Abstract
Transfusion-related acute lung injury (TRALI) is a life-threatening complication of blood product transfusion characterized by sudden onset hypoxemic respiratory failure with bilateral lung infiltrates and non-cardiogenic pulmonary edema developing within 6 hours of transfusion. It is believed to be under-recognized, particularly among preterm neonates in whom co-existing developmental lung disease adds diagnostic complexity. Here we report the case of a preterm neonate who developed TRALI during a blood transfusion following PDA ligation.
Case summary
A 19-day-old ex-25 week (birth weight of 0.82 kilogram) female neonate born to a now gravida 2, para 2 mother, underwent surgical patent ductus arteriosus (PDA) ligation for treatment of a hemodynamically significant PDA causing pressor-dependent hypotension and renal failure. PDA ligation was performed at the bedside without any complication; post-operative chest x-ray (Image 1) showed no evidence of pulmonary infiltrates and improved pulmonary edema compared to pre-operative baseline (Image 2); clinically, the infant was weaned to room air on minimal ventilator settings and came off pre-operative dopamine eight hours after ligation. Post-operative hemoglobin was 10 g/dL thus prompting a 20 mL/kg red blood cell (RBC) transfusion. Ninety minutes after initiation of the RBC transfusion, the infant developed desaturations and decreased breath sounds requiring increasing oxygen requirement and ventilator settings. Capillary blood gas and chest x-ray (Image 3) were obtained showing significant CO2 retention and loss of lung volume with evidence of increasing pulmonary edema. The infant was switched to high frequency oscillating ventilation (HFOV) with improvement in CO2 retention, but without improvement in lung recruitment or pulmonary edema; she developed progressive oliguria, hypotension and worsening hypoxemia, requiring 100% FiO2 that was unresponsive to inhaled nitric oxide. Post-ligation echocardiogram showed absence of ductal shunting and normal left ventricular size and systolic function with patent aorta and pulmonary vessels.
CXR immediately following completion of PDA ligation.
CXR prior to PDA ligation.
CXE immediately following completion of blood transfusion.
At this point, TRALI was suspected given the temporal relationship of clinical deterioration with RBC transfusion. Blood bank was contacted, sepsis workup was initiated (blood, urine, and tracheal cultures) and antibiotics started. The infant remained critically ill with ongoing hypoxic respiratory failure and labile blood pressures requiring volume boluses, dopamine, dobutamine and epinephrine drips as well as corticosteroids for pressor-resistant hypotension. Full supportive management was continued, including high ventilator settings and 100% O2 requirement on HFOV, adequate fluid infusion and BP support.
Thirty six hours following initiation of the RBC transfusion, the infant was able to be weaned off dobutamine and epinephrine drips, and FiO2 requirement was weaned; follow-up chest x-ray (Image 4) demonstrated improving pulmonary edema and better lung expansion. The tracheal culture grew mixed bacterial morphotypes and 3+ Serratia marcescens, which was treated with 10 days of antibiotics; urine and blood cultures were negative. By the 7th post-operative day, dopamine was discontinued and the infant was extubated to nasal intermittent mandatory ventilation (IMV) on 21% FiO2. Two subsequent aliquots were prepared from a different group O, Rh positive, CMV negative parent unit and individually irradiated prior to transfusion. The first of these aliquots was transfused on post-operative day 19 (day +4 from day of collection) and the second on post-operative 32 (day +17 from day of collection). No adverse reactions were associated with either of these transfusions.
CXR at 36 hours post-transfusion.
The donor of the implicated RBC unit was a 25 year-old, never-transfused, nulliparous female. Because of the low likelihood for preformed anti-leukocyte or anti-neutrophil antibodies in this donor, we suspected a possible ‘reverse-TRALI’ mechanism owing to the multiparous status of the mother. Specimens were sent from both the donor and the mother to an outside reference laboratory to test for antibodies directed against both Human Leukocyte Antigens (HLA) and Human Neutrophil Antigens (HNA).
Mother and donor both had negative results for HLA antibody testing. The donor had negative results for anti-HNA testing. The mother had a negative result for Granulocyte Agglutination Assay but positive results (in 9 of 9 donor cells tested) on Granulocyte Immunofluorescence Assay (a flow cytometry-based test). This latter test would have higher sensitivity for detection anti-HNA antibodies compared to the former.
TRALI is an under-recognized, under-reported, serious complication of transfusion of blood products. It is defined by the National Heart, Lung, and Blood Institute (NHLBI) and a Canadian Consensus Conference as new acute lung injury (ALI) or acute respiratory distress syndrome (ARDS) occurring during or within 6 hours of transfusion of blood products [1, 2]. Possible TRALI is used when a clear temporal relationship to an alternative risk factor to ALI/ARDS also exists. TRALI is characterized by sudden onset hypoxemic respiratory insufficiency and bilateral infiltrates on chest x-ray in the absence of pre-existing ALI/ARDS and without evidence of circulatory overload [3].
