Abstract

Objective

Transfusion-related acute lung injury (TRALI) is the leading morbidity and mortality in hemotherapy in the United States. Although it is a serious complication of blood transfusion, it is still underestimated and under-reported because of under-recognition and misdiagnosis. In this report, we present 15 surgical patients who developed pulmonary complications secondary to blood transfusion during the perioperative period.

Methods

A 3-year retrospective analysis of 14,441 patients who received blood transfusion intraoperatively in our Taoyuan center was carried out. 15 patients suspected to be subject to TRALI perioperatively were sorted out for analysis of their clinical characteristics.

Results

All of the 15 patients received inhalational general anesthesia, of whom 10 were anesthetized with sevoflurane, four with desflurane, and one with isoflurane. One patient died on the first postoperative day due to multiorgan failure whereas 14 others who were managed with oxygen therapy or mechanical ventilation recovered uneventfully within 72 hours.

Conclusions

TRALI must be recognized as one of the leading causes of mortality related to blood transfusion. Oxygen support is often sufficient in mild TRALI while ventilatory support is required in severe TRALI. The strategy to minimize such a risk using blood products from male donors or from female donors without history of pregnancy or having a negative leukocyte antibody screening can help reduce severe immune mediated TRALI.

Keywords

acute lung injury; blood transfusionperio; perative care;

---

1. Introduction

Transfusion-related acute lung injury (TRALI) is the leading cause of morbidity and mortality in hemotherapy in the United States. It is a clinical syndrome, that often manifests with respiratory distress, hypotension, and fever within minutes to hours after receiving blood transfusion in spontaneously breathing subjects. On the other hand, those under mechanical ventilation often present with hypoxemia and copious frothy secretion oozing from the endotracheal tube, accompanied additionally by a simultaneous decrease in systemic blood pressure. All blood components could be implicated in TRALI; however, platelet concentrates (PC) derived from fresh frozen plasma (FFP) and whole blood have contributed the largest number of reported cases.1, 2 The incidence of antibody-mediated TRALI was first quoted at a rate of 1/5000blood recipients by Popovsky³; however, in Europe, the Serious Hazards of Transfusion (SHOT) committee reported the incidence as 1/65,000 for FFP, 1/71,000 for PC, and 1/149,000,000 for packed red blood cell (PRBC) with an overall mortality of 21%.⁴ Although it is a serious complication of blood transfusion, it is still underestimated and under-reported because of under-recognition and misdiagnosis. In 2004, the Canadian Consensus Conference proposed the diagnostic criteria for TRALI, primarily consisting of a clinical presentation of tachypnea, cyanosis, and dyspnea with acute hypoxemia within 6 hours following blood transfusion, PaO₂/FiO₂ <300 mmHg, oxygen saturation <90% on room air, bilateral infiltrates consistent with pulmonary edema evidenced by chest radiography in the absence of cardiac failure or intravascular volume overload and absence of other risk factors for acute lung injury.⁵

In this study, we present 15 surgical patients who developed this pulmonary complication after perioperative blood transfusion and also review the literature and discussed prevention of TRALI.

2. Materials and methods

After obtaining approval from the Review Board of Chang Gung Memorial Hospital, we retrospectively analyzed the clinical characteristics of all surgical patients admitted for various procedures under regional or general anesthesia from January 2007 to December 2009, sorted out from our anesthesia quality assurance (QA) database. The QA database was composed of a number of standardized forms (database cards) containing demographic data, patient comorbidities, anesthetic techniques used, and perioperative anesthetic management details such as monitoring, drugs, or fluids administered, transfusion of blood products, and minor or major perioperative complications. On review of the database, we identified the patients who were diagnosed to have TRALI and required special oxygen therapy or ventilatory support postoperatively. Those patients who were age <19 years, admitted for cardiovascular or thoracic surgeries, who did not receive any blood transfusion intraoperatively and who had risk factors for acute lung injury (eg, pneumonia, sepsis, trauma) were excluded, because, in our medical center, pediatric patients rarely received blood transfusion and they should be separately analyzed. Owing to the retrospective nature of the study and the possibility that other less severe forms of TRALI might have not been adequately documented, 15 cases who were considered to have sustained TRALI were based on their clinical presentation and the characteristics in the subsequent clinical course.

