Parvovirus B19 – Revised

1 Current Knowledge about the Pathogen

1.1 Characteristics of Parvovirus B19

The human parvovirus B19 (B19V) was occasionally discovered in a serum with haemagglutinating activity (serum no. 19 in line B) during the testing of sera in 1975 [1]. Years later, this virus was detected in the blood of patients with influenzalike symptoms or with aplastic crises. In 1983, it was successfully established that this virus causes fifth disease (erythema infectiosum) in children [2].

B19V is a member of the family Parvoviridae, subfamily Parvovirinae. Based on its biological characteristics, the subfamily is subdivided into three genera: Parvovirus, Erythrovirus, and Dependovirus. Dependoviruses require so-called helper viruses (adenoviruses, herpes viruses) in order to be able to replicate, while autonomous parvoviruses (genera Parvovirus and Erythrovirus) can replicate independently. In vertebrate, autonomous parvoviruses are wide-spread. Parvoviruses which only replicate in erythroid cells (bone marrow) are assigned to the genus of Erythrovirus. This includes the human B19V and parvoviruses of primates [3, 4, 5].

B19V is subdivided into three different genotypes [6]. Recently, additional parvoviruses (PARV) have been identified in humans [7, 8], which can, however, be clearly distinguished from B19V from the molecular biology point of view. Thus, the nucleotide sequence of PARV4 agrees with that of other parvoviruses in less than 30% of the positions. Therefore, PARV4 has been classified as a new virus species. The recently identified PARV5 differs from PARV4 in only 8–9% of the nucleotide positions and is therefore assigned to the same virus species as PARV4. The human bocavirus has been identified [9]. This virus clearly differs from the above described parvoviruses and has primarily been associated with respiratory infections.

Parvoviruses are non-enveloped, isometric viruses with a diameter of 18–26 nm. The particles consist of 60 copies of the capsid protein and contain single-stranded DNA of positive or negative polarity. The B19V genome has a length of 5,596 nucleotides. On the right and on the left, the encoding sequence of 4,830 nucleotides is flanked by inverted terminal repetitive sequences with a length of 383 nucleotides each. Out of these, 365 nucleotides possess the sequence of a palindrome, which leads to the formation of a hair-pin-like double-stranded structure at both end of the genome (terminal hairpins). DNA strands with positive or negative polarity are distributed in virions with equal frequency.

Replication: At least nine overlapping mRNA transcripts are formed during replication. All transcripts initiate at a common promoter (p6) [10]. There are two groups of spliced mRNAs, which encode for virus structure proteins VP1 and VP2, as well as the two proteins with 11 kDa and 7.5 kDa: There is only one unspliced mRNA species encoding for the non-structure protein NS1 with a molecular weight of 77 kDa.

Structure proteins: The two structure proteins VP1 and VP2 (capsid proteins) are encoded by the 3'-terminal half of the genome. The main structure protein VP2 (58 kDa) differs from VP1 (84 kDa) by a shorter reading frame (it is by 226 N terminal amino acids shorter). As in the case of all other parvoviruses, the surface of B19V consists of 60 copies of the capsid protein. Virus preparations contain 95–96% VP2 and 4–5% VP1. The structure of empty recombinant virus particles was analyzed in detail by X-ray structure analysis [11], and the infectious particles were characterized by cryo-electron microscopy [12].

Non-structure proteins (NS1): A high homology exists between the NS1 proteins of different parvoviruses. Conserved areas show a significant homology with the T-antigen of polyoma viruses and with the E1-protein of papilloma viruses. NS1 is located in the nucleus of B19V-infected cells and is involved in the regulation of gene expression as well as parvovirus DNA synthesis. So far, nothing is known so far about the biological function of the 7.5 kDa and 11 kDa proteins. The gene for the 11 kDa protein is essential for replication in cell culture [13].

Host range: B19V is a human pathogenic virus. Hosts other than humans are not known.

Host cell range: B19V has a narrow host cell range with pronounced tropism towards replicating human erythroid cells. The virus replicates in the bone marrow in the so-called BFU-E (erythroid burst forming units) and CFU-E (erythroid colony forming units) and the erythroid precursor cells.

The P-blood group antigen (globoside, tetra-hexo-seceramide) serves as cellular receptor. Individuals with the rare p-phenotype are resistant to B19V infections [14]. The presence of P-antigen in different tissues only partly reflects the cellular tropism of the B19V. P-antigen can be found on erythroblasts and megakaryocytes as well as on endothelial cells and foetal myocardial cells. However, not all cells expressing the P-antigen are permissive to virus production since the expression of the viral transcripts can be blocked in non-erythroidal cells [15]. An additional receptor may be required for virus entry into the cell [16, 17].

