Common variable immunodeficiency disorders (CVIDs) are a group of primary immunodeficiencies associated with impaired B-cell differentiation and defective primary antibody production (1). CVID is characterized by heterogeneous clinical manifestations including recurrent infections, chronic lung disease, autoimmune disorders, gastrointestinal diseases, and heightened susceptibility to lymphoma (2). CVID is usually sporadic; patients usually have no reported family history of immunodeficiency. Nevertheless, approximately 10% of the patients may have other first-degree relatives with hypogammaglobulinemia or selective IgA deficiency (3). Monogenic causes of the disease, showing an autosomal dominant inheritance pattern, have been identified in less than 10–15% of all CVID cases and include mutations in genes involved in lymphoid organogenesis, B-cell survival, and maturation (4). In most cases, the causative mutations remain unknown, making it difficult to elucidate the pathogenic mechanisms of CVID (5). Next-generation sequencing (NGS) has revealed several genes associated with the pathology (6), including TACI, CD19, BAFFR, CD20, CD81, CD21, LRBA, NFKB2, IL21, NFKB1, IKZF1, IRF2BP2, among others. Several nonsense (7, 8) and less frequent missense mutations (9) have been reported as potential causes of CVID. Heterozygous NFKB1 and NFKB2 variants, collectively account for the most frequent genetic cause in CVID (9).

In particular, NFKB2 is a widely studied transcription factor with key roles in the functioning of the immune system (2, 10), such as regulation of pathways leading to inflammation, cell survival, and differentiation (10). NF-κB2 is the principal protein involved in the noncanonical NF-κB pathway, is evolutionarily conserved, and functions in peripheral lymphoid organ development, B cell development, and antibody production (2). Mutations in the NFKB2 gene can lead to a deficiency in NFKB2 protein, which is associated with CVID (2, 11). The exact mechanisms by which NFKB2 mutations cause CVID are not fully elucidated, but it is thought that NFKB2 plays a role in the development and function of B cells and T cells (12), which may lead to a deficiency in antibodies (13). Efforts have been made to formulate diagnostic criteria to better define this heterogeneous group of pathologic entities, including the ESID/Pan American Group of Immune Deficiency (PAGID, 1999) criteria (14), Ameratunga and colleagues (2013) criteria (15), and the most recent CVID International Consensus (ICON, 2016) criteria (16). Currently, all definitions of CVID exclude patients with known disorders; therefore, if a specific genetic variant is found (NFKB1, NFKB2 etc.), the patient would be excluded from a CVID diagnosis and would then be reclassified as having a CVID-like disorder caused by a specific PID (primary immunodeficiency disorder)/inborn error of immunity (17).

CVID can be divided into two groups according to clinical manifestations. The most frequent group (more than 70% of patients) exclusively presents with infections, whereas the second, less frequent group has a variety of noninfectious inflammatory, autoimmune, and/or lymphoproliferative presentations, which are also associated with systemic immune activation (18, 19). In some cases (approximately 25%), patients in the second group may have inflammatory symptoms as the only symptom, with no obvious susceptibility to infectious disease. Patients with inflammatory/autoimmune features have increased morbidity and mortality; hence, several studies have focused on these groups to understand its heterogeneity (20).

Mycosis fungoides (MF) is a primary cutaneous T-cell lymphoma associated with the clonal proliferation of epidermotropic T-lymphocytes residing in the skin. MF accounts for more than half of all cutaneous lymphomas (21), and its reported frequency varies worldwide (22). The prevalence in the United States is ~0.58 cases per 100,000 person/years (23), and across Europe varies between 0.2-0.38 per 100,000 persons/year (24, 25), while in Uruguay, the prevalence is not well documented. In some cases, MF transforms into a more aggressive lymphoma that requires systemic chemotherapy, and the frequency of such transformations has been reported to range from 8% to 55% (26–29).

Here, we present the case of a family with two brothers who presented with mycosis fungoides as an exclusive symptom of CVID. Diagnosis was made via exome sequencing which revealed pathogenic variants in the NFKB2 gene. Based on this diagnosis and re-evaluation of other family members, the father and brother were also diagnosed with this rare immune and neoplastic syndrome. All CVID-affected family members presented with mycosis fungoides as their only symptom, which is, to the best of our knowledge, the first case series to be reported.

Here, we present the case of a family with CVID with an unusual clinical presentation diagnosed by WES. Three family members harbor a mutation in NFKB2 associated with CVID and they present mycosis fungoides as their only manifestation. To the best of our knowledge, this is the first case series reported in the region.

Patients with CVID have an inherently increased risk of lymphoproliferation, from non-malignant lymphoproliferation to lymphoma. This lymphoproliferative disorder is heterogeneous, with multiple factors including impaired immune surveillance, chronic infection, and genetic predisposition. Compared with the general population of the same age group, patients with CVID had higher rates of lymphoma, leukemia, gastric cancer, and skin cancer. Other known comorbidities of CVID include autoimmune diseases, interstitial lung diseases, bronchiectasis, enteropathies, liver diseases, and granulomatous diseases.