TRALI was defined as a clinical presentation of respiratory distress with the development of pulmonary edema and hypoxemia, evidenced by arterial blood gas parameters of PaO₂/FiO₂ <300 mmHg in the operating room or in the recovery unit within 6 hours after the commencement of blood transfusion. The diagnosis could be further confirmed by chest radiography. Intraoperative fluid balance was defined as the periodic total input (crystalloids, synthetic colloids and blood products) minus total output (basic maintenance, insensible loss, urine, and blood loss) from the time of endotracheal intubation to the end of surgery.

Parameters including age, gender, weight, American Society of Anesthesiologists (ASA) physical status, underlying medical conditions, anesthetic method (general anesthesia vs. regional anesthesia), anesthetic agents, anesthetic time (from the arrival to the departure of the operating suite), type of surgery, type of blood products used, intraoperative fluid balance, and patients’ postoperative progression were reviewed and analyzed when available.

2.1. Statistical analysis

All data were entered into a Microsoft Excel spreadsheet and analyzed using SPSS 17.0 (SPSS Inc., Chicago, IL, USA). All variables were assessed using descriptive statistics for mean and standard deviation of continuous variables.

3. Results

Of a total of 196,762 patients who received anesthesia from 2007 to 2009, 14,441 were identified to have received blood transfusion perioperatively, of whom 15 developed severe pulmonary complications with a prevalence of 0.10%. Patients’ demographic data are shown in Table 1. The mean age of those diagnosed with TRALI, was 59.5 ± 17.8 years with equal gender distribution, and they were of ASA physical status of II or III. A majority (47%) of them had pre-existing cardiovascular disorder such as hypertension or of coronary artery disease. Table 2 lists the characteristics of surgical procedures and anesthetic factors of the TRALI patients. All 15 patients underwent their surgical procedures under inhalational general anesthesia, 10 (67%) with sevoflurane, four (26%) with desflurane, and one (7%) with isoflurane. Five (33%) patients underwent spine surgery, two craniotomy, three abdominal surgery, two gynecological surgery, one orthopedic surgery, and one esophageal reconstruction. Overall, the mean ± SD anesthetic time was 5.8 ± 3.2 hours and an estimated blood loss of 1260.7 ± 1883.0 mL ranging from minimal to an extreme of 7200 mL. The fluid therapy the patients received was adjusted to their body weight and surgical procedure. On average, they had a negative fluid balance of 257.3 ± 777.4 mL intraoperatively.

Table 3 summarizes the clinical scenarios of all 15 TRALI patients. Extubation almost always took place upon completion of the surgical procedures but Patient 4 and6 were extubated 7 and 7.5 hours after transfusion, respectively. In the patients who were noted to have respiratory distress upon extubation, arterial blood gas analysis and chest radiograph were taken to establish the diagnosis of TRALI. Of the 15 patients, five received only PRBC transfusion, two only FFP transfusion, seven received both PRBC and FFP, and one both PRBC and PC. The time at which recognition of respiratory distress was made was between 1 and 7 hours post-transfusion. Postoperatively, similar clinical presentation of dyspnea and hypoxemia, requiring oxygen therapy and ventilatory support was observed. Of the patients who had taken chest radiograph and had arterial blood gas analysis, all except one had radiographic evidence of pulmonary edema and PaO₂/FiO₂ <300 mmHg. Fourteen patients had their symptoms fully resolved and were successfully extubated within 72 hours with the exception of Patient 3 who expired on the day after surgery and blood transfusion due to multiorgan failure despite aggressive treatment in the intensive care unit. Fig. 1 shows his pre- and postoperative chest radiographs. The postoperative chest radiograph revealed bilateral infiltration, suggestive of acute pulmonary edema.

