In nearly half of the cases (49%), the Sanger sequencing method for individual genes was used initially. In 38% of cases, various NGS methods were used. In 9% of cases, the FISH method was used to determine the 22q11.2 deletion. In 3% of cases, the aCGH method, and in 1% of cases, the MLPA method were used.
Variants considered causative were detected in 61.5% (1112/1809) of the probands examined. In addition, 130 different forms of IEI with underlying abnormalities in six chromosomes and defects in 118 genes were identified (Table S1). It is important to note that pathogenic variants in certain genes, such as WAS, STAT1, STAT2, and STAT3, led to different IEI phenotypes depending on the damaging mechanism: loss-of-function (LOF), dominant-negative effect (DN), or gain-of-function (GOF).
Monogenic defects were the cause of IEI in 87.9% (978/1112) of the children. Of these defects, 98.9% (967/978) had germline origins, and 1.1% (11/978) had somatic origins. An additional 11.8% (131/1112) of patients developed IEI due to large chromosomal abnormalities, predominantly 22q11.2 deletion (Table 1).
Detection rate of genetic defects in probands by group according to IEI classification
The highest efficiency in detecting genetic defects (97.6%; 411/421) was observed in the group with syndromic IEI. This group included 128 children with chromosomal abnormalities. Comparable detection rates were observed in the group of phagocytic defects (17 genes)—86.4% (185/214)—and in the group of complement defects (3 genes): 84.9% (45/53). In 73.1% and 72.3% of cases, respectively, the genetic diagnosis was confirmed among patients with predominantly antibody deficiencies (11 genes) (98/134) and those with defects in innate immunity (10 genes) (34/47), respectively. The rate of genetic confirmation in the group with immune dysregulation defects (18 genes) was 60.0% (111/185), and in the group with combined IEI (16 genes), it was 51.1% (117/229). The lowest efficiency of genetic validation was 31.3% (87/278), which was observed among patients with autoinflammatory disorders (14 genes) (Fig. 1). Defects in three genes related to bone marrow failure syndromes were detected in seven children, and defects in three genes leading to IEI somatic phenocopies were detected in 11 children. Three children were confirmed to have IEI involving defects in several genes. In addition, in the studied group, 10 patients with clinical and laboratory features of IEI were found to carry pathogenic variants in six genes and defects in three chromosomes not currently included in the IEI classification (Table S1). The symptoms of 287 probands who were referred for genetic testing and were negative did not fall into any group of IEI and therefore are not shown in Fig. 1.
Monogenic IEI distribution by inheritance mode and detected variants
Of 967 individuals with germline variants in 114 genes, 36.5% (353/967) were confirmed to have IEI with autosomal recessive inheritance mode (64 genes), 34.4% (333/967) with X-linked recessive inheritance mode (12 genes), and 29.1% (281/967) with autosomal dominant inheritance mode (38 genes).
Among patients with AR forms of IEIs, 52.1% (184/353) had compound heterozygous variants, and 47.9% (169/353) had homozygous variants.
Among homozygous IEIs, the most prevalent were patients with Nijmegen syndrome, 28.4% (48/169), with a single “Slavic” variant c.657_661delACAAA in the NBN gene [13] (Fig. S1). None of the parents of these patients reported consanguinity.
The next most frequent homozygous IEI, found in 6.5% (11/169), was familial Mediterranean fever, with the homozygous hotspot variant c.2050A>G (p.M694V) in the MEFV gene (Fig. S1). All these patients had Armenian origins, with only 1/11 resulting from consanguineous marriages. In the entire cohort of patients (21 individuals) with familial Mediterranean fever, the p.M694V variant was found in 28 of 42 alleles (an allele frequency of 66.7%) (Table 2).
