Transfusion-related acute lung injury (TRALI) is a well described complication of blood product transfusion. The initial reports of TRALI were found in the pulmonary and hematology literature (1,2). However, the descriptions of the syndrome as a clinical entity have led to more widespread recognition.
Several reports in the anesthesia literature described TRALI intraoperatively or in the immediate postoperative period (3–8). Though still a rare event, it is probably often unrecognized intraoperatively, in many instances because of its variable presentation (7). This makes diagnosis difficult under general anesthesia, as the usual signs and symptoms may be masked. Yet perioperative TRALI reaction is more than an isolated occurrence. In one series, 31 of 36 (86%) patients developed TRALI intraoperatively with 24 of 36 patients being transfused during the operation and the other 7 transfused in the postoperative period (2). The significance in the perioperative period is further supported by a recent report in the anesthesia literature (9). Furthermore, TRALI is potentially devastating, with a reported mortality rate of 5%–25% (10).
Under anesthesia, children most commonly receive blood during cranial, spinal, and cardiac operations. These are complex operations with significant fluid shifts, and the surgery can impact hemodynamic stability. This complicates diagnosis and treatment of potential adverse reactions to blood products while under anesthesia. In this report, we describe an anesthetized infant undergoing craniosynostosis repair who developed fulminant intraoperative TRALI with the administration of packed red blood cells (PRBC).
A 3-mo, full-term, 7.5-kg child with the diagnosis of metopic craniosynostosis underwent a general anesthetic for cranial remodeling. An uneventful inhaled anesthetic induction was performed using sevoflurane and nitrous oxide in oxygen. Two peripheral 22-gauge IV catheters were inserted, and pancuronium and fentanyl were administered. After direct laryngoscopy, a 3.5-mm uncuffed, oral endotracheal tube was placed and confirmed with bilateral breath sounds and end-tidal carbon dioxide (ETco2). The patient was mechanically ventilated with initial peak inspiratory pressures (PIP) of 20 cm H2O to achieve an ETco2 of 38. Anesthesia was maintained with isoflurane and 50% nitrous oxide in oxygen and fentanyl. A 22-gauge catheter was inserted in the left radial artery for invasive arterial blood pressure monitoring and serial laboratory determination. A precordial Doppler was placed for detection of venous air embolism.
The initial hemoglobin (Hgb) in the operating room was 10.7. The neurosurgeon injected lidocaine with epinephrine 1:200,000 subcutaneously and proceeded with the craniectomy. Blood loss began and continued through the procedure. The patient’s temperature was 36.7°C and it remained stable throughout the procedure. After administration of 350 mL of lactated Ringer’s solution and 100 mL of 5% albumin, the administration of PRBC began. An arterial blood gas (ABG) (pH/CO2/Po2/Hgb) reading at this time was 7.28/50/126/7.7 with an ETco2 of 37 on 40% oxygen. Ventilation was without difficulty with bilateral breath sounds. The inadequate ventilation was thought to be the result of under-ventilation, possibly from atelectasis, and the PIP was increased to 28. Blood transfusion continued in response to blood loss and to maintain arterial blood pressure. After approximately 400 mL of blood loss, 250 mL of PRBC, 250 mL of 5% albumin, and 650 mL of crystalloid administration, the ABG was 7.25/41/96/12.0. The oxygen saturation by pulse oximetry began to decrease to the mid-90s. No change in the signal of the Doppler was noted, but the patient was placed on 100% oxygen, the head was lowered slightly, and the surgeon asked to flood the field with saline.
