British Journal of Haematology, 1996, 93, 707 –713 Transfusion-related acute lung injury due to HLA-A2-specific antibodies in recipient and NB1-specific antibodies in donor blood JUăRGEN BUX, FRANK BECKER , W ERNER SEEGER , DAVID KILPATRICK, JUDITH CHAP MAN AND A LAN WATERS Institute for Clinical Immunology and Transfusion Medicine and Department of Internal Medicine, Justus-Liebig University, Giessen, and Department of Internal Medicine, St Josefs Hospital, Giessen, Germany; Department of Cardiothoracic Surgery, Royal Infirmary, Edinburgh, and Department of Haematology, St Bartholomew’s Hospital, London, U.K Received 21 November 1995; accepted for publication 22 February 1996 Summary Transfusion-related acute lung injury (TRALI) is a hazardous but little-known complication of blood transfusion, characterized by non-cardiogenic lung oedema after blood transfusion Leucoagglutinating antibodies in the donor plasma are considered to play a central role in the pathogenesis of TRALI but no recommended procedure currently exists for their detection, and most of them have not yet been well characterized Serum samples of two patients who have developed TRALI within 30 of blood transfusion and the sera of the involved blood donors were investigated for leucocyte antibodies by granulocyte immunofluorescence, granulocyte agglutination and lymphocytotoxicity assays using typed test cells Suspected specificities of the detected antibodies were confirmed by a luminoimmunoblot assay and the antigen capture assay MAIGA One case was associated with granulocyte agglutinating anti-HLA-A2 antibodies in the recipient’s (i.e patient’s) own blood and the other with donor-related non-agglutinating antibodies directed against the granulocyte-specific antigen NB1 Leucocyte incompatibility between donor and recipient was shown in both cases by crossmatching and typing of the incompatible cells for the appropriate antigen The results show that TRALI is associated not only with donor- but also with recipient-related leucocyte antibodies In addition to leucoagglutinating antibodies, non-agglutinating granulocyte-specific antibodies can be also involved For immunodiagnosis, sera from both must be investigated by a combination of granulocyte and lymphocyte (HLA) antibody screening tests and leucocyte incompatibility verified by crossmatching Transfusion-associatedlung oedema is traditionally attributed to circulatory overload due to over-transfusion or pre-existing myocardial insufficiency However, as early as 1951, Barnard described pulmonary oedema without hypervolaemia associated with blood transfusion A few similar reports of acute respiratory distress due to non-cardiogenic pulmonary oedema followed (summarized by Nordhagen et al, 1986) using a variety of designations Nowadays, the term transfusion-related acute lung injury (TRALI) is frequently used (Popovsky & Moore, 1985) Since Ward’s description of leucoagglutinins in cases of TRALI in 1970 (Ward, 1970), there is increasing evidence that TRALI is associated with the presence of leucocyte, and in particular, granulocyteagglutinating antibodies (Nordhagen et al, 1986) McCullough et al (1986) found an abnormal pulmonary sequestration of 111-indium-labelled granulocytes injected in patients with granulocyte antibodies More recently, Seeger et al, (1990) reproduced TRALI in an ex vivo rabbit lung model using human granulocytes, granulocyte-agglutinating antibodies and rabbit plasma as complement source However, with the exception of anti-NA2 (Yomtovian et al, 1984), anti5b (Nordhagen et al, 1986) and anti-NB2 (Van Buren et al, 1990), leucocyte antibodies have not been well characterized and techniques for their demonstration have varied The role of HLA and non-agglutinating granulocyte antibodies in TRALI is still not clear Additionally, it is not known whether recipient-related antibodies can cause TRALI or not Correspondence: Dr Juă rgen Bux, Institute for Clinical Immunology and Transfusion Medicine, Justus-Liebig