Germline variants in acquired aplastic anemia: current knowledge and future perspectives

Abstract

Aplastic anemia (AA) is a disease characterized by failure of hematopoiesis, bone marrow aplasia, and pancytopenia. It can be inherited or acquired. Although acquired AA is believed to be immune-mediated and random, new evidence suggests an underlying genetic predisposition. Besides confirmed genomic mutations that contribute to inherited AA (such as pathogenic mutations of TERT and TERC), germline variants, often in heterozygous states, also play a not negligible role in the onset and progression of acquired AA. These variants, associated with inherited bone marrow failure syndromes and inborn errors of immunity, contribute to the disease, possibly through mechanisms including gene homeostasis, DNA repair, and immune injury. This article explores the nuanced association between acquired AA and germline variants, detailing the clinical significance of germline variants in diagnosing and managing this condition. More work is encouraged to better understand the role of immunogenic pathogenic variants and whether somatic mutations participate as secondary “hits” in the development of bone marrow failure.

Introduction

Aplastic anemia (AA) is the most common bone marrow failure syndrome and is classified into inherited and acquired forms based on clinical manifestations and genomic background. Inherited AA, often referred to as inherited bone marrow failure syndrome (IBMFS), is represented by Fanconi anemia and dyskeratosis congenita, both of which are clinically rare.¹ Acquired AA is characterized by immune-mediated bone marrow injury, manifested as a hypoplastic, fatty bone marrow, with a profound reduction in the numbers of hematopoietic stem/progenitor cells that leads to defective hematopoiesis and peripheral pancytopenia.² Immunosuppressive therapy (IST) with cyclosporine A, combined or not with antithymocyte globulin, is considered the first-line therapy that achieves hematologic remission in approximately 60%–70% of cases of acquired AA.³ However, immune system involvement alone cannot fully explain the pathogenesis and progression of acquired AA. The specific triggering factors and antigen targets responsible for the hyperactivity of T cells in acquired AA remain unclear.^4,5 Notably, abnormalities in the bone marrow microenvironment and hematopoietic stem cells (HSC) also play a significant role in the development of the disease.⁶

With the advancement and application of next-generation sequencing and whole-exome sequencing,⁷ HSC defects in acquired AA patients have been progressively revealed. Somatic variants, including those involving ASXL1, DNMT3A, TP53, and RUNX1, have been shown to be associated with worse responses to IST, but are still insufficient to explain IST failure in nearly 40% of cases.⁸ Approximately 5%–30% of young patients diagnosed with AA carry IBMFS-associated germline variants, potentially leading to adverse outcomes.^9-11 These variants can be categorized according to American College of Medical Genetics and Genomics guidelines as pathogenic (P), likely pathogenic (LP), variant of uncertain significance (VUS), likely benign (LB), and benign (B).¹² Dr. Neal S. Young’s research indicates that concomitant variants, excluding P/LP variants, are considered acquired AA variants,¹³ which should include recessive heterozygotes and VUS or B/LB variants in dominant genes. Despite lacking typical symptoms due to their mild pathogenicity, individuals carrying germline variants associated with IBMFS and inborn errors of immunity eventually experience immune-mediated bone marrow destruction as deleterious factors accumulate. Our review aims to explore this phenomenon: clinically non-pathogenic variants (B/LB, VUS, or recessive heterozygotes) hold significant implications for acquired AA, as their enrichment is not accidental.

Distinguishing inherited bone marrow failure syndromes in the diagnosis of acquired aplastic anemia

Before defining novel subgroups in acquired AA, it is essential to retrospectively review the characteristics of both acquired AA and IBMFS using their unique features.