The presence of bilateral breath sounds was verified and the saturation increased to 100%, but the patient would not tolerate <100% oxygen to maintain the saturation 96%–100%. There was no wheezing and good chest rise. The ABG at this time was 7.01/56/82/13.0 with persistent mild hypotension. A small dose (1 μg/kg) of epinephrine was administered. Despite clear urine, the possibility of a transfusion reaction was considered. PRBC were stopped, and the blood bank was notified. Within 2 min of stopping the transfusion, the saturation began to decrease again into the low 90s with the patient breathing 100% oxygen. Vigorous hand ventilation was initiated with PIP near 40 cm H2O and breath sounds were present bilaterally, with no wheezing noted. The endotracheal tube was suctioned and the patient’ oxygen saturation decreased into the 60% to 70% range despite ventilation of the patient on 100% oxygen. The surgeon was asked to cover the skull for full access to the airway. On uncovering the patient, the child was noted to be dusky, severely mottled, and severely edematous throughout. The arterial blood pressure and saturation continued to decrease and IV epinephrine (10 μg/kg) was administered. Resuscitation continued with vasopressors, bronchodilators, and crystalloid. Despite these resuscitation efforts, the child could not be ventilated even via a rigid bronchoscope. In the face of an inability to adequately ventilate and oxygenate the child, cardiac function continued to deteriorate. Cardiopulmonary resuscitation was initiated but the child died.
Postmortem examination revealed grossly edematous organs throughout the body. The routine laboratory investigations failed to detect evidence of a transfusion reaction. However, platelet and granulocyte clumps throughout the pulmonary capillary and arterial network were found, suggestive of a TRALI-type reaction (Fig. 1). Further laboratory investigation detected antigranulocyte antibodies in the patient. No other abnormalities were noted. The cause of death was determined to be TRALI.
Transfusion reactions occur in 2% of units administered (11). These include all reactions to blood and blood products, the majority of which are mild and not life-threatening. Mild symptoms, such as urticaria, tachypnea, dyspnea, back pain, and agitation, are generally masked under anesthesia. Therefore, the more serious and later problems, such as hypoxemia, hypotension, hemoglobinuria, oliguria, bronchospasm, shock, and cardiac arrest, may be the only markers to make a presumptive diagnosis or even suspect a transfusion reaction. This complicates and delays diagnosis and treatment.
Most noninfectious transfusion reactions are immune-mediated. Acute hemolytic transfusion reactions are the classic concern under anesthesia, with ABO incompatibility being the most feared because of its frequent morbidity and mortality as well as the association with human error (12). Reactions to blood product administration also occur in the absence of hemolysis. These include anaphylactoid and anaphylactic reactions either antibody mediated or directly through histamine or complement.
Air embolism was suspected early in the patient’s presentation. Routine maneuvers for air embolism, such as lowering the head and having the surgeon flood the field, were undertaken. However, the lack of air detected by Doppler and the lack of any change in ETco2 led us to search for other causes for the respiratory and cardiovascular instability. Although no hemolysis was present in this patient, the patient’s deteriorating cardiovascular and respiratory status, coupled with worsening acidosis in temporal association with blood transfusion, led us to consider the possibility of a transfusion reaction. The transfusion was stopped, the blood bank was notified, laboratory values were sent, and the remaining blood and its tubing were sent to the blood bank. Although TRALI is still a diagnosis of exclusion, the presence of noncardiogenic pulmonary edema and ventilatory problems in proximity to blood product transfusion in patients of any age should raise suspicion. Other markers of TRALI are hypotension and tachycardia (13,14). This differentiates the noncardiogenic pulmonary edema of TRALI from that of circulatory overload, which is associated with hypertension (10). However, hypotension and tachycardia with significant blood loss is usually treated by transfusion, thereby limiting their reliability. Complicating the fluid management is the difficulty in accurately determining blood loss in small children, as well as the possibility of confounding causes of hypotension in craniofacial surgery, such as venous air. Furthermore, once blood loss is sufficient to require blood product transfusion, it is difficult to determine whether blood administration may be the problem instead of the treatment.
This child developed a clinical syndrome consisting of severe and rapidly progressive respiratory failure in temporal relation to the administration of PRBC. Although the possibility of a transfusion reaction was considered during the case, the diagnosis of TRALI was not. TRALI is thought to arise from activation of leukocytes by one of two mechanisms; 1) class I and II antibodies or granulocyte antibodies in the donor reacting with the recipient leukocytes (although recipient antibodies may also react with donor leukocytes) or 2) a two-event model, with the first event being the clinical condition of the patient resulting in pulmonary endothelial activation and neutrophil sequestration and the second event occurring with the transfusion of lipids or antibodies in the blood component that activates primed neutrophils (10). The result of either mechanism is formation of complexes that are filtered in the pulmonary circulation, the release of cytokines and reactive oxygen species that lead to platelet aggregation, and subsequently capillary plugging and capillary leak. This leads to ventilatory failure from capillary leak resulting in pulmonary edema and reduced pulmonary blood flow from plugging of capillaries in the pulmonary bed (2) (Fig. 1).