University, Langhansstr 7, D-35385 Giessen, Germany # 1996 Blackwell Science Ltd Keywords: lung injury, transfusion reaction, granulocyte serology, neutrophil 707 708 Juărgen Bux et al We report two cases of TRALI which were caused by wellcharacterized non-agglutinating NB1-specific antibodies and recipient’s own anti-HLA-A2 antibodies, respectively CASE REPORT A 58-year-old woman with an 18-year history of diabetes mellitus type II was hospitalized for the amputation of the right forefoot because of gangrene The patient had become insulin-dependent and had developed the typical late syndrome of diabetes mellitus with signs of macro- and microangiopathy, retinopathy and nephropathy Mild hypertension was also present but well controlled by medication Since 1988 the patient had suffered from blood flow disorders ofthe left foot which resulted in 1990 in its amputation because of extended forefoot gangrene The patient showed multiple small foci of infected gangrene of the right forefoot but had no fever The chest X-ray at Fig Case Chest X-ray d before blood transfusion Fig Case Chest X-ray shortly after blood transfusion # 1996 Blackwell Science Ltd, British Journal of Haematology 93: 707 –713 Transfusion-related Acute Lung Injury the time of admission is shown in Fig 1; it shows a lung without shadows and the heart is not enlarged The electrocardiogram was unremarkable months before admission to the hospital for surgical preparation, a echocardiogram had been done in which the left ventricle was not enlarged, contractility was normal, and there was no hypokinesia The other cardiac chambers and heart valves were within normal limits The patient was given clindamycine and gentamicin for the infected gangrene, which improved during the following days Since the haemoglobin level of g/ dl on admission dropped to g/ dl d later, it was decided to transfuse units of buffy coat-poor erythrocyte concentrates The transfusion of the first concentrate lasted 90 30 later the patient developed shaking chills and complained of dyspnoea Auscultation was typical for lung oedema Despite treatment with frusemide, prednisolone (1 g) and diphenylhydramine, respiratory distress worsened and the patient had to be intubated The chest X-ray carried out shortly after the event is shown in Fig 2; it shows infiltrates in both lungs indicating a marked interstitial/ alveolar lung oedema but a heart picture within normal limits Hypertension of 180/ 90 mmHg and tachycardia of 140 beats/ were measured By echocardiography a normal left ventricle with good contractility was found Respiratory function did not improve despite dehydration therapy The patient required artificial ventilation with PEEP of + cmH2 O The peak airway pressure was 40 cmH2 O Arterial blood gases on a minute volume of 10 litres with FiO2 of showed a PaO2 of and a PaCO2 of kPa The central venous pressure was 9 cmH2 O The following day a bronchoalveloar lavage was carried out showing a large number of granulocytes Respiratory function had improved so the FiO2 was reduced to on airway pressure of 25 cmH2 O A red blood cell concentrate transfused on the same day was well tolerated On the evening of the second day after the transfusion reaction the patient began to develop fever which continued over the next d with peaks up to 40 8C, and did not improve despite intensive treatment with antibiotics A further erythrocyte concentrate transfused on the third day after the reaction was again well tolerated Since Candida glabrata was found repeatedly in large amounts in the tracheal secretion, amphotericin B was administered on the 6th day after the transfusion reaction Soon after, the patient died of sudden heart failure CASE REPORT A 45-year-old, 80 kg, man with an 8-month history of disabling angina presented for coronary artery surgery due to an occluded right coronary artery and severe stenoses in the left anterior descending and circumflex arteries He did not smoke and had never received blood