Infectious Disease Unit
Department of Medicine, Center for Molecular Medicine
Karolinska Institutet, Solna, Sweden
LABORATORY FINDINGS IN
Anders Lundqvist, MD
Gårdsvägen 4 169 70 Solna
All previously published papers were reproduced with permission from the publisher.
Published and printed by Reproprint AB
169 70 Solna
© Anders Lundqvist, 2006
Published and printed by
All previously published papers were reproduced with permission from the publisher.
Published and printed by Reproprint AB
169 70 Solna
© Anders Lundqvist, 2006
PREFACE – PARVOVIRUS AND FOOTBALL
During the last few years, a lot of my time has been shared between football and
parvovirus B19 (B19), and it has been evident that these interests have several
similarities. First, the forms of these objects, icosahedral and global, respectively, are
rather conserved over time and around the world. Secondly, events surrounding these
objects show considerable diversity. If you are interested in football or coaching a
football team, you may have great knowledge about individual players, such as body
weight, shoe size, running speed and scoring record, but the most important question is:
How does he or she play football? In that dimension, every match is a new,
unpredictable game. What makes team member Zlatan score more beautifully than ever
and why does he miss? These outcomes are matters of coincidence, situation,
environment and previous experiences in combination with the same factors in 21 other
players in the football field.
In a comparable way, the immune system plays football with microorganisms. The
clinical picture of infectious diseases is a synthesis of features belonging to the microbe
and to the immune system. The latter has basic characteristics in common among most
humans, but the individual patterns of reaction against different enemies and in
different situations vary. Consequently, even if a specific infection causes relatively
predictable symptoms in most individuals, we have to expect outliers whose symptoms
are difficult to recognize. For widespread infections like that from B19, which infects
the majority of humans, the total number of individuals with a deviant course may be
high, even if the percentage is low.
Our studies of B19 have focused on clinical and immunological aspects of persistent
infection, with persistence per se of central interest. Hypothetically, an immune system
that cannot eliminate an infecting virus is still annoyed by the infection and responds
with reactions and symptoms that vary according to the host’s individual character. The
ball is on the field and one may guess what will happen but cannot know with certainty.
I want to thank my wife Inger, who is supportive in both football and B19 studies, and
all the football players in my family: Malin, Petter, Karin and Elin. Their inspiration for
life, science and football has been beyond description, just like immunology and
Parvovirus B19 (B19) is the etiological agent of the common childhood disease,
erythema infectiosum, also named fifth disease or slapped cheek disease. About
50% of humans are infected during childhood, but the virus is also transmitted to
adults and about 80 % of the elderly are seropositive. Apart from asymptomatic
infection, which is common, the clinical presentations in erythema infectiosum are
usually slight catarrhal symptoms and fever followed by rash, typically intense on
cheeks and later distributed on the extensor parts of legs and arms. For infected
adults, the rash is less common, whereas the frequency of arthropathy is about 50%
and may last for months or years. The main replicative sites of the virus are
erythropoetic cells, and anaemia is a well-known manifestation of the infection,
especially in immunocompromised patients. Pregnant women, when infected, are at
risk of spontaneous abortion, hydrops fetalis and, later in pregnancy, intrauterine
fetal death. B19 has also been associated with rheumatologic diseases, hepatitis,
myocarditis and neurological manifestations.
In 1993, we used PCR-technique and found, for the first time, B19 DNA in the
bone marrow (BM) of a patient with chronic fatigue, anaemia and leukopenia.
Since then, we have followed her and other patients with chronic symptoms of
fatigue, arthralgia and fever and found that some of them had persistent B19 DNA
in BM for years after primary infection. Lifelong latency, as a natural course of the
disease could account for this persistence, but the prevalence of B19 DNA in BM in
the general population was not previously known. To get a hint of the frequency,
we tested the BM of 100 patients with haematological disorders from whom BM
samples were available and found B19 DNA in four of them. Therefore, we
concluded that B19 DNA in BM is not a general finding in seropositive individuals.
Arthropathy is a common manifestation of B19 infection in adults and sometimes
even meets the official criteria for rheumatoid arthritis. Consequently, the
frequency of B19 DNA in BM of rheumatic patients is of interest, which motivated
us to study a group of 50 patients with various rheumatological diseases. We found
B19 DNA in 13 of 50 patients (26%), a significantly higher level than in our
previous study on patients with haematological disorders.
As is well known, immunocompromised patients are susceptible to B19 infection,
but our hypothesis is that individuals who are otherwise immunocompetent may
have a selective immune deficiency with respect to B19 causing the virus to persist
and evoke symptoms of chronic immunological stimulation. To confirm this
hypothesis, we have studied patients with persistent B19 DNA in BM with regard
to general immunological parameters such as HLA-type, levels of lymphocyte
subpopulations and cytokine profiles. In some aspects we have compared with
expected values in the general population, in others with reactions in acutely or
remotely B19-infected individuals. We also investigated the specific CD8 T-cell
response to B19 antigens with an enzyme-linked immunospot assay for detecting
interferon gamma (IFNγ) production. Although the patients with persistent B19
DNA in BM had no general immunological aberrations, their B19 specific cellular
immune responses diverged from those of controls supporting our hypothesis of a
selective immune deficiency.
Key words: Human parvovirus B19, persistent infection, immune defence.
LIST OF PUBLICATIONS
Lundqvist A, Tolfvenstam T, Bostic J, Söderlund M, Broliden K.
Clinical and laboratory findings in immunocompetent patients with persistent
parvovirus B19 DNA in bone marrow.
Scand J Infect Dis, 1999. 31 (1): p. 11-6.
Lundqvist A, Tolfvenstam T, Brytting M, Stolt CM, Hedman K, Broliden K.
Prevalence of parvovirus B19 DNA in bone marrow of patients with
Scand J Infect Dis, 1999. 31 (2): p. 119-22.
Lundqvist A, Isa A, Tolfvenstam T, Kvist G, Broliden K.
High frequency of parvovirus B19 DNA in bone marrow samples from
J Clin Virol, 2005. 33 (1): p. 71-4.
Isa A, Lundqvist A, Lindblom A Tolfvenstam T, Broliden K.
Cytokine responses in acute and persistent human Parvovirus B19 infection
Clin Exp Immunol, 2006. In press
Isa A, Norbeck O, Hirbod T, Lundqvist A, Kasprowicz V, Bowness P,
Klenerman P, Broliden K, Tolfvenstam T.
Aberrant cellular immune responses in persistently human parvovirus B19
J Med Virol. 2006 Jan; 78 (1): p.129-33.
1.1 Background and taxonomy................................................................1
1.2 The virus............................................................................................2
1.2.5 Culture and animal model.....................................................3
1.3 Epidemiology and transmission........................................................3
1.4 Infection of human cells....................................................................3
1.5 Viral persistence................................................................................4
1.6 Immune responses.............................................................................4
1.7 Clinical manifestations......................................................................6
1.7.1 Asymptomatic infection........................................................6
1.7.2 Erythema infectiosum............................................................6
1.7.3 Arthropathy ...........................................................................8
1.7.4 Other autoimmune diseases...................................................8
1.7.5 Chronic fatigue syndrome and fibromyalgia.........................9
1.7.6 Haematological manifestations.............................................9
1.7.7 Complications associated with pregnancy ..........................10
1.7.8 Other manifestations............................................................11
1.8 Treatment and prophylaxis..............................................................11
1.8.1 Symptomatic treatment........................................................11
1.8.2 Antiviral drugs.....................................................................11
1.8.3 Blood transfusion ................................................................11
1.8.4 Intravenous immunoglobulin (IVIG)..................................12
1.8.5 Cessation of chemotherapy .................................................12
AIMS OF THE STUDY...........................................................................13
MATERIALS AND METHODS .............................................................14
3.1 Study subjects..................................................................................14
3.1.1 Paper I..................................................................................14
3.1.2 Paper II................................................................................14
3.1.3 Paper III...............................................................................16
3.1.4 Paper IV...............................................................................17
3.1.5 Paper V................................................................................18
3.2 Collection of samples......................................................................19
3.2.1 Bone marrow.......................................................................19
3.3 Detection of B19 DNA....................................................................19
3.4 Serological assays...........................................................................20
3.4.1 Specific B19 IgG in serum..................................................20
3.4.2 Specific B19 IgM in serum.................................................20
3.4.4 Epitope type specificity ......................................................21
3.4.5 Neutralizing antibodies.......................................................21
3.5 Immunophenotyping of lymphocytes.............................................22
3.7 Interferon gamma (IFNγ) ELISpot................................................22
3.9 Quantification of Cytokines............................................................23
RESULTS AND DISCUSSION..............................................................24
4.1 Paper I: Clinical and laboratory findings in immunocompetent
patients with persistent parvovirus B19 DNA in bone marrow......24
4.2 Paper II: Prevalence of parvovirus B19 DNA in bone marrow of
patients with haematological disorders...........................................30
4.3 Paper III: High frequency of parvovirus B19 DNA in bone
marrow samples from rheumatic patients.......................................32
4.4Paper IV: Cytokine responses in acute and persistent human
parvovirus B19 infection................................................................35
4.5Paper V: Aberrant cellular immune responses in humans infected
persistently with parvovirus B19....................................................40
Conclusions and future possibilities.........................................................42
Parvovirus B19 – inte bara femte sjukan..................................................43
LIST OF ABBREVIATIONS
Cluster of differentiation
Chronic fatigue syndrome
Enzyme immuno assay
Enzyme linked immunospot assay
Epitope type specificity
Fluorescence activated cell sorter
Hepatitis B virus
Hepatitis C virus
Human immunodeficiency virus
Human leukocyte antigen
Intrauterine fetal death
Major histocompatibility complex
Peripheral blood mononuclear cells
Polymerase chain reaction
Systemic lupus erythematosus
Spot forming cells
Virus capsid protein
Transient aplastic crisis
T helper cell
T cytotoxic cell
1.1 BACKGROUND AND TAXONOMY
Parvum is Latin for small, and viruses in the family of Parvoviridae are small non-
enveloped particles with a diameter of 18 to 26 nm and single stranded DNA. These
pathogens are very common in insects and animals. The virion is stable and resistent to
heat, detergents and radiation. The Parvoviridae family is divided into Densovirinae,
which infect insects, and Parvovirinae, which infect vertebrates (Heegaard et al. 2002).
Three genera are defined in the Parvovirinae group:
Infects humans and other vertebrates and are dependent on other viruses, e.g.,
adenovirus or herpesvirus to infect and replicate. In humans they are not
associated with any known disease.
Is a common pathogen among animals and, e.g., canine parvovirus causes
gastrointestinal infection in dogs, a disease associated with high mortality.
Humans are not susceptible to these viruses.
