Adverse effects of plasma transfusion


  • This work was supported in part by grants from the NHLBI and NIDDK.

Girish N. Vyas, PhD, UCSF Box 0134, 185 Berry Street, Lobby 6, Room 2110-07, San Francisco, CA 94143; e-mail:


Plasma utilization has increased over the past two decades, and there is a growing concern that many plasma transfusions are inappropriate. Plasma transfusion is not without risk, and certain complications are more likely with plasma than other blood components. Clinical and laboratory investigations of the patients suffering reactions after infusion of fresh-frozen plasma (FFP) define the etiology and pathogenesis of the panoply of adverse effects. We review here the pathogenesis, diagnosis, and management of the risks associated with plasma transfusion. Risks commonly associated with FFP include: 1) transfusion-related acute lung injury, 2) transfusion-associated circulatory overload, and 3) allergic and/or anaphylactic reactions. Other less common risks include 1) transmission of infections, 2) febrile nonhemolytic transfusion reactions, 3) red blood cell alloimmunization, and 4) hemolytic transfusion reactions. The effects of pathogen inactivation or reduction methods on these risks are also discussed. Fortunately, a majority of the adverse effects are not lethal and are adequately treated in clinical practice.


acute lung injury


acute respiratory distress syndrome

ATR(s) =

allergic transfusion reaction(s)


brain natriuretic peptide

FNHTR(s) =

febrile nonhemolytic transfusion reaction(s)

IgA-D =


MB =

methylene blue


pathogen-inactivated or reduced plasma


transfusion-associated circulatory overload


transfusion-associated graft versus host disease

vCJD =

Creutzfeldt-Jakob disease


venous thromboembolism

Fresh-frozen plasma (FFP) utilization has increased steadily over the past two decades. In 1991, 2.3 million units of FFP were transfused in the United States versus 3.9 million in 2001.1 By2008, 4.5 million units of FFP were transfused, an 11.8% increase from 2006.2 In addition, FFP use in the United States appears to be disproportionately higher than in other developed countries.1 In the United Kingdom, there has been little change in FFP usage over the past decade, but this is in contrast to decreasing red blood cell (RBC) utilization.3 Indications for FFP transfusion, which are reflected in national and local plasma guidelines, are included in Table 1.4-7 Studies show that FFP is commonly requested for nonbleeding patients with abnormal coagulation studies. Approximately, 30% to 50% of FFP transfusions are prophylactic with or without a planned procedure.3,8-10 Despite this common practice, there is little evidence to show that either 1) prophylactic plasma transfusion is beneficial or 2) modest elevations in international normalized ratio and/or prothrombin time predict bleeding and correct with plasma transfusion.3,11-17 Furthermore, multiple studies have shown that a large proportion (up to 50%) of FFP transfusions do not follow guidelines.3,7,18,19 For pediatric patients, a recent study showed an unchanging rate of FFP use in children's hospitals over an 8-year period even though FFP is no longer recommended in many clinical scenarios.20 Recently, a panel of experts convened by the AABB to develop evidence-based guidelines for plasma transfusion concluded that current evidence supports the use of plasma for massive transfusion and warfarin-related intracranial hemorrhage, but for most other scenarios additional studies are required to establish guidelines.21 In light of increased FFP utilization, a paucity of good data for certain indications, and the high rate of inappropriate transfusions, it is crucial that clinicians understand the risks of FFP transfusion. The incidence of adverse events to plasma found in hemovigilance reports varies widely,22-25 and these data are limited by passive reporting and in many countries nonmandatory reporting. In France, where reporting is mandatory, the incidence of adverse events to plasma was 1:1700 units in 2010.22 In a recently published study of 31,329 plasma transfusions in a large US hospital, a reaction rate of 1:360 plasma units transfused was reported, but some FFP transfusions were excluded from the analysis.26 Risks commonly associated with plasma transfusion include transfusion-related acute lung injury (TRALI), transfusion-associated circulatory overload (TACO), and allergic transfusion reactions (ATRs) while more rare complications include infectious disease transmission, white blood cell (WBC)-associated risks, and RBC alloimmunization. Recent studies have also commented on FFP transfusion and overall morbidity and mortality. Here we review the aforementioned risks associated with FFP, including pathogen-inactivated or reduced plasma (PRP). In this review, FFP refers to plasma frozen within 8 or 24 hours of collection. The described complications also apply to thawed plasma, which is often used in the trauma setting.

