The onset of viremia, defined as verifiable detection of virus particles or genetic material in samples taken from the bloodstream, marks a critical time point in the natural history of HIV infection because it indicates infectivity of the exposed individual with the potential for secondary transmissions1-3 and provides the opportunity to diagnose and confirm the infection from a routinely collected blood sample.4-7
HIV RNA levels rapidly and predictably increase in serial samples from newly infected persons from the earliest quantifiable viral concentration to a peak level approximately coinciding with the time of antibody seroconversion.4,8,9 This phenomenon of ramp-up viremia allows back-projection from a given time point in the pre-antibody seroconversion period to the presumed onset of detectable viremia, assuming unchanged viral replication kinetics during this time frame.8 The ability to perform these calculations provides an important tool to project further closure of the window period (WP) by increasingly sensitive methods of viral detection and is critical for directing research toward the development of rational policy decisions on the implementation of new testing strategies.10
Implicit in the model calculations is the assumption that the beginning of the ramp-up phase of HIV viremia is a discrete event that can be referenced to a single time point after exposure to the virus. The relatively short interval between exposure to HIV and peak viremia supports this assumption, but confirmation from laboratory studies is lacking. To study this question, we took advantage of the availability of serial plasma sample panels from source plasma donors newly infected with HIV. Samples collected before onset of ramp-up viremia (defined as ≥100 copies [cp]/mL) were retested with a highly sensitive qualitative HIV reverse transcriptase (RT) polymerase chain reaction (PCR) assay with increased volume input to look for low-level viremia (LLV) during the preceding very early phase of primary HIV infection.
MATERIALS AND METHODS
HIV-1-Positive Plasma Donor Panels
Plasma donations (range: 600-800 mL) from source plasma donors (ie, paid donors who donate plasma for manufacture of plasma derivatives) were routinely collected at US donor centers operated by Alpha Therapeutic Corporation (ATC, Los Angeles, CA) from August 1996 through December 1998 at approximately twice-weekly intervals. Sample collection was approved by the Biomed Institutional Review Board in San Diego as part of an Investigational New Device application to the US Food and Drug Administration (FDA) for a qualitative HIV-1 RT-PCR assay (UltraQual HIV-1 RT-PCR assay; National Genetics Institute [NGI], Los Angeles, CA) for source plasma donor screening. The median donation interval for the 15 donor panels that were analyzed in this study was 4 days (range: 10th-90th percentile, 2-7 days). The collected plasma was frozen within 8 hours of collection and routinely stored frozen at −20°C or lower for an at least 60-day quarantine period. Samples of plasma from each donation were submitted for infectious disease screening, including serologic assays and HIV RNA testing in pools of 512 with the NGI's UltraQual assay. HIV infection was confirmed by follow-up RNA testing and demonstration of antibody seroconversion on undiluted samples from individual donors. HIV-1-positive donors were notified, counseled, and permanently deferred. Frozen plasma donations from cases with confirmed viremia followed by seroconversion were retrieved to construct panels comprising sequentially drawn plasma samples. Each donation was rapidly thawed and aliquoted, and the aliquots were refrozen at −20°C or lower. Serial donation aliquots were coded and compiled into anonymized panels not linked to individual donors. The records of each donation date and the results of routine and research laboratory tests for each plasma aliquot were entered in a computerized database that included the anonymized study code identifications. Given the anonymized design, the study protocol was approved after expedited review by the University of California, San Francisco (UCSF) Committee of Human Research Internal Review Board.
Panel Inclusion Criteria
Of a collection of 44 available plasma seroconversion panels,8 15 panels were included in the present study, based on the following criteria: (1) they contained at least 2 samples that tested negative for HIV-1 RNA by the pooled donation screening nucleic acid testing (NAT) and the quantitative RNA assays (see section on HIV-1 assays), (2) at least 1 of these RNA-negative donations had been collected >10 days preceding the first donation with detectable RNA, and (3) samples with negative HIV-1 RNA results were followed by at least 1 sample with ≥100 cp/mL of HIV-1 RNA. The 15 selected panels each contained 2 to 7 samples with <100 cp/mL of HIV RNA (70 samples total).