Characteristically they have a tropism for and replicate in erythropoetic cells,
thereby causing anaemia, which is a well-known clinical feature of the
infection. At least three different strains infect humans (Servant et al. 2002).
1. Erythrovirus Genotype 1 Parvovirus B19
2. Erythrovirus Genotype 2 Parvovirus A6/K71
3. Erythrovirus Genotype 3 Parvovirus V9
Among these, B19 is the only one known to cause disease. Although B19 was
identified 1974, it was not associated with disease until 1981 when the
relationship to aplastic crisis in patients with sickle cell anaemia was discovered
(Pattison et al. 1981). Genotypes 2 and 3 were subsequently identified (Nguyen
et al. 1999; Hokynar et al. 2000; Nguyen et al. 2002) and are proposed as the
cause of human disease in case reports, but their importance has yet not been
1.2 THE VIRUS
Parvovirus B19 was first discovered in 1974 in London, when Yvonne Cossart and
colleagues were testing assays for hepatitis B in samples of blood donors (Cossart et al.
1975). Specimen 19 in panel B revealed an unexpected antigen in a counter-
immunoelectrophoresis, and analysis with electron microscopy as well as further
molecular biology studies classified the virus in the Parvoviridae family.
The B19 virion is composed of two distinctive proteins that build an icosahedral
capsid containing 60 capsomeres and have a molecular weight of 5.6x 106and a
diameter of about 23 nm (Corcoran et al. 2004).
Parvovirus B19 DNA is single stranded and consists of about 5600 nucleotides among
which 4830 encode proteins and the rest are palindromic terminal repeats at each end.
The nucleotide sequence variability of B19 is known to be low, i.e., no higher than 1%
when one compares different isolates.
Two proteins provide building blocks in the B19 capsid. VP1 is the larger of these
proteins (84 kDa) but constitutes only about 4% of the capsid structure. VP2 is,
consequently, the major structural protein, forms the remaining ~96% of this structure
and has a molecular weight of 58 kDa. VP1 and VP2 are partially identical except for a
227 amino acid sequence, named the unique region (VP1ur), which ends the N-terminal
part of VP1 (Brown et al. 1994).
The major non-structural protein (NS1), which is hidden inside the capsid, is believed
to function as a regulator of transcription and translation and also, probably, is
important for apoptosis of infected cells (Moffatt et al. 1998). The molecular weight is
77 kDa, and the corresponding genome is well conserved (Erdman et al. 1996).
Another two NS proteins whose molecular weights are 7.5 and 11 kDa have been
described; although their functions are currently unproven, they may be involved in a
signalling pathway (Shade et al. 1986; Deiss et al. 1990).
1.2.5 Culture and animal model
Parvovirus B19 is difficult to culture, and trials to propagate this virus in several
erythropoetic cell lines have failed except for short periods of viability with limited
expression of viral antigens. Since B19 does not infect species other than humans, the
only similar animal model for experimentation is macaque monkeys infected with
simian parvovirus, which resembles B19. Immunocompromised monkeys infected with
the simian virus develop persistent anaemia, and fetal infection with this virus is
associated with hydrops fetalis analogous to that caused by B19 infection in humans
(Brown et al. 1997).
1.3 EPIDEMIOLOGY AND TRANSMISSION
During the early viremic phase of infection, B19 is found in respiratory secretions
indicating a respiratory route of transmission, but there is no known replication in the
mucosal cells. B19 infection is common in that 50% of the population is infected
during childhood (Chorba et al. 1986), and more than 80% of elderly humans are
seropositive (Heegaard et al. 2002). The extent of its contagiousness is rather high; that
is, about 50% of non-immune contacts within a household have been shown to
seroconvert (Chorba et al. 1986). Additionally, the attack rate among personnel
working in child day care centres or schools is described as 20-30% during outbreaks
(Gillespie et al. 1990). B19 infection is distributed worldwide and low seroprevalence
has been found only in occasional isolated tribes in Africa and South America.
However, viremia is rare considering that the frequency of B19 DNA in sera from
blood donors varies from 1:167 to 1:35000 in different studies (Yoto et al. 1995;
Tsujimura et al. 1995). When large pools of plasma are collected to produce
immunoglobulin or factor VIII concentrates for haemophiliacs, the risk is greater,
because B19 is so resistant to heat treatment and detergents that eliminating it from
such products is difficult (Wu et al. 2005). Vertical transmission has been registered in
about one-third of pregnant women with primary B19 infection, and these patients
endure a potential risk of spontaneous abortion and intrauterine fetal death. The annual
seroconversion rate among women of childbearing age is estimated to about 1.5%
(Koch et al. 1989) but may be higher during an epidemic situation (Valeur-Jensen et al.
1.4 INFECTION OF HUMAN CELLS
Characteristic for B19 is the tropism for erythropoetic cells and the binding to P-antigen
that serves as a receptor mediating B19’s internalization into cells and makes
replication possible. Individuals lacking P-antigen (1/200 000) are naturally resistent to
B19 infection. P-antigen is present on the surfaces of haematopoetic precursor cells but
also in several other cell lines, such as endothelial cells, fetal myocytes and placental
trophoblasts (Brown et al. 1993). Yet no replication of the complete virus has been
shown in these latter cell types. However, erythropoetic cells also express a co-receptor,
identified as an alfa-5-betaintegrin, which is involved in viral replication (Weigel-
Kelley et al. 2003). Another co-receptor, Ku80, was recently described and is expressed
in erythropoetic cells and also in lymphocytes (Munakata et al. 2005).
1.5 VIRAL PERSISTENCE
Two well-known manifestations of B19 infection are its potentially chronic course in
immunocompromised individuals and the persistent arthropathy it may cause in adults
with primary infection. Many other diseases with a chronic course have been proposed
as related to B19 based on findings of B19 DNA with PCR technique in various tissues.
Further studies have shown that B19 DNA is often present in, e.g., the synovia and skin
of healthy individuals. Consequently, studying the etiological relationship between B19
and disease is a difficult task, and the clinical relevance of finding B19 DNA must be
established organ-by-organ. BM is of special interest, since it is the known site of viral
replication, and microscopy typically shows erythropoesis with giant pronormoblasts,
which may also indicate B19 infection (Brown et al. 1996; Heegaard et al. 2002). The
frequency of B19 DNA in BM of healthy individuals or autopsied material has varied
from 2.1% (Heegaard et al. 2002) to 17% (Eis-Hubinger et al. 2001).
1.6 IMMUNE RESPONSES
Viral infections, in general, induce activation of both humoral and cellular immune
defence. The humoral response is initiated by production of IgM antibodies followed
by IgG and IgA. As shown, antibodies per se may provide immunity against viral
infections, e.g., prophylactic treatment with commercial immunoglobulin to individuals
exposed to hepatitis A virus (Mosley et al. 1968). However, the neutralizing effect of
antibodies is, in principle, extracellular, and viral infections are by nature intracellular
and need the cellular production system for replication. On the other hand, infected
cells have viral antigens on their surface and, in cooperation with the complement
system, antibody binding to the cell may cause cytolysis and neutralization of viruses
that are set free (Lachmann et al. 1997). The cellular immune response, mediated by
cytotoxic T-cells (CD8), is important in the defence against intracellular pathogens
such as virus and fungi by killing cells that present microbial antigens on their surface
(Zinkernagel 1996). In recent years, many new techniques to study cellular immune
responses have evolved, e.g., intracellular cytokine staining and phenotyping using new
markers for maturation and proliferation detected by FACS. The ELISpot-technique
has made it possible to elucidate aspects of the T-cell response in various infections and
also the variability of reaction in distinct infections in different individuals (Larsson et
The multiple mechanisms of immune reactions are of special interest in our
understanding of persistent infections. Patients who spontaneously recover from
hepatitis B infection usually produce high levels of specific CD8 cells, yet others who
develop chronic infection have a less intense reaction (Chisari et al. 1995). Individual
immunological characteristics including HLA-type of the host determine which viral
epitopes become exposed and, thereby, modulate the immune reaction (Penna et al.
1991). Theoretically, the chronic course of an infectious disease may be related to
selectively impaired immune responses to specific pathogens rather than a generalized
In B19 infection the humoral response typically is initiated by the appearance of
specific IgM antibodies ten days after infection and their disappearance two to four
months later, although the IgM-reaction is sometimes prolonged (Anderson et al. 1985;
Musiani et al. 1995). Viremia peaks at about the same time as IgM-antibodies reach a
detectable level but declines when IgG-antibodies become evident some days later.
Knowledge of the epitope-specificity of the humoral response is important for the
development of diagnostic assays. IgM and IgG antibodies are directed against both
VP1 and VP2, but different types of antigens vary in certain properties. Reactivity to
linear epitopes of VP1 is lost, whereas antibodies against conformational antigens
persist (Kerr et al. 1999). NS1 is an important protein of the virus, and NS1-specific
antibodies have been described in several but conflicting reports. In some studies, a
relationship between NS1-antibodies and persistent infection has been proposed
(Modrow et al. 2002; von Poblotzki et al. 1995; Kerr et al. 2000), a theory that has not
been confirmed by others (Jones et al. 1999). Also IgA-antibodies are produced and
may be important in the mucosal immune reaction (Erdman et al. 1991). IgE-antibodies
have been demonstrated, but their clinical relevance is still under evaluation (Bluth et
The cellular immune defence against B19 has been difficult to investigate and is not as
well characterized as the humoral response. The cellular immune response is probably
important though, since infections with a chronic course have been noted in the
presence of circulating and neutralizing antibodies (Cassinotti et al. 1997; Dobec et al.
2006). Tolfvenstam et al. used peptide pools representing the B19 NS1 protein to
stimulate CD8 cells and were able to demonstrate epitopes associated with cellular
responses (Tolfvenstam et al. 2001). A strong and long-lasting CD8 response against
NS1 has also been demonstrated (Norbeck et al. 2005).
Some studies of cytokine responses to acute B19 infection have shown prolonged up-
regulation of serum IFN? and TNF? during symptomatic B19 virus infection (Kerr et
al. 2001). Furthermore, patients with acute B19-associated arthritis have been
demonstrated with lower levels of IL6, TNF?, and GM-CSF than patients without
arthritis, and B19 related rash has been associated with decreased levels of TGFß1
(Kerr et al. 2004).
1.7 CLINICAL MANIFESTATIONS
My daughter Malin presenting a
typical facial rash of fifth disease. She
also had exanthema on arms and legs
and later on her two sisters showed the
same clinical picture while their
mother during a period of two weeks
suffered from arthralgia.
1.7.1 Asymptomatic infection
Although most of us have antibodies against B19, proving prior exposure to this
infection, rarely is its diagnosis recorded, which reflects the commonly mild clinical
picture of the disease. In an epidemiological study done in England during 1985, of 54
adults who were serologically positive for recent B19 infection, 14 (25%) were
asymptomatic (Woolf et al. 1989). In another study, 32% were without symptoms;
however, fewer B19-infected blacks are diagnosed, possibly because the rash is
difficult to see on dark skin (Chorba et al. 1986). The frequency may be even higher as
reported in an outbreak, including medical students, in which 68% were asymtomatic
(Noyola et al. 2004).
