Mycobacterial diseases in patients with inborn errors of immunity

Abstract

Clinical disease caused by the agent of tuberculosis, Mycobacterium tuberculosis, and by less virulent mycobacteria, such as bacillus Calmette-Guérin (BCG) vaccines and environmental mycobacteria, can result from inborn errors of immunity (IEIs). IEIs underlie more than 450 conditions, each associated with an impairment of the development and/or function of hematopoietic and/or non-hematopoietic cells involved in host defense. Only a minority of IEIs confer predisposition to mycobacterial disease. The IEIs underlying susceptibility to bona fide tuberculosis are less well delineated than those responsible for susceptibility to less virulent mycobacteria. However, all these IEIs share a defining feature: the impairment of immunity mediated by interferon gamma (IFN-γ). More profound IFN-γ deficiency is associated with a greater vulnerability to weakly virulent mycobacteria, whereas more selective IFN-γ deficiency is associated with a more selective predisposition to mycobacterial disease. We review here recent progress in the study of IEIs underlying mycobacterial diseases.

Introduction

The genus Mycobacterium includes more than 190 recognized species displaying physiological and phylogenetic differences and with different growth rates. These gram-positive bacteria are acid- and alcohol- fast, and these characteristics are used to distinguish them from other bacteria. Most environmental mycobacteria (EM) are isolated from soils (e.g. Mycobacterium fortuitum, M. avium complex, M. flavescens), water and aerosols (e.g. M. gordonae, M. simiae), and all humans come into contact with these bacteria via the skin and mucosa. In particular conditions, they can cause rare cases of localized or disseminated clinical disease in humans [1]. Most children worldwide are vaccinated with Mycobacterium bovis bacille Calmette-Guérin (BCG) vaccines [2], which are innocuous in the general population and provide partial protection against disseminated tuberculosis (TB) in childhood, but almost no protection against pulmonary TB in adults. Two mycobacterial species, M. tuberculosis (M.tb) and M. leprae, are particularly virulent and human-tropic. They cause TB and leprosy, respectively, when individuals exposed to these mycobacteria in a sustained manner develop clinical disease [3–6]. M. ulcerans is responsible for Buruli ulcer [7].

Human vulnerability to invasive, disseminated and/or recurrent mycobacterial infections is controlled by a combination of environmental and non-microbial factors, and by genetic and non-genetic host conditions [8, 9]. More than 450 inborn errors of immunity (IEIs) have been described, each impairing the development and/or function of hematopoietic and/or non-hematopoietic cells in host defense [10, 11]. Not all patients with IEIs develop mycobacterial disease following BCG vaccine, infection with EM or even exposure to M.tb. We review the available literature on patients with IEIs and mycobacteriosis, focusing, in particular, on BCG, EM and M.tb. The IEIs underlying clinical disease due to weakly virulent mycobacteria are much better delineated than those underlying bona fide TB, because of universal exposure to environmental mycobacteria and the high rate of BCG vaccination worldwide. By contrast, TB is no longer endemic in the countries in which IEIs are most widely diagnosed. This review does not cover acquired immunodeficiency (such as immunosuppressive treatments, AIDS), inherited conditions affecting the lungs (such as primary ciliary dyskinesia, or pulmonary alveolar proteinosis) and susceptibility to M. leprae and M. ulcerans, as this field has been widely explored and IEIs are rarely diagnosed in the areas in which these two mycobacterial species are endemic.

