Lessons and opportunities from epidemiologic and molecular investigations of infected blood donors

Approximately 10 years ago I wrote an editorial in TRANSFUSION entitled “To thy (reactive) donors be true!”1 My goal was to focus attention on what I still believe is an obligation of those of us responsible for blood donor screening, both in government and in blood collection organizations, to assure that donors who test reactive on infectious disease screening assays are further evaluated by optimal supplemental testing methods and confirmatory algorithms (of index donation and follow-up specimens) to definitively classify their infection status. Donors with results determined to be falsely positive should be reassured, and reinstatement options should be available to offer them the ultimate reassurance—the opportunity to donate again. Donors who are determined to be infected should receive confidential notification and counseling that includes communication of accurate information regarding the health implications and treatment options of their infection and risk of transmission to their contacts, as well as appropriate referrals for follow-up counseling and care.2 Unfortunately my plea had only a modest impact, and my message remains essentially as relevant today as it was 10 years ago. In the United States, we continue to unnecessarily defer more than 10,000 uninfected donors each year, many of whom are given an “indeterminate” test result message due to the combination of nonspecific screening and supplemental serologic tests.3 There is a disturbing lack of state-of-the-art supplemental assays approved by the Food and Drug Administration (FDA) for use in resolving the infection status of reactive blood donors. This shortage is due to both the lack of financial incentives for test manufacturers to develop and commercialize assays with low volume markets and the regulatory burden and timeline required to advance donor infectious disease screening and supplemental tests through the FDA approval process. For example, blood banks in the United States are still required to use human immunodeficiency virus (HIV)-1 viral lysate–based Western blot or immunofluorescence assays that were licensed by the FDA almost 20 years ago, despite well-recognized problems with availability of these tests, high rates of irrelevant indeterminate and even false-positive results, and inadequate sensitivity to early seroconversion relative to contemporary screening assays.3,4 There is no licensed supplemental test for human T-lymphotropic virus (HTLV)-reactive donors, and therefore no reentry option, with many blood centers relying on a single state public health lab that performs a complex series of home brew assays.5 The latest supplemental test for hepatitis C virus (HCV) available in the United States (HCV recombinant immunoblot assay [RIBA]) was approved by the FDA more than 10 years ago and also has documented problems with specificity and sensitivity to early seroconversion and some viral genotypes.6,7 We continue to use a neutralization procedure to confirm hepatitis B surface antigen (HBsAg) reactivity despite the fact that this procedure yields false-positive interpretations due to both specimen cross-contamination and biologic nonspecificity. Fortunately, the recent implementation of nucleic acid testing (NAT) has led the FDA to consider new confirmatory algorithms that integrate NAT results with serologic screening and supplemental test data.8-11 When fully approved and implemented, these algorithms should result in reduced use of outdated supplemental serologic tests, more accurate donor notification messages, and increased opportunities for expedited reinstatement of false-positive donors. Unfortunately, the process of formulation and approval of these algorithms has been slow. Moreover, many in the blood banking industry in the United States still consider reactive donors to be a problem to be avoided rather than a responsibility to be effectively addressed. Hence some US blood collection organizations perform the minimum confirmatory testing possible. Most centers notify false-positive and even confirmed-infected donors (except HIV) by letters, rather than by a personal communication, which presents the donor with limited and often outdated counseling and medical referral information. A significant proportion of US blood banks do not offer eligible donors reinstatement testing even when algorithms have been approved by FDA due to the low yield of actual reinstatement and subsequent donations and concerns over regulatory actions that have occurred after “inappropriate” reinstatements. The situation is very different in other regions of the world. Recently developed, more specific serologic screening assays and more accurate supplemental assays and testing algorithms are approved in the European Union (i.e., CE marked) and by other regulatory authorities and are routinely employed. These testing protocols include inexpensive and highly accurate dual enzyme immunoassay testing algorithms and recombinant/peptide antigen–based supplemental assays that enable better classification of reactive donors into true and false positives, with minimal rates of indeterminate results compared to those reported by US blood banks.12 Many of these tests also subtype HIV, HTLV, and HCV, which in some cases has important implications for donor counseling. Reactive donors, including those who are determined to be truly infected and donors with false-positive results, are generally recalled and notified in person by physicians or trained counselors. The goal of these interactions is both to provide donors with appropriate information and care and to learn from the true-positive donors about how they became infected: the latter can result in better understanding of the effectiveness and potential need for modification of the donor selection and screening process, as well as contribute to public health surveillance. Two articles from the Netherlands in the current issue of TRANSFUSION demonstrate the insights that can be gained from rigorous studies of infected blood donors. In the first study, Van der Bij and coworkers13 report the results of risk factor interviews performed on 593 confirmed-infected donors, representing 77 percent of all infected donors identified in the Netherlands from 1995 through 2003. Of note, 67 to 85 percent of hepatitis B virus (HBV), HCV, and HTLV infections were among new donors (from whom blood is not collected), whereas 79 percent of HIV infections were among repeat donors. Rates of these viral infections among accepted donors were universally lower than in the general population, indicating substantial selection for safer first-time donors. When interviewed by a blood center physician, however, a large proportion of infected donors acknowledged risk behaviors that would have deferred them from donating if they had been disclosed during the screening process. For example, 19 percent of new donors infected with HCV reported previous injection drug use, and 31 percent of repeat donors infected with HIV reported male-to-male sexual contact. Overall, approximately 25 percent of the infected donors (22% of new donors and 28% of repeat donors) reported risk factors that should have led to their deferral had they been disclosed at the time of donation. The authors make a number of recommendations for improving the donor screening process and argue for retention of their policy of not collecting blood units from