Quest for the holy grail: pathogen reduction in low‐income countries

Over the past three decades, great improvements have been made in low- and low-middle-income countries (LICs, LMICs) in the testing of blood donations for human immunodeficiency virus (HIV) and hepatitis B and C viruses (HBV, HCV). In the World Health Organization (WHO) 2013 Global Database survey, more than 95% of countries reported having a policy of screening all donations for HIV, HBV, and HCV by serology.1 However, only two of 46 countries in Africa reported using nucleic acid testing (NAT), and in spite of the testing policies, 13 countries, including six in Africa, reported not being able to test 100% of units collected, possibly because of an irregular supply of test kits, and only 66% of LICs reported that the testing was performed in a quality-assured manner (i.e., with the use of standard operating procedures and participation in an external quality control program). Thus, while there have been recent improvements in blood safety, transfusion-transmitted HIV, HCV, and HBV continue to be a much bigger risk for transfusion recipients in LICs/LMICs than is the case in high-income countries (HICs). According to WHO, as many as 5% to 10% of transfusions lead to a transfusion-transmitted infection (TTI).2, 3 The United Nations Programme on HIV and AIDS (UNIADS) has suggested up to 1% of new HIV infections per year may be attributable to transfusions.4 The reported median proportion of blood donations with positive or reactive results on screening tests for HIV, HBV, and HCV in HICs versus LICs, respectively, are: HIV, 0.003 versus 1.08; HBV, 0.03 versus 3.70; HCV, 0.02 versus 1.03 (although in LICs, unlike HICs, these results are often reactive but not confirmed positive results).1 With respect to the actual residual risk of TTIs, in one study, using mathematical modeling, the residual risks for transmission of HIV, HBV, and HCV in sub-Saharan Africa (SSA) were estimated to be 1, 4.3, and 2.5 per 1000 units, respectively.5 Another study, using data from repeat donors in five SSA countries, estimated the combined residual risk of HIV transmission to be 1 per 29,000 units.6 Additional viruses also threaten the blood supply, including Zika virus, hepatitis E virus (HEV), Dengue virus, and human herpesvirus 8 (HHV-8), for which routine screening is not conducted.7-13 Bacterial contamination, often introduced during blood collection of platelet concentrates and sometimes RBCs, is another substantial infectious disease risk to blood transfusions that is not tested for in LICs/LMICs.14 Screening tests for parasites are rarely performed, and specific threats include leishmaniasis, malaria, and babesiosis.15, 16 Transfusion-transmitted malaria is of particular concern in those LICs/LMICs where malaria is endemic, as blood donations in these regions are not tested for malaria. In a review of 17 published studies from SSA, the median prevalence of malaria among blood donors was found to be 10% (although little is currently known about the rate and severity of transfusion-transmitted malaria in these regions).17 Finally, screening for pathogens, when performed, mitigates risks for only specific, known infections, so transfusion recipients remain susceptible to untested infectious agents. In addition to the risk of TTIs, some transfusion recipients in LICs/LMICs may be at risk of transfusion-associated graft-versus-host disease (TA-GvHD). Most LICs/LMICs do not have access to and/or cannot afford gamma-irradiation or leukoreduction. Since transfusions in these countries often consist of fresh or relatively fresh whole blood (WB), these units may contain viable donor white blood cells (WBCs). Live donor WBCs can cause microchimerism in a transfused immunocompromised patient, where small populations of donor cells engraft and survive in the host. The consequences of TA-GvHD are dire, as mortality rates of 80% to 90% have been reported.18 A low-cost, easy-to-use technology that can inactivate WBCs and a wide array of pathogens, including known, emerging, and unidentified threats, in WB could potentially have a dramatic public health impact in LICs/LMICs. Two methods of pathogen reduction (PR) of WB are in development.19-23 Both target DNA or RNA to prevent pathogen replication: one uses an ultraviolet light–triggered reaction with riboflavin (Mirasol, Terumo BCT), and the other a frangible chemical alkylating agent (Amustaline or S303) in combination with glutathione (Swiss Red Cross in collaboration with Cerus Corp.). Cells with nucleic acids, such as viruses, bacteria, parasites, or WBCs, are inactivated to reduce their infectivity by varying degrees. Plasma and cells without nucleic acids, like RBCs and platelets, suffer only minor cellular damage. PR technologies have been shown to substantially reduce the pathogen burden of WB and are likely to reduce the risk of TTIs and TA-GvHD from WB transfusions.15 With one of the technologies, in vitro spiking experiments, animal models or culture assays showed that PR decreased Trypanosoma cruzi,24 Babesia microti,25 