Table 1. Indications for plasma transfusion
1. Treatment of multiple coagulation factor deficiencies in patients with bleeding or before an invasive procedure.
2. Immediate reversal of warfarin or correction of vitamin K deficiency in patients with bleeding or before an emergency invasive procedure.
3. Disseminated intravascular coagulopathy or consumptive coagulopathy in patients with bleeding.
4. Dilutional coagulopathy (i.e., massive transfusion).
5. Plasma exchange in thrombotic thrombocytopenic purpura.
6. Treatment of coagulation factor deficiencies for which concentrates are unavailable.
7. Management of rare protein deficiencies


In 2003, TRALI emerged as the leading cause of transfusion-related mortality reported to the US Food and Drug Administration (FDA).27 FFP was the most frequently implicated blood product, and the United Kingdom's Serious Hazards of Transfusion (SHOT) hemovigilance data from 2003 showed that TRALI risk per component was 6.9 times higher for FFP than for RBCs.28,29 TRALI is characterized by acute hypoxemia and noncardiogenic pulmonary edema during or within 6 hours of transfusion (Table 2).27,30,31 Most patients recover in 3 days with respiratory support, but 5% to 25% of cases are fatal.32,33 The primary mechanism of TRALI is the accumulation and activation of neutrophils within the pulmonary endothelium. Recent studies indicate that platelets (PLTs) may also play a role.34 In the threshold model of TRALI, recipient and transfusion factors must act together to overcome a certain threshold and induce TRALI.35,36 This model incorporates the “two-hit” hypothesis for TRALI, the first hit being a recipient factor that primes neutrophils on the pulmonary endothelium and the second hit being a mediator within the transfused component that activates primed neutrophils and induces a permeability edema.37 A list of recipient risk factors for TRALI can be found in Table 3. The first described transfusion mediator of TRALI was WBC antibodies. In 1985, a study of 36 cases of TRALI demonstrated antibodies to human WBC antigens in 89% of cases, mostly of donor origin.44 Since then multiple studies have supported the role of donor-derived HLA and HNA (human neutrophil antigen) antibodies in TRALI.45-49 When WBC antibodies are transfused into a patient with the cognate antigen, neutrophils within the pulmonary microvasculature agglutinate and release enzymes, reactive oxygen species, and inflammatory mediators that injure the pulmonary endothelium.47,48 HLA Class II antibodies are implicated more frequently than Class I antibodies and can indirectly activate primed neutrophils via monocyte activation and cytokine release (i.e., tumor necrosis factor-α and interleukin [IL]-1β).29,50-55 Interestingly, in a recent study, little or no risk was associated with HLA Class I antibodies.38 HNA antibodies, specifically against HNA 3a, have also been shown to be potent mediators of TRALI.55,56 An active surveillance study of TRALI in two US hospitals found that the quantity of strong cognate HLA Class II antibodies and volume of HNA antibodies in blood products were predictive transfusion risk factors for TRALI.38 Of note, numerous look-back studies of donors with WBC antibodies have demonstrated that the majority of transfused patients do not develop TRALI even when the cognate antigen is expressed.57-63 Furthermore, occasionally the implicated donor's WBC antibodies do not express specificity for recipient antigens, or donor WBC antibodies are not detected at all.44,45 For these cases, a nonimmune mechanism for TRALI has been described in which bioactive substances that accumulate during storage of cellular components (i.e., lysophosphatidylcholine, nonpolar lipids, and CD40 ligand) can provide the “second hit” to induce lung injury in primed patients.64-67 Additional studies, however, are needed to further characterize this nonimmune pathway for TRALI. Ultimately, the literature supports the fact that the majority of severe and fatal TRALI cases are in fact antibody mediated.28,49 A systematic review of studies reporting on TRALI and donor antibodies demonstrated that 1) the odds ratio (OR) for developing TRALI was 15 for patients who received a transfusion from a donor who tested positive for WBC antibodies, compared to donors who tested negative; and 2) WBC antibodies contributed to approximately 80% of all TRALI cases.68 The increasing incidence of TRALI led blood collecting facilities to implement policies to prevent antibody-mediated TRALI from high-risk products (plasma and PLTs).69 Since most donors implicated in TRALI were multiparous women and approximately 17% of female donors have WBC antibodies (risk of antibodies increases with more pregnancies), the main strategy used to mitigate TRALI was to decrease or stop production of transfusable plasma from all females or females with pregnancy history.49,70 Alternatively, for products typically in short supply, such as PLTs and AB plasma, a strategy of testing female donors with pregnancy history for HLA antibodies has also been employed to limit donor loss.71 Over the past few years, multiple publications have documented a decrease in TRALI after implementation of these strategies. Data from the American Red Cross and SHOT demonstrated a TRALI incidence of 1:51,000 to 65,000 plasma units issued before mitigation versus 1:250,000 to 317,000 after mitigation.29,72 German and Canadian hemovigilance systems also showed a decrease in the number of reported TRALI cases,73,74 and a comparative cohort study from the Netherlands showed a 33% reduction of TRALI cases after implementation of a male-only plasma strategy.75 A retrospective study in three US hospitals showed a 0.0084% risk of TRALI from plasma transfusion (47,756 plasma units transfused) in the 16 months preceding implementation of low-TRALI-risk plasma versus 0% risk in the 16 months after implementation (52,230 plasma units transfused).76 Finally, in the only large active surveillance study of TRALI, TRALI incidence went from 1:4000 blood products transfused before mitigation to 1:12,000 after mitigation.38 As expected, rates with active surveillance were higher and likely better represent true incidence. The consistent downward trend seen over these various reports strongly suggests that TRALI mitigation strategies contributed to the decrease in TRALI. Ultimately, there are now fewer fatal and nonfatal TRALI cases caused by plasma transfusion, and plasma safety with regard to TRALI risk has significantly improved.