The ultimate method to eliminate transfusion reactions is to eliminate transfusion. However, this is difficult in craniofacial reconstruction surgery where fluid requirements vary from 12%–200% of the estimated blood volume (15). Although advances have reduced the need for transfusions (16,17), blood administration during these cases has been considered almost inevitable (18). Because of the mortality rate associated with TRALI, greater efforts at prevention are desirable. At present, the main effort at prevention is to eliminate donors who have been implicated in a TRALI reaction until testing donors can be tested for antibodies to high frequency antigens (10,19). However, this kind of retrospective prevention does nothing to avoid the initial case, in this case, a fatal reaction. The anti-granulocyte antibodies presumed responsible are found most frequently in multiparous females (13). Eliminating female donors with more than 3 pregnancies has been suggested; however, this would negatively impact the availability of blood products and only eliminate the risk from a very small number of donors (10,13). Donor leukocytes have also been implicated in TRALI. Screening of donors does nothing to reduce the reaction from patient antibodies to transfused leukocytes. Leukocyte reduction before storage of blood products has been suggested for reduction of other types of transfusion reactions and may reduce TRALI as well (20). Washing of cellular blood products may also be beneficial, but it is cumbersome and expensive (10). Screening tests would be very useful, but the uncertain sensitivity and the low specificity make them unlikely at present (14). HLA antibody screening of multiparous females and eliminating large plasma volumes from these patients may prove to be beneficial (21). However, this still has limited utility and the cost of instituting this strategy for PRBC is prohibitive.
Since this case, our institution has instituted universal leukocyte reduction. The impact of this on TRALI is not clear. However, a reduction in TRALI from leukocyte reduction has been reported and leukocyte reduction has a role in reducing other types of blood transfusion reactions (20). The patient who donated blood in this case was disqualified from further transfusion. Blood products are not routinely washed because of the expense. Unfortunately, none of these interventions eliminate TRALI. Thus, the lack of specific and effective measures to eliminate TRALI emphasizes the need for greater understanding of this entity.
Complicating TRALI is the fact that no effective treatment exists. Treatment remains supportive in most cases. Recognition and stopping the transfusion are important to reduce propagation of the reaction. Mechanical ventilation strategies to overcome the ventilatory failure are the major interventions. Treatment of hypotension is essential. Steroids and diuretics are probably ineffective and diuretics during hypotension may worsen the patient’s hemodynamic status (22). Emergent extracorporeal membrane oxygenation may lead to increased survival in fulminant cases (23). However, the availability and time needed for this intervention make it unavailable to most practitioners. Thus, recognition, elimination of the initiating agent, and supportive care are the foundations of treatment.
Critical to diagnosing transfusion reactions is the awareness that any abnormality occurring during a transfusion is suspect until proven otherwise. Stopping the transfusion, notifying the blood bank, and ordering the appropriate laboratory tests to exclude a transfusion reaction may help avoid disastrous consequences for the patient. The possibility of a transfusion reaction in children undergoing blood or blood product transfusion should be entertained when ventilatory difficulty develops. Although all routine airway manipulations and interventions should be expeditiously undertaken, a temporal association of blood product administration should alert the provider to the possibility of a transfusion reaction, possibly TRALI. In this case, the diagnosis of TRALI was made at postmortem examination, underscoring the potential gravity of this disease. This is the first case of TRALI under anesthesia in an infant that has been reported; others may have occurred and been less severe and/or unrecognized. The severity of TRALI and the potential gravity of the outcome require greater understanding and recognition and increased efforts to develop appropriate screening mechanisms to reduce TRALI.
