Beginning in the early 1950s, physicians began to publish case reports and small case series describing a new clinical disease syndrome that appeared to be related to the transfusion of whole blood and blood products.1–3 This syndrome, marked by acute onset of hypoxemia, pulmonary edema, and fever, quickly became distinct from the well-described circulatory overload that was known to occasionally result after blood transfusion.4 Although reports were initially sporadic and most patients recovered uneventfully, this new syndrome gripped the attention of the medical community. During the next half-century, this syndrome of transfusion-related acute lung injury (TRALI) stirred the concerns of physicians and surgeons across the globe. The fear of TRALI strongly contributed to the limiting of plasma transfusions and led physicians to resuscitate hemorrhaging patients with red blood cells (RBCs) and crystalloid alone.5,6
The advent of damage control resuscitation in the past decade has swung the pendulum away from the “supranormal” crystalloid resuscitation of the 1980s and back to resuscitation with blood products. However, there is lingering resistance to this paradigm shift in the medical community due to fear of transfusion complications like TRALI. In fact, although TRALI is repeatedly named as the primary cause of transfusion-related death by the Food and Drug Administration,7 the incidences of TRALI and TRALI-related mortality have declined consistently since 2006 based on American Red Cross (ARC) surveillance data.8,9 Accordingly, because blood banks have moved toward male-predominant plasma to reduce exposure to implicated human leukocyte antigen (HLA) antibodies, TRALI appears, anecdotally, to have been reduced to an extremely rare event.8–11 These most recent reports estimate the incidence of TRALI between 1 in 4000 and 1 in 40,000 blood units transfused.
Given our research and quality improvement initiatives in massive transfusion and optimal blood product ratios, we proposed to review our institutional experience with TRALI and its evolution over the course of the past decade. Our hospital and its trauma center are well known for early aggressive use of blood products in the resuscitation of trauma patients, including placing RBCs and liquid plasma on our prehospital helicopters.12 The purpose of this study was to estimate the current incidence of TRALI at a large, urban center.
This article was structured to adhere to the relevant Enhancing the QUAlity and Transparency Of health Research (EQUATOR) guidelines.
Memorial Hermann Hospital (MHH) at Texas Medical Center is an American College of Surgeons–verified level-1 trauma center that is the primary teaching hospital for the University of Texas McGovern Medical School in Houston. MHH is home to the Red Duke Trauma Institute, and is one of only 2 level-1 trauma centers in Houston, Texas, the fourth-largest city in the United States. The hospital is an 850-bed facility within Texas Medical Center with busy trauma, cardiovascular services, neuroscience, orthopedics, women’s health, general surgery, emergency medicine, and organ transplantation services.
After approval from the University of Texas Health Science Center and MHH institutional review boards, the MHH Blood Bank Registry was queried for all transfused patients admitted between September 2002 and March 2013. The requirement for written consent was waived due to the large, retrospective nature of the study and its minimal risk. The MHH Blood Bank routinely collects and investigates all cases of reported, possible, or recalled TRALI. These cases are then independently reviewed by transfusion medicine physicians and classified as either “probable TRALI” or “unrelated etiology” after detailed evaluation of the transfusion reaction. Diagnoses of TRALI were made using the Canadian Consensus panel definition: (1) acute lung injury (ALI) occurring within 6 hours of transfusion; (2) objective or clinical evidence of hypoxia; (3) presence of bilateral interstitial infiltrates on chest radiograph; and (4) absence of circulatory overload.13 Recognizing that passive reporting strategies tend to underestimate the incidence of the disease, we also chose to include patients for whom donor product was recalled for having HLA antibodies with a high reactivity level (and therefore the potential to cause TRALI), even if all criteria in the Canadian Consensus definition were not met. Gulf Coast Regional Blood Center provided all blood products as well as the recall information. All units of blood for each study group were typed, cross-matched, and leukoreduced where applicable and underwent standard infectious disease testing.
