Acute Lung Injury Complicating Blood Transfusion in Post-Partum Hemorrhage: Incidence and Risk Factors
Luciana Teofili1, Maria Bianchi1, Bruno A. Zanfini2, Stefano Catarci2, Rossella Sicuranza2, Serena Spartano1, Gina Zini1 and Gaetano Draisci2
1 Hematology Department, Università Cattolica del Sacro Cuore, Rome, Italy
2 Anesthesiology Department, Università Cattolica del Sacro Cuore, Rome, Italy
Received: June 20, 2014
Accepted: October 7, 2014
Meditter J Hematol Infect Dis 2014, 6(1): e2014069, DOI 10.4084/MJHID.2014.069
This article is available on PDF format at:
| This is an Open Access article distributed
under the terms of the Creative Commons Attribution License
(http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background: We retrospectively investigated
the incidence and risk factors for transfusion-related acute lung injury
(TRALI) among patients transfused for post-partum hemorrhage (PPH).
Postpartum hemorrhage (PPH) constitutes the most frequent cause of maternal mortality in low-income countries.
In developed countries, PPH is the prevalent cause of critical illness
among obstetric patients, and recent evidences suggest that its
incidence is progressively rising.[2,3] In particular,
women with persistent PPH, defined as “the active bleeding exceeding
1000 ml within 24 hours following delivery, that continues despite the
use of initial measures including first-line uterotonic agents and
uterine massage”, are considered at high risk of adverse outcome.
Transfusion-related acute lung injury (TRALI) is a serious transfusion
reaction characterized by non-cardiogenic lung oedema, hypoxemia and
respiratory distress occurring after blood transfusion.[5-7]
The reported incidence of TRALI greatly differs in retrospective and
prospective studies: overall, it is estimated to vary between 0.08% and
15% of patients receiving a blood transfusion.
According to the two-hit hypothesis, TRALI results from a capillary
leak caused by two consecutive events: the adhesion of primed
neutrophils to pulmonary endothelial cells (first hit) and, the
subsequent activation of both cells by antibodies or inflammatory
mediators present in transfused blood (second hit).
Antibodies to class II- human leukocyte antigens and to human
neutrophil antigens, contained in donations from persons with a history
of transfusions or previous pregnancies, are often implicated in the
antibody-mediated TRALI. Moreover, various
substances accumulated during the prolonged storage of RBC or PLT are
suspected to elicit antibody-negative TRALI. TRALI
is particularly observed in critically ill patients: massive
transfusion, mechanical ventilation, sepsis, hematological
malignancies, end stage liver disease and cardiac surgery are all
acknowledged important risk factors for TRALI.[9-15]
Since patients with PPH receive transfusion of great amounts of blood
products, it is conceivable that they might be at high risk for
developing TRALI. Therefore, we retrospectively identified a series of
patients heavily transfused for PPH, and we evaluated among them the
incidence and risk factors for TRALI.
Patients and methods
Clinical data collection and TRALI definition. We identified in the blood bank transfusion database EmoNet (Insiel, Trieste, Italy) the patients needing an urgent transfusion of a minimum of three RBC units in the delivery room, with or without fresh frozen plasma (FFP) and platelet concentrates (PLT). Clinical records of patients were then retrieved and, for each patient, were collected demographics, detailed clinical and obstetric history, laboratory and radiological findings and data on administered drugs. In order to identify cases of TRALI, two anesthesiologists (B.A.Z. and S.C.) independently examined clinical records. Cases, requiring arterial blood gas analysis and special oxygen therapy or ventilatory support, were carefully checked; clinical information was then matched with transfusion history. According to the 2004 consensus criteria, TRALI was diagnosed in the event of new-onset hypoxemia or deterioration demonstrated by a PaO2/FiO2 <300 mmHg within 6 hours after transfusion, with bilateral pulmonary changes in the absence of cardiogenic pulmonary edema. If an alternative risk factor for acute lung injury (ALI) was present, the diagnosis of possible TRALI (pTRALI) was performed.[6,7] Patients were then grouped according to the outcome “TRALI” or “no TRALI”. The following variables were compared between groups: age, smoking, presence of disseminated intravascular coagulopathy – DIC, fibrinogen and hemoglobin prepartum levels, number of previous pregnancies, parity, vaginal or cesarean section delivery, need for surgical repair of the uterus (including uterine balloon tamponade, uterine compression sutures, uterine artery ligation and peripartum hysterectomy), presence of morbidities preexisting to pregnancy, pregnancy-related non-hypertensive and hypertensive disorders, post-delivery hospitalization time, rate of admission at intensive care unit (ICU) and estimated blood loss.