Severe combined immunodeficiency (SCID) underlies BCG disease

SCID is a heterogeneous group of IEIs characterized by a profound deficiency of autologous T cells associated, in some cases, with an absence of B and/or NK cells. Genetic etiologies are classified according to the cellular abnormality and typically designed SCID B+/− NK+/− [10]. Infants with SCID are at high risk of developing life-threatening infections caused by all types of microorganisms (fungi, various viruses, bacteria). Several SCID patients have developed clinical manifestations associated with BCG vaccine (Supplemental Table 1) [12–16]. Fever, ulcer or abscess at the vaccination site, pneumonia, hepatosplenomegaly, osteomyelitis, hepatic abscess, inguinal abscess, and meningoencephalitis have been reported in the context of localized (BCG-itis) or disseminated (BCG-osis) disease. One study of a cohort of 349 BCG-vaccinated SCID patients from 17 countries showed that 51% developed BCG-associated infections [12]. Surprisingly, children vaccinated after the age of one month developed fewer BCG-associated infections (37.9%) than children vaccinated earlier (55.4% BCG-associated infection and 18% BCG-associated mortality) and suffered almost no BCG-associated mortality, although the mechanisms underlying this difference remain unclear. Contrasting with the numerous cases of BCG and SCID, there have been very few reports of an association of SCID with EM infection (disease caused by M. avium complex, M. marinum, M. genavense or M. intracellulare, each in one case) [17–20]. Even fewer reports of associations of SCID with typical mycobacteria have been published, with only two cases known, one caused by M.tb and the other by M. bovis [21, 22]. The rarity of reported cases may reflect the high mortality from other infectious diseases before exposure to mycobacteria. Alternatively, innate immunity may control EM infections, at least partially. The number of T cells is particularly low, suggesting that the severity of mycobacterial infectious disease is directly correlated with the severity of the T-cell defect.

Combined immunodeficiency (CID) underlies BCG, EM and M.tb disease

The use of whole-exome sequencing (WES) in the field of primary immunodeficiencies (PIDs) or IEIs has led to the identification of many new combined immunodeficiencies (CIDs) [10, 11]. These diseases are less severe and profound than SCID. However, affected patients are nevertheless susceptible to several pathogens. The genetic defects of CID patients displaying susceptibility to mycobacteria are annotated in Supplemental table 1. BCG vaccination status was unknown in many reported cases. Mycobacterial disease is observed only rarely in patients with some defects, such as PNP (two patients with BCG), NBS (two patients with TB), SMARCAL1 (one patient with M. avium), MHC class II deficiency (two patients with BCG) and ZAP70 (two patients with BCG), raising questions about whether the relationship between these defects and diseases is truly causal or merely coincidental [3]. Mycobacterial disease (due to BCG and EM) occurs at higher frequency in patients with other types of CID, such as AR FOXN1 (one patient with BCG, from 8)[23], AR FCHO1 (one patient with M. genavense, from 15) [24] and AR POLE2 (one patient with BCG, from one patient with the defect) [25] deficiencies. No cases of TB have been reported in patients with these deficiencies. These deficiencies affect T-cell counts and/or functions, certainly accounting for the susceptibility of these patients to mycobacteria. However, studies of larger numbers of patients are required to test this hypothesis. Patients with defects of CD40L and NF-kB signaling (i.e., patients with defects of CD40L, CD40, IKBKG, NFKBIA, IKBKB, NIK, REL, and CARD11) are more susceptible to mycobacterial disease. Excluding patients with deficiencies of CARD11 and c-rel, more than half the patients identified have suffered from at least one mycobacterial episode (BCG, EM or M.tb). Interestingly, three cases of TB have been reported in the 18 patients with AD LOF CARD11 (none with BCG and EM), suggesting a higher clinical penetrance for tuberculous mycobacteria [26]. The only patient with c-rel deficiency reported to date had TB, and was vaccinated with BCG with no adverse effect [27]. In this group of patients, susceptibility to mycobacteria may be related to a defect of T cell-dependent IL-12/IL-23-producing monocytes, leading to an impairment of IFN-γ secretion by T cells, as shown in NEMO-deficient patients [28]. Overall, patients with these CIDs are more susceptible to disease caused by mycobacteria that are normally non-pathogenic (BCG and EM), and they can develop disease due to M.tb, a feature not seen in SCID patients, probably due to a higher levels of exposure to M.tb itself given that the defect is less severe and has a lower mortality.