Table 2. TRALI Consensus Criteria*
  • * 

    Adapted, with permission, from Kleinman et al.30 Criteria 1 through 4 are consistent with the American-European Consensus Conference definition of acute lung injury. Definition of hypoxemia expanded for TRALI.

  • † 

    Alternative risk factors for TRALI include but are not limited to aspiration, pneumonia, sepsis, multiple fractures, pancreatitis, shock, and cardiopulmonary bypass. If Criteria 1 through 6 are met but there is a temporal association with another ALI risk factor, case is categorized as “possible TRALI.”

1. Acute onset.
2. Hypoxemia: partial pressure of oxygen in arterial blood divided by fraction of inspired oxygen (PaO2/FiO2) ≤ 300 or SpO2 < 90% on room air or other clinical evidence of hypoxemia.
3. Bilateral infiltrates on frontal chest radiograph.
4. No clinical evidence of circulatory overload.
5. Occurs during or within 6 hr of transfusion.
6. No existing ALI before transfusion.
7. No temporal relationship to an alternative risk factor for ALI.
Table 3. Recipient risk factors for TRALI*
Risk factorORPatient cohortAuthor
  • *

     Adapted, with permission, from Sachs.35 Patient risk factors for TRALI identified in clinical studies published after 2004 are listed. All studies used the Consensus Criteria to diagnose TRALI. Adjusted ORs are included.

  •  Peak airway pressure >30 cm H2O within 12 hours after intubation and before transfusion.

  •  Adjusted OR calculated by Sachs.35

  • CABG = coronary artery bypass graft.

Chronic alcohol abuse5.9Hospitalized patientsToy et al.38
Fluid balance pre-transfusion (increment per liter)1.15  
Mechanical ventilation3.6  
Shock before transfusion4.2  
Current smoker3.4  
Liver surgery (transplant)6.7  
IL-8 concentration pre-transfusion, per 10-fold increase3.0  
End-stage liver disease31.7Patients with GI bleedingBenson et al.39
Emergency CABG17.6ICU patientsVlaar et al.40
Hematologic malignancy13.1  
Massive transfusion4.5  
Mechanical ventilation3  
History of heavy alcoholism2.7ICU patientsGajic et al.41
Liver disease2.1  
Patient ageNACardiac surgery patientsVlaar et al.42
Time on cardiopulmonary bypassNA  
None identifiedNALiver transplant patientsBenson et al.43


TACO is similar to TRALI since it is also characterized by acute respiratory distress, hypoxia, and pulmonary edema temporally associated with transfusion. However, TACO is a hydrostatic not permeability edema. Despite these different mechanisms, there is no distinct clinical finding or test that can differentiate TACO from TRALI. Nevertheless, a few features can aid the diagnosis (Table 4).77 Most patients rapidly improve with diuresis, but the mortality rate has been reported as 5% to 15%.84 Older age, younger age, and existing cardiac and/or renal dysfunction are known risk factors.77,85 Until recently, TACO has received little attention in the literature. However, in 2010, TACO was the second leading cause of mortality in the United States, and TACO cases reported to SHOT increased from 18 in 2008 to 34 in 2009 and 40 in 2010.86,87 The reported incidence of TACO ranges from less than 1% of transfusions to 8% of transfusions depending on patient population and identification method (passive or prospective observation).78,88-90 Although TACO has been reported after even a single unit of RBCs,89,90 greater transfusion volume is a risk factor for TACO independent of cardiovascular risk factors as was reported in a recent prospective cohort study.88 Other risk factors included greater plasma volume transfused, FFP ordered for anticoagulant reversal, positive fluid balance, and increased infusion rate.88 Regarding infusion rate, a rate of 1 mL/kg body weight per hour is often cited for patients at risk for TACO, but there is a lack of data on appropriate infusion rates in this setting.91,92 Plasma transfusion is a risk factor for TACO since large volumes are usually needed in adults. The recommended dose of plasma for adults is 10 to 15 mL/kg for coagulation factor replacement, and some data suggest this may even be insufficient.7,93 Narick and coworkers26 recently evaluated the rate of TACO with plasma transfusion in a large US hospital using both passive reporting and active surveillance. The rate of TACO with passive reporting over a 7-year period was 1 in 1566 plasma units transfused; however, during a 1-month period of active surveillance a rate of 1 in 68 was observed. In summary, although TACO is a potentially avoidable complication, more studies are needed to further assess patient risk factors and recommend effective preventive strategies, such as appropriate infusion rates and diuretic use in susceptible patients.

