GATA1 variants or gain-of-function TP53 variants; 20–30% of patients have no identified genetic cause [1]. Ribosomal stress activates p53 and reduces translation of GATA1, which explains the erythroid failure. Macrocytic anaemia with reticulocytopenia appears in 90% of patients in the first year of life, but the spectrum ranges from fetal hydrops to absence of anaemia, and more than half of patients have congenital anomalies, mainly craniofacial, thumb, cardiac and urogenital [1, 2].

The 2024 international consensus simplified the diagnosis (pathogenic variant or compatible phenotype with exclusion of the differentials), adopted the designation "DBA syndrome" and replaced "remission" with "treatment independence" and "silent carrier" with "currently without phenotype" [1]. eADA and HbF sampled before transfusion support the diagnosis, but a normal eADA does not exclude it. Genetic testing must cover deletions, and a negative panel or exome result does not exclude non-coding variants, which genome and RNA sequencing may reveal [1, 40]. TEC, parvovirus B19, ADA2 deficiency, the other inherited bone marrow failure syndromes and, in adults, 5q− syndrome are the main differentials [1].

Treatment rests on transfusions with pre-transfusion haemoglobin ≥9–10 g/dL, corticosteroids from 12 months (2 mg/kg/day, maximum 80 mg/day, maintenance ≤0.3 mg/kg/day), early MRI-guided chelation and HCT from a matched sibling or a 10/10 unrelated donor before 10 years [1]. No medicinal product has a specific authorisation for DBA: deferiprone is used off-label in DBA in the USA and the EU, deferasirox below 2 years and desferrioxamine below 3 years are used outside their labels, and corticosteroids have a generic indication for congenital hypoplastic anaemia on the US labels [90–93, 97]. L-leucine produced a response in 16% in an uncontrolled trial; eltrombopag, sotatercept and bitopertin showed no clinical efficacy; RPS19 gene therapy entered a phase 1 trial in 2026 [11, 115–118].

Predisposition to MDS, AML, colorectal cancer and osteosarcoma (O/E ratio of 5.4 for all cancers) and iron overload determine the prognosis in adults, with median survival of 56 years in DBAR and survival of 70.7% at 50 years in the French registry; surveillance includes a blood count every 3–4 months, annual liver and cardiac MRI from 5 years in transfused patients, monitoring of drug toxicity and colonoscopy from 20 years, every 5 years [1, 10, 76]. In France, the 2023 PNDS follows the same direction, with its own thresholds (pre-transfusion haemoglobin of 9.0–9.5 g/dL, ideal corticosteroid maintenance <0.15 mg/kg/day, liver iron target <36 µmol/g, quarterly blood count and annual immunoglobulins in all patients) [9]. Genetic counselling must take into account variable penetrance, unpredictable severity, gonadal and revertant mosaicism and exclusion of the variant in sibling donors [1, 2]. The other ribosomopathies (5q− syndrome, Treacher Collins syndrome, Shwachman-Diamond syndrome, cartilage-hair hypoplasia, isolated congenital asplenia, dyskeratosis congenita and RNA polymerase III diseases) share the disruption of ribosome biogenesis but differ in the tissue affected, inheritance and treatment [12].

13. References

1. Wlodarski MW, Vlachos A, Farrar JE, Da Costa LM, Kattamis A, Dianzani I, et al. Diagnosis, treatment, and surveillance of Diamond-Blackfan anaemia syndrome: international consensus statement. Lancet Haematol. 2024;11(5):e368-e382. doi:10.1016/S2352-3026(24)00063-2. PMID: 38697731. PMC12424227.
2. Sieff C. DBA Syndrome. 2009 Jun 25 [updated 2025 Jul 31]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2025. Available from: https://www.ncbi.nlm.nih.gov/books/NBK7047/
3. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Diamond-Blackfan anemia (Aase syndrome). p. 458–459.
4. GeneGuidance. Gene curation database (entries RPS19, RPL5, RPL11, RPS26, RPS10, RPL35A, RPS24, RPS17, RPS7, RPL26, RPL15, RPS28, RPS29, RPL9, RPL18, RPL27, RPS15A, RPS27, GATA1, TSR2, HEATR3, RPS14, RPSA, RMRP), including the ClinGen actionability (AC1045) and haploinsufficiency assertions. Accessed 2026-10-17.
5. Orphanet. Diamond-Blackfan anemia (ORPHA:124). https://www.orpha.net/en/disease/detail/124
6. Takasaki K, Kacena MA, Raskind WH, Weiss MJ, Chou ST. GATA1-Related Cytopenia. 2006 Nov 22 [updated 2023 Feb 16]. In: Adam MP, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2025. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1364/
7. Mäkitie O, Vakkilainen S. Cartilage-Hair Hypoplasia – Anauxetic Dysplasia Spectrum Disorders. 2012 Mar 15 [updated 2023 May 11]. In: Adam MP, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2025. Available from: https://www.ncbi.nlm.nih.gov/books/NBK84550/
8. Ulirsch JC, Verboon JM, Kazerounian S, Guo MH, Yuan D, Ludwig LS, et al. The Genetic Landscape of Diamond-Blackfan Anemia. Am J Hum Genet. 2018;103(6):930-947. doi:10.1016/j.ajhg.2018.10.027. PMID: 30503522. PMC6288280.