Transfusion data collection. Blood products were supplied from the blood bank of Policlinico “A. Gemelli”; blood product data were collected using the EmoNet database (Insiel). We evaluated the following transfusion-related risk factors for TRALI: number of units received of RBC, PLT and FFP, number of units received from female donors, mean storage time of transfused units (for RBC and PLT), number of RBC units with a storage time longer than 14 days. All RBC units were in saline, adenine, glucose and mannitol (SAGM) additive solution and were both leukoreduced and buffy-coat removed; PLT concentrates were obtained from buffy-coats of single donors or from apheresis; similarly, FFP units were from both single donors or from apheresis. For the purpose of analysis, we calculated one PLT apheresis unit as five single donor units and one FFP apheresis unit as two single donor units.
Among the 22,344 deliveries occurred at our hospital from January
2005 to December 2011, we identified 71 patients with PPH requiring the
transfusion of at least three RBC units. The mean age + SD was 34 + 5.5
years; 21% of patients had a vaginal delivery and 79% cesarean
delivery. The reported estimated blood loss varied from 300 to 7000 ml.
TRALI diagnosis. We found evidence of a new-onset hypoxemia within 6 hours after transfusion in 15 cases: TRALI were identified in 10 of them, possible TRALI (4 cases) was diagnosed in one patient with pneumonia and in three patients with pre-eclampsia, whilst in one patient with pre-existing valvular heart disease, hypoxemia and pulmonary edema were attributed to transfusion-associated circulatory overload (TACO). Five patients in the TRALI group (36%) and 5 patients in the no-TRALI group (9%) required admission to ICU (p=0.021, Table 2); overall, no patient required ventilatory support for more than 96 hours and no patient died. TRALI cases were not notified to the local hemovigilance office; no immunologic studies in patient and donor samples were performed.
Transfusion-related risk factors. Among 71 patients with PPH, 20 (28% ) were transfused with only RBC, 44 (62%) with RBC and FFP and 7 (10%) with RBC, FFP and PLT. Overall, the number of patients receiving all the three blood products was higher in the TRALI than in no-TRALI group (p=0,025, Table 1). As shown in Table 1, patients with TRALI received a higher number of units of RBC (p=0,008), PLT (p=0,008) and FFP (p=0,034). Overall, patients in the TRALI group received a higher number of blood products from female donors (p=0,047, Table 1). RBC and PLT units transfused to patient with or without TRALI had similar storage times and patients in the TRALI group did not receive higher number of "old" RBC units (i.e. those units stored for more than 14 days) (Table 1).