In-vitro replication: B19V replicates in precursor cells of the red blood cells. Replication was detected in primary cultures of human bone marrow cells, foetal liver cells, umbilical cord cells and peripheral blood cells. Erythropoietin is an important component of the cell culture medium supporting erythroid cell differentiation. Limited virus replication was also described in few permanent erythroid leukaemic cell lines (JK-1 and KU812Ep6) or in one megakaryoblastoid cell line (UT7/Epo, UT7-EpoS1).

B19V replication is associated with a cytopathic effect. Socalled giant pronormoblasts around five times the size of lymphocytes are observed in the bone marrow. Infected cells contain enlarged cell nuclei with chromatin displaced to the marginal area. A similar cytopathic effect is also observed in cultured cells. The virus particles accumulate in paracrystalline fields in the cell nucleus (fig. 1).

Fig. 1.

Infected KU812EP6 cell (Blümel, Boller).

1.2 Infection and Infectious Disease

Experimental infections of voluntary subjects have provided insights into virological, haematological, and clinical findings [18].

The majority of B19V infections take a clinically asymptomatic course. Detection of B19V-specific IgG antibodies indicates that an individual had gone through a B19V infection. Detection of specific IgM or virus DNA by hybridization indicates a present or recent infection. Circulating B19V DNA can be detected in the serum up to several months using polymerase chain reaction (PCR) or other nucleic acid amplification techniques (NAT). It is therefore difficult to interpret positive PCR results and draw conclusions with regard to the clinical situation. Antibodies against NS1 were frequently found in patients with persisting B19V infection [20]. In certain tissues (skin, liver), B19V can be detected throughout an individual's life using sensitive PCR techniques. This means that the DNA sequences detected show a so-called 'bio-portfolio' consisting of the DNA of all parvoviruses with which an individual became infected during his life-time [21].

A number of diseases is caused by B19V:

• –erythema infectiosum (synonyms: morbus quintus, fifth disease)
• –arthritis, arthralgia in 8% of the infected children and 80% of the adults
• –aplastic crisis in patients with chronic haemolytic disorders
• –chronic anaemia (in immunosuppressed patients and patients with leukaemia)
• –myocarditis
• –vasculitis
• –glomerulonephritis
• –foetal anaemia, hydrops fetalis
• –congenital anaemia.

In general, a slightly more serious course of the disease can be expected if the B19V infection occurs at a more advanced age. The three genotypes of B19V do not seem to cause differing clinical symptoms.

Erythema infectiosum (Morbus Quintus, Fifth Disease)

The pathology (patchy maculopapular rash, starting on the cheeks and spreading mainly to the exposed parts of the extremities; differential diagnosis: rubella and enterovirus infections) is caused by antigen-antibody complex formation and plaques in the skin and the joints. Mild fever and malaise are usually observed in children. Adults suffer frequently from rheumatic complications (inflammations of the joints comparable with those of rheumatoid arthritis).

Transient Aplastic Crisis

Transient aplastic crisis (TAC) occurs in patients with underlying haemolytic anaemia, shortened red cell survival period, hereditary sphaerocytosis, or increased red cells production (iron deficiency, acute haemorrhagia). TAC manifests itself by anaemia, reticulocytopenia, and aplasia of the red blood cells. Bone marrow necrosis may occur, and the disease may be lethal. The treatment of choice is transfusion of red blood cells. The disease is self-limiting, and the immune response protects the patient from recurring episodes and new infections. In rare cases, TAC was also observed in immunocompetent persons [22, 23].

Hydrops fetalis and Congenital Infections

B19V infections in non-immune pregnant women may lead to virus transmission to the foetus [24, 25]. The risk of foetopathy is highest in the first and second trimesters. Hydrops fetalis (approximately 20%), foetal death (approximately 9%), and spontaneous miscarriages (5%) have been reported [26, 27]. Infections have been most thoroughly studied in the second trimester. The most affected foetal organs are the liver (red blood cell production), and partly the heart (foetal myocardial cells express the P-antigen). Specific antibodies are formed in the foetus [28]. Untreated anaemia and damage to the heart can cause massive oedema in the foetus and intrauterine foetal death or post partum death of the infant. Treatment is possible by intrauterine transfusion, whereas virus was still detectable in the bone marrow of treated newborns, but not in the blood circulation. Injections of gammaglobulin have also been successful in treatment of hydrops fetalis [29].

Congenital infection may lead to persistent infection.

Chronic Anaemia

In patients with impairment of the immune system, permanent replication of the virus in the bone marrow can occur. This may lead to chronic anaemia (pure red cell aplasia; PRCA). Affected patients include individuals with congenital or acquired immune deficiencies (HIV infections), patients with lymphatic systemic disorders during or following chemotherapy, and patients undergoing iatrogenic immunosuppression (e.g. high-dose chemotherapy, blood stem cell transplantations, organ transplantations, or during treatment of autoimmune diseases). Besides an interruption of immunosuppressive and/or cytostatic treatment, i.v. administration of normal immunoglobulin preparations containing neutralising antibodies is the treatment of choice. In very rare cases, PRCA can also occur in immunocompetent individuals. In principle, B19V should be considered in the diagnostics whenever the origin of anaemia is unclear.

