Autoimmune cytopenias represent a group of heterogenous disorders characterized by immune-mediated destruction of hematopoietic cells. The main mechanism is the production of autoantibodies directed against platelets [autoimmune thrombocytopenia (AIT)], erythrocytes [autoimmune hemolytic anemia (AHA)], and neutrophils [autoimmune neutropenia (AIN)]; however, they can also be directed simultaneously or sequentially against more than two cell lineages (Evans syndrome) [1, 2]. Autoimmune cytopenias present as either complications of other diseases—including infections, malignancies, systemic autoimmune disorders, or immunodeficiencies among others—or manifest as primary diseases [3].

In recent years, there is increasing evidence that primary immunodeficiencies, currently referred to as inborn errors of immunity (IEI), represent a significant predisposing factor of autoimmune cytopenias, particularly in patients with recurrent or chronic forms [4–6]. Among them, common variable immunodeficiency (CVID) is the most common IEI in humans, and autoimmune cytopenias represent the most prevalent autoimmune manifestation of the disease [7, 8]. Notably, more than 20% of CVID patients develop an autoimmune disorder, more often AIT accompanied by AHA, and occasionally by AIN [7]. Disease onset may occur at any age and can precede CVID diagnosis by a mean of 0.5–18.0 years, however, it may also be diagnosed concomitantly or after initiation of CVID treatment [8–10]. Data from the US Immunodeficiency Network cites that among CVID patients with a median age of 16.0 years, 4.5% suffer from AHA and 7.4% from AIT. These patients are also at higher risk of developing one or more non-infectious complications related to CVID [10]. For this reason, quantitative measurement of at minimum the three main immunoglobulins types—IgG, IgM, and IgA—should be considered essential during initial diagnosis of autoimmune cytopenias [6, 11].

Previously unpublished observations from our group revealed that some patients with autoimmune cytopenias exhibited low to undetectable levels of IgG4 immunoglobulin in peripheral blood, without evidence of overt hypogammaglobulinemia. IgG subclass deficiency is a rare type of IEI, identified by low levels of at least one or more IgG subclasses; however, its prevalence and clinical significance are rather unclear [12, 13]. Interestingly, some patients with immunoglobulin subclass deficiency (IgGScD) experience recurrent infections, autoimmunity, and atopy more frequently than the general population [14–17].

Although data in the literature are limited, previous studies suggest that IgG3 deficiency (IgG3D) is the most common IgGScD in adults, while IgG2D is more prevalent among children, with an incidence of approximately 1 in 10,000 individuals. Moreover, some patients displaying concomitant IgG2D and IgA deficiency (IgAD) exhibit a higher risk of developing CVID over their lifetime [14, 16, 18]. According to our published data, the most prevalent IgGScD in patients with IEI was selective IgG4 deficiency (sIgG4D), and affected patients displayed a notably high frequency of TNFRSF13B/TACI defects [19].

The aim of the present study was to investigate the prevalence of hypogammaglobulinemia and IgGScD in patients with autoimmune cytopenias, as well as identify the presence of TNFRSF13B/TACI defects that may contribute to disease pathogenesis and phenotype.

An overview of immunoglobulin and IgG subclass levels is presented in Table 2. No deficiencies in any of the main immunoglobulin types were detected among patients. However, among 84 patients for whom a measurement of IgG subclass was performed, eight (9.5%) displayed sIgG4D.

Clinical and laboratory data of these patients are detailed in Table 3. Interestingly, no patient with sIgG4D displayed a medical history of systemic autoimmune disease.

Three of eight patients with sIgG4D (37.5%) suffered from chronic AIT (patients 2, 5, and 7 in Table 3), while the remaining patients suffered from acute/newly diagnosed AIT (three patients, 37.5%) or AHA (two patients, 25.0%). According to these findings, we estimated the incidence of sIgG4D was more prevalent among patients with AHA (two of eight patients, 25.0%) compared to patients with AIT (6 of 113, 5.3%) (p = 0.030). Moreover, sIgG4D was marginally more common among patients with chronic AIT (3 of 46 patients, 6.5%) compared to patients with acute/newly diagnosed and/or persistent disease [3 of 50 (6.0%) and none from among 17 patients, respectively]. However, these differences were not significant (p > 0.05, in all cases).

