www.thelancet.com Vol 372 August 9, 2008 489
Common variable immunodefi ciency: a new look at an
Miguel A Park, James T Li, John B Hagan, Daniel E Maddox, Roshini S Abraham
Primary immunodefi ciencies comprise many diseases caused by genetic defects primarily aff ecting the immune
system. About 150 such diseases have been identifi ed with more than 120 associated genetic defects. Although
primary immunodefi ciencies are quite rare in incidence, the prevalence can range from one in 500 to one in 500 000
in the general population, depending on the diagnostic skills and medical resources available in diff erent countries.
Common variable immunodefi ciency (CVID) is the primary immunodefi ciency most commonly encountered in
clinical practice, and appropriate diagnosis and management of patients will have a signifi cant eff ect on morbidity
and mortality as well as fi nancial aspects of health care. Advances in diagnostic laboratory methods, including
B-cell subset analysis and genetic testing, coupled with new insights into the molecular basis of immune
dysfunction in some patients with CVID, have enabled advances in the clinical classifi cation of this heterogeneous
Frequent sinopulmonary infections are a character-
istic clinical presentation in many patients with pri-
mary immunodefi ciencies. Antibody-related defects or
humoral primary immunodefi ciencies account for
65% of all primary immunodefi ciencies, whereas
defects in both the cellular and antibody compartments
account for another 15% of the cases. Among the
humoral primary immuno defi ciencies,
variable immuno defi ciency (CVID) generally comprises
antibody defi ciencies that present in either late
childhood or, more typically, early to mid adulthood.
The enormous heterogeneity in the clinical presentation
of CVID poses a challenge to primary-care physicians
who are most likely to encounter patients with the
disorder due to their predisposition to infections.
Delays in recognising CVID are common1–3 in
primary-care settings because of the pervasive
misconceptions that primary immuno defi ciencies are
extremely rare, that the disorders are largely restricted
to children, and that all patients are invariably moribund
or seriously ill at the time of presentation.4
In 1953, Janeway and colleagues5 were the fi rst to
report CVID in a 39-year-old with recurrent
Haemophilus infl uenzae meningitis. Although the
clinical entity of CVID has been known for over fi ve
decades, our understanding of the disease is far from
complete. In this article, we focus on recent
developments in this subject that have the potential to
improve initial clinical and laboratory assessments of
patients, enabling early appropriate diagnosis and
Although selective IgA defi ciency is the com monest
primary immunodefi ciency,
asymptomatic,6 and CVID is the commonest clinically
relevant primary immunodefi ciency. In a European
internet-based database, which included patients’ and
most patients are
research data on primary immunodefi ciencies, 30% of
the patients had CVID.7 Both sexes are aff ected equally,
and the prevalence of CVID ranges from one per 50 000
to one per 200 000 with a reported incidence of one
per 75 000 live births.8–10 Most patients have sporadic
disease, but 10–25% have familial inheritance, typically
with autosomal-dominant inheritance.11–14
The age at presentation of CVID has a bimodal
distribution. A few patients present in mid childhood
but most present in early to mid adulthood, although
some patients present even later. In one of the largest
series of patients with CVID (n=248), which included
patients aged 3–79 years, the mean age at onset of
symptoms was 23 years for males and 28 years for
females; while the mean ages of diagnosis were 29 years
and 33 years, respectively.15
Lancet 2008; 372: 489–502
Division of Allergic Diseases,
Department of Internal
Medicine (M A Park MD,
J T Li MD, J B Hagan MD,
D E Maddox MD), and Division
of Clinical Biochemistry and
Immunology, Department of
Laboratory Medicine and
Pathology (R S Abraham PhD),
Mayo Clinic, Rochester, MN,
Dr Roshini S Abraham,
Department of Laboratory
Medicine and Pathology,
Hilton 210e, Mayo Clinic College
of Medicine, 200 1st St SW,
Rochester, MN-55905, USA
Search strategy and selection criteria
We searched Medline (1950–present), Embase
(1988–present), Web of Science (1993–present), Cochrane
Database of Systematic Reviews (from inception),
Cochrane Central Register of Controlled Trials (from
inception), and SCOPUS with the term “common variable
immunodefi ciency” for the years. We used the terms
“immunologic defi ciency syndromes”, “immune defi ciency”,
“agammaglobulinaemia”, “hypogammaglobulinaemia” in
conjunction with keywords such as “CVID”, “common adj
variable”, “primary”, “humoral”, “BAFF-R”, “TACI”, “ICOS”,
“CD19”. We also used terms for the diseases and recurrent
infections commonly associated with the syndrome:
“sinusitis”, “rhinitis”, “pneumonia”, “asthma”. The searches
were further limited to large epidemiological studies,
cohort studies, clinical trials, meta-analysis, and areas
specifi cally related to diagnosis (laboratory diagnosis,
specifi city and sensitivity, diff erential diagnosis, diagnostic
accuracy, and clinical competence). We also searched the
references of relevant articles.
www.thelancet.com Vol 372 August 9, 2008
CVID has a broad and heterogeneous phenotype (fi gure 1)
that spans sinopulmonary and systemic bacterial infections
and gastrointestinal complications.15,16 Most of 248 patients
with CVID followed up for 1–25 years had recurrent
bronchitis, sinusitis, otitis media, and pneumonia; while a
few had viral hepatitis, severe Herpes zoster infection, and
Giardia enteritis.15 The frequency of infectious presenta-
tion diff ered slightly in paediatric CVID populations, in
which sinusitis is the commonest clinical presentation
followed by otitis media and pneumonia.17
The recurrent infections of both the upper and the
lower respiratory tract are over-represented for en cap-
sulated (H infl uenzae, Streptococcus pneumoniae) or
atypical (Mycoplasma spp) bacteria.8,15,18
25% of patients with CVID have autoimmune events
(fi gure 1). These events in CVID typify the underlying
immune dysregulation in these patients, in whom
specifi c checkpoints for autoreactivity during B-cell
development19 either fail or are circumvented. This
dysregulation leads to the generation of multiple
Villous atrophy and
Red blood cells
Figure 1: Organ systems involved in the pathogenesis of CVID
Left: healthy organs. Right: organ-system involvement. Patients also have increased risk of neoplasia, rheumatoid arthritis, vitiligo, and other autoimmune diseases.
Reproduced with permission from the Mayo Foundation for Medical Education and Research.
www.thelancet.com Vol 372 August 9, 2008 491
autoantibodies against various antigenic targets.20
Autoimmune thrombocytopenic purpura and auto-
immune haemolytic anaemia are the most common
autoimmune consequences, occurring in 5–8% of all
patients with CVID.2,15,21 Some patients have onset of
these disorders before the diagnosis of CVID. Therefore,
CVID needs to be considered in the diff erential diagnosis
of adult-onset autoimmune thrombocytopenic pupura
and autoimmune haemolytic anaemia.22 A third of
patients with CVID have splenomegaly.15,23 Other
autoimmune disorders include the presence of anti-IgA
antibodies,15,24 perni cious anaemia,20,25 and autoimmune
thyroiditis.26 Other less
consequences of CVID include rheumatoid arthritis,
vitiligo, and vasculitis.