Other Diseases

In rare cases, forms of vasculitis, glomerulonephritis or myocarditis can occur in the course of B19V infections [e.g. 30]. The pathogenetic role of B19V has still not been understood. It is not clear whether productive virus replication occurs in the tissues affected (endothelium, myocardium) or whether secondary mechanisms play a role (e.g. antigen/antibody complexes). In addition, serious courses of encephalitis and acute heart failure have been described. However, the etiological role of the virus remains unknown in these cases.

1.3 Epidemiology

B19V is world-wide distributed. According to a new study in Germany, 10–20% of all children under 3 years underwent an infection, 10-to 19-year-old individuals showed antibodies against B19V in 66% and over 65-year-old persons in 75% of the cases examined [31, 32]. B19V infections are mainly observed during late winter to early summer in regions with a moderate climate (Europe). Epidemics occur at intervals of around 4–5 years.

The virus is transmitted mainly via the saliva or by droplet infection.

Nearly all currently characterized virus samples are assigned to genotype 1. Genotype 2 is found sporadically in Europe, the USA, and South America [33, 34, 35, 36, 37, 38]. A study in skin biopsies shows that genotype 2 was widespread in Finland until the 1940s. More than 40% of the B19V DNA-positive samples from patients born before 1940s contained genotype 2. On the other hand, this genotype was found in biopsies of young patients only very rarely [21].

Genotype 3 seems to occur mainly in the North African and West African regions [39], but has also been identified in samples from France [6].

1.4 Detection Methods and Their Significance

Antibody Detection

Commercial tests for the detection of IgG or IgM antibodies against B19V in the serum are offered as ELISA or immunofluorescence tests by various companies. The virus antigen is predominantly of recombinant origin (VP2 and VP1 from baculovirus or E. coli systems). Immunoblot tests equipped with recombinant antigens are also available for the confirmation of positive results.

Presence of IgG antibodies indicates that an individual has recovered from infection and is immune. IgM antibodies indicate a recent recovery from an infection or an existing infection.

Antigen Detection

Monoclonal antibodies are commercially available for research purposes and can be used for the detection of virus and infected tissues.

B19V particles have haemagglutinating properties. Passive haemagglutination tests have been developed for antigen detection and were used for the recognition of highly viraemic donators in blood banks [40, 41]. The haemagglutination test detects only highly viraemic samples (approximately 10⁸-10⁹ genome equivalents/ml and above). Moreover, this test will be negative, as soon as antibodies are formed in the patient, since the antibodies interfere with the haemagglutination. In principle, the haemagglutination test is, therefore, suitable only for the detection of highly viraemic samples from the antibody-negative window phase.

Virus Particle Detection

Identification of parvoviruses in the electron microscope is difficult because of the low particle size. Only highly viraemic samples are detected. In such cases, the identification is possible by means of immuno-electron microscopy, but this is a difficult technique.

Genome Detection

The dot blot is suitable for detection of B19V DNA in the viraemic phase. The PCR allows sensitive detection of B19V DNA. In addition, quantitative PCR methods were established for exact determination of the B19V DNA concentration. In establishing the international WHO standard for B19V DNA, the participating laboratories determined a ratio between International Units (IU) and genome equivalents (geq) of 1:0.6 to 1:0.8 [42]. High-titre samples (10⁶ IU/ml and above) indicate acute infections, whereas low-titre samples (up to 10⁴ IU/ml) are found in the later phase of acute infections or at chronic infections. In serum or plasma, B19V DNA can be detected for 6 months up to several years after acute infection. In certain tissues (e.g. skin, liver) B19V DNA can be detected even lifelong.

2 Blood and Plasma Donors

2.1 Prevalence and Incidence in Donor Populations

Prevalence and incidence of the B19V infection in blood donors reflect those of the general population. Persons over 20 years show antibodies against B19V in 40–60%, persons over 70 years in more than 85%. Examinations of blood donors revealed an average seroprevalence of 60%. The prevalence differs slightly within Europe. In the Scandinavian countries (Finland), a slightly higher prevalence has been observed compared to other European countries [43].

Since 18- to 30-year-old blood donors show a prevalence of only 40–60%, there is a permanent risk that blood donors are undergoing a primary infection at the time of donation. The rate of new infections per year can be estimated at up to 0.5–1% of the blood donors based on the documented infection rates in pregnant women [44].

