On November 29, 2018, experts in the field of infectious diseases, pathogen reduction technologies (PRTs) and other participants from blood centers, academia, and industry gathered at the Food and Drug Administration (FDA) White Oak Campus in Silver Spring, Maryland, for a 2-day public workshop entitled “Pathogen Reduction Technologies for Blood Safety.” The workshop opened with welcome remarks from Dr. Nicole Verdun, Director, Office of Blood Research and Review (OBRR), Center for Biologics Evaluation and Research (CBER), FDA, followed by introductory remarks from Dr. Peter Marks, Director, CBER, FDA. The first day of the workshop focused on blood-borne infectious agents and their impact on blood safety, experiences of the American Red Cross, and other blood establishments in implementing FDA-approved pathogen inactivation (PI) technology for plasma and platelets (PLTs) in the United States and novel PRTs under consideration for whole blood (WB) and red blood cells (RBCs). The second day opened with welcome remarks from Dr. Chintamani Atreya, Associate Director for Research, OBRR, CBER, FDA. The focus was on emerging innovations relevant to PRTs and potential alternatives to PRTs. The workshop concluded with remarks on insights for future research and development in this area for blood and blood product safety from infectious agents. A brief introduction of each session by the session moderator followed by a summary of the speaker presentation as submitted by the moderator and speaker are reported here. The first session titled “Blood-Borne Infectious Agents and Their Impact on Blood Safety” provides a state-of-the-science overview of the risks to blood safety posed by infectious agents. Additionally, this session addresses the strategies used to mitigate these risks in the United States including the introduction of increasingly sensitive laboratory screening testing platforms and PRTs for PLTs and plasma products. In a first part, the session includes a general overview of the evolution of responses to established, emerging, and reemerging transfusion-transmitted infectious diseases in the past 50 years. Further, it addresses the need for ongoing surveillance for and systematic responses to emerging infectious diseases (EIDs), optimally with sensitive metagenomics, multiplexed nucleic acid amplification technology (NAT) and serologic testing strategies in sentinel global donor populations. This is followed by a review of the major policy issues pertaining to the development and implementation of PRTs, which if successfully adopted will provide insurance against known and unknown pathogens that may enter the blood supply. It will be noted that these technologies, if applied to all blood components or WB, may allow for the relaxation of redundant donor laboratory screening, modified donor questioning and/or deferral, and simplified handling of postdonation information while preserving or enhancing the safety of the blood supply. The session ends with an overview of the current status of approved pathogen-reduced (PR) PLT and plasma products in the United States with attention provided to their current effectiveness and safety profile. Major reasons for the slow adoption of the currently approved PR products in the United States are discussed including cost-effectiveness (CE) considerations. Speaker's summary: Blood donor screening began in the 1940s with testing for syphilis, followed in the early 1970s by testing for hepatitis B surface antigen. The discovery of human immunodeficiency virus (HIV), human T-lymphotropic viruses, and hepatitis C virus (HCV) and introduction of progressively more sensitive serological assays targeting virus-specific antibodies and antigens for these “classic” transfusion-transmitted infections (TTIs) in the 1980s and 1990s was effective in interdicting the majority of infectious blood donations.1 Implementation of NAT screening for HIV, HCV, and hepatitis B virus (HBV) further reduced the residual risk of infectious window period donations, such that per-unit risks are less than one in 1,000,000 in the United States (Fig. 1).2-4 We now recognize that in addition to classic TTIs that establish chronic infection, agents that cause acute transient infections may also be TTI at significant rates if there are large epidemics or recurrent seasonal transmission.5-7 Salient examples of EIDs where interventions were implemented in the United States include nationwide screening of donors for Trypanosoma cruzi using a one-time antibody testing strategy,8 NAT testing for West Nile virus (WNV)9, 10 and Zika virus (ZIKV),11, 12 and testing for Babesia microti in endemic regions.13, 14 Testing for bacterial contamination of PLT components was instituted in the 2000s to prevent septic transfusion reactions.15 Donor deferrals were also implemented to reduce other risks including variant Creutzfeldt-Jakob disease and several other agents.16 Fig. 1 lists the EIDs for which interventions were