Chronic pulmonary complications including recurrent
pneumonia are the primary cause of signifi cant
morbidity in patients with CVID (fi gure 1). Many
patients have anomalies of lung parenchyma visible on
chest radiographs and CT scans. The most common
pulmonary CT fi ndings include airway disease,
ground-glass attenuation, nodules, and parenchymal
opacifi cation.27 Pulmonary fi brosis and bronchiectasis
might also present;28 the latter is a common clinical
fi nding in CVID and other immunodefi ciencies.28–30
High-resolution CT is the best diagnostic tool for
bronchiectasis.28 As many as 50% of patients with CVID
may have other pulmonary features presenting with an
obstructive lung phenotype, such as chronic bronchitis
Multisystem granulomas are a well documented cause
of increased morbidity and mortality in patients with
CVID; the lungs are the most commonly aff ected site
(fi gure 1), though other organs, such as liver, skin,
spleen, and gastrointestinal tract can also be involved.15,32
Granulomatous disease in CVID might aff ect about
10–22% of patients1,33–35 and the mean age at onset is
18–34 years.33,34,36 Although granulomas are most
common in adults, up to a third of paediatric patients
with CVID can also have this complication.32
Presentation of granulomatous disease can precede a
diagnosis of CVID34 and the delay in recognition of
CVID can aff ect prognosis. Histologically, the
sarcoidosis. Angiotensin-converting enzyme concen-
trations are high in some patients with CVID.34,36 And
preliminary evidence suggests that a human herpes
virus 8 is a cause of systemic granulomatous disease
and lymphoproliferative disorders in some patients
with CVID.32,37 The prevalence of autoimmunity,
particularly autoimmune haemolytic anaemia, is
higher (>50%) in patients with CVID and granulomatous
disease than in those without (~47%).32,34 There are no
conclusive data showing that immunoglobulin changes
the course of granulomatous complications of
in CVID resemble
Laboratory featuresClinical features Further testing
Selective IgA defi ciencyLow titres of or absent IgA with normal
IgG and IgM concentrations
Usually asymptomaticNeeded if symptomatic
Selective IgG subclass
Low titres of one or more of the IgG
subclasses (G1, G2, G3, and G4); normal
total IgG concentrations unless IgG1 is
Modest to profound
hypogammaglobulinaemia with low
numbers or absence of peripheral-blood
Hypogammaglobulinaemia, EBV infection
is usually the trigger
Usually asymptomaticNeeded if symptomatic
Recurrent infectionsBTK protein expression by fl ow
cytometry;* gene sequencing* if protein
Aberrant response to EBV,
SH2D1A protein expression by fl ow
cytometry,* confi rmation by SH2D1A
Gene sequencing of six genes identifi ed
(Igμ heavy chain, IgVκA2, CD79A, CD79B,
λ5 surrogate light chain, BLNK)†
Protein expression by fl ow cytometry for
CD40L on activated T cells and CD40 on
B cells,* receptor-binding function for
CD40L, confi rmation by gene
sequencing is available for the fi ve genes
implicated (CD40L, NEMO, CD40, AID,
with absent or very low numbers of
peripheral B cells
Normal to high titres of IgM with low
titres of IgG and IgA, very low numbers or
absent class-switched memory B cells
Tends to manifest early in infancy or
childhood, recurrent and severe infections
Recurrent opportunistic sinopulmonary
infections, patients with NEMO defects
have hypohidrotic ectodermal dysplasia
and susceptibility to recurrent
Drugs, haematological malignancies, and other clinical phenotypes that can cause secondary hypogammaglobulinaemia are described in the text. *Tests available clinically in
specialised reference laboratories. †Tests available only in the research setting. BTK=Bruton’s tyrosine kinase. EBV=Epstein-Barr virus. CD40L=CD40 ligand. NEMO=NF-kB
essential modulator. AID=activation-induced cytidine deaminase. UNG=uracil DNA glycosylase.
Table 1: Diff erential diagnosis for patients with suspected CVID (hypogammaglobulinaemia with recurrent infections)
www.thelancet.com Vol 372 August 9, 2008
Gastrointestinal complications are fairly common in
CVID (fi gure 1)—up to 50% of patients with CVID have
chronic diarrhoea with
gastrointestinal diagnoses in patients with CVID include
Crohn’s disease, intestinal granulomatous disease,
intestinal parasitic bacterial or viral infections, coeliac
sprue, and intestinal lymphangiectasia.15 The incidence
of selective IgA defi ciency in patients with coeliac disease
is about ten-times higher than in the general
Results of three large clinical studies including 248,
220, and 176 patients suggest that patients with CVID
have a high risk of neoplastic disease—both
haematological and solid-tumour (breast, prostate,
ovary, skin, and colon). In particular, the incidence of
lymphoma is increased in patients with CVID.15,40,41 The
most common malignancies were non-Hodgkin
lymphoma15,40 and gastric cancers.40 Therefore, accurate
clinical and family histories of neoplasia along with
consideration of surveillance for malignancy, especially
lymphoma and gastric cancer in patients with CVID,
are appropriate. The surveillance approach has to be
applied judiciously without indiscriminate and frequent
use of radiological diagnostic procedures because some
patients with CVID might have increased radio-
Diff erential diagnosis
When fi rst assessing patients with recurrent infections
or with suspected CVID, several alternative diagnoses
must be considered. Due to the heterogeneous clinical
presentation of CVID, investigation of anatomical
anomalies of the lungs and sinus and of asthma and
allergic rhinitis must precede further investiga tion into
immune function. Other causes of hypo gamma-
globulinaemia that need to be ruled out include
protein-losing enteropathy, nephrotic syndrome, haema-
tological malignancies (including chronic lympho cytic
leukaemia, multiple myeloma, primary amyloidosis, and
non-Hodgkin lymphoma), and specifi c therapeutic
drugs. Corticosteroids, gold salts, penicillamine, anti-
malarial drugs, sulfasalazine, fenclofenac, phenytoin,
and carbamazepine among other drugs can cause hypo-
gammaglobulinaemia.43 In patients taking steroids,
hypogammaglobulinaemia is rarely associated with
functional antibody defects.43–45
Several other humoral immunodefi ciencies present
with hypogammaglobulinaemia and recurrent sino-
pulmonary infections, although most clinically relevant
defi ciencies usually present in infancy or childhood.
These include X-linked
autosomal-recessive agamma globulin aemia, X-linked
lymphoproliferative syndrome, and hyper-IgM syn-
dromes (table 1). Each of these immunodefi ciencies
can be eliminated as a diagnostic possibility by
combining clinical assessment with additional labora-
tory testing. A study of 60 male patients with CVID
indicates that defects in SH2D1A, the gene encoding
SH2 domain-containing protein 1A (SH21A; also
known as SLAM-associated protein), associated with
X-linked lymphoproliferative syndrome are very rare in
patients with CVID, and follow-up testing is needed
only for those with other associated clinical features of
SH21A defi ciency.47 The onset of the clinical phenotype
of X-linked lymphoproliferative syndrome is associated
with a history of infection with Epstein-Barr virus.
Panel: Laboratory investigations for CVID in the patients
who are HIV-negative in whom other causes of recurrent
infections have been ruled out
• Complete blood count with diff erential
• Serum immunoglobulins—IgG, IgA, and IgM
• Urine protein analysis (to rule out loss of
immunoglobulins due to nephrotic syndrome)
• IgG subclasses (IgGl to IgG4; useful for patients with
IgA defi ciency or history of recurrent sinopulmonary
• Functional antibody tests
• Protein: diphtheria toxoid, tetanus toxoid,
H infl uenzae B, isohaemagglutinins
• Polysaccharide antigen: S pneumoniae
• T-cell, B-cell, and natural-killer cell quantitation by fl ow
• Possibly test for lymphocyte proliferation for mitogens
and antigens (tetanus toxoid and Candida)
• B-cell subsets by fl ow cytometry* (to determine if there is
a reduction in class-switched memory B cells, and changes
in other B-cell subsets that correlates with certain clinical
• Class-switched memory B cells (CD27+ IgD- IgM-)
• Non-switched memory B cells (CD27+ IgD+ IgM+)
• IgM-memory B cells (CD27+ IgM+ IgDdull)
• Transitional B cells (CD38+++ IgM++)
• Plasmablasts (CD38+++ M-)
• Mature B cells (CD19+ CD21+)
• CD21lo B cells (CD19+ CD21lo)
• Protein expression for BAFF-R,† TACI,† and CD19† on
B cells and ICOS† on activated T cells by fl ow cytometry
• Mutation analysis by gene sequencing for TNFRSF13B†
• Mutation analysis for TNFRS13C and CD19 are presently
available only in specifi c research laboratories
*Tests may not aff ect diagnosis or management decisions but will provide additional
information on the underlying basis for the CVID presentation. †Tests are clinically
available in specialised reference laboratories.
www.thelancet.com Vol 372 August 9, 2008 493
Similarly, the possibility of X-linked agamma-
globulinaemia presenting as CVID should be
considered only in patients who present with less
than 1% of the normal number of B cells, because this
disorder is rare in adulthood.48 Autosomal-recessive
agamma globulin aemia and hyper-IgM defi ciencies are
genetically heterogeneous, with fi ve or six known
defects each. Confi rmatory testing (table 1) will enable
identifi cation of patients with specifi c genetic defects
associated with these diseases. Selective IgA defi ciency
and selective IgG subclass defi ciencies are typically
asymptomatic but should be investigated if there is
clinical evidence of recurrent infections without other
features of CVID.
Diagnosis of CVID
The well-accepted defi nition of CVID includes three key
features: the presence of hypogammaglobulinaemia of
two or more immunoglobulin isotypes (low IgG, IgA, or
IgM), recurrent sinopulmonary infections, and impaired
functional antibody responses.4,8,15 The criteria for
impaired functional antibody responses include absent
isohaemagglutinins, poor responses to protein (diph-
theria, tetanus) or polysaccharide vaccines (S pneumoniae),
or both. In addition to these, there can be other clinical
fi ndings including autoimmunity, granulomatous dis-
ease, and neoplasia.
After obtaining relevant personal and family history
and careful clinical examination, a systematic laboratory
assessment (panel)49,50 should be done.4 Clinical
examination should include assessment of key target
organs, such as pulmonary-function testing, ear nose
and throat review, and CT scans for sinusitis or
bronchiectasis. Testing for gastrointestinal complaints
and haematological anomalies is also useful. For
patients presenting with recurrent bacterial sino-
pulmonary infections, in whom other causes of
infections have been ruled out, one useful step-wise
approach (panel) to the laboratory assessment of such
patients would include (phase 1) a complete blood
count, urine protein analysis, and measurement of
serum immunoglobulin (IgG, IgA, and IgM) con-
centrations. The IgG concentrations in CVID are at
least two SD below the mean for the patients’ age8,18 and,
in most cases, accompanied by low concentrations or
absence of IgA,18 IgM, or both.