There is a short highly viraemic phase (up to 10¹³ virus particles/ml) during primary infection. With the onset of the humoral immune response, virus titres in the blood rapidly drop to <10⁶ geq/ml. Low virus titres are observed in the subsequent months (up to 12 months). In a study in Japanese blood donors, low viraemic titres were observed up to 3 years post infection [45]. It is questionable whether low viraemic donations, already containing virus-specific antibodies are infectious. In a recent study [46], DNA levels fell below 10³ geq/ml within the first 3 months. However low DNA levels persisted for a period of at least 6 months. In this study with donations from Germany and Austria from 2004 to 2006, the average portion of high-titre donations (>10⁵ geq/ml) was 12.7 per 100,000 donations and the portion of donations with <10⁵ geq/ml was 261.5 per 100,000 donations per year. These portions are more or less equivalent to those of a previous study in plasma donors [47]. In this study, highly viraemic donations from the acute phase (≥10⁷ geq/ml) were observed in 1 of 480 cases and low viraemic donations in 1 of 8,000 cases. A number of studies on B19V prevalence in donor populations from various countries have been published confirming the above. A direct comparison of such epidemiological studies, however, is often not possible since the test methods (counter flow electrophoresis, dot-blot, various PCR methods) may differ in their sensitivity and seasonal as well as local outbreaks may influence the results. In a PCR study on blood donors in Scotland, B19V DNA was detected in one of approximately 300 blood donors during the seasonal phase of B19V infection; Yoto and co-workers [48] reported on a frequency of 1:167 donations. Outside such periods, a frequency of 1:10,000 to 1:50,000 could be expected [49, 50, 51]. At CSL Behring, approximately 31 million donations (predominantly from the USA and Germany) were examined in the past 7 years (starting in 2000), and a frequency of 1:6,200 B19V DNA (≥2 × 10⁶ geq/ml) was identified in the donations (Gröner, personal communication).

The portion of the variant genotypes 2 and 3 in the donor collective seems to be very small. In a recent study with samples from 2005–2007, genotype 2 was identified only in one of 232 highly viraemic plasma donations [52]. The prevalence of genotype 3 also seems to be low in the European population. However, it should be noted that the commercial NAT systems currently available do not safely detect genotype 3, and genotype 2 might be only partially detected. Genotype 3 was identified in one plasma donation in the USA.

The DNA of the other human parvoviruses such as PARV4 is frequently found in plasma pools (14% of the pools tested), unlike the DNA of bocaviruses, which was not found in any of the 167 pools tested [53].

2.2 Definition of Exclusion Criteria

No specific exclusion criteria are currently laid down (cf. 2.4).

2.3 Donor Testing and Significance

Testing of the donor or the donations for B19V is neither required according to the Guidelines for the Collection of Blood and Blood Components and for the Use of Blood Products (Haemotherapy) nor according to the recommendations of the Council of Europe nor the WHO recommendations on blood and blood products. Plasma donations intended for the production of pooled virus-inactivated plasma or for the manufacture of anti-D immunoglobulin products are pre-tested for their B19V DNA content in order to prevent the limit value of not more than 10⁴ IU B19V DNA/ml laid down in the European Pharmacopoeia from being exceeded.

The portion of immune donors can be identified using the test for IgG antibodies. The exclusion of donors with IgG antibodies from donating blood is not desirable, since such donations are not expected to transmit B19V (cf. 3.2), and the antibodies in the donations form an important part of immunoglobulin products. Theoretically, an antibody test could be used to detect viraemic donations among those donations stored in quarantine. However, testing with NAT systems is now the preferred testing method, and not haemagglutination testing.

Detection of viraemic donations is usually carried out by detection of virus genome (e.g. by NAT). Due to the partly extremely high DNA concentrations in a single donation, effective measures designed to avoid cross-contaminations are important. In the event of a mini-pool tested positive, the highly viraemic single donation can be identified by further testing. The comparability of the NAT results has improved considerably after introducing a WHO standard for B19V DNA [42, 54]. Since the rare genotypes 2 and 3 are assigned to the same virus species and probably display similar pathogenicity, the European Directorate for the Quality of Medicines (EDQM) has required also these rare genotypes to be detected in a NAT [55].

In highly viraemic donations, infectivity has to be assumed while infectivity of low viraemic blood donations which at the same time contain B19V-specific antibodies is questionable. In such anti-B19V-containing blood donations, the virus particles are probably present in the form of neutralising antigen-antibody complexes.

The haemagglutination test was evaluated in blood banks in Japan and Germany [40, 41, 56]. In principle, it is suitable for the detection of highly viraemic donations. Its sensitivity, however, was limited to high concentrations of ≥10⁸ genome equivalents/ml. Only small amounts of antibodies can lead to inhibition of haemagglutination. This explains that samples containing antibodies and/or medium to low DNA levels were not detected. Only single donations (no mini-pools) can be tested using this method. Transmissions of B19V by antibody-containing SD plasma with a concentration of 10⁷ genome equivalents/ml have been described [57]. Therefore, a fair amount of uncertainty exists as to whether the haemagglutination test really does detect all infectious donations. In most facilities, NAT is currently preferred to the haemagglutination test. Recently, an improved ELISA was tested for the detection of virus antigen. This test showed less interference with B19V-specific antibodies; its sensitivity, however, was limited to ≥10⁸ genome/ml. The antigen test is expected to recognise all 3 genotypes of B19V [58].

