HIV-1 and HCV Infections among Antibody-Negative Blood Donors
http://www.100md.com
《新英格兰医药杂志》
To the Editor: As reported by Stramer et al. (Aug. 19 issue),1 testing for human immunodeficiency virus type 1 (HIV-1) and hepatitis C virus (HCV) by nucleic acid amplification with the use of minipools has markedly reduced the transmission of HIV-1 and HCV through blood products. Testing of individual units is now being considered, but such a procedure would still not completely abolish transmission.2
Earlier this year in Croatia, we studied the case of a regular blood donor who transmitted HIV-1 to a recipient of platelets and a recipient of plasma. The transmission was detected by a look-back investigation performed because of HIV-1 seroconversion of the blood donor by the time of the next donation. The reverse transcriptase and protease sequences of the donor and the two recipients had a bootstrap value of 100 of 100. The implicated donation was negative for anti–HIV-1 antibody (Ortho HIV-1/HIV-2 Ab Capture) and was not examined by minipool testing; however, plasma collected at the time of the donation, preserved at –40°C for nine months, tested negative with the use of the Roche Amplicor HIV-1 Monitor UltraSensitive Method (version 1.5) with a detection limit of 50 HIV-1 RNA copies per milliliter. Thus, complete elimination of risk is not achievable, and as underlined by Goodman in an editorial3 accompanying the report by Stramer et al., implementation of additional measures should be subject to consideration of cost-effectiveness in a global context.
Josip Begovac, M.D.
University Hospital for Infectious Diseases
Zagreb 10000, Croatia
josip.begovac@zg.htnet.hr
Ivanka Mihaljevic, M.Sc.
Croatian Institute of Transfusion Medicine
Zagreb 10000, Croatia
Luc Perrin, M.D.
Geneva University Hospital
Geneva 1211, Switzerland
Dr. Perrin has been involved in the evaluation of assays for the detection of nucleic acids by the polymerase chain reaction for Roche Diagnostics (HIV-1 and HCV) and Abbott Diagnostics (HCV).
References
Stramer SL, Glynn SA, Kleinman SH, et al. Detection of HIV-1 and HCV infections among antibody-negative blood donors by nucleic acid-amplification testing. N Engl J Med 2004;351:760-768.
Fiebig EW, Wright DJ, Rawal BD, et al. Dynamics of HIV viremia and antibody seroconversion in plasma donors: implications for diagnosis and staging of primary HIV infection. AIDS 2003;17:1871-1879.
Goodman JL. The safety and availability of blood and tissues -- progress and challenges. N Engl J Med 2004;351:819-822.
To the Editor: Stramer et al. present the results of nucleic acid–amplification testing of donated blood for HIV-1 and HCV in the United States over a three-year period. In France, nucleic acid–amplification testing for HIV-1 and HCV was implemented for all blood donations in July 2001 with the use of the Chiron TMA HIV-1/HCV Assay or the Roche Cobas AmpliScreen test (in 8 and 24 minipools, respectively). After 30 months of testing, 2 of 6,136,292 units screened were negative for anti-HIV antibody and positive for HIV on nucleic acid–amplification testing; both had been collected before seroconversion. Four donations were negative for anti-HCV antibody and positive for HCV on nucleic acid–amplification testing, but one was discarded because the alanine aminotransferase level was elevated. Of the three donations that were positive for HCV on nucleic acid–amplification testing, one was a long-term, immunologically silent infection; one was in a preseroconversion phase; and one was not investigated.
As it has in the United States, nucleic acid–amplification testing has allowed us to discard HIV-1–seronegative and HCV-seronegative blood donations that are infected; in France, our yield is the same for HIV-1 but six times lower for HCV. This suggests that in France, the HCV incidence rate is lower in blood donors, and probably in the general population, than it is in the United States.1
Syria Laperche, M.D.
Institut National de la Transfusion Sanguine
75015 Paris, France
slaperche@ints.fr
Josiane Pillonel, Ph.D.
Institut de Veille Sanitaire
94415 Saint-Maurice, France
Patrick Herve, M.D., Ph.D.
Etablissement Fran?ais du Sang
75015 Paris, France
References
Pillonel J, Laperche S, Saura C, Desenclos JC, Couroucé AM. Trends in residual risk of transfusion-transmitted viral infections in France between 1992 and 2000. Transfusion 2002;42:980-988.