Diseases of immune dysregulation mostly underlie EM and M.tb disease

Several members of this heterogeneous group of diseases defined by International Union of Immunological Societies (IUIS) are associated with mycobacteriosis, to which AD STAT1 gain-of-function (GOF) must be added (Supplemental Table1) [10, 29]. Indeed, patients with heterozygous STAT1 GOF mutations display a broad spectrum of clinical phenotypes, including chronic mucocutaneous candidiasis (CMC) and autoimmunity [30, 31]. Enhanced STAT1-dependent responses to IFN-α/β, IFN-γ, and IL-27 are observed in patients with STAT1 GOF, due to the impairment of STAT1 nuclear dephosphorylation [32]. More than 400 patients have now been described; the largest study included 274 patients [31] and unexpectedly showed that 6% of these patients had suffered from BCG, EM or M.tb disease (5, 6 and 6 patients, respectively). The precise molecular mechanism underlying this susceptibility remains unknown, as a lack of IFN-γ-mediated immunity and AD STAT1 LOF underlie susceptibility to mycobacteria [33, 34]. Patients with AD STAT3 GOF mutations display early-onset multiorgan autoimmunity, lymphoproliferation, recurrent infections and short stature [35, 36]. STAT3 is highly pleiotropic. More than 50 patients with STAT3 GOF have been identified, three of whom suffered from EM (M. avium and M. abcessus) and M.tb disease. The reasons for this susceptibility remain unclear. AR RIPK1 deficiency (characterized by lymphopenia, arthritis and intestinal inflammation) was recently reported in 13 patients, one of whom presented an episode of M. avium infection [37, 38]. Another defect, AR RASGRP1 deficiency, was recently described in nine patients with autoimmune lymphoproliferation [39]. Two of these patients had TB. In addition, more than 30 patients with RLTPR/CARMIL2 deficiency have been described, and these patients present recurrent infections, eczematous or psoriaform dermatitis, inflammatory bowel disease, and malignancy [40, 41]. Two patients suffered from TB and one from M. chelonae disease. Additional studies of larger numbers of patients are required to understand the susceptibility to mycobacteria observed in these conditions, which probably involve an impairment of T-cell counts or function related to IFN-γ production. BCG disease has been observed in STAT1 GOF cases, but not with the other defects, highlighting the preservation of some antimycobacterial activity and differences in clinical penetrance.

Chronic granulomatous disease (CGD) underlies BCG and M.tb disease

CGD is an IEI characterized by an absence of or the impaired production of reactive oxygen species (ROS), due to defects of the nicotinamide adenine dinucleotide phosphate (NADPH) oxidase complex in phagocytic cells (neutrophils, monocytes, macrophages, and dendritic cells [DCs]) [42, 43]. Monoallelic mutations of the CYBB gene or biallelic mutations of the NCF1, NCF2, NCF4, CYBA and CYBC1 genes underlie CGD [44, 45]. The expression of these CGD genes is controlled by IFN-γ. Patients diagnosed with CGD are highly susceptible to multiple life-threatening bacterial (including mycobacterial) and fungal infections, and, less commonly, parasitic infections. In countries in which TB is an endemic infectious disease and BCG vaccination is mandatory, some CGD patients may develop disease caused by BCG and/or M.tb (Supplemental Table 1) [46–49]. The penetrance for mycobacterial disease is incomplete in this group of patients. BCG infection is often the first clinical manifestation in CGD patients, due to an early vaccination, and it generally presents as lymphadenitis (BCG-itis), although BCG-osis (generalized disease) may also occur [46]. CGD patients present pulmonary disease or recurrent, severe infection with M.tb. Rare cases of EM disease (M. avium, M. flavescens, M. fortuitum, M. gordonae) have been reported, suggesting that the NADPH oxidase complex is redundant in the control of these types of mycobacteria [47]. Strikingly, two hypomorphic mutations of CYBB are responsible for Mendelian susceptibility to mycobacterial disease (MSMD), affecting ROS production in monocyte-derived macrophages (MDMs) and Epstein-Barr virus-transformed B lymphocytes (EBV-B cells), but not in neutrophils, monocytes or monocyte-derived DCs (MDDCs) [50]. These patients suffered from BCG disease or TB. These findings suggest that ROS production in neutrophils, monocytes and DCs is essential for the control of extracellular bacteria, whereas ROS production in macrophages seems to be essential for the control of intracellular bacteria, including mycobacteria.