Table 4. Features to distinguish TRALI from TACO*
  • *

     Adapted, with permission, from Skeate and Eastlund.77 When confronted with a patient who develops hypoxemia and pulmonary edema within 6 hours of transfusion, the best strategy is to evaluate the above features to determine which diagnosis is most supported. Note that although the TRALI Consensus Criteria states that circulatory overload must be excluded to diagnose TRALI, TACO and TRALI may coexist. Elevated posttransfusion BNP (and/or N-terminal pro-BNP) has been shown to be a useful adjunct marker for TACO.81,82 A posttransfusion-to-pretransfusion ratio of greater than or equal to 1.5 with posttransfusion BNP of at least 100 pg/mL has a sensitivity and specificity of 81 and 89%, respectively, for TACO.81 However, one study found that BNP had limited diagnostic value in distinguishing etiology of pulmonary edema after transfusion in critically ill patients.83

Body temperatureIncrease may occurNo change
Blood pressureHypotensionIncrease in systolic blood pressure
Systolic ejection fractionDecreased or normalDecreased (>45% and no severe valvular disease)78
Chest x-rayBilateral infiltratesBilateral infiltrates enlarged heart (vascular pedicle width >70 mm and cardiothoracic ratio >0.55)78
Pulmonary edema fluid/plasma protein ratio79,80≥0.75 (exudate)≤0.65 (transudate)
BNP<200 pg/mL>1200 pg/mL or pre-/posttransfusion BNP ratio of ≥1.581
Clinical examRales on auscultationPeripheral edema, distended neck veins, rales, and S3 may be heard on auscultation
Pulmonary artery occlusion pressure≤18 mmHg>18 mm Hg
Response to diureticMinimalSignificant
WBC countTransient leukopeniaUnchanged
WBC antibodiesCognate donor WBC antibodies support the diagnosis of TRALIDonor WBC antibodies may or may not be present


The incidence of ATRs has been estimated at less than 1% to 3% of all transfusions.94,95 Fortunately, most ATRs are mild and limited to urticaria, pruritis, and/or flushing. Anaphylactic reactions are characterized by systemic symptoms of bronchospasm, angioedema, and/or hypotension and estimated incidence ranges from 1:18,000 to 1:172,000 transfusions.96 ATRs are commonly associated with FFP and PLT transfusions,87 and the rate of ATR to FFP found in two retrospective studies was 1:591 and 1:2184 plasma units transfused.26,94 In general, the offending plasma proteins and/or antigens to which patients react are not easily identifiable, excepting haptoglobin and human immunoglobulin A (IgA). Reports of rare cases of anaphylaxis after transfusion of methylene blue (MB)-treated plasma are mentioned below.