If there is hypogammaglobulinaemia, the next
laboratory tests (phase 2) would include quantitative
fl ow cytometric analysis of T, B, and natural killer cells,
functional antibody responses to protein antigens
(diphtheria toxoid, tetanus toxoid, Haemophilus
infl uenzae –Hib, isohaemagglutinins) and poly-
saccharide antigens (S pneumoniae vaccine; panel). The
importance of assessing functional antibody responses
in these patients cannot be overstated because the
results of these tests can determine whether patients
require immunoglobulin replacement or not.51 If
patients with recurrent infections have normal
concentrations of IgG and IgA, or IgA defi ciency alone,
then IgG subclass concentrations can be measured to
determine if there is an IgG subclass defi ciency.
Lymphocyte proliferative responses to mitogens and
specifi c antigens, such as Candida albicans and tetanus
toxoid, can also be measured, although this test is not
essential for the diagnosis, because only 20% of patients
with CVID have impaired proliferative responses,52–54 and
these are often associated with reduction in the CD4
count associated with a normal to increased CD8 count,
which can alter the ratio of CD4 to CD8.55,56 Numbers of
natural-killer cells, as determined by fl ow cytometry,
might also be low.57
The cellular characteristics of the immune system in
CVID are complex with several numerical and func tional
defects involving B cells, T cells, natural killer cells and
macrophages and monocytes.57–62 The number of B cells
in peripheral blood can be normal or reduced.15,63 T-cell
abnormalities are common including decreases in
number and function,15,64 defects in cytokine produc-
tion,65,66 decreased T-helper-cell function,67 abnormalities
in T-cell signalling,68,69 diminished expression of the
costimulatory molecule CD40 ligand,70 and increased
suppressor T-cell function.70,71
Memory B cellsTo blood
(secondary lymphoid organ)
To bone marrow
(spleen and blood)
Figure 2: B-cell development and diff erentiation
B cells develop in bone marrow from pluripotent haemopoietic stem cells through rearrangement of
immunoglobulin heavy-chain and light-chain genes and initial selection of the repertoire with selection against
autoreactive B cells. Mature B cells expressing both IgM and IgD are exported from bone marrow and enter
secondary lymphoid organs. Affi nity maturation takes place through somatic hypermutation of the variable region
genes in the germinal centre of the secondary lymphoid follicle where isotype class switching also takes place
through class switch recombination, which enables the production of IgG, IgA, and IgE isotypes. B cells selected
through affi nity maturation can become either memory B cells or long-lived plasma cells that home back to the
bone marrow and produce high-affi nity antibodies. HEV=high endothelial venule.
www.thelancet.com Vol 372 August 9, 2008
If the clinical presentation and laboratory assessment
(phase 1 and phase 2) are consistent with a CVID
phenotype, then analysis (panel) for defects in the
memory B-cell compartment and other peripheral B-cell
subsets may provide further information (phase 3 testing;
fi gures 2 and 3), particularly in relating changes in B-cell
subsets, such as class-switched memory B cells, to clinical
features of disease. The number of class-switched
memory B cells (CD27+ IgM– IgD–)72 is low in 50–75% of
patients with CVID;8,50,73,74 although it can also be low or
the cells absent in other humoral immunodefi ciencies,
such as hyper-IgM syndrome.75,76
In the past 5 years, the Paris73 and Freiburg74
classifi cations have attempted to defi ne CVID with
fl ow-cytometry techniques on the basis of the presence
or absence of class-switched memory B cells. Very
recently, however, data from the EUROclass trial35 unifi ed
the two classifi cations and provided clinical links with
results from the immunophenotyping of B-cell subsets
(fi gure 3). The EUROclass data are from a multicentre
European trial that assessed 303 patients with CVID and
showed that severe reduction in the number of
class-switched memory B cells is associated with
granulomatous disease, splenomegaly, and autoimmune
cytopenias.35 Their results also showed that increases in
other B-cell subsets, such as transitional B cells and
CD21lo B cells, were associated with lymphadenopathy
and splenomegaly, respectively.35
The origin and function of the various memory B-cell
subsets in the humoral response has been ardently
debated.77–82 However, there are enough data to prove a
role for these memory-B-cell subsets in generating
antibodies to both T-dependent and T-independent
antigens,83 and that changes in the B-cell memory
compartment, due to an underlying immunodefi ciency,
could have substantial eff ects on the quality and
quantity of the humoral immune response.
In CVID, the subgroups of B-cell defects that can be
classifi ed by fl ow cytometric analysis (panel) allow
categorisation of patients on the basis of the underlying
immune defect, although not all defects are clearly
known. Diff erent clinical laboratories have diff erent
approaches to diagnostic testing for CVID and clinical
associations might therefore diff er with the patients
Classifi cations that take class-switched memory and
non-switched memory B cells into account are useful
for diagnosis.50,84 In one study, a reduction in the
proportion of class-switched memory B cells was
associated with a higher incidence of bronchiectasis,
splenomegaly, and autoimmunity and was clinically
more informative than either serum immunoglobulin
concentrations or arbitrary grouping of patients as
having CVID or specifi c antibody defi ciency.50 In another
study, the absence or presence of IgM memory B cells
(CD27+ IgM+ IgDdull) and anti-IgM pneumococcal
polysaccharide antibody responses subclassifi ed patients
with CVID into those with recurrent bacterial pneumonia
and bronchiectasis and those without pneumonia and
lung lesions, respectively.84 Flow cytometric laboratory
assessment of peripheral B-cell subsets can be made
more amenable to routine clinical diagnostic testing by
use of whole blood instead of isolated peripheral-blood
Genetic defects in CVID
In the past 5 years, investigators have described defects
in four genes associated with CVID—inducible T-cell
costimulator (ICOS),86,87 tumour necrosis factor receptor
superfamily, member 13B (TNFRSF13B, also known as
TACI),88–91 tumour necrosis factor receptor superfamily,
member 13C (TNFRSF13C, also known as BAFFR),92 and
CD19 93 (table 2).94–96
The fi rst reported genetic defect associated with CVID
and characterised by a defi ciency of ICOS, which is
expressed on activated T cells, was identifi ed in nine
individuals with CVID from four apparently unrelated
families.86,87 These nine patients with mutations in ICOS
presented with recurrent
splenomegaly, autoimmune neutropenia, intestinal
lymphoid hyperplasia, and neoplasia.96 ICOS-defi cient
patients have few peripheral B cells, few or no
class-switched memory B cells, and hypogammaglobulin-
Total memory B cells
memory (marginal-zone) B cells
(CD19+ CD27+ IgM+ IgD+)
CD21lo B cells
Mature B cells
Transitional B cells
(CD19+ CD39+++ IgM++)
(CD19+ CD38+++ IgM–)
memory B cells
(CD19+ CD27+ IgM+ IgDdull)
memory B cells
(CD19+ CD27+ IgM– IgD–)
Figure 3: Peripheral blood B-cell subsets
The assessment of peripheral blood B-cell subsets is useful in the diagnosis of CVID. CD19 is a pan-B-cell marker
that allows identifi cation of all B-cell subsets in blood except normal plasma cells. CD27 typically indicates the
memory phenotype and presence or absence of IgM and IgD diff erentiates memory subsets. The expression of
CD38 and IgM distinguishes transitional B-cells and plasmablasts. CD21 is a marker for B-cell activation and is
expressed on mature B cells. Multiparametric fl ow cytometry allows both absolute quantifi cation and proportion
analysis of these subsets.
www.thelancet.com Vol 372 August 9, 2008 495
aemia.87 T cells from ICOS-defi cient patients produce
very little interleukin 10, which may be associated with
the defective formation of germinal centres leading to
impaired B-cell memory (fi gure 4).94
The defect seems to be inherited as an autosomal-
recessive trait because all nine patients have the same
homozygous deletion in ICOS. Heterozygous parents
and siblings of patients defi cient in ICOS are
asymptomatic, and genetic analysis of these nine patients
indicates a common founder ancestor,87 suggesting that
ICOS mutations cause CVID in only a few patients.
About 2% of patients with CVID have defects in this
ICOS is upregulated on both CD4 and CD8 eff ector
and memory T cells97 and activated natural-killer cells,
and it enhances natural-killer-cell function. The
stimulator’s ligand is expressed on lymphoid and non-
lymphoid tissue, whereas ICOS is expressed constitutively
in germinal centres and T-cell zones of spleen, lymph
nodes, and Peyer’s patches.98
Tumour necrosis factor superfamily
Two groups independently identifi ed mutations in
TNFRSF13B in 17 patients with CVID and one with
sIgAD in 2005.89,91 Mutations in TACI, the protein encoded
by TNFRSF13B, are associated with a clinical phenotype
of lymphoproliferation, which may include splenomegaly
or tonsillar hyperplasia, and IgA defi ciency, with
autoimmune thyroiditis being reported in 15% of patients
The clinical relevance of these mutations was recently
further analysed.88,90 Pan-Hammarstrom and colleagues90
studied 424 patients with CVID and sIgAD and
2209 healthy people from Sweden, Germany, and the
USA. The diff erences between patients and healthy
people in prevalence of some mutations in TACI, the
protein encoded by TNFRSF13B, such as the Cys104Arg,
Ala181Glu and 204insAla (also known as Leu69fsX11)
was statistically signifi cant, indicating that these
mutations even in heterozygous states are associated
with CVID; but they do not seem to be associated with
sIgAD. An Arg202His mutation, however, was
signifi cantly associated with with sIgAD. Prevalence of
several other mutations did not diff er signifi cantly
between patients and controls.89
In a second study of mutation in TACI, 212 patients
with CVID were compared with 124 healthy controls.