Leishmania donovani,26 and Plasmodium falciparum27 between a 3.3 and 7.0 log reduction in parasite load. PR interdicts infection in the pre-seroconversion window period for the major TTI virus HIV.28-31 This technology for WB has also been shown to have variable in vitro efficacy against both enveloped and nonenveloped viruses, including most hepatitis viruses (HAV, HBV, HCV, HEV)32 and is likely to reduce disease transmissions. Effectiveness has been demonstrated by preventing the transmission of 6 log of cytomegalovirus in a mouse disease transmission model.33 Among spiking experiments with bacteria performed in RBCs (as opposed to WB) at low titers, PR was 80% effective against Acinetobacter baumannii, and 100% effective against Serratia liquefaciens, and Yersinia enterocolitica.32 At higher titers, PR in RBCs was also shown to reduce but not eliminate a variety of bacterial species by more than 3 to 5 log reduction, with complete inactivation shown for Streptococcus pyogenes (≥5.1 log reduction).30 In addition to TTIs, PR treatment of WB was effective in reducing viable T cells, antigen presentation, cellular activation, and cytokine secretion.34, 35 In this issue of TRANSFUSION, McCullough and Butler present a model describing the use of WB PR with rapid diagnostic tests (RDTs) as an alternative to standard serologic testing in a low-resource setting. RDTs are used alone for TTI testing in some African countries, but because of their low sensitivities are generally considered to be suboptimal as compared to standard serologic assays. However, the use of RDTs together with PR as an alternative to standard serologic testing is an intriguing idea. Based on blood transfusion parameters from the Ugandan Blood Transfusion Service (UBTS), McCullough and Butler estimate that the combination of RDT with PR could reduce the rate of transfusion-transmitted HIV, HBV, HCV, and malaria in Uganda by 100%, 20%, 98%, and 83%, respectively. They conclude that this combination could substantially enhance blood safety in Uganda as well as other LICs. As with many model-based evaluations, this study is an extremely thought-provoking and hypothesis-generating experiment. There are nevertheless several limitations that need to be considered when interpreting the results. The model was based on one country in Africa (Uganda), so the applicability to other LICs/LMICs is unknown. Their model (like all models) required the authors to make several assumptions, which the authors rightly acknowledge may not all be accurate. Empirical data are always best, and sensitivity analyses should be conducted to vary the input parameters, especially when assumptions are used in the model. Except for a single-center randomized trial of 226 participants in Ghana that showed the incidence of transfusion-transmitted malaria was significantly lower in the PR group (1/28 [4%] vs. 8/37 [22%]; p = 0.039),36 in vivo human efficacy estimates for PR of WB are not available. This small trial in Ghana did not assess other TTIs or adverse outcomes that can be evaluated only with larger and longer trials. Further research in this area is critically needed. McCullough and Butler limited their study to an analysis of the effect of RDTs with WB PR on the risk of TTIs with respect to standard serologic testing. Obviously, the very important questions of the feasibility and the costs of implementing this approach to the prevention of TTIs need to be studied, and formal cost-effectiveness studies need to be performed. These should also include cost of this approach as opposed to the costs of introducing standard serologic tests in those regions where only RDTs are used and of introducing NAT in addition to standard serologic testing. If feasible and cost neutral (or even cost saving), WB PR could offer a unified approach to address blood transfusion risks even in LICs/LMICs. PR has been shown to be effective across different classes of pathogens including bacteria, viruses, and parasites, with variable effectiveness. With traditional assay-based testing, high costs and regulatory hurdles are extremely challenging barriers for LICs. Screening infrastructure in low-resource settings, such as Africa, still has major gaps not only for the classic TTIs, that is, HIV, HBV, and HCV, but also for other TTIs such as bacteria and malaria, in addition to emerging and/or unknown infectious agents. With subsequent feasibility and implementation studies and clinical trials, we believe the uptake in LICs/LMICs could occur, especially if additional empirical and modeling studies show it is prudent, feasible, and economical. AART has received travel support from Terumo BCT. AART and HAH have a pending grant proposal to evaluate feasibility of implementation of Mirasol whole blood pathogen reduction in Uganda. Aaron A.R. Tobian, MD, PhD1 E-mail: atobian1@jhmi.edu Heather A. Hume, MD2 E-mail: heather.hume@umontreal.ca 1Johns Hopkins University School of Medicine Baltimore, MD 2CHU Ste Justine, University of Montreal Montreal, QC, Canada

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.