We report a case of HIV seroconversion in a health-care worker, due to successive contaminations (i) from a blood donor to a recipient and (ii) from the recipient to the health-care worker. Unexpectedly, both infections occurred during the HIV serological window period. A nurse sustained a needle-stick injury with a large gauge hollow-bore needle on 24 January 1996. On the same day, she and the source patient were tested for anti-HIV antibodies by Genelavia Mixte (Sanofi Diagnostics Pasteur, Marnes-la-Coquette, France) and Axsym HIV (Abbott laboratories, North Chicago, Illinois, USA) assays. The results were negative for both subjects. According to legal procedure, an HIV serology check was performed on the nurse 3 months after exposure. The presence of anti-HIV antibodies was observed by enzyme-linked immunosorbent assay and was confirmed by HIV-1 Western blot. Because the nurse reported no risk factors for HIV infection, HIV serology was again performed on a frozen sample from the source patient collected on the day of the needle-stick injury. This serum was found to be weakly positive (three times above cut-off) by Genscreen (Sanofi Diagnostics Pasteur), one of the most sensitive anti-HIV screening tests available [1]. p24 antigen was also positive with a high concentration (> 640 pg/ml). These results indicated that the patient was undergoing HIV seroconversion at the time of the nurse's exposure. The source patient, a 89 year-old woman who was hospitalized for a digestive bleeding, received six red cell units and two platelet concentrates between 6 December 1995 and the 28 January 1996. She died on the 2 February 1996. Sixteen out of the 17 implicated donors were traced, retested and found to be anti-HIV negative. The 17th donor did not answer to the inquiry, but HIV serology performed on 15 February 1996 in a hospital laboratory revealed the presence of anti-HIV-1 antibodies. His implicated donation was given on 24 November 1995 and was transfused to the patient on 6 December 1995. A frozen aliquot of this donation was investigated and was found to be anti-HIV-negative by Genscreen, p24 antigen-negative by Vironostika (Organon Teknika, Oss, The Netherlands), but was HIV RNA-positive by PCR (Amplicor, Roche, Branchburg, New Jersey, USA) and an in-house nested PCR performed in pol region. The results from the three subjects are summarized in Table 1.Table 1: . Data from three subjects implicated in two successive HIV contaminations.We describe an exceptional case of transfusion-transmitted HIV infection by a seronegative donor who was undergoing seroconversion and was in fact HIV RNA-positive. According to some predictions [2], this case highlights the capacity of nucleic acid amplification assays to eliminate the infectious but seronegative donors by reducing the length of the pre-seroconversion window period. Nevertheless, the cost and the complexity of such a measure in mass screening donations [3] must be balanced with the risk of transmitting HIV infection by transfusion of screened blood, which today is very small (1 per 1 million donations in France [4]). Because the precise date of the contamination is often unknown for HIV-infected patients, the length of window period is generally estimated by mathematical models [2]. In this transfusion-transmitted HIV infection, the interval from exposure to seroconversion was 49 days. The nurse was also infected by a subject in preserocon-version stage. The mean risk of transmission of HIV after percutaneous exposure has been estimated to be 0.3% with a higher rate when the HIV load in the source patient's blood is high [5]. Because the p24 antigen level was high and the injury was deep, it can be assumed that this was the case for this patient. On the other hand, if the source patient had been tested for p24 antigen in addition to anti-HIV antibodies, we can presume that the nurse would have been treated sooner by chemoprophylaxis. Thus, one of the recommendations for the improvement of the management of HIV post-exposure could be the implementation of a p24 antigen test in the source patient at the time of health-care worker exposure. Acknowledgement The authors thank F. Simon from the Virology Unit of the Hôpital Bichat, Paris, for performing nucleic acid analysis.
Read full abstract
Oct 1, 2005
Darin J Saltzman 
Open Access