9. Haute Autorité de Santé. Protocole National de Diagnostic et de Soins (PNDS): Aplasies médullaires acquises et constitutionnelles. Saint-Denis La Plaine: HAS; 2023 (version of 1 March 2023, published on 2023-05-04; reference centre for acquired and constitutional bone marrow aplasias, filière MaRIH). Available from: https://www.has-sante.fr/jcms/p_3431517
10. Vlachos A, Rosenberg PS, Atsidaftos E, Alter BP, Lipton JM. Incidence of neoplasia in Diamond Blackfan anemia: a report from the Diamond Blackfan Anemia Registry. Blood. 2012;119(16):3815-9. doi:10.1182/blood-2011-08-375972. PMID: 22362038. PMC3335385.
11. Apriligen, Inc. Assessing the safety, tolerability, and efficacy of APR-2020 in pediatric and adolescent subjects with RPS19 deficient Diamond-Blackfan anemia. ClinicalTrials.gov identifier: NCT07476183 [Internet]. Bethesda (MD): National Library of Medicine; updated 2026-07-21. Available from: https://clinicaltrials.gov/study/NCT07476183
12. Ferreira CR, Darling A, Vockley J. Disorders of nucleic acid metabolism, tRNA metabolism and ribosomal biogenesis. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 39, section 39.3 Ribosomal Biogenesis. doi:10.1007/978-3-662-63123-2_39
13. Orphanet. Myelodysplastic syndrome associated with isolated del(5q) chromosome abnormality (ORPHA:86841). https://www.orpha.net/en/disease/detail/86841
14. Orphanet. Familial isolated congenital asplenia (ORPHA:101351). https://www.orpha.net/en/disease/detail/101351
15. Orphanet. Cartilage-hair hypoplasia (ORPHA:175). https://www.orpha.net/en/disease/detail/175
16. Fathi P, Jafari L, Fathi A, Ahmadi A, Karamlou Y, Behfar M, et al. Hematopoietic Stem Cell Transplantation Outcomes in Diamond-Blackfan Anemia Patients Based on Myeloablative Conditioning Regimen With or Without Total Body Irradiation: A Systematic Review and Meta-Analysis. EJHaem. 2025;6(3):e70043. doi:10.1002/jha2.70043. PMID: 40342615. PMC12061074.
17. Aase JM, Smith DW. Congenital anemia and triphalangeal thumbs: a new syndrome. J Pediatr. 1969;74:417–22.
18. Draptchinskaia N, Gustavsson P, Andersson B, Pettersson M, Willig TN, Dianzani I, et al. The gene encoding ribosomal protein S19 is mutated in Diamond-Blackfan anaemia. Nat Genet. 1999;21(2):169-75. doi:10.1038/5951. PMID: 9988267.
19. Farrar JE, Nater M, Caywood E, et al. Abnormalities of the large ribosomal subunit protein, Rpl35a, in Diamond-Blackfan anemia. Blood. 2008;112:1582–92. PMID 18535205.
20. Gazda HT, Sheen MR, Vlachos A, Choesmel V, O'Donohue MF, Schneider H, et al. Ribosomal protein L5 and L11 mutations are associated with cleft palate and abnormal thumbs in Diamond-Blackfan anemia patients. Am J Hum Genet. 2008;83(6):769-80. doi:10.1016/j.ajhg.2008.11.004. PMID: 19061985. PMC2668101.
21. Mirabello L, Khincha PP, Ellis SR, Giri N, Brodie S, Chandrasekharappa SC, et al. Novel and known ribosomal causes of Diamond-Blackfan anaemia identified through comprehensive genomic characterisation. J Med Genet. 2017;54(6):417-425. doi:10.1136/jmedgenet-2016-104346. PMID: 28280134. PMC12265046.
22. Wang R, Yoshida K, Toki T, Sawada T, Uechi T, Okuno Y, et al. Loss of function mutations in RPL27 and RPS27 identified by whole-exome sequencing in Diamond-Blackfan anaemia. Br J Haematol. 2015;168(6):854-64. doi:10.1111/bjh.13229. PMID: 25424902.
23. Sankaran VG, Ghazvinian R, Do R, Thiru P, Vergilio JA, Beggs AH, et al. Exome sequencing identifies GATA1 mutations resulting in Diamond-Blackfan anemia. J Clin Invest. 2012;122(7):2439-43. doi:10.1172/JCI63597. PMID: 22706301. PMC3386831.
24. Gripp KW, Curry C, Olney AH, Sandoval C, Fisher J, Chong JX, et al. Diamond-Blackfan anemia with mandibulofacial dystostosis is heterogeneous, including the novel DBA genes TSR2 and RPS28. Am J Med Genet A. 2014;164A(9):2240-9. doi:10.1002/ajmg.a.36633. PMID: 24942156. PMC4149220.
25. Ludwig LS, Gazda HT, Eng JC, Eichhorn SW, Thiru P, Ghazvinian R, et al. Altered translation of GATA1 in Diamond-Blackfan anemia. Nat Med. 2014;20:748–53. PMID 24952648.
26. O'Donohue MF, Da Costa L, Lezzerini M, Unal S, Joret C, Bartels M, et al. HEATR3 variants impair nuclear import of uL18 (RPL5) and drive Diamond-Blackfan anemia. Blood. 2022;139(21):3111-3126. doi:10.1182/blood.2021011846. PMID: 35213692. PMC9136880.
