Implementation of pathogen inactivation technology: how to make the best decisions?

A decade ago, this journal published an editorial penned by Dr McCullough accompanying the publication of a consensus statement arising from a consensus conference on pathogen inactivation (PI) technologies.1, 2 This consensus conference recommended that “pathogen inactivation should be implemented when a feasible and safe method to inactivate a broad spectrum of infectious agents is available.” In the intervening decade, further implementation of PI for blood components has taken place in numerous jurisdictions, including recent approval of the first PI technology for apheresis platelets (PLTs) in the United States. However, many of the uncertainties identified a decade ago remain. The drivers for implementation of PI have varied from one jurisdiction to another, ranging from avoidance of implementation of bacterial testing entirely to increased risk reduction from bacterial contamination despite the use of culture methods and to the more comprehensive goal of mitigation of unknown or newly emerging pathogens. Notably in the United States, the implementation of PI for components has been jumpstarted by the extensive spread of Zika virus into the western hemisphere and concerns that dengue virus and other pathogens may do likewise, along with its potential as an alternative strategy for bacterial risk reduction. Many blood collectors and national blood organizations have struggled to decide first whether to implement PI and, if so, to what extent. These decisions are challenging because they are not simple. They require establishing comfort with the balance between increased safety from pathogen transmission and an apparent decrease in circulation residency time of the transfused PLTs and a concomitant decrease in the interval between transfusions, which raise questions of decreased efficacy and some remaining concern about the potential toxicity of treated PLTs. At this point in the evolution of PI, we do not understand enough about the damage that PI causes to PLTs to develop strategies to ameliorate it. Thus, for the foreseeable future, the risk decision must continue to balance increased safety from pathogens or passenger white blood cells with PI-mediated reduction in the product quality. Considerations of cost and reimbursement mechanisms are also factored into these decisions and cost may preclude adoption even if other factors argue in favor of PI implementation. From the perspective of the blood operator, the full cost of implementing PI technologies must be considered, especially if there is no offsetting financial support from governments and all costs are borne by the purchasing hospital. In jurisdictions where there are multiple suppliers, cost pressures may be even greater. These costs may not only be related to the PI systems themselves, but include additional staffing costs required to perform the PI procedures, as well as additional costs involved with recruitment, collection, and production arising from modifications necessary to install PI. The hospital customers of the blood operators may also face increased costs, not only for additional products per treatment course, but also for the labor and supplies associated with increased volume of transfusions and the additional cost per treated unit charged by the supplier. The best policies for blood safety are created using relevant data and with consideration of the elements of a risk-based decision-making framework.3-5 A number of countries that have considered the introduction of PI have undertaken studies intended to provide the data necessary to assess the performance of treated products or to decide on which technology provider to adopt. This issue of TRANSFUSION reports on the efforts of Italy to provide data to inform decision making about the introduction of PI technology for the treatment of PLT concentrates.6 The Italian Platelet Technology Assessment Study (IPTAS) was designed to conduct two simultaneous clinical noninferiority studies examining the two Council of Europe–marked PI systems, Intercept (Cerus) and Mirasol (TerumoBCT). It is the first study to compare the two technologies in the same time frame with the same protocol in the same country. This ambitious assessment was designed to study 828 stable hematology-oncology patients undergoing chemotherapy or marrow transplantation in six Italian centers that also had the ability to treat their own PLT concentrates. Both apheresis and whole blood–derived pooled PLT concentrates manufactured by the buffy coat method were used either untreated or after PI treatment. Each participating hospital used only one of the two technologies. Unfortunately, funding ran out before the investigators could complete the study; however, there are informative data for the community to consider arising from the study as reported out. While the ability to assess a noninferiority study with a primary endpoint of WHO Grade 2 or greater bleeding was compromised, the data from 424 evaluable patients were analyzed from the perspective of intent to treat. A collection of secondary outcomes was compared between the PI and control arm using absolute differences or ratios and their 95% confidence intervals; this analysis was done in the absence of any a priori hypothesis. The findings of the IPTAS study are instructive. First the impact of PI on the PLT-related measures was grossly similar between the two technologies and the use of either product was not associated with increased WHO bleeding of Grade 2 or greater compared to untreated PLTs. This observation implies that decisions about choice of technology should not be primarily based on concerns about PLT efficacy in stable hematology patients receiving prophylactic transfusions. It is important to recognize that the apparent lack of impact of PI on bleeding is, at this time, only applicable to stable hematology patients. It is not known whether this suggestion of efficacy is true in actively bleeding patients and concerns have been raised over making assumptions in the absence of data to support use of PI products in this patient population.7 The second notable finding of IPTAS was that the overall use of blood products, both PLTs