Patient-related risk factors. The clinical features of patients grouped according to the diagnosis of TRALI are shown in Table 2. TRALI was not associated to age or smoking. Pre-existing diseases were recorded in 21 (30 %) patients and included: hypothyroidism (5), obesity (4), heart valvular diseases (2), HCV hepatitis (2), asthma (1), HIV infection (1), Poland syndrome (1 pt) , sickle cell disease (1), Marfan syndrome (1 pt), hyperthyroidism (1), type I (1) and type II diabetes mellitus(1), HBV hepatitis (1), liver failure due to Crigler-Najjar disease (1), systemic lupus erythematosus (1); 3 patients had more than one disease. Among patients with pre-existing diseases, seven experienced TRALI (50%) and 14 did not (24%) (p=0.099, Table 2). TRALI did not occur in the two patients with heart valvular disease, whereas one of them had TACO. Non-hypertensive co-morbidities related to pregnancy occurred in 7 patients and included anemia (2), gestational diabetes (2), hypothyroidism (2) and intra-hepatic cholestasis (1), in similar proportions between TRALI and no-TRALI groups (Table 2). Eight patients had pregnancy-related hypertensive disorders: 4 suffered from gestational hypertension, 2 were affected by gestational hypertension with superimposed pre-eclampsia, and two patients had pre-eclampsia. Overall, 5 of them experienced TRALI (Table 2). Indeed, hypertensive disorders were overrepresented among patients with TRALI/possible TRALI (36% versus 5%, p=0.006, Table 2). In contrast, patients with and without TRALI had similar obstetric issues, including parity, vaginal or cesarean section delivery, number of previous pregnancies, estimated blood losses and necessity of surgical intervention (Table 2). In addition, other acknowledged risk factors for PPH such as uterine atony, previous uterine surgery, oxytocin administration or placental abnormalities were equally represented among patients with or without TRALI (data not shown). Finally, patients with TRALI were more frequently admitted to the ICU and had a longer hospitalization (p=0.021 and p<0.0001, respectively, Table 2).
Analysis of combined risk factors. We next examined the concomitant effect of transfusion- and patient-related factors on the risk to develop TRALI, by combining in a multivariate model all factors that in univariate analysis had a significance level of 10% (p < 0.1, Table 3). We found that only pregnancy-related hypertensive disorders were significant predictors for TRALI, with an OR of 27.7 (95% CI 1.27-604.3, p=0.034) (Table 3). In order to ascertain if these underlying conditions were not merely increasing the risk of transfusion rather than risk of TRALI, we compared the amount of RBC, FFP and PLT transfused to patients with and without hypertensive disorders. We found that patients with hypertensive disorders required similar amounts of blood products as patients without hypertensive disorders (p=0.699 for RBC, p=0.685 for FFP and p=0.325 for PLT, respectively), suggesting that the higher risk for TRALI observed in these patients is not dependent on blood volume transfused.
|Table 1. Univariate analysis of transfusion-related risk factors for TRALI. Significant results are in bold. RBC denotes red blood cell; PLT denote platelet; FFP denotes fresh frozen plasma; IQR denotes interquartile range; SD denotes standard deviation|
|Table 2. Univariate analysis of patient-related risk factors for TRALI. * Pregnancy-related hypertensive disorders include three patients with pre-eclampsia and 2 patients with hypertension in the TRALI group and 1 patient with pre-eclampsia and 2 patients with hypertension in the no-TRALI group. Significant results are in bold. ICU denotes intensive care unit; DIC denotes disseminated intravascular coagulation. Other abbreviations as in Table 1.|
|Table 3. Multivariate analysis of transfusion- and patient-related risk factors for TRALI. OR denotes Odds ratio, CI denotes Confidence Interval. Other abbreviations as in Table I. Significant results are in bold.|
Previous hemovigilance studies suggest that 6.7 to 15% of reported TRALI occur among obstetrics-gynecological patients,[19,20] but data so far published within in this setting are scarce or even anecdotal.[21,22]
In our retrospective series of heavily transfused patients, we found an
overall incidence of TRALI/ pTRALI of 16,9%, that is noteworthy. We
investigated a limited number of patients, representing a fraction of
the overall population with PPH admitted to our hospital in the same
period. Nonetheless, our study is the first attempt to identify TRALI
incidence and risk factors in the obstetrics-gynecology population. In
general, critically ill patients have the highest risk to develop
TRALI, with reported incidence ranging from 1,8 to 15%.