transfusions Angiography showed good left ventricular function with anteroapical hypokinesia and ejection fraction of 65% Physical and laboratory examinations were normal Coronary artery surgery was carried out in the standard fashion After rewarming, the patient was easily weaned in sinus rhythm without inotropic support on a low-dose GTN infusion, heart rate of 75 beats/ and central venous pressure of 709 7 mmHg PaO2 was 57 kPa and PaCO2 kPa with FiO2 of 0, and minute volume of l/ 30 after bypass and protamine administration, unit of whole blood was administered After 15 the SAP fell to 95/ 50 mmHg, the central venous pressure rose to 12 mmHg and the heart rate to 85 beats/ The changes were followed closely by the appearance of pulmonary oedema fluid from the tracheal tube The peak airway pressure rose from 11 to 36 cmH2 O Arterial blood gases on a minute volume of litres with FiO2 of showed a PaO2 of and a PaCO2 of kPa Further investigation with flow-directed pulmonary artery catheterization revealed pulmonary artery pressures of 38/ 20 mmHg and wedge pressure of 10 mmHg Despite increasing minute ventilation to 14 l/ (FiO2 0, PEEP cmH2 O), administration of methylprednisolone and optimization of cardiac function (cardiac output of –5 l/ min), pulmonary oedema persisted with deteriorating gas exchange In order not to threaten the integrity of the internal mammary graft by even more increased ventilation, extracorporeal membrane ventilation was instituted with flows of l/ and satisfactory gas exchange (PaO2 of 30 kPa and a PaCO2 of kPa) Ventilation was 10 l/ with a FiO2 of PEEP of cmH2 O Pulmonary function improved during the first postoperative day and ventilatory and inotropic support were weaned The chest X-ray cleared and the patient was extubated on the second post-operative day Subsequent recovery was uneventful The clinical course of this case has been reported previously (Worsley et al, 1991) MATERIAL AND METHODS Serum samples from the donors and patients were collected and tested in the routine manner for red blood cell compatibility Cells for the granulocyte and lymphocyte panel were obtained from normal individuals previously typed for neutrophil antigens (NA1, NA2, NB1, NB2, 5b) and HLA-A, -B and -C locus antigens Lymphocytes and granulocytes were isolated by the method of Boăyum (1968) Serum samples were screened for leucocyte antibodies by a modified granulocyte immunofluorescence test (Verheugt et al, 1977), the granulocyte agglutination test (Lalezari et al, 1979) and the lymphocytotoxicity test (Terasaki & Park, 1982) Suspected antigen specificities were confirmed using appropriate antigen-specific assays (Bux et al, 1996) For confirmation of NB1-specificity of the serum we developed a luminoimmunoblotting technique: 10 granulocytes were lysed with 500 l lysis buffer consisting of 1% Triton-X 100 and the enzyme inhibitors PMSF (2 m M), aprotinin (500 kIE/ ml), EDTA (5 m M) and leupeptin (0 g/ ml) in trisbuffered saline, pH After lysate centrifugation (30 min, 12000 g) and SDS-polyacrylamide gel electrophoresis of the supernatant, the antigens were blotted onto a nitrocellulose membrane Membrane strips were blocked with 5% bovine serum albumin and afterwards incubated with diluted serum samples (1:20 –1:200) The strips were washed with tris-buffered saline, pH 4, containing 05% Tween 20 and then incubated with peroxidaseconjugated rabbit anti-human antibodies (dilution 1:200 000) After another washing step, a chemiluminescent # 1996 Blackwell Science Ltd, British Journal of Haematology 93: 707 713 710 Juărgen Bux et al Fig Case Results of the MAIGA assay Alloantibodies in serum bind only to isolated HLA class I molecules which are from HLA-A2-positive cells but not from HLA-A2-negative cells substrate (ECL, Amersham, U.K.) was added for and the recognized antigens visualized on a X-ray film To confirm the suspected presence of alloantibodies with HLA-A2 specificity in the patient’s serum, the recently described antigen