Context.—Transfusion-associated graft-versus-host disease is a rare, often fatal complication of cellular blood product transfusion. The requirement that at-risk groups receive irradiated products reduces the incidence of transfusion-associated graft-versus-host disease. A comprehensive survey of irradiation practices has not been performed since 1989; meanwhile, new indications for irradiation have emerged.
Objective.—To assess current irradiation practices at College of American Pathologists member institutions. Changes in irradiation practice indicated by comparing results of a survey of irradiation practices in 1989 with those of a survey performed in 2014 may reveal how the field has developed and what areas (if any) remain to be improved.
Design.—A supplemental College of American Pathologists survey was sent out with questions regarding irradiation practices for specific conditions and circumstances. The questions included conditions for which irradiation is generally considered required for the prevention of transfusion-associated graft-versus-host disease as well as those not considered to be a special risk.
Results.—An average of 2100 organizations responded to each question regarding their irradiation practices. Irradiation for transfusion from blood relatives, human leukocyte antigen–matched products, preterm infants, and Hodgkin disease were the most common indications cited. A few organizations had universal irradiation, whereas others irradiated products by floor/unit or by service.
Conclusions.—For some at-risk populations irradiation of cellular blood products is more common than in 1989, whereas for others this practice has been reduced. Although gains have been made since the last national survey of irradiation practices, work remains to eliminate the possibility of transfusion-associated graft-versus-host disease from known at-risk populations.
Transfusion-associated graft-versus-host disease (TA-GVHD) is an uncommon yet highly fatal complication of cellular blood product transfusion including red blood cell, platelet, and granulocyte products. Development of TA-GVHD requires that the blood product contain immunologically competent lymphocytes. Importantly, the recipient must express tissue antigens absent in the donor and the recipient must also be incapable of mounting an effective immune response to destroy the foreign lymphocytes. In cases of TA-GVHD, donor lymphocytes replicate and attack the recipient (host), who in turn is unable to mount an adequate immune response. Transfusion-associated graft-versus-host disease is defined by the US National Healthcare Safety Network as fever, rash, hepatomegaly, diarrhea between 2 days and 6 weeks after transfusion with laboratory evidence of liver dysfunction, pancytopenia, leukocyte chimerism, and findings of TA-GVHD on skin or liver biopsy.1 In spite of the well-documented constellation of signs and symptoms associated with the development of TA-GVHD, the condition likely remains underrecognized. Of particular importance, evidence of TA-GVHD may be confounded by the patient's underlying condition or incorrectly attributed to the effects of chemotherapy or infection.2 Cellular blood products (including leukocyte-reduced blood products) contain viable lymphocytes that can proliferate and result in TA-GVHD. Inactivation of these lymphocytes, usually through gamma irradiation, but also through x-ray and pathogen inactivation technologies, prevents TA-GVHD.
Although donor lymphocytes were implicated in the development of TA-GVHD as long ago as the 1950s,3 a complete understanding of the factors contributing to TA-GVHD remains elusive. Cellular blood product recipients with impaired immunity were initially recognized as being at risk for TA-GVHD. Subsequently, recognition of the development of TA-GVHD as a result of human leukocyte antigen (HLA) similarities between the donor and recipient added to the understanding of the development of the condition.4 In 2015, a systematic review of reported cases of TA-GVHD was published. This review supported the notion that TA-GVHD occurs primarily as a consequence of transfusion of viable donor lymphocytes; however, rather than occurring primarily in immunocompromised patients, this work favored HLA haplotype homozygosity as the dominant mechanism for development of TA-GVHD. Intriguingly, this study found that most patients who went on to develop TA-GVHD “did not have an underlying diagnosis conferring immune compromise.” 5 Nonetheless, identification of individuals at high risk for TA-GVHD, such as patients with specific immune impairments or those receiving products from relatives, and the subsequent requirement that these individuals receive irradiated products, remains an important consideration in relation to reducing the risk of TA-GVHD.