Antibodies to human IgA were first identified in 1968 as high-titered (>1:1000) IgG antibodies, reacting with a panel of purified IgA monoclonal myeloma proteins of both IgA1 and IgA2 subclasses and termed “class-specific” anti-IgA.97,98 Such antibodies cause dramatic anaphylactic reactions mediated by complement activation to small amounts of plasma containing IgA proteins.97-99 The argument that the anaphylaxis is mediated by IgE antibodies was conclusively refuted by studies performed by Homburger and colleagues.100 The correlation of anaphylactic reactions with class-specific anti-IgA is so compelling that such patients must be managed with 1) cellular products extensively washed to remove residual plasma and 2) plasma products from IgA-deficient (IgA-D) donors.101 Responding to this clinical need, the first registry of IgA-D donors was established in San Francisco in 1975. Analysis of sera from 73,569 blood donors revealed IgA deficiency in 113 (1:650) samples, all with normal IgG and IgM levels. Of these, 30 sera had low levels of IgA, while the remaining 83 had no IgA detectable by a more sensitive hemagglutination inhibition assay.102 Class-specific anti-IgA was detected in 13 IgA-D donors and only two had any known history of parenteral injection of plasma proteins. Because isoimmunization to IgA in intrauterine life has been reported,103 it is not surprising that class-specific anti-IgA occur in the serum of IgA-D donors without any parenteral exposure to IgA. Passive transfusion of high-titer anti-IgA provoked no clinical reaction in the recipients.102 This observation of passive transfusion of high-titer anti-IgA to normal patients without provoking a reaction is now affirmed by a larger study carried out by Robitaille and colleagues.104 Currently, the American Rare Donor Program (a joint program with the American Red Cross and the AABB) has an active registry of IgA-D donors in the United States and can provide IgA-D products to recipients nationwide who meet certain clinical and laboratory criteria.105,106 Additional registries exist in Europe, Australia, and most recently in China.107 Depending on the screening method(s) used, the prevalence of IgA deficiency varies widely, for example, highest in Portugal (1:327) and lowest in Japan (1:31,800).107,108 While many published reports continue to document the consistent picture of the anaphylactic reactions caused by class-specific anti-IgA,109 how to manage patients with an anaphylactic reaction is laid out in a clinically useful form by Sandler.101 In contrast with the high-titered class-specific anti-IgA causing serious anaphylactic reactions, low-titered (1:128) antibodies reacting with some of the proteins of either IgA1 or IgA2 subclass are termed anti-IgA of “limited specificity.”98,110,111 These antibodies are characteristically associated with milder ATRs.97

Genetic deletions resulting in deficiencies in other plasma proteins can also result in sensitization and induce anaphylaxis upon exposure to that protein within a blood product. Shimada and coworkers112 identified 7 of 4138 Japanese patients with haptoglobin deficiency, six of whom had severe anaphylactic reactions after transfusion of a small volume of blood product. Haptoglobin IgG and IgE antibodies were detected in the patients' sera, and the authors concluded that both antibody types may have played a role in inducing the reaction. Ultimately, the authors reported that “haptoglobin deficiency is an important risk factor for anaphylactic reactions in Japan.” Although less clear cut, another described case of a protein deficiency which may have induced a transfusion reaction is a patient who had 1) absence of complement component C4 with anti-C4 (with Chido and Rogers specificity) and 2) adverse reactions to plasma transfusions.113

The potential risk of food allergies as a cause of ATR has also been postulated by Erick,114 and anaphylaxis from passive transfer of a peanut allergen in transfused PLTs was recently documented by Jacobs et al.115 Selective protein deficiencies in recipients, however, do not account for most anaphylactic reactions. Furthermore, passive transfer of food allergens likely does not contribute to most ATRs although this is harder to prove. Evidence to support the role of passive IgE-mediated transfer in ATRs is not established, but one recent study did not show a significant difference in IgE levels in apheresis PLTs implicated in allergic reactions versus control PLTs.116,117 Recent studies published by Savage and coworkers117-119 have tried to better characterize mechanisms and risk factors for ATRs with apheresis PLTs. One study demonstrated that certain agonists of basophils and mast cells, such as C5a, brain-derived neurotropic factor, and CCL5 (RANTES), were found in greater concentrations in the supernatant of apheresis PLTs implicated in ATRs than control units.118 However, it is unclear if these agonists are associated with ATRs due to FFP. Another study observed 1616 ATRs among 93,737 transfusions (1.72% incidence) and found that 30% of recipients with an ATR had an ATR rate of more than 5%, which was greater than the overall incidence.119 Furthermore, these 30% of patients accounted for 62.1% of all the ATRs.119 Thus, certain patients are more prone to ATRs. In fact, atopic predisposition in the recipient has been shown to be a risk factor for ATRs to PLTs.117 Recipients with ATRs had a higher total IgE and aeroallergen-specific IgE than matched controls.117 Savage and colleagues119 also showed that certain donors donated PLT products that resulted in an ATR rate of 5.8%, which was greater than overall incidence of ATR in the study (1.72%). Thus, certain donor factors also play a role. Yet, interestingly, in 630 instances where split apheresis PLTs were given to two patients in which one had a reaction, there were only six instances where the patient who received the split product also had a reaction.119 This further supports the importance of recipient factors. Thus, the mechanism of ATRs likely entails a two-event model where both recipient and donor factors must be present. Another possible mechanism for ATRs, which has been proposed, involves the possible activation of anaphylatoxins in the recipient upon infusion of negatively charged PLT microparticles, which are abundant in FFP and PLT units.96 Understanding the mechanism of ATRs is important to recommend appropriate preventive measures. Currently, pretransfusion medication of patients with an antihistamine is a common practice, but two randomized controlled trials indicated that pretransfusion antihistamine medication did not decrease the incidence of ATR.120,121 However, in patients with repeat ATRs premedication with antihistamines and/or corticosteroids should be considered. Although most ATRs do not have major clinical sequelae, prevention is important since these reactions cause patient discomfort and incur extra cost due to reaction workup and product wastage.