Only Cys104Arg and Ala181Glu mutations were
signifi cantly associated with CVID.88 The Arg181Glu
mutation is present in some healthy people but is
signifi cantly rarer than in patients.88 These healthy
controls might develop CVID later in life.15
In families with dominant inheritance of the Ala181Glu
mutation, some people with the mutation are completely
asymptomatic, suggesting that the mutation has
incomplete penetrance.90,91 In a very recent study of
176 patients with CVID by Zhang and co-workers,95
13 patients had heterozygous mutations in TACI. Five of
these 13 patients had the Cys104Arg mutation, while
another two were compound heterozygotes with the
Cys104Arg and Ser144X or Ser194X mutations. Three of
the 13 patients had the Ala181Glu mutation and a fourth
was a compound heterozygote for Ala181Glu and
Leu171Arg mutations.95 The remaining two patients had
a Cys172Tyr and Arg72His mutation respectively. A
previous study did not fi nd signifi cant association for
either Arg72His or Leu171Arg, which were found in two
and one of 212 patients, respectively, but not in any of
124 healthy people.88
In a smaller group of 53 patients with CVID assessed
at the Mayo Clinic (unpublished), we found six patients
with heterozygous TACI mutations: fi ve had the
Ala181Glu and one the Cys104Arg sub stitutions. Of
these patients, two were brothers, indicating a familial
inheritance. From these various studies, about
10–20% of patients with CVID seem to have mutations
Zhang and co-workers95 recently showed that the
presence of a heterozygous mutation in TNFRSF1B
alone does not seem to cause CVID. Whether this
fi nding is due to incomplete penetrance or delayed onset,
or whether additional genetic–environmental factors are
Inheritance B-cell-subset analysis by fl ow cytometryProtein expression on cell surface
TNFRSF13C (<1% of CVID
TNFRSF13B (10–20% of CVID
Autosomal recessiveReduced class-switched and non-switched memory
B cells with increased transitional B cells
Low to absent IgA, autoimmune disease,
lymphoproliferative disease, splenomegaly, reduced
class-switched memory B cells
Reduced class-switched memory B cells, nodular
lymphoid hyperplasia, autoimmunity,
predisposition to neoplastic disease
Decrease in class-switched memory B cells, low
CD21 expression on B cells, normal numbers of
CD20+ mature B cells in peripheral blood
BAFF-R expression is absent on B-cell surface
95% have normal TACI expression on B-cell
surface, <5% have absent TACI expression
ICOS expression on the surface of activated
T cells is absent
Low to absent expression of CD19 protein on
the surface of CD20+ B cells
BAFFR=B-cell-activating-factor-family receptor. *And our unpublished data.
Table 2: Gene defects in CVID, inheritance, and associated laboratory phenotype
www.thelancet.com Vol 372 August 9, 2008
required for the manifestation of an immune phenotype,
is not entirely clear.
Mutation of TNFRSF13C, which encodes BAFF-R, has
been described in only one individual with CVID aged
60 years who had a homozygous 24 base-pair deletion.92
In this patient, the mutation was associated with distinct
anomalies in peripheral B-cell subsets with profound
reduction of both class-switched (CD27+ M– D–) and
Naive T cell
of T cells
• ICOS induced on activated T cells
• Enhances T-cell responses, T-cell–B-cell cooperation
• Induces IL-10 (Th-1 differentiation inhibitor) and IL-17
• Augments isotype class-switching in B cells
of T cells
Activated T cell
Contributes to B-cell
signalling though BCR
APRIL B cell co-receptor
• B-cell survival
• Plasma-cell survival
• Isotype switching
• Isotype switching
Figure 4: Molecules implicated in genetic studies of CVID
(A) ICOS is a positive costimulator (like the constitutively expressed CD28) that enables T-cell interactions with B cells, monocytes, and dendritic cells. (B) BAFF-R and
TACI are cell-surface receptors belonging to the TNF-receptor family that play a part in B-cell diff erentiation and function. BAFF–BAFF-R interactions provide critical
survival signals for diff erentiation of peripheral B cells. The role of BAFF-TACI interactions is less clearly defi ned. TACI signals intracellularly through the TNF
receptor-associated factors (TRAF) to induce nuclear factor-κ-B activation.101 TACI also interacts intracellularly with calcium modulator and cyclophilin ligand (CAML).
Through interaction with APRIL, TACI regulates isotype class switching of immunoglobulins and the antibody response to T-independent antigens.89,94,99,100 (C) CD19 is
a B-cell-specifi c cell-surface marker that is part of the B-cell coreceptor along with CD21 and CD81;93 CD19 is expressed throughout B-cell maturation from pro-B cells
through to plasmablasts, before CD21 or CD81, it is not expressed on plasma cells. Coligation of the B-cell receptor (BCR) with the coreceptor complex of
CD19–CD21–CD81 increases B-cell signalling by several thousand times.102
www.thelancet.com Vol 372 August 9, 2008 497
non-switched memory or marginal-zone (CD27+ M+
D+) B cells with an increase in the transitional B-cell
compartment (CD38+++ M++) and a decrease in
plasmablasts (CD38+++ M–). Preliminary evidence of
other patients with this mutation from clinical
presentation and fl ow cytometric laboratory analysis
needs to be investigated with genetic testing.
TACI and BAFF-R ligands are expressed on
macrophages, monocytes, and dendritic cells.99,100 BAFF-R
interactions provide crucial survival signals for
diff erentiation of peripheral B cells whereas TACI
induces nuclear factor κB activation101 and regulates
class-switching of immunoglobulins and the antibody
response to T-independent antigens (fi gure 4).89,94,99–101
Four patients with CVID from two unrelated families
had homozygous mutations in CD19, resulting in
undetectable CD19 protein expression on B cells in one
patient and reductions in the level of expression in the
other three.93 Three of the four patients were siblings
and were diagnosed as adults, although they had been
symptomatic during childhood. The fourth patient was
diagnosed at age 10 years after recurrent infections
starting in infancy. In patients with this defect, the total
number of B cells (CD20+) in blood is normal with low
or undetectable surface expression of CD19 , and the
numbers of CD27+ memory B cells and CD5+ B cells
are decreased (fi gure 4).93,102 CD19 knockout mice have
defi cient antibody responses to most antigens. CD19 is
expressed on B cells from an early stage of development
and therefore seems to play a part in signalling through
the B-cell receptor even without forming a complex
with CD21 and CD81.
The genetic defects we have described account for only a
few patients with CVID (table 2), and nearly 75% of
patients have no known defect.
Flow cytometry screening for protein expression for
the four proteins implicated is available in specialised
laboratories in the USA and Europe. However, in the case
of mutations in TNFRSF13B, less than 5% of patients
have abnormal levels of protein expression on the cell
surface. The remaining TNFRSF13B mutations are
associated with functional defects. Therefore, fl ow
cytometry for TNFRSF13B expression is likely to be
uninformative in most cases.
Although genetic testing is at the forefront of diagnosis
for primary immunodefi ciencies, its use in CVID has to
be cautious and clinically justifi ed due to both the
expense and the medicolegal ramifi cations for both
patients and family members. At present, the diagnosis
and treatment of CVID do not require specifi c knowledge
of the underlying genetic defect; however, determination
of the genetic defect helps to understand the biology
and epidemiology of the disease. Other reasons for
genetic testing would be to enable early management of
complications associated with specifi c genetic defects
and to develop robust genotype–phenotype correlations
in a specifi c population of patients, which have not
always been reproducible in genetic studies.103 Also,
genetic testing may be helpful in the investigation of
familial cases of CVID, since two of the four genetic
associations (ICOS and CD19) reported so far have a
strong family-based association. In the case of
TNFRSF13B mutations, only a few patients have
evidence of familial bias, and most family members
with mutations are asymptomatic.95
Treatment and clinical management
Treatment of CVID
The main goal of therapeutic management in CVID is to
decrease the morbidity and mortality associated with
recurrent infections. Intravenous immunoglobulin is
eff ective104–106 and is currently the mainstay of therapy for
CVID.8 Intravenous immunoglobulin also reduces the
incidences of pneumonia107 and serious recurrent
bacterial infections15,107 and prevents chronic lung disease
and enteroviral meningoencephalitis.108
Immunoglobulin replacement can be given either
subcutaneously or intravenously.109 The current dosing
recommendations for intravenous immuno globulin
are 300–400 mg/kg body weight, every 3–4 weeks8 with
the IgG concentration maintained above 5 g/L;110
although some patients may benefi t from a higher
trough concentration, closer to 7 g/L. Alternatively,
subcutaneous delivery of immunoglobulin with
slightly more than a quarter of the monthly dose each
week is therapeutically comparable with intra venous
There can be both mild and serious adverse reactions
to the use of intravenous immunoglobulin. The minor
adverse reactions to intravenous immunoglobulin
include headache, nausea, malaise, myalgias, arthralgias,
chills, anxiety, fl ushing, abdominal cramps, rash,
low-grade fever, and leukopenia. Most of these can be
prevented by slowing of intravenous immunoglobulin
infusion, premedication of patients with 500–1000 mg
of paracetamol and 25–50 mg of diphenhydramine
orally, or both. The rarer but more serious side-eff ects of
intravenous immunoglobulin include anaphylaxis,113
acute renal failure,114 stroke,115 myocardial infarction,116
deep venous thrombosis or pulmonary embolus,117,118 and
Patients with CVID who have IgA defi ciency (IgA
<0·07 g/L) typically receive IgA-defi cient blood products.