1.7.2 Erythema infectiosum
The clinical picture of erythema infectiosum was first described in 1799 by the famous
dermatologist Robert Willan and later by Ager (Ager et al. 1966). The manifestations,
however, were not attributed to B19 until 1983, when 31 individuals in a London
outbreak were found to have B19 IgM antibodies (Anderson et al. 1984). This
relationship was also proven experimentally by inducing the disease with inoculation of
the virus intranasally in healthy individuals (Anderson et al. 1985). Later, B19 was
established as the only cause of this disease, which is also called slapped cheek disease,
Stickers disease or, more commonly, fifth disease. Fifth disease is a designation used
after a classification of common childhood
diseases introduced by Cheinisse in 1905. The
typical course is initiated with non-specific
symptoms about one week after infection,
including fever, often mild, malaise, coryza
and myalgia. These prodromal symptoms
usually vanish after a few days but are
replaced with a rash about 15 to 17 days after
(Figure 2). The facial erythema usually starts
on cheeks and includes circumoral pallor, the
slapped cheek appearance (Figure 1). Some
days later, the rash appears on the trunk and
typically the extensor sides of extremities,
often in a lacy or reticular pattern that may
recur for several weeks and vary in intensity
depending on environmental factors such as
heat or stress. The clinical appearance may
vary greatly in different individuals, and the
classical eruption is much more common
among children than in adults (Woolf et al.
Virologic, immunologic and clinical course following B19 infection (modified from
Heegaard et al. 2002).
Of children with B19 infection, about 10% develop arthralgia, but the frequency is
higher in adults, reported as 30% in men and 59 % in women (Torok 1992).
Arthropathy is the most common manifestation of B19 infection in grown-ups. In one
study, five women of 26 patients (six men, 20 women) with B19 arthropathy had
persisting symptoms for two months or more (Woolf et al. 1989). Some patients with
B19 arthropathy also meet criteria for rheumatoid arthritis (RA) according to the
American College of Rheumatology and sometimes test positive for rheumatoid factor
or other autoantibodies (Naides et al. 1988; Kerr et al. 1996). White et al. investigated
153 patients with early synovitis and found B19-specific IgM in 17 of them whereas
other infections were diagnosed in five (White et al. 1985). Other investigators
examined 90 patients with arthritis of unknown origin for the presence of B19 DNA in
synovial fluid and synovial tissue (Cassinotti et al. 1998). The results yielded B19 DNA
in one of 78 synovial fluid samples and in 15 of 90 synovial tissue samples. The 15
patients with B19-positive synovial tissue were further investigated, and nine (60%) of
them had BM aspirates that turned out to be B19 DNA positive. Söderlund et al., on the
other hand, reported B19 DNA in synovia from eight of 29 children with arthritis
(28%) but also in 13 of 27 individuals (48%) in the control group including young
healthy adults with joint trauma. Moreover, when results from these patient and control
groups were summarized, 18 of 20 seropositive individuals also had B19 DNA positive
synovia (Söderlund et al. 1997; Vuorinen et al. 2002).
The most frequently affected joints in patients with B19 arthropathy are
metacarpophalangeal joints, knees, wrists and ankles (Woolf et al. 1989). Therefore,
this virus has been postulated as a possible cause of RA (Takahashi et al. 1998),
although other reports disagree (Kerr et al. 1995; Nikkari et al. 1995; Söderlund et al.
1997). B19 arthropathy is generally not known to include erosions or destruction of
joints (Naides et al. 1990) even though such conditions have been described (Tyndall et
al. 1994). Additionally, an increased seroprevalence for B19 in patients with RA has
been reported (Cohen et al. 1986) but not confirmed (Nikkari et al. 1994). Therefore,
the role of this virus in rheumatic diseases, if any, remains to be clarified. An
interesting point, however, is that HLA DR4, which bears a relationship to RA, has also
been associated with B19 arthropathy (Gendi et al. 1996).
1.7.4 Other autoimmune diseases
B19 has been suggestively implicated in several chronic autoimmune diseases such as
systemic lupus erythematosus (SLE), Wegener`s granulomatosis, Kawasaki’s
syndrome, systemic sclerosis and others (Moore et al. 1999; Nikkari et al. 1994; Nigro
et al. 1994; Pugliese et al. 2006; Ferri et al. 2005). Although the clinical symptoms of
B19 in patients with these diseases may be similar, no statistical proof of any general
etiologic connection has emerged.
1.7.5 Chronic fatigue syndrome and fibromyalgia
The diagnosis of chronic fatigue syndrome (CFS) is based on internationally accepted
criteria, first defined in 1988 and revised in 1994 (Fukuda et al. 1994). That description
depicts a disease with debilitating fatigue of more than six months duration associated
with at least four of the following symptoms: impaired memory or concentration,
tender lymph nodes, myalgia, arthralgia, headache, unrefreshing sleep and
postexertional malaise. The etiology is unknown but among the microbial causes
proposed are herpesviruses (Kawai et al. 1992) and enteroviruses (Chia 2005). Several
case reports have associated CFS with parvovirus B19 (Jacobson et al. 1997; Kerr et al.
2002), and treatment with immunoglobulin has been successful in some cases (Kerr et
al. 2003). On the other hand, a study of seven patients with haematological and
rheumatological symptoms who also fulfilled the criteria for CFS could not
demonstrate any sign of ongoing B19 infection (Ilaria et al. 1995). However, CFS is
very complex and probably has a multifactorial etiology including endocrine,
psychological, neurological and immunological aspects (Maquet et al. 2006). CFS is
connected with various disturbances in the immune system according to analyses of
cytokines and cell lines of various kinds. The IFN? level may be elevated as well as the
IL6 response to stimulation of PBMC. NK-cell activity may be decreased, and
cytotoxic T-cells activated (Barker et al. 1994). B19-related CFS has been associated
with elevated levels of IFN? and TNF? (Kerr et al. 2003).
The diagnostic criteria of fibromyalgia are based on chronic, general diffuse pain and
tenderness in defined points, but a subset of these patients also has symptoms
mimicking CFS and sometimes resembling B19 infection (Wolfe et al. 1990). Case
reports of B19 infection in patients developing fibromyalgia have been published
(Leventhal et al. 1991), but a study of 15 patients with fibromyalgia showed no
association with B19 infection (Berg et al. 1993).
1.7.6 Haematological manifestations
Historically, the first clinical manifestation of B19 observed was transient aplastic crisis
(TAC) in children with haemolytic anaemia. Six hundred children admitted to a
London hospital were examined for B19, and all six patients with laboratory signs of
recent infection turned out to be Jamaican immigrants with sickle cell disease on
presentation with aplastic crisis (Pattison et al. 1981). Later, B19 was identified as the
major cause of aplastic crisis, which is a well-known complication of sickle cell
anaemia, occurring just once during a patient’s life (Serjeant et al. 2001). The
pathogenic process stems from the tropism of B19 for erythropoetic cells, where the
virus replicates and causes apoptosis (Hsu et al. 2004). Usually TAC is a self-limiting
disease, but these patients may be severely ill with symptoms that include acute chest
syndrome, splenic sequestration and BM necrosis, occasionally leading to death
(Smith-Whitley et al. 2004; Lowenthal et al. 1996; Eichhorn et al. 1999). Blood
transfusion is commonly needed, but these patients usually recover within ten days.
B19 may also be the origin of aplastic crisis in other disorders and may be the initial
presentation of a previously undiagnosed chronic haemolytic disease (Eriksson et al.
For patients whose disease involve a compromised immune defence, including HIV or
disorders related to haematological malignancies or chemotherapy, B19 infection may
be associated with chronic BM failure (Abkowitz et al. 1997; Broliden et al. 1998). The
most common manifestation in these cases is pure red cell aplasia (PRCA). Tests of
peripheral blood demonstrate anaemia and reticulocytopenia, and in the BM, reduced
erythropoesis with giant pronormoblasts is typical (Kurtzman et al. 1988). The patient
may recover spontaneously but the anaemia may persist, sometimes for years
(Kurtzman et al. 1989). Treatment is possible with intravenous immunoglobulin,
(IVIG), usually given in doses of 0,4 g /kg for five consecutive days (Mouthon et al.
2005). Immunocompromised patients infected with B19 may present with leukopenia
or thrombocytopenia and occasionally pancytopenia (Smith et al. 1995). In the general
population, B19 infection usually decreases hemoglobin level but not to the point of
clinical importance and without overt anaemia. Occasionally, though, previously
healthy individuals with no known immune deficiency may develop PRCA as well as
agranulocytosis and thrombocytopenia in association with B19 infection.
Apart from bone marrow insufficiency, thrombocytopenia may have an immunological
source, so-called idiopathic thrombocytopenic purpura (ITP). This disorder has been
associated with B19 infection in children (Heegaard et al. 1999; Murray et al. 1994) but
not in adults (van Elsacker-Niele et al. 1996).
1.7.7 Complications associated with pregnancy
About 30 to 40% of pregnant women are seronegative for B19 and susceptible to
infection. The annual seroconversion rate is about 1.5% in women of childbearing age
but may be higher in epidemic situations (Valeur-Jensen et al. 1999; Jensen et al.
2000). Infected pregnant women may have exanthema or arthralgia, but asymptomatic
B19 infection is so common that the diagnosis is often overlooked. The risk of an
adverse outcome of pregnancy after primary maternal infection with this virus has been
estimated at 5 to 10% (Enders et al. 2004; Yaegashi et al. 1998). The fetus is most
vulnerable in the second trimester, a period when the erythropoetic activity increases
rapidly (Yaegashi et al. 1998); however, fetal loss correlated with B19 infection has
occurred throughout pregnancy and been considered responsible for intrauterine fetal
death in the third trimester (Tolfvenstam et al. 2001). The best-documented disorder
attributed to B19 infection during pregnancy is non-immune hydrops fetalis, with a
pathogenetic cause of anaemia in association with heart failure (Yaegashi 2000).
Treatment is possible with intrauterine blood transfusion (Enders et al. 2004) or
immunoglobulin (Selbing et al. 1995).
1.7.8 Other manifestations
B19 has, in case reports, been associated with a large number of diseases including
myocarditis (Lamparter et al. 2003), neurological disorders (Aguilar-Bernier et al.
2006), hepatitis (Sokal et al. 1998) and glomerulonephritis (Mori et al. 2002). This
association is sometimes probable when a typical clinical picture of B19 infection is
followed by complications in specific organs, by the presence of B19-specific IgM
antibodies and by B19-specific IgG seroconversion. However, validating the diagnosis
by using PCR to locate B19 DNA in different tissues is more difficult, since B19 DNA
can also be found in healthy individuals (Söderlund et al. 1997; Vuorinen 2002).