To the Editor: The articles by Stramer et al. and Zou et al.1 provide a comprehensive review of blood and tissue safety with respect to the transmission of virus to recipients. Zou et al. cite evidence that although safety is improved, the examination of single blood units by polymerase-chain-reaction (PCR) nucleic acid–amplification testing (for HIV, HCV, and hepatitis B virus [HBV]), instead of the current pool testing, would result in a cost increase of at least $150 per donor.
To address this issue, my colleagues and I at the New York Blood Center have developed a semiautomated nucleic acid extraction and PCR procedure that promises far greater economy and increased sensitivity and specificity. An Investigational New Drug application for evaluation of this procedure in West Nile virus testing has been submitted to the Food and Drug Administration (FDA). A similar application for evaluation of multiplex testing for other blood-borne viruses will follow. The anticipated cost is in the range of $2 to $5 per virus per unit tested. Furthermore, because of their very low viral loads, testing of single units is most beneficial for West Nile virus.
Alfred M. Prince, M.D.
New York Blood Center
New York, NY 10021
aprince@nybloodcenter.org
References
Zou S, Dodd RY, Stramer SL, Strong DM. Probability of viremia with HBV, HCV, HIV, and HTLV among tissue donors in the United States. N Engl J Med 2004;351:751-759.
To the Editor: Zou et al. and Goodman provide important insights into tissue safety. However, cost–benefit analyses should take into account important differences between tissue donors and blood donors and between tissue recipients and blood recipients. First, screening of tissue donors for high-risk behavior is limited by history taking through next of kin — hence, the higher seroprevalences for blood-borne viruses. Second, in contrast to blood donors, one infected tissue donor could potentially affect more than 100 tissue recipients. Third, the risk tolerance of tissue recipients is lower than that of blood recipients: most tissue is used for procedures that enhance the quality of life, not for lifesaving procedures. Fourth, the life expectancy of tissue recipients is probably greater than that of blood recipients,1 providing ample time for the development of complications from blood-borne viruses. Therefore, informed consent should be required from tissue recipients before the implantation of potentially contaminated tissue.
Neither Zou et al. nor Goodman discusses the most important approach to the enhancement of tissue safety — namely, viral inactivation and sporicidal methods that do not adversely affect the structural and functional integrity of tissue. Such methods would provide an additional safety net against known pathogens (e.g., clostridium2 and HCV3) and emerging pathogens for which there are severe limitations in screening methods.
Marion A. Kainer, M.B., B.S., M.P.H.
Tennessee Department of Health
Nashville, TN 37247
marion.kainer@state.tn.us
William R. Jarvis, M.D.
135 Dune Lane
Hilton Head Island, SC 29928
References
Kleinman S, Marshall D, AuBuchon J, Patton M. Survival after transfusion as assessed in a large multistate US cohort. Transfusion 2004;44:386-390.
Kainer MA, Linden JV, Whaley DN, et al. Clostridium infections associated with musculoskeletal-tissue allografts. N Engl J Med 2004;350:2564-2572.
Hepatitis C virus transmission from an antibody-negative organ and tissue donor -- United States, 2000-2002. MMWR Morb Mortal Wkly Rep 2003;52:273-4, 276.
Dr. Stramer and colleagues reply: We commend Dr. Prince for his efforts to develop nucleic acid tests for blood-donor screening. It is true that elsewhere we have cited a figure of about $150 as the cost of individual testing of tissue-donor samples (the cost of which has since dropped to $65 to $110); we based this figure on the contracted costs of testing single samples — an approach that may be appropriate for samples from tissue donors but that clearly is not applicable to the high-throughput testing used for blood donations, the cost of which has been estimated to be $15 to $20 per sample.1
We acknowledge that it may be possible to develop reagent systems at a cost of a few dollars, but much of the cost of commercial reagents lies in license fees for intellectual property and in the overhead inherent in complying with the stringent requirements of the FDA-regulated Good Manufacturing Practices. The costs of nucleic acid–amplification testing include not only cost recovery for the developed technology and direct costs for the manufacture of reagents and instrumentation, but also the system costs required to ensure lot-to-lot consistency and reagent stability, provide training and customer service, and develop product enhancements. These are daunting tasks in supplying the U.S. and worldwide screening markets and historically have been successfully accomplished only by the most sophisticated diagnostic corporations. Even some of them have had difficulty complying with all of the expectations of the FDA.