2.4 Donor Interviews

Since the course of B19V infections is often asymptomatic, even at high virus loads (up to 10¹³ geq/ml), interviewing the donor is not useful for identification of infected donors.

3 Recipients

3.1 Prevalence and Incidence of Blood-Associated Infections and Infectious Diseases in Recipient Populations

Considering the high frequency of B19V DNA detection in blood donors (cf. 2.1) as well as the number of recipients susceptible to B19V infection, infections by cellular blood products or plasma (fresh frozen plasma; FFP) should be relatively frequent. However, case reports have been published only very rarely [19, 59]. No reports on suspected cases of a B19V infection in connection with the administration of cellular products or FFP have been received at the Paul-Ehrlich-Institut (PEI) since 2004, even though it was pointed out in a previous publication by this working group that the possibility of B19V infection should be taken into account [60]. A B19V infection by cellular blood products seems to be never detected clinically, since such infections occur without any specific symptoms. Symptoms of anaemia, thrombocytopenia or leukopenia occur frequently in the context of a transfusion. In addition, the transfusion of red blood cells per se already effectively prevents a possible emergence of an aplastic crisis caused by B19V. B19V IgG-positive products which are occasionally administered as adjuvant treatment may contribute to the suppression of the infection or the clinical symptoms.

In a retrospective study from the USA [61], 24 cases were identified from a total of 12,529 tested blood donors, in which B19V DNA-positive blood components had been administered to susceptible recipients. None of the donations had high viral burden (<10⁶ IU B19V DNA/ml), and they all simultaneously contained virus-specific Ig. In all 24 cases, no infection (seroconversion) occurred in the recipients. In a German study, no complications were observed after the administration of low viraemic blood components [62].

Transmission by pooled plasma has been described. This plasma was manufactured by means of the solvent/detergent method (SD plasma) [63, 64, 65]. Pooled plasma contains relatively constant B19V-specific antibody levels of approximately 40 IU B19V antibodies/ml [66]. However, such levels are not sufficient to neutralise high-titre virus contaminations entirely. In the study conducted in the USA, transmission cases were observed with SD plasma containing more than 10⁷ genomes/ml, whereas plasma containing 10³⁵ genomes/ml did not show any transmissions [63, 65]. As a response to these reports, the FDA and the European Pharmacopoeia now require a limit of not more than 10⁴ IU B19V DNA for the SD plasma.

A higher prevalence of B19V antibodies was reported in young haemophilia A patients treated on a long-term basis with coagulation factor concentrate [67, 68]. These results indicate that B19V was also transmitted by virus-inactivated factor VIII (FVIII) concentrates. Transmissions have unambiguously been documented in recipients of coagulation factors (FVIII, factor IX (FIX), prothrombin / proconvertin / Stuart factor / antihaemophilic factor B (PPSB)) and fibrin sealant [69]. Transmission of B19V by coagulation factors was documented by means of DNA sequence analysis [70, 71]. There are few reports in which transmission of B19V by immunoglobulins was suspected. However, clear evidence for the transmission chain was never provided in these reports.

3.2 Immune Status (Resistance, Existing Immunity, Immune Response, Age, Exogenous Factors)

The formation of IgG antibodies upon a B19V infection results in immunity. IgG antibodies directed against linear epitopes of VP1 and conformational epitopes of VP2 show neutralising activity [72, 73].

Antibodies from the convalescence phase of an infection with genotype 1 display cross-neutralising activity against the other genotypes [36, 74]. However, a study on liver biopsy probes indicated that a patient can be infected with several genotypes during a life-time [75] since B19V DNA from two genotypes was detected in the liver.

Antibody formation causes a reduction of virus DNA concentration in the blood. However, low DNA titres can persist in the blood for 6–12 months. Kerr et al. [76] have described two cases in which B19V DNA remained detectable over a long period of 57 months despite the presence of IgG antibodies. These patients showed no clinical symptoms. These results challenge the existing view that virus persistence and virus reactivation can only occur in the event of impaired immunity. Studies from the past few years [41, 46] show that low virus DNA levels and the concomitant presence of antibodies in the blood of asymptomatic donors can persist for 6–12 months. Using a sensitive method low B19V DNA levels were detected even in up to 1% of all blood donors [77]. This indicates that B19V is not always completely eliminated after the end of primary infection.

The question of whether a patient with B19V-specific IgG antibodies may be protected against an infection if he receives an i.v. administration of a possibly contaminated product has not yet been systematically investigated. The fact that a B19V infection is reported only very rarely following administration of coagulation products, however, indicates that acquired immunity should be sufficient to protect the patient from a new or symptomatic infection. In a retrospective study [61], 2 donations with high viral burden (>10⁷ geq/ml) were identified, which had been administered to non-susceptible (IgG-positive) recipients. IgG levels in the patient increased four-fold. It was not clear to what extent the virus DNA transiently detected in the recipient's blood reflected only the inoculum or temporarily replicating virus.