HIV-1 Assays and Testing of Panel Samples for Low-Level Viremia
Each sample in the 15 included panels was characterized as to collection date and status for qualitative HIV-1 markers (eg, RNA, antibody) and HIV-1 RNA levels. Serologic testing was performed at ATC's central testing laboratory using FDA-approved commercial donor screening assays according to manufacturers' instructions. Tests included HIV-1/HIV-2 antibody enzyme immunoassays (Abbott Diagnostics, Abbott Park, IL) in conjunction with Western blot confirmation. HIV RNA quantitation was performed at the NGI, using the NGI's proprietary SuperQuant HIV-1 RT-PCR assay. This assay provides quantitative results in the range from 100 to 5 × 106 cp/mL.11 Samples with results >5 × 106 cp/mL were subjected to serial dilution and retesting to obtain quantitative results.
Of the 70 panel samples with <100 cp/mL of HIV RNA, 69 were retested in 4 to 10 replicates, each with 2-mL input volume, using the NGI's UltraQual HIV-1 RT-PCR assay. In addition, 7 of 8 samples with <400 cp/mL were retested in 5 to 7 replicates. The number of replicates tested was determined by available sample volume. The assay was approved in 2001 by the FDA for screening pools of source plasma donations containing up to 512 equal aliquots of individual plasma samples. The stated analytic sensitivity of the assay performed with a 2-mL sample is 1.4 HIV cp/mL at 50% detection and 4 HIV cp/mL at 95% detection. Strict precautions were observed to prevent inadvertent sample contamination during aliquoting, storage, and testing.12 LLV, defined as a positive result by the NGI's UltraQual assay on a panel sample, was confirmed by retesting of coded specimens in a second round of sample submission to the NGI.
In 9 of the 15 investigated panels, we found no evidence of LLV preceding the ramp-up phase of HIV viremia (Fig. 1). In 2 of these panels, none of the retested samples showed LLV; in an additional 7 panels, we found single samples with LLV that immediately preceded the first sample with HIV-1 RNA levels >100 cp/mL. We interpret these samples as most likely representing the beginning of the ramp-up phase of viremia (Figs. 2A-I). In support of this interpretation is the relatively short interval (range: 2-6 days, median = 3 days) in these panels between LLV and ramp-up viremia samples, which is compatible with the reported rate of viral expansion during the latter period.8,9
In contrast, a different distribution of LLV samples was observed in the remaining 6 panels (see Figs. 2J-O). In addition to 5 LLV samples that likely represent the beginning of the ramp-up phase of viremia, we identified in these panels 11 earlier samples with LLV (1-3 per panel) that were separated by 1 or more negative samples from samples that are contiguous with the beginning of the ramp-up phase. This pattern is consistent with intermittent LLV during the presumed nonviremic early period of primary HIV infection (see Discussion section). Such samples were collected 9 to 25 days (median = 18 days) before the first sample with an HIV-1 RNA level >100 cp/mL. The percentage of replicates that tested positive for HIV-1 RNA in these samples was variable, ranging from 10% to 90%, consistent with an approximate viral load of 1 to 10 cp/mL, based on the performance statistics of the qualitative HIV-1 RNA assay used.