27. Fedorova D, Ovsyannikova G, Kurnikova M, Pavlova A, Konyukhova T, Pshonkin A, et al. De novo TP53 germline activating mutations in two patients with the phenotype mimicking Diamond-Blackfan anemia. Pediatr Blood Cancer. 2022;69(4):e29558. doi:10.1002/pbc.29558. PMID: 35084091.
28. Moisan R, Da Costa L, Nguyen LD, Chosseler H, Riquet A, Bruno B, et al. Germline TP53 Mutations Causing Diamond-Blackfan Anemia: A French Report. Pediatr Blood Cancer. 2026;73(7):e70363. doi:10.1002/1545-5017.70363. PMID: 42104732.
29. Da Costa L, Leblanc T, Mohandas N. Diamond-Blackfan anemia. Blood. 2020;136(11):1262-1273. doi:10.1182/blood.2019000947. PMID: 32702755. PMC7483438.
30. Vlachos A, Ball S, Dahl N, Alter BP, Sheth S, Ramenghi U, et al. Diagnosing and treating Diamond Blackfan anaemia: results of an international clinical consensus conference. Br J Haematol. 2008;142(6):859-76. doi:10.1111/j.1365-2141.2008.07269.x. PMID: 18671700. PMC2654478.
31. Ebert BL, Pretz J, Bosco J, Chang CY, Tamayo P, Galili N, et al. Identification of RPS14 as a 5q- syndrome gene by RNA interference screen. Nature. 2008;451(7176):335-9. doi:10.1038/nature06494. PMID: 18202658. PMC3771855.
32. Bolze A, Boisson B, Bosch B, Antipenko A, Bouaziz M, Sackstein P, et al. Incomplete penetrance for isolated congenital asplenia in humans with mutations in translated and untranslated RPSA exons. Proc Natl Acad Sci U S A. 2018;115(34):E8007-E8016. doi:10.1073/pnas.1805437115. PMID: 30072435. PMC6112730.
33. Doty RT, Munday AD, Cottnair H, Funk SE, Young DJ, Dunbar CE, et al. Restricting glycine uptake with bitopertin improves erythropoiesis in preclinical models of Diamond-Blackfan anemia. Blood Red Cells Iron. 2025;1(2):100010. doi:10.1016/j.brci.2025.100010. PMID: 42182038. PMC13197043.
34. Rodríguez-Ruiz L, Lozano-Gil JM, Ocaña-Esparza M, Conesa-Hernández AM, Pérez-Oliva AB, Fuster JL, et al. Spironolactone inhibits the NLRP1 inflammasome and alleviates defective erythropoiesis in Diamond-Blackfan anemia. Hemasphere. 2025;9(4):e70131. doi:10.1002/hem3.70131. PMID: 40248129. PMC12005213.
35. Deng K, Wang Y, Min JC, Liu X, Ueharu H, Myers G, et al. Lipid metabolism transcriptomic signature of defective erythropoiesis in Diamond Blackfan anemia syndrome. Blood Adv. 2026;10(7):2272-2285. doi:10.1182/bloodadvances.2025017318. PMID: 41576344. PMC13059010.
36. Gianferante MD, Wlodarski MW, Atsidaftos E, Da Costa L, Delaporta P, Farrar JE, et al. Genotype-phenotype association and variant characterization in Diamond-Blackfan anemia caused by pathogenic variants in RPL35A. Haematologica. 2021;106(5):1303-1310. doi:10.3324/haematol.2020.246629. PMID: 32241839. PMC8094096.
37. Quarello P, Garelli E, Brusco A, Carando A, Mancini C, Pappi P, et al. High frequency of ribosomal protein gene deletions in Italian Diamond Blackfan anemia patients detected by multiplex ligation-dependent probe amplification (MLPA) assay. Haematologica. 2012;97:1813–7. PMID 22689679.
38. Kuramitsu M, Sato-Otsubo A, Morio T, et al. Extensive gene deletions in Japanese patients with Diamond Blackfan anemia. Blood. 2012;119:2376–84. PMID 22262766.
39. Quarello P, Karimi K, Trajkova S, Garelli E, Samadieh M, Iovino E, et al. DNA Methylation Episignature as a Novel Diagnostic Tool for Diamond-Blackfan Anemia Syndrome. Am J Hematol. 2026;101(2):228-241. doi:10.1002/ajh.70141. PMID: 41247002. PMC12766366.
40. Wen T, Boyden SE, Hocutt CM, Lewis RG, Baldwin EE, Vagher J, et al. Identification of 2 novel noncoding variants in patients with Diamond-Blackfan anemia syndrome by whole genome sequencing. Blood Adv. 2025;9(10):2443-2452. doi:10.1182/bloodadvances.2024015347. PMID: 40029997. PMC12144513.
41. Kim B, Kim HY, Woo S, Yoo KH, Kim HJ. Intrafamilial variability in Diamond-Blackfan anemia with a novel canonical splice-site variant in the RPL11 gene. Lab Med. 2025;56(6):795-798. doi:10.1093/labmed/lmaf040. PMID: 40720759.
42. Leshchynska I, Das D, O'Reilly V, Sipka A, Iyer K, Alankarage D, et al. A three generation family with VACTERL association is found to have a rare form of diamond-blackfan anaemia. Eur J Hum Genet. 2026;34(5):619-629. doi:10.1038/s41431-026-02076-z. PMID: 41851260. PMC13172565.