A registry of over 1000 patients with CVID confirmed lymphoma in 4.1% (32). 62% with the lymphoma subtype, 24.4% had a mature B-cell neoplasm, 11.1% had Hodgkin’s lymphoma, and only 4.4% had a mature T-cell neoplasm. A US-based study of 473 patients with CVID found that 8.2% of lymphomas were B-cell subtypes (33). These data are consistent with previous reports, in which the frequency of lymphomas ranges between 2-8% and of them, the T subtype is clearly less frequent; among them, we found no cases of mycosis fungoides (13, 34). However, mycosis fungoides is very infrequent, and familial presentation is extremely rare. Although its etiology is largely unknown, some evidence suggests that genetic and hereditary factors play a role in the pathogenicity of cutaneous lymphomas. To the best of our knowledge, only a few familial MF have been reported (35–39). Unfortunately, there is not much data on the mutational state of NFKB2 in patients with CVID and lymphoma, since the sequencing of NKFB2 is not part of the diagnosis algorithm of CVID.

Independent studies have reported strong linkage disequilibrium between MF and specific HLA class II allotypes, suggesting significant genetic susceptibility to these diseases (40). A study was conducted based on the Danish Twin Register (a cohort of 42 twins with cutaneous lymphomas (case twins) and their 42 co-twins, 420 age-and sex-matched twin controls (case controls), and 420 co-twin controls over a period of 30 years. The 42 twin pairs comprised 13 monozygotic and 27 dizygotic pairs. Contrary to expectations, neither co-twin was diagnosed with cutaneous lymphoma over the long follow-up period. The first study in a cohort of twins was unable to detect any familial aggregation of these diseases (41).

Father and both sons have a novel stop-gain variant in heterozygosity in the NFKB2 gene compatible with a CVID phenotype. The variant was classified as pathogenic according to the ClinVar and ACMG rules. Although it is a small family, the disease co-segregates with the variant in the family members studied, since all affected individuals have the variant, whereas the only unaffected relative studied (the sister of the father) does not.

The variant is absent in population frequency databases such as GNOMAD and has pathogenic in silico predictors (both conservational and physicochemical). The variant lies towards the end of the gene (at position 853); it has no in-silico prediction of activating the nonsense-mediated-decay (NMD) mechanism, but it compromises the last 48 amino acids of the protein, which are key for proper function. Previous studies have shown that NFKB2 protein that was produced from the R853* mutation was able to localize to the nucleus and activate transcription, suggesting that it was functional (no NMD). However, the authors also found that the protein was degraded more rapidly than wild-type NFKB2 protein (42). An additional study showed that R853* mutated protein was able to bind to DNA and activate transcription but was less effective than the wild-type (43).

Activation of NF-κB2 signaling requires a kinase cascade, resulting in the phosphorylation of p100 (the product of NF-kB2) at two conserved C-terminal Serines (Ser866, Ser870) (44) by a kinase called IKKκ. Then, the lysine at position 855 is ubiquitinated, which acts as a signal for the proteasomal processing of p100 to generate p52 (the active form), which associates with RelB and translocates into the nucleus, where this heterodimeric active complex acts as a transcription factor. The premature heterozygous stop codon affects the region of the protein required for posttranslational modifications, specifically phosphorylation and ubiquitination, leading to proteasomal processing. This likely results in reduced (but not completely blocked) protein activation and nuclear translocation as shown in animal models (2). The pathogenesis mechanism in humans with heterozygous NFKB2 mutations results from the haploinsufficiency condition, affecting only the noncanonical NF-κB pathway.

Phenotype refinement was achieved by PIDOT cytometry. It showed a reduced percentage of switched memory B cells (<70% of age-related normal values), a phenotype described in association with LOF variants in NFKB2, thus supporting the diagnosis (17).

To the best of our knowledge, hereditary MF as the sole manifestation of CVID is rare. It might have an additional genetic component (not only the NFKB2 variant) detectable in the patient’s genome. Hence, we analyzed other genomic variants present in both siblings with low frequencies and potential pathogenic impacts. We found a heterozygous frameshift variant in the ECM1 gene with a population frequency of 0,000015, that falls within the first 45-48% of the protein (depending on the transcript) and is predicted to activate an NMD mechanism and be damaging. Although variants of this gene are related to a dermatological disorder with generalized thickening of the skin and lipoid proteinosis of Urbach and Wiethe (OMIM:247100), it seems unlikely that this variant has an impact on the phenotype. Both brothers were heterozygous for this autosomal recessive disorder and did not have any elements suggestive of this phenotype. Hence, we did not speculate the role of this variant in ECM1 in this case. However, we do not exclude the possibility that the NFKB2 variant might act as a first “driver-hit” that might need a second-hit to determine lymphoma development favored by the immunodeficiency.

Since a molecularly defined entity associated with NFKB2 has not been developed yet it is still named as “Immunodeficiency common variable” (see: https://www.omim.org/entry/615577). We propose to depart from the “umbrella diagnosis” of CVID when a specific cause is found (45).

This case highlights the importance of using WES or WGS for proper diagnosis and treatment of hereditary hematological disorders.

The original contributions presented in the study are included in the article/ Supplementary Material , further inquiries can be directed to the corresponding author/s.

The studies involving humans were approved by Ethics committee of the Institut Pasteur de Montevideo, Mataojo 2020, Montevideo Uruguay. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study. Written informed consent was obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article. Informed consent to participate in the study and its publication was obtained from the participants and their families.

MNS, SG: treated and followed the patient and family. Wrote the manuscript. CS, MB: bioinformatic analysis of whole exome data. Reviewed the manuscript. MBo, SR: followed the patients in the hematological service. AB, JG-L, DL, NT: cytometry analysis. Reviewed the manuscript. VR: Interpretation of variants. Wrote original draft. LS: Bioinformatic analysis, interpretation of variants and funding acquisition. Wrote original draft. All authors contributed to the article and approved the submitted version.