Our understanding of the earliest events in primary HIV infection is still evolving. Initial insights were drawn from in vivo studies in primates infected with simian immunodeficiency virus (SIV).13,14 These SIV studies revealed a sequence of events with particular relevance to sexual transmission of HIV, beginning with viral penetration of surface mucosal epithelium, followed by infection of submucosal CD4+ T cells, dendritic cells, and macrophages and subsequent spread to draining lymph nodes.15 Measurable viremia was observed between 5 and 30 days after the experimental intravaginal exposure. A similar early period of HIV infection marked by undetectable viremia attributable to sequestration of HIV in mucosal and lymphoid tissues is presumed to exist in human infections.16 The length of this phase has not been well established but is generally expected to range between 1 and 4 weeks, based on the time of appearance of acute retroviral syndromes and of HIV viremia in patients with known dates of exposure.4,17
Our finding of seemingly random “blips” of LLV occurring up to ∼3 weeks before the well-described dynamic ramp-up phase of HIV in 6 of 15 newly infected source plasma donors suggests that viremia is not reliably absent during this very early phase of primary HIV infection. A similar blip viremia has been reported in early hepatitis C virus (HCV) infection18 but has not been previously described in primary HIV infection. Conversely, low-level (20-200 cp/mL) “spikes” of HIV viremia were observed by ultrasensitive assays in occasional patients on combination antiretroviral therapy whose plasma viral levels were undetectable (<200 cp/mL) by standard RT-PCR assay but who had detectable latent virus in peripheral blood CD4+ T cells.19 Resting CD4+ T cells infected with very low levels of HIV were detected in long-term seronegative individuals who were exposed to the virus on multiple occasions but neither seroconverted nor developed detectable viremia.20 The phenomenon of intermittent plasma LLV in very early primary HIV infection described here presumably reflects intermittent escape or leakage of small amounts of HIV from mucosal and lymphoid tissues near the entry site into the bloodstream or from rare circulating infected CD4+ T cells. This observation confirms the capacity for early spread and dissemination of HIV after mucosal penetration, a property that needs to be taken into account in vaccine development and in other strategies for prevention and early treatment.
Limitations of the current study are primarily those associated with anonymized opportunity samples. First, demographic data and other relevant donor information such as a history of high-risk exposures (which were denied or the donors would have been deferred), likely source of the infecting virus, time of exposure, and route of transmission are not available. Second, the availability of pre-ramp-up viremia samples is limited, restricting definitive determination of the beginning of blip viremia and precise characterization of the length and peak concentration of pre-ramp-up viral blips.
With these limitations in mind, we estimated the HIV concentration in plasma during viremic blips in the primary eclipse phase to be between 1 and 10 cp/mL. This level of viremia is at least 2 orders of magnitude lower than the reported threshold of 1500 cp/mL for heterosexual transmission,21 which makes it doubtful that infected persons would transmit HIV through sexual contact at this very early stage of infection. Conversely, it is conceivable that plasma collected from a newly infected donor during an early HIV blip could be infectious when given, for example, as a 200- to 250-mL fresh-frozen plasma component transfusion or a 50- to 400-mL platelet product. The infectivity of a red blood cell component derived from a donation given during the viremic blip preceding ramp-up viremia is less likely given the much smaller plasma content (approximately 20 mL) and extended storage at refrigerated temperatures, conditions known to reduce HIV infectivity.22
Our findings add uncertainty to the true beginning of infectious viremia in HIV infection, which may precede ramp-up viremia by several weeks and may extend back close to the time of exposure, causing theoretic concern that infectious blood or plasma donations may escape detection by currently used screening algorithms. Unlike source plasma, which is quarantined past the WP of HIV infection and thus is presumably protected from this concern, cellular blood components have to be released before the donor can be retested. Neither the “minipool” (MP) NAT currently used in blood donor screening2,23,24 nor the individual donation (ID) NAT used in selected high-risk screening and diagnostic settings has reliable sensitivity in the range of 1 to 10 cp/mL and thus would not consistently detect the low HIV concentrations associated with pre-ramp-up phase viremia.4-7 As shown here, even testing of multiple sample replicates with an ultrasensitive screening assay, a practice that is currently not feasible in routine blood donor screening, cannot consistently detect blips of very early viremia. In terms of current WP risk for US blood donations, the estimated residual risk for donations screened with MP-NAT (current practice) is approximately 1 in 2 million and implementation of ID-NAT might moderately reduce this risk to approximately 1 in 5 million.10 The LLV phenomenon described here could contribute another risk of 1 in 5 million, independent of the MP-NAT or ID-NAT risk estimates, assuming an incidence rate of HIV infection among blood donors of approximately 2 per 105 person-years10 and the presence of infectious LLV in 40% of donors for half of an average of 18 days preceding ramp-up viremia. Pathogen reduction methods,25 already applied in the manufacture of plasma components and derivatives and in various stages of trials and development for other blood components, seem to offer the best chance to counter the small theoretic risk of HIV transmission during the very early period of the infection.
The authors thank Barbara Johnson for expert assistance with preparation of the manuscript.