The intraoperative prevalence of TRALI is shown in Table 4. In a total of 196,762 patients who received anesthesia in our center, the prevalence of intra-operative TRALI was 1/13,117. The incidence of TRALI per blood component was also shown. In 13061 patients who received PRBC transfusion, a TRALI did occur in 1005 patients or one in 3464 units of PRBC transfusion. Similarly, the incidence of TRALI associated with FFP transfusion was 1/825 patients and 1/3382 units; associated with PC transfusion was 1/1263 patients and 1/19,908 units; associated with WB transfusion was 1/394 patients and 1/4399 units.

4. Discussion

When postsurgical patients who have received blood transfusion intraoperatively develop acute respiratory distress, it is important to identify the possible causes. Causes can be either cardiogenic or noncardiogenic. Cardiogenic pulmonary edema is characterized by decompensated heart failure associated with a gradual onset and slow resolution. Causes of noncardiogenic pulmonary edema include TRALI, transfusion associated circulatory overload (TACO), and occurrence of negative thoracic pressure, aspiration, atelectasis, pulmonary embolism, and pneumonia. TACO is considered when tachypnea/dyspnea, cyanosis, tachycardia, and hypertension in addition to signs of circulatory overload such as jugular venous distension and elevated pulmonary artery occlusion pressure are observed. Patients' positive fluid balance and quick response to diuresis and ventilatory support can usually serve as evidence to the diagnosis. Negative pressure pulmonary edema on the contrary results from laryngospasm or other forms of upper airway obstruction following extubation. Anaphylactic transfusion reaction can also lead to acute respiratory distress with clinical presentation of respiratory wheezing, hypotension, and skin manifestations.

TRALI is increasingly recognized globally as the leading cause of morbidity and mortality associated with transfusion. The exact pathogenesis of TRALI is not fully understood, but immune mediated (antibody) and nonimmune mediated processes have been described. Approximately 80% of TRALI is associated with antibodies to white blood cells, mainly of human leukocyte antigens (HLA) class II, human neutrophil alloantigens (HNA)-3a and HLA-A2 antigens from the donors.⁶ Most of the donors implicated in TRALI have been multiparous women and the prevalence of HLA class I and II antibodies is highly correlated at parity.7, 8 As to those sufferers of TRALI in whom no antibodies were identified, Silliman et al⁹ have proposed a nonimmune mediated mechanism or “two-hit” mechanism, which involves an initial insult to vascular endothelium leading to priming of neutrophils to the endothelium. Severe infection, surgery, trauma, massive transfusion, general anesthesia could contribute to the first insult. Then a second insult activates adherent neutrophils to release toxic mediators that damage the endothelium causing capillary leak and acute lung injury. The culprit of second insult could be donor-derived anti-leukocyte antibodies or substances that accumulate during blood storage.10, 11, 12

Worldwide studies have been conducted to investigate risk-minimization measures for TRALI. In 2003, the UK National Blood Service introduced a “only male plasma policy”, which covered all high plasma components (FFP and PC containing more than 250 mL)¹³ and in 2006, the American Red Cross (ARC) also chose a “male-predominant plasma strategy.”¹⁴ After implementation of these safety measures alone or in combination with exclusion of female donors with a history of pregnancy or with leukocyte antibodies, SHOT and ARC demonstrated a reduction in TRALI cases and fatality. Funk et al also demonstrated benefits from similar risk minimization measures in Germany.¹⁵ In addition to the risk minimization measures to decrease transfusion related morbidity and mortality, preventive serological measures such as granulocyte immunofluorescence test, granulocyte agglutination test, monoclonal antibody immobilization of granulocyte antigen and new donor neutrophil screening platforms for HNA antibody detection are also under development.¹⁶