Features and specific examinations of inherited bone marrow failure syndrome

Inherited AA or IBMFS is a collection of rare disorders resulting from the inheritance of pathogenic germline variants and includes Fanconi anemia, dyskeratosis congenita, Diamond-Blackfan anemia, Shwachman-Diamond syndrome, severe congenital neutropenia, congenital amegakaryocytic thrombocytopenia, MIRAGE (myelodysplasia, infection, restriction of growth, adrenal hypoplasia, genital phenotypes, and enteropathy) syndrome, and GATA2 deficiency.^14-16 These diseases often have distinctive manifestations that differentiate them from acquired AA, such as developmental abnormalities, physical deformities, mucocutaneous triad (dyskeratosis congenita), pulmonary fibrosis (dyskeratosis congenita), and exocrine pancreatic insufficiency (Shwachman-Diamond syndrome).¹⁷ Thorough investigation into family history and specific testing is crucial. A family history displaying abnormalities heightens suspicion of IBMFS. Positive results in chromosomal breakage tests serve as the diagnostic standard for identifying Fanconi anemia.¹⁸ An extremely short telomere length (<1^st percentile for age) can be used as a highly sensitive and specific marker for the detection of dyskeratosis congenita.¹⁹

Heterogeneity of clonal hematopoiesis

Clonal hematopoiesis can be viewed as the adaptive response of HSC to environmental stress.²⁰ The heterogeneity of clones reflects pathogenic mechanisms that are distinct for acquired AA and IBMFS. In acquired AA, HSC tend to have variants with immune escape and proliferation in overcoming the immune response and the toxic microenvironment. This grants them a competitive advantage in clone formation, which is observed in around 50%-70% of patients.^21-23 Somatic variants, including PIGA, BCOR, and BCORL1, are characterized by an improved response to IST and overall survival, and are considered clinically beneficial. On the other hand, DNMT3A, ASXL1, TP53, RUNX1, and CSMD1 variants are associated with a lower survival rate and progression to myelodysplastic syndromes/acute myeloid leukemia, and are considered clinically unfavorable.^8,21 Comparable features have been observed in IBMFS, and are referred to as somatic compensation. Unlike in acquired AA, the pressures confronting HSC in IBMFS stem primarily from their inherent genetic defects. Consequently, the significance of these clones in IBMFS is centered around compensating for or restoring the original function caused by the intrinsic deficiencies. Somatic variants can be categorized into three types based on the specific compensation mechanisms:²⁴ gene-specific, pathway-specific, and pathway-independent. Gene-specific compensation is observed in variants that fully or partially restore original gene function. It usually occurs in diseases such as Fanconi anemia, dyskeratosis congenita, and SAMD9/9L-related disorders. Pathway-specific compensation maintains pathway integrity through compensatory variants in other genes, such as the acquired promoter variants of telomerase reverse transcriptase (TERT) in dyskeratosis congenita and loss of EIF6 in Shwachman-Diamond syndrome. Pathway-independent compensation involves acquiring growth advantages through alternative pathways or mechanisms, exemplified by TP53 variants in Shwachman-Diamond syndrome, telomere biology disorders, and Diamond-Blackfan anemia, RAS pathway variants in Fanconi anemia, RUNX1 variants in Fanconi anemia, and ASXL1 variants in GATA2-related disorders.^24-28 Somatic compensation endows HSC with proliferative advantages within the hematopoietic microenvironment of IBMFS, leading to two possible outcomes: disease reversal or malignant transformation.