Despite the high mortality associated with TA-GVHD and the fact that irradiation dramatically reduces this risk, relatively few surveys of irradiation for the prevention of TA-GVHD have been undertaken in the United States. One notable work published in 1990 surveyed 2250 American Association of Blood Banks (AABB) institutional members.6 In the approximately 25 years that have passed since this survey was performed, it has become increasingly apparent that certain immune-deficient patients (such as those with human immunodeficiency virus [HIV] or acquired immune deficiency syndrome [AIDS]) are not at risk for the development of TA-GVHD. On the other hand, cases of TA-GVHD have been reported in the context of the administration of new therapies such as fludarabine,7 a treatment not available at the time of the AABB survey. With this in mind, our goal was to assess current irradiation practices through surveys completed by College of American Pathologists (CAP) member institutions. By comparing the differences between data from a survey of irradiation practices in 1989 and those of our own survey (performed in 2014) the objective was to reveal how irradiation practices have developed and what areas (if any) remain to be improved.
MATERIALS AND METHODS
In 2014, supplemental surveys were sent out from the CAP Transfusion Medicine Resource Committee. The survey requested feedback regarding the practices and use of irradiated cellular blood products and was designed to cover common indications for irradiation that reflect generally considered best practices for irradiation as well as irradiation practices that may not currently be supported by the literature. The survey was sent on February 10, 2014, as supplemental questions with the CAP Transfusion Medicine comprehensive proficiency testing survey (JA-2014). This proficiency testing survey was sent to all laboratories who subscribe to the J survey service, 3447 in all. The intended audience was laboratory personnel familiar with the facility's irradiation policies. Respondents were asked to choose yes or no in regard to whether their organization's policy called for irradiation of cellular blood products for conditions such as hematologic malignancies, congenital immunodeficiency syndromes, and hematopoietic progenitor cell transplantation. Questions likewise covered irradiation for intrauterine transfusions, infants, and neonates, HLA-matched products, and transfusions from blood relatives. Additional questions involved irradiation for granulocyte transfusions, drugs that affect T lymphocytes, solid tumors, and solid organ transplants (Table 1). Finally, certain questions were designed to elucidate strategies for the implementation of irradiation policies, such as, “Does your institution perform universal irradiation?”, “Who in your institution determines who should receive irradiated blood products?”, and “Do you irradiate all blood products for the following patient age: 0–1 years, 0–6 years, all children (0–18 years)?” The results were compiled electronically and analyzed using Microsoft Excel (Redmond, Washington). Statistical data were organized into a proficiency survey testing result.
RESULTS
For each of the supplemental questions regarding irradiation practices of cellular blood products a mean number of 2100 organizations participated (range, 986–2853). Ninety-one percent of the proficiency testing surveys were domestic and 9% were international (3585 proficiency testing survey kits were distributed). The most frequent condition requiring irradiated products was transfusion from blood relatives: of 2370 respondents, 1862 (78.6%) replied yes (Table 1). The next most frequent indication was HLA-matched or partially matched products. In this case, 1552 of the 2252 respondents (68.9%) noted that they irradiated for this indication. Other frequent indications were neonatal exchange transfusions, 1460 of 2202 (66.3%); intrauterine transfusions, 1344 of 2122 (63.3%); and preterm/low-birth-weight infants, 1359 of 2200 (61.8%). A common indication cited was hematopoietic progenitor cell transplantation, 1340 of 2136 (62.7%). Only 1210 of 2122 laboratories (57.0%) responded that they require irradiation for congenital immunodeficiency syndrome. Other potentially at-risk populations were less frequently cited as requiring irradiation, including Hodgkin disease, 1113 of 2113 (52.7%); acute leukemia, 1069 of 2115 (50.5%); and lymphoma, 1007 of 2108 (47.8%). Granulocyte transfusions were irradiated per policy in only 995 of 2066 organizations (48.2%). Less common was irradiation for donors undergoing bone marrow or peripheral blood stem cell collection, 936 of 2100 (44.6%); solid organ tumors, 767 of 2079 (39.4%); infants or children with congenital heart disease secondary to possible DiGeorge syndrome, 818 of 2076 (39.3%); aplastic anemia, 765 of 2062 (37.1%); and recipient and donor pair from a genetically homogenous population, 622 of 1986 (31.3%). Lastly, few organizations, 730 of 2027 (36.0%), irradiated product for patients treated with purine analogue drugs, and fewer still, 558 of 1980 (28.2%), irradiated for treatment with alemtuzumab (anti-CD52) or antithymocyte globulin (ATG).