All our 15 patients underwent general anesthesia with inhalational agents and received blood transfusion intraoperatively. They fulfilled the diagnostic criteria of PaO₂/FiO₂ <300 mmHg and chest radiographic evidence when TRALI was diagnosed. Supportive oxygen therapy or ventilatory support was necessary and sufficient for quick and complete recovery within 96 hours. When diuresis was used, it was of limited help. Cardiogenic pulmonary edema was excluded by differential diagnosis because none of our patients had a history of congestive heart failure or valvular disease and, of those who survived, all had a quick resolution of pulmonary edema on their chest radiographs and recovered within 3 days. Other causes of pulmonary complication were considered to be related to endotracheal extubation. TACO was excluded due to the negative intraoperative fluid balance was slight. Negative pressure pulmonary edema secondary to laryngospasm or upper airway obstruction on extubation was excluded by the smooth extubation process shown on our anesthetic records.

In our study, transfusion of blood components was responsible for higher incidence of intraoperative TRALI to what was previously published by SHOT.⁴ TRALI might have been well under-recognized and underestimated in patients undergoing surgery and anesthesia. Of our 15 patents, 12 had at least one organ system dysfunction, seven had hypertension, and two had a history of coronary artery disease, suggesting that systemic diseases could play a role in trigging TRALI. The three patients who were previously healthy without any systemic disease were all females. This may perhaps add information to the existing literature that female gender, systemic disease, and general anesthesia might be associated with increased risk of TRALI. Interestingly, Eder et al¹⁷recently identified that major surgery as one of the predisposing factors for TRALI. Ten of our15 patients received sevoflurane as the sole inhalational agent; however, we were unable to extrapolate which inhalational agents might pose more threat to developing TRALI since sevoflurane was used preferentially in our center for general anesthesia. In terms of surgical procedures, spinal surgeries (33%) were associated with the highest incidence of TRALI. Among the 15 TRALI cases, 14 required oxygen support or brief ventilatory support. The only mortality was associated with multi-organ failure and advanced age. Due to the retrospective nature of the study, the information of the blood donor was not attainable and since autopsy was not performed, whether the mortality was directly associated with TRALI was therefore unclear. This study had a few limitations. First, pediatric population was excluded from the study and this might lead to an over- or underestimation of TRALI in the surgical patients; an overestimation is more likely when all patients are included. Second, only the subjects with severe forms of TRALI were identified and subject to study, leading possibly to an underestimation to the incidence.

Eight out of 15 patients received blood transfusion despite the fact that their estimated blood loss was <20% of their estimated blood volume, suggesting that the intraoperative blood transfusion was unnecessary. In addition, respiratory distress was recognized some 4 to 7 hours after blood transfusion upon completion of the surgery and extubation. It is therefore important to stress that unnecessary blood transfusion could predispose patients to risk of developing TRALI and constant arterial blood gas analysis could alert the surgeons and anesthesiologists for possible occurrence of TRALI. When massive blood transfusion is required, it is imperative to be cautious about the development of TRALI postoperatively.

In conclusion, TRALI must be recognized as one of the leading cause of mortality related to blood transfusion. Identification of the risk factors, such as systemic disease, female gender, and surgical procedure with manipulation of the bones may caution the anesthesiologists and surgeons to alert for TRALI development. Frequently taking blood samples for gas analysis and avoiding unnecessary intraoperative blood transfusion on the part of surgeons and anesthesiologists should therefore be strictly practised. Whenever blood transfusion is essential, the use of fresh blood components may further help reduce the incidence of nonimmune TRALI. Risk minimization strategies such as using plasma from male donors or from female donors without history of conception and blood products with negative leukocyte antibody screening test can help reduce severe immune mediated TRALI. In patients who develop respiratory distress and the diagnosis of TRALI is settled, oxygen support is often sufficient for milder TRALI, while ventilatory support is required in severe TRALI.

Disclosure

The authors report that there are no disclosures relevant to this publication.