Patients with ataxia–telangiectasia (AT) and variants in the ATM gene also constituted 7.1% of the homozygous cases (12/169) (Fig. S1). However, about half of them (5/12; 41.2%) were born in consanguineous marriages. Interestingly, in 46 probands with AT, the c.5932G>T (p.E1978*) variant of the ATM gene, previously associated with a founder effect in Eastern Europe and Russia [19,20], was found in the compound heterozygous state in only nine individuals, with an allele frequency of 9.8% (9/92) (Table 2).
An additional 4.1% of probands with homozygous variants consisted of seven children (7/169) with APECED syndrome who had a previously described hotspot variant c.769C>T (p.R257*) in the AIRE gene [21,22,23] (Fig. S1). None of their parents reported consanguinity. Among the 11 patients in our cohort with this IEI, the frequency of this allele was 77.3% (17/22).
Three percent of IEI with homozygous defects were accounted for five probands (5/169) with deletions of the first three exons of the DCLRE1C gene. The parents of these patients were allegedly unrelated, yet all families reside in the same geographic region, which might point to a new variant with a founder effect in this population.
The remaining 51% of probands with homozygous variants were diagnosed with various rare forms of IEI (fewer than 10 patients per gene).
Recurrent variants were also detected in unrelated probands with Shwachman–Diamond syndrome (SDS), activated PI3K defect syndromes 1 and 2, mevalonate kinase deficiency, PAID syndrome, Schimke syndrome, and combined immunodeficiency with RAG1 defect (Table 2). All these variants have previously been described as hotspot defects in the corresponding genes.
In the group of probands with monogenic IEI, 799 unique germline variants potentially related to the disease were identified. Interestingly, among the 799 variants, 44% (350/799) have not been previously reported. These included variants causative of well-studied, common IEI (Fig. 2).
According to the recommendations of the American College of Medical Genetics [17], 713 variants were classified as pathogenic or likely pathogenic based on bioinformatics analysis, and 83 were originally classified as VUS.
After performing relevant functional tests, we felt comfortable reclassifying 24 VUSs in 13 genes (CORO1A, RAC2, CD40LG, WAS, STAT3 LOF, BTK, FOXP3, XIAP, CYBB, CYBA, MVK, STAT1 GOF, and SERPING1) as pathogenic (Table S3). In addition, 59 unique VUS in 55 patients were considered causative based on a combination of the following data, which added significance to each variant: all probands had clinical and laboratory phenotypes specific to an IEI (e.g., profound T cell lymphopenia in patients with putative SCID variants, high AFP levels in patients with clinical manifestations of AT), VUS found in a compound heterozygous state with a pathogenic or likely pathogenic variant in the AR IEI, consistent segregation analysis of VUS, or the presence of affected siblings with the same genetic variant(s). Interestingly, the same variant in the ATM gene was detected in two unrelated probands with AT: one in compound with a likely pathogenic variant (P262) and the other with another VUS (P264)). Two unrelated probands with ICF1 syndrome had the same homozygous VUS in the DNMT3B gene. (Table S4).
Rare and common IEI representations
The largest group of children with a genetic diagnosis (more than 30 patients with defects in one gene) were patients with well-described primary immunodeficiencies: Wiskott-Aldrich syndrome (WAS), X-linked hyper-IgM syndrome (XHIGM), X-linked agammaglobulinemia (XLA), Nijmegen syndrome (NBS), hereditary angioedema types 1 and 2 (HAE), AT syndrome, SDS, X-linked severe combined immunodeficiency (XSCID) (with defects in the WAS, CYBB, BTK, NBN, SERPING1, ATM, SBDS, and IL2RG genes, respectively). In addition, a large number of patients (120 children) exhibiting clinical manifestations of DiGeorge syndrome were found to have 22q11.2 deletion, with only one patient having TBX1 gene defects as the cause of the DiGeorge syndrome (P418 in Table S1). These 10 forms of IEI were documented in 54% (598/1112) of all probands with confirmed genetic diagnoses.
Rare forms of IEI, in which defects in a gene were found in fewer than 10 probands, accounted for 22% (239/1112) of all children with IEI and were characterized by a high diversity of gene and chromosomal abnormalities that caused 91 different forms of IEI.