first time donors (similar to the applicant donor program used by US commercial source plasma collection programs) and retention of long-term deferral of men who have had sex with men. Although similar risk factor interview studies have been conducted in the United States and other countries, including several recent studies from developing countries, these have generally been limited to infected donors identified during licensure trials or shortly after implementation of new donor screening assays. Studies of risk factor profiles among HIV-infected donors were funded by the CDC for approximately 10 years after implementation of serologic screening in the mid-1980s, whereas studies of HTLV- and HCV-seropositive (and indeterminate) donors, funded by NIH, were conducted in the early 1990s. Insights from these studies led to major changes in the content and process of donor qualification criteria and health interviews, which contributed to significant reductions in the viral prevalence (among first-time donors) and incidence (among repeat donors) rates.14,15 More recently, after implementation of HCV and West Nile virus NAT screening, interview studies were conducted by the American Red Cross in collaboration with the CDC and FDA, to understand the risk exposures of donors detected in the viremic preseroconversion window period.16,17 Unfortunately, none of these studies is ongoing, in contrast to the systematic risk factor interview process described in the study from the Netherlands and in place in a number of other countries with comprehensive hemovigilance programs.18 The second Dutch study, by van de Laar and associates19 integrated epidemiologic and risk factor information with detailed molecular genetic data (from sequencing and phylogenetics analysis) for 99 HCV-infected donors identified in the Netherlands from 1997 through 2002; these data were compared with sequence data from 99 infected local injection drug users (IDUs), matched for age and sex with the infected blood donors. Through rigorous analyses, the investigators demonstrate linkage between modes of acquisition of HCV by donors and HCV genotypes, indicating that separate HCV epidemics have occurred in the Netherlands over the past several decades. Compared to IDUs, the genotype distribution in infected donors was highly diverse. Half of the infected donors harbored IDU-related subtypes 1a and 3a, whereas 24 percent of donors were infected with subtype 1b, which was associated with a history of blood transfusion before donor screening and to various other remote nosocomial exposures. The study by van de Laar and coworkers further demonstrates the value of molecular epidemiologic studies of infected donors as an approach for providing insight into the origin and dynamics of infections at the population level.20-22 Surveillance for molecular variants in blood donors is also vital to assuring that blood screening and supplemental assays are sensitive to circulating strains of blood-borne viruses. Documentation of divergent viral strains in the donor pool can lead to accelerated development and licensure of robust serologic and NAT assays that have enhanced capacity for detection of divergent infections in donor screening and diagnostic applications. In addition, surveillance for viral variants among donors can have implications for assessing the prevalence of drug and vaccine escape mutants and for detecting and monitoring rare variants that may be newly introduced or increasing as a result of “global trafficking.” Similar to the lack of ongoing systematic interviews of infected donors in the United States, the logistics and cost of molecular genotyping and phylogenetic studies has resulted in such studies being performed only on a limited number of infected donors identified shortly after implementation of new screening assays. Impediments to such studies can be overcome, however, with an alternative approach that focuses on the targeted characterization of viral genomes in donors with incident (recently acquired) infections. Combined NAT and serologic screening, supplemented by novel serologic testing strategies such as the “detuned HIV antibody assay,” can be used to identify donors with incident infections.8,22,23 Such donors are of particular interest with respect to blood safety and public health. Identification of donors with acute or recently acquired viral infections allows viral genetic analysis to be focused on actively circulating viruses rather than viral strains that may have been transmitted years or even decades earlier. As such they provide a more current view of the viruses of interest and provide more pertinent data with regard to evaluating different screening and diagnostic methods and estimating the rate of transmission of drug-resistant variants in the general population. Furthermore, by comparing the distribution of variants belonging to different genetic lineages in donors with those in high-risk populations, genetic analysis of incident viral strains allows the molecular epidemiology of these pathogens to be tracked over time and in different geographic areas. The Retrovirus Epidemiology Donor Study—II (REDS-II) has developed a molecular surveillance study with the primary objective of detecting genetically divergent HIV, HCV, and HBV in blood donors.22 Donations representing incident infections will be identified by the major US blood collection organizations, including the six REDS II blood centers, the American Red Cross, Blood Systems, and the New York Blood Center, and the phylogenetically most informative regions of each virus will be sequenced. Strains in recently infected donors will then be compared to strains from representative seropositive prevalent cases in the same population of blood donors, as well as to strains from high-risk groups. This study will allow us to determine whether the distributions of variants are changing over time, vary geographically, or differ relative to donor demographics. These data will allow subtype and/or genotype analyses to determine whether previously rare subtypes are becoming established in US blood donors. In addition, the frequency of mutations associated with antiviral drug resistance (HIV and HBV), vaccine or immune globulin neutralization failure (HBV), or alterations potentially affecting immunoassay detection (HBsAg) will be determined in recently infected low-risk blood donors. This study is scheduled to launch in 2006 and continue for at least 3 years. International sites in China and Brazil that are joining the REDS-II program in 2006 will be included in this molecular surveillance program. As illustrated by the study by van de Laar and associates, it is valuable to examine the association between viral genetic data and likely exposure routes and donor motivations such as test seeking. Longitudinal analysis of these data will help us assess the efficacy of the donor selection and interview processes and the impact of changes to eligibility criteria and interview methods and content. Finally, there is early discussion of extending this program even further by launching a global study of molecular diversity in recently infected blood donors under the auspices of the International Society of Blood Transfusion Working Party on Infectious Diseases. Hopefully, 10 years from now, we will look back and appreciate the benefits of systematic molecular epidemiologic studies of infected blood donors with respect to both enhanced blood safety and contributions to public health.