The incidence of TRALI was estimated with respect to all transfused blood products as well as to plasma alone. The individual cases were then reviewed and compared with specific attention to age and sex of the patient, type and number of blood products transfused, diagnosis/procedure, blood groups, the sex of the blood product donor, and mortality. Finally, the total number of units transfused at our facility during this time period was also obtained, allowing the incidence of TRALI to be estimated. Confidence intervals (CIs) for incidence calculations were determined using exact methods for proportions using STATA Statistical software (version 14; StataCorp, College Station, TX).
Description of TRALI Cases
By the Canadian consensus definition, 7 cases of TRALI were identified at MHH during the time period beginning September 1, 2002 and ending March 31, 2013. An additional 8 patients who had donor product recalled were also identified in the analysis. Although these 8 patients did not meet the full consensus definition of TRALI, all had clinical or laboratory evidence of ALI after transfusion of blood products with highly reactive HLA antibodies. The most common factor that excluded the consensus diagnosis of TRALI was the absence of bilateral interstitial infiltrates on chest radiograph. A broad range of patients was affected across a wide group of medical services. Patients ranged in age from 10–84 years (Table 1). Both genders were affected, with women affected more frequently than men (9 vs 6 cases). TRALI occurred in both nontype-specific (4 cases) and type-specific (11 cases) blood product transfusion. TRALI tended to occur more frequently when the donor was female (8 cases), but it did occur with male donors as well (5 cases in male-only donation, 2 cases in mixed male/female donation). In 1 patient, TRALI was the proximate cause of death; the remaining 14 patients survived to discharge.
TRALI cases were found to occur with each of the major blood components (ie, plasma, platelets, and RBCs), though most occurred in transfusion of plasma alone or in combination with another product (Table 2). When TRALI occurred in the setting of plasma transfusion alone, it occurred more often when the plasma was from a female donor (6 out of 8 cases). From a clinical perspective, the vast majority of patients presented with acute onset of hypoxemia (Table 3). Dyspnea and bilateral interstitial infiltrates on chest radiograph were present in only about half of cases.
Incidence of TRALI
During the study time period, a total of 714,757 units of RBCs, platelets, plasma, and cryoprecipitate were transfused to patients admitted to MHH (Table 4). Seven cases met the Canadian consensus definition for TRALI. An additional 8 cases of ALI did not meet the full consensus definition but were identified to have received blood products with highly reactive HLA antibodies. When considering only the patients who met the full Canadian consensus definition, the incidence of TRALI with respect to all blood products was estimated to be 0.10 cases per 10,000 units of blood product (95% CI, 0.04–0.20). Of these, 5 could be attributed to plasma, with an estimated incidence of 0.28 cases per 10,000 units of plasma (CI, 0.09–0.65). When the additional 8 ALI patients were included in the analysis, the incidence was estimated to be 0.21 cases per 10,000 units of blood product (95% CI, 0.12–0.35). Nine cases could be attributed directly to plasma, with an estimated incidence of 0.50 cases per 10,000 units of plasma (CI, 0.23–0.95).
In the United States, the plasma donor pool was transitioned to male-predominant plasma in 2007 to minimize exposure to highly reactive HLA alloantibodies. For this reason, the incidences were also estimated for the pretransition time period (2002–2006) and the posttransition time period (2008–2013). When considering only the patients who met the full Canadian consensus definition, the 2002–2006 incidence was estimated to be 0.11 cases per 10,000 units of all blood products (CI, 0.02–0.33) and 0.29 cases per 10,000 units of plasma (CI, 0.03–1.04). From 2008 to 2013, these incidences remained similar, at 0.08 cases per 10,000 units of all blood products (CI, 0.02–0.24) and 0.35 cases per 10,000 units of plasma (CI, 0.07–1.03). When the additional 8 ALI patients were included in the analysis, the 2002–2006 incidence was estimated to be 0.34 cases per 10,000 units of all blood products (95% CI, 0.15–0.64), and the incidence that was directly attributable to plasma was estimated to be 0.72 cases per 10,000 units of plasma (95% CI, 0.23–1.7). From 2008 to 2013, the incidence was estimated to be 0.11 cases per 10,000 units of all blood products (95% CI, 0.02–0.35), and the incidence that was directly attributable to plasma was estimated to be 0.35 cases per 10,000 units of plasma (95% CI, 0.07–1.0).