In PPH patients, as well as in general people, the amounts of
transfused products constitute a predictable risk factor for TRALI,
with a significant increase of risk in patients receiving all three
types of blood components. In order to reduce the TRALI risk, female
donors with previous pregnancies or miscarriages as well as donors who
had been previously transfused, are currently no longer eligible for
plasma donations in many countries.[19,23]
Since this policy started in Italy only in mid-2011, many patients
included in this study had probably received antibody-positive blood
products. Accordingly, also in our series we found a possible
implication of blood products from female donors in inducing TRALI. In
addition, our findings suggest that in the obstetric setting the
predisposition to TRALI is also driven by clinical condition of
patients. We found that pregnancy-related hypertensive disorders,
encompassing gestational hypertension and pre-eclampsia, are the most
significant predictors for TRALI, with an increase of risk for
developing this complication of about 27 folds. Severe hypertension
with pulmonary edema can frequently complicate preeclampsia;[18,24,25]
therefore, in pre-eclamptic patients, a careful differential diagnosis
with this condition is mandatory. In particular, among our patients
with pre-eclampsia, three developed the hypoxemia. In these patients,
hypoxemia was not associated with severe hypertension and was not
responsive to diuretic administration, suggesting that an underlying
cause different from pulmonary edema. Several recent studies
demonstrated that various cell types can act as multipliers or
attenuators of TRALI, including platelets, monocytes and T lymphocytes, and endothelial cells themselves.
Clinical manifestations of pregnancy-related hypertensive disorders
reflect a widespread endothelial cell dysfunction, likely due to the
release of soluble factors from the ischemic placenta.[24,25]
Recently, Caudrillier et al. reported that targeting platelet
activation with either aspirin or a glycoprotein IIb/IIIa inhibitor, is
also able to protects mice in an experimental TRALI model.
Therefore, even though our multivariate analysis relies on a low number
of TRALI patients, the mechanisms underlying pregnancy-related
hypertensive disorders support our findings. In fact, a hypothetical
platelet-mediated mechanism could underlie pulmonary endothelial
breakdown in TRALI and organ vascular damage in pre-eclampsia. Notably,
none among our hypertensive patients developing TRALI assumed
antiplatelet agents at the time of delivery. Moreover, TRALI occurred
in all the three hypertensive patients who did not receive
anti-hypertensive therapy but only in two out of five patients
receiving alpha-methyldopa, thus suggesting that a tight control of
hypertension may help to mitigate the TRALI risk.
The main limitation of our study is the lack of an immunologic confirmatory test in TRALI cases, aimed to detect the antibody-mediated nature of lung injury. From a clinical point of view, several pre-existing conditions (such as neurologic or valvular heart diseases) or new occurring diseases (such as infections, inhalation of gastric content during general anesthesia, amniotic fluid embolism) can cause respiratory distress during post-partum. The revision of clinical records allowed us to detect or to rule out frequent causes of ALI, like pneumonia, aspiration or sepsis. About amniotic fluid embolism (AFE), the reported incidence is very low, ranging from 1.9 to 6,1 cases per 100,000 maternities. Accordingly, among 22,344 deliveries occurred at our hospital in the same period of the study, no cases of AFE were observed. Finally, we could reasonably rule out TACO in our TRALI patients. TACO, usually, occurs in elderly people with poor cardiovascular function, often after FFP transfusion to reverse anticoagulation. Importantly, respiratory distress due to TACO rapidly improves with diuretic therapy. In contrast, in our patients with TRALI, the respiratory distress was not ameliorated by diuretic administration, suggesting that the pulmonary edema not be caused by cardiac dysfunctions or elevated systemic vascular resistance. Finally, we would emphasize that all patients included in the study were subjected to continuous monitoring of ECG, p02, blood pressure and breath rate in the delivery room. This procedure allowed the prompt detection of hypoxia and facilitated to define the timing of hypoxia onset in respect to transfusion.
Our findings suggest that patients undergoing massive transfusion in post-partum are highly predisposed to develop TRALI: these observations can be particularly relevant in those countries where female donors with previous pregnancies or miscarriages as well as donors who had been previously transfused are still eligible for plasma donations. The strong association between pregnancy-related hypertensive disorders and TRALI predisposition deserves to be confirmed in a prospective evaluation. Moreover, future studies could definitely ascertain if therapies to treat hypertension and prevent pre-eclampsia also reduce the TRALI risk. In the meanwhile, the close observation of patients suffering from hypertensive disorders after transfusion should be recommended.