3.3 Severity and Course of the Disease

For details on the clinical course of the B19V infection, please refer to section 1.2.

When taking a look at the nearly 100% infection rates in young children who were treated with coagulation factors [78] and the rare case reports on transmissions, it appears that blood-associated infection of immunocompetent children is mostly asymptomatic or associated with unspecific symptoms (fever, flue-like infection) not recognized as B19V infection.

Although serious symptoms of B19V infection are associated with existing underlying diseases or iatrogenic immunosuppression, there are few case reports of an infection of immunocompetent haemophilic patients with severe, partly life-threatening symptoms [79, 80]. table 1 provides a summary of B19V transmission cases after administration of blood products. However, not in all these cases B19V transmission by blood products has clearly been identified by means of DNA sequencing.

Table 1.

Cases of plasma product associated B19V infections

Product	Virus inactivation method	Patient, sex, age in years	Underlying disease	Symptoms	Reference
FIX	dry heat	f, 39	haemophilia B (transmitter)	exanthema, arthralgia, 19 days following i.v. drip	101
		m, 11	haemophilia B	exanthema, arthralgia, 35 days following i.v. drip
		m, 20	haemophilia B	exanthema, fever, 12 days following i.v. drip
FVIII	SD	m, 11	haemophilia A	hypoplastic anaemia	102
FVIII	dry heat	m, 33	mild haemophilia A, non-steroidal immunosuppression, sulfasalazin	hepatomegaly, pancytopenia, high fever, exanthema	79
FVIII	SD or pasteurisation	m, 11	mild haemophilia A, steroidal immunosuppression	fever 12 days following i.v. drip, somnolence, confusion, pancytopenia, liver cystolysis	80
FIX	SD	f, 31	haemophilia B (transmitter)	fever, exanthema, arthralgia, lymphadenopathy 21 days following i.v. drip	103
FVIII	SD	m, 7	mild haemophilia A	aplastic crisis 16 days following i.v. drip	104
Fibrin sealant	dry heat	f, 3 cases (35–42)	cervical carcinoma, ovarial carcinoma, myoma	fever, leukopenia, erythroblastopenia, 6–11 days following administration	105
Plasma	SD	f, 36	myasthenia gravis	exanthema, arthralgia, 14 days following i.v. drip	64
Fibrin sealant	pasteurisation	m, 4 cases	lung cancer	reticulocytopenia	106
		f, 2 cases
		>20
FVIII	dry heat	m, 5	haemophilia A	asymptomatic	70
FEIBA	vapor heat	m, 1	haemophilia A	asymptomatic	70
FVIII	SD	m, 47	mild haemophilia A	fever, arthralgia, 7 days following i.v. drip	71

In summary, clinical symptoms of blood-associated infections are equivalent to those of natural infections. There seems to be an infection risk for seronegative patients who are treated with plasmatic coagulation factors for the first time, since these patients are still susceptible to B19V infection.

An epidemiological study on more than 700 haemophilia patients [81] showed a certain association of B19V-seropositive status with reduced mobility of the joints. However, a causative link to B19V infection is not clear.

3.4 Therapy and Prophylaxis

Vaccine

There are developments of a vaccine. Empty virus particles produced using recombinant DNA technology are used as immunogens. Trials with experimental vaccines indicate that the ratio of the proteins VP1 to VP2 has an essential influence on the formation of neutralising antibodies in animal experiments. An immunogen containing ≥25% VP1 induced neutralising antibodies [82]. First clinical trials have been conducted and seemed promising [83]. Further developments of a vaccine, however, are currently not being pushed forward by the vaccine manufacturers. A vaccine will thus not be available in the foreseeable future.

Passive Immunisation and Treatment with Antibodies

In principle, passive immunisation with antibody products seems possible, but has not been investigated yet. Because of the high B19V infection rate, normal immunoglobulin preparations from large plasma pools generally contain neutralising antibodies. A considerable number of case reports have been published in which an acute or chronic B19V infection could be treated successfully. However, there are currently no immunoglobulin products which have an indication for treatment of acute or persisting B19V infection included in their marketing authorisation.

3.5 Transmissibility

B19 infections are usually transmitted by saliva or droplet infection with close personal contact. The frequency of transmissions by blood donations and blood products is described in 3.1.

3.6 Frequency of Administration, Type and Amount of Blood Products

B19V contamination of cellular blood components is possible.

Contamination of plasma derivatives, especially coagulation products (FVIII, FIX, PPSB) and fibrin sealant with B19V cannot be ruled out. Since coagulation factors have to be administered to patients with haemophilia A or B on a regular basis, this patient group is exposed to a particular risk. Therefore, haemophilia patients who were treated with high doses of plasmatic coagulation factors in the past have an increased risk of B19V infection [81].

