Dear Editor,Juhl et al. in their recent article "Look-back study on recipients of Parvovirus B19 (B19V) DNA-positive blood components" presented identical partial sequencing data of B19V NS1 gene region and NS1-VP1 region from two donor-recipient pairs (D1/R1, D2/R2) to support their claim on transfusion transmission of B19V [1]. While the epidemiological data correlate with the investigators' conclusion, the same data presented can be alternatively interpreted and different conclusions can be drawn.In D1/R1, B19V seroconversion in R1 occurred after blood product transfusion. The two regions interrogated in the B19V DNA products from D1 and R1 were identical. On the other hand, the sequences from the German D1 and R1 differ from the reference strain from a geographically distinct region (Netherlands) by only two nucleotide positions, suggesting the circulating viruses in that part of Europe are highly similar. Without sequence data from other circulating strains unrelated to this potential transmission event, the sequence data from D1 and R1 can potentially be interpreted as either that D1 and R1 are independently infected by two highly similar strains not differentiable by the partial sequences presented, or a single highly prevalent circulating strain. Nosocomial infections from infected staff or patient have been reported previously in the literature [2,3], and Juhl et al. did not present data to exclude potential concurrent B19V infection within the ward in which R1 was a patient.In D2/R2, R2 did not show seroconversion after the alleged transfusion transmitted infection. The investigators proposed two possible reasons: antibody adsorption and permanent protein loss. But alternatively R2 could have been previously infected and the negative serology pre-transfusion could be explained by the exact same proposed reasons. The observed viral load post transfusion could be potentially explained by silent reactivation in a very sick patient. Furthermore, the absence of sequence data unrelated to the event and data excluding concurrent B19V infection within the ward in which R2 was a patient, again mean that independent infection events have not been excluded, and a confirmed transmission between D2/R2 cannot be made.To infer transmission using molecular data, it is necessary to have sequences from unrelated strains to demonstrate the background of variations upon which the relationship among implicated strains is inferred. In highly conserved DNA viruses such as Parvovirus, this could be very difficult, especially when only short partial sequences are interrogated.
References