Value of immune mechanisms

Somatic variants, such as ASXL1, RUNX1, and TP53, lack specificity in bone marrow failure diseases, as well as the variable penetrance of IBMFS, posing a challenge for differentiation. The immune escape mechanisms of paroxysmal nocturnal hemoglobinuria (PNH) and HLA class I allelic gene loss (Figure 1), along with the immune characteristics of TCR-Vβ oligoclonal expansion, collectively serve as markers for acquired AA, providing diagnostic references with high positive predictive value and specificity.^29-31 PNH is characterized by the HSC PIGA clone, where PIGA variants impede the synthesis of glycosylphosphatidylinositol-anchored proteins, resulting in the loss of CD55 and CD59 expression. This deficiency triggers complement activation, leading to hemolysis and thrombosis.^21,32 Meanwhile, the deficiency of glycosylphosphatidylinositol-anchored proteins renders inactive immune response-related cell surface proteins,³³ which is a common immune escape mechanism observed in 20%–60% of patients with acquired AA.^34,35 The detection of PNH clones (>1%) in patients with acquired AA is considered to exclude patients from having IBMFS and has been validated in clinical and research settings.^11,13,36,37 Similarly, the loss of HLA class I allelic genes endows HSC with the capacity to establish clones with cellular immunity in 20% of patients, involving the mechanisms of HLA variants and 6p copy number-neutral loss of heterozygosity (CN-LOH) (also known as 6p uniparental disomy).³¹ CN-LOH describes the phenomenon of heterozygous loss where one parental chromosome region is replaced by the other,^38,39 and occurs frequently in the HLA class I region of chromosome 6p in acquired AA.²¹ HLA variants cause acquired loss of partial HLA class I genes. Both losses are concentrated at sites such as HLA-B*14:02, HLA-A*02:01, and HLA-B*40:02, outlining the classic immune escape landscape in acquired AA.^21,31 In a previous study, 6p CN-LOH showed almost 100% positive predictive value for acquired AA, highlighting its significant diagnostic value.²⁹ Limited usage of TCR-Vβ indicates the oligoclonal expansion of CD8⁺CD28^- T cells,^40-42 suggesting immune dysregulation at the T-cell level under chronic antigen stimulation,⁴³ which is a common occurrence in acquired AA. Successful use of IST results in a significant reduction in the clonal expansion of CD8⁺CD28^- T cells, considered robust evidence for an immune-mediated mechanism.^44,45 Recent evidence indicates that oligoclonal patterns are also present in effector memory CD8⁺CD28^-CD57⁺ T cells. Notably, effector memory T cells pose a persistent threat and play a crucial role in the recognition and recurrence of acquired AA.^30,46

Figure 1.Immune destruction and involvement of germline variants. Antigen-presenting cells (APC) process antigens, presenting them to T cells. CD4⁺Th0 cells differentiate into Th1 and Th17 cells, maintaining a relative balance. Among them, STAT1 promotes the differentiation of Th0 to Th1 cells, leading to the secretion of cytokines interferon gamma (IFNγ) and tumor necrosis factor alpha (TNFα) by CD4⁺ Th1 cells, which bind to their respective receptors (IFNR and TNFR), promoting apoptosis of hematopoietic stem cells (HSC). Additionally, increased IFNγ stimulates CD8⁺ T-cell proliferation and promotes HSC apoptosis through FASLFAS signaling. CD8⁺ T cells also release granzyme and perforin encoded by PRF1, leading to the elimination of APC and negative regulation of immune responses. Furthermore, SAMD9/9L is associated with antitumor and anti-infection responses, although the specific mechanism of action remains unclear. The formidable survival pressure prompts HSC to employ mechanisms such as PIGA, 6p uniparental disomy, and HLA-I mutations to escape from immune responses. *Genes affected by germline variants involved in acquired aplastic anemia.

Contribution of germline variants to acquired aplastic anemia

In patients with acquired AA, genetic reports of individuals characterized by more severe phenotypes or unfavorable treatment responses often show common occurrences of germline heterozygous recessive variants, as well as VUS or B/LB variants in dominant genes.⁷, ⁴⁷ Related to IBMFS and inborn errors of immunity, the genes involved maintain the stability, repair, and renewal of HSC under normal conditions (Figures 1 and 2). When mutated, impaired or exaggerated functions may contribute to the pathogenesis and development of acquired AA (Table 1).^7,47-50 Exploring the prevalence and mechanisms of germline variants in acquired AA will contribute to a more comprehensive understanding of the etiology and risk factors of the disease.