capture assay MAIGA was used (Bux et al, 1993b) Briefly, 10 granulocytes were incubated with 50 l serum and the HLA class I molecule-specific monoclonal antibody TP25.99, obtained from Dr S Ferrone (New York Medical College, New York, U.S.A.) After lysis of the washed granulocytes with the same lysis buffer as used for immunoblotting, the lysate was centrifuged The immune complexes in the supernatant consisting of monoclonal antibody, antigen and, in the positive case, a human antibody were immobilized to the surface of plastic tubes by binding to goat anti-mouse antibodies precoated to the tubes surface After washing the tubes, the binding of a human antibody to the antigen was tested by addition of a peroxidase-conjugated goat anti-human antibody (Fig 3) Detection was done by a chemiluminescence assay using a chemiluminescent substrate (luminol, 4-iodophenol, H2 O2 in tris-buffered saline, pH 0) and a luminometer (Lumat, Beckmann, Bad Wildbad, Germany) In a modification of the MAIGA assay lymphocytes were used instead of granulocytes RESULTS Case Red blood cell incompatibility was not detectable In the blood of the donor, a 28-year-old man who had never received blood transfusions, neither granulocyte-reactive nor lymphocytereactive antibodies were detectable by granulocyte immunofluorescence and agglutination tests and lymphocytotoxicity test (data not shown) In the blood of the patient, a woman who had a history of three pregnancies (three live births) and three blood transfusions years previously, antibodies binding to both granulocytes and lymphocytes could be detected, as shown in Table I The antibodies had a strong agglutinating effect on granulocytes The leucocyte crossmatch between the donor’s granulocytes and lymphocytes and the patient’s serum confirmed the suspected leucocyte incompatibility (Table I) The reaction pattern with the panel lymphocytes in the complement-dependent lymphocytotoxicity test indicated the presence of antibodies with specificity for the HLA-A2 antigen HLA typing of the blood donors showed that the implicated donor was HLA-A2 positive (HLA phenotype: A1,2; B7,44(12); Cÿ; ÿ) whereas the other two donors did not have the HLA-A2 antigen (HLA phenotypes: A1,23(9); # 1996 Blackwell Science Ltd, British Journal of Haematology 93: 707 –713 Transfusion-related Acute Lung Injury 711 Table I Case report 1: serologic results Leucocyte antibody screening results Granuloctyte Patient’s serum: pos./ neg results Donor’s serum: pos./ neg result Immunofluorescence (n 8) Agglutination (n 8) Lymphocytotoxicity (n 51) 1/ 0/ 1/ 0/ 13/ 38 0/ 51 Suspected antibody specificity in the patient’s blood: anti-HLA-A2 Leucocyte crossmatch: patient serum versus donor leucocytes Granulocyte Patient serum Immunofluorescence Agglutination Lymphocytotoxicity Pos () pos ( ) pos ( ) Phenotyping result of donor granulocytes: HLA-A2 B8,44(12); Cw4,ÿ and A3,ÿ; B7,ÿ; Cw7,ÿ) and their red cell concentrates were transfused without any complications shortly after the reaction event In order to confirm the specificity of the HLA-A2 alloantibodies, the serum was tested in the antigen capture assay MAIGA Results are shown in Fig By this assay, antibodies binding only to the HLA class I molecules isolated from HLA-A2-positive granulocytes and lymphocytes could be detected There were no antibodies detectable which bound to class I molecules bearing epitopes other than HLAA2 Further testing of the patient’s serum in MAIGA for antibodies binding to the integrin molecule CD11b/ CD18 and the Fc-gamma-receptors II and III as well as immunoblotting revealed no additional granulocyte antibodies (data not shown) Case Red blood cell incompatibility was not detectable The Table II Case report 2: serologic results Leucocyte antibody screening results Granuloctyte Donor’s serum: pos./ neg results Patient’s serum: pos./ neg result Immunofluorescence (n 14) Agglutination (n 8) Lymphocytotoxicity (n 20) 11/ 0/ 14 0/ 0/ 0/ 20 0/ 20 Suspected antibody specificity: anti-NB1 Granulocyte crossmatch: donor serum versus patient’s granulocytes Granulocyte Donor serum Immunofluorescence Agglutination Pos () neg Phenotyping result of patient granulocytes: NB1 # 1996 Blackwell Science Ltd, British Journal of Haematology 93: 707 713 712 Juărgen Bux et al implicated blood donor, a 41-year-old female, had a history of four pregnancies (one live birth, one neonatal death and two abortions) and therefore a high chance of immunization against leucocyte antigens The leucocyte antibody screening results are shown in Table II The donor serum was screened for granulocyte-reactive antibodies by the granulocyte immunofluorescence test using two panels of a total of 14 test granulocytes and by the granulocyte agglutination test using one panel of eight granulocyte donors HLA antibody screening was done by the lymphocytotoxicity test using a panel of 20 lymphocytes Positive results were found only in the granulocyte immunofluorescence test which indicated the presence of a non-agglutinating granulocyte-specific antibody The granulocyte crossmatch between donor serum and patient’s cells confirmed the suspected incompatibility between donor plasma and patient’s granulocytes due to granulocyte antibodies (Table II) The reaction pattern with the granulocytes of the test panel suggested the presence of NB1-specific antibodies Serotyping of the donor and recipient granulocytes showed that the donor cells had the phenotypes NA1 ,NA2 ; NB1 ÿ and the patient cells NA1 ,NA2 ÿ; NB1 and supported the assumption of NB1-specific alloantibodies in the donor’s blood In order to confirm the NB1 specificity an immunoblot, an antigenspecific test, was carried out The immunoblot results showed the two bands of 58 and 64 kD which are typical for the NB1 antigen (Stroncek et al, 1990) (Fig 4) DISCUSSION Diagnosis of antibody-mediated TRALl is made by clinical Fig Case Confirmation of suspected NB1 antigen specificity by luminoimmunoblot test Lane shows the results of the implicated serum The indicated bands of Mr58 and 64 kD are typical for the granulocyte-specific NB1 antigen The results of the negative control serum is shown in lane presentation and immunological investigations TRALI occurs during, or within few hours after, transfusion (Popovsky et al, 1985) Patients complain of acute dyspnoea or, in patients with ventilatory assistance, the inspiratory airway pressure increases and arterial blood gas analysis shows hypoxia and hypercapnia The chest X-ray demonstrates bilateral pulmonary infiltrates typical of lung oedema in the absence of cardiac enlargement and pulmonary vascular engorgement Normal capillary wedge pressure and no abnormal findings in echocardiography verify the non-cardiogenic origin of the oedema Mechanical ventilation with PEEP became necessary in up to 70% and death has been estimated to occur in 6% of patients (Popovsky et al, 1985) Immunodiagnosismust include testing the donor’s and the recipient’s blood In contrast to the common antibodymediated transfusion reactions, TRALI so far has been considered to be caused by antibodies in the donor plasma (Nordhagen et al, 1986; Popovsky et al, 1985) Due to the relatively low number of leucocytes in the transfused blood components such as buffy coat-depleted erythrocyte concentrates, it has been questioned that TRALI can be caused by leucocyte antibodies in the patient’s blood However, our first case report shows that this does happen Possibly, the transfusion of a blood component containing incompatible leucocytes results in the activation of the recipient’s own granulocytes either as a consequence of activating factors released by the few transfused granulocytes after antibody binding or by an abnormal binding of the recipient’s alloantibodiesto his own granulocytes The latter mechanism is well known in post-transfusion purpura In any case, the leucocyte antibodies are typically formed by multiparous women The present report has identified two further leucocyte antibody