Further molecular analysis of the TNFRSF13B/TACI gene in patients with sIgG4D revealed that only one patient (12.5%, patient 2 in Table 3) carried a pathogenic missense mutation in heterozygous state. The remaining patients carried common intronic and silent polymorphisms (Table 3). The patient carrying the pathogenic TNFRSF13B/TACI mutation (c.542C>A, p.A181E, rs72553883) displayed a persistent and stable disease (AIT) for at least one and a half years, with an initial response to corticosteroid treatment, followed by platelet counts ranging between 60–110/μL. The patient did not have a history of recurrent infections up to the age of 16, and his detailed family history revealed no IEI disease.

Interestingly, two patients with sIgGD4 and autoimmune cytopenias had additional organ-specific autoimmune manifestations: one patient with AIT and diabetes mellitus type I, and another patient with AHA and vitiligo (Table 3). Furthermore, two patients (patients 4 and 6 in Table 3) had a family history of CVID. Patient 4 (Table 3) displayed an acute/newly diagnosed AIT during pregnancy at 21 years of age. This patient was treated with corticosteroids and recovered without relapse and/or recurrent infections. Her daughter was diagnosed with CVID at the age of 15, displaying recurrent and severe respiratory infections with severe hypogammaglobulinemia; she was treated with immunoglobulin replacement therapy, initially intravenously and later subcutaneously. Patient 6 (Table 3) was diagnosed with AHA at 72 years of age, displaying severe and resistant disease. This patient’s niece (daughter of her sister) was diagnosed with CVID at 31 years of age, presenting with 10 years of recurrent respiratory infections and severe hypogammaglobulinemia, which was also treated with immunoglobulin replacement therapy. Neither patient or their relatives carried pathogenic TNFRSF13B/TACI mutations.

To the best of our knowledge, this is the first study in the literature exploring the possible association of autoimmune cytopenias with immunoglobulin subclass levels. We clearly demonstrate that a rather high proportion of patients with autoimmune cytopenias (9.5%) display sIgG4D. Interestingly, some of these sIgG4D patients also had a family history of IEI (two of eight, 25.0%), while one patient also carried a pathogenic TNFRSF13B/TACI defect (without a family history of IEI).

Estimating immunoglobulin levels at the time of autoimmune cytopenia diagnosis is indispensable to rule out an underlying IEI [4–8]. This is particularly essential considering patients with an IEI CVID have at least a 120-fold higher incidence of developing autoimmune cytopenias compared to the general population [9]. Moreover, abnormal immunoglobulin levels have been linked to more severe clinical illness and treatment resistance in AIT, since lower IgM levels have been correlated with a lower platelet count. Elevated IgA levels have been associated with bleeding and a higher risk of resistance to first-line therapy, independent of the platelet count [21].

IgG4D represents a type of IEI, either as an isolated phenomenon (sIgG4D), or in combination with a deficiency in other immunoglobulin subclasses, usually IgG2, IgA, or IgG1 [22, 23]. The precise incidence of sIgG4D in the general population is unknown. In a study of 414 healthy Greek children, Liatsis et al. [24] found no individuals with IgG4D. Meanwhile, in the most recent and comprehensive review on IgG4-related physiology and pathology, Rispens and Huijbers [23] reported that the incidence of sIgG4D in the general population is extremely low. This contrasts with the initial study by Plebani et al. [25], which measured IgG subclasses by radial immunodiffusion and reported undetectable IgG4 levels in approximately 8% of children. Conversely, several studies have reported that sIgG4D patients display a high incidence of recurrent sinopulmonary infections, atopy, chronic diarrhea, and chronic fungal infections, especially candidiasis [22, 23, 26]. Moreover, Koutroumpakis et al. [27] retrospectively analyzed data from 1,193 patients with inflammatory bowel disease (IBD) and found that approximately 20% displayed low IgG4 levels, which were significantly correlated to disease severity and outcome. However, very low IgG4 levels in IBD patients may be induced by either steroid treatment or protein loss [28].

The present research is the only study we are aware of which indicates sIgG4D positively correlates with autoimmunity, including both AIT and AHA (Table 3). Although sIgG4D incidence was higher among patients with AHA (25.0%), this finding should be interpreted with caution, considering fewer AHA patients were enrolled compared to AIT patients (8 versus 113). Furthermore, we found that two patients with sIgG4D had a family history of CVID, suggesting a possible common genetic cause requiring further investigation.