In response to Dr. Begovac and colleagues: the screening tests we used for nucleic acid testing have a sensitivity of less than 14 copies per milliliter, in comparison with available diagnostic tests with a sensitivity of 50 HIV-1 RNA copies per milliliter. Nonetheless, look-back and animal-model data for HIV, HCV, and West Nile virus support the conclusion that nucleic acid testing even of individual donations will not interdict 100 percent of units during the infectious window period.
Dr. Laperche and colleagues' comments demonstrate that the yield of nucleic acid tests for each agent will vary according to local epidemiologic conditions. In areas with a high incidence of HIV infection, such as South Africa and India, HIV-1 RNA detection would probably be very useful but not affordable, whereas in other regions nucleic acid–amplification testing for HCV or HBV could be of great benefit. For agents such as West Nile virus, seasonal and weather-related factors, as well as local mosquito populations and control policies, will influence yield and effectiveness.
Susan L. Stramer, Ph.D.
American Red Cross
Gaithersburg, MD 20878
stramers@usa.redcross.org
Roger Y. Dodd, Ph.D.
American Red Cross
Rockville, MD 20855
Michael P. Busch, M.D., Ph.D.
Blood Systems Research Institute
San Francisco, CA 94188
References
Jackson BR, Busch MP, Stramer SL, AuBuchon JP. The cost-effectiveness of NAT for HIV, HCV, and HBV in whole blood donations. Transfusion 2003;43:721-729.
Dr. Goodman replies: As Begovac and colleagues note, even sensitive methods, such as nucleic acid testing of small pools of donor specimens, cannot eliminate the risk of viral transmission completely.1,2 In fact, it is primarily units of blood containing extremely low amounts of HIV or HCV that account for the current residual risk, which is approximately 1 in 2 million blood units. This residual risk can be addressed in part through nucleic acid–amplification testing of individual donor specimens. However, without technological improvements, such approaches markedly increase the number of tests needed, and sufficient laboratory capacity is not currently available. Technological improvements could improve cost-effectiveness and sensitivity. Although such advances should further reduce risks, testing can never prevent all transmissions.3,4 To take an extreme example: if a single virus particle can transmit disease, one would potentially have to test the entire product to find it.
As Kainer and Jarvis point out and as I point out in my editorial, certain characteristics of tissues make risk reduction challenging and important, including the possibility that a donation could affect 100 recipients. Although it is correct that blood donors can be screened more accurately for behavioral risks than can most tissue donors, tissue donors still have prevalence rates of viral markers below those of the general population, as reported by Zou et al. As Kainer and Jarvis note, many tissue procedures are life-enhancing, and recipients may live for long periods. The latter is also true of many recipients of blood products — for example, children and persons with disorders requiring long-term plasma-based replacement therapies. It is important that both transplanted tissues and blood be safe and, as with other therapies, that the benefits and risks be clearly understood and communicated.
In my editorial, I advocated the development of "more effective ways of removing or inactivating pathogens or achieving sterility while maintaining function." A challenge will be to ensure that such treatments do not create new risks. The structural properties of tissues must be maintained, and processing must not affect tissues in ways that cause harmful inflammatory or immune responses in recipients. Although I am optimistic that such methods can be developed, it remains critical that tissue procurement and manufacturing be carried out with the use of the best available methods to reduce the risks of contamination and disease transmission. The regulations proposed by the FDA include enhanced requirements for validated processes to reduce contamination risks, including risks related to suboptimal processes found in some Centers for Disease Control and Prevention–FDA collaborative investigations of disease transmission.
Jesse L. Goodman, M.D., M.P.H.
Food and Drug Administration
Bethesda, MD 20892
References
Phelps R, Robbins K, Liberti T, et al. Window-period human immunodeficiency viral transmission to two recipients by an adolescent blood donor. Transfusion 2004;44:929-933.
Biswas R, Tabor E, Hsia CC, et al. Comparative sensitivity of HBV NATs and HBsAg assays for detection of acute HBV infection. Transfusion 2003;43:788-798.
Schuttler CG, Caspari G, Jursch CA, Willems WR, Gerlich WH, Schaefer S. Hepatitis C virus transmission by a blood donation negative in nucleic acid amplification tests for viral RNA. Lancet 2000;355:41-42.