DNA repair deficiencies accelerate bone marrow failure

DNA damage repair has a crucial role in maintaining the response of HSC to both external and internal stimuli, as well as their self-renewal process.⁵¹ The Fanconi anemia pathway is one of the important mechanisms involved in DNA damage repair (Figure 2). The prevalence of common heterozygous FANC variants is 8.38% in patients with acquired AA and hematologic malignancies.⁵² In Chinese patients with acquired AA below the age of 40, the prevalence of non-pathogenic heterozygous FANC variants was 45.9%, with LB variants and VUS accounting for 97.56% of all cases.⁷ To date, 22 FANC genes have been identified, and their homozygous states are closely associated with Fanconi anemia, cancer susceptibility, and hereditary breast tumors.¹⁸ Previous studies have suggested that FANC heterozygous carriers typically do not have the congenital anomalies and chromosome breakage observed in Fanconi anemia.⁵³ The risk of developing cancer in heterozygous carriers is comparable to that in the general population, and these states are generally considered non-pathogenic.^54,55 However, in a study integrating AA and hematologic malignancies, heterozygous carriers of FANC variants showed susceptibility to both conditions, suggesting that in certain disease contexts, heterozygosity may be insufficient to maintain proper function.⁵² The underlying mechanism may involve the diminished ability of heterozygous variants to repair DNA, leading to unresolved DNA damage that activates P53 and results in HSC exhaustion. Alternatively, without P53 inhibition, accumulated mutations may trigger immune destruction or evolve into hematologic malignancies.^52,56 FANCA variants are the most common among cases of acquired AA with FANC heterozygous variants, followed by BRCA2 and FANCD2, which have a pattern of variant frequencies similar to that in Fanconi anemia. The proportions of FANCG and FANCC variants are relatively small, possibly due to limited sample sizes. The enrichment of rare variants in FANCN and SLX4, which are less commonly observed in patients with Fanconi anemia, may indicate susceptibility in acquired AA or could be attributed to sample size, necessitating further comprehensive investigations.^7,18,52

Excessive immune responses aggravate hematopoietic stem cell damage

Some patients with acquired AA carry immune-related germline variants, exacerbating the immune response in disease progression. Heterozygous recessive variants of perforin (PRF1) have been identified in acquired AA, accounting for approximately 6% of cases.⁵⁰ Homozygous PRF1 variants are associated with familial hemophagocytic lymphohistiocytosis, a fatal disorder of childhood onset characterized by functional perforin deficiency.⁵⁷ In a previous study, the p. A91V variant was most frequently observed, and four out of the five patients had increased hemophagocytes in the bone marrow, without other typical clinical features of hemophagocytic syndrome.⁵⁰ The p. A91V variant was confirmed to induce sustained inflammation, chronic antigen presentation, and release of inflammatory mediators.⁵⁸ Low perforin caused by PRF1 variants impairs the elimination of antigen-presenting cells, leading to the activation and proliferation of cytotoxic T cells.^50,59 Accompanied by the secretion of interferon-γ, this ultimately results in the destruction of HSC.⁶⁰ Recently, two patients with acquired AA were found to have heterozygous variants in Myb-like SWIRM and MPN domains 1 (MYSM1), a regulator of hematopoiesis and immune cell development, known to cause IBMFS in a homozygous state.^61,62 These patients had loss of HLA-A*02:06, and one also had PNH. Tatsuya et al. suggested that MYSM1 heterozygous defects lead to an excessive immune response to exogenous antigens, increasing susceptibility to acquired AA.⁶¹ The association between acquired AA and immune-related variants appears to be non-incidental. Whole-genome sequencing of patients with AA/PNH revealed that 65% (37/57) carried heterozygous germline variants associated with inborn errors of immunity. Among the 60 variants identified, VUS were predominant, involving 37 autosomal recessive variants and 23 autosomal dominant variants, offering a new perspective on acquired AA.⁶³ In conclusion, heterozygous recessive variants or VUS in dominant genes may lead to aberrant responses or incomplete reactions to antigens, ultimately resulting in the development of acquired AA due to the failure of feedback regulatory mechanisms, as detailed in the preceding discussion.

Table 1.Incidence and functional mechanisms of germline variants in acquired aplastic anemia.

Figure 2.Germline variants involved in genome stability and DNA repair. Telomeres are repeated DNA segments located at the ends of chromosomes, protected by shelterin proteins and extended by telomerase. The recruitment of telomerase by shelterin proteins is a crucial step for telomere elongation, a process in which SBDS also participates. The CST complex (CTC1-STN1-TEN1) has a dual role: recruiting Polα-primase for C-strand synthesis and inhibiting telomere extension, providing intricate regulation of telomeres. In response to DNA damage, the MRN complex serves as a sensor, signaler, and effector to mediate DNA repair. It promotes the recruitment of CTIP, ATR, and ATM, facilitating DNA repair pathways such as microhomology-mediated end joining and non-homologous end joining. ATR and ATM also facilitate the formation of the Fanconi anemia core and are crucial for FANCD2-FANCI monoubiquitylation, ultimately promoting homologous recombinational repair, nucleotide-excision repair, and translesion synthesis DNA repair pathways. *Genes affected by germline variants, in acquired aplastic anemia. Ub: ubiquitylation; HRR: homologous recombinational repair; NER: nucleotide-excision repair; TLS: translesion synthesis; MMEJ: microhomology-mediated end joining; NHEJ: non-homologous end joining.

