Portal de Periódicos da CAPES

Exposed or not exposed—that is the question: evidence for resolving and abortive hepatitis C virus infections in blood donors

2009; Wiley; Volume: 49; Issue: 7 Linguagem: Inglês

10.1111/j.1537-2995.2009.02266.x

1537-2995

Anders Widell, Michael P. Busch,

Liver Disease Diagnosis and Treatment

As a consequence of efforts to achieve increasing safety of blood components and plasma derivatives, the battery of tests applied to blood donors has expanded incrementally and become more sophisticated since the 1970s. Whereas tests such as those for hepatitis B surface antigen and for viral nucleic acids (nucleic acid test [NAT] for human immunodeficiency virus [HIV], hepatitis C virus [HCV], hepatitis B virus [HBV], West Nile virus) directly target viral antigens or molecular constituents of the blood-borne viruses, other screening systems rely on detection of antibodies to the viruses. There are also recently developed “combo tests” that simultaneously detect HIV and HCV viral antigens and antibodies; these tests detect both prevalent (antibody-positive) infections and incident infections with high levels of acute viremia, thereby interdicting donations given during the late ramp-up preseroconversion stages of these infections. Although the combined implementation of these tests generally allows classification of reactive donors into classic infection categories, such as acute infection (NAT/antigen-positive/antibody-negative), chronic carriers (NAT/antigen-positive/antibody-positive), resolved infections (antibody-positive/NAT/antigen-negative), and false-positive profiles, we have come to appreciate that some donors manifest additional categories of infection. This editorial will consider these atypical infection types with specific reference to HCV. One atypical HCV infection pattern is the so-called serosilent HCV carrier, defined as a person who is persistently viremic but antibody-negative. Such serosilent infections are very rare, an observation strongly supported by the absence of transfusion-transmitted HIV, HBV, or HCV infections from persistently antibody-negative donors. In blood donor populations like the United States, HCV antibody testing detects approximately 99% of infected donors and the major value of NAT has been detection of the approximately 1% of infected donors in the viremic preseroconversion window phase of infection.1, 2 The combined application of NAT and serology has also detected rare serosilent HCV carriers. For example, of 90 HCV RNA–positive, antibody-negative donors identified during the first 5 years of NAT screening by the American Red Cross who enrolled into a follow-up study, all but three were window-phase donations where the deferred donor later developed anti-HCV.1, 3 Two of the three long-time serosilent HCV-infected donors (one was HIV-coinfected) maintained high plasma levels of truncated HCV RNA with a deletion covering the structural genes. These HCV deletion mutants appear to represent defective interfering particles, which are common among RNA viruses, but were previously not documented for HCV. Such defective genomes need the support of a smaller coexisting population of wild-type viruses with full-length viral RNA genomes for their persistence. Of interest, all three serosilent HCV subjects were also nonreactive to recall antigens in T-cell immunology tests. Thus donor anti-HCV screening “sweeps the floor” by detecting the large majority of infected units, with NAT providing a quantitatively minor contribution by detection of window-phase and serosilent infected donations. This contribution depends on the level of risk in the donor population. HCV NAT yield may be close to nil in Scandinavian countries, intermediate in the United States, higher in some South European countries, and high in Egypt and other high-endemic settings.4 With respect to HCV antibody tests, the goal of high sensitivity has been reached at some cost in terms of specificity and deferred noninfected donors pay the price. For HCV the specificity of antibody screening tests in low-risk donor populations has been suggested to be 60%, with most enzyme immunoassay (EIA) repeat-reactive, NAT-negative sera testing low-level reactive and either recombinant immunoblot (RIBA)-negative or -indeterminate. Counseling these deferred donors is difficult and many remain concerned and seek expensive medical follow-up, which can add to their anxiety.5 Given that HCV is a virus with six or seven genotypes and a multitude of subtypes, it is rather amazing that a few recombinant antigens, mostly derived from Genotype 1a, are sufficient to detect antibodies generated in the vast majority of individuals infected by the different genotypes and subtypes and in individuals with diverse HLA types and levels of immune responsiveness. It has, however, been observed that some immunodominant epitopes, such as a peptide sequence of NSA4 that is useful for serotyping of HCV by type-specific antibodies, fail to induce antibodies in approximately 10% of patients, even to their own homologous subtype.6 This hole in the Swiss cheese of the adaptive B-cell immune response, occurring in immune-competent persons, is presumably compensated for by a redundancy in other B-cell epitopes in the same individual and hence does not significantly impact either primary screening test performance or the infected person's capacity to clear the virus. This finding of variable and selective immunoreactivity, however, should be remembered when considering T-cell immunity studies. HCV is well known for its propensity to cause chronic infections in the majority of cases, once the early pivotal phase of clearance/pseudo-clearance of 6 months has passed. In cases of clearance of HCV infection with true eradication of viremia, which occurs in 30% to 50% of infections, both CD4 and CD8 cells play important roles. For example, chimpanzee studies7 have shown that experimental