1.8 TREATMENT AND PROPHYLAXIS
1.8.1 Symptomatic treatment
Non-steroidal anti-inflammatory drugs and paracetamol may be of use in cases of
arthralgia (Golstein et al. 1996).
1.8.2 Antiviral drugs
The non-structural protein, NS1, is involved in viral replication, but viral proliferation
is more likely catalyzed by host-related DNA-polymerase (Astell et al. 1997). Unlike
HIV treatment with nucleoside analogues, no such antiviral treatment is available for
B19. However, HAART treatment for HIV infection may be enough to resolve co-
existent chronic B19 infection in an indirect way, since HAART increases the number
and functions of immune cells (Mylonakis et al. 1999).
1.8.3 Blood transfusion
Patients with sickle cell anaemia and (TAC) often need blood transfusions, which yield
a favourable prognosis and commonly resolve the anaemia within ten days. When
hydrops fetalis is diagnosed and associated with B19 infection, intrauterine blood
transfusion is a possible treatment, as described in several reports, and should be
considered (Enders et al. 2004).
1.8.4 Intravenous immunoglobulin (IVIG)
In immunocompromised patients with chronic B19-related anaemia, IVIG is a well-
established treatment and usually given in doses of 0.4 g/ kg consecutively for five
days, as stated above. Sometimes this therapy is curative, but relapses may occur, in
which case repeat doses of IVIG are needed (Broliden 2001). However, one must
consider that IVIG is a possible route of B19 transmission, since the virus is stable and
difficult to eliminate during the procedure of IVIG production (Erdman et al. 1997;
Hayakawa et al. 2002). Nevertheless, for patients with chronic B19 infection but not
obviously impaired immune defence, IVIG has been found successful for treating
chronic fatigue syndrome, arthritis and intrauterine infection (Kerr et al. 2003;
Lehmann et al. 2004; Rugolotto et al. 1999; Selbing et al. 1995).
1.8.5 Cessation of chemotherapy
For patients with B19 infection for whom cessation of immunosuppression is possible
or planned, the withdrawal or switch of chemotherapy may be enough to resolve the
infection (De Renzo et al. 1994; Wong et al. 1999). Yet, for patients with renal failure
treated with erythropoeitin, B19 may cause severe worsening of anaemia, necessitating
a withdrawal of and a contradiction in case management (Lim 2005).
B19 vaccine has been tested in clinical trials and induced strong humoral responses. By
recombinant technology, empty capsids consisting of 25% VP1 and 75% VP2 have
been produced. The immune response to such vaccine can be measured by assessing
levels of neutralizing antibodies (Ballou et al. 2003). However, to receive long-term
effects from the vaccine, cellular immunity must be considered (Isa et al. 2005).
Presumably, certain patients at risk for B19 infection, such as those with sickle cell
anaemia, could benefit from vaccine prophylaxis. Moreover, widespread vaccination
might be worthwhile if the clinical spectrum of B19 becomes related to a larger number
of diseases than is known today.
2 AIMS OF THE STUDY
To investigate the frequency of B19 DNA in bone marrow of patients with
haematological and rheumatological diseases.
To study clinical and immunological parameters in patients with signs of
persistent B19 infection, specifically to define whether persistence is
associated with a general immunodeficiency or a selective failure of the
3 MATERIALS AND METHODS
3.1 STUDY SUBJECTS
The patients included in the five papers and their connections are illustrated in
3.1.1 Paper I
In March 1993, a 36-year-old woman was referred to the Clinic of Infectious Diseases
in Borås, Sweden, after presenting with six months of variable symptoms including
myalgia, fever and exanthema. Routine laboratory analysis had revealed leukopenia
and anaemia. BM aspiration demonstrated megaloblastic erythropoesis of unknown
etiology. B19 serology was IgM negative but strongly positive for IgG. When BM
aspiration was repeated, a sample for PCR analysis of B19 was sent to the laboratory
capriciously and found positive. This result was the beginning of my interest in this
subject and resulted initially in acquiring BM samples from other patients with chronic,
unexplained symptoms such as fatigue, arthralgia, leukopenia and fever. The index case
and nine others are described superficially and retrospectively in Paper I. Results for
patients 1-5 were, in contrast to those for patients 6-10, all B19 DNA positive in BM.
Subsequently, patients 1-4 were found to be persistently infected and are further studied
in Papers IV and V.
3.1.2 Paper II
During the period October 1994 to March 1996, one hundred patients were included in
this study. They were all undergoing BM sampling indicated by haematological
reasons, at the Clinic of Medicine, Södra Älvsborg Hospital, Borås, Sweden. They were
diagnosed with acute lymphatic leukaemia (1), acute myeloid leukaemia (17), chronic
lymphoproliferative disease (23), chronic myeloproliferative disease (15), plasma cell
disease (13), anaemia (11), other cytopenias (11), adenocarcinoma, (3) and other
diagnoses (6). Patients were offered inclusion in the study regardless of symptoms or
diagnosis, simply because BM sampling was planned. We do not know how many
patients chose not to participate or how many BM samples were taken during the study
Patients in the five different Papers. All four
B19 positive patients in Paper I are also
included in studies presented in Paper IV and
V. Seven patients in Paper III are included in
paper IV and one of them also in Paper V
whereas none of the B19 positive patients in
Paper II are included in any other study.
3.1.3 Paper III
During the study period between 1997 and 1999, when 996 patients visited the
Department of Rheumatology in the Clinic of Medicine at Södra Älvsborg Hospital
Borås, Sweden, 130 of them were randomized and invited to participation in the study,
and 50 accepted. The diagnoses of study subjects appear in Table I. Patients with
septic arthritis or crystal-induced tenosynovitis were not included. A healthy control
group had been desirable but was not available.
Distribution of diagnoses in the study group,presented in Paper III, and compared with
all patients attending the clinic during the study period.
Diagnosis/ symptoms Study group Population
Giant cell arteritis
50 100 996 100
= Rheumatoid arthritis
= Systemic lupus erythematosus
= Mixed connective tissue disease
= Undefined connective tissue disease.
3.1.4 Paper IV
Acutely B19 infected patients
Eight previously healthy female adults were prospectively identified and included in
the study after their anti-B19 IgM serology proved to be positive, and B19 DNA was
found by PCR in samples referred to Clinical Virological Laboratory at Karolinska
Hospital. All these women had at least three of the four following symptoms: fever,
arthralgia, fatigue and rash.
Persistently B19 infected patients
Initially 25 patients were included in the study because of fulfilling criteria for
suspected persistent B19 infection based on chronic clinical symptoms and a B19-
positive PCR-sample from BM. Some were recruited after being diagnosed in routine
clinical practice in Clinic of Infectious Diseases, Södra Älvsborg Hospital, Borås,
Sweden (n=16) and some from the study presented in Paper III (n=9) who were
included consecutively regardless of diagnosis or other parameters until we reached the
number 25. Later, criteria were altered to be more restrictive, and two B19-positive BM
samples with at least six months interval were required for our own definition of
persistent infection. In this way three patients were excluded, and 22 subjects remained:
four males and 18 females.
Healthy seropositive individuals
Eighteen B19 IgG positive healthy laboratory workers were included, all negative for
B19 IgM and B19 DNA in serum and without history of recent B19-related symptoms.
Healthy blood donors
Nineteen healthy blood donors served as controls in one the part of the study,
investigating cellular immune response to concanavalin A.
3.1.5 Paper V
Persistently infected patients
Seven females and two males also included in the study described in Paper IV
participated in the study. They were B19 IgG positive and IgM negative but B19 DNA
positive in BM during follow up (two to eight years).
Healthy seropositive individuals
Fourteen healthy individuals were included, seven males and seven females, all were
B19 IgG positive but negative for B19 IgM and DNA in serum.
Healthy seronegative individuals
Three individuals negative for B19 IgG were used as methodological controls.
3.2 COLLECTION OF SAMPLES
3.2.1 Bone marrow
BM puncture was performed with patients given local anaesthesia (xylocain) in the
posterior superior iliac spine. BM samples were delivered to the laboratory in sterile
tubes without additives. BM smears so-collected were examined by light microscopy at
the Clinical Chemical Laboratory, Södra Älvsborg Hospital, Borås, Sweden.
Blood was sampled by venous puncture and delivered to various laboratories for each
1. Sera were placed in sterile tubes without additives for PCR and serology.
2. Blood samples in tubes with EDTA and ACD, respectively, were used for
analysis of HLA-type.
3. Blood in tubes with EDTA were analysed with flow cytometry for
subpopulations of lymphocytes.
4. Blood in tubes with heparin were used for analysis of IFN? production by
3.3 DETECTION OF B19 DNA
Serum and BM samples were tested in two different nested PCRs in Paper I and II.
These assays represented the structural (VP) and the non-structural (NS) proteins of the
B19 virus, as described previously (Broliden et al. 1998). Briefly, in each PCR, the
result was considered positive if at least two of three separate amplifications gave
identical results, and all controls were correctly positive or negative. At least one of the
two PCR:s (VP and NS) was required to be positive for each patient. Serum and BM
samples were analysed after heat treatment at 94°C for 10 minutes, followed by
centrifugation at 12,000 rpm for five minutes. Two µl of the supernatant was used in
the PCR assay. The first amplification consisted of 35 cycles and the nested round 25.
The following primers were used in Paper I and II.
The nested round:
nucleotides 2955 to 2974 and reverse primer 3364 to 3349.
nucleotides 3002 to 3020 and reverse primer 3291 to 3272.
The nested round:
nucleotides 1355 to1374 and reverse primer 1723 to 1703.
nucleotides 1399 to1422 and reverse primer 1682 to 1659.
In Paper III only NS-PCR was used and with a slightly modified outer forward primer.
3.4 SEROLOGICAL ASSAYS
3.4.1 Specific B19 IgG in serum
Enzyme linked immunosorbent assay
1. ELISA; Eurodiagnostica, Malmö, Sweden
2. EIA; DAKO, Glostrup, Denmark
3. EIA; Biotrin International Ltd, Dublin England
I , II
III, IV, V
3.4.2 Specific B19 IgM in serum
Indirect immunofluorescence test
IFA; Biotrin, International Ltd, Dublin, England I
Enzyme Linked Immunosorbent Assay
1. EIA; DAKO, Glostrup, Denmark
2. EIA; Biotrin International Ltd, Dublin, England
I , II, III, IV, V
These assays were kindly performed by Professor Klaus Hedman at the Department of
Virology, Haartman Institute, University of Helsinki, Finland. Avidity of B19
antibodies was measured by protein denaturing EIA:s using recombinant antigens
consisting of VP1 alone, VP2 alone and VP1 and VP2 together (VP1/2) in ratio 1/11.