SAMD9/SAMD9L exaggerate defects of hematopoietic stem cells

Sterile alpha motif domain-containing protein 9/9-like (SAMD9/9L) encode two proteins involved in antitumor and antiviral responses; however, their precise functions and regulation remain elusive.⁶⁴ In 2016, SAMD9 was first identified as the causative gene for MIRAGE syndrome, a multisystem disorder characterized by a predisposition to myelodysplastic syndromes and loss of chromosome 7.¹⁶ Meanwhile, whole-exome sequencing also revealed that SAMD9L was the cause of ataxia-pancytopenia syndrome.⁶⁵ Both syndromes share similarities in their excessive antiproliferative effects due to gain-of-function (GOF) variants, leading to the manifestation of cytopenia or even pancytopenia.^66,67 Pediatric cohorts with SAMD9/9L (GOF) variants tend to develop hematologic malignancies.^{37,6 7,6 8} SAMD9L (GOF) variants were found to impair HSC proliferation and self-renewal, promote inflammation, and exacerbate bone marrow failure.^69,70 Intense survival pressure drives HSC to counteract the variant damage through somatic compensation or CN-LOH,⁶⁹ resulting in transient early-life AA episodes.³⁷ Recent studies have revealed that up to 20% of adult patients with severe AA harbor SAMD9/9L (non-GOF) heterozygous germline variants.⁴⁷ These variants frequently co-occur with typical features of acquired AA, including 6p CN-LOH, PNH clones, and variants in HLA and PIGA genes. Notably, the probability of chromosome 7 abnormalities in these patients was relatively low, 2/40 cases (5%), differently from what is observed in SAMD9/9L-related disorders. The most common types of variants observed are missense, nonsense, and frameshift. Carriers of these variants tend to be younger and exhibit lower levels of neutrophils, reticulocytes, and platelets.⁴⁷ However, our understanding of these variants and their pathogenic mechanisms remains incomplete. The regulation of HSC and the microenvironment by SAMD9/9L is highly complex. Loss-of-function variants are also linked to a predisposition to myelodysplastic syndromes and abnormal expression of specific inflammatory pathways.⁷¹ SAMD9/9L variants in adult acquired AA can be considered to contribute to mild functional disruptions, including dysregulation of HSC proliferation and immune-inflammatory responses, ultimately leading to T-cell destruction and bone marrow failure.

Other variants

With the application of whole-exome sequencing, numerous reports on other germline variants continue to emerge. Heterozygous variants in ribosome maturation protein (SBDS) account for approximately 5% of cases of acquired AA.⁴⁹ Biallelic variants in SBDS are found in over 90% of patients with Shwachman-Diamond syndrome and are a major genetic factor associated with this disorder.⁷² Shwachman-Diamond syndrome is a genetic syndrome characterized by an elevated risk of bone marrow failure, exocrine pancreatic insufficiency, skeletal defects, and hematologic malignancies.⁷³ Patients with acquired AA with heterozygous SBDS variants do not have such manifestations but share pancytopenia and telomere shortening with those having Shwachman-Diamond syndrome.⁴⁹ The telomere shortening may be attributed to the dysfunction in SBDS in interacting with shelterin, a complex that protects telomeres and initiates telomerase recruitment.⁷⁴