and RBCs, was markedly increased by approximately 50 and approximately 25%, respectively, and this finding adds another important consideration to any decision around widespread adoption of PI. Would PLT requirements for a treatment course in a hematology-oncology patient really increase by up to 50% if universal application of PI was applied to a country's PLT inventory? For countries that already experience periodic PLT shortages, this observation raises a daunting prospect. Concern may be somewhat lessened by the experience of countries that have already gone to full implementation of PI for their PLT products. Not all have seen a significant increase in PLT demand8, 9 presumably due in part to hospital routine around when transfusions are ordered and given for hematology-oncology patients and partly because our understanding of PLT dosing is rudimentary as evidenced by the PLADO trial.10 For blood operators some challenges with the use of the current versions of PI systems have been identified. For example, there are relatively tight upper and lower boundaries for PLT content in the treated unit, and this protocol gives rise to a proportion of collected PLTs that cannot be subjected to PI treatment. The systems that are currently in the marketplace may not be configured to match the way that inventory is generated. For example, the Intercept system for apheresis PLT PI is designed to work with single or double collections, but not with the triples that make up an increasing proportion of US collections as they are more cost-effective than doubles or singles. Additionally, current PI systems are not available for PRP PLTs that are still made in some countries including the United States. At present, these issues have not been fully addressed, and they serve as an impediment to full implementation for some blood operators. Additional aspects of decision making for PI use include the role of regulatory authorities who may make their own decisions obviating the need for clinicians or blood operators to decide or creating unanticipated drivers. For example, in the United States, moving toward a requirement to take additional action to reduce bacterial contamination risk in PLTs beyond 3 days of storage may create additional demand to implement PI since the proposed alternative is testing on the day of use. Other considerations around recipient and clinician choice may come into play. In some jurisdictions, PI of PLT products has been mandated by law (e.g., Belgium and Switzerland) and although this requirement may remove an element of product choice, it simultaneously removes the vagaries of managing a dual inventory of treated and untreated PLT components. That said, there are significant concerns among some treaters about the safety profile of treated components owing to the presence of residual additives such as amotosalen or riboflavin by-products and about their clinical efficacy. While the available literature does not support the safety concerns, concerns about a reduction in PLT efficacy are supported by an increase in the frequency of transfusion in several studies, including IPTAS, but without evidence of increased bleeding in the recipients of PI PLTs. These findings contribute to the challenge of decision making. For many countries where the serious adverse events associated with transfusion are very rare owing to other risk mitigation strategies in place, it is difficult for treaters to balance a further increase in risk reduction with possible increased requirement for PLT support at least for patients receiving prophylactic PLT transfusions. The near absence of data on the use of PI-treated components in bleeding patients further hampers clinical decision making. One other notable feature of the IPTAS study is that it brought together the investigators and two companies that aggressively compete in the PI space to work together on IPTAS. The companies involved, Cerus and TerumoBCT, are to be commended for engaging in a collaborative research study even though it is not the norm for technology assessment studies or clinical trials. This example demonstrates a means to counteract a current challenge to the continued progress of transfusion science. For most countries, government investment in transfusion medicine research has declined over the past two decades. This reality is paralleled by either a decline or at best a lack of increased investment in research and development by blood operators in the high-development-index countries as their profits continue to decline with falling demand for RBCs. As organizations collecting and processing blood donations continue to demand the lowest possible price for the supplies used in producing components, the companies that manufacture these products see a decrease both in the price per unit sold and in the total volume of product sold with a concomitant erosion in profits that can be plowed back into research and development. We have inadvertently created a very challenging environment for conducting the studies that are necessary to ensure the best decision making for blood safety. While the IPTAS study provides useful information to feed into decisions around PI implementation for Italy, it highlights the need for more research to help in decision making and for thoughtful analysis of the available data to determine whether findings can be readily extrapolated from one jurisdiction to another. We also clearly need studies to help us understand whether concerns over the use of PI-treated components in bleeding patients are warranted and other determinations of whether there are specific groups of patients for whom PI-treated PLTs would create more risk than they mitigate. The concept of PI of fresh blood components is highly attractive and we must continue to work to resolve the uncertainties around its use. Data are necessary to make the best decisions. The author received research funding for pathogen inactivation studies from TerumoBCT and Macopharma. Dana V. Devine, PhD1,2 e-mail: dana.devine@blood.ca 1Canadian Blood Services 2Centre for Blood Research and the Department of Pathology & Laboratory Medicine University of British Columbia Vancouver, BC, Canada

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.