Some organizations indicated that they still irradiated for populations not generally considered at risk for TA-GVHD, including healthy/term newborns, 951 of 2156 (44.1%); those with solid tumors (undergoing intensive chemotherapy and/or radiotherapy), 818 of 2076 (39.4%); and those undergoing solid organ transplantation and not on anti-CD52 agents, 767 of 2079 (36.9%). Other groups considered to be at low risk for the development of TA-GVHD were less frequently included in organizations' irradiation policies, including patients with HIV/AIDS, 402 of 2063 (19.5%), and patients with autoimmune disorders, 357 of 2044 (17.5%). Some organizations indicated that they irradiated products by floor/unit, most frequently hematology/oncology unit, 1422 of 2135 (66.6%). Others provide irradiated products by service, most frequently oncology service, 1547 of 2277 (67.9%), and hematology service, 1198 of 2277 (52.6%). Finally, certain organizations provide irradiated blood products by patient age, most frequently 0 to 1 years, 844 of 956 (88.3%); and 2.6% (75 of 2853) of organizations had universal irradiation.
DISCUSSION
In the most recent data from the US Food and Drug Administration regarding transfusion-related fatalities, single cases of TA-GVHD were reported in 2010 and 2011, and none were reported from 2012 to 2014.8 Although the populations at risk of TA-GVHD have become more well defined over the years, practices in the United States regarding irradiation of cellular blood products have not been extensively studied since a 1989 AABB survey of its member institutions.6 In this survey, institutions were queried concerning whether they irradiated cellular blood product for various indications, with the responses tabulated as yes, no, or not applicable (N/A). After the N/A responses were omitted from the 1989 survey, 219 of 693 (31.6%) indicated that they irradiated for transfusion from blood relatives (Table 2). In the 1989 survey, 711 of 800 (88.9%) irradiated for first-degree relatives and 95 of 946 (10.0%) irradiated for non–blood relatives. Although the United States does not publish nationwide guidelines in the United Kingdom, the British Committee for Standards in Haematology (BCSH) Blood Transfusion Task Force has published updated guidelines as recently as 2011. Their standards state, “All donations from first- or second-degree relatives should be irradiated . . .” 10 Likewise, the CAP requires that all directed donations between blood relatives be irradiated (TRM.45270). Interestingly, in 1989 only 134 of 432 (31.0%) replied yes as to whether they irradiated for HLA-matched cellular products. By contrast, this was the second most frequent indication for irradiation in our survey, 1552 of 2252 (68.9%) (HLA-matched or partially matched products). The BCSH Blood Transfusion Task Force recommends, “All human leucocyte antigen (HLA)–selected components should be irradiated . . .” 10 whereas the AABB Standards for Blood Banks and Transfusion Services require irradiation of HLA-compatible products (Standard 5.1673.1.3).9 The underlying rationale for irradiation of HLA-matched or directed donation from blood relatives is that the product lymphocytes may recognize the recipient as foreign but the recipient may not recognize the donor lymphocytes as foreign, and this will result in TA-GVHD. In our survey, transfusion for blood relatives was the most frequent condition requiring irradiated products, 1862 of 2370 (78.6%) (Table 1).
In the 1989 AABB survey, 249 of 462 (53.9%) replied yes as to whether preterm infants were provided with irradiated cellular blood products, whereas in 2014, 1359 of 2200 (61.8%) replied that preterm/low-birth-weight infants require irradiation. Of note, the BCSH Blood Transfusion Task Force takes the stance that “[i]t is not necessary to irradiate red cells for routine ‘top-up' transfusions of premature or term infants.” 10 The BCSH cited the fact that there are few reports of TA-GVHD in preterm and full-term infants10 ; however, this does not appear to take into account the potential for developing TA-GVHD if a condition that would put the infant at risk has yet to be diagnosed. Other frequently cited indications for irradiation (not queried in 1989) were neonatal exchange transfusions, 1460 of 2202 (66.3%), and intrauterine transfusions, 1344 of 2122 (63.3%). The BCSH Blood Transfusion Task Force proposes, “All blood for intrauterine transfusion (IUT) should be irradiated,” and “It is essential to irradiate blood for neonatal exchange transfusion (ET) if there has been a previous IUT or if the donation comes from a first- or second-degree relative. For other neonatal ET cases, irradiation is recommended provided this does not unduly delay transfusion.” 