Conserved telomere maintenance component 1 (CTC1) is a part of the CST complex.⁷⁵ Approximately 2.3% of patients with acquired AA and PNH harbor heterozygous variants in CTC1 and do not have any signs and family history of IBMFS.⁴⁸ Homozygous CTC1 variants are common in Coats plus syndrome and dyskeratosis congenita, in which these variants impair the function of the CST complex, leading to telomere dysfunction and genomic instability.^75,76 Of four cases of CTC1-Tier-1 carriers, three (75%) were found to have large PNH clones (clone size >50%), indicating the widespread occurrence of immune escape. While these heterozygous CTC1 variants do not significantly affect telomere length, they still exhibit a tendency to acquired bone marrow failure and PNH, needing further exploration. Moreover, additional cases have offered insights for future investigation. A B/LB variant of TRF1-interacting nuclear factor 2 (TINF2), Ser245Tyr, associated with late-onset recurrent AA, has been identified in three patients with acquired AA.^{7 7,7 8} GOF variants in the signal transducer and activator of the transcription 1 (STAT 1 ) have been reported in two cases of AA and led to increased T helper type 1 cells and the overexpression of interferon-γ.^79,80 Other concomitant variants, including heterozygous variants of RTEL1, DDX41, ZRSR2, NFKB1, and ETV6, have provided further information, but their pathogenicity has yet to be clarified.^81-85

Clinical significance of germline variants in acquired aplastic anemia

Co-existence of germline variants and paroxysmal nocturnal hemoglobinuria clones

PNH, characterized by a clone of PIGA somatic cell variants, is a hallmark of acquired bone marrow failure.⁸⁶ The PNH clone (>1%) detected in AA patients is considered to exclude IBMFS.^11,36,37 However, evidence suggests that patients with acquired AA who carry germline variants can also have clones of PNH. In four cases of CTC1-Tier-1 carriers, three (75%) were found to have large PNH clones (clone size >50%).⁴⁸ The presence of PNH clones was frequently observed in adult patients with severe AA carrying SAMD9/9L variants.⁴⁷ In one AA patient carrying the MYSM1 variant, the PNH clone accounted for 30%.⁶¹ Moreover, some patients with variants in inborn errors of immunity show a tendency to develop PNH clones.⁶³ In conclusion, the co-existence of PNH and germline variants indicates their distinction from IBMFS. Excluding germline variants by the presence of PNH clones may not be entirely conclusive.

Diversity of response to immunosuppressive therapy

Patients with germline variants are generally considered to have a poor response to IST.¹¹ However, specific outcomes may vary due to gene variations and individual differences (Table 2). For example, four out of five cases of acquired AA with heterozygous germline variants in CTC1 demonstrated a response to IST.⁸⁷ A patient carrying a heterozygous variant of MYSM1 showed a response to antithymocyte globulin plus cyclosporine A.⁶¹ Heterozygous variants in SMAD9/9L were found to have a minimal impact on the efficacy of IST, with approximately 85% of patients carrying these variants showing a response to IST plus eltrombopag.⁴⁷ However, no favorable outcomes were achieved with the use of IST in patients carrying other germline variants. In another trial, three out of four AA patients with SBDS heterozygous variants did not respond to IST, and the remaining patient showed only a transient response.⁴⁹ Additionally, three out of five patients with PRF1 heterozygous variants did not respond to IST; the remaining two were dependent on cyclosporine A.⁵⁰ In a group of eight AA patients with FANC heterozygous variants, the efficacy of IST (antithymocyte globulin plus cyclosporine A) was only 25%, whereas a response rate of up to 100% was obtained in the control group (6 patients).⁷

Table 2.Treatment and its efficacy in acquired aplastic anemia with germline variants.

Choice of hematopoietic stem cell transplantation and prognosis

Hematopoietic stem cell transplantation (HSCT) is considered the ultimate and effective approach for cases of severe AA failing to respond to IST and other treatment options.⁸⁸ The impact of recessive germline variants on transplantation in bone marrow failure has long been unknown. An earlier study provided overarching insights: P/LP variants (a homozygous or compound heterozygous state is required when inheritance is recessive) led to poorer post-transplant survival. On the other hand, carriers with single recessive P/LP variants or VUS (defined as acquired AA in our review) did not show a significant difference in post-transplant survival compared to non-carriers.¹¹ The study also identified graft-versus-host disease as the primary cause of mortality after transplantation in these carriers, distinct from the organ failure observed in patients with P/LP variants. Specifically, in patients with acquired AA with FANC heterozygous germline LB variants or VUS who underwent a treatment approach consistent with that of non-carriers, the final hematologic response showed no difference.⁷ Similar conclusions were drawn for carriers with heterozygous FANC P/LP variants in another study.⁸⁹ Both indicated that FANC carriers do not require reduced-intensity conditioning regimens, highlighting their heterogeneity from patients with Fanconi anemia.^7,90