implemented over the past two decades. The emergence of EIDs has proven to be unpredictable, as is their risk to blood safety.5-7 EIDs of concern span all pathogen classes, with 60% being from zoonotic sources. The AABB has developed “Fact Sheets” (available at http://www.aabb.org/tm/eid/Pages/default.aspx) that provide information on agent classification, background on pathogenesis and clinical syndromes, modes of transmission [including vectors/reservoirs], likelihood of transfusion transmission [TT] and information on known transmission cases, feasibility and predicted success of interventions that could be used for donor qualification [questioning], tests available for donor screening, and efficacy of PRTs). Proactive surveillance and research to evaluate responses to potential EID threats to blood safety have been adopted through collaborative initiatives of NIH, FDA, CDC, AABB, and blood research organizations.5-7 An unintended consequence has been the identification of agents found through viral discovery programs using metagenomics technologies that can theoretically be transmitted by transfusion but which, upon subsequent investigation, prove not to be (Fig. 1). The most striking example of this was xenotrophic murine leukemia–related virus, which was reported to be associated with prostate cancer and later chronic fatigue syndrome and to be present in the blood of asymptomatic blood donors.17 Intensive research consuming much time and money subsequently determined that xenotrophic murine leukemia–related virus did not affect humans and was a laboratory contaminant from cell lines that contained this murine virus.18 These experiences led to the US National Heart, Lung and Blood Institute (NHLBI) and FDA to convene workshops focused on proactive but rational and systematic responses to EIDs.6, 7 The Alliance of Blood Operators developed a “risk-based decision-making” process that includes formalized methods for quantifying risk and evaluating interventions.19 The 2015 to 2017 ZIKV pandemic is illustrative of the ongoing challenge of balancing timely and precautionary responses to emerging TTI threats with logistic and economic considerations. As the epidemic expanded in the Americas and the association of ZIKV with severe fetal outcomes emerged along with several cases of probable TT, the FDA mandated implementation of individual-donation (ID)-NAT or PRT of PLTs throughout the United States in 2016.11, 12 After 2 years of ID-NAT with very low yield at an annual cost of $137 million, the FDA ZIKV policy was revised to allow for minipool NAT with ID-NAT triggered during future epidemics, similar to successful screening for WNV. Nonetheless, ZIKV testing now represents a theoretical benefit at a high cost, precipitating consideration of regional screening policies and an urgent need to further define “tolerable risk” in the blood safety arena.20 Speaker's summary: Pathogen inactivation and/or reduction should be viewed in the context of shifting the blood safety paradigm from reactive to proactive thereby providing insurance against known and unknown pathogens that may enter the blood supply or are currently un(der)recognized. Based on the positive experience of PI for plasma derivatives (e.g., no HIV, HCV, or HBV transmission since 1987), it seems reasonable to apply this safety paradigm to blood components. Of note, a consensus conference held in Canada in 2007 issued recommendations in favor of rapid adoption of a PI technology even if it could not be applied to the full range of blood components. Despite these recommendations, PI technology for PLTs has been slow to be adopted in the United States. As described in Table 1, the reasons for this are many but it appears that the predominant impediment has been cost. When evaluating current blood safety risks and the need for additional interventions, it is important to understand that most often these risks are expressed on a per-unit basis and represent the likelihood of the agent surviving during storage in a particular type of blood component (e.g., plasma, PLTs, or RBCs). The risk of TT of an agent and/or the occurrence of symptomatic or serious disease is likely to be less than this per-unit exposure risk. On the other hand, these risk estimates are for single-unit transfusions; risk will be higher (i.e., multiplied by the number of units) for the majority of patients who receive multiple units, either in single exposures or over a treatment course. Assuming that therapeutic product efficacy is maintained and cost issues can be addressed, the goal is to have all blood components (RBCs, PLTs, plasma) or WB (before component separation) treated by PI—this could then allow for the relaxation of redundant donor laboratory screening, modification of donor questioning and/or deferral, simplified handling of postdonation information, and elimination of the need for irradiation of cellular components to prevent transfusion-associated graft-versus-host