43. Ridanpää M, Sistonen P, Rockas S, Rimoin DL, Mäkitie O, Kaitila I. Worldwide mutation spectrum in cartilage-hair hypoplasia: ancient founder origin of the major70A-->G mutation of the untranslated RMRP. Eur J Hum Genet. 2002;10(7):439-47. doi:10.1038/sj.ejhg.5200824. PMID: 12107819.
44. Quarello P, Garelli E, Carando A, Brusco A, Calabrese R, Dufour C, et al. Diamond-Blackfan anemia: genotype-phenotype correlations in Italian patients with RPL5 and RPL11 mutations. Haematologica. 2010;95:206–13. PMID 19773262.
45. Cmejla R, Cmejlova J, Handrkova H, Petrak J, Petrtylova K, Mihal V, et al. Identification of mutations in the ribosomal protein L5 (RPL5) and ribosomal protein L11 (RPL11) genes in Czech patients with Diamond-Blackfan anemia. Hum Mutat. 2009;30:321–7. PMID 19191325.
46. Gianferante DM, Mendez KJW, Cole S, Gadalla SM, Alter BP, Savage SA, et al. Genotype-phenotype associations in individuals with Diamond Blackfan anaemia. EJHaem. 2024;5(6):1117-1124. doi:10.1002/jha2.1031. PMID: 39691264. PMC11647742.
47. Choi YB, Mathew R, Li H, Cada M, Klaassen RJ, Villeneuve S, et al. Steroid Responsiveness and Clinical Outcomes in Diamond-Blackfan Anemia: Analysis From the Canadian Inherited Marrow Failure Registry. Eur J Haematol. 2026;117(1):267-273. doi:10.1111/ejh.70187. PMID: 41948977.
48. Wlodarski MW, et al. Recurring mutations in RPL15 are linked to hydrops fetalis and treatment independence in Diamond-Blackfan anemia. Haematologica. 2018. PMID 29599205.
49. Dashraath P, et al. Non-immune hydrops fetalis caused by Diamond-Blackfan anaemia and a mutation of the RPL15 gene. Lancet. 2023. PMID 37422302.
50. Karger LM, Webb BD, Edelmann L, Liao J, Mehta L. Familial RPL26 Variant Causing Congenital Anomalies Without Hematological Features of Diamond Blackfan Anemia. Am J Med Genet A. 2025;197(5):e63954. doi:10.1002/ajmg.a.63954. PMID: 39710607.
51. Zheng H, Wang K, Wang J, Fang J, Zhou D, Qiu K. Neonatal Diamond-Blackfan anemia with persistent neutropenia caused by an RPS15A variant. Pediatr Investig. 2026. doi:10.1002/ped4.70084. PMID: 42630679. PMC13495045.
52. Dhangar S, Shanmukhaiah C, Shah A, Chellasamy SK, George M, Ghatanatti J, et al. A Rare Variant in RPS26 Gene Causing Variable Phenotype of Diamond-Blackfan Anemia Type 10 (DBA10) in an Indian Family. Indian J Pediatr. 2026;93(2):208-209. doi:10.1007/s12098-025-05933-9. PMID: 41460469.
53. Garelli E, Quarello P, Giorgio E, et al. Spontaneous remission in a Diamond-Blackfan anaemia patient due to a revertant uniparental disomy ablating a de novo RPS19 mutation. Br J Haematol. 2019;185:994–8. PMID 30460677.
54. Venugopal P, Moore S, Lawrence DM, et al. Self-reverting mutations partially correct the blood phenotype in a Diamond Blackfan anemia patient. Haematologica. 2017;102:e506–e509. PMID 28971907.
55. Borja NA, Tekin M. Revertant Mosaicism Obscures Long-Awaited Molecular Confirmation of Diamond-Blackfan Anemia. Am J Med Genet A. 2026;200(10):2381-2383. doi:10.1002/ajmg.a.70209. PMID: 42174374.
56. Cmejla R, Blafkova J, Stopka T, Zavadil J, Pospisilova D, Mihal V, et al. Ribosomal protein S19 gene mutations in patients with diamond-blackfan anemia and identification of ribosomal protein S19 pseudogenes. Blood Cells Mol Dis. 2000;26:124–32. PMID 10753603.
57. Dunlea E, Lynch SA, Angelov D, Cotter M. Germline mosaicism in a family affected by Diamond-Blackfan anaemia. Br J Haematol. 2025;207(6):2629-2631. doi:10.1111/bjh.70173. PMID: 40976991.
58. Vlachos A, Farrar JE, Atsidaftos E, Muir E, Narla A, Markello TC, et al. Diminutive somatic deletions in the 5q region lead to a phenotype atypical of classical 5q- syndrome. Blood. 2013;122:2487–90. PMID 23943650.
59. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Treacher Collins syndrome (mandibulofacial dysostosis). p. 362–363.
60. Positional cloning of a gene involved in the pathogenesis of Treacher Collins syndrome. The Treacher Collins Syndrome Collaborative Group. Nat Genet. 1996;12(2):130-6. doi:10.1038/ng0296-130. PMID: 8563749.
61. Boocock GR, Morrison JA, Popovic M, Richards N, Ellis L, Durie PR, et al. Mutations in SBDS are associated with Shwachman-Diamond syndrome. Nat Genet. 2003;33(1):97-101. doi:10.1038/ng1062. PMID: 12496757.