specificities associated with TRALI The first case report shows that granulocyte agglutinating HLA-A2 antibodies are able to initiate TRALI The role of HLA antibodies remained unclear because TRALI due to HLA antibodies occurred more rarely than could be expected from their frequency determined by lymphocytotoxicity Since it is known that granulocytes play a central role in the pathogenesis of acute respiratory distress syndrome, a strong agglutinating effect on granulocytes which seems to be typical of HLA-A2 (Dausset, 1958), but not of all HLA antibodies, is probably a prerequisite for the initiation of TRALI This is supported by the finding that HLA antibodies only detectable by lymphocytotoxicity did not cause abnormal pulmonary sequestration of injected 111-indium granulocytes (McCullough et al, 1986) Our second case report shows that non-agglutinating antibodies, such as the reported granulocyte-specific anti-NB1 antibody, can also cause TRALI Since the NB1 antigen is expressed on a variable subpopulation of neutrophils from different individuals, and sometimes in the same individual over a period of time (Goldschmeding et al, 1992), agglutination can occur in vivo but must not be necessarily detectable in vitro On the other hand, binding of NB1-specific antibodies to granulocytes can activate granulocytes without agglutination by internalization of NB1 antigens after antibody-mediated crosslinking # 1996 Blackwell Science Ltd, British Journal of Haematology 93: 707 –713 Transfusion-related Acute Lung Injury which may result in oxygen radical release (Goldschmeding et al, 1992; Bux et al, 1993a) Since some granulocyte antibodies can be detected only by agglutination (Nordhagen et al, 1986), or by immunofluorescence as shown in the second case report, for granulocyte antibody screening, immunofluorescence and agglutination tests are, at present, the best means of detection We prefer the granulocyte agglutination to the leucocyte agglutination test because the latter is difficult to standardize and lymphocytes probably not to contribute to agglutination For HLA antibody screening, lymphocytotoxicity is the standard technique Leucocyte incompatibility of blood transfusion has to be verified by crossmatching Specificity of the detected antibodies should be confirmed by antigenspecific tests such as immunoblotting, immunoprecipitation or the antigen capture assay MAIGA, and incompatible cells should be typed for the appropriate antigen Alloimmunization to granulocyte or HLA antigens occurs usually during pregnancy, so that directed blood donations are of increased risk In the United States, acute pulmonary injury has been reported to be the second commonest cause of transfusionassociated death after acute haemolysis due to transfusion of ABO-incompatible blood products (Sazama, 1990) Possibly the greater popularity of directed donations in the U.S.A than in Europe has an impact on the incidence of TRALI On the other hand, TRALI is still little known and probably often misdiagnosed The investigation of a suspected case of TRALI is important in order to prevent further events by identifying a potentially ‘dangerous donor’ or a ‘susceptible’ patient Depending on the component preparation procedure, the donor plasma and platelets should be discarded or the donor removed from the donor panel The patients concerned should consequently be transfused with leucocyte-compatible blood components ACKNOWLEDGMENT Supported in part by a grant (Bu 770/ 3-1) from the Deutsche Forschungsgemeinschaft REFERENCES Barnard, R.D (1951) Indiscriminate transfusion: a critique case report illustrating hypersensitivity reactions New York State Journal of Medicine, 51, 2399–2402 Boăyum, A (1968) Isolation of mononuclear cells and granulocytes from human blood Scandinavian Journal of Clinical and Laboratory Investigation, 21, (Suppl 97), 77 –89 Bux, J (1996) Challenges in the detection and characterisation of clinical relevant neutrophil antibodies and antigens