Wells with antigen were incubated with serum and then washed with 8 M urea in PBST
(PBS containing 0.05% Tween) or only PBST, respectively. All wells were treated for
1 h with alkaline phosphatase-conjugated anti-human IgG and substrate for 30 min.
Denaturation of antigens with urea results in decreased bond between antibodies and
antigen especially in newly infected individuals who have antibodies with lesser avidity
than individuals infected long ago. By using the EIA absorbance quotient between
urea-treated and urea-free samples, the time interval since primary infection in each
patient could be estimated. Avidity levels >25% indicate past infection of more than six
months (Söderlund et al. 1995).
3.4.4 Epitope type specificity
Epitope type specificity (ETS) was, just like the avidity test, used to estimate if
antibodies against B19 were signalling past or acute infection. By using denaturated
B19 antigens, namely self-assembled VP1/2 capsid and VP2 capsid with 8 M urea,
antibody response to VP1/2 antigens have been shown to be more stable than to VP2
in patients with past infection. On the contrary, there is no such difference in acute
infection. Consequently, the extent of antibody responses to VP1/2 relatively VP2
rises over time, resulting in low quotient (<2) for acute infection and elevated scores
(>5) afte more than six months after the infection.
3.4.5 Neutralizing antibodies
Dr. Jennifer Bostic, MedImmune Inc, Gaithersburg, Maryland, USA helpfully
performed analysis of neutralizing antibodies. A megakaryoblastic cell line
(UT-7/Epo), adapted to grow with erythropoietin and semipermissive to B19 infection,
was used and incubated with serum at various dilutions. The neutralizing titre was
defined as the highest dilution that yielded no B19 RNA, which was analysed with
reverse transcriptase-PCR (RT-PCR).
3.5 IMMUNOPHENOTYPING OF LYMPHOCYTES
Subpopulations of lymphocytes in blood from persistently infected patients were
determined by fluorescent activated cell sorter (FACS) and presented in Paper IV. The
samples were stained with combinations of murine monoclonal antibodies directly
conjugated with fluorochromes, fluorescein isothiocytanate (FITC), phycoerythrine
(PE) or peridinin chlorophyll protein (PerCP). Blood was incubated at 4o C for 15
minutes with antibodies in the concentrations recommended by the manufacturer. The
following flurorochrome-conjugated antibodies were used: anti-CD3, -CD4, -CD5,
-CD8, -CD16, -CD19, -CD45RA, -CD45RO, -CD56, and anti-HLA DR (all antibodies
were purchased from BD Biosciences, Mountain View, CA, USA). Cell analysis was
performed by flow cytometer (FACSCalibur, BD Biosciences). Dot plots and quadrant
statistics from three-colour analysis were generated by CellQuest software (BD
Biosciences). The absolute number of blood lymphocytes was determined using a
haematological cell counter (Sysmex K-4500; TOA Medical Electronics Co, Japan).
The results for each subpopulation were expressed as the percentage of lymphocytes
and as the number of cells x 109/L.
Amino acid sequences published by Shade were used as templates for synthesizing a
total of 210 peptides representing NS1, VP1ur and VP2 (Shade et al. 1986). Nonamers
with an overlap of one amino acid were purchased (Mimotopes, Clayton, Victoria,
Australia) except for the VP2-nonamers, which together with all other peptides were
synthesized in-house by F-moc chemistry. Peptides were first dissolved in
dimethylslufoxide (DMSO) and were then diluted with distilled water to a DMSO
concentration of less then 5%.
3.7 INTERFERON GAMMA (IFNγ) ELISPOT
The cellular immune response in general, described in Paper IV, was investigated by
analyzing responses to concanavalin A, and helpfully performed by Associate Professor
Bengt Andersson at Clinical Immunology Laboratory, Sahlgrenska University Hospital,
Sterile microtitre plates with nitrocellulose bottoms (Multiscreen-HA plates; Milipore,
Mulsheim, France) were coated overnight at 4o C with 100 μλ of mouse monoclonal
antibodies to human IFNγ in sterile filtered PBS (final concentration of 15 μg/ml;
Mabtech, Stockholm, Sweden). After washing five times with PBS, 200 μL of
peripheral blood mononuclear cells (PBMC) in cell culture medium were added in
different concentrations (5 x 104, 2 x 104 and 1 x 104 cells per well) in duplicates.
Concanavalin A was added at a final concentration of 25 μg/ml and unstimulated cells
were used as controls. The PBMC were incubated for 48 hours at 37o C in a humid
atmosphere with 5% CO2. Plates were washed and 100 μl of biotinylated mouse
monoclonal anti-IFNγ (final concentration 1 μg/ml; Mabtech) was added and they were
incubated at room temperature for 2 hours. After washing, 100 μl of streptavidin con-
jugated with alkaline phosphatase (final concentration 4 μg/ml) was added. The spots
were developed by BCIP-NBT (Life Technologies) substrate diluted in Tris buffer.
Blue immunospots, considered to represent individual IFNγ secreting cells, were
counted in a dissection microscope.
The cellular immune response to B19 virus, described in Paper V was measured in vitro
by analysing antigen-specific secretion of IFN? with a previously described ELISpot
assay (Larsson et al. 1999), but modified by the use of streptavidin conjugated with
alkaline phosphatase and a corresponding substrate (BioRad, Hercules, CA, USA).
PBMC were mixed with single or pooled peptides at a concentration of 10µg/ml and
plated in triplicates of 2x105 cells/ well. After 18 hours of incubation (37o, 5% CO2,
with 95% humidity), spots were visualized, counted and registered as spot-forming
cells (SFC)/106 PBMC after subtracting the release from negative controls in wells with
PBMC without peptide stimulation. A cut-off level was set, and responses were
determined to be positive or negative. IFN?-release in the ELISpot assay was
previously shown to correlate well with the presence of specific CD8 T-cells and
cytolytic activity (Lalvani et al. 1997; Horton et al. 2004). To ensure that IFN? detected
in the ELISpot originated from CD8-cells, depletion assays in ELISpot were done, as
described previously (Norbeck et al. 2005). The responses in ELISpot were lost by
depletion of the CD8 cells using mini-MACS CD8 microbeads and subsequent
magnetic separation (Miltenyi Biotec, Bergisch Gladbach, Germany) compared to non-
HLA expression was analysed by Consultant Jan Konar at the Tissue Typing
Laboratory, Sahlgrenska University Hospital, Gothenburg, Sweden. HLA class I
expression was determined by serology. Lymphocytes in peripheral blood were
isolated, and HLA was assessed by using a complement dependent cytotoxicity
technique (CDC) (Vartdal et al. 1986). HLA-DR genotyping was performed using a
PCR-SSP technique as described elsewhere (Olerup et al. 1992).
3.9 QUANTIFICATION OF CYTOKINES
The cytokines’ concentrations were longitudinally analysed by multiplex beads assay
(Luminex). We quantified twelve different cytokines: IL1?, IL2, IL4, IL5, IL6, IL8,
IL10, IL12, IL15, GM-CSF, INF? and TNF? in sera from all acutely infected
individuals (Paper IV). Cryopreserved serum samples from persistently infected and
healthy seropositive individuals were analysed on one occasion. A commercially
available multiplex beads immunoassay, based on the Luminex platform (Biosource
International, Inc, Camarillo, CA, USA) was used according to the manufacturer’s
procedure. All samples were run in duplicate. Briefly, beads with defined spectral
properties were conjugated to the analyte-specific capture antibodies. Beads, samples,
standards and controls were pipetted in a filter-bottomed 96-well plate and incubated
for 2 hours while shaking (550 rpm). After three washes, the biotinylated detector
antibodies were added to the beads and incubated for one hour at room temperature.
Streptavidin conjugated to R-phycoerythrin (SA-PE) was added to the wells after
several washings and incubated for 30 minutes. By monitoring the spectral properties
of the beads and the amount of fluorescence associated with PE, the instrument
measures the concentration of the analytes presented in the original specimens. The
data (mean fluorescence intensity) were analyzed using a Luminex reader (Luminex,
Austin, TX), and the mean concentration was calculated as pg/ml serum.
4 RESULTS AND DISCUSSION
4.1 PAPER I: CLINICAL AND LABORATORY FINDINGS IN
IMMUNOCOMPETENT PATIENTS WITH PERSISTENT PARVOVIRUS
B19 DNA IN BONE MARROW
This retrospective report describes our initial experiences of PCR analysis of B19 DNA
in BM samples and how my interest begun. In March 1993, at the Clinic of Infectious
Diseases, Södra Älvsborg Hospital, Borås, Sweden, a 36-year-old woman presented
with a six-month-history of variable symptoms including myalgia, fever and
exanthema. Previously, routine laboratory analysis had revealed leukopenia and
anaemia. BM aspiration demonstrated megaloblastic erythropoesis of unknown
etiology. Parvovirus serology was IgM negative but highly positive for IgG. When BM
aspiration was repeated, a sample for PCR analysis of B19 was sent, to the Virological
Laboratory at thee Swedish Institute for Infectious Disease Control, and found positive.
This unexpected result inspired me to analyse B19 DNA in BM samples in other
patients with chronic, unexplained symptoms such as fatigue, leukopenia and fever.
That original sampling was irregular, intuitive and without systematic order so could be
referred to “randomized biased selection.” Our first doubts and questions, which still
continue, were how to evaluate the relevance of B19 DNA in patients’ BM and whether
further immunogical investigations were clinically indicated. Knowledge of the natural
virological course of B19 was, and remains, limited and several explanations for
positive finding of B19 DNA are possible.
1. Chronic active infection
2. Persistence as a part of the natural course of B19 infection analogous to that
of herpesvirus infection
3. Fragments of B19 DNA without the ability to replicate viruses
The diagnosis of this first patient (case 1) was relatively obvious, since she had several
clinical signs characteristic of B19 infection. She had anaemia associated with
megaloblastic erythropoesis (Figure 4), suffered from arthralgia and had exanthema, a
combination that made B19 infection probable. Initially, her serum was declared IgM
negative, but reanalysis determined that it was positive. Furthermore, when the patient
was confronted with the diagnosis, she recalled an outbreak of fifth disease in the day
care centre where she worked during the same year as she fell ill. Her persistent
symptoms, i.e., anaemia and leukopenia, led to treatment with IVIG that was initiated
in 1995, after she had endured more than two years of disease (Figure 5). The levels of
haemoglobin and leukocytes were improved after IVIG therapy, but her fatigue and
arthralgia persisted, and the virus was not cleared from BM. ETS and avidity were
tested to estimate the duration of the disease, and was helpfully performed by Dr. Maria
Söderlund at the Department of Virology in Helsinki, Finland. The results indicated
past infection of more than six months duration. The patient is further described in
Paper IV and V.