10 Fetuses and neonates have immature immune systems, and cases have been reported in preterm infants after neonatal exchange transfusion and intrauterine transfusions. Apart from their having immature immune systems, another rationale for irradiation of all cellular products for preterm infants and neonates is that it may be unknown at the time if there is a congenital immune deficiency. Transfusion-associated graft-versus-host disease has been reported to occur in patients who are transfused but not recognized as having a congenital immunodeficiency syndrome.11 Unexpectedly, in our study only 1210 of 2122 laboratories (57.0%) responded that they irradiated for congenital immunodeficiency syndrome, compared with 273 of 399 (68.4%) who responded yes in the 1989 AABB survey. Even fewer in our study, 783 of 1991 (39.3%), indicated that they irradiated for infants or children with congenital heart disease secondary to possible DiGeorge syndrome (a population considered to be at risk for developing TA-GVHD as a result of T-lymphocyte impairment). One strategy developed to circumvent the risk of transfusion prior to discovery of a risk factor for development of TA-GVHD (such as a congenital immunodeficiency syndrome) is providing irradiated products by patient age.12 In our survey, for those who irradiated based on age, the most frequently designated age group was 0 to 1 years, 844 of 956 (88.3%). Although healthy/term newborns are not generally considered to be at risk for TA-GVHD, the percentage of organizations that irradiated for this group rose from 107 of 446 (24.0%) in 1989 to 951 of 2156 (44.1%) in 2014, perhaps in part because of the adoption of age-based irradiation guidelines.
The 1989 AABB survey revealed that 336 of 382 respondents (88.0%) indicated that they irradiated for recipients of allogeneic bone marrow transplants, whereas 263 of 323 respondents (81.4%) indicated that they irradiated for autologous bone marrow transplant recipients. Although this was still a relatively frequent indication for irradiation in 2014, only 1340 of 2136 institutions (62.7%) responded that they provided irradiation for hematopoietic progenitor cell transplantation. Based on multiple documented cases of TA-GVHD in this population, the European School of Haematology, the European Group for Blood and Marrow Transplantation, the Foundation for the Accreditation of Cellular Therapy, and the BCSH Blood Transfusion Task Force all recommend irradiated blood products for allogeneic and autologous hematopoietic progenitor cell recipients. Although this was not addressed in the 1989 AABB study, we found that 936 of the 2100 respondents in 2014 (44.6%) indicated that they irradiated for cellular blood products going to donors undergoing bone marrow or peripheral blood stem cell collection. The 2011 BCSH Blood Transfusion Task Force recommends, “Patients undergoing bone marrow or peripheral blood stem cell ‘harvesting' for future autologous re-infusion should receive irradiated cellular blood components during and for 7 d before the bone marrow/stem cell harvest . . .” 10 This precaution is intended to prevent viable allogenic T-lymphocyte collection and the development of subsequent TA-GVHD.
In relationship to hematologic malignancies, in the 1989 survey 169 of 497 (34.0%) responded yes as to whether they irradiated for Hodgkin disease. Despite the fact that cases have been reported in this population,13 only 1113 of 2113 (52.7%) of respondents irradiated for this condition in 2014. For non-Hodgkin lymphoma, 156 of 487 (32.0%) replied that they irradiate in 1989, compared with 1007 of 2108 (47.8%) in 2014. Regarding non-Hodgkin lymphoma, the 2011 BCSH Blood Transfusion Task Force states, “There are relatively few reports of TA-GVHD in non-Hodgkin lymphoma (NHL) and the majority have been in patients with high-grade disease.” 10 No further recommendation was specified for this group. Irradiation for leukemia fell from 299 of 582 (51.4%) in the 1989 survey to 1069 of 2115 (50.5%) in our survey. Although cases of TA-GVHD have been reported in acute leukemia patients, the risk was not considered enough to warrant a recommendation for irradiation in the 2011 BCSH Task Force recommendations. The evidence used to support these recommendations includes that as of 2011, no cases had been reported in this group in the UK.10
Some therapies addressed in the 2014 CAP survey were not included in the 1989 AABB survey. Granulocyte transfusions, which are given to neutropenic patients within 24 hours of collection and contain lymphocytes and have been identified as a potential cause of TA-GVHD, were only irradiated per policy in 995 of 2066 organizations (48.2%). The BCSH Blood Transfusion Task Force recommends, “All granulocyte components should be irradiated before issue and transfused with minimum delay.” 10 Other therapies not included in the AABB survey that have also been identified as potential causes of TA-GVHD are treatment with purine analogue drugs (fludarabine, cladribine, and deoxycoformycin) as well as alemtuzumab (an anti-CD52 agent) and ATG. Purine analog drugs result in profound lymphopenia. In the years since fludarabine became available, cases of TA-GVHD have been reported in patients with chronic lymphocytic leukemia, patients with acute myeloid leukemia, and non-Hodgkin lymphoma patients who received the drug up to 11 months prior to transfusion. The BCSH Blood Transfusion Task Force advises, “Patients treated with purine analogue drugs . . . should receive irradiated blood components indefinitely.” The Task Force likewise recommends that irradiated blood components should be used after alemtuzumab and ATG therapy. At the time of our survey, adoption of irradiation for purine analogues was only seen in 730 of 2027 (36.0%) institutions whereas irradiation for treatment with alemtuzumab or ATG was 558 of 1980 (28.2%). In terms of aplastic anemia, 765 of 2062 (37.1%) responded that this is an indication for irradiation in their institution. The BCSH Blood Transfusion Task Force recommends the use of irradiated blood components for aplastic anemia in patients receiving immunosuppressive therapy with ATG; however, it does not address the use of irradiated blood product in aplastic anemia in any other context.