The incidence of TRALI at our large, urban tertiary medical center during the past decade was low and comparable to previously published studies. Despite the fact that >700,000 units of blood products were administered during this time period, there were only 7 documented cases of TRALI (incidence 0.10 cases per 10,000 units of product or 1 case per 100,000 units of product). When incidence calculations were limited to cases confirmed or suspected to be the result of plasma administration, the incidence was estimated to be 0.28 cases per 10,000 units of plasma (1 case per 35,714 units of plasma). When the analysis was expanded to include all patients with ALI who had received blood with highly reactive HLA antibodies, incidence remained low, at 0.21 cases per 10,000 units of all blood products or 0.50 cases per 10,000 units of plasma.
The true incidence of TRALI is difficult to establish. Historically, this was due to an inconsistency in the clinical definition and the challenges involved in establishing a direct causal relationship with blood products. For this reason, estimates of incidence have varied widely throughout the literature. In 2004, the “Toward and Understanding of TRALI” consensus conference was convened, and agreement was reached regarding clinical and diagnostic criteria necessary for a diagnosis of probable TRALI. In addition, the consensus committee reviewed all available published and unpublished data and reported estimates ranging from 0.11 to 21 per 10,000 units of plasma, from 0.13 to 1.2 per 10,000 units of platelets, and from 0.02 to 2.5 per 10,000 units of RBCs.13 Interestingly, more recent research has suggested that many cases of ALI occurring within 6 hours of blood transfusion are, in fact, attributable to acute respiratory distress syndrome related to trauma or sepsis or some other contemporaneous risk factor for lung injury.14
The recognition of an increased incidence of TRALI that is directly attributable to plasma administration has helped to bolster the hypothesis that TRALI is caused (at least in part) by the presence of HLA alloantibodies within the blood products.15 HLA antibodies occur as a result of sensitization that occurs during pregnancy, transfusion, or transplantation. As a result, the majority of antibody-positive blood donors are multiparous women.16,17 This recognition led to a shift toward male-predominant plasma in 2003 in the United Kingdom. The 10-year report from the Serious Hazards in Transfusion hemovigilance program compared cases of TRALI or probable TRALI that occurred in the United Kingdom before and after the transition. It identified a dramatic decrease in the incidence of TRALI due to plasma (from 0.15 per 10,000 in 1999–2003 to 0.032 per 10,000 in 2004–2006).18 A similar reduction was noted for TRALI due to platelets (from 0.14 per 10,000 in 1999–2003 to 0.058 per 10,000 in 2004–2006).17
This study prompted a similar change in blood product collection and distribution by ARC in 2007.5 The initial ARC Hemovigilance Program report demonstrated an 80% reduction of the incidence in TRALI due to fresh frozen plasma after the implementation of a male-predominant plasma strategy. These findings were supported by a 4-year study at University of California, San Francisco (UCSF) and the Mayo Clinic, where a 68% decrease in incidence was demonstrated between 2006 (before transition to male-predominant plasma) and 2009 (after transition to male-predominant plasma).8 Likewise, a review of national data from ARC published in 2013 compared the incidence of TRALI before and after the 2007 transition.9 It found that the overall incidence of TRALI decreased dramatically, from 18.6 cases per million units of plasma in 2006 to 4.2 cases per million units in 2008–2011. This drop was driven largely by a decrease in incidence in blood groups A, B, and O, for which male donors exceeded 99%. The incidence of TRALI cases was essentially unchanged in AB plasma from 2006 to 2008–2011, for which females continued to comprise 40% of the donor pool. In yet another multicenter study, the combined incidence of TRALI and possible TRALI dropped from about 39 cases per million units in the 16 months before the 2007 transition to 4.2 cases per million units in the 16 months after.19
This shift in the donor pool is particularly important when considering the recent changes in management of patients in hemorrhagic shock. Because clinicians continue to move away from a crystalloid-based resuscitation strategy and toward a balanced blood product transfusion strategy, earlier and increasing volumes of plasma and platelets are being administered nationwide. Interestingly, despite the increased use of blood products (especially plasma and platelets), cases of TRALI remain exceedingly rare. The recent Pragmatic, Randomized Optimal Platelet and Plasma Ratios trial investigated the use of 2 resuscitation strategies for patients with hemorrhage.20 In that study, 680 severely injured patients from 12 level-1 trauma centers in North America who were predicted to require massive transfusion were randomized to a platelet:plasma:RBC product ratios of 1:1:1 or 1:1:2 during active resuscitation. While these groups were already at high risk, given their large transfusion requirements and severe injury, no transfusion-related deaths were identified. In fact, even with an increased use of plasma and platelets in the 1:1:1 group, no safety differences (including TRALI, transfusion-associated circulatory overload, or other transfusion-associated complications) were identified between the 2 groups.