Infectious risks

Risk of infectious disease transmission has been dramatically reduced in the past two decades due to extensive donor medical screening and infectious disease testing. Improvements in test sensitivity, such as nucleic acid testing (NAT), have significantly contributed to this decreased risk. In the United States, the estimated risk for acquiring human immunodeficiency virus (HIV), hepatitis C virus (HCV), and hepatitis B virus (HBV) through transfusion is 1:1,467,000, 1:1,149,000, and 1:280,000 donations, respectively.122,123 To further enhance plasma safety, many blood centers, mainly outside the United States, use 1) donor-retested plasma or 2) PRP. With donor-retested plasma, FFP is quarantined until the donor gives a subsequent donation that tests negative for infectious disease.124 PRP offers good virus protection and can be prepared from solvent/detergent (S/D) treatment of pools of plasma or treatment of single donor units with MB, amotosalen, or riboflavin and UV light. S/D-treated plasma prevents transmission of lipid-enveloped viruses (HIV, HCV, HBV), but does not protect against nonenveloped viruses such as HAV or parvovirus B19. Transmission of these two viruses is prevented by testing plasma units for HAV and parvovirus B19 by NAT, dilution through pooling, and neutralization with antibodies present in the pool.125 With regard to prions, there are no donor screening tests but the S/D treatment process does result in a reduction of abnormal prion protein.125 Creutzfeldt-Jakob disease (CJD) is the best known prion disease in humans but is most likely not transmitted through transfusion.126 However, there have been four possible cases of transfusion-transmitted variant CJD (vCJD) in the United Kingdom, all associated with transfusion of nonleukoreduced RBCs between 1996 and 1999.127 Universal leukoreduction has been implemented in Europe, which decreases the risk of vCJD transmission but does not completely eliminate it.128,129 Although there have been no reported cases of transmission via FFP transfusion, animal studies show that plasma can contain the infective prion.130 In the United States and other non-UK countries, donors with prion-related risk factors are permanently deferred. In the United Kingdom, all children up to age 16 receive plasma imported from areas with low bovine spongiform encephalopathy incidence to reduce vCJD risk and MB treated to reduce other infectious risks.129,131 Finally, bacterial contamination of plasma is rare due to frozen storage but is still reported. Five cases of bacterial contamination of FFP were reported in Canada from 2002 to 2003 and five cases in Germany from 1997 to 2007.132,133 Organisms identified included species of Staphylococcus, Klebsiella, Propionibacterium, and Pseudomonas. Water baths used to thaw plasma are a potential source of contamination, and pseudomonas has been cultured from frozen products thawed in contaminated water baths.134,135 Care must be taken to properly clean and sterilize water baths regularly, and plasma should be transfused as soon as possible after thawing. Malaria from transfusion of previously frozen plasma does not occur, and as discussed below, such is also the case for cytomegalovirus (CMV) transmission.

WBC-associated risks

WBC-associated complications, such as febrile nonhemolytic transfusion reactions (FNHTRs), transfusion-associated graft versus host disease (TA-GVHD), white blood cell (WBC) alloimmunization, and transmission of leukotropic viruses (i.e., CMV, human T-lymphotropic virus), are not typically associated with plasma transfusion since FFP is considered noncellular. However, several studies have shown significant numbers of WBCs contaminating plasma units (1 × 106-3 × 106 WBCs per unit) before freeze, although only a small percentage of viable WBCs remain after the freeze-thaw process.136-139 Destruction of WBCs during freeze-thaw can release bioactive mediators which may mediate FNHTR.140 In 2010, the rate of FNHTRs reported to SHOT for plasma was 0.9 per 100,000 units.87 One retrospective study in a large US hospital reported the rate of FHNTR to plasma transfusion as 1:4476.26 Typically, FNHTRs are clinically insignificant and resolve quickly. TA-GVHD, on the other hand, is usually fatal and is caused by viable lymphocytes within a transfused product which engraft and proliferate within the transfusion recipient. Fortunately, TA-GVHD is rare and has never been reported with FFP. It has been estimated that TA-GVHD can occur with as few as 80,000 transfused lymphocytes, but a thawed plasma unit is unlikely to contain that number of viable lymphocytes.139,141,142 Therefore, irradiation of FFP is not currently recommended.142,143 WBC alloimmunization is a potential risk of plasma transfusion since both dead and viable WBCs within FFP express HLA antigens.139 Regarding CMV transmission, two studies did not detect CMV within frozen plasma units supporting that CMV transmission is highly unlikely.144,145 Prestorage leukoreduction decreases the number of residual WBCs in FFP and further prevents these unlikely complications.