62. Ridanpää M, van Eenennaam H, Pelin K, Chadwick R, Johnson C, Yuan B, et al. Mutations in the RNA component of RNase MRP cause a pleiotropic human disease, cartilage-hair hypoplasia. Cell. 2001;104(2):195-203. doi:10.1016/s0092-8674(01)00205-7. PMID: 11207361.
63. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. Cartilage-hair hypoplasia. p. 37–40.
64. Bolze A, Mahlaoui N, Byun M, Turner B, Trede N, Ellis SR, et al. Ribosomal protein SA haploinsufficiency in humans with isolated congenital asplenia. Science. 2013;340(6135):976-8. doi:10.1126/science.1234864. PMID: 23579497. PMC3677541.
65. Ohga S, Mugishima H, Ohara A, Kojima S, Fujisawa K, Yagi K, et al. Diamond-Blackfan anemia in Japan: clinical outcomes of prednisolone therapy and hematopoietic stem cell transplantation. Int J Hematol. 2004;79:22–30. PMID 14979474.
66. Ball SE, McGuckin CP, Jenkins G, Gordon-Smith EC. Diamond-Blackfan anaemia in the U.K.: analysis of 80 cases from a 20-year birth cohort. Br J Haematol. 1996;94:645–53. PMID 8826887.
67. Lipton JM, Atsidaftos E, Zyskind I, Vlachos A. Improving clinical care and elucidating the pathophysiology of Diamond Blackfan anemia: an update from the Diamond Blackfan Anemia Registry. Pediatr Blood Cancer. 2006;46:558–64. PMID 16317735.
68. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Radial ray defects and thumb hypoplasia. p. 296–299.
69. Chen S, Warszawski J, Bader-Meunier B, Tchernia G, Da Costa L, Marie I, et al. Diamond-Blackfan anemia and growth status: the French registry. J Pediatr. 2005;147:669–73. PMID 16291361.
70. Ragab I, Makkeyah S, Hassan N, Botros M, Da Costa L, Aly NH. Immunodeficiency in children with Diamond Blackfan and Diamond Blackfan like anemia. Blood Cells Mol Dis. 2025;111:102911. doi:10.1016/j.bcmd.2025.102911. PMID: 39923319.
71. Pezzoli F, Guarnieri V, Chiocca E, Pegoraro F, Veltroni M, Quarello P, et al. Persistent Severe Lymphopenia Identified by Newborn Screening Program for Inborn Errors of Immunity in a Child With Diamond-Blackfan Anemia Syndrome. Pediatr Blood Cancer. 2026;73(3):e70145. doi:10.1002/1545-5017.70145. PMID: 41555788.
72. Alter BP, Giri N, Savage SA, Rosenberg PS. Cancer in the National Cancer Institute inherited bone marrow failure syndrome cohort after fifteen years of follow-up. Haematologica. 2018;103:30–9. PMID 29051281.
73. Vlachos A, Rosenberg PS, Atsidaftos E, Kang J, Onel K, Sharaf RN, et al. Increased risk of colon cancer and osteogenic sarcoma in Diamond-Blackfan anemia. Blood. 2018;132(20):2205-2208. doi:10.1182/blood-2018-05-848937. PMID: 30266775. PMC6238192.
74. Lipton JM, Molmenti CLS, Hussain M, Desai P, Florento M, Atsidaftos E, et al. Colorectal cancer screening and surveillance strategy for patients with Diamond Blackfan anemia: Preliminary recommendations from the Diamond Blackfan Anemia Registry. Pediatr Blood Cancer. 2021;68(8):e28984. doi:10.1002/pbc.28984. PMID: 34089224.
75. Lipton JM, Molmenti CLS, Desai P, Lipton A, Ellis SR, Vlachos A. Early Onset Colorectal Cancer: An Emerging Cancer Risk in Patients with Diamond Blackfan Anemia. Genes (Basel). 2021;13(1):56. doi:10.3390/genes13010056. PMID: 35052397. PMC8774389.
76. Lecornec N, de Fontbrune FS, Forcade E, Garçon L, Terriou L, Cougoul P, et al. Hematologic Landscape of Adult Patients With Diamond-Blackfan Anemia Syndrome. Am J Hematol. 2026;101(4):687-696. doi:10.1002/ajh.70197. PMID: 41498485.
77. Yılmaz E, Özer NK, Topal E. Steroid Response Prediction in Children With Diamond-Blackfan Anemia Requires Age at Treatment, Not Diagnosis. Eur J Haematol. 2026. doi:10.1111/ejh.70326. PMID: 42812072.
78. Taskinen M, Ranki A, Pukkala E, et al. Extended follow-up of the Finnish cartilage-hair hypoplasia cohort confirms high incidence of non-Hodgkin lymphoma and basal cell carcinoma. Am J Med Genet A. 2008;146A:2370–5.
79. Mahlaoui N, et al. Isolated congenital asplenia: a French nationwide retrospective survey of 20 cases. J Pediatr. 2011. PMID 20846672.
80. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. Shwachman syndrome. p. 50–53.
81. Marie S, Dewulf JP, Nassogne MC. Disorders of purine and pyrimidine metabolism. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 32, section 32.3 (32.3.2 ADA2 deficiency; 32.3.5 ADA1 overactivity). doi:10.1007/978-3-662-63123-2_32
82. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Immunodeficiency and recurrent infection. p. 464–471.
83. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Laterality sequences. p. 866–869.