4 Blood Products

4.1 Infectious Load of the Starting Material and Test Methods

Considering the fact that an individual high-titre donation can contain up to 10¹³ geq/ml, the contamination of a plasma pool obtained from 10,000 donations with only one high-titre donation will lead a burden of 10⁸ geq/ml in the plasma pool. Investigations of plasma pools in which the starting material was not tested for B19V DNA indeed showed that 60% of all plasma pools contained B19V DNA, and 35% of these pools were highly contaminated with 10⁶-10⁸ genomes/ml [47, 66].

Such high virus contaminations are not cleared using conventional protein purification methods as used for plasma products. Contaminations with B19V were shown by B19V DNA detection in all classes of plasma products (coagulation factors, immunoglobulins, albumins, etc.) [47, 66, 84]. The highest contamination rates were found in coagulation products: 50–90% of the product batches tested were DNA-positive with maximum values of up to 10⁷ geq/ml, while the maximum contaminations for immunoglobulins (up to 10⁴ geq/ml) and albumin (up to 10⁴ geq/ml) were lower because of the additional purification steps [66, 84].

In response to the high contamination rates of plasma pools and final products, many manufacturers have introduced a so-called 'high-titre screening' on a voluntary basis. In this screening procedure, individual high-titre donations are identified by mini-pool testing and sorted out by means of a suitable test algorithm, thereby limiting the maximum load of plasma pools to 10³-10⁴ geq/ml. The Plasma Protein Therapeutics Association (PPTA) suggested a limit of not more than 10⁵ genomes/ml. The high-titre screening seems to be suitable for reducing contamination rates of the plasma pools considerably, with maximum concentrations of not more than 10⁴ genomes/ml in the plasma pool. The reduction of contamination in the starting materials consequently results in a reduction of B19V DNA concentration in the final products even though not all blood products could be tested entirely DNA-negative after the introduction of high-titre screening [47, 85] (Blümel, personal communication). The extent of residual infectious risk from such plasma products remains to be investigated by careful surveillance.

A recently established in vitro neutralisation test showed that the B19V-specific antibodies normally present in plasma pools have an effectively neutralising activity [86].

4.2 Methods for Removal and Inactivation of the Infectious Agent

Various methods of pathogen inactivation in blood components are currently introduced in blood components. Methods using psoralen derivatives (amotosalen + UV irradiation) show effective inactivation of B19V [87], but not of other non-enveloped viruses such as enteroviruses. Effective inactivation of the porcine parvovirus was shown with riboflavin [88] suggesting that B19V could be also inactivated. It was shown that methylene blue/light treatment has no or little effect with regard to inactivation of animal parvovirus, however, marked damage to B19V DNA can be identified after such treatment.

In summary, methods are available for effective inactivation of B19V in blood components. However, it remains uncertain whether the inactivation capacity is sufficient to inactivate entirely high virus loads which can occur in a single donation (10¹³ geq/ml) since the technical limitations of B19V infectivity assays only allow quantification of the inactivation capacity to values up to 5–6 logs. The inactivation capacity can depend on specific conditions of use, which is why a product-specific test is advisable in any case.

4.3 Feasibility and Validation of Procedures for Removal/Inactivation of the Infectious Agent

Up to now, few studies with B19V with high informational value are available. The removal or inactivation of B19V in the course of the production of plasma derivatives cannot be tested directly using B19V, since cultivation of the virus presents major difficulties. Animal parvoviruses are therefore widely used for such tests. The guideline (CPMP/268/95 'Virus Validation Studies: the Design, Contribution and Interpretation of Studies Validating the Inactivation and Removal of Viruses') lists all possible model viruses used for this purpose: canine, murine, bovine, and porcine parvovirus. Antibodies against bovine parvovirus seem to be present in human serum, or antibodies with cross-reactivity between the human and the bovine parvovirus, since bovine parvovirus can be inactivated in human plasma. Therefore, plasma or the appropriate intermediate product must first be tested for suitability if bovine parvovirus is used.

A small number of cell culture systems has been recently developed allowing the direct examination of B19V inactivation. It has been shown that B19V can rapidly be inactivated during pasteurisation of albumin (at 60 °C for 10 h), whereas animal parvoviruses largely survive this procedure unharmed [89, 90]. In the manufacture of some coagulation factor products and immunoglobulin products, pasteurisation at 60 °C for 10 h was also found effective for inactivation of B19V (Gröner et al., personal communication). However, stabilisers may cause a delay in pasteurisation kinetics, and B19V inactivation is not effective in specific cases (Gröner et al., personal communication) [91]. The B19V capsid [70] is destroyed at heating, whereas the B19V DNA remains undamaged and is expelled from opened capsids [92].

Other heat methods seem to inactivate B19V at least partly. In the so-called steam method (vapor heat), the intermediate product is treated at a reduced residual humidity with hot vapor at 60 °C and/or up to 80 °C for 1–10 h. In some cases, this method also showed effective B19V inactivation [93]. However, due to the product-specific conditions and effects, inactivation may not be effective in each case and a product-specific examination of the manufacturing step is required.