Sibling donor transplantation is considered the preferred treatment for acquired AA over unrelated donor transplantation.^91,92 However, it is important to note that family members or unrelated individuals carrying germline variants could be donors for HSCT, introducing the possibility of transplanting imperfect HSC to patients. Previous findings in other hematologic diseases have indicated that donor-derived pathogenic germline mutations could increase the risks of malignancies and engraftment failure. Certain variants, such as DDX41, have been linked to a higher incidence of acute graft-versus-host disease.⁹³ However, the impact of donor germline variants on HSCT in AA remains inadequately explored. There is one reported case of a patient who died of graft failure after having undergone HSCT from a histocompatible sibling with an unrecognized mutation of telomerase RNA component (TERC).⁹⁴ Another report described the case of a 43-yearold male patient who experienced remission for 4 years after receiving a transplant from a sibling carrying heterozygous variants in FANCJ Arg814Cys and TINF2 Leu429Val.⁹⁵ These cases suggest that the strategy for donor selection requires urgent attention. Sibling transplantation has been performed for decades for the treatment of autosomal recessive genetic diseases such as Fanconi anemia. In China, haploidentical HSCT was conducted more frequently than unrelated donor transplantation in acquired AA,^96,97 which suggests a higher incidence of transmission of germline variants. The occurrence of graft-versus-host disease⁹⁷ may be associated with acquired AA variants, whether originating from patients or donors. Nevertheless, the impact of these variants on HSCT engraftment, survival, and the long-term risks have not been comprehensively evaluated.¹¹ Further studies are needed to answer these questions and refine the selection criteria for HSCT.

Outlook and future directions

Considerations on immunogenic pathogenic variants

In the current conceptual framework, the diagnosis of acquired AA primarily emphasizes the exclusion of pathogenic variants associated with IBMFS,⁹⁸ with less attention given to other systems or isolated cases. Recently, whole-genome sequencing showed that 65% of AA/PNH patients carry germline variants associated with inborn errors of immunity.⁶³ In acquired AA, PIGA is considered a beneficial variant, often indicating a more favorable response to IST.^8,21 However, P/ LP immunogenic germline variants are also frequently associated with PNH, exhibiting higher autoimmune activity and poorer response to IST.⁶³ For instance, patients with NFKB1 haploinsufficiency do not tolerate cyclosporine A plus eltrombopag treatment.⁸⁴ The association of complement germline variations with PNH⁹⁹ further underscores the importance of identifying relevant variants in AA/PNH. The next consideration should be whether pathogenic immunogenic variants should be incorporated into acquired bone marrow failure. This may require a multifaceted evaluation, considering distribution frequency, efficacy, and pathogenicity, to provide a more precise definition of acquired AA.

Enhanced identification and comprehensive understanding

Many non-pathogenic variants remain associated with acquired AA. Current research is revealing the correlation between acquired AA and variants in genes related to IBMFS and inborn errors of immunity, although more exploration is still needed to identify others. For instance, the identification of variants corresponding to FZR1 deficiency, proven to induce AA, holds significant meaning in this context.¹⁰⁰ Recently, machine-learning models have made strides in predicting acquired versus inherited AA with an accuracy rate of 89%. However, specific identification of germline variants, such as SAMD9/9L, and other heterozygous recessive variants remains limited.¹³ Carriers of these variants exhibit distinct clinical features such as younger age, lower counts of neutrophils, reticulocytes, platelets, and CD34⁺ cells, suggesting the potential for using machine-learning algorithms for discrimination.^{7,4 7} Applying machine-learning models in this area will bring very interesting results. The mechanism through which germline variants cause acquired AA is not yet fully understood. Some germline variants may contribute to acquired AA through a co-occurring heterozygous mode, such as FANC and telomere biology disorders, SAMD9/9L, etc.^11,47,101 This also provides another perspective, suggesting the cumulative effect of non-pathogenic factors. As previously discussed, somatic compensation is a distinctive phenomenon in IBMFS. It is worth exploring whether similar mechanisms exist in the subset of acquired AA with germline variants. The pathogenesis of the acquired AA in these cases may be linked to specific clonal populations, aligning with the concept of a genetic “second hit”. Nonetheless, the relationship between germline variants in acquired AA and the response to conventional treatment has not been fully established. FANC heterozygous carriers were found to have a poor response to IST, but the underlying mechanisms remain unclear.⁷

