Did intravenous immune globulin really transmit parvovirus b19?

Hayakawa et al (2002) did not present some important clinical data for their patient. According to one of the authors, a serum sample obtained on 11 January was positive for B19 IgG antibody, prior to the administration of i.v.Ig from 27 January to 31 January. Additionally, no B19 DNA was detected in serum samples drawn on 14 February and 8 March, and no B19 IgM was detected in samples obtained on 16 February and 8 March. Thus, the clinical course of the patient was not simple. This information should be considered during any examination of the aetiology of the patient's disease, especially the assessment of a possible causal relationship between i.v.Ig infusion and the patient's clinical course.

The authors noted that the two implicated i.v.Ig lots were positive for B19 DNA. However, our in-house B19 polymerase chain reaction (PCR) did not detect B19 DNA in these lots. Our testing was conducted using the PCR enzyme-linked immunosorbent assay system (digoxigenin labelling and detection; Roche diagnostics). Total DNA was extracted from 100-µl samples of each lot. Primer B1 (3187–3206: CAAAAGCATGTGGAGTGAGG) and primer B2 (3558–3539: GTGCTGTCAGTAACCTGTAC) were used for the amplification. The PCR products were analysed by solution hybridization to biotin-labelled probe B (3310–3339: TAGCTGCCACAATGCCAGTGGAAAGGAGGC). The detection limit of our PCR is 50 copies/ml. The methods and performance of the PCR used by Hayakawa et al (2002) should be disclosed because the authors' conclusion that i.v.Ig-transmitted B19 was based on the detection of B19 DNA from the implicated lots.

There was no comparison of the B19 DNA sequences from the patient's serum and the DNA sequences in the implicated i.v.Ig lots.

Transmission of B19 through solvent/detergent (S/D)-treated pooled plasma has been reported (Davenport et al, 2000). Eighteen of the 19 healthy subjects infused with three lots that had B19 NAT (nucleic acid amplification technique) titres > 10⁷ GE (genome equivalents)/ml became positive for B19 IgG. None of the 81 subjects infused with S/D plasma lots with B19 NAT titres < 10⁴ GE/ml became seropositive for B19. These results suggest that there is a relationship between the B19 DNA content and the risk of B19 transmission in plasma products that contain immunoglobulin, and that a relatively high dose of B19 is required for infection in IgG-containing products. The amount of B19 required for infection would have been readily detected by our PCR assay if it had been present in these i.v.Ig lots.

The manufacturing process of our i.v.Ig product includes a pasteurization step which inactivates 2·1 logs of canine parvovirus. It has recently been reported that B19 was rapidly inactivated by pasteurization of an albumin solution while an animal parvovirus survived (Blümel et al, 2002). These findings cannot be directly extrapolated to our i.v.Ig product. However, in our preliminary experiment using KU812 cells, no B19 infectivity was detected after 2 h of heat treatment at 60°C in a 33% sorbitol solution (without immunoglobulin) used as a stabilizer during pasteurization of our i.v.Ig preparation.