Transfusion Medicine Reviews (in press) Bux, J., Dickmann, J.O., Stockert, U & Mueller-Eckhardt, C (1993a) Influence of granulocyte antibodies on granulocyte function Vox Sanguinis, 64, 220 –225 713 Bux, J., Kober, B., Kiefel, V &Mueller-Eckhardt, C (1993b) Analysis of granulocyte-reactive antibodies using an immunoassay based upon monoclonal antibody-specific immobilization of granulocyte antigens (MAIGA) Transfusion Medicine, 3, 157 –162 Dausset, J (1958) Iso-leuco-anticorps Acta Haematologica, 20, 156 – 166 Goldschmeding, R., van Dalen, C.M., Faber, N., Calafat, J., Huizinga, T.W.J., van der Schoot, C.E., Clement, L.T & von dem Borne, A.E.G.Kr (1992) Further characterization of the NB1 antigen as a variably expressed 56 –62 kD GPI-linked glycoprotein of plasma membranes and specific granules of neutrophils British Journal of Haematology, 81, 336 –345 Lalezari, P., Jiang, A &Lee, S (1979) A microagglutination technique for the detection of leucocyte agglutinins NIAID Manual of Tissue Typing Techniques (ed by J G Ray), pp 20 –22 NIH Publication No 80-545 McCullough, J., Clay, M., Hurd, D., Richards, K., Ludvigsen, C & Forstrom L (1986) Effect of leucocyte antibodies and HLA matching on the intravascular recovery, survival, and tissue localization of 111-indium-granulocytes Blood, 67, 522 –528 Nordhagen, R., Conradi, M & Droămtorp, S.M (1986) Pulmonary reaction associated with transfusion of plasma containing anti-5b Vox Sanguinis, 51, 102 –107 Popovsky, M.A & Moore, S.B (1985) Diagnostic and pathogenetic considerations in transfusion-related acute lung injury Transfusion, 25, 573 –577 Sazama, K (1990) Reports of 355 transfusion-associated deaths: 1976 through 1985 Transfusion, 30, 583 –590 Seeger, W., Schneider, U., Kreusler, B., von Witzleben, E., Walmrath, D., Grimminger, F & Neppert, J (1990) Reproduction of transfusion-related acute lung injury in an ex vivo lung model Blood, 76, 1438–1444 Stroncek, D.F., Skubitz, K.M & McCullough, J.J (1990) Biochemical characterization of the neutrophil-specific antigen NB1 Blood, 75, 744 –755 Terasaki, P.I & Park, M.S (1982) Microdroplet lymphocyte cytotoxicity test NIAID Manual of Tissue Typing Techniques, 1979 –1980 (ed by J G Ray), pp 92 –103 NIH Publication No 77-545 Van Buren, N.L., Stroncek, D.F., Clay, M.E., McCullough, J & Dalmasso, A.P (1990) Transfusion-related acute lung injury by an NB2 granulocyte-specific antibody in a patient with thrombotic thrombocytopenic purpura Transfusion, 30, 42 –45 Verheugt, F.W.A., von dem Borne, A.E.G.Kr., Decary, S & Engelfriet, C.P (1977) The detection of granulocyte alloantibodies with an indirect immunofluorescence test British Journal of Haematology, 36, 533 –544 Ward, H.N (1970) Pulmonary infiltrates associated with leukoagglutinin transfusion reactions Annals of Internal Medicine, 73, 689 –694 Worsley, M.H., Sinclair, C.J., Campanella, C., Kilpatrick, D.C & Yap, P.L (1991) Non-cardiogenic pulmonary oedema after transfusion with granulocyte antibody containing blood: treatment with extracorporeal membrane oxygenation British Journal of Anaesthesiology, 67, 116 –119 Yomtovian, R., Press, C., Engman, H., Kline, W., Clay, M & McCullough, J (1984) Severe pulmonary hypersensitivity associated with passive transfusion of a neutrophil-specific antibody Lancet, i, 244 –246 # 1996 Blackwell Science Ltd, British Journal of Haematology 93: 707 –713 ... before blood transfusion Fig Case Chest X-ray shortly after blood transfusion # 1996 Blackwell Science Ltd, British Journal of Haematology 93: 707 –713 Transfusion- related Acute Lung Injury the... with transfusion of plasma containing anti-5b Vox Sanguinis, 51, 102 –107 Popovsky, M.A & Moore, S.B (1985) Diagnostic and pathogenetic considerations in transfusion- related acute lung injury Transfusion, ... Witzleben, E., Walmrath, D., Grimminger, F & Neppert, J (1990) Reproduction of transfusion- related acute lung injury in an ex vivo lung model Blood, 76, 1438–1444 Stroncek, D.F., Skubitz, K.M & McCullough,