The need for exclusive use of IgA-defi cient preparations
has been controversial.120–122 However, patients who have
anti-IgA antibodies24 and meet clinical criteria for
replacement therapy should also be considered for
IgA-depleted immunoglobulin products because of a
potential increased risk of anaphy lactic reactions.11,121,122
The use of subcutaneous immunoglobulin replacement
www.thelancet.com Vol 372 August 9, 2008
might lower the risk of many of the side-eff ects associated
with intravenous immunoglobulin. The minor reactions
of subcutaneous immunoglobulin replacement are local
infl ammation at the infusion site and, rarely, fever, chills,
and cold sweats.112,123
Antimicrobial drugs also play an integral part in the
treatment of CVID,8 because intravenous immuno-
globulin alone is not enough to prevent or eradicate all
active infections.124 The use of fl uoroquinolones and
amoxicillin clavulanate are eff ective in managing the
sinopulmonary infections in CVID. The role of
antimicrobial prophylaxis in addition to intravenous
immunoglobulin has not been defi nitively established
and needs further study.125
Autoimmunity and neoplasia associated with CVID
are commonly treated as per standard clinical practice in
patients without CVID,8 although theoretical risks of
additional immunosuppression exist. In particular, the
treatment of autoimmune and granulomatous diseases
in CVID presents a profound therapeutic challenge.
Corticosteroids are eff ective in combating many of the
autoimmune and granulomatous manifestations, but
the side-eff ects of corticosteroids may limit its long-term
effi cacy. New-generation monoclonal antibodies have
been used to treat some of the autoimmune and
granulomatous complications in CVID, but no systematic
double-blind, randomised clinical trials have investigated
their effi cacy and safety. Several case reports describe the
clinical usefulness of monoclonal antibodies. For
example, infl iximab has been used for Crohn’s disease
associated with CVID126 and also for caseating
granulomatous disease.127,128 Rituximab has been used
with some success in medically refractory severe
autoimmune thrombolytic purpura129 and autoimmune
haemolytic anaemia.130 Etanercept has been used to treat
scarring alopecia caused by sarcoidal granulomas in
patients with coexisting juvenile rheumatoid arthritis
Vaccinations, surveillance, and education
Because CVID is most commonly diagnosed in
adulthood, many patients are likely to have previously
received live vaccines for infectious diseases. Additional
immunity is provided for patients treated with
intravenous or subcutaneous immunoglobulin because
circulating antibodies are present in these preparations.
However, the measles-mumps-rubella and varicella
vaccines are not recommended in patients receiving
replacement immunoglobulin therapy, because the
vaccines may be inactivated by the presence of
neutralising antibodies.8,132 Inactivated vaccines can be
given to patients with CVID but these may not be
eff ective because of the underlying antibody defi ciency.8
Because infl uenza is unlikely to be represented in the
replacement-immunoglobulin, the inactivated-subunit
infl uenza vaccine is commonly recommended yearly as
Lifetime surveillance for cancer and autoimmunity
after the diagnosis of CVID is important so that eff ective
intervention can be started early if necessary.
Surveillance should include physical assessments and
blood tests when appropriate. Endoscopy may also be
useful in this screening protocol for the detection of
gastric carcinoma or mucosa-associated lymphoid-tissue
lymphoma. Patients should be assessed by physicians
familiar with their immunodefi ciency history every
6–12 months.8 As with all chronic diseases, history and
the physical examination should guide appropriate
investigations. For those patients who remain healthy
on replacement immunoglobulin, we recommend, at a
minimum, age-appropriate cancer screening as
endorsed for the general healthy population, such as
colonoscopy, prostate examination, pap smears, and
mammograms. There are currently no formal
recommendations on the frequency of radiography for
Finally, education on and raising the awareness of the
medical and social implications CVID are crucial for both
patients and family members, for whom the rarity and
substantial complexity of this disease can pose a
signifi cant emotional and fi nancial burden. There are
several resources available in the USA and Europe that
provide educational and social support for patients
with primary immunodefi ciencies and their families;
these include the Immune Defi ciency Foundation and
the Jeff rey Modell Foundation.
Summary and recommendations
Although the 2005 practice guideline for the diagnosis
and management of primary immunodefi ciency was
developed for allergy and immunology specialists and
not family doctors, it off ers helpful information on the
clinical and laboratory assessment of patients with
several kinds of immunodefi ciency disorders, including
humoral immunodefi ciencies. Algorithm 2 in the
2005 practice guidelines8 off ers a global diagnostic
algorithm for the assessment of antibody-related primary
immunodefi ciencies, although it does not provide
specifi c information on the clinical usefulness of new
laboratory tests, such as fl ow-cytometric B-cell subset
analysis and genetic testing in the assessment of CVID.
In summary, clinicians should consider CVID as a
possible diagnosis when assessing patients with
frequent bacterial sinopulmonary infections. There
should be judicious use of laboratory tests when
assessing such patients to prevent unnecessary testing
and expense. Many of the new and advanced laboratory
tests, such as peripheral-blood B-cell-subset studies,
specifi c protein analysis, and genetic testing for
CVID-associated mutations, are now clinically available
in specialised centres to aid in the diagnosis and
management of CVID. Because most patients with
primary immunodefi ciencies are fi rst seen by family
doctors, the goal of these recommendations for clinical
For the Immune Defi ciency
Foundation see http://www.
For the Jeff rey Modell
Foundation see http://www.
For the 2005 practice guidelines
www.thelancet.com Vol 372 August 9, 2008 499
and laboratory assessment of patients with CVID is to
promote prompt and accurate diagnosis in this
MAP and RSA contributed equally to the preparation and writing of
this Seminar. MAP participated in the planning, writing, and editing of
the paper and approved the submitted version. JTL, JBH, and DEM
participated in the conception, reviewing and editing of the paper and
approved the submitted version. RSA was involved in the conception
and preparation of the paper at all stages, including the reference
search, writing the text, preparation of fi gures, and editing.
Confl ict of interest statement
MAP and JBH are coinvestigators in a multicentre extension study on
the safety and effi cacy of IgPro10 in patients with primary immune
defi ciency sponsored by ZLB Behring; MAP, JBH, and DEM are
coinvestigators in a phase III open-label, prospective, multicentre
study of the effi cacy, tolerability, safety, and pharmacokinetics of
immune globulin subcutaneous IgPro20 in patients with primary
immunodefi ciency sponsored by ZLB Behring; MAP, JTL and JBH
were coinvestigators in ZLB03_002CR, a multicentre study of the
effi cacy, safety, and pharmacokinetics of IgPro10 in patients with
primary immunodefi ciency, sponsored by ZLB Behring—none of the
investigators received personal funding for their involvement in the
above studies. RSA received a USIDNET (US Immunodefi ciency
Network funded by the NIH) and Jeff rey Modell Foundation travel
scholarship to attend the IUIS/WHO meeting on primary
immunodefi ciencies in 2007.
1 Cunningham-Rundles C. Common variable immunodefi ciency.
Curr Allergy Asthma Rep 2001; 1: 421–29.
2 Hermaszewski RA, Webster AD. Primary
hypogammaglobulinaemia: a survey of clinical manifestations
and complications. Q J Med 1993; 86: 31–42.
3 Winkelstein JA, Marino MC, Johnston RB Jr, et al. Chronic
granulomatous disease: report on a national registry of
368 patients. Medicine (Baltimore) 2000; 79: 155–69.
4 Cunningham-Rundles C. Immune defi ciency: offi ce evaluation
and treatment. Allergy Asthma Proc 2003; 24: 409–15.
5 Janeway CA, Apt L, Gitlin D. Agammaglobulinemia.
Trans Assoc Am Physicians 1953; 66: 200–02.