The pathogenesis of TRALI is thought to involve a two-part mechanism starting with sequestration and priming of neutrophils in the lung microvasculature due to recipient factors [4, 5]. These primed neutrophils are then activated by anti-human leukocyte antigen (anti-HLA) antibodies or anti-human neutrophil antigen (anti-HNA) antibodies present in the donor. This triggers an inflammatory cascade leading to non-cardiogenic pulmonary edema. Not all TRALI is caused by anti-HLA and anti-HNA antibodies. Silliman, et al. previously reported lipids that rapidly prime neutrophil oxidase are generated during routine blood storage [6]. Reverse TRALI is a small subset of TRALI whereby circulating anti-HLA or anti-HNA antibodies in the recipient activate leukocyte antigens in the donated blood product. A case report exists describing directed donation between mother and child resulting in TRALI via maternal circulating anti-HLA antibodies activating leukocyte antigens in the donated blood from her child [7]. Recipient factors that increase the risk of developing TRALI include recent surgery (especially thoracic, vascular, and transplant), massive transfusion, active infection, critical illness, shock, positive fluid balance, and higher peak airway pressure while being mechanically ventilated [8, 9]. Donor factors that increase the risk of a TRALI event include female sex, increased parity, increased volume of anti-HLA or anti-HNA antibodies, and high-plasma-volume blood products.
Treatment of TRALI is largely supportive. The primary therapeutic modalities include supplemental oxygen and positive end expiratory pressure via non-invasive or invasive mechanisms. Several pharmacologic strategies have also been used including corticosteroids, surfactant, prostaglandin E, NSAIDs, anti-endotoxin antibodies, and anti-TNF antibodies, but the evidence for their use is limited and equivocal [10]. Recovery to the previous respiratory baseline typically occurs within 24–48 hours of symptom onset, with longer resolutions reported among those requiring invasive mechanical ventilation [11].
The literature on TRALI among neonates, especially preterm neonates is markedly limited. It is most likely underdiagnosed given the difficulties of recognizing this phenomenon amidst the concurrent respiratory disease common to preterm neonates [12].
Without the presence of identifiable antibodies in the patient’s circulation, TRALI is often a diagnosis of exclusion. This case serves to highlight the challenges of recognizing TRALI in the preterm neonatal population. A broad differential was discussed before TRALI was considered. Once the timeline of deterioration was reviewed, it became apparent that the patient developed rapid onset hypoxic respiratory failure during the RBC transfusion with significant bilateral pulmonary infiltrates noted immediately post-transfusion that had not previously been seen on the post-operative chest x-ray. Additionally, an echocardiogram was obtained showing normal cardiac function and confirmed that the flash pulmonary edema was non-cardiogenic in nature. This made the diagnosis of possible TRALI strongly suspected. Ultimately, our patient required significant supportive care, including invasive mechanical and oscillating ventilation, corticosteroids, volume boluses, and vasopressors before showing signs of improvement 36 hours following the transfusion. This course of improvement also fits with the clinical course expected of TRALI.
Like most extremely low birth weight neonates, our patient had underlying pulmonary edema and developmental lung disease with baseline respiratory insufficiency owing to prematurity. Additionally, the patient had undergone PDA ligation, so some degree of decompensation was expected. However, the presenting deterioration was respiratory rather than cardiovascular in nature. A component of post-PDA ligation syndrome contributing to the hypotension made worse by TRALI is a plausible explanation of severe hypotension in this situation. Serratia tracheitis or pneumonia could be an alternative explanation for the acute respiratory failure; however one would not expect this degree of pulmonary infiltrates and sudden acute deterioration to develop in such a short time frame. Thus, the positive tracheal culture was more likely a contributing factor increasing the risk for TRALI. Overall, this patient had multiple risk factors for TRALI, including recent thoracic surgery, possible infection, mechanical ventilation and critical illness. Although the negative antibody testing of the donor reduces the likelihood of confirmed TRALI, absence of anti-HLA and anti-HNA antibodies does not totally rule it out, since not all reported TRALI cases have identifiable HLA-antibodies in donor plasma [13] and biologically active lipids that accumulate during storage of blood products have also been implicated to cause TRALI [6]. The +28 day status of the implicated unit at time of transfusion may be compatible with this mechanism.
The reference laboratory testing results suggest an antibody directed against a high-incidence HNA absent on maternal neutrophils but present on donor neutrophils. Alternatively, a neutrophil-directed autoantibody or the presence of multiple anti-HNA alloantibodies is possible. Although the available testing is qualitative in nature and therefore not designed to identify antibody specificity, results would support a possible reverse-TRALI mechanism. In our case, we suspect that pre-existing cardiopulmonary disease in the neonate represented the first hit, and transfusion of residual donor neutrophils in the presence of residual circulating, maternally derived, anti-HNA antibodies representing the second hit.
Conclusion
Recognition and diagnosis of TRALI among preterm neonates will remain a challenging proposition. However, clinicians should be aware of the possibility of this phenomenon since this is a serious cause of morbidity and mortality in patients who receive blood product transfusions. Raising awareness of this diagnosis may result to identification of donors who have identifiable anti-leukocyte antibodies and potentially prevent future occurrence of such a life-threatening complication.
Disclosure statement
All authors have nothing to disclose.