Limitations to the present study include the single center and retrospective design of the study. While cases were collected prospectively, other data obtained were collected via a trauma registry database and electronic medical record. In addition, the population is not homogenous, but rather, it reflects the diverse patient population that received blood products at our facility. However, the most relevant limitation is that we were unable to account for investigational and reporting biases. Cases in the present study include both clinically reported TRALI cases that met the consensus definition, as well as serologically recalled cases. Many of the serologically recalled cases did not meet the full consensus definition for TRALI. Moreover, patients who died before any cause–effect may have been missed by our hospital reporting and investigation process. Indeed, TRALI is generally considered to be underreported throughout the literature due to the wide variation in clinical presentation and the complex, incompletely-understood mechanisms by which it operates.21
TRALI remains a major cause of transfusion-related morbidity, as documented by the Food and Drug Administration. However, the risk to an individual patient or even a high-risk population of patients is incredibly low (estimated incidence 0.04–0.20 cases per 10,000 units of product). Even at a high-volume level-I trauma center that is known for aggressive use of blood products, new cases of TRALI occurred rarely. When TRALI did occur, only 3 of 7 were trauma patients, and all patients survived to discharge. Although the risks associated with transfusion are small, a restrictive transfusion policy continues to be recommended best practice. However, fear of TRALI should not deter physicians from even the most aggressive blood product resuscitation policy when clinically indicated.
Name: David E. Meyer, MD, MS.
Contribution: This author helped with data analysis and interpretation, drafting the manuscript, and critical manuscript revisions.
Name: Jacob W. Reynolds, MD.
Contribution: This author helped with study conception and design, data acquisition, data analysis and interpretation, and drafting the manuscript.
Name: Rhonda Hobbs, MT (ASCP).
Contribution: This author helped with data acquisition.
Name: Yu Bai, MD, PhD.
Contribution: This author helped with study conception and design, data acquisition, and data analysis and interpretation.
Name: Beth Hartwell, MD.
Contribution: This author helped with study conception and design, data acquisition, and data analysis and design.
Name: Matthew J. Pommerening, MD.
Contribution: This author helped with analysis and interpretation of the data.
Name: Erin E. Fox, PhD.
Contribution: This author helped with statistical analysis and interpretation of the data.
Name: Charles E. Wade, PhD.
Contribution: This author helped with study conception and design, and data analysis and interpretation.
Name: John B. Holcomb, MD.
Contribution: This author helped with study conception and design, data analysis and interpretation, and critical manuscript revisions.
Name: Bryan A. Cotton, MD, MPH.
Contribution: This author helped with study conception and design, data acquisition, analysis and interpretation of the data, drafting the manuscript, and critical manuscript revisions.
This manuscript was handled by: Avery Tung, MD, FCCM.