Interestingly, we also observed that a patient with chronic AIT carried a pathogenic TNFRSF13B/TACI defect (rs72553883, Table 3). It is currently well-recognized that monoallelic (heterozygous) or biallelic (homozygous or compound heterozygous) TNFRSF13B/TACI defects represent the most common alterations in CVID, which mainly affect disease phenotype and outcome [29]. The defect found in our patient—TNFRSF13B-p.A181E (rs72553883)—is located within the transmembrane domain (TM) domain of the TACI receptor. Along with the TNFRSF13B-p.C104R defect (rs34557412), they have a more profound effect on CVID phenotype [29, 30]. TACI is a receptor on B cells that binds BAFF and APRIL, playing a crucial role in B cell survival, class-switch recombination, and, importantly, immune tolerance [31]. As mentioned above, one of the sIgG4D study patients displayed the TNFRSF13B/TACI-p.A181E defect. Previous studies demonstrate that this defect impairs TACI signaling—particularly through pathways like NF-kappaB and MAPK—leading to a defective deletion or anergy of autoreactive B cells, increasing susceptibility to autoimmunity [31, 32]. Moreover, impaired TACI function can elevate BAFF levels, which further promotes the survival of autoreactive B cells, a mechanism also implicated in diseases like systemic lupus erythematosus (SLE) [33].

Additional in vitro studies in the murine equivalent of the p.A181E defect, namely the mTACI-p.A144E, suggest that this alteration disrupts TACI signaling and impairs TACI-dependent B-cell functions [34]. The contribution of this defect in the emergence of autoimmunity is still obscure, considering that TACI-knockout mice mainly exhibit a prominent autoimmune phenotype and signs of lymphoproliferation, rather than hypogammaglobulinemia [35]. Although the p.A181E defect is not fully penetrant, meaning not all carriers develop disease, it significantly increases the risk of conditions such as CVID, SLE, Sjögren’s syndrome, and autoimmune cytopenias, by promoting chronic immune activation and loss of B cell tolerance [34].

In the same context, Ma et al. [36] examined CVID-related genetic defects through NGS in pediatric patients with refractory AIT, as a screening tool for suspected CVID. Notably, five of the eight patients characterized as highly suspicious for CVID carried TNFSRF13B/TACI mutations (two with biallelic defects and three with monoallelic defects). Peng et al. [37] analyzed two families with familial AIT and suggested that a specific TNFRSF13B/TACI defect (p.G76S) should be considered as a predisposing factor for AIT development. TNFRSF13B/TACI mutations, along with defects in other genes implicated in IEI pathogenesis, have also been identified in patients with Evans syndrome [5]. Considering other TNFRSF13B/TACI defects detected in our patients (Table 3), they are highly common in the general population [19]. Functional studies have not shown any impact on TACI protein function, signaling pathways or B cell tolerance mechanisms. Additionally, genome-wide association studies and population analyses do not link these variants to CVID or its autoimmune manifestations [29]. Therefore, without evidence of impaired receptor function, disrupted immune regulation, or disease segregation, these variants are classified as benign polymorphisms rather than pathogenetic mutations [19]. Further studies should clarify the possible contribution of TNFRSF13B/TACI defects to the emergence and prognosis of autoimmune cytopenias, even among patients without profound hypogammaglobulinemia.

According to Morell’s old assumption [38] that “IgG subclasses deficiencies represent an indicator of more basic immunologic abnormalities”, we suggest that measurement of IgG4 levels could be utilized as a biomarker in patients with autoimmune cytopenias. Moreover, further genetic analyses would be beneficial, distinguishing among patients with primary autoimmune cytopenias and cases of suspected CVID, for optimal follow-up and management.

In conclusion, we identified a high incidence of sIgG4D among patients with autoimmune cytopenias, whom also displayed a notable frequency of either a family history of IEI or pathogenic TNFRSG13B/TACI mutations. Our findings suggest that measuring IgG4 levels should be considered a useful biomarker in patients with autoimmune cytopenias. Additional studies and meta-analyses are necessary to validate our results and clarify the importance of sIgG4D in patients with autoimmune cytopenias, for their disease pathogenesis and prognosis.