disease (TA-GVHD). Potential blood screening changes include eliminating syphilis, T. cruzi, cytomegalovirus (CMV), and Babesia testing; modifying the menu of HBV tests; eliminating off-season WNV and ZIKV testing; and eliminating ID-NAT. A fully PI-treated blood supply would shape the response to threats from new enzyme immunoassays in that there would be less pressure to develop laboratory screening assays. Additional important considerations in evaluating the role of PI in blood safety policy are that not all infectious agents are inactivated by PI technology (nonenveloped viruses and prions show variable resistance) and each manufacturer's process must be independently evaluated for quantitative levels of inactivation of numerous known pathogens as well as for therapeutic efficacy of the treated component and potential adverse effects in the recipient. The health care reimbursement system must also be able to accommodate the cost. Speaker's summary: Pathogen-reduced PLTs manufactured using a synthetic psoralen compound (amotosalen) are approved by the FDA for use by all patient demographics.21-24 Currently this is the only PLT PR manufacturing system approved by the FDA in the United States. It requires ultraviolet (UV)-A light activation of the psoralen photochemical to enable it to function as the inactivation agent.21-25 Approval is limited to single-donor PLTs collected using either of two apheresis devices and stored in a PLT additive solution, PAS-C, or in autologous donor plasma, depending on the apheresis device used for manufacture. Both PR products have a 5-day shelf life at 20 to 24°C.23 The psoralen product currently is being evaluated in PIPER, a Phase IV postmarketing study. Other manufacturing systems are under varying degrees of development.26, 27 One of these systems uses a different light-activated photochemical, riboflavin, and is currently being evaluated in the United States in a Phase III randomized clinical trial, MIPLATE.28 A third PR technology uses a shorter wavelength of UV light (UVC), as the sole mechanism of inactivation.27, 29, 30 It, too, is being evaluated in CAPTURE, a Phase III clinical trial in Europe. Despite FDA approval and the acknowledged benefits of the technology, however, the medical field has been slow to adopt and integrate PR technology into day-to-day hospital operations. Many of these concerns have been addressed in published studies. Importantly, clinical reports have shown PR PLTs to be clinically acceptable.31-36 FDA has provided draft guidance, but to date the Agency has stopped short of encouraging use of PR technology.21 Thus, it is left up to individual hospitals as to whether they adopt, or refrain from, use of PR PLT technology. Currently the biggest ongoing credible blood-borne pathogen threat to the nation's blood supply comes from bacterial contamination.37 While PR technology can address this, there are other options for mitigating risk of bacterial contamination of PLTs, including performing secondary bacterial cultures and point-of-release immunologic testing. However, should a new viral or other nonbacterial agent threaten the national blood supply, the time to ramp up adequate PR manufacturing infrastructure to meet such a nonbacterial threat would likely be substantial.38 More widespread adoption of PR technology now would do much to ameliorate the concern over this scenario. Overall, the use of PR technology is slowly increasing, and data addressing many of the above-listed concerns are being reported, at least in abstract form.31-36 However, the lack of an extensive degree of published US data, especially for pediatric and transplant recipients, coupled with the absence of a strong FDA endorsement of the technology and the high cost of this technology, has hampered widespread acceptance of PR PLTs.30 The possibility of another blood-borne threat to the safety of the national blood supply seems inevitable. How well we mitigate that threat may well depend on how these issues regarding PR blood products are resolved. It is critical that early adopters of PR technology in the United States publish their experience with utilization of PR PLTs for patient care, especially their pediatric experience. Speaker's summary: Two methods of PR plasma are currently licensed and available in the United States, solvent/detergent (S/D)-treated plasma (SD plasma; Octaplas, Octapharma) and amotosalen/UV-treated plasma (Intercept plasma, Cerus Corporation). Plasma treated with riboflavin and UV (Mirasol plasma; Terumo BCT) is part of a similar system being developed for other components and is also included in this summary. Each of the techniques results in a reduction of the content and/or activity of the pro- and anticoagulant proteins in plasma. In general, these reductions do not exceed 20% to 30%, and for many proteins the reduction is less than this. The most notable reductions are in fibrinogen and Factors (F)VIII and FV across all platforms; F IX