1. Busch MP, Satten GA. Time course of viremia
and antibody seroconversion following human immunodeficiency virus exposure. Am J Med
. 1997;102:117-124; discussion 125-126.
2. Ling AE, Robbins KE, Brown TM, et al. Failure of routine HIV
-1 tests in a case involving transmission with preseroconversion blood components during the infectious window period
3. Kopko PM, Fernando LP, Bonney EN, et al. HIV
transmissions from a window-period platelet donation. Am J Clin Pathol
4. Lindback S, Thorstensson R, Karlsson AC, et al. Diagnosis of primary HIV
-1 infection and duration of follow-up after HIV
exposure. Karolinska Institute Primary HIV
Infection Study Group. AIDS
5. Hecht FM, Busch MP, Rawal B, et al. Use of laboratory tests and clinical symptoms for identification of primary HIV
6. Holodniy M, Busch M. Establishing the diagnosis of HIV
infection. In: Dolin R, Masur H, Saag M, eds. AIDS Therapy
. 2nd ed. New York: Churchill Livingstone (Elsevier); 2002:3-20.
7. Nguyen KA, Busch MP. Evolving strategies for diagnosing human immunodeficiency virus infection. Am J Med
8. 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
9. Little SJ, McLean AR Spina CA et al. Viral dynamics of acute HIV
-1 infection. J Exp Med
10. Busch MP, Glynn SA, Stramer SL, et al. A new strategy for estimating risks of transfusion-transmitted viral infections based on rates of detection of recently infected donors. Transfusion
11. Lin HJ, Pedneault L, Hollinger FB. Intra-assay performance characteristics of five assays for quantification of human immunodeficiency virus type 1 RNA in plasma. J Clin Microbiol
12. Sninsky JJ, Kwok S. The application of quantitative polymerase chain reaction to therapeutic monitoring. AIDS
. 1993;7(Suppl 2):S29-S34.
13. Lifson JD, Nowak MA, Goldstein S, et al. The extent of early viral replication is a critical determinant of the natural history of simian immunodeficiency virus infection. J Virol
14. Nowak MA, Lloyd AL, Vasquez GM, et al. Viral dynamics of primary viremia
and antiretroviral therapy in simian immunodeficiency virus infection. J Virol
15. Pope M, Haase AT. Transmission, acute HIV
-1 infection and the quest for strategies to prevent infection. Nat Med
16. Kahn JO, Walker BD. Acute human immunodeficiency virus type 1 infection. N Engl J Med
17. Gaines H, von Sydow M, Pehrson PO, et al. Clinical picture of primary HIV
infection presenting as a glandular-fever-like illness. BMJ
18. Glynn SA, Wright DJ, Kleinman SH. Dynamics of viremia
in early hepatitis C virus infection. Transfusion
. (In press).
19. Finzi D, Blankson J, Siliciano JD, et al. Latent infection of CD4+ T cells provides a mechanism for lifelong persistence of HIV
-1, even in patients on effective combination therapy. Nat Med
20. Zhu T, Corey L, Hwangbo Y, et al. Persistence of extraordinarily low levels of genetically homogeneous human immunodeficiency virus type 1 in exposed seronegative individuals. J Virol
21. Quinn TC, Wawer MJ, Sewankambo N, et al. Viral load and heterosexual transmission of human immunodeficiency virus type 1. Rakai Project Study Group. N Engl J Med
22. Busch MP, Operskalski EA, Mosley JW, et al. Factors influencing human immunodeficiency virus type 1 transmission by blood transfusion. Transfusion Safety Study Group. J Infect Dis
23. Busch MP, Kleinman SH, Jackson B, et al. Committee report. Nucleic acid amplification testing of blood donors for transfusion-transmitted infectious diseases: report of the Interorganizational Task Force on Nucleic Acid Amplification Testing of Blood Donors. Transfusion
24. Delwart EL, Kalmin ND, Jones TS, et al. First report of human immunodeficiency virus transmission via an RNA-screened blood donation. Vox Sang
25. Dodd RY. Pathogen inactivation: mechanisms of action and in vitro efficacy of various agents. Vox Sang
. 2002;83(Suppl 1):267-270.