6 Schroeder HW Jr. Genetics of IgA defi ciency and common
variable immunodefi ciency. Clin Rev Allergy Immunol 2000;
7 Eades-Perner AM, Gathmann B, Knerr V, et al. The European
internet-based patient and research database for primary
immunodefi ciencies: results 2004–06. Clin Exp Immunol 2007;
8 Bonilla FA, Bernstein IL, Khan DA, et al. Practice parameter for
the diagnosis and management of primary immunodefi ciency.
Ann Allergy Asthma Immunol 2005; 94 (5 suppl 1): S1–63.
9 Fasth A. Primary immunodefi ciency disorders in Sweden: cases
among children, 1974–1979. J Clin Immunol 1982; 2: 86–92.
10 McCluskey DR. Prevalence of primary hypogammaglobulinemia
in Northern Ireland. Proc R Coll Physicians Edinb 1989;
11 Hammarstrom L, Vorechovsky I, Webster D. Selective IgA
defi ciency (SIgAD) and common variable immunodefi ciency
(CVID). Clin Exp Immunol 2000; 120: 225–31.
12 Schroeder HW Jr, Schroeder HW 3rd, Sheikh SM. The complex
genetics of common variable immunodefi ciency. J Investig Med
2004; 52: 90–103.
13 Vorechovsky I, Cullen M, Carrington M, Hammarstrom L,
Webster AD. Fine mapping of IGAD1 in IgA defi ciency and
common variable immunodefi ciency: identifi cation and
characterization of haplotypes shared by aff ected members of
101 multiple-case families. J Immunol 2000; 164: 4408–16.
14 Vorechovsky I, Zetterquist H, Paganelli R, et al. Family and
linkage study of selective IgA defi ciency and common variable
immunodefi ciency. Clin Immunol Immunopathol 1995;
15 Cunningham-Rundles C, Bodian C. Common variable
immunodefi ciency: clinical and immunological features of
248 patients. Clin Immunol 1999; 92: 34–48.
16 Hermans PE, Diaz-Buxo JA, Stobo JD. Idiopathic late-onset
immunoglobulin defi ciency: clinical observations in 50 patients.
Am J Med 1976; 61: 221–37.
17 Ogershok PR, Hogan MB, Welch JE, Corder WT, Wilson NW.
Spectrum of illness in pediatric common variable
immunodefi ciency. Ann Allergy Asthma Immunol 2006; 97: 653–56.
18 Conley ME, Notarangelo LD, Etzioni A. Diagnostic criteria for
primary immunodefi ciencies: representing PAGID
(Pan-American Group for Immunodefi ciency) and ESID
(European Society for Immunodefi ciencies). Clin Immunol 1999;
19 Tsuiji M, Yurasov S, Velinzon K, Thomas S, Nussenzweig MC,
Wardemann H. A checkpoint for autoreactivity in human IgM+
memory B cell development. J Exp Med 2006; 203: 393–400.
20 Brandt D, Gershwin ME. Common variable immune defi ciency
and autoimmunity. Autoimmun Rev 2006; 5: 465–70.
21 Cunningham-Rundles C. Hematologic complications of primary
immune defi ciencies. Blood Rev 2002; 16: 61–64.
22 Michel M, Chanet V, Galicier L, et al. Autoimmune
thrombocytopenic purpura and common variable
immunodefi ciency: analysis of 21 cases and review of the
literature. Medicine (Baltimore) 2004; 83: 254–63.
23 Di Renzo M, Pasqui AL, Auteri A. Common variable
immunodefi ciency: a review. Clin Exp Med 2004; 3: 211–17.
24 Horn J, Thon V, Bartonkova D, et al. Anti-IgA antibodies in
common variable immunodefi ciency (CVID): diagnostic workup
and therapeutic strategy. Clin Immunol 2007; 122: 156–62.
25 Knight AK, Cunningham-Rundles C. Infl ammatory and
autoimmune complications of common variable immune
defi ciency. Autoimmun Rev 2006; 5: 156–59.
26 Scharenberg AM, Hannibal MC, Torgerson TR, Ochs HD,
Rawlings DJ. Common variable immunodefi ciency overview.
Gene Reviews 2006; http://www.ncbi.nlm.nih.gov/books/
bv.fcgi?rid=gene.chapter.cvid (accessed March 12, 2008).
27 Tanaka N, Kim JS, Bates CA, et al. Lung diseases in patients with
common variable immunodefi ciency: chest radiographic, and
computed tomographic fi ndings. J Comput Assist Tomogr 2006;
28 Kainulainen L, Varpula M, Liippo K, Svedstrom E,
Nikoskelainen J, Ruuskanen O. Pulmonary abnormalities
in patients with primary hypogammaglobulinemia.
J Allergy Clin Immunol 1999; 104: 1031–36.
29 Curtin JJ, Webster AD, Farrant J, Katz D. Bronchiectasis in
hypogammaglobulinaemia—a computed tomography assessment.
Clin Radiol 1991; 44: 82–84.
30 Obregon RG, Lynch DA, Kaske T, Newell JD Jr, Kirkpatrick CH.
Radiologic fi ndings of adult primary immunodefi ciency disorders.
Contribution of CT. Chest 1994; 106: 490–95.
31 Martínez García MA, de Rojas MD, Nauff al Manzur MD, et al.
Respiratory disorders in common variable immunodefi ciency.
Respir Med 2001; 95: 191–95.
32 Morimoto Y, Routes JM. Granulomatous disease in common
variable immunodefi ciency. Curr Allergy Asthma Rep 2005;
33 Bates CA, Ellison MC, Lynch DA, Cool CD, Brown KK, Routes JM.
Granulomatous-lymphocytic lung disease shortens survival in
common variable immunodefi ciency. J Allergy Clin Immunol 2004;
34 Mechanic LJ, Dikman S, Cunningham-Rundles C. Granulomatous
disease in common variable immunodefi ciency. Ann Intern Med
1997; 127: 613–17.
35 Wehr C, Kivioja T, Schmitt C, et al. The EUROclass trial: defi ning
subgroups in common variable immunodefi ciency. Blood 2007;
36 Fasano MB, Sullivan KE, Sarpong SB, et al. Sarcoidosis and
common variable immunodefi ciency: report of 8 cases and review
of the literature. Medicine 1996; 75: 251–61.
37 Wheat WH, Cool CD, Morimoto Y, et al. Possible role of human
herpesvirus 8 in the lymphoproliferative disorders in common
variable immunodefi ciency. J Exp Med 2005; 202: 479–84.
www.thelancet.com Vol 372 August 9, 2008
38 Cataldo F, Marino V, Ventura A, Bottaro G, Corazza GR.
Prevalence and clinical features of selective immunoglobulin
A defi ciency in coeliac disease: an Italian multicentre study.
Gut 1998; 42: 362–65.
39 Klemola T. Defi ciency of immunoglobulin A. Ann Clin Res 1987;
40 Kinlen LJ, Webster AD, Bird AG, et al. Prospective study of cancer
in patients with hypogammaglobulinaemia. Lancet 1985;
41 Mellemkjaer L, Hammarstrom L, Andersen V, et al. Cancer risk
among patients with IgA defi ciency or common variable
immunodefi ciency and their relatives: a combined Danish and
Swedish study. Clin Exp Immunol 2002; 130: 495–500.
42 Vorechovsky I, Scott D, Haeney MR, Webster DA. Chromosomal
radiosensitivity in common variable immune defi ciency.
Mutat Res 1993; 290: 255–64.
43 Jaff e EF, Lejtenyi MC, Noya FJD, Mazer BD. Secondary
hypogammaglobulinemia. Immunol Allergy Clin North Am 2001;
44 Hamilos DL, Young RM, Peter JB, Agopian MS, Ikle DN, Barka N.
Hypogammaglobulinemia in asthmatic patients. Ann Allergy 1992;
45 Lack G, Ochs HD, Gelfand EW. Humoral immunity in
steroid-dependent children with asthma and
hypogammaglobulinemia. J Pediatr 1996; 129: 898–903.
46 Conley ME, Howard V. Clinical fi ndings leading to the diagnosis
of X-linked agammaglobulinemia. J Pediatr 2002; 141: 566–71.
47 Eastwood D, Gilmour KC, Nistala K, et al. Prevalence of SAP gene
defects in male patients diagnosed with common variable
immunodefi ciency. Clin Exp Immunol 2004; 137: 584–88.
48 Weston SA, Prasad ML, Mullighan CG, Chapel H, Benson EM.
Assessment of male CVID patients for mutations in the Btk gene:
how many have been misdiagnosed? Clin Exp Immunol 2001;
49 Bossuyt X, Moens L, Van Hoeyveld E, et al. Coexistence of
(partial) immune defects and risk of recurrent respiratory
infections. Clin Chem 2007; 53: 124–30.
50 Alachkar H, Taubenheim N, Haeney MR, Durandy A,
Arkwright PD. Memory switched B cell percentage and not serum
immunoglobulin concentration is associated with clinical
complications in children and adults with specifi c antibody
defi ciency and common variable immunodefi ciency. Clin Immunol
2006; 120: 310–18.