- Khan KS, Wojdyla D, Say L, Gulmezoglu AM, Van Look PFA. WHO analysis of causes of maternal death: a systematic review. Lancet 2006;367:1066-74 http://dx.doi.org/10.1016/S0140-6736(06)68397-9
- Neligan PJ, Laffey JG. Clinical review: Special populations--critical illness and pregnancy. Crit Care 2011; 15: 227 http://dx.doi.org/10.1186/cc10256 PMid:21888683 PMCid:PMC3387584
- Kramer MS, Berg C, Abenhaim H, et al. Incidence, Risk Factors, and Temporal Trends in Severe Postpartum Hemorrhage. Am Obstet Gynecol. 2013; 209:449. e1-7
- Abdul-Kadir R, McLintock C, Ducloy AS, et al. Evaluation and management of postpartum hemorrhage: consensus from an international expert panel. Transfusion. 2014;54:1756-68. http://dx.doi.org/10.1111/trf.12550 PMid:24617726
- Vlaar AP and Juffermans NP. Transfusion-related acute lung injury: a clinical review. Lancet 2013; 382:984-94 http://dx.doi.org/10.1016/S0140-6736(12)62197-7
- Kleinman S, Caulfield T, Chan P, et al. Toward an understanding of transfusion-related acute lung injury: statement of a consensus panel. Transfusion 2004; 44:1774-89 http://dx.doi.org/10.1111/j.0041-1132.2004.04347.x PMid:15584994
- Toy P, Popovsky MA, Abraham E, et al. Transfusion-related acute lung injury: definition and review. Crit Care Med 2005; 33:721-6 http://dx.doi.org/10.1097/01.CCM.0000159849.94750.51 PMid:15818095
- Bux J, Sachs UJ. The pathogenesis of transfusion-related acute lung injury (TRALI). Br J Haematol 2007; 136:788-99. http://dx.doi.org/10.1111/j.1365-2141.2007.06492.x PMid:17341264
- Silliman CC, Boshkov LK, Mehdizadehkashi Z, et al. Transfusion-related acute lung injury: epidemiology and a prospective analysis of etiologic factors. Blood 2003;101:454-62 http://dx.doi.org/10.1182/blood-2002-03-0958 PMid:12393667
- Gajic O, Rana R, Winters JL, et al. Transfusion-related acute lung injury in the critically ill: prospective nested case-control study. Am J Respir Crit Care Med 2007; 176:886-91 http://dx.doi.org/10.1164/rccm.200702-271OC PMid:17626910 PMCid:PMC2048675
- Vlaar AP, Binnekade JM, Prins D, et al. Risk factors and outcome of transfusion-related acute lung injury in the critically ill: a nested case-control study. Crit Care Med 2010; 38: 771-8 http://dx.doi.org/10.1097/CCM.0b013e3181cc4d4b PMid:20035217
- Benson AB, Moss M , Silliman CC. Transfusion-related acute lung injury (TRALI): a clinical review with emphasis on the critically ill. Br J Haematol 2009; 147:431-3 http://dx.doi.org/10.1111/j.1365-2141.2009.07840.x PMid:19663827
- Benson AB, Austin GL, Berg M, et al. Transfusion-related acute lung injury in ICU patients admitted with gastrointestinal bleeding. Intensive Care Med 2010; 36: 1710-7 http://dx.doi.org/10.1007/s00134-010-1954-x PMid:20658125 PMCid:PMC3727900
- Vlaar A P, Hofstra, JJ, Determann RM, et al. The incidence, risk factors, and outcome of transfusion-related acute lung injury in a cohort of cardiac surgery patients: a prospective nested case-control study. Blood 2011; 117:4218-25 http://dx.doi.org/10.1182/blood-2010-10-313973 PMid:21325598