84. Correction to "DNA Methylation Episignature as a Novel Diagnostic Tool for Diamond-Blackfan Anemia Syndrome". Am J Hematol. 2026;101(8):2100. doi:10.1002/ajh.70378. PMID: 42178501.
85. Kanezaki R, Toki T, Okuse R, Kobayashi A, Sato M, Kimura S, et al. Functional impact of a deep intronic variant in the RPS19 gene detected in a case of Diamond-Blackfan anemia syndrome. Haematologica. 2026;111(5):1876-1880. doi:10.3324/haematol.2025.300131. PMID: 41568468. PMC13136793.
86. Orphanet. Transient erythroblastopenia of childhood (ORPHA:98871). https://www.orpha.net/en/disease/detail/98871
87. Greenhalgh KL, Howell RT, Bottani A, et al. Thrombocytopenia–absent radius syndrome: a clinical genetic study. J Med Genet. 2002;39:876–81.
88. Saudubray JM, García-Cazorla Á. Clinical approach to inborn errors of metabolism in paediatrics. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 1, section 1.6.5 Haematology (Table 1.37).
89. Nayak AR, Dass J, Parikh H, Tripathi S, Gudapati P, Verghese R, et al. Infantile Pure Red Cell Aplasia Secondary to Deficiency of Adenosine Deaminase2 (DADA2) Syndrome-Time to Think Beyond Diamond Blackfan Anemia. Pediatr Blood Cancer. 2025;72(6):e31656. doi:10.1002/pbc.31656. PMID: 40095445.
90. PredniSONE tablets (ANDA080352): prescribing information [Internet]. Bethesda (MD): DailyMed, National Library of Medicine; revised 2023-08-31 (set id 0060b86b-0b79-4c54-a7be-97683a933a05); and Prednisolone (ANDA215673), revised 2024-02-16 (set id 070f1937-50a5-457f-bef5-4e597014e26d). Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=0060b86b-0b79-4c54-a7be-97683a933a05
91. Novartis Pharmaceuticals Corporation. EXJADE (deferasirox) tablets for oral suspension: prescribing information [Internet]. Silver Spring (MD): U.S. Food and Drug Administration; revised 2026-01-28 (set id 3495a70c-870c-4968-940e-8baea152cf85, version 49). Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=3495a70c-870c-4968-940e-8baea152cf85
92. Chiesi USA, Inc. FERRIPROX (deferiprone) tablets: prescribing information [Internet]. Silver Spring (MD): U.S. Food and Drug Administration; revised 2026-01-28 (set id 946718ff-df43-1baa-74d1-e5bcfbe8ad86, version 15). Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=946718ff-df43-1baa-74d1-e5bcfbe8ad86
93. European Medicines Agency. Ferriprox (deferiprone): EPAR — product information [Internet]. Amsterdam: EMA; authorised 1999-08-25; latest EC decision 2026-04-20. Available from: https://www.ema.europa.eu/en/documents/product-information/ferriprox-epar-product-information_en.pdf
94. Lecornec N, Castex MP, Réguerre Y, Moreau P, Marie I, Garçon L, et al. Agranulocytosis in patients with Diamond-Blackfan anaemia (DBA) treated with deferiprone for post-transfusion iron overload: a retrospective study of the French DBA cohort. Br J Haematol. 2022;199:285–8. PMID 35852515.
95. Henter JI, Karlen J. Fatal agranulocytosis after deferiprone therapy in a child with Diamond-Blackfan anemia. Blood. 2007;109:5157–9. PMID 17344464.
96. Papneja K, Bhatt MD, Kirby-Allen M, Arora S, Wiernikowski JT, Athale UH. Fanconi Syndrome Secondary to Deferasirox in Diamond-Blackfan Anemia: Case Series and Recommendations for Early Diagnosis. Pediatr Blood Cancer. 2016;63(8):1480-3. doi:10.1002/pbc.25995. PMID: 27082377.
97. Novartis Pharmaceuticals Corporation. DESFERAL (deferoxamine mesylate) for injection: prescribing information [Internet]. Bethesda (MD): DailyMed, National Library of Medicine; published 2025-06-18 (set id a7174843-5965-49fc-b842-f7eff7b48bbc). Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=a7174843-5965-49fc-b842-f7eff7b48bbc
98. Torchio F, Lecalvez B, Garelli E, Mallebranche C, Carando A, Castex MP, et al. Management of Iron Overload in Infants and Toddlers With Diamond-Blackfan Anemia Syndrome: A French-Italian Study. Am J Hematol. 2026;101(8):1856-1865. doi:10.1002/ajh.70354. PMID: 42108919. PMC13331634.