GATA2 deficiency underlies EM and M.tb infections

GATA2 is a component of the GATA-binding transcription factor, which plays an important role in the survival and self-renewal of hematopoietic stem cells and in the specification of both the lymphoid and myeloid lineages [51]. Germline monoallelic mutations of GATA2 confer a broad spectrum of diseases, including myelodysplasia (MDS), chronic and acute myeloid leukemia (AML), familial MDS/AML, Emberger syndrome, and monocytopenia with mycobacterial infections (MonoMAC) [52–55]. Patients with MonoMAC typically display progressive monocytopenia, with or without a loss of DCs, B and NK cells [51, 55, 56]. Mycobacterial infections in these patients are caused principally by EM and M.tb (Supplemental Table 1) [56]. The most frequently identified EM are M. kansasii and M. avium, followed by M. fortuitum and M. abscessus. Rare infections caused by M. szulgai, M. malmoense, M. genavense, M. sherisii, and M. fukienense have also been reported in sporadic cases [56]. One recent study of a cohort of MonoMAC patients reported a mean age at first mycobacterial infection of 22.5 years [56]. This probably reflects the age-dependent decline of immunity in these patients, also accounting for the absence of BCG disease following vaccination in these patients. The penetrance for mycobacterial infection was incomplete at the age of 40 years, with no major difference observed between mycobacterial disease and other clinical phenotypes. Other factors, such as modifier genes in humans, environmental exposure, somatic mutations and the type of microbe, probably contribute to the heterogeneity of this syndrome. BCG vaccination was documented in this recent cohort, but the notion of BCG vaccination and its secondary effects is missing from many studies. Only one patient with BCG-osis has been reported. BCG disease is very rare in patients with GATA2 deficiency, by contrast to those with CGD. These findings suggest that GATA2 is not essential for the control of BCG but is necessary for the control of EM and M.tb.

Mendelian susceptibility to mycobacterial disease (MSMD), BCG and EM

Our understanding of the pathogenesis of mycobacterial diseases has been greatly improved by the study of patients with MSMD (Table 1 and Figure 1). These patients have inherited defects conferring a selective predisposition to clinical disease caused by weakly virulent mycobacteria, such as BCG vaccines and EM [1, 33]. Many different EM species have been isolated from patients, including M. abscessus, M. asiaticum, M. avium, M. avium complex, M. bohemicum, M. chelonae, M. elephantis, M. fortuitum, M. genevense, M. gordonae, M. kansasii, M. mageritense, M. peregrinum, M. porcium, M. scrofulaceum, M. simiae, M. szulgai, M. triplex, M. tilburgii, M. smegmatis, and M. szulgai [57]. The more virulent species M.tb has also been implicated in clinical disease in some patients [3]. A wide range of clinical symptoms, from localized to persistent, disseminated infections with impaired granuloma formation, has been observed in MSMD. Macrophage activation syndrome has been reported in some patients with uncontrolled mycobacterial disease [58]. Salmonella infections have occurred in about half the reported cases. A large proportion of MSMD patients suffer from CMC [33, 57]. Other severe infections have more rarely been reported, including viral, parasitic and bacterial diseases. Standard hematological and immunological screening results for classic PIDs are generally normal in MSMD patients [33, 57]. Disorders of 18 genes (IFNG, IFNGR1, IFNGR2, STAT1, IL12B, IL12RB1, IL12RB2, IL23R, RORC, TBX21, IRF8, SPPL2A, ISG15, TYK2, JAK1, ZNFX1, NEMO, CYBB) have been discovered, defining up to 33 genetic disorders, due to high levels of allelic heterogeneity [33, 59–67]. All the genetic disorders underlying MSMD are directly related to IFN-γ, the macrophage-activating factor (MAF) crucial for the destruction of intracellular mycobacteria by myeloid cells. The clinical severity of MSMD varies considerably between and even within affected kindreds, and increases with decreasing levels of IFN-γ. MSMD involves both myeloid and lymphoid cells, with selective impairments of the function and/or development of different subsets of these cells. “Isolated” MSMD was the first group of IEIs conferring a selective predisposition to one or a few infectious agents to be described [33, 68, 69]. “Syndromic” MSMD is the combination of the mycobacterial infection phenotype with another, equally common phenotype, infectious or otherwise [70–75]. The identification of new genetic etiologies of MSMD, although elusive, should extend the list of molecules known to be involved in the control of human IFN-γ immunity.

Table 1 :

Mycobacterium species isolated from patients with Mendelian susceptibility to mycobacterial diseases (MDM) or isolated tuberculosis (TB)