Laboratory parameters associated to case 1 in the first paper. She is also described in
Paper IV and V. The figure show variations in hemoglobin (Hb, g/l) leukocyte count
in blood (WBC, leucocytes x 109/l) and in time corresponding treatment with
immunoglobulin (♦) Initially the treatment was given intravenously (IVIG), one dose
of 30 g in December 1994, three doses during three days in May 1995 and five doses
during five days in September 1995. Subcutaneous treatment with immunoglobulin
(SCIG) was given in doses of 40 ml twice a month from April to October 1996. B19
positive PCR samples in bone marrow are also shown (+).
BM preparation in case 1 presented in Paper I. Megaloblastic erythropoetic
cells are shown with white arrows and normal erythropoetic cells with black
The patient recorded as case 2 suffered from arthralgia fever and fatigue subsequently
to acute respiratory symptoms. She was working in a childrens’ day care centre, so
stimulated by the results from case 1, we examined her BM searching for B19 DNA
and the result from the PCR assay was positive. Her symptoms, including those of
chronic fatigue syndrome, persisted for several years, but her haematological status
remained normal and, except for B19 DNA in the BM, no other obvious clinical signs
of B19 infection appeared. Treatment with immunoglobulin was given and to some
extent subjectively connected to improvement. BM samples have been B19-DNA
negative several times but periodically positive again. Although the explanation could
be reinfection, variations in viral load combined with limited sensitivity in the PCR
procedure are more likely. ETS and avidity assay demonstrated a prior immune
reaction more than six months before, but the initial respiratory symptoms may or may
not have been related to B19 infection.
The patient designated as case 3 also worked in a childrens’ day care centre, but her
medical history was more complicated than the others’. In childhood she had symptoms
of motor and sensory dysfunction, and by the age of twelve, she underwent a muscle
biopsy, examined on the suspicion of hereditary muscle dystrophy but proving to be
negative. However, her symptoms persisted and propagated slowly. In 1991,
neurophysiologic examination demonstrated motor and severe sensory neuropathy of
unspecified origin. She was admitted to the Clinic of Infectious Diseases in 1994,
because of a long-lasting ulcer on her right foot, probably related to neuropathy. The
patient also experienced recurrent episodes of erythema nodosum since 1979, and when
it reappeared in 1995, rheumatologic investigation was initiated. Additionally, for many
years, she had suffered from periods of facial erythema and non-arthritic arthralgia in
knees, wrists and shoulders. Results of testing for rheumatoid factor and antinuclear
antibodies were positive. After thorough investigation in the Department of
Rheumatology at Sahlgrenska Hospital, Gothenburg, in 1997, including examination of
saliva secretion and lip biopsy, she was diagnosed with primary Sjögren’s syndrome.
In 1996, arthralgia, facial erythema, anaemia and probably the patient’s profession as
preschool teacher all contributed to the decision to look for B19 DNA in her BM.
When that result proved to be positive, the current question about relevance was
Various neurological manifestations associated with B19 infection are well described in
several reports (Barah et al. 2003; Aguilar-Bernier et al. 2006; Kerr et al. 2002), and
IVIG treatment is considered of potential value (Nigro et al. 1994). Therefore, this
patient’s slow but serious progressive symptoms motivated tentative treatment with
IVIG, which was given in doses of 0.4g / kg five days followed by one dose monthly
for six months. No obvious alteration in clinical status was noted initially but her
subsequent symptoms of erythema nodosum occurred much less frequently, and the
neuropathy, which according to neurophysiologic tests, had worsened between 1991
and 1996 was without evident changes between 1996 and 2000. After therapy ceased,
follow-up analysis of B19 DNA in BM performed in 1996, in 1997 and most recently
in 2000 all showed positive results, illustrating the failure of viral clearance.
Interestingly, the very first muscle biopsy from 1962 was found, re-tested and declared
positive for B19 DNA. With respect to the difficulty of establishing accredited analysis
of 40-year-old, formalin-fixed samples, the finding was astonishing. Was the patient
already persistently infected in childhood? Was the chronic immune stimulation by this
virus the pathogenetic origin of her disease manifestations including erythema
nodosum, facial erythema, autoantibodies and progressive neuropathy?
The profession of the patient case 4 represents was, not surprisingly, pre-school teacher.
She was admitted to the clinic in October 1996 after four weeks of continuous fever
and arthralgia in the knees. Her routine haematology status was normal, and serology
concerning cytomegalovirus, Epstein-Barr virus and human immunodeficiency virus
was negative. Serological testing for B19 demonstrated highly positive results for B19
IgG and slightly positive B19 IgM indicating primary B19 infection. The symptoms
proceeded without relief, and persistent B19 DNA remained in BM samples in
November 1996 and March 1997. Treatment with IVIG 0.4 g/ kg was given once in
May 1997 and additional three doses three weeks later. In connection with the latter
doses, the patient became more acutely ill with high fever but no focal sites of
symptoms for some days. Subsequent BM sampling in July was negative, and the
arthralgia vanished, but the inconvenience of fatigue and fever has continued over the
years fulfilling criteria for CFS. B19 DNA in BM was again positive in September
1997 and in 1998 but negative when analysed in 1999. That year, she was struck down
by a miscarriage in week 15 of pregnancy, but another pregnancy ended successfully in
2000 when an acute Caesarean was performed because of pregnancy-related toxaemia.
The child was born healthy without signs of infection. Although the patient had another
miscarriage in 2002, in 2003 she gave birth to a healthy girl after a pregnancy without
Primary B19 infection during pregnancy is associated with miscarriage but also
hydrops fetalis, mainly in the second trimester, and intrauterine fetal death later in
pregnancy (Norbeck et al. 2002; Skjoldebrand-Sparre et al. 2000). B19 has also been
linked with preeclampsia and eclampsia (Selbing et al. 1995; Yeh et al. 2004). If and
how chronic B19 infection might affect the foetus is not known but should be
considered, since persistent infection logically might be related to some sort of
impaired or deficient immune response to B19 in the mother.
The patient known here as case 5 had fever and fatigue for ten months previous to
examination for B19 DNA in BM, which proved to be positive. Sampling 11 months
later was negative, and in follow up the patient was asymptomatic.
The patients designated as cases 6 to 10 were examined for similar reasons as those
constituting cases 1-5, chronic symptoms of unknown origin, but with components
consistent with known manifestations of parvovirus B19. They were all B19 DNA
negative in BM as summarized in Table 2.
Clinical and virological findings in five patients with negative B19 DNA in BM
presented in paper I.
Case Clinical presentation
6 Leukopenia, anaemia, fever
7 Chronic leukopenia
8 Anaemia and fever
10 Recurrent arthralgia and fever
The conclusions from these case reports in which B19 DNA was consistently found in
BM are and were difficult to draw, but we established that some individuals were B19
DNA positive in BM and others were not. One explanation could be that life-long viral
persistence in BM is the natural course after primary infection and that our method was
not sensitive enough to detect all instances. If so, viral load could be of interest, and a
positive result in our qualitative test could indicate a relatively high extent of viral
replication. Actually, a recent study with a sensitive quantitative PCR has demonstrated
B19 persistence in serum for more than one year after acute infection (Lindblom et al.
2005). Our qualitative PCR has turned out to have sensitivity about 103copies/ ml.
Another explanation for positive PCR results could be fragments of DNA remaining
from primary infection, even over long periods of time. This possibility was not
confirmed, since B19 DNA from some of the patients (cases 1 through 4) was later
sequenced, and the complete B19 virus genome was present in each patient
CFS is a diagnosis based on internationally accepted criteria, defined first in 1988 and
revised in 1994, describing a disease with debilitating fatigue lasting more than six
months and associated with at least four of the following symptoms: impaired memory
or concentration, tender lymph nodes, myalgia, arthralgia, headache, unrefreshing sleep
and postexertional malaise (Fukuda et al. 1994). That patients with B19 infection
frequently develop this syndrome is well known, but when larger populations of
patients with CFS were investigated, no sign of B19 was found. Undoubtedly, this
disease has a broadly heterogeneous aetiology (Maquet et al. 2006).
Numerous viral infections, not only that from B19, have been discussed as possible
causes of a full range of autoimmune inflammatory diseases. One theory regarding
pathogenesis has been a “hit and run phenomenon” in which virus acts as a trigger
factor that initiates an immunological reaction. Subsequently, this reaction becomes
self-generating and, possibly by molecular mimicry, continues to evoke a destructive
autoimmune response even after the virus has been eliminated (Holtzman et al. 2005;
Rouse et al. 2002). Another hypothesis, illustrated by case 3, is that persistent viral
infection causes autoimmune manifestations from chronic immune stimulation. An
example is the association between chronic hepatitis B virus infection and periarteritis
nodosa, in which the deposition of immune complexes may cause vasculitis in various
organs (Trepo et al. 2001). The autoimmune disease Sjögren’s syndrome has been
linked with such viral infections as that by Epstein-Barr virus (Wen et al. 1996) herpes
virus 6 (Newkirk et al. 1994) and HTLV1 (Nakamura et al. 1997). An association with
B19 has been investigated in two small studies but with negative results (De Stefano et
al. 2003; De Re et al. 2002). The patient in case 3 was diagnosed with Sjögren’s
syndrome, yet her B19 DNA analysis was negative when a sample from the salivary
gland was examined with PCR. However, immunological manifestations secondary to
infectious diseases may not necessarily be related to a viral presence in the affected
In conclusion, this retrospective report contains the results of our first use of PCR to
search for B19 DNA in BM samples. The positive findings in patients with relevant,
chronic symptoms raised questions that were important to answer and became the basis
for our further studies.
Prevalence of parvovirus B19 DNA in bone marrow of patients with haematological
How common is B19 DNA in BM samples in a larger population?
High frequency of parvovirus B19 DNA in bone marrow samples from rheumatic
How common is B19 DNA in patients with rheumatologic diseases?
Cytokine responses in acute and persistent human Parvovirus B19 infection
Do patients with persistent B19 infection have any general immunological
characteristics in common?
Aberrant cellular immune responses in persistently human parvovirus B19-infected
Do patients with persistent B19 infection have any specific B19-related immunological
characteristics in common?
4.2 PAPER II: PREVALENCE OF PARVOVIRUS B19 DNA IN BONE
MARROW OF PATIENTS WITH HAEMATOLOGICAL DISORDERS
Paper I presented the histories of patients with chronic symptoms and positive findings
of B19 DNA in BM samples. These findings led to questions concerning the frequency
of B19 DNA in the general population. If, speculatively, B19 persistence in BM was
the natural course of primary B19 infection, the frequency of positive test results ought
to be rather high, since the seroprevalence in the adult population is over 50% and more
than 80% in the elderly. A large study including healthy individuals would have been
desirable, but BM puncture is an invasive method limiting the possibilities. That was
the reason for investigating B19 DNA in BM of patients with haematological disorders.