RBC alloimmunization

The United Kingdom, Germany, and Council of Europe require that a single plasma unit contain less than 6.0 × 109 RBCs/L before freezing.146 In the United States, there is no standard for acceptable RBC concentration in plasma units. Residual RBCs and RBC fragments within plasma units can potentially cause RBC alloimmunization, and identification of anti-D, -E, -Jka, and -Fya after plasma transfusion has been reported.146-149 After the freeze-thaw process, most RBCs are fragmented, which decreases their immunogenicity.4 Since the complication of RBC alloimmunization is rare, there is currently no requirement to provide D– plasma to D– patients.

Hemolytic transfusion reactions

To prevent hemolytic transfusion reactions (HTRs), transfusion services provide ABO-compatible FFP to patients. However, occasionally ABO-compatible plasma is unavailable due to inventory limitations or incompatible plasma is erroneously provided due to specimen or patient identification errors. Fortunately, a severe HTR with a unit of ABO-incompatible plasma is less likely than with a unit of ABO-incompatible RBCs since the clinical effect of transfusing a small volume of isohemagglutinins relative to an adult recipient's RBC volume is usually insignificant.141 Nevertheless, transfusion of an ABO-incompatible plasma unit may cause a HTR, especially if the donor has high-titer isohemmagglutinins. There have been multiple case reports of HTRs after transfusion of a single unit (approx. 200 mL) of ABO-plasma incompatible PLTs (i.e., group O PLT to a group A patient).149-151 Thus, even small volumes of ABO-incompatible plasma can potentially cause a HTR, and transfusion of ABO-incompatible FFP should be avoided. Guidelines for FFP use in the UK recommend that if ABO-compatible FFP is not available, FFP of a different ABO group may be used if it does not contain high titer anti-A or anti-B. The UK Blood Services tests donations for “high-titer” antibodies.4 Donations with low titers are labeled to indicate a low risk of causing hemolysis; however, hemolysis can still occur with these units.4


S/D-treated plasma and MB-treated plasma are widely used in Europe. In the United States, S/D plasma is FDA licensed but is not generally available.152 Pathogen inactivation or reduction methods can cause loss of some coagulation factors.125 Compared to FFP, MB plasma has reduced fibrinogen, Factor (F)V, FVIII, FIX, and FXI activity, which may impact efficacy.153 S/D plasma has decreased activity of protein S, FVIII, FV, and plasmin inhibitor, but multiple studies have documented its clinical effectiveness and safety.125,153 The reduced activity of protein S in S/D plasma may, however, be associated with venous thromboembolism (VTE). One study reported VTE in 7 of 68 thrombotic thrombocytopenic purpura patients receiving plasma exchange with S/D plasma.154 Despite this potential risk of S/D plasma, other benefits besides viral protection include a lower rate of allergic reactions, febrile reactions, and TRALI.155,156 Klein and colleagues157 reported an overall reaction rate per unit of S/D plasma transfused as 0.66% and most complications were minor (i.e., hives, chills). Another study, however, reported no reactions to S/D plasma after transfusion of 5064 S/D units to 894 recipients,158 and Norway's hemovigilance system reported 14 mild adverse events with transfusion of 47,690 S/D-treated plasma units (1:3400) in 2008.159 Overall, studies and hemovigilance reports support that acute reactions are less common with S/D plasma than FFP, and most notably there have been no documented cases of TRALI (meeting Consensus Criteria) associated with S/D plasma despite transfusion of approximately 10 million units.125,153 The decreased incidence of TRALI may be explained by in vitro studies which show that HLA antibodies are undetectable in S/D plasma units likely due to the dilutional effect of pooling and/or neutralization of HLA antibodies by soluble HLA antigens in the plasma pool.125,155,160,161 Similarly, pooled S/D-treated plasma products may cause fewer allergic reactions due to dilution of ATR mediators. Removal of PLT microparticles may also contribute to a decreased risk of ATR with S/D-treated plasma.95 With regard to MB plasma, there have been three recent reports of anaphylactic reactions from the residual MB in MB-treated plasma.162,163 Although this is a rare complication of MB plasma, it should be considered when a patient has an anaphylactic reaction during or after MB plasma transfusion. Finally, S/D plasma (but not MB plasma) contains no residual cells or cell fragments preventing WBC-associated risks and RBC alloimmunization.4,153,164