and FXI, protein C, and large von Willebrand factor multimers with Mirasol; and protein S and antiplasmin with Octaplas.39-45 There have been few investigations of these treatments on complement components; in an analysis of Intercept plasma, C3a was found to be reduced.45 It has been noted that the reductions, while significant, usually resulted in contents within the reference range.39 The pooled nature of SD plasma greatly reduces the variability in content that can easily be demonstrated between different donors' plasma samples.46 Cryoprecipitate may be prepared from Intercept and Mirasol plasma units to meet the minimum content requirements, although the effect of the treatment is still evident.47, 48 In vitro analyses of the clotting system have generally demonstrated substantial retention of clinically relevant function.42, 44 As might be expected from the reduced contents noted, the resulting fibrin strands are thinner (with resulting increased clot density and reduced permeability) with longer lag time for formation or prolonged time to lysis.49 Multiple clinical trials have demonstrated expected outcomes with the use of these PR plasma samples. Prophylactic transfusion of Intercept plasma into congenitally deficient patients yielded expected increases in the deficient factor in circulation with anticipated half-lives.39 Use of large volumes of Intercept plasma in plasma exchanges for thrombotic thrombocytopenic purpura or idiopathic thrombocytopenic purpura resulted in expected outcomes without generating adverse events or (new) coagulopathy.50-52 Use of large volumes (approx. 2 L) of Intercept plasma in liver transplantation yielded the same outcomes as with quarantine plasma.50-53 The current formulation of Octaplas has not been reported to be associated with thrombotic events when used in large volumes as had been seen with the original version of SD plasma.54 A theoretical question has been raised whether the reduced content these components might place massive transfusion recipients at increased risk of inadequate hemostasis and death.55 One in vitro mixing study suggested that a 50% plasma replacement would be necessary before altering coagulation kinetics.56 Several large, historically controlled experiences with Intercept plasma in trauma situations, however, have failed to show any impact on the need for other blood components, time to discharge or mortality.32 These PR plasma samples have not been associated with increased adverse events after transfusion.57 Because of the pooled basis of SD plasma, it is believed to carry a reduced risk of transfusion-related acute lung injury (TRALI) because of dilution of potentially offending antibodies. There have been no TRALI cases reported in more than 10 million Octaplas units transfused.58, 59 If the existing risk of TRALI is greater than one in 5000, removal of this risk in itself makes use of SD plasma cost-effective.60 This pooling, however, does increase the risk of early and wide dissemination of a nascent nonenveloped virus.61 To date, there has been little uptake of PR plasma in the United States outside of Octaplas for patients with severe allergic reactions to single-donor plasma. This is probably due to the perception of viral safety (and lack of bacterial contamination risk) in standard plasma and the increased cost of these newer plasmas. Widespread introduction likely will follow only after implementation of similar systems for PLTs and RBCs despite admonishments from a consensus conference and demonstration of the importance of plasma transmission of new pathogens. Four years have passed since approval of a PRT devices for PLTs and plasma products by the FDA. The five presentations of this session addressed the implementation of these devices in the United States, their impact on PLT quality, the availability of PRT plasma in the United States, and the health economic considerations. Experiences from the nation's largest blood product supplier62 and the blood bank at the NIH Clinical Center63 were documented. Both reports stressed the relevance of strict volume and cell limits, not required without PRT, and their effect on collection procedures and failures. The approaches were almost diametrically opposed, reflecting the different donor settings: while the American Red Cross preferred small-volume over large-volume kits (2/3 vs. 1/3), the NIH Blood Bank exclusively used dual-storage kits. The American Red Cross is boosting PRT PLTs to meet the steadily increasing demand. The NIH Blood Bank has transitioned to 100% PRT PLT production, which was well received by the attending physicians and nurses. Quality variables are expected to change, as PRT affects all treated cells. The risks must be monitored and balanced while the technologies for PRT and PLT additive solution (PAS) continue to evolve.64 PRT plasma from individual donors, although FDA-licensed devices are available, had not been introduced in patient care by the end of 