Gene	Disease	Mycobacterium species
IFNGR1	AR complete	BCG M. abscessus, M. avium, M. avium intracellulare, M. chelonae, M. fortuitum, M. scrofolaceum, M. spp. M. tuberculosis
	AR partial	BCG M. abscessus, M. avium, M. avium complex, M. szulgai M. tuberculosis
	AD partial	BCG M. abscessus, M. asiaticum, M. avium complex, M. avium, M. avium intracellulare, M. bohemicum, M. chelonei, M. gordonae, M. kansasii, M. scrofulaceum M. tuberculosis
IFNGR2	AR complete	BCG M. abscessus, M. avium, M. avium complex, M. fortuitum, M. kansasii, M. porcium, M. simiae, M. smegmatis, M. szulgai
	AR partial	BCG M. fortuitum, M. simiae
	AD partial	BCG
STAT1	AR complete	BCG M. abscessus, M. avium, M. kansasii, M. malmoense, M. scrofulaceum, M. spp.
	AR partial	BCG M. avium, M. szulgai
	AD partial	BCG M. avium M. tuberculosis
IFNG	AR complete	BCG
IL12RB1	AR complete	BCG M. avium, M. avium intracellulare complex, M. intracellulare, M. chelonae, M. fortuitum, M. fortiutum-chelonae complex, M. genevense, M. gordonae, M. kansasii, M. simiae, M. tilburgii, M. triplex M. tuberculosis
IL12RB2	AR complete	BCG M. tuberculosis
IL23R	AR complete	BCG
IL12B	AR complete	BCG M. chelonae M. simiae M. tuberculosis
TYK2	AR complete	BCG M. intracellulare M. tuberculosis
	AR partial	BCG M. avium complex M. tuberculosis
SPPL2A	AR complete	BCG
TBX21	AR complete	BCG
ISG15	AR complete	BCG
RORC	AR complete	BCG M. tuberculosis
IRF8	AR complete	BCG
	AD partial	BCG
JAK1	AR partial	M. gordonae, M. malmoense, M. scrofulaceum
ZNFX1	AR complete	BCG M. tuberculosis
PDCD1	AR complete	M. tuberculosis
IKBKG	XR partial	BCG M. tuberculosis
CYBB	XR partial	BCG M. tuberculosis

AD : autosomal dominant ; AR : autosomal recessive ; BCG: Bacille Calmette-Guerin ; XR : X-linked recessive

Figure 1:

Schematic representation of the cooperation between phagocytes/dendritic cells and T lymphocytes/NK cells during a mycobacterial infection. The molecules, encoded by genes found mutated in patients with isolated or syndromic MSMD and isolated TB in highlighted in red and are involved in IFN-γ mediated immunity.

Monogenic IEIs conferring a predisposition to TB

Most patients display TB as their sole clinical phenotype. The genetic study of MSMD paved the way for the demonstration that TB can also be genetic, as described in a dozen patients with AR IL-12Rβ1 and TYK2 deficiencies due to rare mutations (found in no more than 1/600,000 individuals) (Table 1, Figure 1 and Supplemental Table 1) [3, 75–81]. IL-12/IL-23-mediated IFN-γ production is impaired in the cells of these patients. Most of the TB patients were vaccinated with BCG with no adverse effects, revealing incomplete clinical penetrance for MSMD, but susceptibility to the more pathogenic species M.tb. However, these rare defects cannot account for the status of TB as a global public health problem. Two recent studies showed a strong enrichment in individuals homozygous for a common variant of TYK2 (P1104A) (a genotype found in 1/600 Europeans and 1/5,000 people elsewhere, with the exception of sub-Saharan Africa and Eastern Asia) in two cohorts of TB patients: one originating from countries outside Europe in which TB is currently endemic and the other, a large cohort of European ancestry (United Kingdom (UK) Biobank) [62, 82]. Homozygous carriers also displayed a predisposition to MSMD, albeit to a lesser extent [62]. Studies of ancient DNA have revealed that the frequency of TYK2 P1104A has decreased significantly over the last 2000 years, due to the strong negative selection pressure exerted by TB [83]. The catalytic activity of TYK2 P1104A is impaired, and the leukocytes of patients homozygous for P1104A have a specific impairment in IL-23-dependent IFN-γ mediated immunity [62]. The estimated penetrance for TB is high (at least 50%) in areas of endemic disease, whereas that for MSMD is much lower, at no more than 0.5%, consistent with M.tb being much more virulent than BCG and EM. Homozygosity for P1104A TYK2 may account for ~1% of the TB cases in Europeans and ~0.33% of cases in most other regions of the world [82]. It has been estimated that TB may have killed up to one billion people in Europe over the last 2,000 years [82], suggesting that about 10 million people may have died due to homozygosity for TYK2 P1104A. Interestingly, homozygosity for this variant has also been shown to be highly protective (ORs from 0.1 to 0.3) against various inflammatory or autoimmune disorders [84, 85], justifying the use of TYK2 inhibitors to treat some of these conditions. However, it will be important to evaluate the risk of TB before and during such treatment.