Transfusion-associated morbidity and mortality

Plasma transfusion has been associated with increased morbidity in different patient populations. In trauma patients who survive their initial injury, one study showed a 2.1 and 2.5% increased risk of multiorgan failure and acute respiratory distress syndrome (ARDS), respectively, for every unit of FFP given.165 Another study in non–massively transfused trauma patients (<10 RBC units within 12 hr of admission) similarly found increased complications with increasing volumes of plasma transfused. Patients transfused with more than 6 units of plasma had a 12-fold increase in ARDS, sixfold increase in multiorgan dysfunction syndrome, and fourfold increase in pneumonia and sepsis.166 In addition, FFP transfusion has been associated with multiorgan failure in pediatric liver transplant patients, and acute lung injury (ALI) or ARDS in critically ill adult patients.167,168 Plasma transfusion has also been associated with nosocomial infection in surgical and trauma patients.166,169,170 Finally, Puetz and coworkers20 recently reported that FFP transfusion may be associated with venous thrombosis in children. With regard to mortality, plasma transfusion has been associated with improved survival in trauma patients. Two recent reviews summarized the findings of 11 retrospective studies evaluating the effects of aggressive plasma transfusion on mortality in massively transfused trauma patients.171,172 Most studies showed improved survival with increased plasma-to-RBC ratios, but the optimal ratio varied between studies. Although this evidence supports increased plasma use in massive transfusion, prospective randomized controlled trials are needed to confirm the efficacy of this practice. For civilian trauma patients not massively transfused, Inaba and colleagues166 reported no improvement in survival with plasma transfusion whereas Spinella and coworkers173 showed decreased mortality with increased plasma to RBC ratios for combat-related injuries with or without massive transfusion. Studies have associated plasma transfusion with increased mortality in nontrauma settings. Church and coworkers174 described a dose-dependent association between FFP and increased mortality in children with ALI. Interestingly, recent studies have shown increased mortality or morbidity with transfusion of ABO-compatible but nonidentical plasma. A large retrospective study in Sweden showed that exposure to ABO-compatible but nonidentical plasma was associated with increased 14-day mortality after transfusion, especially in group O patients receiving AB plasma.175 However, the association was not dose dependent. Another study did not show increased mortality with transfusion of ABO-compatible, nonidentical plasma, but did report increased complications including ARDS and sepsis.176 Plausibility exists since soluble donor antigens in plasma may result in formation of immune complexes with recipient antibodies causing immune modulation. Although the described studies suggest a possible association between plasma transfusion and morbidity and/or mortality, which help guide further study, proving a cause–effect relationship remains a challenge as is discussed in a recent article on establishing causation in transfusion medicine.177


The risks associated with plasma transfusions have changed over the years. Risk of infectious disease transmission has been significantly reduced with donor testing and pathogen reduction strategies bringing noninfectious complications to the forefront. Hemovigilance systems play an important role in helping to identify areas of concern so that appropriate mitigation strategies can be developed. Early in the past decade, hemovigilance systems revealed TRALI to be a major cause of transfusion-associated morbidity and mortality from plasma transfusion. Knowledge of TRALI's pathogenesis led to policy changes for plasma collection that evidence indicates has decreased the risk of TRALI from plasma transfusion. With risk of TRALI from plasma transfusion now decreased, more focus is being placed on TACO for which plasma transfusion appears to carry a greater risk. However, more studies are needed to better understand TACO to make appropriate recommendations for prevention. The same can be said for allergic reactions. Although plasma transfusion is safer today than in the past, zero risk is not attainable and clinicians must be aware of the potential hazards that accompany the transfusion of plasma. Finally, a recent review by Vamvakas and Blajchman178 described different strategies to reduce transfusion-related mortality, one of which was “avoidance of unnecessary transfusions through evidence-based transfusion guidelines.” Unfortunately, much of the current practice of plasma transfusion is not based on sound evidence. Clinical trials examining restrictive versus liberal plasma transfusion, similar to those done with RBC transfusions, are needed in many settings for which plasma is currently transfused. Ultimately, better evidence-based guidelines for plasma transfusion will increase transfusion safety by minimizing inappropriate transfusion.


This work was supported in part by NIH Grants NHLBI (RO1 HL053384) and NIDDK (P30 DK026743).


The authors declare that they have no conflict of interest regarding this manuscript.