Their BM was examined for diagnostic purposes or as follow-up after chemotherapy;
that is, sampling BM for B19 DNA analysis was not a motive. A total of 100 patients
was enrolled in the study between October 1994 and April 1996 at the Haematology
Department, Clinic of Medicine, Södra Älvsborg Hospital, Borås, Sweden. Regrettably,
neither the total number of BM punctures performed during the study period nor count
of patients denied inclusion in the study was recorded. Moreover the population was
heterogeneous including patients with such varied diseases as acute lymphatic
leukaemia (1), acute myelocytic leukaemia (17), chronic lymphoproliferative disease
(23), chronic myeloproliferative disease (15), plasma cell disease (13), anaemia (11),
other cytopenias (11), adenocarcinoma (3) and other (6).
The main purpose of the study was to examine the frequency of B19 DNA in BM, but
ETS and avidity were also assessed to establish if the primary infection was more
recent than six months. Results indicated borderline values in the patient designated
case 1 and signs of acute infection in that of case 2. They were IgM- positive indicating
recent infection, but B19 PCR was negative in BM as well as in serum. Both patients
had, related to their haematological diseases, elevations of polyclonal IgM in serum;
therefore, falsely positive B19 IgM cannot be excluded, or is even probable. In general,
results from serological assays are questionable in patients with haematological
disorders; additionally, our analysis of avidity in these cases is of limited value.
Problems with serological testing are further illustrated by the fact that only one of nine
myeloma patients was IgG-positive. Myeloma is characterized by the production of the
monoclonal M-component, but the production of polyclonal antibodies against various
infectious agents is depressed. Consequently, when studying seroprevalence, the true
history of exposure to B19 may be underestimated. Nevertheless, the seroprevalence in
these patients was 59%, corresponding to the expected figure in an adult population.
However, patients with haematological diseases are at risk for acquiring B19 infection
by transfusion and immunoglobulin treatment (Cohen et al. 1997; Hayakawa et al.
Four patients were positive for B19 DNA in BM, whereas none was positive in serum.
Symptoms, possibly related to B19 infection, such as arthralgia, exanthema and
anaemia, were a part of the medical history for three of them, but they had no obvious
signs of ongoing B19 infection at the time of BM puncture. Obviously, it is difficult to
determine if manifestations like anaemia are B19 related or not in patients with
As mentioned, patients at a haematological department are at risk for acquiring B19
infection, but they might also be prone to developing persistent infection. When one
attempts to approximate the prevalence of B19 DNA in BM in the general population,
the result may be an underestimation. In a previously published study, B19 was
detected in the BM of four patients with chronic B19-related arthropathy but not found
in any of six seropositive healthy controls (Foto et al. 1993). The frequency of B19 in
BM samples was described in two studies published 1997. Liu detected B19 DNA by
in situ hybridization, which was positive in seven of 81 (8.6%) AIDS patients (Liu et al.
1997), and Cassinotti reported positive B19 DNA in four of 45 healthy bone donors
(Cassinotti et al. 1997). Later, Heegaard presented data from 153 healthy BM donors of
whom 73 % were IgG-positive and four (2.6%) B19 positive in BM (Heegaard et al.
2002). These figures are coherent with our results. One question is evident, though:
What is the sensitivity of the PCR method? In Heegaard`s study, the sensitivity was
stated as 200-2000 viral copies /ml sample volume, and in our laboratory, it has later
specified as 1000 copies/ ml. Consequently, one can not exclude that the true general
prevalence of B19 DNA in BM is actually higher than in those studies and that only
high viral loads have been detected so far. Disregarding these aspects, our main
conclusions from this study were that B19 DNA is not a general finding in seropositive
individuals and that persistence of B19 DNA in BM is not a common course after
4.3 PAPER III: HIGH FREQUENCY OF PARVOVIRUS B19 DNA IN BONE
MARROW SAMPLES FROM RHEUMATIC PATIENTS
We previously presented a case report, Paper I, describing patients with various chronic
symptoms and with persistent findings of B19 DNA in BM. Paper II included 100
patients with haematological disorders, and from the results we concluded that viral
persistence is not a general finding in seropositive individuals. Next, because
arthropathy is a well-established clinical symptom of B19 infection, especially in adults
(Torok 1992), and numerous reports have associated this virus with rheumatic
manifestations (Moore et al. 1999; Nikkari et al. 1994; Nigro et al. 1994; Pugliese et al.
2006; Ferri et al. 2005), we were inspired to investigate the frequency of B19 in
patients with rheumatic diseases. PCR analysis of BM samples was used, since we
know that erythropoetic cells are the predominant targets for B19 infection. Of patients
who visited the Department of Rheumatology at Södra Älvsborg Hospital, Borås,
Sweden, during the years 1997 to 1999, 130 were randomly chosen and invited to
participate in this study. The various diagnoses of the 50 patients who accepted are
summarized in Table 1.
B19 IgG serology was positive in 41 of those 50 patients (86 %), but all were B19
IgM negative. By PCR, 13 of the 50 patients (26%) were B19 DNA-positive in BM,
and the frequency was higher, i.e., seven of 22 (32%), in the subset of patients with
RA. When compared to our previous results in the study of haematological patients
(4% positive), the difference was significant (Figure 6).
paper IIpaper III
Number of patients
The figure illustrates the difference between the results in Paper II including patients
with haematological diseases in which four of 100 patients (4%) were B19 DNA
positive and in Paper III in which B19 DNA was positive in 13 of 50 patients (26%)
with rheumatological diseases. All results refer to B19 DNA in bone marrow analysed
Regretfully, we retrospectively are aware that a contemporary study comparing the two
study groups had been more powerful when interpreting results. However, the BM
samples were collected under similar conditions, and the PCR technique used was
mainly identical. In the haematology study, though, two PCR assays were used to
detect DNA representing structural (VP) as well as non-structural proteins (NS1), and
in the present study just the latter one was applied that might have underestimated the
figures. Furthermore, the parameters of gender and age differed, with female
dominance and younger patients evaluated in the rheumatology study. In our attempt to
look for B19 DNA in patients with various diagnoses, broad inclusion criteria were
used within the rheumatology spectrum, and the heterogeneous population was a
disadvantage, making statistical analysis doubtful. However, of 22 patients with RA
enrolled in the study, 19 were rheumatoid factor-positive, and 20 had erosive arthritis in
X-ray. B19 DNA in BM was positive in seven of 22 RA patients and interestingly, in
spite of the small population, the frequency of rheumatoid factor differed significantly
between B19-positive and -negative patients, as shown in Table 3. A difference was
also noted in that erosive arthritis was less common in the subset of RA patients with
B19 DNA in BM. The difference was not significant, but of interest, since it is known
that patients with established B19 arthropathy sometimes fulfil the criteria for RA but
do not develop joint destruction. Neither were other differences significant but cannot
be excluded since the number of patients was small. A larger study investigating B19
DNA in BM relative to erosive arthritis and rheumatoid factor in RA patients would be
Clearly, laboratory tests that detect B19 DNA are insufficient to prove viral etiology in
diseases of different kind, since B19 DNA is also found in healthy controls. In general,
the principle difference in diagnosing infections is dividing microbes into those that
colonize or persist in healthy individuals (E. coli, Candida, herpesvirus) and those that
are more obligatorily associated with disease (Legionella, Pneumocystis jiroveci, HIV).
In the latter group, detection of the microbe is usually enough for diagnosis, but in the
former, one must implement indirect methods like IgM serology, pathological
preparations and, most important, clinical picture. Previously, B19 was classified as an
abortive virus that causes an acute infection and is subsequently eliminated from the
body by the immune defence. Accordingly, detection of the virus would be diagnostic.
However, several studies have now demonstrated that B19 can persist irrespective of
clinical symptoms (Söderlund et al. 1997). Consequently, other diagnostic tools are
required for establishing a viral etiology such as immuno-histochemistry and
quantitative PCR. The origin of the sample, in which virus is detected, is probably of
importance when assessing B19 DNA presence and the relevance has to be established
organ by organ. With the methodology available, we concluded that in a large
proportion of rheumatic patients BM samples were B19 DNA-positive yet that these
B19-positive RA-patients tested positive for RA-factor less often than B19-negative
individuals. Only with larger studies will such results yield a statistically valid
Characteristics of a subgroup of patients presented in Paper III, the 22 RA patients, and
related to the presence of B19 DNA in bone marrow.
Parameters in 22 RA patientsB19 DNA pos
B19 DNA neg
Erosive arthritis in X-ray
Fisher’s exact test was used and demonstrated that presence of rheumatoid factor was
more common in RA patients without B19 DNA in BM than in patients with positive
B19 DNA samples. The differences in other parameters were not significant.
= rheumatoid arthritis
= not significant
4.4 PAPER IV: CYTOKINE RESPONSES IN ACUTE AND PERSISTENT
HUMAN PARVOVIRUS B19 INFECTION
This paper is divided into two parts. The first records the cytokine response after
primary B19 infection in eight acutely infected individuals over a period of 2.5 years.
The second describes 22 individuals with signs of persistent B19 infection studied for
immunological parameters. This latter part is particularly important in the context of the
present thesis, which questions whether persistent B19 infection can be referred to as a
general immune deficiency or a selective impairment in immunological competence to
eliminate B19. We also included 18 healthy seropositive individuals as controls for
comparison of cytokine responses.
Dominance of a Th1 cytokine response in acute B19 infection
The cytokine profile associated with acute B19 infection was studied in eight female
patients, followed for about 20-130 weeks. Diagnosis was determined by typical
clinical symptoms associated with positive B19 IgM serology and detection of B19
DNA in serum. In only two of the eight patients, pro-inflammatory cytokines (IL1ß,
GM-CSF, IL6, TNF?, and chemokine IL8 were elevated within the first weeks of acute
infection. Thereafter, the concentrations decreased to low or undetectable levels. With
respect to immunologic Th1 and Th2 responses, all eight patients had elevated levels of
one or more Th1-related cytokines (IL2, IL12 and IL15) but no rise of IFN? (except in
one patient at weeks 1-3). This is surprising and notable, since IFN? is usually
associated with the Th1 response. The possibility that a technical artefact caused our
results was considered but is not probable, since we demonstrated elevated levels of
IFN? in patients with persistent infection. The Th2 response, defined as detection of the
cytokines IL4, IL5 and IL10, was low during the entire study period. Th2 cytokines are
regarded as important in the immune system for development of immunoglobulin
production; however, the levels of B19 IgG were normal. Evidently, then, the
production of IgG antibodies may be induced by other mechanisms not studied here.