99. European Medicines Agency. Exjade (deferasirox): EPAR — product information [Internet]. Amsterdam: EMA; authorised 2006-08-28; latest EC decision 2026-02-04. Available from: https://www.ema.europa.eu/en/documents/product-information/exjade-epar-product-information_en.pdf
100. Novartis Pharmaceuticals Corporation. JADENU (deferasirox) tablets; JADENU Sprinkle (deferasirox) granules: prescribing information [Internet]. Silver Spring (MD): U.S. Food and Drug Administration; revised 2025-12-11 (set id fee89140-fff1-4443-9f42-24ac004fcda1, version 26). Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=fee89140-fff1-4443-9f42-24ac004fcda1
101. U.S. Food and Drug Administration, Office of Orphan Products Development. Search Orphan Drug Designations and Approvals: designations with "Diamond" and "Blackfan" (CD34+ cells transduced with a lentiviral vector carrying a codon-optimized RPS19 sequence, Apriligen, 2020-10-21; Interleukin-3 human recombinant, Immunex, 1991-05-20, designation withdrawn or revoked) [Internet]. Silver Spring (MD): FDA; accessed 2026-10-17. Available from: https://www.accessdata.fda.gov/scripts/opdlisting/oopd/
102. European Medicines Agency. EU/3/21/2488: orphan designation for treatment of Diamond-Blackfan anaemia (autologous CD34+ cells transfected with a lentiviral vector containing codon optimised RPS19 gene; APR-2020) [Internet]. Amsterdam: EMA; designated 2021-08-20. Available from: https://www.ema.europa.eu/en/medicines/human/orphan-designations/eu-3-21-2488
103. European Medicines Agency. EU/3/21/2519: orphan designation for treatment of Diamond-Blackfan anaemia (autologous CD34+ enriched cells transduced with a self-inactivating lentiviral vector containing the codon-optimized RPS19 gene) [Internet]. Amsterdam: EMA; designated 2021-11-12. Available from: https://www.ema.europa.eu/en/medicines/human/orphan-designations/eu-3-21-2519
104. Giménez Y, Palacios M, Sánchez-Domínguez R, Zorbas C, Peral J, Puzik A, et al. Lentivirus-mediated gene therapy corrects ribosomal biogenesis and shows promise for Diamond Blackfan anemia. JCI Insight. 2024;9(10):e171650. doi:10.1172/jci.insight.171650. PMID: 38775150. PMC11141922.
105. Diaz-de-Heredia C, Bresters D, Faulkner L, Yesilipek A, Strahm B, Miano M, et al. Recommendations on hematopoietic stem cell transplantation for patients with Diamond-Blackfan anemia. On behalf of the Pediatric Diseases and Severe Aplastic Anemia Working Parties of the EBMT. Bone Marrow Transplant. 2021;56(12):2956-2963. doi:10.1038/s41409-021-01449-w. PMID: 34462566.
106. Roy V, Perez WS, Eapen M, Marsh JC, Pasquini M, Pasquini R, et al. Bone marrow transplantation for Diamond-Blackfan anemia. Biol Blood Marrow Transplant. 2005;11:600–8. PMID 16041310.
107. Strahm B, Loewecke F, Niemeyer CM, Albert M, Ansari M, Bader P, et al. Favorable outcomes of hematopoietic stem cell transplantation in children and adolescents with Diamond-Blackfan anemia. Blood Adv. 2020;4(8):1760-1769. doi:10.1182/bloodadvances.2019001210. PMID: 32343795. PMC7189291.
108. Miano M, Eikema DJ, de la Fuente J, Bosman P, Ghavamzadeh A, Smiers F, et al. Stem Cell Transplantation for Diamond-Blackfan Anemia. A Retrospective Study on Behalf of the Severe Aplastic Anemia Working Party of the European Blood and Marrow Transplantation Group (EBMT). Transplant Cell Ther. 2021;27(3):274.e1-274.e5. doi:10.1016/j.jtct.2020.12.024. PMID: 33781541.
109. Volt F, Akhoudas M, Kenzey C, Cappelli B, Scigliuolo GM, Cacace F, et al. Umbilical cord blood transplantation in children with Diamond-Blackfan anemia. Bone Marrow Transplant. 2026;61(6):705-710. doi:10.1038/s41409-026-02852-x. PMID: 41957272.
110. Kuppens GZL, Kiakou E, de Zwart L, Miano M, van Beers EJ, Bierings M, et al. Survival After Hematopoietic Stem Cell Transplantation in Diamond-Blackfan Anemia Syndrome: The Role of Iron Overload-A Systematic Review. Pediatr Blood Cancer. 2026;73(11):e70664. doi:10.1002/1545-5017.70664. PMID: 42720436.
111. Yu H, Anderson EJ, Schiff DE, Gloude NJ. Irradiation-Free Reduced Intensity Conditioning Stem Cell Transplantation for Young Patients With Diamond-Blackfan Anemia Syndrome Is Well-Tolerated and Effective. Pediatr Blood Cancer. 2026;73(3):e70007. doi:10.1002/pbc.70007. PMID: 41257311.
112. Qudeimat A, Suryaprakash S, Madden R, Srinivasan A, Wlodarski MW, Bhoopalan SV. Reduced toxicity conditioning for hematopoietic stem cell transplantation in children with Diamond-Blackfan anemia. Haematologica. 2024;109(10):3404-3407. doi:10.3324/haematol.2024.285147. PMID: 38813735. PMC11443391.
113. Sakakibara T, Kobayashi S, Ito T, Tachibana M, Ito E. Post-Transplant Cyclophosphamide-Based Related Haploidentical Transplantation for Adult Diamond-Blackfan Anemia: Long-Term Survival and Review. J Hematol. 2026;15(4):214-218. doi:10.14740/jh2220. PMID: 42730145. PMC13568743.