Novel therapeutic approaches

Currently, there is a limited understanding of germline variants in acquired AA, and the treatment framework primarily adheres to the general protocols established for acquired AA. When an IST regimen fails to have satisfactory efficacy, other drugs with different mechanisms of action should be considered. Androgens have shown promising response rates and safety in conditions such as dyskeratosis congenita, Fanconi anemia and acquired AA.^102,103 Treatment with androgens increases TERT expression and improves telomere maintenance,¹⁰⁴ thus being potentially beneficial to patients with short telomeres and acquired AA variants. However, the use of androgens for Shwachman-Diamond syndrome is unconventional. When considering androgen therapy for heterozygous SBDS patients, the potential benefits must be evaluated carefully. Some cases involving TINF2 variants¹⁰⁵ suggested that augmenting telomerase activity alone may not effectively address recruitment failure. Patients with acquired AA with FANC heterozygous variants appear to inherit some characteristics of FANC and exhibit a lower response to IST compared to non-variant individuals.⁷ The administration of androgens has achieved positive results in the remission of FANC,¹⁰⁶ in which it is thought to stabilize telomere and genomes,¹⁰³ and may play a role in the remission of FANC pathway disorders. Further exploration is required to determine the viability of using androgens in acquired AA patients with heterozygous FANC variants, including potential combination with eltrombopag. Itacitinib, a selective JAK1 inhibitor, has shown preliminary efficacy in the treatment of acute graft-versus-host disease.¹⁰⁷ Promising results have been observed with itacitinib in a patient with STAT 1 (GOF) variant, and evidence of STAT 1 overexpression has also been found in patients with non-mutated acquired AA, suggesting a potential therapeutic approach targeting STAT1 in AA.^79,80 However, the effectiveness of itacitinib beyond S TAT 1 (GOF) AA patients requires further validation. More research and data are required to refine the management of patients with acquired AA.

Conclusion

In acquired AA, individuals carrying certain germline heterozygous recessive variants, along with VUS or B/LB variants in dominant genes, constitute a distinct subgroup. They manifest typical features of acquired AA, with variations in disease severity and treatment response. The variants they carry induce and aggravate AA through various mechanisms such as gene homeostasis, DNA repair, and immune injury, representing risk factors for the development of acquired AA. It is important not to prematurely dismiss the effectiveness of IST, and a comprehensive evaluation of the patient’s variant type and condition should be made. When deciding on HSCT, caution should be exercised in considering sibling and haploidentical donors. There is currently a lack of relevant research to provide specific recommendations for this particular group of patients. There is an urgent need for a more rigorous and authoritative definition of acquired AA, determining whether immune-related pathogenic variants should be encompassed. Exploring further mechanisms in acquired AA carriers may unveil novel mechanisms and treatment approaches, advancing our understanding of AA and promoting the development of precision medicine.

Footnotes

• Received September 16, 2023
• Accepted April 9, 2024

Correspondence

*PW, WJ and TL contributed equally as first authors

D. Wu wudijiong@zcmu.edu.cn

B. Ye 13588453501@163.com

Disclosures

No conflicts of interest to disclose.

Funding

The present study was supported by the National Natural Science Foundation of China (N. 82174138, 82304937), Zhejiang Scientific Research Fund of Traditional Chinese Medicine (N. 2020ZB085, 2022RC216), Project of Academic Inheritance Studio of Famous and Aged Chinese Medicine Experts in Zhejiang Province (N. GZS2021022), and the Hematological Diseases Program of Chinese Association of Traditional Chinese Medicine (N. 202169-003).

Acknowledgments

We thank Dr. Didier Blaise for his intellectual advice. We also thank Dr. Charmaine Fay Carcallas Soco and Dr. Daisi Zheng for their advice and assistance with the preparation of the figures. The figures were drawn using Figdraw (www.figdraw.com).