- Toy P, Gajic O, Bacchetti P, et al.: Transfusion-related acute lung injury: incidence and risk factors. Blood 2012; 119:1757-67 http://dx.doi.org/10.1182/blood-2011-08-370932 PMid:22117051 PMCid:PMC3286351
- SPSS Inc., SPSS Base 10.0 Applications Guide, SPSS Inc., USA, 1999
- Li G, Rachmale S, Kojicic M, et al. Incidence and transfusion risk factors for transfusion-associated circulatory overload among medical intensive care unit patients. Transfusion 2011; 51:338-43 http://dx.doi.org/10.1111/j.1537-2995.2010.02816.x PMid:20723173 PMCid:PMC3006039
- Milne F, Redman C, Walker J, et al. The pre-eclampsia community guideline (PRECOG): how to screen for and detect onset of pre-eclampsia in the community. BMJ. 2005; 330:576-80. http://dx.doi.org/10.1136/bmj.330.7491.576 PMid:15760998 PMCid:PMC554032
- Chapman CE, Stainsby D, Jones H, et al. 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; 49:440-52 http://dx.doi.org/10.1111/j.1537-2995.2008.01948.x PMid:18980623
- Ozier Y, Muller JY, Mertes PM, et al. Transfusion-related acute lung injury: reports to the French Hemovigilance Network 2007 through 2008. Transfusion 2011; 51:2102-10 http://dx.doi.org/10.1111/j.1537-2995.2011.03073.x PMid:21382042
- Michala L, Madhavan B, Win N et al. Transfusion-related acute lung injury (TRALI) in an obstetric patient. Int J Obstet Anesth 2008; 17:66-9 http://dx.doi.org/10.1016/j.ijoa.2007.07.002 PMid:18162202
- Lee AJ, Koyyalamudi PL, Martinez-Ruiz R. Severe transfusion-related acute lung injury managed with extracorporeal membrane oxygenation (ECMO) in an obstetric patient., J Clin Anesth 2008; 20:549-52 http://dx.doi.org/10.1016/j.jclinane.2008.05.019 PMid:19019654
- Arinsburg SA, Skerrett DL, Karp JK, et al.: Conversion to low transfusion-related acute lung injury (TRALI)-risk plasma significantly reduces TRALI. Transfusion 2012; 52:946-52 http://dx.doi.org/10.1111/j.1537-2995.2011.03403.x PMid:22060800
- Mustafa R, Ahmed S, Gupta A et al. A comprehensive review of hypertension in pregnancy. J Pregnancy 2012; 105918 PMid:22685661 PMCid:PMC3366228
- Steegers EA, von Dadelszen P, Duvekot JJ, Pijnenborg R. Pre-eclampsia. Lancet. 2010;376:631-44. http://dx.doi.org/10.1016/S0140-6736(10)60279-6
- Caudrillier A, Kessenbrock K, Gilliss BM, et al. Platelets induce neutrophil extracellular traps in transfusion-related acute lung injury. J Clin Invest 2012; 122:2661-71 http://dx.doi.org/10.1172/JCI61303 PMid:22684106 PMCid:PMC3386815
- Fung YL, Kim M, Tabuchi A, et al. Recipient T lymphocytes modulate the severity of antibody-mediated transfusion-related acute lung injury. Blood 2010; 116:3073-9 http://dx.doi.org/10.1182/blood-2010-05-284570 PMid:20616220
- Sachs UJ. A threshold model for the susceptibility to transfusion-related acute lung injury. Transfus Clin Biol 2012; 19:109-16 http://dx.doi.org/10.1016/j.tracli.2012.03.003 PMid:22677431
M, Berg C, Brocklehurst P, et al. Amniotic fluid embolism incidence,
risk factors and outcomes: a review and recommendations. BMC Pregnancy
Childbirth 2012; 12:7 http://dx.doi.org/10.1186/1471-2393-12-7 PMid:22325370 PMCid:PMC3305555