Busch MP, Tobler LH, Gerlich WH, Schaefer S, Giachetti C, Smith R. Very low level viremia in HCV infectious unit missed by NAT. Transfusion 2003;43:1173-1174.
Earlier this year in Croatia, we studied the case of a regular blood donor who transmitted HIV-1 to a recipient of platelets and a recipient of plasma. The transmission was detected by a look-back investigation performed because of HIV-1 seroconversion of the blood donor by the time of the next donation. The reverse transcriptase and protease sequences of the donor and the two recipients had a bootstrap value of 100 of 100. The implicated donation was negative for anti–HIV-1 antibody (Ortho HIV-1/HIV-2 Ab Capture) and was not examined by minipool testing; however, plasma collected at the time of the donation, preserved at –40°C for nine months, tested negative with the use of the Roche Amplicor HIV-1 Monitor UltraSensitive Method (version 1.5) with a detection limit of 50 HIV-1 RNA copies per milliliter. Thus, complete elimination of risk is not achievable, and as underlined by Goodman in an editorial3 accompanying the report by Stramer et al., implementation of additional measures should be subject to consideration of cost-effectiveness in a global context.
Josip Begovac, M.D.
University Hospital for Infectious Diseases
Zagreb 10000, Croatia
josip.begovac@zg.htnet.hr
Ivanka Mihaljevic, M.Sc.
Croatian Institute of Transfusion Medicine
Zagreb 10000, Croatia
Luc Perrin, M.D.
Geneva University Hospital
Geneva 1211, Switzerland
Dr. Perrin has been involved in the evaluation of assays for the detection of nucleic acids by the polymerase chain reaction for Roche Diagnostics (HIV-1 and HCV) and Abbott Diagnostics (HCV).
References
Stramer SL, Glynn SA, Kleinman SH, et al. Detection of HIV-1 and HCV infections among antibody-negative blood donors by nucleic acid-amplification testing. N Engl J Med 2004;351:760-768.
Fiebig EW, Wright DJ, Rawal BD, et al. Dynamics of HIV viremia and antibody seroconversion in plasma donors: implications for diagnosis and staging of primary HIV infection. AIDS 2003;17:1871-1879.
Goodman JL. The safety and availability of blood and tissues -- progress and challenges. N Engl J Med 2004;351:819-822.
To the Editor: Stramer et al. present the results of nucleic acid–amplification testing of donated blood for HIV-1 and HCV in the United States over a three-year period. In France, nucleic acid–amplification testing for HIV-1 and HCV was implemented for all blood donations in July 2001 with the use of the Chiron TMA HIV-1/HCV Assay or the Roche Cobas AmpliScreen test (in 8 and 24 minipools, respectively). After 30 months of testing, 2 of 6,136,292 units screened were negative for anti-HIV antibody and positive for HIV on nucleic acid–amplification testing; both had been collected before seroconversion. Four donations were negative for anti-HCV antibody and positive for HCV on nucleic acid–amplification testing, but one was discarded because the alanine aminotransferase level was elevated. Of the three donations that were positive for HCV on nucleic acid–amplification testing, one was a long-term, immunologically silent infection; one was in a preseroconversion phase; and one was not investigated.
As it has in the United States, nucleic acid–amplification testing has allowed us to discard HIV-1–seronegative and HCV-seronegative blood donations that are infected; in France, our yield is the same for HIV-1 but six times lower for HCV. This suggests that in France, the HCV incidence rate is lower in blood donors, and probably in the general population, than it is in the United States.1
Syria Laperche, M.D.
Institut National de la Transfusion Sanguine
75015 Paris, France
slaperche@ints.fr
Josiane Pillonel, Ph.D.
Institut de Veille Sanitaire
94415 Saint-Maurice, France
Patrick Herve, M.D., Ph.D.
Etablissement Fran?ais du Sang
75015 Paris, France
References
Pillonel J, Laperche S, Saura C, Desenclos JC, Couroucé AM. Trends in residual risk of transfusion-transmitted viral infections in France between 1992 and 2000. Transfusion 2002;42:980-988.