1. Barnard RD. Indiscriminate transfusion: a critique of case reports illustrating hypersensitivity reactions. N Y State J Med. 1951;51:2399–2402.
2. Brittingham TE. Immunologic studies on leukocytes. Vox Sang. 1957;2:242–248.
3. Philipps E, Fleischner FG. Pul monary edema in the course of a blood transfusion without overloading the circulation. Dis Chest. 1966;50:619–623.
4. Ward HN. Pulmonary infiltrates associated with leukoagglutinin transfusion reactions. Ann Intern Med. 1970;73:689–694.
5. Cotton BA, Guy JS, Morris JA Jr, Abumrad NN. The cellular, metabolic, and systemic consequences of aggressive fluid resuscitation strategies. Shock. 2006;26:111–116.
6. Hess JR, Hiippala S. Optimizing the use of blood products in trauma care. Crit Care. 2005;9(suppl 5):S10–S14.
8. Eder AF, Herron RM Jr, Strupp A, et al. Effective reduction of transfusion-related acute lung injury risk with male-predominant plasma strategy in the American Red Cross (2006-2008). Transfusion. 2010;50:1732–1742.
9. Eder AF, Dy BA, Perez JM, Rambaud M, Benjamin RJ. The residual risk of transfusion-related acute lung injury at the American Red Cross (2008-2011): limitations of a predominantly male-donor plasma mitigation strategy. Transfusion. 2013;53:1442–1449.
10. Eder AF, Herron R, Strupp A, et al. Transfusion-related acute lung injury surveillance (2003-2005) and the potential impact of the selective use of plasma from male donors in the American Red Cross. Transfusion. 2007;47:599–607.
11. Toy P, Gajic O, Bacchetti P, et al.; TRALI Study Group. Transfusion-related acute lung injury: incidence and risk factors. Blood. 2012;119:1757–1767.
12. Holcomb JB, Donathan DP, Cotton BA, et al. Prehospital transfusion of plasma and red blood cells in trauma patients. Prehosp Emerg Care. 2015;19:1–9.
13. Kleinman S, Caulfield T, Chan P, et al. Toward an understanding of transfusion-related acute lung injury: statement of a consensus panel. Transfusion. 2004;44:1774–1789.
14. Toy P, Bacchetti P, Grimes B, et al. Recipient clinical risk factors predominate in possible transfusion-related acute lung injury. Transfusion. 2015;55:947–952.
15. Silliman CC, Fung YL, Ball JB, Khan SY. Transfusion-related acute lung injury (TRALI): current concepts and misconceptions. Blood Rev. 2009;23:245–255.
16. Triulzi DJ, Kleinman S, Kakaiya RM, et al. The effect of previous pregnancy and transfusion on HLA alloimmunization in blood donors: implications for a transfusion-related acute lung injury risk reduction strategy. Transfusion. 2009;49:1825–1835.
17. Densmore TL, Goodnough LT, Ali S, Dynis M, Chaplin H. Prevalence of HLA sensitization in female apheresis donors. Transfusion. 1999;39:103–106.
18. Chapman CE, Stainsby D, Jones H, et al.; Serious Hazards of Transfusion Steering Group. Ten years of hemovigilance reports of transfusion-related acute lung injury in the United Kingdom and the impact of preferential use of male donor plasma. Transfusion. 2009;49:440–452.
19. Arinsburg SA, Skerrett DL, Karp JK, et al. Conversion to low transfusion-related acute lung injury (TRALI)-risk plasma significantly reduces TRALI. Transfusion. 2012;52:946–952.
20. Holcomb JB, Tilley BC, Baraniuk S, et al. Safety and effectiveness of damage control resuscitation: results of the Pragmatic, Randomized Optimal Platelet and Plasma Ratios (PROPPR) trial. JAMA. 2015;313:471–482.
21. Silliman CC, Boshkov LK, Mehdizadehkashi Z, et al. Transfusion-related acute lung injury: epidemiology and a prospective analysis of etiologic factors. Blood. 2003;101:454–462.