2018. Similar to lyophilized plasma, which is not available from single-donor sources,65 a PRT plasma alternative pooled from many donors does exist: S/D-treated plasma66 has a history of worldwide use since 1992. Five randomized controlled trials showed no difference in efficacy, but trial sizes ranged from 49 to 293 patients for a total of only 552 patients. No TRALI has been reported from passively collected data, which may not reflect all incidence. Cost-effectiveness estimates for PRT PLTs and plasma were reported.67 These products are considered no less cost-effective than other widely adopted interventions in the context of blood safety technologies. A budget gap is likely to remain until PRTs become available for WB or RBCs. Low- and middle-income countries may require a mixed approach of needs assessment and targeted interventions.68 While PRT PLTs have been implemented nationwide in some countries69 in an effort to improve patient safety, reimbursement was noted as a key factor in the United States stalling the quicker implementation of PRT PLT and plasma transfusions. Speaker's summary: Red Cross implemented PI technology with the treatment of apheresis PLTs in March 2015. The program was Initiated in Puerto Rico under a clinical study. The organization initiated routine production of PR PLTs in July 2016; it will have 15 manufacturing sites producing pathogen inactivated products by early 2019. The most significant challenge with during the initial experience was that a limited percentage of PLT collections were eligible for PI based on approved guard bands. There was also a goal to do no harm to the PLT supply due to the growing demand for single-donor PLTs by converting triples to doubles or doubles to singles to qualify for PI. Finally, a licensed INTERCEPT kit for PAS triples does not yet exist. The initial performance against guard bands was unsatisfactory; more than 30% apheresis PLTs produced in Red Cross are from triples leaving only doubles or singles to qualify. Most of the single- and double-plateletpheresis units failed to natively meet PI input requirements because of targeted programing set points to maximize yield on collections devices and the default values for first-time donors. Red Cross initiated evaluation of mitigation strategies designed to increase percentage of units that qualify for PI. Volume reduction was deployed by removing product volume from a homogenous mixture of Amicus-PAS apheresis PLTs meet Intercept guard bands. Red Cross developed and validated a software tool that aids staff in selecting options for volume reduction. Additional mitigations were developed in advance of an INTERCEPT triple kit. Staff split triple collections into three individual storage containers before PRT and use single (small-volume or large-volume) INTERCEPT kits to treat products individually. Large-volume doubles may be split into two small-volume or large-volume kits; the collections team adjusted Amicus settings to optimize storage volumes to 625 mL for doubles and 780 mL for triples, which included a minimum 10-mL volume buffer and a minimum PLT yield of 3.4. Approximately 65% of PLT products met the guard bands during the operational trial. Presplitting largely obviated the need for conducting volume reduction. The number of PR units currently labeled is less than 50% due to combination of demand, staffing, aggregates, and units exceeding 24 hours. Additional observations included a radical shift in type of PI kit used from predominantly dual-storage kits to small-volume kits; the use of large-volume kits remained the same. Before implementation of mitigations, the split rate of PI products fell to 1.30. After implementation of mitigations the split rate of PI products increased to 2.10. Programming of Amicus devices was standardized, and collection volume increased; triples became eligible and products were not downgraded. The labor required for PRT is greater than using the BacT/ALERT system with a single bottle. Time studies compared both processes. Demonstrated an 11.1% increase in the time required to complete tasks for unmitigated PI products as compared to the traditional process. There was a 22% increase in time required to complete tasks for mitigated PI products compared to traditional process. Productivity for the unmitigated PI process was poor because the volume of products eligible for treatment was low with batch sizes of two to four products. Productivity improved 52% after implementation of mitigation steps due to significant increase in products eligible for treatment with Increased batch sizes of eight to 12 products. Pathogen reduction of 100% products is challenging but not impossible based on current guard bands. Mitigations required to meet guard bands are feasible but labor-intensive and time-consuming. Implementation of PRT will require adjustment of set points and collection variables on apheresis devices. Speaker's summary: The NIH Clinical Center at the