To the Editor: The articles by Stramer et al. and Zou et al.1 provide a comprehensive review of blood and tissue safety with respect to the transmission of virus to recipients. Zou et al. cite evidence that although safety is improved, the examination of single blood units by polymerase-chain-reaction (PCR) nucleic acid–amplification testing (for HIV, HCV, and hepatitis B virus [HBV]), instead of the current pool testing, would result in a cost increase of at least $150 per donor.
To address this issue, my colleagues and I at the New York Blood Center have developed a semiautomated nucleic acid extraction and PCR procedure that promises far greater economy and increased sensitivity and specificity. An Investigational New Drug application for evaluation of this procedure in West Nile virus testing has been submitted to the Food and Drug Administration (FDA). A similar application for evaluation of multiplex testing for other blood-borne viruses will follow. The anticipated cost is in the range of $2 to $5 per virus per unit tested. Furthermore, because of their very low viral loads, testing of single units is most beneficial for West Nile virus.
Alfred M. Prince, M.D.
New York Blood Center
New York, NY 10021
aprince@nybloodcenter.org
References
Zou S, Dodd RY, Stramer SL, Strong DM. Probability of viremia with HBV, HCV, HIV, and HTLV among tissue donors in the United States. N Engl J Med 2004;351:751-759.
To the Editor: Zou et al. and Goodman provide important insights into tissue safety. However, cost–benefit analyses should take into account important differences between tissue donors and blood donors and between tissue recipients and blood recipients. First, screening of tissue donors for high-risk behavior is limited by history taking through next of kin — hence, the higher seroprevalences for blood-borne viruses. Second, in contrast to blood donors, one infected tissue donor could potentially affect more than 100 tissue recipients. Third, the risk tolerance of tissue recipients is lower than that of blood recipients: most tissue is used for procedures that enhance the quality of life, not for lifesaving procedures. Fourth, the life expectancy of tissue recipients is probably greater than that of blood recipients,1 providing ample time for the development of complications from blood-borne viruses. Therefore, informed consent should be required from tissue recipients before the implantation of potentially contaminated tissue.
Neither Zou et al. nor Goodman discusses the most important approach to the enhancement of tissue safety — namely, viral inactivation and sporicidal methods that do not adversely affect the structural and functional integrity of tissue. Such methods would provide an additional safety net against known pathogens (e.g., clostridium2 and HCV3) and emerging pathogens for which there are severe limitations in screening methods.
Marion A. Kainer, M.B., B.S., M.P.H.
Tennessee Department of Health
Nashville, TN 37247
marion.kainer@state.tn.us
William R. Jarvis, M.D.
135 Dune Lane
Hilton Head Island, SC 29928
References
Kleinman S, Marshall D, AuBuchon J, Patton M. Survival after transfusion as assessed in a large multistate US cohort. Transfusion 2004;44:386-390.
Kainer MA, Linden JV, Whaley DN, et al. Clostridium infections associated with musculoskeletal-tissue allografts. N Engl J Med 2004;350:2564-2572.
Hepatitis C virus transmission from an antibody-negative organ and tissue donor -- United States, 2000-2002. MMWR Morb Mortal Wkly Rep 2003;52:273-4, 276.
Dr. Stramer and colleagues reply: We commend Dr. Prince for his efforts to develop nucleic acid tests for blood-donor screening. It is true that elsewhere we have cited a figure of about $150 as the cost of individual testing of tissue-donor samples (the cost of which has since dropped to $65 to $110); we based this figure on the contracted costs of testing single samples — an approach that may be appropriate for samples from tissue donors but that clearly is not applicable to the high-throughput testing used for blood donations, the cost of which has been estimated to be $15 to $20 per sample.1
We acknowledge that it may be possible to develop reagent systems at a cost of a few dollars, but much of the cost of commercial reagents lies in license fees for intellectual property and in the overhead inherent in complying with the stringent requirements of the FDA-regulated Good Manufacturing Practices. The costs of nucleic acid–amplification testing include not only cost recovery for the developed technology and direct costs for the manufacture of reagents and instrumentation, but also the system costs required to ensure lot-to-lot consistency and reagent stability, provide training and customer service, and develop product enhancements. These are daunting tasks in supplying the U.S. and worldwide screening markets and historically have been successfully accomplished only by the most sophisticated diagnostic corporations. Even some of them have had difficulty complying with all of the expectations of the FDA.