In the so-called dry-heat treatment (100 °C for 30 min or 80 °C for 72 h), partial inactivation of B19V has been described. However, inactivation very strongly depends on the residual humidity which can fluctuate between 0.3 and 2%. While B19V was inactivated considerably faster than porcine parvovirus at a high residual humidity (1–2%), both viruses were equally stable at a low residual humidity [94, 95, 96].

Low pH values also lead to effective inactivation of B19V when other conditions such as a suitable temperature and time for inactivation were met [97]. Again, B19V proved to be less stable than animal parvovirus.

Another effective inactivation method described for B19V is irradiation, especially UV-C irradiation [98, 99]. Plasma products manufactured with this inactivation method are still in the development stage.

B19V can also be removed using special virus filters (nanofiltration). However, since parvovirus particles are very small, the applicability of such filters can be limited. Filtration of high-molecular-weight coagulation factors such as fibrinogen or von Willebrand factor containing complexes currently is difficult. Parvovirus-removing filters can be used for small molecules such as FIX, antithrombin or even IgG and PPSB. However, the effectiveness of these methods also depends on the exact product conditions. In spiking experiments, residual amounts of virus were sometimes found in the filtrate. In some cases, two filters had to be arranged in sequence in order to obtain effective virus reduction of more than 4 log steps.

Highly contaminated pools (≥10⁷ genomes/ml) were always identified as starting material in the few examined transmission cases where the viral loads in the plasma pools were retrospectively quantified [70, 71]. In one case of transmission by a product which had been treated with SD for virus inactivation, the inoculated amount of genome equivalents was identified as 10⁴ genomes [71]. Even heat inactivation methods (dry heat, vapor heat method) did not suffice to eliminate the infection risk by plasma products form high-burden plasma pools entirely [70]. Recently the FDA published a recommendation to limit the B19V DNA content in the plasma pools to 10⁴ IU/ml [100].

B19V DNA has been detected in immunoglobulin products, IVIG, and IMIG, although DNA concentrations were markedly lower than in coagulation factors. There seems to be no infection risk from such normal immunoglobulin preparations due to the high content of B19V-neutralising antibodies. In addition, immunoglobulin products have been used successfully for treatment of acute B19V infection.

Albumin is regarded safe with respect to B19V infections. Even though low concentrations of B19V DNA have been found in albumin, B19V is effectively inactivated by pasteurisation as required by the European Pharmacopoeia [90]. B19V DNA was also identified in recombinant coagulation factors which contained human albumin as stabiliser [78]. Pasteurisation does not eliminate the virus DNA, but the detection of B19V DNA in albumin does not reflect infectious virus.

5 Assessment

Infections with B19V are usually harmless. Serious cases rarely occur and are in most cases restricted to immunosuppressed patients with impaired erythropoiesis and to foetuses. B19V is endemic in Germany. However, not all blood donors have been infected during childhood. Therefore, there is a risk of viraemic donations from blood donors undergoing primary infection. In spite of this, B19V infections following the administration of blood components have not been reported in Germany. Low viraemic donations containing virus-specific immunoglobulin do not seem to be infectious, whereas highly viraemic donations are considered to involve a risk of infection.

If in specific cases (e.g. B19V antibody-negative pregnant women in the first and second trimenon, HIV-infected individuals or patients who underwent high-dose chemotherapy following stem cell transplantation) it is considered necessary to avoid the risks associated with B19V transmission, there is the possibility to use blood components which have been tested negative by means of NAT. In order to obtain more data on B19V transmissions, physicians shall still be made aware of such risks and are requested to observe their patients and to report any suspected cases of a B19V transmission by blood components or plasma derivatives.

For plasma products, in particular coagulation factor products and fibrin sealants, there is an infection risk if highly viraemic donations (>10⁶ IU/ml/single donation) are not excluded from production and the capacity of methods for virus inactivation/removal is limited. High-titre screening significantly reduces the starting load of plasma pools, and, consequently reduces the load in finished products. Therefore, high-titre screening should be consistently performed for all plasma pools for fractionation. Such a screening and the implementation of an effective B19V-reducing manufacturing step can widely reduce the infection risk caused by coagulation factors, even if a theoretical residual risk cannot always be safely ruled out.

Areas Requiring Further Research

Research is required with regard to the development of effective methods of B19V elimination/inactivation and the characterisation of these methods by model viruses. Since the model viruses do not always exactly reflect the behaviour of B19V, an improvement of the currently available cell culture systems for B19V is of interest, in order to study virus inactivation in a product-specific context and to obtain information on the infection risk of plasma products as exactly as possible.

The further development of WHO standards with regard to variant genotypes 2 and 3 will provide the basis for a reliable rejection of highly contaminated blood and blood plasma donations by means of NAT. Further studies on the distribution and the clinical relevance of these genotypes are desirable. More research is also necessary regarding the role of immunoglobulins in B19V safety.