114. European Medicines Agency. Trecondi (treosulfan): EPAR — product information [Internet]. Amsterdam: EMA; authorised 2019-06-20; latest renewal 2024-01-05; text version consulted: December 2025. Available from: https://www.ema.europa.eu/en/documents/product-information/trecondi-epar-product-information_en.pdf
115. Vlachos A, Atsidaftos E, Lababidi ML, Muir E, Rogers ZR, Alhushki W, et al. L-leucine improves anemia and growth in patients with transfusion-dependent Diamond-Blackfan anemia: Results from a multicenter pilot phase I/II study from the Diamond-Blackfan Anemia Registry. Pediatr Blood Cancer. 2020;67(12):e28748. doi:10.1002/pbc.28748. PMID: 33025707. PMC8273758.
116. Duncan BB, Lotter JL, Superata J, Barranta ME, Machado T, Darden I, et al. Treatment of refractory/relapsed Diamond-Blackfan anaemia with eltrombopag. Br J Haematol. 2024;204(5):2077-2085. doi:10.1111/bjh.19357. PMID: 38462764. PMC11212774.
117. Young DJ, Machado TR, Cowling BC, Doty RT, Venugopal S, Wu CO, et al. Results from the first clinical trial of bitopertin for steroid-relapsed/refractory Diamond-Blackfan anemia. Blood Red Cells Iron. 2026;2(3):100090. doi:10.1016/j.brci.2026.100090. PMID: 42583025. PMC13459865.
118. Northwell Health. Safety and efficacy study of sotatercept in adults with transfusion dependent Diamond Blackfan anemia. ClinicalTrials.gov identifier: NCT01464164 [Internet]. Bethesda (MD): National Library of Medicine; results posted 2021-12-21. Available from: https://clinicaltrials.gov/study/NCT01464164
119. Vlachos A. The use of trifluoperazine in transfusion dependent DBA. ClinicalTrials.gov identifier: NCT03966053 [Internet]. Bethesda (MD): National Library of Medicine; results posted 2022-12-14. Available from: https://clinicaltrials.gov/study/NCT03966053
120. Taylor AM, Macari ER, Chan IT, Blair MC, Doulatov S, Vo LT, et al. Calmodulin inhibitors improve erythropoiesis in Diamond-Blackfan anemia. Sci Transl Med. 2020;12(566):eabb5831. doi:10.1126/scitranslmed.abb5831. PMID: 33087503. PMC7709865.
121. Erratum for the Research Article "Calmodulin inhibitors improve erythropoiesis in Diamond-Blackfan anemia". Sci Transl Med. 2026;18(835):eaef4957. doi:10.1126/scitranslmed.aef4957. PMID: 41637529.
122. Gotlib JR. Pilot lenalidomide in adult Diamond-Blackfan anemia patients w/ RBC transfusion-dependent anemia. ClinicalTrials.gov identifier: NCT01034592 [Internet]. Bethesda (MD): National Library of Medicine; updated 2017-08-02. Available from: https://clinicaltrials.gov/study/NCT01034592
123. Northwell Health. Mobilization of CD34+ peripheral blood stem cells in patients with Diamond Blackfan anemia syndrome (DBAS). ClinicalTrials.gov identifier: NCT07186179 [Internet]. Bethesda (MD): National Library of Medicine; updated 2026-04-27. Available from: https://clinicaltrials.gov/study/NCT07186179
124. Payne EM, Virgilio M, Narla A, Sun H, Levine M, Paw BH, et al. L-Leucine improves the anemia and developmental defects associated with Diamond-Blackfan anemia and del(5q) MDS by activating the mTOR pathway. Blood. 2012;120(11):2214-24. doi:10.1182/blood-2011-10-382986. PMID: 22734070. PMC3447780.
125. Voit RA, Liao X, Caulier A, Antoszewski M, Cohen B, Armant M, et al. Regulated GATA1 expression as a universal gene therapy for Diamond-Blackfan anemia. Cell Stem Cell. 2025;32(1):38-52.e6. doi:10.1016/j.stem.2024.10.012. PMID: 39532107. PMC11698655.
126. Bhoopalan SV, Mayuranathan T, Liu N, Mayberry K, Yao Y, Zhang J, et al. Preclinical development of lentiviral vector gene therapy for Diamond-Blackfan anemia syndrome. Mol Ther. 2025;33(7):3086-3100. doi:10.1016/j.ymthe.2024.12.020. PMID: 39673126. PMC12265966.
127. Taher AT, Iolascon A, Matar CF, et al. Recommendations for pregnancy in rare inherited anemias. Hemasphere. 2020;4:e446. PMID 32885142.
128. Faivre L, Meerpohl J, Da Costa L, Marie I, Nouvel C, Gnekow A, et al. High-risk pregnancies in Diamond-Blackfan anemia: a survey of 64 pregnancies from the French and German registries. Haematologica. 2006;91:530–3. PMID 16537118.
129. Orfali KA, Ohene-Abuakwa Y, Ball SE. Diamond Blackfan anaemia in the UK: clinical and genetic heterogeneity. Br J Haematol. 2004;125:243–52. PMID 15059149.
130. Ridanpää M, Jain P, McKusick VA, Francomano CA, Kaitila I. The major mutation in the RMRP gene causing CHH among the Amish is the same as that found in most Finnish cases. Am J Med Genet C Semin Med Genet. 2003;121C(1):81-3. doi:10.1002/ajmg.c.20006. PMID: 12888988.
131. DBAS UK — Diamond Blackfan Anaemia Syndrome UK (registered charity no. 1083179). Home page [Internet]. United Kingdom: DBAS UK; accessed 2026-10-17. Available from: https://diamondblackfan.org.uk/