In response to Dr. Begovac and colleagues: the screening tests we used for nucleic acid testing have a sensitivity of less than 14 copies per milliliter, in comparison with available diagnostic tests with a sensitivity of 50 HIV-1 RNA copies per milliliter. Nonetheless, look-back and animal-model data for HIV, HCV, and West Nile virus support the conclusion that nucleic acid testing even of individual donations will not interdict 100 percent of units during the infectious window period.
Dr. Laperche and colleagues' comments demonstrate that the yield of nucleic acid tests for each agent will vary according to local epidemiologic conditions. In areas with a high incidence of HIV infection, such as South Africa and India, HIV-1 RNA detection would probably be very useful but not affordable, whereas in other regions nucleic acid–amplification testing for HCV or HBV could be of great benefit. For agents such as West Nile virus, seasonal and weather-related factors, as well as local mosquito populations and control policies, will influence yield and effectiveness.
Susan L. Stramer, Ph.D.
American Red Cross
Gaithersburg, MD 20878
stramers@usa.redcross.org
Roger Y. Dodd, Ph.D.
American Red Cross
Rockville, MD 20855
Michael P. Busch, M.D., Ph.D.
Blood Systems Research Institute
San Francisco, CA 94188
References
Jackson BR, Busch MP, Stramer SL, AuBuchon JP. The cost-effectiveness of NAT for HIV, HCV, and HBV in whole blood donations. Transfusion 2003;43:721-729.
Dr. Goodman replies: As Begovac and colleagues note, even sensitive methods, such as nucleic acid testing of small pools of donor specimens, cannot eliminate the risk of viral transmission completely.1,2 In fact, it is primarily units of blood containing extremely low amounts of HIV or HCV that account for the current residual risk, which is approximately 1 in 2 million blood units. This residual risk can be addressed in part through nucleic acid–amplification testing of individual donor specimens. However, without technological improvements, such approaches markedly increase the number of tests needed, and sufficient laboratory capacity is not currently available. Technological improvements could improve cost-effectiveness and sensitivity. Although such advances should further reduce risks, testing can never prevent all transmissions.3,4 To take an extreme example: if a single virus particle can transmit disease, one would potentially have to test the entire product to find it.
As Kainer and Jarvis point out and as I point out in my editorial, certain characteristics of tissues make risk reduction challenging and important, including the possibility that a donation could affect 100 recipients. Although it is correct that blood donors can be screened more accurately for behavioral risks than can most tissue donors, tissue donors still have prevalence rates of viral markers below those of the general population, as reported by Zou et al. As Kainer and Jarvis note, many tissue procedures are life-enhancing, and recipients may live for long periods. The latter is also true of many recipients of blood products — for example, children and persons with disorders requiring long-term plasma-based replacement therapies. It is important that both transplanted tissues and blood be safe and, as with other therapies, that the benefits and risks be clearly understood and communicated.
In my editorial, I advocated the development of "more effective ways of removing or inactivating pathogens or achieving sterility while maintaining function." A challenge will be to ensure that such treatments do not create new risks. The structural properties of tissues must be maintained, and processing must not affect tissues in ways that cause harmful inflammatory or immune responses in recipients. Although I am optimistic that such methods can be developed, it remains critical that tissue procurement and manufacturing be carried out with the use of the best available methods to reduce the risks of contamination and disease transmission. The regulations proposed by the FDA include enhanced requirements for validated processes to reduce contamination risks, including risks related to suboptimal processes found in some Centers for Disease Control and Prevention–FDA collaborative investigations of disease transmission.
Jesse L. Goodman, M.D., M.P.H.
Food and Drug Administration
Bethesda, MD 20892
References
Phelps R, Robbins K, Liberti T, et al. Window-period human immunodeficiency viral transmission to two recipients by an adolescent blood donor. Transfusion 2004;44:929-933.
Biswas R, Tabor E, Hsia CC, et al. Comparative sensitivity of HBV NATs and HBsAg assays for detection of acute HBV infection. Transfusion 2003;43:788-798.
Schuttler CG, Caspari G, Jursch CA, Willems WR, Gerlich WH, Schaefer S. Hepatitis C virus transmission by a blood donation negative in nucleic acid amplification tests for viral RNA. Lancet 2000;355:41-42.
Busch MP, Tobler LH, Gerlich WH, Schaefer S, Giachetti C, Smith R. Very low level viremia in HCV infectious unit missed by NAT. Transfusion 2003;43:1173-1174.