Results from previous studies of the cytokine profile in acute B19 infection contrast
with ours. Kerr investigated 84 patients with acute B19 infection and found a mixed
Th1/Th2 response with elevated levels of IFN?, TNF? and IL4 (Kerr et al. 2004). In
that study, however, sampling was done only once and at different time points after
infection than we used, perhaps accounting for the discrepancy. Production of IFN? is
crucial for the cellular immune response, and we previously noted that B19 antigen
evoked IFN? responses in CD8 cells of five acutely B19 infected patients (Norbeck et
al. 2005). The present study shows elevated levels of IL12 and IL15, both known to be
important in cellular immunity. IL12 stimulates IFN? production by CD8 cells and Th1
cells and IL15 differentiation of CD8 cells into mature effector and memory cells (Kidd
2003; Yajima et al. 2006).
Since we had demonstrated persistent infection of parvovirus B19 DNA in numerous
patients on more than one occasion during a period of several years, we were interested
in investigating different immunological parameters for general aberrations that might
be responsible. Initially 25 persistently infected individuals were included in the study.
They were deemed chronically infected with parvovirus B19 on the basis of B19 DNA
detected in BM and with long-lasting symptoms such as chronic fatigue, arthralgia and
myalgia. Some were recruited after being diagnosed in routine clinical practice at the
Clinic of Infectious Diseases, Södra Älvsborg Hopital, Borås, Sweden (n=16), and
others from the study presented in Paper III (n=9). These were included consecutively,
regardless of diagnosis or other parameters until the number reached 25. Later, we
altered the criteria to be more restrictive, and two B19 positive BM samples taken at
least six months apart were required for our own definition of persistent infection. In
this way three patients were excluded and 22 remained. No general criteria exist for a
precise definition of chronic B19 infection, and the natural course of B19 presence in
BM following acute infection is unknown. However, B19 DNA in BM is not a general
in seropositive individuals according to other studies (Paper II; Heegaard et al. 2002).
By revising our criteria to demand two positive samples taken at an interval of more
than six months, the presence and persistence of B19 virus were confirmed, and the risk
of accidental occurrence could be minimised. Characteristics of included patients are
briefly shown in Table 4.
Recent infection might account for viral presence in BM, and as much as 25-68% of
B19 infections are reported to be asymptomatic (Heegaard et al. 2002; Noyola et al.
2004). In the present study six patients were IgM-positive indicating recent infection. In
four of them the probable point of disease onset could be approximated and associated
with symptoms consistent with B19 infection. The other two patients had no recent
indicators as to the start of disease; furthermore, they were IgM-positive more than
once in samples taken at intervals of several years. Usually B19 IgM serology is
negative within a couple of months after primary infection (Anderson et al. 1985).
Persistent IgM may illustrate chronic infection but sometimes represents a false
positive reaction, which is fairly common in IgM assays in general, and particularly in
patients with rheumatologic diseases (Johnson et al. 2004). One of our patients who
was positive for IgM in 1999 as well as in 2003 suffered from SLE; still, she was B19
DNA-positive in BM on both occasions and viral relevance must be considered. In
several reports, SLE has been associated with B19 infection (Trapani et al. 1999; Hsu et
al. 2001), but contradictory results have also been published (Bengtsson et al. 2000).
Nevertheless, B19 infection and SLE share clinical features, and in individual cases, the
connection may be real. This patient with B19 infection and SLE was recruited from
the study investigating B19 DNA in BM from rheumatology patients (Paper III) and
illustrates a weakness of the study presented in Paper IV. The population of the study
was heterogeneous, and the patients enrolled had different origins, clinical practise and
a clinical study, respectively. A larger study with homogenous groups of patients
positive for B19 DNA in BM compared with matched patients with negative B19 DNA
would have been desirable but was at this time not possible to perform.
Table 4.??????????????????????????????????????????????????????????????????????????????????????????????????? ?? ????????????? ????
??? ????????????????????????????????????????????????????????????????????????????????????????????????? ?????????????????????????????????????????????? ??????????????????????????????? ?????????????????????????????????????????????????????????????????????????????????????? ?????????????????????????????????????????????? ???????????????????????????????????????????????????????????????????????????? ??????????????????????????????????????????????????????????????????????????????????????
MHC class I molecules (HLA A, B and C) are presented on the surface of nearly all
cells in the human body, whereas MHC class II (HLA DP, DQ, DR) are seen on
lymphocytes and dendritic cells, the latter presenting antigens to lymphocytes
following phagocytosis. Briefly, MHC class II, associated with antigens, incite
immunological activity in lymphocytes whereas MHC class I, likewise associated with
antigens, presented on infected cells designate them as targets for components of the
activated immune system. Viral antigens are presented on the surface of cells they
infect by binding to a groove in the MHC molecule, and the binding strength may vary
for different HLA type. Furthermore, some antigens are HLA-restricted and presented
exclusively in individuals with specific HLA types (Tolfvenstam et al. 2001).
Consequently individuals who are homozygous in HLA type may provoke a more
narrow antiviral immune reaction than seen in those who are HLA heterozygotic, since
the infected cells in the latter group may present more antigen epitopes.. In conclusion,
the MHC system is of great importance in immune defence, and the HLA type might be
related to our questions concerning B19 persistence in a subset of individuals.
Previously, Kerr described a relationship between symptomatic B19 infection and HLA
DR4 (Kerr et al. 2002), which is of interest since that allele is associated with RA.
Lacking access to matched controls, we compared the distribution of HLA class I
antigens in our patients to that in previously unpublished data from the Tobias registry,
which contained information for 40928 Swedish BM donors. The great number of
possible results in each variable makes a large control group of great value but BM
donors represent a selective part of the general population. Even so, HLA A9 was
shown to be more common in the 22 individuals, persistently infected with B19, than in
the 40928 controls (36% and 18% respectively, p=0.047, Fisher’s exact test), an
association that not has been described before. No differences were shown for HLA B.
Results concerning HLA DR did not show any significant differences when compared
with a previous study including a healthy Swedish control group (Berlin et al. 1997).
Immune phenotyping of lymphocytes revealed deviating results according to reference
values in some patients, but no uniform aberration in lymphocyte count was found to
explain viral persistence (Table 5). Six patients had subnormal levels of B-cells (CD19)
but their levels of immunoglobulin were normal. The finding of subnormal lymphocyte
levels in some patients was not obviously related to viral persistence, considering
inhomogeneous results in different patients and that most patients did not have any
aberration. Alternatively, the pathogenetic pathway of persistence may differ among
individuals. However, our more functional assay, measuring the production of IFNγ
after antigenic stimulation by concanavalin A, failed to demonstrate any functional
deficiency in T-cell response in these 22 persistently B19 infected individuals.
An explanation for viral persistence could be an imbalance of Th1/Th2 cytokines as
proposed by Spanakis concerning hepatitis C (Spanakis et al. 2002). Our measurements
of cytokine levels, though, did not reveal any definitive difference when samples from
the B19-infected patients were compared to those from healthy seropositive
individuals. The exception was IFN?-level, which was higher in B19 infected subjects
than in the controls, but this outcome could be an expression of disease rather than B19
infection per se.
Although we have not demonstrated any uniform deficiency in immunological function
in patients with persistent B19 DNA in BM, on the other hand, our results have not
excluded that possibility. Our methods to measure and describe the complicated system
of defence against infectious diseases are limited, and sometimes we are studying the
map rather than reality. Nevertheless, one of the major issues of my thesis is the
question whether viral persistence is related to B19-specific immunodeficiency or to a
general immunodeficiency of the host, and the former option is still without counter-
evidence. Further support for this former theory is also described in Paper V.
Absolute size of lymphocyte subpopulations (109/l) in the 22 patients presented in
CD3 and CD4
CD3 and CD8
CD16 and CD56 NK
= Cluster of differention
= Natural killer cells
= T helper cells
= T cytotoxic cells
= B cells
ELISpot results, IFN? production in SFC per million PBMC with and without
concanavalin A stimulation in patients presented in paper IV.
Study patients n=22
= concanavalin A
= not significant according to Mann-Whitney
40 Download full-text
4.5 PAPER V: ABERRANT CELLULAR IMMUNE RESPONSES IN
HUMANS INFECTED PERSISTENTLY WITH PARVOVIRUS B19
In our previous studies, the persistence of B19 DNA in BM was demonstrated in
patients with chronic symptoms. We also concluded that B19 DNA in BM is not a
general finding in seropositive individuals. Additionally, patients with viral persistence
tested immunologically have not had uniform general deficiencies in immune function.
Yet, irrespective of its relevance for clinical manifestations, the immunological
background of persistence is of interest. Broader implications are also possible; for
example, after hepatitis B infection in human adults, about 5% develop a persistent
disease, possibly related to low levels of cytotoxic T-lymphocytes directed to hepatitis
B antigen, indicating a selective immune deficiency (Nayersina et al. 1993). A similar
pathogenetic mechanism of B19 infection might be responsible for viral persistence and
such a mechanism would be well worth investigating.
Therefore, the study presented in Paper V was initiated and included nine subjects
persistently infected with B19. The criteria for persistent infection were defined as at
least two B19 DNA-positive BM samples identified at least six months apart. No other
selection criteria were used when choosing among subjects in the larger group of
patients with B19 persistence described in Paper IV. Fourteen healthy individuals were
also enrolled in the study as a control group. They were B19 IgG-positive but IgM- and
PCR-negative in serum. Achieving BM samples from these individuals to analyse by
PCR would have been of great importance but wasn’t possible for ethical reasons.
Additionally three individuals who were B19 IgG-seronegative were used as controls
for assay specificity.
The ELISpot technique was used to analyze T-cell responses ex vivo to stimulation by
several antigens of B19 virus. First, 210 peptides consisting of sequences with 15-20
amino acids were synthesized. The peptides overlapped by 10 amino acids and covered
about 92% of the entire translated proteins of the B19 genome such as NS1, VP1 and
VP2. Peptides were then mixed in diverse pools, each pool covering a sequence of
about 60-65 amino acids. PBMC were separated from blood by Ficoll-Paque and
assayed within 8 hours after venesection. A more extended delay would have the cells’
functionality. Triplicates of 2x105 PBMCs from each individual were incubated with
different pools of peptides, and T-cell responses were measured as IFNγ production
expressed by the number of spot-forming cells (SFC) per million PBMC counted in a
stereomicroscope. At the same time, SFC in wells with non-stimulated cells was
counted, and the values subtracted from results of cells exposed to antigen to eliminate
non-specific IFNγ production from the final results.
The baseline for a positive response was defined by values exceeding two standard
deviations given by values of SFC in all wells of a 96-well plate (Millipore) in an
experiment with samples from one patient. Negative controls were excluded as well as
the three highest and three lowest to rule out outliers. The cut-off line for a positive
response was defined by values exceeding baseline with 20 SFC. This value was
chosen on the basis of previous results in healthy individuals and acutely infected