Neutralizing anti-IFN-γ autoantibodies

Neutralizing anti-IFN-γ autoantibody (nAIGA) production causes an adult-onset immunodeficiency, the incidence of which differs between ethnic groups, characterized by vulnerability to mycobacterial infection, including disseminated EM infection in particular (Supplemental Table 1)[86]. The immunodeficiency resulting from nAIGA production is an autoimmune phenocopy of the genetic forms of MSMD, as for anti-IL-6 autoantibody (aAb) production leading to pyogenic bacterial infections and IEIs of the IL-6 circuit, anti-T_H17-cytokine aAbs underlying fungal infections and IEIs of the IL-17 circuit, and anti-type I IFN antibodies responsible for severe coronavirus disease 2019 (COVID-19) and IEIs of the type I IFN circuit [86–88]. More than 520 patients producing nAIGAs have been described [89, 90]. Only two of the reported patients were children or adolescents [91], and only five patients (two South Africans, two UK residents and one US Caucasian) originated from places other than South Asia, East Asia or Southeast Asia [92, 93]. Almost all the patients with nAIGAs had EM infections; about half had salmonellosis, ~30% had at least one episode of cutaneous VZV infection, ~ 20% had cryptococcosis, ~15% had histoplasmosis, ~15% had talaromycosis, and ~9% had burkholderiosis [94]. However, determinations of the exact proportions of infectious phenotypes in patients with nAIGAs are complicated by ascertainment biases, as almost all early reports confirmed the presence of isolated or disseminated EM infection, whereas a recent study focusing on Talaromyces marneffei reported that only 12.7% of cases with nAIGAs presented with EM infection [95]. Other autoantibodies can be detected in these patients, such against GM-CSF, IL-10 or IFN-α1 but the relation with mycobacterial disease remains unknown [94]. HLA-DRB1*15:02/DQB1*05:01 and HLA-DRB1*16:02/DQB1*05:02, two haplotypes highly prevalent in Asia, are strongly associated with nAIGAs. Almost all patients with nAIGAs genotyped to date carry at least one risk haplotype [96, 97]. Some synergic effects are observed in patients carrying both risk haplotypes [97]. However, the only HLA-typed patient originating from a continent other than Asia to date carries neither of these two haplotypes, suggesting that this HLA association may be ethnicity-related [98]. Nevertheless, it remains unclear whether these HLA haplotypes are causal.

Conclusion

A large number of new IEIs have been discovered: some without (e.g. antibody deficiency, neutropenia, complement or asplenia) and a few with susceptibility to mycobacteria. The latter have been shown to affect mostly T cells, and myeloid cells, such as monocytes. Patients with earlier, more severe and broader defects are more vulnerable to BCG (i.e., SCID). By contrast, milder, more specific defects with a later onset tend to be more associated with clinical disease due to M.tb. However, since the first description of a small French cohort of patients with BCG disease 25 years ago [99], tremendous progress has been made towards understanding interindividual clinical variability in susceptibility to non-pathogenic mycobacteria. In particular, the discovery and functional characterization of 33 genetic etiologies of MSMD have revealed the essential, non-redundant role of IFN-γ in antimycobacterial immunity. These genetic defects probably account for MSMD in more than 50% of patients with this condition. By contrast, the genetic etiology is known for only a small percentage of patients with isolated TB. Rare and common genetic etiologies of TB impairing IFN-γ mediated immunity can account for, at most, 1% of European TB patients. MSMD and TB are a collection of single-gene IEIs, which can be allelic, with different clinical penetrances (correlated with pathogen virulence). Overall, the available data show that IFN-γ-mediated immunity is the crucial antimycobacterial circuit, acting as a genetically controlled continuous trait determining the type and outcome of mycobacterial infections [66]. However, further studies are required, as no genetic etiology has yet been identified for 50% of MSMD and 99% of TB patients. New genetic etiologies in patients with IEIs and mycobacterial disease are likely to reflect a connection to IFN-γ-mediated immunity. However, it remains possible that not all new IEIs underlying MSMD and TB will be linked to IFN-γ. Will MSMD and TB turn out to be allelic, or will new pathways be discovered? Similar questions apply to genetic inheritance. Are all cases monogenic? Incomplete penetrance is another important issue that we will need to address in the future.