depletion of CD4 cells at the time of HCV challenge leads to near uniform development of chronic infection. Once chronic infection is established, immune modulating (interferon) and antiviral treatment is needed to achieve viral clearance; current therapy results in permanent sustained response (persistently negative RNA >6-12 months off therapy) in 50% to 80% of cases. Both spontaneous and treatment-induced clearance of HCV viremia is followed by a waning of humoral immune markers. Such so-called seroreversion usually occurs over decades, although it may occur relatively shortly after seroconversion in cases of very rapid clearance. Seroreversion eventually leads to negativity by screening EIAs, but there is also a period during which weak EIA reactivity and single antibody reactivity on RIBA may be the only sign of previous exposure to HCV.8 Such cases are difficult to discriminate from nonspecific EIA and indeterminate immunoblot reactivity, and hence it is challenging to unequivocally classify subjects with these patterns regarding their past HCV infection status. Additional aberrant host responses to the HCV exist. Several years ago, Koziel and colleagues9 demonstrated that persons with no risk other than occupational exposure in a hospital laboratory showed adaptive HCV cellular immune markers without signs of conventional antibody responses to HCV or viremia. Several recent studies support this finding. In a study of 10 hospital staff members exposed by needle-stick injuries from HCV-infected patients, three developed new HCV-specific T-cell responses within months of the exposure, while two others already had such responses at the time of exposure, implying previous exposures; neither anti-HCV nor HCV RNA developed in any of the cases.10 Furthermore, studies in multiple exposed injection drug users (IDUs) have shown that active IDUs with anti-HCV but without HCV viremia have the highest rates (94%) and broadest T-cell responses, evidenced by both HCV antigen–induced proliferation and interferon (IFN)-γ production by enzyme-linked immunosorbent spot (ELISpot) assays; in contrast those with antibody and persistent viremia had lower rates of T-cell reactivity (45%), while those IDUs lacking both HCV viremia and anti-HCV displayed T-cell responses to HCV antigens at an intermediate rate (62%).11 Other studies have demonstrated that the breadth of HCV cellular immune responses to different regions of the HCV proteome is narrower among low-risk controls than those with multiple risk exposure who did not develop anti-HCV nor HCV viremia.12 This issue of TRANSFUSION includes two articles13, 14 on blood donors who were deferred after having reactive HCV screening EIA results and single antigen antibody reactivity in confirmatory RIBAs. Although indeterminate HCV immunoblot patterns have been generally dismissed as representing nonspecific reactivity unrelated to HCV exposure, these same patterns can transiently occur toward the end of the seroconversion window phase after acute infection, often showing reactivity to either NS3 (protease-helicase) or nucleocapsid/core, as well as after clearance of viremia as seroreversion evolves. In high-risk settings such as in IDU clinics, these RIBA-indeterminate patterns normally indicate either an acute infection (if RNA is detectable) or past infection (if RNA is absent). In contrast, in blood donor settings immunoblot indeterminate results in the context of negative HCV NAT have usually been considered as false positivity, although they present a dilemma leading to a confusing and often painful donor counseling session. On follow-up these donors rarely change in immunoblot reactivity and remain negative for viral RNA in blood, often perpetuating donor anxiety year after year, which is further exacerbated by the fact that the donor is permanently or indefinitely deferred from donation. In the study by Hitziger and colleagues,13 the authors recalled 72 previously deferred donors with absent viremia and an indeterminate pattern for anti-HCV on EIA and immunoblot. They used more recent versions of anti-HCV screening and non-Chiron immunoblot assays as well as in vitro functional T-cell assays. These assays involved stimulation of donor T cells with HCV proteins, including classical T-cell proliferation assays and assays that measured release of interferon-γ by T cells using the ELISpot format. In both types of T-cell assay recombinant HCV proteins derived from Genotype 1b were used. The investigators observed lower residual antibody levels correlated with increasing time since initial deferral, suggesting waning antibody levels over time as documented by others, although this finding was not definitive since the longitudinal pattern of each individual was not studied on serial samples with the same assay. In contrast, T-cell responses did not wane with time. T-cell responses were stronger with the proliferation assay than with ELISpot; actually all ELISpot-reactive cases were included among those with proliferative T-cell responses. Importantly, 56% of these ex–blood donors, initially identified by an indeterminate humoral (B-cell) response, manifested T-cell–driven HCV reactivity to their repertoire of HCV antigens, thus supporting the conclusion that these donors had a true exposure to HCV at some point in the past. The percentage was the same for the subsets of initially immunoblot-indeterminate donors with persistent versus waning antibody reactivity by EIA. Another observation in this study was that when comparing two immunoblot tests, concordant reactivity to HCV core antigen was seen in only 20 of 65 sera, whereas discordant immunoblot reactivity to core antigens was seen in 13 sera and concordant negativity in 32 sera. In a similar manner, concordant positive reactivity to NS3/helicase occurred in 18 and discordant reactivity in 12 cases. The RIBA immunoblot assay from Chiron was not included in this comparison, but