Gender distribution
Males were twice as prevalent as females in the study cohort, with a 2:1 (748:364) ratio. Among the main forms of IEI, the ratios of males to females were as follows: 1.3:1 (26:20) in AT patients, 1:1 (24:24) in the NBN gene defect group, 1:1.1 (56:64) among children with 22q.11.2 deletion, 1:1 (21:22) among those with SBDS gene defects, 1.5:1 (26:17) among those with defects in the SERPING1 gene. Defects in the IL2RG gene were found only in males, consistent with the XL mode of inheritance. Female patients with defects in the WAS, BTK, CYBB, and IKBKG genes and clinical manifestations of Wiskott-Aldrich syndrome (P124 in Table S1), agammaglobulinemia (P578 in Table S1), chronic granulomatous disease (P843 in Table S1), and combined immunodeficiency (P491 in Table S1), respectively, were determined to have nonrandom X-chromosome inactivation.
Di(poly)genic IEI
In three probands, defects in multiple genes led to complex IEI phenotypes. One patient (P1110 in Table S1) had a combination of pathogenic variants in the ATM and NFKB1 genes and symptoms of AT and severe immune dysregulation resistant to immunosuppressive treatment. Another patient (P1112 in Table S1) had defects in the MVK and MEFV genes and an autoinflammatory phenotype that had features of both MKD and FMF. In a female patient with typical proteasome-associated type 1 syndrome (P1111 in Table S1), heterozygous pathogenic variants were found in three genes encoding different subunits of the proteasome: PSMB8, PSMA5, and PSMC5 [24].
Large chromosomal aberrations as a cause of IEIs
Deletion of 22q11.2 caused DiGeorge syndrome in 120 children, and one patient had DiGeorge syndrome 2 due to deletion of chromosome 10 (P451 in Table S1). Seven probands were confirmed to have Jacobsen syndrome, with terminal deletion of chromosome 11. Three children with syndromic features and clinical and laboratory signs of immunodeficiency were found to have defects in chromosomes 7, 21, and 18 (P1107, P1108, and P1109, respectively, in Table S1). These chromosomal anomalies are not currently included in the IEI classification [1].
Several patients with monogenic IEI had large chromosomal abnormalities encompassing all or part of an IEI gene.
In four probands, microdeletions included the entire CTLA4 gene (P649 in Table S1) or the entire NFKB1 gene (P606, P609, and P616 in Table S1). Two patients had deletions that included part of a gene in a compound heterozygote state with a pathogenic variant (NBAS (P948 in Table S1) and DCLRE1C (P68 in Table S1)). Four children had additional phenotypic features, presumably caused by large chromosomal aberrations combined with pathogenic variants in a known IEI gene on a paired chromosome (BTK, CYBB, STAT1 GOF, and ATM). In one case, a paternally inherited variant of the USB1 gene appeared in the patient in a homozygous state due to uniparental disomy of chromosome 16 (P771 in Table S1) [25].
Familial cases
After segregation analysis, at least one relative with relevant IEI symptoms and the same genetic variant(s) was identified for 101/1112 (9.0%) probands with 23 different forms of IEI. In 60/101 families (59.4%), they were patients under the age of 18, and in the rest (40.6%) - adult relatives of the probands. Both pediatric and adult relatives with IEI were detected in five families. None of the newly diagnosed patients had an IEI diagnosis before.
Prenatal/preimplantation diagnostics
We conducted 60 tests of embryos/embryonic chorions in 57 families of patients with IEI. In 20% of cases (12/60), we detected variants previously identified in the families in a pathogenic combination, and 11 of the 12 pregnancies were terminated at the parents’ request. In addition, 15% of the embryos (9/60) had carrier status. In 65% of cases, (39/60), no known familial genetic defects were identified. A total 80% of pregnancies (48/60) resulted in births of immunologically healthy children.