51 Buckley RH. Primary immunodefi ciency or not? Making the
correct diagnosis. J Allergy Clin Immunol 2006; 117: 756–58.
52 Eisenstein EM, Jaff e JS, Strober W. Reduced interleukin-2 (IL-2)
production in common variable immunodefi ciency is due to
a primary abnormality of CD4+ T cell diff erentiation.
J Clin Immunol 1993; 13: 247–58.
53 Jaff e JS, Eisenstein E, Sneller MC, Strober W. T-cell abnormalities in
common variable immunodefi ciency. Pediatr Res 1993;
33 (1 suppl): S24–27.
54 Jaff e JS, Strober W, Sneller MC. Functional abnormalities of
CD8+ T cells defi ne a unique subset of patients with common
variable immunodefi ciency. Blood 1993; 82: 192–201.
55 Baumert E, Wolff -Vorbeck G, Schlesier M, Peter HH.
Immunophenotypical alterations in a subset of patients with
common variable immunodefi ciency (CVID). Clin Exp Immunol
1992; 90: 25–30.
56 Holm AM, Sivertsen EA, Tunheim SH, et al. Gene expression
analysis of peripheral T cells in a subgroup of common variable
immunodefi ciency shows predominance of CCR7(-)
eff ector-memory T cells. Clin Exp Immunol 2004; 138: 278–89.
57 Aspalter RM, Sewell WA, Dolman K, Farrant J, Webster AD.
Defi ciency in circulating natural killer (NK) cell subsets in
common variable immunodefi ciency and X-linked
agammaglobulinaemia. Clin Exp Immunol 2000; 121: 506–14.
58 Bayry J, Hermine O, Webster DA, Levy Y, Kaveri SV. Common
variable immunodefi ciency: the immune system in chaos.
Trends Mol Med 2005; 11: 370–76.
59 Brouet JC, Chedeville A, Fermand JP, Royer B. Study of the B cell
memory compartment in common variable immunodefi ciency.
Eur J Immunol 2000; 30: 2516–20.
60 Goldacker S, Warnatz K. Tackling the heterogeneity of CVID.
Curr Opin Allergy Clin Immunol 2005; 5: 504–09.
61 Moratto D, Gulino AV, Fontana S, et al. Combined decrease of
defi ned B and T cell subsets in a group of common variable
immunodefi ciency patients. Clin Immunol 2006; 121: 203–14.
62 Taubenheim N, von Hornung M, Durandy A, et al. Defi ned blocks
in terminal plasma cell diff erentiation of common variable
immunodefi ciency patients. J Immunol 2005; 175: 5498–503.
63 Cunningham-Rundles C. Clinical and immunologic analyses of
103 patients with common variable immunodefi ciency.
J Clin Immunol 1989; 9: 22–33.
64 Eibl MM, Wolf HM. Common variable immunodefi ciency: clinical
aspects and recent progress in identifying the immunological
defect(s). Folia Microbiol 1995; 40: 360–66.
65 Cunningham-Rundles C, Bodian C, Ochs HD, Martin S,
Reiter-Wong M, Zhuo Z. Long-term low-dose IL-2 enhances
immune function in common variable immunodefi ciency.
Clin Immunol 2001; 100: 181–90.
66 Punnonen J, Kainulainen L, Ruuskanen O, Nikoskelainen J,
Arvilommi H. IL-4 synergizes with IL-10 and anti-CD40 MoAbs to
induce B-cell diff erentiation in patients with common variable
immunodefi ciency. Scand J Immunol 1997; 45: 203–12.
67 Reinherz EL, Geha R, Wohl ME, Morimoto C, Rosen FS,
Schlossman SF. Immunodefi ciency associated with loss of T4+
inducer T-cell function. N Engl J Med 1981; 304: 811–16.
68 Boncristiano M, Majolini MB, D’Elios MM, et al. Defective
recruitment and activation of ZAP-70 in common variable
immunodefi ciency patients with T cell defects. Eur J Immunol
2000; 30: 2632–38.
69 Majolini MB, D’Elios MM, Boncristiano M, et al. Uncoupling of
T-cell antigen receptor and downstream protein tyrosine kinases
in common variable immunodefi ciency.
Clin Immunol Immunopathol 1997; 84: 98–102.
70 Farrington M, Grosmaire LS, Nonoyama S, et al. CD40 ligand
expression is defective in a subset of patients with common
variable immunodefi ciency. Proc Natl Acad Sci USA 1994;
71 North ME, Webster AD, Farrant J. Primary defect in CD8+
lymphocytes in the antibody defi ciency disease (common variable
immunodefi ciency): abnormalities in intracellular production of
interferon-gamma (IFN-gamma) in CD28+ (‘cytotoxic’) and CD28-
(‘suppressor’) CD8+ subsets. Clin Exp Immunol 1998; 111: 70–75.
72 Tangye SG, Liu YJ, Aversa G, Phillips JH, de Vries JE.
Identifi cation of functional human splenic memory B cells by
expression of CD148 and CD27. J Exp Med 1998; 188: 1691–703.
73 Piqueras B, Lavenu-Bombled C, Galicier L, et al. Common
variable immunodefi ciency patient classifi cation based on
impaired B cell memory diff erentiation correlates with clinical
aspects. J Clin Immunol 2003; 23: 385–400.
74 Warnatz K, Denz A, Drager R, et al. Severe defi ciency of switched
memory B cells (CD27(+)IgM(-)IgD(-)) in subgroups of patients
with common variable immunodefi ciency: a new approach to
classify a heterogeneous disease. Blood 2002; 99: 1544–51.
75 Durandy A, Peron S, Fischer A. Hyper-IgM syndromes.
Curr Opin Rheumatol 2006; 18: 369–76.
76 Lee WI, Torgerson TR, Schumacher MJ, Yel L, Zhu Q,
Ochs HD. Molecular analysis of a large cohort of patients with
the hyper immunoglobulin M (IgM) syndrome. Blood 2005;
77 Agematsu K, Nagumo H, Yang FC, et al. B cell subpopulations
separated by CD27 and crucial collaboration of CD27+ B cells and
helper T cells in immunoglobulin production. Eur J Immunol
1997; 27: 2073–79.
78 Fecteau JF, Cote G, Neron S. A new memory CD27-IgG+ B cell
population in peripheral blood expressing VH genes with low
frequency of somatic mutation. J Immunol 2006; 177: 3728–36.
79 Klein U, Kuppers R, Rajewsky K. Human IgM+IgD+ B cells, the
major B cell subset in the peripheral blood, express V kappa
genes with no or little somatic mutation throughout life.
Eur J Immunol 1993; 23: 3272–77.
80 Klein U, Kuppers R, Rajewsky K. Evidence for a large
compartment of IgM-expressing memory B cells in humans.
Blood 1997; 89: 1288–98.
www.thelancet.com Vol 372 August 9, 2008 501
81 Weller S, Braun MC, Tan BK, et al. Human blood IgM “memory”
B cells are circulating splenic marginal zone B cells harboring a
prediversifi ed immunoglobulin repertoire. Blood 2004;
82 Weller S, Faili A, Garcia C, et al. CD40-CD40L independent Ig
gene hypermutation suggests a second B cell diversifi cation
pathway in humans. Proc Natl Acad Sci USA 2001; 98: 1166–70.
83 Tangye SG, Good KL. Human IgM+CD27+ B cells: memory
B cells or “memory” B cells? J Immunol 2007; 179: 13–19.
84 Carsetti R, Rosado MM, Donnanno S, et al. The loss of IgM
memory B cells correlates with clinical disease in common variable
immunodefi ciency. J Allergy Clin Immunol 2005; 115: 412–17.
85 Ferry BL, Jones J, Bateman EA, et al. Measurement of peripheral
B cell subpopulations in common variable immunodefi ciency
(CVID) using a whole blood method. Clin Exp Immunol 2005;
86 Grimbacher B, Hutloff A, Schlesier M, et al. Homozygous loss of
ICOS is associated with adult-onset common variable
immunodefi ciency. Nat Immunol 2003; 4: 261–68.
87 Salzer U, Maul-Pavicic A, Cunningham-Rundles C, et al. ICOS
defi ciency in patients with common variable immunodefi ciency.
Clin Immunol 2004; 113: 234–40.
88 Castigli E, Wilson S, Garibyan L, et al. Reexamining the role of
TACI coding variants in common variable immunodefi ciency and
selective IgA defi ciency. Nat Genet 2007; 39: 430–31.
89 Castigli E, Wilson SA, Garibyan L, et al. TACI is mutant in
common variable immunodefi ciency and IgA defi ciency.
Nat Genet 2005; 37: 829–34.
90 Pan-Hammarstrom Q, Salzer U, Du L, et al. Reexamining the role
of TACI coding variants in common variable immunodefi ciency
and selective IgA defi ciency. Nat Genet 2007; 39: 429–30.