the authors speculated that had this third immunoblot been included further patterns of discordant reactivity would have emerged. The bottom line remains that both immunoblot assay design and choice of antigens play a role in detection of HCV-specific antigens by these assays and that “monoreactivity” to a viral protein may be a “volatile” and somewhat artifactual finding. This apparent lack of reproducible patterns of reactivity between anti-HCV immunoblots from different manufacturers should be kept in mind when the outcome of even more complex HCV-specific T-cell assays developed in different research laboratories are interpreted. The study by Bes and coworkers14 used a slightly different approach to estimate the rates of remote HCV exposure in immunoblot-indeterminate donors as evidenced by HCV-specific T-cell reactivity. These investigators reviewed data from ongoing anti-HCV screening of a half-million blood donors. The authors used anti-HCV RIBA as their confirmatory assay and viral RNA was also assessed. Study subjects were divided into RIBA-indeterminate/RNA-negative, RIBA-positive/RNA-negative (resolved infection), RIBA-positive/RNA-positive (ongoing infection), and RIBA-negative controls (including donors with and without anti-HCV reactivity in screening tests). These authors used core, NS3, and helicase (a part of NS3) antigens from the same manufacturer as the Hitziger group and applied ELISpot and Th1 cytokines (IFN-γ, interleukin [IL-2]) and Th2 cytokine IL-10 release into medium in short-term cultures as T-cell immune response variables. Among RIBA-indeterminate donors, 48% showed ELISpot reactivity, a similar rate to that observed among subjects with ongoing infection (46%); interestingly an even higher percentage of T-cell reactivity was observed in subjects with resolved infection (83%). These findings are remarkably consistent with the T-cell proliferation data in the study by Hitziger and coworkers.13 IFN-γ release into supernatant paralleled the ELISpot pattern for core and NS3, whereas significant release of IL-2 was only seen for core. No significant difference was observed for the IL-10. It is noteworthy that 13% of the negative controls also showed T-cell reactivity to one or more HCV antigens, and the authors speculate that more subjects may have encountered HCV, although this reactivity could also be interpreted as background reactivity rates of these complex assays. Furthermore, the reactivities to B- and T-cell antigens developed independently and a monoreactivity to core antigens by B cells could well be linked with a T-cell response to NS3 antigen. Some caution should be paid to the fact that both these groups used recombinant antigens from the same manufacturer and of a single genotype,11, 13, 14 which may introduce a bias. As was seen with differential antibody responses to different immunoblot assays and NS4 serotyping kits,6 there are probably Swiss cheese–like holes in the T-cell recognition system. Thus the absence of reactivity may be due to mismatches between the antigen presented to the host and the one used in the assays. Neither of the two studies presented any HLA-restricted peptide data or any T-cell receptor staining using tetramers, which could impact their findings since T-cell responses are generally HLA-restricted.15 Also, although the donors in the present studies were all HCV RNA–negative in serum, there have been recent reports of “occult” hepatitis C infections, based on persistent low levels of HCV RNA found in the liver or in peripheral mononuclear blood cells of seropositive subjects, which can occur after spontaneous virus clearance or after successful treatment.16 There are also sporadic reports of occult HCV infections in patients with no peripheral anti-HCV or HCV RNA in plasma, but ongoing liver injury and HCV RNA in liver and/or blood cells.9, 10 Such patients may also display typical HCV T-cell responses and a subset of them (approx. 40%) has recently been shown to also have immunoglobulin G antibodies to a 15-amino-acid peptide of the N-terminal part of HCV core.17 It is unlikely that persons with sporadic occult HCV infection are infectious and in particular that they could infect recipients via donated antibody/RNA-negative blood. The absence of reports of transfusion transmission of HCV by persistently antibody- and RNA-negative donors is reassuring. Nonetheless, the reported occurrence of occult HCV infections warrants further studies of HCV RNA—and antibody-negative donors with elevated alanine aminotransferase using additional T-cell and B-cell assays as described above. In summary, assays for T-cell memory to HCV antigens, as employed in the current TRANSFUSION articles, provide additional tools that can help in understanding the basis for indeterminate HCV immunoreactivity. However, these assays are not yet ready to be employed for counseling of deferred HCV RNA–negative blood donors. The positive and negative predictive values of such assays have yet to be established, and the most relevant specific antigens and assay designs have yet to be defined. Future development in research settings will be required to establish which, if any, of these test formats will be suitable for clinical use. Options include the more simple, straightforward, and widely used ELISpot method or more complex assays such as nonradioactive labeled (e.g., 5,6-carboxyfluorescein diacetate ester) T-cell proliferation assays,18 receptor staining techniques using MHC-restricted Class I or II tetramers,15 or intracellularly stained cytokines trapped by brefeldin of HCV antigen–stimulated cells.19 Given the diverse range of host responses to different viruses, both for B- and T-cell antigens, combinations of these assays may well be needed to fully sort out the exposure status of donors with atypical patterns of reactivity on serologic screening and confirmatory assays. The authors claim no conflict of interest.