The BCSH Blood Transfusion Task Force also takes the position that “[i]t is not necessary to irradiate blood components for patients undergoing routine surgery, those with solid tumors, HIV infection, autoimmune diseases or after solid organ transplantation . . .” and some of these groups were less frequently included in individual organizations' irradiation policies in 2014 when compared with the 1989 survey results. For instance, 116 of 473 of those polled in 1989 (24.5%) indicated that they irradiated for HIV/AIDS, whereas in 2014 only 402 of 2063 of the responding organizations (19.5%) replied yes to whether this was an indication. Likewise, in 1989, 136 of 337 institutions (40.4%) irradiated for recipients of solid organ transplantation, whereas irradiation for the same group (not on anti-CD52 agents) fell slightly to 767 of 2079 (36.9%) in 2014. Curiously, the converse was true for patients with solid organ tumors, where the percentage of those providing irradiated cellular blood product for the prevention of TA-GVHD nearly doubled, from 95 of 474 (20.0%) in 1989 to 818 of 2076 (39.4%) in 2014. A potential contributor to this increase in irradiation for solid organ tumors is the adoption of the strategy to provide irradiated products by floor or unit, most frequently hematology/oncology, 1422 of 2135 (66.6%), or by service, most commonly oncology, 1547 of 2277 (67.9%).
For some conditions where irradiation of cellular blood product is widely considered required in order to prevent TA-GVHD, there were improvements between 1989 and 2014. Based on the results of our survey, irradiation of cellular blood products for HLA-matched cellular products, preterm/low-birth-weight infants, Hodgkin disease, and non-Hodgkin lymphoma increased (compared with the 1989 AABB survey). Whereas this represents potential progress in irradiation practices (particularly in the case of Hodgkin disease), irradiation of cellular blood product for the prevention of TA-GVHD would appear to remain underused in other groups well documented to be at risk for TA-GVHD. For conditions where there is a general consensus that the irradiation of cellular blood product should be required in order to prevent TA-GVHD, the percentage of those who irradiated actually dropped. Surprisingly, irradiation for leukemia fell slightly when compared with the 1989 AABB survey, as did irradiation for congenital immunodeficiency syndromes. There was slight progress in terms of not mandating irradiation of cellular blood product for conditions that have no known risk for TA-GVHD. For instance, TA-GVHD has not been reported in HIV/AIDS patients. Accordingly, the percentage of those providing irradiated products for HIV/AIDS patients fell in relation to the results obtained in 1989. Likewise, fewer institutions indicated that patients undergoing solid organ transplantation were required to receive irradiated cellular blood products.