91 Salzer U, Chapel HM, Webster AD, et al. Mutations in
TNFRSF13B encoding TACI are associated with common
variable immunodefi ciency in humans. Nat Genet 2005;
92 Warnatz K, Salzer U, Gutenberger S. Finally found: human
BAFF-R defi ciency causes hypogammaglobulinemia.
Clin Immunol 2005; 115 (suppl 1): 820.
93 van Zelm MC, Reisli I, van der Burg M, et al. An
antibody-defi ciency syndrome due to mutations in the CD19 gene.
N Engl J Med 2006; 354: 1901–12.
94 Salzer U, Grimbacher B. TACItly changing tunes: farewell to a yin
and yang of BAFF receptor and TACI in humoral immunity?
New genetic defects in common variable immunodefi ciency.
Curr Opin Allergy Clin Immunol 2005; 5: 496–503.
95 Zhang L, Radigan L, Salzer U, et al. Transmembrane activator and
calcium-modulating cyclophilin ligand interactor mutations in
common variable immunodefi ciency: clinical and immunologic
outcomes in heterozygotes. J Allergy Clin Immunol 2007;
96 Warnatz K, Bossaller L, Salzer U, et al. Human ICOS defi ciency
abrogates the germinal center reaction and provides a monogenic
model for common variable immunodefi ciency. Blood 2006;
97 Broides A, Conley ME. The role of inducible co-stimulator (ICOS)
in immunodefi ciency. Clin Immunol 2004; 113: 221–23.
98 Greenwald RJ, Freeman GJ, Sharpe AH. The B7 family revisited.
Annu Rev Immunol 2005; 23: 515–48.
99 Mackay F, Ambrose C. The TNF family members BAFF and
APRIL: the growing complexity. Cytokine Growth Factor Rev 2003;
100 Mackay F, Schneider P, Rennert P, Browning J. BAFF and APRIL:
a tutorial on B cell survival. Annu Rev Immunol 2003; 21: 231–64.
101 Castigli E, Geha RS. Molecular basis of common variable
immunodefi ciency. J Allergy Clin Immunol 2006; 117: 740–46.
102 Carter RH, Fearon DT. CD19: lowering the threshold for
antigen receptor stimulation of B lymphocytes. Science 1992;
103 Chanock SJ, Manolio T, Boehnke M, et al. Replicating
genotype-phenotype associations. Nature 2007; 447: 655–60.
104 Cunningham-Rundles C, Siegal FP, Smithwick EM, et al. Effi cacy
of intravenous immunoglobulin in primary humoral
immunodefi ciency disease. Ann Intern Med 1984; 101: 435–39.
105 Nolte MT, Pirofsky B, Gerritz GA, Golding B. Intravenous
immunoglobulin therapy for antibody defi ciency.
Clin Exp Immunol 1979; 362: 237–43.
106 Roifman CM, Lederman HM, Lavi S, Stein LD, Levison H,
Gelfand EW. Benefi t of intravenous IgG replacement in
hypogammaglobulinemic patients with chronic sinopulmonary
disease. Am J Med 1985; 79: 171–74.
107 Busse PJ, Razvi S, Cunningham-Rundles C. Effi cacy of
intravenous immunoglobulin in the prevention of pneumonia in
patients with common variable immunodefi ciency.
J Allergy Clin Immunol 2002; 109: 1001–04.
108 Quartier P, Debre M, De Blic J, et al. Early and prolonged
intravenous immunoglobulin replacement therapy in childhood
agammaglobulinemia: a retrospective survey of 31 patients.
J Pediatr 1999; 134: 589–96.
109 Weiler CR. Immunoglobulin therapy: history, indications, and
routes of administration. Int J Dermatol 2004; 43: 163–66.
110 Roifman CM, Levison H, Gelfand EW. High-dose versus
low-dose intravenous immunoglobulin in
hypogammaglobulinaemia and chronic lung disease. Lancet
1987; 1: 1075–77.
111 Chapel HM, Spickett GP, Ericson D, Engl W, Eibl MM,
Bjorkander J. The comparison of the effi cacy and safety of
intravenous versus subcutaneous immunoglobulin replacement
therapy. J Clin Immunol 2000; 20: 94–100.
112 Gardulf A, Andersen V, Bjorkander J, et al. Subcutaneous
immunoglobulin replacement in patients with primary antibody
defi ciencies: safety and costs. Lancet 1995; 345: 365–69.
113 Burks AW, Sampson HA, Buckley RH. Anaphylactic reactions
after gamma globulin administration in patients with
hypogammaglobulinemia: detection of IgE antibodies to IgA.
N Engl J Med 1986; 314: 560–64.
114 Sati HI, Ahya R, Watson HG. Incidence and associations of acute
renal failure complicating high-dose intravenous
immunoglobulin therapy. Br J Haematol 2001; 113: 556–57.
115 Caress JB, Cartwright MS, Donofrio PD, Peacock JE Jr.
The clinical features of 16 cases of stroke associated with
administration of IVIg. Neurology 2003; 60: 1822–24.
116 Stamboulis E, Theodorou V, Kilidireas K, Apostolou T. Acute
myocardial infarction following intravenous immunoglobulin
therapy for chronic infl ammatory demyelinating polyneuropathy
in association with a monoclonal immunoglobulin G paraprotein.
Eur Neurol 2004; 51: 51.
117 Alliot C, Rapin JP, Besson M, Bedjaoui F, Messouak D.
Pulmonary embolism after intravenous immunoglobulin.
J R Soc Med 2001; 94: 187–88.
118 Go RS, Call TG. Deep venous thrombosis of the arm after
intravenous immunoglobulin infusion: case report and literature
review of intravenous immunoglobulin-related thrombotic
complications. Mayo Clin Proc 2000; 75: 83–85.
119 Kato E, Shindo S, Eto Y, et al. Administration of immune globulin
associated with aseptic meningitis. JAMA 1988; 259: 3269–71.
120 Brennan VM, Salome-Bentley NJ, Chapel HM. Prospective audit
of adverse reactions occurring in 459 primary antibody-defi cient
patients receiving intravenous immunoglobulin.
Clin Exp Immunol 2003; 133: 247–51.
121 de Albuquerque Campos R, Sato MN, da Silva Duarte AJ. IgG
anti-IgA subclasses in common variable immunodefi ciency and
association with severe adverse reactions to intravenous
immunoglobulin therapy. J Clin Immunol 2000; 20: 77–82.
122 Sandler SG, Mallory D, Malamut D, Eckrich R. IgA anaphylactic
transfusion reactions. Transfus Med Rev 1995; 9: 1–8.
123 Gardulf A, Nicolay U, Asensio O, et al. Rapid subcutaneous IgG
replacement therapy is eff ective and safe in children and adults
with primary immunodefi ciencies—a prospective, multi-national
study. J Clin Immunol 2006; 26: 177–85.
124 Lindberg K, Gustafson R, Samuelson A, Rynnel-Dagoo B.
Impact of IgG replacement therapy and antibiotic treatment on
the colonization of non-encapsulated Haemophilus infl uenzae in
the nasopharynx in patients with hypogammaglobulinaemia.
Scand J Infect Dis 2001; 33: 904–08.
125 Sewell WA, Buckland M, Jolles SR. Therapeutic strategies in
common variable immunodefi ciency. Drugs 2003; 63: 1359–71.
www.thelancet.com Vol 372 August 9, 2008
126 Nos P, Bastida G, Beltran B, Aguas M, Ponce J. Crohn’s disease
in common variable immunodefi ciency: treatment with
antitumor necrosis factor alpha. Am J Gastroenterol 2006;
127 Hatab AZ, Ballas ZK. Caseating granulomatous disease in
common variable immunodefi ciency treated with infl iximab.
J Allergy Clin Immunology 2005; 116: 1161–62.
128 Thatayatikom A, Thatayatikom S, White AJ. Infl iximab treatment
for severe granulomatous disease in common variable
immunodefi ciency: a case report and review of the literature.
Ann Allergy Asthma Immunol 2005; 95: 293–300.
129 Carbone J, Escudero A, Mayayo M, et al. Partial response to
anti-CD20 monoclonal antibody treatment of severe immune
thrombocytopenic purpura in a patient with common variable
immunodefi ciency. Ann N Y Acad Sci 2005; 1051: 666–71.
130 Wakim M, Shah A, Arndt PA, et al. Successful anti-CD20
monoclonal antibody treatment of severe autoimmune hemolytic
anemia due to warm reactive IgM autoantibody in a child with
common variable immunodefi ciency. Am J Hematol 2004;
131 Smith KJ, Skelton H. Common variable immunodefi ciency
treated with a recombinant human IgG, tumour necrosis
factor-alpha receptor fusion protein. Br J Dermatol 2001;
132 Atkinson WL, Pickering LK, Schwartz B, et al. General
recommendations on immunization: recommendations of the
Advisory Committee on Immunization Practices (ACIP) and the
American Academy of Family Physicians (AAFP).
MMWR Recomm Rep 2002; 51: 1–35.