1. Levy GJ, Shabot MM, Hart ME, et al. Transfusion-associated noncardiogenic pulmonary edema: report of a case and a warning regarding treatment. Transfusion 1986;26:278–81.
2. Popovsky MA, Moore SB. Diagnostic and pathogenetic considerations in transfusion-related acute lung injury. Transfusion 1985;25:573–7.
3. Florell SR, Velasco SE, Fine PG. Perioperative recognition, management, and pathologic diagnosis of transfusion-related acute lung injury. Anesthesiology 1994;81:508–10.
4. Latson TW, Kickler TS, Baumgartner WA. Pulmonary hypertension and noncardiogenic pulmonary edema following cardiopulmonary bypass associated with an antigranulocyte antibody. Anesthesiology 1986;64:106–11.
5. Ebert JP, Grimes B, Niemann KM. Respiratory failure secondary to homologous blood transfusion. Anesthesiology 1985;63:104–6.
6. De Wolf AM, Van den Berg BW, Hoffman HJ, Van Zundert AA. Pulmonary dysfunction during one-lung ventilation caused by HLA-specific antibodies against leukocytes. Anesth Analg 1987;66:463–7.
7. Weber JG, Warner MA, Moore SB. What is the incidence of perioperative transfusion-related acute lung injury? Anesthesiology 1995;82:789.
8. Keifer JC, Kingsley CP, Roth MT, et al. Transfusion-related acute lung injury (TRALI) complicating colectomy for ulcerative colitis. Anesthesiology 1998;89:1020–3.
9. Toy P, Gajic O. Transfusion-related acute lung injury. Anesth Analg 2004;99:1623–4.
10. Silliman CC, Ambruso DR, Boshkov LK Transfusion-related acute lung injury (TRALI). Blood 2005;105:2266–73.
11. Baker RJ, Monichen SL, Nyhus LM. Transfusion reaction: a reppraisal of surgical incidence and significance. Proc Inst Med Chic 1969;27:214–5.
12. Spiess BD. Blood transfusion: the silent epidemic. Ann Thorac Surg 2001;72:S1832–7.
13. Popovsky MA, Haley NR. Further characterization of transfusion-related acute lung injury: demographics, clinical and laboratory features, and morbidity. Immunohematol 2001;17:157–9.
14. Dry SM, Bechard KM, Milford EL, et al. The pathology of transfusion-related acute lung injury. Am J Clin Pathol 1999;112:216–21.
15. Scholtes JL, Thauvoy C, Moulin D, Gribomont BF. Craniofaciosynostosis: anesthetic and perioperative management: report of 71 operations. Acta Anaesthesiol Belg 1985;36:176–85.
16. Ririe DG, David LR, Glazier SS, et al. Surgical advancement influences perioperative care: a comparison of two surgical techniques for sagittal craniosynostosis repair. Anesth Analg 2003;97:699–703.
17. Barone CM, Jimenez DF. Endoscopic craniectomy for early correction of craniosynostosis. Plast Reconstr Surg 1999;104:1965–75.
18. Faberowski LW, Black S, Mickle JP. Blood loss and transfusion practice in the perioperative management of craniosynostosis repair. J Neurosurg Anesthesiol 1999;11:167–72.
19. Insunza A, Romon I, Gonzalez-Ponte ML, et al. Implementation of a strategy to prevent TRALI in a regional blood centre. Transfus Med 2004;14:157–64.
20. Yazer MH, Podlosky L, Clarke G, Nahirniak SM. The effect of prestorage WBC reduction on the rates of febrile nonhemolytic transfusion reactions to platelet concentrates and RBC. Transfusion 2004;44:10–5.
21. Wallis JP. Transfusion-related acute lung injury. Transfus Med 1999;9:175.
22. Looney MR, Gropper MA, Matthay MA. Transfusion-related acute lung injury: a review. Chest 2004;126:249–58.
23. Nouraei SM, Wallis JP, Bolton D, Hasan A. Management of transfusion-related acute lung injury with extracorporeal cardiopulmonary support in a four-year old child. Br J Anaesth 2003;91:292–4.