Despite recognition of some of the underlying risks predisposing patients to TA-GVHD, rare cases of TA-GVHD continue to be reported, simply based on the chance that a random blood donor will share enough similarity in HLA type that the recipient will fail to recognize the donor lymphocytes as foreign (thus allowing them to proliferate and attack the host).14 The risk of such an event increases in populations where there is less HLA diversity, such as occurs in Japan.15 This understanding has led to more stringent guidelines for irradiation of cellular blood product in Japan than have generally been adopted elsewhere in the world. Interestingly, only 31.3% of respondents in our survey indicated that they irradiated when the recipient and donor pair came from a genetically homogeneous population. Cases of TA-GVHD have been reported in patients who were not initially considered to be at risk of TA-GVHD; however, in retrospect the underlying condition predisposing the patient to developing TA-GVHD was not identified prior to the patient's receiving cellular blood product. In order to mitigate this risk, some have advocated for universal irradiation of cellular products.16 In our study we found that 2.6% of organizations currently universally irradiate cellular products, theoretically reducing the risk of TA-GVHD to near zero.
Notwithstanding careful selection of questions aimed at elucidation of current practices for irradiation of cellular blood product, there were several limitations to this study. Although a laboratory director was required to sign off on this survey, the survey was not designed to determine the specific role or the level of training of the respondents. Thus, it is difficult to be certain to what extent the answers may reflect a misunderstanding regarding a particular institution's irradiation practices. To the extent that this is the case, respondents' lack of familiarity with their own institutions' indications for irradiation of cellular blood products cannot be excluded as contributors to the unexpectedly low rates of irradiation for conditions or circumstances widely thought to put patients at risk for developing TA-GVHD. Importantly, there are differences in the 2 irradiation practice surveys performed approximately 25 years apart. For instance, the 1989 AABB survey allowed response in an N/A category, a choice that was not included in the CAP supplemental survey. In order to compare the 2 studies, the percentages of respondents who answered yes or no had to be recalculated with the N/A responses omitted. Those who were unsure of their irradiation practices or misunderstood the question could choose N/A, whereas in the 2014 survey they were forced to choose yes or no for most queries. Slight differences in the wording of questions or grouping of diagnostic categories may also make it challenging to directly compare the 2 surveys. The 1989 AABB survey divided irradiation for relatives into all family members, first-degree blood relatives, and non–blood relatives, whereas our study simply used the term blood relatives. This could conceivably skew the comparison between practices then and now. In our survey, we included partially matched products (platelets) along with HLA-matched products, possibly capturing more of those who irradiated platelets when compared with 1989. In the most recent survey, we specified irradiation for acute leukemia and non-Hodgkin lymphoma (“and other hematologic malignancies”); in 1989 these conditions were grouped with the more general diagnoses of leukemia and non-Hodgkin lymphoma. This difference in wording could possibly reveal more organizations that irradiated for leukemias in general in 1989 when compared with “acute leukemia” in 2014. The opposite is true for non-Hodgkin lymphoma with the addition of “other hematologic malignancies” to this answer choice in 2014, which could potentially have swayed more respondents to choose yes for this category than would have been inclined to do so in 1989.
The United States does not provide national guidelines on the use of irradiated blood components. Although certain North American organizations, such as the AABB, the CAP, and the Foundation for the Accreditation of Cellular Therapy, provide suggestions, in the absence of countrywide standards indications for irradiation are left up to individual institutions to decide. Comparing responses collected in 1989 and 2014 with one another and then contrasting the responses with the 2011 BCSH Blood Transfusion Task Force guidelines, a mixed picture emerges. Although some strides have been made, a surprising number of organizations do not currently irradiate cellular blood product for groups considered to be at risk for the development of TA-GVHD. Unexpectedly, for some at-risk groups fewer organizations required irradiation in 2014 when compared with 1989. There was a slight decrease in irradiation for a few groups who were not considered at risk, whereas the percentage of those irradiating blood product for solid organ tumors (a group also not considered to be at risk) nearly doubled. Meanwhile, a small cohort of organizations in the United States have adopted universal irradiation practices. Intriguingly, despite the high fatality rate associated with the development of TA-GVHD and findings that would appear to reflect that irradiation of cellular blood products is less than uniform for at-risk populations, there are relatively few recent reports of deaths attributed to TA-GVHD in the United States.8 One possible explanation for this apparent contradiction is hinted at in a recent systematic review of TA-GVHD that suggested that HLA homozygosity may be more important in relationship to the development of TA-GVHD than other previously identified risks.5
Following publication of the results of the 1989 AABB survey, irradiation practices in the United States were described as being “markedly heterogeneous.” 16 Since that time, our findings suggest that cellular blood product irradiation practices remain widely disparate across institutions. As understanding of the populations at risk for TA-GVHD evolves, new recommendations aimed at targeting the most vulnerable patients may be in order.