National Institutes of Health is the nation's largest hospital devoted entirely to clinical research.63 Approximately 1600 studies are in progress, focusing on Phase I and II clinical trials. The Department of Transfusion Medicine is a full-service blood bank providing blood donation, clinical apheresis, transfusion service, human leukocyte antigen (HLA) and infectious disease testing for the hospital, and cellular engineering70 in support of cell therapies.71 In 2016, we transfused 3930 PLT, 4561 RBC, 614 plasma, and 59 granulocyte products to 668 patients. All blood components are 25 Gy irradiated; RBC products are also leukoreduced72, 73 since 2009 and none is transfused older than 35 days74, 75 since 2014. We transitioned to PR PLTs (100% collected by apheresis) with PAS in January 2016. Within 1 month of the device's approval by the FDA, the NIH Clinical Center decided in January 2015 to implement PRT76 and concurrently a PAS for all PLT products, finally bringing this long-anticipated technology77, 78 to the bedside in the United States. A retrospective evaluation of 1007 successful collections during 6 months in early 2015 showed 99.7% of the collections that met any guard bands specified by PRT fell within the guard band of the PRT dual-storage kit. We decided to exclusively use this kit and prepared to adjust the variables for approximately 5% of our collections to meet the guard band specifications. If successful, we anticipated a loss of less than 1% of collections due to failures to meet the guard band. Once agreements with the device providers were signed (Intercept, Cerus Corporation; and Intersol, Fresenius Kabi), the initial tasks of the implementation team involved writing of validation plans and standard operating procedures; ordering, installing, and validating required equipment; and reconfiguring space to house equipment and fit the new work flow. Computer upgrades to accommodate changes were made by July. Training for PRT began in August and for PAS in September. Adjusting collection variables and validation of PRT and PAS processes continued for the next 3 months. The first apheresis PLT product with PRT and PAS was released on January 11, 2016. A dual inventory of PLTs produced by the new or the previous processes lasted for less than 1 week, because we promptly transitioned all our PLT collections to PRT and PAS. The fine tuning of collection variables was critical and needs to be monitored and maintained continuously. This remains an ongoing task for our donor staff during each PLT collection. We are closing in on our original goal of less than 1% guard band failures (Table 2) while producing 100% PRT PLTs from all plateletpheresis collections at the NIH Clinical Center. To bridge shortages of supply or serve patients with rare HLA antibodies, we must import PLTs that remain almost invariably produced without PRT. Their irradiation with 25 Gy is required before they can enter the inventory for release to patients. We do not provide bacterial testing upon release and retired our previous precaution to prevent bacterial contamination: 4 to 5 mL sampling within 24 hours of collection and release into inventory after 12 hours of negative culture in single standard aerobic bottles (BACTEC, Becton, Dickinson and Co.) while monitoring the culture for 7 days. With PRT products, the elimination of irradiation (possible without variance notification since March 17, 2016)79 and bacterial culture resulted in substantial savings of consumables and handling, which were however exceeded by the costs of the new technologies and the increased hands-on time for the production staff. Reports of transfusion reactions ranged from nine to 19 annually without discernible trend, certainly no increase from 2015 to 2018; no severe transfusion reactions occurred. Education and notification are important for the acceptance by hospital staff. Within 2 weeks before the first new product was released for transfusion, we notified several external customers and provided the revised circular of information; there were no calls received. The prescribers of the NIH Clinical Center were informed through the office of the deputy director for clinical care with a focus on improved patient safety; there were a few calls. The nursing staff was informed through the nursing education leadership. We provided photos of current versus new bags, highlighted the lack of irradiation labels for new bags only, and explained the new electronic transfusion documentation. Some institutions are reluctant to introduce PRT PLTs for neonates, children, and pregnant women because of the theoretical risks associated with the toxicity of psoralen and its photo products, for which there has been no evidence to date in the doses applied. The few concerns raised at NIH, for example, why the cost increase was justified, abated once the ZIKV occurred in the continental United States and PRT PLTs were recognized as safe without any action needed.